Recovery Plan for the Mexican Spotted Owl (Strix occidentalis)

December 1995

Recovery PlanRecovery PlanRecovery PlanRecovery PlanRecovery Planfor thefor thefor thefor thefor the

Mexican Spotted OwlMexican Spotted OwlMexican Spotted OwlMexican Spotted OwlMexican Spotted Owl(((((Strix occidentalis lucidaStrix occidentalis lucidaStrix occidentalis lucidaStrix occidentalis lucidaStrix occidentalis lucida)))))

Plan de Recuperaciondel Tecolate Moteado Mexicano

(Strix occidentalis lucida)

Recovery PlanRecovery PlanRecovery PlanRecovery PlanRecovery Planfor thefor thefor thefor thefor the

Mexican Spotted OwlMexican Spotted OwlMexican Spotted OwlMexican Spotted OwlMexican Spotted Owl(((((Strix occidentalis lucidaStrix occidentalis lucidaStrix occidentalis lucidaStrix occidentalis lucidaStrix occidentalis lucida)))))

Plan de Recuperaciondel Tecolate Moteado Mexicano

(Strix occidentalis lucida)

Nancy M. Kaufman, Regional Director,U.S. Department of the Interior, Fish and Wildlife Service,

Southwestern Region

Primary Authors:Primary Authors:Primary Authors:Primary Authors:Primary Authors:

William M. Block; Fernando Clemente; Jack F. Cully; James L. Dick, Jr.; Alan B. Franklin;Joseph L. Ganey; Frank P. Howe; W.H. Moir; Steven L. Spangle; Sarah E. Rinkevich; Dean L. Urban;

Robert Vahle; James P. Ward, Jr.; and Gary C. White

Other Contributors:Other Contributors:Other Contributors:Other Contributors:Other Contributors:

Kevin J. Cook (Technical Editor): Brian Geils (Disturbance Analyses);Kate W. Grandison (Meeting Facilitator); Tim Keitt (Landscape Analyses); Juan F. Martinez-Montoya

(Mexican Recovery Units); Art Needleman and Jill Simmons (Layout and Design);Joyce V. Patterson (Illustrator); Tom Spalding (State of Arizona); Rich Teck (Forest Modeling);Steven Thompson (Southwestern Tribal Liaison); Brenda E. Witsell (Administrative Assistant).

1995

Approved:__________________________________________________ Date:_____________Regional Director, U.S. Fish and Wildlife Service

DISCLAIMER:DISCLAIMER:DISCLAIMER:DISCLAIMER:DISCLAIMER:

This Recovery Plan is not intended to provide details on all aspects of Mexican spotted owl management.The Recovery Plan outlines steps necessary to bring about recovery of the species. The Recovery Plan is nota “decision document” as defined by the National Environmental Policy Act (NEPA). It does not allocateresources on public lands. The implementation of the recovery plan is the responsibility of Federal and Statemanagement agencies in areas where the species occurs. Implementation is done through incorporation ofappropriate portions of the Recovery Plan in agency decision documents such as forest plans, park manage-ment plans, and State game management plans. Such documents are then subject to the NEPA process forpublic review and selection of alternatives.

LITERATURE CITATIONS:LITERATURE CITATIONS:LITERATURE CITATIONS:LITERATURE CITATIONS:LITERATURE CITATIONS:

This document should be referenced in literature citations as follows:

USDI Fish and Wildlife Service. 1995. Recovery plan for the Mexican spotted owl: Vol.I.Albuquerque, New Mexico. XXXpp.

A fee for additional copies of this document will be charged depending on the number of pagesand postage. Additional copies of this document may be purchased from:

Fish and Wildlife Reference Service5430 Grosvenor Lane, Suite 110Bethesda, MD 20814(303) 492-64031-800-582-3421

Mexican Spotted Owl Recovery Plan

MEXICAN SPOMEXICAN SPOMEXICAN SPOMEXICAN SPOMEXICAN SPOTTTTTTED OTED OTED OTED OTED OWL RECOWL RECOWL RECOWL RECOWL RECOVERVERVERVERVERY PLY PLY PLY PLY PLANANANANAN

Table of ContentsTable of ContentsTable of ContentsTable of ContentsTable of Contents

List of List of List of List of List of TTTTTablesablesablesablesables ................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................ vii vii vii vii viiList of FList of FList of FList of FList of Figurigurigurigurigureseseseses ........................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................... viiiviiiviiiviiiviiiEEEEExxxxxecutivecutivecutivecutivecutive Se Se Se Se Summarummarummarummarummaryyyyy ................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................... ixixixixixAAAAAcknocknocknocknocknowledgmentswledgmentswledgmentswledgmentswledgments ........................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................ xiiixiiixiiixiiixiii

PPPPPARARARARART I: PLT I: PLT I: PLT I: PLT I: PLAN DEAN DEAN DEAN DEAN DEVELVELVELVELVELOPMENTOPMENTOPMENTOPMENTOPMENT

A. RECOA. RECOA. RECOA. RECOA. RECOVERVERVERVERVERY PLY PLY PLY PLY PLANNINGANNINGANNINGANNINGANNING ................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................ 11111B. LISTINGB. LISTINGB. LISTINGB. LISTINGB. LISTING ............................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................ 22222

The Present or Threatened Destruction, Modificiation, or Curtailment of its Habitat or Range .....................................................................................................2

Overutilization for Commercial, Recreational, Scientific, or EducationalPurposes ..........................................................................................................................3

Disease or Predation ........................................................................................................ 3Inadequacy of Existing Regulatory Mechanisms ...................................................................3Other Natural or Manmade Factors Affecting Its Continued Existence ................................ 3

C. PC. PC. PC. PC. PAST AND CURRENT MANAAST AND CURRENT MANAAST AND CURRENT MANAAST AND CURRENT MANAAST AND CURRENT MANAGEMENT OF GEMENT OF GEMENT OF GEMENT OF GEMENT OF THETHETHETHETHEMEXICAN SPOMEXICAN SPOMEXICAN SPOMEXICAN SPOMEXICAN SPOTTTTTTED OTED OTED OTED OTED OWLWLWLWLWL .................................................................................................................................................................................................................................................................................................................................................................................................................................................................. 44444Fish and Wildlife Service ...................................................................................................... 4Forest Service ....................................................................................................................... 4

Forest Service Southwestern Region (Region 3) ................................................................ 4Forest Service Rocky Mountain Region (Region 2) .......................................................... 5Forest Service Intermountain Region (Region 4) .............................................................. 5

Other Federal Agencies ........................................................................................................ 6National Park Service ....................................................................................................... 6Bureau of Land Management ........................................................................................... 6Department of Defense ................................................................................................... 6

States ................................................................................................................................... 7Arizona ............................................................................................................................ 7New Mexico .................................................................................................................... 7Colorado ......................................................................................................................... 7Utah ................................................................................................................................ 7Texas ............................................................................................................................... 8

Tribes .................................................................................................................................. 8Mescalero Apache Tribe ................................................................................................... 8White Mountain Apache Tribe......................................................................................... 8San Carlos Apache Tribe .................................................................................................. 9Jicarilla Apache ................................................................................................................ 9Navajo Nation ................................................................................................................. 9

Mexico ................................................................................................................................ 9D. CONSIDERAD. CONSIDERAD. CONSIDERAD. CONSIDERAD. CONSIDERATIONS IN PLTIONS IN PLTIONS IN PLTIONS IN PLTIONS IN PLAN DEAN DEAN DEAN DEAN DEVELVELVELVELVELOPMENTOPMENTOPMENTOPMENTOPMENT ................................................................................................................................................................................................................................................................................... 1111111111

Recovery Units .................................................................................................................... 11The Current Situation ......................................................................................................... 11Recovery Plan Duration ....................................................................................................... 12Conservation Plans for Other Spotted Owl Subspecies ......................................................... 13

i


Ecosystem Management ..................................................................................................... 15Economic Considerations ................................................................................................... 16Human Intervention and Natural Processes ........................................................................ 17Differences Between the Draft and Final Recovery Plans ..................................................... 17

PPPPPARARARARART II: BIOLT II: BIOLT II: BIOLT II: BIOLT II: BIOLOGICAL AND ECOLOGICAL AND ECOLOGICAL AND ECOLOGICAL AND ECOLOGICAL AND ECOLOGICAL BAOGICAL BAOGICAL BAOGICAL BAOGICAL BACKCKCKCKCKGRGRGRGRGROUNDOUNDOUNDOUNDOUND

A. GENERAL BIOLA. GENERAL BIOLA. GENERAL BIOLA. GENERAL BIOLA. GENERAL BIOLOGY AND ECOLOGY AND ECOLOGY AND ECOLOGY AND ECOLOGY AND ECOLOGICAL RELOGICAL RELOGICAL RELOGICAL RELOGICAL RELAAAAATIONSHIPSTIONSHIPSTIONSHIPSTIONSHIPSTIONSHIPSOF OF OF OF OF THE MEXICAN SPOTHE MEXICAN SPOTHE MEXICAN SPOTHE MEXICAN SPOTHE MEXICAN SPOTTTTTTED OTED OTED OTED OTED OWLWLWLWLWL ............................................................................................................................................................................................................................................................................................................................................................................. 1919191919Taxonomy .......................................................................................................................... 19Description ........................................................................................................................ 21Distribution and Abundance .............................................................................................. 21Habitat Associations ........................................................................................................... 26

Nesting and Roosting Habitat......................................................................................... 26Foraging Habitat ............................................................................................................ 27

Territoriality and Home Range ........................................................................................... 27Vocalizations ...................................................................................................................... 28Interspecific Competition ................................................................................................... 28Feeding Habitats and Prey Ecology ..................................................................................... 28Reproductive Biology ......................................................................................................... 29Mortality Factors ................................................................................................................ 31

Predation ........................................................................................................................ 31Starvation ....................................................................................................................... 31Accidents ........................................................................................................................ 31Disease and Parasites ....................................................................................................... 32

Population Biology ............................................................................................................. 32Survival .......................................................................................................................... 32Reproduction ................................................................................................................. 33Environmental Variation ................................................................................................. 33Population Trends ........................................................................................................... 33

Movements ......................................................................................................................... 33Seasonal Movements. ...................................................................................................... 33Natal Dispersal ............................................................................................................... 33

Landscape Pattern and Metapopulation Structure ............................................................... 34Conclusions ....................................................................................................................... 34

B.B.B.B.B. RECORECORECORECORECOVERVERVERVERVERY UNITY UNITY UNITY UNITY UNITSSSSS ........................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................ 3636363636United States ...................................................................................................................... 36

Colorado Plateau ............................................................................................................ 36Southern Rocky Mountains - Colorado. .......................................................................... 40Southern Rocky Mountains - New Mexico...................................................................... 40Upper Gila Mountains .................................................................................................... 44Basin and Range - West .................................................................................................. 46Basin and Range - East ................................................................................................... 46

Mexico ............................................................................................................................... 49Sierra Madre Occidental - Norte ..................................................................................... 49Sierra Madre Oriental- Norte .......................................................................................... 50Sierra Madre Occidental - Sur........................ ................................................................. 50Sierra Madre Oriental - Sur ............................................................................................. 51Eje Neovolcanico ............................................................................................................ 51

ii


C. DEFINITIONS OF FOREST COC. DEFINITIONS OF FOREST COC. DEFINITIONS OF FOREST COC. DEFINITIONS OF FOREST COC. DEFINITIONS OF FOREST COVER VER VER VER VER TTTTTYPYPYPYPYPESESESESES ................................................................................................................................................................................................................................................................................................................. 5252525252Literature Review................................................................................................................ 52

Ponderosa Pine ............................................................................................................... 53Mixed-conifer ................................................................................................................. 53Spruce-fir ....................................................................................................................... 54Other Forest Types of Interest ......................................................................................... 54

Chihuahua Pine .......................................................................................................... 54Quaking Aspen ........................................................................................................... 55Riparian Forests .......................................................................................................... 55

Plan Definitions ................................................................................................................. 55Ponderosa Pine Forest ..................................................................................................... 55Pine-oak Forest ............................................................................................................... 55Mixed-conifer Forest ....................................................................................................... 56High-elevation Forests, Including Spruce-fir Forest ......................................................... 57Quaking Aspen ............................................................................................................... 57

Key to Forest Types............................................................................................................. 57D. CONCEPTUD. CONCEPTUD. CONCEPTUD. CONCEPTUD. CONCEPTUAL FRAMEAL FRAMEAL FRAMEAL FRAMEAL FRAMEWWWWWORK FOR RECOORK FOR RECOORK FOR RECOORK FOR RECOORK FOR RECOVERVERVERVERVERYYYYY.................................................................................................................................................................................................................................................................... 5959595959

Ecosystem or Landscape Management ................................................................................ 59Fire ................................................................................................................................. 60Other Natural Disturbances ............................................................................................ 61Degradation of Riparian Forests ...................................................................................... 65

Timber Harvest and Silvicultural Practices .......................................................................... 65Historical Perspectives .................................................................................................... 65

Past Practices .............................................................................................................. 65Forest Plans ................................................................................................................ 66Habitat Trends ............................................................................................................ 66

Data Availability ..................................................................................................... 66Trends in Forest Land Base and Timber Volume ...................................................... 67Trends in Forest Types ............................................................................................ 67Trends in Size-class Distibutions ............................................................................. 68Summary of Recent Habitat Trends ........................................................................ 68

Silvicultural Practices and Forest Management ................................................................ 68Silviculture ................................................................................................................. 69

Even-aged Management .......................................................................................... 69Uneven-aged Management...................................................................................... 70Development and Maintenance of Stratified Mixtures ............................................. 71

Conclusions .................................................................................................................... 71Grazing .............................................................................................................................. 72Recreation .......................................................................................................................... 73

Types of Recreation ........................................................................................................ 73Camping .................................................................................................................... 74Hiking ....................................................................................................................... 74Off-road Vehicles ....................................................................................................... 74Rock Climbing ........................................................................................................ . 74Wildlife Viewing and Photography ........................................................................... . 74Recreation Summary .................................................................................................. .74

Summary ........................................................................................................................... .75

iii


PPPPPARARARARART III: RECOT III: RECOT III: RECOT III: RECOT III: RECOVERVERVERVERVERYYYYY

A. DELISTINGA. DELISTINGA. DELISTINGA. DELISTINGA. DELISTING .................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................... 7676767676The Delisting Process ........................................................................................................ 76Delisting Criteria ............................................................................................................... 76

Monitoring Population Trends ....................................................................................... 77Other Considerations of Population Monitoring ........................................................ 79

Monitoring Habitat Trends ............................................................................................ 79Long-term Management Plan ........................................................................................ 80Delisting at the Recovery Unit Level .............................................................................. 80

Moderating and Regulating Threats. .......................................................................... 81Habitat Trends Within Recovery Units ....................................................................... 81

B. GENERAL APPRB. GENERAL APPRB. GENERAL APPRB. GENERAL APPRB. GENERAL APPROOOOOAAAAACHCHCHCHCH ................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................ 8282828282Assumptions and Guiding Principles.................................................................................. 82

Assumptions. ................................................................................................................. 83Guiding Principles ......................................................................................................... 84

General Recommendations ................................................................................................ 84Protected Areas .............................................................................................................. 84

Protected Activity Center (PAC) ................................................................................ 84Guidelines ............................................................................................................. 84Rationale ............................................................................................................... 89

Steep Slopes (Outside of PACs) .................................................................................. 89Guidelines ............................................................................................................. 89Rationale ............................................................................................................... 89

Reserved Lands .......................................................................................................... 90Guidelines ............................................................................................................. 90Rationale ............................................................................................................... 90

Restricted Areas ............................................................................................................. 90Existing Conditions ................................................................................................... 91Reference Conditions ................................................................................................. 91

Nesting and Roosting target/threshold conditions .................................................. 91Coarse Filter .............................................................................................................. 93Fine Filter .................................................................................................................. 94

Overriding Guidelines ........................................................................................... 94Specific Guidelines ................................................................................................ 94Rationale ............................................................................................................... 95

Riparian Communities ............................................................................................... 95Guidelines ............................................................................................................. 95Rationale ............................................................................................................... 95

Other Forest and Woodland Types ................................................................................. 95Grazing Recommendations ................................................................................................ 96

Grazing Guidelines ........................................................................................................ 96Rationale for Grazing Guidelines ................................................................................... 97

Recreation Recommendations............................................................................................ 98Guidelines ..................................................................................................................... 98

Recovery Unit Considerations ........................................................................................... 98Colorado Plateau ........................................................................................................... 98

Potential Threats ........................................................................................................ 99Southern Rocky Mountains - Colorado.......................................................................... 99

iv


Potential Threats ........................................................................................................ 99Southern Rocky Mountains - New Mexico..................................................................... 99

Potential Threats ....................................................................................................... 100Upper Gila Mountains .................................................................................................. 100

Potential Threats ....................................................................................................... 100Basin and Range - West ................................................................................................ 101

Potential Threats ....................................................................................................... 101Basin and Range - East ................................................................................................. 102

Potential Threats ....................................................................................................... 102Mexico ......................................................................................................................... 103

Potential Threats ....................................................................................................... 104Sierra Madre Occidental - Norte ........................................................................... 104Sierra Madre Oriental - Norte ............................................................................... 104Sierra Madre Occidental - Sur ............................................................................... 104Sierra Madre Oriental - Sur .................................................................................. 104Eje Neovolcanico .................................................................................................. 104

C. MONITC. MONITC. MONITC. MONITC. MONITORING PRORING PRORING PRORING PRORING PROCEDURESOCEDURESOCEDURESOCEDURESOCEDURES..................................................................................................................................................................................................................................................................................................................................................................................................................... 105105105105105Habitat Monitoring .......................................................................................................... 105

Macrohabitat ................................................................................................................ 105Microhabitat ................................................................................................................ 106

Population Monitoring ..................................................................................................... 107Target Population ......................................................................................................... 107Sampling Units ............................................................................................................. 107Sampling Procedures..................................................................................................... 108

Stratification ............................................................................................................. 108Selection of Quadrats ................................................................................................ 108Sampling Within Quadrats ....................................................................................... 108Banding Birds ........................................................................................................... 109

Statistical Analysis ......................................................................................................... 109Estimation of Capture Probability ............................................................................. 109Estimation of Density Per Quadrat ............................................................................ 110Estimation of Density Per Stratum ............................................................................ 110Variance Estimators ................................................................................................... 111Cormack-Jolly-Seber Models ..................................................................................... 111

Personnel ...................................................................................................................... 111Training .................................................................................................................... 112

Computerized Data Entry and Summarization.............................................................. 112Costs ............................................................................................................................ 113Potential Experiments ................................................................................................... 113Alternative Designs for Population Monitoring ............................................................. 114

Drawing New Samples of Quadrats Each Year ........................................................... 114Conducting Surveys Less Often than Yearly ............................................................... 115Adaptive Sampling .................................................................................................... 115

Conclusions ..................................................................................................................... 115D. AD. AD. AD. AD. ACTIVITCTIVITCTIVITCTIVITCTIVITYYYYY-SP-SP-SP-SP-SPECIFIC RESEARECIFIC RESEARECIFIC RESEARECIFIC RESEARECIFIC RESEARCHCHCHCHCH ............................................................................................................................................................................................................................................................................................................................................................................................ 116116116116116

Role of the Scientific Process ............................................................................................ 116Important Ascpects of Experimental Design ...................................................................... 117Limitations in Past Research on the Mexican Spotted Owl ................................................ 118Research Needs ................................................................................................................. 118

Dispersal ...................................................................................................................... 119

v


Genetics. .......................................................................................................................... 120Habitat ............................................................................................................................. 120Population Biology ........................................................................................................... 120Threats to Recovery .......................................................................................................... 120Other Ecosystem Components ......................................................................................... 120

E. SUMMARE. SUMMARE. SUMMARE. SUMMARE. SUMMARY OF RECOY OF RECOY OF RECOY OF RECOY OF RECOVERVERVERVERVERYYYYY ........................................................................................................................................................................................................................................................................................................................................................................................................................................................ 121121121121121

PPPPPARARARARART IVT IVT IVT IVT IV: RECO: RECO: RECO: RECO: RECOVERVERVERVERVERY PLY PLY PLY PLY PLAN IMPLEMENTAN IMPLEMENTAN IMPLEMENTAN IMPLEMENTAN IMPLEMENTAAAAATIONTIONTIONTIONTION

A. IMPLEMENTING LA. IMPLEMENTING LA. IMPLEMENTING LA. IMPLEMENTING LA. IMPLEMENTING LAAAAAWS, REGULWS, REGULWS, REGULWS, REGULWS, REGULAAAAATIONS, AND ATIONS, AND ATIONS, AND ATIONS, AND ATIONS, AND AUTHORITIESUTHORITIESUTHORITIESUTHORITIESUTHORITIES .............................................................................................................. 124124124124124Endangered Species Act .................................................................................................... 124

Section 4 ...................................................................................................................... 124Section 5 ...................................................................................................................... 125Section 6 ...................................................................................................................... 125Section 7 ...................................................................................................................... 125Section 8 ...................................................................................................................... 126Section 9 ...................................................................................................................... 126Section 10 .................................................................................................................... 126

National Forest Management Act ...................................................................................... 126National Environmental Policy Act ................................................................................... 128Migratory Bird Treaty Act ................................................................................................. 128Tribal Lands ..................................................................................................................... 128State and Private Lands ..................................................................................................... 128Mexico ............................................................................................................................. 128

B. IMPLEMENTB. IMPLEMENTB. IMPLEMENTB. IMPLEMENTB. IMPLEMENTAAAAATION OTION OTION OTION OTION OVERSIGHTVERSIGHTVERSIGHTVERSIGHTVERSIGHT ....................................................................................................................................................................................................................................................................................................................................................................................... 129129129129129Recovery Unit Working Teams.......................................................................................... 129Continuing Duties of the Recovery Team ......................................................................... 120Centralized Spotted Owl Information Repository ............................................................. 130

C. STEPDOC. STEPDOC. STEPDOC. STEPDOC. STEPDOWN OUTLINEWN OUTLINEWN OUTLINEWN OUTLINEWN OUTLINE ...................................................................................................................................................................................................................................................................................................................................................................................................................................................................................... 131131131131131D.D.D.D.D. IMPLEMENTIMPLEMENTIMPLEMENTIMPLEMENTIMPLEMENTAAAAATION AND COST SCHEDULETION AND COST SCHEDULETION AND COST SCHEDULETION AND COST SCHEDULETION AND COST SCHEDULE ................................................................................................................................................................................................................................................................................... 137137137137137

GLGLGLGLGLOSSAROSSAROSSAROSSAROSSARYYYYY ......................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................... 145145145145145

LITERALITERALITERALITERALITERATURE CITEDTURE CITEDTURE CITEDTURE CITEDTURE CITED ............................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................... 152152152152152

APPAPPAPPAPPAPPENDIX A: ENDIX A: ENDIX A: ENDIX A: ENDIX A: The RThe RThe RThe RThe Recoecoecoecoecovvvvvererererery y y y y TTTTTeam and Associateseam and Associateseam and Associateseam and Associateseam and Associates ................................................................................................................................................................................................................................................................................... 163163163163163

APPAPPAPPAPPAPPENDIX B: Schedule of MENDIX B: Schedule of MENDIX B: Schedule of MENDIX B: Schedule of MENDIX B: Schedule of Meetingseetingseetingseetingseetings ................................................................................................................................................................................................................................................................................................................................................................................................. 165165165165165

APPAPPAPPAPPAPPENDIX C: Schedule of FENDIX C: Schedule of FENDIX C: Schedule of FENDIX C: Schedule of FENDIX C: Schedule of Field ield ield ield ield VisitsVisitsVisitsVisitsVisits ............................................................................................................................................................................................................................................................................................................................................................................. 166166166166166

APPAPPAPPAPPAPPENDIX D: List of EENDIX D: List of EENDIX D: List of EENDIX D: List of EENDIX D: List of English and Latin Nnglish and Latin Nnglish and Latin Nnglish and Latin Nnglish and Latin Namesamesamesamesames .................................................................................................................................................................................................................................................................................................. 167167167167167

APPAPPAPPAPPAPPENDIX E: Agencies and PENDIX E: Agencies and PENDIX E: Agencies and PENDIX E: Agencies and PENDIX E: Agencies and Persons Commenting on Dersons Commenting on Dersons Commenting on Dersons Commenting on Dersons Commenting on Draft Rraft Rraft Rraft Rraft Recoecoecoecoecovvvvvererererery Py Py Py Py Planlanlanlanlan ........................................................................... 170170170170170

vi


List of TablesList of TablesList of TablesList of TablesList of Tables

TTTTTable ES.1 able ES.1 able ES.1 able ES.1 able ES.1 Overview of management categories by vegetation type for lands notadministratively reserved................................................................................................................. xi

TTTTTable I.D.1able I.D.1able I.D.1able I.D.1able I.D.1 Summary of relative Mexican spotted owl population size, timber harvest threat, and firethreat by Recovery Unit .................................................................................................................. 12

TTTTTable II.A.1.able II.A.1.able II.A.1.able II.A.1.able II.A.1. Historical records and minimum numbers of Mexican spotted owlsfound during planned surveys and incidental observations, reported by Recovery Unitand land ownership .................................................................................................................. 23-24

TTTTTable II.B.1.able II.B.1.able II.B.1.able II.B.1.able II.B.1. Land-ownership patterns (thousands of hectares) in Recovery Unitswithin the United States ................................................................................................................. 41

TTTTTable II.B.2.able II.B.2.able II.B.2.able II.B.2.able II.B.2. Land-ownership patterns (thousands of hectares) in Recovery Unitswithin Mexico ................................................................................................................................ 50

TTTTTable II.D.1.able II.D.1.able II.D.1.able II.D.1.able II.D.1. Changes in the area (ha x 1,000 [ac x 1,000]) and distribution offorest types from the 1960s to 1980s on commercial forest lands within Arizonaand New Mexico ............................................................................................................................ 67

TTTTTable II.D.2.able II.D.2.able II.D.2.able II.D.2.able II.D.2. Changes in the density (trees/ha [trees/ac]) and distribution oftree size classes from the 1960s to 1980s on commercial forest lands within Arizonaand New Mexico ............................................................................................................................ 68

TTTTTable III.B.1able III.B.1able III.B.1able III.B.1able III.B.1 Target/threshold conditions for mixed-conifer and pine-oak forestswithin restricted areas ..................................................................................................................... 92

TTTTTable IVable IVable IVable IVable IV.1.1.1.1.1 Implementation and cost schedule ..................................................................... 139-144

vii


List of FiguresList of FiguresList of FiguresList of FiguresList of Figures

FFFFFigurigurigurigurigure II.A.1.e II.A.1.e II.A.1.e II.A.1.e II.A.1. Geographic range of the spotted owl ...................................................................... 20

FFFFFigurigurigurigurigure II.A.2.e II.A.2.e II.A.2.e II.A.2.e II.A.2. Current distribution of Mexican spotted owls in the United States based on planned surveys and incidental observations recorded from 1990 through 1993............... 22

FFFFFigurigurigurigurigure II.A.3. e II.A.3. e II.A.3. e II.A.3. e II.A.3. Current distribution of Mexican spotted owls in Mexico based onplanned surveys and incidental observations recorded from 1990 through 1993 .............................. 25

FFFFFigurigurigurigurigure II.A.4.e II.A.4.e II.A.4.e II.A.4.e II.A.4. Geographic variability in the food habits of Mexican spotted owlsrepresented as relative frequencies of (a)(a)(a)(a)(a) woodrats, (b)(b)(b)(b)(b) voles, (c)(c)(c)(c)(c) gophers, (d)(d)(d)(d)(d) birds,(e)(e)(e)(e)(e) bats, and (f(f(f(f(f ) ) ) ) ) reptiles .................................................................................................................. 30

FFFFFigurigurigurigurigure II.B.1.e II.B.1.e II.B.1.e II.B.1.e II.B.1. Recovery Units within the United States ................................................................. 37

FFFFFigurigurigurigurigure II.B.2.e II.B.2.e II.B.2.e II.B.2.e II.B.2. Recovery Units within the Republic of Mexico ........................................................ 38

FFFFFigurigurigurigurigure II.B.3e II.B.3e II.B.3e II.B.3e II.B.3 Colorado Plateau Recovery Unit .............................................................................. 39

FFFFFigurigurigurigurigure II.B.4e II.B.4e II.B.4e II.B.4e II.B.4 Southern Rocky Mountains - Colorado Recovery Unit ............................................ 42

FFFFFigurigurigurigurigure II.B.5e II.B.5e II.B.5e II.B.5e II.B.5 Southern Rocky Mountains - New Mexico Recovery Unit ....................................... 43

FFFFFigurigurigurigurigure II.B.6e II.B.6e II.B.6e II.B.6e II.B.6 Upper Gila Mountains Recovery Unit ..................................................................... 45

FFFFFigurigurigurigurigure II.B.7e II.B.7e II.B.7e II.B.7e II.B.7 Basin and Range - West Recovery Unit .................................................................... 47

FFFFFigurigurigurigurigure II.B.8e II.B.8e II.B.8e II.B.8e II.B.8 Basin and Range - East Recovery Unit ..................................................................... 48

FFFFFigurigurigurigurigure II.D.1e II.D.1e II.D.1e II.D.1e II.D.1 Historical record of number of total fires and number ofnatural fires. Data from USFS Southwestern Region ...................................................................... 62

FFFFFigurigurigurigurigure II.D.2e II.D.2e II.D.2e II.D.2e II.D.2 Historical record of fires over four hectares in size. Data fromUSFS Southwestern Region ............................................................................................................ 62

FFFFFigurigurigurigurigure II.D.3e II.D.3e II.D.3e II.D.3e II.D.3 Five-year running averages of area burned. Data from USFSSouthwestern Region ...................................................................................................................... 63

FFFFFigurigurigurigurigure III.A.1e III.A.1e III.A.1e III.A.1e III.A.1 Hypothetical curve of the statistical power to detect a trend in a population ........... 78

FFFFFigurigurigurigurigure III.B.1e III.B.1e III.B.1e III.B.1e III.B.1 Conceptualization of the Recovery Plan and needs for delistingof the Mexican spotted owl depicting the interdependency of populationmonitoring, habitat monitoring, and management recommendations ............................................. 83

FFFFFigurigurigurigurigure III.B.2e III.B.2e III.B.2e III.B.2e III.B.2 Generalization of protection stratgies by forest/vegetation type ............................... 85

FFFFFigurigurigurigurigure III.B.3e III.B.3e III.B.3e III.B.3e III.B.3 Examples of protected activity center (PAC) boundaries fromthe Lincoln National Forest ............................................................................................................ 87

viii


Volume 1/Executive Summary

EXECUTIVE SUMMARYEXECUTIVE SUMMARYEXECUTIVE SUMMARYEXECUTIVE SUMMARYEXECUTIVE SUMMARY

INTRODUCTIONINTRODUCTIONINTRODUCTIONINTRODUCTIONINTRODUCTION

The Mexican spotted owl was listed as athreatened species on 15 April 1993. Twoprimary reasons were cited for the listing: his-torical alteration of its habitat as the result oftimber management practices, specifically theuse of even-aged silviculture, plus the threat ofthese practices continuing, as provided in Na-tional Forest Plans. The danger of catastrophicwildfire was also cited as a potential threat foradditional habitat loss. Concomitant with thelisting of the Mexican spotted owl, a RecoveryTeam was appointed by FWS SouthwesternRegional Director John Rogers to develop aRecovery Plan. This report constitutes theRecovery Plan for the Mexican spotted owl.

This Recovery Plan provides a basis formanagement actions to be undertaken by land-management agencies and Indian Tribes toremove recognized threats and recover thespotted owl. Primary actions will be taken by theUSDA Forest Service, USDI Bureau of LandManagement, USDI Fish and Wildlife Service,USDI Bureau of Indian Affairs, and sovereignAmerican Indian Tribes. The Fish and WildlifeService will oversee implementation of theRecovery Plan through its authorities under theEndangered Species Act.

The Team made every effort to identify andconsider all sources of information in developingthis plan. Previous plans developed for thenorthern spotted owl (Thomas et al. 1990, Bartet al. 1992) and the California spotted owl(Verner et al. 1992) were considered in thedevelopment of this Recovery Plan. The Teamanalyzed data that had not been evaluatedpreviously and re-analyzed data when appropri-ate to ensure that information was consistent orto address questions not considered in previousanalyses of those data.

RECOVERY GOALRECOVERY GOALRECOVERY GOALRECOVERY GOALRECOVERY GOAL

The purpose of this Recovery Plan is tooutline the steps necessary to remove the Mexi-

can spotted owl from the list of threatenedspecies.

THE RECOVERY PLANTHE RECOVERY PLANTHE RECOVERY PLANTHE RECOVERY PLANTHE RECOVERY PLAN

The Recovery Plan contains five basicelements:

1. A recovery goal and a set of delistingcriteria that, when met, will allow theMexican spotted owl to be removed fromthe list of threatened species.

2. Provision of three general strategies formanagement that provide varying levelsof habitat protection depending on theowl’s needs and habitat use.

3. Recommendations for population andhabitat monitoring.

4. A research program to address criticalinformation needs to better understandthe biology of the Mexican spotted owland the effects of anthropogenic activi-ties on the owl and its habitat.

5. Implementation procedures thatspecify oversight and coordinationresponsibilities.

Each of these elements is described brieflybelow.

Delisting CriteriaDelisting CriteriaDelisting CriteriaDelisting CriteriaDelisting Criteria

The primary threat to the Mexican spottedowl leading to its listing as a threatened specieswas the alteration of its habitat in Arizona andNew Mexico as the result of timber manage-ment, specifically even-aged management.Mexican spotted owls use a variety of habitats,but are typically associated with multi-canopiedstands of mature mixed-conifer and ponderosapine-Gambel oak forests. Past logging usingeven-aged shelterwood prescriptions that in-cluded short rotations and the removal of large

ix



volumes of timber greatly simplified standstructures which adversely affected >300,000 ha(800,000 ac) of spotted owl habitat (Fletcher1990). Existing Forest Plans call for continueduse of shelterwood harvests, potentially leadingto continued loss of owl habitat. However, theTeam recognizes and is encouraged by recentefforts to amend existing Forest Plans to de-emphasize the use of even-aged silviculture andincorporate the management guidelines providedwithin this Recovery Plan.

The range of the Mexican spotted owl wasdivided into six Recovery Units in the UnitedStates and five in Mexico. Recovery Units werebased on various factors including biotic prov-inces, the spotted owl’s ecology, and manage-ment considerations. If delisting criteria are met,Recovery Units can be delisted separately. Thefollowing criteria must be met for delisting to beconsidered: (1) the population in the three mostpopulated Recovery Units must be stable orincreasing after 10 years of monitoring; (2)scientifically-valid habitat monitoring protocolsare designed and implemented to assess (a) grosschanges in habitat quantity across the range ofthe Mexican spotted owl, and (b) habitat modifi-cations and habitat trajectories within treatedstands; and (3) a long-term management plan isin place to ensure appropriate management forthe spotted owl and its habitat. If these threecriteria are met, then the Mexican spotted owlcan be delisted within any Recovery Unit ifthreats have been moderated or regulated, and ifhabitat trends are stable or increasing.

Levels of ProtectionLevels of ProtectionLevels of ProtectionLevels of ProtectionLevels of Protection

General recommendations are proposed forthree levels of management: protected areas,restricted areas, and other forest and woodlandtypes (Table ES.1). Protected areas include a 243ha (600 ac) “Protected Activity Center” (PAC)placed at known or historical nest and/or roostsites, slopes >40% in mixed-conifer and pine-oakforests that have not been harvested within thepast 20 years, and administratively reservedlands. Harvest of trees >22.4 cm (9 inches) dbh(diameter at breast height) is not allowed withinprotected areas, but light underburning ispermitted on a case-specific basis as needed to

reduce fuels. Also, a fire risk-abatement programis proposed to allow the treatments of fuels usinga combination of fuel removal and fire. Thismanagement can be conducted initially within10% of the PACs, after which time the effective-ness of the program should be evaluated. Similarmanagement can be conducted on steep slopes,but with no areal restrictions.

Restricted areas include ponderosa pine-Gambel oak and mixed-conifer forests andriparian environments. Target/threshold criteriaare provided to define the proportion of thelandscape that should be in or approachingconditions suitable for nesting and roosting. Theremainder of the landscape should be managedin such a way to allocate stands to ensure asustained provision of nest and roost habitatthrough time. Broad guidelines for ripariansystems emphasize the maintenance and restora-tion of riparian areas to ensure a mix of size andage classes.

Other forest and woodland types includeponderosa pine and spruce-fir forests, pinyon-juniper woodlands, and aspen groves that are notincluded within PACs. No specific guidelines areproposed, but general recommendations aregiven to manage these areas for landscape diver-sity within natural ranges of variation.

Population and Habitat MonitoringPopulation and Habitat MonitoringPopulation and Habitat MonitoringPopulation and Habitat MonitoringPopulation and Habitat Monitoring

The Recovery Plan provides a detailedprogram to monitor spotted owl populationsand habitats. Both are key components of thedelisting criteria. Population monitoring isrestricted to the three most populated RecoveryUnits because their spotted owl populationsmeet sample size criteria for the monitoringdesign. Further, these Recovery Units comprisethe core Mexican spotted owl population andthe Team assumes that their population statusreflects that of the entire population. The designpresented in the Recovery Plan entails the use ofmark-recapture methodology on random quad-rats to estimate key population parameters. Theobjectives of the habitat monitoring are (a) totrack gross changes in habitat quality and quan-tity using remote sensing technology, and (b) toevaluate whether treatments meet the desiredgoal of setting stands on trajectories to becomereplacement habitat.

x



TTTTTable ES.1.able ES.1.able ES.1.able ES.1.able ES.1. Overview of management categories by vegetation type for lands notadministratively reserved.

TTTTTimber Himber Himber Himber Himber HarararararvvvvvestestestestestVVVVVegetationegetationegetationegetationegetation IIIIIn Nn Nn Nn Nn Nestestestestest S S S S Slopelopelopelopelope WWWWWithin Pithin Pithin Pithin Pithin Pastastastastast MMMMManagementanagementanagementanagementanagement TTTTTypeypeypeypeype ArArArArArea?ea?ea?ea?ea?11111 >>>>> 40% 40% 40% 40% 40% 20 20 20 20 20 YYYYYears?ears?ears?ears?ears? CCCCCategorategorategorategorategoryyyyy

Any Yes ProtectedMixed-conifer No Yes No ProtectedPine-Oak No Yes No ProtectedMixed-conifer No Yes Yes RestrictedPine-Oak No Yes Yes RestrictedMixed-conifer No No RestrictedPine-Oak No No RestrictedRiparian No No RestrictedPonderosa Pine No No Other typesSpruce-Fir No No Other typesPinyon-Juniper No No Other typesAspen No No Other typesOak No No Other types

1Refers to land contained within a protected activity center.

Activity-specific Research ProgramActivity-specific Research ProgramActivity-specific Research ProgramActivity-specific Research ProgramActivity-specific Research Program

The Recovery Team made extensive use ofscientific data. During the process of gatheringand evaluating these data, it became evident thatadditional information was needed to refine therecovery measures. Past research efforts empha-sized inductive approaches to gather informationon basic life history needs of the spotted owl.Although these research efforts provided somekey information, more rigorous and directedapproaches will be needed to address questionson dispersal, genetics, habitat, populations, andeffect of management on spotted owls and otherecosystem attributes.

Implementation MeasuresImplementation MeasuresImplementation MeasuresImplementation MeasuresImplementation Measures

Recovery Plans are not self-implementingunder the Endangered Species Act. Thus, animplementation schedule is provided thatoutlines steps needed for the execution of therecovery measures. These implementationguidelines include the formation of an inter-agency working team for each Recovery Unit tooversee implementation of the recovery measuresthat encompass four broad areas: resource

management programs, active managementactions, monitoring, and research.

CONCLUSIONCONCLUSIONCONCLUSIONCONCLUSIONCONCLUSION

The Recovery Plan is based largely on finaland preliminary results of field studies of spottedowl habitat use, population biology, and distri-bution. The Team relied on information pub-lished in the both scientific and “gray” literature.If data were available but unanalyzed, the Teammade every reasonable effort to conduct thoseanalyses. Reanalyses of data were conductedwhen the Team wished to address questions notaddressed by those who collected the data. Thus,this Recovery Plan represents the current state-of-knowledge on the Mexican spotted owl.

The Recovery Plan recommendations are acombination of (1) protection of both occupiedhabitats and unoccupied areas approachingcharacteristics of nesting habitat, and (2) imple-mentation of ecosystem management withinunoccupied but potential habitat. The goal is toprotect conditions and structures used by spot-ted owls where they exist and set other stands ona trajectory to grow into replacement nesthabitat or to provide conditions for foraging and

xi



dispersal. By necessity this Plan is a hybridapproach because the status of the Mexicanspotted owl as a threatened species requires somelevel of protection until the subspecies isdelisted. These constraints modify ways andopportunities to manage ecosystems withinlandscapes where owls occur or might occur inthe future. The Plan advocates applying ecosys-tem management in two slightly different ways.Within unoccupied mixed-conifer and pine-oakforest on <40 % slope, we provide both general(coarse filter) and specific (fine filter) guidelinesto provide a sustainable quantity of replacementnest habitat across the landscape. Within otherunoccupied forest and woodland types (e.g.,ponderosa pine, spruce-fir, aspen, and pinyon-juniper), general guidance is provided for man-aging the landscape to meet multiple resourceobjectives including spotted owl foraging anddispersal habitat.

Management priority should focus onactions to alleviate threats to Mexican spottedowls; thereafter, or in coordination with alleviat-ing threats, other management priorities (e.g.,creating replacement owl habitat) should bepursued. Two primary threats that should be thefocus of such management priorities are cata-strophic wildfire and widespread use of even-aged silviculture.

Heavy accumulations of ground and ladderfuels have rendered many Southwestern forestsvulnerable to stand-replacing fires. Such firesrepresent a real and immediate threat to theexistence of spotted owl habitat. The manage-ment guidelines are intended to provide landmanagers with flexibility to reduce these fuellevels and abate fire risks. Fire managementshould be given the highest priority.

Even-aged silviculture within potential owlhabitat is regarded as a threat because it tends tosimplify stand structure and move stands awayfrom containing owl habitat characteristics. TheTeam recognizes, however, that such regenera-tion cuts may provide useful tools in specialcircumstances to manage for spotted owls andother ecosystem objectives. Any use of even-agedmanagement should be done sparingly and onlyafter careful deliberation to ensure that it repre-sents the best approach to meet managementobjectives.

Under proposed delisting criteria the owlcould be delisted within 10 years, rendering theprotection measures in this Recovery Planobsolete. At that time, sufficient knowledgeshould be available to design a strategy for long-term conservation of the Mexican spotted owl.Many of the ecosystem management guidelinesprovided in this Plan will provide a foundationfor development of the long-term strategy. Informulating the recommendations, the Teamassumes that population and habitat status willbe monitored in conjunction with implementa-tion of these management guidelines. Therefore,the management guidelines are not meant tostand alone. Monitoring provides objectivecriteria to assess the efficacies of the managementguidelines. Without both habitat and populationmonitoring, the status of the owl cannot beassessed and it should not be delisted. We furtherassume that existing management constraints onvegetative manipulations (such as size of open-ings and maintenance of hiding and thermalcover for other species) will remain in place. Thisassumption is especially critical for vegetationtypes--ponderosa pine, pinyon-juniper, aspen,and spruce-fir--for which we provide no specificmanagement recommendations.

The Recovery Plan presents realistic goals forrecovery of the species and its ultimate delisting.The goals are flexible in that they require localland managers to make site-specific decisions.Success of the plan, however, hinges on thecommitment and coordination among thevarious Federal and State land-managementagencies, sovereign Indian nations, and theprivate sector to ensure that the plan is followedand executed as intended by the Team.

xii


ACKNOWLEDGMENTSACKNOWLEDGMENTSACKNOWLEDGMENTSACKNOWLEDGMENTSACKNOWLEDGMENTS

Many people helped in the completion of this Plan. A list of people who commented on the draftRecovery Plan is provided in Appendix E. In addition, we thank everyone who contributed valuabletime and information to the Recovery Team, and apologize for any omissions in our special thanks tothe following:

SSSSSupporupporupporupporupport of the teamt of the teamt of the teamt of the teamt of the team: Nancy M. Kaufman, John G. Rogers, Jim Young, Susan MacMullin, JamieRappaport-Clark (Fish and Wildlife Service [FWS] Region 2); Denver Burns (Rocky Mountain Station[RMS], Colorado); Charles W. Cartwright, Jim Lloyd, John Kirkpatrick, Larry Henson (Forest Service[FS] Region 3).

PPPPPrrrrroooooviding data to the teamviding data to the teamviding data to the teamviding data to the teamviding data to the team: Craig Allen, Mike Britten, Dick Braun, Larry Henderson, ClivePinnock (National Park Service [NPS]); George Alter, Cheryl Carothers, Larry Cordova, Larry Cosper,Katherin Farr, Keith Fletcher, Heather Hollis, Patrick D. Jackson, Mike Manthei, Renee Galeano-Popp, Danney Salas, John Shafer, Dennis Watson (FS); Mike Bogan, Cindy Ramotnik (NationalBiological Service, [NBS]); Erik Brekke (Bureau of Land Management, [BLM]); Russell Duncan (SWField Biologists); R.J. Gutiérrez, David Olson, Mark Seamans (Humboldt State Univ. [HSU]); CharlesJohnson, Richard Reynolds (RMS); Susan Eggen-McIntosh (Southern Station New Orleans); PatMehlhop (Natural Heritage Program, NM); Hildy Reiser (Holloman AFB, NM); Steve Speich (Damesand Moore Consultants); Roger Skaggs (Glenwood, NM); Luis Tarango (Colegio De Postgraduados,San Luis Potosi, Mexico); Jim Tress (SWCA Inc.); Dave Willey (N. Arizona Univ. [NAU]); RickWinslow (Navajo Fish and Wildl. Dept.).

Consultation and PConsultation and PConsultation and PConsultation and PConsultation and Prrrrresentationsesentationsesentationsesentationsesentations: Cristen Allen, Don Reimer (DR Systems); Bruce Anderson,Douglas A. Boyce, Paul Boucher, Cecelia Dargan, Don DeLorenzo, Jim Ellenwood, Leon Fager, MaryLou Fairweather, Leon Fisher, Keith Fletcher, Heather Hollis, Marlin Johnson, Milo Larson, BobLeaverton, Mike Leonard, Jim Lloyd, Mike Manthei, Chris Mehling, Terry Myers, Danney Salas, SteveServis, George Sheppard, Tom Skinner, Josh Taiz, Rich Teck, Borys Tkacz, Jon Verner, Dennis Watson(FS); Bill Austin, Steve Chambers, George Devine, Jennifer Fowler-Propst, Marcos Gorreson, SonjaJahrsdoerfer, Michele James, Sue Linner, Britta Muiznieks, Carol Torrez (FWS); George Barrowclough(Am. Museum of Natural Hist.); Ernie Beil, Sheridan Stone (Ft. Huachuca, AZ); Erik Brekke (BLM);Tina Carlson (Natural Heritage Program); Mario Cirett (Ecological Center, Mexico); Martha Coleman,Tanya Shenk (Colorado State Univ.); Gerald Craig (CO Division of Wildlife); Norris Dodd (AZ Gameand Fish Dept.); Russell Duncan (SW Field Biologists); Jeff Feen, Tim Wilhite (San Carlos Apache);Eric Forsman (FS Pacific NW Res. Stn.); Brian Geils, John Lundquist, Richard Reynolds (RMS); R.J.Gutiérrez, Dave Olson, Mark Seamans (HSU); Steve Haglund (Bureau of Indian Affairs); Craig Hauke(NPS); Tod Hull, Rich VanDemark (Applied Ecosystem Management Inc.); Terry Johnson (indepen-dent researcher); Joe Jojola (White Mtn. Apache); Norman Jojola (Mescalero Apache); CameronMartinez (Northern Pueblos); Marcelo Marquez, Luis Tarango (Colegio De Postgraduados, San LuisPotosi, Mexico); John McTague (NAU); Armando Cortez Ortiz, Jose Luis Omelasde Anda (NationalInstitute of Statistics and Geographic Info., Mexico); Peter Stacey (Univ. of Nevada); AlejandroVelazques (Silviculturist, Mexico); Jerry Verner (FS Pacific SW Res. Stn.); Bob Waltermire (FWS/NBS); Sartor O. Williams, III (NM Dept. Game & Fish); Kendall Young (New Mexico State. Univ.);Mary Price (Univ. Calif., Riverside); Carl Marti (Weber State College); Andy Carey (FS Pacific NWRes. Stn.).

xiii


xiv

Conducting FConducting FConducting FConducting FConducting Field ield ield ield ield TTTTTrrrrripsipsipsipsips: Bruce Anderson, Don DeLorenzo, Heather Green, Dave Nelson,Danney Salas, Steve Servis (FS); Russell Duncan (SW Field Biologists); Joe Jojola (White Mtn.Apache); San Carlos Apache Tribe, White Mountain Apache Tribe (providing the team access to theirlands); Steve Speich (Dames and Moore Consultants); Sheridan Stone (Ft. Huachuca, AZ); LuisTarango (Colegio De Postgraduados, San Luis Potosi, Mexico).

GGGGGatheratheratheratherathering and Analyzing Ding and Analyzing Ding and Analyzing Ding and Analyzing Ding and Analyzing Dataataataataata: Damon Brown, Charlotte Corbett, Laura Duncan, Jill Dwyer,Jeff Jenness, Rudy King, Sharon Masek, Mike Nelson, Janie Sandoval, Peter Scott, Wayne Shepperd,Doug Spaeth, Mike Stanghellini, Gerry Wildeman, Randy Wilson, Rick Winslow, Karen Yarnell(RMS); Martha Coleman, Kevin Kalin (CSU); Susan DeRosier, Earl Hill, Karen Holmstrom, RichTeck (FS); Susan Eggen-McIntosh (FS Southern Stn., New Orleans); Bob Waltermire, Barb White(FWS/BS); Dave Willey (NAU).

Logistical SLogistical SLogistical SLogistical SLogistical Supporupporupporupporupporttttt: Bekki King, Karen Montano (FWS); Kim Ross (RMS).


Volume I

Volume I, Part I

Recovery Plan Development


Volume I/Part I

peer review of either the entire document orportions treating subjects within their specificareas of expertise. A list of reviewers is providedin Appendix E.

Recovery plans are neither self-implementingnor legally binding. Rather, approved recoveryplans effectively constitute FWS policy on thatlisted species or group of species, thereby guidingthe Service in conducting various processesrequired under the Act, such as section 7 consul-tation, conservation planning under section 10,and other procedures. In most cases, recoveryplans are followed by other Federal agencies incompliance with the mandate under sections2(c)(1) and 7(a)(1) of the Act to utilize theirauthorities in carrying out programs for theconservation of endangered and threatenedspecies. In addition, State and local governmentsusually follow the recommendations of recoveryplans in their species conservation efforts.

Section 4(f)(1)(B) of the Act specifies thecontents of a recovery plan:

“(i) a description of such site-specificmanagement actions as may be neces-sary to achieve the plan’s goal for theconservation and survival of thespecies” (III.B);

“(ii) objective, measurable criteria which,when met, would result in adetermination...that the species beremoved from the list” (III.A);

“(iii) estimates of the time required and thecost to carry out those measuresneeded to achieve the plan’s goal andto achieve intermediate steps towardthat goal” (IV.C and IV.D).

A. RECOVERY PLANNINGA. RECOVERY PLANNINGA. RECOVERY PLANNINGA. RECOVERY PLANNINGA. RECOVERY PLANNING

The USDI Fish and Wildlife Service (FWS)added the Mexican spotted owl to the List ofThreatened and Endangered Wildlife (50 CFR17.11) as a threatened species, effective on 15April 1993. Section 4(f)(1) of the EndangeredSpecies Act of 1973, as amended (Act) (16U.S.C. 1531), requires the Secretary of theInterior (usually delegated to the Director of theFWS) to “...develop and implement (recovery)plans for the conservation of endangered speciesand threatened species...unless he finds that sucha plan will not promote the conservation of thespecies.”

To develop scientifically credible recoveryplans for listed species, the FWS may appointrecovery teams comprised of scientists andresource specialists with expertise either on thespecies being considered or with other relevantexpertise. In the case of the Mexican spottedowl, the FWS appointed the Mexican SpottedOwl Recovery Team (Recovery Team). A list ofRecovery Team members and their areas ofexpertise can be found in Appendix A. A chro-nology of Recovery Team activities is provided inAppendices B and C.

Recovery teams present recovery plans to theFWS as their recommendation on the stepsnecessary to remove a species from the List ofThreatened and Endangered Wildlife and Plants.Removal from the list, or “delisting,” means thespecies is no longer in need of protection underthe Act and is therefore considered “recovered.”If deemed acceptable, the Director of the FWSRegion assigned the lead for that species ap-proves the plan.

The FWS, pursuant to requirements undersection 4(f )(4) of the Act, published a Notice ofAvailability of the Draft Mexican Spotted OwlRecovery Plan in the Federal Register on March27, 1995 (60 FR 15787). In addition to thisgeneral solicitation for information and publiccomment, the FWS sent copies of the draftRecovery Plan to numerous Federal and Stateagencies, Indian Tribes, county governments,environmental and industry groups, and otherswho had expressed interest in the Mexicanspotted owl. Finally, specific professional organi-zations and individuals were asked to provide

1


Volume I/Part I

The Mexican spotted owl is one of threespotted owl subspecies (see II.A). Under section3 of the Act, the term “species” includes “...anysubspecies of fish or wildlife...”. Although theMexican spotted owl is a subspecies, it is some-times referred to as a “species” in this documentwhen discussed in the context of the Act or otherlaws and regulations. An “endangered species” isdefined under the Act as “...any species which isin danger of becoming extinct throughout all ora significant portion of its range....” A threat-ened species is one “...which is likely to becomean endangered species in the foreseeable futurethroughout all or a significant portion of itsrange.” Section 4 (A)(1) of the Act lists fivefactors that can, either singly or collectively,result in listing as endangered or threatened:

“(A) the present or threatened destruction,modification, or curtailment of itshabitat or range;

(B) overutilization for commercial, recre-ational, scientific, or educationalpurposes;

(C) disease or predation;

(D) the inadequacy of existing regulatorymechanisms;

(E) other natural or man-made factorsaffecting its continued existence.”

The final rule listing the Mexican spottedowl as a threatened species (final rule) (58 FR14248) provides a detailed discussion of theprimary factors (A and D) leading to the deter-mination of threatened status. It should be notedthat the Recovery Team summarizes the finalrule here for information purposes only. TheRecovery Team’s assessment of the currentsituation with regard to the subspecies’ statusand threats is reflected in Part III. The followingbriefly summarizes the factors leading to thespecies’ listing, as discussed in the final rule:

THE PRESENT OR THREATENEDTHE PRESENT OR THREATENEDTHE PRESENT OR THREATENEDTHE PRESENT OR THREATENEDTHE PRESENT OR THREATENEDDESTRUCTION, MODIFICATION,DESTRUCTION, MODIFICATION,DESTRUCTION, MODIFICATION,DESTRUCTION, MODIFICATION,DESTRUCTION, MODIFICATION,

OR CURTAILMENT OF ITSOR CURTAILMENT OF ITSOR CURTAILMENT OF ITSOR CURTAILMENT OF ITSOR CURTAILMENT OF ITSHABITAT OR RANGEHABITAT OR RANGEHABITAT OR RANGEHABITAT OR RANGEHABITAT OR RANGE

Past, current, and future timber-harvestpractices in the Southwestern Region (Region 3)of the USDA Forest Service (FS) were cited asthe primary factors leading to listing the Mexi-can spotted owl as a threatened species. The finalrule stated that the Southwestern Region of theFS managed timber primarily under ashelterwood harvest regime. This harvest methodproduces even-aged stands rather than theuneven-aged, multi-layered stands most oftenused by Mexican spotted owls for nesting androosting. In addition, the shelterwood silvicul-tural system calls for even-aged conditions inperpetuity. Thus, stands already changed from“suitable” to “capable” would not be allowed toreturn to a “suitable” condition; and acreageslated for future harvest will be similarly ren-dered perpetually unsuitable for Mexican spottedowl nesting and roosting.

The final rule stated that “...significantportions of spotted owl habitat have been lost ormodified,” and cited Fletcher (1990) in estimat-ing that 420,000 ha (1,037,000 ac) of habitatwere converted from “suitable” to “capable.” Ofthis, about 78.7%, or 330,000 ha (816,000 ac),was a result of human activities, whereas theremainder was converted naturally, primarily bywildfire. According to the final rule, forest plansin the FS Region 3 allowed for up to 95% ofcommercial forest (59% of suitable spotted owlhabitat) to be managed under a shelterwoodsystem. The loss of lower- and middle-levelriparian habitat plus habitat lost to recreationdevelopments were also cited in the final rule asfactors in habitat loss.

B. LISTINGB. LISTINGB. LISTINGB. LISTINGB. LISTING

2


Volume I/Part I

OVERUTILIZATION FOROVERUTILIZATION FOROVERUTILIZATION FOROVERUTILIZATION FOROVERUTILIZATION FORCOMMERCIAL, RECREATIONAL,COMMERCIAL, RECREATIONAL,COMMERCIAL, RECREATIONAL,COMMERCIAL, RECREATIONAL,COMMERCIAL, RECREATIONAL,SCIENTIFIC, OR EDUCATIONALSCIENTIFIC, OR EDUCATIONALSCIENTIFIC, OR EDUCATIONALSCIENTIFIC, OR EDUCATIONALSCIENTIFIC, OR EDUCATIONAL

PURPOSESPURPOSESPURPOSESPURPOSESPURPOSES

The final rule stated that scientific researchhas the greatest potential for overutilization ofthe Mexican spotted owl, whereas birding,educational field trips, and agency “show me”trips are likely to increase as the owl becomesbetter known. The effects of these activities,either chronically or acutely, are unknown.

DISEASE OR PREDATIONDISEASE OR PREDATIONDISEASE OR PREDATIONDISEASE OR PREDATIONDISEASE OR PREDATION

The final rule stated that great horned owlsand other raptors are predators of Mexicanspotted owls. It also implied that forest manage-ment created ecotones favored by great hornedowls, thus creating an increased likelihood ofcontact between the two species.

INADEQUACY OF EXISTINGINADEQUACY OF EXISTINGINADEQUACY OF EXISTINGINADEQUACY OF EXISTINGINADEQUACY OF EXISTINGREGULATORY MECHANISMSREGULATORY MECHANISMSREGULATORY MECHANISMSREGULATORY MECHANISMSREGULATORY MECHANISMS

The final rule discussed various Federal andState laws and agency management policies,concluding that existing regulatory mechanismswere inadequate to protect the Mexican spottedowl. For further discussion on extant regulatorymechanisms, refer to Part IV.

OTHER NATURAL OR MANMADEOTHER NATURAL OR MANMADEOTHER NATURAL OR MANMADEOTHER NATURAL OR MANMADEOTHER NATURAL OR MANMADEFACTORS AFFECTING ITSFACTORS AFFECTING ITSFACTORS AFFECTING ITSFACTORS AFFECTING ITSFACTORS AFFECTING ITSCONTINUED EXISTENCECONTINUED EXISTENCECONTINUED EXISTENCECONTINUED EXISTENCECONTINUED EXISTENCE

The final rule cited wildfires as a past andfuture threat to spotted owl habitat. The poten-tial for increasing malicious and accidentalanthropogenic harm to the species was also citedas a possible threat. In addition, the final rulerecognized the potential for the barred owl toexpand its range into that of the Mexican spot-ted owl, resulting in possible competition and/orhybridization. It was speculated that habitatfragmentation may encourage and hasten thisexpansion.

3


Volume I/Part I

C. PAST AND CURRENT MANAGEMENTC. PAST AND CURRENT MANAGEMENTC. PAST AND CURRENT MANAGEMENTC. PAST AND CURRENT MANAGEMENTC. PAST AND CURRENT MANAGEMENTOF THE MEXICAN SPOTTED OWLOF THE MEXICAN SPOTTED OWLOF THE MEXICAN SPOTTED OWLOF THE MEXICAN SPOTTED OWLOF THE MEXICAN SPOTTED OWL

Prior to the proposed listing of the Mexicanspotted owl, some Federal agencies and involvedStates had conferred special status on the subspe-cies (e.g., State “threatened,” FS “sensitive,”FWS “candidate”) in recognition of its rarity,habitat preferences and threats to those habitattypes, and/or need of special managementconsiderations. This section summarizes thespecial status assigned to the subspecies and theresulting conservation efforts.

FISH AND WILDLIFE SERVICEFISH AND WILDLIFE SERVICEFISH AND WILDLIFE SERVICEFISH AND WILDLIFE SERVICEFISH AND WILDLIFE SERVICE

The FWS listed the entire spotted owlspecies as a Category-2 candidate in its 6 January1989 Notice of Review (54 FR 554). Category-2candidates are those species that the FWSbelieves may qualify for listing as threatened orendangered but for which insufficient informa-tion is available to support the required rule-making process. The northern subspecies waslisted as threatened in 1990; the California andMexican subspecies remained in Category-2candidate status.

The Mexican spotted owl was proposed forlisting as a threatened species on 4 November1991 (56 FR 56344) as a result of a status reviewprompted by a petition to list the subspecies.Following publication of the listing proposal, theFWS attempted to develop a conservationagreement with involved Federal agencies toconserve the Mexican spotted owl. This effortwas unsuccessful, so the final rule was publishedon 16 March 1993 (58 CFR 14248). Criticalhabitat was not determinable at the time oflisting.

Since the listing of the subspecies, the FWShas been conducting the processes associatedwith listed species under the Act, such as section7 consultation on Federal actions that may affectthe subspecies, issuance of research permitsunder Section 10, and recovery planning undersection 4, including funding of several researchprojects.

Two petitions to delist the species have beenreviewed by the FWS. In both cases, delistingwas determined to be “not warranted” because

the petitions failed to present substantial scien-tific and commercial information to supporttheir assertion that the species should bedelisted. Notices of those findings, includingdiscussions of the issues raised in the petitions,were published in the Federal Register on 23September 1993 (58 FR 49467) and 1 April1994 (59 FR 15361).

The FWS proposed critical habitat for theMexican spotted owl on 7 December 1994 (59FR 63162), and published the final criticalhabitat rule on 6 June 1995 (60 FR 29914).Since that time, the FWS has been in consulta-tion with action agencies on the effects ofproposed and ongoing actions on critical habitat.

FOREST SERVICEFOREST SERVICEFOREST SERVICEFOREST SERVICEFOREST SERVICE

The primary administrator of lands support-ing Mexican spotted owls in the United States isthe FS. Most spotted owls have been foundwithin FS Region 3 (including 11 NationalForests in Arizona and New Mexico). The RockyMountain (Region 2, including two NationalForests in Colorado) and Intermountain (Region4, including three National Forests in Utah)Regions support fewer spotted owls.

Forest Service Southwestern RegionForest Service Southwestern RegionForest Service Southwestern RegionForest Service Southwestern RegionForest Service Southwestern Region(Region 3)(Region 3)(Region 3)(Region 3)(Region 3)

Prior to the listing of the Mexican spottedowl, FS Region 3 issued detailed guidelines forits management. Those guidelines were issued asInterim Directive Number 1 (ID No. 1) in June1989, then revised and reissued as ID No. 2approximately one year later. Although ID No. 2expired in December 1991, FS Region 3 hascontinued managing under those guidelines.

Interim Directive Number 2 guidelinesrequired establishing management territoriesaround all nesting and roosting spotted owls, aswell as territorial owls detected at night forwhich daytime locations were not recorded. Allmanagement territories except those on theLincoln and Gila National Forests had a 182-ha

4


Volume I/Part I

(450 ac) core area surrounded by 627 ha (1,550ac) of the “best available” habitat, extending thearea to 809 ha (2,000 ac) per managementterritory. On the Lincoln and Gila NationalForests, the 182-ha (450 ac) cores were aug-mented by an additional 425 ha (1,050 ac) ofhabitat, for a total management territory size of607 ha (1,500 ac).

Except for road construction, habitat degra-dation was not allowed within managementterritory cores. In the remainder of the manage-ment territory management activities, includingtimber harvest, were limited to 209-314 ha(516-775 ac). The FS guidelines provided noprotection for unoccupied habitat except inwilderness areas and administratively restrictedlands.

The FS Region 3 has been in the process ofamending forest plans through the NationalEnvironmental Policy Act (NEPA) process toincorporate the management recommendationscontained in this Recovery Plan. The RecoveryTeam commends that effort.

Forest Service Rocky Mountain RegionForest Service Rocky Mountain RegionForest Service Rocky Mountain RegionForest Service Rocky Mountain RegionForest Service Rocky Mountain Region(Region 2)(Region 2)(Region 2)(Region 2)(Region 2)

Region 2 of the FS formed a task force in1992 to begin developing management guide-lines for the Mexican spotted owl. These man-agement guidelines are still in draft form andhave not been formally approved and adopted byFS Region 2. However, management activitiescontinue to be examined on a case-by-case basis,and ID No. 2 may be used as a general guideline.

Forest Service Intermountain RegionForest Service Intermountain RegionForest Service Intermountain RegionForest Service Intermountain RegionForest Service Intermountain Region(Region 4)(Region 4)(Region 4)(Region 4)(Region 4)

Prior to the listing of the Mexican spottedowl, biologists from FS Region 4 and Utah’sother land and wildlife management agencies,plus owl researchers, formed the Utah MexicanSpotted Owl Working Group (Working Group).The Working Group meets annually to identifyand address issues pertaining to the managementand conservation of Mexican spotted owls inUtah (Kate Grandison, FS, Cedar City, UT, pers.comm.). The Utah Mexican Spotted OwlTechnical Team (Technical Team) was formed by

the Working Group to focus on spotted owlissues such as (1) potential impacts to Mexicanspotted owls in southern Utah; (2) currentresearch and future research needs and priorities;(3) inventory and monitoring protocols; (4)management suggestions suitable for applicationby all land management agencies in southernUtah; and (5) dissemination of informationfrom the Working Group and Technical Team tomanagement and administrative levels. The goalsof the Technical Team, which is composed ofbiologists from the FWS, FS, USDI Bureau ofLand Management (BLM), USDI National ParkService (NPS), Utah Division of Wildlife Re-sources (UDWR), and a researcher/technicalconsultant, are to provide land and wildlifemanagers with the information necessary toensure the protection of Mexican spotted owlsand to suggest strategies for managing spottedowl habitat in Utah.

The Technical Team developed “Suggestionsfor Management of the Mexican Spotted Owl inUtah.” These suggestions were sent to lineofficers for approval on 5 August 1994. Manage-ment territories on the Manti-LaSal NationalForest were established using these suggestions,although ID No. 2 has also been adopted byRegion 4. Interim Directive No. 2 was modifiedin March 1994 in Region 4 to change the surveyprotocol to include only potential breedinghabitat in canyon areas below 2,590 m (8,500ft).

According to the suggestions, managementterritory size should be 1,330 ha (3,350 ac) with355-ha (875 ac) core areas of canyon habitat. Inaddition, a 0.8-km (0.5 mi) protection areacentered on the nest site was established toprotect the nest stand and surrounding areas.Habitat degradation is not allowed in the man-agement territory areas. A “potential dispersalarea” extends 58 km (35.8 mi) beyond theperimeter of the management territory. This areacan be used for timber harvest, but post-harvestconditions must meet a reasonable facsimile ofthe 50-11-40 dispersal rule developed by Tho-mas et al. (1990). Forest Service Region 4continues to manage under ID No. 2, with theabove modifications, except where superseded bythe “Suggestions for Management of MexicanSpotted Owls in Utah.”

5


Volume I/Part I

OTHER FEDERAL AGENCIESOTHER FEDERAL AGENCIESOTHER FEDERAL AGENCIESOTHER FEDERAL AGENCIESOTHER FEDERAL AGENCIES

National Park ServiceNational Park ServiceNational Park ServiceNational Park ServiceNational Park Service

Several NPS-administered units are knownto support Mexican spotted owls, includingZion, Capitol Reef, and Canyonlands NationalParks plus Glen Canyon National RecreationArea in Utah; Mesa Verde National Monumentin Colorado; Grand Canyon National Park plusSaguaro, Walnut Canyon, and ChiricahuaNational Monuments in Arizona; BandelierNational Monument in New Mexico; andGuadalupe Mountains National Park in Texas.

The National Park Service Organic Actprotects all wildlife on National Parks andMonuments. However, no specific managementguidelines are in place for Mexican spotted owls,and the effectiveness of applying general lawsand policies for spotted owls is difficult toevaluate.

Bureau of Land ManagementBureau of Land ManagementBureau of Land ManagementBureau of Land ManagementBureau of Land Management

The BLM has developed managementpolicies specifically for the Mexican spotted owlin Colorado and New Mexico. The Coloradoguidelines state that “...in areas with a confirmednest or roost site, surface management activitieswill be limited and will be determined on a case-by-case basis to allow as much flexibility aspossible outside of the core area.” The BLM inColorado has management guidelines for oil andgas development where Mexican spotted owls areknown to occur. No surface occupancy is al-lowed within 0.4 km (0.25 mi) of a nest or roostsite, and restrictions on other associated activitiesapply between 1 February and 31 July. Spottedowl management policy by the BLM in NewMexico establishes and preserves cores of habitatwherever the owl is found. The BLM determinesthe size of the cores on a case-by-case basis.

The BLM in Colorado follows the surveytechniques of the FS Region 3 spotted owlprotocol. Management territories have not beendesignated for known birds. Surveys are con-ducted in areas of potential habitat whereprojects are planned that may be in conflict withspotted owl management.

The BLM in Utah has no specific internalguidelines on management practices for Mexicanspotted owls. However, agency personnel didparticipate in producing “Suggestions for theManagement of Mexican Spotted Owls inUtah.” The BLM will incorporate managementprescriptions for the Mexican spotted owl and itspotential habitat into resource managementplans as they are updated over the next severalyears.

The BLM in Arizona has no specific guide-lines for managing Mexican spotted owls.However, the standard BLM procedure forassessing impacts on threatened or endangeredspecies will be followed for projects proposed inspotted owl habitat. Guidelines for protectingthe owl or its habitat would then be developedon a site-specific basis (Ted Corderey, BLM,Endangered Species Coordinator, PhoenixOffice, pers. comm.).

Department of DefenseDepartment of DefenseDepartment of DefenseDepartment of DefenseDepartment of Defense

The Fort Huachuca Military Reservation(Post) in southeastern Arizona is the only mili-tary land known to support nesting Mexicanspotted owls. On the Post, military activity inspotted owl habitat is generally confined tovarious foot maneuvers, although the Army isconsidering expanding some tank maneuversinto higher elevations where the owl occurs(Sheridan Stone, Fort Huachuca Military Reser-vation, pers. comm.). One spotted owl site hasbeen popular with birders for a number of years,but the effect of this activity is unknown. TheArmy also considers wildfire to be a potentialthreat and assesses the possibility of wildfireignition when designing military activities onthe Post.

Wintering Mexican spotted owls have beenfound on Fort Carson, near Colorado Springs,Colorado, and breeding owls are present on theFremont Military Operating Area, which in-cludes FS and BLM lands designated for con-ducting military maneuvers. Finally, low-levelmilitary air operations in some areas have beenidentified as actions that may affect Mexicanspotted owls. Such operations are likely toincrease in the next several years, and the De-partment of Defense is currently funding studies

6


Volume I/Part I

of the effects of these activities on spotted owls(M. Hildegard Reiser, Holoman Air Force Base,pers. comm.).

STATESSTATESSTATESSTATESSTATES

ArizonaArizonaArizonaArizonaArizona

The Mexican spotted owl is listed as “threat-ened” on the list of “Threatened Native Wildlifein Arizona” (Arizona Game and Fish Depart-ment 1988). The Arizona Game and FishDepartment has authority to manage wildlifeunder provisions of Arizona Revised Statute 17,the goal of which is to maintain State’s naturalbiotic diversity by listing and protecting threat-ened and endangered species. “Threatened”species is defined as “...those species or subspe-cies whose continued presence in Arizona couldbe in jeopardy in the near future. Serious threatshave been identified and populations are (a)lower than they were historically or (b) extremelylocal and small.”

Threatened status provides no special protec-tion to species, although it does provide amechanism through which the state can allocateHeritage Program grants to fund research forspecially designated species. However, generalArizona wildlife rules make it unlawful “...unlessotherwise prescribed...for a person to...take,possess, transport, buy, sell or offer or expose forsale wildlife, except as expressly permitted ....”

New MexicoNew MexicoNew MexicoNew MexicoNew Mexico

The State of New Mexico confers no specialstatus on spotted owls. However, New MexicoStatute 17-2-14 makes it unlawful “...for anyperson to take, possess, trap or ensnare, or in anymanner to injure, maim or destroy birds of theorder Strigiformes.” However, permits may beobtained to take owls for purposes of Indianreligion, scientific study, or falconry. In addition,persons who commercially raise poultry or gamebirds may legally kill any owl that has killed theirstock.

ColoradoColoradoColoradoColoradoColorado

The Mexican spotted owl was listed asthreatened by the Colorado Division of Wildlife(CDOW) in 1993. “Threatened” wildlife isdefined as “...any species or subspecies of wildlifewhich, as determined by the Colorado WildlifeCommission, is not in immediate jeopardy ofextinction but is vulnerable because it exists insuch small numbers or is so extremely restrictedthroughout all or a significant portion of itsrange that it may become endangered.” Threat-ened status protects wildlife species by makingit unlawful “...for any person to take, possess,transport, export, process, sell or offer forsale...any species or subspecies of [threatened]wildlife....” In addition, the CDOW is legisla-tively mandated to “...establish such programsincluding acquisition of land...as are deemednecessary for management of...threatenedspecies.” An interagency working groupcoordinates spotted owl inventories throughoutColorado.

UtahUtahUtahUtahUtah

The UDWR included the Mexican spottedowl as a sensitive species on its 1987 Native UtahWildlife Species of Special Concern list (UDWR1987). “Sensitive” wildlife is defined as “...anywildlife species which, although still occurring innumbers adequate for survival, whose populationhas been greatly depleted, is declining in num-bers, distribution, and/or habitat (S1); or occursin limited areas and/or numbers due to a re-stricted or specialized habitat (S2).” A manage-ment program, including protection or enhance-ment, is needed for these sensitive species.

The owl’s status was elevated to “Threat-ened” in the revised draft list in 1992 (UDWR1992). According to UDWR definition, “threat-ened” species include “...any wildlife species,subspecies, or population which is likely tobecome an endangered species within the fore-seeable future throughout all or a significantportion of its range in Utah or the world.”

Both sensitive and threatened species receive“protected” status under Utah’s wildlife codes.For species under protected status, “...[A] person

7


Volume I/Part I

may not take...protected wildlife or their parts;an occupied nest of protected wildlife; or anegg of protected wildlife.” Nor may a person“...transport,...sell or purchase...or possessprotected wildlife or their parts.”

TexasTexasTexasTexasTexas

Few Mexican spotted owls are documentedfor Texas, and most of the location records are inGuadalupe National Park. Thus, the State ofTexas has no spotted owl program. However,Texas Parks and Wildlife Code Section 64.002provides protection for nongame birds byprohibiting killing, trapping, transportation,possession of parts, and the like. Destruction ofeggs and nests of nongame birds is also prohib-ited.

TRIBESTRIBESTRIBESTRIBESTRIBES

Tribal beliefs and philosophies guide re-source management on Tribal lands. SeveralTribes consider owls a bad omen; however, Tribalbeliefs also dictate that all living creatures areessential parts of nature and, as such, they arerevered and protected. For example, the EldersCouncil of San Carlos Apache Tribe expressedthe traditional view that owls and their homesshould not be disturbed.

Mexican spotted owl habitat or potentialhabitat exists on 10 Indian reservations in theSouthwest. Eight of the Tribes have conductedspotted owl surveys, and five Tribes have locatedspotted owls on their lands. Two other Tribeshave historical spotted owl records.

Reservations were established for the benefitof the Tribes and their members. Tribal lands areheld in “trust” by the Federal Government. Theyare not considered public lands or part of thepublic domain. Tribes are sovereign governmentswith management authority over wildlife andother Tribal land resources. Many Tribes main-tain professionally staffed wildlife and naturalresources management programs to ensureprudent management and protection of tribalresources, including threatened and endangeredspecies.

The FWS is aware of spotted owl conserva-tion efforts on five Indian reservations: theMescalero Apache, Fort Apache, San CarlosApache, Jicarilla Apache, and Navajo Nation.

Mescalero Apache TribeMescalero Apache TribeMescalero Apache TribeMescalero Apache TribeMescalero Apache Tribe

The Mescalero Apache Tribe in New Mexicoactively manages their forest while managing forall federally listed or proposed threatened orendangered species that may exist on the reserva-tion, including the Mexican spotted owl. This isaccomplished through developing strategies foridentifying and managing habitat determined bythe Tribe to be necessary to ensure protection.The Mescalero has been working with the FWSin development of a conservation strategy for thesubspecies on reservation lands.

White Mountain Apache TribeWhite Mountain Apache TribeWhite Mountain Apache TribeWhite Mountain Apache TribeWhite Mountain Apache Tribe

The Tribe recently developed a conservationplan for Mexican spotted owls on the reserva-tion. Areas containing spotted owls are placed inone of two land-management categories, termedDesignated Management Areas (DMAs). Areassupporting “clusters” of four or more territoriesare considered “Category-1” DMAs. In theseareas, spotted owl habitat concerns drive man-agement prescriptions; timber harvest is asecondary objective. Category-1 DMAs rangefrom about 2,430-4,050 ha (6,000-10,000 ac),and contain 57% of known spotted owl sites onthe reservation.

“Category-2” DMAs include areas support-ing 1-3 owl territories. Habitat outside theterritories is managed only secondarily forspotted owls, with other resource objectivesgiven priority. No timber harvest is allowed in30-ha (75 ac) patches around owl activity cen-ters. A seasonal restriction on potentially disturb-ing activities is provided in a 202-ha (500 ac)area, and timber prescriptions within this areashould be designed to improve habitat integrity.

The Tribe continues to survey their lands forspotted owls. If more owl sites are detected,Category-1 and Category-2 DMAs may beestablished upon approval by the Tribal Council.

8


Volume I/Part I

San Carlos Apache TribeSan Carlos Apache TribeSan Carlos Apache TribeSan Carlos Apache TribeSan Carlos Apache Tribe

Spotted owl surveys on the San CarlosApache Reservation have been conducted ac-cording to the FS Region 3 Mexican SpottedOwl Inventory Protocol. Mexican spotted owlhabitat has been identified and delineatedthroughout the reservation. A joint Tribal/Bureau of Indian Affairs interdisciplinary teamevaluates effects of actions on spotted owls. Anypotential impact on spotted owls or owl habitatis deferred until compliance with the Act andassociated regulations is attained.

Preliminary discussions between the SanCarlos Apache and the FWS have taken placeregarding development of specific spotted owlmanagement guidelines. Approximately 90% oftribally identified nesting, roosting, and foraginghabitats are on lands inoperable for timberharvest and therefore are not in the commercialtimber base.

Jicarilla Apache TribeJicarilla Apache TribeJicarilla Apache TribeJicarilla Apache TribeJicarilla Apache Tribe

The Jicarilla Apache Tribe has developed aspotted owl conservation plan, approved by theJicarilla Tribal Council and accepted by theFWS. No resident owls have been detected todate on the reservation; however, in the eventresident owls are detected, the Tribe has pro-posed to designate a 405-ha (1,000 ac) manage-ment territory. Uneven-aged timber manage-ment will be allowed to continue in all but 40 ha(100 ac) of the territory. In the absence ofconfirmed resident owls, all mixed-conifer standsof 10 ha (25 ac) or greater are treated as roost-ing/nesting sites, and timber harvest will not beallowed. A seasonal restriction around any activenest sites that are found is also proposed.

Navajo NationNavajo NationNavajo NationNavajo NationNavajo Nation

Mexican spotted owl management on theNavajo Nation, and particularly on the NavajoNation Commercial Forest, currently adheres toFS Region 3’s ID No. 2. The FS Region 3Mexican Spotted Owl Inventory Protocol isfollowed for all timber sales on the commercialforest and for any project or disturbance, on or

off the commercial forest, that may impactspotted owls. The current Navajo spotted owlinventory program is limited to areas wheretimber sales or other projects are planned.

The Navajo Nation is developing a multi-species conservation plan, including manage-ment guidelines for spotted owl conservation, inconjunction with their 10-year plan for manag-ing commercial forest. Upon completion of themulti-species conservation plan, the NavajoNation may apply to the FWS for a section10(a)(1)(B) permit, which will allow limitedincidental take to occur provided an adequatehabitat conservation plan is implemented.

MEXICOMEXICOMEXICOMEXICOMEXICO

The Mexican spotted owl is listed as athreatened species under Mexico’s OfficialMexican Norm (NOM) (NOM-059-ECOL-1994). Threatened species are defined as thosewhich could face danger of extinction if theconditions that cause deterioration or modifica-tion of their habitats, or decline of their popula-tions, prevail.

Species listed under NOM are affordedcertain protections:

1. Possession, use, or derivation of profitfrom live wildlife or plants, whetheroriginating in captivity or in the wild, areprohibited.

2. Use and exploitation of the habitats oflisted species are prohibited in someStates.

Some use of threatened species is allowed forscientific and recovery purposes. For example,specimens and their parts, products, and by-products can be removed from their naturalenvironment for scientific purposes underpermits issued by legal authorities, with theunderstanding that specimens or their partscannot be used for commercial purposes. Inaddition, specimens can be removed from thewild for the purpose of captive breeding uponapproval of the Mexican government.

Under NOM, recovery plans have beenprepared for sea turtles and the monarch butter-

9


Volume I/Part I

fly. The Government and other institutions haveshown interest in the conservation of otherspecies such as manatees, yellow-headed parrots,and Mexican spotted owls, but currently norecovery plans are in place for these species.

Social, economic, and political systems differin Mexico from those in the U.S. Concomi-tantly, land ownership patterns differ and influ-ence natural resource management. Mexicanlands are classified into three types of tenancy:

1. Federal lands include all lands adminis-tered under Federal Government institu-tions. Federal lands include protectednatural areas such as Reserves of theBiosphere, National Parks, and Areas ofProtection of Natural Resources. Pro-tected natural areas comprise 3% of thetotal area of Mexico’s five recovery units.

2. Ejidal lands are allotted by the MexicanGovernment to a person or communityfor agriculture, forestry, mining, andother uses. Thus, lands within ejidos areintensively managed for natural resourceuse. Ejidos comprise approximately17.5% of the area within the Mexicanrecovery units.

3. Private lands are possessed under a“certificate of inaffectability.” Anyprotection afforded these lands is at thediscretion of the landowner. Approxi-mately 79.5% of the Mexican recoveryunits is comprised of private land.

10


Volume I/Part I

D. CONSIDERATIONS IND. CONSIDERATIONS IND. CONSIDERATIONS IND. CONSIDERATIONS IND. CONSIDERATIONS INRECOVERY PLAN DEVELOPMENTRECOVERY PLAN DEVELOPMENTRECOVERY PLAN DEVELOPMENTRECOVERY PLAN DEVELOPMENTRECOVERY PLAN DEVELOPMENT

This section describes various consider-ations, other than the basic biology of theMexican spotted owl, that were integral indevelopment of this Recovery Plan.

RECOVERY UNITSRECOVERY UNITSRECOVERY UNITSRECOVERY UNITSRECOVERY UNITS

The Mexican spotted owl is a widespreadsubspecies that occurs in a wide variety ofhabitats (see Part II). In addition, the threatsfaced by the subspecies, the management re-gimes employed by various agencies and in eachcountry, and the protective mechanisms availablein different portions of the subspecies’ range arevariable. Finally, spotted owl densities, foodhabits, degree of isolation, and other aspects ofthe subspecies’ biology differ somewhat amongportions of its range. For these reasons, theRecovery Team partitioned the Mexican spottedowl range into distinct recovery units. Sixrecovery units were designated in the UnitedStates: Colorado Plateau, Southern RockyMountains - Colorado, Southern Rocky Moun-tains - New Mexico, Upper Gila Mountains,Basin and Range - West, and Basin and Range -East (Figs. II.B.1 and II.B.3–II.B.8). Five recov-ery units were established in Mexico: SierraMadre Occidental - Norte, Sierra Madre Occi-dental - Sur, Sierra Madre Oriental - Norte,Sierra Madre Oriental - Sur, and EjeNeovolcanico (Figs. II.B.2). For a completedescription of the recovery units and the basesfor their designation see II.B.

Whereas some management recommenda-tions apply to the subspecies rangewide, delin-eating recovery units allowed specific recommen-dations to be prioritized appropriately withineach portion of the subspecies’ range. In addi-tion, some criteria for delisting the subspeciesapply at the recovery-unit level. This approachallows delisting of the Mexican spotted owl byrecovery unit when certain rangewide populationand habitat criteria are met and when regionalmanagement plans or other sufficient regulatorymechanisms are implemented.

THE CURRENT SITUATIONTHE CURRENT SITUATIONTHE CURRENT SITUATIONTHE CURRENT SITUATIONTHE CURRENT SITUATION

In developing this Recovery Plan, the Recov-ery Team considered various aspects of thecurrent spotted owl population, habitat, andthreats. Two salient points emerged. First, theRecovery Team assumes that the current popula-tion size and distribution are adequate forproviding a reference point for assessing futurechanges in the population, since no undisputableevidence is available indicating that the popula-tion is declining or is significantly below histori-cal levels. This is a critical assumption that mustbe tested through the population monitoringrequired by this Recovery Plan. If the monitor-ing data demonstrate that the population isstable or increasing, the assumption of adequatepopulation size will be validated. Conversely, ifmonitoring data show a decreasing population,the situation will need to be reexamined andcorrective measures must be developed. Thus,the population and habitat monitoring require-ments are essential parts of this Recovery Plan; ifthese monitoring efforts are not conducted, themanagement recommendations provided hereincannot stand alone.

A second consideration involves variations inboth spotted owl densities and threats facedthroughout the subspecies’ range. Spotted owldensities are greatest in the center of the subspe-cies’ range and they decrease toward the rangeperiphery. In addition, the main threats identi-fied during the listing process were forestrypractices and wildfire risk, both of which varyacross the subspecies’ range. Table I.D.1. illus-trates the Recovery Team’s appraisal.

The Upper Gila Mountains, Basin andRange - West, and Basin and Range - EastRecovery Units have significant owl populationswith the potential of being seriously impacted byfire and/or forestry practices (Table I.D.1). Thisconclusion does not imply that the other recov-ery units are not important, but leads to therecommendations that (1) recovery effortsconcentrate on recovery units with the highestowl populations and where significant threats

11


Volume I/Part I

exist; (2) management within recovery unitsshould emphasize alleviating the greatest threats;and (3) the management recommendations inPart III should be tailored to the owl populationand the threats existing in the specific area underanalysis.

The Recovery Team believes the risk ofextirpation of Mexican spotted owls under thenear-term management recommendations is low.This belief is based on two points:

1. Implementation of the managementrecommendations within this RecoveryPlan (see Part III) will protect occupiedhabitat, protect other habitat that canpresumably be occupied in the nearfuture, and allow for “replacement”habitat to develop and be sustained onthe landscape. Habitat monitoring asrequired by this Recovery Plan shouldprovide data on habitat trends through-out Recovery Plan duration.

2. The population will be monitored overthe life of the Recovery Plan, thusproviding insight as to whether thecurrent “baseline” population is suffi-cient to maintain the subspecies overtime and testing the assumption thatthe “baseline” population is adequate.The Recovery Team did not make theassumption that the “baseline” popula-tion is adequate lightly, but reasonedthat the Mexican spotted owl is welldistributed throughout its historicalrange, suggesting that no significantextirpations have occurred.

RECOVERY PLAN DURATIONRECOVERY PLAN DURATIONRECOVERY PLAN DURATIONRECOVERY PLAN DURATIONRECOVERY PLAN DURATION

Any management plan must specify the timeperiod over which the plan is to be imple-mented. The Recovery Team decided that a 10-year period is appropriate for the MexicanSpotted Owl Recovery Plan (assuming thedelisting criteria specified in III.A are met) forseveral reasons:

1. Ten years allows adequate time to moni-tor the trends in population and habitat.The charge of the Recovery Team wasto develop a plan that would lead torecovery of the subspecies. In developingthe delisting criteria specified in III.A,the Recovery Team reasoned that thepopulation must be stable or increasingbefore the subspecies could be consid-ered for delisting. The Recovery Teamfurther determined that a monitoringperiod of 10 years would provide infor-mation about population trends thatcould be used with a reasonably highlevel of confidence. The five-year moni-toring period the Act requires after aspecies is delisted will further increaseconfidence in trend information.

2. A 10-year period should be sufficienttime to fill some of the major gaps inexisting knowledge, and accommodatepossible changes in future conditions.Many aspects of Mexican spotted owlbiology remain unknown or poorlyunderstood. Consequently, the effects of

12

TTTTTable I.D.1.able I.D.1.able I.D.1.able I.D.1.able I.D.1. Summary of relative Mexican spotted owl population size, timber harvest threat, andfire threat by U.S. Recovery Unit.

ThrThrThrThrThreat Seat Seat Seat Seat Significanceignificanceignificanceignificanceignificance

RRRRRecoecoecoecoecovvvvvererererery Uy Uy Uy Uy Unitnitnitnitnit PPPPPopulationopulationopulationopulationopulation FFFFFiririririreeeee TTTTTimberimberimberimberimber

Colorado Plateau low moderate lowSouthern Rocky Mtns-CO low high lowSouthern Rocky Mtns-NM low high highUpper Gila Mountains high high highBasin and Range - West high high lowBasin and Range - East high high high


Volume I/Part I

different resource-management practiceson the fitness of individuals and onpopulation persistence remain unclear. Abetter understanding of these relation-ships is needed before a viable long-termmanagement plan can be developed.Implementing the Recovery Plan in-cludes conducting the research activitiesrecommended in III.D; if these studiesare started immediately, 10 years shouldbe adequate to complete the majority ofthem.

3. Uncertainty about the future couldrender this Recovery Plan inadequate,unacceptable, or otherwise obsolete.Future events and developments couldhave social, economic, environmental,and other ramifications that cannot bepredicted. To try to plan beyond the nextdecade or so would require an unjustifiedconfidence in our ability to predict thestate of our society and the environment.

4. Consistency with the requirements of theAct. The Act requires that the status oflisted species be reviewed every five years.This Recovery Plan constitutes an in–depth status review of the Mexicanspotted owl, so a formal status reviewshould be conducted in years five and 10of Recovery Plan implementation.Unless new information or other devel-opments render this Recovery Planobsolete in the interim, the 10-yearpoint should mark the end of thisRecovery Plan and implementation of alonger-term management strategy.

Several reviewers of the draft version of thisRecovery Plan pointed out that this relativelyshort Recovery Plan duration fails to take intoaccount the long-term processes that haveinfluenced and will continue to influencedynamic ecosystems. The Recovery Team be-lieves, however, that the management recom-mended for the next few years was developedwith consideration of the long- and short-termeffects of these near-term management recom-mendations.

Based on the foregoing points, the RecoveryTeam recommends a Recovery Plan duration of10 years unless data indicate that earlier revisionis appropriate, or that the applicable recommen-dations be continued beyond that time. Themonitoring and research to be conducted duringthe life of the Recovery Plan will resolve muchuncertainty surrounding the Mexican spottedowl. The uncertainty about the future can neverbe resolved, but a better understanding ofMexican spotted owl natural history will en-hance our ability to create a long-term manage-ment plan.

CONSERVATION PLANSCONSERVATION PLANSCONSERVATION PLANSCONSERVATION PLANSCONSERVATION PLANSFOR OTHER SPOTTED OWLFOR OTHER SPOTTED OWLFOR OTHER SPOTTED OWLFOR OTHER SPOTTED OWLFOR OTHER SPOTTED OWL

SUBSPECIESSUBSPECIESSUBSPECIESSUBSPECIESSUBSPECIES

Several conservation strategies have beendeveloped for the other spotted owl subspecies.Perhaps the best known subspecies is the north-ern spotted owl of the Pacific Northwest andnorthwestern California. The northern subspe-cies was listed as threatened in June 1990,resulting in extensive conflict between conserva-tion of the subspecies and economic and socialinterests of the Pacific Northwest, particularlythe timber industry.

The first management strategy was initiatedby the FS in the late 1970s. That approach,which continued within some portions of thesubspecies’ range until 1990, was to manageindividual spotted owl territories, called SpottedOwl Habitat Areas (SOHAs), or, earlier, SpottedOwl Management Areas. Each SOHA consistedof certain acreages that varied according tolocation. Those territories were establishedaccording to certain clustering and spacingguidelines, and the general prescription for theterritories was to restrict timber harvest so that aminimum “suitable habitat” acreage standard wasmaintained in the territories. However, in certaincircumstances some harvest was allowed, such assalvage harvest.

In 1989, in response to increasing contro-versy over the spotted owl issue, the difficultythe issue was causing land-management agencies,and the proposed listing of the northern spottedowl as a threatened species, the InteragencyScientific Committee (ISC) was established. The

13


Volume I/Part I

ISC produced “A Conservation Strategy for theNorthern Spotted Owl” (Thomas et al. 1990),which recommended significant changes inspotted owl management on public lands in thePacific Northwest. Briefly, the ISC delineatedlarge blocks of owl habitat, called HabitatConservation Areas (HCAs). The goal was todelineate, where possible, HCAs known tosupport or with the potential to support at least20 spotted owl pairs. However, where 20-pairHCAs were not possible, HCAs of 1 to 19 pairswere delineated. The HCAs were spaced certaindistances from one another, depending on theirsizes. The HCAs were to be managed so that nohabitat degradation occurred within them,protecting existing habitat and allowing previ-ously disturbed areas to return to a suitablecondition. The ISC envisioned eventual HCAswhere owl pairs could freely interact withoutsignificant disruption of habitat continuitybetween territories.

In addition, the ISC recommended manag-ing the areas between HCAs, termed the “forestmatrix,” according to the “50-11-40 rule.” Thisrule prescribed that at least 50% of the forestedarea within each quarter-township was to con-tain trees averaging a minimum of 28 cm (11 in)in diameter and with at least 40% crown closure.The idea was that these conditions would allowmovement of owls between HCAs, therebyallowing genetic flow and demographic rescue ofsubpopulations. The ISC also recommendedretention of 28-ha (70-acre) areas within theforest matrix to possibly provide future nesting/roosting sites.

In 1991, the Secretary of the Interior ap-pointed the Northern Spotted Owl RecoveryTeam and charged it with developing a recoveryplan for that subspecies. That recovery plan wasclosely modeled after the ISC plan. The HCAnetwork was modified based on updated infor-mation, resulting in a network of DesignatedConservation Areas (DCAs). Timber harvest inDCAs was generally not allowed in suitablehabitat. Silvicultural treatments designed toencourage spotted owl habitat were limited to nomore than 5% of a DCA in the first five years ofplan implementation. Management recommen-dations for areas outside DCAs deviated fromthe ISC approach by providing greater justifica-

tion for specific recommendations and consider-ation of economic efficiency of implementation.

The Northern Spotted Owl Recovery Planwas never implemented. Instead, the most recentmanagement strategy for the northern spottedowl resulted from analyses and recommendationsformulated by the Forest Ecosystem Manage-ment Assessment Team (FEMAT). The FEMATwas appointed by President Clinton to developalternative plans for management of late-succes-sional ecosystem components including, but notspecific to, the northern spotted owl. OnlyFederal land management was addressed in theFEMAT alternatives.

The President selected “Option 9” developedby the FEMAT, which calls for a series of LateSuccessional Reserves (LSRs) correspondingroughly to the HCAs under the ISC plan.Outside the LSRs, the forest matrix includesRiparian Reserves (various sized buffers alongclass 1-3 streams); green-tree retention require-ments (where 15% of each watershed is managedfor late successional forest and at least 15% ofeach harvest unit is retained in the latest succes-sional forest available); and preservation of 40 ha(100 ac) around all owl sites known as of 1January 1994. The goal of this matrix prescrip-tion is to accommodate dispersing and “floater”owls, as well as other species dependent on oldand mature forest conditions.

The FEMAT plan also establishes AdaptiveManagement Areas (AMAs) in California,Oregon, and Washington. These AMAs varyfrom less than 40,500 ha (100,000 ac) to nearly200,000 ha (500,000 ac). The managementobjectives for each AMA also vary, but they aregenerally established to develop and test tech-niques for active forest management that providea wide range of resource values including forestproducts, late-successional forest habitat, andhigh-quality recreation.

The range of the California spotted owlabuts the range of the northern subspecies innortheastern California, ranging south throughthe Sierra Nevada, west through the “transverseranges” of Southern California, then north alongthe Coast Range to Monterey County. TheCalifornia spotted owl was first managed by theFS using the SOHA system described for thenorthern subspecies. The portion of the subspe-

14


Volume I/Part I

cies’ range in Southern California is still man-aged under that system while its effectiveness isassessed. However, the current managementregime in the Sierra Nevada is described in the“Assessment of the Current Status of the Califor-nia Spotted Owl, with Recommendations forManagement” (Caspow Plan) (Verner et al.1992b). The FS implemented the Caspow Planon a temporary basis through an environmentalassessment under the National EnvironmentalPolicy Act (NEPA) on 1 March 1993. The FShas since released a draft environmental impactstatement under the NEPA that analyzes severalalternatives for California spotted owl manage-ment, one of which is the Caspow Plan.

In the Sierra Nevada, the Caspow Planrecommends management emphasizing adequateamounts and distribution of suitable owl habitatthrough improved habitat and resource manage-ment practices rather than through protection oflarge blocks of habitat. The strategy seeks toprotect known owl nest or roost sites, retain thelarger and older components of forest structure,and address the problems of fire suppression andfuel loading. Additional objectives includerapidly recovering nesting and roosting habitatfollowing disturbance, maintaining existingcanopy layers, promoting tree growth by thin-ning in middle and lower canopy layers, andreducing vertical fuel ladders. Habitat availablefor timber management is classified by structuralcondition and utility for various life historyrequirements, and the resultant habitat classesare managed with restrictions on structuralmodifications. Long-term management propos-als also focus on spotted owl nesting and roost-ing habitat as the target conditions for silvicul-tural activities.

To formulate the management strategycontained in this Recovery Plan, the RecoveryTeam examined the extensive efforts to protectthe California and northern spotted owl subspe-cies. These efforts have, so far, experiencedvarying degrees of success and controversy.Moreover, the three subspecies exhibit differ-ences in habitat use, habitat distribution, andthreats. The Mexican Spotted Owl RecoveryPlan combines, in the Recovery Team’s opinion,the applicable recommendations from other

planning efforts with those uniquely applicableto the Mexican subspecies.

ECOSYSTEM MANAGEMENTECOSYSTEM MANAGEMENTECOSYSTEM MANAGEMENTECOSYSTEM MANAGEMENTECOSYSTEM MANAGEMENT

The development of ecosystem managementin the history of conservation plans for thenorthern spotted owl was described by Meslow(1993). Results of population and habitatresearch were incorporated into landscapedesigns involving reserves (HCAs) and intersti-tial matrices, both of which are basic attributesof landscape ecology (Diaz and Apostol 1993).Further considerations of the ecosystem manage-ment approach were extended to sympatricFederal- and State-listed species. It was clear thatsingle-species management for the northernspotted owl would have numerous impacts onthe many eligible but yet to be listed species(USDI 1992; Block et al. 1995). Thomas et al.(1993) described the relationship between theISC plan for the northern spotted owl and thelikelihood of viability for a suite of other speciesclosely associated with late-successional forest.Verner et al. (1992) described numerous linksbetween the California spotted owl and associ-ated ecosystem components that it uses andrequires to survive. Assessments of other species,such as northern goshawks (Reynolds et al.1992) have also underscored the need to managelarge landscapes to provide adequate prey andthe diversity of habitats needed by those species.

Despite growing academic and professionalawareness of the need to manage entire ecosys-tems, the Recovery Team is charged with devel-opment of a recovery plan for a single species.However, as the FWS and other land-manage-ment agencies move toward managing entireecosystems, they are recognizing that single-species management will never protect all of theorganisms that comprise the ecosystems uponwhich target species depend. Furthermore, amanagement plan for one species may conflictwith a management plan for a sympatric speciesin absence of careful integration of the twoplans. Block and Brennan (1993) noted that themanagement recommendations of the ISC forthe northern spotted owl were firmly based inhabitat management. In addition to habitat,however, both ecosystem-oriented and popula-

15


Volume I/Part I

tion-level considerations must be wed in conser-vation planning (Gutiérrez 1994).

The recovery team considered the interac-tion of populations, habitats, and ecosystems inthe development of this Recovery Plan. TheRecovery Team recognizes that numerous habi-tats exist within the range of the Mexican spot-ted owl and that not all of those habitats areimportant to the subspecies. The Recovery Teamconcentrated its management recommendationson habitats known to be important to the owl(see Parts II and III), while allowing otherecosystem management objectives, such asconservation of other species, to drive manage-ment of habitats where spotted owls are a sec-ondary concern.

The Recovery Team believes that it is impor-tant to evaluate the effects of implementingRecovery Plan management recommendationson other endangered, threatened, sensitive,candidate, or other species of concern. In addi-tion, it is important that the recommendationsfor Mexican spotted owl management be com-pared with the recommendations in otherspecies’ recovery or management plans. If con-flicts are identified, they need to be resolved byappropriate land managers and/or scientists.

An important objective in management offorested ecosystems should be to address foresthealth problems, return forested ecosystems toconditions within their natural range of varia-tion, and work toward sustainable and resilientecosystems. The goals of this Recovery Plan andecosystem management principles are compat-ible. Proper ecosystem management will providefor landscapes in which spotted owls and otherecosystem components persist within the rangeof their evolutionary adaptations. The metricused to measure progress should be the amountof acreage successfully treated to meet a desiredresult, and not commodity-based measures suchas “board feet” or “animal unit months.” Com-modities will undoubtedly be a byproduct offorested ecosystem management, but should notbe the driving consideration.

ECONOMIC CONSIDERATIONSECONOMIC CONSIDERATIONSECONOMIC CONSIDERATIONSECONOMIC CONSIDERATIONSECONOMIC CONSIDERATIONS

Protecting threatened and endangeredspecies can conflict with other resource objec-

tives. These conflicts can become more intensewhen species conservation efforts restrict eco-nomic returns from lands people depend uponfor their livelihoods and communities dependupon for their very existence. Whether conflictsbetween species conservation and economicreturn are real or perceived, human concernsshould be considered so long as the conservationgoal is achieved.

As mentioned previously, the RecoveryTeam’s charge was to develop a plan that wouldlead to recovery of the Mexican spotted owl.However, specific cause-effect relationships ofmany management activities on individual owlsand pairs or in relation to population processesare not entirely clear. Given these uncertainties,it may be tempting to take a conservative ap-proach to recovery by recommending cessationof all anthropogenic activities for which effectsof the activity on the target species are poorlyunderstood. However, recommendations forresource management should be based onestablished information. The absence of neededinformation should stimulate research, and theresults of that research should guide manage-ment. The only way to understand the cause-and-effect relationships between managementactions and specific resources is by studyingthem, preferably through controlled experi-ments.

The recommendations contained hereinallow most land-management activities to occurprovided that the effects of those activities areevaluated during the recovery period. In addi-tion, the Recovery Plan recommends thatscientific monitoring of the Mexican spotted owlpopulation and its habitat should accompanythose activities to assess their impact on spottedowl populations. If warranted, these activitiescan be altered or eliminated if monitoring orresearch indicates a significant risk to the spottedowl population. In other cases, restrictions onhuman activities are recommended where datashow a high likelihood that the spotted owl’spersistence may be significantly compromised ifcertain land-management practices continue.

Obviously, the decision on which activitiesmust be altered or eliminated and which mayproceed if closely monitored cannot be madewith absolute certainty. Such decisions require

16


Volume I/Part I

professional judgement of the most qualifiedscientists using the best available data. The FWSselected the Recovery Team members with thatfact in mind. However, the FWS also intends touse the expertise of others who may contributeuseful information to improve management ofthe Mexican spotted owl.

Any conservation plan, regardless of species,must include the considerations discussed abovewhen uncertainties exist. The FWS is confidentthat this Recovery Plan, given its inherentflexibility, has a high likelihood of leading to therecovery of the Mexican spotted owl withoutcausing unacceptable levels of economic andsocial hardship during its implementation.

HUMAN INTERVENTIONHUMAN INTERVENTIONHUMAN INTERVENTIONHUMAN INTERVENTIONHUMAN INTERVENTIONAND NATURAL PROCESSESAND NATURAL PROCESSESAND NATURAL PROCESSESAND NATURAL PROCESSESAND NATURAL PROCESSES

Much criticism directed at Mexican spottedowl management centers on the concept thattoday’s southwestern forests are in an unnaturalstate; that grazing, fire suppression, forestrypractices, and other anthropogenic processeshave led to forest conditions much denser thanthose existing during presettlement times. Aconcurrent increase in mixed-conifer forests isalso believed to have occurred. These pointslead some to the conclusion that the Mexicanspotted owl population is at an all-time (andunsustainable) high. The Recovery Team isunaware of data that clearly support that conclu-sion, and questions whether stands recentlyconverted to mixed-conifer forest possess thestructural characteristics utilized by the subspe-cies. The Recovery Team acknowledges thathumans have had a pronounced influence oncontemporary forest conditions; however, theeffects of human activities on the Mexicanspotted owl population are unknown. Even ifone accepts that the spotted owl population inmixed-conifer forest is unnaturally high, onemust also consider that other habitats that mayhave been important historically, such as lower-and middle-elevation riparian areas, have beendramatically reduced. These two trends may beoffsetting, and the net gain or loss of spotted owlcarrying capacity can only be speculated upon.

It would be imprudent, if not impossible, todevelop a management plan for a species byspeculating on its status in the distant past;rather, the appropriate approach is to acknowl-edge that we are dealing with a drastically alteredlandscape and that a return to presettlementconditions is impossible. In that light, theRecovery Team acknowledges that humans havea major role to play in management of thespotted owl and the forests of the Southwest.The Recovery Team believes that a viable forest-products industry is critical in carrying out themanagement actions recommended in thisRecovery Plan, making it an essential agent ofplan implementation.

DIFFERENCES BETWEENDIFFERENCES BETWEENDIFFERENCES BETWEENDIFFERENCES BETWEENDIFFERENCES BETWEENTHE DRAFT AND FINALTHE DRAFT AND FINALTHE DRAFT AND FINALTHE DRAFT AND FINALTHE DRAFT AND FINAL

RECOVERY PLANSRECOVERY PLANSRECOVERY PLANSRECOVERY PLANSRECOVERY PLANS

The Recovery Team considered all com-ments received on the draft Mexican SpottedOwl Recovery Plan. In addition, the draftRecovery Plan underwent extensive peer reviewfrom both purposely selected reviewers and“blind” reviewers selected by certain scientificsocieties (see Appendix E). These reviews led to afinal Recovery Plan that differs substantiallyfrom the draft version. We do not attempt todetail every difference between the two versionsof the Recovery Plan, but discuss these differ-ences in general terms.

Part II of the draft Recovery Plan containeda great deal of technical information. In theinterest of making the Recovery Plan an easierdocument to use, several of those chapters wereplaced in a companion volume to this RecoveryPlan. The information contained in thosechapters was integral to Recovery Plan develop-ment, so the main points in each are summa-rized in Part II of this final Recovery Plan.

Part III has changed substantially from thedraft version. Much of the background andjustification discussion has been moved to PartII, so that the current Part III deals strictly withthe management recommendations and delistingcriteria. This allows land managers to more easilypull the specific recommendations out of thePart III text. In addition, the management

17


Volume I/Part I

recommendations have changed considerably, inthat they are now in a more descriptive, ratherthan prescriptive, context. The changes were inresponse to comments from numerous landmanagers who expressed their concern that manyof the tools available to achieve land-manage-ment objectives were overly constrained. TheRecovery Team recognizes that the best approachis to describe the desired conditions on thelandscape, while providing land-managementprofessionals the flexibility to choose the tools toachieve the stated objectives.

The most significant change in Part IV isthat the responsibility for implementing some ofthe tasks recommended in this Recovery Planhas been distributed among different entities. Inaddition, the estimated costs of implementingspecific recovery tasks is provided.

18

Biological and Ecological Background

Volume I, Part II


Volume I/Part II

A. GENERAL BIOLOGYA. GENERAL BIOLOGYA. GENERAL BIOLOGYA. GENERAL BIOLOGYA. GENERAL BIOLOGYAND ECOLOGICAL RELATIONSHIPS OFAND ECOLOGICAL RELATIONSHIPS OFAND ECOLOGICAL RELATIONSHIPS OFAND ECOLOGICAL RELATIONSHIPS OFAND ECOLOGICAL RELATIONSHIPS OF

THE MEXICAN SPOTTED OWLTHE MEXICAN SPOTTED OWLTHE MEXICAN SPOTTED OWLTHE MEXICAN SPOTTED OWLTHE MEXICAN SPOTTED OWLSarah E. Rinkevich, Joseph L. Ganey, James P. Ward Jr.,

Gary C. White, Dean L. Urban, Alan B. Franklin,William M. Block, and Fernando Clemente

(Ridgway 1914) and the subspecific name waschanged to lucida to conform to taxonomicstandards. Monson and Phillips (1981) regardedthe Mexican spotted owl in Arizona as S. o.huachucae, noting that they were paler than S. o.lucida. However, this taxonomic designation wasnot followed by the AOU (1957).

The Mexican subspecies is geographicallyisolated from both the California and northernsubspecies. Using electrophoresis to examineallozyme variation, Barrowclough and Gutiérrez(1990) found a major allelic difference betweenthe Mexican spotted owl and the two coastalsubspecies. This difference suggests that theMexican spotted owl has been isolated geneti-cally from the other subspecies for considerabletime, has followed a separate evolutionaryhistory, and could therefore be considered aseparate species (Barrowclough and Gutiérrez1990:742).

Northern spotted owls are known to hybrid-ize with barred owls. Hybrids have been foundin Washington and Oregon (Hamer et al. 1992),and in California (Alan Franklin, HumboldtState Univ., Arcata, CA, pers. comm.). Thehybrids can be identified by their plumage,vocalizations, and morphology (Hamer et al.1992). Closely related species occasionallyhybridize naturally, especially where habitatdisruption has led to contact between speciespreviously isolated geographically (Short 1965,1972). Hybridization has not been reported inthe Mexican subspecies. The possibility ofhybridization exists in Mexico where barredowls, fulvous owls, and spotted owls overlap indistribution. No evidence currently exists docu-menting actual sympatry among these species,however.

This section presents a summary ofVolume 2, which examines aspects of the biologyand ecological relationships of Mexican spottedowls in more detail. This is not an exhaustivetreatment of the owl’s biology and ecology, but isintended to provide an overview of biologicalelements germane to recovering the Mexicanspotted owl. Although gaps still exist, ourunderstanding of the Mexican spotted owl’snatural history has increased with recent researchas well as data analyses accomplished by theRecovery Team.

A wealth of information exists for the north-ern and California spotted owls (Thomas et al.1990, Bart et al. 1992, Verner et al. 1992a,Gutiérrez et al., 1995). Although different insome respects, many aspects of the owls’ biologyand ecology are similar among the three subspe-cies. Thus, where appropriate, information fromthese subspecies was used for comparison orwhere data were limited regarding the Mexicanspotted owl.

TAXONOMYTAXONOMYTAXONOMYTAXONOMYTAXONOMY

Three species within the genus Strix occurnorth of Mexico: spotted (S. occidentalis), barred(S. varia), and great gray owls (S. nebulosa).Mexican spotted, barred, and fulvous owls (S.fulvescens) occur in Mexico. The Mexican spot-ted owl (S. o. lucida) is one of three subspecies ofspotted owl recognized by the American Orni-thologists’ Union (AOU) in its last checklist thatincluded subspecies (AOU 1957:285). The othertwo subspecies are the northern (S. o. caurina)and the California spotted owl (S. o. occidentalis)(AOU 1957; Figure.II.A.1).

The Mexican subspecies was first describedfrom a specimen collected at Mount Tancitaro,Michoacan, Mexico and named Syrniumoccidentale lucidum (Nelson 1903). The spottedowl was later assigned to the genus Strix

19


Volume I/Part II

FFFFFigurigurigurigurigure II.A.1. e II.A.1. e II.A.1. e II.A.1. e II.A.1. Geographic range of the spotted owl.

20


Volume I/Part II

DESCRIPTIONDESCRIPTIONDESCRIPTIONDESCRIPTIONDESCRIPTION

The spotted owl is mottled in appearancewith irregular white and brown spots on itsabdomen, back, and head. The spots of theMexican spotted owl are larger and more numer-ous than in the other two subspecies, giving it alighter appearance. Strix occidentalis translates as“owl of the west” and lucida means “light” or“bright.” Unlike most owls, spotted owls havedark eyes. Several thin white bands mark anotherwise brown tail.

Adult male and female spotted owls aremostly monochromatic in plumage characteris-tics, but the sexes can be readily distinguished byvoice (see below). Juveniles, subadults, andadults can be distinguished by plumage charac-teristics (Forsman 1981, Moen et al. 1991).Juvenile spotted owls (hatchling to approxi-mately five months) have a downy appearance.Subadults (5 to 26 months) closely resembleadults, but have pointed retrices with a purewhite terminal band (Forsman 1981, Moen et al.1991). The retrices of adults (>27 months) haverounded tips, and the terminal band is mottledbrown and white.

Although the spotted owl is often referred toas a medium-sized owl, it ranks among thelargest owls in North America. Of the 19 speciesof owls that occur in North America, only 4 arelarger than the spotted owl (Johnsgard 1988).Like many other owls, spotted owls exhibitreversed sexual dimorphism (i.e., females arelarger than males). Adult male Mexican spottedowls (n = 37) average 519 + 32.6 (SD) g(18.5 oz), and adult females (n = 31) average 579+ 31.2 g (20.7 oz) (Kristan et al., in prep.).There appears to be clinal variation among thethree subspecies in a number of morphologicalcharacteristics measured, with size decreasingfrom north to south (Kristan et al., in prep.).

DISTRIBUTIONDISTRIBUTIONDISTRIBUTIONDISTRIBUTIONDISTRIBUTIONAND ABUNDANCEAND ABUNDANCEAND ABUNDANCEAND ABUNDANCEAND ABUNDANCE

The Recovery Team gathered and examinedinformation on the distribution and abundanceof Mexican spotted owls through 1993. Datafrom surveys conducted after 1993 were notavailable for our analyses.

We used the information collected to (1)document historical and current range of thissubspecies, (2) help formulate Recovery Unitboundaries, and (3) provide a template foranalyses at the landscape scale. Descriptions ofRecovery Units are provided in the followingchapter (II.B).

The Mexican spotted owl currently occupiesa broad geographic area, but does not occuruniformly throughout its range (Figure II.A.2).Instead, the owl occurs in disjunct localities thatcorrespond to isolated mountain systems andcanyons. In the United States, 91% of the owlsknown to exist between 1990 and 1993 occur onlands administered by the FS (Table II.A.1).Other lands currently occupied by Mexicanspotted owls in the United States include, NPS(4%), BLM (2%), Tribal (2%), and DOD (1%).We know that more owls occur on Tribal landsthan indicated here, but specific information onnumbers of owls known on Tribal lands was notmade available to the Team. Owl distributionaccording to land ownership is unavailable forMexico. Eighty-nine percent of the owls knownto exist between 1990 and 1993 in Mexico werein the States of Sonora and Chihuahua (TableII.A.1, Figure II.A.3). However, most surveyefforts in Mexico were restricted to these states,and these numbers do not necessarily reflectactual trends in distribution.

The current owl distribution mimics itshistorical extent, with a few exceptions. The owlhas not been reported recently along majorriparian corridors in Arizona and New Mexico,nor in historically documented areas of southernMexico. Riparian communities and previouslyoccupied localities in the southwestern UnitedStates and southern Mexico have undergonesignificant habitat alteration since the historicalsightings (USDI 1993). However, the amount ofeffort devoted to surveying these areas is un-known and future surveys may documentspotted owls there. Surveys conducted to relocatespotted owls in northern Colorado near FortCollins and Boulder, where records exist fromthe early 1970s and 1980s, have been unsuccess-ful. Surveys conducted in the Book Cliffs of east-central Utah, where owls were recorded in 1958,have also been unsuccessful. Although historical(pre-1990) data provide some information about

21


Volume I/Part II

FFFFFigurigurigurigurigure II.A.2.e II.A.2.e II.A.2.e II.A.2.e II.A.2. Current distribution of Mexican spotted owls in the United States based on plannedsurveys and incidental observations recorded from 1990 through 1993.

22


Volume I/Part II

TTTTTable II.A.1.able II.A.1.able II.A.1.able II.A.1.able II.A.1. Historical records and minimum numbers of Mexican spotted owls found duringplanned surveys and incidental observations, reported by Recovery Unit and land ownership. RecoveryUnits are described in Part II.B.

NNNNNumber of umber of umber of umber of umber of 11111 N N N N Number ofumber ofumber ofumber ofumber ofooooowl rwl rwl rwl rwl recorecorecorecorecordsdsdsdsds o o o o owl siteswl siteswl siteswl siteswl sites

RRRRRecoecoecoecoecovvvvvererererery Uy Uy Uy Uy Unitnitnitnitnit beforbeforbeforbeforbefore 1990e 1990e 1990e 1990e 1990 1990-1993 1990-1993 1990-1993 1990-1993 1990-1993

UNITED STUNITED STUNITED STUNITED STUNITED STAAAAATESTESTESTESTESColorado Plateau

FS 21 16BLM 6 10NPS 34 23Tribal 20 13 2

New Mexico State 1 0Unknown 3 5 0

SSSSSubtotalubtotalubtotalubtotalubtotal 87 87 87 87 87 62 62 62 62 62Southern Rocky Mountains – Colorado

FS 2 8BLM 0 6NPS 0 0Tribal 1 -- 2

Unknown 3 17 0SSSSSubtotalubtotalubtotalubtotalubtotal 20 20 20 20 20 14 14 14 14 14

Southern Rocky Mountains – New MexicoFS 25 34NPS 3 0New Mexico State 1 0Private 4 0Unknown 3 8 0

SSSSSubtotalubtotalubtotalubtotalubtotal 41 41 41 41 41 34 34 34 34 34Upper Gila Mountains

FS 138 424BLM 5 0NPS 5 0Tribal 0 -- 2

Private 1 0Unknown 3 104 0

SSSSSubtotalubtotalubtotalubtotalubtotal 253253253253253 424424424424424Basin and Range – West

FS 82 97NPS 13 0Tribal 0 -- 2

DOD 9 6Private 8 0Unknown 3 57 0

SSSSSubtotalubtotalubtotalubtotalubtotal 169169169169169 103103103103103

23


Volume I/Part II

NNNNNumber of umber of umber of umber of umber of 11111 NNNNNumber ofumber ofumber ofumber ofumber ofooooowl rwl rwl rwl rwl recorecorecorecorecordsdsdsdsds ooooowl siteswl siteswl siteswl siteswl sites

RRRRRecoecoecoecoecovvvvvererererery Uy Uy Uy Uy Unitnitnitnitnit beforbeforbeforbeforbefore 1990e 1990e 1990e 1990e 1990 1990-19931990-19931990-19931990-19931990-1993

Basin and Range – EastFS 18 111BLM 1 0NPS 6 10Tribal 2 -- 2

FWS 1 0Private 2 0

SSSSSubtotalubtotalubtotalubtotalubtotal 30 30 30 30 30 121121121121121

UUUUUnited Snited Snited Snited Snited States tates tates tates tates TTTTTotalotalotalotalotal 600600600600600 758758758758758

MEXICOMEXICOMEXICOMEXICOMEXICOSierra Madre Occidental – Norte

Sonora 8 9Chihuahua 10 8 4

SSSSSubtotalubtotalubtotalubtotalubtotal 18 18 18 18 18 17 17 17 17 17Sierra Madre Oriental – Norte

Coahuila 2 0Sierra Madre Occidental – Sur

Sinaloa 1 0Durango 2 0Aguacalientes 0 1 4

Zacatecas 0 0 4

San Luis Potosi 1 0Guanajuato 1 0

SSSSSubtotalubtotalubtotalubtotalubtotal 5 5 5 5 5 1 1 1 1 1Sierra Madre Oriental – Sur

Coahuila 4 0Nuevo Leon 4 1Tamaulipas 0 0 4

SSSSSubtotalubtotalubtotalubtotalubtotal 8 8 8 8 8 1 1 1 1 1Eje Neovolcanico

Jalisco 1 0Colima 1 5 0Michoacan 1 0Puebla 1 5 0

SSSSSubtotalubtotalubtotalubtotalubtotal 2 2 2 2 2 0 0 0 0 0

MMMMMexico exico exico exico exico TTTTTotalotalotalotalotal 35 35 35 35 35 19 19 19 19 19

1 These values do not necessarily indicate numbers of owls or owl sites because multiple records may exist from thesame site through time.

2 Additional owls are known to exist on many Tribal lands, but the exact number is unavailable.3 Locations of these records were insufficient for assigning a land ownership.4 Additional sightings have been reported from 1994 surveys.5 Unverified record not included in totals (see text).

TTTTTable II.A.1.able II.A.1.able II.A.1.able II.A.1.able II.A.1. continued

24


Volume I/Part II

FFFFFigurigurigurigurigure II.A.3.e II.A.3.e II.A.3.e II.A.3.e II.A.3. Current distribution of Mexican spotted owls in Mexico based on planned surveysand incidental observations recorded from 1990 through 1993.

25


Volume I/Part II

about past distribution of spotted owls, we stressthat it is not sufficient to allow us to estimatechanges in the number or distribution of owlsfrom historical to present time.

Most current observations of Mexicanspotted owls are from the Upper Gila MountainsRU (see II.B). This unit can be considered acritical nucleus for the subspecies because of itscentral location within the owl’s range and itsseemingly high number of owls. Other areaslikely to be important population centers in-clude the sky islands of southeastern Arizona andthe Sacramento Mountains of central NewMexico (Basin and Range RUs; see II.B). Al-though information on owl numbers permits aview of the current distribution, it is not com-plete enough to provide a reliable estimate oftotal population size.

Mexican spotted owls occur at higher densi-ties in mixed-conifer forests than in pine-oak,pine, and pinyon-juniper forest types (Skaggsand Raitt 1988, White et al. 1995). A combinedestimate of Mexican spotted owl density on twostudy areas in the Upper Gila Mountains RU issimilar to estimates reported for other spottedowl subspecies.

In summary, the Mexican spotted owl isdistributed discontinuously throughout itsrange, with its distribution largely restricted tomontane forests and canyons. Although futureefforts will undoubtedly discover additionalowls, their documented spatial distribution inthe United States is not likely to change greatly.The converse is true for Mexico, where plannedsurveys have begun only recently.

HABITAT ASSOCIATIONSHABITAT ASSOCIATIONSHABITAT ASSOCIATIONSHABITAT ASSOCIATIONSHABITAT ASSOCIATIONS

Mexican spotted owls nest, roost, forage, anddisperse in a diverse array of biotic communities.Mixed-conifer forests are commonly usedthroughout most of the range (Johnson andJohnson 1985, Skaggs and Raitt 1988, Ganey etal. 1988, Ganey and Balda 1989a, Rinkevich1991, Willey 1993, Fletcher and Hollis 1994,Seamans and Gutiérrez, in press). In general,these forests are dominated by Douglas-fir and/or white fir, with codominant species includingsouthwestern white pine, limber pine, andponderosa pine (Brown et al. 1980). The under-

story often contains the above coniferous speciesas well as broadleaved species such as Gambeloak, maples, boxelder, and New Mexico locust.In southern Arizona and Mexico, Madrean pine-oak forests are also used commonly (Ganey andBalda 1989a, Duncan and Taiz 1992, Ganey etal. 1992, Tarango et al. 1994). These forests aretypically dominated by an overstory of Chihua-hua and Apache pines in conjunction withspecies such as Douglas-fir, ponderosa pine, andArizona cypress. Evergreen oaks are typicallyprominent in the understory (Brown et al.1980).

Habitat-use patterns vary throughout therange and with respect to owl activity (seebelow). In the northern portion of the range,including southern Utah, southern Colorado,and far northern Arizona and New Mexico, owlsoccur primarily in steep-walled, rocky canyons(Kertell 1977, Reynolds 1990, Rinkevich 1991,Willey 1993). Along the Mogollon Rim inArizona and New Mexico, habitat use is lessrestricted, and spotted owls occur in mixed-conifer forests, ponderosa pine-Gambel oakforests, rocky canyons, and associated riparianforests (Ganey and Balda 1989a, Ganey et al.1992, Fletcher and Hollis 1994, Seamans andGutiérrez, in press, Peter Stacey, Univ. ofNevada, Reno, pers. comm.). South of theMogollon Rim and into Mexico a still widervariety of habitat types are used, includingmixed-conifer, Madrean pine-oak, and Arizonacypress forests, encinal oak woodlands, andassociated riparian forests (Ganey and Balda1989a, Duncan and Taiz 1992, Ganey et al.1992, Tarango et al. 1994). Much of this re-gional variation in habitat use likely results fromdifferences in regional patterns of habitat andprey availability.

Nesting and Roosting HabitatNesting and Roosting HabitatNesting and Roosting HabitatNesting and Roosting HabitatNesting and Roosting Habitat

Mexican spotted owls nest and roost prima-rily in closed-canopy forests or rocky canyons. Inthe northern portion of the range (southernUtah and Colorado), most nests are in caves oron cliff ledges in steep-walled canyons. Else-where, the majority of nests appear to be in trees(Fletcher and Hollis 1994).

26


Volume I/Part II

Forests used for roosting and nesting oftencontain mature or old-growth stands withcomplex structure (Skaggs and Raitt 1988,Ganey and Balda 1989a, 1994; McDonald et al.1991, Seamans and Gutiérrez, in press). Theseforests are typically uneven-aged, multistoried,and have high canopy closure. Nest trees aretypically large in size (SWCA 1992, Fletcher andHollis 1994, Seamans and Gutiérrez, in press),whereas owls roost in both large and small trees(Ganey 1988, Rinkevich 1991, Willey 1993,Zwank et al. 1994, Peter Stacey, Univ. ofNevada, Reno, pers. comm.). Tree species usedfor nesting vary somewhat among areas andhabitat types, but available evidence suggests thatDouglas-fir is the most common species of nesttree (SWCA 1992, Fletcher and Hollis 1994,Seamans and Gutiérrez, in press). A wider varietyof trees are used for roosting, but again Douglas-fir is the most commonly used species (Ganey1988, Fletcher and Hollis 1994, Zwank et al.1994, Peter Stacey, Univ. of Nevada, Reno, pers.comm.).

Several hypotheses have been proposed toexplain why spotted owls nest in closed-canopyforests (reviewed by Carey 1985, Gutiérrez1985). Barrows (1981) suggested that spottedowls are relatively intolerant of high tempera-tures and roost and nest in shady forests becausethey provide favorable microclimatic conditions.Ganey et al. (1993) observed that Mexicanspotted owls produced more metabolic heat thangreat horned owls, and were less able to dissipatethat heat. This may lead these owls to seek outcool microsites during periods of high ambienttemperature. Mexican spotted owls typically nestand roost in closed-canopy forests or deep shadycanyons; both situations provide cool microsites(Kertell 1977, Ganey et al. 1988, Rinkevich1991, Ganey and Balda 1989a, Willey 1993).

Foraging HabitatForaging HabitatForaging HabitatForaging HabitatForaging Habitat

Little is known about patterns of habitat useby foraging owls. The only available data de-scribe habitat use by eight owls occupying fivehome ranges on three study areas in northernArizona (Ganey and Balda 1994). In general,owls foraged more than or as expected inunlogged forests, and less than or as expected in

selectively logged forests. Expected values werebased on relative occurrences of habitats. How-ever, patterns of habitat use varied among studyareas and individuals, and even between pairmembers in some cases, making generalizationsdifficult. Both high-use roosting and high-useforaging sites had more big logs, higher canopyclosure, and greater densities and basal areas ofboth trees and snags than random sites. Owlsclearly used a wider variety of forest conditionsfor foraging than they used for roosting (Ganeyand Balda 1994). We caution, however, aboutextending these results too far given the limitednumber of owls sampled, and the variabilityobserved among owls and sites.

TERRITORIALITYTERRITORIALITYTERRITORIALITYTERRITORIALITYTERRITORIALITYAND HOME RANGEAND HOME RANGEAND HOME RANGEAND HOME RANGEAND HOME RANGE

Home range is defined as the area used by ananimal during its normal activities (Burt 1943)whereas territory is a defended area within anindividual’s home range (Nice 1941). Territoriesare typically smaller than home ranges, but theexact relationship between the territory and thehome range is generally not known. Fidelity toterritories is apparently high in Mexican spottedowls, with most owls remaining on the sameterritory year after year.

Home-range size of Mexican spotted owls,as estimated by monitoring movements ofradiotagged owls, appears to vary considerablyamong habitats and/or geographic areas (Ganeyand Dick 1995). Differences in sampling meth-ods among studies make comparisons difficult,however. Minimum convex polygon homerange estimates of home range-size variedfrom (1) 924 - 1,487 ha (2,282 - 3,672 acres)for individuals (n = 11) on three study areas inColorado Plateau RU (Willey 1993); (2) 261 -1,053 ha (645 - 2,601 acres) for individuals(n = 25) and 381 - 1,551 ha (941 - 3,831 acres)for pairs (n = 10) on five study areas in theUpper Gila Mountains RU (Ganey and Balda1989b, Ganey and Block, unpublished data,Peter Stacey, Univ. of Nevada, Reno, pers.comm.); and (3) 452 - 937 ha (1,116 - 2,314acres) for individuals (n = 20) and 573 - 1,401ha (1,415 - 3,461 acres) for pairs (n = 8) on twostudy areas in the Basin and Range-East RU

27


Volume I/Part II

(Zwank et al. 1994, Ganey and Block, unpub-lished data).

VOCALIZATIONSVOCALIZATIONSVOCALIZATIONSVOCALIZATIONSVOCALIZATIONS

The spotted owl, being primarily nocturnal,is more often heard than seen. It has a widerepertoire of calls (Forsman et al. 1984, Ganey1990) that are relatively low-pitched and com-posed of pure tones (Fitton 1991). Sexes can bedistinguished by calls; males have a deeper voicethan females and generally call more frequently.The most common vocalization, used moreoften by males, is a series of four unevenly-spaced hoots. Females frequently use a clearwhistle ending with an upward inflection as wellas a series of sharp barks. Forsman et al. (1984)described 14 calls for the northern spotted owl,of which 10 were reported by Ganey (1990) forArizona birds.

Mexican spotted owls call mainly fromMarch - November and are relatively silent fromDecember - February (Ganey 1990). Callingactivity increases from March through May(although nesting females are largely silentduring April and early May), and then declinesfrom June through November (Ganey 1990).On a daily basis, calling activity is greatestduring the 2-hour period following sunset, withsmaller peaks 4-8 hours after sunset and justbefore sunrise. Owls called more than expectedduring the last quarter and new moon phases ofthe lunar cycle; and they called most frequentlyon calm, clear nights when no precipitation wasfalling (Ganey 1990).

INTERSPECIFIC COMPETITIONINTERSPECIFIC COMPETITIONINTERSPECIFIC COMPETITIONINTERSPECIFIC COMPETITIONINTERSPECIFIC COMPETITION

Several other species of owls occur within therange of the Mexican spotted owl. Interferencecompetition, where individuals physicallyinterfere with each other, probably does notoccur to any great extent between the Mexicanspotted owl and other owl species. However,exploitative competition, where individualscompete for similar resources such as prey ornest sites, may occur. Competition might begreatest between spotted and great horned owlsbecause both species are relatively large andwidely distributed. Preliminary data from a

telemetry study in northern Arizona suggest thatthese owls overlap broadly in diet and space, butexhibit some habitat segregation (Ganey andBlock, unpublished data). If Mexican spottedand barred owls are sympatric in Mexico, thencompetition might also occur between theseclosely related species. In general, however, moreresearch is needed to assess the potential occur-rence and importance of interspecific competi-tion between spotted and other owls.

FEEDING HABITS AND PREYFEEDING HABITS AND PREYFEEDING HABITS AND PREYFEEDING HABITS AND PREYFEEDING HABITS AND PREYECOLOGYECOLOGYECOLOGYECOLOGYECOLOGY

Forsman (1976) described spotted owls as“perch and pounce” predators. They typicallylocate prey from an elevated perch by sight orsound, then pounce on the prey and capture itwith their talons. Spotted owls have also beenobserved capturing flying prey such as birds andinsects (Verner et al. 1992b). They hunt prima-rily at night (Forsman et al. 1984, Ganey 1988),although infrequent diurnal foraging has beendocumented (Forsman et al. 1984, Laymon1991, Sovern et al. 1994).

Mexican spotted owls consume a variety ofprey throughout their range but commonly eatsmall- and medium-sized rodents such aswoodrats, peromyscid mice, and microtine voles.Spotted owls also consume bats, birds, reptiles,and arthropods. The diet varies by geographiclocation (Ward and Block 1995; Figure. II.A.4).For example, spotted owls dwelling in canyonsof the Colorado Plateau take more woodrats,and fewer birds, than do spotted owls from otherareas (Ward and Block 1995, Figure. II.A.4). Incontrast, spotted owls occupying mountainranges with forest-meadow interfaces, as foundwithin the Basin and Range - East, SouthernRocky Mountains - Colorado, and Upper GilaMountains RUs, take more voles (Ward andBlock 1995, Figure. II.A.4). Regional differencesin the owl’s diet likely reflect geographic varia-tion in population densities and habitats of boththe prey and the owl.

The Team was unable to link consumptionof specific prey and successful reproduction bythe Mexican spotted owl, with two possibleexceptions. First, fecundity of spotted owlsoccupying the Sacramento Mountains (Basin

28


Volume I/Part II

and Range - East RU) appeared to be associatedwith trends in abundance of peromyscid mice(Ward and Block 1995). Second, the predomi-nance of woodrats in the diet throughout muchof the owl’s range suggests that this prey mayinfluence the owl’s fitness. Other studies haveshown positive associations between larger prey(e.g., woodrats) in the diet of northern andCalifornia spotted owls and their reproductivesuccess (Barrows 1987, Thrailkill and Bias1989). In most cases, however, total prey biom-ass may be more influential on the owl’s fitnessthan the abundance of any particular preyspecies.

Habitat correlates of the owl’s common preyemphasize that each prey species uses a uniquemicrohabitat. For example, deer mice are ubiqui-tous in distribution, occupying areas withvariable conditions, whereas brush mice arerestricted to communities with a strong oakcomponent and dry, rocky substrates with sparsetree cover. Mexican woodrats are typically foundin areas with considerable shrub or understorytree cover, little herbaceous cover, and high logvolumes. Mexican voles are found in areas withhigh herbaceous cover, primarily grasses. Long-tailed voles are associated with high herbaceouscover, primarily forbs, many shrubs, and limitedtree cover. Thus, to provide a diverse prey base,managers should provide diverse habitats forprey species. Managing habitat for a diversity ofprey species may help buffer against populationfluctuations of individual prey species andprovide a more constant food supply for the owl.

REPRODUCTIVE BIOLOGYREPRODUCTIVE BIOLOGYREPRODUCTIVE BIOLOGYREPRODUCTIVE BIOLOGYREPRODUCTIVE BIOLOGY

Knowledge of the annual reproductive cycleof the Mexican spotted owl is important both inan ecological context, and for placing seasonalrestrictions on management or on other activitiesthat may occur within areas occupied by spottedowls. Data on the reproductive cycle of theMexican spotted owl are limited compared toinformation on the northern and Californiasubspecies. Therefore, although the followingdiscussion is based primarily on observations ofthe Mexican spotted owl, data from the othersubspecies are provided to fill some informationgaps.

Mexican spotted owls nest on cliff ledges,stick nests built by other birds, debris platformsin trees, and in tree cavities (Johnson andJohnson 1985, Ganey 1988, SWCA 1992,Fletcher and Hollis 1994, Seamans andGutiérrez, in press). Spotted owls have one ofthe lowest clutch sizes among North Americanowls (Johnsgard 1988). Females normally layone to three eggs, two being most common. Re-nesting following nest failure is unusual, but hasbeen observed in Mexican spotted owls (Kroel1991, David Olson, Humboldt State Univ.,Arcata, CA, pers. comm.). Mexican spotted owlsbreed sporadically and do not nest every year(Ganey 1988). In good years most of the popu-lation will nest, whereas in other years only asmall proportion of pairs will nest successfully(Fletcher and Hollis 1994). Reasons for thispattern are unknown.

Mexican spotted owls have distinctannual breeding periods, but reproductivechronology varies somewhat across the range ofthe owl. In Arizona, courtship apparently beginsin March with pairs roosting together during theday and calling to each other at dusk (Ganey1988). Eggs are laid in late March or, moretypically, early April. Incubation begins shortlyafter the first egg is laid, and is performedentirely by the female (Ganey 1988). Femalenorthern spotted owls incubate for approxi-mately 30 days (Forsman et al. 1984), andMexican spotted owls appear to incubate for asimilar period (Ganey 1988). During incubationand the first half of the brooding period, thefemale leaves the nest only to defecate, regurgi-tate pellets, or to receive prey delivered by themale, who does most or all of the foraging(Forsman et al. 1984, Ganey 1988).

The eggs usually hatch in early May (Ganey1988). Females brood their young almostconstantly for the first couple of weeks after theeggs hatch but then begin to spend time huntingat night, leaving the owlets unattended for up toseveral hours (Eric Forsman, FS, Corvallis, OR,pers. comm.). Nestling owls generally fledge fourto five weeks after hatching in early to mid-June(Ganey 1988). Owlets usually leave the nestbefore they can fly, jumping from the nest on tosurrounding tree branches or the ground(Forsman et al. 1984, Ganey 1988). Owlets that

29


Volume I/Part II

FFFFFigurigurigurigurigure II.A.4.e II.A.4.e II.A.4.e II.A.4.e II.A.4. Geographic variability in the food habits of Mexican spotted owls presented as relativefrequencies of (a)(a)(a)(a)(a) woodrats, (b)(b)(b)(b)(b) voles, (c)(c)(c)(c)(c) gophers, (d)(d)(d)(d)(d) birds, (e)(e)(e)(e)(e) bats, and (f(f(f(f(f ))))) reptiles. Point values arefrom single studies or averages among the number of studies shown in parenthesis (a)(a)(a)(a)(a). Vertical bars are95% confidence intervals showing sampling and interstudy variation within a recovery unit. Recoveryunit acronyms are COPLAT-Colorado Plateau; SRM-CO-Southern Rocky Mountain-Colorado; SRM-NM-Southern Rocky Mountain-New Mexico; UPGIL-Upper Gila Mountain; BAR-W-Basin andRange - West; BAR-E-Basin and Range - East; SMO-N-Sierra Madre Occidental-Norte.

30


Volume I/Part II

end up on the ground will often climb back up atree to a safe roost site. The mobility and forag-ing skills of owlets improve gradually during thesummer. Within a week after leaving the nest,most owlets can make short, clumsy flightsbetween trees. Three weeks after leaving the nest,owlets can hold and tear up prey on their own(Forsman et al. 1984).

Fledglings depend on their parents for foodduring the early portion of the fledgling period.Hungry owlets give a persistent, raspy “beggingcall,” especially when adults appear with food orcall nearby (Forsman et al. 1984, Ganey 1988).Begging behavior declines in late August, butmay continue at low levels until dispersal occurs,usually from mid September to early October(Ganey and Block, unpubl. data, Peter Stacey,Univ. of Nevada, Reno, pers. comm., DavidWilley, Northern Arizona Univ., Flagstaff, pers.comm.).

MORTALITY FACTORSMORTALITY FACTORSMORTALITY FACTORSMORTALITY FACTORSMORTALITY FACTORS

Several mortality factors (discussed below)have been identified as potentially importantwith respect to the Mexican spotted owl. Al-though a number of owls have been recoveredfollowing mortality and examined by both fieldbiologists and laboratory personnel, in generallittle is known about the extent or importance ofthese mortality factors.

PredationPredationPredationPredationPredation

Predation, particularly by avian predators,may be a common mortality factor of spottedowls. Potential avian predators of Mexicanspotted owls include great horned owls, northerngoshawks, red-tailed hawks, and golden eagles.Some of these predators occupy the same generalhabitats as the spotted owl, but there is littledirect evidence that they prey on spotted owls toany great extent (Gutiérrez et al. 1995). Ganey(1988) reported one instance of apparent greathorned owl predation on an adult spotted owl,and Richard Reynolds (FS, Fort Collins, CO.,pers. comm.) reported a golden eagle preying ona spotted owl. Preliminary results from radio-tagged Mexican spotted owls indicate that both

adults and juveniles are preyed upon (Willey1993, Ganey and Block, unpubl. data), but inmost cases the identity of the predator wasunknown. Further, in southern Arizona,procyonid mammals were observed attemptingto raid cliff site nests occupied by spotted owls(Russell Duncan, Southwestern Field Biologists,Tucson, AZ, pers. comm.). Thus, the extent towhich Mexican spotted owls are preyed upon isunknown at this time.

StarvationStarvationStarvationStarvationStarvation

Starvation is likely another common sourceof mortality. Juvenile northern spotted owls maybe more vulnerable to starvation than adults(Gutiérrez et al. 1985, Miller 1989), because oftheir poor hunting skills. Starvation may alsoresult from low abundance or availability of prey,which could affect both adults and juveniles.Both adult and juvenile owls radio-tagged inArizona have been found dead of apparentstarvation (Ganey and Block, unpublished data),and two of seven radio-tagged juveniles in Utahdied of starvation (Willey 1993). Most instancesof starvation occurred from late fall throughwinter, when prey resources were reduced inabundance and availability (Willey 1993, Blockand Ganey, unpublished data). In addition,starvation may predispose young or even adultsto predation.

AccidentsAccidentsAccidentsAccidentsAccidents

Accidents may be another mortality factor.For example, instances of spotted owls being hitby cars have been documented (Roger Skaggs,New Mexico State Univ, Las Cruces, pers.comm; Russell Duncan, Southwestern FieldBiologists, Tucson, AZ, pers. comm.). Owlsflying at night might also collide withpowerlines, tree branches, or other obstacles.This might be particularly true for birds migrat-ing or dispersing through unfamilar terrain.Again, little information is available on howfrequently this might occur.

31


Volume I/Part II

Disease and ParasitesDisease and ParasitesDisease and ParasitesDisease and ParasitesDisease and Parasites

Little is known about how disease andparasites contribute to mortality of spotted owls.Hunter et al. (1994) found a larval mite and liceon 2 of 28 museum specimens of Mexicanspotted owls examined for parasites, and 6 of 18live owls examined had hippoboscid fly larvae intheir ears. Some of the live owls examined alsohad lice. Hunter et al. (1994) reached no conclu-sions concerning mortality and ectoparasites inspotted owls, but did suggest that larval infesta-tions in their ears could affect the owls’ hearing.Because hearing is important for foraging atnight, such infestations could impact the birds’ability to hunt effectively.

In general, however, spotted owls may beadapted to high parasite loads. In a survey ofblood parasites in all three subspecies of spottedowls, Gutiérrez (1989) found an infection rate of100 percent. Although disease and parasitescould predispose owls to death by starvation,predation, or accident, no evidence exists docu-menting disease and parasites as direct mortalityfactors within the Mexican subspecies.

POPULATION BIOLOGYPOPULATION BIOLOGYPOPULATION BIOLOGYPOPULATION BIOLOGYPOPULATION BIOLOGY

The Mexican Spotted Owl population for aspecific area can be modeled with the simpleequation

Nt+1 = Nt + Bt - Dt + It - Et ,

where Nt is the population size at time t, Bt is thenumber of new birds recruited into the popula-tion (births), Dt is the number of birds dying, It

is the number of birds immigrating into thepopulation, and Et is the number of birds emi-grating from the population. The combinedeffect of births, deaths, immigrations, andemigrations dictate the viability of the popula-tion, and hence its long-term persistence.

SurvivalSurvivalSurvivalSurvivalSurvival

Annual survival rates of adult Mexicanspotted owls is 0.8-0.9 based on short-termpopulation and radio-tracking studies andlonger-term monitoring studies (White et al.

1995). These annual survival estimates can beviewed as the probability of an individual surviv-ing from one year to the next or as the propor-tion of individuals that will survive from oneyear to the next. A variety of different estimatorsof adult survival using different types and sets ofdata gave similar results.

Juvenile survival is considerably lower ( 0.06-0.29) than adult survival. Juvenile survival alsoappears more spatially variable, although thisconclusion reflects only two population studyareas and two radio-telemetry studies spanningtwo years or less.

We strongly suspect that estimates of juve-nile survival from the population studies whichutilize mark-recapture methods are biased lowbecause of (1) a high likelihood of permanentdispersal (emigration) from the study area, and(2) a lag of several years before marked juvenilesreappear as territory holders, at which point theyare first detected for recapture. Juvenile northernspotted owls have a high dispersal capability(reviewed in Thomas et al. 1990). If Mexicanspotted owl juveniles have a similar dispersalcapability, we expect that a substantial portion ofmarked juveniles will emigrate from the respec-tive study areas. However, estimates from theradio-telemetry studies roughly corroborated thelow estimates from the population studies. Biasesin the radio-telemetry estimates of juvenilesurvival can result if radios significantly affecttheir survival. Whether radios or their attach-ment affect survival of northern spotted owls isdebatable (Paton et al. 1991, Foster et al. 1992).Concerning the second point, Franklin (1992)found a lag of 1-4 years between the time whenjuvenile northern spotted owls were banded andsubsequently recaptured. If this process is similarfor Mexican spotted owls, then the currentpopulation studies may be of insufficient dura-tion to adequately estimate juvenile survival.

In summary, current survival estimates arebased primarily on studies of insufficient dura-tion or studies not explicitly designed to estimatesurvival. In most cases, the data are too limitedto support or test the assumptions of the estima-tors used. However, the age- and sex-specificestimates of survival calculated here are useful atthis point as qualitative descriptors of the life-history characteristics of Mexican spotted owls.

32


Volume I/Part II

That is, Mexican spotted owls exhibit high adultand relatively low juvenile survival. In thisrespect, Mexican spotted owl survival probabili-ties appear similar to northern (see review inBurnham et al. 1994) and California spottedowls (Noon et al. 1992).

ReproductionReproductionReproductionReproductionReproduction

Reproductive output of Mexican spottedowls, defined as the number of young fledgedper pair, varies both spatially and temporally(White et al. 1995). Mexican spotted owls mayhave a higher average reproductive rate (1.001fledged young per pair) than the California(~0.712; Noon et al. 1992) and the northernspotted owl (~0.715; Thomas et al. 1993). Allthree subspecies exhibit temporal fluctuations inreproduction, although the amplitude of thosefluctuations may be greatest for the Mexicanspotted owl.

Environmental VariationEnvironmental VariationEnvironmental VariationEnvironmental VariationEnvironmental Variation

Environmental conditions greatly affectreproduction and/or survival of nestlingsthrough fledging and to adulthood. However,adult survival rates appear to be relatively con-stant across years, as suggested by high pairpersistence rates (White et al. 1995). Such lifehistory characteristics are common for K-selected species, for which populations remainrelatively stable even though recruitment ratesmight be highly variable. With no recruitment,the population declines at the rate of 1 minusadult survival, or the adult mortality rate.

Population TrendsPopulation TrendsPopulation TrendsPopulation TrendsPopulation Trends

We have inadequate data to estimate popula-tion trends in Mexican spotted owls. We havelittle confidence in our estimates of populationtrend that include estimates of juvenile survivalbecause these estimates of juvenile survival areprobably biased low. Further, the populationstudies from which parameter estimates werederived have not been conducted for a suffi-ciently long period to capture temporal varia-tion. Population trend was also evaluated with

occupancy data (White et al. 1995), but again issuspect. Changes in occupancy rate probablycorrespond more with how monitoring of owlswas performed rather than reflecting true changein the owl population. As a complicating factor,a nonrandom sample of all existing Mexicanspotted owl territories was monitored, thuslimiting possible inferences.

MOVEMENTSMOVEMENTSMOVEMENTSMOVEMENTSMOVEMENTS

Seasonal MovementsSeasonal MovementsSeasonal MovementsSeasonal MovementsSeasonal Movements

Seasonal movement patterns of Mexicanspotted owls are variable. Some radio-trackedowls are year-round residents within an area,some remain in the same general area but showshifts in habitat-use patterns, and some migrateconsiderable distances (20-50 km [12-31 miles])during the winter (Ganey and Balda 1989b,Ganey et al. 1992, Willey 1993, Ganey andBlock unpublished data). In general, migratingowls move to more open habitats at lowerelevations (Ganey et al. 1992, Willey 1993).Willey (1993), however, observed one owl thatmigrated to coniferous forest at a higher eleva-tion than the owls’ breeding-season range.

Natal DispersalNatal DispersalNatal DispersalNatal DispersalNatal Dispersal

Little is known about habitat use byjuveniles during natal dispersal. Seven juvenilesradio-tracked in southern Utah (Willey 1993)dispersed over distances ranging from 24 to 145km (15 to 90 miles). These owls apparentlymoved through a variety of habitats includingspruce-fir and mixed-conifer forests, pinyon-juniper woodland, mountain shrublands, desertscrublands and desert grasslands. Another fivejuvenile owls were radio-tagged in the San MateoMountains of New Mexico in 1993 (PeterStacey, Univ. of Nevada, Reno, pers. comm.).Two of these apparently moved to an adjacentmountain range before their signals were lost. Ofthe remaining three, one was relocated thefollowing year within the San Mateo Mountains.Fates of the other two juveniles were unknown.

33


Volume I/Part II

LANDSCAPE PATTERNLANDSCAPE PATTERNLANDSCAPE PATTERNLANDSCAPE PATTERNLANDSCAPE PATTERNAND METAPOPULATIONAND METAPOPULATIONAND METAPOPULATIONAND METAPOPULATIONAND METAPOPULATION

STRUCTURESTRUCTURESTRUCTURESTRUCTURESTRUCTURE

Keitt et al. (1995) examined the spatialpattern of forest habitat patches across the rangeof the Mexican spotted owl. Their objective wasto gauge the extent to which the owl mightbehave as a metapopulation in the classical senseof a set of local populations linked by infrequentdispersal. Such a finding, if verified, wouldsuggest that population dynamics of owls in onelocal population might be influenced by factors,including management activities, that affectednearby populations. Conversely, if local popula-tions are functionally discrete, then those popu-lations could be treated separately with someconfidence that actions in one part of the owl’srange would not greatly affect other populations.

Keitt et al. (1995) concluded that the owlprobably behaves as a classical metapopulationover much of its range. That is, the level ofhabitat connectivity is such that many habitatsare “nearly connected” at distances correspond-ing to their best empirical estimates of the owl’sdispersal capability. At this scale, the landscapeconsists of a set of large, more-or-less discretehabitat clusters. For example, most of theMogollon Rim functions as a single cluster, thesouthern Rockies as another single cluster, andso on. This suggests that owls could dispersewithin habitat clusters with very high probabil-ity, and disperse between clusters at very lowprobability. Thus, we would expect owls todisperse within clusters most of the time butbetween clusters only rarely which is consistentwith the definition of a metapopulation. Thisfinding suggests that the Plan should incorporaterecommendations that maintain (or increase)habitat connectivity across the owl’s range.Habitat connectivity buffers a population fromstochastic variability through time by providingthe opportunity for local population failures tobe “rescued” by immigration from other popula-tions.

Keitt et al. (1995) also attempted to identifythose habitat clusters most important to overalllandscape connectivity. They first ranked habi-tats to emphasize the importance of large patches

in the landscape, and second, they modified thisapproach to emphasize positional effects (i.e.,small clusters that are important because they actas “stepping stones” or bridges between largerhabitat clusters).

In the first analysis, the largely contiguoushabitat of the Mogollon Rim emerged as mostimportant overall, because of its large area. In theanalysis emphasizing cluster position, a few smallclusters emerged as particularly important. Theseincluded several fragments of the Cibola Na-tional Forest (Mt. Taylor and Zuni Mountains)that may serve as stepping stones between other,larger clusters. These small patches may warrantparticular management attention; they maysupport few owls but may nevertheless beimportant to overall landscape connectivity.

CONCLUSIONSCONCLUSIONSCONCLUSIONSCONCLUSIONSCONCLUSIONS

In many ways, the Mexican spotted owlappears to be quite similar to both the northernand California spotted owls with respect togeneral behavioral patterns and ecology. Forexample, all three subspecies are most commonin forests of complex structure, prey mainly onnocturnally-active small mammals, and sharesimilar vocalizations, reproductive chronologies,and population characteristics. However, impor-tant differences exist between the Mexicanspotted owl and the other subspecies. Thedistributional pattern of the Mexican spottedowl is more disjunct than that of the othersubspecies, with the possible exception of theCalifornia spotted owl population in the moun-tain ranges of southern California (Noon andMcKelvey 1992). The Mexican subspecies alsoappears to use a wider range of habitat typesthan the other subspecies. These unique aspectsof the ecology of the Mexican spotted owlrequire unique approaches to its management.For example, threats to owl habitat and manage-ment proposed to address those threats may welldiffer among the diverse habitats occupied byMexican spotted owls. In addition, because ofits’ disjunct distributional pattern, dispersalamong subpopulations of Mexican spotted owlsis an important consideration. Thus, habitatmanagement plans may need to consider notonly areas occupied by owls but also intervening

34


Volume I/Part II

areas, even where such areas are very differentin habitat structure from those typicallyoccupied by spotted owls.

We have learned a great deal about theMexican spotted owl in the last decade, butsignificant information gaps still remain. Moststudies of the owl to date have been descrip-tive rather than experimental (III.D). Al-though we have identified patterns with

respect to some aspects of the owls’ ecology (e.g.habitat use), cause and effect relationships havenot been documented. Further, many aspects ofspotted owl demography and population struc-ture remain unclear. These considerationssuggest that much additional research is needed,and that management recommendations in thenear term must deal with high levels of uncer-tainty.

35


Volume I/Part II

B. RECOVERY UNITSB. RECOVERY UNITSB. RECOVERY UNITSB. RECOVERY UNITSB. RECOVERY UNITSSarah E. Rinkevich, Joseph L. Ganey, William H. Moir,

Frank P. Howe, Fernando Clemente, and Juan F. Martinez-Montoya

The Mexican spotted owl inhabits diverseforest types scattered across an even more physi-cally diverse landscape. Further, human activitiesvary dramatically throughout the owl’s range.These variations limit our ability to approach astatus assessment on a rangewide basis. Con-sequently, we divided the range of the owl into11 geographic areas called “Recovery Units”(hereafter RUs). Six RUs were recognized withinthe United States: Colorado Plateau, SouthernRocky Mountains - Colorado, Southern RockyMountains - New Mexico, Upper Gila Moun-tains, Basin and Range - West, and Basin andRange - East (Figure. II.B.1). Five RUs wererecognized in Mexico: Sierra Madre Occidental -Norte, Sierra Madre Oriental - Norte, SierraMadre Occidental - Sur, Sierra Madre Oriental -Sur, and Eje Neovolcanico (Figure. II.B.2).

UNITED STATESUNITED STATESUNITED STATESUNITED STATESUNITED STATES

Recovery Units were identified based on thefollowing considerations (in order of impor-tance): (1) physiographic provinces, (2) bioticregimes, (3) perceived threats to owls or theirhabitat, (4) administrative boundaries, and (5)known patterns of owl distribution. It is impor-tant to note that owl distributional patterns werea minor consideration in RU delineation, andthat RUs do not necessarily represent discretepopulations of owls. In fact, movement ofindividuals between RUs has been documented(Ganey and Dick 1995).

Four major physiographic provinces wereused in delineating RUs in the United States: theColorado Plateau, Basin and Range, SouthernRocky Mountains, and Upper Gila Mountains(Hammond 1965, Wilson 1962, USGS 1970,Bailey 1980). Biotic regimes were based onclassifications by Bailey (1980) and Brown et al.(1980). Administrative boundaries were usedwhere management practices differed betweenjurisdictions (e.g., Southern Rocky MountainsRUs). The following narratives describe domi-nant physical and biotic characteristics, patterns

of owl distribution and habitat use, and thedominant patterns of land ownership and landuse within each RU.

Colorado PlateauColorado PlateauColorado PlateauColorado PlateauColorado Plateau

The Colorado Plateau RU (Figure. II.B.3)coincides with the Colorado Plateau Physi-ographic Province (USGS 1970). It includesmost of south-central and southern Utah plusportions of northern Arizona, northwestern NewMexico, and southwestern Colorado. Majorlandforms include interior basins and highplateaus dissected by deep canyons, includingthe canyons of the Colorado River and itstributaries (Williams 1986).

Grasslands and shrub-steppes dominate theColorado Plateau at lower elevations, but wood-lands and forests dominate the higher elevations(Bailey 1980, West 1983). Pinyon pine andvarious juniper species comprise the primary treetypes in the woodland zone. A montane zoneextends over areas on the high plateaus andmountains (Bailey 1980). Forest types in thiszone include ponderosa pine, mixed-conifer, andspruce-fir. Conifers may extend to lower eleva-tions in canyons. Deciduous woody speciesdominate riparian communities, and are mostcommon along major streams.

The Mexican spotted owl reaches the north-western limit of its range in this RU. Owlhabitat appears to be naturally fragmented inthis RU, with most owls found in disjunctcanyon systems or on isolated mountain ranges.In southern Utah, breeding owls primarilyinhabit deep, steep-walled canyons and hangingcanyons. These canyons are typically surroundedby terrain that does not appear to supportbreeding spotted owls. Owls also apparentlyprefer canyon terrain in southwestern Colorado,particularly in and around Mesa Verde NationalPark. In northern Arizona and New Mexico,owls have been reported in both canyon andmontane situations. Recent records of spottedowls exist for the Grand Canyon and Kaibab

36


Volume I/Part II

FFFFFigurigurigurigurigure II.B.1.e II.B.1.e II.B.1.e II.B.1.e II.B.1. Recovery Units within the United States.

37


Volume I/Part II

FFFFFigurigurigurigurigure II.B.2.e II.B.2.e II.B.2.e II.B.2.e II.B.2. Recovery Units within the Republic of Mexico.

38


Volume I/Part II

FFFFFigurigurigurigurigure II.B.3.e II.B.3.e II.B.3.e II.B.3.e II.B.3. Colorado Plateau Recovery Unit.

39


Volume I/Part II

Plateau in Arizona, as well as for the ChuskaMountains, Black Mesa, Fort Defiance Plateau,and the Rainbow/Skeleton Plateau on theNavajo Reservation. In addition, records exist forthe Zuni Mountains and Mount Taylor in NewMexico.

Federal lands account for 44% of this RU(Table II.B.1). Tribal lands collectively total30%, with the largest single entity being theNavajo Reservation. Private ownership accountsfor 19%, and State lands just 8%. Most Mexicanspotted owls have been located on NPS lands inthis RU, followed by FS and then BLM lands(Ward et al. 1995).

Recreation ranks first among land uses inNational Parks within this RU. Activities such ashiking, camping, hunting, rock climbing, andmountain biking occur in owl habitat. Many ofthese activities plus off-road vehicle recreationalso occur on BLM and FS lands throughout theColorado Plateau. Various commercial enter-prises relevant to industry and agriculture takeplace on these lands. Particularly important arelivestock grazing, timber cutting, coal anduranium mining, oil and natural gas pumping,and continued exploration for these and otherresources. Access roads, drill pads, pipelines, andloading and storage areas accompany all of theseactivities.

Southern Rocky Mountains - ColoradoSouthern Rocky Mountains - ColoradoSouthern Rocky Mountains - ColoradoSouthern Rocky Mountains - ColoradoSouthern Rocky Mountains - Colorado

This RU (Figure. II.B.4) falls partly withinthe Southern Rocky Mountains PhysiographicProvince (USGS 1970) and partly within theColorado Plateau Ecoregion (Bailey 1980). TheColorado - New Mexico state line delimits thesouthern boundary of this RU because land-usepractices and potential threats on Federal landsdiffer between these states. High mountainranges characterize the RU (Curtis 1960);dominant ranges include the San Juan Moun-tains of southwestern Colorado, the Sangre deCristo Mountains, and the Front Range.

Vegetation ranges from grasslands at lowelevations through pinyon-juniper woodlands,interior shrublands, ponderosa pine, mixed-conifer and spruce-fir forests, to alpine tundraon the highest peaks (Daubenmire 1943).

The Mexican spotted owl reaches the north-eastern limit of its range in this RU. Foundprimarily in canyons in this RU, the owls appearto occupy two disparate canyon habitat types.The first is sheer, slick-rock canyons containingwidely scattered patches (up to 1 ha in size) ofmature Douglas-fir in or near canyon bottoms orhigh on the canyon walls in short, hangingcanyons. The second consists of steep canyonscontaining exposed bedrock cliffs either close tothe canyon floor or, more typically, several tiersof exposed rock at various heights on the canyonwalls (Reynolds 1993). Mature Douglas-fir,white fir, and ponderosa pine dominate canyonbottoms and both north- and east-facing slopes.Ponderosa pine grows on the more xeric south-and west-facing slopes, with pinyon-junipergrowing on the mesa tops.

Federal lands encompass 55% of the RU,with the majority administered by the FS,followed by the BLM and NPS (Table II.B.1).Approximately 40% of the land is privatelyowned, 3% is State administrated, and <1% isTribal land. Owls have been located on FS,BLM, NPS, and Tribal lands (Ward et al. 1995).

Land-use practices throughout the RUinclude timber cutting, grazing, mining, oil andnatural gas pumping, plus all the associatedfacilities development such as access roads,pipelines, and staging and storage areas. Recre-ational activities include downhill and cross-country skiing, off-road driving, rock climbing,backpacking, camping, hiking, and mountainbiking. Road, water, and urban developmentmay also affect spotted owl habitat in this RU.

Southern Rocky MountainsSouthern Rocky MountainsSouthern Rocky MountainsSouthern Rocky MountainsSouthern Rocky Mountains- New Mexico- New Mexico- New Mexico- New Mexico- New Mexico

This RU (Figure. II.B.5) coincides with theSouthern Rocky Mountains PhysiographicProvince (USGS 1970) and the Rocky MountainForest Province (Bailey 1980). The landscapeincludes a system of high ranges separated bydeep structural basins of the northern RioGrande rift (Williams 1986). Major rangesinclude the Sangre de Cristo and Jemez Moun-tains.

40

TTTTTable II.B.1.able II.B.1.able II.B.1.able II.B.1.able II.B.1. Land ownership patterns (thousands of hectares) in Recovery Units (RU) within the United States.

LLLLLAND STAND STAND STAND STAND STAAAAATUSTUSTUSTUSTUS CPCPCPCPCP11111 SRM-COSRM-COSRM-COSRM-COSRM-CO22222 SRM-NMSRM-NMSRM-NMSRM-NMSRM-NM33333 UGMUGMUGMUGMUGM44444 BR-WBR-WBR-WBR-WBR-W55555 BR-EBR-EBR-EBR-EBR-E66666

Federal LandsFederal LandsFederal LandsFederal LandsFederal LandsFS 2,503 5,546 1,220 3,520 2,238 495BLM 7,672 2,227 520 145 3,359 1,829NPS 1,678 140 14 17 172 59

TTTTTotal Fotal Fotal Fotal Fotal Federalederalederalederalederal 11,85311,85311,85311,85311,853 7,913 7,913 7,913 7,913 7,913 1,7541,7541,7541,7541,754 3,6823,6823,6823,6823,682 5,7695,7695,7695,7695,769 2,3832,3832,3832,3832,383

State LandsState LandsState LandsState LandsState Lands

AZ 957 0 0 20 2,905 0NM 300 0 200 192 84 856UT 862 0 0 0 0 0CO 11 454 0 0 0 0

TTTTTotal Sotal Sotal Sotal Sotal Statetatetatetatetate 2,129 2,129 2,129 2,129 2,129 454 454 454 454 454 200 200 200 200 200 212 212 212 212 212 2,989 2,989 2,989 2,989 2,989 856 856 856 856 856

Tribal Lands 8,026 85 463 941 1,922 382

Private Lands 5,013 5,717 2,141 3,520 3,712 2,586

Other Lands7 16 105 24 70 1,759 1,101

TTTTTOOOOOTTTTTALALALALAL 27,03727,03727,03727,03727,037 14,27414,27414,27414,27414,274 4,5824,5824,5824,5824,582 8,4258,4258,4258,4258,425 16,15116,15116,15116,15116,151 7,3087,3087,3087,3087,308

1 Colorado Plateau RU2 Southern Rocky Mountain - Colorado RU3 Southern Rocky Mountain - New Mexico RU4 Upper Gila Mountains RU5 Basin and Range - West RU6 Basin and Range - East RU7 Other Lands include U.S. Department of Defense, Bureau of Reclamation, U.S. Fish and Wildlife Service Refuge Lands, etc.

41

FFFFFigurigurigurigurigure II.B.4.e II.B.4.e II.B.4.e II.B.4.e II.B.4. Southern Rocky Mountains - Colorado Recovery Unit.

42


Volume I/Part II


Volume I/Part II

FFFFFigurigurigurigurigure II.B.5.e II.B.5.e II.B.5.e II.B.5.e II.B.5. Southern Rocky Mountains - New Mexico Recovery Unit.

43


Volume I/Part II

Vegetation within the unit has been modi-fied by past logging, grazing, surface mining,fuelwood gathering, and fire suppression (Will-iams 1986, Van Hooser et al. 1993). Ponderosapine, mixed-conifer, and spruce-fir forests arewidespread at higher elevations. Juniper savannaand montane grasslands dominate lower eleva-tions (Brown et al. 1980). In some areas, mesatops dominated by ponderosa pine and juniperare dissected by steep canyons. Vegetation oncanyon slopes and bottoms includes a variety ofconiferous and deciduous trees.

In general, owls inhabit steep terrain andcanyons in this RU. They typically occur inmixed-conifer forests on steep slopes in theSangre de Cristo Mountains, and in the JemezMountains they occupy canyons incised intovolcanic rock. Patches of mixed-conifer forestwhich appear to contain attributes of owl habitatexist throughout northern New Mexico.

Privately owned lands comprise 47% of thetotal land within this RU (Table II.B.1). Federallands account for 38%, numerous Pueblos andTribal lands 10%, and State-administered lands4%. Mexican spotted owls have been foundprimarily on FS lands, with several records inBandelier National Monument as well (Johnsonand Johnson 1985:5).

Dominant land-use practices within this RUinclude timber cutting and livestock grazing.Products such as vigas (small- to medium-diameter trees, generally 30-35cm dbh, used fortraditional southwest ceiling beams), latillas(small-diameter trees, generally 10cm dbh apsensaplings, used for decorative southwest ceilingsor fences), and fuelwood are harvested forpersonal use. Recreational activities in northernNew Mexico include skiing, off-road driving,hiking, camping, and hunting. Other land usesinclude oil, natural gas, and mineral develop-ment, and pipeline corridors.

Upper Gila MountainsUpper Gila MountainsUpper Gila MountainsUpper Gila MountainsUpper Gila Mountains

The Upper Gila Mountains RU (Figure.II.B.6) is based on the Upper Gila MountainsForest Province (Bailey 1980). Williams (1986)refers to this area as the Datil-Mogollon Section,part of a physiographic subdivision transitional

between the Basin and Range and ColoradoPlateau Provinces. This complex area consistsof steep mountains and deep entrenched riverdrainages dissecting high plateaus. TheMogollon Rim, a prominent fault scarp, bisectsthe unit.

McLaughlin (1986) described a “Mogollon”floral element in this region. The vegetation is azonal pattern of grasslands at lower elevationsupward through pinyon-juniper woodlands,ponderosa pine, mixed-conifer, and spruce-firforests at higher elevations. Many canyonscontain stringers of deciduous riparian forests,particularly at low and middle elevations. Thisunit contains the largest contiguous ponderosapine forest in North American, an unbrokenband of forest 25 to 40 miles wide and approxi-mately 300 miles long extending from north-central Arizona to west-central New Mexico(Cooper 1960).

Mexican spotted owls are widely distributedand use a variety of habitats within the UpperGila Mountains RU. Owls are most common inmixed-conifer forests dominated by Douglas-firand/or white fir and canyons with varyingdegrees of forest cover (Ganey and Balda 1989a,Ganey and Dick 1995). Owls also occur inponderosa pine-Gambel oak forest, where theyare typically found in stands containing well-developed understories of Gambel oak.

Federal lands, mostly FS, encompass 44% ofthis RU (Table II.B.1). Tribal lands account for11%, privately owned lands 42%, and Statelands 3%. The greatest concentration of theknown Mexican spotted owl population occurswithin this RU, and most known owl locationsoccur on FS and Tribal lands (Ward et al. 1995).Many spotted owls are found within wildernessareas in this RU with the Gila Wildernesssupporting the largest known wilderness popula-tion.

The major land use within this RU is timberharvest. All of the National Forests as well as theFort Apache and San Carlos Indian Reservationshave active timber management programs.Fuelwood harvest, including both personal andcommercial harvest, occurs across much of thisunit. Livestock grazing is ubiquitous on FS landsand widespread over large portions of the FortApache and San Carlos Indian Reservations. In

44

45

FFFFFigurigurigurigurigure II.B.6.e II.B.6.e II.B.6.e II.B.6.e II.B.6. Upper Gila Mountains Recovery Unit.

addition, recreational activities such as hiking,camping, and hunting attract many people tothis RU.

Basin and Range - WestBasin and Range - WestBasin and Range - WestBasin and Range - WestBasin and Range - West

The Basin and Range Area Province (USGS1970, Bailey 1980) provided the basis for twoRUs (Figure. II.B.7). We subdivided the Basinand Range area into eastern and western unitsusing the Continental Divide as the partitionbetween these units. The division was based ondifferences in climatic and floristic characteristicsbetween these areas. The Basin and Range - Westflora is dominated by Madrean elements whilethe Basin and Range - East unit shows moreRocky Mountain affinities (Brown et al. 1980).

Geologically, the Basin and Range - WestRU exhibits horst and graben faulting (Wilson1962) with numerous fault-block mountainsseparated by valleys. Complex faulting andcanyon carving define the physical landscapewithin these mountains. These ranges include,but are not limited to, the Chiricahua,Huachuca, Pinaleno, Bradshaw, Pinal, SantaCatalina, Santa Rita, Patagonia, Santa Teresa,Atascosa, Mule, Dragoon, Peloncillo, Mazatzal,and Rincon Mountains.

Vegetation ranges from desert scrubland andsemi-desert grassland in the valleys upwards tomontane forests. Montane vegetation includesinterior chaparral, encinal woodlands, andMadrean pine-oak woodlands at low and middleelevations, with ponderosa pine, mixed-conifer,and spruce-fir forests at higher elevations (Brownet al. 1980). Isolated mountain ranges aresurrounded by Sonoran and Chihuahuan desertbasins.

Mexican spotted owls occupy a wide range ofhabitat types within this RU. The majority ofowls occur in isolated mountain ranges wherethey inhabit encinal oak woodlands, mixed-conifer and pine-oak forests, and rocky canyons(Ganey and Balda 1989a, Duncan and Taiz1992, Ganey et al. 1992).

Federal lands encompass 36% of this RU,mostly administered by the BLM followed bythe FS and a small portion by the NPS (TableII.B.1). Privately owned lands amount to 22%,State lands 19%, Tribal lands (San Carlos

Apache Reservation) 12%, and DOD lands11%. Within this RU the Mexican spottedowl occupies primarily FS lands, and the major-ity occur within the Coronado National Forest.DOD lands also support the owl on FortHuachuca Army Base in the HuachucaMountains.

Recreation dominates land use within thisunit. Activities such as hiking, birdwatching,camping, off-road driving, skiing, and huntingare particularly popular. Livestock grazing iswidespread but most intensive at low and middleelevations. Urban and rural development andmining modify portions of the Basin and Range- West landscape. Timber harvest occurs mainlyon the Prescott National Forest and the SanCarlos Apache Indian Reservation. According tothe Coronado National Forest Land Manage-ment Plan, timber cutting is used sparingly toenhance wildlife and recreational values. Militarytraining maneuvers take place in and aroundMexican spotted owl habitat on Fort HuachucaArmy Base.

Basin and Range - EastBasin and Range - EastBasin and Range - EastBasin and Range - EastBasin and Range - East

We delineated the Basin and Range - EastRU (Figure. II.B.8) based on the Basin andRange Area Province (USGS 1970) and theDesert and Steppic Ecoregions (Bailey 1980).This RU is characterized by numerous parallelmountain ranges separated by alluvial valleys andbroad, flat basins. Williams (1986) refers to theRio Grande Rift as the separation between theBasin and Range physiographic province and theColorado Plateau and Upper Gila Mountainsphysiographic provinces. The climate featuresmild winters, as indicated by the presence ofbroad-leaved evergreen plants at relatively highelevations (USDA 1991).

Regional vegetation ranges from Chihua-huan desert scrubland and Great Basin grass-lands at low elevations, through Great Basinwoodland (pinyon-juniper) at middle elevationsto petran montane coniferous forests at highelevations (Brown et al. 1980). Montane habitatincludes ponderosa pine, mixed-conifer, andspruce-fir forests and is patchily distributedthroughout the higher mountain ranges. Cot-tonwood bosques as well as other riparian

46


Volume I/Part II


Volume I/Part II

FFFFFigurigurigurigurigure II.B.7.e II.B.7.e II.B.7.e II.B.7.e II.B.7. Basin and Range - West Recovery Unit.

47


Volume I/Part II

FFFFFigurigurigurigurigure II.B.8.e II.B.8.e II.B.8.e II.B.8.e II.B.8. Basin and Range - East Recovery Unit.

48


Volume I/Part II

vegetation exist along the Rio Grande corridor.Montane and especially riparian communitieshave been altered considerably by human activi-ties.

Mexican spotted owls occur in the isolatedmountain ranges scattered across this RU. Theyare most common in mixed-conifer forest butare also found in ponderosa pine forest andpinyon-juniper woodland (Skaggs and Raitt1988). The owl has been found within mixed-conifer canyon habitat in the Guadalupe Moun-tains (McDonald et al. 1991).

Of the Basin and Range - East RU land area,private lands encompass 35%, Federal lands48%, State lands 12%, and Tribal lands 5%(Table II.B.1). The Mescalero Apache IndianReservation comprises the largest portion of theTribal lands. The majority of known Mexicanspotted owls are located on FS lands, with somefound on NPS and Tribal lands.

Dominant land uses within this RU includetimber management and livestock grazing.Recreational activities such as off-road driving,skiing, hiking, camping, and hunting are alsolocally common within the RU.


Conserving its natural resources has been asignificant challenge for Mexico. To meet thechallenge, the National System of ProtectedAreas was formed; in March of 1988, the Gen-eral Law of Ecological Balance and Environmen-tal Protection was implemented. A total of 5,992km2 (almost 600,000 ha) has been decreed asProtected Natural Areas within the RUs. Thisexpanse has been classified into nine categoriesaccording to the management objectives and thelegal uses of particular areas. The categoriesinclude: (1) Biosphere Reserves, (2) SpecialBiosphere Reserves, (3) National Parks, (4)National Monuments, (5) National MarineParks, (6) Areas of Protection of Natural Re-sources, (7) Areas of Protection of Land andAquatic Wildlife, (8) Urban Parks, and (9) AreasSubject to Ecological Conservation. Overall,there are three types of land tenancy exist inMexico: (1) Federal lands, which includedifferent institutions of the Federal Governmentsuch as Protected Natural Areas; (2) ejidal land,

which includes land allotted by the MexicanGovernment to a person or community, foragriculture, forestry, mining, or other uses; and(3) private land.

The five RUs in Mexico include SierraMadre Occidental - Norte, Sierra Madre Orien-tal - Norte, Sierra Madre Occidental -Sur, SierraMadre Oriental - Sur, and Eje Neovolcanico(Figure. II.B.2). Three major physiographicprovinces were used in the delineation: SierraMadre Occidental, Sierra Madre Oriental, andSistema Volcanico Transversal (Cuanalo et al.1989). Criteria used to delineate RUs in Mexicowere similar to that used to conform the RUs inthe United States. These criteria, listed in orderof importance, were: (1) distribution of thespotted owl, (2) local vegetation, (3) physi-ographic features, (4) administrative boundaries,and (5) potential threats to the conservation ofthe owl and its habitat.

Owl distribution is disjunct across Mexico.Williams and Skaggs (1993) report spotted owlsat 53 locations in 11 mainland Mexican States.Although vegetation types differ throughouteach RU, oak and pine-oak forest types appearedto be commonly associated with owl habitat inmost or all RUs. These oak species includedQuercus resinosa, Q. gentryi, Q. eduardii, Q.grisea, Q. chihuahuensis, Q. potosina/Q. laeta, andQ. coccolobifolia. Further, Pinus teocote was themost common pine occurring on upper mesasand occasionally on north-facing slopes in someareas where owls were found. Land uses withinall RUs include timber cutting, cattle and sheepgrazing, fuelwood gathering, and clearing for-ested areas for agriculture. Although, these landuses are practiced in different amounts through-out each RU, the majority occur within ejidos.The following narratives describe dominantphysical and biotic attributes, distribution ofowls, and land administration and ownershipof each unit.

Sierra Madre Occidental - NorteSierra Madre Occidental - NorteSierra Madre Occidental - NorteSierra Madre Occidental - NorteSierra Madre Occidental - Norte

Covering an enormous area, the SierraMadre Occidental - Norte includes parts of theStates of Chihuahua, Sinaloa, Durango, andSonora. In general, this area is characterized byisolated mountain ranges surrounded by both

49


Volume I/Part II

TTTTTable II.B.2.able II.B.2.able II.B.2.able II.B.2.able II.B.2. Land ownership patterns (thousands of hectares) in Recovery Units within Mexico.__________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________Land OwnershipLand OwnershipLand OwnershipLand OwnershipLand Ownership SMOSMOSMOSMOSMOcNcNcNcNcN11111 SMOSMOSMOSMOSMOrNrNrNrNrN22222 SMOSMOSMOSMOSMOcScScScScS33333 SMOSMOSMOSMOSMOrSrSrSrSrS44444 ENVENVENVENVENV55555

_____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________Ejidos6 4,783 28 1,220 235 441PNAs7 46 0 38 250 274Private 14,100 7,506 2,075 1,630 5,306

TTTTTotalotalotalotalotal 18,92918,92918,92918,92918,929 7,5347,5347,5347,5347,534 3,3333,3333,3333,3333,333 2,1152,1152,1152,1152,115 6,0216,0216,0216,0216,021

1 Sierra Madre Occidental - Norte RU2 Sierra Madre Oriental - Norte RU3 Sierra Madre Occidental - Sur RU4 Sierra Madre Oriental - Sur RU5 Eje Neovolcanico RU6 Ejidos are lands allocated by the Mexican Government to a person or community, to be

used for agriculture, forestry, mining, etc.7 Protected Natural Areas

narrow and wide valleys. Vegetation communi-ties consist of pine-oak forest, tropical deciduousforest, oak forest, microphyll shrub, and grass-land.

Mexican spotted owls have been reported inthe northern and western portions of this RU. Arecent study in Sonora found 12 sites in isolatedmountain ranges (Cirett-Galan and Diaz 1993).The owls occupied canyons and slopes withvarious exposures, and most were found in pine-oak forest. In portions of Chihuahua, 25 owlswere located at 13 different localities in severalmountain ranges (Tarango et al. 1994). Mostowls were found in small, isolated patches ofpine-oak forest in canyons.

Records for the State of Sinaloa are limited.There are at least two records from the highRancho Liebre Barranca, near the Sinaloa-Durango State line (Williams and Skaggs 1993).These sites were described as deep canyonscontaining pine-oak and subtropical vegetation(Alden 1969).

Private lands comprise 74%, ejidos 25%, andFederal lands 1% of the total land within thisRU (Table II.B.2). Chihuahua has two NationalParks: Cascadas de Bassaseachic, and Cumbresde Majalca. This RU also includes La MichiliaBiosphere Reserve, located in Durango. Bio-sphere Reserves are protected areas withrelatively unaltered landscapes and containendemic, threatened, or endangered species.

Sierra Madre Oriental - NorteSierra Madre Oriental - NorteSierra Madre Oriental - NorteSierra Madre Oriental - NorteSierra Madre Oriental - Norte

The Sierra Madre Oriental - Norte includesthe central portion of the State of Coahuila. Thisarea is characterized by broad mountain rangessurrounded by valleys. Vegetation consists ofgrasslands, mesquite woodland, dwarf oakgroves, submontane shrubland, desert shrubland,crasicaule shrub, and pine-oak and oak forests.

Two owl records are reported for this RU. Atone of these sites an owl was observed roostingin a canyon bottom under a dense canopy ofmaples and oaks. Vegetation in the other canyonwas described as “garden-like,” containing pines,oaks, and madrones (Williams and Skaggs1993).

Lands in this RU are almost entirely pri-vately owned (Table II.B.2). Private lands en-compass over 99% and ejidos comprise < 1% ofthe total land area. No Protected Natural Areasof any category exist in this RU.

Sierra Madre Occidental - SurSierra Madre Occidental - SurSierra Madre Occidental - SurSierra Madre Occidental - SurSierra Madre Occidental - Sur

The Sierra Madre Occidental - Sur RUincludes parts of the States of Durango,Zacatecas, San Luis Potosi, Aguascalientes,Jalisco, Nayarit, Queretaro, and Guanajuato. Ingeneral, this area is characterized by isolated

50


Volume I/Part II

This RU is comprised of 77% private prop-erty, 12% Federal lands, and 11% ejidos (TableII.B.2). The Federal lands include one NationalPark, Cumbres de Monterrey in Nuevo Leon.Natural Monument Cerro de la Silla, in NuevoLeon, is also within this RU. Natural Monu-ments possess one or more elements of nationalsignificance. These elements may be sites ornatural objects that have been placed underabsolute protection because of their unique andexceptional makeup, aesthetic interest, and/orhistorical or scientific value.

Eje NeovolvanicoEje NeovolvanicoEje NeovolvanicoEje NeovolvanicoEje Neovolvanico

The Eje Neovolcanico RU covers portionsof many States including Jalisco, Michoacan,Guanajuato, Queretaro, Hidalgo, Mexico,Guerrero, Puebla, Morelaos, Tlaxcala Veracruz,and Oaxaca. This RU is characterized by volca-nic cones severely dissected by ravines. Thearea also includes rounded hills, slopes, andplateaus. Vegetation communities include pine-oak forest, grassland, low tropical deciduousforest, crasicaule shrub, oak forest, juniper forest,pine forest, mesquite woodlands, and desertshrublands.

Mexican spotted owls have been reported inJalisco on the volcano of Cerro Nevado deColima (Voacan de Nieve). Vegetation in thisarea consists of pine-oak forest. One Mexicanspotted owl was collected near the city ofUruapan in the State of Michoacan at Cerro deTancitaro. However, this area is now urbanizedand no longer contains owl habitat. Althoughother states in this RU appear to contain suitableowl habitat, Jalisco is the only State known tohave recent records of spotted owls.

This unit is comprised of 88% private lands,5% Federal lands, and 7% ejidos (Table II.B.2).This RU includes 19 National Parks, 1 SpecialBiosphere Reserve, and 1 Area of Protection ofLand and Aquatic Wildlife.

mountains, valleys, and severely dissected can-yons and gorges. Vegetation includes mesquitewoodland, submontane shrub, grasslands, pine-oak forest, crasicaule shrub, low tropical decidu-ous forest, and desert shrubland.

Records exist for owls in La Michilia Bio-sphere Reserve. In addition, Mexican spottedowls have recently been found in Aguascalientesnear the border of Zacatecas, in the Sierra Fria(Williams and Skaggs 1993). Owl records alsoexist within Guanajuato State.

Private lands comprise 62%, ejidos 37%, andFederal lands 1% of this RU (Table II.B.2).Federal lands include two National Parks: ElClimatario in Queretaro State, and Gogorron inthe State of San Luis Potosi. In addition, this RUincludes Mariposa Monarca Sanctuary, a SpecialBiosphere Reserve. The Special BiosphereReserves have one or more ecosystems, arerelatively unaltered by anthropogenic activities,and contain endemic, threatened, or endangeredspecies.

Sierra Madre Oriental - SurSierra Madre Oriental - SurSierra Madre Oriental - SurSierra Madre Oriental - SurSierra Madre Oriental - Sur

The Sierra Madre Oriental - Sur includesparts of the States of Coahuila, Nuevo Leon, andTamaulipas. This RU is characterized by longridges with sharp pinnacles, narrow valleys, and afew plateaus. Vegetation consists of pine forest,submontane shrublands, dwarf oak, and desertrosetofilo shrublands.

Mexican spotted owls have been found inthe southern portions of Coahuila (Williams andSkaggs 1993) and in Tamaulipas (Ward et al.1995). The owls were found in oak, pine,juniper, and mixed-conifer forests. They werereported to use cliff sites for nesting and roost-ing.

Five locations have been reported in NuevoLeon. These sites were described as pine-oak andmixed-conifer forests with large cliffs havingnortheast exposures.

51


Volume I/Part II

C. DEFINITIONS OF FOREST COVER TYPESC. DEFINITIONS OF FOREST COVER TYPESC. DEFINITIONS OF FOREST COVER TYPESC. DEFINITIONS OF FOREST COVER TYPESC. DEFINITIONS OF FOREST COVER TYPESJames L. Dick, Jr., Joseph L. Ganey, and William H. Moir

This Recovery Plan proposes specificguidelines for several forest cover types, includ-ing mixed-conifer, pine-oak, and riparian forests(III.B). This is based on: (1) considerable evi-dence that these cover types are of specificimportance to the Mexican spotted owl in termsof providing habitat for nesting, roosting, andforaging activities (Ganey and Dick 1995); and(2) the Team’s desire to target guidelines for themost appropriate habitats and avoid imposingrestrictions where specific guidelines to protectspotted owl habitat are unwarranted.

Numerous treatments deal with the conceptsof classifying vegetation to cover or habitat types(e.g., Daubenmire 1952, 1968, Pfister 1989).These concepts will not be reviewed in anydepth here. In general, we accept the view thatthe basic unit of classification of climax vegeta-tion is the plant association (Küchler 1964,Daubenmire 1968, Pfister 1989). These associa-tions are defined using information on presentspecies composition and successional pathways.The problem with applying guidelines to plantassociations is that many forests in the southwestmay and should not be in or even near a climaxcondition because of the frequency and intensityof disturbance events in these forests. For ex-ample, in an analysis of Mexican spotted owlhabitat on the Alpine Ranger District, Apache-Sitgreaves National Forest, the Team determinedthat habitat classifications based on current andclimax vegetation gave very different results.Based on current vegetation, important roostingand nesting habitat typed out as mixed-coniferforest, whereas a classification based on potentialnatural vegetation (PNV) typed many of theseareas as spruce-fir forest. This points out theneed for clear, operational definitions of covertypes to be used when applying guidelines underthis Plan.

In this section, we first review some of therelevant literature on forest cover types in thesouthwest, and then provide operational defini-tions and a simple key that should allow landmanagers to classify lands in a manner compat-ible with the recommendations provided in this

Plan. Our intent is not to provide a comprehen-sive classification scheme here or to supplantextant classification schemes. Rather, our intentis to provide guidance to land managers chargedwith applying Recovery Plan guidelines and tofacilitate uniform application of guidelines acrossadministrative boundaries.

LITERATURE REVIEWLITERATURE REVIEWLITERATURE REVIEWLITERATURE REVIEWLITERATURE REVIEW

Extensive literature exists on both vegetationclassification in general and on classificationsystems for southwestern forests. Our intent isnot to review that literature exhaustively, but topresent an overview of some classificationsystems currently in use. This information willprovide the background for discussion of foresttype definitions relevant to this Plan.

Soil temperature (STR) and moisture (SMR)regimes provide one possible approach to classi-fying forest types. STR and SMR may be used toconceptualize three major groups of cover typesin the southwestern United States. Ponderosapine forests typically occur where the STR isfrigid or (in southern Arizona and southwesternNew Mexico) mesic and where the SMR is ustic.Spruce-fir forests everywhere occur on soils ofcryic SMR, whereas mixed-conifer forests areuniquely udic and frigid in their SMR and STR,respectively. The implication is that these soilparameters partition the soil environment intothree mutually exclusive but all encompassingclasses (USDA 1991). These classes are generallyconsistent with the three major forest typegroups mentioned.

Most vegetation-classification schemes,however, are based on either exisiting vegetationor on a combination of existing vegetation andknowledge of successional potential (Layser andSchubert 1979). Eyre (1980) discussed thepractice of defining forest cover types on thebasis of “present occupancy of an area by treespecies.” He further described the practice ofnaming forest types after the dominant treespecies. Dominance was determined by relativeproportions of basal area, and the type name was

52


Volume I/Part II

usually confined to one or two species. An addedrequirement was that a species must contributeat least 20% of the total basal area to be used inthe type name.

Numerous authors have expanded on thisapproach, and developed and refined classifica-tion systems for southwestern forests based notonly on existing vegetation but also on estimatedsite potential. These treatments are discussedbelow.

Ponderosa PinePonderosa PinePonderosa PinePonderosa PinePonderosa Pine

The ponderosa pine forest type occurs inwhat Moir (1993) described as the Lower Mon-tane Coniferous Forest. Forests in this zone aredominated by pines, sometimes co-occurringwith junipers and oaks. The climate is sometimesborderline for forests, with moisture becominglimiting in the upper portions of the soil profileduring part of the long growing season. Moir(1993) included the following series in thisgeneral forest type: Ponderosa pine - Gambeloak, Ponderosa pine - silverleaf oak, Ponderosapine - pinyon pine - Gambel oak, Ponderosapine - pinyon pine - gray oak, and Chihuahuapine.

Layser and Schubert (1979) described thePonderosa Pine Series as being generally domi-nated by the Rocky Mountain variety of ponde-rosa pine (var. scopulorum), except in southeast-ern Arizona, where Pinus ponderosa var. arizonicadominates. This series occurs in areas that aregenerally too warm or too dry for Douglas-firand/or true firs. Gambel oak is often a long-livedseral species. The Ponderosa Pine Series in thesouthwest is generally more complex than thatdescribed for the northern Rocky Mountains,because of the additional associated tree speciesand the presence of two varieties of ponderosapine (Layser and Schubert 1979). Hanks et al.(1983), Alexander et al. (1984a), Alexander andRonco (1987), DeVelice et al. (1986), andFitzhugh et al. (1987) provide further discussionof the Ponderosa Pine Series and associatedhabitat types and phases in the southwesternUnited States.

Mixed-ConiferMixed-ConiferMixed-ConiferMixed-ConiferMixed-Conifer

Mixed-conifer forests in the southwesternUnited States generally approximate to theUpper Montane Coniferous Forest discussed byMoir (1993). Mixed-conifer forests are mostcommon between approximately 2,440 and3,050 m (8,000-10,000 feet) in elevation, butmay occur higher or lower depending on topog-raphy and aspect. In particular, mixed-coniferforest may extend to lower elevations in canyonsystems and cold-air drainages.

Southwestern mixed-conifer forests areamong the most complex forest types known,exhibiting great variation in tree composition(USDA 1983). Overstory species in these forestsinclude Rocky Mountain Douglas-fir, white fir,Rocky Mountain ponderosa pine, quakingaspen, southwestern white pine, limber pine,and blue spruce. Forests in any successional stagemay be mixed-conifer if tree regeneration indi-cates any of the above tree species will assumedominance in time. Some stands may consist ofonly two species, whereas others may contain asmany as eight associates (USDA 1983). Gambeloak and/or silverleaf oak may share overstory orunderstory dominance with the conifers inmixed-conifer forests. Again, one of the keyattributes of southwestern mixed-conifer is itsinherent variability and diversity.

At the warm/dry end of the environmentalcontinuum, mixed-conifer forest typicallyintergrades with ponderosa pine forest. WhereDouglas-fir, white fir, or blue spruce, eithersingly or in combination, constitute less than5% cover or are considered “accidental” in latesuccessional stands, these stands are not includedin the mixed-conifer forest classification.

At the cold/wet end of the environmentalcontinuum, mixed-conifer forest typicallyintergrades with subalpine spruce-fir forest.Where corkbark fir (Abies lasiocarpa var.arizonica), subalpine fir (A. lasiocarpa var.lasiocarpa), or Englemann spruce, either singlyor in common, constitute more than 5% of thecover or are not considered “accidental,” theforest is subalpine and no longer consideredmixed-conifer.

In addition to this general description,numerous authors have discussed aspects of

53


Volume I/Part II

southwestern mixed-conifer forests or classifica-tion of southwestern forests. Pearson (1931)described a “Douglas-fir Zone.” He stated thatalthough Douglas-fir is generally regarded as thecharacteristic tree of this type, it rarely occurs inpure stands. Instead, it commonly occurs instands with white fir, limber or Mexican whitepine, and blue spruce. Western yellow pine (e.g.,ponderosa pine) is common in the lower portionof the type, and Engelmann spruce is commonin the upper portion. Quaking aspen is commonthroughout the type.

Choate (1966) also described Douglas-firforests in New Mexico as seldom growing inpure stands. He also stated that it mixes withponderosa pine at lower elevations and with truefirs and spruce at the upper limits. White fir andquaking aspen are common associates through-out the Douglas-fir type.

USDA (1992) described old-growth at-tributes by cover types, including a “mixed-species group” Forest Cover Type, which in-cluded the Douglas-fir, white fir, blue spruce,and limber pine forest cover types. They de-scribed these mixed-species stands as having arich diversity of vegetation, typically including atleast three tree species.

Moir (1993) described mixed-conifer forestsas upper montane coniferous forests featuringDouglas-fir, white fir, several tall pine species,blue spruce, and quaking aspen. He included thefollowing series in this general forest type: BlueSpruce, White fir - Douglas-fir, Douglas-fir -Southwestern White Pine, White Fir - Douglas-fir - Ponderosa Pine, Douglas-fir - Limber Pine -Bristlecone Pine, Douglas-fir - Gambel Oak, andDouglas-fir - Silverleaf Oak. These forests arevery productive because of ample precipitationand soils that are well watered throughout thelong growing season.

Fletcher and Hollis (1994) described mixed-conifer forest cover types as those dominated byDouglas-fir and/or white fir, usually containingvarying amounts of ponderosa pine, southwest-ern white pine, and/or limber pine. Hardwoodspecies, including rocky mountain maple (Acerglabrum), boxelder (A. negundo), bigtooth maple(A. grandidentatum), Gambel oak, quakingaspen, and other hardwood species may also bepresent. Douglas-fir and/or white fir typically

comprise at least 40 percent and hardwoodspecies less than 40 percent of the stand basalarea. Conifers typical of higher elevations, suchas Engelmann spruce, blue spruce, and/or sub-alpine fir may occur as “accidentals,” or provideless than about 5% cover in late successionalstands.

Layser and Schubert (1979), Moir andLudwig (1979), Alexander et al. (1984a, b),Alexander and Ronco (1987), Youngblood andMauk (1985), DeVelice et al. (1986), andFitzhugh et al. (1987) all discussed classificationof forest types in general or mixed-conifer foresttypes in particular in the southwestern UnitedStates. These treatments vary somewhat, possiblybecause of regional differences in forest types. Ageneral consensus, however, indicates thatmixed-conifer forest types generally fall in thefollowing four series: Abies concolor, Psuedotsugamenziesii, Pinus flexilis, or Picea pungens.

Spruce-FirSpruce-FirSpruce-FirSpruce-FirSpruce-Fir

Spruce-fir forests in the southwestern UnitedStates generally coincide with the SubalpineConiferous Forest discussed by Moir (1993).These are high-elevation forests occurring oncold sites. They have short growing seasons,heavy snow accumulations, and strong ecologicaland floristic affinities to cold forests of higherlatitudes. Dominant trees include Engelmannspruce, subalpine and/or corkbark fir, or some-times bristlecone pine. Moir and Ludwig (1979)included the Picea engelmannii and Abieslasiocarpa Series in the general spruce-fir foresttype. Moir (1993) included the following seriesin this general forest group: Bristlecone pine,Engelmann spruce - bristlecone pine, corkbarkfir - Engelmann spruce, Corkbark fir - Engel-mann spruce - white fir, Engelmann spruce -Douglas-fir, and Engelmann spruce - limberpine.

Other Forest Types of InterestOther Forest Types of InterestOther Forest Types of InterestOther Forest Types of InterestOther Forest Types of Interest

Chihuahua PineChihuahua PineChihuahua PineChihuahua PineChihuahua Pine

The Pinus leiophylla Series is described byLayser and Schubert (1979). This series typically

54


Volume I/Part II

contains a diverse mixture of conifers andevergreen oaks. The conifer component isextensive enough to characterize this series asforest rather than woodland. Dominant conifersare typically Chihuahua pine, Apache pine, andP. ponderosa var. arizonica (Layser and Schubert1979). This series is Madrean in affinity and,within the U.S., is restricted to central andsouthern Arizona and southwestern New Mexico(Brown et al. 1980). Moir (1993) included thistype in the Lower Montane Coniferous Forestgroup.

Quaking AspenQuaking AspenQuaking AspenQuaking AspenQuaking Aspen

Quaking aspen is a special feature of westernlandscapes. It is a major seral species in thefollowing series; Abies lasiocarpa, Picea pungens,and Abies concolor. It is a minor seral species inthe Picea engelmannii, Pseudotsuga menziesii, andPinus ponderosa Series (Larson and Moir 1986).As such, quaking aspen should be a commoncomponent of these landscapes under naturaldisturbance regimes.

Riparian ForestsRiparian ForestsRiparian ForestsRiparian ForestsRiparian Forests

Numerous authors have discussed classifica-tion and ecology of riparian forests in the south-western United States (e.g., Pase and Layser1977, Layser and Schubert 1979, Brown et al.1980, Medina 1986, Szaro 1989). In general,southwestern riparian forests are dominated byvarious species of broadleaved deciduous treesand shrubs. Trees common in adjacent uplands,such as conifers, oaks, and quaking aspen, mayoccur in association with riparian trees, butgenerally do not dominate the site (Brown1982).

PLAN DEFINITIONSPLAN DEFINITIONSPLAN DEFINITIONSPLAN DEFINITIONSPLAN DEFINITIONS

Our classification scheme is primarilyconcerned with a subset of the available foresttypes in the southwestern United States. We areinterested in both potential and existing vegeta-tion. Consequently, our scheme is a hybrid ofclassification schemes based on potential vegeta-tion (series, association and habitat type) andforest cover types based on existing vegetation.

Three terms used in our definitions requireclarification here. These are “pure,” “majority,”and “plurality.” Various definitions exist todescribe what constitutes a pure stand. Daniel etal. (1979) described pure stands as those where>90% of the dominant or codominant trees areof a single species. A stand may have an under-story of other species without changing the puredesignation. The key to this concept is thedistinction between the dominant and codomi-nant species and the understory component. Incontrast, Eyre (1980) defined a pure stand as onewhere >80% of the stocking is by one species.

For purposes of this plan, we use the termpure to refer to any stand where a single speciescontributes >80% of the basal area of dominantand codominant trees. We use the term majorityto refer to the situation where a single speciescontributes >50% of the basal area (Eyre 1980).We use the term plurality to refer to the situa-tion where a species (or group of species ofinterest) comprises the largest proportion of amixed-species stand (Eyre 1980).

With these definitions and concepts inmind, definitions for specific forest cover typesare provided below.

Ponderosa Pine ForestPonderosa Pine ForestPonderosa Pine ForestPonderosa Pine ForestPonderosa Pine Forest

We define the Ponderosa Pine Forest Typeas:

1. Any forested stand of the Pinus ponderosaSeries not included in the Pine-OakForest Type (see below), or;

2. Any stand that qualifies as pure (Eyre1980) ponderosa pine, regardless of theseries or habitat type.

Pine-oak ForestPine-oak ForestPine-oak ForestPine-oak ForestPine-oak Forest

A number of habitat types exist in thesouthwestern United States that could be de-scribed as pine-oak. Most of the stands relevantto recovery of the Mexican spotted owl fallwithin two series, the Pinus ponderosa Series andthe Pinus leiophylla Series. Present evidence,however, suggests that the former series includes

55


Volume I/Part II

many areas that could never attain the type offorest structure sought by spotted owls forroosting and nesting. Therefore, in an attempt toavoid needlessly restricting management optionson lands not used to any great extent by thespotted owl, we propose the following opera-tional definition for pine-oak forest under thisPlan:

1. Any stand within the Pinus leiophyllaSeries.

2. Any stand within the Pinus ponderosaSeries that meets the following criteriasimultaneously:

a) Habitat types that reflect Quercusgambelii or a Quercus gambelii phaseof the habitat type.

b) The stand is located in either theUpper Gila Mountains RecoveryUnit, the Basin and Range-WestRecovery Unit, or the Zuni Moun-tains or Mount Taylor regions of theColorado Plateau Recovery Unit.

c) >10% of the stand basal area or2.3 m2/ha (10 ft2/ac) of basal areaconsists of Gambel oak > 13 cm(5 in) diameter at root collar.

3. Any stand within the Basin and Range-West Recovery Unit of any other seriesthat meets the following criteriasimultaneously:

a) A plurality (Eyre 1980) of the basalarea exists in yellow pines (ponde-rosa, Arizona, Apache, or Chihua-hua).

b) >10% of the stand basal area or2.3 m2/ha (10 ft2/ac) of basal areaconsists of any oaks > 13 cm (5 in)diameter at root collar.

Mixed-conifer ForestMixed-conifer ForestMixed-conifer ForestMixed-conifer ForestMixed-conifer Forest

Natural variability is high within this foresttype and has been increased by both natural andhuman-caused disturbances. Despite this vari-ability, an extant classification scheme based onseries and habitat types (Hanks et al. 1983,Layser and Schubert 1979, Alexander et al. 1984a, b; Alexander and Ronco 1987, Youngbloodand Mauk 1985, DeVelice et al. 1986, Fitzhughet al. 1987) is available. This classification systemis in widespread use and has multiple-agencysupport. Given that background, we proposeusing that system as a starting point in definingmixed-conifer forest, with some added refine-ments. Specifically, we propose that:

1. The definition of mixed-conifer forestgenerally be confined to the followingseries (Layser and Schubert 1979) andassociated habitat types (after authorslisted above): Abies concolor, Pseudotsugamenziesii, Pinus flexilis, or Picea pungens.

Within this framework, we provide thefollowing exceptions to the general guidelinelisted above:

1. Any stand within the Pinus aristata,Picea engelmannii, or Abies lasiocarpa Series not having a plurality (Eyre 1980)of basal area of any of Pinus aristata,Picea engelmannii, Abies lasiocarpa, orPinus ponderosa, singly or in combina-tion, should also be defined as mixed-conifer.

2. Stands that can be described as “pure” forconiferous species other than Douglas-fir, white fir, southwestern white pine,limber pine, or blue spruce should beexcluded from the broad category ofmixed-conifer for the purposes of Planimplementation regardless of the seriesor habitat type. By pure, we mean thatone species comprises 80% or more ofthe dominant and codominant trees(Eyre 1980).

56


Volume I/Part II

3. Stands of mixed species with >50% ofthe basal area consisting of quakingaspen should be defined as quakingaspen for the purposes of Plan imple-mentation regardless of the series orhabitat type.

High-elevation Forests,High-elevation Forests,High-elevation Forests,High-elevation Forests,High-elevation Forests,Including Spruce-fir ForestIncluding Spruce-fir ForestIncluding Spruce-fir ForestIncluding Spruce-fir ForestIncluding Spruce-fir Forest

We define this forest type as:

1. Any stand of the Pinus aristata, Piceaengelmannii, or Abies lasiocarpa Seriesthat meets the following criteria:

a) The majority (Eyre 1980) of standbasal area consists of any of the threespecies listed above, either singly orin combination, or;

b) Any stand that qualifies as a purestand (Eyre 1980) of any of thesespecies, regardless of the series orhabitat type.

Quaking AspenQuaking AspenQuaking AspenQuaking AspenQuaking Aspen

We propose that any stands with >50% ofthe basal area consisting of quaking aspen bedefined as quaking aspen.

KEY TO FOREST COVER TYPESKEY TO FOREST COVER TYPESKEY TO FOREST COVER TYPESKEY TO FOREST COVER TYPESKEY TO FOREST COVER TYPES

1. Trees deciduous and broadleaved, oftenconfined to floodplain, drainageway, orcanyon bottom (Layser and Schubert1979) . . . . . . . . . . . . . . RiparRiparRiparRiparRiparian Fian Fian Fian Fian Forororororestestestestest

1. Dominant trees evergreen and needle-leaved . . . . . . . . . . . . . . . . . . . . . . . . . ..2

2. Series = Psuedotsuga menziesii, Abiesconcolor, Pinus flexilis, or Piceapungens . . . . . . . . . . . . . . . . . . . . . .3

2. Series not as above . . . . . . . . . . . . .5

3. >80% of dominant and codominanttrees are species other than Pseudotsugamenziesii, Abies concolor, Pinusstrobiformis, Pinus flexilis, or Piceapungens . . . . . . . . Classify bClassify bClassify bClassify bClassify by dominanty dominanty dominanty dominanty dominant

speciesspeciesspeciesspeciesspecies

3. Stand not as above . . . . . . . . . . . . . . . .4

4. Populus tremuloides contributes>50% of stand basal area. . . . . . . . . . QQQQQuaking Aspen Fuaking Aspen Fuaking Aspen Fuaking Aspen Fuaking Aspen Forororororestestestestest

4. Not as above. . . . . . . MMMMMixixixixixed-conifered-conifered-conifered-conifered-coniferFFFFForororororestestestestest

5. Series = Pinus leiophylla . . . . . PPPPPine-oakine-oakine-oakine-oakine-oakFFFFForororororestestestestest

5. Series not as above . . . . . . . . . . . . . . . . 6

6. Series = Pinus ponderosa . . . . . . . . . 7

6. Series not as above . . . . . . . . . . . . 10

7. Habitat type or phase includes Quercusgambelii . . . . . . . . . . . . . . . . . . . . . . . . 8

7. Not as above. . . . . . . . . . . . . .PPPPPonderonderonderonderonderosaosaosaosaosaPPPPPine Fine Fine Fine Fine Forororororestestestestest

8. Area is located within Upper GilaMountains Recovery Unit, Basin andRange-West Recovery Unit, or thesoutheastern portion of the ColoradoPlateau Recovery Unit (Zuni Mtns.,Mt. Taylor). . . . . . . . . . . . . . . . . . . 9

8. Area not located as above. . . . . . . . . .. . . .. . . . . . . PPPPPonderonderonderonderonderosa Posa Posa Posa Posa Pine Fine Fine Fine Fine Forororororestestestestest

9. >10% of stand basal area or 2.3 m2/ha(10 ft2/ac) consists of Quercus gambelii >13 cm (5 in) diameter at root collar. . . .

PPPPPine-oak Fine-oak Fine-oak Fine-oak Fine-oak Forororororestestestestest

9. Not as above . . . . . . . . . . . . . PPPPPonderonderonderonderonderosaosaosaosaosaPPPPPine Fine Fine Fine Fine Forororororestestestestest

57


Volume I/Part II

10. Series = Pinus aristata, Piceaengelmannii, or Abies lasiocarpa. . . . . . . . . . . . . . . . . . . . . . . . . . 11

10. Series not as above . . . . . . . . . . . 13

11. Stand can be defined as pure for eitherPinus aristata, Picea engelmannii, or Abieslasiocarpa . . . . . . . . . . S S S S Sprprprprpruce-fir Fuce-fir Fuce-fir Fuce-fir Fuce-fir Forororororestestestestest

11. Stand not as above . . . . . . . . . . . . . . . 12

12. Pinus aristata, Picea engelmannii,or Abies lasiocarpa contribute >50%of stand basal area, either singly orin combination. . . . . . . S S S S Sprprprprpruce-firuce-firuce-firuce-firuce-fir

FFFFForororororestestestestest

12. Stand not as above . . . . . . . . . . . . .. . . . . . . . . . .MMMMMixixixixixed-conifer Fed-conifer Fed-conifer Fed-conifer Fed-conifer Forororororestestestestest

13. Stand located in Basin and Range-WestRecovery Unit . . . . . . . . . . . . . . . . . . 14

13. Stand not located as above . . . . . . OOOOOthertherthertherther

14. A plurality of stand basal area iscontributed by Pinus ponderosa,Pinus engelmannii, or Pinusleiophylla, either singly or incombination . . . . . . . . . . . . . . . . .15

14. Stand not as above . . . . . . . . . OOOOOthertherthertherther

15. >10% of stand basal area or 2.3 m2/ha(10ft2 ac) consists of any oak > 13 cm (5in) diameter at root collar. . . . . . . . . . . .. .. . . . . . . . . . . . . . . . . P P P P Pine-oak Fine-oak Fine-oak Fine-oak Fine-oak Forororororestestestestest

15. Stand not as above . . . . . . . . . . . . OOOOOthertherthertherther

58


Volume I/Part II

The goal of this Recovery Plan is to recoverthe Mexican spotted owl so that it no longerrequires protection under the EndangeredSpecies Act. We submit that this can be bestachieved by ensuring a mosaic of all successionalstages, now and in the future, throughout alandscape comprised of all known habitat typesused by the owl. In mixed-conifer, pine-oak andriparian forests, this habitat mosaic must containstands adequate for all life-history requirements,including nesting and roosting.

We agree with the widely held belief thatconditions within some southwestern forestsdeviate substantially from those existing prior toEuropean settlement. Moreover, forests through-out the U.S. range of the owl are at high riskfrom fire, insects, and disease. The mechanismsresponsible for current condition are not com-pletely known, but synergistic effects of pasttimber harvest, overgrazing, and fire suppressionare plausible explanations. The intent of thisRecovery Plan is not to cast blame on anyparticular aspect of past management, but tooutline the appropriate steps needed to ensurepersistence of the Mexican spotted owl. Thus,the basis to maintain owl populations is toensure that adequate habitat quality and quantitywill be sustained through time. These conditionsalso must be within the natural range of varia-tion.

We recognize, however, that major knowl-edge gaps preclude accurate descriptions of thenatural range of variation and presettlementconditions. We cannot verify that fewer owlsexist today than 100 years ago, or vice versa. Weknow little about habitat quality and howcontemporary landscapes and ecosystem condi-tions contribute to owl fitness and populationpersistence. Thus, management of the owlshould proceed in an iterative fashion. We mustuse the best available knowledge to guide currentmanagement, recognizing that new informationfrom research and monitoring is critical for thedevelopment of long-term management plans.

This Recovery Plan details a short-term (10-15 years) strategy aimed at maintaining owlhabitat where it exists and initiating a process todevelop a forested landscape that includesreplacement habitat. We presently have insuffi-cient knowledge to design a strategy that willanswer all long-term considerations, such asallocation of stand structures in space and time.Under proposed delisting criteria (III.A), the owlcould be delisted within this 10-15 years, render-ing this plan obsolete.

To achieve the recovery goals outlined in thisPlan, management must emulate natural ecosys-tem processes and landscape mosaics that bal-ance natural variability and secure the landscapeagainst catastrophic habitat loss. Our recommen-dations assume that population status andhabitat condition will be monitored in conjunc-tion with recovery efforts for the Mexicanspotted owl (Part III). The management recom-mendations are not meant to stand alone with-out such monitoring.

ECOSYSTEM OR LANDSCAPEECOSYSTEM OR LANDSCAPEECOSYSTEM OR LANDSCAPEECOSYSTEM OR LANDSCAPEECOSYSTEM OR LANDSCAPEMANAGEMENTMANAGEMENTMANAGEMENTMANAGEMENTMANAGEMENT

Volume 2 summarizes current knowledge ofthe Mexican spotted owl’s basic natural and lifehistories, but a brief reiteration is appropriatehere. First, the owl is found in a number ofdifferent habitat types ranging from slickrockcanyons to cool, mesic forests. Second, the owlhas relatively large home ranges, typically con-taining mosaics of vegetation types and differentseral stages and conditions within those types.Third, the owl takes numerous species of preyand each of these species has unique habitatrequirements. These factors considered simulta-neously stress the need to consider managementacross spatial scales ranging from sites to land-scapes and to provide the diversity of conditionsrequired for the owl’s life history. Consequently,we submit that management for Mexican spot-ted owls must be viewed within the context ofmanaging ecosystems.

D. CONCEPTUAL FRAMEWORK FOR RECOVERYD. CONCEPTUAL FRAMEWORK FOR RECOVERYD. CONCEPTUAL FRAMEWORK FOR RECOVERYD. CONCEPTUAL FRAMEWORK FOR RECOVERYD. CONCEPTUAL FRAMEWORK FOR RECOVERYWilliam H. Moir, James L. Dick Jr., William M.

Block, James P. Ward Jr., Robert Vahle,Frank P. Howe, and Joseph L. Ganey

59


Volume I/Part II

How can ecosystem management be appliedto the Mexican spotted owl? Ecosystem manage-ment should sustain biotic diversity and thenatural processes and landscape mosaics thatgenerate that diversity (cf., Jensen andBourgeron 1993, Franklin 1993, 1994; Diazand Apostol 1993, Kaufmann et al. 1994,Williams 1994). The current emphasis inecosystem management is to use the filterapproach described by Hunter (1991). Two“sizes” of filters, coarse and fine, are used.

The objective of the coarse filter approach isto maintain the natural array of conditions thatexist within the biotic and physical limits of thelandscape. This would include special as well ascommon habitats. Ideally, the array of condi-tions provided by using the coarse-filter ap-proach should maintain most plants and animalsadapted to natural conditions (Hunter 1991).This should include most of the habitat condi-tions needed by the owl and its prey.

In some cases, however, a fine filter may berequired for specific habitats, or habitat ele-ments, that fall through the coarse filter. Withrespect to the Mexican spotted owl, the coarsefilter is probably sufficient for most foraginghabitats, but a fine filter may be needed toprovide nest and roost sites. For example, theowl prefers or needs particular landforms (suchas steep-walled canyons), particular structures(including snags, large trees, large logs, cavitiesand other platforms for nesting), mature forests,and specialized microhabitats. Thus, a fine-filteranalysis is required to identify and ensure con-tinuing availability of the owl’s specific habitatneeds.

In summary, the coarse filter approach isused to manage the overall landscape, and, ifproperly applied, should suffice to maintain thenatural array of conditions on that landscape.The fine filter is used to provide specializedhabitats or habitat elements within that overalllandscape.

Two themes of the recovery measures areconsistent with these principles of ecosystemmanagement. The first theme is that the generalrecommendations of this Recovery Plan provideconditions for the owl across the landscape. Thislandscape should provide nesting, roosting,foraging, and dispersal macrohabitats in the

short term. This theme emphasizes protectingand monitoring owl populations and habitats.

The second theme acknowledges that ecosys-tems are temporally dynamic, and that provi-sions are needed to ensure owl habitat in thelong term. As nest sites change and are aban-doned, new nest sites should develop and be-come occupied. Allocation of mid- to late-seralforests needed by spotted owls, and other spe-cies, in future decades requires knowledge offorest disturbances, risks, and rates of successionat different spatio-temporal scales. We outlinebelow some disturbances, risks, and tools thatshould be considered in managing present andfuture owl habitat.

FireFireFireFireFire

Fire is the most rapidly acting of naturaldisturbances. A crown fire can quickly consumeforests across vast tracts. After a large crown fire,habitat components for nesting, roosting, andforaging are reduced or eliminated. Small-scalenatural fires and prescribed burns, however, canreduce fuel loadings and create small openingsand thinned stands that increase horizontaldiversity and reduce the spread of catastrophicfire. Small-scale fires and lightning also createsnags, canopy gaps, and large logs, plus theyperpetuate understory shrubs, grasses, and forbswhich are important habitat components to theowl, its prey, and other wildlife. Under naturalfire regimes prior to 1890 these small firesoccurred frequently (Moody et al. 1992).

The risk of catastrophic fires is widespread inSouthwestern forests and woodlands (Moody etal 1992). Fuel accumulations and forests over-stocked with trees place spotted owl habitat atrisk with respect to stand-replacing fires. FiguresII.D.1-3 show the changing fire record from1910 to 1992, based on records compiled at theFS Southwestern Regional Office. Because FSburn policies changed during this period and theuse of prescribed natural fire increased, interpre-tation of these records is not straightforward. Ingeneral, however, the figures document anincrease in both area burned per year and in arealost to catastrophic, stand-replacing fires.

The number of total natural and human-caused fires generally declined after 1981 (Figure

60


Volume I/Part II

II.D.1), but the number of large fires (> 4 ha[>10 acres]) increased at the same time (FigureII.D.2). Figure II.D.3 shows the trend clearly;from 1985 to 1992 the number of hectares thatburned increased. If the influence of two excep-tional fire years (1974 and 1979) is removed, thetrend shown in Figure II.D.3 remains; that is,the number of large fires increased.

Moody et al. (1992) estimated that about303,500 ha [750,000 acres] of mixed-coniferforest within FS Region 3 needed treatment toreduce fire risk in the next 10 years. Unmanagedand unplanned conversion of large areas offorests or woodlands to early seral conditions bywildfire can disrupt management goals to main-tain existing and to provide future spotted owlhabitat (USDA 1993c).

Characteristics of many nest and roost sitesof spotted owls place them at high fire risk.Some nest/roost locations at special topographiclocations (such as steep-walled canyons orisolated places) may be fire refugia, however.Taking cue from these, one promising manage-ment tactic is to isolate nest/roost sites from theadjoining high-risk forest by reducing flamma-bility and fire spread in a buffer around the site.This must be done, of course, without compro-mising the site itself as nest/roost habitat.

Inevitably, severe climatic conditions willoccur in the future, and extreme fire years arepossible (Swetnam and Betancourt 1990). Giventhe present conditions of Southwestern forests,extreme fire years could result in holocaustic firesthroughout large portions of the owl’s range.Because the resulting damage to owl habitatwould be irreparable in the forseeable future,efforts to limit large-scale catastrophic fires are ofutmost importance for owl conservation.

Increased use of fire and other tools will beneeded to reduce the amount of forest at highrisk from stand-replacing fires. The RecoveryTeam encourages proactive fire managementprograms which assume active roles in fuelsmanagement and understanding the ecologicalrole of fire. An example of such a program is theone employed by the Gila National Forest.

The Recovery Team recognizes that firetechnology may not be at the level of sophistica-tion needed to maintain owl habitat and createnew habitat. Although we advocate broadscale

use of fire in the Southwest, we also stress theneed to approach the use of fire in an adaptivemanagement context. Prescriptions that main-tain key structural features of owl and small preyhabitats should be developed and tested. Thesefeatures include large trees (which are often fireresistant), snags, logs, and understory hardwoodtrees. Treatments to produce or maintain suchhabitat components must be assessed by moni-toring to evaluate if treatment objectives weremet in both short and long terms. Wholesale useof fire without understanding or monitoring itseffects on habitat may render areas unusable byowls, and may also miss opportunities to im-prove our knowledge of fire effects. Fire andwildlife personnel should work together to refinefire prescriptions compatible with maintenanceof important habitat elements.

Other Natural DisturbancesOther Natural DisturbancesOther Natural DisturbancesOther Natural DisturbancesOther Natural Disturbances

The vegetative communities that providehabitat for the Mexican spotted owl are dynamicassemblages of living plants, snags, logs, andnumerous organisms active in decay and nutri-ent-cycling processes. Herbivory, disease, andstructural change caused by bacteria, fungi,insects, and vertebrates are natural agents ofchange in forest and woodland communities andoccur at scales ranging from individual trees tolandscapes.

These disturbances contribute to the forma-tion of complex landscape mosaics in whichwoodlands and forests consist of aggregates oftransient patches and gaps. Added to this patchi-ness are changes of the structural elements of owlhabitat caused by disturbances at scales largerthan gaps. Climate change, pollutants, and otherextensive events will produce effects of magni-tudes that are poorly understood (Davis 1989).Although management scarcely influences theprimary determinants of vegetation pattern(geology, climate, and genetics), managementcan affect vegetation by manipulating the extent,severity, and frequency of disturbance.

Land managers should recognize that naturaldisturbances can create and maintain diverse andproductive ecosystems that always include,somewhere on the landscape, an adequateamount and distribution of the vegetative

61


Volume I/Part II

FFFFFigurigurigurigurigure II.D.1.e II.D.1.e II.D.1.e II.D.1.e II.D.1. Historical record of number of total fires and number of natural fires. Data fromUSFS Southwestern Region.

FFFFFigurigurigurigurigure II.D.2.e II.D.2.e II.D.2.e II.D.2.e II.D.2. Historical record of fires over four hectares in size. Shown are numbers of fires andarea burned. Data from USFS Southwestern Region.

62


Volume I/Part II

FFFFFigurigurigurigurigure II.D.3.e II.D.3.e II.D.3.e II.D.3.e II.D.3. Five-year running averages of area burned. Data from USFS Southwestern Region

elements that are the required habitat for theMexican spotted owl. The word “adequate” iscrucial. Adequacy is derived from publiclyacceptable landscape descriptions (desiredconditions), together with use of the best succes-sion, allocation, and landscape-dynamic modelsto guide managers in how to get there. Adequacyis tested by ongoing monitoring and adaptivemanagement (III.C) and should not be assumedin the absence of monitoring.

Insects and microorganisms can be beneficialas well as destructive agents of plant succession(Dinoor and Eshed 1984, Knauer 1988,Dickman 1992, Haack and Byler 1993). Theseorganisms may produce large-scale communitychanges after periods of climatic stress that“predispose” forests to insects or pathogenicoccurrences (Colhoun 1979). Several groups offorest insects occasionally develop epidemicpopulations that severely damage mature foresttrees over large areas. Among the defoliatinginsects, the western spruce budworm killsunderstory white fir and Douglas-fir and thinsthe crowns of overstory trees (Archambault et al.1994). Outbreaks of western spruce budworm

occur every decade or so and extend widelyacross the landscape. Perhaps as a response to fireexclusion policies, recent budworm outbreakshave tended to be regionally synchronous withthe maturing of host species over large areas(Swetnam and Lynch 1993). As a complicatingfactor, trees that suffer declining vigor frommultiple years of defoliation by budworms maylose their resistance to more injurious wood-boring insects and ultimately die. Bark beetlesare important wood-boring insects in pinyon,ponderosa pine, Douglas-fir, and Englemannspruce. During outbreaks (about every 7 to 10years), these insects kill groups of mature trees.In longer outbreaks (usually those followingdroughts), mortality groups coalesce and damageappears to be widespread. Bark beetle popula-tions are most likely to increase where host treesare stressed as a result of sublethal fire damage,dwarf mistletoe infection, or where abundantgreen slash is available from thinning orblowdown.

The principal forest pathogens are rootdisease fungi and dwarf mistletoe. Armillariaroot disease is widespread across the forests of

63


Volume I/Part II

the Southwest. In a few locations it behaves as anaggressive killing agent (Marsden et al. 1993),but in most stands it acts to remove trees weak-ened by lightning or insects. Other root diseasesare caused by Heterobasidion annosum andPhellinus schweinitzii.

The most common tree disease in South-western forests is caused by parasitic seed plantsof the genus Arceuthobium, the dwarf mistletoes.About one-half to two-thirds of the stands inthese forests are infested by dwarf mistletoe.Infected trees become stunted, develop witches’brooms, and are eventually killed by this orother mortality agents. Both root disease andmistletoe typically occur as “centers” or “patches”and create slowly but continously expandingcanopy gaps. These agents increase ecosystemdiversity by producing snags, logs, and, in thecase of mistletoe, witches brooms. They also actsynergistically with forest insects.

The relationship between fire and dwarfmistletoe is complex. Brooms caused by dwarfmistletoe provide fuel continuity from ground totree crown. By maintaining seral trees in foreststands, fire increases the opportunity for mistle-toe infection because the seral trees are morecommonly hosts than climax trees. Similarcomplex relationships exist between fire, barkbeetles, and western spruce budworms.

White pine blister rust is caused by an exoticfungus that was recently introduced into theSacramento Mountains. It has the potential tokill most of the southwestern white pine in themixed-conifer forests (Hawksworth and Conklin1990) where the greatest concentration ofMexican spotted owls occurs. Although south-western white pine is seldom the most frequenttree species of a stand, it is an important seral,dominant, or codominant species in most areas.This tree produces large seeds and readily fillsgaps opened by mortality of other trees tobudworm, bark beetles, root disease, and mistle-toe. Therefore, the short-term effects of whitepine blister rust may be negative, since a strongreordering of forest tree composition may takeplace. A number of actions can be taken to“control” the rust and reduce its impacts, butthey are expensive and their effectiveness andpossible side effects are unknown. In the long-term, a genetic balance between the rust and

white pines may occur, as it did with Pinusmonticola in the northern Rockies (Ledig 1992).

Various other arthropods and saprophyticfungi are also important agents of deteriorationand decay of snags and logs. Although theseagents generally do not kill trees directly, theiractivity (decay) can lead to stem breakage andtree death. They are thus important in determin-ing the condition and persistence of coarsewoody debris within forest stands.

The cumulative impacts of these disturbanceagents on owl habitat depends on a number offactors, some of which are subject to manipula-tion. In general, these and other kinds of distur-bances affect forest nutrient and water cycles,solar penetration to the understory, and plantand animal food webs. The response of under-story vegetation, fungal-small mammal relation-ships (Maser et al. 1978), and owl prey tovarious disturbance factors can be positive ornegative, depending on numerous site factorsand the successional stage of the affected vegeta-tion. Because these processes are interactive andaffect a number of vegetation attributes, simpleassessments are inadequate. Several vegetationmanagement tools, including various kinds ofsilviculture, risk-abatement for fire or insect/disease damage, prescribed burning, and directpopulation control are appropriate in variouscombinations.

These disturbance agents should be consid-ered in developing management strategies forowl recovery. Managers must recognize that theorganisms discussed above and their effects arenot necessarily or even primarily bad. Certainnatural processes may interfere with short-termpriorities of forest management; but the perpetu-ation of forest conditions that support thosepriorities may depend on natural processescontinuing in the long term. Moreover, conflict-ing priorities, or even second- or third-levelpriorities may benefit from these organisms.Evaluations should be based on the role theseorganisms play in directing succession toward, oraway from, desired future conditions at differentspatiotemporal scales.

Managers, in consultation with specialists,can use these organisms to strategic advantage increating, enhancing, or maintaining habitats forowls (and associated biota) in accord with

64


Volume I/Part II

landscape goals. For example, dwarf mistletoecreates nest sites for owls in Douglas-fir. In someplaces, outbreaks of western spruce budwormeliminate understory host trees, helping toreduce fuel ladders that carry fires into treecrowns. These biotic agents of mortality havethinning effects on tree overstories. Such thin-ning affects nutrient and hydrological cycles,understory vegetation, and availability of prey toowls.

In summary, we encourage resource manag-ers to work with forest insect and disease special-ists to develop ecological assessments of thesekinds of disturbances at various scales(Kaufmann et al. 1994). Understanding thescientific basis of forest change and evolution iscrucial to successful management of forestecosystems, and therefore to recovery of thespotted owl.

Degradation of Riparian ForestsDegradation of Riparian ForestsDegradation of Riparian ForestsDegradation of Riparian ForestsDegradation of Riparian Forests

Riparian forests may also function as impor-tant components of ecosystems supportingspotted owls. These communities, particularlymature, multi-layered forests, could be impor-tant linkages between otherwise isolated sub-populations of spotted owls. They may serve asdirect avenues of movement between mountainranges or as stopover sites where drainages bisectlarge expanses of landscape that otherwise wouldbe inhospitable to dispersing owls. Further,historical evidence exists that spotted owls oncenested in such habitats.

Many riparian ecosystems have deterioratedin the Southwest (Cooperrider 1991, Bock et al.1993, USDI 1994), and the loss of riparianhabitat was one of the reasons for listing the owl(Part I). Dick-Peddie (1993) estimated frommap and air photo data that 96% of the RioGrande riparian area in New Mexico has beenlost to urbanization, agriculture, water impound-ments, and other modifications. Gallery foreststhat once extended into woodlands, grasslands,and deserts have significantly declined or dete-riorated, adversely affecting numerous wildlifepopulations (Minckley and Clark 1984, Skovlin1984, Minckley and Rinne 1985, Bock et al.1993, USDI 1994). Efforts to improve riparianand watershed conditions (DeBano and Schmidt

1989a, 1989b) could facilitate movements ofspotted owls between distant geographic loca-tions and perhaps even provide nesting habitat.A wide variety of other organisms would alsobenefit from healthier riparian systems.

TIMBER HARVEST ANDTIMBER HARVEST ANDTIMBER HARVEST ANDTIMBER HARVEST ANDTIMBER HARVEST ANDSILVICULTURAL PRACTICESSILVICULTURAL PRACTICESSILVICULTURAL PRACTICESSILVICULTURAL PRACTICESSILVICULTURAL PRACTICES

Historically, the principal objectives of forestmanagement were to derive economic gain andcommodities from forests. Silviculture has greatpotential as a tool for meeting other objectives,however, such as maintaining and developingMexican spotted owl habitat, alleviating fire risk,minimizing impacts of insects and disease, andenhancing various ecological values. In thissection, we review past timber-harvest practicesin the Southwest and contrast those practiceswith alternatives. Our focus is the potentialeffects of these practices on Mexican spottedowls.

Historical PerspectivesHistorical PerspectivesHistorical PerspectivesHistorical PerspectivesHistorical Perspectives

Past PracticesPast PracticesPast PracticesPast PracticesPast Practices

The primary factors leading to the listing ofthe Mexican spotted owl were adverse modifica-tion of its habitat as the result of even-agedmanagement and plans to continue this harvestmethod as detailed in existing Forest Plans.Fletcher (1990) reported the loss of >325,000 ha(800,000 acres) of spotted owl habitat within FSRegion 3 as the result of human activities,primarily forest management. Silvicultureemphasized even-aged systems which tended tosimplify stand structure and harvest a dispropor-tionate share of large trees. The Team used pastforest inventory data to evaluate the change inthe size-class distribution of trees from the 1960sto the 1980s. The trend that emerged from ouranalysis was a substantial increase in the densityof trees 12.7-32.8 cm (5-12.9 in) dbh, but alarge decrease in numbers of trees >48.3 cm (19in) dbh (see below). As discussed by Ganey andDick (1995), large trees are an important com-ponent of spotted owl habitat; thus, the 20%decrease in numbers of trees >48.3 cm (19 in)

65


Volume I/Part II

dbh removed a key habitat component of theMexican spotted owl. The simplification of standstructure is not so easily quantified. Given thatmostly even-aged management was used, how-ever, the conclusion of stand simplification isreasonable.

Forest PlansForest PlansForest PlansForest PlansForest Plans

Existing Forest Plans and their underlyingstandards and guidelines are fairly explicit withrespect to the silvicultural practices to be usedand the expected timber volumes to be extracted.These Forest Plans articulate classic even-agedmanagement regimes with regeneration treat-ments occurring at 120-year intervals, intermedi-ate treatments employed to maintain open standconditions, and disease-control treatments asconditions warrant. Thus, management calledfor fairly frequent entries into a stand. Further,this management system stressed simple standstructures, decreased residual densities, andelimination of large, slow-growing, but highvalue trees (primarily ponderosa pine and Dou-glas-fir). Salvage, sanitation, fuel reductions, andfuelwood harvest as specified in Forest Planscombined to reduce numbers of snags, anothercorrelate of spotted owl habitat. In summary,even-aged management as specified in ForestPlans is incompatible with maintaining anddeveloping spotted owl habitat. The Team isencouraged, however, by recent efforts by FSRegion 3 to amend forest plans to incorporatethe recommendations proposed in this RecoveryPlan, and to emphasize uneven-aged manage-ment as the preferred silvicultural system in theRegion.

Habitat TrendsHabitat TrendsHabitat TrendsHabitat TrendsHabitat Trends

Historical and current trends in spotted owlhabitat are presently unknown. Numerousfactors underlie this lack of knowledge, but thepaucity of reliable vegetation data is the mostglaring explanation. This lack of credible datahas not precluded rampant speculation onhabitat trend, however. In general, habitat trendis perceived in two divergent ways. One view isthat past timber harvest within the forest typesused by Mexican spotted owls has caused a

dramatic decline in habitat quantity and quality.Indeed, the conclusion of historical habitat losscoupled with projections for additional habitatloss were the primary factors for listing thesubspecies (Part I). The contrary view suggeststhat many years of fire exclusion within South-western forests has allowed mixed-conifer foresttypes to increase at the expense of meadows andfire-disclimax species such as quaking aspen andponderosa pine (USDA 1993b, Johnson 1994).Further, Southwestern ponderosa pine forests areknown to be generally denser today than theywere in pre-settlement times (Covington andMoore 1992, 1994a, b, c). Based on this infor-mation, USDA (1993b) concluded that habitatsuitability for the Mexican spotted owl hadincreased. Because of these conflicting views, theTeam attempted a quantitative evaluation ofhabitat trend with respect to the Mexican spot-ted owl.

Data Availability.Data Availability.Data Availability.Data Availability.Data Availability.—Limited sources of data areavailable for assessing habitat trend. Withinforested types, forest inventories from the 1960s(Choate 1966, Spencer 1966) and the 1980s(Conner et al. 1990, Van Hooser et al. 1993)have been compared by USDA (1993b) andJohnson (1994) and are used, in part, for ouranalyses. We admit, however, that differences indefinitions and in how data were collected makecomparisons between the 1960s and 1980s datatenuous, at best (Van Hooser et al. 1993). Thesedifferences include: (1) changes in definitions ofvegetation types; (2) changes in the landbasebeing sampled, (e.g., changes in wildernessdesignation); and (3) changes in samplingintensity.

The following comparisons are limited tocommercial forest lands within the States ofArizona and New Mexico on a per hectare basis.Thus, all forest types are included but, unlikeUSDA (1993b) and Johnson (1994) we do notextrapolate the data to unsampled forested landssuch as wilderness areas. Therefore, our analysesfocus on changes on commercial forest landswhere data exist. Because of differences in landdesignations (i.e., commercial timber landbecoming wilderness between the two samplingperiods), comparisons of raw values are poten-tially misleading. Thus, our comparisons are

66


Volume I/Part II

primarily restricted to evaluations of propor-tions. To compare stand structure, we usedrelative frequencies of trees by size class. Wereiterate that caution is warranted when inferringconclusions from these data, but submit thatsome gross generalizations are possible.

Trends in Forest Landbase and TimberTrends in Forest Landbase and TimberTrends in Forest Landbase and TimberTrends in Forest Landbase and TimberTrends in Forest Landbase and TimberVolume.—Volume.—Volume.—Volume.—Volume.—Total forested land increased from4,516,000 to 4,750,000 ha (11,160,000 to11,738,000 acres) from the 1960s to the 1980s,roughly a 5% increase. The commercial forestlandbase decreased by approximately 15%(624,000 ha [1,541,000 acres]), however, andreserved forested lands increased by 858,000 ha(2,119,000 acres). Growing stock (i.e., theharvestable volume) on commercial lands de-creased from 12,707 MMCF to 11,549 MMCF.This decrease is not surprising given the volumeof timber harvested on commercial forest landsand the decrease in the amount of commercialforest lands from the 1960s to the 1980s.

Trends in Forest Types.Trends in Forest Types.Trends in Forest Types.Trends in Forest Types.Trends in Forest Types.—Within the commer-cial landbase, mixed-conifer forests comprisedapproximately 11% of total area in the 1960s

TTTTTable II.D.1.able II.D.1.able II.D.1.able II.D.1.able II.D.1. Changes in the area (ha X 1,000 [acres x 1,000]) and distribution of forest types fromthe 1960s to 1980s on commercial forest lands within Arizona and New Mexico. Data from Choate(1966), Spencer (1966), Conner et al. (1990), Van Hooser et al. (1993).__________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

LandbaseLandbaseLandbaseLandbaseLandbase PPPPPrrrrroporoporoporoporoportion oftion oftion oftion oftion of LandbaseLandbaseLandbaseLandbaseLandbase PPPPPrrrrroporoporoporoporoportion oftion oftion oftion oftion of Change in Change in Change in Change in Change inFFFFForororororest est est est est TTTTTypeypeypeypeype in 1960sin 1960sin 1960sin 1960sin 1960saaaaa 1960s Landbase1960s Landbase1960s Landbase1960s Landbase1960s Landbasebbbbb in 1980sin 1980sin 1980sin 1980sin 1980saaaaa 1980s Landbase1980s Landbase1980s Landbase1980s Landbase1980s Landbasebbbbb P P P P Prrrrroporoporoporoporoportiontiontiontiontionccccc

__________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________Ponderosa Pine 3,234 78 2,530 72 -6

[7,992] [6,252]Mixed-conifer 475 11 709 20 9

[1,173] [1,752]Spruce-fir 257 6 201 6 0

[635] [496]Quaking aspen 180 4 81 2 -2

[446] [201]TTTTTotalotalotalotalotal 4,146 4,146 4,146 4,146 4,146 100 100 100 100 100 3,523 3,523 3,523 3,523 3,523 100 100 100 100 100

[10,246][10,246][10,246][10,246][10,246] [8,701][8,701][8,701][8,701][8,701]__________________________________________________________________________________________

a Landbase in hectares [acres] covered by the forest type.b Proportion of the total forested landbase belonging to the forest type.c (Prop. 1960s landscape)-(Prop. 1980s landscape).

67

and 20% of the total area in the 1980s, a 9%increase (Table II.D.1). Possible explanations forthis change include: (1) increasing invasion ofmixed-conifer species (presumably Douglas-firand white fir) into other types, such as mead-ows; (2) more liberal definitions of mixed-conifer (i.e., includes types previously classifiedas something else); (3) quaking aspen giving wayto other species in the absence of fire; and (4)selective harvest of ponderosa pine, leavingresidual forests composed primarily of otherconifer species. Any of these reasons may explainthe perceived changes of forest type; probably allof these and other factors contributed to somedegree. We speculate that classification changesaccount for most of the change, and that selec-tive removal of ponderosa pine and the succes-sion of quaking aspen stands to mixed-coniferare also plausible short-term explanations.Conversely, we have difficulty accepting thatencroachment of mixed-conifer species intoother forested types was responsible for morethan a relatively small portion of this changewithin the twenty-year period. Thus, any gener-alizations concerning changes in forest types andany actions proposed to reverse these trends


Volume I/Part II

TTTTTable II.D.2.able II.D.2.able II.D.2.able II.D.2.able II.D.2. Changes in the density (trees/ha [trees/acre]) and distribution of tree size classes fromthe 1960s to 1980s on commercial forest lands within Arizona and New Mexico. Data from Choate(1966), Spencer (1966), Conner et al. (1990), Van Hooser et al. (1993).

TTTTTrrrrreeeeeeeeee DDDDDensityensityensityensityensity PPPPPrrrrroporoporoporoporoportion oftion oftion oftion oftion of DDDDDensityensityensityensityensity PPPPPrrrrroporoporoporoporoportion oftion oftion oftion oftion of Change inChange inChange inChange inChange in DDDDDensityensityensityensityensitySSSSSizizizizize Classe Classe Classe Classe Class in 1960sin 1960sin 1960sin 1960sin 1960saaaaa 1960s 1960s 1960s 1960s 1960s TTTTTotal otal otal otal otal bbbbb in 1980sin 1980sin 1980sin 1980sin 1980saaaaa 1980 1980 1980 1980 1980 TTTTTotalotalotalotalotalbbbbb PPPPPrrrrroporoporoporoporoportiontiontiontiontionccccc ChangeChangeChangeChangeChangeddddd

2.5-12.5 cm 146.2 62.5 134.2 53.7 -8.8 -8.3[1.0-4.9 in] [59.2] [54.3]12.6-32.8 cm 70.3 30.0 98.5 39.4 9.4 40.2[5.0-12.9 in] [28.5] [39.9]32.9-48.0 cm 12.1 5.2 13.0 5.2 0.0 7.5[13.0-18.9 in] [4.9] [5.3]>48 cm 5.4 2.3 4.3 1.7 -0.6 -20.4[>19 in] [2.2] [1.7]

a Tree density in no./ha [no./acre]b Proportion of the total number of trees within that size class.c (Prop. 1960s total)-(Prop. 1980s total).d ([Prop. 1960s total]-[Prop. 1980s total])/(Prop. 1960s total).

must acknowledge this uncertainty, and shouldconsider all plausible explanations for thesetrends.

Trends in Size-class Distributions.Trends in Size-class Distributions.Trends in Size-class Distributions.Trends in Size-class Distributions.Trends in Size-class Distributions.—We noteda change in the size-class distribution of trees oncommercial forest lands of Arizona and NewMexico (Table II.D.2). Sapling-sized trees (2.5-12.5 cm [1-4.9 in] dbh) decreased in bothabsolute density and in relative contribution tothe size-class distribution; trees 12.6-31 cm(5-12 in) dbh increased in density by 40% andin relative proportion of the size class distribu-tion by >9%; and trees in the 31-48 cm (13-19in) size class increased in density but not inrelative proportion of the tree distribution.Finally, the density of large trees (>48 cm [19 in]dbh) decreased from 2.3 to 1.7 trees/ha (0.9 to0.7 trees/ac), a 20% decline. This decrease inlarge trees would be expected given past timberharvest practices which emphasized harvest ofthe large trees. Possible explanations for theincrease in smaller stems include the growth ofregeneration, limited pre-commercial thinning,fire suppression, and the lack of interest by theforest industry in the smaller-sized stems.

Summary of Recent Habitat Trends.Summary of Recent Habitat Trends.Summary of Recent Habitat Trends.Summary of Recent Habitat Trends.Summary of Recent Habitat Trends.—Ouranalyses indicate that between the 1960s and

1980s (1) total forested acres increased, (2)mixed-conifer types apparently covered more ofthe landbase (but see above cautions regardingthis conclusion), and (3) densities of large treesdeclined. Although the amount of total forestedland has increased and the amount of mixed-conifer forest may have increased, we doubt thatthe amount of Mexican spotted owl habitat hasincreased concomitantly. Given the 20-yr periodbetween inventories, most of these additionalacres are likely in early successional stages andunlikely to possess the habitat characteristicsused by spotted owls. Conversely, the 20%decrease in the density of large trees is an alarm-ing negative trend with respect to a very criticalcomponent of spotted owl habitat.

Silvicultural PracticesSilvicultural PracticesSilvicultural PracticesSilvicultural PracticesSilvicultural Practicesand Forest Managementand Forest Managementand Forest Managementand Forest Managementand Forest Management

Four common forest structures occur natu-rally or by silvicultural efforts. These structuresare even-aged, balanced uneven-aged, irregularuneven-aged, and even-aged/uneven-agedstratified mixtures. Even-aged stands, for themost part, are characterized by most trees beingapproximately the same age. The general con-vention is that the spread of ages within thestand are within approximately 20% of the

68


Volume I/Part II

specified rotation age. Two types of uneven-agedstands are balanced and irregular stands. In bothcases, at least three distinct age classes exist. Abalanced uneven-aged stand equates to each ageclass occupying roughly equal areas. Distributionby diameter class approximates a reverse, j-shaped curve. Irregular uneven-aged stands havesome age and associated diameter classes missingacross the possible range of ages and diameters.A two-storied stand, one with two distinct age-class and diameter distributions, is neither even-or uneven-aged, but is intermediate between thetwo. Stratified mixtures occur where trees areessentially even-aged, but differences in growthrates and shade tolerance among tree speciesresult in multiple canopy strata. This structurealso occurs when selective regeneration of shade-tolerant species or high site productivity leads toheterogeneous age and diameter distributions.In much of the mixed-conifer type in the South-west, the stratified-mixture of stand structureappears to be relevant to habitats used by spottedowls.

SilvicultureSilvicultureSilvicultureSilvicultureSilviculture

Silviculture has been variously defined as (1)the art of producing and tending a forest, (2) theapplication of knowledge of silvics in the treat-ment of a forest, and (3) the theory and practiceof controlling forest establishment, composition,structure and growth (Smith 1986). In a generalsense, silviculture is the practice of managingforest establishment, composition, structure, andgrowth to meet stated objectives. Thus, silvicul-ture should be regarded as a system of treatmentsand not solely the practice of removing treesfrom a stand.

In the Southwest, two broad classificationsof silvicultural systems, based on methods ofreproduction and resulting age-class mixes offorested stands, are even-aged and uneven-agedmanagement. These are reviewed below; again,our focus is the potential these systems have fordeveloping spotted owl habitat.

Even-aged Management.Even-aged Management.Even-aged Management.Even-aged Management.Even-aged Management.—Even-aged manage-ment has been used commonly in Southwesternforests. Reasons for its popularity are based bothon ecology and economics. Ecologically, even-

aged systems favor species with limited shadetolerance, such as quaking aspen, oaks, andlodgepole pine. Shade tolerance is the ability toreproduce and grow under the shade of larger,taller trees. Shade-tolerant species typicallyinclude true firs in Southwestern forests. Ponde-rosa pine is considered to be of intermediateshade tolerance, but tending toward the intoler-ant side. Economically, even-aged managementis more efficient when considering short-termcosts of site utilization, sale preparation, trans-portation systems, harvesting, and slash reduc-tion. Further, even-aged systems are easier tomodel, administer, and track over time.

Regeneration methods within even-agedsystems of the Southwest include shelterwood,clearcutting, and seed tree methods. Theshelterwood method typically has a series ofcuttings. The first treatment in mature stands isto stimulate cone and seed production forregeneration. This is followed by a series oftreatments that remove the larger, older stems asregeneration matures. Variations on the generalmethod include irregular shelterwood andgroup-shelterwood. Clearcutting involves theremoval of the entire stand in one cutting.Reproduction is obtained artificially by seedingor planting, or naturally by seeding from adja-cent stands. This method is appropriate forshade-intolerant species. In appearance,clearcutting is indistinguishable from the cop-pice-forest method of regeneration for quakingaspen, where reproduction is obtained fromsuckering of sub-terrain clones. The seed-treemethod resembles clearcutting except that a fewtrees are left to provide a source of seed withinthe treated area. Of these three, the shelterwoodmethod is used most commonly in the South-west; clearcut and seed-tree methods are usedinfrequently.

Variations of even-aged management areused throughout the Southwest, but all share thefollowing characteristics. A predetermined timefor regeneration of the stand is set a priori; thisregeneration time can vary. The FS Region 3generally schedules regeneration treatments from100 to 120 years of stand age, an age when treesare expected to reach 45.7 cm (18 in) dbh as themaximum size. The management objective is tomaximize total volume over time while provid-

69


Volume I/Part II

ing a “saw-timber” sized product. Residual standdensity can be controlled by thinning at theo-retically scheduled entries in the stand. Thisenables the capture of mortality on a semi-regular basis and provides intermediate revenuesfrom timber harvest.

Even-aged stand structures are not used toany great extent by the Mexican spotted owl.Further, with its intent to promote uniformity intree age, size, spacing, and density, even-agedmanagement generally would not be a preferredsystem for short-term development of spottedowl habitat. Even-aged management may beappropriate to maintain quaking aspen withinthe mixed-conifer type, however. Quaking aspenis typically even-aged in its earlier stages of standdevelopment. Because of its extreme shadeintolerance and requirements for elevated soiltemperatures for sprouting, quaking aspenshould be managed under even-aged systems.Later seral stages of quaking aspen that include amixed-conifer understory appear both as asimple mixture of an even-aged overstory anduneven-aged understory, or as a stratified mix-ture. As decadent quaking aspen stands arereplaced by the shade-tolerant conifers, spottedowl nesting/roosting habitat begins to develop.In summary, even-aged management has limitedpotential with respect to developing the types ofstand structures used by spotted owls. Neverthe-less, even-aged management is a critical tool tomeet specific objectives, and can meet certainecosystem objectives if employed at the properscale. Perhaps the primary objection to its use inthe past was its uniform, widescale applicationacross the Southwest.

Uneven-aged Management.Uneven-aged Management.Uneven-aged Management.Uneven-aged Management.Uneven-aged Management.—Uneven-agedmanagement entails the removal of timber in allsize classes on a periodic basis so that regenera-tion is continuously established over time, andstand size-class distribution is regulated. Un-even-aged management is loosely based on thepremise that density control across a range ofdiameter classes will ensure growth of stems overtime to a set maximum diameter, while ensuringregeneration of species at regular intervals overtime. Only one reproduction method, theselection method, is used with uneven-agedmanagement. The selection method provides

openings in the stand to enable regeneration tooccur. Simultaneous with the selection andharvest of trees to provide growing space forregeneration, trees across all diameter classes arethinned to ensure the desired distribution ofsize- and age-classes within the stand.

Two variations of the selection method areindividual tree selection and group selection.Individual tree selection, as the name implies,involves the removal of single, scattered trees.This method generally favors shade-tolerantspecies, but this is also a function of the residualstocking levels. Group selection entails theremoval of a small patch of trees; the width ofthe patch is usually less than twice the height ofthe dominant (i.e., largest) tree. This is some-what analogous to a very small clearcut, but thedifference between the group selection and theclearcut method is in the spatial scale of applica-tion. Group selection is used to create a balanceof age- or size-classes in small contiguous groupsresulting in a mosaic within a stand. In contrast,even-aged methods are typically applied to anentire stand. Group selection can be used topromote the establishment and growth of shade-intolerant trees since there is opportunity toreduce localized residual densities and theamount of area shaded.

Group selection offers a number of advan-tages for the development of potential spottedhabitat over single-tree selection techniques.Application of the group selection method couldprovide a mosaic of many small even-aged ortwo-storied groups across a forest stand. Regen-eration of shade-intolerant species is possiblewhere a reproduction source, either clones orseeds, is present. With respect to insect anddisease problems, management options increasefor both suppression and prevention, especiallyin mixed-species stands. Edge effects found atgroup interfaces can provide structural featuresand openings that mimic gap-phase regenera-tion, and provide early-seral vegetation for preyspecies (Ward and Block 1995). In some cases,group-selection methods may result in lessresidual damage to the stand as the result oflogging activities than single-tree selection.

Uneven-aged silvicultural practices predomi-nate on Tribal lands where commercial timberharvest exists. Reasons for the emphasis on this

70


Volume I/Part II

silvicultural system include aesthetics, providingforest cover over all lands simultaneously, theperception that it provides a more even-flow ofproducts than even-aged management, and thefact that it allows continuous regeneration.

We have not been able to assess the effects ofclassic uneven-aged management on Mexicanspotted owl habitat because we were unable toacquire data for most areas where uneven-agedmanagement is practiced on a large scale. How-ever, based upon our understanding of theapplication of uneven-aged systems, standdensity is often kept at a fairly low level, seldomexceeding 18 m2/ha (80 ft2/acre) of basal area.These low residual stand densities allow forregeneration and growth of ponderosa pine.Uneven-aged systems, whether they retainindividual trees or groups of trees, allow for thedevelopment of multiple canopy levels, a keycomponent of Mexican spotted owl habitat.However, Ganey and Dick (1995) demonstrateclearly that owl habitat typically also includessignificant numbers of large trees. These largetrees may not be retained where uneven-agedmanagement is applied in this fashion.

In summary, uneven-aged management hassome promise for providing stands exhibitingcharacteristics of spotted owl habitat. As cur-rently practiced, however, uneven-aged manage-ment results in large acreages of low-densitystands, numerous road openings, and the even-tual eradication of large diameter stems. Al-though neither the short- or the long-termeffects of these applications on spotted owls areknown, this type of application may not be thebest option for producing spotted owl habitat.

Development or maintenance of stratifiedDevelopment or maintenance of stratifiedDevelopment or maintenance of stratifiedDevelopment or maintenance of stratifiedDevelopment or maintenance of stratifiedmixtures.mixtures.mixtures.mixtures.mixtures.—Stratified mixtures can originate invarious ways, including stand-replacing events(e.g., crown fire, even-aged management) thatmay or may not leave remnant stems. Theestablishment of stratified mixtures is not likelyin mixed-conifer types within the short timeframe (10-15 years) of this Recovery Plan.Maintenance of such stands could be consideredin the short-term, however, and development ofstratified mixtures could be incorporated inlonger-term management plans.

Management for stratified mixtures mustconsider the mix of species along the continuumof shade tolerance. Thus, any regenerationefforts should be designed to have enoughopenings of sufficient size for seedling/sproutingestablishment and release. Openings can beaccomplished by group selection cuts favoringretention of shade-intolerant species, or selectionof individual trees adjacent to stems that couldprovide a seed source for regeneration. Onetechnique to consider is small-scale seed tree cuts(perhaps to be thought of as group seed-treeselection cuts), which would provide both a seedsource and trees for ultimate snag development.This method could maintain shade-intolerantspecies, but would not be intrusive enough toproduce even-aged structure throughout thestand. Within stratified mixtures, intermediatetreatments including pre-commercial andcommercial thinning could be beneficial inincreasing the growth of residual trees. At somepoint, however, further treatments should bedeferred, and natural stand maturation andsuccession should be allowed to proceed untileither (1) the stand is no longer spotted owlhabitat; or (2) the stand can be replaced byhabitat (preferably occupied by spotted owls)that has been developed elsewhere.

ConclusionsConclusionsConclusionsConclusionsConclusions

Clearly, recent forest management practicesand those detailed in existing Forest Plans arenot beneficial to Mexican spotted owls. Relianceon traditional forest management and silvicul-tural techniques may no longer be possible, notonly with respect to the conservation of theMexican spotted owl but also with respect tomaintaining other ecosystem attributes. Newapproaches must be developed that ensure thelong-term provision of owl habitat and themaintenance of ecosystem structure and func-tion. Traditional approaches will still have theirrole, but perhaps used in slightly different waysand with different intensities. Innovative appli-cations of uneven-aged management may beparticularly useful in developing and maintain-ing spotted owl habitat. In addition, particularapplications of uneven-aged management maybe useful in maintaining habitat conditions for

71


Volume I/Part II

the owl where they exist. In some cases, theapplication of even-aged management systemsmay also be appropriate, so future forest man-agement should not preclude the use of even-aged management.

GRAZINGGRAZINGGRAZINGGRAZINGGRAZING

Grazing by livestock and wildlife (e.g., elk,deer) occurs throughout the range of the Mexi-can spotted owl. Depending on the intensity,grazing has the potential to influence habitatcomposition and structure, and affect foodavailability and diversity for the owl. However,predicting the magnitude of grazing effects onspotted owls and their habitat, and evaluatingmanagement options requires a better under-standing of the relationship between spotted owlhabitat and grazing.

Specific studies that document the effects oflivestock and wildlife grazing on spotted owlhabitat have not been conducted. Until specificinformation is available, the potential effects onthe owl of grazing and trampling of vegetationmust be identified and considered to the extentpossible. For example, livestock and wildlife maynot impact spotted owl roost and nest sitesimmediately, but could alter riparian habitats byreducing, eliminating, or suppressing regenera-tion. In time, reduced regeneration could limitthe development of overstory structure neededfor nesting, roosting, and other life historyneeds, as well as jeopardize the sustainability ofthese habitat types.

Grazing can alter a plant community di-rectly, indirectly, or both. Direct alterations maybe as obvious as plant removal by consumptionor as subtle as removal by trampling. Indirectalterations may be as straightforward as loss ofseed source or as insidious as damaged soil(Dwyer et al. 1984, Kauffman and Krueger1984, Fleischner 1994). Moderate to heavygrazing can reduce plant density, cover, biomass,vigor, and regeneration ability. Collectively, thesefactors can alter the relative composition andstructure of grass, forb, shrub, and tree compo-nents in an area (Hanley and Page 1982,Zimmerman and Neuenschwander 1984, Schulzand Leininger 1990, Milchunas and Lauenroth1993). Within conifer forests, grazing can

remove or greatly reduce grasses and forbs,thereby allowing large numbers of coniferseedlings to become established because ofreduced competition for water and nutrients andreduced allelopathy. Establishment of largenumbers of seedlings coupled with the reductionin light ground fuels (i.e., grasses and forbs) mayact synergistically with fire suppression tocontribute to dense overstocking of ladder fuels.This dense overstocking can alter forest structureand composition and degrade spotted owl andprey habitats while increasing risks of stand-replacing fires.

Beyond the effects of grazing on plants,livestock activity can increase duff layers, acceler-ate decomposition of woody material, producecompacted soils, damage stream banks andchannels, and damage lake shores (Kennedy1977, Blackburn 1984, Kauffman and Krueger1984, Skovlin 1984, Clary and Webster 1989).The combination of these changes to the bioticand physical landscapes also affects plant com-munity composition, structure, and vigor. Ifsuch changes occur in or near areas used byspotted owls, then grazing can influence the owl.Those influences can be manifested by altering(1) prey availability, (2) susceptibility of spottedowl habitat to fire, (3) the health and conditionof riparian communities; and (4) development ofhabitat. We summarize below the major sus-pected influences of grazing on Mexican spottedowls.

1. For the Mexican spotted owl, preyavailability is determined by the distribu-tion, abundance, and diversity of preyand by the owl’s ability to capture it.Grazing may influence prey availabilityin dissimilar ways. For example, grazingthat reduces dense grass cover can createfavorable habitat conditions for deermice while creating unfavorable condi-tions for voles, meadow jumping mice,and shrews (Medin and Clary 1990,Schultz and Leininger 1991). Thischange might decrease prey diversity(Medin and Clary 1990, Hobbs andHuenneke 1992). A diverse prey base canprovide a more predictable food resourcefor the owls over time, because popula-

72


Volume I/Part II

tions of many small mammals fluctuateasynchronously. Conversely, short-termremoval of grass and shrub cover mayimprove conditions for the owl to detectand capture prey. Long-term loss ofgrasses, forbs, and shrubs may promotetree growth and cover that could decreaseprey abundance. Thus, grazing can poseecological tradeoffs.

2. Grazing that significantly reduces herba-ceous ground cover and increases shrubsand small trees can decrease the potentialfor beneficial low-intensity ground fireswhile increasing the potential fordestructive high-intensity vertical fires(Zimmerman and Neuenschwander1984). Low-intensity ground firesprevent fuel accumulation, stimulatenutrient cycling, promote grasses andforbs, discourage shrubs and trees, andperpetuate the patchiness that supportssmall mammal diversity. Catastrophicfire reduces or eliminates foraging,wintering, dispersal, roosting, andnesting habitat components.

3. Excessive grazing in riparian areas canreduce or eliminate important shrub,tree, forb, and grass cover, all of which insome capacity support the owl or itsprey. Excessive grazing can also physicallydamage stream channels and banks(Ames 1977, Kennedy 1977, Kauffmanet al. 1983, Blackburn 1984, Slovkin1984, Clary and Webster 1989, Platts1990.) Deterioration of riparian vegeta-tion structure can allow channel widen-ing. This event, in turn, elevates waterand soil temperatures and thus evapora-tion and lowering of water tables, plus itsignificantly increases the potential foraccelerated flood damage (Platts 1990).These processes alter the microclimateand vegetative development of riparianareas, potentially impairing its use byspotted owls.

4. Excessive grazing, sustained for longperiods, can inhibit or retard an area’s

ability to produce or eventually matureinto habitat for the owl or its prey. Thiswill probably prove to be an inevitableconsequence of the events and processesdescribed above.

The potential for grazing to influencevarious components of spotted owl habitatcannot be ignored. However, current predictionsof grazing effects on plant communities as theyrelate to the owl are inexact. Thus, the integra-tion of spotted owl needs and grazing manage-ment will require coordination, and an interac-tive and adaptive approach between protection,restoration, and management.

RECREATIONRECREATIONRECREATIONRECREATIONRECREATION

Recreational activities may affect Mexicanspotted owls directly by disturbing nests, roosts,or foraging sites. Disturbance may occur indi-rectly through altered habitat caused by tram-pling of vegetation, soil damage, or both. Devel-oping new recreation facilities or expandingexisting facilities, such as campgrounds andtrails, may alter spotted owl habitat and habitatuse and perpetuate disturbance impacts causedby recreation.

If a given recreational activity does not causehabitat alteration, the Team assumes that thatactivity generally has relatively low impactpotential with respect to spotted owls. However,exceptions may exist in local situations or certainRUs where the level of recreational activities ishigh. Essentially, the determining factor of anactivity’s impact on spotted owls is a combina-tion of its location, intensity, frequency, andduration rather than simply its character.

Types of RecreationTypes of RecreationTypes of RecreationTypes of RecreationTypes of Recreation

Recreational activities fall into several catego-ries; the number, size, and intensity of suchactivities will vary with location. The followinggeneral categories include most widespreadrecreational activities that might affect spottedowls and their habitat.

73


Volume I/Part II

CampingCampingCampingCampingCamping

Although the effects of camping on spottedowls have not been studied, disruption of nest-ing, roosting, and foraging activities is a distinctpossibility. The character of camping variesdramatically, however. One person may campalone in a small tent, whereas others may campin groups with motorhomes. The disparity incharacter does not necessarily translate to distur-bance potential, however. One person campingin a nest grove could be more disruptive than 12people camping in a foraging area. Therefore,blanket generalizations about the impacts ofcamping activities are inappropriate. Theseactivities should be assessed on a case-by-casebasis, considering factors such as the location ofthe activity relative to the owls, the number ofindividuals involved, the type of group involved,and the frequency and duration of the activity.

HikingHikingHikingHikingHiking

Hiking is typically a short-term activity, andmay bring a person into and out of an owl’spresence relatively quickly. Most spotted owlsappear to be relatively undisturbed by smallgroups (<12 people) passing nearby. Largergroups are probably more disruptive, but themore serious threat of disturbance probablyarises where there is steady hiking traffic. Populartrails through spotted owl habitat may attractenough hikers to disturb owls. Certain kinds ofhiking activities may degrade portions of spottedowl habitat, disrupt crucial behaviors, increasesusceptibility of owls to predation, or causeabandonment of a nest area or key roost grove.The potential for hikers to disturb owls isprobably greatest where hiking is concentrated innarrow canyon bottoms occupied by nesting orroosting owls. Again, we argue that blanketstatements about the effects of hikers on owls areinappropriate, and recommend evaluation on acase-by-case basis as described above.

Off-road VehiclesOff-road VehiclesOff-road VehiclesOff-road VehiclesOff-road Vehicles

Both motorized and nonmotorized vehiclesmay degrade or destroy spotted owl habitat,particularly meadow and shrub habitats vital to

the owl’s prey. Noise produced by vehicles andthe vehicle riders may disturb spotted owls atimportant nesting and roosting sites.

Rock-climbingRock-climbingRock-climbingRock-climbingRock-climbing

In some portions of its range, the spottedowl nests and roosts in shallow recesses and cavesassociated with canyon walls and cliffs. Rock-climbing activities in the vicinity of cliff-dwell-ing spotted owls could disturb the owls, particu-larly during the nesting season. This problemcould be partially alleviated by invoking seasonalclosures in areas of conflict. Again, case-by-caseevaluations of activities and their potential fordisturbance seem most appropriate.

Wildlife Viewing and PhotographingWildlife Viewing and PhotographingWildlife Viewing and PhotographingWildlife Viewing and PhotographingWildlife Viewing and Photographing

Because birders and wildlife photographersactively seek spotted owls, their encounters maybe more disruptive than the accidental encoun-ters associated with other recreational activities.Such recreationists often make repeated visitsand may follow birds that flush. They oftenemploy hooting or mousing techniques to attractthe owls, and these behaviors, practiced toexcess, may disrupt owls’ territorial, mating, andnesting activities.

Recreation SummaryRecreation SummaryRecreation SummaryRecreation SummaryRecreation Summary

Incidental encounters between spotted owlsand people pursuing some recreational activityare relatively insignificant in most cases. In othercases, there may be significant effects. These arerelatively uncommon, and are typically localized.Consequently, these situations will usuallyimpact one or at most a few pairs of owls, andare not likely to impact large portions of the owlpopulation. We believe that these situations arebest evaluated on a case-by-case basis.

In a broader sense, the construction ofrecreation facilities, the loss of habitat to makeroom for recreation facilities, the collective effectof recreation traffic, and the compoundingeffects of recreation in concert with other site-specific disturbance factors make recreationmanagement an important consideration fordelisting.

74


Volume I/Part II

SUMMARYSUMMARYSUMMARYSUMMARYSUMMARY

Part III of this Recovery Plan outlinesmanagement guidelines to alleviate threats to thespotted owl. These recommendations are basedlargely upon the Team’s evaluation of the biologyof the owl as detailed in Volume II. From theseanalyses, the Team has drawn the followingconclusions.

Mexican spotted owls generally occupyremnants of the landscape that have experiencedminimal human disturbance. We acknowledgethat exceptions to this generalization occur.These remnants include inaccessible canyons,steep slopes, wilderness, and other environmentsnot heavily modified by humans. Persistence ofowls depends partly on these remnant patches,but these environments alone may be insuffi-cient to ensure long-term conservation of theMexican spotted owl. A key point here is thatnot all human activies are detrimental to spottedowls. In fact, if directed appropriately, somehuman activities can be used to the owl’s benefit.Consequently, management must focus on

creating new habitat to replace remnants thatbecome no longer appropriate for the owl.Creation of replacement habitat hinges onunderstanding patterns of natural variation andmodifying human activities that might conflictwith the development of habitat. Natural varia-tion across the landscape results from uniquebiophysical conditions at each location on theland. Further, effects of human activities areequally variable across the landscape. Althoughwe cannot ascribe strict cause-effect relationshipsof natural processes and human activities onMexican spotted owls, we can draw certaininferences about their probable impacts. Theprevious section detailing the conceptual frame-work underlying the recovery measures providesthe rationale for those inferences. Thus, themanagement recommendations (Part III) werebased on two interrelated sets of information: (1)basic knowledge of Mexican spotted biology;and (2) understanding how various naturalprocesses and human activities modify theenvironment to maintain, develop, and alterspotted owl habitat.

75

Recovery

Volume I, Part III


Volume I/Part III

A. DELISTINGA. DELISTINGA. DELISTINGA. DELISTINGA. DELISTING

Removing a species or subspecies fromthreatened status becomes a primary manage-ment objective the moment listing is finalized.Listing a species as threatened affords moreprotection to the species than it would normallyreceive through other laws governing wildlife.Specifically, “threatened” status implies thathuman activities and/or natural disturbancespose greater than normal risks to the entirespecies or subspecies rather than just to individu-als. Greater protection can manifest itself asmore explicit and careful regulation of humanactivities. In some measure, a Recovery Planreconciles human needs and desires with thesurvival needs of the threatened species orsubspecies. If successful, the reconciliationprocess leads to an arrangement to accommodateboth people and the threatened species. Ulti-mately, careful regulation of human activitiescombines with careful management of naturalresources to allow removing the species fromthreatened status, or “delisting.” Just as listing aspecies requires a process of information gather-ing and assessment, delisting requires a similarprocess.

THE DELISTING PROCESSTHE DELISTING PROCESSTHE DELISTING PROCESSTHE DELISTING PROCESSTHE DELISTING PROCESS

Section 4 of the Act governs the listing,delisting, and reclassification of species, thedesignation of critical habitat, and recoveryplanning. Regulations implementing listing,delisting, reclassification, and critical habitatdesignation are codified at 50 CFR 424.

The process of delisting a species or subspe-cies is essentially the same as that of listing: aproposed rule describing the justification for theaction is published in the Federal Register; apublic comment period is opened, includingpublic hearings if requested; and, within oneyear of the proposal, either a final rule delistingthe species or a notice withdrawing the proposedrule is published in the Federal Register.

In considering whether to delist a species,the same five factors considered in the listingprocess (see Part I) are evaluated. While empha-sis may be given to those factors leading to the

species’ listing, all of the factors must be evalu-ated in making a delisting determination.

Section 4(c)(2) of the Act directs the FWS toconduct, at least once every five years, a reviewof all listed species and determine for eachspecies whether it should be removed from thelist, reclassified from endangered to threatened,threatened to endangered, or remain in itscurrent status. This Recovery Plan lists criteriaonly for delisting the Mexican spotted owl. Anydecision to reclassify the subspecies to endan-gered status will be made by the FWS either as aresult of the aforementioned mandatory reviewor at any other time information becomesavailable indicating that reclassification is appro-priate.

Section 4(g) of the Act directs the FWS toimplement a system in cooperation with theStates to monitor effectively for not less than fiveyears the status of a species or subspecies that hasbeen delisted due to recovery. The provisions ofthe Act do not apply to the delisted speciesduring this monitoring period. However, theFWS could relist a species, through the standardlisting process, should monitoring indicate thatthe species will decline without the Act’s protec-tion.

DELISTING CRITERIADELISTING CRITERIADELISTING CRITERIADELISTING CRITERIADELISTING CRITERIA

We recognize that we lack data and authorityto prescribe and implement monitoring strate-gies for Mexico. Thus, our recommendationsbelow apply only to the U.S. range of theMexican spotted owl. We recommend thatMexican authorities develop similar delistingcriteria and monitoring schemes for delisting inMexico.

Five specific criteria must be met before theMexican spotted owl can be delisted in the U.S.The first three criteria, which operate at amultiple-RU level, must be satisfied before thelast two criteria, which operate at the RU level,apply. These are the three overriding criteria:

1. The populations in the Upper GilaMountains, Basin and Range - East, and

76


Volume I/Part III

Basin and Range - West RUs must beshown to be stable or increasing after 10years of monitoring, using a study designwith a power of 90% to detect a 20%decline with a Type I error rate (�) of0.05.

2. Scientifically-valid habitat monitoringprotocols are designed and implementedto verify that (a) gross changes inmacrohabitat quantity across the U.S.range of the Mexican spotted owl arestable or increasing, and (b) microhabitatmodifications and trajectories withintreated stands meet the intent of theRecovery Plan.

3. A long-term, U.S.-rangewide manage-ment plan is in place to ensure appropri-ate management of the subspecies andadequate regulation of human activityover time.

Once these three criteria are satisfactorilyachieved, delisting may occur in any U.S. RUthat meets the final two criteria:

4. Threats to the Mexican spotted owlwithin the RU are sufficiently moderatedand/or regulated.

5. Habitat of a quality to sustain persistentMexican spotted owl populations isstable or increasing within the RU.

These criteria are, by design, redundant anddependent. Meeting one criterion, to somedegree, requires meeting all or some portion ofthe other criteria. Integrating the criteria isunavoidable but nevertheless desirable. Progresson one translates to progress on all.

Monitoring Population TrendsMonitoring Population TrendsMonitoring Population TrendsMonitoring Population TrendsMonitoring Population Trends

For a statistically valid monitoring design, wesuggest the quadrat sampling scheme describedin III.C. The three RUs where populationmonitoring is required for delisting represent thebulk of the known Mexican spotted owl popula-tion in the U.S. No population delisting criteria

are applied to the remaining U.S. RUs becausethey would be difficult to monitor because ofthe small, fragmented nature of the populations.

A premise for our population monitoringapproach is that the existing Mexican spotted owlpopulation in the U.S. is adequate. This premisewill be tested by monitoring population trends.If the results of monitoring indicate that theU.S. population is stable or increasing over thenext 10 to 15 years (assuming 10 years prior todelisting followed by the required 5 years afterdelisting), the Team is willing to accept that thecurrent population will remain viable in theforeseeable future and to assume that the popula-tion is recovered. That is, the Team believes thatif the current population is able to maintainitself, or to increase, then the population hasexhibited evidence that it is of ample size topersist.

Our basis for the parameters included in thedelisting criteria are as follows. The annual rateof change of the population within a RU can beestimated as ��= Nt+1/Nt . A population is stable if��= 1, decreasing if�� < 1, and increasing if� > 1. A 20% reduction over a 10-year periodimplies a value of � = 0.978; i.e., �10 = 0.80.

To conclude that a population is stable, wefail to reject the null hypothesis that � = 1, oralternatively, that the 95% confidence intervalon � includes 1. If we fail to reject this nullhypothesis, we want to ensure that the possiblerate of decline is very small. Thus, we suggest aType II error rate of 0.10, and for a 15-yearperiod, the annual estimate of�� is0.98523 = 0.8(1/15).

For this statistical test of trend, continuedpersistence of the Mexican spotted owl popula-tion means the Type II error rate is more impor-tant than the Type I error rate. That is, a Type Ierror means that we mistakenly conclude thatthe population is declining when it is not.Although costly measures might be taken toreverse our incorrect perception of the trend inthe owl population, the persistence of thepopulation is not threatened. In contrast, a TypeII error means that we conclude the populationis stable or increasing when it is really declining.Thus, persistence of the population could be injeopardy because measures would not be takento correct the decline. Therefore, we emphasize

^ ^ ^

77


Volume I/Part III

that a low Type II error rate of �� = 0.10 (poweris 1 - � = 0.90) must be met to delist the species.

Several biological reasons lead us to select atime span of 10-15 years for monitoring. Themean life span (MLS) of Mexican spotted owlsthat reach adulthood falls within this range.MLS is calculated as 1/(-log(S)), with S repre-senting the adult survival rate. Using S = 0.8889(SE = 0.0269), survival rates calculated from thedemographic study areas, the MLS is about 8.5years. Calculating confidence intervals for MLSyields 16.6 years as an upper age limit. Popula-tion turnover rates provide another biologicalargument for the time span (x) required fordelisting. For example, we can estimate the timethat it takes 90% of the youngest members ofthe adult population to completely turn over, orfor 90% of the existing young adult birds to die.Given the adult S of 0.8889, solving for x in0.8889x = 1 - 0.90 gives x = 19.6 years for 90%of a given cohort of young birds to turnover. A50% turnover would be 5.9 years, which wouldcorrespond to the median life span. For x equals10 years, 70% of the young adult populationwill have turned over.

The time duration for the monitoring andmagnitude of change required to detect a popu-lation decline are related. Thomas (1990) arguedthat the minimum viable population size de-pended on the temporal variation expected in apopulation. Species with much temporal varia-tion in their population size might normallyexhibit a 20% decline over a short period. We donot expect Mexican spotted owl populations todisplay much temporal variation. The mostvariable aspect of their population biology isprobably recruitment, and years of little or norecruitment may occur. However, because of thehigh adult survival rate, the decline in thepopulation during a year of no recruitmentwould still only be 11%. Thus, two consecutiveyears of no recruitment would result in a 21%decline. But the fecundity estimates presented byWhite et al. (1995) suggest that no recruitmentis unlikely. Thus, we conclude that a 20%decline over a 10-year period indicates thepopulation is truly declining and is not the resultof normal temporal variation.

The choice of a Type II error rate of 0.10 issomewhat arbitrary. However, this value interacts

with the choice of a 20% decline over the 10-year period. Figure III.A.1 depicts a hypotheticalcurve for power as a function of the size of theeffect being detected (labeled Detectable EffectSize in the graph). We could specify that a 15%change is detectable with a 67% power, or that a25% change is detectable with a 94% power.These statements are all equivalent in terms ofthe effort required for the monitoring protocol(as shown by the graph). This is because therelationship between the detectable differenceand the power to detect this difference is fixedby the monitoring effort (normally considered asthe sample size of the statistical procedure).Thus, we have suggested that a 90% power todetect a 20% decline over 10 years is a reason-able point to fix the function that relates powerand magnitude of the detectable effect.

In summary, we believe 10 years is a reason-able time span for monitoring because morethan half of the adult population has turnedover. Further, we expect that the populationwould have been subjected to adequate environ-mental variation during this 10-year period.Once the species is delisted, the additional fiveyears of monitoring as required under the Act

Figure III.A.1.Figure III.A.1.Figure III.A.1.Figure III.A.1.Figure III.A.1. Hypothetical curve of thestatistical power to detect a trend in a popula-tion.

78


Volume I/Part III

should provide further assurance that the popu-lation is not declining.

We believe that the delisting criterionproposed here provides positive incentives toland-management organizations to vigorouslypursue the proposed population monitoringsystem. Delisting of the species depends onproviding clearly specified evidence that thepopulation is stable or increasing. The soonerthe responsible land-management organizationsbegin the population monitoring, the sooner theowl can be delisted.

Other Considerations of PopulationOther Considerations of PopulationOther Considerations of PopulationOther Considerations of PopulationOther Considerations of PopulationMonitoringMonitoringMonitoringMonitoringMonitoring

The proposed procedure for populationmonitoring only monitors the territorial popula-tion of owls. Because nonterritorial owls (“float-ers”) do not respond to the usual methods oflocating them (i.e., calling), the only method ofmonitoring nonterritorial birds is via radio-tracking. However, radio-tracking nonterritorialbirds would require large samples of juveniles tobe marked with radios, and these radios replacedon the birds as necessary to maintain the batter-ies as long as the individual remained in thenonterritorial population. Placing radios onspotted owls may alter their behavior and/orsurvival (Paton et al. 1991, Foster et al. 1992),making such an approach of questionable value.Thus, the Team concludes that no viable methodof monitoring nonterritorial birds is available.

An alternative approach to monitoringpopulations was considered, that of using demo-graphic study areas. We decided against usingdemographic studies for three reasons.

First, demographic study areas suffer from adeficiency that is not inherent in the quadratplace procedure described in III.C. Demo-graphic study areas are chosen at the beginningof the monitoring period and must remain inplace to provide appropriate data to meet theirobjectives. Because these study areas must bepermanently delimited, management practiceson them may not reflect those occurring onother lands. In contrast, quadrats can be ran-domly replaced in the sample to ensure thathabitat changes and management practices

adequately reflect those occurring on landsoutside of the quadrats.

Second, the cost of demographic study areasprobably exceeds the cost of our proposedquadrat monitoring approach. The cost ofconducting five demographic studies for theCalifornia spotted owl is roughly equivalent tothe estimated cost of quadrat monitoring (J.Verner, FS, PSW, Fresno, pers. comm). Morethan five demographic study areas would beneeded for a valid population monitoringscheme, thus putting the cost of demographicstudy areas well above the costs of the proposedquadrat sampling procedure.

Finally, the two existing demographic studyareas were not randomly selected from all pos-sible demographic areas (thus not providing adefendable sample). Thus, results from theexisting demography studies apply only to theplace where these studies were done. Further,even if a demographic study approach was used,these existing study areas may not be included inthe random sample needed for a statistically-defensible monitoring scheme.

The problems outlined above with respect tousing demographic studies for monitoring donot negate the usefulness of such studies. Demo-graphic studies were designed to understandaspects of spotted owl population biology andprovide a wealth of information on populations,habitat characteristics, and parameters of owlfitness. They were not designed to monitorlarge-scale population trends.

Monitoring Habitat TrendsMonitoring Habitat TrendsMonitoring Habitat TrendsMonitoring Habitat TrendsMonitoring Habitat Trends

Ganey and Dick (1995) demonstrate thatthe Mexican spotted owl uses specific habitatcharacteristics. These features vary geographi-cally, but within pine-oak and mixed-coniferforests spotted owls use areas that contain largetrees, snags, high log volume, multistoried standstructure, and other specific attributes. Presently,habitat trends for the Mexican spotted owl areunknown, and the subject of conflicting specula-tion (II.D). Clearly, adequate habitat of suffi-cient quality must exist into the future to ensurepopulation viability. Consequently, habitatmonitoring is an essential part of the recovery

79


Volume I/Part III

process and the bird should not be delisted untilmonitoring can ensure unequivocally thatsufficient habitat exists to support a viablepopulation of spotted owls.

Habitat monitoring should address twoaspects: persistence of forest types that owlsprefer (macrohabitat) and specific habitat at-tributes within those types (microhabitat). Thisroughly corresponds to the coarse and fine filtersdescribed in II.D.

The first task, then, is to quantify large-scalechanges in macrohabitat across the range of thebird. Given existing high fire risks and thecurrent state of southwestern forests, we expectthat net macrohabitat change over the next 10years will be negative. Although some areas willdevelop into habitat as the result of successionand past management activities, the Teamassumes that (1) most acreage currently on atrajectory to become habitat will not do soduring the life of the plan, and (2) some habitatwill be lost to fire during the next 10-15 years(II.D). Thus, the Team anticipates a slightdecline in the total acreage of spotted owlmacrohabitat during the short term. Unfortu-nately, we cannot specify a priori a thresholdlevel of habitat loss that the spotted owl popula-tion can safely endure. That relationship canonly be evaluated by combining the results ofpopulation and habitat monitoring.

The second task for habitat monitoring is toevaluate whether or not management prescrip-tions were implemented effectively, and whethertreated stands will remain or become owl habitatin the near future. Prescriptions pertain prima-rily to the use of prescribed fire and varioussilvicultural tools. Monitoring to meet thisobjective would entail pre-treatment sampling tomeasure existing habitat attributes, and post-treatment sampling to verify that the prescrip-tion met the intent of the treatment. Attributesto be sampled include both those typicallymeasured during stand examinations and alsoadditional variables not typically measured butthat are strong correlates of owl presence (e.g.,canopy cover, log volume). The general designfor measuring owl habitat can be modified tomonitor other ecosystem attributes as well. Thatis, additional variables can be measured besidesthose needed for spotted owls as required for

other ecosystem management objectives. Thesetypes of coordinated efforts will be crucial tomeeting the monitoring needs inherent to bothecosystem and adaptive management.

Long-term Management PlanLong-term Management PlanLong-term Management PlanLong-term Management PlanLong-term Management Plan

As described in Part I, this Recovery Plan isintended to guide management for Mexicanspotted owls over the next 10-15 years. If imple-mented as recommended, significant researchwill be conducted during this period and impor-tant new information will become available,specifically data on owl biology, populationstructure, and effects of certain managementpractices on owl habitat. Further, the guidelinesthat we propose for managing restricted areas(III.B) will provide a foundation upon whichlong-term management might be based. Evalua-tion of this approach and the informationprovided through research and monitoring willbe integral to developing and refining a long-term plan for managing the Mexican spottedowl. Such a plan will be required before delistingcan be considered.

Delisting at the RU LevelDelisting at the RU LevelDelisting at the RU LevelDelisting at the RU LevelDelisting at the RU Level

The Team recommends that once the popu-lation and habitat are shown to be stable orincreasing, delisting should be considered at theRU level. When delisting is considered, atten-tion must focus on the resolution of knownthreats and the identification of emerging threatsthat could potentially compromise populationviability. Similarly, spotted owl habitat must bemonitored in each RU to determine trends.Monitoring will reveal habitat decline, improve-ment, or relative stability. The reasoning is that ifthe threats are removed or adequately regulatedand if habitat trends are stable or showingimprovement, protection under the Act will nolonger be necessary. Conversely, a habitat declineor a lack of adequate regulatory mechanisms(other than provided by the Act) would warrantcontinued protection under the Act.

The reasoning behind monitoring popula-tion levels within three RUs and habitat in allRUs is as follows. A viable core population will

80


Volume I/Part III

exist in the three RUs if the population is shownto be stable or increasing. Therefore, if habitatrangewide is also stable or increasing, the corepopulation is provided the opportunity ofexpanding its area and greatly increasing thepersistence probability of the subspecies. Asdiscussed by Keitt et al. (1995), some keyunoccupied habitat patches are potentiallysignificant in the expansion of the core popula-tion.

Moderating and Regulating ThreatsModerating and Regulating ThreatsModerating and Regulating ThreatsModerating and Regulating ThreatsModerating and Regulating Threats

Threats to be moderated include those thatneed site-specific treatment to alleviate them.The primary threat throughout the forested U.S.range of the Mexican spotted owl is the threat ofwidescale, stand-replacing fire. For threats to beconsidered as moderated, reasonable progressmust have been made in removing the threatsand adequate assurance, in the form of the long-term management plan described above, mustexist that those programs will continue as neces-sary.

Threats to be regulated include those result-ing from agency management programs or otheranthropogenic activities that are either ongoing

or reasonably certain to occur. A partial listing ofthreats besides fire include:

1. Timber or fuelwood harvest that eitherdirectly affects habitat within a territoryor indirectly affects the owl by collateralactivity adjoining owl territories;

2. urban and rural land development;

3. livestock and wildlife grazing;

4. recreation involving both consumptiveand nonconsumptive activities.

Habitat Trends Within Recovery UnitsHabitat Trends Within Recovery UnitsHabitat Trends Within Recovery UnitsHabitat Trends Within Recovery UnitsHabitat Trends Within Recovery Units

For the spotted owl to be delisted within anyRU, the following conditions must be met. First,threats to the continued loss of habitat and keyhabitat components must be moderated andregulated as detailed in the previous section.Second, habitat trends must be monitored toassess gross changes in habitat quantity withineach RU. Third, effects of modifying activitieswithin existing and potential spotted owl habitatmust be monitored to ensure that existinghabitat is maintained and potential habitat isprogressing towards becoming replacementhabitat.

81


Volume I/Part III

The Recovery Plan recommendations are acombination of (1) protection of both occupiedhabitats and unoccupied areas approachingcharacteristics of nesting habitat, and (2) imple-mentation of ecosystem management withinunoccupied but potential habitat. The goal is toprotect conditions and structures used by spot-ted owls where they exist and to set other standson a trajectory to grow into replacement nesthabitat or to provide conditions for foraging anddispersal. By necessity this Plan is a hybridapproach because the status of the Mexicanspotted owl as a threatened species requires somelevel of protection until the subspecies isdelisted. These constraints modify ways andopportunities to manage ecosystems withinlandscapes where owls occur or might occur inthe future. We are applying ecosystem manage-ment in two slightly different ways. Withinunoccupied mixed-conifer and pine-oak foreston <40% slope, we provide both general (coarsefilter) and specific (fine filter) guidelines toprovide a sustainable quantity of replacementnest habitat across the landscape. Within otherunoccupied forest and woodland types (e.g.,ponderosa pine, spruce-fir, aspen, and pinyon-juniper), general guidance is provided for man-aging the landscape to meet multiple ecosystemmanagement objectives including spotted owlforaging and dispersal habitat.

Management priority should focus onactions to alleviate threats to Mexican spottedowls; thereafter, or in coordination with alleviat-ing threats, other management priorities (e.g.,creating replacement owl habitat) should bepursued. Two primary threats that managersshould focus on are catastrophic wildfire and thewidespread use of even-aged silviculture.

Heavy accumulations of ground and ladderfuels have rendered many Southwestern forestsvulnerable to stand-replacing fires. Such firesrepresent real and immediate threats to theexistence of spotted owl habitat. The manage-ment guidelines that follow are intended toprovide land managers with flexibility to reducethese fuel levels and abate fire risks. Fire manage-ment should be given the highest priority.

Even-aged silviculture within potential owlhabitat is regarded as a threat because it tends tosimplify stand structure and move stands awayfrom containing structures used by owls. Werecognize, however, that such regeneration cutsmay provide useful tools in certain circumstancesto manage for spotted owls and other ecosystemobjectives. Any use of even-aged managementshould be done sparingly and only after carefuldeliberation to ensure that it represents the bestapproach to meet management objectives.

Under proposed delisting criteria the owlcould be delisted within 10 years, rendering theprotection measures in this Recovery Planobsolete. At that time, we anticipate havingsufficient knowledge to design a strategy forlong-term conservation of the Mexican spottedowl. Many of the ecosystem managementguidelines provided in this Plan will provide afoundation for development of the long-termstrategy. In formulating our recommendations,we assume that population and habitat statuswill be monitored in conjunction with imple-mentation of these management guidelines. ThisRecovery Plan is analogous to a three-leggedstool (Figure III.B.1); therefore, the manage-ment guidelines are not meant to stand alone.Monitoring provides objective criteria to assessthe efficacies of the management guidelines.Without both habitat and population monitor-ing, the status of the owl cannot be assessed andit should not be delisted. We further assume thatexisting management constraints on vegetativemanipulations (such as size of openings andmaintenance of hiding and thermal cover forother species) will remain in place. This assump-tion is especially critical for vegetation types--ponderosa pine, pinyon-juniper, aspen, andspruce-fir--for which we provide no specificmanagement recommendations.

ASSUMPTIONS AND GUIDINGASSUMPTIONS AND GUIDINGASSUMPTIONS AND GUIDINGASSUMPTIONS AND GUIDINGASSUMPTIONS AND GUIDINGPRINCIPLESPRINCIPLESPRINCIPLESPRINCIPLESPRINCIPLES

The recommendations proposed here arebased on several key assumptions about habitatrequirements of the Mexican spotted owl, and a

B. GENERAL APPROACHB. GENERAL APPROACHB. GENERAL APPROACHB. GENERAL APPROACHB. GENERAL APPROACH

82


Volume I/Part III

number of guiding principles. These are enumer-ated below.

AssumptionsAssumptionsAssumptionsAssumptionsAssumptions

1. Spotted owl distribution is limitedprimarily by the availability of habitattypes used for nesting and/or roosting.

2. Habitats used for nesting/roosting alsoprovide adequate conditions for foragingand dispersal activities. Thus, providingnesting/roosting habitat partially meetsother survival requirements as well. Inturn, some stand structures not used fornesting/roosting may provide adequateconditions for other activities such asforaging and dispersal. These includesome stands in younger seral stages thantypical nesting/roosting habitat.

3. Nesting/roosting habitat in forest envi-ronments is typified by certain structuralfeatures, including large trees and lateseral characteristics, which are commonin, but not restricted to, old-growthforests.

4. Forested nesting/roosting habitat istypically found in mixed-conifer, pine-oak, and riparian forests. Other habitattypes are used primarily for foraging,dispersal, or wintering. Thus, the distri-bution of nesting/roosting habitat isnaturally discontinuous. Further, thepotential distribution of such habitat isquite limited in some areas.

5. The presence of shade-intolerant speciesin many spotted owl nest/roost standssuggests that these areas are dynamic andhave developed over time, often frommore open stands. Disturbance eventsleading to forest canopy gaps may beimportant in maintaining shade-intoler-ant species, particularly in mixed-coniferstands.

6. Existing stand structures used by Mexi-can spotted owls for nesting/roostinggenerally have not been a target ofplanned silvicultural treatments. Wheresuch conditions exist in managed stands,they are more than likely an unplanned

Figure III.B.1.Figure III.B.1.Figure III.B.1.Figure III.B.1.Figure III.B.1. Conceptualization of the Recovery Plan and needs for delisting of the Mexicanspotted owl depicting the interdependency of population monitoring, habitat monitoring, and man-agement recommendations.

83


Volume I/Part III

rather than purposeful result. Theexistence of Mexican spotted owl nest-ing/roosting habitat usually results fromthe lack of recent alteration of foreststructures in certain landscapes.

Guiding PrinciplesGuiding PrinciplesGuiding PrinciplesGuiding PrinciplesGuiding Principles

1. Silvicultural applications must be evalu-ated over time by rigorous monitoringprocedures to assess their effectiveness inmanaging or creating owl habitat.

2. Obtaining large trees is a function ofboth time and site productivity. Simi-larly, many late seral characteristicstypical of owl habitat, such as broken-topped trees, snags, large downed logsand the sharing of growing space amongmultiple shade-tolerant and intolerantspecies, are attained primarily throughtime.

3. Although this Recovery Plan represents ashort-term strategy, management actionsrecommended herein will have long-term consequences. Therefore, careshould be taken to preserve futureoptions while evaluating the effectivenessof proposed treatments.

GENERAL RECOMMENDATIONSGENERAL RECOMMENDATIONSGENERAL RECOMMENDATIONSGENERAL RECOMMENDATIONSGENERAL RECOMMENDATIONS

General management recommendations foruse throughout the range of the Mexican spottedowl are given here. These general recommenda-tions apply primarily to forested areas, andaspects of the recommendations are more appli-cable to some locations than others. Because theseverity of potential threats varies among RUs,the general guidelines should be prioritized andapplied accordingly. Specific managementpriorities are emphasized in sections on indi-vidual RUs, as warranted by the differencesamong RUs.

Three levels of habitat management aregiven in this Recovery Plan: protected areas,restricted areas, and other forest and woodlandtypes (Figure III.B.2). Protected areas receive thehighest level of protection under this plan, other

forest and woodland types the lowest. Guidelinesproposed in this Recovery Plan take precedenceover other agency management guidelines inprotected areas. Guidelines for restricted areasare less specific and operate in conjunction withecosystem management and existing manage-ment guidelines. We propose no owl-specificguidelines for lands not included in protectedand restricted areas; these areas will continue tobe managed under existing guidelines, assumingthat the emphasis is towards ecosystem manage-ment.

One guideline that applies to all areas withany potential for owl use is to inventory forspotted owls before implementing any manage-ment action that will alter habitat structure. Ifresults of past inventory efforts can demonstrateunequivocally that no spotted owls have beendetected within a given area or habitat and thatthe probability of detecting a bird there is small,then future surveys may not be needed. Undersuch circumstances, concurrence must begranted by the Recovery Team through theappropriate RU working team.

Protected AreasProtected AreasProtected AreasProtected AreasProtected Areas

Protect all Mexican spotted owl sites knownfrom 1989 through the life of the Recovery Plan(Protected Activity Centers), all areas in mixed-conifer and pine-oak types (defined in II.C) withslope >40% where timber harvest has not oc-curred in the past 20 years, and all legally andadministratively reserved lands. Specific guide-lines and the rationale for these guidelines areprovided below.

Protected Activity Center (PAC)Protected Activity Center (PAC)Protected Activity Center (PAC)Protected Activity Center (PAC)Protected Activity Center (PAC)

Guidelines.—Guidelines.—Guidelines.—Guidelines.—Guidelines.—Eight specific guidelines pertain tothe designation and implementation of PACs.These guidelines supersede steep slope guide-lines; that is, steep slopes occurring within PACsshould be managed under PAC guidelines.

1. Establish PACs at all Mexican spottedowl sites known from 1989 through thelife of the Recovery Plan, including newsites located during surveys. PACs shouldalso be established at any historical siteswithin the Colorado Plateau, SouthernRocky Mountains - Colorado, and

84


Volume I/Part III

FFFFFigurigurigurigurigure III.B.2.e III.B.2.e III.B.2.e III.B.2.e III.B.2. Generalization of protection strategies by forest/vegetation type. Proportions are notto scale.

85


Volume I/Part III

Southern Rocky Mountains - NewMexico RUs. Identify the activity centerwithin each PAC. “Activity center” isdefined as the nest site, a roost grovecommonly used during the breedingseason in absence of a verified nest site,or the best roosting/nesting habitat ifboth nesting and roosting informationare lacking. Site identification should bebased on the best judgement of a biolo-gist familiar with the area. Delineate anarea no less than 243 ha (600 ac) aroundthis activity center using boundaries ofknown habitat polygons and/or topo-graphic boundaries, such as ridgelines, asappropriate (Figure III.B.3). The bound-ary should enclose the best possible owlhabitat, configured into as compact aunit as possible, with the nest or activitycenter located near the center. Thisshould include as much roost/nesthabitat as is reasonable, supplemented byforaging habitat where appropriate. Forexample, in a canyon containing mixed-conifer on north-facing slopes andponderosa pine on south-facing slopes, itmay be more desirable to include someof the south-facing slopes as foraginghabitat than to attempt to include 243ha (600 ac) of north-slope habitat. Inmany canyon situations, oval PACs maymake more sense than, for example,circular PACs; but oval PACs could stillinclude opposing canyon slopes asdescribed above. All PACs should beretained for the life of this RecoveryPlan, even if spotted owls are not locatedthere in subsequent years. A potentialexception to this rule is described in #8below. Feedback on PAC delineationshould be provided to managers throughRU working groups (see Part IV). PACboundaries may not overlap.

2. No harvest of trees >22.4 cm (9 in) dbhis allowed in PACs. Harvest of any treesis only permitted as it pertains to 5below.

3. Fuelwood harvest within PACs should bemanaged in such a way as to minimize

effects on the owl, its prey, and theirhabitats. The most effective managementto meet these objectives may be toprohibit such harvest. However, werecognize that it may be virtually impos-sible to enforce such a prohibition andrestrict access to all PACs for fuelwoodharvest. When fuelwood harvest in PACsis unavoidable, we advocate the use ofvarious forms of management that canregulate access to PACs and to the typesof fuels harvested. Potential forms offuelwood management include roadclosures, prohibiting harvest of impor-tant tree species such as oaks, prohibitingharvest of key habitat components suchas snags and large downed logs (>30 cm[12 inch] midpoint diameter), andencouraging the harvest of small diam-eter conifers in accord with 5c below.Prohibiting fuelwood harvest of keyhabitat components such as oaks, snags,and large logs should be applied bothinside and outside of PACs to ensure thatthese special components remain on thelandscape.

4. Road or trail building in PACs shouldgenerally be avoided but may be allowedon a case-specific basis if pressing man-agement reasons can be demonstrated.

5. Implement a program consisting ofappropriate treatments to abate fire risk.The intent of this program is to assessthe combined effects of thinning and fireon spotted owls and their habitat. Theprogram should be structured as follows:

a) Select up to 10% of the PACs withineach RU that exhibit high fire riskconditions. Nest sites must be knownwithin these PACs. Ideally, a pairedsample of PACs should be selected toserve as control areas.

b) Within each selected PAC,designate 40 ha (100 acres)centered around the nest site.This nest area should includehabitat that resembles the structural

86


Volume I/Part III

FFFFFigurigurigurigurigure III.B.3. e III.B.3. e III.B.3. e III.B.3. e III.B.3. Examples of protected activity center (PAC) boundaries from the Lincoln NationalForest. Prepared by D. Salas and L. Cole, Lincoln NF.

87


Volume I/Part III

and floristic characteristics of thenest site. These 40 ha (100 acres) willbe deferred from the treatmentsdescribed below.

c) Within the remaining 203 ha (500acres), combinations of thinningtrees <22.4 cm (9 inches) dbh,treatment of fuels, and prescribed firecan be used to reduce fire hazard andto improve habitat conditions forowl prey. Habitat components thatshould be retained or enhancedinclude large logs (>30 cm [12inches] midpoint diameter), grassesand forbs, and shrubs. These habitatcomponents are strong correlates ofthe presence of many key preyspecies of the owl. Emphasis of thespatial configuration of treatmentsshould be to mimic natural mosaicpatterns.

d) Treatments can occur only duringthe nonbreeding season (1 Septem-ber-28 February) to minimize anypotential deleterious effects on theowl during the breeding season.

e) Following treatments to 10% of thePACs, effects on the owl, preyspecies, and their habitats should beassessed. If such effects are non-negative, an additional sample ofPACs may be treated. If negativeeffects are detected, these effectsmust be carefully evaluated. If theycan be ameliorated by modifyingtreatments, those modificationsshould occur prior to treatment ofadditional PACs. If not, no addi-tional treatments should be permit-ted.

6. Within the remaining PACs, lightburning of ground fuels may be allowedwithin the 500 acres surrounding the100-acre PAC centers (5b above), follow-ing careful review by biologists and fuelsmanagement specialists on a case-specificbasis. Burns should be designed and

implemented to meet the objectivesnoted in 5c above. Burns should be doneonly during the nonbreeding season(1 September-28 February).

7. Within PACS treated to reduce fire risk,either by the use of prescribed fire aloneor in conjunction with mechanicalremoval of stems and ground fuels, pre-and post-treatment assessments (i.e.,monitoring) of habitat conditions andowl occupancy must be done. Specifichabitat characteristics that should bemonitored include fuel levels, canopycover, snag basal area, volume of largelogs (>30 cm [12 inch] midpoint diam-eter), and live tree basal area.

8. If a stand-replacing fire occurs within aPAC, timber salvage plans must beevaluated on a case-specific basis. In allcases, the PAC and a buffer extending400 m from the PAC boundary must besurveyed for owls following the fire. Aminimum of four visits, spaced at leastone week apart, must be conductedbefore non-occupancy can be inferred. Ifthe PAC is still occupied by owls or ifowls are nearby (i.e., within 400 m of thePAC boundary), then the extent andseverity of the fire should be assessed andreconfiguration of the PAC boundariesmight be considered through section 7consultation. If no owls are detected,then section 7 consultation should beused to evaluate the proposed salvageplans. If informal consultation cannotresolve the issue within 30 days, theappropriate RU working team should bebrought into the negotiations.

Salvage logging within PACs shouldbe the exception rather than the rule.The Recovery Team advocates thegeneral philosophy of Beschta et al.(1995) for the use of salvage logging. Inparticular: (1) no management activitiesshould be undertaken that do notprotect soil integrity; (2) actions shouldnot be done that impede natural recoveryof disturbed systems; and (3) salvage

88


Volume I/Part III

activities should maintain and enhancenative species and natural recoveryprocesses. Further, any salvage shouldleave residual snags and logs at levels andsize distributions that emulate thosefollowing pre-settlement, stand-replacingfires. Scientific information applicable tolocal conditions should be the basis fordetermining those levels.

RationaleRationaleRationaleRationaleRationale.....—The primary objective to beachieved by these guidelines is to protect the bestavailable habitat for the Mexican spotted owl,while maintaining sufficient flexibility for landmanagers to abate high fire risks and to improvehabitat conditions for the owl and its prey. Weassume that the best available owl habitat is thatwhich is currently occupied by owls, or thatoccupied by owls in the recent past (since 1989).The median size of the adaptive kernel contourenclosing 75% of the foraging locations for 14pairs of radio-marked owls was 241 ha (595 ac).Therefore, a 243 ha (600 ac) PAC shouldprovide a reasonable amount of protectedhabitat and should provide for the nest site,several roost sites, and the most proximal andhighly used foraging areas. We assume thatexisting management guidelines and thosediscussed below for areas outside of PACs willensure the existence of additional habitat appro-priate for foraging.

The intent of these guidelines is not topreserve these PACs forever, but rather toprotect them until it can be demonstrated thatwe can create replacement habitat through activemanagement. We describe below in the sectioncovering restricted areas the approach for manag-ing to create replacement habitat. Once landmanagers demonstrate that they can createreplacement habitat, and when monitoringindicates that populations and habitats are stableor increasing, PACs could be abolished inconjunction with delisting the owl.

The Team recognizes that protection statuscarries some risk with respect to probabilities ofcatastrophic fire. The reason for the proposedmanagement within PACs is to encourage aproactive approach to reduce fuel risks andsimultaneously enhance prey habitat. If theseobjectives are achieved, existing owl habitat will

be maintained and in some cases enhanced,while identified risks of catastrophic fire will belessened.

Salvage logging in PACs should be allowedonly if sound ecological justification is providedand if the proposed actions meet the intent ofthis Recovery Plan, specifically to protect exist-ing habitat and accelerate the development ofreplacement habitat. Fires within PACs are notnecessarily bad. In many cases, patchy fires willresult in habitat heterogeneity and may benefitthe owl and its prey. In such cases, adjustmentsto PAC boundaries are probably unnecessary andsalvage should not be done. Salvage should beconsidered in PACs only when the fire is exten-sive in size and results in the mortality of asubstantial proportion of trees.

Steep Slopes (outside of PACs)Steep Slopes (outside of PACs)Steep Slopes (outside of PACs)Steep Slopes (outside of PACs)Steep Slopes (outside of PACs)

GuidelinesGuidelinesGuidelinesGuidelinesGuidelines.....—Within mixed-conifer and pine-oak types, allow no harvest of trees >22.4 cm (9inches) on any slopes >40% where timberharvest has not occurred in the past 20 years.(Mixed-conifer and pine-oak types found onsteep slopes that have been treated within thepast 20 years are managed under restricted areaguidelines below). These guidelines also apply tothe bottoms of steep canyons. Thinning of trees<22.4 cm (9 inches) dbh, treatment of fuels, andfire are allowed, as discussed in 5c above. Noseasonal restrictions apply, however. Prescribednatural fire is also permitted as is the creation offire breaks on a case-specific basis.

On steep slopes treated to reduce fire risk,either by the use of prescribed fire alone or inconjunction with removal of stems and groundfuels, pre- and post-treatment monitoring ofhabitat conditions should be done. Specifichabitat characteristics to be measured includefuel levels, snag basal area, volume of large logs(>30 cm midpoint diameter), and live tree basalarea.

RRRRRationaleationaleationaleationaleationale.....—The objective of prohibiting timberharvest but allowing treatment of fuels andburning is to retain additional habitat withexisting conditions similar to owl nesting/roosting habitat while reducing fire risks. These

89


Volume I/Part III

conditions appear to be found commonly inmature/old-growth stands, and such stands arenow found most commonly on steep slopesbecause past management practices have largelyoccurred on slopes <40%. We have restrictedthese guidelines only to the mixed-conifer andpine-oak types because existing informationindicates that the owl favors these types fornesting and roosting (Ganey and Dick 1995).

These guidelines depart somewhat fromrecent management for steep slopes on south-western FS lands. R. Fletcher (FS SouthwesternRegion, Albuquerque, NM, comment submittedon draft Recovery Plan) noted that only about1,215 ha (3,000 acres) of steep slopes have beentreated since 1987. Our guidelines emphasizethat greater acreage should be treated throughthinning and fire if threats of catastrophic fireare to be decreased on steep slopes. We haveexcepted steep slopes that been harvested in therecent past because many of these areas may notcurrently exhibit the forest structure spotted owlsuse for nesting or roosting. Guidelines forrestricted areas apply to these lands.

Reserved LandsReserved LandsReserved LandsReserved LandsReserved Lands

GuidelinesGuidelinesGuidelinesGuidelinesGuidelines.....—Encourage the use of prescribednatural fire where appropriate in Wilderness,Research Natural Areas, and other reservedlands.

RationaleRationaleRationaleRationaleRationale.....—Prescribed natural fire may bebeneficial to owl habitat in several ways. First, itcan aid in reducing fuel loads and risk of cata-strophic wildfire resulting in loss of habitat overlarge areas. Second, it can create a diverse land-scape with considerable horizontal heterogene-ity. This seems to be relatively characteristic ofmany areas occupied by spotted owls and alsoprovides for a diverse prey base. Third, it cancreate conditions that maintain shade-intolerantspecies such as ponderosa pine or Gambel oak inthe landscape. Prescribed fires should be usedcarefully in spotted owl habitat, however; andthe results should be monitored to evaluate theeffects on habitat components suspected to beimportant to the spotted owl and its prey, suchas large snags and logs.

Restricted AreasRestricted AreasRestricted AreasRestricted AreasRestricted Areas

Not all lands can or should receive equalprotection. We provided guidelines above toprotect all occupied nesting and roosting habitat,as well as unoccupied steep slopes and reservedlands. Potential exists, however, for the owl touse other, unoccupied areas. Thus, we provideadditional guidelines to maintain and developpotential nesting and roosting habitat now andinto the future. The guidelines that we presentare stratified by broad vegetative cover types:mixed-conifer forest, pine-oak forest, and ripar-ian areas. Definitions for pine-oak and mixed-conifer forests as applicable to these recoverymeasures are given in II.C.

For the most part, these guidelines apply toplanning areas. Planning areas can be diversityunits, sale planning areas, or ecological areas, allplaces where management activities are consid-ered and evaluated. The intent is to spreadactivities over the landscape rather than concen-trating them in particular areas. Managementwithin restricted mixed-conifer and pine-oakforests is derived from concepts of ecosystemmanagement. Ecosystem management, however,requires ecological assessments at hierarchies ofspatial scales (Kaufmann et al. 1994:6). Thus,although management is applied to planningareas, it is crucial that the impacts are assessed atlarger spatial scales (e.g., landscape, subregional,and regional scales).

The underlying objective of the followingguidelines is to manage the landscape to main-tain and create replacement owl habitat whereappropriate, while providing a diversity of standconditions and stand sizes across the landscape.As noted previously, we assume that the primarylimiting factor for Mexican spotted owls is theamount of nesting habitat. A logical conclusionfrom this premise is that the landscape should bemanaged to sustain owl nesting habitat welldistributed spatially. Because various naturalprocesses lead to the development, maturation,and senescence of such stands through time,management should allocate stands in such away as to mimic the natural landscape. We alsoassume that providing a continuous supply ofnesting and roosting habitat requires that re-maining stands be in various stages of ecological

90

succession. The landscape mosaic resulting fromsuch an allocation should ensure adequatenesting, roosting, and foraging habitat for theowl, and habitats for its variety of prey.

Existing ConditionsExisting ConditionsExisting ConditionsExisting ConditionsExisting Conditions

Ideally, assessments of existing conditionsshould follow the spatial hierarchy presented byKaufmann et al. (1994:6). At the very least,existing distributions of seral stages should beassessed at the planning level, landscape, subre-gional, and regional scales (sensu Kaufmann et al.1994). We recognize that information may beinadequate to conduct assessments at largerspatial scales, but this constraint should beameliorated as resource agencies continue toacquire appropriate data. Existing vegetativeconditions within mature-old stands must alsobe assessed to determine the treatment potentialswithin those stands. However, given the highfrequency of recent stand-altering disturbances,many areas are likely deficient in mature to old-growth forests. Thus, any treatments to thesestands should be applied judiciously, if at all.

Reference ConditionsReference ConditionsReference ConditionsReference ConditionsReference Conditions

Nesting and roosting target/thresholdNesting and roosting target/thresholdNesting and roosting target/thresholdNesting and roosting target/thresholdNesting and roosting target/thresholdconditionsconditionsconditionsconditionsconditions.....—Forested stands used by spottedowls have certain structural features in common.These conditions do not, nor can they, occureverywhere. For example, many south-facingslopes may never attain this type of forest struc-ture. It is impossible for us to imagine everypossible management scenario, and this limitsour ability to formulate specific guidelines thatwould be appropriate to all situations. Ourintent here is to protect appropriate nestinghabitat structure where it exists and manageother stands to develop the needed structure.

Although our knowledge of spotted owlhabitat is incomplete, nesting/roosting standsexhibit certain identifiable features, includinghigh tree basal area, large trees, multi-storiedcanopy, high canopy cover, and decadence in theform of downed logs and snags (Ganey and Dick1995). Further, these stands often contain aconsiderable hardwood component generallyprovided by Gambel oak in ponderosa pine-Gambel oak forests and by various species (e.g.,

oaks, maples, box elder, aspen) in mixed-coniferforests.

We used tree basal area, large tree (>45.7 cm[18 in] dbh) density, and tree size-class distribu-tion as the variables to define target/thresholdconditions (Table III.B.1). Other variables suchas snags and downed logs are important as well.We assume that if the basal area and tree densitylevels given in Table III.B.1 exist, adequateamounts of snags and downed logs (and otherhabitat elements) should be present.

The values provided in Table III.B.1 repre-sent targets in that they define the desiredconditions to be achieved with time and man-agement. They also represent threshold condi-tions in that they define minimal levels thatmust be maintained. That is, activities can occurwithin stands that exceed these conditions, butthe outcome of such activities cannot lower thestands below the threshold levels unless large-scale ecosystem assessments demonstrate thatsuch conditions occur in a surplus across thelandscape (see below). Note that all values mustbe met simultaneously for a stand to meet target/threshold conditions.

We used two primary types of informationto define target/threshold conditions. First, weused quantitative descriptions of site- and stand-level habitat conditions. Second, we estimatedthe proportion of the landscape that couldsustain those conditions through time. A similarapproach was provided for managing northerngoshawk habitat in the southwest (Reynolds etal. 1992). Thus, our approach is not withoutprecedence.

Despite repeated attempts by the RecoveryTeam to obtain data from land-managementagencies and researchers, only limited data wereavailable for our analyses. We used nest-site datacollected by SWCA (1992) which included plotmeasurements centered (1) at each nest location,(2) a random location within each nest stand,and (3) a random location within a stand adja-cent to the nest stand (see Ganey and Dick[1995] for more detailed information). We alsoused FS stand inventory data provided by theCoconino, Apache-Sitgreaves, and LincolnNational Forests. These data consisted of stand-level data stratified by nest, core, and territorystands. Core and territory delineations were

91Volume I/Part III


Table III.B.1.Table III.B.1.Table III.B.1.Table III.B.1.Table III.B.1. Target/threshold conditions for mixed-conifer and pine-oak forests within restricted areas. Forest types are defined in II.C.

92


Volume I/Part III

based on FS management guidelines for theMexican spotted owl provided by ID No. 2 (seePart I). We had no way to assess the accuracy ofthe data. Different methods were used to collectthe SWCA data from those used to collect theFS data, thus direct comparisons are tenuous.Neither data set was collected specifically toaddress our objectives; thus, the data were lessthan optimal for our purposes. Further, numer-ous people collected data and we cannot assessinter-observer variation (Block et al. 1987). Asobvious as these points may be, they greatlylimited the inferences that we could make basedon our analyses. Thus, we relied on our bestprofessional judgement to evaluate the analysesand formulate our recommendations.

We explored several analyses to derive target/threshold conditions including empiricalunivariate and multivariate analyses, and model-ing. Most analyses converged on a set a valuesthat were validated by existing data on spottedowl nesting habitat.

We used the following approach. First, weused the SWCA (1992) data to characterize neststands on the basis of tree basal area, density oflarge trees, and the distribution of stand densityby size classes. Next, we used the available FSstand data to identify the percentage of thecontemporary landscape that simultaneouslymeets these values. Third, we modeled foreststands under various post-disturbance/standinitiation conditions using the Forest VegetationSimulator (Wykoff et al. 1982, Dixon 1991,Edminster et al. 1991). This model allowed us topredict the amount of time that a stand wouldbe in each successional stage, including theamount of time the stand would retain or exceedthe characteristics required for nesting androosting. Knowledge of how long a stand meetsor exceeds target conditions was used to estimatethe proportion of the landscape that should meetor exceed these stand conditions at a given time.

Analyses were conducted separately formixed-conifer and pine-oak forests. We alsoprovide two sets of values for mixed-coniferforest that reflect different target/thresholdvalues which are applied to different proportionsof the landscape (Table III.B.1). We reiteratethat all target/threshold values must be metsimultaneously. For example, within mixed-

conifer forests in all RUs except Basin and Range- East, 25% of the landscape should consist ofstands that have >32 m2/ha (150 ft2/acre) of treebasal area, and include >49 trees/ha (20 trees/acre) that are >45.7 cm (18 inches) dbh. Man-agement should strive for an even distribution ofstand density across all sizes classes with no lessthan 10% of the distribution of stand density ineach of the upper three size classes: 30.5-45.7 cm(12-18 inches), 45.7-61.0 cm (18-24 inches),and >61.0 cm (24 inches). Also, 10% of the totallandscape (a subset contained within the 25%discussed above), should have >39 m2/ha (170ft2/acre) of basal area in addition to the largetrees and distribution of trees by size class.Target/threshold conditions for mixed-coniferforests in the Basin and Range - East RU differslightly in that landscape percentages are 20%and 10%. The areal percentage for Basin andRange - East RU is lower (20% compared to25%) because of the high density of owls in theSacramento Mountains which effectively places alarge proportion of the landscape in protectedstatus. Target/threshold conditions apply to only10% of the pine-oak forest (Table III.B.1).Target/threshold conditions for pine-oak forestsalso require that >4.6 m2/ha (20 ft2/acre) of oakbasal area be present, and that all oaks >13 cm [5inches] dbh be retained (Table III.B.1).

Coarse FilterCoarse FilterCoarse FilterCoarse FilterCoarse Filter

We recognize that most project planningoccurs at limited spatial scales such as 4,050 ha(10,000 acre) blocks. This limited spatial scaleprecludes ecological assessments at larger scales.Because of this limitation, the areal percentagesprovided in Table III.B.1 should be regarded asminimum levels for a given planning area. If adeficit occurs within the planning area, addi-tional stands should be identified that (1) havethe site potential to reach target conditions and(2) whose current conditions most closelyapproach those conditions. Those stands shouldthen be managed to achieve target conditions asrapidly as possible. However, if the proportion ofthe planning area that meets target conditions isgreater than the percentages in Table III.B.1,none of those stands can be lowered belowthreshold conditions until ecosystem assessmentsat larger spatial scales (landscape, subregion,

93


Volume I/Part III

region) demonstrate that target conditionsexceed the required areal percentages (TableIII.B.1) at these larger scales. This does notpreclude use of treatments to reduce fire risks orlessen insect or disease problems nor does itpreclude management to meet other ecosystemobjectives as long as stand-level conditionsremain at or above the threshold values given inTable III.B.1.

Fine FilterFine FilterFine FilterFine FilterFine Filter

Overriding GuidelinesOverriding GuidelinesOverriding GuidelinesOverriding GuidelinesOverriding Guidelines.....—Management activitiesthat influence the owl and its habitat should beconducted according to the following overridingguidelines:

1. Manage mixed-conifer and pine-oakforest types to provide continuousreplacement nest habitat over spaceand time. Treatment of a particularstand depends on its capability to attainthe desired stand conditions. Targetstand structure would be the describedconditions for nesting and roostinghabitat (Table III.B.1) but only theportion of the landscape that can besustained through time should be inthat condition.

2. Incorporate natural variation, such asirregular tree spacing and various stand/patch sizes, into management prescrip-tions and attempt to mimic naturaldisturbance patterns.

3. Maintain all species of native vegetationin the landscape, including early seralspecies. To allow for variation in existingstand structures and provide speciesdiversity, both uneven-aged and even-aged systems may be used as appropriate.

4. Allow natural canopy gap processes tooccur, thus producing horizontal varia-tion in stand structure.

Specific GuidelinesSpecific GuidelinesSpecific GuidelinesSpecific GuidelinesSpecific Guidelines—The following guidelinesare intended to minimize threats to the Mexicanspotted owl, retain and enhance important butdifficult-to-replace habitat elements, and providemanagement flexibility.

1. Emphasis should be placed on uneven-aged management systems. Existingstand conditions will determine whichsilvicultural system is appropriate.

2. Extend rotation ages for even-agedstands to >200 years. Silvicultural pre-scriptions should explicitly state whenvegetative manipulation will cease untilrotation age is reached. This age maydepend on site quality, but ceasingactivity at 140 years and allowing 60years for unaltered stand maturation andsenescence seems reasonable.

3. Within pine-oak types, emphasis shouldbe placed on management that retainsexisting large oaks and promotes thegrowth of additional large oaks.

4. Retain all trees >61 cm [24 in] dbh.

5. Retain hardwoods, large down logs, largetrees, and snags.

6. Management priority should be placedon reducing identified risks to spottedowl habitat. The primary existing threatis catastrophic wildfire. Thus, westrongly encourage the use of prescribedand prescribed natural fire to reducehazardous fuel accumulations. Thinningfrom below may be desirable or necessarybefore burning to reduce ladder fuels andthe risk of crown fire. Such thinningmust emphasize irregular tree spacing.

7. No stand that meets threshold condi-tions can be treated in such a way as tolower that stand below those conditionsuntil ecosystem assessments can docu-ment that a surplus of these stands existat larger landscape levels (e.g., no lessthan the size of a FS District). This doesnot preclude use of treatments to reducefire risks or lessen insect or diseaseproblems, nor does it preclude manage-ment to meet other ecosystem objectivesas long as stand-level conditions remainat or above the threshold values given inTable III.B.1.

94


Volume I/Part III

RationaleRationaleRationaleRationaleRationale.—.—.—.—.—The collective goal of these generaland specific guidelines is to provide spotted owlhabitat that is well distributed over space andtime. To accomplish this goal requires maintain-ing or creating stand structures typical of nestingand roosting habitats and sustaining them insufficient amounts and distribution to support ahealthy population of Mexican spotted owls. Afew guidelines merit further comment.

Retaining large trees is desirable because theyare impossible to replace quickly and becausethey are common features of nesting and roost-ing habitats for the owl. Fire, viewed as a naturalformative process rather than as a destructiveanthropogenic process, can be used advanta-geously to maintain or improve spotted owlhabitat.

The guidelines presented above should notbe misconstrued as onetime management events.For example, large trees and snags are requiredby the spotted owl and will continue to beneeded by the owl in the future. Further, theapproach outlined above provides a foundationfor the development of a long-term managementstrategy. Once the owl is delisted, we expect thatthis general template can be evaluated, fine-tuned, and possibly applied to PACs and steepslopes.

Riparian CommunitiesRiparian CommunitiesRiparian CommunitiesRiparian CommunitiesRiparian Communities

GuidelinesGuidelinesGuidelinesGuidelinesGuidelines.—.—.—.—.—The goals of these guidelines areto maintain healthy riparian ecosystems wherethey exist and initiate restoration measures toreturn degraded areas to healthy conditions.

1. Maintain riparian broad-leaved forests ina healthy condition where they occur,especially in canyon-bottom situations.Where such forests are not regeneratingadequately, active management may benecessary. Possible actions to restorethese forests may include reducinggrazing pressure, establishing riparianexclosures to manage forage use better,and shifting to winter grazing seasons.

2. Restore lowland riparian areas. Spottedowls once nested in riparian galleryforests. Conceivably, restored riparianforests could contribute additional

nesting habitat in the future and couldalso create a landscape that is moreeffectively connected for dispersing owls.

3. Emphasize a mix of size and age classesof trees. The mix should include largemature trees, vertical diversity, and otherstructural and floristic characteristics thattypify natural riparian conditions.

RationaleRationaleRationaleRationaleRationale.—.—.—.—.—We assume that riparian forestsprovide important habitat for spotted owls.Many riparian systems within the range of theMexican spotted owl are extremely degraded asthe result of past management practices. Becausemany of these systems are degraded and littledocumentation of recent owl use exists, we havelittle empirical information upon which toprovide specific guidelines. Thus, our underlyingpremise is that if riparian systems are restored tomore natural conditions, the needs of the owl(and numerous other species) will be satisfied.This is particularly true in canyon-bottomsituations at middle and lower elevations wherelittle other typical nesting or roosting habitatmay be available. We know that canyon bottomsare used extensively by the owl, thus it is impor-tant to preserve and increase the quality of suchhabitat. We anticipate that PACs will includesome of the best of this type of habitat that stillexists, but increasing the quantity and distribu-tion of healthy riparian habitats provides thepotential for increasing spotted owl habitat.Furthermore, maintenance of existing healthyriparian systems and restoration of those that aredegraded will benefit numerous riparian-depen-dent flora and fauna, and ecosystem healthacross the landscape.

Other Forest and Woodland TypesOther Forest and Woodland TypesOther Forest and Woodland TypesOther Forest and Woodland TypesOther Forest and Woodland Types

We propose no specific guidelines for severalforest and woodland community types wherethey occur outside PACs. These include ponde-rosa pine, spruce-fir, pinyon-juniper, and aspenas defined in II.C. We emphasize, however, thatthe lack of specific management guidelineswithin this plan does not imply that we regardthese types as unimportant to the Mexicanspotted owl.

95


Volume I/Part III

The Team’s rationale for these recommenda-tions is based on extant information on thenatural history of the Mexican spotted owl assummarized in Part II.A and detailed in VolumeII. These forests and woodlands are not typicallyused for nesting and roosting. However, theymay provide habitat for foraging and possibly forboth dispersing and wintering spotted owls. Ourgrasp of the owl’s natural history regarding thesebehaviors is incomplete, so we do not fullyunderstand the structural features the owlrequires for pursuing these activities in theseforest and woodland types. Furthermore, someof the best foraging habitat should be protectedin PACs. All of these circumstances allow us tobe less restrictive in these community typeswithout harming the owl or compromising itsprimary habitat.

With the exception of the acreage of thesetypes contained within PACs, we assume thatthe remaining lands are used primarily forforaging, wintering, migration, and dispersal.Thus, we contend that existing and plannedmanagement for these types will maintain orimprove habitat for these needs of the owl. Thiscontention is based largely on the assumptionthat existing old-growth areas will be maintainedacross the landscape, silvicultural practices willfavor selection over regeneration cuts, andmanagement will be guided by ecosystemapproaches that strive to provide sustainableconditions across the landscape that fall withinthe natural range of variation.

Guidelines developed for protected andrestricted areas may have useful applicationswhen judiciously administered in these otherforest and woodland types. Such guidelinesinclude managing for landscape diversity, mim-icking natural disturbance patterns, incorporat-ing natural variation in stand conditions, retain-ing special features such as snags and large trees,and utilizing fires as appropriate. We also em-phasize the need for proactive fuels managementwhere appropriate. Decreasing fire risks withinthese types, particularly ponderosa pine forests,will also decrease fire risks to adjoining protectedand restricted areas by minimizing the probabil-ity of large landscape-level crown fires that couldimpinge upon occupied or potential nestinghabitat.

96

GRAZING RECOMMENDATIONSGRAZING RECOMMENDATIONSGRAZING RECOMMENDATIONSGRAZING RECOMMENDATIONSGRAZING RECOMMENDATIONS

The explicit goals of managing grazing inspotted owl habitat reflect the four manifestedinfluences of grazing discussed in II.D. Thoseinfluences were (1) altered prey availability, (2)altered susceptibility to fire, (3) degeneration ofriparian plant communities, and (4) impairedability of plant communities to develop intospotted owl habitat. The goals then become (1)to maintain or enhance prey availability, (2) tomaintain potential for beneficial ground fireswhile inhibiting potential for destructive stand-replacing fire, (3) to promote natural andhealthy riparian plant communities, and (4) topreserve the processes that ultimately developspotted owl habitat.

The Team strongly advocates field monitor-ing and experimental research related to theimpacts of grazing on the Mexican spotted owl.Only through monitoring and research can we(1) develop a comprehensive understanding ofhow grazing affects the habitat of the owl and itsprey; (2) determine the effectiveness of currentgrazing standards and guidelines as they relate tothe owl’s needs; and (3) devise grazing strategiesthat can benefit the owl and its prey.

Grazing GuidelinesGrazing GuidelinesGrazing GuidelinesGrazing GuidelinesGrazing Guidelines

The following guidelines should be appliedto all protected and restricted areas:

1. Monitor grazing use by livestock andwildlife in “key grazing areas.” Keygrazing areas are primarily riparian areas,meadows, and oak types. Monitoringshould begin by determining currentlevels of use plus current composition,density, and vigor of the plants. Ulti-mately, monitoring should detect anychange in the relative composition ofherbaceous and woody plants. The intentis to maintain good to excellent rangeconditions in key areas while accommo-dating the needs of the owl and its prey.


Volume I/Part III

2. Implement and enforce grazing utiliza-tion standards that would attain good toexcellent range conditions within the keygrazing areas. Use standards (e.g., FSRegion 3, Range Analysis Handbook)that have been developed for localgeographic areas and habitat types--particularly in key habitats such asriparian areas, meadows, and oak types--that incorporate allowable use levelsbased on current range condition, keyspecies, and the type of grazing system.Establish maximum allowable use levelsthat are conservative and that willexpedite attaining and maintaining goodto excellent range conditions. Thepurpose of establishing these use levels isto ensure allowable use of plant speciesto maintain plant diversity, density,vigor, and regeneration over time.Additionally, a primary purpose is tomaintain or restore adequate levels ofresidual plant cover, fruits, seeds, andregeneration to provide for the needs ofprey species and development of futureowl foraging and dispersal habitat.

3. Implement management strategies thatwill restore good conditions to degradedriparian communities as soon as possible.Strategies may include reductions ingrazing levels and increased numbers ofexclosures (i.e., fencing) to protectriparian plant cover and regeneration,and to prevent damage to stream banksand channels (Clary and Webster 1989,Platts 1990). In many cases, degradedriparian areas may require complete restfor periods from a few years to 15 yearsfor the area to recover (Kennedy 1977,Rickard and Cushing 1982, Clary andWebster 1989). Additional strategies mayinclude the use of riparian pastures,limited winter use, double rest-rotation,and other methods that emphasizeriparian vegetation and stream bank/channel recovery (Platts 1990). Riparianrestoration projects that include the useof exclosures need not require exclosuresalong the entire drainage course at one

time. Rather, systematic use of exclosuresthat protect the most sensitive portionsof riparian habitats is encouraged.Riparian areas can also benefit fromprotection of adjacent upland areas(Bryant 1982). Placement of exclosures(controls) and areas open to grazing(treatments) should be designed topermit determination of effects onseveral ecological responses (e.g., vegeta-tion, erosion, water quality, prey avail-ability).

Rationale for Grazing GuidelinesRationale for Grazing GuidelinesRationale for Grazing GuidelinesRationale for Grazing GuidelinesRationale for Grazing Guidelines

Some effects of excessive grazing on vegeta-tion and habitat features are predictably nega-tive, particularly in riparian communities.However, the collective effects of grazing areneither always predictable nor always negative.Effects depend on site-specific factors such as thegrazing system, condition of the plant commu-nity prior to livestock grazing, soil types, climate,community composition of plant species, andthe presence or absence of aggressive exotic plantspecies. Succinctly, predictability is inexact; andwithout predictability the Team cannot givedetailed and specific recommendations.

We suggest that, when implemented andenforced, general guidelines and the standardsthey prescribe will promote and maintain goodto excellent range conditions over time andacross communities used by the owl. Despite ourimprecise knowledge of how grazing affectsspotted owl habitat, the collective body ofgeneral knowledge regarding the impacts ofgrazing on wildlife mandates prudence. TheTeam believes that understanding how grazingaffects the owl is paramount, and we stronglyurge that specific grazing practices and levels ofgrazing use be carefully evaluated through anexperimental approach (Bock et al. 1993).

Habitats in protected and restricted areasshould receive high-priority management atten-tion relative to grazing. Any riparian communi-ties of potential importance for spotted owldispersal and wintering habitat should alsoreceive high-priority attention. Such attentionwill not only benefit spotted owls but manyother species in the Southwest as well (Hubbard

97


Volume I/Part III

1977). Fundamental to the guidelines forgrazing is the assumption that individual actionshave collective effects. For example, the short-term goal of exclusion by fencing is to protectriparian plants and to prevent physical damageto stream banks and channels (Clary andWebster 1989, Platts 1990). The long-term goalis that such short-term protection ultimatelyallows spotted owl habitat to develop. Implicit inthis rationale is that excessive grazing sustainedfor long periods not only deteriorates potentialor actual spotted owl habitat but it also inevita-bly leads to a deterioration of the very qualitiesthat make an area attractive for grazing in thefirst place.

We also assert that attainment and mainte-nance of good to excellent conditions in keygrazing areas will translate to better conditions inthe uplands. Most native and exotic ungulatespreferentially graze within key areas such asmeadows and riparian areas. We assume that ifthese key areas exhibit ecologically good condi-tions, upland forests and woodlands should alsobe in good condition. Thus, negative effects ofgrazing that lead to the establishment of ladderfuels and “dog-hair” thickets may be amelio-rated.

RECREATIONRECREATIONRECREATIONRECREATIONRECREATIONRECOMMENDATIONSRECOMMENDATIONSRECOMMENDATIONSRECOMMENDATIONSRECOMMENDATIONS

GuidelinesGuidelinesGuidelinesGuidelinesGuidelines

The following guidelines should be appliedto all protected and restricted areas:

1. No construction, either of new facilitiesor for expanding existing facilities,should take place within PACs duringthe breeding season, 1 March through31 August. Any construction withinPACs during the nonbreeding seasonshould be considered on a case-specificbasis. Modifications to existing facilitiespertaining to public safety and routinemaintenance are excepted.

2. Managers should, on a case-specific basis,assess the presence and intensity ofallowable recreational activities within

PACs. Spatial and temporal restrictionsshould be considered for new activities.

3. Seasonal closures of specifically desig-nated recreational activities should beconsidered where appropriate.

RECOVERY UNITRECOVERY UNITRECOVERY UNITRECOVERY UNITRECOVERY UNITCONSIDERATIONSCONSIDERATIONSCONSIDERATIONSCONSIDERATIONSCONSIDERATIONS

We review below primary threats within eachrecovery unit. Some threats are ubiquitous acrossthe range of the spotted owl, whereas others arelimited to one or few RUs. To place these threatsin perspective, we review below relevant infor-mation from Part II.B for each RU. Manage-ment priorities within each RU should focus onthe threats identified below.

One management consideration that appliesto all RUs is the potential for migration anddispersal of spotted owls within and among RUs.Admittedly, we know very little of the prevalenceof such movements, nor do we know much ofthe habitats used. We suspect, however, thatmovements of birds may be important to geneflow and the maintenance of a metapopulationstructure (Keitt et al. 1995). Thus, efforts shouldbe made to preserve options by maintaining andenhancing potential avenues for migration anddispersal. This could be particularly important inspecific canyons, riparian areas, and mountainranges that might provide links within, between,and among RUs.

Colorado PlateauColorado PlateauColorado PlateauColorado PlateauColorado Plateau

The Colorado Plateau is the largest of allU.S. RUs. It encompasses the southern half ofUtah, much of northern Arizona, most ofnorthwestern New Mexico, and a small portionof southwestern Colorado. In the northwesternportion of this RU, owls have been located insteep-walled canyons with apparent concentra-tions in the areas of Zion N.P., Capitol ReefN.P., western Abajo Mountains, andCanyonlands N.P. Historical records are availablefrom forested habitats on the Kaibab Plateau ofnorthern Arizona. In the southeastern portion ofthis RU, owls occur in both steep-sloped, mixed-conifer forested canyons and steep-walled

98


Volume I/Part III

canyons on the Navajo Indian Reservation;known concentrations occur in the Black Mesaarea and the Chuska Mountains. Owls have alsobeen located in mixed-conifer habitats in theZuni Mountains and on Mount Taylor (FigureII.B.3).

Owl distribution in this RU appears to behighly fragmented. This distributional patternmay be natural or the result of inadequate surveyeffort in some parts of the RU. Extensive surveyshave, however, been completed in the southernUtah portion of this RU. Here, breeding owlshave been found only in canyons where theynest and roost in caves and on ledges. In south-ern Utah, no breeding owls have been located inhundreds of thousands of hectares surveyed inmixed-conifer or other forest types in areas withless than 40% slope. Therefore, we recommendthat surveys in southern Utah emphasize steepslopes and rocky canyons.

Potential ThreatsPotential ThreatsPotential ThreatsPotential ThreatsPotential Threats

Levels of recreational activity are high andincreasing in some areas of this RU, such assouthern Utah. Some activities may potentiallylead to habitat alteration or direct disturbance ofowls. Furthermore, owls in southern Utah nestand roost in canyons, and the physical structureof canyons tend to magnify disturbances andlimit escape/avoidance routes for owls. Potentialthreats listed in order of severity for the north-western portion of this RU include recreation,overgrazing, and road development withincanyons, and catastrophic fire and timber harvestwithin upland forests potentially used for forag-ing, dispersal, and wintering. For the southeast-ern portion, threats include timber harvest,overgrazing, catastrophic fire, oil, gas, andmining development, and recreation (UtahMexican Spotted Owl Technical Team 1994).

Southern Rocky Mountains - ColoradoSouthern Rocky Mountains - ColoradoSouthern Rocky Mountains - ColoradoSouthern Rocky Mountains - ColoradoSouthern Rocky Mountains - Colorado

Lying completely within the State of Colo-rado, the Southern Rocky Mountains - Coloradorepresents the northeastern extreme of theMexican spotted owl’s range. Few owls have beendetected in this portion of its range, and itsnatural history in Colorado is poorly under-stood. However, the condition of a species being

scarce at the periphery of its range is not un-usual. Nesting and roosting habitat may beprimary concerns, but wintering habitat may bean important factor in this RU. Although verylittle is known about wintering habitat, somedata suggest that birds may winter at lowerelevations that include a wider range of condi-tions than found in breeding habitats.


In order of severity, potential threats for theSouthern Rocky Mountains - Colorado RU arecatastrophic fire, recreation, urbanization, timberharvest, and road construction. Less severethreats include land exchange, oil and gasleasing, mineral development, and grazing.Singly, these factors may have low impact, buthigh synergistic consequences. For example,much of the urban development in SouthernRocky Mountains - Colorado currently occurs atelevations lower than those occupied by breed-ing owls; but development increases recreationalaccess to public lands. Road construction orexpansion causes initial disturbance, recreationfacilities extend the disturbance, and theimproved access increases the contact betweenpeople and spotted owls. The initial activitymay directly affect wintering habitat. Thedevelopment threat is considered to be of lowto moderate severity and is highest along theFront Range.

Southern Rocky Mountains -Southern Rocky Mountains -Southern Rocky Mountains -Southern Rocky Mountains -Southern Rocky Mountains -New MexicoNew MexicoNew MexicoNew MexicoNew Mexico

Ranking as the smallest U.S. RU, the South-ern Rocky Mountains - New Mexico supportsone of the smallest known populations ofMexican spotted owls as well. Existing data aretoo incomplete to cite even a crude estimate ofthe RU’s spotted owl population or its density.

The inability to provide crude populationestimates may be partially related to the inad-equacy of existing survey protocols. Owl surveycrews, following the FS Region 3 survey proto-col, have speculated that spotted owls in the areamay not respond to calling surveys as predictablyas they do in other RUs. They base this opinionon the lack of response to nighttime calling inareas where owls or young have been observed

99


Volume I/Part III

during subsequent daytime visits. Given thatsurveys are an important step in designatingPACs and the types of management permissible,it is critical that surveys have a high probabilityof detecting owls. Further, an ineffective, orpartially effective, survey protocol leaves the FSand other agencies ill-equipped to managepotential threats to the Mexican spotted owl inthe RU.

Recent survey efforts on the Santa Fe Na-tional Forest suggest that some areas formerlyoccupied by owls appear vacant now despite thefact that the habitat has not been altered appre-ciably (T. Johnson, Las Alamos, NM, pers.comm.). This perceived, but unconfirmed,population decline indicates the importance ofprotecting unoccupied habitat.


The most serious threat to spotted owls inSouthern Rocky Mountains - New Mexico RU iswildfire and, in localized areas, timber harvest.Fire may not be as serious in canyon systems as itis in other areas because the open structure ofsteep-slope woodlands associated with canyonsare not conducive to conflagration. However,dense mixed-conifer and ponderosa pine forestsoutside of canyons may present the greatest firehazards. Although these areas may not containowls, fires initiated in these forests may continueinto the forested canyon habitats. Personnel atSanta Fe National Forest have instituted anaggressive prescribed fire program in the JemezMountains as a way to reduce the risk of exten-sive fire. Though useful, prescribed fire shouldbe used conservatively in spotted owl habitat.

Timber harvest levels appear greatest on theCarson National Forest where few spotted owlshave been confirmed. Timber harvest levels onSanta Fe National Forest have been reduced inareas where spotted owls are known to occur.Isolation of spotted owl pairs and small popula-tions distributed over large areas of fragmentedlandscape prompt concern because if they arelost, the species disappears from entire land-scapes it once inhabited. Since the spotted owlwas listed, planned timber sales have mostlyavoided the owl’s habitat. However, in theVallecitos Federal Sustained Yield Unit in Carson

National Forest, sales are still being planned inpotential spotted owl habitat despite uncon-firmed sightings of spotted owls in 1993.

Lesser threats to the spotted owl includehuman activities that produce extremely local-ized effects, but that may ultimately prove tohave a large collective impact. Among them areunregulated fuelwood harvesting, grazing (par-ticularly in riparian areas), and recreation devel-opments at ski areas. All of these activities havethe potential to degrade spotted owl habitatincluding habitat for the owl’s prey.

Upper Gila Mountains Upper Gila Mountains Upper Gila Mountains Upper Gila Mountains Upper Gila Mountains

The Upper Gila Mountains RU contains thelargest known number of Mexican spotted owlswith approximately 55% of known spotted owlterritories (Ward et al. 1995). The owl alsoappears to be more continuously distributedacross this RU than any other (II.B). The appar-ent gap in owl distribution in the center ofFigure II.B.6 reflects incomplete informationfrom Tribal lands rather than an actual disconti-nuity within the owl’s range.


Spotted owls throughout the RU are foundprimarily in mixed-conifer and pine-oak forests(Ganey and Dick 1995), often in conjunctionwith canyon terrain. The primary threats tospotted owls and their habitat are timber harvestand catastrophic fire, not necessarily in thatorder. Both threats could destroy forest habitatwith the structural features used by spotted owls,and both could operate over large spatial scales.

Other threats within this RU include indis-criminate fuelwood cutting and overgrazing byboth wildlife and livestock. These threats are notas widespread or severe as the threats discussedabove, but they can be significant in some areas.Fuelwood cutting is a problem in some areasprimarily because people remove (usually ille-gally) large oaks. These trees appear to be criticalto owls in some areas or habitats, particularly thepine-oak type (Ganey et al. 1992). Fuelwoodharvest can also result in loss of large snags anddown logs. Both of these habitat components arealso apparently important to the owl, eitherdirectly or indirectly through effects on the prey

100


Volume I/Part III

base (Ganey and Dick 1995, Ward and Block1995).

Overgrazing is suspected to be detrimental insome areas and can affect both habitat structureand the prey base. Effects on the prey base aredifficult to quantify, but removal of herbaceousvegetation can reduce both food and coveravailable to small mammals (Ward and Block1995). This may be especially true with respectto voles, which are often associated with densegrass cover. Direct effects on habitat are obviousin some places, particularly with respect tobrowsing on young Gambel oak. In some areas,oak is regenerating well but unable to growbeyond the sapling stage because of this brows-ing. Coupled with loss of large oaks to fuelwoodharvest, maintenance of most oak stems in asapling stage suggests a very real possibility thatlarge oak trees will not be replaced over largeareas, resulting in the loss of an importanthabitat component. Grazing effects on habitatare also potentially significant in canyon-bottomriparian areas. We do not attribute these effectssolely to livestock. Forage resources are shared bylivestock and wild ungulates, and reducingnumbers of both will likely be necessary to bringforage use to reasonable levels.

Basin and Range - WestBasin and Range - WestBasin and Range - WestBasin and Range - WestBasin and Range - West

Sprawling across southern Arizona andextreme southwestern New Mexico, the Basinand Range - West RU ranks as the second largestRU in the United States. Though it probablydoes not support as large a spotted owl popula-tion as the Upper Gila Mountains RU, theknown population ranks third highest in theUnited States despite limited survey efforts inmany areas. Therefore, the Team regards theBasin and Range - West RU as an important unitfor the recovery effort.


The Team perceives limited threats overall tospotted owls in the Basin and Range - West RUas the result of human activities. Very littletimber harvest occurs in this RU, though sometimber is cut in the Bradshaw Mountains of thePrescott National Forest and on the San Carlos

Apache Reservation. The primary threats tospotted owls within this RU are catastrophicwildfire, recreation, and grazing. We detail belowthe nature and extent of these and other poten-tial threats.

Historical efforts to suppress fire haveallowed fuel loads to accumulate to dangerouslevels within most of the wooded and forestedvegetation types in this RU. For example, the1983 fire in the Animas Mountains removed theconiferous forest from the higher elevations.Recent wildfires in the Pinaleno, Rincon,Chiricahua, and Huachuca Mountains also attestto the volatile situation in this region. We viewthe potential for catastrophic wildfire as theprimary threat to spotted owls in the Basin andRange - West RU.

Many mountain ranges in the Coronado,Prescott, and Tonto National Forests are usedheavily for recreation. This is partly because oftheir proximity to large urban areas (Tucson andPhoenix) and partly because of their interna-tional reputation for exceptional birding. Effectsof recreation include development of roads,campgrounds, and trails, and also extraordinaryuse of those facilities. For example, a number ofareas within the Coronado National Forest (e.g.,Madera Canyon in the Santa Ritas, Garden andRamsey Canyons in the Huachucas, and theSouth Fork of Cave Creek in the Chiricahuas)are world renowned for birding and receivethousands of visitors per year. Scheelite Canyonon the Fort Huachuca Army Base is visited oftenby birders specifically to view the pair of spottedowls that occur there. The Mexican spotted owl,in fact, is one of the more popular species soughtby birders in this region.

Cattle grazing occurs throughout the RU.Impacts are greatest in the high desert grasslands,desert scrub, and riparian habitats found be-tween mountain ranges. Moderate grazingpressures occur within mid-elevational encinaland pinyon-juniper woodlands; and grazingpressures are evident within higher elevationalcanyon stringers of pine-oak, mixed-conifer, andriparian forests. Perhaps the primary threat ofgrazing is to the low-elevation riparian forests.These forests may represent critical linkagesamong the mountain ranges. Modified anddegraded riparian forests may inhibit dispersal

101


Volume I/Part III

ring in the Sacramento Mountains have beenexposed to various disturbances for more than acentury. Natural disturbances include forest firesplus insect and disease outbreaks. Humandisturbances include timber and fuelwoodharvest, grazing, land development, and recre-ation. The cumulative effects of these naturaland anthropogenic disturbances have resulted ina landscape that differs from that existing priorto European settlement. The threat of thesedisturbances to the owl’s persistence cannot bequantified at this time, but certain incongruitiesare detectable. For example, owl density isrelatively high on FS lands, but fecundity isquite variable over time and annual survival isunknown. Thus, even though the currentpopulation density may be high, we knownothing of population trends. Further, givenexisting forest conditions in this RU, threats ofwidescale habitat loss are real and immediate.Consequently, active management is needed toalleviate these threats while ensuring that ad-equate habitat will exist well into the future.


The Team categorized potential threats tospotted owl recovery according to magnitude.Major threats pose immediate potential forcausing declines in spotted owl populations, andminor threats present no such immediacy. Majorthreats, in order of potential effects, include (1)catastrophic, stand-replacement fires, (2) someforms of timber harvest, (3) fuelwood harvest,(4) grazing, (5) agriculture or development forhuman habitation, and (6) forest insects anddisease. Minor threats are activities not currentlyextensive in time or space but that have beenconsidered potential threats to the owl. Theseinclude (1) certain military operations, (2) otherhabitat alterations (e.g. power line and roadconstruction, noxious weed control), (3) mining,and (4) recreation.

Existing dense forest conditions makes muchof the Basin and Range - East RU vulnerable tocatastrophic fire. Such stand-replacing fires havebeen documented in the Sacramento Mountainssince the 1950s and continue to the present(e.g., the Burgett and Bridge fires in 1993 and1994, respectively). Similar fires occurred in the

among mountain ranges and gene flow amongowl subpopulations.

Land ownership within the Basin and Range- West is a mosaic of public and private lands(II.B). Most major mountain ranges fall underFederal jurisdiction, with some privateinholdings and other lands administered by theSan Carlos Apache and White Mountain Apachetribes. Much of the shrublands and grasslandsbetween mountain ranges are administered bythe FS or BLM, but a fair portion is privatelyowned. Many of these private lands are used forcattle, which graze both in upland and adjoiningriparian communities. Grazing in ripariancommunities is a concern because of the poten-tial for negative impacts on areas that can pro-vide dispersal habitat among mountain ranges.This mosaic pattern of jurisdiction by multiplelandholders may impede coordinated manage-ment efforts for the owl.

Most land development within the RU isrelated to enhancing recreation opportunities.These include developing or expanding camp-grounds (e.g., Twilight Campground in thePinaleno Mountains, the loop turn at JohnHand Lake in the Chiricahuas) or enlargingroads for safety (e.g., widening of MountLemmon highway in the Santa Catalina Moun-tains). Developments such as these often requireremoving trees, potentially altering owl habitat.Further, a rapidly increasing human populationin the southwest portends increasing urbandevelopment which can potentially encroachupon owl habitat and also impact groundwaterregimes, potentially impacting riparian systems.

Evergreen oak and pinyon-juniper wood-lands receive the most pressure from fuelwoodharvest. Historical harvest of mature mesquitestands within the shrubland and grassland areasmay have contributed to the demise of theriparian forest, thus continued harvest of maturemesquite in these areas may be a concern.

Basin and Range - EastBasin and Range - EastBasin and Range - EastBasin and Range - EastBasin and Range - East

The Basin and Range - East RU lies mostlywithin New Mexico and supports the secondlargest known number of Mexican spotted owlsin the United States. It adjoins four U.S. andone Mexican RUs. Mexican spotted owls occur-

102


Volume I/Part III

Smokey Bear Ranger District (Hanks and Dick-Peddie 1974) and other mountain ranges in theBasin and Range - East RU (Plummer andGowsell 1904, Moody et al. 1992). Failure toaddress this potential for fire by reducing fuellevels and fuel continuity will inevitably lead tomore and larger fires resulting in the continuedloss of owl habitat.

Past timber harvest practices have left a fewremnant old-growth stands and residual pocketsof pre-harvest trees in the Sacramento Moun-tains. Trees older than 200 ybh (years at breastheight) can be found in these remnant standsand pockets. Many of these stands, however, aresmall (<4 ha [10 ac]) and exist as smaller grovesamid the younger coniferous forests. Our obser-vations indicate that these remnant patches arecritical to the Mexican spotted owl, particularlyfor nesting and roosting. This situation is similarto spotted owl use of second-growth redwood innorthern California. Both cases should beviewed as exceptional and regionally dependentprocesses. Regardless, many of these olderpatches are on the verge of senescence anddecline. Few patches are on a trajectory toreplace remnant patches as they are lost in theshort term. Thus, timber harvest in the immedi-ate future must avoid altering these remnantpatches in such a way as to accelerate theirdecline. Rather, forest management must striveto create replacement patches as quickly asconditions allow to ensure that these uniquehabitat patches are sustained through time.

Insects, plant pathogens, and dwarf mistletoecomprise a third important agent of forestdisturbance in the Sacramento Mountains(Plummer and Gowsell 1904, Stevens and Flake1974, Hessburg and Beatty 1986, Hawksworthand Conklin 1990, Archambault et al. 1994).Principal organisms are western spruce bud-worm, round-headed beetle, white pine blisterrust, dwarf mistletoe, Phellinus schweinitzii, andother fungi. These organisms operate at scalesranging from single trees to landscapes. Not onlyare insects and tree diseases fundamental deter-minants of forest structure and function(Attiwell 1993, Haack and Byler 1993), butforest structure and composition influencepopulation levels of these organisms. As a result,the dense forest conditions existing in the

Sacramento Mountains have allowed someinsects and diseases to increase from endemic toepidemic levels. Clearly, forest management thatdecreases forest density, primarily by thinningfrom below, will help to control populations ofsome of these organisms.

Grazing by domestic livestock and elk in thisRU has altered botanical cover, specifically plantcomposition and structure. Range managementhas been oriented toward domestic livestock andother wildlife goals, but not for the owl. Regard-less of its past orientation, grazing can affect owlhabitat and prey populations in conflicting andpoorly understood ways (II.D). Effects of grazingare largely manifested in meadow and riparianareas, but effects within forests cannot be easilydiscounted. Implementation of the grazingrecommendations (provided above) are neededto understand and address potential effects ofgrazing on the spotted owl.

Agriculture and concentrated human devel-opments occur in the Rio Grande Valley and to alesser extent in the Sacramento Mountains. Bothmay affect dispersing or wintering owls byreducing the spatial extent of habitat. Manage-ment that emphasizes the restoration of riparianforests may benefit both resident birds in theSacramento Mountains and birds migratingbetween mountain ranges.

At present, the Team considers the impactsof recreation to be of minor importance to theRU’s spotted owls; but we have no studies ordocumentation to substantiate our position.Recreation noise from motorcycles and snowmo-biles has been implicated as a potential threat.Indirect habitat disturbance from recreation mayoccur on a local scale but is also undocumented.Other activities include camping, hiking,birding, hunting, off-road vehicle use,snowmobiling, and skiing. Many private landinholdings are summer homes or camps and arealso a source of recreation in spotted owl habitat.

MexicoMexicoMexicoMexicoMexico

Presently, limited information on the biologyof the Mexican spotted owl and on land manage-ment activities within Mexican RUs precludesthe provision of extensive management recom-mendations. However, the information available

103


Volume I/Part III

indicates that spotted owls use forest types nottypically found in the United States and thatland-management practices differ substantiallyin Mexico from those in the U.S. The Teamproposes that the general recommendations beapplied within the Mexican RUs where appro-priate. The Team strongly recommends that RUworking teams (see Part IV) develop manage-ment recommendations for Mexico more fully.

The Mexican spotted owl is typically foundin montane habitats where the vegetation isdominated by pines and oaks. Although spottedowls in the U.S. also use pine-oak forests, theyvary somewhat in composition and structurefrom similar forests in Mexico. However, withinMexican pine-oak forests, spotted owls appear tofavor canyons, as they do in many of the areasused in the U.S.

Because social and economic systems inMexico differ from those in the U.S., activitiesthat take place within potential spotted owlhabitat differ somewhat between the two coun-tries. Whereas grazing, fire, timber harvest, andfuelwood harvest are threats common to bothcountries, other threats are unique to Mexicoand some to specific RUs.


Sierra Madre Occidental - NorteSierra Madre Occidental - NorteSierra Madre Occidental - NorteSierra Madre Occidental - NorteSierra Madre Occidental - Norte.—.—.—.—.—The pri-mary threat is land conversion for subsistenceagriculture. Impacts of this threat include theloss of spotted owl habitat, soil loss, and erosion.A related threat is overgrazing by livestock.Historically and through the present, forestlands have been cleared to create pastures forcattle; the cumulative effects of these practiceshave modified spotted owl habitat. Althoughextensive timber harvest occurs within this RU,most harvest occurs in the uplands and is not adirect threat to spotted owls found in canyons.Whether or not timber harvest indirectly affectsthe owl is unknown, but could be a concernworth addressing through research.

Sierra Madre Oriental - NorteSierra Madre Oriental - NorteSierra Madre Oriental - NorteSierra Madre Oriental - NorteSierra Madre Oriental - Norte.—.—.—.—.—Timberharvest and grazing are the primary threatswithin this RU. The main effects of these threatsis to further fragment an already disjunct popu-lation.

Sierra Madre Occidental - SurSierra Madre Occidental - SurSierra Madre Occidental - SurSierra Madre Occidental - SurSierra Madre Occidental - Sur.—.—.—.—.—Primarythreats within this RU are fuelwood harvest,timber harvest, charcoal production, grazing,and agricultural development. Fuelwood harvestoften entails cutting snags and the harvest ofriparian plant species, both of which are criticalcomponents of spotted owl habitat. Timberharvest in itself is not a direct threat to spottedowl habitat. However, harvest methods thatentail rolling logs from higher to lower sitesresult in soil loss and erosion, thereby affectingthe habitat of the owl and its prey. Fire is con-sidered only a moderate threat because of thedisjunct distribution of owls in canyons andbecause limited efforts towards fire suppressionhave allowed natural fire regimes to persist. Fire,however, can possibly destroy spotted owlhabitat under the right conditions. Livestockgraze throughout the year within spotted owlhabitat, and the cumulative effect of this grazingaffects prey habitat and spotted owl habitatstructure. Type conversion of forests for agricul-ture also occurs within this RU.

Sierra Madre Oriental - SurSierra Madre Oriental - SurSierra Madre Oriental - SurSierra Madre Oriental - SurSierra Madre Oriental - Sur.—.—.—.—.—The potentialthreats in this RU parallel those within SierraMadre Occidental - Norte RU. The primarydifference is that this RU probably contains thebest and most extensive habitat within any of theMexican RUs, thus threats can occur over amuch greater area.

Eje NeovolcanicoEje NeovolcanicoEje NeovolcanicoEje NeovolcanicoEje Neovolcanico.—.—.—.—.—The primary threats in thisRU are those associated with population expan-sion and industrial development. Specifically,industrial development can lead to loss of habitatand an increase in pollution. Other threatsinclude management to control insects anddisease, fire suppression, grazing, and agricul-tural development.

104


Volume I/Part III

The Team has assimilated, reviewed, andanalyzed data generated by the Mexican SpottedOwl Monitoring Program of FS Region 3. Wehave also compiled and reviewed data from theBLM and the FS Region 4 in Utah, and FSRegion 2 in Colorado. Here, we offer an alterna-tive design for monitoring the Mexican spottedowl population within the three core RUs. Wealso provide recommendations for monitoringhabitat throughout the owl’s range. This pro-posed monitoring program will evaluate popula-tion and habitat trends as required by the criteriafor delisting the species (III.A). The philosophyof our proposed monitoring scheme is to mea-sure the critical variables--changes in owl num-bers and changes in habitat--needed for delistingthe species.

For the purposes of recovery and under-standing effects of land management activitieson the Mexican spotted owl, monitoring shoulddetermine with adequate reliability temporalchanges in the owl population and its habitatwhen in fact such changes are occurring. Aneffective monitoring program requires measuringchanges in habitat quantity, estimating popula-tion size of territorial owls, and determining keydemographic parameters including survival,recruitment, and reproduction, all of whichinfluence population size.

Habitat monitoring will rely heavily on bothremote-sensing of habitat across the range of thebird and field measurements of habitat variablesbefore and after treatments. Population monitor-ing is based on mark-recapture theory andCormack-Jolly-Seber modeling approaches,similar to Pollock’s robust design approach(Lefebre et al. 1982, Pollock 1982, Kendall andPollock 1992). To our knowledge, a systematichabitat monitoring approach as extensive andintensive as the one we propose has never beenimplemented. In contrast, a prototype design forpopulation monitoring was implemented inOlympic National Park, Washington (Noon etal. 1993, E. Seamen pers. comm.) to monitor anorthern spotted owl population.

Accurate and efficient protocols for bothhabitat and population monitoring require pilotstudies to estimate recapture probabilities, and toestimate variances associated with each of thepopulation parameters and each of the habitatvariables. For population monitoring, theseestimates can then be used to determine optimalquadrat size and numbers of quadrats requiredwithin predefined strata and RUs. Funding wasallocated in 1994 to further refine the proposedpopulation design with a small field trial involv-ing four quadrats. Results of this field trialvalidated the study design provided qualifiedpersonnel conduct the work (May et al., inpress). A larger pilot study is needed to refineparameter estimates for actual implementationof the proposed design. For habitat monitoring,separate pilot studies will be needed to establishsampling designs, including sample size require-ments.

HABITAT MONITORINGHABITAT MONITORINGHABITAT MONITORINGHABITAT MONITORINGHABITAT MONITORING

The habitat delisting criterion states thathabitat monitoring must be implemented (1) totrack changes in the quantity of macrohabitatand (2) to verify that microhabitat changeswithin treated stands meet the intent of theRecovery Plan. Thus little, if any, owl habitat canbe lost if this goal is to be met. Further, habitatquality cannot decline significantly. A concern ofthe Team is that habitat quality cannot beadequately assessed, particularly with remote-sensing data. To alleviate this concern, ourrecommendations also include field measure-ments of microhabitat characteristics withintreated stands. However, we reiterate thatmacrohabitat quantity should also be monitoredon a rangewide basis.

MacrohabitatMacrohabitatMacrohabitatMacrohabitatMacrohabitat

The purpose of rangewide monitoring is totrack gross changes in habitat as the result ofdisturbance, from both natural (e.g., fire) and

C. MONITORING PROCEDURESC. MONITORING PROCEDURESC. MONITORING PROCEDURESC. MONITORING PROCEDURESC. MONITORING PROCEDURES

105


Volume I/Part III

anthropogenic (e.g., timber harvest, prescribedfire) causes. Given the extent of the area to bemonitored, remote-sensing technology will berequired. An imagery baseline should be estab-lished within six months of Recovery Planapproval using LANDSAT Thematic Mapperimagery (currently available in the FS Region 3Geometronics Remote-Sensing Laboratory). Theimagery for potential owl habitat and surround-ing areas should be aggregated, georeferenced,and merged with vector map data to provideimage maps. These image maps may be usedboth as general planning tools and as a baselinefor detecting change. The 30-m (98-ft) spatialresolution of the LANDSAT Thematic Mapperimagery is adequate to detect both anthropo-genic and natural change across large land areas.Changes that can be detected include wildfirescars, timber harvests, gross changes in foresthealth, and possibly the addition or removal ofroads, large developments and other culturalfeatures, and fluctuations in grazing practices.

In a standard change-detection analysis, twoimage data sets are co-registered and then sub-tracted from each other. The resulting differenceimage highlights changes across the area coveredby the image data sets. An additional image dataset will need to be purchased in the future toperform change detection. The additional dataset may be either LANDSAT Thematic Mapper(TM) data or LANDSAT Multispectral Scanner(MSS) data. A TM data set would be preferred,but MSS data could be substituted and wouldprovide adequate spectral and spatial resolutionto perform useful change detection.

The baseline TM data set can also be used todevelop a generalized regional vegetation covermap, using standard supervised and unsuper-vised image classification techniques with TMbands 4, 3, and 2. The literature indicates thatthe 4,3,2 band combination is most useful forvegetation analysis. Developing a vegetation mapwould also require the integration of 1:250,000digital elevation data to account for the effects ofvarying slope aspects and elevation on vegetationpatterns.

Texture analysis techniques may be used toassess vegetation structure and density across

large areas using either the LANDSAT TM orMSS data. While texture analysis applications invegetation analysis have appeared frequently inthe literature, the methodologies are not aswidely accepted as image classification tech-niques, so they must be considered experimental.

Given the resolution possible with the toolsand information available, remote sensingmonitoring techniques will provide an estimateof macrohabitat trend. Within five years ofcreating the imagery baseline, participatingagencies should produce a report assessingchanges in vegetation composition, structure,and density. This will provide an interim check-point to determine if the delisting criteria can bemet at the end of 10 years.

MicrohabitatMicrohabitatMicrohabitatMicrohabitatMicrohabitat

Microhabitat monitoring is required becauseremote sensing is largely insensitive to subtleintra-stand changes that may enhance or degradeowl habitat. Microhabitat monitoring will entailmeasuring habitat variables before and aftersilvicultural or prescribed fire treatments de-signed to maintain, improve, or create owlhabitat. This monitoring is to verify that treat-ments (silviculture, fire) are meeting their statedobjectives. We acknowledge that many treatedstands will not meet the desired future conditionin 10 years, but the trajectory on which a standis placed can be modeled to evaluate if it ismoving towards owl habitat. This knowledge isneeded to demonstrate that any short-term lossesin macrohabitat will be partially offset as standsmature into owl habitat. If adequate acreage ofvegetation is moving towards owl habitat, ourconfidence in long-term habitat stability willbe enhanced.

Sampling units will be treated stands. Withinthese stands, an adequate number of vegetationsampling points must be established. The exactnumber of sampling points needed will bedictated by the most variable characteristic(likely snag density; Bull et al. 1990). Pointsshould be sampled prior to initiating a treat-ment, resampled following the treatment afterallowing adequate time for the area to equilibrate

106


Volume I/Part III

from temporary disturbance effects, and then atfive-year intervals. The variables measured canthen be input into a vegetation model to esti-mate stand characteristics at different points intime.

At a minimum the following variablesshould be measured and assessed: (1) tree diam-eters by species, (2) tree basal area by species, (3)size-class distributions of trees, (4) log volume bysize class, (5) canopy cover, (6) snag diameter,and (7) snag basal area. We strongly advocatethat additional variables be included that mightbe relevant to monitoring other ecosystemattributes. The return in critical monitoringinformation derived by expanding the variablesmeasured would far outweigh any additionalcosts, assuming that the new variables are nothighly correlated with the variables suggestedabove.

POPULATION MONITORING POPULATION MONITORING POPULATION MONITORING POPULATION MONITORING POPULATION MONITORING

Monitoring habitat as a singular effort willnot adequately reveal the true status of the owlpopulation. Relatively long-lived birds with ahigh (~0.89) adult survival, Mexican spottedowls may live 16 years or more once they reachadulthood. However, an intense period ofmortality during the first year could producepopulation consequences that habitat monitor-ing would not detect. Habitat quality coulddecline from various natural processes or anthro-pogenic activities, yet the territorial populationwould remain unchanged because of site fidelityamong existing birds and recruitment of floaters.Young might still be produced, but would notsurvive to be recruited into the territorial popu-lation because of poor habitat quality, limitedhabitat availability, or because their inexperiencewould not allow them to survive and disperseduring their first year.

A limitation of this proposed monitoringscheme (and all known approaches) is that onlyterritorial birds are monitored. The total orproportional number of floaters (sensu Franklin1992) remains undetermined and unmonitoredrelative to the target population.

Because nonterritorial birds are not directlymonitored, we want to guard against an unde-tected decline in the total population of spottedowls when the territorial population remainsstable. We suggest the following procedures toevaluate trends in the nonterritorial population.First, the age of birds that establish territorieswill indicate the size of the nonterritorial popu-lation. If new territorial birds are only one yearold, then they have never existed as floaters inthe population. Thus, a decline in the age ofterritorial birds suggests that the nonterritorialpopulation is low or declining (Franklin 1992).Second, the presence of unfilled territories wouldsuggest that an inadequate floater populationexists, and hence that a decline in the populationis taking place.

In the following, we outline a suitableframework and statistical estimation approachfor monitoring owl populations in 3 RUs.However, critical design and sampling details,such as sample sizes and delineation of strata,have been omitted, and must be developed by animplementation team as data become availableto make those decisions.

Target PopulationTarget PopulationTarget PopulationTarget PopulationTarget Population

The target population for the abundanceestimate is territorial Mexican spotted owls(exclusive of floaters) in the Upper Gila Moun-tains, Basin and Range - West, and Basin andRange - East RUs. Thus, all potential owl habitatin these 3 RUs must be included in the samplingframe. All land management jurisdictions areencouraged to cooperate in providing access andresources to monitor the entire owl population.

Sampling UnitsSampling UnitsSampling UnitsSampling UnitsSampling Units

Sampling units will consist of 50 to 75 km2

(19 to 29 mi2) quadrats randomly allocated tohabitat strata. Quadrats will be defined based onecological boundaries such as ridge lines andwatersheds to reduce edge effects. Selection ofquadrat boundaries must emphasize edges thatare unlikely to traverse owl territories, so that theerrors of including a territory in multiple quad-

107


Volume I/Part III

rats or in no quadrat do not occur. The exactnumber of quadrats and their size will dependon the specifics of implementing the monitoringscheme and results of pilot studies.

In general, the population monitoringscheme will require: (1) determining strata thatrepresent different owl densities or habitat typesoccupied by owls within each RU; (2) determin-ing quadrat size which should be sufficientlylarge to reduce edge effects and small enough toallow a minimum of four surveys per quadrat inthe survey season limited to 1 April to 30 August(initial approximation of quadrat size is 50 to75 km2 [19 to 29 mi2] [May et al., in press]);(3) defining the sampling frame of quadrats ineach stratum such that quadrats are relativelyequal in size and have boundaries selected tominimize edge effect; (4) selecting a randomsample of quadrats from each stratum the firstyear, and then randomly replacing 20% of thesampled quadrats each year with quadratsrandomly selected from the currently unsampledquadrats (with quadrats in the initial samplepotentially removed, and then possibly includedin the sample again at a later time); and (5)developing protocols for conducting fieldsurveys (likely different from past FS protocolsbecause of the different goal of the proposedprocedure from past goals).

Sampling ProceduresSampling ProceduresSampling ProceduresSampling ProceduresSampling Procedures

StratificationStratificationStratificationStratificationStratification

LANDSAT multispectral scanner imagerywith 30-m spatial resolution would be suitablefor defining habitat strata within each RU, andthus the sampling frame of quadrats for eachstratum. Approximate density of territorial owlswithin strata can then be used to allocate surveyeffort (number of quadrats) to strata. Theoptimum allocation of survey effort is one thatwould minimize erroneous estimation of spottedowl abundance. Optimal allocation of quadratsto strata will probably not be in proportion tostrata size. More likely, optimal allocation willmean that a higher percentage of quadrats instrata with high owl densities will be sampled.

Optimal allocation might also take into accountthe cost per quadrat because of potential differ-ences in the cost of measuring quadrats withindifferent strata. Strata should be computed bothas projected (ignoring topography) and assurface (incorporating topography) areas becausedifferences in topography affect vegetation type.

Selection of QuadratsSelection of QuadratsSelection of QuadratsSelection of QuadratsSelection of Quadrats

Within strata, quadrats will be randomlyselected for inclusion in the sample. We suggestthat 80% (randomly selected) of the previousyear’s quadrats be revisited the following year.The other 20% of the previous year’s quadratsshould be removed from the sample, and areplacement sample of quadrats not sampled theprevious year should be substituted. Quadratsremoved one year can included in the sample infollowing years provided that it is randomlyselected and that at least one year has elapsedsince it was last sampled. This procedure meansthat the average number of years that a particularquadrat remains in the sample is 4.5 years.Inclusion of new quadrats into the sample eachyear guards against management practices withinthe sampled quadrats not being representative ofthose practices occurring elsewhere. This rota-tion of quadrats included in the sample willprovide a smaller variance because observationsare correlated across time and many quadrats willbe sampled repeatedly, but still provides arepresentative sample of the available quadrats.

Sampling Within QuadratsSampling Within QuadratsSampling Within QuadratsSampling Within QuadratsSampling Within Quadrats

Sampling procedures within a quadrat willinclude (1) assigning survey stations to ensureadequate coverage and a standardized density ofsuch stations among quadrats; (2) allowing for aminimum of four complete (i.e., all call pointssampled) surveys through each quadrat; (3)conducting nighttime surveys from surveystations to map general locations of spotted owlsand to estimate per-visit detection probabilities;and (4) conducting daytime (auxiliary) surveysand mousing to find roosting and nestingspotted owls, determine if a mate not detected

108


Volume I/Part III

Statistical AnalysisStatistical AnalysisStatistical AnalysisStatistical AnalysisStatistical Analysis

Noon et al. (1993) provide the followingstatistical discussion.

Estimation of Capture ProbabilityEstimation of Capture ProbabilityEstimation of Capture ProbabilityEstimation of Capture ProbabilityEstimation of Capture Probability

The described spotted owl surveys willprovide data regarding the number of territorialowls detected and determined to have theiractivity centers within the survey plots. It isunlikely that all owls with activity foci lyingwithin these plots will be detected, and captureprobabilities will therefore be less than 1. Cap-ture is defined as either physically capturing anduniquely marking an individual or resighting itsunique color band combination without physi-cally recapturing it. Thus, the per-visit captureprobabilities must be estimated to “correct” thecount statistics and reflect the true number ofterritorial birds with activity centers withinquadrat boundaries.

Per-visit capture probabilities can be esti-mated using data on the capture histories ofindividual owls on the quadrats. The foursurveys will be conducted during a relativelyshort period, so it is appropriate to use capture-recapture models for closed populations (e.g.,Otis et al. 1978, White et al. 1982, Pollock et al.1990). Per-visit capture probabilities may vary byvisits, strata, RUs, and years. However, substan-tial gains in the precision of capture probabilityestimates would be achieved if they could beestimated using data pooled over visits, strata,RUs, or years. Standardized survey protocolwithin and among quadrats using field crewswith communal training should decrease thevariation in per-visit capture probabilities.

Heterogeneity of per-visit capture probabili-ties across individuals should be examined. Ifheterogeneity is found, estimators developedunder model Mh of Otis et al. (1978), such asthe jackknife (Burnham and Overton 1978,1979) and Chao’s (1987, 1988, 1989) estimatorare appropriate. If heterogeneity is not serious,models with data pooled over visits, strata, RUs,and years should be considered.

Capture probability estimates resulting fromthese modeling efforts will pertain to the prob-

during nighttime survey is present, determinenumber of young present, and capture andcolor-mark all spotted owls found. For eachspotted owl found, the center of its activity areamust be determined as either in or out of thequadrat.

Proper timing of surveys maximizes effi-ciency in locating spotted owls. We recommendthat nighttime surveys be conducted within 3hours following sunset. Owls detected at duskare near diurnal roosts and thus provide anoptimal starting point for confirming pairs andreproduction during daytime surveys. Similarly,daytime surveys should begin at or near sunrise(preferably just before) as owls are returning totheir roosts.

Banding BirdsBanding BirdsBanding BirdsBanding BirdsBanding Birds

Marking individual birds with FWS legbands and color bands for visual identificationprovides greater validity in the estimation of theowl population size on the quadrat because theassumptions of the mark-recapture methods canbe tested. Conceivably, owl population size onquadrats with high densities of owls might beunderestimated without banding because twodifferent birds might be counted as only one.Conversely, on quadrats with low densities, asingle bird might be counted as 2 birds, biasingthe population estimate high. Individuallymarking birds will eliminate some of this poten-tial bias. Second, banding birds is necessary toestimate annual survival on quadrats that aresampled for two consecutive years. Third,capturing birds allows more careful aging ofindividuals; hence, the resulting age structuredata are more useful in assessing the impact offloaters in the population. Finally, minimumestimates of dispersal and emigration from thequadrat can be assessed with banded birds thatare located off the quadrat. Although the cost ofthe population estimation procedure may beincreased by up to 40% by individually markingbirds (May et al., in press), the Team feels theadditional information and rigor provided bymarking birds is justified.

109


Volume I/Part III

Estimation of Density Per QuadratEstimation of Density Per QuadratEstimation of Density Per QuadratEstimation of Density Per QuadratEstimation of Density Per Quadrat

Each quadrat requires two parts to theestimation process. First, estimation of apparentdensity by

Dia = = ,

which is the total number of spotted owlsdetected and having activity centers in quadrat i(ni), based on both survey (ni) and auxiliary (ni)visits divided by the area of the quadrat (ai).

Next, apparent density is adjusted for thosespotted owls that maintain an activity centerwithin the quadrat but were not detected duringa survey visit. The adjustment for the survey visitis the reciprocal of pk , the probability of aspotted owl being captured at least once on agiven survey based on k visits to quadrat i. Notethat p

k pertains only to the n

i animals detected

during the survey point visits of quadrat i, not tobirds detected on auxiliary visits. If auxiliaryvisits are made to determine pair or reproductivestatus, any additional pair members that aredetected contribute to the density estimate forquadrat i. As discussed previously, their captureis conditional upon an initial capture from asurvey point. To adjust the count of the ni owlsdetected on quadrat i during auxiliary visits, wedivide the count by the probability of detecting apair member during k auxiliary visits (p

k ), given

detection and capture of its mate from surveypoints. Thus, we would estimate density onquadrat i as

Di = .

Estimation of Density Per StratumEstimation of Density Per StratumEstimation of Density Per StratumEstimation of Density Per StratumEstimation of Density Per Stratum

Once we have density estimates for eachquadrat within a stratum, they can be combinedinto an overall mean estimate of density for thestratum. Because quadrats are not the same size,weighting of the quadrat density estimates byarea is essential, so that

Dj =

ability of detecting and capturing an individualspotted owl during a single survey visit. Capturehistory data, and hence the capture probabilityestimates, will be restricted to those spotted owlswith a nest or focus of activity within the areabeing sampled. The density estimation proce-dure (see below) actually requires an estimate ofthe probability of detecting and capturing aspotted owl at least once during an entire season.Given owl presence, the estimated probability ofdetecting and capturing an owl during theseason using surveys is given by

pk = 1 - (1 - p)k,

where pk is the probability of detecting andcapturing a spotted owl at least once duringk visits, and p is the single-visit capture probabil-ity.

The above scheme for estimating captureprobability should work well with owls initiallydetected from survey points within the quadrat.However, these capture probabilities are notapplicable, by themselves, to other pair membersfound during daytime visits or to spotted owlslocated other than by calling from survey points.To include data from pair members locatedduring daytime visits, we consider their captureprobability as the product of the probability ofcapturing a pair member from survey pointstimes the probability of capturing the other pairmember during a daytime visit given capture ofone mate from survey points. The first probabil-ity in this product is obtained using capturehistories (i.e., capture probabilities) of birdsdetected from survey points as previously de-scribed. The second, conditional probabilitymust be obtained using data from daytime visitsand captures only.

Auxiliary or daytime visits can be viewed inthe context of removal modeling (e.g., Zippin1956, 1958; Otis et al. 1978; Ward et al. 1991).The primary purpose of daytime visits is todetermine pair and breeding status of birdsdetected from survey points. Auxiliary visits mayor may not be terminated after capture of themate, depending on whether data on reproduc-tive success is adequately obtained.

^

^

i=1 ,

mj

^

mj

i=1

ai

aiDi

�

ni

pk

^ +n

i¨·

pk p

k^ ¨

ai

110

ni + ninia

i

·

ai

¨

�

^

¨·

^

^

¨

¨

·


Volume I/Part III

where mj is the number of quadrats in stratum j.To obtain population size estimates for eachstratum (j), multiply by the area of the stratum

(Aj) so that Nj = Dj · Aj. Note that Aj may changethrough time as habitat changes and quadrats aremoved from one stratum to another.

An estimate of overall abundance for RU u is

then Nu = � Nj , where mu is the number ofstrata in RU u.

Variance EstimatorsVariance EstimatorsVariance EstimatorsVariance EstimatorsVariance Estimators

With stratified-random sampling, we usuallyhave simple variance equations because thestratum means of totals are independent betweenstrata. Here, this is not the case because correc-tions for spotted owls not seen are commonacross strata and induce the need for covarianceterms. These variance equations will need to bedeveloped. Specific closed-formed solutions maynot be possible for the variance estimates. Thus,estimating the variance components by boot-strap methods may be more feasible.

Cormack-Jolly-Seber ModelsCormack-Jolly-Seber ModelsCormack-Jolly-Seber ModelsCormack-Jolly-Seber ModelsCormack-Jolly-Seber Models

Cormack-Jolly-Seber (CJS) models(Lebreton et al. 1992) can be used to estimateage-specific apparent survival (�a) and recruit-ment to the territorial population (B). Wesuggest using CJS modeling procedures asoutlined by Lebreton et al. (1992), Burnhamand Anderson (1992), Pollock et al. (1990), andBurnham et al. (1987). This approach is demon-strated by White et al. (1995) for the analysis ofdata from the demographic study areas. Wesuggest that data from all quadrats within astratum (or even larger area) can be pooled toestimate apparent survival and recruitment.

Both the apparent survival (�) and recruit-ment (B) are biased estimates of true survival (S)and true recruitment rates because the quadratsdo not have geographic closure. For survival,� = S - E, where E is the emigration rate off thequadrats. For recruitment, B = R + I, where R istrue recruitment and I is immigration onto thequadrats. If the area of the combined quadratswere used in a single demographic study area,the bias of � and B would be smaller because the

probability of birds emigrating off of and immi-grating onto a large single area would be smallerthan for a collection of small quadrats represent-ing the same area. However, if we assume thatthis bias is somewhat constant across time, thentests for changes in ��and B across time withmodels such as �T as demonstrated by Burnhamet al. (1994) provide a potent tool to assesschanges in these population parameters throughtime. The optimal size of quadrats is dictated bykeeping them large enough that reasonableestimates of the number of territorial birdspresent can be accomplished, while smallenough so that an adequately large sample ofquadrats is possible to estimate precisely theamong-quadrat variation.

We would not suggest that ��and B be usedto compute�� in a Leslie matrix model as wasdone with the demographic study areas byWhite et al. (1995). Biases caused by emigrationand immigration make any estimate of � com-puted from these parameters biased as well.Furthermore, the main objective of the quadratsurveys is to provide an unbiased estimate of thetotal number of territorial owls so that anunbiased estimate of � can be obtained as ��= Nt+1/Nt.

Estimates of juvenile apparent survivalobtained from the CJS model with banding datafrom juvenile spotted owls will probably not beuseful from the pooled quadrat survey databecause the emigration rate of this populationsegment will be quite high as they disperse awayfrom their natal territories.

PersonnelPersonnelPersonnelPersonnelPersonnel

Quality work cannot be completed withoutcapable people who desire to perform well. Owlsurveys are difficult to conduct. To achieveaccurate survey results requires a certain combi-nation of physical and mental traits. The idealcandidate for spotted owl survey work must bephysically capable of negotiating difficult terrainand doing so after dark. The mental demandsinclude the intellectual capacity to understandthe nuances of the work, the perseverance tosucceed under adverse conditions, the ability tofollow directions, and the discipline to be

^

^mu

j=1

^

^ ^ ^

111


Volume I/Part III

patient. Rigorous training to certify people toconduct monitoring is a critical step to ensurethat qualified people implement the procedures.

Restrictions on duration of work day, nighttime work, and camping near survey sites canlead to inefficiency. Effective inventory andmonitoring may often require personnel tosurvey between dusk and 2300 hrs and prior tosunrise the next morning if an owl is detectedthe previous night. Not permitting camping atsites or not allowing more than 8-hour workdays is an ineffective survey strategy. Thus, costand effort for determining occupancy or repro-duction in a given territory may be doubled ortripled. Not allowing personnel to survey alongmarked ridge lines at night (i.e., off roads) mayresult in inadequate survey of an area. Compe-tent, qualified, eager personnel can conduct suchactivities safely and with desired results as dem-onstrated by May et al. (in press).

TrainingTrainingTrainingTrainingTraining

Training is the most important mechanismfor ensuring quality data and standardization.The current certification program employed bythe FS should continue, but in a more intensefashion. High-quality photographic media,including video tapes of proper procedures,should be incorporated. Use of a map, compass,and GPS system, and recording spatial informa-tion with Universal Transverse Mercator (UTM)coordinates (Grubb and Eakle 1988) are critical.Additional training for recording information ondata forms, maps, and in an electronic data baseis required (see below). Certified owl biologistsshould be tested routinely on their ability tocomplete data forms and plot locations correctly.A standardized procedure for storing all informa-tion also needs to be developed and enforced.

All training must be reinforced with ad-equate (4-day minimum) field exercises followedby periodic reinforcement of learned skills.Although initial skills may be provided with thecertification process, reinforcement throughfeedback on procedures and results is required.An electronic data entry program will help tostandardize inputs and reinforce proper docu-mentation procedures. Such a routine will alsoindicate progress of the monitoring program and

identify personnel or administrative units thatneed additional training. Further, additionaltraining should include periodic visits by pro-gram supervisors to review field procedures.Incorrect observations may not necessarily bedetected on data forms.

We also suggest a greater emphasis onidentification of spotted owl age classes (juvenile,subadult, and adult) as described by Forsman(1981) and Moen et al. (1991). This valuableinformation may be obtained if observers takebinoculars on surveys. Information on agestructure may prove useful for identifyingchanges in demographic trends.

All survey routes and results need to besummarized in a standardized manner on mediathat can be entered into a Geographic Informa-tion System (GIS). Thus, field personnel willrequire training on use of map, compass, andGPS. In addition, a standardized map systemand symbol set is paramount. We recommend a7.5" USGS topographic map. This map type isreadily available in paper and digital form. Wealso insist that field personnel record UTMcoordinates. This will allow rapid updating ofdigital maps of owl locations.

Computerized Data Entry andComputerized Data Entry andComputerized Data Entry andComputerized Data Entry andComputerized Data Entry andSummarizationSummarizationSummarizationSummarizationSummarization

A major weakness of past inventory andmonitoring programs has been the lack ofaccessibility to data; as a result, few summariesand analyses were prepared. This scarcity of dataexamination appears to be due to the lack of acentral, accessible, computerized database wherefield forms are regularly entered for computeranalysis. Field workers submitting data forms butnot receiving feedback from their efforts nor acopy of the master database for them to reviewleads to errors that are difficult to rectify retroac-tively. We suggest that field workers who collectthe data should also be responsible for data entryinto a standardized computer form. The benefitswould be twofold.

First, a computerized data entry form wouldguarantee that only admissible codes are usedbecause invalid codes would not be accepted bythe computer, and correct entries would be

112


Volume I/Part III

needed in all data fields before the user couldproceed. Quality control would be facilitated viaan interactive computer data entry interface.With such a data entry program, data fromdifferent jurisdictions of all involved land man-agement entities would be compatible.

Second, once the data have been entered,summaries can be produced with standardsummary programs. At a minimum, field work-ers should be able to produce summaries of datathey entered, and make comparisons with pastyears and maybe other geographic areas. Themain reason for this instant feedback is toencourage field personnel to examine their owndata plus get a temporal and spatial perspectiveof existing data. Field workers would have amuch better picture of how their data fit into theoverall effort and would have access to data inthe master database. Simple graphs and tabularsummaries should be available via a menusystem. This feedback would also promotegreater cooperation in future surveys and makesthe field worker feel a part of the completeprocess.

Creation of a master database on an acces-sible computer network (such as the World WideWeb [WWW] on Internet) has another, lessapparent, benefit for the program. From thismaster database, region-wide summaries couldbe generated. Annual summaries would helpdetect trends in the data. Sophisticated statisticalanalyses could be programmed to implementtests for trends in the data. Safeguards would benecessary to limit access to the data, particularlysensitive site locations, to only authorizedpersons. Finally, scrutiny by outside reviewerswould improve the integrity of the database.

To implement the above scheme, two piecesof software need to be written. The first is thedata entry system, for which extensive errorchecking should be coded into the software.Data entered by a field worker would be ap-pended to the master data file only after passinga stringent series of integrity checks. The secondis a data summary program that would be menudriven and allow the user to summarize his/herown data plus other data of interest. We presumethat modern PC computer software systems andaccess to a computer network should make thissoftware development fairly easy. Once the field

season is completed, each land managemententity and their respective subunits should beable to obtain graphical summaries as well asstatistical summaries in tabular form.

CostsCostsCostsCostsCosts

The cost for implementing the populationmonitoring scheme should include hiring aprincipal investigator to design this survey andcoordinate sampling efforts. Field crew leaderswill be necessary for supervising study logisticsand field technicians will be required to conductsurveys. In addition, while models exist forestimating total population size through time,models of multiple capture probabilities requiresome independent work. All field personnelhired to conduct the pilot study and subsequentmonitoring program must be qualified andtrained.

Our initial estimate of the costs to fullyimplement the proposed monitoring scheme isapproximately $1.2-1.5 million per year. Basedon the delisting criteria, monitoring mustcontinue for a minimum of 15 years.

Costs for implementing macrohabitatmonitoring are unknown. However, much of theneeded remote-sensing coverage exists or is beingobtained as a tool for implementing ecosystemmanagement. Thus, additional costs attributableto the spotted owl should be minimal. Costs ofimplementing microhabitat sampling are diffi-cult to estimate without knowledge of thenumber of plots to be sampled. We assume,however, that sampling an area pre- and post-treatment is already required as a standard partof activity implementation; consequently, thecost attributable to owl monitoring would entailthose associated with measuring additionalvariables specific to the owl. Thus, the total costsof habitat monitoring should be relativelyminimal beyond that already required or in theprocess of being developed independent of thespotted owl.

Potential ExperimentsPotential ExperimentsPotential ExperimentsPotential ExperimentsPotential Experiments

Many habitat variables important to Mexi-can spotted owls cannot be monitored by remotesensing. Further, it is important to ensure that

113


Volume I/Part III

adequate habitat is provided for key prey as well.Thus, we propose some potential experiments torelate habitat conditions to owl populationdynamics where key habitat characteristics wouldbe measured on the ground. On-the-groundmonitoring of relevant habitat characteristicswould quantify their change at a local (i.e.,within quadrat) scale and relate them to owlpopulation dynamics.

Population monitoring based on randomlyselected quadrats provides the opportunity toconduct experiments to extend our knowledge ofthe impact of habitat manipulation on Mexicanspotted owl population dynamics. We proposethese experiments to produce credible, defen-sible, and reliable results (sensu Murphy andNoon 1991). Quadrats within the monitoringdesign may serve as experimental units forexamining the effects of future managementsuch as fires, grazing, timber harvest, and recre-ation.

Given that a treatment is identified prior toits occurrence, vegetation measurements can takeplace on the site of the expected treatment and asecond, control quadrat that is selected based onits similarity to the expected treatment quadrat.This experimental design is not a true experi-ment, because the treatment is not randomlyallocated to one of the pair of quadrats. How-ever, this quasi-experiment is still more powerfulin developing cause-and-effect relationshipsbetween habitat manipulations and owl popula-tion dynamics than the more common correla-tive designs used by past researchers (see III.Dfor further details on experimental design).Further, the capability to replicate the treatmentexists because of the extensive number of quad-rats that will be required for measuring changesin population size.

Areas where planned treatments result insome form of habitat alteration provide excellentopportunities for quasi-experiments. Vegetationmeasures should be taken immediately beforeand after the habitat-modifying event, andthereafter at 5-year intervals. Vegetation mea-surements that seem especially important toexamine are tree size-class distribution, log size-class distribution, canopy cover, and shrub cover.Results from these experiments, coupled withresults of population monitoring, will provide

the basis for a predictive model of spotted owlhabitat quality (assuming that owl densityreflects habitat quality). Data on apparent owlsurvival and reproduction will also be available,which may relate to habitat quality more directlythan owl density.

Alternative Designs for PopulationAlternative Designs for PopulationAlternative Designs for PopulationAlternative Designs for PopulationAlternative Designs for PopulationMonitoringMonitoringMonitoringMonitoringMonitoring

Drawing New Sample of Quadrats Each YearDrawing New Sample of Quadrats Each YearDrawing New Sample of Quadrats Each YearDrawing New Sample of Quadrats Each YearDrawing New Sample of Quadrats Each Year

Instead of drawing an initial sample ofquadrats from the sampling frame and monitor-ing these same quadrats through time, an alter-native approach would be to draw a completelynew random sample of quadrats each year. Forrepeated sampling of a set of quadrats to belegitimate, normal activities that occur in spot-ted owl habitat should continue during themonitoring program, provided these activitiesmeet the requirements of section 7(a)(2) of theAct by not likely jeopardizing the continuedexistence of the Mexican spotted owl. The mainadvantage of a new sample each year is that itguards against the potential for land managers tomanage areas within the quadrats differentlythan the remainder of the landscape. The priceof this protection is relatively great as illustratedby these four points: (1) the logistics of conduct-ing the surveys each year would increase becauseof the new quadrats; (2) age-specific apparentsurvival rates and recruitment to the territorialpopulation could not be estimated with the CJSanalysis because birds would not be marked onthe same area each year; (3) quasi-experiments todetect the relationship between habitat manipu-lations and owl population dynamics would notbe possible; and (4) higher sampling intensitieswould be required because this design is lessefficient for estimating change. Our proposeddesign is intermediate between sampling thesame set of quadrats each year and a completelynew sample each year. We obtain the benefitsfrom both alternatives in that the correlation ofmeasurements for a specific quadrat across yearsis used to lower the overall variance of ourpopulation estimate, making the design moreefficient than complete replacement each year,yet we are guarding against the potential for

114


Volume I/Part III

sampled quadrats to be managed differently thanother areas.

Conducting Surveys Less Often Than YearlyConducting Surveys Less Often Than YearlyConducting Surveys Less Often Than YearlyConducting Surveys Less Often Than YearlyConducting Surveys Less Often Than Yearly

Instead of surveying quadrats each year,effort and cost could be saved by conducting thesurveys at longer intervals, such as every 5 years.An advantage of this approach is that costs willbe lowered, and possibly more precise estimatesof population size could be obtained by poolingmoney to conduct a few very good surveysinstead of more frequent surveys with lowereffort per survey. The main disadvantage of thisapproach is that age-specific apparent survivalrates and recruitment to the territorial popula-tion could not be estimated with the CJS analy-sis because birds would not be marked frequentlyenough to obtain these estimates. For example,given an estimate of 0.89 for adult survival, only56% of the initial population would still be aliveafter 5 years, resulting in small sample sizes ofrecaptured birds, and hence poorer precision ofthe survival estimates. Further, reproductive andannual survival rates and their variation acrossyears are needed to realistically evaluate popula-tion viability. Finally, our ability to detectrelationships between habitat manipulations andpopulation dynamics would be greatly decreasedbecause this approach is more sensitive to vari-ability introduced by the years chosen for sam-pling.

Adaptive SamplingAdaptive SamplingAdaptive SamplingAdaptive SamplingAdaptive Sampling

Thompson (1992) has developed an adaptivesampling scheme to improve the efficiency ofsampling clustered populations, such as isprobably the case for Mexican spotted owls.Thompson’s scheme is theoretically appealingbecause more effort is applied to areas whereowls are located. Under this approach, quadratsadjoining a quadrat that contains some thresholdnumber of spotted owls would also be sampled.Unfortunately, we cannot envision how tohandle the logistics of adding some unknownnumber of quadrats to the sample when surveycrews must be hired, trained, and outfitted withequipment and vehicles prior to sampling. Wesuspect the logistical overhead of this approachmay make it impractical for monitoring owls onquadrats. However, as the theory and applicationof the adaptive sampling scheme is developedfurther, an innovative application of the tech-nique may be possible with our proposed quad-rat monitoring scheme.

CONCLUSIONCONCLUSIONCONCLUSIONCONCLUSIONCONCLUSION

The technology and expertise are available tomonitor trends in Mexican spotted owl habitatand population size. Clearly, the objectives anddesign of the monitoring program must bedefined explicitly and they must be attainable.To implement the process, knowledgeable,dedicated people must be assigned the task.Adequate training and constant feedback mecha-nisms are critical aspects to a successful monitor-ing program as tenable conclusions can be basedonly on reliable data.

115


Volume I/Part III

The primary objectives of our proposedresearch program are to (1) enhance understand-ing of Mexican spotted owl biology and (2)assess how land management practices affect theowl population’s viability. These types of infor-mation are necessary to complement recoveryefforts outlined in this plan. The research pro-gram described here is different from the moni-toring program outlined in III.C. Whereas bothprograms are necessary, specific research needsmay or may not be related to monitoring. Indeveloping this chapter, we realized that readersof this plan have a variety of backgrounds. Thus,to establish a common framework for the discus-sion of a research program for the Mexicanspotted owl, we first outline the role of thescientific process in research and some importantaspects of study design. We then discuss somelimitations with previous research, and suggestfuture research questions and processes thatshould be examined.

ROLE OF THE SCIENTIFICROLE OF THE SCIENTIFICROLE OF THE SCIENTIFICROLE OF THE SCIENTIFICROLE OF THE SCIENTIFICPROCESSPROCESSPROCESSPROCESSPROCESS

Research and the reliability of knowledgegained from research depend on appropriateapplication of the scientific method. Reliableknowledge can be defined as “the set of ideasthat agree or are consistent with the facts ofnature,” whereas “unreliable knowledge is the setof false ideas mistaken for knowledge”(Romesburg 1981). Three primary scientificmethods have been used in scientific research(Romesburg 1981): (1) induction that involvesthe use of repeated observations to discover lawsof association; (2) retroduction where a “best-guess” hypothesis is developed to explain a law ofassociation or some set of observations; and (3)hypothetico-deductive (HD) where a priori hy-potheses are developed and tested, and a decisionmade about whether to reject the hypotheses. Itis generally accepted in science that applicationof the HD method provides the best avenue forgaining reliable knowledge (Platt 1964, Popper1965, Romesburg 1981, 1991). Steps used in

D. ACTIVITY-SPECIFICD. ACTIVITY-SPECIFICD. ACTIVITY-SPECIFICD. ACTIVITY-SPECIFICD. ACTIVITY-SPECIFICRESEARCHRESEARCHRESEARCHRESEARCHRESEARCH

the HD method can be reiterated as followsfrom Nichols (1991): “(1) suggest a hypothesisto explain some phenomenon of interest, (2)deduce a testable prediction from that hypoth-esis, (3) devise and carry out a suitable test, and(4) use observations from the test to decidewhether the prediction is met.” When observa-tions and predictions match, the hypothesis iscorroborated; when they do not match, thehypothesis has been falsified and can be dis-carded. Rejection of hypotheses is key to the HDmethod. Corroboration of hypotheses can resultfrom poor experimental designs (e.g., lowpower). Therefore, knowledge in the HDmethod is gained more through falsification ofhypotheses than through corroboration.

Most research related to natural resourcemanagement and conservation has relied prima-rily on induction and retroduction (Romesburg1981, 1991). Induction can provide us withreliable knowledge about associations such as theassociation of Mexican spotted owls with forestshaving certain structural characteristics. How-ever, this method does not provide the mecha-nism for understanding the processes thatunderlie this association nor does it providereliable knowledge about cause and effect.Whereas we can describe the structure of forestsused by spotted owls, we cannot ascertain whichstructural characteristics are “important,” or why,without application of the HD method. Inshort, we can describe patterns through induc-tion but need the HD method to understandwhy those patterns occur and which componentsof those patterns are “important.” In terms ofmanagement, understanding why a pattern hasoccurred and what caused it are important forpredicting effects when observed patterns arechanged.

Romesburg (1981) argued that retroductiondoes not provide reliable knowledge because ofthe inability of this method to falsify hypothesesand the large number of alternative hypothesesthat could equally explain the same conclusions.However, both induction and retroduction areuseful for describing relationships and develop-

116


Volume I/Part III

ing hypotheses to be further tested using the HDmethod. Unreliability of retroduction can beexacerbated when untested hypotheses areintegrated into our knowledge base as dogma inthe form of scientific “rules.”

While induction and the HD methodprovide a general framework for gaining reliableknowledge, design of appropriate studies iscrucial to the application of this method inspecific situations. This applies to both describ-ing and understanding patterns in nature. Anymanagement plan, including the MexicanSpotted Owl Recovery Plan, is a complex hy-pothesis whose rejection or corroboration isdetermined by the success or failure of the planover the long term.

IMPORTANT ASPECTS OFIMPORTANT ASPECTS OFIMPORTANT ASPECTS OFIMPORTANT ASPECTS OFIMPORTANT ASPECTS OFEXPERIMENTAL DESIGNEXPERIMENTAL DESIGNEXPERIMENTAL DESIGNEXPERIMENTAL DESIGNEXPERIMENTAL DESIGN

The ability to confidently infer results from asample to a population of interest and thestrength of that inference are entirely dependenton study design. Reliability of knowledge andthe ability to make correct inferences are directlyproportional; the stronger the inference one canmake, the more reliable the knowledge stem-ming from that inference. The strongest infer-ence in understanding patterns is achievedthrough controlled experiments, with thestrength of inference diminishing the further agiven study design departs from the experimen-tal (HD) approach. However, inferences can beweakened even in experimental studies if thedesign is not valid. With Mexican spotted owls,we would like to extend inferences to a largerpopulation than the one from which wesampled. This larger population may be acrossthe range of the owl, within a certain recoveryunit, or on a Ranger District within a NationalForest. The ability to extend conclusions from astudy to a larger area or a longer time perioddepends directly on how the study was designedand implemented.

For our purposes, study designs can becharacterized as either descriptive or experimen-tal (following Eberhardt and Thomas 1991).Descriptive studies employ survey sampling,whereas experimental studies use treatment and

control groups. Necessary components of thedesign in both cases include: (1) randomizationwhere samples are randomly selected in a de-scriptive study, or treatments and controls arerandomly assigned in an experiment; and (2)replication of experimental units through bothspace and time. Randomization removes subjec-tive biases that may be found in descriptivestudies and guards against systematic differencesother than treatment effects in experiments.Randomization also allows for stronger inferenceto a larger population and is the theoretical basisfor employing statistical tests. Replication allowsfor estimation of experimental error, a prerequi-site for employing statistical tests. If eitherrandomization or replication are omitted froman experimental design, inferences will be greatlyweakened. True replication should not be con-fused with “pseudoreplication” wheresubsampling of experimental units is confusedwith replication of experimental units (Hurlbert1984). For example, a habitat study that mea-sures 100 vegetation plots within each of fourowl home ranges represents a sample of four, not400. Frequently, researchers are guilty ofpseudoreplication by reporting a sample size of400. Inferences from such a study apply only tothe 4 owls studied and not to a larger popula-tion. Thus, adequate replication must occur atthe level of the experimental unit (in this case,number of home ranges) to apply to a largerpopulation.

A major difficulty in doing field experimentsis that they are performed in an uncontrolled,“noisy” environment (Eberhardt and Thomas1991). Therefore, pre- and post-treatmentmeasurement periods in both control andtreatment groups are needed to reduce the effectsof external variation and to ensure that a treat-ment effect can be adequately measured. Addi-tional important design features necessary infield experiments include the choice of experi-mental units (e.g., owls, owl sites), local control(amount of balancing and blocking of experi-mental units), and the choice of the design (e.g.,complete block, incomplete block, factorial).

An important consideration when designingand implementing studies that involve testingstatistical hypotheses is the power of the statisti-cal test used (the probability of rejecting the null

117


Volume I/Part III

hypothesis when it is false). Failure to reject anull hypothesis is due to either (1) the nullhypothesis was indeed “true” or (2) there wasinsufficient power to reject it. Thus, powershould be as high as possible (>90%) to corrobo-rate that an unfalsified null hypothesis wasactually not false. Power is dependent on acombination of the severity of the treatmentapplied, sample size, and experimental error. If atreatment is subtle, then a larger sample will benecessary to achieve the same power as if a severetreatment was used. This is important becausebiological questions often involve chronic(subtle) effects rather than acute (severe) effects.For example, the effects of a given land manage-ment practice may have a slight effect on adultsurvival rates which may in turn have strongeffects on population viability. If an experimenttesting such an effect has low power, then it maybe tempting to state that the practice does notsignificantly affect survival rates when in fact itdoes. Repercussions from an experiment lackingsufficient power would then be misleading andresult in false confidence in the health of thepopulation.

LIMITATIONS IN PASTLIMITATIONS IN PASTLIMITATIONS IN PASTLIMITATIONS IN PASTLIMITATIONS IN PASTRESEARCH ON THE MEXICANRESEARCH ON THE MEXICANRESEARCH ON THE MEXICANRESEARCH ON THE MEXICANRESEARCH ON THE MEXICAN

SPOTTED OWLSPOTTED OWLSPOTTED OWLSPOTTED OWLSPOTTED OWL

Previous research on Mexican spotted owlshas been largely descriptive and has relied oninduction and retroduction; our current knowl-edge concerning underlying ecological processesis, therefore, limited. However, previous researchon Mexican spotted owls has provided a goodfoundation to describe the natural history of thespecies and to generate hypotheses for experi-mental tests with the HD method. Additionallimitations on conclusions from previous re-search result from (1) lack of randomization inselecting experimental units and study areas, (2)lack of true replication (including small samplesizes), and (3) lack of experiments. The follow-ing discussion is not meant as criticism ofspecific research studies or scientists. Many ofthe previous studies have been hampered byinadequate funding and logistical constraintsbeyond the investigators’ control. These factors

are not unique to research on the Mexicanspotted owl; they are common to research onnumerous species, including both the northernand California spotted owls.

Lack of randomization and replication hashampered the ability to infer from particularsamples to the general. In a number of studies,pseudoreplication has also been confused withtrue replication, weakening inferences evenfurther. For example, most of the habitat studiesusing radiotelemetry have suffered frompseudoreplication. These studies typicallysampled few (4-10) birds, but sampled habitatcharacteristics within these few birds’ homeranges extensively. In testing hypotheses, thenumber of subsamples were used, rather than thenumber of owls, to estimate error terms used instatistical tests. Such pseudoreplication lends anincorrect perception of adequate power tostatistical tests which may lead to incorrectconclusions. However, repetition of home-rangestudies over additional areas has strengthenedinferences concerning certain habitat associa-tions.

Controlled experiments have not been usedin research on Mexican spotted owls. Lack ofexperiments is probably related to the need toquickly identify basic aspects of spotted owlnatural history and apply this information tomanagement situations. However, experimentsare critical for defining the impacts of currentand proposed management activities on Mexicanspotted owls.

RESEARCH NEEDSRESEARCH NEEDSRESEARCH NEEDSRESEARCH NEEDSRESEARCH NEEDS

Several management issues and questionsmust be resolved to better understand andimplement recovery measures for the Mexicanspotted owl. Communication and collaborationbetween people with strong research skills andpeople with strong management skills will be akey component in this process. Managers needto better understand the methods, problems anduncertainties involved with research. Research-ers, on the other hand, must rely on managers toidentify appropriate questions, political and legalconstraints, and to develop appropriate imple-mentation of knowledge derived from research

118


Volume I/Part III

results. Too often researchers design and imple-ment studies that do not adequately addressmanagement problems. People having bothresearch and management skills will hopefullybridge the gap between management and re-search disciplines. Both time and money areshort. Clearly, all research questions cannot beanswered within a short time frame. Therefore,we advocate that a series of crucial experimentsbe implemented that address questions mostrelevant to the needs of management agencies.The following example of such an experimentaddresses the question, “what structural featuresin forest habitat are needed to maintain highfitness in Mexican spotted owls,” where fitness issome function of survival and reproduction:

• Determine appropriate statistical hy-potheses (predictions) and responsevariables to be tested. Testing fitnessdirectly through survival and reproduc-tive rates may not be feasible because ofthe prohibitively large samples needed todetect chronic effects and ethical prob-lems in purposely affecting survival ofthe owls. However, appropriate hypoth-eses from the initial question is thatdecline in foraging use and prey avail-ability in altered habitats would directlyaffect fitness.

• Determine the extent and magnitude oftreatments to apply to forested habitat.For example, testable research hypothesescould be that the extent of large trees insites affects foraging use by owls and preyabundance. Treatments may be nested sothat the same experimental units can beused in repeated experiments, assumingthat treatments can be decided uponbeforehand and applied consecutively. Inaddition, treatments need not be “nega-tive” by removing habitat componentsbut can be “positive” by treating previ-ously impacted habitats. Thus, carefulplanning is needed at this stage.

• Randomly select n spotted owl sites sothat sufficient power can be achieved todetect differences in foraging by owls

between treatments. Attach radio-transmitters to owls within sites.

• Collect pre-treatment data and definehigh-use areas by owls within all sites.

• Randomly assign treatment and controlclassifications to the n owl sites.

• Apply the treatment to high-use foragingareas within treatment sites only.

• Collect post-treatment data.

• Test for differences between treatmentand control groups.

• Continue the same procedure withadditional treatments.

While not ideal, such an experiment illus-trates the principles of scientific experimentaldesign necessary to achieve reliable knowledgeconcerning spotted owl habitat use and, indi-rectly, fitness, as a function of forest structure.Such crucial experiments are difficult to design,require commitments of funding, and scientificimagination because of ethical constraints andthe limitations on allowable habitat alterationsproposed in this plan. However, these types ofexperiments also more rapidly answer pressingmanagement questions.

We recommend research on the followingquestions about Mexican spotted owls that stillneed answers. Clearly, a large number of researchquestions could be developed that address allaspects of Mexican spotted owl biology forwhich knowledge is lacking. However, we posewhat we believe are the most crucial questionsthat need to be addressed in terms of immediatemanagement problems and the recovery of theowl. Studies designed to answer these questionswill be descriptive, experimental, or a combina-tion of both.

DispersalDispersalDispersalDispersalDispersal

Dispersal is a key process in metapopulationtheory and to maintain genetic diversity between

119


Volume I/Part III

isolated subpopulations (Keitt et al. 1995). Keyquestions include:

• Are subpopulations within and betweenRecovery Units connected?

• What habitats and large-scale habitatconfigurations do dispersing juvenilesrequire to maintain adequate survivalrates during dispersal?

GeneticsGeneticsGeneticsGeneticsGenetics

Mexican spotted owl populations are natu-rally fragmented across their range. Genetics canprovide insight into historical connectionsbetween subpopulations. Therefore, questionson genetics also relate to dispersal. Key questionsinclude:

• Are subpopulations within and betweenRecovery Units genetically isolated andto what degree?

• What is the extent of genetic interchangeacross the entire range of the owl?

HabitatHabitatHabitatHabitatHabitat

Mexican spotted owls use a variety ofhabitats ranging from canyons to forested areas(Ganey and Dick 1995). Key questions include:

• To what extent is habitat use determinedby various factors, such as prey availabil-ity, temperature regulation, and/oravoidance of predators?

• What habitat components confer highfitness?

• How do land management activities,specifically grazing, timber harvest, fire,and recreation use, proximately affecthabitat use and ultimately affect fitness?

Population BiologyPopulation BiologyPopulation BiologyPopulation BiologyPopulation Biology

Currently, little is known about Mexicanspotted owl populations. Key questions can beaddressed with our proposed monitoring plan:

• Is the Mexican spotted owl populationstable, increasing, or declining?

• Are some subpopulations increasingwhile others are decreasing within therange of the owl?

Threats to RecoveryThreats to RecoveryThreats to RecoveryThreats to RecoveryThreats to Recovery

Perceived threats need to be examined inrelation to current management strategies toexamine whether these strategies are appropriateand to develop appropriate management strate-gies. Key questions include:

• What management strategies can beemployed to reduce to possibility ofcatastrophic loss of owl habitat by firewhile maintaining important habitatcomponents?

• To what extent does disturbance fromrecreation, vehicles, etc. affect use of sitesby spotted owls?

• How does grazing affect prey abundancein habitats used by spotted owls forforaging?

Other Ecosystem ComponentsOther Ecosystem ComponentsOther Ecosystem ComponentsOther Ecosystem ComponentsOther Ecosystem Components

Implementation of the recovery measures forthe Mexican spotted owls will directly andindirectly affect numerous ecosystem attributes.Research is needed to determine the extent ofthese effects on biotic and abiotic components,and ecosystem processes and function. Keyquestions are:

• What are the effects of this recovery planon other vertebrates?

• What are the effects of implementing thePlan on nonvertebrates?

• What are the effects of implementing theRecovery Plan on plant communitystructure and composition?

• What are the effects of implementing theRecovery Plan on abiotic ecosystemprocesses (e.g., hydrological systems)?

• What are the effects of implementing theplan on ecosystem structure and func-tion?

• How might the recovery plan be adjustedto mitigate potentially deleterious effectson other ecosystem attributes?

120


Volume I/Part III

The ultimate goal of this Recovery Plan isto “recover” the Mexican spotted owl fromthreatened status. This action is referred to as“delisting” and is governed by section 4 of theAct. Delisting the Mexican spotted owl willrequire reexamination of the same five factorsconsidered during every listing process. Inaddition, five specific criteria have been devel-oped to aid the delisting determination. Three ofthese criteria pertain to the entire range of theowl and two refer to a recovery unit level. Therangewide delisting criteria are:

1. The populations in the Upper GilaMountains, Basin and Range-East, andBasin and Range - West RUs must beshown to be stable or increasing after 10years of monitoring, using a study designwith a power of 90% to detect a 20%decline with a Type I error rate of 0.05.

2. Scientifically-valid habitat monitoringprotocols are designed and implementedto assess (a) gross changes in habitatquantity across the range of the Mexicanspotted owl, and (b) whether microhabi-tat modifications and trajectories withintreated stands meet the intent of theRecovery Plan.

3. A long-term, U.S.-rangewide manage-ment plan is in place to ensure appropri-ate management of the subspecies andadequate regulation of human activityover time.

Once the above three criteria are met,delisting may occur in any RU that meets thefinal two criteria:

4. Threats to the Mexican spotted owlwithin the RU are sufficiently moderatedand/or regulated.

5. Habitat of a quality to sustain persistentMexican spotted owl populations isstable or increasing within the RU.

Recovery of the Mexican spotted owl hingeson successful implementation of three inter-related programs: population and habitat moni-toring, management guidelines, and research.These aspects are not intended to stand alone;thus, all programs must be implemented simul-taneously. For example, monitoring provides ameasure of the effectiveness of the managementguidelines. Without such monitoring, we willhave no basis for determining whether manage-ment guidelines lead to the desired outcomes,and thus whether the bird should be delisted.Research is needed to answer key questionsrelevant to the Mexican spotted owl, particularlyhow implementation of management recom-mendations will affect the Mexican spotted owland its habitat. The knowledge derived from thisresearch will provide a scientific basis for revisingshort-term guidelines and developing a long-term management plan.

We have proposed a quadrat samplingscheme and provide detailed considerations fordetermining spotted owl population trendswithin the Upper Gila Mountains, Basin andRange - East, and Basin and Range - West RUs.Population monitoring is not required for otherrecovery units because of sampling constraintsposed by smaller population sizes. The suggestedscheme provides a statistically valid means forassessing population change Initial cost esti-mates for the owl monitoring scheme will rangefrom $1.2 to $1.5 million per year.

Habitat monitoring is needed to estimatetrends in the quantity and quality of the owl’shabitat through time. Rangewide monitoring ofthe owl’s habitat should be conducted in con-junction with population monitoring. Becauseof the areal extent over which monitoring will berequired, we propose the use of satellite imageryfor tracking gross losses in habitat. We alsopropose that field sampling be conducted inconjunction with planned management treat-ments. Treatments include the use of prescribedfire, thinning, and silviculture. Monitoringshould be done prior to and immediately follow-ing the treatment, and then at five-year intervals.The objective of this sampling is to determine

E. SUMMARY OF RECOVERYE. SUMMARY OF RECOVERYE. SUMMARY OF RECOVERYE. SUMMARY OF RECOVERYE. SUMMARY OF RECOVERY

121


Volume I/Part III

changes to microhabitat features and also toverify that vegetation was placed or continues ona trajectory to become replacement habitat.

Threats to be moderated include those thatneed site-specific treatment to alleviate them,such as the reduction of the risks of catastrophicfire. For threats to be considered moderated,reasonable progress must have been made toremove identified threats and there must beadequate assurance that management programswill continue. Long-term management plans areneeded to guide management after the bird isdelisted. Threats to be regulated include thoseresulting from agency management programs orother anthropogenic activities that are eitherongoing or reasonably certain to occur. Thesetypes of threats include wildfire hazard, timberharvest, urban or rural land development,grazing, and recreation.

A primary focus of this Recovery Plan is toprovide recommendations that will moderate orregulate threats over the short term (10-15years). Conceptually, this requires the presenceof a mosaic of successional stages throughout alandscape comprised of the different habitatsused by Mexican spotted owls. The arrangementand diversity of these habitats must promote theowl’s persistence. The short-term strategy isaimed at protecting existing owl habitat andinitiating a process to develop replacementhabitat. Although the approach is not com-pletely synonymous with ecosystem manage-ment, implementation of the recommendationsshould sustain biotic diversity and naturalprocesses by managing several forest and wood-land systems used by the owl. Recommendationsinclude management of mixed-conifer and pine-oak forests, and riparian areas. Ponderosa pineand spruce-fir forests are also considered to alimited degree.

Several potential threats to the owl wereidentified by examining the best available infor-mation on the owl’s biology, and by evaluatingecological disturbance patterns and currentconditions throughout the owl’s range. Primarythreats include catastrophic fire, timber andfuelwood harvest, grazing, and recreation. Themagnitude of a threat’s influence on the owl canvary according to temporal and spatial setting.For this reason, general recommendations were

developed by habitat type which apply through-out the owl’s range, and are emphasized accord-ing to the magnitude of the threats within eachRU. The recommendations were also designedto provide different levels of protection depend-ing on the owl’s use of a particular habitat, thenature of the threats, and management potential.Our intent was to offer the most specific recom-mendations that the best available informationwould permit while allowing land managersflexibility for implementing the recommenda-tions.

Three areas of management are providedunder the general recommendations: protectedareas, restricted areas, and other forest andwoodland types. Protected areas receive thehighest level of protection. Recovery planguidelines take precedence over other manage-ment guidelines in protected areas. Guidelinesfor restricted areas are less specific and operate inconjunction with existing management guide-lines. Specific guidelines are not proposed forother forest and woodland types.

Protected areas are all occupied nest or roostareas, all areas with slope >40% where timberharvest has not occurred in the past 20 years,and all legally administered reserved lands.Protection of owl nest and roost areas will beestablished by designating an area of protectionaround an activity center (PAC). This willrequire (1) inventory of spotted owls beforeplanning any management activity that will alterstand structure; (2) delineating PAC areas of 243ha (600 ac) for all known Mexican spotted owlsites, including sites located prior to proposedmanagement activities; and (3) light burning inPACs if considered necessary and prudent toreduce risk of catastrophic loss. Further, a fireabatement program is proposed to allow treat-ment of small fuels within PACs and minimizeprobabilities of catastrophic fire. The purpose ofPACs is to provide refugia habitat until it can bedemonstrated reliably that owl habitat can becreated through management. In addition,harvest of trees <22.4 cm (9 inches) dbh is notallowed on slopes >40% where timber harvesthas not occurred in the past 20 years. However,light burning and prescribed natural fire man-agement is permitted. Prescribed natural fire is

122


Volume I/Part III

also encouraged on reserved lands (e.g., wilder-ness, Research Natural Areas) where appropriate.

We recommend that management activitiesbe restricted on some lands outside of protectedareas because patterns of owl use can be expectedto change over time. The guidelines dependupon forest or woodland type. Silviculturalprescriptions should emphasize measures to placestand conditions on a trajectory to become owlhabitat where appropriate. Stands that currentlymeet or exceed threshold conditions are subjectto more stringent restrictions than other stands.Specific management prescriptions should be sitespecific and will vary according to short- orlong-term objectives.

Short-term guidelines should not be miscon-strued as onetime management events. Forexample, large trees and snags are used by thespotted owl and will continue to be neededbeyond the life of the plan. Long-term guide-lines are recommended for those activities andnatural processes that combine to influence theowl and its habitat beyond the life expectancy ofthis Recovery Plan.

In addition, riparian communities should bemanaged by maintaining broad-leaved forests inhealthy condition where they occur, especially incanyon-bottoms. Restoration may be necessarywhere such forests are not regenerating ad-equately. Conceivably, restored riparian forestscould contribute additional nesting, winteringand dispersal habitat in the future. A mix ofplant size and age classes should be emphasizedin this community, to include large mature trees,vertical diversity, and other structural characteris-tics.

No specific guidelines are recommended inforest or woodland types not typically used bythe owl for nesting. These include ponderosapine, spruce-fir, pinyon-juniper, and quakingaspen in areas outside of PACs. However, somerelevant management of these communities mayproduce desirable results for owl recovery.Examples of guidelines include managing forlandscape diversity, mimicking natural distur-bance patterns, incorporating natural variationin stand conditions, retaining special featuressuch as snags, and utilizing fire in an appropriatemanner.

Livestock and wildlife grazing may influencespotted owls by altering (1) prey availability, (2)fire risk of some habitats, (3) riparian plantcommunities, and (4) development of spotted

owl habitat. The Team strongly advocates fieldmonitoring and experimental research related toimpacts of grazing on the Mexican spotted owl.Other specific guidelines include (1) monitoringgrazing use by livestock and key wildlife species(e.g. elk, deer), (2) implementing and enforcinggrazing utilization standards that attain good toexcellent range use standards, and (3) protectingor restoring riparian communities. These guide-lines are emphasized in protected, restricted, andriparian areas.

Several guidelines for managing recreation inprotected, restricted, and riparian areas arerecommended. These include: (1) no construc-tion, either of new facilities or for expandingexisting facilities, is allowed within PACs duringthe breeding season; (2) construction during thenonbreeding season should be considered on acase-specific basis; (3) managers should, on acase-specific basis, assess the presence andintensity of allowable recreational activitieswithin PACs; and (4) seasonal closures of specifi-cally designated recreation activities should beconsidered in extreme circumstances.

Several important questions regarding theowl’s ecology, and in particular about the effectsof different management activities on the owl’spopulation viability, still remain. The Teamrecommends additional research on Mexicanspotted owl dispersal, genetics, habitat ecology,and population biology. Key information that isvital for refining recovery strategies include (1)the degree of demographic and genetic isolationamong subpopulations; (2) the relationshipbetween fitness and specific habitat components;(3) population trend. Communication andcollaboration between researchers and managerswill be paramount for obtaining necessaryinformation.

This Recovery Plan presents realistic goalsfor recovery of the Mexican spotted owl and itsultimate delisting. The goals are flexible in thatthey allow local land managers to make site-specific decisions about management for recov-ery. The success of the recovery process hingeson commitment and coordination amongFederal and State land management agencies,sovereign Indian Nations, and the private sectorto ensure that the plan is followed and executedas intended by the Team.

123

Recovery Plan Implementation

Volume I, Part IV

Volume I/Part IV


Part IV discusses laws, regulations, policies,and authorities directly relevant to implementingthe recovery recommendations included in PartIII. An approach to implementation oversight isalso recommended. Finally, a stepdown outlineof recovery tasks and an implementation sched-ule are provided.

This Mexican Spotted Owl Recovery Plan isbased or predicated upon laws that designatespecific legal authority and responsibility togovernment agencies for managing publicresources, including wildlife and wildlife habitat.The following summarizes relevant laws andauthorities applicable to implementation of thisRecovery Plan.

ENDANGERED SPECIES ACTENDANGERED SPECIES ACTENDANGERED SPECIES ACTENDANGERED SPECIES ACTENDANGERED SPECIES ACT

Section 2(c)(2) of the Act expresses thepolicy of Congress that “...all Federal depart-ments and agencies shall seek to conserve endan-gered species and threatened species and shallutilize their authorities in furtherance of thepurposes of [the] Act.” Section 7(a)(1) of theAct requires Federal agencies to “...utilize theirauthorities in furtherance of the purposes of theAct by carrying out programs for the conserva-tion of endangered species and threatenedspecies....” Thus, Congress clearly intendedconservation of endangered and threatenedspecies to be considered in implementation ofFederal programs and actions. In addition, otherFederal laws and regulations require consider-ation of endangered and threatened species inprogram implementation, including the Na-tional Forest Management Act (NFMA) and theNational Environmental Policy Act (NEPA).

Implementation of the Act is the responsibil-ity of the Secretary of the Interior for listedterrestrial species. The Secretary generally del-egates implementation authority to the FWS.The following sections of the Act are relevant toimplementation of species recovery efforts:

Section 4Section 4Section 4Section 4Section 4

Section 4 includes the listing and recoveryprovisions of the Act, which are discussed indetail in Part I. Section 4(b) of the Act providesfor designation of critical habitat for endangeredand threatened species. Regulations governingcritical habitat designation are codified at 50CFR 424. Protection of critical habitat is admin-istered under section 7 of the Act (discussedbelow). Critical habitat is defined under section3(5)(A) of the Act as:

“(i) the specific areas within the geographi-cal area occupied by the species...onwhich are found those physical orbiological features (I) essential to theconservation of the species and (II)which may require special managementconsiderations or protection; and

(ii) specific areas outside the geographicalarea occupied by the species...upon adetermination by the Secretary that suchareas are essential for the conservation ofthe species.”

Section 4(d) of the Act provides for pro-mulgation of special rules for threatenedspecies only. This allows the Secretary to issueregulations as deemed necessary for the conserva-tion of such species. Special rules can be usefulin enacting regulatory provisions uniquelyapplicable to the species at hand, and can bepromulgated to avoid unnecessary regulatoryburden. For example, the FWS is considering aspecial 4(d) rule to allow small landowners in thePacific Northwest to harvest timber and conductother activities without risk of violating theprohibition of incidentally taking (see definitionunder Section 9, below) northern spotted owls.

A. IMPLEMENTING LAWS, REGULATIONS,A. IMPLEMENTING LAWS, REGULATIONS,A. IMPLEMENTING LAWS, REGULATIONS,A. IMPLEMENTING LAWS, REGULATIONS,A. IMPLEMENTING LAWS, REGULATIONS,AND AUTHORITIESAND AUTHORITIESAND AUTHORITIESAND AUTHORITIESAND AUTHORITIES

124

Volume I/Part IV



Section 5 directs the Secretary to utilizefunds and authorities of other laws in acquisitionof lands, as deemed appropriate for conservationof endangered and threatened species.


This section authorizes cooperation with theStates in conservation of threatened and endan-gered species. Among its provisions is the au-thority to enter into management agreementsand cooperative agreements and to allocate fundsto the States that have entered into such agree-ments.


Section 7 and its implementing regulationsat 50 CFR 402 govern cooperation betweenFederal agencies. Federal agencies must, inconsultation with and with the assistance of theSecretary, ensure that any action they fund,authorize, or carry out is not likely to jeopardizethe continued existence of listed species or resultin the destruction or adverse modification of alisted species’ designated critical habitat. Regula-tions at 50 CFR 402 provide the followingdefinitions:

“‘Jeopardize the continued existence of ’means to engage in an action that reason-ably would be expected, directly or indi-rectly, to reduce appreciably the likelihoodof both survival and recovery of a listedspecies in the wild by reducing the repro-duction, numbers, or distribution of thatspecies.”

“‘Destruction or adverse modification’means a direct or indirect alteration thatappreciably diminishes the value of criticalhabitat for both the survival and recoveryof a listed species.”

Section 7 requires action agencies to assessthe effects of proposed actions on listed speciesand their critical habitat. If, as a result of that

assessment, the agency determines that an actionmay affect a listed species or its critical habitat,the agency must enter into consultation with theFWS. That consultation may result in a biologi-cal opinion from the FWS, in which a determi-nation is made as to whether jeopardy to thespecies and/or destruction or adverse modifica-tion of its critical habitat are likely to result fromthe agency action.

If a biological opinion concludes that jeop-ardy to the species and/or adverse modificationof its critical habitat are not likely to result froma proposed action, the action may proceed. TheFWS may provide conservation recommenda-tions to the agency on ways to minimize oravoid potential adverse effects on listed speciesand/or critical habitat. Implementation of theseconservation recommendations is at the actionagencies’ discretion. In cases where the action islikely to result in the incidental taking of aspecies (see definition under “Section 9,” below),the Service may provide reasonable and prudentmeasures to minimize the amount or extent ofincidental take. The terms and conditions thataccompany and implement any reasonable andprudent measures are nondiscretionary and mustbe implemented. However, reasonable andprudent measures and their implementing termsand conditions cannot alter the basic design,location, scope, duration, or timing of theaction; and they may involve only minorchanges.

If a biological opinion determines thatjeopardy and/or adverse modification is likely toresult from a proposed action, the FWS and theaction agency develop reasonable and prudentalternatives, if any, to the proposed action.Reasonable and prudent alternatives refer toalternative actions that are consistent with theintended purpose of the proposed action, thatcan be implemented within the action agency’slegal authority, that are economically and tech-nologically feasible, and that the FWS believeswill not result in jeopardy to listed species ordestruction or adverse modification of criticalhabitat. If no reasonable or prudent alternativescan be identified, the action agency may apply tothe Endangered Species Committee for anexemption to the prohibition of jeopardy and/or

125

Volume I/Part IV


scribed above. In addition, section 10(a)(1)(B)allows permits for incidental taking that mayresult from an activity, provided an applicantsubmits a conservation plan that specifies:

“(i) the impact which will likely resultfrom such taking;

(ii) what steps the applicant will take tominimize and mitigate such impacts,and the funding that will be availableto implement such steps;

(iii) what alternative actions to such takingthe applicant considered and thereasons why such alternatives are notbeing utilized; and

(iv) such other measures that the [FWS]may require as being necessary orappropriate for purposes of the plan.”

NATIONAL FORESTNATIONAL FORESTNATIONAL FORESTNATIONAL FORESTNATIONAL FORESTMANAGEMENT ACTMANAGEMENT ACTMANAGEMENT ACTMANAGEMENT ACTMANAGEMENT ACT

The NFMA governs Forest Service Manage-ment on National Forest System lands. Section219.19 (Fish and wildlife resources) states:

“Fish and wildlife habitat shall be managedto maintain viable populations of exist-ing native and desired nonnative verte-brate species in the planning area. Forplanning purposes, a viable populationshall be regarded as one which has theestimated numbers and distribution ofreproductive individuals to ensure itscontinued existence is well distributed inthe planning area. In order to ensure thatviable populations will be maintained,habitat must be provided to support, atleast, a minimum number of reproduc-tive individuals and that habitat must bewell distributed so that those individualscan interact with others in the planningarea.”

In formulating alternatives during projectplanning, the following is required in regard tofish and wildlife habitat:

destruction or adverse modification of criticalhabitat.


Section 8 authorizes international coopera-tion in conservation of endangered and threat-ened species. Included under this section is theauthority to provide financial assistance toforeign countries to assist in their conservationefforts.


Section 9 covers prohibited acts in regard tolisted species. Of relevance to the Mexicanspotted owl is the prohibition of taking indi-viduals. “Take” is defined as “...to harass, harm,pursue, shoot, wound, kill, trap, capture, orcollect, or to attempt to engage in any suchconduct.” Permits for direct taking of threat-ened species may be issued for scientific pur-poses, to enhance propagation or survival, incases of economic hardship, for zoologicalexhibition, or for educational purposes (50 CFR17.32).

Taking of spotted owls is most likely tooccur through “incidental take.” “Incidentaltake” is defined as taking that results from, but isnot the purpose of, carrying out an otherwiselawful activity. Incidental taking of spotted owlsmay result from such activities as timber harvest,if that activity results in habitat loss to an extentthat an individual spotted owl’s normal behaviorpatterns are impaired. In cases where incidentaltaking will not result in jeopardy to a listedspecies, the FWS may issue an incidental takestatement in a biological opinion on a proposedFederal action, thereby removing the takeprohibition. Relief from the taking prohibitionfor non-Federal activities is discussed under“section 10” below.


Section 10 authorizes the FWS to issuepermits for takings otherwise prohibited undersection 9. Such permits may be issued for re-search purposes and the other situations de-

126

Volume I/Part IV


“Each alternative shall establish objectives forthe maintenance and improvement ofhabitat for management indicator species...tothe degree consistent with overall multiple-use objectives of the alternative. To meet thisgoal, management planning for the fish andwildlife resources shall meet the require-ments set forth [as follows:]

(1) In order to estimate the effects of eachalternative on fish and wildlife popula-tions, certain vertebrate and/or inverte-brate species present in the area shall beidentified and selected as managementindicator species and the reasons for theirselection will be stated. These speciesshall be selected because their populationchanges are believed to indicate theeffects of management activities. In theselection of management indicatorspecies, the following categories shall berepresented where appropriate:

* Endangered and threatened plant andanimal species identified on State andFederal lists for the planning area;

* Species with special habitat needs thatmay be influenced significantly byplanned management programs;

* Species commonly hunted, fished, ortrapped;

* Nongame species of special interest;and

* Additional plant or animal speciesselected because their populationchanges are believed to indicate theeffects of management activities onother species of selected major biologi-cal communities or on water quality.

“On the basis of available scientificinformation, the interdisciplinary teamshall estimate the effects of changes invegetation type, timber age classes,community composition, rotation age,and year-long suitability of habitat

related to mobility of managementindicator species. Where appropriate,measures to mitigate adverse effects shallbe prescribed.

(2) Planning alternatives shall be stated andevaluated in terms of both amount andquality of habitat and of animal popula-tion trends of the management indicatorspecies.

(3) Biologists from State fish and wildlifeagencies and other Federal agencies shallbe consulted in order to coordinateplanning for fish and wildlife, includingopportunities for the reintroduction ofextirpated species.

(4) Access and dispersal problems of hunt-ing, fishing, and other visitor uses shallbe considered.

(5) The effects of pest and fire managementon fish and wildlife populations shall beconsidered.

(6) Population trends of the managementindicator species will be monitored andrelationships to habitat changes deter-mined. This monitoring will be done incooperation with State fish and wildlifeagencies, to the extent practicable.

(7) Habitat determined to be critical forthreatened and endangered species shallbe identified, and measures shall beprescribed to prevent the destruction oradverse modification of such habitat.Objectives shall be determined forthreatened and endangered species thatshall provide for, where possible, theirremoval from listing as threatened andendangered species through appropriateconservation measures, including thedesignation of special areas to meet theprotection and management needs ofsuch species.”

127

Volume I/Part IV


NATIONAL ENVIRONMENTALNATIONAL ENVIRONMENTALNATIONAL ENVIRONMENTALNATIONAL ENVIRONMENTALNATIONAL ENVIRONMENTALPOLICY ACTPOLICY ACTPOLICY ACTPOLICY ACTPOLICY ACT

The NEPA requires Federal agencies toprepare Environmental Impact Statements (EIS)or Environmental Assessments (EA) for imple-mentation of agency actions and issuance ormodification of agency policies and guidance.Impacts of the proposed action or policy amend-ment on endangered and threatened speciesmust be evaluated. If a deciding official deter-mines that no significant impact will result froman action or policy amendment, a Finding of NoSignificant Impact is issued. If an agency deter-mines that a significant impact will result fromthe proposed action or policy amendment, anEIS must be prepared. An EIS addresses a rangeof alternatives. It is released for public reviewand comment, after which an alternative isselected and a Record of Decision is signed bythe deciding official.

MIGRATORY BIRD TREATY ACTMIGRATORY BIRD TREATY ACTMIGRATORY BIRD TREATY ACTMIGRATORY BIRD TREATY ACTMIGRATORY BIRD TREATY ACT

Prior to listing the Mexican spotted owl asthreatened, the Migratory Bird Treaty Act(MBTA) provided the only Federal protectionfor the subspecies other than that afforded byland-management agencies. Under the provi-sions of the MBTA, it is unlawful to pursue,hunt, take, capture, or kill in any manner anymigratory bird unless permitted by regulations.The MBTA applies in both the U.S. andMexico. Because the Mexican spotted owlexhibits migratory behavior in some areas it isincluded on the list of birds protected under theMBTA.

TRIBAL LANDSTRIBAL LANDSTRIBAL LANDSTRIBAL LANDSTRIBAL LANDS

The Recovery Team encourages adoption ofthe recovery recommendations by all Tribesadministering lands that support Mexicanspotted owl habitat. Tribal land-managementregulations and programs, including those forconservation of species, typically require enact-ment by Tribal Councils.

STATE AND PRIVATE LANDSSTATE AND PRIVATE LANDSSTATE AND PRIVATE LANDSSTATE AND PRIVATE LANDSSTATE AND PRIVATE LANDS

Although relatively few Mexican spottedowls are known on State and private lands, theTeam recommends that States continue and/orbegin a program to inventory forested areas forthe presence of Mexican spotted owls. TheRecovery Team is unaware of any State laws orregulations that govern management of spottedowl habitat on State or private lands. The Recov-ery Team recommends incorporating the recov-ery recommendations into State wildlife andforest practices laws and regulations. In addition,the Recovery Team encourages the FWS toevaluate the importance of State and privatelands to the Mexican spotted owl, and to con-sider promulgating a special rule under section4(d) of the Act that specifies habitat-alteringactivities that can be allowed on private landswithout violating the prohibition of incidentallytaking Mexican spotted owls.


The Recovery Team is unfamiliar with thelaws, regulations, and authorities that are avail-able or appropriate for implementing the recov-ery recommendations in Mexico. As recom-mended later in Part IV, the Recovery Teamexpects the FWS to arrange a meeting withMexican officials to discuss the Recovery Planand its implementation.

128

Volume I/Part IV


B. IMPLEMENTATIONB. IMPLEMENTATIONB. IMPLEMENTATIONB. IMPLEMENTATIONB. IMPLEMENTATIONOVERSIGHTOVERSIGHTOVERSIGHTOVERSIGHTOVERSIGHT

RECOVERY UNITRECOVERY UNITRECOVERY UNITRECOVERY UNITRECOVERY UNITWORKING TEAMSWORKING TEAMSWORKING TEAMSWORKING TEAMSWORKING TEAMS

The Team strongly recommends formationof interagency working teams whose responsibil-ity would be to oversee the implementation ofthe Recovery Plan. These Recovery Unit Work-ing Teams would coordinate with and report tothe Recovery Team, which would evaluate anyWorking Team recommendations before passingthem on to the FWS. Working Teams for eachU.S. Recovery Unit should be appointed by theFWS as subunits under the Recovery Teamumbrella. Recovery Team members may alsoserve on Recovery Unit Working Teams if thatarrangement is agreeable. Membership of theWorking Teams should include, at a minimum,one representative from each of the following:

1. Each involved FWS Ecological ServicesField Office

2. Each involved FS Region

3. Each involved State

4. Each involved Indian Reservation

5. Any other involved agency (e.g., BLM,NPS).

Each Working Team should have a researchscientist among its membership. That personmay be affiliated with one of the agencies listedabove, or may be independent. In addition tothe above, other interested persons approved bythe Recovery Team and the FWS should beallowed to participate if they so request. Suchparticipants may include a representative from aconservation organization, a representative fromthe timber or other affected industry, a represen-tative from an interested county or other localgovernment agency, and others as appropriate.Such a diverse membership would allow ideas ofvarying viewpoints to be discussed and wouldallow local interested parties to participate in

plan implementation and resolution of localissues. Working Teams for each Mexican Recov-ery unit should be similarly composed.

Once the FWS formulates a membershiplist, that list should be submitted to the Recov-ery Team for review. The Recovery Team wouldthen request the FWS’s Southwest RegionalDirector’s approval. Travel costs for each memberwould be borne by the member’s agency ororganization.

The functions of the Recovery Unit WorkingTeams should include the following:

1. Provide technical assistance to agenciesand landowners on such issues as projectdesigns, spotted owl management plandevelopment, and Recovery Plan compli-ance. The Recovery Team stronglyencourages conducting Recovery Planimplementation workshops to providebiologists, foresters, and other land-management personnel a commonworking knowledge of the provisions ofthis Recovery Plan. For example, a work-shop to develop procedures for delineat-ing PACs would encourage consistentapplication of recovery recommenda-tions. Specific workshop recommenda-tions are provided in IV.C.

2. Provide guidance and interpretation onimplementation of the recommendationscontained in this Recovery Plan.

3. Provide research assistance by procuringfinancial and logistic support, screeningresearch proposals for importance andrelevance, recommending to the Recov-ery Team prioritization of researchproposals, and other functions.

4. Recommend Recovery Plan revisionsbased on research results that mayenhance recovery efforts in thatspecific RU.

129

Volume I/Part IV


5. Prioritize areas to be inventoried withinthe RU.

6. Promote communication betweenvarious local interests and help resolveconflicting interpretations of the Recov-ery Plan provisions.

7. Monitor plan implementation andreport problems, successes, and generalrecovery progress to the FWS and theRecovery Team at least annually.

CONTINUING DUTIES OF THECONTINUING DUTIES OF THECONTINUING DUTIES OF THECONTINUING DUTIES OF THECONTINUING DUTIES OF THERECOVERY TEAMRECOVERY TEAMRECOVERY TEAMRECOVERY TEAMRECOVERY TEAM

The Recovery Team recommends that it becontinued throughout Recovery Plan implemen-tation. Once the final Recovery Plan is com-plete, the Recovery Team should meet at leasttwice per year for the first two years and annu-ally thereafter. The purpose of these meetingswould be to hear and discuss plan implementa-tion reports with the Recovery Unit WorkingTeams, and to report to the FWS on the progressof the recovery effort. The Team would alsoconsider recommendations from Recovery UnitWorking Teams and decide what recommenda-tions should be brought forward to the FWS aspotential revisions to the Recovery Plan.

CENTRALIZED SPOTTED OWLCENTRALIZED SPOTTED OWLCENTRALIZED SPOTTED OWLCENTRALIZED SPOTTED OWLCENTRALIZED SPOTTED OWLINFORMATION REPOSITORYINFORMATION REPOSITORYINFORMATION REPOSITORYINFORMATION REPOSITORYINFORMATION REPOSITORY

The Recovery Team recommends that acentral Mexican spotted owl data facility bemaintained throughout the life of the RecoveryPlan. The main purpose of such a facility wouldbe to house a spotted owl GIS database, includ-ing data assembled through the monitoringprogram, inventory program, and other pro-grams recommended in this Recovery Plan. Inaddition, the facility would maintain and peri-odically update a Mexican spotted owl bibliogra-phy.

Such a facility would be a valuable resourcefor biologists, land managers, researchers, andothers who may need information throughoutthe plan implementation period. Considerableinformation, assembled as a result of develop-ment of this plan, is already stored in a GISsystem maintained by the National BiologicalService’s Midcontinent Ecological ScienceCenter (formerly the National Ecology ResearchCenter) in Fort Collins, Colorado; continuanceof that arrangement is recommended by theTeam. In addition, a considerable “LiteratureCited” section is included in this plan, whichshould provide a good start to development of aMexican spotted owl bibliography.

130

Volume I/Part IV


C. STEPDOWN OUTLINEC. STEPDOWN OUTLINEC. STEPDOWN OUTLINEC. STEPDOWN OUTLINEC. STEPDOWN OUTLINE

This section lists specific tasks that need tobe implemented according to the recoveryrecommendations in Part III, plus Recovery Planoversight provisions discussed earlier in Part IV.This list is in a stepdown format, as required inthe FWS recovery planning guidelines. Each taskis also listed in Table IV.D.1, where the respon-sible parties for task implementation and theestimated costs of carrying out the tasks areprovided. Tasks are categorized as follows:

1. Resource Management Programs. Many ofthe recovery recommendations relate tospotted owl considerations that shouldbeincorporated into planning for otherresource management objectives such astimber harvest, recreation, and manage-ment of other species.

2. Active Management. These recoverytasks are to be implemented actively.They include forest health enhancementand protection, riparian restoration, anddevelopment of a long-term spotted owlmanagement plan.

3. Monitoring. These recommendationsrelate to monitoring the spotted owlpopulation and habitat.

4. Research. These recommendationsinclude research studies designed toincrease life-history knowledge of thesubspecies and to test the effects of landmanagement activities on spotted owls.

5. Oversight, Review, Evaluation, andRevision. These tasks are necessary tomonitor the Recovery Plan’s effectivenessand to determine if and when RecoveryPlan revision is necessary.

1.1.1.1.1. RRRRResouresouresouresouresource Mce Mce Mce Mce Management Panagement Panagement Panagement Panagement Prrrrrogramsogramsogramsogramsograms

11. Incorporate recovery recommendations(Part III) into land managementprograms.

111. Conduct the NEPA process toamend appropriate land manage-ment guidance and policy docu-ments (Federal lands).

1111. FS1112. BLM1113. NPS1114. DOD

112. Incorporate recovery recommenda-tions into Tribal management plans.

1121. White Mountain Apache1122. Mescalero Apache1123. San Carlos Apache1124. Navajo1125. Other tribes

113. Incorporate recovery recommenda-tions into State regulations pertain-ing to timber harvests and otheractivities on State and private lands.

1131. Arizona1132. New Mexico1133. Utah1134. Colorado

114. Incorporate recovery recommenda-tions into Mexican policydocuments.

1141. Arrange a meeting betweenFWS, Recovery Team, andMexican representatives todiscuss provisions of theRecovery Plan.

131

Volume I/Part IV


1142. Conduct actions necessaryto officially adopt RecoveryPlan recommendations intoMexican law and/or policy,as appropriate.

12. Conduct pre–project Mexican spotted owlinventories in project areas.

121. Federal agencies

1211. FS1212. BLM1213. NPS1214. DOD

122. Tribes


123. States1231. Arizona1232. New Mexico1233. Utah1234. Colorado

124. Mexico

2.2.2.2.2. AAAAActivctivctivctivctive Me Me Me Me Managementanagementanagementanagementanagement

21. Develop and/or implement foresthealth improvement and protectionprograms.

211. Federal lands

2111. FS2112. BLM2113. NPS2114. DOD2115. Other Federal agencies

212. Tribal lands

2121. White Mountain Apache2122. Mescalero Apache

2123. San Carlos Apache2124. Navajo2125. Other tribes

213. State and private lands


214. Mexico

22. Actively manage riparian habitat(e.g., restore degraded areas).

221. Lowland riparian

2211. BLM2212. State of Arizona2213. State of New Mexico2214. State of Utah2215. State of Colorado2216. Mexico

222. Middle to upper elevation riparian

2221. Federal lands

22211. FS22212. BLM22213. NPS22214. DOD22215. Other Federal

agencies

2222. Tribes

22221. White MountainApache

22222. Mescalero Apache22223. San Carlos Apache22224. Navajo22225. Other tribes

2223. Mexico

132

Volume I/Part IV


23. Develop and implement a long–term,range wide management plan.

231. Establish and support a Federal/Tribal/State/Mexican team todevelop the plan.

232. Develop a draft management plan.233. Conduct peer/public review.234. Produce final management plan.235. Develop appropriate implementa-

tion documents.

2351. Joint Federal agency EIS2352. White Mountain Apache2353. Mescalero Apache2354. San Carlos Apache2355. Navajo2356. Other tribes2357. State MOUs


2358. Mexico

3. M3. M3. M3. M3. Monitoronitoronitoronitoronitoringinginginging

31. Implement the population monitoringprogram detailed in Part III.

311. Secure funding for the entiremonitoring period (up to 15years).

312. Appoint a principle investigator.313. Develop detailed study method-

ology/protocols.314. Conduct Recovery Team/peer

review of program.315. Conduct a pilot study.316. Evaluate and revise methodology/

protocols.317. Implement the monitoring

program.

32. Implement the habitat monitoringprogram detailed in Part III.

321. Macrohabitat

3211. Acquire appropriate remotesensing imagery.

3212. Conduct necessary ground-truthing, imagery classifica-tion, geo-referencing, etc.

3213. Acquire remote sensingimagery at year 5.

3214. Conduct necessary ground-truthing, imager classifica-tion, geo-referencing, etc.

3215. Conduct change-detectionanalysis.

3216. Acquire remote sensingimagery at year 10.

3217. Conduct necessary ground-truthing, imagery classifica-tion, geo-referencing, etc.

3218. Conduct change-detectionanalysis.

322. Microhabitat (ongoing)

3221. Take pre-treatment mea-surements of relevanthabitat variables.

3222. Design treatment(s) toaccomplish spotted owlhabitat or other ecosystemmanagement goals.

3223. Conduct treatment3224. Take post-treatment mea-

surements at year 1 ofimportant habitat variables.

3225. Compare pre- and post-treatment data to determinewhether objectives oftreatment were met.

3226. Measure habitat variables atyear 5.

3227. Determine whether treatedstands are on appropriatetrajectories.

4.4.4.4.4. ResearchResearchResearchResearchResearch

41. Implement the research recommendationsoutlined in Part III.

411. Conduct dispersal studies.

133

Volume I/Part IV


4111. Examine connectivity ofsubpopulations within andbetween RUs.

4112. Determine habitat configu-rations that best facilitatedispersal and enhancesurvival rates of dispersingjuveniles.

412. Conduct studies on genetics.

4121. Determine whether and towhat degree subpopulationsare genetically isolated.

4122. Determine the extent andpatterns of gene flow acrossthe landscape.

413. Conduct habitat studies.

4131. Study the extent to whichhabitat use is influencedby prey availability,microclimatic factors, orpresence of predators.

4132. Determine which habitatcomponents influenceindividual fitness andpopulation persistence.

414. Study the effects of land-use prac-tices on spotted owls and/or spottedowl habitat.

4141. Determine the effects ofvarious silvicultural andtimber–harvest practices onspotted owl habitat.

4142. Determine the effects oflivestock and wildlifegrazing on spotted owlhabitat and prey.

4143. Determine the effects ofprescribed fire on spottedowl habitat and prey.

4144. Determine the effects ofrecreational activities onspotted owl habitat.

415. Study the effects of humandisturbance on spotted owls.

4151. Determine the effects ofnoise-producing activitieson nesting spotted owls.

4152. Determine the effects ofsuburban and rural develop-ment on habitats andpopulations of spotted owls.

416. Study the effects of Recovery Planimplementation on other ecosystemcomponents.

4161. Vertebrates and vertebratecommunities

4162. Invertebrates andinvertebrate communities

4163. Plants and plantcommunities

4164. Abiotic features(e.g. hydrological systems)

4165. Ecosystem structure andfunctioning

42. Conduct general inventories in areas thathave not previously been inventoried forspotted owls.

421. Federal lands

4211. FS4212. BLM4213. NPS4214. DOD4215. Other Federal agencies

422. Tribal lands


423. State and private lands

4231. Arizona4232. New Mexico

134

Volume I/Part IV


4233. Utah4234. Colorado

424. Mexico

43. Maintain a centralized Mexican spottedowl information facility.

431. Establish and maintain a Mexicanspotted owl GIS database.

4311. Establish and annuallyupdate spotted owl locationrecords and inventorycoverages.

4312. Establish and periodicallyupdate spotted owl habitatcoverages at varying spatialscales.

432. Develop and periodically update aspotted owl bibliography.

433. Distribute information to landmangers and others who request it.

5. Oversight, Review, Evaluation,5. Oversight, Review, Evaluation,5. Oversight, Review, Evaluation,5. Oversight, Review, Evaluation,5. Oversight, Review, Evaluation,and Revisionand Revisionand Revisionand Revisionand Revision

51. Oversee and monitor Recovery Planimplementation.

511. Conduct section 7 consultation onany Federal actions that may affectMexican spotted owls.

5111. Conduct a workshopbetween Recovery Team andFWS consultation biologistson evaluation of projectsfor Recovery Plancompliance.

5112. Consult programmaticallyon each agency’s incorpora-tion of the Recovery Planinto land managementpolicy and guidancedocuments.

5113. Review projects for compli-ance with the Recovery

Plan.

512. Form Recovery Unit WorkingTeams for each Recovery Unit.

5121. Appoint working Teammembers

5122. Develop charter, protocolsfor agreeing upon recom-mendations to be made toRecovery Team(e.g., voting protocols).

5123. Conduct training sessionwith Recovery Team toensure understanding andconsistent interpretation ofthe Recovery Plan.

5124. Conduct Recovery Planimplementation workshopswith biologists and otherland-managementpersonnel.

5125. Convene approximatelyquarterly or as needed.

5126. Working Team Leadersattend all Recovery Teammeetings.

513. Retain Recovery Team throughoutthe life of the Recovery Plan.

5131. Convene Recovery Teamsemi–annually for a minimum of two years afterRecovery Plan adoption.

5132. Convene Recovery Teamannually thereafter.

52. Oversee Research

521. Recovery Unit Working Teamsshould review and prioritize re-search proposals and make recom-mendations to the Recovery Team.

522. Recovery Unit Working Teamsshould annually update the FWSand the Recovery Team on plannedstudies.

135

Volume I/Part IV


53. Review, evaluate, and revise recovery planas appropriate.

531. Recovery Unit Working Teamsshould review plan implementationat least annually, reporting theresults to the FWS and theRecovery Team.

532. Recovery Unit Working Teamsshould review research and suggestplan revisions, if any, to the FWSand Recovery Team.

54. Recovery Unit Working Teams shouldprovide technical assistance whenrequested.

541. Provide land managers with techni-cal assistance in designing projectsto minimize impacts on spottedowls.

542. Provide technical assistance inprocuring funding and logisticsupport for research projects.

543. Provide technical assistance indeveloping spotted owl manage-ment plans.

544. Provide technical assistancein developing conservationagreements.

545. Provide other technical assistanceas needed.

55. Conduct Mexican spotted owl statusreviews.

56. State and private lands.

561. Conduct assessment of Mexicanspotted owl status on State andprivate lands.

562. Promulgate rule under 4(d) of theEndangered Species Act to providefor Mexican spotted owl conserva–tion on State and private lands.

136

Volume I/Part IV


Table IV.D.1 displays estimated costs andan approximate schedule for implementing therecovery tasks listed in the stepdown outlineprovided in IV.C. More detailed information onthe recommended actions is provided in Part III.The following material explains relevant detailsabout Table IV.D.1:

TTTTTaskaskaskaskask: This column lists specific tasks recom-mended in Part III. The format of this column issimilar to that used in the stepdown outline inIII.C, with each task under one of five generaltask categories (preceded by an Arabic numeral).In some cases “subtasks” are included if theRecovery Team wished to identify specificintermediate actions to accomplish an ultimateobjective. Please refer to the stepdown outline inIII.C for a more detailed description of eachtask. Part III provides yet more detail, such assuggested methodologies and rationales.

TTTTTask Nask Nask Nask Nask No.o.o.o.o.: This column lists the task numbersas developed in the stepdown outline (IV.C).

PPPPP: This column assigns priority numbers asfollows:

1: 1: 1: 1: 1: Tasks that must be completed to achievethe delisting criteria detailed in III.A.(Example: Population monitoring); tasksrequired by law (Example: Section 7consultation); and other tasks essential toRecovery Plan implementation (Example:amendment of agency planning docu-ments).

2:2:2:2:2: Tasks that should be done to help attainthe recovery objective. (Example: Restora-tion of degraded riparian areas).

3:3:3:3:3: Tasks that should be done to implementthe Recovery Plan efficiently or to other-wise enhance spotted owl management.(Example: general spotted owl inventory).

DurDurDurDurDur.....: The approximate duration (in years) ofeach task. Items that are expected to take less

than one year are assigned the number “1.”Tasks that are ongoing are labeled “cont.” (con-tinuous). Some tasks can be done to varyingdegrees or intensities, particularly researchprojects. In those cases, the duration is labeled“tbd” (to be determined).

RRRRResp. Pesp. Pesp. Pesp. Pesp. Pararararartytytytyty: Assigns lead responsibility of eachtask to a specific party. This does not necessarilymean that the indicated entity has sole responsi-bility for completion of a specific task; theRecovery Team recommends that agencies,Tribes, and others work cooperatively on recov-ery tasks whenever possible.

The following abbreviations are used:

AA = As appropriate1

Ac = Action agencyAll = All involved2

AZ = State of ArizonaBLM = Bureau of Land ManagementCO = State of ColoradoDOD = Department of DefenseFS = Forest ServiceFWS= Fish and Wildlife ServiceMA = Mescalero ApacheMEX = MexicoNAV = NavajoNM = State of New MexicoNPS = National Park ServicePI = Principle InvestigatorRT = Recovery TeamSCA = San Carlos Apachetbd = to be determinedUT = State of UtahWMA = White Mtn. ApacheWT = Working Team3

1 Used in situations such as under “Other Federalagencies.”

2 All parties involved in a cooperative effort, such as thepopulation monitoring program.

3 Used both for all Recovery Unit Working Teamscollectively, or for the appropriate WT for a RecoveryUnit.

D. IMPLEMENTATION ANDD. IMPLEMENTATION ANDD. IMPLEMENTATION ANDD. IMPLEMENTATION ANDD. IMPLEMENTATION ANDCOST SCHEDULECOST SCHEDULECOST SCHEDULECOST SCHEDULECOST SCHEDULE

137

Volume I/Part IV


Cost EstimatesCost EstimatesCost EstimatesCost EstimatesCost Estimates: The figures in this columnrepresent the estimated costs (x$1,000) ofcarrying out the recommended tasks in eachfiscal year (FY) indicated. Estimated costs arerounded to the nearest $1,000, i.e., a projectestimated at $200 will show “0”; a projectestimated at $500 will show “1”, etc. Some ofthe tasks assigned “NA” will be so labeledbecause no additional cost attributable to Mexi-can spotted owl recovery will be incurred. Theseinclude activities that are either already part ofland management programs or those that can bepaid for through commercial receipts (e.g. foresthealth enhancement/protection projects). Nocost estimates are given on tasks for Mexicobecause the Recovery Team was unable to obtainthe information.

Obviously, it is impossible to accuratelypredict the costs of many tasks. For example, thecost to carry out recommended research activi-ties can vary widely depending on the studydesign, the duration of the study, and otherfactors. Similarly, the fiscal year(s) under whichthe costs are placed may or may not be the fiscalyear in which the cost is actually incurred; again,it is impossible predict when a project will beundertaken. Finally, in cases such as pre-projectinventories, costs can only be estimated on a per-unit basis (e.g., $1.25/acre).

138

TTTTTable IVable IVable IVable IVable IV.D.1.D.1.D.1.D.1.D.1 Implementation and Cost Schedule

139

TTTTTable IVable IVable IVable IVable IV.D.1.D.1.D.1.D.1.D.1, continued

140


141


142


143


144


GLOSSARYGLOSSARYGLOSSARYGLOSSARYGLOSSARY

Acronyms and AbbreviationsAcronyms and AbbreviationsAcronyms and AbbreviationsAcronyms and AbbreviationsAcronyms and Abbreviations

Act - Endangered Species Act of 1973, asamended

AIC - Akaike’s Information CriteriaAK - Adaptive kernelAOU - American Ornithologists’ UnionANOVA - Analysis of variance: a statistical

evaluation procedureAZ - ArizonaBLM - Bureau of Land ManagementCJS - Cormack-Jolly-Seber: a population modelCO - ColoradoCSA - Coconino Study Area: a demographic

study areaDOD - Department of DefenseEA - Environmental AssessmentEIS - Environmental Impact StatementFEMAT - Forest Ecosystem Management

Assessment TeamFS - Forest Service (USDA Forest Service)FWS - Fish and Wildlife Service (U.S. Fish and

Wildlife Service; USDI Fish and WildlifeService)

FY - Federal budget fiscal year; 1 October to 30September

GIS - Geographic Information SystemGSA - Gila Study Area: a demographic study

areaHCA - Habitat Conservation AreaISC - Interagency Scientific CommitteeLMP - Land Management PlanLRT - Likelihood ratio testsLSR - Late Successional ReserveMANOVA - Multivariate analysis of varianceMCP - Minimum convex polygonMBTA - Migratory Bird Treaty ActNBS - National Biological ServiceNEPA - National Environmental Policy ActNFMA - National Forest Management ActNM - New MexicoNPS - National Park Service (USDI National

Park Service)PAC - Protected Activity CenterRU - Recovery UnitSISA - Sky Island Study AreaSOHA - Spotted Owl Habitat AreaSOMA - Spotted Owl Management Area

TES - Terrestrial Ecosystem SurveyUSDA - United States Department of Agricul-

tureUSDI - United States Department of InteriorUSGS - United States Geological SurveyUT - Utah

TTTTTererererermsmsmsmsms

Adaptive kernel (AK) - refers to a method ofestimating home-range size. This methodinvolves estimating a bivariate probabil-ity distribution from the observed animallocations, and can be used to computethe area containing a specified propor-tion of those locations.

Adaptive management - refers to a process inwhich policy decisions are implementedwithin a framework of scientificallydriven experiments to test predictionsand assumptions inherent in manage-ment plans.

Algorithm - a mathematical formula for solvinga problem.

Basal area - the cross-sectional area of a tree stemnear its base. Generally measured atbreast height (including bark).

Biomass - with respect to individuals, this refersto the weight (mass) of a plant or ananimal. With respect to areas or commu-nities, refers to the total mass of livingorganisms in that area or community atany given time. With respect to owl diet,used to refer to the relative contributionof one species (or group) of prey animalsto the overall diet.

Birth-pulse population - a population assumedto have a discrete point in time duringwhich all offspring are produced.

Bonferroni confidence interval - a family ofsimultaneous confidence intervals inwhich the width of each interval isadjusted downward to account for theestimation of simultaneous intervals.Basically, allows for multiple compari-sons without inflating the Type I errorrate.

145Volume I/Glossary


Bosque - a discrete grove or thicket of trees,particularly in lowland or riparian areasof the Southwestern United States andMexico; for example a cottonwoodbosque or a mesquite bosque.

Canopy - a layer of foliage, generally the upper-most layer, in a forest stand. Can be usedto refer to mid- or understory vegetationin multi-layered stands.

Canopy closure - an estimate of the amount ofoverhead tree cover (also canopy cover).

Clearcut - an area where the entire stand of treeshas been removed in one cutting.

Climax species - any species that is characteristicof a plant community that throughnatural processes reaches the apex of itsdevelopment after sufficient time. Theopposite of seral species.

Closed population - a population that receivesno immigrants from other populations,and from which no individuals emigrateto other populations.

Cohort - individuals of the same age, resultingfrom the same birth-pulse.

Commercial forest land - forested land deemedtentatively suitable for the production ofcrops of timber, that has not beenwithdrawn administratively from timberproduction (see reserved land).

Confidence interval - an interval constructedaround a parameter estimate, in whichthat estimate should occur with a speci-fied probability, such as 95% of the time.

Connectivity - an estimate of the extent towhich intervening habitats connectsubpopulations of spotted owls.

Cordillera - a mountain range or chain.Cordilleran - of or relating to a range of moun-

tains.Dbh - diameter at breast height, a standard

measure of tree size.Demography - the quantitative analysis of

population structure and trend.Demographic stochasticity - fluctuations in

population size driven by randomfluctuations in birth and death rates.

Dispersal - The movement of organisms fromtheir birth place to another locationwhere they produce offspring.

Disturbance - significant alteration of habitatstructure or composition. May be natural(e.g. fire) or human-caused events (e.g.timber harvest).

Early seral stage - an area that is in the earlystages of ecological succession.

Ecological succession - the orderly progression ofan area through time from one vegetativecommunity to another in the absence ofdisturbance. For example, an area mayproceed from grass-forb through aspenforest to mixed-conifer forest.

Ecosystem - an interacting biophysical system oforganisms and their environment.

Emigration - permanent movement of individu-als away from a population.

Encinal - of or relating to oaks, particularly plantcommunities dominated by live oaks.

Environmental stochasticity - random variationin environmental attributes, such asweather patterns or fire regimes.

Even-aged forest - used to refer to forests com-posed of trees with a time span of <20yrs between oldest and youngest indi-viduals.

Even-aged management - the application of acombination of actions that result in thecreation of stands in which trees areessentially all of the same age. Cuttingmethods that produce even-aged standsinclude clearcuts, seed-tree cuts, andshelterwood cuts.

Fecundity - a statistical parameter of productiv-ity determined by the number of same-gender offspring produced by each adultin a population. Thus, either maleoffspring produced per male adult orfemale offspring produced per femaleadult is a measure of fecundity.

Fire regime - a description of the frequency,severity, and extent of fires that typicallyoccur in an area or vegetation type.

Floater - a member of a spotted owl populationthat does not hold, maintain, or defend aterritory (see Franklin 1992).

Forb - a broadleaved, herbaceous plant; forexample, columbine.

Fragmentation - the process of reducing the sizeand connectivity of habitat patches.



Fuel ladder - dead or living fuels that connectfuels on the forest floor to the canopy,and promote the spread of surface firesto tree crowns.

Fuel loads - the amount of combustible materialpresent per unit area.

Fuels - combustible materials.Fuelwood - wood, either green or dead, har-

vested for purposes of cooking or spaceheating, and usually measured in cords.(1 cord = 128 cubic feet.)

Geographical Information System (GIS) - acomputer system capable of storing andmanipulating spatial data.

Gene flow - the movement of genetic materialamong populations.

Genetic stochasticity - random changes in genefrequencies within a population, mayresult from factors such as inbreedingand mutation.

Graminoids - any plants of the grass family inparticular and also those plants in otherfamilies that have a grass-like form orappearance (for example, sedges).

Group-selection cutting - removal during harvestof groups of trees.

Habitat - suite of existing environmental condi-tions required by an organism for sur-vival and reproduction. The place wherean organism typically lives.

Habitat fragmentation - see fragmentation.Habitat mosaic - the mixture of habitat condi-

tions across a landscape.Habitat type - see vegetation type.Hanging canyon - a side canyon, the mouth of

which lies above the floor of a largercanyon to which it is tributary.

Home range - the area used by an animal in itsday-to-day activities.

Immigration - the movement of individualsfrom other areas into a given area.

Intermountain Region - an administrative regionof the USDA Forest Service, lyingbetween the Pacific Coastal and RockyMountain Ranges and including Utah,Nevada, southern Idaho, and parts ofWyoming and Montana.

Lambda - the finite rate of change in populationsize. If lambda is greater than 1, thepopulation trend is increasing; if lambda

equals 1, the population trend is stable;if lambda is less than 1, the populationtrend is decreasing.

Land Management Plan (LMP) - a plan writtenfor the management of a NationalForest. These plans were mandated bythe National Forest Management Act of1976.

Late seral stage forest - a forest in the latter stagesof development, usually dominated bylarge, old trees.

Leslie matrix - a two-dimensional array ofnumbers representing age- or stage-specific estimates of birth and deathrates, used to project population age (orstage) structure through time.

Life table - mathematical table of age- or stage-specific birth and death rates of a popu-lation.

Macrohabitat - landscape-scale features that arecorrelated with the distribution of aspecies; often used to describe seral stagesor discrete arrays of specific vegetationtypes.

Madrean - pertaining to Mexico’s Sierra Madrecordillera, or to plant species or commu-nities whose primary affinity is to thatregion (see also Petran).

Madrean pine-oak forest - forests in which anyof several pines characterize the over-story, and midstory oaks are mostlyevergreen species. Many of the dominantspecies are Madrean in affinity. SeeMarshall (1957) for descriptions. Thishabitat was included as Pine-oak byFletcher and Hollis (1994).

Mesic - of or relating to conditions betweenhydric and xeric or the specific quality ofbeing adapted to conditions between wetand dry.

Metapopulation - systems of local populationsconnected by dispersing individuals.

Microhabitat - habitat features at a fine scale;often identifies a unique set of localhabitat features.

Microtine - any vole of the genus Microtus.Migration - the seasonal movement from one

area to another and back.Minimum convex polygon (MCP) - a method

used to estimate home-range size. This



method involves forming a polygon byconnecting the outermost animal loca-tions with a series of convex lines, thencomputing the area of that polygon.

Mixed-conifer forest type - overstory species inthese forests include Rocky MountainDouglas-fir, white fir, Rocky Mountainponderosa pine, quaking aspen, south-western white pine, limber pine, andblue spruce. Refer to II.C for a moreprecise discussion and definition ofmixed-conifer forest type.

Model - a representation of reality, based on a setof assumptions, that is developed andused to describe, analyze, and understandthe behavior of a system of interest.

Monitoring - the process of collecting informa-tion to track changes of selected param-eters over time.

Mousing - a technique used to assess reproduc-tive status of a pair of spotted owls.Entails feeding mice to adult owls andobserving the owls’ subsequent behavior.

Multi-layered (or multi-storied) stands - foreststands with >2 distinct canopy layers.Applied to forest stands that containtrees of various heights and diameters,and therefore support foliage at variousheights in the vertical profile of thestand.

Null hypothesis - a hypothesis stating that thereis no difference between units beingcompared.

Other forest and woodland types - vegetationtypes that are neither “restricted” orwithin PACs (see definitions of thoseterms) as to management recommenda-tions provided in this Recovery Plan.

Old growth - an old forest stand, typicallydominated by large, old trees, withrelatively high canopy closure and a highincidence of snags, as well as logs andother woody debris.

Overstory - the highest limbs and foliage of atree, and consequently extending andrelating to the upper layers of a forestcanopy.

Pellet - a compact mass of undigested materialremaining after preliminary digestionand eliminated by regurgitation rather

than by defecation.Peromyscid - any mouse in the genus Peromyscus

of the family Muridae (formerlyCricetidae).

Petran - pertaining to the Rocky Mountain area.Used to identify plant associations orspecies that have their primary affinity tothe Rocky Mountain area (see alsoMadrean).

Physiographic province - a geographic region inwhich climate and geology have givenrise to a distinct array of land forms andhabitats.

Pine-oak forest type - stands within the Pinusponderosa and Pinus leiophylla series thatexhibit a pine overstory and oak under-story. Refer to II.C for these criteria anda more precise discussion and definitionof pine-oak forest type.

Precommercial thinning - the practice of remov-ing some of the smaller trees in a standso that remaining trees will grow faster.

Prescribed fire - a fire burning under specifiedconditions; may result from eitherplanned or unplanned ignitions.

Ponderosa pine forest type - any forested standof the Pinus ponderosa Series not in-cluded in the pine-oak forest typedefinition, or any stand that qualifies aspure (i.e., any stand where a singlespecies contributes >80 % of the basalarea of dominant and codominant trees)ponderosa pine, regardless of the series orhabitat (see also Eyre 1980). Refer toPart II.C for a more precise discussionand definition of ponderosa pine foresttype.

Population - a collection of individuals thatshare a common gene pool.

Population density - the number of individualsper unit area.

Population persistence - the capacity of a popu-lation to maintain sufficient numbersand distribution over time.

Population viability - the probability that apopulation will persist for a specificperiod of time, despite demographic andenvironmental stochasticity.

Power - with respect to statistical comparisons,refers to the probability of not making a



Type-II error.Protected Activity Center (PAC) - an area

established around an owl nest (orsometimes roost) site, for the purpose ofprotecting that area. Management ofthese areas is largely restricted to manag-ing for forest health objectives.

Protected areas - as used in this Plan, refers toareas that are protected, and where mostmanagement activities are very restrictedor disallowed. Includes Protected Activ-ity Centers.

Recovery - as provided by the EndangeredSpecies Act and its implementing regula-tions, the process of returning a threat-ened or endangered species to the pointat which protection under the Endan-gered Species Act is no longer necessary.

Recovery Plan - as provided by the EndangeredSpecies Act, a plan for management of athreatened or endangered species thatlays out the steps necessary to recover aspecies (see “Recovery”).

Recovery Team - a team of experts appointed bythe Fish and Wildlife Service whosecharge is development of a RecoveryPlan (see “Recovery Plan”).

Recovery Unit (RU) - a specific geographic area,identified mainly from physiographicprovinces, used to evaluate the status ofthe Mexican spotted owl.

Recruitment - the addition of individuals to apopulation from birth and immigration.

Reserved lands - lands that have been adminis-tratively withdrawn from commercialactivities, such as wilderness areas orresearch natural areas.

Restricted Areas - as used in this Plan, refers toareas that are not protected (see Pro-tected Areas), but where specific guide-lines for management activities areproposed.

Riparian - of or relating to a river; specificallyapplied to ecology, “riparian” describesthe land immediately adjoining anddirectly influenced by streams. Forexample, riparian vegetation includes anyand all plant-life growing on the landadjoining a stream and directly influ-enced by that stream.

Riparian forests - forests along rivers, streams,and other wetland environments, typi-cally characterized by the presence ofriparian-obligate plants such as cotton-woods, willows, sycamores, or alders.Descriptions are provided by Dick-Peddie (1993) and others.

Rocky Mountain Region - An administrativeregion of the USDA Forest Service,including Colorado, Nebraska, SouthDakota, and parts of Wyoming.

Rotation - the planned number of years betweenregeneration of a forest stand and finalharvest of that stand.

Salvage - see sanitation salvage.Sanitation salvage - removal of dead, damaged,

or susceptible trees primarily to preventthe spread of pests or pathogens and topromote forest health.

Seed-tree cut - an even-aged regeneration cuttingin which only a few seed trees are re-tained per hectare. Shelterwood cutsretain more seed trees.

Seral species - any plant or animal that is typicalof a seral community (stage).

Seral stage - Any plant community whose plantcomposition is changing in a predictableway; for example, an aspen communitychanging to a coniferous forest commu-nity.

Shelterwood cut - an even-aged regenerationcutting in which new tree seedlings areestablished under the partial shade ofremnant seed trees.

Silviculture - the practice of controlling theestablishment, composition, and growthof forests.

Single-tree selection cutting - a cutting methodbased on removal of individual trees,rather than groups of trees (see alsogroup selection cutting).

Sink - in a population sense, refers to a popula-tion whose death rate exceeds its birthrate. Such a population is maintained byimmigration from other populations (seesource), and is not expected to contrib-ute to long-term population mainte-nance.



Slash - the residue left on the ground afterlogging, including logs, uprootedstumps, branches, twigs, leaves, and bark.

Southwestern Region - an administrative unit ofthe USDA Forest Service, includingArizona and New Mexico; and an ad-ministrative unit of the USDI Fish andWildlife Service , including Arizona,New Mexico, Texas and Oklahoma.

Snag - a standing dead tree.Source - in a population sense, refers to a popu-

lation where birth rate exceeds deathrate. Such a population produces anexcess of juveniles that can disperse toother populations (see sink).

Spruce-fir forest type - high-elevation forestsoccurring on cold sites with short grow-ing seasons, heavy snow accumulations,and strong ecological and floristicaffinities to cold forests of higher lati-tudes. In general, dominant trees includeEnglemann spruce, subalpine and/orcorkbark fir, or sometimes bristleconepine. Refer to Part II.C for a moreprecise discussion and definition ofspruce-fir forest type.

Stand - any homogeneous area of vegetationwith more or less uniform soils, land-form, and vegetation. Typically used torefer to forested areas.

Stochastic - random or uncertain.Stringers - narrow bands of trees that extend into

confined areas of suitable habitat such asin ravines.

Subpopulation - a well-defined set of individualsthat comprises a subset of a larger,interbreeding population (see alsometapopulation).

Survivorship - the proportion of newbornindividuals that are alive at any givenage.

Team - the Mexican Spotted Owl RecoveryTeam

Technical Team - the Utah Technical Team; aninteragency team charged with providingmanagement suggestions for the Mexicanspotted owl.

Terrestrial Ecosystem Survey (TES) - a system ofecosystem classification, inventory,mapping, and interpretation based upon

terrestrial vegetation and environmentalfactors, used by the USDA Forest Ser-vice, Southwestern Region. Ecosystemsare defined by combinations of potentialvegetation, soils, and climates. Land ispartitioned into mapping units basedupon inventory data, classification, andair photo interpretation.

Territory - the area that an animal defendsagainst intruders of its own species. Notsynonymous with home range, as partsof the home range are typically sharedwith other individuals.

Toe clipping - a procedure by which smallanimals are captured alive and marked asindividuals for later recapture recogni-tion by clipping off portions of one ormore toes in unique combinations.

Trap-night - a standardized measurement oftrapping effort in wildlife studies; equalsone trap set for night. For example, onetrap set for 10 nights and 10 traps set forone night both equal 10 trap-nights.

Turnover - in a population sense, refers to therate at which individuals that die arereplaced by other individuals.

Type-I error - the error made when a nullhypothesis that is true is inappropriatelyrejected, as when concluding that twosamples from a single population comefrom two different populations.

Type-II error - the error that is made when a nullhypothesis that is false is not rejected, aswhen concluding that two samples fromdifferent populations came from a singlepopulation.

Understory - any vegetation whose canopy(foliage) is below, or closer to the groundthan, canopies of other plants. Theopposite of overstory.

Uneven-aged management - the application of acombination of actions needed to simul-taneously maintain continuous tall forestcover, recurring regeneration of desirablespecies, and the orderly growth anddevelopment of trees through a range ofdiameter or age classes. Cutting methodsthat develop and maintain uneven-agedstands are single-tree selection and groupselection.



Vegetation types - a land classification systembased upon the concept of distinct plantassociations. Vegetation or habitat types(plant associations) have been docu-mented for western forests, and keys totheir identification are available. Theprimary vegetation (or habitat) typesused by Mexican spotted owls are dis-cussed in II.C.

Viability - ability of a population to persistthrough time (see population viability).

Vital rates - collective term for age- or stage-

specific demographic rates, such as birthand death rates, of a population.

Vole - any small rodent in the genus Microtus,Clethrionomys, or Phenacomys, all in thefamily Muridae.

Witches broom - a mass of profuse and denselypacked twigs representing abnormalgrowth of a tree branch. Often resultsfrom infection by dwarf mistletoe.

Xeric - of or relating to perennially dry condi-tions or the specific quality of beingadapted to dry conditions.



Alden, P. 1969. Finding the birds in westernMexico: a guide to the states of Sonora,Sinaloa, and Nayarit. Univ. Arizona Press,Tucson. 138pp.

Alexander, B. G., Jr., and F. Ronco, Jr. 1987.Classification of the forest vegetation on theNational Forests of Arizona and New Mexico.USDA For. Serv. Res. Note RM-469. RockyMtn. For. and Range Exper. Stn. Fort Collins,Colo.

——., ——., E. L. Fitzhugh, and J. A. Ludwig.1984a. A classification of forest habitat typesof the Lincoln National forest, NewMexico.USDA For. Serv. Gen. Tech. Rep.RM-104. Rocky Mtn. For. and Range Exper.Stn. Fort Collins, Colo.

——., ——., A. S. White, and J. A. Ludwig.1984b. Douglas-fir habitat types of northernArizona. USDA For. Serv. Gen. Tech. Rep.RM-108. Rocky Mtn. For. and Range Exper.Stn. Fort Collins, Colo.

American Ornithologists’ Union. 1957. Check-list of North American birds. Fifth ed. Am.Ornithologists’ Union, Washington, D.C.691pp.

Ames, C. R. 1977. Wildlife conflicts in riparianmanagement: grazing. Pages 49-51 in R. R.Johnson and D. A. Jones, eds. Importance,preservation and management of riparianhabitat: a symposium. U.S. For. Serv. Gen.Tech. Rep. RM-43, Ft. Collins, Colo.

Archambault, S., T. W. Swetnam, and A. M.Lynch. 1994. Western spruce budwormoutbreak history in the SacramentoMountains, New Mexico, U.S.A.Unpublished manuscript.

Arizona Game and Fish Department. 1988.Threatened native wildlife in Arizona. Ariz.Game and Fish Dep., Phoenix.

Attiwill, P. M. 1993. The disturbance of forestecosystems: the ecological basis forconservative management. For. Ecol. andManage. 63:247-300.

Bailey, R. G. 1980. Descriptions of theecoregions of the United States. USDA For.Serv. Misc. Pub. 1391, Intermountain Region,Ogden, Utah. 77pp.

LITERATURE CITEDLITERATURE CITEDLITERATURE CITEDLITERATURE CITEDLITERATURE CITED

Barrowclough, G. F., and R. J. Gutiérrez. 1990.Genetic variation and differentiation in thespotted owl (Strix occidentalis). Auk 107:737-744.

Barrows, C. W. 1981. Roost selection by spottedowls: an adaptation to heat stress. Condor83:302-309.

——. 1987. Diet shifts in breeding andnonbreeding spotted owls. J. Raptor Res.21:95-97.

Bart, J., R. G. Anthony, M. Berg, J. H. Beuter,W. Elmore, J. Fay, R. J. Gutiérrez, T. H.Heintz, Jr., R. S. Holtshausen, K. Lathrop, K.Mays, R. Nafziger, M. Pagel, C. Sproul, E. E.Starkey, J. C. Tappeiner, and R. Warren. 1992.Recovery plan for the northern spotted owl.Final draft. USDI Fish and Wildl. Serv.,Portland, Oregon. 194pp.

Beschta, R. L., C. A. Frissell, R. Gresswell, R.Hauer, J. R. Karr, G. W. Minshall, D. A.Perry, J. J. Rhodes. 1995. Wildfire and salvagelogging. Unpubl. Rep., Corvallis, Oreg.

Blackburn, W. E. 1984. Impacts of grazingintensity and grazing systems on vegetationcomposition and production. Pages 927-984in National Research Council and NationalAcademy of Sciences. Developing strategiesfor rangeland management: a report preparedby the committee on developing strategies forrangeland management. Westview Press,Boulder, Colo.

Block, W. M., and L. A. Brennan. 1993. Thehabitat concept in ornithology: theory andapplication. Current Ornith. 11:35-91.

——., D. M. Finch, and L. A. Brennan. 1995.Single-species versus multiple-speciesapproaches for managment. Pages 461-476 inT. E. Martin and D. M. Finch, eds., Ecologyand management of neortropical migratorybirds: a synthesis and review of critical issues.Oxford Univ. Press, Oxford, UK.

——., K. A. With, and M. L. Morrison. 1987.On measuring bird habitat: influence ofobserver variability and sample size. Condor89:241-251.

Bock, C. E., V. A. Saab, T. D. Rich, and D. S.Dobkin. 1993. Effects of livestock grazing onneotropical landbirds in western North

152Volume I/Literature Cited


America. Pages 296-309 in D. M. Finch andP. W. Stangel, eds. Status and management ofneotropical migratory birds. USDA For. Serv.Gen. Tech. Rep. RM-229, Ft. Collins, Colo.

Brown, D. E., ed. 1982. Biotic communities ofthe American southwest-United States andMexico. Desert Plants 4:1-342.

——., C. H. Lowe, and C. P. Pase. 1980. Adigitized systematic classification forecosystems with an illustrated summary of thenatural vegetation of North America. USDAFor. Serv. Gen. Tech. Rep. RM-73, Ft.Collins, Colo. 93pp.

Bryant, L. D. 1982. Response of livestock toriparian zone exclusion. J. Range Manage.35:780-785.

Bull, E. L., R. S. Holthausen, and D. B. Marx.1990. How to determine snag density. West.J. Appl. For. 5:56-58.

Burnham, K. P., and D. R. Anderson. 1992.Data-based selection of an appropriatebiological model: the key to modern dataanalysis. Pages 16-30 in D. R. McCulloughand R. H. Barrett, eds. Wildlife 2001:populations. Elsevier Applied Science, NewYork, N.Y.

——., and W. S. Overton. 1978. Estimation ofthe size of a closed population when captureprobabilities vary among animals. Biometrika65:625-633.

——., and ——. 1979. Robust estimation ofpopulation size when capture probabilitiesvary among animals. Ecology 60:927-936.

——., D. R. Anderson, and G. C. White. 1994.Estimation of vital rates of the northernspotted owl. Colo. Coop. Fish and Wildl. Res.Unit, Colorado State Univ., Fort Collins.44pp.

——., ——., ——., C. Brownie, and K. H.Pollock. 1987. Design and analysis methodsfor fish survival experiments based on release-recapture. Am. Fisheries Soc. Monogr. 5:1-437.

Burt, W.H. 1943. Territoriality and home rangeconcepts as applied to mammals. J. Mammal.24:346-352.

Carey, A. B. 1985. A summary of the scientificbasis for spotted owl management. Pages100-114 in R. J. Gutiérrez and A. B. Carey,tech. eds. Ecology and management of the

spotted owl in the Pacific Northwest. USDAFor. Serv. Gen. Tech. Rep. PNW-185. PacificNorthwest Res. Stn. Portland, Oregon.119 pp.

Chao, A. 1987. Estimating the population sizefor capture-recapture data with unequalcatchability. Biometrics 43:783-791.

——. 1988. Estimating animal abundance withcapture frequency data. J. Wildl. Manage.52:295-300.

——. 1989. Estimating population size forsparse data in capture-recapture experiments.Biometrics 45:427-438.

Choate, G. A. 1966. New Mexico’s forestresource. USDA For. Serv. Resour. Bull. INT-5. Intermountain Res. Stn. Ogden, Utah.

Cirett-Galan, J. M., and Diaz, N. 1993. Estatusy distribución del búho manchado Mexicano(Strix occidentalis lucida) en Sonora, México.Centro Ecológico de Sonora, Hermosillo.66pp.

Clary, W. P., and B. F. Webster. 1989. Managinggrazing of riparian areas in the IntermountainRegion. USDA For. Serv. Gen. Tech. Rep.INT-263, Ogden, Utah. 11pp.

Colhoun, J. 1979. Predisposition by theenvironment. Pages 75-96 in J. G. Horsfalland E. B. Cowling, eds. Plant disease, anadvanced treatise. Vol. IV: How pathogensinduce disease. Academic Press,New York, N.Y.

Connor, R. C., J. D. Born, A. W. Green, and R.A. O’Brien. 1990. Forest resources of Arizona.USDA For. Serv. Resour. Bull. INT-69,Ogden, Utah.

Cooper, C. F. 1960. Changes in vegetation,structure, and growth of southwestern pineforests since white settlement. Ecol. Monogr.30:129-164.

Cooperrider, A. 1991. Conservation ofbiodiversity on western rangelands. Pages 40-53 in W. E. Hudson, ed. Landscape linkagesand biodiversity. Island Press, Washington,D.C.

Covington, W. W., and M. M. Moore. 1992.Postsettlement changes in natural fire regimes:implications for restoration of old-growthponderosa pine forests. Pages 81-99 in M. R.Kaufmann, W. H. Moir, and R. L. Bassett,eds. Old-growth forests in the Southwest and



Rocky Mountain regions: proceedings of aworkshop. U.S. For. Serv. Gen. Tech. Rep.RM-213, Ft. Collins, Colo.

——., and ——. 1994a. Southwesternponderosa forest structure: changes sinceEuro-American settlement. J. For.92(1):39-47.

——., and ——. 1994b. Postsettlement changesin natural fire regimes and forest structure:ecological restoration of old-growthponderosa pine forests. J. Sustainable For.2:153-181.

——., and ——. 1994c. Postsettlement changesin natural fire regimes and forest structure:ecological restoration of old-growthponderosa pine forests. Pages 153-181 in R.N. Sampson and D. L. Adams, eds. Assessingforest ecosytem health in the inland west.Haworth Press, Inc.

Cuanalo de la Cerda, H., E. Ojeda-Trejo, A.Santos-Ocampo, and C. A. Ortíz-Solorio.1989. Provincias, regiones y subrregionesterrestres de México. Colegio dePostgraduados, Centro de Edafología,Chapingo, México.

Curtis, B. F. 1960. Major geologic features ofColorado. Pages 1-30 in Geol. Soc. of Amer.,Rocky Mtn. Assoc. of Geol., Colo Sci. Soc.

Daniel, T. W., J. A. Helms, and F. S. Baker.1979. Principles of silviculture. Second ed.McGraw-Hill, New York, N.Y. 500pp.

Daubenmire, R. 1943. Vegetation zonation inthe Rocky Mountains. Bot. Rev. 9:325-393.

——. 1952. Forest vegetation of northern Idahoand adjacent Washington, and its bearing onconcepts of vegetation classification. Ecol.Monogr. 22:301-330.

——. 1968. Plant communities: a textbook ofplant synecology. Harper and Row. NewYork, N.Y. 300pp.

Davis, M. B. 1989. Lags in vegetation responseto greenhouse warming. Climatic Change15:75-82.

Dawson, W. R., J. D. Ligon, J. R. Murphy, J. P.Myers, D. Simberloff, and J. Verner. 1987.Report of the scientific advisory panel on thespotted owl. Condor 89:205-229.

DeBano, L. F., and L. J. Schmidt. 1989a.Improving southwestern riparian areasthrough watershed management. USDA For.

Serv. Gen. Tech. Rep. RM-182. 33pp, Ft.Collins, Colo.

——., and ——. 1989b. Interrelationshipsbetween watershed condition and riparianhealth in southwestern United States. Pages45-52 in R. E. Gresswell, B. A. Barton, and J.L. Kershner, eds. Practical approaches toriparian resource management. USDI Bur.Land Manage., Billings, Montana.

DeVelice, R. L., J. A. Ludwig, W. H. Moir, andF. Ronco, Jr. 1986. A Classification of foresthabitat types of Northern New Mexico andSouthern Colorado. USDA For. Serv. Gen.Tech. Rep. RM-131. Rocky Mtn. For. andRange Exper. Stn. Fort Collins, Colo.

Diaz, N., and D. Apostol. 1993. Forestlandscape analysis and design: a process fordeveloping and implementing landmanagement objectives for landscape patterns.USDA For. Serv., Pacific Northwest Region,ECO-TP-043-92. 91pp.

Dickman, A. 1992. Plant pathogens and long-term ecosystem changes. Pages 499-520 in G.C. Carroll and D. T. Wicklow, eds. The fungalcommunity: its organization and role in theecosystem. Marcel Dekker, New York, N.Y.

Dick-Peddie, W. A. 1993. New Mexicovegetation, past, present and future. Univ.New Mexico Press, Albuquerque. 244pp.

Dinoor, A., and N. Eshed. 1984. The role andimportance of pathogens in natural plantcommunities. Ann. Rev. Phytopathol.22:443-466.

Dixon, G. 1991. Central Rockies Gemgynvariant of the forest vegetation simulator.Unpub. Rep. USDA For. Serv. WO-TM Serv.Center, Ft. Collins, Colo. 11pp.

Duncan, R. B., and J. D. Taiz. 1992. Apreliminary understanding of Mexicanspotted owl habitat and distribution in theChiricahua Mountains and associated sub-Mogollon mountain ranges in southeasternArizona. Pages 58-61 in A. M. Barton and S.A. Sloane, eds. Proc. Chiricahua Mtns. Res.Symp. Southwest Parks and MonumentsAssoc., Tucson, Arizona.

Dwyer, D. D., J. C. Buckhouse, and W. S. Huey.1984. Impacts of grazing intensity andspecialized grazing systems on the use andvalue of rangeland: summary and



recommendations. Pages 867-884 in NationalResearch Council and National Academy ofSciences. Developing strategies for rangelandmanagement: a report prepared by thecommittee on developing strategies forrangeland management. Westview Press,Boulder, Colo.

Earhart, C. M., and N. K. Johnson. 1970. Sex,dimorphism, and food habits of NorthAmerican owls. Condor 72:251-264.

Eberhardt, L. L., and J. M. Thomas. 1991.Designing environmental field studies. Ecol.Monogr. 61:53-73.

Edminster, C. B., H. T. Mowrer, R. L.Mathasen, T. H. Shuler, W. K. Olsen. F. G.Hawksworth. 1991. GENGYM: a variabledensity stand table projection systemcalibrated for mixed-conifer and ponderosapine stands in the Southwest. USDA For.Serv. Res. Pap. RM-297, Rocky Mtn. For.Range Exp. Stn, Ft. Collins, Colo. 32pp.

Eyre, F.H. 1980. Forest cover types of the UnitedStates and Canada. Society of AmericanForesters, Bethesda, Md. 148pp.

Fitton, S. D. 1991. Vocal learning and callstructure of male northern spotted owls innorthwestern California. M.S. Thesis.Humboldt State Univ. Arcata, Calif. 32pp.

Fitzhugh, E. L., W. H. Moir, J. A. Ludwig, andF. Ronco, Jr. 1987. Forest habitat types in theApache, Gila, and part of the Cibola NationalForests, Arizona and New Mexico. USDAFor. Serv. Gen. Tech. Rep. RM-145. RockyMtn. For. and Range Exper. Stn.Fort Collins, Colo.

Fleischner, T. L. 1994. Ecological costs oflivestock grazing in western North America.Cons. Biol. 8:629-644.

Fletcher, K. W. 1990. Habitats used, abundanceand distribution of the Mexican spotted owl,Strix occidentalis lucida, on National Forestsystem lands. USDA For. Serv., SouthwesternRegion, Albuquerque, N.M. 55pp.

——., and H. E. Hollis. 1994. Habitats used,abundance, and distribution of the Mex- icanspotted owl (Strix occidentalis lucida) onNational Forest System lands in theSouthwestern Region. USDA For. Serv.,Southwestern Region, Albuquerque, N.M.

Forest Ecosystem Management AssessmentTeam. 1993. Forest ecosystem management:an ecological, economic, and socialassessment. Unpubl. Rep. U.S. Gov. Print.Off. (1993 794-478), Washington, D. C.

Forsman, E. D. 1976. A preliminaryinvestigation of the spotted owl in Oregon.M.S. Thesis, Oregon State Univ., Corvallis.127pp.

——. 1981. Molt of the spotted owl. Auk98:735-742.

——., E. C. Meslow, and H. M. Wight. 1984.Distribution and biology of the spotted owl inOregon. Wildl. Monogr. 87:1-64.

Foster, C. C., E. D. Forsman, E. C. Meslow, G.S. Miller, J. A. Reid, F. F. Wagner, A. B. Carey,and J. B. Lint. 1992. Survival andreproduction of radio-marked adult spottedowls. J. Wildl. Manage. 56:91-95.

Franklin, A. B. 1992. Population regulation innorthern spotted owls: theoreticalimplications for management. Pages 815-827in D. R. McCullough and R. H. Barrett, eds.Wildlife 2001: populations. Elsevier AppliedScience, New York, N.Y.

——., J. P. Ward, R. J. Gutiérrez, and G. I.Gould, Jr. 1990. Density of northern spottedowls in northwestern California. J. Wildl.Manage. 54:1-10.

Franklin, J. F. 1993. Preserving biodiversity:species, ecosystems, or landscapes? Ecol.Applic. 3:202-205.

——. 1994. Ecosystem management: anoverview. Ecosystem management:applications for sustainable forest and wildliferesources. Stevens Point, Wisc. 26pp.

Ganey, J. L. 1988. Distribution and habitatecology of Mexican spotted owls in Arizona.M.S. Thesis, Northern Arizona Univ.,Flagstaff. 229 pp.

——. 1990. Calling behavior of spotted owls innorthern Arizona. Condor 92:485-490.

——., and R. P. Balda. 1989a. Distribution andhabitat use of Mexican spotted owls inArizona. Condor 91:355-361.

——., and ——. 1989b. Home-rangecharacteristics of spotted owls in northernArizona. J. Wildl. Manage. 53:1159-1165.



——., and ——. 1994. Habitat selection byMexican spotted owls in northern Arizona.Auk 111:162-169.

——., and J. L. Dick, Jr. 1995. Habitatrelationships of the Mexican spotted owl:Current knowledge. Chapter 2 (25pp.) inUSDI 1995. Final recovery plan for theMexican spotted owl. Volume II. U.S. Fishand Wildl. Serv. Albuquerque, N.M.

——., R. P. Balda, and R. M. King. 1993.Metabolic rate and evaporative water loss ofMexican spotted and great horned owls.Wilson Bull. 105:645-656.

——., R. B. Duncan, and W. M. Block. 1992.Use of oak and associated woodlands byMexican spotted owls in Arizona. Pages 125-128 in P. F. Ffolliott, G. J. Gottfried, D. A.Bennett, V. M. Hernandez C., A. Ortega-Rubio, and R. H. Hamre, eds. Ecology andmanagement of oak and associatedwoodlands: perspectives in the southwesternUnited States and northern Mexico. USDAFor. Serv. Gen. Tech. Rep. RM-218, Ft.Collins, Colo.

——., J. A. Johnson, R. P. Balda, and R. W.Skaggs. 1988. Status report: Mexican spottedowl. Pages 145-150 in R. L. Glinski et al., eds.Proc. Southwest Raptor Manage. Symp. andWorkshop. Natl. Wildl. Feder. Sci. and Tech.Ser. 11.

Grubb, T. G., and W. L. Eakle. 1988. Recordingwildlife locations with the UniversalTransverse Mercator (UTM) grid system.USDA For. Serv. Res. Note RM-483, Ft.Collins, Colo. 3pp.

Gutiérrez, R. J. 1985. An overview of recentresearch on the spotted owl. Pages 39-49 inR. J. Gutiérrez and A. B. Carey, tech. eds.Ecology and management of the spotted owlin the Pacific Northwest. USDA For. Serv.Gen. Tech. Rep. PNW-185. Pacific NorthwestRes. Stn. Portland, Oreg. 119 pp.

——. 1989. Hematozoa from the spotted owl. J.Wildl. Dis. 25:614-618.

——. 1994. Conservation planning: lessonsfrom the spotted owl. Pages 51-58 in W. W.Covington and L. F. Debano, eds. Sustainableecological systems: implementing anecological approach to land management.

USDA For. Serv. Gen. Tech. Rep. RM-247,Ft. Collins, Colo.

——, A. B. Franklin, and W. S. LaHaye. 1995.Spotted owl. in A. Poole, P. Stettenheim, andF. Gill, eds. The birds of North America.Acad. Nat. Sciences and Am. Ornithol.Union. Washington, D.C.

——., ——., W. LaHaye, V. J. Meretsky, and J.P. Ward. 1985. Juvenile spotted owl dispersalin northwestern California: preliminaryresults. Pages 60-65 in R. J. Gutiérrez and A.B. Carey, eds. Ecology and management ofthe spotted owl in the Pacific Northwest.USDA For. Serv. Gen. Tech. Rep. PNW-185,Portland, Oreg.

Haack, R. A., and J. W. Byler. 1993. Insects andpathogens, regulators of forest ecosystems. J.For. 91(9):32-37.

Hamer, T. E., E. D. Forsman, A. D. Fuchs, andM. L. Waters. 1992. Hybridization betweenbarred and spotted owls. Unpubl. abstract.Raptor Res. Foundation Annu. Meeting.

Hammond, E. H. 1965. Physical subdivisions.USDI Geological Survey.

Hanks, J. P., and W. A. Dick-Peddie. 1974.Vegetation patterns of the White Mountains,New Mexico. Southwest. Nat. 18:371-382.

——., E. L. Fitzhugh, and S. R. Hanks. 1983. Ahabitat type classification system forponderosa pine forests of northern Arizona.USDA For. Serv. Gen. Tech. Rep. RM-97.Rocky Mtn. For. and Range Exper. Stn. FortCollins, Colo.

Hanley, T. A., and J. L. Page. 1982. Differentialeffects of livestock use on habitat structureand rodent populations in Great Basincommunities. Calif. Fish and Game68:160-174.

Hawksworth, F. G., and D. Conklin. 1990.White pine blister rust in New Mexico. Pages43-44 in J. Hoffman and L. H. Spiegel, eds.Proc. 38th Annu. Western International For.Disease Work Conf. U.S. For. Serv., For. PestManage., Boise, Idaho.

Hessburg, P. F., and J. S. Beatty. 1986.Incidence, severity, and growth lossesassociated with ponderosa pine dwarfmistletoe on the Lincoln National Forest,New Mexico. USDA For. Serv. SouthwesternRegion, Forest Pest Manage. R-3 86-5. 30pp.



Hobbs, R. J., and L. F. Huenneke. 1992.Disturbance, diversity, and invasion:implications for conservation. Cons. Biol.6:324-337.

Hubbard, J. P. 1977. Importance of riparianecosystems: biotic considerations. Pages 14-18in R. R. Johnson and D. A. Jones, eds.Importance, preservation and management ofriparian habitat: a symposium. USDA For.Serv. Gen. Tech. Rep. RM-43, Ft. Collins,Colo.

Hunter, M. L., Jr. 1991. Coping with ignorance:the coarse-filter strategy for maintainingbiodiversity. Pages 266-281 in K. A. Kohm,ed. Balancing on the brink of extinction.Island Press, Washington, D.C.

Hunter, J. E., R. J. Gutiérrez, A. B. Franklin,and D. Olson. 1994. Ectoparasites of thespotted owl. J. Raptor Res. 28:232-235.

Hurlburt, S. H. 1984. Pseudoreplication and thedesign of ecological field experiments. Ecol.Monogr. 54:187-211.

Jensen, M. E., and P. S. Bourgeron. 1993.Ecosystem management: principles andapplications. USDA For. Serv., NorthernRegion. 388pp.

Johnsgard, P. A. 1988. North American owls:biology and natural history. Smithsonian Inst.Press, Washington, D.C. 295pp.

Johnson, J. A., and T. H. Johnson. 1985. Thestatus of the spotted owl in northern NewMexico. Unpubl. rep. New Mexico Dep.Game and Fish, Santa Fe. 39pp.

Johnson, M. 1994. Changes in Southwesternforests: stewardship implications. J. For.92(12):16-19.

Kauffman, J. B., and W. C. Krueger. 1984.Livestock impacts on riparian ecosystems andstreamside management implications, areview. J. Range Manage. 37:430-438.

——., ——., and M. Vavra. 1983. Effects of lateseason cattle grazing on riparian plantcommunities. J. Range Manage. 36:685-691.

Kaufmann, M. R., R. T. Graham, D. A. Boyce,W. H. Moir, L. Perry, R. T. Reynolds, R. L.Bassett, P. Mehlhop, C. B. Edminster, W. M.Block, and P. S. Corn. 1994. An ecologicalbasis for ecosystem management. USDA For.Serv. Gen. Tech. Rep. RM-246, Ft. Collins,Colo. 22pp.

Keitt, T. H., A. B. Franklin, and D. L. Urban.1995. Landscape analysis and metapopulationstructure. Chapter 3 (16pp.) in USDI 1995.Mexican spotted owl recovery plan.Volume II. USDI Fish and Wildl. Serv.Albuquerque, N.M.

Kendall, W. L., and K. H. Pollock. 1992. Therobust design in capture-recapture studies.Pages 31-43 in D. R. McCullough and R. H.Barrett, eds. Willdife 2001: populations.Elsevier Applied Sci., New York, N.Y.

Kennedy, C. E. 1977. Wildlife conflicts inriparian management: water. Pages 52-58 inR. R. Johnson and D. A. Jones, eds.Importance, preservation and management ofriparian habitat: a symposium. USDA For.Serv. Gen. Tech. Rep. RM-43,Ft. Collins, Colo.

Kertell, K. 1977. The spotted owl at ZionNational Park, Utah. West. Birds 8:147-150.

Knauer, K. H. 1988. Forest health throughsilviculture and integrated pest management, astrategic plan. USDA For. Serv. 26pp.

Knopf, F. L., R. R. Johnson, T. Rich, F. B.Samson, and R. C. Szaro. 1988 Conservationof riparian ecosystems in the United States.Wilson Bull. 100:272-284.

Kristan, D. M., R. J. Gutiérrez, A. B. Franklin,and W. S. LaHaye. In prep. Comparativemorphology of the spotted owl: subspecificand regional variation. Auk.

Kroel, K. and P. J. Zwank. 1991. Home rangeand habitat use characteristics of the Mexicanspotted owl in the southern SacramentoMountains, New Mexico. Unpubl. rep.submitted to USDA For. Serv., LincolnNational Forest. New Mexico Coop. Fish andWildl. Unit, Las Cruces.

Küchler, A. W. 1964. The potential naturalvegetation of the conterminous United States.Amer. Geogr. Soc. Spec. Publ. 361. 116pp.

Larson, M., and W. H. Moir. 1986. Forest andwoodland habitat types (plant associations) ofsouthern New Mexico and central Arizona(north of the Mogollon Rim). USDA For.Serv. Southwestern Region. Albuquerque,N.M.

Laymon, S. A. 1991. Diurnal foraging byspotted owls. Wilson Bull. 103:138-140.



Layser, E. F., and G. H. Schubert. 1979.Preliminary classification for the coniferousforest and woodland series of Arizona andNew Mexico. USDA For. Serv. Res. Pap. RM-208. Rocky Mtn. For. and Range Exper. Stn.Fort Collins, Colo.

Lebreton, J. D., K. P. Burnham, J. Clobert, andD. R. Anderson. 1992. Modeling survival andtesting biological hypotheses using markedanimals: a unified approach with case studies.Ecol. Monogr. 62:67-118.

Ledig, F. T. 1992. Human impacts on geneticdiversity in forest ecosystems. Oikos 63:87-108.

Lefebvre, L. W., D. L. Otis, and N. R. Holler.1982. Comparison of open and closed modelsfor cotton rat population estimates. J. Wildl.Manage. 46:156-163.

Marsden, M. A., C. G. I. Shaw, and M.Morrison. 1993. Simulation of managementoptions for stands of Southwestern ponderosapine attacked by Armillaria root disease anddwarf mistletoe. USDA For. Serv. Res. Pap.RM-308, Ft. Collins, Colo. 5pp.

Maser, C., J. M. Trappe, and D. C. Ure. 1978.Implications of small mammal myco- phagyto the management of western coniferousforests. Trans. North Am. Wildl. Nat. Resour.Conf. 43:78-88.

May, C. A., M. Z. Peery, R. J. Gutiérrez, M. E.Seamans, and D.R. Olson. In press.Feasability of a random quardat study designto estimate changes in density of Mexicanspotted owls. U. S. For. Serv. Res. Pap. Ft.Collins, Colo.

McDonald, C. B., J. Anderson, J. C. Lewis, R.Mesta, A. Ratzlaff, T. J. Tibbitts, and S. O.Williams. 1991. Mexican spotted owl (Strixoccidentalis lucida) status report. USDI Fishand Wildl. Serv., Albuquerque, N.M.

McLaughlin, S. P. 1986. Floristic analysis of thesouthwestern United States. Great Basin Nat.46:46-65.

Medin, D. E., and W. P. Clary. 1990. Bird andsmall mammal populations in a grazed andungrazed riparian habitat in Idaho. USDAFor. Serv. Res. Pap. INT-425, Ogden,Utah. 8pp.

Medina, A. L. 1986. Riparian plantcommunities of the Fort Bayard watershed in

southwestern New Mexico. Southwestern Nat.31:345-359.

Meslow, E. C. 1993. Spotted owl protection:unintentional evolution toward ecosystemmanagement. Endang. Species Update10:34-38.

Milchunas, D. G., and W. K. Lauenroth. 1993.Quantitative effects of grazing on vegetationand soils over a global range of environments.Ecol. Monogr. 63:327-366.

Miller, G. S. 1989. Dispersal of juvenilenorthern spotted owls in western Oregon.M.S. Thesis, Oregon State Univ., Corvallis.

Minckley, W. L., and T. O. Clark. 1984.Formation and destruction of a Gila Rivermesquite bosque community. Desert Plants6:23-30.

——., and J. N. Rinne. 1985. Large woodydebris in hot-desert streams: an historicalreview. Desert Plants 7:142- 153.

Moen, C. A., A. B. Franklin, and R. J. Gutiérrez.1991. Age determination of subadultnorthern spotted owls in northwestCalifornia. Wildl. Soc. Bull. 19:489-493.

Moir, W. H. 1993. Alpine tundra and coniferousforest. Pages 47-84 in W. A. Dick-Peddie.New Mexico vegetation, past, present andfuture. Univ. of New Mexico Press,Albuquerque. 244pp.

——., and J. A. Ludwig. 1979. A classificationof spruce-fir and mixed-conifer habitat typesof Arizona and New Mexico. USDA For. Serv.Res. Pap. RM-207. Rocky Mtn. For. andRange Exper. Stn. Fort Collins, Colo.

Monson, G., and A. R. Phillips. 1981.Annotated checklist of the birds of Arizona.Univ. of Arizona Press, Tucson. 240pp.

Moody, R., L. Buchanan, R. Melcher, and H.Wistrand. 1992. Fire and forest health. USDAFor. Serv. Southwestern Region. Albuquerque,N.M. 23pp.

Murphy, D. D., and B. R. Noon. 1991. Copingwith uncertainty in wildlife biology. J. Wildl.Manage. 55:773-782.

Nelson, E. W. 1903. Descriptions of new birdsfrom southern Mexico. Proc. Biol. Soc. Wash.16:151-160.

Nice, M. M. 1941. The role of territory in birdlife. Am. Midl. Nat. 26:441-487.



Nichols, J. D. 1991. Science, populationecology, and the management of the Americanblack duck. J. Wildl. Manage. 55:790-799.

Noon, B. R., L. L. Eberhardt, A. B. Franklin, J.D. Nichols, and K. H. Pollock. 1993. Asurvey design to estimate the number ofterritorial spotted owls in Olympic NationalPark. Final Rep. submitted to Olympic Natl.Park. 31pp.

——., K. S. McKelvey, D. W. Lutz, W. S.LaHaye, R. J. Gutiérrez, and C. A. Moen.1992. Estimates of demographic parametersand rates of population change. Pages 175-186 in J. Verner, K. S. McKelvey, B. R. Noon,R. J. Gutiérrez, G. I. Gould, Jr., and T. W.Beck, eds. The California spotted owl: atechnical assessment of its current status.USDA For. Serv. Gen. Tech. Rep. PSW-133,Albany, Calif.

——., and ——. 1992. Stability properties ofthe spotted owl metapopulation in southernCalifornia. Pages 187-206 in J. Verner, K. S.McKelvey, B. R. Noon, R. J. Gutiérrez, G. I.Gould, Jr., and T. W. Beck, eds. TheCalifornia spotted owl: a technical assessmentof its current status. USDA For. Serv. Gen.Tech. Rep. PSW-133, Albany, Calif.

Otis, D. L., K. P. Burnham, G. C. White, andD. R. Anderson. 1978. Statistical inferencefrom capture data on closed animalpopulations. Wildl. Monogr. 62:1-135.

Pase, C. P., and E. F. Layser. 1977. Classificationof riparian habitat in the southwest. Page 5-9in R. R. Johnson and D. A. Jones, tech.coords. Importance, preservation, andmanagement of riparian habitat: Asymposium. USDA For. Serv. Gen. Tech. Rep.RM-43. Rocky Mtn. For. and Range Exper.Stn. Fort Collins, Colo.

Paton, P. W. C., C. J. Zabel, D. L. Neal, G. N.Steger, N. G. Tilghman, and B. R. Noon.1991. Effects of radio tags on spotted owls. J.Wildl. Manage. 55:617-622.

Pearson, G. A. 1931. Forest types in thesouthwest as determined by climate and soil.USDA For. Serv. Tech. Bull. 247.Washington, D.C.

Pfister, R. D. 1989. Basic concepts of usingvegetation to build a site classification system.Pages 22-28 in D. E. Ferguson, P. Morgan,

and F. D. Johnson, compilers. Proceedings-land classifications based on vegetation:Applications for resource management.USDA For. Serv. Gen. Tech. Rep. INT-257.Intermountain Res. Stn., Ogden, Utah.Platt, J. R. 1964. Strong inference. Science46:347-353.

Platts, W.S. 1990.Managing fisheries andwildlife on rangelands grazed by livestock: aguidance and reference document forbiologists. Nevada Dep. of Wildl. 448pp.

Plummer, F. G., and M. G. Gowsell. 1904.Forest conditions in the Lincoln Forest,Reserve, NM. USDI, Geol. Surv. Prof.Pap. 33.

Pollock, K. H. 1982. A capture-recapture designrobust to unequal probability of capture. J.Wildl. Manage. 37:757-760.

——., J. D. Nichols, C. Brownie, and J. E.Hines. 1990. Statistical inference for capture-recapture experiments. Wildl. Monogr.107:1-97.

Popper, K. R. 1965. The logic of scientificdiscovery. Harper and Row, New York, N.Y.

Powell, D. S., J. L. Faulkner, D. Darr, Z. Zhu,and D. W. MacCleery. 1993. Forest resourcesof the United States. USDA For. Serv., RockyMtn. For. Rng. Exp. Stn., Ft. Collins, Colo.

Reynolds, R. T. 1990. Distribution and habitatof Mexican spotted owls in Colorado:preliminary results. USDA For. Serv., RockyMtn. For. Rng. Exp. Stn., Laramie, Wyoming.10pp.

——. 1993. Ecology and dispersal of theMexican spotted owl in Colorado, FY 1993-94. Research Proposal. USDA For. Serv.,Rocky Mtn. For. Rng. Exp. Stn., Fort Collins,Colo.

——., R. T. Graham, M. H. Reiser, R. L.Bassett, P. L. Kennedy, D. A. Boyce, Jr., G.Goodwin, R. Smith, and E. L. Fisher. 1992.Management recommendations for thenorthern goshawk in the southwestern UnitedStates. USDA For. Serv. Gen. Tech. Rep. RM-217. Ft. Collins, Colo. 90pp.

Rickard, W. H., and C. E. Cushing. 1982.Recovery of streamside woody vegetation afterexclusion of livestock grazing. J. RangeManage. 35:360-361.



Ridgway, R. 1914. The birds of north andmiddle America. U.S. Natl. Mus. Bull. 50:1-882.

Rinkevich, S. E. 1991. Distribution and habitatcharacteristics of Mexican spotted owls inZion National Park, Utah. M.S. Thesis,Humboldt State Univ., Arcata, Calif. 62pp.

Romesburg, H. C. 1981. Wildlife science:gaining reliable knowledge. J. Wildl. Manage.45:527-534.

——. 1991. On improving the natural resourcesand environmental sciences. J. Wildl. Manage.55:744-756.

Schulz, T. T., and W. C. Leininger. 1990.Differences in riparian vegetation structurebetween grazed areas and exclosures. J. RangeManage. 43:295-299.

——., and ——. 1991. Nongame wildlifecommunities in grazed and ungrazed montaneriparian sites. Great Basin Nat. 51:286-292.

Seamans, M. E., and R. J. Gutiérrez. In press.Breeding habitat ecology of the Mexicanspotted owl in the Tularosa Mountains, NewMexico. Condor.

Short, L. L. 1965. Hybridization in the flickers(Colaptes) of North America. Bull. Amer.Mus. Nat. Hist. 129:309-428.

——. 1972. Hybridization, taxonomy and avianevolution. Ann. Missouri Bot. Gard.59:447-453.

Skaggs, R. W., and R. J. Raitt. 1988. A spottedowl inventory of the Lincoln National Forest,Sacramento Division, 1988. Unpubl. rep.New Mexico Dep. Game and Fish, Santa Fe.12pp.

Skovlin, J. M. 1984. Grazing impacts onwetlands and riparian habitat: a review of ourknowledge. Pages 1001-1004 in NationalResearch Council and National Academy ofSciences. Developing strategies for rangelandmanagement: a report prepared by thecommittee on developing strategies forrangeland management. Westview Press,Boulder, Colo.

Smith, D. M. 1986. The practice of silviculture.Second ed. John Wiley and Sons, New York,N.Y. 527pp.

Sovern, S. G., E. D. Forsman, B. L. Biswell, D.N. Rolph, and M. Taylor. 1994. Diurnal

behavior of the spotted owl in Washington.Condor 96:200-202.

Spencer, J. S., Jr. 1966. Arizona’s forests. USDAFor. Serv. Resour. Bull. INT-6, Ogden, Utah.

Stevens, R. E., and H. W. Flake, Jr. 1974. Aroundheaded pine beetle outbreak in NewMexico. USDA For. Serv. Res. Note RM-259,Ft. Collins, Colo. 4pp.

SWCA, Inc. 1992. Nest tree and nest treehabitat characteristics of the Mexican spottedowl in Arizona and New Mexico. Unpubl. rep.USDA For. Serv., Southwestern Region,Albuquerque, N.M.

Swetnam, T. W., and J. L. Bettancourt. 1990.Fire--southern oscillation relations insouthwestern United States. Science249:1017-1020.

——., and A. M. Lynch. 1993. Multicentury,regional-scale patterns of western sprucebudworm outbreaks. Ecol. Monogr. 63:399-424.

Szaro, R. C. 1989. Riparian forest and scrublandcommunity types of Arizona and NewMexico. Desert Plants 9: 70-138.

Tarango, L. A., R. Valdez, and P. J. Zwank.1994. Mexican spotted owl distribution andhabitat characterizations in southwesternChihuahua, Mexico. Final report, ContractNo. 80-516.6-66. New Mexico Dep. Gameand Fish, Santa Fe. 69pp.

Thomas, C. D. 1990. What do real populationdynamics tell us about minimum viablepopulation sizes. Conserv. Biol. 4:324-327.

Thomas, J. W., E. D. Forsman, J. B. Lint, E. C.Meslow, B. R. Noon, and J. Verner. 1990. Aconservation strategy for the spotted owl. U.S.Gov. Print. Off., Washington, D.C. 427 pp.

——., M. G. Raphael, R. G. Anthony, E. D.Forsman, A. G. Gunderson, R. S.Holthausen, B. G. Marcot, G. H. Reeves, J.R. Sedell, and D. M. Solis. 1993. Viabilityassessments and management considerationsfor species associated with late- successionaland old-growth forests of the PacificNorthwest. USDA For. Serv., Portland, Oreg.529pp.

Thompson, S. K. 1992. Sampling. John Wileyand Sons, New York, N.Y. 343pp.

Thornbury, W. 1965. Regional geomorphologyof the United States. John Wiley and Sons,



New York. N.Y. 609pp.Thrailkill, J., and M. A. Bias. 1989. Diets of

breeding and nonbreeding California spottedowls. J. Raptor Res. 23:39-41.

USDA Forest Service. 1983. Silvicultural systemsfor the major forest types of the United States.Agri. Handbook 445. U. S. Govt. PrintingOffice. Washington, D.C. vii, 114pp.

——. 1990. Management guidelines andinventory and monitoring protocols for theMexican spotted owl in the SouthwesternRegion. Federal Register 55:27278-27287.

——. 1991. General ecosystem survey, generalterrestrial vegetation, existing vegetation.USDA For. Serv. Southwestern Region andTechnology Application Center, Univ. NewMexico, Albuquerque. Map at scale1:1,000,000.

——. 1992. Recommended old growthdefinitions and descriptions and old growthallocation procedure. Southwestern Region.Albuquerque, N.M.

——. 1993a. Facts about the Mexican spottedowl. USDA For. Serv., Southwestern Region,Albuquerque, N.M.

——. 1993b. Changing conditions insouthwestern forests and implications on landstewardship. USDA For. Serv., SouthwesternRegion, Albuquerque, N.M.

——. 1993c. Forest health restoration initiative.USDA For. Serv., Southwestern Region,Albuquerque, N.M. 19pp.

——. 1994. Biological assessment andevaluation: livestock grazing. Unpubl. Rep.,USDA For. Serv., Southwest Regional Office,Albuquerque, N.M. 12pp.

USDI. 1994. The impact of federal programs onwetlands, Vol.II, A report to Congress by theSecretary of the Interior. Washington, D.C.

USDI Fish and Wildlife Service. 1992. Recoveryplan for the northern spotted owl. Final draft.U.S. Fish and Wildl. Serv., Portland, Oreg.194pp.

USDI Fish & Wildlife Service. 1993.Endangered and threatened wildlife andplants; final rule to list the Mexican spottedowl as a threatened species. Federal Register58: 14248-14271.

USDI Geological Survey. 1970. The nationalatlas of the United States of America.

Washington, D.C. 417pp.Utah Division of Wildlife Resources. 1987.

Native Utah wildlife species of specialconcern. Utah Div. Wildl. Resour., Salt LakeCity.

——. 1992. Native Utah wildlife species ofspecial concern, draft. Utah Div. Wildl.Resour., Salt Lake City.

Utah Mexican Spotted Owl Technical Team.1994. Suggestions for management ofMexican spotted owls in Utah. Utah Div.Wildl. Resour., Salt Lake City. 103pp.

Van Hooser, D. D., D. C. Collins, and R. A.O’Brien. 1992. Forest resources of NewMexico. USDA For. Serv. Inter. Res. Stn.,Ogden, Utah. 80pp.

Verner, J., K. S. McKelvey, B. R. Noon, R. J.Gutiérrez, G. I. Gould, Jr., and T. W. Beck,eds. 1992a. The California spotted owl: atechnical assessment of its current status.USDA For. Serv. Gen. Tech. Rep. PSW-133,Albany, Calif. 285pp.

——., ——., ——., ——., ——., and ——.1992b. Assessment of the current status of theCalifornia spotted owl, withrecommendations for management. Pages 3-26 in J. Verner, K. S. McKelvey, B. R. Noon,R. J. Gutierrez, G. I. Gould, Jr., and T. W.Beck, eds. The California spotted owl: atechnical assessment of its current status.USDA For. Serv. Gen. Tech. Rep. PSW-133,Albany, Calif.

Walker, L. W. 1974. The book of owls. Alfred A.Knopf, New York, N.Y. 255pp.

Ward, J. P., Jr., and W. M. Block.1995. Mexicanspotted owl prey ecology. Chapter 5 (48pp.)in USDI Fish and Wildlife Service. 1995.Mexican spotted owl recovery plan.Volume II. Albuquerque, N.M.

——., A. B. Franklin, S. Rinkevich, and F.Clemente. 1995. Distribution and abundance.Chapter 1 (14pp.) in USDI Fish and WildlifeService. 1995. Mexican spotted owl recoveryplan. Volume II. Albuquerque, N.M.

——., ——., and R. J. Gutiérrez. 1991. Usingsearch time and regression to estimateabundance of territorial spotted owls. Ecol.Applic. 1:207-214.

West, N. E. 1983. Western intermountainsagebrush steppe. Pages 351-374 in N. E.



West, ed. Temperate deserts and semi-deserts.Elsevier Press, New York, N.Y.

White, G. C., D. R. Anderson, K. P. Burnham,and D. L. Otis. 1982. Capture-recapture andremoval methods for sampling closedpopulations. USDE Los Alamos NationalLab., Los Alamos, N.M. 235pp.

White, G. C, A. B. Franklin, and J. P. Ward, Jr.1995. Population biology. Chapter 2 (25pp.)in USDI Fish and Wildlife Service. 1995.Mexican spotted owl recovery plan.Volume II. Albuquerque, N.M.

Wykoff, W. R., N. L. Crookston, and A. R.Stage. 1982. User’s guide to stand prognosismodel. USDA For. Serv. Gen. Tech. Rep.INT-133, Intermounain Res. Stn., Ogden,Utah. 112pp.

Willey, D. W. 1993. Home-range characteristicsand juvenile dispersal ecology of Mexicanspotted owls in southern Utah. Unpubl. rep.Utah Div. Wildl. Resour., Salt Lake City.

Williams, J. L. 1986. New Mexico in maps.Univ. New Mexico Press, Albuquerque.409pp.

Williams, J. T. 1994. Fire’s role in support ofecosystem management. The BiswellSymposium, Walnut Creek, Calif. USDA For.Serv., Fire and Aviation Management. 4pp.

Williams, S. O., III, and R. W. Skaggs. 1993.Distribution of the Mexican spotted owl (Strixoccidentalis lucida) in Mexico. Unpubl. rep.,New Mexico Dep. Game and Fish, Sante Fe.38pp.

Wilson, E. D. 1962. A resume of the geology ofArizona. Univ. Arizona Press, Tucson. 140pp.

Young, K. E., A. B. Franklin, and J. P. Ward, Jr.1993. Infestation of northern spotted owls byhippoboscid (Diptera) flies in nortwesternCalifornia. J. Wildl. Dis. 29:278-283.

Youngblood, A. P., and R. L. Mauk. 1985.Coniferous forest habitat types of central andsouthern Utah. USDA For. Serv. Gen. Tech.Rep. INT-187. Intermountain Res. Stn.Ogden, Utah.

Zimmerman, G. T., and L. F. Neuenschwander.1984. Livestock grazing influences oncommunity structure, fire intensity, and firefrequency within the Douglas-fir/ninebarkhabitat type. J. Range Manage. 37:104-110.

Zippin, C. 1956. An evaluation of the removalmethod of estimating animal populations.Biometrics 12:163-189.

——. 1958. The removal method of populationestimation. J. Wildl. Manage. 22:82-90.

Zwank, P. J., K. W. Kroel, D. M. Levin, G. M.Southward, and R. C. Rommé. 1994. Habitatcharacteristics of Mexican spotted owls insouthwestern New Mexico. J. Field Ornithol.65: 324-334.



APPENDIX AAPPENDIX AAPPENDIX AAPPENDIX AAPPENDIX A

THE RECOVERY TEAMTHE RECOVERY TEAMTHE RECOVERY TEAMTHE RECOVERY TEAMTHE RECOVERY TEAMAND ASSOCIATESAND ASSOCIATESAND ASSOCIATESAND ASSOCIATESAND ASSOCIATES

RECOVERY TEAM:RECOVERY TEAM:RECOVERY TEAM:RECOVERY TEAM:RECOVERY TEAM:

William M. Block, Team Leader;Wildlife Biologist.

Education: B.A., Economics, San Diego StateUniversity, 1974; B.S., Wildlife Biology,Michigan State University, 1981; M.S.,Wildlife Biology, Humboldt StateUniversity, 1985; Ph.D., WildlandResource Science, University of Califor-nia, Berkeley, 1989.

Current Position: Project Leader, ResearchWildlife Biologist, USDA Forest Service,Rocky Mountain Forest and RangeExperiment Station, Flagstaff, Arizona.

Fernando Clemente, Wildlife Biologist.Education: B.S., Animal Science, University

of Chapingo, Mexico, 1977; M.S.,Wild Animal Nutrition, Colegio DePostgraduados, Mexico, 1984; Ph.D.,Range and Wildlife Management, NewMexico State University, 1992.

Current Position: Head, Department of WildlifeScience, and Wildlife Professor, ColegioDe Postgraduados, Campus San LuisPotosi, Mexico.

James L. Dick, Jr., Silviculturist.Education: B.S., Forestry, University of Mon-

tana, 1967; M.S., Forest Resources,Pennsylvania State University, 1972.

Current Position: Forester, Recreation Staff Unit,USDA Forest Service, SouthwesternRegion, Albuquerque, New Mexico.

Alan B. Franklin, Spotted Owl Researcher.Education: B.S., Wildlife Biology, Cornell

University, 1979; M.S., Wildlife Biology,Humboldt State University, 1987;Doctoral Candidate, Department ofFisheries and Wildlife, Colorado StateUniversity.

Current Position: Project Leader, Humboldt

State University Foundation, HumboldtState University, Arcata, California.

Joseph L. Ganey, Spotted Owl Researcher.Education: B.S., Wildlife Biology, Humboldt

State University, 1981; M.S., Biology,Northern Arizona University, 1988;Ph.D., Zoology, Northern ArizonaUniversity, 1991.

Current Position: Research Wildlife Biologist,USDA Forest Service, Rocky MountainForest and Range Experiment Station,Flagstaff, Arizona.

W. H. Moir, Ecologist.Education: Ph.D., Botany and Soils,

Washington State University, 1965.Current Position: Research Ecologist, USDA

Forest Service, Rocky Mountain Forestand Range Experiment Station, FortCollins, Colorado.

Sarah E. Rinkevich, Wildlife Biologist.Education: B.S., Wildlife and Fisheries Science,

University of Arizona, 1987; M.S.,Wildlife Biology, Humboldt StateUniversity, 1991.

Current Position: Fish and Wildlife Biologist,USDI Fish and Wildlife Service, NewMexico Ecological Services State Office,Albuquerque, New Mexico.

Dean L. Urban, Landscape Ecologist.Education: B.A., Botany and Zoology, Southern

Illinois University, 1978; M.A., Zoology(Wildlife Ecology), Southern IllinoisUniversity, 1981; Ph.D., Ecology,University of Tennessee, 1986.

Current Position: Assistant Professor, School ofEnvironment, Duke University.

James P. Ward, Jr., Spotted Owl Researcher.Education: B.S., Wildlife Biology, Humboldt

State University, 1985; M.S., NaturalResources (wildlife science emphasis),Humboldt State University, 1990;Doctoral Candidate, Department ofBiology, Colorado State University.

163Volume I/Appendix A


Current Position: Wildlife Biologist, USDAForest Service, Rocky Mountain Forestand Range Experiment Station, FortCollins, Colorado.

Gary C. White, Population Ecologist.Education: B.S., Fisheries and Wildlife Biology,

Iowa State University, 1970; M.S.,Wildlife Biology, University of Maine-Orono, 1972; Ph.D., Zoology, OhioState University, 1976; Post Doctorate,Wildlife Biology, Utah State University,1976-77.

Current Position: Professor, Department ofFishery and Wildlife, Colorado StateUniversity, Fort Collins, Colorado.

RECOVERY TEAM-RECOVERY TEAM-RECOVERY TEAM-RECOVERY TEAM-RECOVERY TEAM-FISH AND WILDLIFE SERVICEFISH AND WILDLIFE SERVICEFISH AND WILDLIFE SERVICEFISH AND WILDLIFE SERVICEFISH AND WILDLIFE SERVICE

LIAISON:LIAISON:LIAISON:LIAISON:LIAISON:

Steven L. Spangle, Fish and Wildlife Biologist —Regional Listing Coordinator,USDI Fish and Wildlife Service, South-western Regional Office, Albuquerque,New Mexico.

CONSULTANTS:CONSULTANTS:CONSULTANTS:CONSULTANTS:CONSULTANTS:

Pat Christgau, Coordinator, Mexican SpottedOwl Management, Arizona Game andFish Department, Phoenix, Arizona.

Jack F. Cully, Jr., Assistant Unit Leader-Wildlife,National Biological Service, KansasCooperative Fish and Wildlife ResearchUnit, Kansas State University, Manhat-tan, Kansas.

Frank P. Howe, Utah Partners in Flight

Coordinator, Utah Division of WildlifeResources, Salt Lake City, Utah.

Tim Keitt, Graduate Research Assistant,Department of Biology, Universityof New Mexico, Albuquerque, NewMexico

Tom Spalding, Deputy Director, ArizonaDepartment of Game and Fish,Phoenix, Arizona.

Steve Thompson, Biological Technician, ForestResources Program, San Carlos ApacheTribe, San Carlos, Arizona.

Robert Vahle, Program Manager, ArizonaGame and Fish Department,Region 1, Pinetop, Arizona.

MEETING FACILITATOR:MEETING FACILITATOR:MEETING FACILITATOR:MEETING FACILITATOR:MEETING FACILITATOR:

Kate W. Grandison, Intermountain RegionalSpotted Owl Coordinator,Dixie National Forest, Cedar City, Utah.

ADMINISTRATIVE ASSISTANT:ADMINISTRATIVE ASSISTANT:ADMINISTRATIVE ASSISTANT:ADMINISTRATIVE ASSISTANT:ADMINISTRATIVE ASSISTANT:

Brenda Witsell, Biological Technician, USDAForest Service, Rocky Mountain Forestand Range Experiment Station, Flagstaff,Arizona.

164Volume I/Appendix A


APPENDIX BAPPENDIX BAPPENDIX BAPPENDIX BAPPENDIX B

SCHEDULE OF TEAM MEETINGSSCHEDULE OF TEAM MEETINGSSCHEDULE OF TEAM MEETINGSSCHEDULE OF TEAM MEETINGSSCHEDULE OF TEAM MEETINGS

March 29 to April 4, 1993Albuquerque, New Mexico

April 27-30, 1993Flagstaff, Arizona

May 17-21, 1993Sierra Vista, Arizona

June 23-25, 1993Alamogordo, New Mexico

July 12-16, 1993Flagstaff, Arizona

August 16-19, 1993Pinetop, Arizona

September 14-16, 1993Fort Collins, Colorado

October 13-15, 1993Albuquerque, New Mexico

January 10-14, 1994Flagstaff, Arizona

February 22-25, 1994Fort Collins, Colorado

March 14-16, 1994Albuquerque, New Mexico

April 25-29, 1994Aguascalientes, Mexico

May 23-27, 1994Cedar City, Utah

June 27 to July 1, 1994Flagstaff, Arizona

August 8-12, 1994Flagstaff, Arizona

September 7-9, 1994Fort Collins, Colorado

September 29, 1994Albuquerque, New Mexico

October 18-19, 1994Phoenix, Arizona

February 13-17, 1995Phoenix, Arizona

June 19-23, 1995Phoenix, Arizona

July 17-21, 1995Flagstaff, Arizona

August 14-18, 1995Albuquerque, New Mexico

165Volume I/Appendix B


APPENDIX CAPPENDIX CAPPENDIX CAPPENDIX CAPPENDIX C

SCHEDULE OF FIELD VISITSSCHEDULE OF FIELD VISITSSCHEDULE OF FIELD VISITSSCHEDULE OF FIELD VISITSSCHEDULE OF FIELD VISITS

DATEDATEDATEDATEDATE PLACEPLACEPLACEPLACEPLACE COORDINATORCOORDINATORCOORDINATORCOORDINATORCOORDINATOR

April 4, 1993 Walnut Canyon/Bar M Joe Ganey,Canyon, Coconino NF, Arizona Heather Green

May 20, 1993 Huachuca, Santa Rita and, Russell DuncanPatagonia Mountains, Arizona Steve Spiech

June 23, 1993 Sacramento Mountains, Danney SalasLincoln NF, New Mexico Pat Ward

August 8, 1993 Overflight, Gila NF, Bruce AndersonNew Mexico Steve Servis

August 20, 1993 Fort Apache Indian Reservation, Joe JojolaArizona

April 4, 1994 Sierra Fria, National Wildlife CouncilAguascalientes, Mexico

May 11-13, 1994 San Carlos Apache Indian Reservation, Steve ThompsonArizona Tim Wilhite

May 5, 1994 Zion National Park, Utah Sarah Rinkevich

166Volume I/Appendix C


APPENDIX DAPPENDIX DAPPENDIX DAPPENDIX DAPPENDIX D

A REFERENCE FOR ENGLISHA REFERENCE FOR ENGLISHA REFERENCE FOR ENGLISHA REFERENCE FOR ENGLISHA REFERENCE FOR ENGLISHAND LATIN NAMESAND LATIN NAMESAND LATIN NAMESAND LATIN NAMESAND LATIN NAMES

English names were used in the text ofthis Recovery Plan to make smoother readingand to improve comprehension among peoplewho are not familiar with Latin names. Namesare arranged in alphabetical order of families forplants because this is the conventional way ofarranging them in most commonly accessibletree and wildflower manuals. Species within thefamilies are listed alphabetically by Latin names.Animal names are listed phylogenetically, ortaxonomically, because this system prevails inmost commonly accessible books of birds andmammals, the principal species reported here.Family names are also provided for the animalsas an aid for further investigation.

PLANTSPLANTSPLANTSPLANTSPLANTS

AceraceaeAceraceaeAceraceaeAceraceaeAceraceae

Canyon Acer(Bigtooth) maple grandidentatum

Boxelder Acer negundo

ChenopodiaceaeChenopodiaceaeChenopodiaceaeChenopodiaceaeChenopodiaceae

Shadscale Atriplex sp.

CupressaceaeCupressaceaeCupressaceaeCupressaceaeCupressaceae

Arizona cypress Cupressus arizonicaJuniper Juniperus sp.

EricaceaeEricaceaeEricaceaeEricaceaeEricaceae

Madrone Arbutus sp.Manzanita Arctostaphylos sp.

FabaceaeFabaceaeFabaceaeFabaceaeFabaceae

Mesquite Prosopis sp.

FagaceaeFagaceaeFagaceaeFagaceaeFagaceae

Arizona white oak Quercus arizonicaQuercus chihuahuensisQuercus coccolobifolia

Emory oak Quercus emoryiGambel oak Quercus gambelii

Quercus gentryiGray oak Quercus griseaSilverleaf oak Quercus hypoleucoides

Quercus laetaQuercus potosinaQuercus resinosa

Netleaf oak Quercus rugosaWavyleaf oak Quercus undulata

PapilionoideaePapilionoideaePapilionoideaePapilionoideaePapilionoideae

New Mexico locust Robinia neomexicana

PinaceaePinaceaePinaceaePinaceaePinaceae

White fir Abies concolorBlue Spruce Picea pungensPinyon pine Pinus edulisLimber pine Pinus flexilisWestern white pine Pinus monticolaPonderosa pine Pinus ponderosaAztec pine Pinus teocoteSouthwestern Pinus strobiformis

white pineDouglas-fir Pseudotsuga menziesiiRedwood Sequoia sempervirens

PlatanaceaePlatanaceaePlatanaceaePlatanaceaePlatanaceae

Arizona Sycamore Platanus wrightii

SalicaceaeSalicaceaeSalicaceaeSalicaceaeSalicaceaeNarrowleaf cottonwood Populus angustifoliaTrembling aspen Populus tremuloides

ViscaceaeViscaceaeViscaceaeViscaceaeViscaceae

Dwarf mistletoe Arceuthobium sp.

ZygophyllaceaeZygophyllaceaeZygophyllaceaeZygophyllaceaeZygophyllaceae

Creosotebush Larrea sp.

167Volume I/Appendix D


ANIMALSANIMALSANIMALSANIMALSANIMALS

INVERTEBRATESINVERTEBRATESINVERTEBRATESINVERTEBRATESINVERTEBRATES

TortricidaeTortricidaeTortricidaeTortricidaeTortricidae

Western Choristoneuraspruce budworm occidentalis

ScolytidaeScolytidaeScolytidaeScolytidaeScolytidae

Round-headed beetle Dendroctonusadjunctus

DanaidaeDanaidaeDanaidaeDanaidaeDanaidae

Monarch Butterfly Danaus plexippus

VERTEBRATESVERTEBRATESVERTEBRATESVERTEBRATESVERTEBRATES

BirdsBirdsBirdsBirdsBirds

AccipitridaeAccipitridaeAccipitridaeAccipitridaeAccipitridae

Golden eagle Aquila chrysaetosNorthern goshawk Accipter gentilisRed-tailed hawk Buteo jamaicensis

StrigidaeStrigidaeStrigidaeStrigidaeStrigidae

Great horned owl Bubo virginianusSpotted owl Strix occidentalisCalifornia spotted owl S. o. occidentalisMexican spotted owl S. o. lucidaNorthern spotted owl S. o. caurinaBarred owl Strix variaFulvous owl Strix fulvescensTawny owl Strix aluco

PsittacidaePsittacidaePsittacidaePsittacidaePsittacidae

Yellow-headed parrot Amazona ochrocephala

MammalsMammalsMammalsMammalsMammals

SoricidaeSoricidaeSoricidaeSoricidaeSoricidae

Masked shrew Sorex cinereusVagrant shrew Sorex vagrans

Montane shrew Sorex monticoulsDusky shrew Sorex obscurusWater shrew Sorex palustris

LeporidaeLeporidaeLeporidaeLeporidaeLeporidae

Desert cottontail Sylvilagus auduboniiEastern cottontail Sylvilagus floridanusNuttall’s cottontail Sylvilagus nuttallii

SciuridaeSciuridaeSciuridaeSciuridaeSciuridae

Northern Glaucomys sabrinaflying squirrel

Golden-mantled Spermophilusground squirrel lateralis

Rock squirrel Spermophilusvariegatus

Gray-collared Tamias cinereicolluschipmunk

Gray-footed chipmunk Tamias canipesLeast chipmunk Tamias minimusCliff chipmunk Tamias dorsalisRed squirrel Tamiasciurus

hudsonicus

GeomyidaeGeomyidaeGeomyidaeGeomyidaeGeomyidae

Botta’s pocket gopher Thomomys bottaeSouthern pocket Thomomys umbrinus

gopherNorthern pocket Thomomys talpoides

gopher

HeteromyidaeHeteromyidaeHeteromyidaeHeteromyidaeHeteromyidae

Great Basin Perognathus parvuspocket mouse

MuridaeMuridaeMuridaeMuridaeMuridae

Pinyon mouse Peromyscus trueiBrush mouse Peromyscus boyleiCanyon mouse Peromyscus crinitisRock mouse Peromyscus difficilisDeer mouse Peromyscus

maniculatusWhite-footed mouse Peromyscus leucopusBushy-tailed woodrat Neotoma cinereaDesert woodrat Neotoma lepida



Mexican woodrat Neotoma mexicanaStephens woodrat Neotoma stephensiWhite-throated Neotoma albigula

woodratLong-tailed vole Microtus longicaudusMeadow vole Microtus

pennsylvanicusMexican vole Microtus mexicanusMontane vole Microtus montanus

DipodidaeDipodidaeDipodidaeDipodidaeDipodidae

Meadow Zapus hudsoniusjumping mouse

Western Zapus princepsjumping mouse

MustelidaeMustelidaeMustelidaeMustelidaeMustelidae

Long-tailed weasel Mustela frenata

TrichechidaeTrichechidaeTrichechidaeTrichechidaeTrichechidae

Manatee Trichechus manatus



APPENDIX EAPPENDIX EAPPENDIX EAPPENDIX EAPPENDIX E

AGENCIES AND PERSONS COMMENTINGAGENCIES AND PERSONS COMMENTINGAGENCIES AND PERSONS COMMENTINGAGENCIES AND PERSONS COMMENTINGAGENCIES AND PERSONS COMMENTINGON DRAFT RECOVERY PLANON DRAFT RECOVERY PLANON DRAFT RECOVERY PLANON DRAFT RECOVERY PLANON DRAFT RECOVERY PLAN

Critical review of planning documents bythose that must implement them and otherswith relevant expertise is essential to producingmanagement plans that are scientifically credibleand feasible to implement. This appendix listsagencies and persons who reviewed the draftMexican Spotted Owl Recovery Plan and pro-vided comments on the document. Thesecomments are in large part responsible forproduction of a final Recovery Plan that theRecovery Team believes is much improved overthe draft version. Many comments were directlyresponsible for Recovery Plan revision, whilemany comments that were not incorporatedprovoked considerable thought and lively discus-

sion during the Recovery Plan-revision period.The Recovery Team is grateful to those whospent valuable time contributing to this finalRecovery Plan.

PEER REVIEWPEER REVIEWPEER REVIEWPEER REVIEWPEER REVIEW

The following persons were specifically askedto review the draft Recovery Plan or portionsthereof, as indicated. “Parts Reviewed” referredto below relates to the draft Recovery Plan, notthis document. Each persons’ affiliation is listed.In addition, scientific and professional organiza-tions that requested an individual’s review are soindicated.

PPPPPararararartststststsCommentorCommentorCommentorCommentorCommentor RRRRReviewevieweviewevieweviewededededed AffiliationAffiliationAffiliationAffiliationAffiliation OOOOOrganizationrganizationrganizationrganizationrganization

Forsman, E.D. All Pacific Northwest Research American Ornithologists’Station, U.S. Forest Service UnionCorvallis, OR

Gutiérrez, R.J. I, II Humboldt State UniversityArcata, CA

Holthausen, R. I,II, III U.S. Forest ServiceCorvallis, OR

King, R. II, III Rocky Mountain Forest andRange Experiment StationU.S. Forest ServiceFort Collins, CO

LaHaye, W. II.F, II.G Humboldt State University,Arcata, CA

Meslow, E.C. All Wildlife Management Institute The Wildlife Society;Corvallis, OR Wildlife Management

Institute

Morrison, M.L. II.G, II.H University of ArizonaTucson, AZ

170Volume I/Appendix E


PPPPPararararartststststsCommentorCommentorCommentorCommentorCommentor RRRRReviewevieweviewevieweviewededededed AffiliationAffiliationAffiliationAffiliationAffiliation OOOOOrganizationrganizationrganizationrganizationrganization

Pollock, K.H. II.E, III.A, North Carolina State UniversityIII.D Raleigh, NC

Raphael, M.G. I, II, III Pacific Northwest ResearchStation, U.S. Forest ServiceOlympia, WA

Stacey, P. II.F, II.G, University of NevadaIII Reno, NV

Verner, J. II.G, III Pacific Southwest ResearchStation, U.S. Forest ServiceFresno, CA

AGENCY REVIEWAGENCY REVIEWAGENCY REVIEWAGENCY REVIEWAGENCY REVIEW

The following agencies provided commentson the draft Recovery Plan. Federal agencies arelisted first, followed by State and County agen-cies. Agency name is followed by signator; otherreviewers are listed in ( ) when identified by theagency.

National Park Service, Southern Arizona Group;Jerry Belson, General Superintendent.(Benson, L.).

U.S. Bureau of Indian Affairs, Albuquerque AreaOffice; Patrick A. Hayes, Area Director.(Schwab, B.).

U.S. Bureau of Indian Affairs, Southern UteAgency; Charles A. Recker, Superinten-dent. (Friedley, J.; Recker, T.).

U.S. Bureau of Land Management, Utah; MatMillenbach, State Director.(Stringer, W.).

U.S. Fish and Wildlife Service, Arizona Ecologi-cal Services State Office; Sam F. Spiller,State Supervisor. (James, M.; Muiznieks,B.; Palmer, B.).

U.S. Fish and Wildlife Service, New MexicoEcological Services State Office; JenniferFowler-Propst, State Supervisor.(Torres, C.).

U.S. Fish and Wildlife Service, Mountain-PrairieRegion; James M. Lutey, Acting AssistantRegional Director, Ecological Services.

U.S. Forest Service, Southwestern Region;Charles W. Cartwright, Jr., RegionalForester. (Beyerhelm, C.; Birkland,C.; Briggs, A.; Brown, A.; Casper, L.;Cassidy, R.; Dargan, C.; Deaver, R.;DeLorenzo, D.; Derby, J.; Ellenwood, J.;Ewers, S.; Fletcher, R.; Gerritsma, J.;Green, H.; Herron, M.; Higgins, B.;Holbrook, C.; Hollis, H.; Holmstrom,D.; Johnson, D.; Kill, D.; Lucero, L.;MacIvor, J.; Madril, A.; Manthei, M.;Martinez, J.; Menasco, K.; Nelson, J.;Randall-Parker, T.; Rethlake, K.; Rolf, J.;Schaal, L.; Shafer, J.; Sheppard, G.;Spoerl, P.; Stahn, R.; Taylor, C,; Vigil,G.; Wistrand, H.; Zumwalt, M.).

U.S. Forest Service, Intermountain Region; DaleN. Bosworth, Regional Forester. (Botts,J.; Egnew, A.; Grandison, K.; Gray, S.;Hayman, R.).



U.S. Forest Service, Rocky Mountain Region;Elizabeth Estill, Regional Forester.(Player, R.).

Arizona Game and Fish Department; Duane L.Shroufe, Director. (Johnson, T.).

Arizona State Land Department; M. JeanHassell, Commissioner.

Otero County Commission; Richard L. Zierlein,Chairman.

San Juan County Commission;Ty Lewis, Chairman.

OTHER GROUPSOTHER GROUPSOTHER GROUPSOTHER GROUPSOTHER GROUPSAND INDIVIDUALSAND INDIVIDUALSAND INDIVIDUALSAND INDIVIDUALSAND INDIVIDUALS

Applied Ecosystem Management, Flagstaff, AZ;Tod Hull. (On behalf of: NorthernArizona Loggers Association; PrecisionPine and Timber Company; ReidheadBrothers Lumber Company; StoneForest Industries).

Defenders of Wildlife, Washington, DC; RobertM. Ferris, Director, Species ConservationDivision; Gregory J. Sater, WildlifeCounsel, Legal Division.

Forest Conservation Council, Santa Fe, NM;John Talberth, Executive Director.

Southwest Center for Biological Diversity, SilverCity, NM; Kieran Suckling, ExecutiveDirector. (Attachment signed by DennisMorgan, Research Associate, SouthwestCenter for Biological Diversity; KieranSuckling, Executive Director, SouthwestCenter for Biological Diversity; SharonGalbreath, Chairperson, Grand CanyonChapter, Sierra Club; Samuel Hitt,Director, Forest Guardians; Joanie Berde,Carson Forest Watch; Tom Ribe, PublicForestry Foundation; Tom H. Wootten,Conservation Chair, Mesilla ValleyAudubon Society; Mary Lou Jones, ZuniMountain Coalition; Joseph Feller; DaveHenderson, Southwest Forest Alliance;Charles Babbitt, President, MaricopaAudubon Society; John Talberth, Execu-tive Director, Forest ConservationCouncil; Jim Powers, Prescott NationalForest Friends; Gary Simpson, NorthernNew Mexico Chapter, WildernessWatch; Eleanor G. Wooten, Vice Presi-dent, T & E, Inc.; Gwen Wardwell,Director, Rio Grande Chapter, SierraClub; Mike Siedman; Jeff Burgess.).

Stone Forest Industries, Flagstaff, AZ; Steve C.Bennett, Regional Manager.

Mark Herron, Santa Fe, NM.

Terry Johnson, Los Alamos, NM.

Dennis R. Kingsbury, Munds Park, AZ.


Mexican Spotted Owl Recovery PlanSupporting Documents

Volume II


MEXICAN SPOTTED OWL RECOVERY PLANMEXICAN SPOTTED OWL RECOVERY PLANMEXICAN SPOTTED OWL RECOVERY PLANMEXICAN SPOTTED OWL RECOVERY PLANMEXICAN SPOTTED OWL RECOVERY PLANVolume II - Technical Supporting InformationVolume II - Technical Supporting InformationVolume II - Technical Supporting InformationVolume II - Technical Supporting InformationVolume II - Technical Supporting Information

This volume consists of chapters on aspects of Mexican spotted owl natural history that weredeveloped during preparation of the Mexican Spotted Owl Recovery Plan (Recovery Plan). Thesetreatises provide much of the information upon which the Recovery Plan, especially the recoveryrecommendations in Part III of Volume I, are based. Much of the material in Volume II is highlytechnical in nature and, although important, was not considered appropriate for inclusion in a workingimplementation document like Volume I of the Recovery Plan. The Mexican Spotted Owl RecoveryTeam and the Fish and Wildlife Service believe that these technical papers should, however, be availableto anyone who would like a detailed account of subject matter contained herein.

Since the material detailed in Volume II was integral in developing the Recovery Plan, the salientpoints from each of the Volume II chapters are summarized in Part II of Volume I. This makesVolume I a stand-alone document containing the most relevant information needed to implement theRecovery Plan and understand the reasons behind the management recommendations containedtherein.

Citations of material contained in this volume should read as follows:

[Author(s)] 1995. Pages [ - ] in USDI Fish and Wildlife Service. Mexican Spotted Owl RecoveryPlan, Volume II.

CONTENTSCONTENTSCONTENTSCONTENTSCONTENTS

Chapter 1: DChapter 1: DChapter 1: DChapter 1: DChapter 1: Distristristristristribution and Aibution and Aibution and Aibution and Aibution and Abundancebundancebundancebundancebundance James P. Ward, Jr., Alan B. Franklin, Sarah E. Rinkevich, and Fernando Clemente ............... 14 pages

Chapter 2: PChapter 2: PChapter 2: PChapter 2: PChapter 2: Population Bopulation Bopulation Bopulation Bopulation BiologiologiologiologiologyyyyyGary C. White, Alan B. Franklin, and James P. Ward, Jr. ..................................................... 25 pages

Chapter 3: Landscape Analysis and MChapter 3: Landscape Analysis and MChapter 3: Landscape Analysis and MChapter 3: Landscape Analysis and MChapter 3: Landscape Analysis and Metapopulation Setapopulation Setapopulation Setapopulation Setapopulation StrtrtrtrtructuructuructuructuructureeeeeTim Keitt, Alan B. Franklin, and Dean Urban ................................................................... 16 pages

Chapter 4: HChapter 4: HChapter 4: HChapter 4: HChapter 4: Habitat Rabitat Rabitat Rabitat Rabitat RelationshipselationshipselationshipselationshipselationshipsJoseph L. Ganey and James L. Dick, Jr. ............................................................................... 42 pages

Chapter 5: PChapter 5: PChapter 5: PChapter 5: PChapter 5: Prrrrrey Eey Eey Eey Eey EcologcologcologcologcologyyyyyJames P. Ward, Jr. and William M. Block ............................................................................ 48 pages

Volume II i


Volume II/List of Tables

List of TablesList of TablesList of TablesList of TablesList of Tables

Chapter 1Chapter 1Chapter 1Chapter 1Chapter 1

TTTTTable 1.1. able 1.1. able 1.1. able 1.1. able 1.1. Historical records and minimum numbers of Mexican spottedowls found during planned surveys ................................................................................................. 2-3


TTTTTable 2.1.able 2.1.able 2.1.able 2.1.able 2.1. Time periods and number of capture histories from threestudy areas used in estimating Mexican spotted owl survival. ............................................................. 5

TTTTTable 2.2. able 2.2. able 2.2. able 2.2. able 2.2. Apparent survival and recapture probability from threestudy areas used in estimating Mexican spotted owl survival .............................................................. 6

TTTTTable 2.3.able 2.3.able 2.3.able 2.3.able 2.3. Description of radio-telemetry studies conducted on adultand subadult Mexican spotted owls ................................................................................................... 8

TTTTTable 2.4.able 2.4.able 2.4.able 2.4.able 2.4. Estimates of true survival for Mexican spotted owlsbased on radio-telemetry data. .............................................................................................. 9

TTTTTable 2.5.able 2.5.able 2.5.able 2.5.able 2.5. Number of management territories checked one or moretimes during the Mexican spotted owl monitoring program of FS Region 3 .................................... 10

TTTTTable 2.6.able 2.6.able 2.6.able 2.6.able 2.6. Persistence of a Mexican spotted owl pair on a territoryfor formal and informal monitoring data ........................................................................................ 10

TTTTTable 2.7.able 2.7.able 2.7.able 2.7.able 2.7. Persistence of a pair of Mexican spotted owls on a territoryfor the five recovery units contained in the FS Region 3 monitoring data base ................................ 11

TTTTTable 2.8.able 2.8.able 2.8.able 2.8.able 2.8. Estimates of number of young fledged/pair and fecundityon two demographic study areas ..................................................................................................... 13

TTTTTable 2.9.able 2.9.able 2.9.able 2.9.able 2.9. Mean number of young fledged from 1989-1993,based on FS monitoring data per pair ............................................................................................. 14

TTTTTable 2.10. able 2.10. able 2.10. able 2.10. able 2.10. Mean number of young produced/pair from 1989-1993,based on monitoring data ............................................................................................................... 15

TTTTTable 2.11.able 2.11.able 2.11.able 2.11.able 2.11. Mean number of young produced per territory from 1989-1993,based on monitoring data ............................................................................................................... 15

TTTTTable 2.12.able 2.12.able 2.12.able 2.12.able 2.12. Parameters used to estimate ��for Mexican spottedowls on the GSA and CSA study areas ............................................................................................. 20

TTTTTable 2.13.able 2.13.able 2.13.able 2.13.able 2.13. Estimates and standard errors of ��for female Mexicanspotted owls on the CSA and GSA study areas ................................................................................ 20

ii


TTTTTable 2.14.able 2.14.able 2.14.able 2.14.able 2.14. Estimates of population parameters for Mexicanspotted owls on the GSA and CSA study areas ................................................................................ 21

TTTTTable 2.15.able 2.15.able 2.15.able 2.15.able 2.15. Number of Mexican spotted owl territories checkedand the percent of them occupied by a pair of owls for 1989-1993.................................................. 22


TTTTTable 4.1.able 4.1.able 4.1.able 4.1.able 4.1. Percent of nest and roost sites in various vegetationtypes in different Recovery Units ...................................................................................................... 4

TTTTTable 4.2.able 4.2.able 4.2.able 4.2.able 4.2. Home-range sizes of radio-marked Mexican spotted owls ............................................ 6-7

TTTTTable 4.3. able 4.3. able 4.3. able 4.3. able 4.3. Size of nocturnal activity centers of radio-tagged pairs ofMexican spotted owls in the Upper Gila Mountains and Basin andRange-East Recovery Units ............................................................................................................... 8

TTTTTable 4.4. able 4.4. able 4.4. able 4.4. able 4.4. Use of forest types for foraging by radio-tagged Mexican spottedowls on three study areas in northern Arizona. ................................................................................ 10

TTTTTable 4.5. able 4.5. able 4.5. able 4.5. able 4.5. Selected characteristics of nest stands of Mexican spotted owls inthe Upper Gila Mountains and Basin and Range-East Recovery Units ............................................. 12

TTTTTable 4.6. able 4.6. able 4.6. able 4.6. able 4.6. Habitat characteristics sampled at Mexican spotted owl nest siteson three Ranger districts, Apache-Sitgreaves National Forest, Upper GilaMountains Recovery Unit ............................................................................................................... 14

TTTTTable 4.7.able 4.7.able 4.7.able 4.7.able 4.7. Habitat characteristics at Mexican spotted owl nest andrandomly located sites in the Tularosa Mountains, New Mexico; UpperGila Mountains RU. ....................................................................................................................... 15

TTTTTable 4.8a. able 4.8a. able 4.8a. able 4.8a. able 4.8a. Habitat characteristics of Mexican spotted owl nest sitesin the Upper Gila Mountains Recovery Unit................................................................................... 20

TTTTTable 4.8b. able 4.8b. able 4.8b. able 4.8b. able 4.8b. Habitat characteristics of Mexican spotted owl nest sitesin the Basin and Range-East Recovery Unit .................................................................................... 21

TTTTTable 4.9.able 4.9.able 4.9.able 4.9.able 4.9. Habitat characteristics on circular plots within home rangesof radio-tagged Mexican spotted owls ............................................................................................. 23

TTTTTable 4.10. able 4.10. able 4.10. able 4.10. able 4.10. Habitat characteristics at Mexican spotted owl roost andrandom sites in the Tularosa Mountains, New Mexico; Upper GilaMountains RU ............................................................................................................................... 24

TTTTTable 4.11a. able 4.11a. able 4.11a. able 4.11a. able 4.11a. Selected characteristics of roost sites used by radio-taggedMexican spotted owls in the Colorado Plateau Recovery Unit. ........................................................ 27

TTTTTable 4.11b.able 4.11b.able 4.11b.able 4.11b.able 4.11b. Selected characteristics of roost sites used by radio-taggedMexican spotted owls in the Upper Gila Mountains Recovery Unit ................................................. 28

Volume II/List of Tables iii


iv

TTTTTable 4.11c. able 4.11c. able 4.11c. able 4.11c. able 4.11c. Selected characteristics of roost sites used by radio-taggedMexican spotted owls in the Basin and East-Range Recovery Unit .................................................. 29

TTTTTable 4.12.able 4.12.able 4.12.able 4.12.able 4.12. Seasonal roost site characteristics of radio-tagged Mexicanspotted owls in ponderosa pine-Gambel oak forest, Arizona (UpperGila Mountains Recovery Unit) ...................................................................................................... 31

TTTTTable 4.13.able 4.13.able 4.13.able 4.13.able 4.13. Characteristics of roost sites used by two migrant Mexicanspotted owls on their winter range .................................................................................................. 32


TTTTTable 5.1. able 5.1. able 5.1. able 5.1. able 5.1. Relative frequency of prey items found in the diet of Mexicanspotted owls occurring in the northern portion of the subspecies' range ............................................ 3

TTTTTable 5.2. able 5.2. able 5.2. able 5.2. able 5.2. Relative frequency of prey items found in the diet of Mexicanspotted owls occurring in the central portion of the subspecies' range ............................................... 4

TTTTTable 5.3. able 5.3. able 5.3. able 5.3. able 5.3. Relative frequency of prey items found in the diet of Mexicanspotted owls occurring in the southern portion of the subspecies' range ............................................ 5

TTTTTable 5.4. able 5.4. able 5.4. able 5.4. able 5.4. Percent of prey biomass in the diet of Mexican spotted owlsoccurring in the northern portion of the subspecies' range ................................................................ 6

TTTTTable 5.5. able 5.5. able 5.5. able 5.5. able 5.5. Percent of prey biomass in the diet of Mexican spotted owlsoccurring in the central portion of the subspecies' range ................................................................... 7

TTTTTable 5.6. able 5.6. able 5.6. able 5.6. able 5.6. Percent of prey biomass in the diet of Mexican spotted owlsoccurring in the southern portion of the subspecies' range ................................................................ 8

TTTTTable 5.7.able 5.7.able 5.7.able 5.7.able 5.7. Animal species consumed by Mexican spotted owls in18 geographic areas ........................................................................................................................ 10

TTTTTable 5.8. able 5.8. able 5.8. able 5.8. able 5.8. Factors influencing trends observed in Mexican spottedowl diets ................................................................................................................................... 14-15

TTTTTable 5.9.able 5.9.able 5.9.able 5.9.able 5.9. Factors influencing production of Mexican spotted owlsin northern Arizona, the Sacramento Mountains, New Mexico, and theTularosa Mountains, New Mexico, during 1991, 1992, and 1993 ............................................. 17-18

TTTTTable 5.10.able 5.10.able 5.10.able 5.10.able 5.10. Prey comprising �10% of relative frequency or biomassin the diet of Mexican spotted owls ................................................................................................. 21

TTTTTable 5.11.able 5.11.able 5.11.able 5.11.able 5.11. Descriptive statistics of selected habitat variablescharacterizing habitats of common mammalian prey of Mexican spotted owls ........................... 32-33

TTTTTable 5.12.able 5.12.able 5.12.able 5.12.able 5.12. Habitat characteristics used by deer mice in threevegetation communities of the Sacramento Mountains, New Mexico ............................................ 34

Volume II/List of Tables


TTTTTable 5.13.able 5.13.able 5.13.able 5.13.able 5.13. Habitat characteristics used by Mexican woodrats in twovegetation communities of the Sacramento Mountains, New Mexico .............................................. 35

TTTTTable 5.14.able 5.14.able 5.14.able 5.14.able 5.14. Habitat characteristics used by long-tailed voles in twovegetation communities of the Sacramento Mountains, New Mexico .............................................. 36

TTTTTable 5.15.able 5.15.able 5.15.able 5.15.able 5.15. Habitat characteristics used by Mexican voles in threevegetation communities of the Sacramento Mountains, New Mexico .............................................. 37

TTTTTable 5.16.able 5.16.able 5.16.able 5.16.able 5.16. Habitat characteristics used by brush mice in xeric forestsof the Sacramento Mountains, New Mexico. ................................................................................... 39

Volume II/List of Tables v


List of FiguresList of FiguresList of FiguresList of FiguresList of Figures


FFFFFigurigurigurigurigure 1.1. e 1.1. e 1.1. e 1.1. e 1.1. Recovery unit boundaries and current distribution of Mexicanspotted owls in the United States ..................................................................................................... 8

FFFFFigurigurigurigurigure 1.2. e 1.2. e 1.2. e 1.2. e 1.2. Current distribution of Mexican spotted owls in Mexico ............................................... 9

FFFFFigurigurigurigurigure 1.3. e 1.3. e 1.3. e 1.3. e 1.3. Density of (a)(a)(a)(a)(a) all three subspecies compared among regionsand (b)(b)(b)(b)(b) Mexican spotted owls among three forest types in the SacramentoMountains, New Mexico ................................................................................................................ 11


FFFFFigurigurigurigurigure 2.1.e 2.1.e 2.1.e 2.1.e 2.1. Location of Coconino, Gila, and Sky Islands Mexican spottedowl population study areas in Arizona and New Mexico ................................................................... 5

FFFFFigurigurigurigurigure 2.2.e 2.2.e 2.2.e 2.2.e 2.2. Changes in occupancy of formal and informal monitoring territoriesin the FS Region 3 monitoring database ......................................................................................... 23


FFFFFigurigurigurigurigure 3.1. e 3.1. e 3.1. e 3.1. e 3.1. Dispersal-distance relationship for radio-marked juvenile spotted owls ......................... 4

FFFFFigurigurigurigurigure 3.2.e 3.2.e 3.2.e 3.2.e 3.2. The relationship between correlation length and minimumjoining distance. .............................................................................................................................. 8

FFFFFigurigurigurigurigure 3.3.e 3.3.e 3.3.e 3.3.e 3.3. Landscape mosaics of discrete clusters of Douglas-fir and ponderosapine habitat types. ........................................................................................................................... 9

FFFFFigurigurigurigurigure 3.4. e 3.4. e 3.4. e 3.4. e 3.4. Change in correlation length due to cluster removal as a functionof distance ...................................................................................................................................... 10

FFFFFigurigurigurigurigure 3.5.e 3.5.e 3.5.e 3.5.e 3.5. Maps with clusters colored to illustrate their rank importance,weighted by (uncorrected for) patch area. ....................................................................................... 11

FFFFFigurigurigurigurigure 3.6.e 3.6.e 3.6.e 3.6.e 3.6. Patch importance to overall connectedness, normalized for patch area. ........................ 12


FFFFFigurigurigurigurigure 4.1. e 4.1. e 4.1. e 4.1. e 4.1. Diameter distributions of live trees sampled in nest stands inthe Upper Gila Mountains and Basin and Range - East Recovery Units ........................................... 11

FFFFFigurigurigurigurigure 4.2a.e 4.2a.e 4.2a.e 4.2a.e 4.2a. Tree species composition within nest- and random-standplots, Upper Gila Mountains Recovery Unit ................................................................................... 17

FFFFFigurigurigurigurigure 4.2b.e 4.2b.e 4.2b.e 4.2b.e 4.2b. Tree species composition within nest- and random-standplots, Basin and Range-East Recovery Unit ..................................................................................... 18

viVolume II/List of Figures


Volume II/List of Figures vii

FFFFFigurigurigurigurigure 4.3. e 4.3. e 4.3. e 4.3. e 4.3. Diameter distributions of live trees sampled on nest,nest stand, and random stand plots ................................................................................................. 19


FFFFFigurigurigurigurigure 5.1.e 5.1.e 5.1.e 5.1.e 5.1. Cumulative distributions of prey in the diet ofMexican spotted owls ..................................................................................................................... 11

FFFFFigurigurigurigurigure 5.2.e 5.2.e 5.2.e 5.2.e 5.2. Geographic variability in the food habits ofMexican spotted owls ..................................................................................................................... 12

FFFFFigurigurigurigurigure 5.3.e 5.3.e 5.3.e 5.3.e 5.3. Reproductive success of Mexican spotted owls in theSacramento Mountains, New Mexico, northern Arizona, and TularosaMountains, New Mexico ................................................................................................................ 19

FFFFFigurigurigurigurigure 5.4.e 5.4.e 5.4.e 5.4.e 5.4. Biomass of (a)(a)(a)(a)(a) common prey occurring inmixed-conifer forests, (b)(b)(b)(b)(b) frequencies of peromyscid mice consumedby Mexican spotted owls, and (c)(c)(c)(c)(c) average number of owl young produced ...................................... 20

FFFFFigurigurigurigurigure 5.5.e 5.5.e 5.5.e 5.5.e 5.5. Average biomass of common prey of the Mexican spotted owl ..................................... 26

FFFFFigurigurigurigurigure 5.6. e 5.6. e 5.6. e 5.6. e 5.6. Trends in average density of common preyof the Mexican spotted owl ............................................................................................................. 29


Volume II/Chapter 1

CHAPTER 1: Distribution and AbundanceCHAPTER 1: Distribution and AbundanceCHAPTER 1: Distribution and AbundanceCHAPTER 1: Distribution and AbundanceCHAPTER 1: Distribution and Abundanceof Mexican Spotted Owlsof Mexican Spotted Owlsof Mexican Spotted Owlsof Mexican Spotted Owlsof Mexican Spotted Owls

James P. Ward, Jr., Alan B. Franklin, Sarah E. Rinkevich,and Fernando Clemente

Knowledge of the distribution and abun-dance of Mexican spotted owls can provideinsight into the subspecies’ geographic limits andhabitat requirements. For example, standardizedsurveys for northern spotted owls among differ-ent habitats have provided evidence of the owl’saffinity for older, densely layered forests(Forsman et al. 1977; 1987, Thomas et al. 1990,Blakesley et al. 1992). In addition, distributionand abundance patterns often provide a founda-tion for more intensive natural and life historystudies.

For this recovery plan, we gathered andexamined information on the distribution andabundance of Mexican spotted owls accumulatedthrough 1993. We used this information to (1)document historical and current extent of thissubspecies, (2) help formulate recovery unitboundaries, and (3) provide a template forlandscape-scale analyses.

SOURCES OF INFORMATIONSOURCES OF INFORMATIONSOURCES OF INFORMATIONSOURCES OF INFORMATIONSOURCES OF INFORMATION

The quality and quantity of informationregarding the distribution and abundanceMexican spotted owls varies by source. Histori-cal accounts exist from museum collections andanecdotal observations by early natural historiansfrom throughout the owl’s range (reviewed inMcDonald et al. 1991). These early observationsare useful for documenting the owl’s knownhistorical range. However, haphazard and fre-quently unknown methods by which the histori-cal information was obtained confound anyattempt to infer change in the owl’s abundancefrom historical to present time. Modern ac-counts exist from incidental observations pro-vided by amateur and professional biologists andfrom organized surveys conducted by naturalresource management or research personnel.Incidental observations are similar in quality tohistorical accounts, frequently lacking sufficientinformation for estimating population param-eters or testing empirical hypotheses. However,

when combined with results of planned surveysincidental observations can be used to documentthe current extent of the subspecies’ range.Results from planned surveys and demographicstudies have provided the best available data onthe owl’s abundance.

Planned surveys for Mexican spotted owls inthe United States have been conducted by land-management agencies since 1989 and by re-searchers in Mexico since 1992. Survey protocolswere reviewed in 1990 and a more formalprogram was subsequently developed to locateMexican spotted owls (USDA Forest Service1990). Owl demographic studies began in theSky Island Mountains of Arizona in 1990(Duncan et al. 1993), and in northern Arizona(Olson et al. 1993) and in the Tularosa Moun-tains of west-central New Mexico in 1991(Seamans et al. 1993).

To document the current (1990-1993)distribution of the Mexican spotted owl, wedefined an owl site as a visual sighting of at leastone adult spotted owl or as a minimum of twoauditory detections in the same vicinity in thesame year. Observations prior to 1990 areconsidered historical records for the purposes ofthis report. The methods and limitations of thesedata are discussed further in the White et al.1995.

HISTORICAL DISTRIBUTIONHISTORICAL DISTRIBUTIONHISTORICAL DISTRIBUTIONHISTORICAL DISTRIBUTIONHISTORICAL DISTRIBUTION

We compiled 600 and 35 historical recordsof Mexican spotted owls in the United Statesand Mexico, respectively (Table 1.1). We refer tothese as records and not as independent sitesbecause several observations may have beentallied for the same site. Incomplete informationof the owls’ locations prevented us from assign-ing each record to an individual site. Thus, theserecords cannot be used to estimate historicalabundance and are presented only to show theapproximate extent of the owl’s distributionbefore 1990.

1


Volume II/Chapter 1

FS 21 16BLM 6 10NPS 34 23Tribal 20 13b

New Mexico State 1 0Unknownc 5 0

SubtotalSubtotalSubtotalSubtotalSubtotal 8787878787 6262626262

FS 82 97NPS 13 0Tribal 0 --b

DOD 9 6Private 8 0Unknownc 57 0


FS 2 8BLM 0 6NPS 0 0Tribal 1 --b

Unknownc 17 0


Table 1.1.Table 1.1.Table 1.1.Table 1.1.Table 1.1. Historical records and minimum numbers of Mexican spotted owls found during plannedsurveys, and incidental observations by Recovery Unit and land ownership.

Number of owl sitesNumber of owl sitesNumber of owl sitesNumber of owl sitesNumber of owl sites1990 - 19931990 - 19931990 - 19931990 - 19931990 - 1993

FS 25 34NPS 3 0New Mexico State 1 0Private 4 0Unknownc 8 0


Number of Number of Number of Number of Number of owl recordsowl recordsowl recordsowl recordsowl recordsbefore 1990before 1990before 1990before 1990before 1990aaaaa

Recovery UnitRecovery UnitRecovery UnitRecovery UnitRecovery Unit

Colorado Plateau

Southern Rocky Mountains - Colorado

Southern Rocky Mountains - New Mexico

Upper Gila MountainsFS 138 424BLM 5 0NPS 5 0Tribal 20 --b

Private 1 0Unknownc 104 0


Basin and Range - West

2

UNITED STATESUNITED STATESUNITED STATESUNITED STATESUNITED STATES


Volume II/Chapter 1

Coahuila 4 0Nuevo Leon 4 1Tamaulipas 0 0d


600600600600600 758758758758758

FS 18 111BLM 1 0NPS 6 10Tribal 2 --b

FWS 1 0Private 2 0


Table 1.1.Table 1.1.Table 1.1.Table 1.1.Table 1.1. (continued)

Basin and Range - East


Sonora 8 9Chihuahua 10 8d

Sinaloa 1 0


Recovery UnitRecovery UnitRecovery UnitRecovery UnitRecovery UnitNumber of owl recordsNumber of owl recordsNumber of owl recordsNumber of owl recordsNumber of owl records

before 1990before 1990before 1990before 1990before 1990aaaaa

Number of owl sitesNumber of owl sitesNumber of owl sitesNumber of owl sitesNumber of owl sites1990 - 19931990 - 19931990 - 19931990 - 19931990 - 1993

United States TotalUnited States TotalUnited States TotalUnited States TotalUnited States Total

Sierra Madre Occidental - Norte

Sierra Madre Oriental - NorteCoahuila 2 0

Sierra Madre Occidental - SurDurango 2 0Aguascalientes 0 1d

Zacatecas 0 0d

San Luis Potosi 1 0Guanajuato 1 0

SubtotalSubtotalSubtotalSubtotalSubtotal 44444 11111Sierra Madre Oriental - Sur

Eje NeovolcanicoJalisco 1 0Colima 1e 0Michoacan 1 0Puebla 1e 0


3535353535 1919191919Mexico TotalMexico TotalMexico TotalMexico TotalMexico TotalaValues do not connote numbers of owls nor owl sites because multiple records may exist from the same site through time.bAdditional owls are known to exist on many Tribal lands but the exact number is unavailable.cLocations of these records were insufficient for assigning a land ownership.dAdditional sightings have been reported from 1994 surveys.eUnverified record not included in totals (see text).

3

UNITED STATES, continuedUNITED STATES, continuedUNITED STATES, continuedUNITED STATES, continuedUNITED STATES, continued


Volume II/Chapter 1

The general vicinity of all historical recordsis presented in the following section according toRU and land ownership. This informationshould help land managers to identify areas thatmay be occupied by Mexican spotted owls.Historical records for owls in the United Stateswere compiled from several sources which arecited below. In contrast, much of the historicalinformation for Mexico was taken from a com-prehensive summary by Williams and Skaggs(1993).

Colorado Plateau Recovery UnitColorado Plateau Recovery UnitColorado Plateau Recovery UnitColorado Plateau Recovery UnitColorado Plateau Recovery Unit

Prior to 1990, Mexican spotted owls wererecorded from Zion (Kertell 1977, Rinkevich1991), Canyonlands, Capitol Reef (Sue Linner,FWS, Salt Lake City Utah, pers. comm.), MesaVerde (Reynolds and Johnson 1994), and GrandCanyon National Parks (McDonald et al. 1991),Glen Canyon National Recreation Area (Behle1960), and the Cedar City, Richfield, and VernalDistricts of the BLM (Behle 1981; Sue Linner,FWS, Salt Lake City, Utah, pers. comm.). Table1.1 presents more detail regarding these records.The physical attributes of these historical owlsites include steep-sided, narrow-walled, orhanging canyons. The biotic attributes includeconiferous overstory in canyon bottoms with anunderstory comprised by Gambel oak, bigtoothmaple, boxelder, and scattered aspen groves nearwater (McDonald et al. 1991).

Historical accounts also place the Mexicanspotted owl on the Kaibab Plateau in Arizona(Ganey and Balda 1989, North Kaibab RangerDistrict, unpublished data), plus many sites inNew Mexico. These sites include Fence Lake(State), Frances Canyon (BLM), the ZuniMountains and Mount Taylor (Cibola NF), andwithin the Zuni and Navajo Nations (McDonaldet al. 1991, New Mexico Natural Heritage DataBase, Nature Conservancy, Albuquerque, NM).Generally, vegetation types reported for theseareas include montane coniferous forests withincanyon settings. The owl has been observed insteep-walled canyons with minimal vegetation aswell as in forested, steep-sloped canyons onBlack Mesa and in the Chuska Mountains.

Southern Rocky Mountains -Southern Rocky Mountains -Southern Rocky Mountains -Southern Rocky Mountains -Southern Rocky Mountains -Colorado Recovery UnitColorado Recovery UnitColorado Recovery UnitColorado Recovery UnitColorado Recovery Unit

Eighteen historical records of spotted owlsexist within this unit (Webb 1983, Reynolds1989). Most of these owls were found along theColorado Front Range extending northward toFort Collins. Two additional observations, oneeach from Rio Grande and San Juan NationalForests, plus one from the Southern Ute Reser-vation were recorded during 1989 surveys(Reynolds and Johnson 1994; Table 1.1).

Historical owl locations in this recovery unitoccurred in steep-sided canyons. These canyonsare typically broader with walls that are not asvertical as sites occupied by owls in southernUtah (Colorado Plateau RU). Northern aspectsof these canyons contain mixed-conifer forest,while southern aspects contain ponderosa pineand pinyon-juniper. Canyon bottoms containGambel oak and boxelder. Owl sightings insouthwestern Colorado were generally in can-yons that cut into mesas covered with pinyonand juniper. These canyon bottoms containedmixed-conifer or ponderosa pine-Gambel oakforests (McDonald et al. 1991).

Southern Rocky Mountains -Southern Rocky Mountains -Southern Rocky Mountains -Southern Rocky Mountains -Southern Rocky Mountains -New Mexico Recovery UnitNew Mexico Recovery UnitNew Mexico Recovery UnitNew Mexico Recovery UnitNew Mexico Recovery Unit

Mexican spotted owls are known historically(Table 1.1) from private and National Forestlands in the San Juan, Sangre de Cristo, andJemez Mountains, and near Taos and Sante Fe,New Mexico (Johnson and Johnson 1985,McDonald et al. 1991). Incidental observationsbetween 1979 and 1984 established the presenceof the owl at nine additional sites in the JemezMountains, Sante Fe National Forest (Johnsonand Johnson 1985). The owl has also beenobserved in Bandelier National Monument andnear Morhy Lake on State lands (Johnson andJohnson 1985). Historical spotted owl locationsthroughout all of New Mexico (including theBasin and Range - East RU) have been describedas “deep, narrow, timbered canyons with coolshady places, at elevations ranging from 6,500[1,982 sic] to 9,000 ft [2,744 m],” (McDonaldet al. 1991, after Ligon 1926).

4


Volume II/Chapter 1

Upper Gila Mountains Recovery UnitUpper Gila Mountains Recovery UnitUpper Gila Mountains Recovery UnitUpper Gila Mountains Recovery UnitUpper Gila Mountains Recovery Unit

Prior to a statewide survey conducted from1984 through 1988, only one Mexican spottedowl was recorded from the northern Arizona(San Francisco Peaks) portion of this RU (Huey1930). In contrast, several owls were reported forthe National Forest lands in the MogollonHighlands of east-central Arizona and west-central New Mexico (Ligon 1926, Skaggs 1988).Forests in those reports include the Apache-Sitgreaves, Gila, and Cibola National Forests(Table 1.1). A few observations were also re-corded on BLM land near Bitter Creek, GrantCounty and at the Gila Cliff Dwellings NationalMonument (NPS), both in New Mexico. Fol-lowing their survey of Arizona, Ganey and Balda(1989) reported 69 sites in the Arizona portionof the Mogollon Rim and in northern Arizona,including the Coconino, Kaibab, Tonto, andApache-Sitgreaves National Forests. Another 44records exist in the Arizona Heritage Data Base(Arizona Game and Fish Department, Phoenix,AZ). However, the latter records are distributedamong habitats similar to those reported byGaney and Balda (1989) and include several ofthe same sites.

Site characteristics for the historical Arizonalocations (Ganey and Balda 1989) reflect thefeatures described for other RU’s: mountainslopes with mixed-coniferous forest, steep-walledcanyons, or ponderosa pine-Gambel oak forest atelevations ranging from 1,525 to 2,925 m(5,000 to 9,590 ft). In southern New Mexico,Skaggs (1988) found cliffs present at 15 of the18 historical sites which he examined. However,it is unclear how many of these sites were locatedin the Upper Gila Mountains RU. Skaggs (1988)also noted a well developed understory ofbigtooth maple and Gambel oak dominated bymixed-conifer.

Basin and Range - West Recovery UnitBasin and Range - West Recovery UnitBasin and Range - West Recovery UnitBasin and Range - West Recovery UnitBasin and Range - West Recovery Unit

Historical records for Mexican spotted owlsin this RU include observations in the Huachucaand Chiricahua Mountains during the 1890s(reviewed in McDonald et al. 1991). These birdswere observed in a foothills-oak woodland and

in a fir tree in Pinery Canyon, respectively. Twoother sightings were recorded in lowland ripariancommunities including an owl nesting in cotton-woods northwest of Tucson in 1872 and near theSalt River in 1910 (Bendire 1892, Phillips et al.1964, McDonald et al. 1991).

More recent surveys found the owl occurringat 84 sites throughout southern Arizona (Ganeyand Balda 1989). Owls were located in rockycanyons or in several forest types at elevationsranging from 1,125 to 2,930 m (3,690 to 9,610ft) in the Atascosa-Pajarito, Santa Rita, SantaCatalina, Patagonia, Whetstone, Galiuro,Huachuca, Chiricahua, Pinaleno, Superstition,Sierra Ancha, Mazatzal, and Bradshaw Moun-tains, Arizona. Below 1,300 m (4,264 ft), spot-ted owls were found in steep canyons containingcliffs and stands of live oak, Mexican pine andbroad-leaved riparian vegetation (Ganey andBalda 1989). Above 1,800 m (5,904 ft) owlswere found in mixed-conifer and pine-oakforests. Mid-elevation observations included siteswith Arizona cypress and the other forest typespreviously mentioned. The Arizona HeritageData Base reports 78 additional records in manyof the same mountain ranges from 1974 to1989. Historical records on private land includeobservations near Animas Peak, Black BillSpring, and at the Gray Ranch, in New Mexico(Skaggs 1988).

Basin and Range - East Recovery UnitBasin and Range - East Recovery UnitBasin and Range - East Recovery UnitBasin and Range - East Recovery UnitBasin and Range - East Recovery Unit

Historical locations of Mexican spotted owlsoccur on lands of several jurisdictions in NewMexico: the Organ Mountains and near BitterCreek (BLM); the Sandia, Manzano, Sacra-mento, and Guadalupe Mountains in the Cibolaand Lincoln National Forests; and CarlsbadNational Park (Skaggs 1988). The owl has alsobeen found in Guadalupe National Park and onprivate land in the Davis Mountains of Texas(McDonald et al. 1991, Steve Runnels, TheHeard Natural Science Museum and WildlifeSanctuary, McKinney, TX, pers. comm.). Oneobservation each was also reported on the landsof the Mescalero Apache and at the SantoDomingo Pueblo (New Mexico Natural HeritageData Base, Nature Conservancy, Albuquerque,

5


Volume II/Chapter 1

NM). Physical and biotic characteristics werenot documented for these various sites. However,minimal notations describe mixed-conifer forestson eastern slopes of the Sacramento Mountains(Skaggs 1988). Canyons were mentioned forsites near the New Mexico-Texas border of theGuadalupe Mountains (McDonald et al. 1991).

Sierra Madre Occidental -Sierra Madre Occidental -Sierra Madre Occidental -Sierra Madre Occidental -Sierra Madre Occidental -Norte Recovery UnitNorte Recovery UnitNorte Recovery UnitNorte Recovery UnitNorte Recovery Unit

More than half of all historical records ofMexican spotted owls occurring in Mexico havebeen reported in this Recovery Unit (Table 1.1).Owls have been recorded from eight locations inthe State of Sonora prior to 1990. These birdsoccurred in the Sierras Pinitos, Azul, de los Ajos,San Luis, Aconchi, Oposura, and Huachinera(Williams and Skaggs 1993). The eighth recordwas reported from a ridge north of La Mesa,Mexico. All of these areas are physiographicallyand biotically similar to the Sky Island Moun-tains of southeastern Arizona (Cirett and Diaz1993). General elevations at or near thesesightings range from 1,950-2,340 m (6,500 to7,800 ft). Descriptions of sites at or near therecorded owls vary. Some descriptions includedense oaks, pine forest, pine-oak woodland,cliffs, and a spring with alders and sycamores(Williams and Skaggs 1993).

Ten historical records have been reportedfrom the State of Chihuahua (Table 1.1). Owlshave been collected, observed, or heard nearSierra Carcay, Arroyo Tinaja, Pacheco, SierraAzul, Colonia Garcia, Sierra del Nido, RanchoLa Estancia, Yaguirachic, Pinos Altos, andVosagota (Williams and Skaggs 1993). Eleva-tions at or near these sightings range from1,710-2,700 m (5,700 to 9,000 ft). Most ofthese observations were made in or near pine-oak woodlands (Williams and Skaggs 1993).Owls have also been reported in canyons withoaks and madrone, and in subalpine forestcomprised of Douglas-fir, true firs, oak, alder,pines, and chokecherry.

In Sinaloa, one spotted owl was observednear Rancho Liebre Barranca (Williams andSkaggs 1993). This area consists of deeplydissected barrancas with pine and oak forest at

higher altitudes and a mixture of temperate andtropical forest in canyon bottoms.

Sierra Madre Oriental -Sierra Madre Oriental -Sierra Madre Oriental -Sierra Madre Oriental -Sierra Madre Oriental -Norte Recovery UnitNorte Recovery UnitNorte Recovery UnitNorte Recovery UnitNorte Recovery Unit

Spotted owls have only been reported fromtwo sites within this Recovery Unit. Bothrecords are from the Sierra la Madera of centralCoahuila (Table 1.1). One owl was observedroosting in a “cliff-lined canyon bottom under adense canopy of maples and oaks” in Canada elAgua (Williams and Skaggs 1993). Another owlwas observed and heard in a “garden-like” arroyocontaining pines, oaks, and madrones (Williamsand Skaggs 1993). Elevations of these observa-tions were approximately 1,900 and 2,100 m(6,200 and 7,000 ft), respectively.

Sierra Madre Occidental -Sierra Madre Occidental -Sierra Madre Occidental -Sierra Madre Occidental -Sierra Madre Occidental -Sur Recovery UnitSur Recovery UnitSur Recovery UnitSur Recovery UnitSur Recovery Unit

Four historical records exist from the Statesof Durango, San Luis Potosi, and Guanajuato(Table 1.1). In Durango, spotted owls have beenobserved near Espinazo del Diablo in mixed-conifer forest and on two occasions in theMichilia Biosphere Reserve. One of the owlsfound in the reserve was observed roosting in alarge oak that was in a cool, wet ravine. Thesecond owl was found in a pine-oak forest.Remaining records are of two spotted owlscollected at approximately 2,400 m (8,000 ft)near Cerro Campanario, San Luis Potosi, andanother that was collected in the State ofGuanajuato (Williams and Skaggs 1993).

Sierra Madre Oriental -Sierra Madre Oriental -Sierra Madre Oriental -Sierra Madre Oriental -Sierra Madre Oriental -Sur Recovery UnitSur Recovery UnitSur Recovery UnitSur Recovery UnitSur Recovery Unit

In this Recovery Unit, eight records ofspotted owls have been reported from two States,Coahuila and Nuevo Leon (Table 1.1). Owlshave been heard and observed on several occa-sions east of Saltillo, Coahuila. Elevations ofthese records are generally higher than otherhistorical sightings and range from 2,700-3,060m (9,000 to 10,200 ft). The vegetation at thesesites has been described as oak-pine-conifer

6


Volume II/Chapter 1

woodland or as oak-pine woodland. Cliffs arepresent at all sites and chaparral or desert scrubvegetation can also be seen at two of these sites(Williams and Skaggs 1993). In addition, two ofthe four historical records from Nuevo Leon arefrom the mountains east of Saltillo and south ofMonterrey. At one of these sites a spotted owlwas heard calling from an oak-pine-coniferforest, approximately 2,550 m (8,500 ft) inelevation. The second was heard calling from an“oak-pine woodland mixed with Tamaulipanthorn woodland and scrub and palms on thecanyon cliffs” (Williams and Skaggs 1993). Theelevation at this site was 1,350 m (4,500 ft), thelowest to be recorded for spotted owls inMexico. The remaining two Nuevo Leon recordsare from the slopes of Cerro Potosi (Williamsand Skaggs 1993). One female spotted owl wascollected at 2,250 m (7,500 ft) in 1946. Anotherowl was heard in 1978 from a pine woodlandnear the summit at 3,630 m (12,100 ft), thehighest elevation recorded for a Mexican spottedowl.

Eje Neovolcanico Recovery UnitEje Neovolcanico Recovery UnitEje Neovolcanico Recovery UnitEje Neovolcanico Recovery UnitEje Neovolcanico Recovery Unit

Only two confirmed records of Mexicanspotted owls exist for the southernmost RecoveryUnit (Table 1.1). Specimens of spotted owls havebeen collected from the States of Jalisco andMichoacan. Two other records from the States ofColima and Puebla have been published(Enriquez-Rocha et al. 1993) but have not beenverified (Williams and Skaggs 1993).

In Jalisco, the owls were found on the northslope of Cerro Nevado de Colima in a park-likepine forest with broad-leaved oaks, and mesicground flora near 2,400 m (8,000 ft) elevation(Williams and Skaggs 1993). In Michoacan, theholotype of the subspecies and only existingState record was collected in 1903 from CerroTancitaro above 1,950 m (6,500 ft). Detailsregarding the spotted owls allegedly collected inColima and Puebla have not been reported(Enriquez-Rocha et al. 1993). We mention theselatter records because they suggest possibleextensions of the owl’s range. However, we havenot included them in the totals presented inTable 1.1.

CURRENT DISTRIBUTIONCURRENT DISTRIBUTIONCURRENT DISTRIBUTIONCURRENT DISTRIBUTIONCURRENT DISTRIBUTIONAND ABUNDANCEAND ABUNDANCEAND ABUNDANCEAND ABUNDANCEAND ABUNDANCE

Number of SitesNumber of SitesNumber of SitesNumber of SitesNumber of Sites

Surveys for Mexican spotted owls conductedfrom 1990 through 1993 indicate that thespecies persists in most locations reported priorto 1989. Notable exceptions include riparianhabitats in the lowlands of Arizona and NewMexico, and all previously occupied areas in thesouthern States of Mexico. As a result of plannedsurveys, additional sightings have been reportedfor all recovery units. New locations will un-doubtedly be reported following future surveys.

The current known range of the Mexicanspotted owl extends north from Aguascalientes,Mexico, through the mountains of Arizona, NewMexico, and western Texas to the canyons ofsouthern Utah, southwestern Colorado, and theFront Range of central Colorado (Figures 1.1and 1.2). Results from planned surveys andincidental observations conducted during 1990through 1993 indicate one or more owls havebeen observed at a minimum of 758 sites in theUnited States and 19 sites in Mexico (Table 1.1).

The greatest concentration of the knownsites in the United States occurs in the UpperGila Mountains Recovery Unit (55.9%) fol-lowed by the Basin and Range-East (16.0%),and Basin and Range-West (13.6%), ColoradoPlateau (8.2%), Southern Rocky Mountains -New Mexico (4.5%), and Southern RockyMountains - Colorado (1.8%) Recovery Units.Thus, fewer owl sites are currently known tooccur north of the Upper Gila MountainsRecovery Unit (12.7%) than to the south of thisrecovery unit (29.6%). In Mexico, the majorityof spotted owls have been documented in theSierra Madre Occidental - Norte RU (89.5%).However, the number of identified sites within agiven RU depends on survey intensity for whichwe have little reliable data. Therefore, the per-centages of sites within a given RU may notreflect the true relative abundance of Mexicanspotted owls.

7


Volume II/Chapter 1

Figure 1.1.Figure 1.1.Figure 1.1.Figure 1.1.Figure 1.1. Recovery unit boundaries and current distribution of Mexican spotted owls in theUnited States based on planned surveys and incidental observations recorded from 1990 through 1993.

FPO

8


Volume II/Chapter 1

FPO

Figure 1.2.Figure 1.2.Figure 1.2.Figure 1.2.Figure 1.2. Current distribution of Mexican spotted owls in Mexico based on planned surveys andincidental observations recorded from 1990 through 1993.

9


Volume II/Chapter 1

Number of OwlsNumber of OwlsNumber of OwlsNumber of OwlsNumber of Owls

A reliable estimate of the number of Mexicanspotted owls throughout its entire range is notavailable. Fletcher (1990) calculated that 2,074owls existed in Arizona and New Mexico in1990 using information gathered by the FS,Southwestern Region. McDonald et al. (1991)modified Fletcher’s (1990) calculations reportinga total of 2,160 owls in the United States. If oneassumes that all 758 sites included in this recov-ery plan were occupied by owl pairs, then at least1,516 adult or subadult owls were known toexist in the United States and 38 adult or sub-adult owls in Mexico from 1990 through 1993.These numbers are not reliable estimates ofcurrent population size because no measures ofbias or precision can be produced. Further, theamount of survey effort devoted to derivingthese numbers cannot be reliably calculated, noris an accurate measure available for areas orhabitats surveyed. Thus, we did not believe itwould be useful to estimate the size of theMexican spotted owl population given thelimited quality of data currently available. Atbest, our total numbers reported in Table 1.1represent a range for the minimum number ofowls known to exist during some portion of afour year period in the United States and Mexico(777 individuals if each site was occupied by asingle owl to 1,554 individuals if each site wasoccupied by a pair).

DensityDensityDensityDensityDensity

The abundance of any terrestrial organism ismore appropriately presented as density, thenumber of individuals per unit of area (Caughley1977), hereafter referred to as “crude density”(Franklin et al. 1990). Because the Mexicanspotted owl occupies a variety of habitatsthroughout its range, ecological density, thenumber of individuals per area of usable habitat,(Tanner 1978) would be a more meaningfulmeasure of abundance. At this time, rangewideestimates of either crude or ecological density ofMexican spotted owls cannot be provided for thesame reasons that population numbers cannot beestimated reliably.

Two estimates of crude density (D) existfrom studies conducted at either end of theUpper Gila Mountains Recovery Unit. Theseestimates were reported for the Coconino StudyArea (CSA) in northern Arizona and for the GilaStudy Area (GSA) in west-central New Mexicoby Gutiérrez et al. (1994). Density of adult andsubadult owls in both studies was estimatedusing a count of individuals divided by the sizeof the study area (CSA: 484 km2 [187 mi2];GSA: 323 km2 [125 mi2]). Identity of owls wasestablished by capturing and marking or bydirect observation at daytime roosts whenmarking was not possible. Methods were similarto those reported for northern spotted owls byFranklin et al. (1990) and Ward et al. (1991) andresulted only in density of territorial individuals.Only the 1993 estimates of density (D) arepresented here because the boundaries of theCSA study area were shifted between 1991 and1992.

D of Mexican spotted owls in 1993 was0.120 owls/km2 (0.310 owls/mi2) in the CSAand 0.180 owls/km2 (0.464 owls/mi2) in theGSA. The CSA has more ponderosa pine-Gambel oak forest (72.6%) and less mixed-conifer forest (14.4%) than the GSA (22.3%and 28.5%, respectively; Gutiérrez et al. 1994).The larger proportion of mixed-conifer maypartially explain higher owl density in the GSA.For comparison (Figure 1.3a), 1993 density ofCalifornia spotted owls in the San BernardinoMountains, California (LaHaye and Gutiérrez1994) was 0.118 owls/km2 (0.305 owls/mi2) anddensity of northern spotted owls in northwesternCalifornia (Franklin, unpublished data) was0.272 + 0.004 (SE) owls/km2 (0.703 + 0.009owls/mi2). Survey methods used in these latertwo studies were identical to those used in theCSA and GSA. Naive density estimates, thenumber of owls counted divided by study areasize, were used in this comparison (Figure 1.3a)except for the northern spotted owl population.In the latter case, a Jolly-Seber model was usedto estimate numbers of owls and an associatedsampling variance. The two density estimatorsare similar for spotted owls (Ward et al. 1991).

Combining the CSA and GSA density dataand weighting by area provides an averageestimate of 0.144 owls/km2 (0.372 owls/mi2)

^

^

^

10


Volume II/Chapter 1

Figure 1.3. Figure 1.3. Figure 1.3. Figure 1.3. Figure 1.3. Density of (a)(a)(a)(a)(a) all three subspecies compared among regions and (b)(b)(b)(b)(b) Mexican spottedowls among three forest types in the Sacramento Mountains, New Mexico, based on Skaggs and Raitt(1988). Sources of estimates for northern Arizona (CSA) and west-central New Mexico (GSA) are fromGutiérrez et al. (1994); southern California are from LaHaye and Gutiérrez (1994); and northwesternCalifornia are from Franklin (unpublished data). Vertical bars are 95% confidence intervals.

FPO

FPO

11


Volume II/Chapter 1

Arizona and New Mexico where it historicallyoccurred, nor in historically-documented areasof southern Mexico. Riparian communities andpreviously occupied localities in the southwest-ern United States and southern Mexico haveundergone significant habitat alteration since thehistorical sightings (USDI 1994). However, theamount of effort devoted to surveying these areasis poorly known and future surveys may docu-ment spotted owls. Surveys conducted to relo-cate spotted owls have been unsuccessful innorthern Colorado near Fort Collins and Boul-der, where records exist from the early 1970s and1980s, and in the Book Cliffs of east-centralUtah where owls were recorded in 1958.

The majority of Mexican spotted owlscurrently known to exist occur in the Upper GilaMountains Recovery Unit. This unit can beconsidered a critical nucleus for the subspeciesbecause of its central location within the owl’srange and its seemingly high density of owls.Other areas likely to be important include theSky-islands of southeastern Arizona and theSacramento Mountains, New Mexico (Basin andRange RUs). Throughout its range, most (91%)Mexican spotted owls occur on public landadministered by the FS.

Density estimates of Mexican spotted owlscontrasted among forest types in the SacramentoMountains and between two areas in the UpperGila Mountains RU suggest that mixed-conifersupports more owls compared to pine-oak, pine,and pinyon-juniper forest types. Mexican spot-ted owl densities reported from three areas aresimilar to those reported for California spottedowls occurring in the San Bernardino Moun-tains, California and slightly less than the den-sity of northern spotted owls occurring innorthwestern California.

Limited information inhibits reliable estima-tion of the absolute number of Mexican spottedowls. However, it is apparent from currentpatterns in distribution and habitat use that thesubspecies is rare relative to other raptors and isdistributed discontinuously throughout itsrange. Species existing under such conditions areconsidered vulnerable to extirpation (seeDawson et al. 1987 for discussion relevant tospotted owls). Although future efforts willundoubtedly discover additional owls, the extent

within the Upper Gila Maintains RU. However,this estimate should not be extrapolated to alarger area because (1) the CSA and GSA werenot randomly selected and (2) studies were notsufficiently replicated. Both of these problemscould severely bias extrapolated estimates be-cause the areas studied are not necessarily repre-sentative samples of the entire recovery unit orsubspecies’ range.

In another study, Skaggs and Raitt (1988)examined the density of Mexican spotted owlsamong three forest types in the SacramentoMountains (Basin and Range - East RU). Eigh-teen 23.1- km2 (9-mi2) quadrats, six in eachforest type, were surveyed for spotted owls.Quadrats were classified as pinyon-juniperwoodland, pine, or mixed-conifer forest accord-ing to the most common tree species within thequadrat. Owl density averaged across the threeforest types (x = 0.126 owls/km2 [0.325owls/mi2]) was similar to the average estimatefrom the two southwestern demographic studies.When partitioned by forest type, analysis ofthese data by the Team showed significantlyhigher densities (F = 16.93, df = 2, P = 0.0001)in the mixed-conifer (x = 0.275 owls/km2,SE = 0.046 [0.704 owls/mi2, SE = 0.117]) com-pared to the pine-dominated (x = 0.080owls/km2, SE = 0.028 [0.204 owls/mi2,SE = 0.073]) and pinyon-juniper habitats(x = 0.022 owls/km2, SE = 0.036 [0.056 owls/mi2, SE = 0.038]). Density was not statisticallydifferent between the latter two forest types(Figure 1.3b). Forest type explained 69.3% ofthe variation in owl density using this ANOVAmodel. Skaggs and Raitt (1988) showed similarresults using density of sites rather than owldensity.


The Mexican spotted owl currently occupiesa broad geographic area, but it does not occuruniformly throughout its range. Instead, the owloccurs in disjunct localities that correspond toisolated mountain systems and canyons. Thisdistribution mimics most historical locations,with a few exceptions. The owl has not beenreported along major riparian corridors in

12


Volume II/Chapter 1

and total number of this subspecies in theUnited States will likely not change in magni-tude enough to alter this conclusion. The con-trary is true for Mexico where planned surveyshave begun only recently.

Consequently, current strategies devel-oped to conserve Mexican spotted owls will belimited to basic knowledge about the owl’sdistribution. More specific recommendationswill require additional information on the totalpopulation size, population structure, or interac-tions among subpopulations.


Behle, W. H. 1960. The birds of southeasternUtah. Univ. Utah Biol. Surv. 12:1-56.

–—. 1981. The birds of Northeastern Utah.Occas. Publ. No. 2, Utah Mus. of Nat. Hist.,Univ. of Utah. Salt Lake City, Utah. 136pp.

Bendire, C. E. 1892. Life histories of NorthAmerican birds. U.S. Nat. Mus. Spec. Bull. 1.

Blakesley, J. B., A. B. Franklin, and R. J.Gutiérrez. 1992. Spotted owl roost and nestsite selection in northwestern California. J.Wildl. Manage. 56:388-392.

Caughley, G. 1977. Analysis of vertebratepopulations. John Wiley and Sons, London,England. 234pp.

Cirett-Galan, J. M., and E. R. Diaz. 1993.Estatus y distribucion del buho manchadoMexicano (Strix occidentalis lucida) en Sonora,Mexico. Centro Ecológico de Sonora. 66pp.

Dawson, W. R., J. D. Ligon, J. R. Murphy, J. P.Myers, D. Simberloff, and J. Verner. 1987.Report of the scientific advisory panel on thespotted owl. Condor 89:205-229.

Duncan, R. B., S. M. Speich, and J. D. Taiz.1993. Banding and blood sampling study ofMexican spotted owls in southeastern Arizona.Annual report submitted to Coronado Nat.For., Tucson, Ariz. 21pp.

Enriquez-Rocha, P., J. L. Rangel-Salazar, and D.W. Holt. 1993. Presence and distribution ofMexican owls: a review. J. Raptor Res.27:154-160.

Fletcher, K. W. 1990. Habitats used, abundanceand distribution of the Mexican spotted owl,Strix occidentalis lucida, on National Forest

system lands. U.S. For. Serv., SouthwesternRegion, Albuquerque, N.M. 55pp.

Forsman, E. D., E. C. Meslow, and M. J. Strub.1977. Spotted owl abundance in young versusold-growth forests, Oregon. Wildl. Soc. Bull.5:43-47.

Forsman, E. D., C. R. Bruce, M. A. Walter, andE. C. Meslow. 1987. A current assessment ofthe spotted owl population in Oregon.Murrelet 68:51-54.

Franklin, A. B., J. P. Ward, R. J. Gutiérrez, andG. I. Gould, Jr. 1990. Density of northernspotted owls in northwestern California. J.Wildl. Manage. 54:1-10.

Ganey, J. L., and R. P. Balda. 1989. Distributionand habitat use of Mexican spotted owls inArizona. Condor 91:355-361.

Gutiérrez, R. J., D. R. Olson, and M. E.Seamans. 1994. Demography of two Mexicanspotted owl populations in Arizona and NewMexico. Report submitted to Humboldt StateUniv. Foundation, Arcata, Calif. 29pp.

Huey, L. M. 1930. Notes from the vicinity ofSan Francisco Mountain, Arizona. Condor32:128.

Johnson, J. A., and T. H. Johnson. 1985. Thestatus of the spotted owl in northern NewMexico. New Mexico Dept. Game and Fish,Sante Fe, N.M. 39pp.


LaHaye, W. S., and R. J. Gutiérrez. 1994. BigBear spotted owl study, 1993. Reportsubmitted to Calif. Dept. of Fish and Game,Non-game Bird and Mammal Wildl. Manage.Div., Sacramento, Calif. 12pp.

Ligon, J. S. 1926. Habits of the spotted owl(Syrnium occidentalis). Auk 43:421-429.

McDonald, C. B., J. Anderson, J. C. Lewis, R.Mesta, A. Ratzlaff, T. J. Tibbitts, and S. O.Williams. 1991. Mexican spotted owl (Strixoccidentalis lucida) status report. U.S. Fish andWildl. Serv., Albuquerque, N.M. 85pp.

Olson, D. R., M. E. Seamans, and R. J.Gutiérrez. 1993. Demography of twoMexican spotted owl (Strix occidentalis lucida)populations in Arizona and New Mexico:preliminary results, 1992. Humboldt StateUniv., Arcata, Calif.

13


Volume II/Chapter 1

Phillips, A. R., J. T. Marshall, Jr., and G.Monson. 1964. The birds of Arizona. Univ. ofAriz. Press., Tucson, Ariz.

Reynolds, R. T., 1989. Preliminary survey of thedistribution and habitat of Mexican spottedowls in Colorado. U.S. For. Serv. Rocky Mtn.For. and Range Exp. Stn., Laramie, Wyo.19pp.

——., and C. L. Johnson. 1994. Distributionand ecology of the Mexican spotted owl inColorado: annual report. U.S. For. Serv.Rocky Mtn. For. and Range Exp. Stn., Ft.Collins, Colo. 14pp.


Seamans, M. E., D. R. Olson, and R. J.Gutiérrez. 1993. Demography of twoMexican spotted owl (Strix occidentalis lucida)populations in Arizona and New Mexico:preliminary results, 1992. Humboldt StateUniv., Arcata, Calif.

Skaggs, R. W., 1988. Status of the Spotted Owlin southern New Mexico: 1900-1987. NewMexico Dept. Game and Fish, Sante Fe, N.M.

——., and R. J. Raitt. 1988. A spotted owlinventory of the Lincoln National Forest,Sacramento Division, 1988. New MexicoDept. Game and Fish, Sante Fe, N.M. 12pp.

Tanner, J. T. 1978. Guide to the study of animalpopulations. Univ. Tennessee Press, Knoxville,Tenn. 186pp.

Thomas, J. W., E. D. Forsman, J. B. Lint, E. C.Meslow, B. R. Noon, and J. Verner. 1990. Aconservation strategy for the spotted owl. U.S.Gov. Print. Off., Washington, D.C. 427pp.

USDA Forest Service. 1990. Managementguidelines and inventory and monitoringprotocols for the Mexican spotted owl in theSouthwest Region. Federal Register55:27278-27287.

USDI. 1994. The impact of federal programs onwetlands. Vol II., A report to Congress by theSecretary of the Interior. U. S. Dept. ofInterior, Washington, D.C.

Ward, J. P., Jr., A. B. Franklin, and R. J.Gutiérrez. 1991. Using search time andregression to estimate abundance of territorialspotted owls. Ecol. Applic. 1:207-214.

Webb, B. 1983. Distribution and nesting re-quirements of montane forest owls in Colorado. Part IV: spotted owl (Strix occidentalis).Colo. Field Ornithol. 17:2-8.

White, G.C., A. Franklin, and J.P. Ward, Jr.,1995. Population biology. Pages ___ - __ inRecovery plan for the Mexican spotted owl.Vol. . USDI, U.S. Fish and Wildl. Serv.,Albuquerque, N.M.

Williams, S. O., III, and R. W. Skaggs. 1993.Distribution of the Mexican spotted owl (Strixoccidentalis lucida) in Mexico. Unpubl. Tech.Rep., New Mexico Dept. Game and Fish,Sante Fe, New Mexico. 38pp.

14

Volume II/Chapter 2


CHAPTER 2: Population BiologyCHAPTER 2: Population BiologyCHAPTER 2: Population BiologyCHAPTER 2: Population BiologyCHAPTER 2: Population BiologyGary C. White, Alan B. Franklin,

and James P. Ward, Jr.

The development of recovery guidelinesand criteria for Mexican spotted owl populationsrequires understanding the life history traits andpopulation processes for this species. We exam-ined the characteristics and trends of Mexicanspotted owl populations at three spatial scales:range-wide, regional, and local. Range-widescales correspond to characteristics and processesoccurring across the U.S. geographic range ofthe subspecies. Regional scales correspond torecovery units that were delineated according tobroad ecological patterns (see Rinkevich et al.1995). Local scales roughly correspond tosubpopulations that occur within each recoveryunit. Ideally, the same population characteristicsand processes should be examined over differenttemporal and spatial scales. However, temporalcomparisons were restricted because historicalinformation on Mexican spotted owl popula-tions does not exist. Thus, we could not comparehistorical and current populations to assess theimpacts of past and present management activi-ties.

In this section, we first review the sourcesof information available for making inferencesabout Mexican spotted owl populations. Wethen quantitatively describe the life historycharacteristics of the owl using age- and sex-specific survival probabilities and fecundity rates.These characteristics were estimated from datacollected during radio-telemetry studies, bandingstudies, and a regional FS monitoring program.We examined population trends first by estimat-ing the finite rate of population change (�) on alocal scale using our estimates of survival andfecundity from selected studies and then byevaluating temporal trends in occupancy rates ofFS management territories (MTs) on a regionalscale. Finally, we attempted to examine relation-ships of vegetation type on reproductive outputat a range-wide scale. Through this stepwiseprocedure, we evaluated the current state ofMexican spotted owl populations.

SOURCES OF INFORMATIONSOURCES OF INFORMATIONSOURCES OF INFORMATIONSOURCES OF INFORMATIONSOURCES OF INFORMATION

Since 1989, organized surveys for Mexicanspotted owls have been conducted by personnelof the FS, BLM, NPS, Tribes, State wildlifeagencies, and by private researchers. Mostsurveys followed the procedures described for theFS Region 3 (USDA Forest Service 1990).Under the FS Region 3 system, two types ofsurveys, inventory and monitoring, were con-ducted for different purposes. Inventories weregeneral surveys used to detect the presence ofspotted owls within a defined area. Monitoringspecifically assessed temporal changes in siteoccupancy and reproduction by spotted owls inmanagement territories (MT). Both proceduresrequired adherence to a standard survey proto-col. In addition, two types of monitoring,“formal” and “informal,” were utilized, withformal monitoring following guidelines ofUSDA Forest Service (1990). MTs were moni-tored each year from 1989 through 1993 withthe formal procedures if they had (1) no previ-ous management activity, (2) activity5-20 years prior to monitoring, or (3) recentactivity within 5 years of monitoring. Formalmonitoring resulted in more survey effort perMT than for informal monitoring (althoughsome informal monitoring sites were visitedmore often than required by formal monitor-ing). Informal monitoring could be conducted atany site not formally monitored in any year. Thegreatest amount of effort for surveys was devotedto formal monitoring, followed by informalmonitoring, and then inventory.

A limitation of the FS database was that theMTs surveyed were not randomly sampled fromall possible MTs. MTs were added to the data-base as owls were found. We do not expectexcessive bias in estimated fecundity or persis-tence (as defined in Life History Parameterssection) from this nonrandom sample becausethese parameters are probably not directly linkedto the sample selection procedures. We do expect

1Volume II/Chapter 2

Volume II/Chapter 2


System (GIS) were required to document owldistribution and conduct habitat analyses atvarious spatial scales. Unfortunately, the originaldata and GIS-compatible coordinates were notentered into a computerized database. Thus, theTeam attempted to compile and create its owndatabase (referred to as the Team database) fromoriginal data forms and plotted locations tosupplement summaries provided by the FSRegion 3 Office (referred to as the FS database).

Attempts to create the Team database metwith limited success. Entering a portion of thedata forms by a professional data-entry servicefailed because standard codes and recordinginstructions were not followed by data collectors.Further, various details were not recorded andmaps were missing. In a second attempt, theTeam formed and trained a set of crews to visitFS Ranger Stations and Supervisor Offices toenter data from forms and maps and to recordUniversal Transverse Mercator (UTM) coordi-nates of owl locations associated with surveys.The same information was collected from landmanagement agencies throughout the owl’srange for the four-year period (1990-1993).Year-end summaries from the FS Region 3 werecompared to output from the Team database toverify similarity for the period 1990-1993.

We found many discrepancies between theTeam and FS Region 3 databases, the greatestbeing fewer territories with complete reproduc-tive information in the Team’s database. The FSdatabase contained 1,984 records, whereas theTeam’s database contained only 377 observationswhere valid estimates of reproductive output(number of fledged young) were available. Avalid estimate of reproductive output was one forwhich appropriate protocols (see Forsman 1983,USDA Forest Service 1990) substantiated theestimate or for which young were reported.Thus, we were able to locate only a fraction ofthe data presumably available.

When the two databases were merged, wefound 11 records in the Team database that hadno corresponding record in the FS database.Further, the FS database has no fecundityestimate for 45 records for which the Team’sdatabase had a valid fecundity estimate. Of the321 records that matched, 271 records hadidentical values for reproductive output. Eight of

significant biases in the estimates of density,abundance, and occupancy rate because includ-ing an MT in the sample is directly linked tothese parameters.

In 1990, a study of Mexican spotted owlpopulation dynamics was initiated in the SkyIslands of southeastern Arizona (Duncan et al.1993). This study complemented inventory andmonitoring efforts on the Coronado NationalForest. In 1991, two other studies on populationdynamics were initiated: one on the CoconinoNational Forest in northern Arizona (Gutiérrezet al. 1993, 1994) and the other on the GilaNational Forest in west-central New Mexico(Gutiérrez et al. 1993, 1994). These two studieswere conducted independently from inventoryand monitoring efforts, and thus used methodsdifferent than the formal monitoring protocol.In all three population studies, owls were lo-cated, captured, individually marked for futurerecognition, and the number of fledged youngwere recorded. These population studies offerthe most reliable information currently availablefor estimating abundance,,,,, rates of reproductionand survival, and for deriving short-term esti-mates of population change. However, inferencesfrom these studies are somewhat limited becausetheir study areas were not randomly selectedfrom a defined sampling frame. Even with thislimitation, estimates from these studies are usefulin our preliminary evaluation of Mexican spot-ted owl population biology.

Most inventory and monitoring work wasconducted by personnel of the FS Region 3. Bydirection (Forest Service Manual 2676.2), surveydata were recorded on standard forms and mapsfollowing field observations. MTs were assignedusing the survey results. Occupancy and repro-ductive status of Mexican spotted owls weresummarized by MT at the close of each fiscalyear (30 September). These summaries did notinclude all information recorded on field forms,such as that used to determine occupancy,reproductive status, or spatial coordinates of owllocations. Original data recorded on field formswere necessary for verifying occupancy andreproductive status. Without this type of verifi-cation, reliability of previous assignments cannotbe demonstrated. Further, spatial coordinatescompatible with a Geographic Information

2

Volume II/Chapter 2


the remaining records had greater values in theTeam database, whereas 42 had greater values inthe FS database. For the 321 matching records,the FS database had a significantly greaterfledgling rate (P < 0.001) than the Team data-base. The expected bias was that the FS databasewould have a lower estimate of reproductiveoutput because many of the records with zeroyoung reported were not properly substantiatedby mousing data as was done in the Teamdatabase. However, changes by the FS occurredduring summarization procedures, among themselected territories with zero young recordedwere changed to unconfirmed (Keith Fletcher,FS, Albuquerque, NM, pers. comm.). Therefore,the FS database has a higher estimate of repro-ductive output because a number of territorieswith valid values of zero were eliminated fromthe calculation. This would introduce an overes-timate of reproductive output. Whether the twobiases in the FS database cancel each other isunknown.

A potential problem with the Team data-base was the incomplete availability of dataforms; many of the needed forms were missing.The last form used to complete the estimate ofreproduction for many of the MTs may not havebeen available for data entry. As a result, theTeam database has a lower estimate of reproduc-tion than the FS database. The Team’s effort tovalidate the FS estimates of fecundity does notaccomplish this objective. Rather, the Teamdatabase appears too incomplete to be useful forestimating fecundity. Therefore, we have usedestimates of fecundity from the FS database withthe understanding that these estimates may bebiased for two reasons: inadequate validation ofthe fecundity and occupancy results, and non-random sampling. The benefit of the Teamobtaining raw data from the FS is that this efforthas demonstrated severe problems with thehandling of data collection, data management,and data analysis for the monitoring program.By discovering these problems, the problems arecorrectable in the future.

LIFE HISTORY PARAMETERSLIFE HISTORY PARAMETERSLIFE HISTORY PARAMETERSLIFE HISTORY PARAMETERSLIFE HISTORY PARAMETERS

An organism’s life history is the combina-tion of birth and death processes exhibited in itsnatural environment (Partridge and Sibley1991). The optimal trade-off in survival andreproduction, resulting in fitness, should bemaximized by the life history favored by naturalselection. Under optimality theory, theorganism’s life history is a finely tuned result ofadaptations to its environment; major perturba-tions to an organism’s environment couldeventually lead to its extinction. However,behavioral plasticity may allow organisms toadjust life-history strategies to current environ-mental conditions (Hansen and Urban 1992).When examining recovery of a species after pastand present environmental perturbations ofvarying magnitudes, one must question whetherthat species will perish or persist. Therefore,examining the life history characteristics of theMexican spotted owl is paramount to under-standing how it responds to changes in itsenvironment.

An organism’s life history can be quantita-tively described in terms of age- and sex-specificsurvival, age- and sex-specific fecundity rates,longevity, and age at first reproduction (Stearns1992). In the following section, we outlinecurrent estimates of life history parameters forthe Mexican spotted owl, using a variety ofestimators calculated with different sources ofdata. These parameters can be used in two waysto make inferences about Mexican spotted owlpopulations: (1) to estimate the finite rate ofpopulation change (�), and (2) to examinecomponents of fitness within populations (Roff1992).

Age- and Sex-Specific SurvivalAge- and Sex-Specific SurvivalAge- and Sex-Specific SurvivalAge- and Sex-Specific SurvivalAge- and Sex-Specific Survival

We estimated age- and sex-specific survivalfor Mexican spotted owls using mark-recaptureestimators with data from banded owls inpopulation studies, binomial survival estimatorswith data from radio-tagged owls studied atvarious locations, and survival estimators fromdata collected during the FS monitoring pro-gram. Where feasible, we recognized four age

3

Volume II/Chapter 2


classes: juveniles (J), first-year subadults (S1),second-year subadults (S2) and adults (A), all asdescribed by Forsman (1981) and Moen et al.(1991). Juveniles (age, x = 0 years) were fledgedyoung-of-the-year. First-year subadults (x = 1year), second-year subadults (x = 2 years), andadults (x � 3 years) have similar body plumagebut were distinguished by different retrix charac-teristics (Moen et al. 1991). Not all studiesdifferentiated the two subadult age classes, sothese two age classes were usually lumped into asingle subadult (S) age class for the purpose ofestimating survival. We distinguished betweensubadults and adults because subadults usuallyhave lower reproductive rates than adults.

Mark-recapture Survival Estimators fromMark-recapture Survival Estimators fromMark-recapture Survival Estimators fromMark-recapture Survival Estimators fromMark-recapture Survival Estimators fromPopulation StudiesPopulation StudiesPopulation StudiesPopulation StudiesPopulation Studies

Mark-recapture estimators yield maximumlikelihood estimates of apparent survival (�) andrecapture (or resighting) probability (p), whichare asymptotically unbiased, normallydistributed, and have minimum variance(Lebreton et al. 1992). An importantconsideration with apparent survival is that1- � = death + permanent emigration. Forapparent survival to accurately reflect truesurvival (S), permanent emigration over thecourse of the study must be close to zero. Emi-gration is difficult to quantify, requiring the useof large samples of radio-marked birds.

Another limitation of the mark-recapturedata is that only territorial birds are marked,because nonterritorial birds (“floaters”) do notrespond to the capture and sighting methodsemployed. Even though marked juveniles enterthe floater population, they are not recaptureduntil they become territorial, which may nothappen because they emigrate from the studyarea. Floaters that never become territorial arenever included in the data to estimate survival,even if they remain on the study area. As a result,juvenile and subadult survival estimates pro-duced from birds banded as juveniles are biasedlow because of the entry into the floater popula-tion and/or emigration from the study area.However, estimates of survival generated fromsubadults and adults marked as territorial birds,hence with negligible emigration, are unbiased if

inference is only to territorial birds. Thus,inferences from the mark-recapture data onlyapply to the territorial population because onlybirds from the territorial population are marked.

We examined mark-recapture data fromcolor-banded Mexican spotted owls derived fromthe three population study areas (Figure 2.1):(1) a 484-km2 (187-mi2) area located on theCoconino National Forest (Upper Gila Moun-tains RU) denoted as the CSA (Gutiérrez et al.1994); (2) a 323-km2 (125-mi2) area located onthe Gila National Forest (Upper Gila MountainsRU) denoted as the GSA (Gutiérrez et al. 1993);and (3) an area encompassing portions of theSky Island Mountains in southeastern Arizona(Basin and Range - West RU) denoted as theSISA (Duncan et al. 1993). A total of 148 adult,44 subadult, and 238 juvenile capture historiescompiled during 3- and 4-year periods wereutilized in the subsequent analyses (Table 2.1).

Methods used to estimate survival from themark-recapture data are detailed in Burnham etal. (1994) and Lebreton et al. (1992). Two sets ofparameters are estimated with these models: � isthe probability of a bird remaining alive and onthe study area, and p is the probability that thebird will be resighted after initial capture. Both �and p are indexed to provide specific estimateswith respect to time, age, sex, and area. Weemployed three analyses: (1) goodness-of-fittesting to the Cormack-Jolly-Seber (CJS) modelsusing computer program RELEASE (Burnhamet al. 1987) and JOLLY (Pollock et al. 1990); (2)examination of a wide variety of models pro-gressing from a biologically realistic global modelto the simplest model using program SURGE(Lebreton et al. 1992); and (3) selection of themost parsimonious model with Akaike’s Infor-mation Criteria (AIC) and likelihood ratio tests(LRT) (Lebreton et al. 1992).

Goodness-of-fit tests of data to the CJSmodel include Test 2, which tests primarily forlack of independence, and Test 3, which prima-rily tests for heterogeneity in survival and recap-ture probabilities (Burnham et al. 1987). Wecould only use Test 2 because of the shortduration of the studies. Test 2 of the GSA andSISA data did not indicate lack of fit of the data(GSA: �2 = 0.008, 1 df, P = 0.93; SISA:�2= 0.702, 1 df, P = 0.40). Test 2 was not

4

Volume II/Chapter 2


Figure 2.1.Figure 2.1.Figure 2.1.Figure 2.1.Figure 2.1. Location of Coconino (CSA), Gila (GSA), and Sky Islands (SISA) Mexican spotted owlpopulation study areas in Arizona and New Mexico.

FPO

Table 2.1.Table 2.1.Table 2.1.Table 2.1.Table 2.1. Time periods and number of capture histories from three study areas used in estimatingMexican spotted owl survival.

1990-19931991-1993Age Class Age Class Age Class Age Class Age Class When BWhen BWhen BWhen BWhen Bandedandedandedandedanded

41 52 55

18 15 11

95 84 59

GGGGGila Sila Sila Sila Sila Study Artudy Artudy Artudy Artudy Areaeaeaeaea(GSA)(GSA)(GSA)(GSA)(GSA)

Coconino SCoconino SCoconino SCoconino SCoconino Study Artudy Artudy Artudy Artudy Areaeaeaeaea(CSA)(CSA)(CSA)(CSA)(CSA)

SSSSSky Iky Iky Iky Iky Island Ssland Ssland Ssland Ssland Study Artudy Artudy Artudy Artudy Areaeaeaeaea(SISA)(SISA)(SISA)(SISA)(SISA)

AAAAAdultsdultsdultsdultsdults

SSSSSubadultsubadultsubadultsubadultsubadults

JJJJJuvuvuvuvuvenilesenilesenilesenileseniles

1991-1993

5

Volume II/Chapter 2


Other than indicating that adult females mayhave higher survival than males, the � for adultfemales on the CSA was probably an artifact oflimited number of capture occasions. Such aresult is possible from a study with only threecapture occasions. Therefore, we used the nearestmodel to the selected model� i.e., the modelwith the next lowest AIC. This model {�J*g, �A,S,pS1, pA,S2} (AIC = 229.53,K = 5) included separate estimates of juvenilesurvival for the two study areas and a singleestimate for subadults and adults, with no sexeffects (Table 2.2). Recapture probabilities weremodeled the same as the previous model.

Estimated differences in juvenile survivalbetween the two study areas may representdifferences in survey effort around the studyareas and differences in study area size ratherthan real differences in the rates. For moreexplanatory details, refer to the PopulationTrends section.

We could not obtain reliable estimates fromthe SISA. For all the preliminary models exam-ined, survival estimates became bounded at theupper limit of 1 indicating problems with theestimation procedure. The selected model (�, p)had a single survival estimate (� = 1.0000,se (�) = 0.0000) and a single estimate of recap-ture probability (p = 0.2366, se (p) = 0.0344)with no age, sex, or time effects on either esti-mate. Such a conclusion is not biologicallyrealistic; and given that the protocol used in thisstudy did not stress resighting of marked birds(as evidenced by the low estimate of p), we donot feel these results are useful.

se (se (se (se (se (�)))))

computable for the CSA because only one birdreleased in 1991 was not recaptured in 1992 butwas captured in 1993.

For model selection, the GSA and CSAdata sets were modeled together because theyemployed similar designs and methodologies(Gutiérrez et al. 1993, 1994), both occur in theUpper Gila Mountains RU, and we wanted toimprove the precision of survival estimates bycombining the data into one analysis. Theprocedures used allowed testing the assumptionthat survival and recapture rates were the samefor these two study areas. The SISA data set wasmodeled separately from the CSA and GSA datasets because it was conducted by a separate set ofinvestigators using somewhat different protocols,and because this study occurred in the Basin andRange - West RU. In modeling the GSA andCSA data, we considered the global model to be{�g*s*a, pg*s*a} (AIC = 249.05, K = 28 parameters);it included study area (or group, g), sex (s), andage (a) effects. We did not include time effectsbecause only two recapture periods were presentin the data. The selected model{�J*g,�S,A(GSA),AM(CSA), �AF(CSA), pS1, pS2,A} (AIC =223.24, K = 6) included (1) separate estimates ofjuvenile (J) survival for each study area, (2) asingle estimate for subadults (S) and adults (A)of both sexes on the GSA combined with sub-adults of both sexes and male adults (AM) on theCSA, and (3) a single estimate for adult females(AF) on the CSA. No significant study areaeffects in terms of recapture probabilities (p)were found but there was an age effect (S1 � S2and A age classes). However, the estimate of �for CSA females was 1.000 ( se �� = 0.000),which was unsuitable for modeling purposes.

Table 2.2.Table 2.2.Table 2.2.Table 2.2.Table 2.2. Apparent survival (�), recapture probability (p), and their sampling standard errorsestimated for Mexican spotted owls between 1991 and 1993 on the Gila Study Area (GSA), NewMexico, and the Coconino Study Area (CSA), Arizona.

Age-class (SAge-class (SAge-class (SAge-class (SAge-class (Study artudy artudy artudy artudy area)ea)ea)ea)ea)

JJJJJuvuvuvuvuvenile (GSA)enile (GSA)enile (GSA)enile (GSA)enile (GSA)

JJJJJuvuvuvuvuvenile (CSA)enile (CSA)enile (CSA)enile (CSA)enile (CSA)

SSSSSubadult & Aubadult & Aubadult & Aubadult & Aubadult & Adult (GSA & CSA)dult (GSA & CSA)dult (GSA & CSA)dult (GSA & CSA)dult (GSA & CSA)

0.1524

0.1524

0.0176

se (se (se (se (se (ppppp)))))ppppp

0.7147a

0.7147a

0.9818

0.0366

0.0785

0.0269

0.2861

0.8889

�

0.0643

aRecapture probabilities for juveniles represent probability of recapture as an S1 or S2 individual.

^̂̂̂̂^̂̂̂̂ ^̂̂̂̂ ^̂̂̂̂

^ ^

^

^

^ ^

^ ^^

^̂̂̂̂ ^̂̂̂̂

6

Volume II/Chapter 2


^

Adult and subadult age classes were pooledfor analyses because of the small sample ofsubadults in the radio-marked samples and forcomparability with the mark-recapture estimates.Sex-specific estimates of adult and subadultsurvival were computed by pooling data acrossstudy areas. Pooling across study areas wasnecessary because of low sample sizes withineach study area. In addition, data on both sexesfrom the Coconino National Forest were com-bined for comparison with the mark-recaptureestimates from the CSA population study onthat forest. The pooling of data across studyareas and years prevented us from examiningspatial and temporal variation in the Kaplan-Meier estimates of survival. In all Kaplan-Meiermodels, the annual sampling period started on 1June because most birds were radio-marked justafter this date, this coincided with approximatelythe middle of the sampling period for thepopulation studies, and this was the “birth” dateused for fledglings in population modeling(Franklin 1992). In all models, individualstracked beyond the annual sampling period wererecycled back to the beginning. For example, anindividual that was radio-marked in July andthen had radio failure in August of the followingyear would be initially added in July, added againin the following June, and then censored inAugust. In this way, the number of entries in themodel exceeded the actual number of individualstagged and the only fates for an individual wereto die or be censored from the analysis.

Estimates of S for adults and subadultscombined were close to identical for both sexeswhen study areas were pooled (Table 2.4). Onthe Coconino National Forest, Kaplan-Meier (S)and mark-recapture (�) estimates were comparedfor adult/subadult (both sexes combined) andjuvenile age classes using a Wald test (Carrolland Ruppert 1988, Hosmer and Lemeshow1989):

��

Estimates from mark-recapture data werenot significantly different from those estimatedfrom the radio-telemetry data (�2 = 0.950, 1 df,P = 0.330). Estimates for juvenile survival from

^

Binomial Survival Estimators fromBinomial Survival Estimators fromBinomial Survival Estimators fromBinomial Survival Estimators fromBinomial Survival Estimators fromRadio-tracking StudiesRadio-tracking StudiesRadio-tracking StudiesRadio-tracking StudiesRadio-tracking Studies

Binomial estimators allow estimation oftrue survival (S) where 1 - S = mortality. Esti-mates of S can be derived from radio-telemetrydata and are preferable to estimates of apparentsurvival when permanent emigration occurs,radios do not influence survival, and sufficientnumbers of radios are available for preciseestimates. We estimated annual survival forsubadult and adult Mexican spotted owls basedon 73 radio-marked individuals from six studiesat four geographic locations (Table 2.3). We alsoused 22 juveniles from two independent studies,one in Colorado (R. Reynolds, unpub. data) andone in Utah (Willey 1992a, b). We selected theKaplan-Meier product limit estimator (Kaplanand Meier 1958) as modified by Pollock et al.(1989) to analyze the radio-telemetry data. Thisestimator allowed for staggered entry of indi-viduals during the sampling period and the useof right-censored data (exact fates of individualsunknown due to radio failure) without incurringthe biases due to censoring discussed by Whiteand Garrott (1990). In addition, this estimator isnonparametric and, therefore, does not requirean underlying hazard function that must bemathematically tractable (Pollock et al. 1989).

Five important assumptions underlie theKaplan-Meier estimator: (1) individuals havebeen randomly sampled; (2) survival of themarked animal is independent of other individu-als; (3) attached radios do not influence survival;(4) censoring is random and unrelated to anindividual’s fate; and (5) newly marked individu-als have the same survival rate as previouslymarked individuals. We were unable to test theseassumptions because of low sample sizes anddesign of the studies. However, we felt thatassumption (1) was probably not met because ofthe nature of the studies whereas assumptions(2)-(5) probably were met. However, controversyexists whether radios and their attachment(assumption 3) affect survival (Paton et al. 1991,Foster et al. 1992), although the studies reportedhere used tail-mounted radios rather than thebackpacks discussed by the references. Thisassumption cannot be tested with just radio-tracking data.

(S - ��

var (S) + var (��^^

^

^

7

Table 2.3.Table 2.3.Table 2.3.Table 2.3.Table 2.3. Description of radio-telemetry studies conducted on adult and subadult Mexican spotted owls in the Upper Gila (Norther AZ, CoconinoNF, AZ), Basin and Range - East (Lincoln NF, NM), Southern Rocky Mountains Colorado (Rocky Mts, CO), and Colorado Plateau (Zion NP, UT)RUs.

FPO

8

of inventory the preceding year. No differencesin the persistence rate across years were detectedfor the informal monitoring data (1989-1993;�2 = 2.543, 4 df, P = 0.637), or the formalmonitoring data (1989-1993; �2 = 3.233, 4 df,P = 0.520). We found no significant differencesfor year (�2 = 4.2546, 4 df, P = 0.373), monitor-ing type (�2 = 0.7133, 1 df, P = 0.398), or theirinteraction (�2 = 1.3162, 4 df, P = 0.859) usinga logistic regression model (Hosmer andLemeshow 1989).

The overall persistence rate based on pairsversus no pairs from 1989-1993 formal andinformal monitoring types (0.805, SE = 0.0138)was used to estimate adult survival (Sa). If pairsare assumed to remain on the same territory, lackof a pair on a territory that was previouslyoccupied infers the death of one or both mem-bers of the pair. This is a difficult assumption toaccept because pairs have been observed tochange nest locations, but it allows some infer-ence about adult survival. The dichotomy ofpairs versus no pairs was used because thisassumption is the simplest one possible forestimating survival from persistence. Single birds(i.e., persistence of a single) could be included inthe analysis to estimate survival, but wouldrequire an assumption that single males persistedon the territory at the same rate as single fe-males. Another bias is failure to detect a pairwhen both birds are present. Using the di-chotomy of pairs versus no pairs, and assuming

the radio-telemetry data (Table 2.4) were notsignificantly different from estimates from mark-recapture data collected on the CSA (�2 = 0.752,1 df, P = 0.386) or the GSA (�2 = 2.749, 1 df,P = 0.097).

Survival Estimators from FS Survival Estimators from FS Survival Estimators from FS Survival Estimators from FS Survival Estimators from FS Region 3Region 3Region 3Region 3Region 3Monitoring StudiesMonitoring StudiesMonitoring StudiesMonitoring StudiesMonitoring Studies

The FS has inventoried Mexican spottedowl MTs since 1984, and conducted informaland formal monitoring since 1989 (Table 2.5).Summaries of monitoring data were supplied bythe FS (see Sources of Information section). Foreach MT, presence of owls (single male, singlefemale, single unknown, unknown age and sex,or pair, or else unchecked) and result of repro-duction (0, 1, 2, 3 young, unchecked, or uncon-firmed) were noted.

Persistence of a pair can be estimated fromoccupancy data gathered during the FS spottedowl monitoring program. Pair persistence rate isdefined here as the probability that a territorycontaining a pair of owls in one year will containa pair in the succeeding year (but see discussionof biases below). An individual territory must bemonitored both years to compute this statistic.The overall persistence rate was 80.5% for 824territories monitored with formal and informalprotocols (Table 2.6). Note that some territorieshad persistence estimates for their first yearbecause some of these MTs were surveyed as part

Table 2.4. Table 2.4. Table 2.4. Table 2.4. Table 2.4. Estimates of true survival (S) for Mexican spotted owls based on radio-telemetrydata.

FPO

Volume II/Chapter 2


9

Volume II/Chapter 2


Table 2.5. Table 2.5. Table 2.5. Table 2.5. Table 2.5. Number of management territories checked one or more times during the Mexicanspotted owl monitoring program of FS Region 3.

Table 2.6.Table 2.6.Table 2.6.Table 2.6.Table 2.6. Persistence of a Mexican spotted owl pair on a territory for formal and informal moni-toring data, with “persistence” defined as the probability a pair of owls will exist on a territory giventhat a pair was on the same territory the previous year. Data are from the monitoring program of FSRegion 3.

FPO

FPO

10

Volume II/Chapter 2


4 df, P = 0.624), or the interaction between thetwo (�2 = 13.1995, 15 df, P = 0.587). Therefore,a reduced model with only recovery unit wasanalyzed; and again, no differences were foundfor recovery unit (�2 = 7.0271, 4 df, P = 0.134).

Survival Estimates from Band Return DataSurvival Estimates from Band Return DataSurvival Estimates from Band Return DataSurvival Estimates from Band Return DataSurvival Estimates from Band Return Data

Records from the FWS Banding Laboratorywere checked to determine if adequate bandingdata were available for estimation of survival. Allrecords obtained from the Laboratory were partof studies reported above, so did not provideadditional information.

Age- and Sex-specific FecundityAge- and Sex-specific FecundityAge- and Sex-specific FecundityAge- and Sex-specific FecundityAge- and Sex-specific Fecundity

Fecundity (mx) can be defined as the meanannual number of live births of a given sex by aparent of that same sex over an interval of age(Caughley 1977), e.g., the number of femaleyoung per adult female. For Mexican spottedowls, we defined live births as the number ofyoung fledging from the nest because the num-ber of live births at hatching was not measured.To estimate fecundity, we initially used thenumber of total young (e.g. of both sexes)fledged per pair as the response variable from the

that both adults survive at the same rate andindependent of one another, then (Sa)

2 = 0.805,or Sa = 0.897 (se (Sa) = 0.0077). This estimate isvery close to the observed estimate of apparentsurvival from the population studies (0.889).

Two biases of opposite direction are pos-sible for an estimate of adult survival based onpair persistence. A pair of owls may leave aterritory for reasons other than the death of onemember of the pair, resulting in an estimate ofsurvival biased low. In contrast, one member of apair may die, but the remaining member may beable to obtain another mate and stay on theterritory. Another possibility is that both mem-bers of a pair die, but the territory is occupied bya new pair. In these situations, the estimate ofsurvival is biased high. For this reason, we notethe similarity of persistence-based survival (Sa) tomark-recapture estimates (�) but we relied solelyon the mark-recapture estimates to estimatepopulation trends.

The 1989-1993 formal and informalpersistence data were also analyzed by recoveryunit (Table 2.7) using a logistic regression modelthat included year (1989-93), recovery unit andthe interaction between the two factors. Nodifferences were found for year (�2 = 5.9688,4 df, P = 0.202), recovery unit (�2 = 2.6129,

Table 2.7. Table 2.7. Table 2.7. Table 2.7. Table 2.7. Persistence of a pair of Mexican spotted owls on a territory for the five recovery unitscontained in the FS Region 3 monitoring database for 1989-1993, with “persistence” defined as theprobability a pair of owls will exist on a territory given that a pair was on the same territory the previ-ous year. Survival rate is the probability that both members of the pair survived the year, and is thesquare root of persistence.

FPO

^

^

^

^ ^ ^

11

two appropriate population studies (GSA andCSA) and the FS monitoring database. Thethird population study (SISA) was included inthe analysis of the FS monitoring databasebecause the methods used and territories moni-tored were identical. We analyzed the twopopulation studies separately from the monitor-ing database to obtain fecundity estimatesdirectly related to the mark-recapture estimatesof survival. Estimates for number of youngfledged were converted into fecundities bydividing the means in half and adjusting thestandard errors accordingly. This procedureassumed a 1:1 sex ratio at fledging and that siteswhere reproductive data were gathered wereindependent across years.

Fecundity Estimates from Population StudiesFecundity Estimates from Population StudiesFecundity Estimates from Population StudiesFecundity Estimates from Population StudiesFecundity Estimates from Population Studies

We used the general linear models proce-dure in SAS (PROC GLM; SAS Institute Inc.1985) to examine differences in time, age, andstudy areas for number of young fledged by maleand female Mexican spotted owls. Averagenumber of young fledged for all years differedbetween the GSA and CSA (F = 3.73; 1, 151 df;P = 0.055) but average number of young peryear for combined study areas did not differamong years of study (F = 1.58; 2, 151 df;P = 0.210). Based on a priori linear contrasts,number of young fledged by first-year subadultswas different from second-year subadults andadults combined for both males (F = 3.45; 1,151 df; P = 0.065) and females (F = 16.24; 1,151 df; P < 0.001; Table 2.8).

Fecundity Estimates from FS Region 3Fecundity Estimates from FS Region 3Fecundity Estimates from FS Region 3Fecundity Estimates from FS Region 3Fecundity Estimates from FS Region 3Monitoring DatabaseMonitoring DatabaseMonitoring DatabaseMonitoring DatabaseMonitoring Database

An objective of the FS Region 3 monitor-ing program was to monitor reproduction (Table2.9). Differences in mean reproduction acrossmonitoring methods and years were tested withanalysis of variance (ANOVA). Reproduction is acategorical variable, because only values of 0, 1,2, and 3 young are observed. However, ANOVAtechniques are still appropriate because of thelarge sample sizes involved (even though samplesizes are unequal); and hence, the cell meansbeing approximately normally distributed as a

result. Another possible procedure might be alog-linear analysis. However, the null hypothesisof a log-linear model would be that the distribu-tions are the same, compared to the null hypoth-esis of ANOVA that the means are the same.Thus, a log-linear analysis may reject the nullhypothesis even when the mean fecundity ratesdo not differ. We did not use a log-linear analysishere because we are primarily interested in themean rates.

All data for 1989-1993 were used to test fordifferences between formal and informal moni-toring types (Table 2.9). The year effect wassignificant (F = 14.09; 4, 685 df; P < 0.001), butmonitoring method (F < 0.001; 1, 685 df;P = 0.956) and the interaction (F = 1.23; 4,685 df; P = 0.295) were not significant. Theaverage number of young fledged/pair acrossyears was 1.006 (n = 695, SE = 0.037), giving afecundity estimate of 0.503 (SE = 0.018).

Differences in mean reproduction amongrecovery units with years included in the modelwas tested for formal and informal monitoringdata for 1989-1993. The year effect was signifi-cant (F = 3.03; 4, 672 df; P = 0.017), as wererecovery unit (F = 12.55; 4, 672 df; P < 0.001),and their interaction (F = 2.86; 14, 672 df;P < 0.001). Mean reproductive rates are given inTable 2.10 for each year and recovery unit, plusthe recovery unit mean across the years 1989-1993. Reproduction estimates from the FSmonitoring database were compared to estimatesfrom the population studies (CSA and GSA) forthe same two National Forests, i.e., Coconinoand Gila. The ANOVA model included year(1991-1993), National Forest (Coconino, Gila),and monitoring method (population studyversus formal and informal monitoring), plus allthe interactions of these three factors. None ofthe factors was significant (P � 0.300) except theinteraction between forest and monitoringmethod (F = 5.24; 1, 267 df; P = 0.023). Table2.11 presents the four means that produced thisinteraction.

Although the reasons behind this interac-tion are unknown, we speculate that two expla-nations are possible for this difference. First,differences in following the monitoring proto-cols between personnel in the Coconino andGila National Forests may have led to the

Volume II/Chapter 2


12

Table 2.8.Table 2.8.Table 2.8.Table 2.8.Table 2.8. Sample size (n) and estimates (mean and SE) of number of young fledged/pair and fecundity (female young fledged/female) from the GSAa

and CSAa Study Areas, 1991-1993.

FPO13

Volume II/Chapter 2


observed differences between the populationstudies and other monitoring systems. Monitor-ing approaches in the population studies are thesame because each was conducted by the sameresearch group using standardized methods andexperienced personnel. A second possibility isthat the population study areas are not represen-tative of the surrounding habitat for their respec-tive forests; and thus, the observed difference isreal and caused by habitat differences on eachforest. Alternatively, monitoring sites may not berepresentative of the forest because managementterritories were selected because of proposedtimber sales (with better quality owl habitat) orhistorical locations (with easy access).

Effects of Forest Type on LifeEffects of Forest Type on LifeEffects of Forest Type on LifeEffects of Forest Type on LifeEffects of Forest Type on LifeHistory TraitsHistory TraitsHistory TraitsHistory TraitsHistory Traits

We examined the effects of forest type onboth reproduction and persistence, which can beviewed as an indirect measure of survival. Weused two sources of information to examine theeffects of forest type on reproduction: the FSRegion 3 monitoring database and data fromSkaggs and Raitt (1988). For persistence, weused data from the FS Southwestern Regionmonitoring database. These were the onlysources of data available to the Team whichcoupled habitat information with data used toestimate life history traits.

Table 2.9.Table 2.9.Table 2.9.Table 2.9.Table 2.9. Number of Mexican spotted owl pairs checked (n), mean, and standard error (SE) fornumber of young fledged per pair for 1989-1993, based on formal and informal monitoring methods.

FPO

Effects on ReproductionEffects on ReproductionEffects on ReproductionEffects on ReproductionEffects on Reproduction

The FS Region 3 monitoring databaseincluded for each MT the percent (recorded tothe nearest 25%) of the core area consisting ofthe following forest types: mixed conifer, pine-oak, ponderosa pine, pinyon-juniper, oak,Arizona cypress, sycamore, other riparian, andunsuitable for owls. We computed the Pearsoncorrelations (r) between each of these forest typevariables and number of young produced on anMT for each year, given that a pair of owls waspresent. Significant correlations were found formixed-conifer (r = -0.131, P < 0.001), pine-oakr = 0.084, P = 0.011), other riparian (r = 0.062,P = 0.064), and unsuitable (r = 0.098,P = 0.003), with none of the remaining variablessignificant (P > 0.154). This analysis suggestedthat the more mixed-conifer present, the lowerthe reproductive rate (a negative correlation),and the more unsuitable forest type, the greaterthe reproductive rate (a positive correlation).When the forest type variables were used in astep-wise regression to predict number of youngfledged, mixed-conifer and unsuitable were bothselected (P � 0.017), while none of the remain-ing variables was included (P > 0.150). Theregression explained only a very small amount(2.3%) of the variation in the number of youngfledged, with the signs of both variables in theopposite direction of what we expected based onradio-telemetry studies (Ganey and Dick 1995).

14

Volume II/Chapter 2


Table 2.10.Table 2.10.Table 2.10.Table 2.10.Table 2.10. Number of Mexican spotted owl pairs checked (n), mean, and standard error (SE) fornumber of young produced per pair for 1989-1993 by recovery unit, based on formal and informalmonitoring methods.

FPO

Table 2.11. Table 2.11. Table 2.11. Table 2.11. Table 2.11. Sample size (n), mean, and standard error (SE) for the number of Mexican spotted owlyoung produced per territory for 1991-1993, based on population studies (CSA and GSAa) and formaland informal monitoring methods in the Coconino and Gila National Forests.

FPO

15

Volume II/Chapter 2


When we examined the database to determinethe distribution of territories with mixed-conifer,we found that 276 territories with 100% mixed-conifer cover occurred in the Basin and Range -East RU. These 276 territories were 79.3% ofthe territories that had 100% mixed-conifer.However, only 6 territories with <100% mixed-conifer occurred in the Basin and Range - East,or 1.1% of <100% mixed-conifer. Because ofthis inequitable distribution of mixed-coniferterritories across recovery units, the comparisonof mixed-conifer becomes a comparison ofrecovery units, not forest type, with nearly totalconfounding of recovery unit and mixed-coniferhabitat. As shown in Table 2.10, the Basin andRange - East RU has lower fecundity than all theother recovery units. When the recovery unit isincluded in the step-wise regression, the Basinand Range - East unit is included in the regres-sion, but the mixed-conifer variable is excluded.Possibly, mixed-conifer may be the underlyingcause of the differences between recovery units.However, so many other factors, such as climate,are confounded with this variable that such aninference is highly questionable.

A similar problem occurs with the unsuit-able forest type variable. Only 9 of the 907territories contained unsuitable forest type andall 9 occur in the Tonto NF in the Upper GilaMountains RU. Why these 9 territories hadbetter than average fecundity is not clear, but theunbalanced distribution of the unsuitable foresttype across territories implies that the conclusionthat the unsuitable forest type increases fecun-dity is inappropriate.

Because comparisons of forest type withinterritories are confounded with recovery unit,the FS Southwestern Region monitoring data-base is unsuitable to make such comparisons. Totest for the effect of forest type on fecundity, asample that is not confounded must be taken(such as the Skaggs and Raitt data discussednext), where the differences between territories isjust the forest type, and these differences are notconfounded by differences in climate, etc. Otherproblems are inherent in this approach. Thespatial scale at the territory level may not be thescale at which owls perform habitat selection.Overall forest configuration may not be asimportant as specific habitat characteristics

within the territory, such as a nest tree or roostsites needed by the nesting pair. Managementterritory boundaries are assigned subjectively bybiologists, and not based on empirical datadelimiting areas actually used by owls. Hence,the percentages of the various habitats includedin the territory reflects the biologist’s perception,not what owls may actually be using. Further,unless habitat is manipulated, cause and effectcannot be inferred from the kind of analysisperformed here, even though it is tempting todo so.

We reanalyzed data presented in Skaggs andRaitt (1988), who surveyed 18 quadrats, each 23km2 (9 mi2) in size. Six quadrats were placedrandomly into three areas, each of which wasdominated by either pinyon-juniper (PJ),ponderosa pine (PP), or mixed-conifer (MC)forest types. We used a t-test to test whether thenumber of young fledged per pair was differentbetween PP and MC forest types using quadratsas replicates. We did not include the PJ foresttype in the analysis because no pairs were foundin the quadrats dominated by this type. Wefound no significant difference in the number ofyoung fledged per pair (t = 1.58; 8 df; P = 0.152)although numbers of owls differed by forest type(Ward et al. 1995), suggesting that demographicprocesses may differ between the forest types.The estimate for MC forest types (x = 0.88fledged young per pair, SE = 0.281) was greaterthan the PP type (x = 0.25, SE = 0.25). Lack ofsignificance may have been due to low powerresulting from small sample sizes, thus we feelthe study should be repeated with larger num-bers of quadrats.

Effects on PersistenceEffects on PersistenceEffects on PersistenceEffects on PersistenceEffects on Persistence

Logistic regression models predictingpersistence from the forest type variables werealso constructed. Only mixed-conifer (� = -0.0047, P = 0.076), ponderosa pine(� = 0.0072, P = 0.090), and oak (� = 0.0346,P = 0.003) appeared important in predictingpersistence, with the remainder of the habitatvariables not significant (P > 0.167). As withreproduction, the model for mixed-conifersuggests that increasing amounts of this foresttype in the defined core area of MTs results in

^

^ ^

16

Volume II/Chapter 2


SurvivalSurvivalSurvivalSurvivalSurvival

Annual survival rates of adult Mexicanspotted owls is ~0.8-0.9 based on short-termpopulation and radio-tracking studies andlonger-term monitoring studies. These annualsurvival estimates can be viewed as the probabil-ity of an individual surviving from one year tothe next or as the proportion of individuals thatwill survive from one year to the next. A varietyof different estimators of adult survival usingdifferent types and sets of data gave similarresults. Juvenile survival is considerably lower(~0.06-0.29) than adult survival. Juvenile sur-vival also appears more variable spatially, al-though this conclusion reflects only two popula-tion study areas and two radio-telemetry studiesspanning two years or less.

We strongly suspect that estimates ofjuvenile survival from the population studies arebiased low because of (1) a high likelihood ofpermanent dispersal (emigration) from the studyarea, especially the smaller GSA, and (2) a lag ofseveral years before marked juveniles reappear asterritory holders, at which point they are firstdetected for recapture. Concerning the firstpoint, juvenile northern spotted owls have ahigh dispersal capability (reviewed in Thomas etal. 1990). If Mexican spotted owl juveniles havea similar dispersal capability, we expect that asubstantial portion of marked juveniles willemigrate from the respective study areas. How-ever, estimates from the radio-telemetry studyroughly corroborated the low estimates from thepopulation studies. Biases in the radio-telemetryestimates of juvenile survival can result if radiossignificantly affect their survival. Whether radiosor their attachment affect survival of northernspotted owls is debatable (Paton et al. 1991,Foster et al. 1992). Concerning the secondpoint, Franklin (1992) found a lag of 1-4 yearsbetween the time when juvenile northernspotted owls were banded and subsequentlyrecaptured. If this process is similar for Mexicanspotted owls, then the current population studiesmay be of insufficient duration to adequatelyestimate juvenile survival.

In summary, our survival estimates arebased primarily on studies of insufficient dura-tion or studies not explicitly designed to estimate

decreasing persistence, and ultimately fitness. Asdiscussed previously for the fecundity analysis,the inequitable distribution of mixed coniferhabitat across recovery units makes this analysisinappropriate.


Environmental conditions may greatlyaffect reproduction and/or survival of nestlingsthrough fledging and to adulthood. However,adult survival rates appear to be relatively con-stant across years, as suggested by high pairpersistence rates. Such life history characteristicsare common for K-selected species, for whichpopulations remain relatively stable even thoughrecruitment rates might be highly variable. Withno recruitment, the population only declines atthe rate of 1 minus adult survival, or the adultmortality rate.

Undoubtedly, long-lived organisms experi-ence a range of environmental conditionsthrough time. Conditions that result in simulta-neously low survival and reproduction can loweraverage population persistence rates dramatically,when examined over a period that is shortrelative to the organism’s life span. The converse,simultaneously high survival and reproduction,that would raise the expected population persis-tence dramatically, is also possible. However, themagnitude of the effect of such a sudden de-crease or increase on average persistence willnaturally decline as the observation of a givenpopulation is extended in time, while the prob-ability of detecting such events will increase withobservation time. In addition, dispersal amongsubpopulations can greatly influence the persis-tence of relatively isolated populations (Keitt etal. 1995). Successful dispersal may also be a rareand time-dependent event or a density-depen-dent event. We currently have little informationon the frequency of immigration and its associ-ated influence on population persistence. Thus,reliable conclusions on the persistence of Mexi-can spotted owl populations must await addi-tional study. Without reliable projections on theowl’s persistence, we can only summarize conclu-sions about the owl’s survival and reproductionbased on short-term but current knowledge.

17

Volume II/Chapter 2


owl reproduction and forest type are still lacking,even at the coarse-grained scale that we exam-ined. Further, forest type affects more than justreproduction; so additional data are needed toevaluate how forest type affects survival, andultimately fitness.

POPULATION TRENDSPOPULATION TRENDSPOPULATION TRENDSPOPULATION TRENDSPOPULATION TRENDS

We estimated trends in Mexican spottedowl populations two ways: as the finite rate ofpopulation change, lambda (�), and as trends inthe rate of occupancy of spotted owl territories.

Finite Rate of Population ChangeFinite Rate of Population ChangeFinite Rate of Population ChangeFinite Rate of Population ChangeFinite Rate of Population Change

Lambda was computed for female Mexicanspotted owls from the age-specific survival andfecundity rates obtained from two populationstudy areas, the GSA and CSA (see Life HistoryParameters section). Lambda is a useful metricbecause it measures both the direction andmagnitude of change in population trends.Direction of population trends can be character-ized as stationary (� = 1), declining (� < 1) orincreasing (� > 1). The magnitude in change isexpressed as the annual rate of change, R, whereR = � - 1 for a birth-pulse population. From apopulation management perspective, the statisti-cal hypothesis is � < 1 versus the null hypothesisthat the population is either stationary or in-creasing (�� 1). This is a one-sided test of theform:

where Z is normally distributed with = 0,2 = 1.

We used a Leslie matrix (Leslie 1945) asmodified by Usher (1972) to compute estimatesof � based solely on the estimates of age-specificfecundity and survival probabilities obtainedfrom the two study areas. The form of the matrixfollowed Usher (1972):

survival. In most cases, the data were too limitedto support or test the assumptions of the estima-tors used. However, the age- and sex-specificestimates of survival calculated here are useful atthis point as qualitative descriptors of the life-history characteristics of Mexican spotted owls.That is, Mexican spotted owls exhibit high adultand relatively low juvenile survival. In thisrespect, Mexican spotted owl survival probabili-ties appear similar to northern (see review inBurnham et al. 1994) and California spottedowls (Noon and McKelvey 1992).

ReproductionReproductionReproductionReproductionReproduction

Reproductive output of Mexican spottedowls, defined as the number of young fledgedper pair, varies both spatially and temporally.Mexican spotted owls may have a higher averagereproductive rate (1.001 fledged young per pair)than the California (~0.712; Noon andMcKelvey 1992) and the northern spotted owl(~0.715; Thomas et al. 1990). Both of the othersubspecies exhibit temporal fluctuations inreproduction similar to the Mexican subspecies.

Effects of Forest TypeEffects of Forest TypeEffects of Forest TypeEffects of Forest TypeEffects of Forest Type

We feel the data collected by Skaggs andRaitt (1988) were more appropriate for examin-ing the effects of forest type on reproductionthan our analysis of the monitoring data. Theirstudy was designed to look at the relationshipbetween forest type, density and reproductionwhereas the monitoring program was not.However, the Skaggs and Raitt data lackedsufficient statistical power to detect differencesin reproductive output between the three foresttypes. This problem arose primarily becausesignificantly fewer spotted owls were found inthe pinyon-juniper and pine forest types than inmixed-conifer forests (Ward et al. 1995). Theprimary problem in using data on forest typesfrom MTs is that the data are confounded formaking the desired comparisons. Further, thecore delineations were subjective. In addition,forest types included within MT boundaries maynot have been used by the owls for which thegiven MT was established. For these reasons, wefeel that definitive data linking Mexican spotted

Z = ,

,

1 - �^

^

se (�)

�0m1 �1m2

��0 �1

18

Volume II/Chapter 2


where �0 = juvenile survival, �1 = survival for thesubadult and adult age classes combined,m1 = S1 fecundity, and m2 = fecundity for S2and adult age classes combined. The form of thematrix assumed a birth-pulse population with apost-breeding census and a projection interval ofone year (Noon and Sauer 1992). Lambda wasestimated as the dominant eigenvalue associatedwith the right eigenvector derived from thematrix using power analysis (Caswell 1989). These(�) was estimated using the delta method(Seber 1982, Alvarez-Buylla and Slatkin 1994),which included the sampling covariances for theestimated survival probabilities.

Parameter estimates used to compute �(Table 2.12) were taken from only the twopopulation studies because all of the requiredparameters were estimated from data within thesame spatial and temporal scales. Survival esti-mates were based on the mark-recapture estima-tors. Estimates of � were computed separatelyfor the two studies because of significant differ-ences in fecundity and juvenile survival betweenthe two areas.

We obtained estimates of �� which weresignificantly lower than 1 for the GSA but werenot significantly different from 1 for the CSA(Table 2.13). These estimates of �� representtrends in populations for only the places andtimes of study, and are based on only 3 years ofdata collection. We believe the estimate of ��from GSA is <1 because of bias in the estimateof �J due to emigration from the study area.GSA is about 1/2 the size of CSA, so that thebias of juvenile emigration is greater for thissmaller study area. The bias of �J from juvenileemigration from the study area is a function ofstudy area size. E. Forsman (FS, Pacific North-west Forest and Range Experiment Station,Olympia, WA, pers. comm.) found juvenileemigration rates for northern spotted owls of~60% using radio tracking.

We also estimated the parameter valuesnecessary to obtain a value for � of 1, or astationary population, given that the otherparameter estimates were the same and wereunbiased and precise (Table 2.14). These resultssuggest that estimate of juvenile survival for theGSA must increase substantially (i.e., remove thebias in the estimate from emigration from the

study area) for the population to be stationary.However, the estimate of juvenile survival for theCSA was close to that expected for a stationarypopulation on the GSA. The difference injuvenile survival between the two areas may beattributed to 2 causes. First, survey effort maydiffer between the 2 study areas. FS personnelconduct more Mexican spotted owl surveys inthe area surrounding the CSA than around theGSA and hence more color-banded juveniles arereported (M. Seamans, Humboldt State Univ.,Arcata, CA, pers. comm.). Second, as discussedabove, the CSA is almost twice as large as theGSA, so that juveniles may be less likely todisperse off the CSA.

The large value for S1 fecundity (given that� = 1) for the GSA results from � being insensi-tive to that parameter, so that a biologicallyunreasonable value is needed to obtain � = 1.That is, the change needed in S1 fecundity tomake � = 1 must be so large that the value isbiologically impossible.

Occupancy Rate as Measure ofOccupancy Rate as Measure ofOccupancy Rate as Measure ofOccupancy Rate as Measure ofOccupancy Rate as Measure ofPopulation ChangePopulation ChangePopulation ChangePopulation ChangePopulation Change

From the monitoring database summarizedby the FS, we evaluated the occupancy rate ofterritories, expressed as the percent of territoriesoccupied by owls, as a measure of trends in theMexican spotted owl population. For the formaland informal monitoring data (Table 2.15), theoverall occupancy rate of territories was 63.8%pairs, 15.0% not detected (absent), and 21.2%with a single bird, or presence of a bird or birds.The presence of just a single bird or no detec-tions may result from inexperienced crewsperforming the surveys, and hence a biasedestimate of the actual number of pairs. A logisticregression model of pair versus no pair accordingto year (1989-1993), monitoring type (informaland formal), and the interaction of these twovariables suggests a significant difference inoccupancy rate for monitoring type(�2 = 3.1840, 1 df, P = 0.074), but no differ-ences in the other terms (P > 0.288). Presum-ably, the best quality territories were incorpo-rated into the formal monitoring system with apair occupancy rate of 67.6% versus 61.0% for

19

Volume II/Chapter 2


Table 2.12.Table 2.12.Table 2.12.Table 2.12.Table 2.12. Parameters used to estimate ��for Mexican spotted owls on the GSAa and CSAa studyareas.

FPO

FPO

Table 2.13. Table 2.13. Table 2.13. Table 2.13. Table 2.13. Estimates and standard errors of ��for female Mexican spotted owls on the CSAa andGSAa study areas. The Z statistic and probability level are for a one-sided test of the null hypothesis of��versus the alternative hypothesis ��1.

20

Volume II/Chapter 2


informal monitoring; and hence, formal moni-tored territories were more likely to be occupied.However, the formal monitoring system looksmore intensively at the sites for owls, so thisdifference in occupancy rate may be because ofsearch procedures, which we suggest as the mostlikely scenario.

We examined differences in occupancy rate(Table 2.15) with a logistic regression model thatincluded recovery unit, monitoring type (infor-mal and formal), and the interaction of thesetwo variables for the years 1989-1993. Type ofmonitoring was not significant (�2 = 0.0191,1 df, P = 0.890), but recovery unit(�2 = 22.7979, 4 df, P < 0.001) and the interac-tion of recovery unit and type of monitoring(�2 = 16.1657, 4 df, P = 0.003) were significant.

As defined and used here, occupancy rate isa rather artificial parameter because the selectionof territories for inclusion in the monitoringdatabase depends on judgement of humanobservers. MT boundaries are set subjectively, sothey do not necessarily represent owl homeranges. Further, MTs may encompass more thanone pair (May et al. in press), although thispossibility is not supposed to occur. Changes inoccupancy rate probably correspond more withthe addition of new MTs to the list of thosealready monitored by the FS, level of effort used

to monitor the MTs (i.e., number of MTsmonitored that meet formal monitoring proto-cols), and other administrative factors ratherthan true change in the owl population. As acomplicating factor, MTs are not a randomsample of all existing Mexican spotted owlterritories. As can be inferred from Figure 2.2,the percent of MTs occupied has dropped since1989 because all new MTs added to the moni-toring system are initially occupied. Habitatchanges induced by forest alterations over thenext several decades will make some of thecurrently occupied MTs unsuitable. Thus, weexpect that occupancy rate will decline forexisting territories. New MTs will probably beadded to compensate for loss of existing territo-ries, so change in occupancy rate does notprovide a valid inference about changes in theowl population.


We have little confidence in our estimatesof population trends for the following reasons.First, accurate and precise estimates of �� dependon the accuracy and precision of the parameterestimates used in the calculations. Estimates ofjuvenile survival may be biased low for reasonsstated previously and the time over whichparameters have been estimated is insufficient.

Table 2.14.Table 2.14.Table 2.14.Table 2.14.Table 2.14. Estimates of population parameters for Mexican spotted owls on the GSAa and CSAa

study areas, and the value of the parameter that gives ��= 1, provided that all other parameters remainat their original estimate.

FPO

21

Volume II/Chapter 2


changes can be caused by change in survival ratesor change in reproduction, or both. The currentmonitoring program only examines reproduc-tion. Occupancy rate and pair persistence arelogically flawed approximations to survival.Thus, with the current monitoring system,drastic changes in the owl population couldoccur and not be detected; or if detected, thecurrent monitoring system would not yield thestatistical rigor necessary to substantiate theconclusion strongly enough to withstand thecriticism of opponents to the suggested finding.Therefore, an improved monitoring system mustbe developed for future work (USDI 1995).

Table 2.15. Table 2.15. Table 2.15. Table 2.15. Table 2.15. Number of Mexican spotted owl territories checked (n) and the percent of them occu-pied by a pair of owls for 1989-1993, based on formal and informal monitoring methods.

FPO

Second, the population studies from whichparameter estimates were derived have not beenconducted for a sufficiently long period tocapture temporal variation. If one considers ��asan average rate of change over the study period,then three years is insufficient to adequatelyestimate �� from both biological and statisticalperspectives. Third, rates of change estimatedthrough occupancy provide little information onhow Mexican spotted owl populations arechanging for the reasons stated previously.However, the analysis does illustrate why rates ofpopulation change using occupancy are inad-equate for estimating trends in Mexican spottedowl populations. If nothing else, we believe theanalytical procedures and framework that wehave used throughout this section should pro-vide a template for future research as well asindicate priorities for future research efforts.

Theoretical Problems with CurrentTheoretical Problems with CurrentTheoretical Problems with CurrentTheoretical Problems with CurrentTheoretical Problems with CurrentMonitoring ProceduresMonitoring ProceduresMonitoring ProceduresMonitoring ProceduresMonitoring Procedures

Much effort has been expended by person-nel of FS Region 3 in collecting the monitoringdata summarized here. Unfortunately, themonitoring effort was inadequate for detectingimportant changes in the population dynamicsof the Mexican spotted owl because the appro-priate parameters were not measured.

The primary goal of the monitoring pro-gram should be the detection of significantchanges in population levels of the owl. Such

22

Volume II/Chapter 2


Figure 2.2.Figure 2.2.Figure 2.2.Figure 2.2.Figure 2.2. Changes in occupancy of formal and informal monitoring territories in the FS Region 3monitoring database.

FPOFPOFPOFPOFPO

23

Volume II/Chapter 2



Alvarez-Buylla, E. R., and M. Slatkin. 1994.Finding confidence limits on populationgrowth rates: three real examples revised.Ecology 75:255-260.

Burnham, K. P., D. R. Anderson, and G. C.White. 1994. Estimation of vital rates of thenorthern spotted owl. Colo. Coop. Fish andWildl. Res. Unit, Colorado State Univ., Fort

Collins. 44pp.——., ——., ——., C. Brownie, and K. H.

Pollock. 1987. Design and analysis methodsfor fish survival experiments based on release-recapture. Am. Fisheries Soc. Monogr. 5:1-437.

Carroll, R. J., and D. Ruppert. 1988. Transfor-mation and weighting in regression. Chapmanand Hall, New York, N.Y. 249pp.

Caswell, H. 1989. Matrix population models.Sinauer Assoc., Sunderland, Mass. 328pp.

Caughley, G. 1977. Analysis of vertebratepopulations. John Wiley and Sons, New York,N.Y. 234pp.

Duncan, R. B., S. M. Speich, and J. D. Taiz.1993. Banding and blood sampling study ofMexican spotted owls in southeastern Arizona.Unpubl. report, USDA For. Serv., CoronadoNat. For., Tucson, Ariz. 21pp.

Forsman, E. D. 1981. Molt of the spotted owl.Auk 98:735-742.

——. 1983. Methods and materials for locatingand studying spotted owls. USDA For. Serv.Gen. Tech. Rep. PNW-162. 8pp.

Foster, C. C., E. D. Forsman, E. C. Meslow, G.S. Miller, J. A. Reid, F. F. Wagner, A. B. Carey,and J. B. Lint. 1992. Survival and reproduc-tion of radio-marked adult spotted owls. J.Wildl. Manage. 56:91-95.

Franklin, A. B. 1992. Population regulation innorthern spotted owls: theoretical implica-tions for management. Pages 815-827 in D.R. McCullough and R. H. Barrett, eds.Wildlife 2001: populations. Elsevier AppliedScience, New York, N.Y.

Ganey, J. L. 1988. Distribution and habitatecology of Mexican Spotted Owls in Arizona.M.S. Thesis, Northern Arizona Univ., Flag-staff. 229 pp.

——., and J.L. Dick, Jr. 1995. Habitat relation-ships of the Mexican spotted owl. Pages xxx -xxxx in USDI. Recovery plan for the Mexicanspotted owl. Vol. 2. USDI Fish and Wildl.Serv., Albuquerque, N.M.

Gutiérrez, R. J., D. R. Olson, and M. E.Seamans. 1993. Demography of two Mexicanspotted owl Strix occidentalis lucida popula-tions. Humboldt State University. Arcata,California. 16pp.

——., ——., and ——. 1994. Demography oftwo Mexican spotted owl populations inArizona and New Mexico. Rep. submitted toHumboldt State Univ. Foundation, Arcata,Calif. 29pp.

Hansen, A. J., and D. L. Urban. 1992. Avianresponse to landscape pattern: the role ofspecies’ life histories. Landscape Ecol. 7:163-180.

Hosmer, D. W., and S. Lemeshow. 1989. Ap-plied logistic regression. John Wiley and Sons,New York, N.Y. 307pp.

Kaplan, E. L., and P. Meier. 1958. Nonparamet-ric estimation from incomplete observations.J. Am. Statist. Assoc. 53:457-481.

Keitt, T., A.B. Franklin, and D.L. Urban. 1995.Landscape analysis and metapopulationstructure. Chapter 3 (16pp.) in USDI. Recov-ery plan for the Mexican spotted owl. Vol. II.USDI Fish and Wildl. Serv., Albuquerque,N.M.

Kroel, K. and P. J. Zwank. 1991. Home rangeand habitat use characteristics of the Mexicanspotted owl in the southern SacramentoMountains, New Mexico. Unpubl. rep.,USDA For. Serv., Lincoln National Forest.New Mexico Coop. Fish and Wildl. Unit.

Lebreton, J. D., K. P. Burnham, J. Clobert, andD. R. Anderson. 1992. Modeling survival andtesting biological hypotheses using markedanimals: a unified approach with case studies.Ecol. Monogr. 62:67-118.

Leslie, P. H. 1945. On the use of matrices incertain population mathematics. Biometrika35:183-212.

May, C.A., M.Z. Perry, R.J. Gutiérrez, M.E.Seamans, and D.R. Olson. In press. Feasibilityof a random quadrat study design to estimatechanges in density of Mexican spotted owls.U.S. For. Serv. Res. Pap.

24

Volume II/Chapter 2


Moen, C. A., A. B. Franklin, and R. J. Gutiérrez.1991. Age determination of subadult north-ern spotted owls in northwest California.Wildl. Soc. Bull. 19:489-493.

Noon, B. R., and K. S. McKelvey. 1992. Stabil-ity properties of the spotted owlmetapopulation in southern California. Pages187-206 in J. Verner, K. S. McKelvey, B. R.Noon, R. J. Gutiérrez, G. I. Gould, Jr., and T.W. Beck, eds. The California spotted owl: atechnical assessment of its current status.USDA For. Serv. Gen. Tech. Rep. PSW-133.

——., and J. R. Sauer. 1992. Population modelsfor passerine birds: structure, parameteriza-tion, and analysis. Pages 441-464 in D. R.McCullough and R. H. Barrett, eds. Wildlife2001: Populations. Elsevier Applied Science,New York, N.Y.

Partridge, L., and R. Sibly. 1991. Constraints inthe evolution of life histories. Philos. Trans. R.Soc. Lond., Ser. B 332:3-13.

Paton, P. W. C., C. J. Zabel, D. L. Neal, G. N.Steger, N. G. Tilghman, and B. R. Noon.1991. Effects of radio tags on spotted owls. J.Wildl. Manage. 55:617-622.

Pollock, K. H., J. D. Nichols, C. Brownie, and J.E. Hines. 1990. Statistical inference forcapture-recapture experiments. Wildl.Monogr. 107:1-97.

——., Winterstein, C. M. Bunck, and P. D.Curtis. 1989. Survival analysis in telemetrystudies: the staggered entry design. J. Wildl.Manage. 53:7-15.

Rinkevich, S.E., J.L. Ganey, W.H. Mow, F.P.Howe, F. Clemente, and J.F. Martinez-Montoya. 1995. Recovery units. Pages 36-51in USDI. Recovery plan for the Mexicanspotted owl. Vol. I. USDI Fish and Wildl.Serv., Albuquerque, N.M.

Roff, D. A. 1992. The evolution of life histories:theory and analysis. Chapman and Hall. NewYork, N.Y. 535pp.

SAS Institute, Inc. 1985. SAS language guide forpersonal computers, Version 6. SAS Institute,Cary, N.C. 429pp.

Seber, G. A. F. 1982. Estimation of animalabundance. Second ed. C. Griffin, London,U.K. 654pp.

Skaggs, R. W., and R. J. Raitt. 1988. A spottedowl inventory of the Lincoln National Forest,Sacramento Division, 1988. New MexicoDep. Game and Fish, Santa Fe. 12pp.

Stearns, S. C. 1992. The evolution of life histo-ries. Oxford Univ. Press, Oxford, U.K. 249pp.


USDA Forest Service. 1990. Managementguidelines and inventory and monitoringprotocols for the Mexican spotted owl in theSouthwestern Region. Federal Register55:27278-27287.

USDI. 1995. Mexican Spotted Owl RecoveryPlan. Vol. 1. USDI Fish and Wildl. Serv.,Albuquerque, N.M.

Usher, M. B. 1972. Developments in the Lesliematrix model. Pages 29-60 in J. N. R. Jeffers,ed. Mathematical models in ecology.Blackwell Scientific Publ., Oxford, U.K.

Ward, J.P., Jr., S.E. Rinkevich, A.B. Franklin,and F. Clemente. 1995. Distribution andabundance. Chapter 1 (14pp.) in USDI.Recovery plan for the Mexican spotted owl.Vol. II. USDI Fish and Wildl. Serv.,Albuquerque, N.M.

White, G. C., and R. A. Garrott. 1990. Analysisof wildlife radio-tracking data. AcademicPress, San Diego, Calif. 383pp.

Willey, D. W. 1992a. Semi-annual report: homerange characteristics of Mexican spotted owlsin the canyonlands geographic province;winter 1991-1992. Unpubl. rep., Utah Div.Wildl. Resour., Salt Lake City.

——. 1992b. Movements and habitat ecology ofMexican Spotted Owls in southern Utah.Final Rep. submitted to Utah Div. Wildl.Resour., Salt Lake City.

——. 1993. Home-range characteristics andjuvenile dispersal ecology of Mexican spottedOwls in southern Utah. Unpubl. rep., UtahDiv. Wildl. Resour., Salt Lake City.

25


Volume II/Chapter 3

CHAPTER 3: Landscape AnalysisCHAPTER 3: Landscape AnalysisCHAPTER 3: Landscape AnalysisCHAPTER 3: Landscape AnalysisCHAPTER 3: Landscape Analysisand Metapopulation Structureand Metapopulation Structureand Metapopulation Structureand Metapopulation Structureand Metapopulation StructureTim Keitt, Alan Franklin, and Dean Urban

Unlike the northern spotted owl (seeThomas et al. 1990), the range-wide populationof the Mexican spotted owl is naturally frag-mented into geographically distinct subpopula-tions. Thus far, we have discussed populations atthe local scale within Recovery Units whereindividuals within subpopulations interact tosome degree. Understanding population struc-ture at larger scales is important, even thoughdata are extremely limited at these scales for theMexican spotted owl. Hanski and Gilpin (1991)identified two additional scales beyond the localscale: (1) a metapopulation scale where individu-als infrequently move between subpopulationsand typically cross unsuitable habitat to reach anadjacent subpopulation, and (2) a geographicalscale where individuals have little likelihood ofmoving to most portions of the geographic rangeof their species. By definition, metapopulationsare systems of local populations connected bydispersing individuals (Hanski and Gilpin1991). The metapopulation concept has beenapplied to the conservation of northern(Schaeffer 1985) and California (Noon andMcKelvey 1992) spotted owl populations.However, on a geographical scale, the rangewidepopulation of Mexican spotted owls may becomposed of a several discrete metapopulationsrather than of a single integratedmetapopulation. Understanding to what degreeMexican spotted owl populations followmetapopulation dynamics is essential for manag-ing the subspecies on all three scales.

METAPOPULATION MODELSMETAPOPULATION MODELSMETAPOPULATION MODELSMETAPOPULATION MODELSMETAPOPULATION MODELS

Theoretical work has proposed severalmodels of metapopulations based on differentmechanisms for persistence. These have beenclassified as the Levins model, source-sink, core-satellite, patchy, and non-equilibriummetapopulations (Harrison 1991). An importantconsideration when reviewing these models isthat empirical evidence for the existence of the

proposed mechanisms in natural populations isdebatable (Doak and Mills 1994).

The Levins Metapopulation ModelThe Levins Metapopulation ModelThe Levins Metapopulation ModelThe Levins Metapopulation ModelThe Levins Metapopulation Model

Levins (1969) first introduced themetapopulation concept with a simple model.This model incorporated three essential require-ments for persistence of a subdivided popula-tion: (1) density-dependent dynamics of popula-tions, (2) asynchronous dynamics of localsubpopulations in that not all subpopulationshave the same rate of change at the same time,and (3) dispersal between subpopulations.Dispersal is an essential component of this, andall other, metapopulation models and providesthe mechanism for recolonizing areas where localsubpopulations have died out.

Source-sink and Core-satelliteSource-sink and Core-satelliteSource-sink and Core-satelliteSource-sink and Core-satelliteSource-sink and Core-satelliteMetapopulation ModelsMetapopulation ModelsMetapopulation ModelsMetapopulation ModelsMetapopulation Models

In source-sink models (Pulliam 1988),source areas with self-propagating (typicallyincreasing) populations provide a flow of recruitsto sink areas where populations are not self-reproducing (and may be declining). Withoutthe net flow of immigrants from the sourceareas, populations in sinks would not persist.Source-sink mechanisms can be applied tocontiguously distributed populations wherehabitat conditions can dictate whether a popula-tion segment is a source or a sink. This mecha-nism can also be applied to a metapopulation ofdiscrete subpopulations where habitat or otherconditions dictate whether such an area is asource or a sink.

The core-satellite model builds on thesource-sink model by having a central corepopulation which acts as a source surrounded bya number of smaller sink populations (Harrison1991). Persistence of the satellite populationsdepends upon the central source population.

1


Volume II/Chapter 3

Patchy Metapopulation ModelsPatchy Metapopulation ModelsPatchy Metapopulation ModelsPatchy Metapopulation ModelsPatchy Metapopulation Models

Patchy populations in the context of thismodel are characterized by a high dispersalpotential where dispersal takes place on a spatialscale greater than the local events causing sub-population dynamics (Harrison 1991). Thepatchy metapopulation model is distinguishedfrom the Levins model in that the averageindividual can belong to more than one sub-population in its lifetime (excluding its natalsubpopulation). The result, in effect, is a well-mixed version of the Levins model.

Non-equilibrium MetapopulationNon-equilibrium MetapopulationNon-equilibrium MetapopulationNon-equilibrium MetapopulationNon-equilibrium MetapopulationModelModelModelModelModel

Populations characterized by this modelessentially follow the same dynamics as theLevins model except that continuousrecolonization of extinction-prone subpopula-tions has been disrupted. Essentially, this modelrepresents metapopulations in decline eventhough one or more subpopulations may bestable. Instability of a formerly stablemetapopulation may result from factors thataffect one or more subpopulations or whencontiguous populations are fragmented intodiscrete subpopulations. These factors caninclude natural (e.g., fire, disease) or anthropo-genic (e.g., logging, urban development) distur-bances.

DISPERSALDISPERSALDISPERSALDISPERSALDISPERSAL

In all metapopulation models, dispersal is akey component. Dispersal acts as a bridgebetween subpopulations at the metapopulationscale to provide immigrants to otherwise isolatedhabitat patches. The fate of these recruits de-pends on the status of the subpopulation. If thehabitat patch has been unoccupied, then the newrecruits “rescue” it from extinction. If the patchis not saturated with territorial breeders, therecruits might bolster the breeding population.If the patch is saturated, the new recruits mightpersist as nonbreeding “floaters,” perhaps buffer-ing the population against future fluctuations.

Adult and subadult Mexican spotted owlsappear to be relatively sedentary once theycome to occupy a given site. Juvenile Mexicanspotted owls, however, almost always dispersefrom their natal sites (Willey 1993, Hodgsonand Stacey 1994, Reynolds and Johnson 1994).If metapopulation dynamics apply to Mexicanspotted owl populations, these dynamics prob-ably hinge on the flow of juvenile spotted owlsbetween subpopulations.

Dispersal of young Mexican spotted owls ispoorly understood. Several studies have at-tempted to examine juvenile dispersal throughradio-telemetry (Willey 1993, Hodgson andStacey 1994, Reynolds and Johnson 1994); butsample sizes have been small. Total distancesmoved by 7 juveniles radio-marked in Utah(Willey 1993) ranged from 32 to 98 km (20-61miles) (median = 41.8 km [26 miles]). Four ofthese juveniles moved back to within 8 km(5 miles) of their natal area just prior to thebreeding season. Hodgson and Stacey (1994)also reported 2 juveniles dispersing between theSan Mateo and Black Mountain Ranges of NewMexico, distances of 45 and 58 km (28-36miles). In addition, dispersing juveniles crossedlarge expanses of habitat typically consideredunsuitable for resident spotted owls. Reynoldsand Johnson (1994) radio-marked 6 juveniles,none of which were relocated beyond their natalsite. The radio-telemetry data suggest thatjuvenile owls have the dispersal capability to actas recolonizers between many of the subpopula-tions and to provide genetic links betweensubpopulations. Thus far, none of the radio-marked juveniles have been recruited into aresident, territorial subpopulation.

With source-sink and core-satellitemetapopulation mechanisms, juvenile owls fromsource areas must not only reach isolated sub-populations but do so in sufficient numbers tostabilize sink populations. We examined straight-line distances moved by 25 juveniles bandedfrom 1991 through 1993 on the populationstudy areas (CSA and GSA) which had beenrecaptured as territorial subadults and adults(R.J. Gutiérrez, D. Olsen, and M. Seamans,Humboldt State Univ., Arcata, CA, pers.comm.). From these data, we generated a prob-ability function by fitting an exponential curve

2


Volume II/Chapter 3

to distances moved and the cumulative probabil-ity that juveniles had moved at least thosedistance (Figure 3.1). This curve represents theprobability that a juvenile will disperse at least agiven distance and be recruited into the territo-rial population. This curve can also be viewed interms of proportions. For example, about 60%of the juveniles would be expected to disperse atleast 10 km (6.2 miles) and be recruited into thepopulation. We provide this information as anillustration of the type of data needed to assessthe role of dispersal in metapopulation dynam-ics. The data we used were not designed togenerate such a curve and may be artificiallytruncated because of lower search effort forbanded recruits in the larger areas surroundingeach of the study areas. Such an effect can beresponsible for the exponential nature of therelationship in Figure 3.1 and underestimate truedispersal distances.

APPLICATION OF LANDSCAPEAPPLICATION OF LANDSCAPEAPPLICATION OF LANDSCAPEAPPLICATION OF LANDSCAPEAPPLICATION OF LANDSCAPEECOLOGYECOLOGYECOLOGYECOLOGYECOLOGY

Landscape ecology is concerned with thedevelopment, implications, and dynamics oflandscape pattern. We use the term “pattern”generously, but here we are concerned especiallywith three components of pattern: (1) the sizedistribution of habitat patches; (2) the spatialorientation of these patches with respect to eachother, that is, their juxtaposition; and (3) theirmutual distance relationships as these mightinfluence spotted owl dispersal.

With respect to metapopulation dynamics,landscape analysis is concerned with distancerelationships among habitat patches as thesemight affect the relative isolation of certainpatches or regions of patches within the owl’srange where patches may correspond to distinctsubpopulations. The issue of so-called patch“connectedness” is central to current ideas aboutmetapopulation dynamics. By conventionaldefinition, two patches are functionally con-nected if individuals can disperse between themwith a probability or frequency above someminimum threshold value. We will furtherdefine connectedness to incorporate patch areaas well as between-patch distance relationships

(see below); thus, a patch may be highly con-nected if it is near several small patches or nearone large patch. Specifically, we have two con-cerns about patch connectedness:

1. Can we identify habitat patches thatbecause of their area and position withinthe landscape might play a particularlyimportant role in overall landscapeconnectedness? We expect large habitatpatches to be important in this analysis.

2. Are there habitat patches that might becritical to landscape connectednesschiefly because of their spatial position,that is, despite their smaller size? Thesemight function as dispersal conduits orsmall “stepping stones” within thelandscape.

The goal of this set of questions is to identifyhabitat patches that might influence patterns inowl distribution well beyond their immediatelocation. In the case of a small stepping-stonepatch, this importance might be despite itshaving rather modest owl populations. Thus,critical patches identified in the landscapeanalysis might warrant special consideration inthe Recovery Plan even though their localpopulations might not elicit any special concern.

Scale and Resolution in LandscapeScale and Resolution in LandscapeScale and Resolution in LandscapeScale and Resolution in LandscapeScale and Resolution in LandscapeAnalysisAnalysisAnalysisAnalysisAnalysis

Landscape analysis typically connotes ratherlarge spatial scales. In our analyses, we haveidentified two spatial scales of interest. At asmaller scale we are concerned with areas such asa single National Forest, a scale defined in largepart by administrative criteria. At a larger scalewe are concerned with the geographic range ofthe owl as we defined it in this plan. We willrefer to these scales as “forest” and “rangewide.”Importantly, habitat data available at these twoscales is of very different resolution. Data at theforest scale are of higher spatial resolution(smaller grain size) and often of higher informa-tion content as well, while data at the rangewidescale is coarser-resolution and makes fewer

3


Volume II/Chapter 3

FFFFFigurigurigurigurigure 3.1e 3.1e 3.1e 3.1e 3.1 Dispersal-distance relationship for radio-marked juvenile spotted owls.

FPO

distinctions about habitat details. For example,many forests have digital elevation models(DEMs) with a nominal resolution of 30 m (98ft) and habitat data corresponding to timber-management compartments resolved on theorder of tens of hectares. At this scale, habitatsmay be defined in terms of tree-species composi-tion, size-class distributions, or other details. Bycontrast, the data we have available forrangewide analyses have a spatial resolution of 1km (0.62 miles); habitats at this scale are definedas gross cover types (e.g., pinyon-juniper). Dataat these two scales lend themselves to the sameanalytic techniques but require rather differentinterpretation because of their different resolu-tion and information content.

Data AvailabilityData AvailabilityData AvailabilityData AvailabilityData Availability

In keeping with the foregoing discussion ofscale and resolution, we have attempted toacquire data at two scales and resolution forlandscape analyses. At a finer scale, we soughtUSGS DEMs with 30-m (98 ft) resolution and

habitat maps of similar spatial resolution andcomparatively high information content (e.g.,species composition, size-class distributions). Wewere largely unsuccessful in this effort and, thus,have been unable to address questions on spatialaspects of habitat use by owls at this scale.

At a coarser scale, we acquired two sets ofhabitat data. Both sets were derived fromAVHRR satellite imagery and have a spatialresolution of 1 km2 (0.39 miles2). One set is theEROS Land Cover classification (Loveland et al.1991), which recognizes 159 cover classes acrossthe conterminous United States. The seconddataset is derived from the EROS set but wasreclassified by the FS into a smaller number ofrecognized Forest Cover Types (Powell et al.1993, Zhu and Evans 1992, Evans and Zhu1993). This latter set also includes a companioncoverage of forest density (percent canopy cover)derived by resampling the AVHRR-based datawith higher-resolution Thematic Mapper (TM)imagery at 30-m (98 ft) resolution and correlat-ing the percent of cells forested at 30-m (98 ft)resolution to a greenness index (NDVI) at 1-km

4


Volume II/Chapter 3

resolution (Zhu 1994). The FS has also created apreliminary classification of Mexico’s forest covertypes, based on the same AVHRR imagery(Evans et al. 1992). No forest density coverageexists for Mexico at this time. Examination ofthese coverages revealed that the EROS classifi-cation included too much detail and somedetails that were rather suspect from a biogeo-graphic standpoint, both of which arguedagainst its use. The FS coverages span the entirerange of the owl within the United States and atan appropriate information content (number ofclasses). For these reasons, we have elected tobase our rangewide analyses on the FS coverages.At this scale, we also have 1-km2 (0.39 miles2)resolution DEMs as provided by EROS. Becausewe lack comparable data for Mexico, our analy-ses are restricted to the United States.

The lack of data at higher resolution (e.g.,30-m scale) has important implications in ouranalyses, in that it precludes analysis of thosefeatures too fine-grained to be resolved in thecoarse-resolution data we have been forced touse. For example, the 1-km scale data cannotresolve small canyons that are important to owlsin parts of Utah. Neither do we have data thatcan resolve details about owl microhabitat, noreven subtle distinctions among forest covertypes. As we note below, this lack of data doesnot render our analyses pointless, but it doesemphasize the need to repeat and corroboratethese analyses with finer-resolution data. On theother hand, our coarse-resolution approachallows us to analyze landscape pattern overvirtually all of the owl’s range within the UnitedStates.

Landscape ConnectednessLandscape ConnectednessLandscape ConnectednessLandscape ConnectednessLandscape Connectedness

Our approach derives from graph theory andpercolation theory (Gardner et al. 1992) andfocuses on the so-called “radius of gyration” ofthe largest subgraph in a graph representing alandscape of habitat patches. In this, a “graph” isa set (map) of habitat clusters, and each “cluster”is a collection of grid cells that are defined to befunctionally connected. In practice, we definethis connection based on a “minimum joiningdistance” such that cells that are within thisdistance are functionally connected. The radius

of gyration is defined as the mean Euclideandistance between each cell of a habitat clusterand the centroid of that cluster. Habitat clustersare formed by specifying a joining distance andthen identifying groups of cells that are spatiallydiscrete at that distance scale. As one increasesthis joining distance, the habitat map coalescesinto increasingly larger clusters as isolatedpatches are subsumed into nearby clusters. Aconvenient means of indexing this process is asthe radius of the largest cluster in the graph. Aweighted index for the map can be computed bysumming the radii of all clusters in the map,weighting each radius by the proportion of theentire map it comprises (i.e., its relative size).This weighted sum is the map’s “correlationlength” (or connectedness length) and has asunits the distance units of the original map. Oneinterprets correlation length in terms of how far,on average, an animal could traverse the mapwithout straying off “habitat” cells intononhabitat.

The ultimate goal here is not to computelandscape connectedness in itself, but rather toidentify those patches (clusters) that contributemost significantly to overall habitat connected-ness. One means to this end is to estimate thereduction in connectedness that would result if apatch were removed from the landscape. Here,this estimate is computed as the reduction incorrelation length of the landscape. Thesereductions are estimated by systematicallyremoving each cluster and recomputing theconnectedness index. The analysis itself proceedsin steps as follows:

1. Define a binary raster map of “habitat”versus “nonhabitat.”

2. Specify a distance threshold at whichpatches may join.

3. Perform the cluster-removal experiment,saving each cluster’s effect on the correla-tion length.

4. Normalize this effect for each cluster bydividing its effect by its total area, andsave this value as well.

5


Volume II/Chapter 3

The normalization in step (4) adjusts thepatch’s effect for its area, which identifies thosepatches whose effect on connectedness is largerelative to their size. The result of the analysis isa ranking of habitat patches in terms of thereduction in connectedness each elicits, that is,each patch’s contribution to overall connected-ness.

This procedure is then repeated for a rangeof threshold distances to estimate the conse-quences of varying assumptions about thedispersal capabilities of owls. We used distancesfrom 0 to 100 km (0-62 miles) in 5-km (3.1mile) intervals. This range spans our best empiri-cal estimate of the owl’s dispersal range, which ison the order of 50 km (31 miles). If the patchranks change considerably at different joiningdistances, this would suggest a need to improvethe accuracy of our estimates of owl dispersaldistances so that we could tailor the analysis tomaximize its biological significance.

Finally, the entire analysis is repeated for avariety of baseline habitat maps. This gives us anestimate of how robust the results are to assump-tions about what constitutes “potential owlhabitat.” At this spatial scale and with the dataavailable, we have limited opportunities to devisealternative definitions of “owl habitat.” We havedevised three alternative habitat maps:

1. A map wherein all cells assigned as“Douglas-fir” (i.e., mixed conifer) orponderosa pine (including some mixed-conifer as well as pine-oak) cover typesare defined to be potential owl habitat;

2. A map including all Douglas-fir andponderosa pine cells, plus any pinyon-juniper cells that have greater than 50%canopy cover as estimated in the FSForest Density coverage;

3. A map including all Douglas-fir andponderosa pine types, plus those pinyon-juniper cells that are within a 2-km (1.2miles) buffer of Douglas-fir or pinetypes, i.e., pinyon-juniper near betterowl habitat.

These alternative definitions vary in terms ofhow narrowly or generously “habitat” is defined.Other definitions could certainly be devised, andwe do not argue that ours are the only (nor eventhe best) possibilities. We will argue, however,that these are sufficient to indicate whether ourconclusions are robust to the definition ofhabitat or whether we need to invest more effortin generating more realistic base habitat maps.

Results of Landscape AnalysesResults of Landscape AnalysesResults of Landscape AnalysesResults of Landscape AnalysesResults of Landscape Analyses

Results of these analyses provide insight intotwo keys aspects of habitat connectedness: (1)the relationship between minimum joiningdistance and overall connectedness as indexed bycorrelation length; and (2) contributions ofindividual habitat patches to overall connected-ness. We present each of these in turn. Becausethe results were qualitatively similar for each ofthe baseline habitat maps, we present here onlythe results for the mixed-conifer/ponderosa pinebase map; but we return to this issue later in ourdiscussion.

Landscape Connectedness and MinimumLandscape Connectedness and MinimumLandscape Connectedness and MinimumLandscape Connectedness and MinimumLandscape Connectedness and MinimumJoining DistanceJoining DistanceJoining DistanceJoining DistanceJoining Distance

Correlation length exhibits a profoundlynonlinear relationship with minimum joiningdistance in all cases we examined (Figure 3.2).The inflection in this relationship illustrates thatthe landscape changes from being largely “un-connected” to largely “connected” over a narrowrange of distances. In the habitat maps weanalyzed, this transition occurred over distancesof 40-60 km (25-37 miles). This result variedonly slightly for alternative habitat maps, sug-gesting that this qualitative result is rather robustto these definitions.

The relationship between cluster size andjoining distance varies systematically with thespatial resolution with which habitats are de-fined. This can be seen most clearly by contrast-ing the curve for “all clusters” of ponderosa pineand Douglas-fir as compared to the curve for thelargest 254 of these clusters (Figure 3.2). Includ-ing more (smaller) clusters shifts the curve to theleft, effectively joining the landscape at closer

6


Volume II/Chapter 3

distances. Presumably, if we were to include verysmall clusters the landscape might be function-ally connected at extremely small joining dis-tances, which is something of a scaling artifact.This result does underscore the need to morefully characterize the habitat affinities anddispersal behavior of owls. If we could limitcluster sizes and joining (dispersal) distances tobiologically reasonable values, this analysiswould be more confidently focused.

All habitat maps coalesced into a very fewlarge clusters at joining distances of approxi-mately 60 km (37 miles) or more. This indicatesthat at these distances, the entire landscape isessentially connected (Figure 3.3). Yet even atjoining distances of 100 km (62 miles), a fewclusters remain spatially discrete, and by implica-tion, functionally isolated from the rest of thehabitat in the landscape.

Rank Patch Contributions to ConnectednessRank Patch Contributions to ConnectednessRank Patch Contributions to ConnectednessRank Patch Contributions to ConnectednessRank Patch Contributions to Connectedness

The influence of each patch on overallconnectedness was tallied for only the largest254 clusters in each map. In fact, the mapscontained thousands of clusters but many weresingle cells; so it proved to be computationallyimpractical to compute every case. Patch influ-ence on correlation length was strongly depen-dent on joining distance: the average influencewas greatest at intermediate distance scales andmuch less at shorter or longer scales (Figure 3.4).To highlight influential patches spatially, weproduced a habitat map in which each patch iscolor-coded according to its mean rank over alldistance classes (Figure 3.5a). A similar mapemphasizes patch importance to connectednessat the intermediate distance scale of 45 km (28miles) (Figure 3.5b). In both maps, “hot” colorsindicate highly ranked patches (those contribut-ing substantially to landscape connectedness)while “cool” colors indicate patches of low rank(those with little effect on connectedness).

In general, patch effects on connectednessdepended on patch area, in that large patcheshad the greatest influence on overall connected-ness. Patch ranks, normalized for area, areillustrated as averaged over all distances (fig.3.6a) and at 45-km (28-mile) joining distance(Figure 3.6b), using the same color scheme as in

Figure 3.5. These figures emphasize the impor-tance of a few clusters in joining the large habitatblock along the Mogollon Rim to the largecluster farther northeast.

DISCUSSIONDISCUSSIONDISCUSSIONDISCUSSIONDISCUSSION

Distance Relationships in LandscapeDistance Relationships in LandscapeDistance Relationships in LandscapeDistance Relationships in LandscapeDistance Relationships in LandscapeConnectednessConnectednessConnectednessConnectednessConnectedness

The nonlinear relationship in Figure 3.2makes intuitive sense if one envisions the processthat generates this relationship. At small joiningdistances, small and nearby patches are joinedinto somewhat larger clusters, but the landscapestill consists mostly of disjoint clusters. At aparticular distance, a large subset of the clustersjoins into a single large cluster. Once this hashappened, further small accretions to the clusterdo not change its total area appreciably. From afunctional standpoint, these later additions areredundant links to an “already connected”cluster. This same process explains why thegraph of average influence of patch removal(Figure 3.4) shows a peak at these intermediatedistance classes; for shorter or longer distancesthese patches can have little impact on overallconnectedness.

An important result of this analysis is thatthe strongly nonlinear domain of this relation-ship (i.e., the inflection point in Figure 3.2)coincides with our best current estimate of thedispersal capabilities of spotted owls based onfield studies, approximately 50 km (31 miles).By implication, if owls’ dispersal ranges weremuch less than this estimate (say, 20 km [12miles]) then much of their natural range wouldbe functionally unconnected (most patcheswould be isolated). Reciprocally, if owls coulddisperse much farther than our current estimate(say, 100 km [62 miles]), then most of thelandscape would be functionally connected.Likewise, the change in this relationship with theinclusion of more, smaller patches suggests thatowls’ use of very small patches as dispersalconduits or stepping stones could influence theseresults. If owls can use very small patches, thenthe landscape might be more connected than our

7


Volume II/Chapter 3

Figure 3.2.Figure 3.2.Figure 3.2.Figure 3.2.Figure 3.2. The relationship between correlation length and minimum joining distance. The 4 linesillustrate the efects of alternative definitions of “potential owl habitat.”

FPO

results suggest, presuming, of course, that suchsmall patches exist.

The radio-telemetry data suggest that juve-nile owls have the dispersal capability to providepopulation and genetic links between subpopu-lations. To date, none of the radio-markedjuveniles has been recruited into a resident,territorial subpopulation. In attempting to judgewhether owl dispersal is sufficient to maintainpopulation and genetic connectedness, we arefaced with the logistical problem that ecologi-cally significant dispersal events might occurquite infrequently (1-2 per generation) andwould likely go unrecorded by even the mostintensive monitoring program. Thus, our datacan suggest that such dispersal capabilities mightexist, but these data cannot prove that suchdispersal actually occurs. Importantly, neithercan our lack of data prove that such dispersaldoes not occur.

Clearly these results point to a need forbetter understanding of the dispersal behaviorand distance relationships for spotted owls. Onesource of uncertainty in our analyses is that the

clustering is based on boolean distance decisions(i.e., patches are connected if their distance isstrictly within the joining distance). But animaldispersal is probabilistic, exhibiting some sort ofdecreasing probability with increasing distance(recall Figure 3.1). This distinction exaggeratesour results relative to how dispersal probablyoccurs with real animals.

A second source of uncertainty stems fromour limited understanding of owl dispersalbehavior. Our analyses are based on assumptionsabout “reachability” with the tacit assumptionbeing that if habitat is reachable in terms ofabsolute distance, then owls can and will dispersethere. But there are many plausible reasons whythis might not be so (e.g., avoidance behavior,excessive mortality during dispersal); so itremains that we need to temper our interpreta-tions with additional considerations of owldispersal.

8


Volume II/Chapter 3

Figure 3.3.Figure 3.3.Figure 3.3.Figure 3.3.Figure 3.3. Landscape mosaics of discrete clusters of Douglas-fir and Ponderosa pine habitat types, as(a)(a)(a)(a)(a) largest 254 clusters, and at (b)(b)(b)(b)(b) 30-km, and (c)(c)(c)(c)(c) 60-km joining distances. Colors have no significancebeyond labelling discrete clusters.

FPO

9


Volume II/Chapter 3

Figure 3.4.Figure 3.4.Figure 3.4.Figure 3.4.Figure 3.4. Change in correlation length due to cluster removal as a function of distance, averagedover the largest 254 clusters.

FPO

Patch Contributions to LandscapePatch Contributions to LandscapePatch Contributions to LandscapePatch Contributions to LandscapePatch Contributions to LandscapeConnectednessConnectednessConnectednessConnectednessConnectedness

Rank scores for patch contributions tooverall landscape connectedness make intuitivesense when viewed as landscape mosaics (Figures3.5 and 3.6). The ranks uncorrected for areashow the expected result that large clusters thatare centrally located have the highest ranks,while small or isolated clusters are lower-ranked.Area-normalized ranks emphasize the positionalaspects of patch contributions and Figure 3.6illustrates a few patches (in red) that seem to actas bridges spanning larger habitat areas. Some ofthese bridges are rather small yet could beimportant links in landscape connectedness.

An important caveat to bear in mind is thatthese analyses are all based on habitat, not onowl densities. Thus, sparsely populated habitatclusters in the northern reaches of the owl’srange receive equal weight in the analysis ascompared to habitats currently supporting muchlarger owl populations, for example, along the

Mogollon Rim. Thus, the clusters in Coloradothat appear important to connectedness (redpatches in the upper-right regions of Figure 3.5)are actually connecting habitat that supportsessentially no owls. If habitat clusters wereweighted according to present-day owl abun-dance, these same analyses would provide quitedifferent results. Densely populated patcheswould increase in importance while moresparsely populated patches would be down-weighted. While this is rather easy to anticipatein general, such a weighted analysis wouldrequire much better spatial information on owlabundance across its range than the inconsistentcoverage currently available.

These considerations bear strongly on theultimate goal of a conservation plan. If our goalis to provide a template that might sustain owlmetapopulations well into the future, then ananalysis weighted on “potential habitat” seemsmost appropriate. Conversely, a strategy topreserve current populations would seem toargue for an analysis heavily weighted onpresent-day owl abundance. Our results suggest

10


Volume II/Chapter 3

FPO

Figure 3.5.Figure 3.5.Figure 3.5.Figure 3.5.Figure 3.5. Maps with clusters colored to illustrate their rank importance, weighted by (uncorrectedfor) patch area. Hot colors (reds) are highly ranked; cool colors (blue-violet), low ranked. Base habitatmap is ponderosa pine/Douglas fir. (a)(a)(a)(a)(a) Patch importance averaged of all distance classes. (b) (b) (b) (b) (b) Impor-tance at 45-km joining distance.

11


Volume II/Chapter 3

FPO

Figure 3.6. Figure 3.6. Figure 3.6. Figure 3.6. Figure 3.6. Patch importance to overall connectedness, normalized for patch area. Color scheme andbase map, and panels (a)(a)(a)(a)(a) and (b)(b)(b)(b)(b), are the same as in Figure 3.5.

12


Volume II/Chapter 3

that these two strategies might lead to quitedifferent recommendations.

Importantly, our rankings of patch impor-tance to overall connectedness are based on asimple patch-removal algorithm in which eachpatch is removed singly from the existing land-scape. Clearly, these results could be quitedifferent if the algorithm provided for morecomplex scenarios as conditional removals. Forexample, if patch i has already been removed,then the removal of patch j might take on muchgreater importance. Such conditional scenarioswould be more realistic in the sense that land-scape dynamics driven by land-use managementare time-structured (sequential), conditional, andtypically act on multiple patches during anysingle management episode. Such complicatedscenarios might be undertaken on a smaller scale(e.g., for a National Forest) if sufficient data wereavailable for the analysis. Complicated, condi-tional scenarios are probably not feasible forrangewide analyses.

Sensitivity to Habitat DefinitionSensitivity to Habitat DefinitionSensitivity to Habitat DefinitionSensitivity to Habitat DefinitionSensitivity to Habitat Definition

Two empirical biases emerge in consideringalternative definitions of what constitutes “po-tential owl habitat” in these analyses. One bias isdue to the spatial scale at which habitat patchesare resolved. As smaller patches are included,overall landscape connectedness tends to increaseso long as these small patches are liberallysprinkled across the landscape. Similarly, forhabitat definitions that are increasingly “gener-ous” toward owls, more potential owl habitatoccurs in the landscape and so connectednessalso tends to increase (consider a map thatincludes some pinyon-juniper relative to themixed-conifer/pine landscape). These biasesappear in our analyses as a result of data avail-ability. If higher-resolution data were available,more patches could possibly be delineated. Atthe same time, however, given higher-resolutiondata we could define owl habitat more strin-gently and thus some patches would be rede-fined as no longer usable by owls; owl habitatwould decrease in abundance and overall con-nectedness would decrease accordingly. Thus, thetwo empirical biases are somewhat compensat-

ing. But both biases point to a need to improveour ability to discriminate usable owl habitatfrom the surrounding matrix.

Other Uncertainties andOther Uncertainties andOther Uncertainties andOther Uncertainties andOther Uncertainties andConsiderationsConsiderationsConsiderationsConsiderationsConsiderations

A final consideration of our results mustaddress any uncertainties or biases that mightresult from the algorithm itself. One potentialbias that emerges can be seen in the figuresillustrating patch importance to connectedness.Because our approach indexes connectedness asthe mean size of the largest cluster, a biasemerges whereby patches that are peripheral inthe landscape can form clusters with a very largeradius. Thus, the important patches in Figure3.5 tend to form a ring around the landscape as awhole, partly because these patches are large butalso because their joining creates a cluster thathas a radius nearly as large as the entire mosaic.This bias would not occur if the index of con-nectedness counted total area in a way that didnot emphasize among-cell distances. For ex-ample, an index that estimated “total connectedarea” rather than the effective size of this areamight yield a slightly different estimate of patchimportance. Unfortunately, such indices haveproven to be computationally unfeasible thus far.We continue to explore alternative algorithmsfor indexing habitat connectedness.


Application of metapopulation theory to theMexican spotted owl is rather speculative, givenour limited understanding of within-populationdynamics much less between-population dynam-ics. If metapopulation models actually representrealistic abstractions of real-world processes, thena number of different metapopulation modelsmay apply to different geographic regions withinthe range of the Mexican spotted owl. Forexample, one could envision a core-satellitemodel applying to the southern portion of theMexican spotted owl range with the Upper GilaRU acting as a core source population with thesmaller surrounding mountain ranges in Basin

13


Volume II/Chapter 3

and Range - West and Colorado Plateau RUsacting as satellite sinks. On the other hand, theSky Island mountain ranges in southern Arizonacould also follow the classic metapopulationdynamics proposed by Levins (1969). A numberof different scenarios could be envisioned,which, unfortunately, we cannot corroborateeasily empirically. The distribution of geographi-cally isolated subpopulations, however, suggeststhat some form of interaction between thesesubpopulations is plausible. Clearly, intensive,long-term studies over large areas will be re-quired to understand the structure and dynamicsof Mexican spotted owl population at these largescales.

Although the landscape analyses are explor-atory, some general conclusions can still bedrawn from the results. These conclusionsconcern apparent connectedness of variousregions of the owl’s range and the relative impor-tance to particular habitat clusters to overalllandscape connectedness.

Regardless of the underlying habitat mapused in analyses, results consistently show a fewregions that appear functionally isolated atintermediate joining distances similar to thedispersal range of owls. For example, large blocksof southern Utah persist as discrete habitatclusters at joining distances of 40 km (25 miles);the Lincoln National Forest also appears isolatedat this spatial scale. We might test the hypothesisthat these subpopulations are discrete, possiblyby exploring genetic similarities between theseand more central (connected) populations suchas those along the Mogollon Rim. Likewise,basic population analyses of these populationsmight also indicate their degree of functionalconnectedness. For example, spatialdiscontinuities in population density, age struc-ture, or other parameters might suggest thatthese populations do not interact to the sameextent as other, more contiguous populations.

The relative importance of individual habitatpatches, when uncorrected for patch area,suggests the intuitive approach of protectingthose patches that currently support the highestowl densities, such as along the Mogollon Rim.But our results also indicate a high importancefor patches farther north in the owl’s range,

patches which currently do not support appre-ciable owl populations. Our reaction to thisresult depends in part on whether our goal is topreserve present-day populations or to protectthe capability for populations to expand in thefuture. The discrepancy between these twostrategies is most pronounced in the northernreaches of the owl’s current range.

Correcting cluster importance for areaindicates the contribution, per unit area, of eachof the habitat clusters. This correction suggeststhat while the cluster along the Mogollon Rim iscrucial to overall connectedness, the many standsmaking up this cluster are not so important onan individual basis. Thus, this cluster wouldlikely continue to play an important role in thelandscape so long as its internal continuity ismaintained, which would require reiterating ouranalyses on a finer spatial scale and resolution asmanagement of this region proceeds. Conversely,a few clusters emerge as being more importantper unit area than their uncorrected importancewould suggest, such as the stepping-stonesevident in Figure 3.6 as red patches. Suchpatches warrant special attention in land useplanning because of their potential to affectregional populations despite their small size andperhaps modest owl populations. Especially,these patches should be a focus of monitoringefforts so that their use by owls can be assessed.Note that these patches typically would not betargeted in field studies for the simple reasonthat they would not appear to support large owlpopulations.

Finally, we should emphasize that theseresults are exploratory and therefore subject tocorroboration. One form of verification couldcome via analyses of owl subpopulations acrossthe region. Metapopulation theory suggests thatisolated habitats should show higher year-to-yearvariability in population density than wouldbetter-connected patches, which would be bettersubsidized by dispersal. A second form of cor-roboration would entail analyses similar to oursbut using alternative indices of connectednessand alternative definitions of suitable owl habi-tat. Especially, we would like to see these analy-ses repeated with higher-resolution habitat datacapable of resolving the those features mostimportant to owls. In any case, it is clear that we

14


Volume II/Chapter 3

need to invest much more effort toward definingthe dispersal capabilities and behavior of thespotted owl across its range.

In closing, we would like to revisit therationale that underlies our somewhat theoreti-cal, habitat-based approach. While it might beappealing to invoke analyses that rely on owlpopulation data, we were forced to accept at thebeginning that we lacked adequate data for suchan analysis across the owl’s range. Similarly,detailed population models such as those devel-oped for the northern owl are appealing becauseof their biological richness; but we clearly lackthe data to parameterize such a model for theMexican spotted owl. We see both these popula-tion-based approaches as a useful adjunct to ourapproach, but one that we cannot supportempirically with data currently available. Instead,we developed an approach that makes very fewassumptions about owl biology, and we testedour results to determine whether violation ofthese assumptions would change our resultssubstantially. Our approach was to use as gener-ous a definition of owl habitat as possible givenour data; our major conclusions seem largelyunaffected by alternative definitions of habitat,although this needs to be verified with finer-resolation data. The other important assumptionwe made was that owls have a dispersal-distancerelationship such that dispersal probabilitydecreases with increasing distance, and that theiraverage dispersal is in the neighborhood of 50km or so (i.e., not << 10 km, and not >> 100km). Within this domain, our results are alsorobust. Thus, while there remain a number ofquestions still to be resolved concerning land-scape-scale patterns and owl metapopulations,the results we present here are a useful and validfirst approximation.


Doak, D. F., and L. S. Mills. 1994. A useful rolefor theory in conservation. Ecology75:615-626.

Evans, D. L., and Z. Zhu. 1993. AVHRR forforest mapping: national applicationsand global implications. Pages 76-79 inA. Lewis, ed. Looking to the future withan eye on the past: proceedings of the1993 ACMS/ASPRS. ASPRS.

——., Z. Zhu, S. Eggen-McIntosh, P. G. May-oral, and J. L. Omelas de Anda. 1992.Mapping Mexico’s forest lands withadvanced very high resolution radiom-eter. USDA For. Serv. Southern For. Exp.Stn., New Orleans, La.

Gardner, R. H., M. G. Turner, V. H. Dale, andR. V. O’Neill. 1992. A percolationmodel of ecological flows. Pages 259-269in A. J. Hansen and F. di Castri, eds.Landscape boundaries. Springer-Verlag,New York, N.Y.

Hanski, I., and M. Gilpin. 1991.Metapopulation dynamics: briefhistory and conceptual domain. Biol. J.Linnean Soc. 42:3-16.

Harrison, S. 1991. Local extinction in ametapopulation context: an empiricalevaluation. Biol. J. Linnean Soc. 42:73-88.

Hodgson, A., and P. Stacey. 1994. Habitat useand dispersal by Mexican spotted owls.U.S. For. Serv., Rocky Mtn. For. Rng.Exp. Stn., Flagstaff, Ariz.

Levins, R. 1969. Some demographic and geneticconsequences of environmental heteroge-neity for biological control. Bull.Entomol. Soc. Amer. 15:237-240.

Loveland, T. R., J. W. Merchant, D. O. Ohlen,and J. F. Brown. 1991. Development of aland-cover characteristics database for theconterminous U.S. PhotogrammetricEng. and Remote Sensing 57:1453-1463.

Noon, B. R., and K. S. McKelvey. 1992. Stabil-ity properties of the spotted owlmetapopulation in southern California.Pages 187-206 in J. Verner, K. S.

15


Volume II/Chapter 3

McKelvey, B. R. Noon, R. J. Gutiérrez, G. I.Gould, Jr., and T. W. Beck, eds. TheCalifornia spotted owl: a technicalassessment of its current status. USDAFor. Serv. Gen. Tech. Rep. PSW-133,Pac. Southwest Res. Stat., Albany, Calif.

Powell, D. S., J. L. Faulkner, D. Darr, Z. Zhu,amd D. W. MacCleery. 1993. Forestresources of the United States. USDAFor. Serv. Gen Tech Rep. RM-234,Rocky Mtn. For. Range Exp. Stat., Ft.Collins, Colo.

Pulliam, H. R. 1988. Sources, sinks, and popu-lation regulation. Am. Nat. 132:652-661.

Reynolds, R. T., and C. L. Johnson. 1994.Distribution and ecology of the Mexicanspotted owl in Colorado: annual report.USDA For. Serv., Rocky Mtn. For. Rng.Exp. Stn., Ft. Collins, Colo.

Schaffer, M. L. 1985. The metapopulation andspecies conservation: the special case ofthe northern spotted owl. Pages 86-99 in

R. J. Gutiérrez and A. B. Carey, eds.Ecology and management of the north-ern spotted owl in the Pacific Northwest.USDA For. Serv. Gen. Tech. Rep. PNW-185, Portland, Oreg.

Thomas, J. W., E. D. Forsman, J. B. Lint, E. C.Meslow, B. R. Noon, and J. Verner.1990. A conservation strategy for thespotted owl. U.S. Gov. Print. Off.,Washington, D.C. 427pp.

Willey, D. W. 1993. Home-range characteristicsand juvenile dispersal ecology of Mexi-can spotted Owls in southern Utah.Unpubl. Rep., Utah Div. Wildl. Resour.,Salt Lake City.

Zhu, Z. 1994. Forest density mapping in thelower 48 states: a regression procedure.USDA For. Serv. Res. Pap. SO-280,Southern For. Exp. Stat., New Orleans,Louis.

——., and D. L. Evans. 1992. Mappingmidsouth forest distributions withAVHRR data. J. For. 90:27-30.

16

Volume II/Chapter 4


CHAPTER 4: Habitat Relationships of theCHAPTER 4: Habitat Relationships of theCHAPTER 4: Habitat Relationships of theCHAPTER 4: Habitat Relationships of theCHAPTER 4: Habitat Relationships of theMexican Spotted Owl: Current KnowledgeMexican Spotted Owl: Current KnowledgeMexican Spotted Owl: Current KnowledgeMexican Spotted Owl: Current KnowledgeMexican Spotted Owl: Current Knowledge

Joseph L. Ganey and James L. Dick, Jr.

The Mexican spotted owl was listed asthreatened primarily due to concerns over loss ofhabitat (USDI 1993). Consequently, under-standing the habitat relationships of the Mexicanspotted owl is critical to developing soundmanagement plans for this species. Here, wesummarize both recent and historical informa-tion on habitat relationships of Mexican spottedowls. Because most historical information on theowl’s habitat contains little quantitative informa-tion, we rely mainly on recent information(1985 - present time).

Our primary objectives in this treatment areto: (1) evaluate and describe patterns of habitatuse by Mexican spotted owls; (2) evaluate anddescribe patterns of habitat selection by Mexicanspotted owls; (3) identify specific habitats orhabitat components that appear to be particu-larly important to the owl; and (4) identify areaswhere further information on habitat relation-ships of this owl are most needed. We use theterms habitat, habitat use, and habitat selectionas defined by Block and Brennan (1993). Thus,habitat as used here refers to “the subset ofphysical environmental factors that a speciesrequires for its survival and reproduction” (Blockand Brennan 1993:36). Habitat use refers to “themanner in which a species uses a collection ofenvironmental components to meet life requi-sites”, and habitat selection refers to “dispropor-tional use of environmental conditions” (Blockand Brennan 1993:38).

Patterns of habitat use evaluated here arelargely descriptive. We evaluated patterns ofhabitat selection by comparing use of vegetationtypes or specific habitat components to theoccurrence of those types or components.Ecological patterns and processes vary withspatial scale (Urban et al. 1987, O’Neill et al.1988, Turner 1989), however, and consideringpatterns of resource use at only one scale canyield misleading results (Porter and Church1987, Orians and Wittenberger 1991, Block andBrennan 1993). Therefore, we used a number ofdata sources, both published and unpublished,

to examine habitat use and/or selection at fivespatial scales. At some scales we could evaluatehabitat selection, whereas at others we couldonly describe patterns of habitat use. Scalesexamined are described below, arrayed fromcoarsest to finest scale:

1. Landscape scale - habitat use across theentire range of the Mexican spotted owl.

2. Home-range scale - patterns of habitatuse within owl home ranges as definedby locations of radio-tagged owls.

3. Stand scale - use of relatively homoge-neous units of forest vegetation withinowl home ranges.

4. Site scale - habitat use proximate to nest,roost, and foraging sites.

5. Tree scale - use of individual roost andnest trees.

In some cases, we reanalyzed existing datasets to provide more detailed information or toexplore different questions than those asked byoriginal investigators. Methods varied amongstudies, and will be discussed in conjunctionwith the specific data examined. Some of thedata used were from ongoing studies, and someof the analyses are preliminary. Therefore, whilethe information presented here represents thecurrent state of knowledge on habitat relation-ships of Mexican spotted owls, we caution thatadditional trends may emerge with further datacollection and analysis.

1

Volume II/Chapter 4


PATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATTHE LANDSCAPE SCALETHE LANDSCAPE SCALETHE LANDSCAPE SCALETHE LANDSCAPE SCALETHE LANDSCAPE SCALE

Literature ReviewLiterature ReviewLiterature ReviewLiterature ReviewLiterature Review

Available historical information on habitatuse by Mexican spotted owls in Arizona andNew Mexico was reviewed in Johnson andJohnson (1985), Ganey (1988), Ganey et al.(1988), and Skaggs (1988), and summarizedacross the range of the owl in McDonald et al.(1991). Most reports of Mexican spotted owls inboth popular and technical literature are anec-dotal and contain little detailed information.These reports were typically results of faunalsurveys that reported areas where they foundspecies. As such, they were neither systematicnor complete samples, and are thus biasedtowards the few places where spotted owls werelocated. While these reports provide someinformation on areas where owls were located,they should not be viewed as the ultimate wordon habitats used by spotted owls.

With these cautions in mind, the generalpicture that emerges is that owls occurred inhabitats ranging from low elevation riparianforests (Bendire 1892, Phillips et al. 1964, andpossibly Woodhouse 1853:63) to high elevationconiferous forests, but were most common inhigh elevation coniferous and mixed coniferous-broadleaved forests, often in canyons. In perhapsthe most detailed historical account of theMexican spotted owl, Ligon (1926:422) de-scribed its typical haunts as “deep, narrow,timbered canyons where there are always coolshady places.”

Several recent studies describe habitats usedby Mexican spotted owls. For example, Kertell(1977), Rinkevich (1991), and Willey (1992)reported on habitats occupied by Mexicanspotted owls in southern Utah (Colorado PlateauRU). All reported that owls were typically foundin narrow, steep-walled canyons, and all sug-gested that distribution was restricted by theavailability of such canyons. Reynolds (1993)also reported finding owls only in “steep-walled,deeply-cut canyons characterized or dominatedby exposed rocky slopes and tiers of rock cliffs”in Colorado (Colorado Plateau and Southern

Rocky Mountains-Colorado Recovery Units; seeUSDI 1995 for discussion of specific RecoveryUnits [RUs] within the range of the Mexicanspotted owl).

Ganey and Balda (1989a) recorded covertype at 55 roost sites in northern Arizona, ofwhich 53 were located in the Upper Gila Moun-tains RU. Of these, 92.4% were located inmixed-conifer forest. The remaining 7.6% wereclassified as ponderosa pine, but more likely werein ponderosa pine-Gambel oak forest. Most werelocated in steep canyons or on montane slopes.Also in this RU, Seamans and Gutiérrez (inpress) recorded cover type at 79 roost and 28nest sites in the Tularosa Mountains, NewMexico. These owls roosted and nested primarilyin mixed-conifer forests containing an oakcomponent, usually on the lower third of north-facing slopes (Seamans and Gutiérrez in press:fig. 1).

In the Basin and Range-West RU, Ganeyand Balda (1989a) recorded cover type at 64roost sites. Most were in mixed-conifer (48.4%)or Madrean pine-oak (29.7%) forest, with14.1% in encinal (evergreen oak), and 7.8% inponderosa pine forest. At 19 roost and/or nestsites observed by Duncan and Taiz (1992) in thisRU, 31.6% were in mixed-conifer forest, 31.6%in Madrean pine-oak forest, 26.3% in Arizonacypress forest, and 10.5% in encinal. In bothstudies, most owls were observed in montanecanyons.

Skaggs and Raitt (1988) surveyed 42,105 ha(104,000 acres) in the Basin and Range-EastRU. Survey areas were divided into 18 plots of23.3-km2 (9-mi2) prior to survey, with all plotsclassified by the dominant forest type within theplot (6 each in mixed-conifer, ponderosa pine, orpinyon-juniper). They found 33 occupied sites;72.7% in mixed-conifer, 18.2% in ponderosapine, and 9.1% in pinyon-juniper. Even in plotsclassified as ponderosa pine or pinyon-juniper,the owls typically roosted in pockets of mixed-conifer forest (Roger Skaggs, New Mexico StateUniv., Las Cruces, NM, pers. comm.). Kroel(1991:39) also noted that owls in this RUroosted primarily (79% of observed roosts) inmixed-conifer forest, with limited use of ponde-rosa pine forest and pinyon-juniper woodland(14 and 4% of all roosts, respectively).

2

Volume II/Chapter 4


Little recent information exists regardinghabitat use by spotted owls in Mexico. However,Tarango et al. (1994) reported finding Mexicanspotted owls in isolated patches of pine-oakforest in steep canyons in the Sierra MadreOccidental-Norte.

With regard to all of the above studies, it isimportant to note that definitions of cover typesmay have varied slightly among observers, andwe cannot be certain that cover types are com-pletely consistent with definitions used in thisPlan.

Recovery Team Analysis of RecentRecovery Team Analysis of RecentRecovery Team Analysis of RecentRecovery Team Analysis of RecentRecovery Team Analysis of RecentInventory DataInventory DataInventory DataInventory DataInventory Data

Recent inventories by land managementagencies, particularly the U. S. Forest Service(FS), have generated considerable informationon owl locations and habitats used. We used thisinformation to evaluate use of vegetation (orcover) types across the range of the subspecies.Crews visited FS District and Supervisors offices,collated inventory and monitoring data from1990 (when survey efforts were standardizedthroughout the Region) to 1993 (when thesedata were collated), and entered the informationgathered into a data base. These same crews alsocollated records from the NPS, BLM, Triballands, State wildlife agencies, Fort HuachucaMilitary Reservation, and independent research-ers in both the United States and Mexico for thesame time period. Thus, we attempted to gatherall existing information on current distributionand habitat use of the Mexican spotted owlacross its range.

Vegetation type at nest or roost sites wasentered directly from field data forms, with typesgenerally corresponding to Series level designa-tions described in Brown et al. (1980). We couldnot assess the accuracy or consistency of habitatclassification across the range of the owl. Alllocations for which vegetation type informationwas missing or ambiguous were omitted fromanalysis.

We used only visual observations (such asbirds at roost and nest sites) to assess habitat usebecause habitat cannot be determined for distantowls heard at night. Most roost or nest locations

entered in this database could not be assigned toa particular “management territory.” As a result,we are uncertain how many unique pairs of owlswere represented, or how many roost or nestsites might represent a single pair. This poten-tially serious lack of independence in the datarendered statistical comparisons among RUs orbetween roost and nest sites meaningless. There-fore, we simply present summaries of vegetationtypes used for roosting and nesting by RecoveryUnit. We could not compare habitat use withhabitat occurrence at this scale, because of theproblems discussed above regarding lack ofindependence, because no geographical informa-tion system (GIS) coverage was available docu-menting areas surveyed, and because we wereunable to obtain a rangewide vegetation typecoverage. The closest we could come to arangewide vegetation type coverage covered onlythe U. S. portion of the range of the Mexicanspotted owl, and had a minimum resolution of 1km2 (Keitt et al. 1995). This scale is not appro-priate for evaluating roost and nest sites, whichrequire a much finer scale of resolution.

Mexican spotted owls used a variety ofvegetation types across their range (Table 4.1).Although the range of vegetation types usedvaried among RUs, mixed-conifer forest washeavily used in most RUs. In contrast, encinalwas used only in the Basin and Range-West, andpine-oak forest was used primarily in the Basinand Range-West and Sierra Madre Occidental-Norte. The pine-oak forest found in these RUs isMadrean in affinity and differs in both speciescomposition and habitat structure from theponderosa pine-Gambel oak forest found inother RUs (Brown et al. 1980, Ganey et al.1992). Ponderosa pine forest was used rarely, andpinyon-juniper woodland was used primarily byowls in the Colorado Plateau. Riparian forestwas used in several RUs, but at relatively lowlevels.

We could not statistically compare RUsbecause of the problems discussed above. Basedon discussions with researchers and managersfamiliar with local situations, however, webelieve that the observed differences in patternsof habitat use among RUs are both real andecologically important.

3

Table 4.1.Table 4.1.Table 4.1.Table 4.1.Table 4.1. Percent of nest (top row) and roost (bottom row) sites in various vegetation types in different Recovery Units. Based on analysis of inventoryand monitoring data collected since 1990. Vegetation types follow Series in Brown et al. (1980).

FPO4

PATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATTHE HOME-RANGE SCALETHE HOME-RANGE SCALETHE HOME-RANGE SCALETHE HOME-RANGE SCALETHE HOME-RANGE SCALE

Home-Range SizeHome-Range SizeHome-Range SizeHome-Range SizeHome-Range Size

Several studies have examined home-rangesize and/or habitat use within home ranges ofradio-tagged Mexican spotted owls. Minimumconvex polygon (MCP; Mohr 1947) and 95%adaptive kernel (AK; Worton 1989, wherepossible) estimates of home-range size from thesestudies are presented in Table 4.2. Estimates arepresented separately for each study area. Home-range size appeared to vary considerably amongstudy areas, even within a restricted geographicarea (Table 4.2). For example, Zwank et al.(1994) found that home-range size differedsignificantly between two drainages in theSacramento Mountains, New Mexico. Rangeswere smaller in the Rio Penasco watershed (n = 5owls), where mixed-conifer forest comprised 65-86% of home ranges, than in the Sixteen Springswatershed (n = 4 owls), where mixed-coniferforest comprised only 6-42% of home ranges,with most of the remainder consisting of ponde-rosa pine forest and pinyon-juniper woodland(Zwank et al. 1994: table 1). Similarly, Ganeyand Balda (1989b) observed large differences inhome-range size among owls on the San Fran-cisco Peaks and in Walnut Canyon. Althoughthese areas are separated by only approximately16 km (10 mi), habitat composition differsbetween the two areas (Ganey and Balda 1994:table 3). These results suggest that home-rangesize of Mexican spotted owls may vary amongcover types. However, small numbers of owlstracked in all studies, as well as differences insampling intervals and seasons covered limit ourability to make comparisons among study areas.Consequently, we caution that these numbersrepresent general estimates of areas used byspotted owls, and as such do not support sweep-ing generalizations about differences betweenareas and/or cover types.

Size of Activity CentersSize of Activity CentersSize of Activity CentersSize of Activity CentersSize of Activity Centers

Gutiérrez et al. (1992) suggested that thesmallest area encompassing 50% of nocturnal

foraging locations (the 50% adaptive kernel)could define a foraging activity center. Becauseof the great variability in available estimates ofhome-range size for Mexican spotted owls, wetook a more conservative approach and defined anocturnal activity center based on the areaenclosed by the adaptive kernel contour encom-passing 75% of foraging locations. This activitycenter was defined only for territories whereboth pair members were radio-tagged. In gen-eral, owls appeared to forage primarily within arelatively small portion of the home range,suggesting high concentration of activity (Table4.3). This pattern appeared to hold regardless ofRU or study area (but see above cautions regard-ing comparisons among studies).

Habitat Composition and UseHabitat Composition and UseHabitat Composition and UseHabitat Composition and UseHabitat Composition and Use

Three studies quantified habitat compositionwithin MCP home ranges of radio-tagged owls(Willey 1993, Ganey and Balda 1994, Zwank etal. 1994). Home ranges were most variable interms of vegetation types in the ColoradoPlateau RU, encompassing types ranging frommixed-conifer forest to mountain shrub andgrassland (Willey 1993: table 4). Home rangeswere dominated by mixed-conifer and ponderosapine forests in the Upper Gila Mountains RU(Ganey and Balda 1994), and by mixed-coniferforest, ponderosa pine forest, and pinyon-juniperwoodland in the Basin and Range-East RU(Zwank et al. 1994).

Ganey and Balda (1994) compared use ofvegetation types by foraging, radio-tagged owlsto the area of those types within the MCP homerange (a measure of relative availability), usingchi-square tests and Bonferroni confidenceintervals (Neu et al. 1974, Byers et al. 1984).This comparison involved eight owls represent-ing five pairs on three study areas.

Observed patterns of habitat use by foragingowls were complex. In relation to area of differ-ent forest types within their home ranges, allindividual owls used forest types nonrandomly.All forest types were used by foraging owls. Ingeneral, individual owls foraged significantlymore than or as expected in unlogged forests andsignificantly less than or as expected in selec-

Volume II/Chapter 4


5

FPO

Table 4.2.Table 4.2.Table 4.2.Table 4.2.Table 4.2. Home-range sizes (ha) of radio-marked Mexican spotted owls. Part A shows ranges of pairs with both members radio-tagged; part B showsranges for individual owls. Study areas represent mesic mixed-conifer forest on the San Francisco Peaks (SFP), White Mountains (WM), and SacramentoMountains (SM-MC); a mixture of mixed-conifer, ponderosa pine, and xeric pinyon-juniper in the Sacramento Mountains (SM-XE); a rocky canyon(Walnut Canyon, WC); ponderosa pine-Gambel oak forest near Bar-M Canyon (BMC); rugged canyons with mixed forests in the San Mateo Mountains(SANMAT); a mixture of mesic mixed-conifer forest and pinyon-juniper in rocky canyons along Elk Ridge (MANTI) and in Zion National Park (ZION);and xeric pinyon-juniper in rocky canyons in Capitol Reef and Canyonlands National Parks (CNP). Shown are the means (± SD) for 100% minimumconvex polygon (MCP) and 95% adaptive kernel (95% AK) estimates.

6

Table 4.2. Table 4.2. Table 4.2. Table 4.2. Table 4.2. (continued)

FPO

7

Table 4.3.Table 4.3.Table 4.3.Table 4.3.Table 4.3. Size (mean ± standard deviation) of nocturnal activity centers of radio-tagged pairs of Mexcan spotted owls in the Upper Gila Mountains andBasin and Range-East Recovery units. Activity centers defined as the area included in the adaptive kernel contour enclosing 75% of owl foraging locations.

FPO

8

tively logged forest types (Table 4.4). Thiscomparison is overly simplistic, however. Bothlogged and unlogged forests contained a widerange of stand structures. Some logged standswere used for foraging, and some unloggedstands were not used. In addition, two of theunlogged forest types (virgin mixed-coniferforest on rocky slopes and ponderosa pine-oak-juniper forest) were found primarily on rockyslopes interspersed with significant rock outcropsand cliffs, and owls appeared to forage in theserocky areas as well as in the forest (Ganey andBalda (1994:165). Thus, these findings do notindicate that all unlogged stands provide suitableforaging conditions and that all logged stands donot. Rather, they suggest that patterns of habitatuse by foraging owls are complex and not ame-nable to simplistic explanations. At this time, wecannot explain why some logged forests are usedand others are not.

Differences in patterns of habitat use amongand within study areas also suggest that patternsof habitat use for foraging are complex. Use ofparticular forest types sometimes varied consid-erably among individual owls within a study area(Table 4.4), and even between members of amated pair (Ganey and Balda 1994: table 3). Inshort, there is considerable variability in patternsof habitat use for foraging. As a result, it isdifficult to generalize about patterns of habitatselection by foraging owls based on currently-available data.

In contrast to the variable patterns observedamong foraging owls, patterns of habitat use byroosting owls were relatively consistent amongstudy areas and individual owls. Habitat use forroosting was not compared statistically to habitatoccurrence because of small sample sizes forsome individual owls. All owls roosted primarilyin unlogged mixed-conifer forests, but someused unlogged ponderosa pine forest as well(Ganey and Balda 1994; table 3). Very littleroosting occurred in logged stands, particularlyduring the breeding season.

Seasonal MovementsSeasonal MovementsSeasonal MovementsSeasonal MovementsSeasonal Movements

Most Mexican spotted owls appear to remainin the same general area throughout the year,whereas others migrate in winter, usually tolower elevations. Year-round residents often uselarger ranges during the nonbreeding seasonthan during the breeding season (Kroel 1991,Willey 1993, Ganey and Block unpublisheddata), and there appear to be shifts in use of areaand habitat for some owls (Ganey and Balda1989b, Kroel 1991, Willey 1993, Ganey andBlock unpublished data). No quantitative resultsare yet available that describe wintering habitatsused by owls remaining in the same areathroughout the year.

Most samples of Mexican spotted owlsstudied using radiotelemetry appear to havesome individuals who are either migratory ornomadic in winter, but this generally represents aminority of the population. For example, Willey(1993:15) reported that two of 11 (18.2%)radio-tagged owls on the Colorado Plateau RUmigrated during winter. Both apparently left thebreeding area during October and returnedduring February. One moved up in elevation towinter in coniferous forest, whereas the otherwintered in mountain shrub habitat. Both owlsmoved 20-25 km between breeding seasonranges and winter ranges.

In the Upper Gila Mountains RU, two ofeight owls (25%) in one study (Ganey and Balda1989b) and 2 of 13 (15.4%) in another (Ganeyand Block unpublished data) migrated duringthe winter. All left the breeding-season rangebetween November and January, and returned inMarch or April. Wintering areas of two owlswere never located. The remaining two owlsmigrated approximately 50 km, from ponderosapine-Gambel oak forest at approximately 2290m (7500 ft) in elevation to pinyon-juniperwoodland at approximately 1370 m (4490 ft) inelevation (Ganey et al. 1992, Ganey and Blockunpublished data). The wintering area waslocated in the Basin and Range-West RU,providing evidence that some individuals maymove seasonally between RUs.



Table 4.4.Table 4.4.Table 4.4.Table 4.4.Table 4.4. Use of forest types for foraging by radio-tagged Mexican spotted owls on three study areas in northern Arizona. Data summarized fromGaney and Balda (1994).

FPO

10

In the Basin and Range-East RU, neitherZwank et al. (1994, n = 9 owls) nor Ganey andBlock (unpublished data, n = 15 owls) observedmigration during the nonbreeding season, buttwo of eight radio-tagged owls (25%) in anotherstudy moved during the winter. These owlsmoved downslope from mixed-conifer forests tothe interface between pinyon-juniper woodlandand desert scrub (Roger Skaggs, New MexicoState Univ., Las Cruces, NM; pers. comm.).

PATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATTHE STAND SCALETHE STAND SCALETHE STAND SCALETHE STAND SCALETHE STAND SCALE

Limited data were available on habitat use byMexican spotted owls at the stand scale duringpreparation of this Recovery Plan. Analysis of FSdata for some nesting stands in the Upper GilaMountains and Basin and Range-East RUssuggests that owls typically nest in relativelydense stands with high basal areas of live trees, a

wide range of tree sizes, suggesting an uneven-aged structure (Figure 4.1), and a large treecomponent (Table 4.5). The data were also usedto estimate diminution quotients, or q-factors,for nesting stands in both RUs (Table 4.5). Thisparameter is useful in uneven-aged managementof timber stands. It describes the ratio of numberof trees in any diameter class to the number inthe next-lowest diameter class, and thus describesthe relative shape of the diameter distribution(Daniel et al. 1979). In both RUs, q-factorsaveraged <1.4.

PATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATPATTERNS OF HABITAT USE ATTHE SITE SCALETHE SITE SCALETHE SITE SCALETHE SITE SCALETHE SITE SCALE

Several studies have examined characteristicsof specific sites used by Mexican spotted owls,such as nest and roost sites. These are discussedbelow, by site type.

Figure 4.1.Figure 4.1.Figure 4.1.Figure 4.1.Figure 4.1. Diameter distributions of live trees sampled in nest stands in the Upper Gila Mountains(UGM; n = 13 stands) and the Basin and Range - East (BR-E; n = 44 stands) Recovery Units. Shown ispercentage of total live tree basal area by 4 in (10 cm) size classes. Data from the FS stand data base.

FPO



Table 4.5.Table 4.5.Table 4.5.Table 4.5.Table 4.5. Selected characteristics of nest stands of Mexican spotted owls in the Upper Gila Mountains (n = 13 stands) and Basin and Range-East(n = 44 stands) Recovery Units. Data from the stand data bases for the Apache-Sitgreaves (Upper Gila Mountains) and Lincoln (Basin and Range-East)National Forests. Values shown are means; no estimates of variability among stands were available.

FPO

12

Nest SitesNest SitesNest SitesNest SitesNest Sites

Armstrong et al. (1994) and Arizona Gameand Fish Department (unpublished data)sampled habitat characteristics at nest sites onthe Apache-Sitgreaves National Forest, UpperGila Mountains RU (Table 4.6). They sampledboth tree and cliff nests, but the majority ofnests sampled were in trees. While limited inscope, their results suggest that owls typicallynested in large trees in closed-canopy stands.

Ruess (1995) documented current standstructure on 0.04-ha (0.1-ac) plots at 11 nestand 9 roost sites representing an unknownnumber of owl pairs in ponderosa pine-Gambeloak forest in northcentral Arizona. He alsoattempted to estimate what these sites wouldhave looked like in terms of forest structure in1876, before effective fire suppression began inthis area, using methods developed byCovington and Moore (1994). With a fewexceptions, Ruess (1995) pooled roost and nestsites when presenting data. Thus, both site typeswill be discussed together here.

Comparisons of current and estimatedpresettlement conditions on these sites suggestthat pronounced increases have occurred in treedensity, basal area, and canopy cover. Currentdensity of trees above breast height averaged1708.8 ± 1009.1 (SD) trees/ha (691.8 ± 408.5trees/ac), versus an estimated presettlementdensity of 0-225 trees/ha (0-91 trees/ac; [Ruess1995:12, no mean presented]). Current basalarea averaged 66.7 m2/ha (290.7 ft2/ac; [Ruess1995:12, no estimate of variability provided]),versus an estimated average basal area of 17.8m2/ha (77.6 ft2/ac; [Ruess 1995:12, no estimateof variability provided]) circa 1876. Canopycover, modeled based on projection of mappedtree crowns, was estimated at 44.8 ± 12.9% atpresent, versus 2.2 ± 2.9% circa 1876 (Ruess1995:22). Variability in species composition andstructural variables was relatively high for bothestimated 1876 conditions and current condi-tions.

Two detailed treatments of nesting habitat ofthe Mexican spotted owl are currently available(SWCA 1992, Seamans and Gutiérrez in press).Seamans and Gutiérrez (in press) compared

habitat characteristics between 27 plots (0.04 ha;0.1 ac) centered on nest trees and 27 randomplots from throughout their study area (TularosaMountains, Gila National Forest, New Mexico;Upper Gila Mountains RU). The nest plotsrepresented nest sites of 27 pairs of owls. Ran-dom plots were centered on a randomly-selectedtree �27.3 cm diameter at breast height (dbh),the minimum diameter among the 27 nest trees.This was an attempt to minimize the potentialbias associated with centering nest plots on largetrees and random plots on trees of any size.

Owls nested in mixed-conifer/oak forestsmore than expected by chance, and in pine-oakand pinyon-juniper forests less than expected(Seamans and Gutiérrez in press: fig. 1). Mostnests were located on the lower third of slopes,and the mean slope aspect at nest sites wasnortherly. Nest plots differed significantly fromrandomly-located plots within the study area fora number of variables (Table 4.7). In a discrimi-nant function analysis, nest plots were bestseparated from random plots by variance in treeheight, canopy closure, and basal area of maturetrees (defined as stems >45.8 cm dbh); all weregreater on nest than on random plots (Table4.7). Cross-validation analyses indicated that theresults of the discriminant analysis were stable,and the discriminant function successfullyclassified 84.6% of a sample of 13 owl nest sitesfrom other mountain ranges outside the studyarea (Seamans and Gutiérrez in press).

Seamans and Gutiérrez (in press) also com-pared nest plots to 27 plots randomly locatedwithin nest stands. Nest plots did not differsignificantly from random plots within the samestand.

SWCA (1992) sampled habitat characteris-tics at 84 nests on FS lands in Arizona and NewMexico. They sampled habitat characteristicswithin circular plots (0.2 ha; 0.5 ac), eachcentered on and including either a nest tree, arandomly selected tree within the nest stand, ora randomly selected tree in a stand within 0.8km (0.5 mi) of the nest. These will be referred toas nest, nest-stand, and random-stand plots.SWCA (1992) concluded that owls selected nestsites based primarily on the availability of asuitable nest tree. Hardwood snag basal area andcanopy cover also emerged as potentially impor-

Volume II/Chapter 4


13

Table 4.6. Table 4.6. Table 4.6. Table 4.6. Table 4.6. Habitat characteristics sampled at Mexican spotted owl nest sites on three Ranger districts, Apache-Sitgreaves National Forest, Upper GilaMountains Recovery Unit. Shown are means and standard errors (in parentheses). Nests sampled were found in trees (n = 30) or on cliffs (n = 4); somevariables are relevant to only one of the two situations.

FPO

14

tant factors in their analysis. This data set wasreanalyzed by both Zhou (1994) and the Team,using different methods. These reanalyses arediscussed below.

Reanalysis of SWCA (1992) by Zhou (1994)Reanalysis of SWCA (1992) by Zhou (1994)Reanalysis of SWCA (1992) by Zhou (1994)Reanalysis of SWCA (1992) by Zhou (1994)Reanalysis of SWCA (1992) by Zhou (1994)

Zhou (1994) conducted three, two-grouplinear discriminant function analyses betweennest, nest-stand, and random-stand plots. Heconcluded (Zhou 1994:96-102) that:

1. Mexican spotted owl nest stands differedfrom randomly-selected stands in thevicinity. Nest stands typically had a widerange of tree diameters and heights, largemaximum tree diameter, and high treebasal area. Nest stands also had higherspecies richness than random stands.

2. Mexican spotted owls also selected formicrosites within nest stands. Nest plotswere located on steeper slopes, hadgreater live tree basal area, and were morelikely to be found on north or eastaspects than nest-stand plots.

3. Diameter and height distributions hadsimilar shapes on all three plot types, butthe spread of the distribution differedamong plot types. With respect todiameter distributions, the nest and nest-stand plots had greater percentages oflarge trees. With respect to tree height,nest and nest-stand plots had greaterspread to the distribution than random-stand plots. This wider spread suggests atendency for the owl to nest in multi-storied stands, as wider spread to theheight distribution increases the prob-ability that the stand is multi-storied.

4. Mexican spotted owls nested in largetrees. The mean nest tree was located inthe 92nd diameter percentile and the79th height percentile.

5. The linear combination of habitatvariables was more successful in classifying habitat than reliance on interpreta-tion of coefficients for single variables.

Table 4.7.Table 4.7.Table 4.7.Table 4.7.Table 4.7. Habitat characteristics at Mexican spotted owl nest (n = 27) and randomly located sites(n = 27) in the Tularosa Mountains, New Mexico; Upper Gila Mountains RU. Shown are means andstandard deviation (in parentheses). Data from Seamans and Gutiérrez (in press).

FPO



Volume II/Chapter 4


Reanalysis of SWCA (1992) by theReanalysis of SWCA (1992) by theReanalysis of SWCA (1992) by theReanalysis of SWCA (1992) by theReanalysis of SWCA (1992) by theRecovery TeamRecovery TeamRecovery TeamRecovery TeamRecovery Team

The Team also reanalyzed the data fromSWCA (1992). The Team was interested inpatterns of habitat use within particular geo-graphic regions, and in how similar nest siteswere among regions. This reanalysis was re-stricted to sites in the Upper Gila Mountainsand Basin and Range-East RUs because thesewere the only RUs well represented among nestssampled (n = 44 and 26 sites, respectively).Because sample sizes were small, sites werepooled among habitat types within each RU.Habitat characteristics were first comparedamong plot types within RUs, to see if nest sitesdiffered from nest stands or locally-available,randomly-selected stands. Characteristics of nestplots were next compared between RUs, to seehow similar nesting habitat was in differentgeographic areas.

We used chi-square tests to evaluate differ-ences in tree species composition between plottypes (and/or RUs), and Kolmogorov-Smirnovtests to compare diameter distributions. Becausethe Kolmogorov-Smirnov test compares onlytwo groups at a time, three separate tests werenecessary to compare all three plot types. Toavoid inflating the Type I error rate, we parti-tioned the error among these three comparisonsusing a Bonferroni adjustment, and used analpha level of P <0.016 for significance.

Other habitat variables were comparedamong plots or between RUs using univariateANOVAs. Where these comparisons weresignificant, Scheffe’s multiple range test was usedto determine where the differences occurred.Variables included diameter at breast height(dbh) of center tree, density (trees/ha) of livetrees and snags >12 cm (4.7 in) dbh, basal area(m2/ha) of live trees and snags >12 cm dbh, andvolume of logs (m3/ha) >12 cm in large-end-diameter. Basal area was calculated for eachindividual tree or snag based on dbh, thensummed within each plot. Log volume wascalculated using diameter and length measures,assuming a cylindrical shape. Log volumes areoverestimated (perhaps greatly) because diameterwas measured at the large end.

CCCCComparisons omparisons omparisons omparisons omparisons WWWWWithin Rithin Rithin Rithin Rithin Recoecoecoecoecovvvvvererererery Uy Uy Uy Uy Units.—nits.—nits.—nits.—nits.—Treespecies composition differed significantly amongplot types in both RUs. In the Upper GilaMountains RU, nest plots contained greaterproportions of Douglas-fir and Gambel oak andless ponderosa pine than did random-stand plots(Figure 4.2). In the Basin and Range-East RU,nest and nest-stand plots contained more whitefir and less ponderosa pine than did random-stand plots (Figure 4.2).

Diameter distributions in both RUs weresignificantly different between nest plots andboth nest-stand and random-stand plots, butwere not significantly different between nest-stand and random-stand plots. Nest plots typi-cally had lower proportions of basal area in thesmallest size classes than did random-stand plots(Figure 4.3). In general, basal area was moreevenly distributed across size classes in neststands than in random stands, suggesting a trendtoward uneven-aged stands. This trend was moreevident in the Basin and Range-East RU than inthe Upper Gila Mountains RU. Grouping oftrees into four diameter classes, representing“young,” “mid-aged,” “mature,” and “old” trees(USDA Forest Service 1993), also indicates thatnest plots contained relatively fewer trees in thesmallest size class and more trees in the largesttwo size classes than other plots (Table 4.8a, b;note that these comparisons refer to percentagesof total trees, not to absolute numbers). In bothRUs, nest trees were significantly larger(P <0.0001) than randomly-selected trees (Table4.8). Only 2.4% of 21,951 trees and snagssampled had a dbh �61.4 cm (24.2 in), themean diameter for nest trees.

In the Upper Gila Mountains RU, plot typesdiffered significantly in snag density(P <0.0001), snag basal area (P = 0.0003), livetree basal area (P <0.0001), and log volume(P = 0.0006), but not in live tree density(P = 0.30). For snag density, nest plots differedfrom both nest-stand and random-stand plots;and nest-stand plots also differed from random-stand plots. For snag basal area, nest plots andnest-stand plots differed from random-standplots, but not from each other. For live tree basalarea, nest plots differed from both nest-standand random-stand plots, but we found nodifference between the latter two plot types. For

16

Volume II/Chapter 4


FPO

Figure 4.2a.Figure 4.2a.Figure 4.2a.Figure 4.2a.Figure 4.2a. Tree species composition within nest- and random-stand plots. Data reanalyzed fromSWCA (1992). Upper Gila Mountains Recovery Unit (n = 44 sites).

17

Volume II/Chapter 4


FPO

Figure 4.2b.Figure 4.2b.Figure 4.2b.Figure 4.2b.Figure 4.2b. Basin and Range-East Recovery Unit (n = 26 sites).

18

Volume II/Chapter 4


FPO

Figure 4.3.Figure 4.3.Figure 4.3.Figure 4.3.Figure 4.3. Diameter distributions of live trees sampled on nest, nest stand, and random stand plots.Shown is percentage of total live tree basal area by 4 in (10 cm) size classes for (a)(a)(a)(a)(a) Upper Gila Moun-tains, and (b)(b)(b)(b)(b) Basin and Range-East. Data reanalyzed from SWCA (1992).

19

Volume II/Chapter 4


log volume, both nest plots and nest-stand plotsdiffered from random-stand plots, but not fromeach other (Table 4.8a).

Of the five characteristics discussed above,only live tree basal area and log volume differedsignificantly (P = 0.0121 and 0.0061, respec-tively) between plot types in the Basin andRange-East RU (all other P values >0.05). Bothvariables differed only between nest plots andrandom-stand plots, not between nest plots andnest-stand plots or between nest-stand andrandom-stand plots (Table 4.8b).

CCCCComparisons Betwomparisons Betwomparisons Betwomparisons Betwomparisons Between Reen Reen Reen Reen Recoecoecoecoecovvvvvererererery Uy Uy Uy Uy Units.—nits.—nits.—nits.—nits.—Speciescomposition of nest sites differed significantlybetween RUs (�2 = 1669, df = 4, P <0.00001).Nest sites in the Basin and Range-East RU weredominated by white fir and Douglas-fir, those inthe Upper Gila Mountains RU by ponderosapine, Douglas-fir, and Gambel oak (Figure 4.2).Diameter distributions also differed betweenRUs in comparisons of nest plots (Figure 4.3).No differences were found between RUs for nesttree dbh, live tree basal area or density, snag basalarea or density, or log volume.

Table 4.8a.Table 4.8a.Table 4.8a.Table 4.8a.Table 4.8a. Habitat characteristics sampled at 44 Mexican spotted owl nest sites in the Upper GilaMountains Recovery Unit, as well as at randomly located plots within the nest stand and in a randomlyselected stand within 0.5 mi of the nest site. Data reanalyzed from SWCA (1992). Values shown aremean (± standard deviation).

FPO

20

Volume II/Chapter 4


Roost sitesRoost sitesRoost sitesRoost sitesRoost sites

Many studies have described characteristicsof roost sites. Some of these descriptions arebased on plot-level sampling, whereas others arebased on sampling of microsites (such as anindividual tree). Microsite descriptions will bediscussed under “Patterns of Habitat Use at theTree scale”; plot-level data are discussed below.

In most of the studies described here, thenumber of plots measured exceeded the numberof owls studied. Thus, these plots cannot beconsidered totally independent samples, and

apparent levels of significance may be inflated(Hurlbert 1984). In most cases, significancelevels are so high that we believe pseudoreplica-tion is not a major problem. This is furthersuggested by the fact that results are comparablebetween these studies and another (Seamans andGutiérrez in press) that did not involve pseu-doreplication (see below).

Rinkevich (1991; see also Rinkevich andGutiérrez in review) and Willey (1993) sampledhabitat characteristics within roost-centeredcircular plots of 0.04 ha (0.1 ac) each within theColorado Plateau RU. All roost sites were

Table 4.8b.Table 4.8b.Table 4.8b.Table 4.8b.Table 4.8b. Habitat characteristics sampled at 26 Mexican spotted owl nest sites in the Basin andRange-East Recovery Unit, as well as at randomly located plots within the nest stand and in a ran-domly selected stand within 0.5 mi of the nest site. Data reanalyzed from SWCA (1992). Valuesshown are mean (± standard deviation).

FPO

21

located in narrow gorges and/or canyons.Rinkevich (1991) used discriminant functionanalysis to compare owl roost sites to randomlylocated sites in Zion National Park, Utah. Owlsites had higher absolute humidity, more vegeta-tion strata, narrower canyon width, and higherpercent ground litter than random sites(Rinkevich and Gutiérrez in review). She alsocompared randomly located plots (n = 54)within canyons where owls were heard to plots(n = 44) within canyons where owls were notheard. Canyons occupied by owls had higherhumidity and snag basal area than canyonswhere owls were not heard (Rinkevich andGutiérrez in review). The number of plots inthese analyses is greater than the number of owls(or canyons in the second analysis), which isunknown.

Willey (1993) sampled habitat characteristicson 129 plots representing roost sites of 14 radio-tagged owls on three study areas in southernUtah. Habitat features at roost sites were com-pared to characteristics measured at 30-50random points scattered within each home range(Willey 1993:10). Seven variables were analyzedusing discriminant function analysis (Willey1993:10). Temperature, slope, vegetation canopycover, and number of ledges and fir trees dis-criminated between roosting plots and randomplots (Willey 1993: table 5). Univariate analysesprovided similar results, suggesting that “owlsused narrow canyon roosts characterized by cooldaytime temperatures, steep slopes, and relativelydense overhead cover. Roosts typically possessedlarge trees in juxtaposition with caves and ledges,providing a more complex habitat architecturethan the surrounding habitat” (Willey 1993:16).The number of plots also exceeded the numberof owls in this study.

Ganey (1988, see also Ganey and Balda1994) sampled habitat characteristics on 167circular plots (0.04 ha; 0.1 ac) within four owlhome ranges (defined using the MCP method)in the Upper Gila Mountains RU. Plots repre-sented high-use roosting and foraging sites, andrandomly-selected sites within owl ranges. Thisstudy also had more plots than owls. In addition,roost plots were always tree-centered, whereasonly some foraging and randomly-selected plotswere tree-centered. This could create a positive

bias for tree-related variables, such as tree den-sity, tree basal area, and canopy cover, on roostplots.

Plot type was misclassified 33% of the timein a 3-group discriminant function analysis(Ganey 1988; all classification rates refer tojackknifed classification). Most misclassificationoccurred between foraging and roosting sites.Two-group analyses had higher rates of success-ful classification and were easier to interpret. Thefunction that resulted in maximum separation ofroosting and foraging sites correctly classified76% of the sites. Variables entering the equationwere canopy closure and snags/ha; both weregreater on roosting sites (Table 4.9). A compari-son of foraging and random sites resulted in84% successful classification. Variables enteringthe discriminant function were total basal areaand big down logs/ha (defined as logs >30.5 cm[12.0 in] in diameter); both were greater onforaging sites. Comparing roosting and randomsites, 90% of all sites were successfully classified.Variables entering the discriminant functionwere total basal area, snags/ha, canopy closure,and big down logs/ha; all were greater on roost-ing sites.

Using data from the plots sampled by Ganey(1988), the Team conducted Kolmogorov-Smirnov tests on diameter distributions asdescribed under nest sites (see above). We foundno difference in diameter distributions betweenroosting sites and either foraging or randomsites. Diameter distributions were significantlydifferent between foraging and random sites. Ingeneral, foraging sites had fewer small trees andmore trees in the largest size classes than randomsites (Table 4.9). Again, note that these arerelative comparisons, and refer to percentages oftotal trees rather than to absolute numbers oftrees.

Seamans and Gutiérrez (in press) comparedhabitat characteristics sampled on 0.04-ha (0.1-ac) circular plots at 78 roost sites and 71 randomsites, Tularosa Mountains, Upper Gila Moun-tains RU. Roost plots were centered on the roosttree (one plot each from 78 separate owls), andrandom plots were centered on randomly-selected trees throughout the study area. Owlsroosted in mixed-conifer/oak forest more thanexpected by chance, and in pine-oak forest and



Table 4.9.Table 4.9.Table 4.9.Table 4.9.Table 4.9. Habitat characteristics sampled on 0.04 ha (0.1 ac) circular plots within home ranges of radio-tagged Mexican spotted owls inhabitingmixed-conifer and ponderosa pine forests, northern Arizona (Upper Gila Mountains Recovery Unit). Data from Ganey and Balda (1994); n = sixowls occupying four home ranges. Shown are means and standard deviations (in parentheses).

FPO

23

Volume II/Chapter 4


pinyon-juniper woodland less than expected.Most roosts were located on the lower third onslopes. Roost sites differed significantly fromrandom plots for several variables (Table 4.10).Roost sites were best separated from randomsites in a discriminant function analysis bycanopy closure and height variance; both weregreater on roost than on random sites. Cross-validation analyses indicated that the discrimi-nant analysis results were stable (Seamans andGutiérrez in press).

A. Hodgson and P. Stacey also evaluatedroost sites in the Upper Gila Mountains RU(Peter Stacey, Univ. of Nevada, Reno, NV; pers.comm.). They compared habitat characteristicssampled on 0.04-ha (0.1-ac) circular plotsbetween 55 roost sites and 69 random sites inthe San Mateo Mountains, New Mexico (thenumber of plots is also greater than the numberof owls in this study). Owls typically roosted inor near canyon bottoms, in relatively densestands of mixed-conifer forest containing signifi-cantly more Douglas-fir, Gambel oak, andlimber pine than random sites. Deciduous treesaccounted for >28% of total basal area and>50% of total tree density. Roost trees averaged31 cm (11.6 in) dbh, with most roosting occur-ring in Douglas-fir (54%) or Gambel oak (21%).

In a second comparison, Hodgson andStacey restricted their analysis to roost andrandom sites in mixed-conifer forest (n = 55 and36, respectively). Within this forest type, roost

sites contained greater densities and basal areasof Gambel oak and lower densities and basalareas of conifers than random sites. Differencesbetween roost and random sites were generallygreatest with respect to trees from 15-30 cm(5.9-11.8 in) dbh. Roost sites had greater densi-ties of deciduous trees and lower densities ofconiferous tree in this size-class than randomplots within mixed-conifer forest (Peter Stacey,Univ. of Nevada, Reno, NV; pers. comm.).

Tarango et al. (1994) sampled seven roostsites in Chihuahua, Mexico (Sierra MadreOccidental-Norte RU), using 0.04-ha (0.1-ac)circular plots. Roosts were typically located inmulti-layered pine-oak forests on the lowerportions of north-facing slopes. Oaks dominatedmost roost sites by density, comprising 46.6% ofthe trees present, on average (Tarango et al.1994:1).

Foraging SitesForaging SitesForaging SitesForaging SitesForaging Sites

The only available data on foraging sitescome from Ganey (1988) and Ganey and Balda(1994), discussed above. Relative to randomsites, foraging sites within owl home ranges hadgreater total basal areas and more big down logs/ha (Table 4.9). Relative to roosting sites, forag-ing sites had lower canopy closure and fewersnags/ha (Table 4.9). In a discriminant functionanalysis, foraging sites were not as readily distin-guished from random sites as roosting sites and

Table 4.10.Table 4.10.Table 4.10.Table 4.10.Table 4.10. Habitat characteristics at Mexican spotted owl roost (n = 78) and random sites in theTularosa Mountains, New Mexico; Upper Gila Mountains RU. Shown are means and standarddeviations (in parenthesis). Data from Seamans and Gutiérrez (in press).

FPO

24

Volume II/Chapter 4


were also more variable along a discriminantfunction axis (Ganey 1988, fig. 12). Both ofthese results suggest greater variability in forag-ing habitat than in roosting habitat, consistentwith results at the home-range scale (Ganey andBalda 1994). As noted above for roosting sites,foraging sites sampled in this study representedintensively used habitat and probably do notrepresent the full range of conditions used forforaging.

PATTERNS OF HABITAT USEPATTERNS OF HABITAT USEPATTERNS OF HABITAT USEPATTERNS OF HABITAT USEPATTERNS OF HABITAT USE AT A TREE SCALE AT A TREE SCALE AT A TREE SCALE AT A TREE SCALE AT A TREE SCALE

Several studies have examined characteristicsof trees and other microsites, such as cliff ledgesor caves, used by owls for nesting or roosting.This information is primarily descriptive, andwith the exception of SWCA (1992), Ruess(1995), and Seamans and Gutiérrez (in press)provides no basis for comparing used and avail-able trees.

Nest TreesNest TreesNest TreesNest TreesNest Trees

Nest trees were described, with varying levelsof detail, by SWCA (1992), Armstrong et al.(1994), Fletcher and Hollis (1994), Ruess(1995), Seamans and Gutiérrez (in press), andArizona Game and Fish Department (unpub-lished data). All of these studies were conductedon FS lands in Arizona and New Mexico, andsome nests may be represented in �2 studies.

SWCA (1992) sampled 84 nest trees; 81%were conifers and 19% were hardwoods. Fiftypercent of all nests were in Douglas-fir trees,with 20% and 19% occurring in Gambel oakand white fir, respectively (SWCA 1992:17).Eight percent (n = 7) of all nests occurred insnags; five of these (57%) were Gambel oaksnags (SWCA 1992:17).

Nest trees averaged 63.3 cm (24.9 in) dbh,and ranged from 17-127 cm (6.7-50.0 in;SWCA 1992). Nest structures in living oak trees(n = 14, 17%) were located either in a brokentop (n = 3) or a side cavity (n = 11). Nest struc-tures in live conifers included broken top cavities(n = 5), old raptor nests (n = 14), witchesbrooms (n = 25), stick platforms on “bayonet

limbs” (n = 12), stick nests in a multiple-toppedtree (n = 4), and a squirrel nest (n = 1). All snagnests were either broken top (n = 5), or cavity(n = 3), and one snag contained both nest types(SWCA 1992:21).

Nest trees were significantly more likely tohave a deformed crown than randomly-selectedtrees (SWCA 1992:30), although 65% of all nesttrees had a normal crown form. Types of de-formed crown included broken top (19%),multiple top (5%), dead top (10%), and dyingtop (1%). These percentages are based on 81nest trees, with three unaccounted for (SWCA1992: table 9).

Fletcher and Hollis (1994) reported onmicrosite characteristics of 248 nests locatedduring FS inventory and monitoring activitiesthroughout Arizona and New Mexico. It isimpossible to tell how many different pairs ofowls these nests represent, and some of thesesites may also be included in samples discussedelsewhere (SWCA 1992, Ruess 1995, Seamansand Gutiérrez in press). Furthermore, not allnests could be assigned to a particular RecoveryUnit, limiting regional analyses. Therefore, wesimply summarize some general patterns result-ing from this data set. In many cases, the num-bers presented here were recalculated fromsummary data in Fletcher and Hollis (1994). Inthose cases the page or figure number where thedata were found is cited. Sample sizes varyamong values reported, because not all variableswere sampled at each site.

Of the 248 nests sampled, 90.3 and 9.7%were in trees and cliffs, respectively (Fletcher andHollis 1994: fig. 28). Most nests fell within afairly narrow elevational band, with 72% fallingbetween 1982 and 2287 m (6500-7500 ft),85.4% falling between 1982 and 2591 m (6500-8500 ft), and 95.5% falling between 1829 and2591 m (6000-8500 ft), respectively (Fletcherand Hollis 1994: fig. 29). Forty-three percent ofthe cliff nests were found below 1982 m (6500ft).

Almost 50% of 236 nests where aspect wasrecorded were located on north or northeastaspects (Fletcher and Hollis 1994:48). Slopeaveraged 44 ± 40 (SD)%, with 34.5% of allnests found on slopes >40% (Fletcher and Hollis1994:49). Almost half of all nests were located

25

Volume II/Chapter 4


on the lower third of slopes, with the remaindersplit almost evenly between middle and upperslopes (Fletcher and Hollis 1994: fig. 50).Roughly 50% of the nests on upper slopes werecliff nests or nests in Gambel oak (Fletcher andHollis 1994: fig. 50).

Dominant cover types recorded at 237 nestsites were: 80.2% mixed-conifer, 15.2% pine-oak, 2.1% ponderosa pine, 1.7% riparian, and0.8% other (Fletcher and Hollis 1994: fig. 35).Of 224 tree nests, 57% were in Douglas-fir, 16%in Gambel oak, 13% in white fir, 9% in ponde-rosa pine, and 5% in other species (Fletcher andHollis 1994: fig 40).

Nest tree diameter was recorded at 204 nestsites. Trees < 15.2 cm (<6 in) in dbh accountedfor 2% of all nests. Relative frequencies of theother size classes were: 13.7% in trees from 15.2-30.5 cm (6-12 in) dbh, 22.6% in trees from30.5-45.7 cm (12-18 in) dbh, 19.1% in treesfrom 45.7-61 cm (18-24 in) dbh, and 42.7% intrees >61 cm (>24 in) dbh (Fletcher and Hollis1994: fig. 41). Forty-five percent of tree nestswere classified as “witches broom”, with 31.3%in cavities (including broken tops), 14.7% in“debris platforms”, and 8.9% in “other sticknests” (Fletcher and Hollis 1994: fig. 36).

Six of 11 nests (54.5%) sampled by Ruess(1995: fig 7) were in Gambel oak, with theremainder in ponderosa pine. Ruess (1995) didnot report mean diameters for nest trees, butfound four nests (36.4%) in trees ofpresettlement origin (>115 yrs in age) despitethe fact that such trees accounted for only 0.5%of total trees on his study area (Ruess 1995: table3, fig. 7).

Seamans and Gutiérrez (in press) reported78% of 27 nests in Douglas-fir, 11% in white fir,7% in ponderosa pine, and 4% in southwesternwhite pine. With respect to nest structure, 61%were located in dwarf mistletoe infections,10.5% in old squirrel nests, 10.5% in old raptornests, 7% in debris collections, 7% in treecavities, and one nest (4%, n = 28 nests for thesecalculations) was on a cliff. Nest trees averaged60.6 ± 22.4 cm (23.9 ± 17.6 in) in dbh and164 ± 44.8 years in age. Nest trees were signifi-cantly larger and older than randomly sampledtrees within the nest vicinity (Seamans andGutiérrez in press).

Roost TreesRoost TreesRoost TreesRoost TreesRoost Trees

Several researchers have described character-istics of roost trees and a very small area aroundthem. Results are summarized by study area andRU, where possible, in Table 4.11. In general,roost characteristics appeared to be relativelyvariable among study areas (Table 4.11), suggest-ing that greater variability exists among roosttrees than among trees used for nesting. Canopyclosure was more consistent among study areasthan most other characteristics sampled, and wasrelatively high on most study areas. Canopyclosure was <65% on only one study area(Canyonlands, Colorado Plateau RU, Table4.11a). This may reflect the fact that >85% ofthe roost sites sampled on that study area werelocated on cliffs or in pinyon-juniper woodland(Table 4.11a).

Roost tree characteristics appeared to berelatively similar among similar habitat types.For example, mean roost tree diameter was moreconsistent among the mesic mixed-conifer sites(San Francisco Peaks, White Mountains, andSacramento Mountains: mixed-conifer) thanbetween these sites and other areas. Similarly,characteristics were more similar among themore xeric sites (Sacramento Mountains: xericmixed forest and Bar-M watershed) than be-tween these and other areas (Table 4.11). Oneclear pattern that emerges from these studies isthat owls roost in smaller trees, on average, thanthose used for nesting. Ruess (1995: fig. 7),however, noted that three of nine roosts ob-served in ponderosa pine-Gambel oak forestwere >115 yrs old, despite the fact that only0.5% of all trees sampled in the area were ofsuch age (Ruess 1995: Table 3). This suggeststhat old, large trees may also be important forroosting in some areas.

Fletcher and Hollis (1994) reported sum-mary characteristics sampled at 433 roost siteslocated during FS inventory and monitoringefforts in Arizona and New Mexico. This sampleis discussed separately here because roost sitescould not generally be assigned to particularRUs, as was done for the data sets summarized inTable 4.11. This data set is subject to all the

26

Volume II/Chapter 4


Table 4.11a.Table 4.11a.Table 4.11a.Table 4.11a.Table 4.11a. Selected characteristics of roost sites used by radio-tagged Mexcan spotted owls in theColorado Plateau Recovery Unit. Values shown are means (± standard deviation) for continuousvariables, % for categorical variables. Source: D.W. Willey (unpublished data).

FPO

27

Volume II/Chapter 4


Table 4.11b.Table 4.11b.Table 4.11b.Table 4.11b.Table 4.11b. Selected characteristics of roost sites used by radio-tagged Mexican spotted owls in theUpper Gila Mountains Recovery Unit. Values shown are mean (± standard deviation) for continuousvariables, % for categorical variables.

FPO

28

Volume II/Chapter 4


Table 4.11c.Table 4.11c.Table 4.11c.Table 4.11c.Table 4.11c. Selected characteristics of roost sites used by radio-tagged Mexican spotted owls in theBasin and East-Range Recovery Unit. Values shown are mean (± standard deviation) for continuousvariables, % for categorical variables.

FPO

29

Volume II/Chapter 4


limitations discussed relative to nest sites (seeabove) described by Fletcher and Hollis (1994).

Ninety-five percent of all roosts were intrees, with the remainder on cliffs or in caves(Fletcher and Hollis 1994: fig. 42). Percent slopeaveraged 46 ± 34%, with 45% of all roost sitesoccurring on slopes >40% (Fletcher and Hollis1994: fig. 46). Most (57%) of these roost siteswere found on the lower third of slopes (Fletcherand Hollis 1994: fig 47).

Of 367 roost sites where cover type wasrecorded, 65.7% were in mixed-conifer, 27.2%in pine-oak, 2.7% in riparian, 1.4% in ponde-rosa pine, and 3.0% in other cover types(Fletcher and Hollis 1994: fig. 49). A variety oftrees provided roost perches, including Douglas-fir (38.3%), Gambel oak (17.9%), ponderosapine (11.9%), white fir (11.4%), other oakspecies (8.0%), and other species (12.4%;Fletcher and Hollis 1994: fig. 53; n = 402). Trees<15.2 cm (<6 in) dbh accounted for 8.8% of allroosts, with 41.7, 24.1, 11.1, and 14.3% in treesfrom 15.2-30.5, 30.5-45.7, 45.7-61, and >61 cm(6-12, 12-18, 18-24, and >24 in) in dbh, respec-tively (Fletcher and Hollis 1994: fig. 54).

Ganey and Block (unpublished data) evalu-ated seasonal differences in roost site characteris-tics in ponderosa pine-Gambel oak forest (Table4.12). Canopy closure was greater at roosts usedduring the breeding season. Owls used Gambeloak significantly more often during the breedingseason, and tended to roost more often on theupper third of slopes during the breeding seasonand on the middle third during the nonbreedingseason.

Ganey and Block (unpublished data) alsoreported some characteristics of winter roostsused by two owls that migrated to lower eleva-tions (Table 4.13). Both owls roosted primarilyin pinyon-juniper woodland, on the middle andupper portions of slopes. They typically perchedlow in short juniper trees, well hidden by densefoliage and near the center of the tree. Thesewinter roosts differed greatly in structure andspecies composition from typical summerroosting habitat.

WINTERING HABITATWINTERING HABITATWINTERING HABITATWINTERING HABITATWINTERING HABITAT

Present knowledge of wintering habitat ofMexican spotted owls comes primarily fromradiotelemetry studies (see Patterns of HabitatUse at the Home Range Scale) and opportunisticobservations of wintering adults. Radiotelemetrystudies indicate that many owls remain on theirbreeding areas throughout the year, whereassome migrate off of the study area. Wherewintering areas of migrants have been located,they are typically in lower elevation woodland orscrub habitats with more open structure thantypical breeding habitat. However, one owl inthe Colorado Plateau RU migrated upwards inelevation to winter in coniferous forest (Willey1993).

Opportunistic sightings of spotted owlsduring the winter also suggest that part of thepopulation moves to lower elevations. Forexample, owls have been sighted in lower SabinoCanyon, outside of Tucson, Arizona, and on golfcourses in Tucson in recent winters (RussellDuncan, Southwestern Field Biologists, Tucson,AZ, pers. comm.). An adult owl banded on theGila National Forest, Upper Gila MountainsRU, was recovered during winter 1995 nearDeming, New Mexico, Basin and Range-EastRU. This bird had apparently traveled approxi-mately 160 km (100 mi) from the area where itwas located during the breeding season, fromhigh-elevation forest to Chihuahuan desert(Mark Seamans, Humboldt State Univ., Arcata,CA, pers. comm.).

In summary, available evidence on winteringhabitat is limited, but suggests that the bulk ofthe owl population is nonmigratory. Wheremigration does occur, it typically involvesmovement to lower, warmer, and more openhabitats. In some cases, migration involvesmovement between adjacent Recovery Units.Little quantitative data exists to describe typicalwintering habitat for either migrants or year-round residents.

DISPERSAL HABITATDISPERSAL HABITATDISPERSAL HABITATDISPERSAL HABITATDISPERSAL HABITAT

Very little is known about habitat use eitherby adults during migration or by juveniles

30

Volume II/Chapter 4


Table 4.12.Table 4.12.Table 4.12.Table 4.12.Table 4.12. Seasonal roost site characteristics of radio-tagged Mexican spotted owls in ponderosapine-Gambel oak forest, Arizona (Upper Gila Mountains Recovery Unit). Data from Ganey andBlock (unpublished). Values shown are mean (± SD) for continuous variables, % for categorical vari-ables.

FPO

31

Volume II/Chapter 4


during dispersal. Willey (1993) monitored sevendispersing juvenile owls in Utah. These juvenilesapparently moved through a variety of habitattypes, including several that might generally beconsidered too open for use by spotted owls.None of these juveniles survived to reproduce.

A. Hodgson and P. Stacey radio tagged fivejuveniles in the San Mateo Mountains, NewMexico (Upper Gila Mountains RU). Twojuveniles apparently dispersed across opengrassland to the Black Range, but their ultimatefate is not known (Peter Stacey, Univ. of Nevada,Reno, pers. comm.). Three other juvenilesapparently remained in the San Mateo Moun-tains. No information is available on habitatsused by these owls.

ADDITIONAL STUDIESADDITIONAL STUDIESADDITIONAL STUDIESADDITIONAL STUDIESADDITIONAL STUDIES

In addition to the above studies, two addi-tional studies are treated separately here becausethey either were not specific to a single scale

(Johnson 1989, Dames and Moore 1990) or didnot distinguish between site types (Dames andMoore 1990). Johnson (1989) compared habitatcharacteristics sampled at one roost and one nestsite with characteristics sampled at 20 pointswithin the same stand. Thus, this provides acomparison of site characteristics with overallcharacteristics of the surrounding stand. Meantree height, overstory basal area, understory basalarea, overstory density, and snag density were allgreater at the roost and nest sites than in thesurrounding stand. Despite the greater under-story basal area at roost and nest sites, understorydensity was greater within the stand, suggestingthat the understory at the roost and nest sitescontained fewer but larger trees (Johnson1989:12). Johnson (1989:14-15) further notedthat the roost and nest sites ranked high (relativeto the stand) for evenness indices for both treespecies and diameter, and suggested that small-scale diversity may be an important factor in

Table 4.13.Table 4.13.Table 4.13.Table 4.13.Table 4.13. Characteristics of roost sites used by two migrant Mexican spotted owls on their winterrange. Both owls bred in ponderosa pine-Gambel oak forest in the Upper Gila Mountains RecoveryUnit, and wintered in pinyon-juniper woodland on the Basin and Range-West Recovery Unit. Datafrom Ganey and Block (unpublished).

FPO

32

Volume II/Chapter 4


habitat selection by spotted owls (see alsoJohnson and Johnson 1988, Johnson 1990).

Dames and Moore (1990) reported onhabitat characteristics in areas occupied byMexican Spotted Owls in Arizona and NewMexico (Upper Gila Mountains and Basin andRange East-RUs). Sampling methods andsampled area differed between states (Dames andMoore 1990:10), and most of the area sampledwas based on owl presence in the area, ratherthan on specific evidence of use by owls foreither foraging, roosting, or nesting. Results ofhypothesis tests regarding habitat characteristicsin this study were inconclusive. In both states,the most consistent feature within and amongareas sampled was variability (Dames and Moore1990: executive summary).


Habitat RelationshipsHabitat RelationshipsHabitat RelationshipsHabitat RelationshipsHabitat Relationships

Several patterns are evident upon evaluationof current knowledge regarding habitat relation-ships of Mexican spotted owls. The first is thatmost information is limited to relatively finespatial scales. For example, we have considerableinformation about roost and nest trees, and roostand nest sites, but have little information onstands used by owls, habitat composition of owlhome ranges, or landscape configurations usedby spotted owls. Second, most information onowl habitat use relates to the breeding season,and to habitats used for nesting and/or roosting.We know little about habitat use during winteror dispersal periods, or about what constitutesadequate foraging habitat. Third, most informa-tion on owl habitat use and selection comesfrom correlative studies that do not demonstratecause-and-effect relationships. Fourth, ourknowledge of owl habitat-use patterns comesfrom a very short time period (mainly 1984-present).

All of these factors limit our ability to definewhat constitutes spotted owl habitat, or whatdesired future conditions should be for spottedowls. What we can do is describe features ofbreeding-season roosting and nesting habitatused by owls at this time. In most cases, histori-

cal information is not adequate to know whethercurrently-occupied sites were also occupied inthe past, or to allow us to draw conclusionsabout structural features of habitats used in thepast.

Spotted owls typically nest or roost either indeep, rocky canyons, or in any of several forestcover types. The relative use of canyons versusforests varies among regions. For example, owlsin parts of the Colorado Plateau RU are foundexclusively in deep, rocky canyons, whereas owlsin many other RUs are found primarily in forests(although these forests are often in canyons).

Where owls occur in forests, they typicallyselect large, old trees for nesting. Nest and roostsites are found primarily in mixed-conifer forest(Table 4.1), although pine-oak forests are alsoused in some areas, such as parts of the UpperGila Mountains, Basin and Range-West, andSierra Madre Occidental-Norte RUs (Table 4.1;see also Ganey and Balda 1989a, Duncan andTaiz 1992, Ganey et al. 1992, Fletcher andHollis 1994, Tarango et al. 1994, Seamans andGutiérrez in press). Nest and roost sites typicallycontain structurally-complex, uneven-agedforests, with a variety of age- and/or size- classes,a large tree component, many snags and downlogs, and relatively high basal area and canopyclosure (Tables 4.6-4.11; see also SWCA 1992,Armstrong et al. 1994, Ganey and Balda 1994,Ruess 1995, Seamans and Gutiérrez in press).Diversity of tree species and diameters appears tobe high in many owl nesting and roosting areas(Johnson and Johnson 1988, Johnson 1989,Johnson 1990, Seamans and Gutiérrez in press).Many roost and nest sites are found in canyonbottoms or low on canyon slopes (Table 4.11;see also Ganey and Balda 1989a, Fletcher andHollis 1994, Tarango et al. 1994:36, Seamansand Gutiérrez in press). Many have a conspicu-ous broadleaved component, in the form ofriparian trees or especially various oaks (Ganeyand Balda 1989a, Duncan and Taiz 1992, Ganeyet al. 1992, SWCA 1992, Tarango et al. 1994,Ruess 1995, Seamans and Gutiérrez in press).

The reasons why spotted owls nest and roostin structurally-complex, diverse forests and deepcanyons have not been conclusively demon-strated. Barrows (1981) suggested that owls seekdense forest stands as protection from high

33

Volume II/Chapter 4


daytime temperatures. This explanation isattractive with respect to Mexican spotted owls,because the most apparent common denomina-tor between the types of forests and deep, rockycanyons used is that both situations provide coolmicrosites (Kertell 1977, Ganey et al. 1988,Ganey and Balda 1989a, Rinkevich 1991, Willey1993). Further, there is some evidence that,relative to the great horned owl, which is foundin hotter and drier areas, the spotted owl hasdifficulty dissipating metabolic heat at hightemperatures (Ganey et al. 1993).

Carey (1985) and Gutiérrez (1985) alsohypothesized that northern spotted owls mightseek dense old, closed-canopy forests becausesuch areas supported higher prey densities, orbecause the owls were better able to avoidpredators in such areas. Present information,however, suggests that owls forage in a widervariety of forest types than are used for roosting(Ganey and Balda 1994, see also Ward and Block1995). Further, although owls in one studyroosted primarily in unlogged mixed-coniferforests, they did not show a strong pattern ofselection for such forests when foraging (Table4.4). This suggests that the association betweenspotted owls and mixed-conifer forest may bedriven more by roosting and/or nesting behaviorthan by foraging behavior (Ganey and Balda1994). We currently have no information withwhich to test the hypothesis that spotted owlsare better able to avoid predators in complexforests or deep, rocky canyons.

In summary, at present we suspect thatselection of typical nesting and roosting habitatis driven primarily by microclimatic consider-ations. Prey availability may also be an importantconsideration, however, and prey density in andaround an area with microclimatic conditionstypical of nest/roost habitat may determinewhether or not that site is used by owls. Preyavailability may also determine how large an areaowls must use to meet their energetic needs(Carey et al. 1992, Verner et al. 1992, Zabel etal. 1995). However, we suspect that the primaryfactor limiting spotted owl distribution is thepresence on the landscape of habitat suitable forroosting and nesting.

In some areas, the types of forests used forroosting and nesting are primarily restricted to

canyon situations. This is particularly true atlower elevations, where mixed-conifer forest isgenerally found only in canyon bottoms or onnorth-facing canyon slopes. Thus, the distribu-tion of roosting and nesting habitat, and ofspotted owls in these areas, is naturally frag-mented and discontinuous. Opportunities forincreasing the amount of spotted owl habitat insuch areas are limited, and management effortswould be better focused on preserving andenhancing habitat where it exists. Replacementhabitat in such areas may need to develop in situfollowing stand-replacing disturbances.

In other areas, such as high-elevation mixed-conifer forest, it may be possible to developspotted owl habitat over more of the landscape.An example of such an area is the SacramentoMountains (Lincoln National Forest, Basin andRange-East RU). Spotted owls are abundant andwidely distributed in mixed-conifer forests acrossthis range (Skaggs and Raitt 1988, Fletcher andHollis 1994: fig. 24). Much of this area wassubject to relatively intensive railroad loggingearly this century (Glover 1984), but theseforests have recovered quickly and attainedstructural complexity (Table 4.8b), demonstrat-ing that development of replacement habitat ispossible after management under some circum-stances. In areas such as this, managers mightcombine protection of existing habitat withattempts to develop replacement habitat overtime, in a more dynamic approach to habitatmanagement.

Comparisons With Other SubspeciesComparisons With Other SubspeciesComparisons With Other SubspeciesComparisons With Other SubspeciesComparisons With Other Subspeciesof Spotted Owlsof Spotted Owlsof Spotted Owlsof Spotted Owlsof Spotted Owls

There are many similarities in habitat-usepatterns of the three subspecies (northern,California, and Mexican) of spotted owls. Forexample, all three appear to be most common instructurally-complex forest environments,although floristic composition of habitats usedvaries both within and between subspecies’ranges (for other subspecies see reviews inThomas et al. 1990, Gutiérrez et al. 1992). Boththe California (Gutiérrez et al. 1992) and Mexi-can subspecies most commonly nest in mixed-conifer forest, followed by forest types domi-

34

Volume II/Chapter 4


nated by oaks or conifers and oaks. Both theCalifornia (Gutiérrez et al. 1992) and Mexicansubspecies appear to use a wider variety ofhabitat conditions for foraging than for roostingand nesting. These consistencies in habitat usepatterns between subspecies occupying differentgeographic areas and habitat types furtherstrengthen our conclusion that spotted owls areseeking particular types of habitat features fornesting and roosting.

Habitat TrendsHabitat TrendsHabitat TrendsHabitat TrendsHabitat Trends

Because the listing of the owl was basedpartially on projected declines in owl habitat, itis worth discussing what we know about trendsin owl habitat. The following discussion focusesseparately on various habitat conditions used byspotted owls. It is largely restricted to trends inroosting and nesting habitat, because most ofour information relates to such areas, and be-cause such habitat is thought to limit spottedowl distribution. Finally, because little historicalinformation exists with respect to the type ofmicrosite conditions that describe roosting andnesting habitat of spotted owls, this discussion isnecessarily largely qualitative.

Rocky CanyonsRocky CanyonsRocky CanyonsRocky CanyonsRocky Canyons

Mexican spotted owls are found primarily inrocky canyons in parts of their range, such assouthern Utah (Kertell 1977, Rinkevich 1991,Willey 1992, 1993, Utah Mexican spotted owlTechnical Team 1994). In other areas, part of thepopulation also inhabits rocky canyons, butthese are generally more heavily forested thanthe slickrock canyons found in parts of theColorado Plateau RU (Ganey and Balda 1989a,USDA Forest Service 1993, Fletcher and Hollis1994). There is little evidence for change inhabitat quality or loss of habitat in slickrockcanyons. Canyon-bottom vegetation may havebeen degraded by grazing in some cases (seebelow), and some habitat may have been lostbeneath large reservoirs along the ColoradoRiver and its tributaries. The latter change couldhave decreased connectivity among remainingpopulations. Otherwise, we suspect that habitat

trends are relatively stable in areas where owls arefound primarily in slickrock canyons.

Riparian ForestsRiparian ForestsRiparian ForestsRiparian ForestsRiparian Forests

Historically, owls were found in low-eleva-tion riparian forests (Bendire 1892, Phillips et al.1964, and possibly Woodhouse 1853). Theseforests have undergone extensive modificationbecause of recreation, flood control, livestockgrazing, and modification of natural water tables(Knopf et al. 1988; see also Kennedy 1977,Kauffman and Krueger 1984, Minckley andClark 1984, Skovlin 1984, Minckley and Rinne1985, Platts 1990, Schulz and Leininger 1990,Dick-Peddie 1993). Collectively, these activitiesand other factors have changed the speciescomposition and structure of riparian forests. Inextreme cases, riparian forests have disappearedentirely. No breeding spotted owls have beendocumented in lowland riparian forests in recenttimes. Surveys of such habitat have been far fromexhaustive, and owls may still inhabit someremnant riparian forests. Nevertheless, theoverall trend with respect to spotted owls breed-ing in lowland riparian forest habitat has clearlybeen negative.

Spotted owls also commonly occur incanyon-bottom riparian forests at higher eleva-tions, interspersed with other forest types. Inmany cases these forests have also been degraded(Schulz and Leininger 1991, see also Fleischner1994). Management to retain and enhance theseriparian forests would likely be beneficial tospotted owls (as well as many other plants andanimals).

Coniferous ForestsConiferous ForestsConiferous ForestsConiferous ForestsConiferous Forests

Trends in coniferous forests are more diffi-cult to evaluate. It is clear that changes haveoccurred in southwestern forests. These stemprimarily from three sources: disruption ofnatural disturbance regimes, grazing, and timberharvest. Many decades of fire suppression havedisrupted natural disturbance regimes (Cooper1960, Madany and West 1983, Stein 1988,Savage and Swetnam 1990, Covington andMoore 1992, 1994, Harrington and Sackett1992, Johnson 1995). The absence of frequent,

35

Volume II/Chapter 4


low-intensity fire, coupled with widespreadovergrazing by livestock in the late 1800s,reduced competition between herbaceousvegetation and tree seedlings. These effectsproduced a good seedbed for conifer regenera-tion, and generally resulted in increases in treedensities on forested lands (Rummel 1951,Madany and West 1983, Zimmerman andNeuenschwander 1984, Stein 1988, Savage andSwetnam 1990, Covington and Moore 1992,1994, Harrington and Sackett 1992, Johnson1995, Ruess 1995). Such increases in tree densitycan not only alter stand structure, but can alsolead to declines in shade-intolerant tree species,and drive ecological succession from one foresttype to another. For example, some ponderosapine forests appear to be converting to mixed-conifer forest.

Timber harvest in many areas has also alteredstand structure and sometimes species composi-tion. Timber harvest can occur in differentforms, with different intensity, and with differ-ent effects on stand structure. In many cases,however, the net effect of timber harvest is adecrease in old trees and at least a short-termdecrease in tree density and basal area. In theserespects timber harvest tends to work oppositethe effects of disruptions in natural disturbanceregimes described above. Where timber harvesttargets shade-intolerant species, however, itcould further the trend of ecological successiontowards shade-tolerant species.

The net effect of these processes on amountand quality of spotted owl habitat is difficult todetermine. USDI (1993:14251), citing Fletcher(1990), postulated that spotted owl habitat haddecreased in amount due to timber harvest.Fletcher and Hollis (1994:29) estimated that1,068,500 ac of forested habitat in Arizona andNew Mexico had been rendered unsuitable forspotted owls due to human activities, primarilytimber harvest, through 1993. Conversely, Hull(1995:7) reported that “thousands of acres haveconverted from ponderosa pine to mixed-conifer,” and argued that the amount of Mexicanspotted owl habitat had increased. For severalreasons, we submit that either viewpoint isdifficult to substantiate with presently-availabledata.

First, the claim that thousands of acres ofponderosa pine have converted to mixed-coniferis unsubstantiated. Johnson (1995: fig.1) shows alarge increase in acreage of mixed-conifer inArizona and New Mexico between 1966 and1986, based on forest inventories conducted inthose years. Johnson (1995:1) also presentsnumbers suggesting that much of the increase inmixed-conifer forest is due to invasion of mead-ows, rather than conversion of ponderosa pineforest. Further, this information is extrapolatedfrom data presented in the original inventories(Choate 1966, Spencer 1966, Conner et al.1990, Van Hooser et al. 1993), and no informa-tion is provided on how that extrapolation wasdone. Therefore, it is impossible to assess theaccuracy of the figures presented. Methods anddefinitions may also have changed betweeninventories. Referring to comparisons betweenthe 1986 inventory and earlier inventories, VanHooser et al. (1993:1) state: “The changes indefinitions and survey standards make detailedcomparisons with previous inventory resultsunwise.”

Second, the idea that all mixed-conifer forestis spotted owl habitat is not supported by theavailable data. As noted previously, owls roostand nest primarily in structurally-complexforests with particular features, including large,old trees. Recent invasion of meadows by mixed-conifer forest is unlikely to have created this typeof habitat. If that process proceeds, however, itcould create owl habitat in the future.

Finally, most available data on historicalforest structure and increases in forest densityrelate to ponderosa pine forest (USGS 1904,Woolsey 1911, Harrington and Sackett 1984,Covington and Moore 1992, 1994). As we haveshown here, this forest type is not typically usedfor roosting and nesting by spotted owls. Datapresented by Ruess (1995) suggest that similarchanges have occurred in ponderosa pine-Gambel oak forest, which is used by spottedowls. No such data exist for mixed-conifer forest,however, which is the primary type used byspotted owls for roosting and especially fornesting. It seems logical to assume that increasesin density have also occurred within this foresttype. Without quantitative data on changes inthis forest type, however, determining whether

36

Volume II/Chapter 4


or not such changes have been favorable tospotted owls is essentially impossible.

In summary, conflicting speculation existsregarding trends in spotted owl habitat inconiferous forest. Some authors speculate thattimber harvest has reduced the amount of owlhabitat, whereas others speculate that fire sup-pression has increased the amount of owl habi-tat. Data presented here suggest that spotted owlhabitat is complex. The types of historicalevidence available do not allow for any clearanalysis of trends in such habitat. We recognizethat we cannot return to the past to collect suchdata, but we can learn from this situation. Ourinability to evaluate habitat trends stronglyemphasizes the need for accurate and compre-hensive inventory and monitoring of forestresources, so that in the future changes in foresthabitat can be assessed over time.

Research ConsiderationsResearch ConsiderationsResearch ConsiderationsResearch ConsiderationsResearch Considerations

Clearly, much remains to be learned abouthabitat relationships of the Mexican spotted owl.Gutiérrez et al. (1992) outlined a number ofresearch considerations relative to habitat rela-tionships of the California spotted owl. Theseare all relevant to the Mexican subspecies as well.We briefly summarize some additional consider-ations here.

Quantitative data on patterns of owl habitatuse are largely or completely lacking for somegeographic regions, habitat types, and spatialscales. Particularly striking is the lack of quanti-tative data on habitat relationships at the standand landscape scales. Some estimates of habitatconditions measured on small plots, such ascanopy closure, may not be representative ofconditions at the stand scale. Further, manymanagement actions are planned at a stand scale,and implementation of ecosystem managementapproaches will require more attention to habitatcomposition and pattern at the landscape scale.Thus, it seems critical to obtain better informa-tion at these (as well as other) scales. The Team’sefforts to evaluate habitat use patterns at thelandscape scale were frustrated by the lack ofsuitable GIS coverages. Development of suchcoverages would greatly facilitate future analyses.

Further, any credible attempts to monitoramounts of spotted owl habitat or trends in suchhabitat will require both better data on whatconstitutes spotted owl habitat at various spatialscales, and better data on forest structure acrossthe landscape.

Most studies of habitat use-patterns havebeen correlative rather than experimental, whichlimits our ability to draw conclusions abouthabitat selection by Mexican spotted owls.Further, even where correlates of owl occupancyhave been identified, these characteristics havenot been linked to owl fitness. Demographicstudies of Mexican spotted owls are underway ina few areas. Far greater efforts will be required toobtain the data necessary to determine whichhabitats or areas contain self-supporting popula-tions, and to evaluate habitat compositionwithin those areas.

In the meantime, forest management isongoing, and we assume that this will continue.Although controlled experiments in forestmanagement are exceedingly difficult to designand conduct, forest management activities couldprovide a great opportunity to learn more aboutthe response of spotted owls to habitat configu-rations at various scales. Experimental silvicul-tural prescriptions could be developed andapplied, guided by current knowledge of habitatconditions. Such knowledge should be supple-mented by studies of owl habitat in other areasand habitats, and at other scales. McKelvey andWeatherspoon (1992) provide an example of aconceptual approach to integrating silviculturewith knowledge of stand structures used byspotted owls.

Finally, little attention has been paid to theecology and habitat relationships of the owl'sprincipal prey species (but see Ward and Block1995), and to how forest management mightinfluence population levels of these species.Management actions could indirectly affectspotted owls, either positively or negatively,through effects on their prey species. Therefore,these species should also be considered in futuremanagement planning and research activities.

37

Volume II/Chapter 4



Armstrong, J. A., P. C. Christgau, and K.Fawcett. 1994. Habitat characteristics ofMexican spotted owl nest sites on the AlpineRanger District, Apache National Forest,Arizona. Draft Report. Arizona Game andFish Dept., Phoenix. 102pp.

Barrows, C. W. 1981. Roost selection by spottedowls: an adaptation to heat stress. Condor83:302-309.

Bendire, C. E. 1892. Life histories of NorthAmerican birds. U.S. Nat. Mus. Spec. Bull. 1.

Block, W. M., and L. A. Brennan. 1993. Thehabitat concept in ornithology: theory andapplication. Current Ornith. 11:35-91.

Brown, D. E., C. H. Lowe, and C. P. Pase. 1980.A digitized systematic classification for ecosys-tems with an illustrated summary of thenatural vegetation of North America. U.S.For. Serv. Gen. Tech. Rep. RM-73. RockyMtn. For. and Range Exper. Stn. Fort Collins,Colo. 93pp.

Byers, C. R., R. K. Steinhorst, and P. R.Krausmann. 1984. Clarification of a tech-nique for analysis of utilization- availabilitydata. J. Wildl. Manage. 48:1050-1053.

Carey, A. B. 1985. A summary of the scientificbasis for spotted owl management. Pages 100-114 in: R. J. Gutiérrez and A. B. Carey, eds.Ecology and management of the spotted owlin the Pacific Northwest. U.S. For. Serv. Gen.Tech. Rep. PNW-185. Pacific Northwest Res.Stn. Portland, Oreg.

——., S. P. Horton, and B. L. Biswell. 1992.Northern spotted owls: influence of prey baseand landscape character. Ecol. Monogr.62:223-250.

Choate, G. A. 1966. New Mexico’s forest re-source. U.S. For. Serv. Resour. Bull. INT-5.Intermountain Res. Stn. Ogden, Utah.

Conner, R. C., J. D. Born, A. W. Green, and R.A. O’Brien. 1990. Forest resources of Arizona.U.S. For. Serv. Resour. Bull. INT-69. Inter-mountain Res. Stn. Ogden, Utah.

Covington, W. W., and M. M. Moore. 1992.Postsettlement changes in natural fire regimes:implications for restoration of old-growthponderosa pine forests. Pages 81-99 in: M. R.

Kaufmann, W. H. Moir, and R. L. Bassett,eds. Old-growth forests in the Southwest andRocky Mountain regions: proceedings of aworkshop. U.S. For. Serv. Gen. Tech. Rep.RM-213. Rocky Mtn. For. and Range Exper.Stn. Fort Collins, Colo.

——., and M. M. Moore. 1994. Southwesternponderosa forest structure: Changes sinceEuro-American settlement. J. For. 92:39-47.

Cooper, C. F. 1960. Changes in vegetation,structure, and growth of southwestern pineforests since white settlement. Ecol. Monogr.30:129-164.

Dames and Moore, Inc. 1990. Spotted owlstudy. Final report submitted to SouthwestPine Assoc. Dames and Moore, Inc., Tucson,Ariz. 79pp.

Daniel, T. W., J. A. Helms, and F. S. Baker.1979. Principles of silviculture. Second ed.McGraw-Hill, New York, N.Y. 500pp.

Dick-Peddie, W. A. 1993. New Mexico vegeta-tion, past, present and future. Univ. NewMexico Press, Albuquerque, N.M. 244pp.

Duncan, R. B., and J. D. Taiz. 1992. A prelimi-nary understanding of Mexican spotted owlhabitat and distribution in the ChiricahuaMountains and associated sub-Mogollonmountain ranges in southeastern Arizona.Pages 58-61 in: A. M. Barton and S. A.Sloane, eds. Proc. Chiricahua Mtns. Res.Symp. Southwest Parks and MonumentsAssoc., Tucson, Ariz.

Fleischner, T. L. 1994. Ecological costs oflivestock grazing in North America. Conserva-tion Biology 8:629-644.

Fletcher, K. W. 1990. Habitats used, abundance,and distribution of the Mexican spotted owl(Strix occidentalis lucida) on National ForestSystem lands in the Southwestern Region.U.S. For. Serv., Southwestern Region. Albu-querque, N.M. 56pp.

——., and H. E. Hollis. 1994. Habitats used,abundance, and distribution of the Mexicanspotted owl (Strix occidentalis lucida) onNational Forest System lands in the South-western Region. U.S. For. Serv., SouthwesternRegion. Albuquerque, N.M. 86pp.

Ganey, J. L. 1988. Distribution and habitatecology of Mexican spotted owls in Arizona.

38

Volume II/Chapter 4


M.S. Thesis, Northern Arizona Univ. Flag-staff, Ariz. 229pp.

——., and R. P. Balda. 1989a. Distribution andhabitat use of Mexican spotted owls in Ari-zona. Condor 91:355-361.

——., and R. P. Balda. 1989b. Home rangecharacteristics of spotted owls in northernArizona. J. Wildl. Manage. 53:1159-1165.

——., and R. P. Balda. 1994. Habitat selectionby Mexican spotted owls in northern Arizona.Auk 111:162-169.

——., J. A. Johnson, R. P. Balda, and R. W.Skaggs. 1988. Status report: Mexican spottedowl. Pages 145-150 in R. L. Glinski et al., eds.Proc. Southwest Raptor Manage. Symp. andWorkshop. Natl. Wildl. Feder. Sci. and Tech.Ser. 11.

——., R. B. Duncan, and W. M. Block. 1992.Use of oak and associated woodlands byMexican spotted owls in Arizona. Pages 125-128 in: P. F. Ffolliott, G. J. Gottfried, D. A.Bennett, V. M. Hernandez, C., A. Ortega-Rubio, and R. H. Hamre, eds. Ecology andmanagement of oak and associated wood-lands: perspectives in the southwestern UnitedStates and northern Mexico. U.S. For. Serv.Gen. Tech. Rep. RM-218. Rocky Mtn. For.and Range Exper. Stn. Fort Collins, Colo.

——., R. P. Balda, and R. M. King. 1993.Metabolic rate and evaporative water loss ofMexican spotted and great horned owls.Wilson Bull. 105:645-656.

Glover, V. J. 1984. Logging railroads of theLincoln National Forest, New Mexico. U.S.For. Serv., Southwestern Region. Albuquer-que, N.M.

Gutiérrez, R. J. 1985. An overview of recentresearch on the spotted owl. Pp. 39-49 in: R.J. Gutiérrez and A. B. Carey, eds. Ecology andmanagement of the spotted owl in the PacificNorthwest. U.S. For. Serv. Gen. Tech. Rep.PNW-185. Pacific Northwest Res. Stn.Portland, Oreg.

——., J. Verner, K. S. McKelvey, B. R. Noon,G. N. Steger, D. R. Call, W. S. LaHaye, B. B.Bingham, and J. S. Senser. 1992. Habitatrelations of the California spotted owl. Pages79-98 in: J. Verner, K. S. McKelvey, B. R.Noon, R. J. Gutiérrez, G. I. Gould, Jr., and T.W. Beck, eds. The California spotted owl: a

technical assessment of its current status. U.S.For. Serv. Gen. Tech. Rep. PSW-133. PacificSouthwest Res. Stn. Albany, Calif.

Harrington, M. G., and S. S. Sackett. 1992. Pastand present fire influences on southwesternponderosa pine old-growth. Pp. 44-50 in:Kauffmann, M. R., W. H. Moir, and R. L.Bassett, tech. coordinators. Old-growth forestsin the southwest and Rocky Mountain re-gions. U. S. For. Serv. Gen. Tech. Rep. RM-213. Rocky Mtn. For. and Range Exper. Stn.Fort Collins, Colo.

Hull, T. 1995. Comments submitted to theRegional Director, U.S. Fish and Wildl. Serv.regarding the draft recovery plan for theMexican spotted owl, by Applied EcosystemManagement, Inc. Flagstaff, Ariz.

Hurlbert, S. H. 1984. Pseudoreplication and thedesign of ecological field experiments. Ecol.Monogr. 54:187-211.

Johnson, J. A., and T. H. Johnson. 1985. Thestatus of the spotted owl in northern NewMexico. New Mexico Dep. Game and Fish,Santa Fe. 39pp.

——., and T. H. Johnson. 1988. Timber typemodel of spotted owl habitat in northern NewMexico. New Mexico Dep. Game and Fish,Santa Fe. 23pp.

——. 1989. Jemez Mountains spotted owlsurvey and habitat evaluation. Santa Fe Natl.For. Santa Fe, N.M. 19pp.

Johnson, T. H. 1990. Revised timber type modelof spotted owl habitat in northern NewMexico. Santa Fe National Forest. Santa Fe,N.M. 10pp.

Johnson, M. A. 1995. Changes in Southwesternforests: stewardship implications. U.S. For.Serv., Southwestern Region. Albuquerque,N.M. 6pp.

Kauffman, J. B., and W. C. Krueger. 1984.Livestock impacts on riparian ecosystems andstreamside management implications, areview. J. Range Manage. 37:430-438.

Keitt, T. H., A. B. Franklin, and D. L. Urban.1995. Landscape analysis and metapopulationstructure. Chapter 3 (16 pp.) in: USDI 1995.Mexican spotted owl recovery plan. U. S. Fishand Wildl. Serv. Albuquerque, NM.

Kennedy, C. E. 1977. Wildlife conflicts inriparian management: water. Pages 52-58 in:

39

Volume II/Chapter 4


R. R. Johnson and D. A. Jones, eds. Impor-tance, preservation and management ofriparian habitat: a symposium. U.S. For. Serv.Gen. Tech. Rep. RM-43. Rocky Mtn. For.and Range Exper. Stn. Fort Collins, Colo.


Knopf, F. L., R. R. Johnson, T. Rich, F. B.Samson, and R. C. Szaro. 1988. Conservationof riparian ecosystems in the United States.Wilson Bull. 100:272-284.

Kroel, K. W. 1991. Home range and habitat usecharacteristics of the Mexican spotted owls(sic) in the southern Sacramento Mountains,New Mexico. M.S. Thesis. New Mexico StateUniv., Las Cruces. 63pp.

Ligon, J. S. 1926. Habits of the spotted owl(Syrnium occidentale). Auk 43:421-429.

Madany, M. H., and N. E. West. 1983. Live-stock grazing-fire regime interactions withinmontane forests of Zion National Park, Utah.Ecology 64:661-667.

McDonald, C. B., J. Anderson, J. C. Lewis, R.Mesta, A. Ratzlaff, T. J. Tibbitts, and S. O.Williams. 1991. Mexican spotted owl (Strixoccidentalis lucida) status report. U.S. Fish andWildl. Serv. Albuquerque, N.M. 85pp.

McKelvey, K. S., and C. P. Weatherspoon. 1992.Projected trends in owl habitat. Pages 261-275 in: J. Verner, K. S. McKelvey, B. R.Noon, R. J. Gutiérrez, G. I. Gould, Jr., and T.W. Beck, eds. The California spotted owl: atechnical assessment of its current status. U.S.For. Serv. Gen. Tech. Rep. PSW-133. PacificSouthwest Res. Stn. Albany, Calif.

Minckley, W. L., and T. O. Clark. 1984. Forma-tion and destruction of a Gila River mesquitebosque community. Desert Plants 6:23-30.

——., and J. N. Rinne. 1985. Large woodydebris in hot-desert streams: an historicalreview. Desert Plants 7:142- 153.

Mohr, C. O. 1947. Table of equivalent popula-tions of North American small mammals.Amer. Midl. Nat. 37:223-249.

Neu, C. W., C. R. Byers, and J. M. Peek. 1974.A technique for analysis of utilization-avail-ability data. J. Wildl. Manage. 38:541-545.

O’Neill, R. V., B. T. Milne, M. G. Turner, andR. H. Gardner. 1988. Resource utilization

scales and landscape pattern. Landscape Ecol.2:63-69.

Orians, G. H., and J. F. Wittenberger. 1991.Spatial and temporal scales in habitat selec-tion. Am. Nat. 137(Supplement):s29-s49.

Phillips, A. R., J. T. Marshall, Jr., and G.Monson. 1964. The birds of Arizona. Univ. ofArizona Press, Tucson. 212pp.

Platts, W. S. 1990. Managing fisheries andwildlife on rangelands grazed by livestock: aguidance and reference document for biolo-gists. Nevada Dep. of Wildl. 448pp.

Porter, W. F., and K. E. Church. 1987. Effects ofenvironmental pattern on habitat preferenceanalysis. J. Wildl. Manage. 51:681-685.

Reynolds, R. T. 1993. Ecology and dispersal ofthe Mexican spotted owl in Colorado, FY1993-94. Research Proposal. U.S. For. Serv.Rocky Mtn. For. and Range Exper. Stn. FortCollins, Colo.


——., and R. J. Gutiérrez. In review. Mexicanspotted owl habitat characteristics in ZionNational Park. J. Raptor Res.

Ruess, B. J. 1995. Changes in Mexican spottedowl habitat within ponderosa pine/Gambeloak communities since Euro-Americansettlement. M.S. Thesis. Northern ArizonaUniversity, Flagstaff. 42 pp.

Rummel, R. S. 1951. Some effects of livestockgrazing on ponderosa pine forest and range incentral Washington. Ecology 32:594-607.

Savage, M., and T. W. Swetnam. 1990. Early19th century fire decline following sheeppasturing in a Navajo ponderosa pine forest.Ecology 32:594-607.

Schulz, T. T., and W. C. Leininger. 1990. Differ-ences in riparian vegetation structure betweengrazed areas and exclosures. J. Range Manage.43:295-299.

——., and W. C. Leininger. 1991. Nongamewildlife communities in grazed and ungrazedmontane riparian sites. Great Basin Naturalist51:286-292.

Seamans, M. E., and R. J. Gutiérrez. In press.Breeding habitat ecology of the Mexican

40

Volume II/Chapter 4


spotted owl in the Tularosa Mountains, NewMexico. Condor.

Skaggs, R. W. 1988. Status of the spotted owl insouthern New Mexico: 1900-1987. NewMexico Dep. Game and Fish, Santa Fe. 13pp.

——., and R. J. Raitt. 1988. A spotted owlinventory of the Lincoln National Forest,Sacramento Division, 1988. New MexicoDep. Game and Fish, Santa Fe. 12pp.

Skovlin, J. M. 1984. Grazing impacts on wet-lands and riparian habitat: a review of ourknowledge. Pages 1001-1004 in NationalResearch Council and National Academy ofSciences. Developing strategies for rangelandmanagement: a report prepared by the com-mittee on developing strategies for rangelandmanagement. Westview Press, Boulder, Colo.

Spencer, J. S., Jr. 1966. Arizona’s forests. U. S.For. Serv. Resourc. Bull. INT-6. Intermoun-tain Res. Stn. Ogden, Utah.

Stein, S. J. 1988. Explanations of theimbalanced age structure and scattered distri-bution of ponderosa pine within a high-elevation mixed coniferous forest. For. Ecol.and Manage. 25:139-153.

SWCA, Inc. 1992. Nest tree and nest treehabitat characteristics of the Mexican spottedowl in Arizona and New Mexico. Final Rep.submitted to U.S. For. Serv., SouthwesternRegion. Albuquerque, N.M. 31pp.

Tarango, L. A., R. Valdez, and P. J. Zwank.1994. Mexican spotted owl distribution andhabitat characterizations in southwesternChihuahua, Mexico. Final report, ContractNo. 80-516.6-66. New Mexico Dep. Gameand Fish, Santa Fe. 69pp.


Turner, M. G. 1989. Landscape ecology: theeffect of pattern on process. Annu. Rev. Ecol.Syst. 20:171-197.

Urban, D. L., R. V. O’Neill, and H. H. Shugart,Jr. 1987. Landscape ecology: a hierarchicalperspective can help scientists understandspatial patterns. BioScience 37:119-127.

USDA Forest Service. 1993. Facts about theMexican spotted owl. U. S. For. Serv. South-western Region, Albuquerque, N.M.

USDI Fish and Wildlife Service. 1993. Endan-gered and threatened wildlife and plants; finalrule to list the Mexican spotted owl as athreatened species. Federal Register 58:14248-14271.

——. 1995. Mexican spotted owl RecoveryPlan. Albuquerque, N.M.

USGS. 1904. Forest conditions in the SanFrancisco Mountains Reserve. U. S. Geologi-cal Survey Prof. Paper 22. Washington, D.C.

Utah Mexican spotted owl technical team. 1994.Suggestions for management of Mexicanspotted owls in Utah. Utah Div. Wildl.Resources. Salt Lake City, UT. 103pp.

Van Hooser, D. D., R. A. O’Brien, and D. C.Collins. 1993. New Mexico’s forest resources.U.S. For. Serv. Resour. Bull. INT-79.Intermtn. Res. Stn., Ogden, Utah. 80pp.

Verner, J., R. J. Gutiérrez, and G. I. Gould, Jr.1992. The California spotted owl: generalbiology and ecological relations. Pp. 55-78 in:Verner, J., K. S. McKelvey, B. R. Noon, R. J.Gutiérrez, G. I. Gould, Jr., and T. W. Beck,eds. The California spotted owl: a technicalassessment of its current status. U.S. For. Serv.Gen. Tech. Rep. PSW-133. Pacific SouthwestRes. Stn. Albany, Calif.

Ward, J. P., Jr., and W. M. Block. 1995. Mexicanspotted owl prey ecology. Chapter 5 (48 pp.)in: USDI Fish and Wildlife Service. 1995.Mexican spotted owl recovery plan.Albuquerque, N.M.

Willey, D. W. 1992. Movements and habitatecology of Mexican spotted owls in southernUtah. Final report. Utah Div. Wildl. Re-sources, Salt Lake City. 23pp.

——. 1993. Home-range characteristics andjuvenile dispersal ecology of Mexican spottedOwls in southern Utah. Rep. submitted toUtah Div. Wildl. Resources, Salt Lake City.44pp.

Woodhouse, S. W. 1853. Report on the naturalhistory of the country passed over by theexploring expedition under the command ofBrev. Capt. L. Sitgreaves, U.S. TopographicalEngineers, during the year 1851. [and]Zoology: Birds. Pages 33-40, 58-105 in Capt.L. Sitgreaves, ed. Report of an expeditiondown the Zuni and Colorado Rivers. RobertArmstrong, Public Printer, Washington, D.C.

41

Volume II/Chapter 4


Woolsey, T. S., Jr. 1911. Western yellow pine inArizona and New Mexico. U. S. For. Serv.Bull. 101. Washington, D.C.

Worton, B. J. 1989. Kernel methods for estimat-ing the utilization distribution in home-rangestudies. Ecology 70:164-168.

Zabel, C., K. McKelvey, and J. P. Ward, Jr. 1995.Influence of primary prey on home-range sizeand habitat-use patterns of Spotted Owls(Strix occidentalis). Can. J. Zool. 73:433-439.

Zhou, B. 1994. Quantifying Mexican spottedowl habitat with the Johnson system ofdistributions. M.S. Thesis. Northern ArizonaUniv., Flagstaff. 169pp.

Zimmerman, G. T., and L. F. Neuenschwander.1984. Livestock grazing influences on com-munity structure, fire intensity, and firefrequency within the Douglas-fir/ninebarkhabitat type. J. Range Manage. 37:104-110.

Zwank, P. J., K. W. Kroel, D. M. Levin, G. M.Southward, and R. C. Rommé. 1994. Habitatcharacteristics of Mexican spotted owls insouthwestern New Mexico. J. Field Ornithol.65: 324-334.

42

Volume II/Chapter 5


CHAPTER 5: Mexican Spotted Owl Prey EcologyCHAPTER 5: Mexican Spotted Owl Prey EcologyCHAPTER 5: Mexican Spotted Owl Prey EcologyCHAPTER 5: Mexican Spotted Owl Prey EcologyCHAPTER 5: Mexican Spotted Owl Prey EcologyJames P. Ward, Jr., and William M. Block

In addition to shelter, water, and otherrequirements, habitat must also provide spottedowls with food. Certain trees within specificforest communities may meet nesting, roosting,and perching needs, but the tree componentalone may not necessarily sustain the animalspecies upon which the owls prey. Conservingappropriate habitat for the owl includes conserv-ing habitat for a suite of prey species.

The distribution and abundance of preyoften influence the distribution, abundance, andreproduction of raptors (Newton 1979). Inowls, reproductive success is often correlatedwith prey abundance (Craighead and Craighead1956, Southern 1970, Lundberg 1976,Wendland 1984, Korpimäki and Norrdahl1991). The postulated mechanism behind thisrelationship is energetically based. Male owlsmust provide enough food to their female matesduring incubation and brooding to preventabandonment of nests or young (Johnsgard1988). Accordingly, ecologists suspect thatspotted owls select habitats partially because ofthe availability of prey (Carey 1985, Thomas etal. 1990, Verner et al. 1992). Understanding thenatural history of the spotted owl’s primary preyis vital information for the Recovery Plan be-cause it provides resource planners and managerswith another tool for evaluating an area’s abilityto support spotted owls.

This section summarizes information aboutspotted owl-prey relationships and the ecology ofthe owl’s prey. Specifically, our objectives are to:(1) describe the diet of the Mexican spotted owl;(2) identify prey that may influence owl fitness;and (3) quantify habitat correlates of the owl’sprimary prey.

Ideally, relationships among the Mexicanspotted owl, its prey, and the prey’s habitatshould be examined across different spatial andtemporal scales. In reality, the available informa-tion permits only a limited view of the owl’s preyecology. For example, we could describe abun-dance and distribution of the owl’s commonprey among different vegetation communities,but could not provide a direct link between owl

habitat use and prey availability. Although thelatter information is preferred for prioritizinghabitat conservation, our approach relies uponconserving a general mixture of habitats for theowl and its prey throughout major portions ofthe owl’s range. In time, information from morespecific studies should be used to refine thegeneral findings presented here.

METHODS FORMETHODS FORMETHODS FORMETHODS FORMETHODS FORDETERMINING OWL DIETSDETERMINING OWL DIETSDETERMINING OWL DIETSDETERMINING OWL DIETSDETERMINING OWL DIETS

The dietary habits of raptors can be deter-mined both directly and indirectly. Observationsof prey capture and of prey taken to roosts ornests provide direct evidence of a raptor’s diet.However, such observations are difficult toobtain from nocturnal foragers like owls andoffer little opportunity for quantitative analysis.Regurgitated pellets of undigested materials (fur,feathers, bones, chitinous exoskeletons) offer anindirect, alternative method for analyzing thefeeding habits of owls (Errington 1930, Gladinget al. 1943, Marti 1987).

Spotted owl prey are identified by examiningthe contents of pellets collected below roosts andnests. Prey remains are identified to species,genus, or a less specific prey category and tallied.Diets are then quantified using two measures,relative frequency and percent biomass (Forsmanet al. 1984, Marti 1987). Relative frequency ofprey is expressed as the number of individualitems of a given prey species or group divided bythe total number of all individual items found ina sample. Biomass (g) is the total number ofitems of a given prey species or group multipliedby the average mass (g) for that species or group.Commonly, both relative frequency and biomassare expressed as percentages. Both measures canbe used to compare owl diets among samplingunits such as reproductive groups, locations,seasons, and so on. The two measures providedifferent information. Measures of relativefrequency indicate the proportion of each preytype in the owl’s diet by number, whereas,

1

percent biomass indicates the proportion of eachprey type by weight.

Pellets may provide biased measures of owldiets. Pellets are typically collected opportunisti-cally below roosting owls or from known roostgroves, not randomly or systematically. The biaspotentially resulting from nonrandom samplingcould not be evaluated. We assumed that itemsfound in pellets reflected the true proportions ofprey species in owl diets. The methods we usedto quantify Mexican spotted owl diet are compa-rable to other studies conducted on the northernand California subspecies (see reviews by Tho-mas et al. 1990 and Verner et al. 1992). Accord-ingly, we restricted analyses and inferences torelative comparisons.

We compiled and analyzed prey remains ofMexican spotted owls as reported in 13 studiesconducted since 1977 (Tables 5.1-5.6). A totalof 11,164 prey items was examined. These dataconsisted of 25 data sets from 18 geographicareas throughout the owl’s range, and includedmost published and unpublished information ofMexican spotted owl diet through 1993 (Tables5.1-5.6). Kertell (1977) was not used because ofthe small number of items reported.

Each data set differed in the number of owls,years, and pellets examined. In some cases, thenumber of owls studied or pellets collected werenot recorded. Thus, we could not use owls as thesampling unit or separate differences among owlterritories in region-wide analyses. However,each data set contained a known number ofidentified prey items. For analysis, we treatedeach prey item as an observation and each dataset as a dietary sample. Following this logic, thenumber of observations in a sample (i.e., samplesize) corresponds to the total number of preyitems identified in each data set. We justified theuse of prey items as an observational unit insteadof pellets because a pellet is difficult to define(i.e., broken pellets are often collected or mul-tiple pellets are stored together and break apartduring storage) and a single pellet may containmultiple prey items. Being an uncertain andinconsistent measure of sample size, no pellettotal is given here.

The methods used to identify remains andtally prey numbers followed Forsman et al.(1984). Samples abbreviated as CAPRF2,

ZION2, CANYL, BAR-M, MOGAZ2,COCODM, GILADM, and SACMT2 (seeTables 5.1-5.6 for acronym definition) wereanalyzed using standardized procedures de-scribed and implemented by DeRosier and Ward(1994) to facilitate comparison. According toboth sources, remains were keyed to specieswhen possible using skulls and appendicularskeletal parts. Specialists were consulted toidentify less common or unusual remains,particularly bats, birds, reptiles, and inverte-brates.

Deviations from our standard identificationprocedures were necessary for the SCCOLsample where invertebrate parts were not identi-fied (Charles Johnson, Rocky Mountain Re-search Station, Fort Collins, CO, pers. comm.),and in the ZION1 sample where appendicularparts were not used to tally prey (SarahRinkevich, FWS, Albuquerque, NM, pers.comm.).

Owl diets were quantified using relativefrequency and percent biomass for 11 preygroups: woodrats, white-footed (peromyscid)mice, voles, pocket gophers, rabbits, bats, otheror unidentified small mammals (mostly murids),other or unidentified medium mammals (mostlysciurids), birds, reptiles, and arthropods (Tables5.1-5.6). We use the term peromyscid mice herein place of white-footed mice to represent theapproximate 15 North American species of thegenus Peromyscus. Confusion sometimes followsdiscussion of white-footed mice because this isalso the common name for P. leucopus.

Several sources were used to estimate preymass (Appendix 5a). Biomass estimates given bythe original authors were retained unless betterestimates were available. Estimates of prey massfrom areas nearest to where pellets were collectedwere used whenever possible. In the absence ofbetter data, general references for mass wereused. Averages, weighted according to propor-tions of species in owl diets, were used to esti-mate biomass for less specific taxa such as“woodrat species” or “unidentified bat.”



3

Table 5.1. Table 5.1. Table 5.1. Table 5.1. Table 5.1. Relative frequency of prey items found in the diet of Mexican spotted owls occurring in the northern portion of the subspecies' range. Valueswere calculated from totals pooled across owl territories and years.

4

Table 5.2.Table 5.2.Table 5.2.Table 5.2.Table 5.2. Relative frequency of prey items found in the diet of Mexican spotted owls occurring in the central portion of the subspecies' range. Valueswere calculated from totals across owl territories and years.

5

Table 5.3.Table 5.3.Table 5.3.Table 5.3.Table 5.3. Relative frequency of prey items found in the diet of Mexican spotted owls occurring in the southern portion of the subspecies' range. Valueswere calculated from totals pooled across owl territories and years.

6

Table 5.4.Table 5.4.Table 5.4.Table 5.4.Table 5.4. Percent of prey biomass in the diet of Mexican spotted owls occurring in the northern portion of the subspecies' range. Values were calculatedfrom totals pooled across owl territories and years.

7

Table 5.5.Table 5.5.Table 5.5.Table 5.5.Table 5.5. Percent of prey biomass in the diet of Mexican spotted owls occurring in the central portion of the subspecies' range. Values were calculatedfrom totals pooled across owl territories and years.

8

Table 5.6.Table 5.6.Table 5.6.Table 5.6.Table 5.6. Percent of prey biomass in the diet of Mexican spotted owls occurring in the southern portion of the subspecies' range. Values were calculatedfrom totals pooled across owl territories and years.

GENERAL FOOD HABITSGENERAL FOOD HABITSGENERAL FOOD HABITSGENERAL FOOD HABITSGENERAL FOOD HABITS

The Mexican spotted owl eats a variety ofanimals throughout its range but usually takessmall and medium-sized mammals (Table 5.7).As a group, mammals are taken more frequently(x = 82.1%; cv = 13.3%; n = 25 data sets) thanbirds (x = 4.8%; cv = 78.4%; n = 25), reptiles(x = 0.5%; cv = 192.5%; n = 25) or arthropods(x = 12.6%; cv = 79.0%, n = 23). Arthropodscould be present in the owl’s diet because theseitems were eaten by the owl’s prey prior tocapture. Mammal use is even greater whenbiomass is considered (x = 95.8%, x = 3.9% forbirds, x = 0.1% for reptiles, and x = 0.2% forarthropods). A cumulative plot of diet frequen-cies (Figure 5.1a) indicates that 90% of an“average” Mexican spotted owl diet wouldcontain 30% woodrats; 28% peromyscid mice;13% arthropods; 9% microtine voles; 5% birds;and 4% medium-sized rodents, mostly diurnalsciurids. A cumulative plot of diet biomass(Figure 5.1b) indicates that, on average, 90% ofa Mexican spotted owl diet is comprised of 53%woodrats; 13% rabbits; 9% peromyscid mice;9% birds; and 6% medium-sized mammals suchas diurnal sciurids. These rangewide patterns,however, are not consistent among RUs.

To evaluate geographic variation of Mexicanspotted owl food habits, we examined diet databy recovery unit. Frequencies of prey weretreated as binomial, either the prey species beingcompared was consumed during a successfulforaging event or some other species was con-sumed. We transformed these data for paramet-ric analyses using an arcsine-squareroot transfor-mation (Freeman and Tukey 1950) and testedthe hypothesis of no difference in diets amongRUs using a one-way analysis of variance(ANOVA). A Tukey’s test was used to conductmultiple comparisons among RUs when theANOVA was significant. We conducted the testsfor each of the 11 prey groups separately. We didnot conduct an ANOVA on percent biomassestimates because these data were comprised ofboth frequency and prey mass, and our estimatesof mass for prey were not significantly differentamong RUs except for voles (F = 4.4, d.f. = 6,15, P = 0.009). Thus, prey mass would have

been roughly equivalent among RUs and com-parisons of percent biomass would have beenequivalent to those of relative frequency.

Our analyses indicated significant differencesin Mexican spotted owl diets among geographiclocation (Figure 5.2). Woodrats were taken moreoften in the Colorado Plateau compared to theUpper Gila Mountains and Sierra MadreOccidentalis - Norte (F = 4.6, d.f. = 6, 18,P = 0.005; Figure 5.2a). Voles were consumedmore frequently by owls in the Basin andRange - East, Southern Rocky Mountains -Colorado, and Upper Gila Mountains RUs thanin the Colorado Plateau and Basin andRange - West RUs (F = 8.3, d.f. = 6, 18,P < 0.001; Figure 5.2a). Pocket gophers wereeaten more often in the Upper Gila Mountainscompared to the Colorado Plateau (F = 4.5,d.f. = 6, 18, P = 0.006; Figure 5.2c); and birdswere consumed more often by owls in theSouthern Rocky Mountains - New Mexico,Basin and Range - West, and Upper Gila Moun-tains RUs than in the Colorado Plateau (F = 5.8,d.f. = 6, 18, P = 0.002; Figure 5.2d). Bats andreptiles were consumed more frequently in theBasin and Range - West RU compared to theUpper Gila Mountains RU (F = 2.9, d.f. = 6, 18,P = 0.038; Figure 5.2e and F = 3.3, d.f. = 6, 18,P = 0.024; Figure 5.2f, respectively). Of the fivegroups that did not differ significantly amongRUs (peromyscid mice, rabbits, other smallmammals, other medium mammals, andarthropods), only peromyscid mice (x = 27.2%,cv = 42% among units) and arthropods (x =13.8%, cv = 70%) were frequent food items.

Interregional trends in prey consumptioncould be attributed to other factors like temporalvariation in the owl’s diet within each study area,or because the breeding status of owls differedamong studies. Diet frequencies from each studywere pooled across years and owl pairs regardlessof reproductive status because this informationwas unavailable in most cases. However, the yearin which diet samples were collected and thereproductive success of a sufficient number ofowls were known for three studies (SACMT2,GILADM, COCODM). These three data setswere collected from two of the RUs, Basin andRange - East (SACMT2; Tables 5.3 and 5.6) andUpper Gila Mountains (GILADM, COCODM;



Volume II/Chapter 5


Table 5.7.Table 5.7.Table 5.7.Table 5.7.Table 5.7. Animal species consumed by Mexican spotted owls as determined from examination of25 data sets collected from 18 geographic areas.

FPO

10

Volume II/Chapter 5


Figure 5.1.Figure 5.1.Figure 5.1.Figure 5.1.Figure 5.1. Cumulative distributions of prey in the diet of Mexican spotted owls presented as (a)(a)(a)(a)(a)relative frequency (%) of items and (b)(b)(b)(b)(b) percent biomass. Values are averages across 25 data sets con-ducted throughout the owl's range and dashed lines are 95% confidence limits. Prey groups areWRAT-woodrats; MICE-peromyscid mice; ARTH-arthropods; VOLE-voles; BIRD-birds; OMMM-other medium-sized mammals; BATS-bats; OSMM-other small-sized mammals; RABB-rabbits;GOPH-gophers; REPT-reptiles.

11

Volume II/Chapter 5


Figure 5.2.Figure 5.2.Figure 5.2.Figure 5.2.Figure 5.2. Geographic variability in the food habits of Mexican spotted owls presented as relativefrequencies of (a)(a)(a)(a)(a) woodrats, (b)(b)(b)(b)(b) voles, (c)(c)(c)(c)(c) gophers, (d)(d)(d)(d)(d) birds, (e)(e)(e)(e)(e) bats, and (f)(f)(f)(f)(f) reptiles. Point values arefrom single studies or averages among the number of data sets shown in parenthesis. Vertical bars are95% confidence intervals showing sampling and inter-data set variation within a recovery unit. Recov-ery Unit acronyms are COPLAT-Colorado Plateau; SRM-CO-Southern Rocky Mountains - Colo-rado; SRM-NM-Southern Rocky Mountains - New Mexico; UPGIL-Upper Gila Mountains; BAR-W-Basin and Range - West; BAR-E-Basin and Range - East; and SMO-N-Sierra Madre Occidental -Norte.

12

Volume II/Chapter 5


Tables 5.2 and 5.5). Owl reproductive successwas determined similarly in all three studiesusing methods described by Forsman (1983). Weused data from these studies to evaluate theinfluence of geography, time, and owl reproduc-tive status on the interregional trends in the owl’sdiet.

To quantify the effect of these three factors,we analyzed the differences in relative frequencyof a given prey group in the owl’s diet amongstudy area (SACMT2, COCODM, orGILADM; geographic variation), year (1991-1993; temporal variation), and number of owlyoung produced (0, 1, 2, or 3; variation in owlreproductive status) using a three-factorANOVA. Prey groups that were common toowls in any of the three study areas were ana-lyzed separately (woodrats, peromyscid mice,voles, pocket gophers, rabbits, other medium-sized mammals, and arthropods). We definecommon prey arbitrarily as those species contrib-uting >10% of diet frequency or biomass, whichincludes the majority of species identified in the90% cumulative averages (Figure 5.1). Statisticalsignificance of each factor was used to quantifythe importance of these factors in determiningregional diet trends.

In 3 of 7 tests, consumption of commonprey varied primarily by geographic location(Table 5.8). In the other 4 tests, consumption ofwoodrats, peromyscid mice, medium-sizedmammals, and arthropods was influenced by aninteraction between geographic location andyear. However, significance values indicated thatthe consumption of woodrats and arthropodswas influenced more by the owls’ location thanthe particular year (Table 5.8). Prevalence ofgeographic variation in the three-factor ANOVAresults supports our claims that the owl’s feedinghabitats differ among regions, as discussed above(Figure 5.2).

Diet differences of the Mexican spotted owllikely result from a combination of habitatdifferences among RUs for both the owl and itsprey (See Zoogeography and Macrohabitats ofCommon Prey). Landscape approaches tomanagement that maintain conditions forcommon prey of any predator are assumed tohave beneficial effects (Reynolds et al. 1992).However, the reliability of such conservation

strategies must be firmly based on ecologicallinks among prey, predator, and environmentalconditions. Thus, it is crucial to consider rela-tionships among prey abundance and persistenceof owl populations, and among prey abundance,availability, and habitat conditions.

RELATIVE IMPORTANCERELATIVE IMPORTANCERELATIVE IMPORTANCERELATIVE IMPORTANCERELATIVE IMPORTANCEOF PREYOF PREYOF PREYOF PREYOF PREY

Prey that positively influences owl survival,reproduction, or numbers may increase thelikelihood of persistence of Mexican spotted owlpopulations. Although no information is avail-able to quantify effects of food on spotted owlsurvival and density, previous studies haveexamined correlates between this owl’s diet andreproduction. For example, Barrows (1987)suggested that larger prey (e.g., woodrats) wastaken in greater frequency by owls with young.Thrailkill and Bias (1989) reported a similarpattern for California spotted owls occurring inthe central Sierra Nevada. In contrast, Ward(1990) observed a different pattern for northernspotted owls in northwestern California. Hefound that large prey was taken in relativelyequal frequency by breeding and nonbreedingowls, presumably because woodrats were acommon food resource for owls regardless ofbreeding status. These different results mayreflect variation in prey availability, samplingbias and variation, temporal variation, or trueregional differences in owl diets.

To evaluate if any particular prey groupcould increase persistence of the Mexicanspotted owl, we analyzed owl reproductivesuccess (average number of young fledged) as afunction of diet, year, and geographic locationusing a three-factor ANOVA. This approachwould allow us to quantify variation in owlreproduction that could be attributed to thefrequency of a particular prey in the owl’s diet.Owl reproduction related to consumptionfrequency of a given prey group would imply arelative importance of that prey. In using thisapproach, we further assumed that prey speciesthat were positively related to reproduction alsoenhanced the owl’s survival.

13

Table 5.8.Table 5.8.Table 5.8.Table 5.8.Table 5.8. Factors influencing trends observed in Mexican spotted owl diets. Data are from three studies conducted in northern Arizona, the Sacra-mento Mountains, New Mexico, and the Tularosa Mountains, New Mexico, during breeding seasons of 1991, 1992, and 1993.

FPO

14

Table 5.8. Table 5.8. Table 5.8. Table 5.8. Table 5.8. (continued)

FPO

15

This analysis was conducted by stratifyingand comparing variation in the number ofyoung owls produced among three study areas(SACMT2, COCODM, GILADM), threedifferent breeding seasons (1991-1993), andthree amounts of prey consumption, low(< 33%), high (� 66%), or medium. Seven preygroups were examined in separate ANOVAs toensure independence among diet frequencies.These included woodrats, peromyscid mice,voles, gophers, rabbits, other medium-sizedmammals, and arthropods. Study area and yearwere included as blocking factors to account forspatial and temporal effects that might alterna-tively explain patterns in the owl’s reproduction,respectively. This analysis differed from the 3-factor ANOVA used to examine geographicvariation in the owl’s diet where number ofyoung was treated as a predictor variable ratherthan a response.

Results from each ANOVA suggested thatthe owl’s reproductive success was not influencedby a single prey species but rather by manyspecies in combination. None of the specificprey groups significantly influenced owl repro-ductive success when diet frequencies wereexamined separately (Table 5.9). Time was asignificant factor influencing the owl’s reproduc-tion (Table 5.9) when consumption of woodratsor voles was examined. Prey abundance is knownto fluctuate through time. Thus, effects attrib-uted to temporal variation may also representassociated changes in prey populations. How-ever, it is more likely that the owl’s reproductivesuccess was influenced by total prey biomassconsumed in a given year, rather than by a singleprey species. For example, when frequencies ofthe three most common prey groups, woodrats,peromyscid mice and voles were combined andanalyzed, diet was a nearly significant predictorof the owl’s reproductive success (F = 2.930,d.f. = 2, 98, P = 0.058). More owl young wereproduced when moderate to high amounts ofthese common prey were consumed in all threestudy areas (Figure 5.3).

Contrary to the above results, certain preyspecies may be more important in certainregions of the owl’s range. For example, otherinformation from Ward et al. (unpublished)suggests that reproductive success of Mexican

spotted owls in the Sacramento Mountains mayincrease when deer mice populations irrupt. In1991, deer mouse biomass averaged 0.911 kg/hain mixed-conifer forests (Figure 5.4a) and thenumber of young produced corresponded to thenumber of peromyscid mice consumed (Figure5.4b). In 1992 and 1993, owl reproductivesuccess decreased corresponding with a reductionin deer mouse abundance and the frequency ofperomyscid mice consumed by owls (Figure5.4c). Reproduction among owls dwelling insteep-walled canyons of the Colorado Plateaumay also depend on a specialized diet becausethese owls consume a greater number ofwoodrats compared to other localities (Table5.1). These results may be exceptions, as most ofour findings support maintenance of severalcommon prey species, rather than enhancingpopulations of a few.

ZOOGEOGRAPHY ANDZOOGEOGRAPHY ANDZOOGEOGRAPHY ANDZOOGEOGRAPHY ANDZOOGEOGRAPHY ANDMACROHABITATSMACROHABITATSMACROHABITATSMACROHABITATSMACROHABITATS

OF COMMON PREYOF COMMON PREYOF COMMON PREYOF COMMON PREYOF COMMON PREY

As detailed earlier, Mexican spotted owls usea variety of prey. Species regarded as “common”are those comprising �10% of the owl diet byrelative frequency or biomass within a given RU.Prey commonly consumed by the owl variesgeographically (Table 5.10). This geographicvariation can be attributed to two primaryfactors: the geographic range of the prey and thedegree of sympatry in macrohabitats of the owland its prey. Below, we provide an overview ofbasic macrohabitat associations of common owlprey.

MAMMALSMAMMALSMAMMALSMAMMALSMAMMALSBatsBatsBatsBatsBats

As a group, bats are common prey in por-tions of the Colorado Plateau, Upper GilaMountains, and Basin and Range - West RUsand they are consumed occasionally by owls inall RUs. If taken from outside roosts or nurserycolonies, bats may represent a low-cost, opportu-nistic food source for the owls. No particularspecies appears to be used in great abundance;but when considered as a group, the importance



Table 5.9.Table 5.9.Table 5.9.Table 5.9.Table 5.9. Factors influencing production of Mexican spotted owls in northern Arizona, the Sacramento Mountains, New Mexico, and the TularosaMountains, New Mexico, during 1991, 1992, and 1993. Consumption of common prey was categorized according to the frequency of that prey in theowls' diet; low (<33%), medium (33% - 65.9%), and high (��66%). Higher order interactions of effects were nonsignificant (P > 0.05) in tests onperomyscid mice or could not be quantified in tests with other prey because of limited sample sizes.

FPO

17


FPO

18

of bats to the owl may be considerable (Tables5.7 and 5.10).

Bats occupy many macrohabitats, rangingfrom arid shrublands to spruce-fir forest. Inparticular, rock crevices and tree snags arecommonly used for roosting and raising young.These microhabitat components often occur inhabitats used by Mexican spotted owls forroosting and nesting.

RabbitsRabbitsRabbitsRabbitsRabbits

Cottontail rabbits are common prey accord-ing to diet biomass for all RUs except in theColorado Plateau (Table 5.10). Rabbits canprovide a great amount (50 to 400 g;1.8-14.1 oz) of food per captured individual.Desert and eastern cottontails are more com-monly associated with xeric vegetation typessuch as pinyon-juniper and oak woodlands,although Ward (personal observation) has

observed eastern cottontails within mixed-conifer forests of the Sacramento Mountains,New Mexico. Only scant information describesthe specific habitat associations of cottontails inNew Mexico. Hoffmeister (1986) notes that inArizona both desert and eastern cottontailsinhabit pinyon-juniper, whereas eastern cotton-tails also use oak woodland. Typically, thesehabitats are relatively open with a well developedgrass understory. In contrast, Nuttall’s cotton-tails are found in more mesic grassy or rockyareas near pine, pine-oak, mixed-conifer, andspruce-fir forests in the northern portion of theowl’s range.

Pocket GophersPocket GophersPocket GophersPocket GophersPocket Gophers

Pocket gophers are common prey within theBasin and Range - West and Upper Gila Moun-tains RUs. Pocket gophers typically inhabitmeadows and meadow edges, although they also

Figure 5.3.Figure 5.3.Figure 5.3.Figure 5.3.Figure 5.3. Reproductive success (mean number of young produced) as a function of low (<33%),medium (33-65.9%), or high (� 66%) consumption of woodrats, peromyscid mice, and voles, byMexican spotted owls occurring in the Sacramento Mountains, New Mexico, northern Arizona, andTularosa Mountains, New Mexico. Vertical bars are standard errors. Number of owl territories isshown in parentheses.

19


Volume II/Chapter 5

Figure 5.4.Figure 5.4.Figure 5.4.Figure 5.4.Figure 5.4. Biomass (kg/ha) of (a)(a)(a)(a)(a) common prey occurring in mixed-conifer forests, (b)(b)(b)(b)(b) frequenciesof peromyscid mice consumed by Mexican spotted owls stratified by number of owl young produced,and (c)(c)(c)(c)(c) average (�95% CI) number of owl young produced in the Sacramento Mountains, NewMexico (Ward et al. unpublished). Bars in (b)(b)(b)(b)(b) represent variation among owl pairs (standard errors).



Table 5.10.Table 5.10.Table 5.10.Table 5.10.Table 5.10. Prey comprising �10% of relative frequency (X) or biomass (O) in the diet of Mexican spotted owls.

FPO

21

Volume II/Chapter 5


occur within woodlands and forests. Botta’spocket gopher, northern pocket gopher, andsouthern pocket gopher occur within the rangeof the Mexican spotted owl. Botta’s pocketgopher is the most widespread, being found inmost vegetation types. The northern pocketgopher occurs in montane habitats of the north-ern part of the owls’ range. The southern pocketgopher is found in Basin and Range - West andMexican RUs. Both Hoffmeister (1986) andFindley et al. (1975) noted that Botta’s pocketgopher is ubiquitous whenever soil is suitable forburrowing, whereas the northern pocket gopheris more typically found in parks and meadowswithin montane forests. Findley et al. (1975)stated that the southern pocket gopher inhabitsshallow, rocky soils of pine forests.

Peromyscid MicePeromyscid MicePeromyscid MicePeromyscid MicePeromyscid Mice

Eight peromyscid mice occur within therange of the Mexican spotted owl. Only twospecies, the deer mouse and brush mouse areconsumed regularly by owls in all RUs (Table5.10). A third species, the canyon mouse, islikely consumed by owls dwelling in the Colo-rado Plateau RU and the rock mouse has beenreported in the diet of owls occurring in theBasin and Range - East RU (Table 5.7).

The deer mouse is widespread, inhabiting allvegetation types except high-elevation tundra(Bailey 1931, Hall and Kelson 1959, Armstrong1977, Goodwin and Hungerford 1979). Highreproductive success of spotted owls in theSacramento Mountains, New Mexico, (Basinand Range - East RU) has been recorded duringirruptions of deer mice in mixed-conifer forests(See Abundance and Distribution of CommonPrey, Sacramento Mountains).

More restricted in distribution, the brushmouse typically inhabits areas with extensiverock and shrub cover in pinyon-juniper, riparian,oak, and pine-oak woodlands (Wilson 1968,Armstrong 1979, Svoboda et al. 1988).Goodwin and Hungerford (1979) found thatbrush mice inhabit rocky slopes in centralArizona’s pine-oak forests, but rarely use purestands of pine.

The canyon mouse occupies the canyonwalls, cliffs, and steep rocky slopes of northern

Arizona, Utah, northwestern New Mexico, andwestern Colorado (Armstrong 1979, Johnsonand Armstrong 1987). Findley et al. (1975)reported that the canyon mouse is often foundin pinyon-juniper woodlands at the base of cliffs,although Johnson and Armstrong (1987) notedthat vegetation associations are of limited impor-tance relative to the presence of suitable rockysubstrates.

The rock mouse is found in association withrocky substrates generally above 1,900 m (6,230ft) elevation (Findley et al. 1975, Cornely et al.1981). This species occurs in most of the UnitedStates portion of the owl’s range and is oftensympatric with deer mice and brush mice (Wil-son 1968, Cornely et al. 1981, Ribble andSamson 1987).

WoodratsWoodratsWoodratsWoodratsWoodrats

Mexican, bushy-tailed, desert, and white-throated woodrats are consumed by Mexicanspotted owls. For the owl, woodrats provide alarge mass of food per capture. This prey is aprimary food source for owls through most of itsrange but particularly in the canyon habitats ofthe Colorado Plateau RU, where all four speciesof woodrats are sometimes found (Table 5.10).

The Mexican woodrat is perhaps the mostcommon woodrat found within the range of theMexican spotted owl. It occurs within all RUs,although populations are disjunct because of thespecies’ montane distribution (Cornely andBaker 1986). The altitudinal range of the Mexi-can woodrat begins in the lower pine zone andextends upward through mixed-conifer forestswhere Findley et al. (1975) reported they reachtheir greatest abundance. Hoffmeister (1986; seealso Goodwin and Hungerford 1979) regardedMexican woodrats as rock dwellers within thesevegetation types, infrequently extending intopinyon-juniper woodlands. Armstrong (1972),however, reported that this species typically usespinyon-juniper woodland in western Coloradoand scrub-like oaks and mountain mahoganyvegetation along the eastern foothills northwardsalmost to Wyoming. The range of Mexicanwoodrats in Utah is restricted to the southeasternportion of the state, east of the Colorado River.

22

Volume II/Chapter 5


Presumably, the species uses similar habitats inUtah as it does in western Colorado.

Bushy-tailed woodrats are a common dietcomponent within the Colorado Plateau andboth Southern Rocky Mountain RUs. Thisspecies is a cordilleran mammal of westernColorado, northcentral and northwestern NewMexico, northeastern Arizona, and eastern Utah(Durrant 1952, Hoffmeister 1986, Findley et al.1975, Armstrong 1972, 1979). Finley (1958)reported that bushy-tailed woodrats use a varietyof vegetation types, primarily woodland andshrublands. Their distribution may depend onthe presence of suitable rock outcrops ratherthan specific vegetation (Finley 1958,Hoffmeister 1986). At higher elevations, theseoutcrops interrupt open forests of Douglas-fir,aspen, or ponderosa pine containing a well-developed shrub understory. Typically, lowerelevation sites are dominated by pinyon andjuniper.

The desert woodrat is consumed by owls inthe Colorado Plateau RU. This species is mostabundant in the Arizona strip where it inhabits avariety of plant communities including creosotebush and cacti to junipers and pine (Hoffmeister1986). Desert woodrats frequently nest increvices of cliffs and rock outcrops withinjuniper or shadscale plant communities of Utahand Colorado, below 2,000 m (6,560 ft) eleva-tion (Finley 1958).

White-throated woodrats are common preyof owls occurring in the Upper Gila Mountainsand Basin and Range - West RUs and are lessfrequently consumed by owls in other RUs. Thiswoodrat species is typically distributed below theconifer belt, although it can be found in pinyon-juniper woodlands (Hoffmeister 1986).

VolesVolesVolesVolesVoles

Four species of voles are common prey of theMexican spotted owl including the Mexican (orMogollon vole [after Frey and LaRue 1993]),mountain, meadow, and long-tailed voles (Table5.10). Three of these species can be orderedalong an environmental moisture gradient fromsemi-arid (Mexican vole), mesic (mountainvole), to hydric (meadow vole). Long-tailed voles

inhabit environments along the entire gradient(Getz 1985).

The Mexican vole is common within thegreatest number of RUs, including Basin andRange - East, Southern Rocky Mountains - NewMexico, and Upper Gila Mountains RUs. It isfairly widely distributed in Arizona and NewMexico, but it is confined to the southeast partof Utah and to southwest Colorado. Thisspecies occurs in the widest range of habitats ofany microtine and is generally associated withxeric grassy locations extending from pinyon-juniper to spruce-fir zones (Armstrong 1972,Findley and Jones 1962, Findley et al. 1975,Finley et al. 1986, Hoffmeister 1986, Frey andLaRue 1993).

The two species associated with wet condi-tions (mountain and meadow voles) generallyoccur in the northern portion of the owl’s range.Mountain voles are common prey in the Colo-rado Plateau and both Southern Rocky Moun-tain RUs. In these areas, the mountain voleoccupies forest meadows ranging in elevationfrom open pine-oak to spruce-fir forests.Armstrong (1977) found mountain voles indense grass cover with a sparse overstory. Thisvole’s geographic range includes most montaneregions along the north-south axes of Coloradoand Utah, northern New Mexico, and the WhiteMountains in Arizona. Meadow voles occur inboth Southern Rocky Mountains RUs wherepermanent water is provided by springs andmarshes (Armstrong 1972, Findley et al. 1975,Finley et al. 1986).

The long-tailed vole occurs within theSouthern Rocky Mountains - New Mexico,Upper Gila Mountains, and Basin and Range -East RUs. Findley et al. (1975) reported that it isassociated with meadows and forest edge, beingmost common in the mixed-conifer and spruce-fir zones but also using mixed-conifer stringersfound along canyons in the ponderosa pinezone. Armstrong (1972, 1977) noted that thismicrotine requires a grassy understory less thanother vole species because it is often foundwithin forests supporting minimal grass cover.

23

Volume II/Chapter 5


BirdsBirdsBirdsBirdsBirds

Birds are common prey for the owls in theSouthern Rocky Mountains - New Mexico,Upper Gila Mountains, and Basin andRange - West RUs where numerous species havebeen identified in pellets (Table 5.7). Theimportance of birds to spotted owls is uncertain.Birds do contribute to the diversity of prey takenby owls and may provide food resources whensmall mammals are less abundant. However, useof birds as prey is likely seasonal because many ofthe passerine species consumed by owls aremigratory. All species identified in the owls’ dietto date are forest dwellers.

ArthropodsArthropodsArthropodsArthropodsArthropods

Arthropods are common prey of spottedowls in the Colorado Plateau, Southern RockyMountains - Colorado, Southern Rocky Moun-tains - New Mexico, and Upper Gila MountainsRUs. Many species are consumed (Table 5.7).The importance of arthropods to spotted owls isalso uncertain. If one considers biomass alone,arthropods contribute little to the owl’s diet.However, arthropods may provide a low cost,high quality food taken opportunistically.Description of macrohabitats of arthropodsconsumed by spotted owls is beyond the scopeof this document. Active management of arthro-pod populations for owl recovery is probably notnecessary.

ABUNDANCE ANDABUNDANCE ANDABUNDANCE ANDABUNDANCE ANDABUNDANCE ANDDISTRIBUTION DISTRIBUTION DISTRIBUTION DISTRIBUTION DISTRIBUTION OFOFOFOFOF

COMMON PREY COMMON PREY COMMON PREY COMMON PREY COMMON PREY

The availability of prey to Mexican spottedowls depends on prey abundance, the vulnerabil-ity of the prey to capture by the owl, and theprobability that the owl and its prey occur in thesame habitat. All of these factors vary by habitatcondition. In addition, the amount of energyavailable to the owl will vary according to thetype, size, and condition of the prey. Thus, it isuseful to convert prey numbers into values thatreflect energy input. Because rodents are themost common prey of the spotted owl, we

assume that most of the difference in energycontent among mammalian prey can be attrib-uted to body mass.

Although no information exists on thevulnerability of different prey species to captureby Mexican spotted owls, estimates of commonprey abundance and mass within several differentvegetation communities are available fromresearch conducted in northern Arizona (Blockand Ganey, unpublished) and the SacramentoMountains, New Mexico (Ward et al., unpub-lished). These two studies provide estimates ofbiomass and microhabitat associations (See PreyHabitat) for the owl’s common prey in fourdifferent vegetation communities used by owlsfor foraging, ponderosa pine-Gambel oak,mixed-conifer, high-elevation meadows, andponderosa pine-pinyon-juniper woodlands. Themethods and results of both studies are brieflydescribed below.

Northern ArizonaNorthern ArizonaNorthern ArizonaNorthern ArizonaNorthern Arizona

MethodsMethodsMethodsMethodsMethods

Small mammal populations were sampledusing live-trapping and mark-recapture tech-niques from November 1990 through December1992 within home ranges of five owl pairs. Thepurpose of the sampling was to estimate biomassof the owl’s common prey and determine theprey’s distribution. Common prey were deter-mined from owl pellets collected during thisstudy.

The study area consisted of ponderosa pine-Gambel oak forest, although each trapping gridwas unique with respect to the relative composi-tion and structure of the vegetation. Ancillarytrapping was conducted within the winter rangeof two owls that migrated downward to pinyon-juniper woodland in the Verde Valley. Thistrapping was confined to five sets of smallertrapping grids (described below) for a total of574 trap nights.

Trapping grids were randomly established atgeneral foraging areas identified by radio telem-etry (Ganey and Block, unpublished data). Trapswere arrayed in 10 x 10 or 2 x 10 grids with20-m (65.5 ft) spacings between stations. The

24

Volume II/Chapter 5


larger grids were used to estimate both densityand habitat correlates; the smaller grids wereused only to assess habitat use by species (seePrey Habitat). Large Sherman live traps (size8 x 9 x 23 cm [3 x 3.5 x 9 in]) were placed ateach grid station; extra-large Sherman live traps(size 10 x 18 x 60 cm (4 x 4.5 x 15 in) wereplaced at alternate stations. Two sizes of trapswere used to minimize the potential bias againstcapturing larger prey such as woodrats in thesmaller traps.

Grids were trapped from 3-7 nights duringeach trapping session. Traps were left openduring the day to sample diurnal sciurids. Thegoal was to continue trapping until 90% of allcaptures were recaptures to approach assump-tions of population closure. This goal wastypically reached between five and seven nights.During periods of inclement weather, however,we relaxed this goal to minimize trappingmortalities. Each grid was trapped for 5-7sessions during the study. Total trapping effortwas 49,911 trapnights (adjusted for closed andunoccupied, or otherwise unavailable traps).When an animal was captured, it was identifiedto species, weighed, and marked by toe clipping.

Abundance was estimated as the number ofindividuals per effective sampling area (ha) usingclosed population estimators (Program CAP-TURE; Otis et al. 1978, White et al. 1982).Although the number of individuals captured atsome grids during some seasons was insufficientfor producing unbiased estimates, we consideredthe closed population estimators in ProgramCAPTURE more appropriate than minimumnumbers alive per grid area. The former permit-ted estimating capture probabilities and sam-pling variances plus population size; the latterwould not.

Biomass of common prey, expressed askilograms per hectare (kg/ha), was calculated as aproduct of prey density and average mass. Thedelta method (Goodman 1960) was used toestimate the sampling variance of biomass fromthe variances associated with sampling prey

density and mass.

General DistributionGeneral DistributionGeneral DistributionGeneral DistributionGeneral Distribution

The three primary prey species captured inponderosa pine-Gambel oak forests were deermouse, brush mouse, and Mexican woodrat.Other prey species captured included pinyonmouse, white-throated woodrat, Stephens’woodrat, Mexican vole, rock squirrel, gray-collared chipmunk, and cliff chipmunk. Brushmice and white-throated woodrats were capturedwith the greatest relative frequency in pinyon-juniper woodlands. The trapping efforts did notsample prey species such as pocket gophers,cottontail rabbits, birds, and arthropods. Fur-ther, only the three primary species were cap-tured in sufficient numbers to permit densitycalculations or estimates of habitat correlates.Thus, our analyses of population size and habitatuse address only these three species.

Population AbundancePopulation AbundancePopulation AbundancePopulation AbundancePopulation Abundance

The deer mouse was the most abundantspecies captured, followed by the brush mouseand Mexican woodrat. A seasonal trend ofdecreased prey biomass during winter was noted(Figure 5.5a). Prey abundance and biomass alsovaried by year (Figure 5.5a) and among owlterritories (Block and Ganey, unpublished data).

Sacramento MountainsSacramento MountainsSacramento MountainsSacramento MountainsSacramento Mountains

MethodsMethodsMethodsMethodsMethods

The Sacramento Mountains are located insouthcentral New Mexico. An investigation ofthe abundance and distribution of commonmammalian prey began in 1991 and is ongoing(Ward et al., unpublished data). Common preywere determined from owl pellets collectedduring this study. Abundance was estimatedamong three general vegetation communitiesusing mark-recapture methodology.

The three communities included mesicforests, xeric forests, and meadows. Mesic forestswere a mixture of Douglas-fir, white fir, south-western white pine, ponderosa pine, andEnglemann spruce (i.e., mixed conifer). Mead-ows consisted of forbs and grasses and were

25

Volume II/Chapter 5


Figure 5.5.Figure 5.5.Figure 5.5.Figure 5.5.Figure 5.5. Average biomass of common prey (kg/ha) of the Mexican spotted owl in (a)(a)(a)(a)(a) pine-Gambel oak habitats of northern Arizona (Block and Ganey unpublished data) and (b)(b)(b)(b)(b) in three habi-tats of the Sacramento Mountains, New Mexico. Vertical bars are standard errors. Lower horizontalbars in (b)(b)(b)(b)(b) separate sampling error from errors associated with variation among sites. Number of siteswithin each habitat are shown in parenthesis.

26

Volume II/Chapter 5


associated with drainage bottoms adjacent tomesic forests. Xeric forests included ponderosapine, pinyon, juniper, and low-growing oaks.

The same two sizes of live-traps described forthe northern Arizona study were used to captureprey in the Sacramento Mountains. Traps werealso arranged similarly, with the followingexceptions. In 1991, four 13-ha (32.1-ac)trapping grids were established as part of a pilotstudy. Two grids were placed in the mesic forestand two in the xeric forest. In 1992 and subse-quent years, six 4-ha (9.9-ac) trapping grids wereplaced in each of the mesic and xeric foresttypes. Traps were arrayed as 11 x 11 station grids.In addition, six 1.8-ha (4.5-ac) grids wereestablished in meadows. Each of the latter gridsconsisted of 105 large Sherman traps spaced15 m (49.2 ft) apart in a 5 x 21 array.

All grid sites were selected randomly from alist of known Mexican spotted owl territories.Grids were placed as close to a nest or roost areaas possible while maintaining homogeneity at ascale analogous to a forest stand. In addition,each grid was � 0.8 km (0.5 mi) from a nest orroost to ensure that the site was available for useby spotted owls.

All captured individuals were marked withuniquely numbered ear tags in both ears. Loss ofboth tags during an eight-night trapping sessionwas less than 1% for all marked species. All othermethods of sampling and data collection weresimilar to the study in northern Arizona (Blockand Ganey, unpublished data). We thereforeconsidered the results of both studies to becomparable. Prey abundance and biomass wereestimated using the same procedures describedfor the northern Arizona study. Tests for spatialand temporal differences in prey biomass wereconducted using a two-factor ANOVA.

General DistributionGeneral DistributionGeneral DistributionGeneral DistributionGeneral Distribution

Common prey included woodrats,peromyscid mice, voles, arthropods, cottontailrabbits, and other medium-sized mammals suchas long-tailed weasels, red squirrels, and chip-munks (Tables 5.3 and 5.6). Cottontails andother medium-sized mammals were consumedinfrequently but larger estimates of mass from afew individuals resulted in larger biomass calcu-

lations from these prey groups. However, theexact sizes of consumed individuals were unde-termined and the range of mass for larger preyconsumed by owls could have been considerable.For example, the size of a cottontail taken by aspotted owl could range 50 to 400 g (1.8-14.1oz) whereas a peromyscid mouse could range 10to 40 g (0.4 to 1.4 oz). This potentially results inupwardly biased estimates of biomass for largerspecies in the owls’ diet. For this reason, we areuncertain about the role of cottontails and othermedium-sized prey as common food resourcesfor the owl. In contrast, arthropods comprised11.4% of the diet but were not consideredenergetically important because of their smallmass (1-2 g [0.04-0.07 oz]).

Five species common in the owl’s diet werecaptured regularly during live-trapping. Thedistribution of each species varied by vegetationcommunity. Deer mice were found in all threecommunities. Brush mice were restricted to thexeric forest type and were primarily associatedwith areas containing shrub-form oaks. Long-tailed voles and Mexican voles were more com-mon in meadows, but they also occupied thetransition zones between meadows and mesicforests. Mexican voles were also found infre-quently in xeric forests. Mexican woodratsoccupied mesic forests and the ecotone betweenthese forests and meadows, as well as xericforests. Medium-sized mammals such as thegray-footed chipmunk were found in mesicforests, the edges between mesic forests andmeadows, and rarely in xeric forests. Cottontailrabbits and red squirrels were only encounteredin mesic forests. However, our sampling proce-dures were inadequate for estimating abundanceand distribution of these species and ofarthropods.

Population AbundancePopulation AbundancePopulation AbundancePopulation AbundancePopulation Abundance

Total biomass (kg/ha) of five common prey(Figure 5.5b) varied annually over the threesummers 1991-1993 (F = 12.7, d.f. = 2, 32,P < 0.001) and also seasonally during the sum-mer-fall-winter period of 1993-1994 (F = 17.0,d.f. = 2, 21, P < 0.001). Fluctuations in preybiomass also varied by vegetation communityduring both annual (F = 5.5, d.f. = 2, 32,

27

Volume II/Chapter 5


P = 0.009) and seasonal cycles (F = 5.8, d.f. = 2,21, P = 0.010). The general trend was that preybiomass was moderately high during the summer(when owls feed their young) of 1991, peakingin 1992, then decreasing to moderate levels in1993 (Figure 5.5b). Further, temporal trendsdiffered by vegetation community. Prey biomassdecreased in mesic forests and meadows overconsecutive summers but increased then de-creased in xeric forests during this same period(Figure 5.5b). This relationship was evident by astatistically significant interaction of vegetationcommunity and year on prey biomass (F = 3.3,d.f. = 3, 32, P = 0.032). Whereas prey biomasspooled across all communities decreased fromthe summer to fall of 1993 before increasingslightly in the winter (Figure 5.5b), seasonaltrends differed among communities (F = 17.0,d.f. = 2, 21, P < 0.001). These patterns alsoshowed an interactive influence of season andvegetation community on prey biomass (F = 3.0,d.f. = 2, 21, P = 0.044). Thus, abundance ofpotential food resources for the Mexican spottedowl in the Sacramento Mountains are tempo-rally variable and habitat dependent.

The variations in total prey biomass werealso reflected in the population dynamics ofdifferent species (Figure 5.6a). For example, deermice densities within mesic forests peaked in thesummer of 1991 and declined through thesummer of 1993 while maintaining low densitiesin the meadows (Figure 5.6b). Deer mice inxeric forests maintained lower, relatively stabledensities that peaked in summer of 1992 (Figure5.6c). In contrast, Mexican voles maintainedlow, stable populations in the mesic forests butincreased dramatically in meadows and moder-ately in xeric forests in the summer of 1992before crashing in both habitats during thesummer of 1993 (Figure 5.6).

In summary, the two studies indicate thatthe owl’s food resources are quite variable amongvegetation communities and through time.Arranging the four vegetation communitiesexamined in these two studies in descendingamount of summer prey biomass indicates thatmeadows > mixed-conifer forest > ponderosapine-pinyon-juniper-oak woodlands > ponderosapine-Gambel oak forest. When consideringother factors that influence the availability of

prey, mixed-conifer forests likely provide thegreatest amount of food during summer periods.Rearranging the same communities according towinter prey biomass indicates that meadows >ponderosa pine-pinyon-juniper-oak woodlands >ponderosa pine-Gambel oak forest > mixed-conifer forest. Accounting for the availability ofprey, woodlands with a mixture of ponderosapine, pinyon-juniper, and oaks provide moreprey to owls during winter months than theother three communities. However, temporalpeaks of prey cycles are not correlated amongthese communities. That is, when prey areabundant in mixed-conifer forest one year andlow a subsequent year, an opposite pattern mayoccur in a different vegetation community. Theasynchrony of abundance among prey species,vegetation communities, and time may provide abuffer against the effects of extreme oscillation inprey cycles. This implies maintaining a mixtureof vegetation communities within the owl’sforaging range.

Results of both studies also showed that theowl’s food is most abundant during the summerwhen young are being raised. Decreases in preybiomass occur from late-fall through winter.Seasonal decreases, like these, are typical of smallmammal populations. Unfortunately, explana-tions for fluctuations in small mammal popula-tions are equally variable and include artificialrandom patterns in population data, weatherchanges, behavioral mechanisms, predation, andage-related effects on demographic processes(e.g. fecundity, dispersal; see reviews by Conleyand Nichols 1978, Finerty 1980).

Although the reasons for seasonal declines inprey and their ramifications on the owl have notbeen quantified, conditions that increase winterfood resources will likely improve conditions forthe owl. For example, Hirons (1985) has shownthat large body reserves of fat and protein areessential during incubation by female tawny owlsfor successful reproduction. Abundant preypopulations during winter and early springperiods increase the likelihood of egg laying anddecrease the rate of nest abandonment (Hirons1985).

Obviously, there will be little recourse forenhancing prey populations if factors like pre-cipitation, temperature, or amount of snow

28

Volume II/Chapter 5


Figure 5.6.Figure 5.6.Figure 5.6.Figure 5.6.Figure 5.6. Trends in average density (no./ha) of common prey of the Mexican spotted owl in (a)(a)(a)(a)(a)mixed-conifer forest, (b)(b)(b)(b)(b) montane meadows, and (c)(c)(c)(c)(c) ponderosa pine-pinyon-juniper-oak woodlands,Sacramento Mountains, New Mexico. Summer (1992-93) values are based on 6 spatial replicates;summer (1991), fall, and winter values are based on 2 replicates.

29

cover are the major causes of decline in preypopulations. The degree to which habitat ma-nipulation can ameliorate against weather effectsor enhance prey availability requires investiga-tion and is encouraged. Successful manipulationmay provide a potent tool for recovering Mexi-can spotted owls or other predators in the future.

PREY HABITATPREY HABITATPREY HABITATPREY HABITATPREY HABITAT

Ensuring adequate food for the owl requiresconserving and possibly restoring habitat of theowl’s prey. Ideally, to succeed, those features ofthe environment that consistently cause increasesin abundance and availability of desired preyspecies should be identified. In reality, fewhabitat studies are so revealing or precise (seeVerner et al. 1986). However, habitat studiesoften do identify patterns and descriptive corre-lates of animal distribution or abundance.Although crude, this type of information fre-quently is all that is available for predicting theoutcome of planning decisions or managementprescriptions.

In the absence of cause-effect relationshipsamong the owl’s prey and its habitat, we presenthabitat correlates for the distribution of severalcommon prey species. This information wasdetermined from the same studies of preyabundance conducted in northern Arizona(Block and Ganey, unpublished data) and theSacramento Mountains (Ward et al., unpub-lished data).

Northern Arizona - Pine-oak ForestNorthern Arizona - Pine-oak ForestNorthern Arizona - Pine-oak ForestNorthern Arizona - Pine-oak ForestNorthern Arizona - Pine-oak Forest

Field Methods and Statistical AnalysesField Methods and Statistical AnalysesField Methods and Statistical AnalysesField Methods and Statistical AnalysesField Methods and Statistical Analyses

Habitat-sampling plots were established as a5-m [16.5-ft) radius centered at each trappingstation (n = 1,260). Cover by grass, forbs, rock,dead woody debris of three size classes (<1 cm[0.34 in], 1-10 cm [0.39-3.9 in], >10 cm [3.9in]), and live woody vegetation at four heightstrata (<1 m [3.3 ft], 1-2 m [3.3-6.6 ft], >2-5 m[6.6-16.4 ft], >5 m [16.4 ft]) were estimated asthe percentage of 10 point intercepts (at 1-m[3.3 ft] intervals along a randomly orientedtransect) covered by each of these variables. Tree

diameters were measured with a dbh tape;heights were measured with a clinometer. Shruband slash pile heights were measured with ameter stick. Mid-point diameters and lengths oflogs within the plot were measured with ameasuring tape. Numbers of trees and shrubs byspecies were recorded. Slope was measured witha clinometer and aspect with a compass.

Deer mice, brush mice, and Mexicanwoodrats were captured in sufficient numbers topermit habitat analyses. Stations where eachspecies was captured were contrasted with thosewhere it was not captured using analysis ofvariance with owl territory as a blocking factor.Two other analyses were also conducted. Logisticregression was used to examine differencesbetween stations where each species was and wasnot captured that may not have been identifiedin the ANOVA. Stepwise multiple linear regres-sion (Draper and Smith 1981) was used toevaluate relationships between small mammalpopulation levels and habitat characteristics. Thedependent variable was the number of trappingstations/grid where the species was captured.Independent variables were habitat characteris-tics aggregated across the grid.

ResultsResultsResultsResultsResults

Deer mice used more open sites, on gentlerslopes, and with less shrub and midstory canopy,smaller densities of Gambel oak trees and shrubs,but more slash piles and greater litter depth thanstations where it was not captured (Table 5.11).In contrast, brush mice and Mexican woodratsboth used sites characterized by greater slopes,low vegetation cover, sparser tree canopy (>5 m[16.4 ft]) cover, more Gambel oak shrubs, andgreater log volume than areas where they werenot captured (Table 5.11). Further, brush miceused areas with greater Gambel oak tree densityand basal area and less ponderosa pine basal areathan found at stations where it was not captured.Tree basal area at sites where Mexican woodratswere captured was significantly less and totalrock cover was significantly greater than at siteswhere it was not captured.

Generally, results from logistic regressionanalyses corroborated results of the univariateanalyses. Deer mice were associated with areas of



31

Table 5.11. Table 5.11. Table 5.11. Table 5.11. Table 5.11. Descriptive statistics (means with SE in parentheses) of selected habitat variables characterizing habitats of common mammalian prey ofMexican spotted owls (4 territories) in ponderosa pine-Gambel oak forests, northern Arizona, 1990-1992.


FPO

32

little slope, Gambel oak, and mid-story (2-5 m[6.6-16.4 ft]) vegetation. Brush mice and Mexi-can woodrats used areas with greater shrubdensity, rock cover, log volume, and Gambel oakcover.

In the multiple regression analysis, novariable entered the model for deer mice, sug-gesting that it is a habitat generalist. Two vari-ables, shrub density and grass height, entered theregression model for brush mice. These twovariables alone explained >96% of the variationin the data set. For Mexican woodrats, shrubdensity and slope entered the model and ex-plained >92% of the variation in numbers. Thus,both brush mice and Mexican woodrats wereapparently more stenotypic in habitat than deermice and closely associated with shrub coverprovided by Gambel oak and New Mexicanlocust.

Sacramento MountainsSacramento MountainsSacramento MountainsSacramento MountainsSacramento Mountains

Field Methods and Statistical AnalysesField Methods and Statistical AnalysesField Methods and Statistical AnalysesField Methods and Statistical AnalysesField Methods and Statistical Analyses

Habitat characteristics were measured withincircular plots of 5-m [16.4-ft] radius centered atabout 80% of the trap stations (n = 1,416).Stations were stratified by vegetation commu-nity: mesic forest (n = 583), xeric forest(n = 578), and meadow (n = 255). Within eachplot, methods similar to those described abovewere used with the following exception: treebasal area and density were estimated with aplotless method using 10- and 20-factor prismsrather than the plot methods used in the ponde-rosa pine-Gambel oak forest of northern Ari-zona. Also, cover of woody vegetation by heightstrata was not recorded in the SacramentoMountains.

Deer mice, brush mice, Mexican woodrats,long-tailed voles, and Mexican voles were cap-tured in sufficient numbers to permit habitatanalyses. Statistical analyses for each species weredone separately by vegetation community.Habitat characteristics of trap stations where aspecies was captured were contrasted withstations where they were not captured usingStudent’s t-test. A random sample of unusedstations equal to the number of used stations was

selected to meet the assumption of equal samplesize except for Mexican voles in meadows, whichwere captured at >75% of the trap stations.Consequently, a subset of used stations equal tothe available number of unused stations wasrandomly selected (n = 62) for conducting theanalysis. Logistic regression was also used todetermine the relative value of variables indistinguishing between stations where theanimal was and was not captured. This secondanalysis was used to detect other variables thatmight identify habitat correlates not identifiedby use of Student t-tests.

ResultsResultsResultsResultsResults

MMMMMesic Fesic Fesic Fesic Fesic Forororororests.ests.ests.ests.ests.—Microhabitat analyses werepossible for deer mice, Mexican woodrats, andlong-tailed and Mexican voles. Deer mice werecaptured in areas with little herbaceous cover andextensive exposed soil (Table 5.12). Results fromthe logistic regression indicated that deer miceused areas with less herbaceous cover and greaterdensities of live conifer trees. Mexican woodratsused areas that had greater shrub but less herba-ceous cover (Table 5.13). Shrub cover was theonly variable included in the stepwise logisticregression model of habitat use by Mexicanwoodrats. Long-tailed voles occurred in areascharacterized by less slope, less tree cover andexposed soil but greater herbaceous cover, fewerstumps, greater shrub numbers, and fewerconifer snags (Table 5.14). According to thelogistic regression model, long-tailed voles usedareas with less exposed soil, greater shrub density,and less slope. Mexican voles used sites with lessshrub, tree and exposed ground cover, greaterherbaceous cover, fewer shrubs, fewer coniferseedlings and saplings, lower density and lessbasal area of deciduous trees (Table 5.15).Mexican voles used sites with less tree cover andlower deciduous tree density according to thelogistic regression model.

XXXXXeric Feric Feric Feric Feric Forororororests.ests.ests.ests.ests.—Habitat analyses werepossible for deer mice, brush mice, Mexicanwoodrats, and Mexican voles. As in mesic forests,deer mice occupying xeric forests were capturedin areas with more exposed soil than atnoncapture stations (Table 5.12). In contrastwith mesic forests, deer mice in xeric forests were



Table 5.12.Table 5.12.Table 5.12.Table 5.12.Table 5.12. Habitat characteristics used by deer mice in three vegetation communities of the Sacramento Mountains, New Mexico. Only those variablesthat differed significantly between trap stations where deer mice were and were not captured are reported.

FPO

34

Table 5.13. Table 5.13. Table 5.13. Table 5.13. Table 5.13. Habitat characteristics used by Mexican woodrats in two vegetation communities of the Sacramento Mountains, New Mexico. Only thosevariables that differed significantly between trap stations where woodrats were and were not captured are reported.

FPO

35

Table 5.14. Table 5.14. Table 5.14. Table 5.14. Table 5.14. Habitat characteristics used by long-tailed voles in two vegetation communities of the Sacramento Mountains, New Mexico. Only thosevariables that differed significantly between trap stations where long-tailed voles were and were not captured are reported.

FPO

36

Table 5.15. Table 5.15. Table 5.15. Table 5.15. Table 5.15. Habitat characteristics used by Mexican voles in three vegetation communities of the Sacramento Mountains, New Mexico. Only thosevariables that differed significantly between trap stations where Mexican voles were and were not captured are reported.

FPO

37

found on flatter areas with less rock but greatershrub cover, and greater density and basal area oflive conifer trees (t-test, P < 0.05). Logisticregression identified slope, density of live conifertrees, litter depth, and number of shrub speciesas useful variables for distinguishing deer mousehabitat. Brush mice were found in areas havingless tree cover, greater rock cover, shallower litterdepth, greater densities of shrubs, gray oak, andconifer seedlings and saplings, but fewer conifertrees (Table 5.16). Logistic regression resultsindicated that brush mice used areas with shal-lower litter depth, greater shrub density, andfewer conifer trees. Mexican woodrats werecaptured in areas with greater shrub cover anddensity, and greater cover by gray oak (Table5.13). Shrub cover was the only variable thatdiscriminated capture from noncapture sites bylogistic regression. Mexican voles were found atsites with greater herbaceous cover and height,and greater density and basal area of conifersnags (Table 5.15). Logistic regression identifiedtwo variables, density of conifer snags and heightof herbaceous cover, as habitat discriminates forMexican voles occupying xeric forests.

MMMMMeadoeadoeadoeadoeadows.ws.ws.ws.ws.—Deer mice, long-tailed voles,and Mexican voles were captured frequentlyenough to permit analyses. As in both foresttypes, deer mice occupying meadows werecaptured in areas with more exposed soil (Table5.12). These mice also used areas with steeperslopes, shallower litter depth, and more shrubspecies (Table 5.12). Logistic regression indi-cated that only steeper slopes could be used todistinguish between capture and noncapture sitesof deer mice. Long-tailed voles were captured atstations with less herbaceous cover, greater shrubdensity, and a greater number of shrub species(Table 5.14). Even though cover by herbaceousvegetation was less at capture stations, it stillaveraged 90.0% (SE = 2.8). Shrub density andherbaceous cover were the two variables thatseparated capture from noncapture sites usinglogistic regression. Mexican voles were found onflatter areas with greater herbaceous cover andheight, greater litter depth, and less exposedground (Table 5.15). Logistic regression identi-fied slope, height of herbaceous vegetation, andlitter depth as habitat variables associated withthe presence of Mexican voles in meadows.

DISTURBANCE EFFECTS ONDISTURBANCE EFFECTS ONDISTURBANCE EFFECTS ONDISTURBANCE EFFECTS ONDISTURBANCE EFFECTS ONOWL PREY AND HABITATOWL PREY AND HABITATOWL PREY AND HABITATOWL PREY AND HABITATOWL PREY AND HABITAT

The distribution and abundance of thespotted owl’s prey are influenced by both naturaland anthropogenic factors. Though all factors tosome degree have formative character, many aremore accurately described as disturbance factors.Definitions of disturbance and the magnitude ofassociated effects vary according to ecologicalscale and conditions. Here, we briefly discussfire, tree harvesting, and livestock grazing be-cause these activities operate at spatial scaleslikely to influence spotted owls. Specifically, theymay determine whether an owl occupies andreproduces in a given area. Important to thisdiscussion is the element of human control overthese disturbance activities.

We know little about direct cause-effectrelationships of most natural and anthropogenicdisturbance factors on owl prey populations.Correlative information allows us to infer someeffects but most published research is from areasoutside the range of the Mexican spotted owl.Their applicability to southwestern conditions isuncertain. Furthermore, effects of disturbancewill vary by species, time, and space. Conse-quently, it is important to view disturbance ateach of these scales. The issue becomes evenmore complicated when one considers thesynergistic and cumulative effects of multipledisturbances. The latter scenario is more likelythe norm than the exception.

FireFireFireFireFire

Generalizing about the effects of fire on theowl’s prey is impossible for a number of reasonsincluding variations in fire characteristics and inprey habitat. Fire intensity, size, and behavior areinfluenced by numerous factors such as vegeta-tion type, moisture, fuel loads, weather, season,and topography. Data presented in the previoussections illustrate how macrohabitat and micro-habitat associations of the owl’s prey vary bothgeographically within a species and amongspecies.

Fire can effectively alter vegetation structureand composition thereby affecting small mam-



Table 5.16. Table 5.16. Table 5.16. Table 5.16. Table 5.16. Habitat characteristics used by brush mice in xeric forests of the Sacramento Mountains, New Mexico. Only those variables that differedsignificantly between trap stations where brush mice were and were not captured are reported.

FPO

39

Volume II/Chapter 5


mal habitats. Population responses by smallmammals to fire-induced changes in theirhabitat vary. For example, deer mouse popula-tions might increase immediately following fireand then decrease through time (Peterson et al.1985, Kaufman et al. 1988). Based on limitedsampling efforts restricted to one location,Campbell et al. (1977) noted that populations ofperomyscid mice decreased immediately follow-ing fire in an Arizona ponderosa pine forest thatremoved one-fourth (moderately burned) totwo-thirds (severely burned) of the basal area;populations then returned to prefire numberstwo years following the burn. Further, no differ-ences were found in rodent populations betweenmoderately and severely burned areas. Theyconcluded that the effects of the fire that theystudied were short-term, and the short-termpositive numerical responses by mice wereattributed to an increase in forage, particularlygrasses and forbs.

However, we suspect that effects of moreintense stand-replacing fires that dramaticallyalter forest structure and move the system toearlier seral stages would have longer-term effectson rodent populations. Likely, early successionalspecies (such as the deer mouse) and those thatrequire open habitats with a well developedherbaceous understory (such as microtine volesand pocket gophers) would benefit. In contrast,species that require a wooded or forested over-story would exhibit population declines. The neteffect of such fires on spotted owls is unclear. Afire that removes the tree canopy would likelyrender the area unusable for foraging because ofhow spotted owls forage. But if the spatial extentof crown loss is limited, a mosaic is created thatcould provide a diversity of prey for the owl andactually be beneficial.

Clearly, research is required to determine theeffects of fire on spotted owl prey. Because owlprey species evolved in ecosystems where fire wasa natural process, we assume that these speciessurvive and some even benefit from the occur-rence of fire. Fire has been excluded from mostsouthwestern ecosystems during the 20th cen-tury, resulting in systems where fire behaviormay deviate substantially from natural condi-tions. Effects of fire on small mammals underpresent environmental conditions are unclear.

Timber Harvest andTimber Harvest andTimber Harvest andTimber Harvest andTimber Harvest andFuelwood RemovalFuelwood RemovalFuelwood RemovalFuelwood RemovalFuelwood Removal

Numerous silvicultural methods are used toharvest timber in the Southwest. These includeboth even-aged (predominantly shelterwoodsystems) and uneven-aged (e.g., single tree,group selection systems) management. Further,these systems are applied differently according tosite characteristics and management objectives.Given the various scenarios for silviculture,generalizing about their effects on prey popula-tions is not possible.

Tree removal, whether to harvest sawlogs orfuelwood, will affect natural ecosystem processesin numerous obvious and obscure ways. Cer-tainly, removal of mast-producing trees (e.g.,pinyon pine, juniper, oak) reduces food availabil-ity for several of the owl’s prey species. Also,removal of tree biomass from the site will inter-rupt both nutrient cycling and energy flow. Theeffects of altering these processes on owl preypopulations is unknown, but the disturbancemay likely benefit some species while negativelyaffecting others. Tree removal and accompanyingsite disturbance during and following removal,plus residual disturbance such as soil compac-tion, increased erosion, and creation of slashpiles, will directly alter habitats of many preyspecies. Block and Ganey (unpublished data)found more deer mice in areas with slash thanareas without slash. In the same general area,Goodwin and Hungerford (1979) noted thatbrush mice and Mexican woodrats used longwindrows of slash following logging. Block andGaney (unpublished data) sampled the sameareas 18-20 years after the Goodwin andHungerford study and captured few woodratsand brush mice. This suggests a temporarybenefit of slash piles in that they may provide ashort-term habitat component that is not used asthe slash becomes compressed and decomposes.

As noted above, effects of tree removal onsmall mammals varies by prescription and sitecharacteristics. The effects are somewhat scaledependent. Prescriptions for even-aged,shelterwood cuts or clearcuts effectively returnlarge areas, the size of stands or greater, to earlierseral stages. Small mammal community structure

40

Volume II/Chapter 5


is correlated to plant seral stage (Fox 1990,Kirkland 1990). Thus, following even-agedmanagement, populations of some species willrespond positively and others negatively. As timeprogresses and vegetation proceeds throughsuccession, population dynamics of species willchange in response to changing habitat condi-tions and community structure. Areas subjectedto even-aged management may not provideappropriate foraging habitat for the spotted owl(see Ganey and Dick 1995). This means changesto small mammal populations within the har-vested areas may not affect the owl prey basedirectly because those animals may not beavailable to the owl. However, the same smallmammal populations may provide a source ofindividuals that disperse into adjacent owlforaging habitat.

Uneven-aged management would likely beused over large areas and does not create smallstands, but rather it creates groups or clumps.Mosaics of habitat provide diverse plant commu-nities and other conditions that, collectively, cansupport a rich diversity of fauna. Populations ofdifferent species may respond variably to aspectsof the mosaic pattern, such as conditions withineach patch type, patch size and shape, andinterspersion and juxtaposition of these patches.Mosaic patterns resulting from timber manage-ment prescriptions such as single-tree or groupselection cuts may in some ways mimic naturaldisturbance patterns and create canopy gaps.This does not imply that effects of silvicultureare equivalent to those of natural disturbance,only that the resultant spatial patterns are some-what similar.

Clearly, research is needed to determinecause-effect relationships of tree removal onspotted owl prey populations, the hunting abilityof the owl, and the mosaic patterns which bestconserve owl populations. Such research willentail experiments conducted at varying spatialand temporal scales to understand the truemagnitude of the effects. Until these experi-ments are conducted, effects of tree removal onprey habitat and populations must be based onspeculation and conjecture. Future managementconducted within a scientific and experimentalcontext may provide a means for establishing

cause-effect relationships (see USDI 1995:Activity-Specific Research Section, Part III.E.).

GrazingGrazingGrazingGrazingGrazing

The effects of livestock and wildlife grazingon spotted owl prey populations and theirhabitats is also a complex issue. Impacts can varyaccording to grazing species; degree of use,including numbers of grazers, grazing intensity,grazing frequency, and timing of grazing; habitattype and structure; and plant or prey speciescomposition. It is well documented and intuitivethat repetitive, excessive grazing of plant com-munities by livestock can significantly alter plantspecies density, composition, vigor, regeneration,above or below ground phytomass, soil proper-ties, nutrient flow, water quality, and ultimatelylead to desertification when uncontrolled(Kauffman and Krueger 1984, Orodho et al.1990, Vallentine 1990, Milchunas andLauenroth 1993). These effects can have bothdirect and indirect adverse impacts on animalspecies that are dependent on plants for foodand cover. However, moderate to light grazingcan benefit some plant and animal species undercertain conditions and in certain environments,maintain communities in certain seral stages,and increase primary productivity (Reynolds1980, Hanley and Page 1982, Kauffman andKrueger 1984, McNaughton 1993). Further,direct influences of livestock on plant communi-ties are not always reflected in small mammalcommunities (Grant et al. 1982). Thus, anygeneralizations presented here should not beconstrued as absolute; there are exceptions.

No studies document the direct and indirecteffects of livestock and wildlife grazing on theMexican spotted owl or its prey (see reviews byUSDA Forest Service 1994, Utah MexicanSpotted Owl Technical Team 1994). We foundonly one study that specifically investigated theeffects of livestock grazing on an upland forestowl, the tawny owl (Putnam 1986). However,the design and limited sampling effort of thisstudy prevents extrapolation to the Mexicanspotted owl. Interpretive extrapolations arefurther hampered because most livestock-effectsstudies on small mammals have been conducted

41

Volume II/Chapter 5


in shrub-steppe, grassland, or riparian communi-ties of arid or semi-arid regions (Kauffman andKrueger 1984, Milchunas and Lauenroth 1993).Except for the possible use of these areas fordispersing or wintering, such areas are nottypically used by Mexican spotted owls.

Despite the dearth of specific informationabout spotted owls and grazing, there exists someknowledge regarding the effects of livestockgrazing on small mammals frequently consumedby spotted owls and regarding mesic or montaneplant communities inhabited by the owl’s prey.Relevant studies include Hanley and Page(1982), Medin and Clary (1990), Schultz andLeininger (1991), and Szaro (1991), and arebriefly summarized below.

In a 250,000-ha (619,250-ac) area of theGreat Basin, Hanley and Page (1982) found thatlivestock grazing (primarily cattle with anaverage of 0.16 animal unit month, April toOctober during each of four consecutive years)generally increased shrub composition anddecreased perennial herbs and grasses (primarilytall bunchgrasses). However, these effects de-pended upon community type. Shrubs increasedin grazed wet meadows and willow-ripariancommunities, whereas tree-form willows de-creased by 60%. Herbaceous layers were 30-50cm (11.8-19.7 in) deep in ungrazed meadowscompared to the “mowed” appearance of grazedmeadows. Reduction of graminoids and forbsincreased structural diversity in mesic habits byincreasing shrubs but decreased structuraldiversity in shrub-dominated, xeric habitats. Theauthors postulated that the loss of perennialgrasses and herbs caused lower numbers of desertwoodrats, and long-tailed and mountain voles ingrazed meadows by decreasing food sources andcover. The authors also reported a greater num-ber of Great Basin pocket mice, deer mice, andleast chipmunks in mesic habitats grazed bylivestock compared to ungrazed, mesic habitats.However, these same three species decreased inxeric habitats grazed by livestock.

Schultz and Leininger (1991) examined theeffects of cattle exclusion along one 5-km (3.1-mi) length of a riparian community innorthcentral Colorado. During the summers of1985 and 1986, plant and small mammalcommunities were sampled for comparison

between adjacent grazed areas and threeexclosures established in 1956 and 1959. Plantcommunities varied from grass-herb-sedgelowlands to ponderosa pine uplands. Observeddifferences included greater vascular vegetationcover, graminoid cover, and shrub cover in theexclosures. Litter accumulation in the exclosureswas nearly twice that of grazed areas, and willowcanopy cover was 8.5 times greater (Schultz1990). Deer mice were significantly moreabundant in grazed areas and western jumpingmice were significantly more abundant inungrazed exclosures. Seven species of smallmammals were found in grazed and exclosedareas, although species composition in each areawas slightly different. Deer mice, least chip-munks, masked shrews, dusky shrews, andwestern jumping mice were found in both areas.Long-tailed and mountain voles were not ob-served in grazed areas. Northern pocket gophersand golden-mantled ground squirrels were notfound in exclosures.

Medin and Clary (1990) compared smallmammal populations between a riparian habitatseasonally grazed by cattle to an adjacent 122-ha(302.2-ac) riparian exclosure which had beenprotected from livestock grazing for the previous14 years. They found that mountain voles werefour times more abundant within the ungrazedexclosure as compared to the grazed riparianhabitat. In addition, vagrant shrews, watershrews, and northern pocket gophers were onlytrapped within the ungrazed habitat. Conse-quently, they noted that small mammal speciesrichness and species diversity were higher withinthe ungrazed exclosure despite the fact that totalpopulation density of small mammals was athird higher in the grazed portion due to thehigh numbers of deer mice.

Szaro (1991) examined the effects of grazingalong the Rio de las Vacas, a montane stream at8,528 ft (2,600 m) elevation in the San PedroParks Wilderness Area, New Mexico. Terrestrialfauna within two 900 x 50 m (2,952 x 164 ft)livestock exclosures were sampled and comparedto downstream private lands that were continu-ously grazed. Small mammals were sampled eachmonth, June through September, in 1985 and1986. He found greater amounts of herbaceousvegetation in the exclosures. Greater numbers of

42

Volume II/Chapter 5


small mammal captures were observed in bothexclosures in 1985 compared to grazed areasdownstream. Small mammal captures werereduced and indistinguishable in 1986 but somecattle trespass had occurred into the controlexclosures. Fewer species were captured in thegrazed areas. Mountain voles and deer mice werefound in both grazed and ungrazed situations,whereas least chipmunks and golden-mantledground squirrels were found only in ungrazedareas.

Other studies have shown similar results:lack of a numerical decrease by deer mice tograzing (Reynolds 1980), and significant de-crease in voles caused by grazing induced loss ofcover in mesic habitats (Grant et al. 1982). Ifthese general patterns can be applied to uplandhabitats of the Southwest, we would expectmoderate to heavy grazing to decrease popula-tions of voles and improve conditions for deermice in meadow habitats. Such decreases couldnegatively influence spotted owls occupyingareas in the Upper Gila Mountains, Basin andRange - East, and portions of other RUs wherevoles are common prey or used as alternativefood sources when other prey species are dimin-ished.

Increases in deer mouse abundance inmeadows probably would not offset decreases invole numbers because voles provide greaterbiomass per individual and per unit of area. Lossof perennial grass cover in xeric communitiesused by owls, specifically, ponderosa pine forestand pinyon-juniper woodlands, due to grazingmay reduce deer mice and Mexican vole popula-tions. The reduction of these prey species inxeric communities could be more critical thanthat in meadows if xeric habitats are necessaryfor winter foraging by the owls.

Finally, high intensity grazing in ripariancommunities during the fall and winter seasonswhere grass seedheads may be totally removedcan cause significant short-term decreases insmall mammal populations (Kauffman et al.1983). Continued heavy grazing in upland orlowland riparian communities could thereforegreatly reduce the potential for utilization byforaging, dispersing or wintering spotted owls.


Mexican spotted owls consume a variety ofprey throughout their range but commonly eatsmall and medium-sized rodents. However, theowl’s food habits vary according to geographiclocation. For example, spotted owls dwelling incanyons of the Colorado Plateau take morewoodrats, and fewer voles and birds than dospotted owls from other areas. In contrast,spotted owls occupying mountain ranges withforest-meadow interfaces, such as the Basin andRange - East, Southern Rocky Mountains -Colorado, and Upper Gila Mountains RUs, takemore microtine voles The differences in dietlikely reflect geographic variation in populationdensities and habitats of both the prey and theowl.

No strong rangewide relationships appearedin our analyses of the owl’s diet and reproduc-tion. The relationship was positive and nearlysignificant when comparing the prevalence ofthe three most common prey (peromyscid mice,woodrats, and voles) in the diet and owl repro-duction, implying that multiple species influencethe owl's fitness. However, this generalizationmay not apply to owls in the Sacramento Moun-tains where the owl's reproduction appears mostinfluenced by deer mouse abundance. In addi-tion, the predominance of woodrats, both in dietfrequency and biomass throughout much of theowl’s range, suggests that a single prey mayinfluence the owl’s fitness. Other studies haveshown positive associations between larger prey(e.g. woodrats) in the diet of northern andCalifornia spotted owls and its reproductivesuccess (Barrows 1987, Thrailkill and Bias1989). The lack of more specific results shouldnot imply that a simple relationship betweendiet and reproduction or other fitness measuresdoes not exist. Rather, those relationships aremore complex than are evident by the availableinformation and analytical approaches used inproducing this report. In most cases, total preybiomass is likely more influential on the owl’sfitness than the abundance of any particular preyspecies. Other factors worth exploring would belag effects such as the effects of winter diet onbreeding potential, prey diversity, or synergistic

43

Volume II/Chapter 5


effects of diet with factors like owl density,weather, and habitat.

Studies conducted in four vegetation com-munities demonstrate that abundance of theowl’s food varies according to habitat and time.In the Sacramento Mountains, the greatest preybiomass is found in high-elevation meadowsoccurring along riparian corridors. Commonprey species occupying these meadows are long-tailed voles, Mexican voles, and deer mice.However, abundance alone does not necessarilyconnote availability. Availability infers co-occurrence of owl and prey plus coincidentalvulnerability to predation and ability to capture.Successful capture of meadow-dwelling rodentsmay be restricted to areas near forest edgescoinciding with the presence of foraging perches.Rather, meadow habitats may play an indirectrole by producing high densities of prey thatbecome available to owls following dispersal intoadjacent forests. Owls in the Sacramento Moun-tains consume a moderate-to-large proportion ofvoles during years of high vole density.

Summer prey biomass in mesic (mixed-conifer) forests can be greater than in xeric(ponderosa pine-pinyon-juniper-oak) forests ofthe Sacramento Mountains for two reasons.First, all five prey species common to this owloccur in mesic forest, whereas only four occur inxeric forests, where long-tailed voles are absent.This vole can provide an average mass of 32 g(1.1 oz) to an owl and it is the second mostabundant species occurring in the mesic forests.Second, deer mice dwelling in mesic forest canattain great summer densities during certainyears. This same pattern has not been observedin the seemingly more stable xeric forests. Preycomposition and abundance in ponderosa pine-Gambel oak forests of northern Arizona aresimilar to the xeric forests of the SacramentoMountains. In both areas, deer mice are ubiqui-tous and the Mexican woodrat and brush mouseare patchy in distribution and abundance. Incontrast, Mexican voles are apparently lessabundant in pine-oak forests of northern Ari-zona compared to xeric forests in the SacramentoMountains. Whether low densities of voles inthis forest type are natural or the result of pastmanagement activities is unknown. However,decreases in herbaceous biomass resulting from

unnaturally dense forest conditions may explainlow numbers of voles in these forests.

Arranging the four vegetation communitiesaccording to summer prey biomass indicates thatmeadows > mixed-conifer forest > ponderosapine-pinyon-juniper-oak woodlands > ponderosapine-Gambel oak forest. When consideringother factors that influence the availability ofprey, mixed-conifer forests likely provide thegreatest amount of food during summer periods.Rearranging the same communities according towinter prey biomass indicates that meadows >ponderosa pine-pinyon-juniper-oak woodlands >ponderosa pine-Gambel oak forest > mixed-conifer forest. Accounting for the availability ofprey, woodlands with a mixture of ponderosapine, pinyon-juniper, and oaks provide moreprey to owls during winter months than theother three communities.

Prey biomass is usually greater in all commu-nities during summer periods when owls raisetheir young. However, temporal peaks of preycycles are not correlated among vegetationcommunities. That is, when prey are abundantin mixed-conifer forest one year and low asubsequent year, an opposite pattern may occurin a different vegetation community. Theasynchrony of abundance among prey species,vegetation communities, and time may provide abuffer against the effects of extreme oscillation inprey cycles. This implies the importance ofmaintaining a mixture of vegetation communi-ties to ensure a diverse and abundant prey basewithin the owls foraging range.

An important concept exemplified by ouranalyses is that the habitat of each prey species isunique. This finding clearly indicates a need forproviding a variety of conditions which are usedby the different species of prey. For example, inthe ponderosa pine-Gambel oak forests ofnorthern Arizona, deer mouse abundance showslittle variation according to forest structure andcomposition whereas Mexican woodrat andbrush mouse abundance are strongly correlatedto understory characteristics, specifically logvolume and shrub cover. Further, Gambel oakdensity is greater within habitats of the woodratand brush mouse than occurs randomly in theforest. These habitat components are rarelyconsidered in planning forest management

44

Volume II/Chapter 5


activities but should be to provide appropriatehabitat conditions for these prey species. Obvi-ously, conserving habitat for a diversity of preymay help buffer against population fluctuationsof individual prey species and provide a lessstochastic food resource for the owl.

The consequences of three common distur-bances (fire, timber/fuelwood harvest, andgrazing) on the owl’s prey and habitat dependson many factors. Often ecological tradeoffsresult, making exact predictions difficult. Someprey species may increase, while others decreasefor a given disturbance. More detailed predic-tions about the influences of these disturbancesmust await more specific research. In general,management practices that lead to discerniblereductions in total prey biomass or diversity overlarge areas will not promote owl recovery.


Armstrong, D. M. 1972. Distribution of mam-mals in Colorado. Univ. Kansas Mus.Nat. Hist. Mongr. 3:1-415.

——.1977. Ecological distribution of smallmammals in the Upper Williams ForkBasin, Grand County, Colorado. South-west. Nat. 22:289-304.

——.1979. Ecological distribution of rodents inCanyonlands National Park, Utah. GreatBasin Nat. 39:199-205.

Bailey, V. 1931. Mammals of New Mexico. N.Amer. Fauna 53:1-412.

Barrows, C. W. 1987. Diet shifts in breeding andnonbreeding spotted owls. J. Raptor.Res. 21:95-97.

Campbell, R. E., M. B. Baker, Jr., P. F. Ffolliott,R. R. Larson, and C. C. Avery. 1977.Wildfire effects on a ponderosa pineecosystem: an Arizona case study. USDAFor. Serv. Res. Pap. RM-191. 12pp.

Carey, A. B. 1985. A summary of the scientificbasis for spotted owl management. Pages50-54 in R. J. Gutiérrez, and A. B.Carey, eds. Ecology and management ofthe spotted owl in the Pacific Northwest.USDA For. Serv. Gen. Tech. Rep. PNW-185. 118pp.

Conley, W., and J. D. Nichols. 1978. The use ofmodels in small mammal population

studies. Pages 14-35 in D. P. Synder, ed.Populations of small mammals undernatural conditions. Spec. Publ. No. 5,Pymatunigh Laboratory of Ecology,Univ. of Pittsburg, Penn.

Cornely, J. E., and R. J. Baker. 1986. Neotomamexicana. Mammal. Species 262:1-7.

——., D. J. Schmidly, H. H. Genoways, and R.J. Baker. 1981. Mice of the GenusPeromyscus in Guadalupe National Park,Texas. Occas. Papers Mus. Texas TechUniv. 74:1-35.

Craighead, J. J., and F. C. Craighead. 1956.Hawks, owls and wildlife. Stackpole,Harrisburg, Penn. 443pp.

DeRosier, S., and J. P. Ward, Jr. 1994. Protocolsfor analyzing pellets regurgitated byMexican spotted owls. Unpubl. Rep.,USDA Forest Service, Rocky MountainResearch Station, Fort Collins, Colo-rado. 13pp.

Draper, N. R., and H. Smith. 1981. Appliedregression analysis. Second ed. JohnWiley and Sons, New York, N.Y. 709pp.

Duncan, R. B., and R. Sydner. 1990. Bats inspotted owl pellets in southern Arizona.Great Basin Nat. 50:197-200.

Durrant, S. D. 1952. Mammals of Utah, tax-onomy and distribution. Univ. KansasPubl., Mus. Nat. Hist. 6:1-549.

Errington, P. L. 1930. The pellet analysismethod of raptor food habits study.Condor 32:292-296.

Findley, J. S., and C. J. Jones. 1982. Distribu-tion and variation of voles of the genusMicrotus in New Mexico and adjacentareas. J. Mamm. 43:154-166.

——., A. H. Harris, D. E. Wilson, and C. Jones.1975. Mammals of New Mexico. Univ.of New Mexico Press, Albuquerque,N.M. 360pp.

Finerty, J. P. 1980. The population ecology ofsmall mammals: mathematical theoryand biological fact. Yale Univ. Press, NewHaven, Conn. 234pp.

Finley, R. B., Jr. 1958. The wood rats of Colo-rado: distribution and ecology. Univ.Kansas Mus. Nat. Hist. Pub. 10:213-552.

45

Volume II/Chapter 5


——., J. R. Choate, and D. F. Hoffmeister.1986. Distributions and habitats of volesin southeastern Colorado and northeast-ern New Mexico. Southwest. Nat.31:263-266.

Forsman, E. D. 1983. Methods and materials forlocating and studying spotted owls. U.S.For. Serv. Gen. Tech. Rep. PNW-162.8pp.

——., C. Meslow, and H. Wight. 1984. Distri-bution and biology of the spotted owl inOregon. Wildl. Monogr. 87. 64pp.

Fox, B. J. 1990. Changes in the structure ofmammal communities over successionaltime scales. Oikos 59:321-329.

Freeman, M. F., and J. W. Tukey. 1950. Transfor-mations related to the angular and thesquare root. Ann. Math Statist. 21:607-611.

Frey, J. A., and C. T. LaRue. 1993. Notes on thedistribution of the mogollon vole(Microtus mogollonensis) in New Mexicoand Arizona. Southwest. Nat. 38:176-178.

Ganey, J.L., and J.L. Dick 1995. Habitat rela-tionships of the Mexican spotted owl.Chapter 4 (42 pp.) in Recovery plan forthe Mexican spotted owl. Vol. II. USDIFish and Wildl. Serv., Albuquerque,NM.

——. 1992. Food habits of Mexican spottedowls in Arizona. Wilson Bull. 104:321-326.

Getz, L. L. 1985. Habitats. Pages 287-309 in R.H. Tamarin. ed.. Biology of new worldMicrotus. Amer. Soc. MammalogistsSpec. Publ. No. 8. 839pp.

Glading, B. D., F. Tillotson, and D. M. Selleck.1943. Raptor pellets as indicators of foodhabits. Calif. Fish and Game 29:92-121.

Goodman, L. A. 1960. On the exact variance ofproducts. J. Amer. Stat. Assoc. 55:708-713.

Goodwin, J. G., Jr., and C. R. Hungerford.1979. Rodent population densities andfood habitats in Arizona ponderosa pineforests. USDA For. Serv. Res. Pap. RM-214. 12pp.

Grant, W. E., E. C. Birney, N. R. French, andD. M. Swift. 1982. Structure and pro-

ductivity of grassland small mammalcommunities related to grazing inducedchanges in vegetative cover. J. Mammal.63:248-260.

Hall, E. R., and K. R. Kelson. 1959. The mam-mals of North America. Ronald Press,Co., N.Y. 2 vols. 1083pp.

Hanley, T. A., and J. L. Page. 1982. Differentialeffects of livestock use on habitat struc-ture and rodent populations in GreatBasin communities. Calif. Fish andGame 68:160-174.

Hirons, G. J. M. 1985. The importance of bodyreserves for successful reproduction inthe Tawny owl (Strix aluco). J. Zool.Lond. (B) 1:1-20.

Hoffmeister, D. F. 1986. Mammals of Arizona.Univ. of Arizona Press, Tucson, Ariz.602pp.

Johnsgard, P. A. 1988. North American owls:biology and natural history. SmithsonianInst. Press, Washington, D.C. 295pp.

Johnson, D. W., and D. M. Armstrong. 1987.Peromyscus crinitus. Mammal. Species287:1-8.

Kauffman, J. B., and W. C. Krueger. 1984.Livestock impacts on riparian ecosystemsand streamside management implica-tions, a review. J. Range Manage.37:430-438.

——., ——., and M. Vavra. 1983. Effects of lateseason cattle grazing on riparian plantcommunities. J. Range Manage. 36:685-691.

Kaufman, G. A., D. W. Kaufman, and E. J.Finck. 1988. Influence of fire andtopography on habitat selection byPeromyscus maniculatus andReithrodontomys megalotis in ungrazedtall-grass prairie. J. Mammal. 66:342-352.


Kirkland, G. L., Jr. 1990. Patterns of initialsmall mammal community change afterclearcutting of temperate North Ameri-can forests. Oikos 59:313-320.

Korpimäcki, E., and K. Norrdahl. 1991. Nu-merical and functional responses of

46

Volume II/Chapter 5


kestrels, short-eared owls, and long-earedowls to vole densities. Ecology 72:814-826.

Lundberg, A. 1976. Breeding success and preyavailability in a Ural owl (Strix uralensisPall.) population in central Sweden.Zoon 4:65-72.

Marti, C. D. 1987. Raptor food habits studies.Pages 67-80 in B. A. Giron Pendelton,ed. Raptor Management TechniquesManual. National Wildlife Institute,Washington, D.C. 394pp.

McNaughton, S. J. 1993. Grasses and grazers,science and management. Ecol. Applic.3:17-20.

Medin, D. E., and W. P. Clary. 1990. Bird andsmall mammal populations in a grazedand ungrazed riparian habitat in Idaho.USDA For. Serv. Res. Pap. INT-425.8pp.

Milchunas, D. G., and W. K. Lauenroth. 1993.Quantitative effects of grazing on vegeta-tion and soils over a global range ofenvironments. Ecol. Monogr. 63:327-366.

Newton, I. 1979. Population ecology of raptors.Vermillion, S.D. 399pp.

Orodho, A. B., M. J. Trlica, and C. D. Bonham.1990. Long-term heavy-grazing effectson soil and vegetation in the four cornersregion. Southwest. Nat. 35:9-14.

Otis, D. L., K. P. Burnham, G. C. White, andD. R. Anderson. 1978. Statistical infer-ence from capture data on closed animalpopulations. Wildl. Monogr. 62. 135pp.

Peterson, S. K., G. A. Kaufman, and D. W.Kaufman. 1985. Habitat selection bysmall mammals of the tall-grass prairie:experimental patch choice. Prairie Nat.17:65-70.

Putnam, R. J. 1986. Grazing in temperateecosystems, large herbivores and theecology of the New Forest. Timber Press,Portland, Oreg. 210pp.

Reichenbacher, F. W., and R. B. Duncan. 1992.Diet of Mexican spotted owls on Na-tional Forest system lands in Arizona andNew Mexico. Unpubl. Rep., USDA For.Serv., Southwestern Regional Office,

Albuquerque, N.M. 77pp.Reynolds, R. T., R. T. Graham, M. H. Reiser, R.

L. Bassett, P. L. Kennedy, D. A. Boyce,Jr., G. Goodwin, R. Smith, and E. L.Fisher. 1992. Management recommenda-tions for the northern goshawk in thesouthwestern United States. USDA For.Serv. Gen. Tech. Rep. RM-217. 90pp.

Reynolds, T. D. 1980. Effects of some differentland management practices on smallmammal populations. J. Mammal.61:558-561.

Ribble, D. O., and F. B. Samson. 1987. Micro-habitat associations of small mammals insoutheastern Colorado, with specialemphasis on Peromyscus (Rodentia).Southwest. Nat. 32:291-303.

Rinkevich, S. E. 1991. Distribution and habitatcharacteristics of Mexican spotted owlsin Zion National Park, Utah. M.S.Thesis. Humboldt State Univ., Arcata,Calif. 62pp.

Schulz, T. T., and W. C. Leininger. 1990. Differ-ences in riparian vegetation structurebetween grazed areas and exclosures. J.Range Management 43:294-298.

——. 1991. Nongame wildlife communities ingrazed and ungrazed montane ripariansites. Great Basin Nat. 51:286-292.

Southern, H. N. 1970. The natural control of apopulation of Tawny owls (Strix aluco). J.Zool. 162:197-285.

Svoboda, P. L., D. K. Tolliver, and J. R. Choate.1988. Peromyscus boylii in the San LuisValley, Colorado. Southwest. Nat.33:239-240.

Szaro, R. C. 1991. Wildlife communities ofsouthwestern riparian ecosystems. Pages174-201 in Rodiek, J. E., and E. G.Bolen eds. Wildlife and habitats inmanaged landscapes. Island Press, Wash-ington, D.C. 219pp.

Tarango, L. A. 1994. Mexican spotted owl distribu-tion and habitat characterizations in south-western Chihuahua, Mexico. M.S. Thesis,New Mexico State Univ., Las Cruces, N.M.

Thomas, J. W., E. D. Forsman, J. B. Lint, E. C.Meslow, B. R. Noon, and J. Verner. 1990.A conservation strategy for the spotted owl.U.S. Gov. Print. Off., Washington, D.C.427pp.

47

Volume II/Chapter 5


Thrailkill, J., and M. A. Bias. 1989. Diets of breed-ing and nonbreeding California spottedowls. J. Raptor Res. 23:39-41.

USDA Forest Service. 1994. Biological assessmentand evaluation: livestock grazing. Unpubl.Rep., USDA For. Serv., Southwest Region,Albuquerque, N.M. 12pp.

USDI Fish and Wildlife Service. 1995. RecoveryPlan for the Mexican Spotted Owl. Vol. 1.USDI Fish and Wildl. Serv., Albuquerque,N.M. 178pp.

Utah Mexican Spotted Owl Technical. Team. 1994.Suggestions for management of Mexicanspotted owls in Utah. Utah Div. of Wildl.Resour., Salt Lake City, Utah. 103pp.

Vallentine, J. F. 1990. Grazing management. Aca-demic Press, Inc., San Diego, Calif. 533pp.

Verner, J., K. S. McKelvey, B. R. Noon., R. J.Gutiérrez, G. I. Gould, Jr., and T. W. Beck,eds. 1992. The California spotted owl: atechnical assessment of its current status.USDA For. Serv. Gen. Tech. Rep. PSW-133.285pp.

——., M. L. Morrison, and C. J. Ralph. 1986.Wildlife 2000: modeling habitats of terres-trial vertebrates. Univ. Wisc. Press, Madison,Wisc. 470pp.

Wagner, P. W., C. D. Marti, and T. C. Boner. 1982.Food of the spotted owl in Utah. J. RaptorRes. 16:27-28.

Ward, J. P., Jr. 1990. Spotted owl reproduction, dietand prey abundance in northwest California.M.S. Thesis, Humboldt State Univ., Arcata,Calif. 70pp.

Wendland, V. 1984. The influence of prey fluctua-tions on the breeding success of the Tawnyowl (Strix aluco). Ibis 126:285-295.

White, G. C., D. R. Anderson, K. P. Burnham, andD. L. Otis. 1982. Capture-recapture andremoval methods for sampling closedpopulations. USDE Los Alamos NationalLab. Rep. LA-8787-NERP. Los Alamos,N.M. 235pp.

Wilson, D. E. 1968. Ecological distribution of thegenus Peromyscus. Southwest. Nat. 13:267-274.

APPENDIX 5APPENDIX 5APPENDIX 5APPENDIX 5APPENDIX 5aaaaa

Sources for estimates of prey mass used in calculatingpercent prey biomass in Mexican spotted owl diets.

Anderson, S. 1972. Mammals of Chihuahua,taxonomy and distribution. Bull. Amer.Museum Nat. Hist. 148:149-410.

Armstrong, D. M. 1972. Distribution of mammalsin Colorado. Mongr. Univ. Kansas Mus.Nat. Hist. 3:1-415.

Berna, H. J. 1990. Seven bat species from the KaibaPlateau, Arizona, with a record of Eudermamaculatum. Southwest. Nat. 35:354-356.

Block, W. M., and J. L. Ganey. Unpublished smallmammal masses collected in NorthernArizona (1990-1992). USDA For. Serv.,Rocky Mtn. For. Range Exp. Sta., Flagstaff,Arizona.

Bogan, M., and C. Ramotnik. Unpublished smallmammal masses collected in Utah (1990-1992). USDI Fish and Wildl. Serv., Mus. ofSouthwest Biol., Univ. of New Mexico,Albuquerque, N.M.

Burt, W. H., and R. P. Grossenheider. 1976. A fieldguide to the mammals. Hougton MifflinCo., Boston, Mass. 289 pp.

Chapman, J. A., and G. A. Feldhammer. 1982. Wildmammals of North America. JohnsHopkins Univ. Press, Baltimore, Maryl.1147 pp.

Dunning, J. B. ed. 1993. CRC handbook of avianmasses. CRC Press, Boca Raton, Flor. 371pp.

Ganey, J. L. Unpublished masses used to determineprey biomass in diet of Mexican spotted owlspresented in 1992 Wilson Bull. publication.

Johnsgard, P. A. 1988. North American owls, biologyand natural history. Smithsonian InstitutionPress, Wash. D. C. 285 pp.

Marti, C. D. 1974. Feeding ecology of four sympat-ric owls. Condor 76:45-61.

Severson, K. E., and B. J. Hayward. 1980. Rodentweights in modified pinyon-juniper wood-lands of southwestern New Mexico. GreatBasin Nat. 48:554-557.

Steenhof, K. 1983. Prey weights for computingpercent biomass in raptor diets. Raptor Res.17:15-27.

Ward, J. P., Jr., S. DeRosier, and W. M. Block.Unpublished masses of small mammalscollected in the Sacramento Mountains,New Mexico (1991-1993). Rocky Mtn. For.Range Exp. Sta., Ft. Collins, Colo.

48