A Guide to AIM-NAMF Indicators, their Computation, and ... · Biodiversity and Riparian Habitat Indicators ... Riparian Habitat, and Biodiversity Benchmark Development ... NAMF indicator

1

A Guide to AIM-NAMF Indicators, their Computation, and Example Applications

DRAFT – DO NOT CITE

Date Compiled: November 8, 2017

Table of Contents Introduction .................................................................................................................................................. 3

Intended Applications ............................................................................................................................... 3

AIM-NAMF Analysis and Reporting Workflow .......................................................................................... 4

Field Method Overview ............................................................................................................................. 5

Indicators for Lotic Systems .......................................................................................................................... 9

Overview ................................................................................................................................................... 9

Water Quality .......................................................................................................................................... 11

Background ......................................................................................................................................... 11

Example Applications (the big picture) ............................................................................................... 11

Indicators ............................................................................................................................................ 11

pH .................................................................................................................................................... 11

Specific Conductance ...................................................................................................................... 12

Water Temperature ........................................................................................................................ 12

Total Nitrogen and Phosphorous .................................................................................................... 13

Turbidity .......................................................................................................................................... 13

Precision of Water Quality Indicators ................................................................................................. 13

Watershed Function – Instream Habitat ................................................................................................ 14

Background ......................................................................................................................................... 14

Example Applications (the big picture) ............................................................................................... 15

Stream Channel Form, Function, and Habitat Indicators ................................................................... 15

Pool depth, length, and frequency ................................................................................................. 15

Thalweg ........................................................................................................................................... 17

Instream Habitat Complexity .......................................................................................................... 17

Large Woody Debris ........................................................................................................................ 19

Streambed Particle Sizes ................................................................................................................. 21

Relative Bed Stability ...................................................................................................................... 22

2

Bank Stability and Cover ................................................................................................................. 23

Bank Angle ...................................................................................................................................... 24

Floodplain Connectivity .................................................................................................................. 24

Precision of Watershed Function – Instream Habitat Indicators ........................................................ 25

Biodiversity and Riparian Habitat ........................................................................................................... 26

Background ......................................................................................................................................... 26

Example Applications (the big picture) ............................................................................................... 26

Biodiversity and Riparian Habitat Indicators ...................................................................................... 26

Benthic Macroinvertebrates ........................................................................................................... 26

Percent Canopy Cover ..................................................................................................................... 29

Riparian Habitat Complexity and Cover .......................................................................................... 29

Precision of Biodiversity and Riparian Habitat Indicators .................................................................. 33

Covariates ............................................................................................................................................... 34

Background and Example Applications ............................................................................................... 34

Covariate Description and Computations ........................................................................................... 34

Bankfull Width and Wetted Width ................................................................................................. 34

Flood-prone Width .......................................................................................................................... 34

Slope ................................................................................................................................................ 35

Sinuosity .......................................................................................................................................... 35

Precision of Covariates ........................................................................................................................ 35

Benchmarks: From Indicator Values to Management Decisions ................................................................ 36

What are Benchmarks and Why are They Needed? ............................................................................... 36

Approaches to Setting Benchmarks ........................................................................................................ 36

Limitations to Benchmark Approaches ................................................................................................... 38

Understanding Benchmarks Available in the Benchmark Tool ............................................................... 38

Water Quality Benchmark Development ............................................................................................ 40

Watershed Function, Instream Habitat, Riparian Habitat, and Biodiversity Benchmark Development ............................................................................................................................................................ 41

Appendix A. BLM AIM AquADat Local Feature Class Metadata ................................................................. 51

Appendix B. Macroinvertebrate Model Metadata ..................................................................................... 59

Literature Cited ........................................................................................................................................... 60

3

Introduction

Intended Applications

This document is intended to help BLM resource specialists understand the types of indicators resulting from collection of the Assessment, Inventory, and Monitoring (AIM) field methods for wadeable lotic systems (TR 1735-2). It includes descriptions of how to calculate, interpret, and apply individual indicators to assess resource condition and trend, as well as discussions of indicator limitations. The provided information can be used by practitioners prior to collecting aquatic AIM data, such as when making decisions about what contingent indicators to collect, or as a reference document during data analysis and reporting. The content was developed with the assumption that users will have some level of familiarity with the AIM-National Aquatic Monitoring Framework (NAMF) and the associated field methods (TR 1735-1 and TR 1735-2). Therefore, we do not provide a rationale for why the core and contingent indicators were selected or detailed descriptions of field methods. If you are brand new to the AIM-NAMF, the utility of this document will be increased by a review of the following three resources:

1. Technical Reference 1735-1: provides a detailed rationale for the development of the AIM-NAMF and the selection of core and contingent indicators

2. Technical Reference 1735-2: detailed field method protocol for wadeable lotic systems

3. AIM Implementation Website: walks practitioners through a series of implementation steps for AIM project from planning to design and analysis

The technical reference can be read cover to cover as a guide to the detailed metadata for AIM-NAMF indicator computation and application, but its greatest values is as a quick reference guide. Below are examples of using the document as a quick reference guide:

1. Indicator applications and computation: If you want to understand applications or computation specifics for an indicator of interest, use the table of contents to quickly reference the page number for the indicator of interest. Note that the indicators are organized by the BLM’s Fundamental of Land Health.

2. Brief description of core and contingent methods and indicators: If you do not have much experience with the AIM-NAMF methods or indictors, this document will provide a brief description of each method and the associated indicators (see pages 5-10 and Appendix A).

3. Contingent indicator selection: If you are just starting an aquatic AIM project, you can

use this document to determine which contingent indicators you may want to collect in addition to the core indicators (see the table of contents for indicator page numbers).

4. Determining benchmarks: If you are using the Benchmark Tool to make condition determinations and you want to better understand the default benchmarks, how the

4

benchmark are being applied in the tool, and whether they are appropriate, see pages 34-50.

5. Limitations of indicators to assess resource condition and trend: If you want to

understand indicator precision, the amount of departure from a benchmark you will be able to detect, and how long it might take to determine if conditions have changed, see the indicator precision descriptions under each indicator of interest. Note that this content is still in development.

AIM-NAMF Analysis and Reporting Workflow

The AIM analysis and reporting workflow consists of three main steps: preparing for analysis, conducting analysis, and interpreting results (Fig. 1). Understanding this general process and the resources available to assist with analysis and reporting will facilitate more effective and efficient communication among interdisciplinary teams making management decisions. This document addresses steps associated with ‘preparing for analysis’. Specifically, step 2 (indicator computation), step 3 (develop benchmarks), and step 5 (understand pertinent data) (Fig. 1). These steps collectively work toward determining the condition of an individual stream reach, which might be the end goal of an analysis. However, if one seeks to combine condition estimates for multiple stream reaches to make inference to a larger area (e.g., BLM perennial streams in the Price Field Office, UT), more complex statistical analyses are required and we recommend consultation with the National AIM team (steps 9 and 10). When preparing for and conducting analyses with AIM-NAMF data, there are two tools that can be of assistance: AquADat and the AIM-NAMF Benchmark Tool. AquADat is the BLM’s national database for AIM-NAMF, which stores the indicators described herein. AIM-NAMF data can be obtained for analysis from AquADat using instructions linked here. Practitioners can then pair the data downloaded from AquADat with other analysis tools, such as the AIM-NAMF Benchmark Tool. The Benchmark Tool assists with assigning default (described in the Benchmark Section) or custom benchmarks for determining the condition of individual sampled stream reaches. Instructions for using the Benchmark Tool are contained within the tool.

5

Figure 1. Analysis and reporting workflow to support assessments of resource condition and trend using the AIM Strategy.

Field Method Overview

Technical reference 1735-1 identified core and contingent methods that should be used for AIM monitoring. Collectively, the core and contingent methods and associated indicators provide multiple lines of evidence for quantifying the chemical, physical, and biological condition and trend of lotic resources. The core methods represent the minimum measurements required to reporting on the attainment of BLM lotic fundamentals of rangeland health in a quantitative manner (i.e., water quality, watershed function and instream habitat, biodiversity and riparian habitat quality, and ecological processes). Beyond the BLM fundamental of land health, the core and contingent indicators have broad applicability to monitoring and assessment needs, including Clean Water Act attainment, the establishment of baseline conditions, assessing the efficacy of restoration or reclamation actions, and resource management plan (RMP) effectiveness monitoring. The use of AIM monitoring to evaluate RMP effectiveness was codified in IM 2016-139 and all three of the Bureau’s aquatic programs (Riparian, Water Resources, and Fisheries and Aquatic Resources) are working to integrate AIM monitoring methods into their manuals and handbooks. These AIM monitoring methods include 11 core methods for wadeable perennial streams and rivers specified in the field protocol (TR-1735-2) to characterize water quality (e.g., pH), instream habitat (e.g., bank stability and cover), and biodiversity and riparian habitat quality (e.g. macroinvertebrate biological integrity) (Table 1). The core methods can be complemented by a set of contingent field methods (e.g., nutrients, turbidity, bank angle, vegetative cover, composition, and structure) to address more specific management questions, as well as several

6

covariates (e.g., flood-prone width, slope, and sinuosity) to help determine the potential of a stream or river to support a set of water quality, geomorphic, or biological conditions. The general field sampling design is to collect data along a length of stream called a reach. Reach lengths are scaled to by the size of the stream and are measured as 20 times the average bankfull width, with a minimum of 150 m and a maximum of 4 km. Along the sample reach, 21 equally spaced transects consisting of 11 main transects and 10 intermediate transects, oriented perpendicular to the thalweg, are temporarily established (Fig. 2). Field measurements consist of individual point measurements made at the center of the reach (e.g., water quality indicators), measurements at each of the 11 main transects (e.g., canopy cover) or all 21 transects (e.g., bank stability and cover), or measurements taken throughout the entire reach (e.g., slope and large woody debris) (Table 1). All field measurements are taken during base flow conditions between June 1st and September 30th (i.e., index period). For detailed descriptions of indicator specific field methods, refer to TR 1735-2 (Miller et al., 2017). AIM monitoring and assessments generally seek to make inference to an individual stream reach or a population of stream reaches through use of statistically appropriate sample designs (i.e., site selection). Regardless of the scope of inference, the unit of replication is the stream reach, and multiple sample reaches or visits per sample reach are required to derive average indicator estimates and associated confidence intervals. Thus, where the field protocol prescribes multiple measurements for a given indicator throughout a reach (Table 1), the intent is to improve the accuracy of reach-scale indicator values (e.g., bank stability), and the individual measurements are not intended as statistical replicates. The use of multiple measurements per reach as replicates is subject to pseudo-replication, where the replicates are not statistically independent of each other (Hurlbert 1984). Pseudo-replication can lead to artificially low variance estimates and the detection of differences when they really do not exist (i.e., type II errors).

Figure 2. Typical reach setup with main transects (A–K; black lines) and 10 intermediate transects (gray lines) oriented perpendicular to the thalweg. Reach lengths are equal to 20 x bankfull width or a minimum of 150 m and a maximum of 4 km. Circled letters represent alternating benthic macroinvertebrate sampling locations for the reachwide protocol (USEPA 2009).

7

Table 1. Summary of core, contingent, and covariate field methods. Fundamentals of land health Field method Detailed field method

Bio

dive

rsity

/ rip

aria

n ha

bita

t Macroinvertebrate biological integrity 8 Surber/kicknet from multiple fast water habitats or 1 Surber/kicknet per each of 11 transects

Ocular est. of riparian vegetative type, cover, and structure

Ocular cover, structure, and type estimates for left and right bank plots at 11 transects

Quantitative estimates of riparian vegetative composition and cover1 Left and right bank greenline-based quadrats.2

Canopy cover 6 densiometer readings at 11 transects (four measurements at mid-channel and one at each bank)

Wat

er q

ualit

y

pH In-situ: multi-parameter sonde

Specific conductance In-situ: multi-parameter sonde

Temperature In-situ: multi-parameter sonde or thermistor

Total nitrogen and phosphorous1 Grab sample for lab analysis

Turbidity1 In-situ turbidometer or grab sample for lab analysis

Wat

ersh

ed fu

nctio

n - i

nstre

am h

abita

t

Pool depth, length, and frequency Measure all qualifying pools within the reach

Large woody debris (LWD) Size class counts of qualifying LWD over the entire reach

Bed particle size distribution 10 substrate particles measured at equal distances across the active channel (scour line to scour line) at 21 transects

Bank stability and cover

Left and right bank at 21 transects. Combination of bank type (erosional/depositional), bank cover (50% covered or not), and unstable bank features (fracture, slumping, sloughing, eroding)

Floodplain connectivity Bankfull and floodplain height measured at 11 transects

Instream fish habitat complexity1 Ocular estimates of instream of concealment features (e.g., LWD, veg., undercuts, boulders) at 11 plots

Bank angle1 Left and right bank at 11 transects

Thalweg depth profile1 100+ inter-transect measurements of thalweg water depth

8

Table 1 continued. Summary of core, contingent, and covariate field methods.

Land health standard

Field method Detailed field method

Cov

aria

tes/

othe

r

Bankfull width One measurement at each of 11 transects

Wetted width One measurement at each of 21 transects

Flood-prone width Two measurements of the floodplain valley width at riffles closest to the bottom and top of reach

Slope Elevation change over entire reach length using transit and stadia rod, clinometer, or hydrostatic level

Reach length GPS points of top and bottom of the reach and stick & tape meas. along the thalweg of the stream

Photos Photo points

Human influence Ocular est. of human activities on left and right banks at each of 11 transects

1Contingent indicator 2methods for the quantification of riparian vegetative cover and composition are pending. In the interim, use the multiple indicator monitoring (MIM) methods.

9

Indicators for Lotic Systems Overview

A suite of indicators can be computed from the core and contingent methods (Table 2). The large number of indicators results from the fact that many indicators can be computed from one field method. For example, AquADat contains five different streambed particle size indicators that are all derived from field pebble counts. Indicators are both computed and stored within the BLM’s AquADat database. AquADat metadata and brief descriptions of each indicator are found in Appendix A. Indicators are grouped by the BLM’s fundamentals of land health and can be used to assess standard attainment in addition to a variety of other applications including: assessing resource management plan effectiveness, quantifying the efficacy of restoration or reclamation actions, and completing grazing or other use-based permit renewals.

Table 2. Core, contingent, and covariate field methods and associated indicators. A brief description of each indicator listed in the right most column can be found in Appendix A.

Fund-amentals of land health

Field method Method type Indicators computed from field measurements (abbreviations explained in Appendix A)

Bio

dive

rsity

/ rip

aria

n ha

bita

t Macroinvertebrate biological integrity Core1 InvasiveInvertSp, ObservedInvertRichness,

ExpectedInvertRichness ,OE_Macroinvertebrate

Ocular est. of riparian vegetative type, cover, and structure

Core VegComplexity, RiparianVegComplexity, InvasiveWoody, NativeWoody, InvasiveHerb, NativeHerb, SedgeRush

Quantitative est. of riparian vegetative composition and cover

Contingent2 Capacity to store and report this data is pending development - Use MIM excel app in the interim

Canopy cover Core PctOverheadCover, BankOverheadCover

Wat

er q

ualit

y1

pH Core pH Specific conductance Core SpecificConductance

Temperature Core InstantTemp, Capacity to store and report continuous temperature data (contingent indicator) is pending development

Total nitrogen and phosphorous Contingent TotalNitrogen, TotalPhosphorous

Turbidity Contingent Turbidity

10

Table 2 continued. Core, contingent, and covariate field methods and associated indicators. A brief description of each indicator listed in the right most column can be found in Appendix A.

Fund-amentals of land health

Field method Method type Indicators computed from field measurements (abbreviations explained in Appendix A)

Wat

ersh

ed fu

nctio

n - i

nstre

am h

abita

t

Pool depth, length, and frequency Core ResPoolDepth, PctPools, PoolFreq

Large woody debris Core Frequency of LWD, Volume of LWD, component of RelativeBedStability

Bed particle size distribution Core

PctFines, D16, D84, D50, GeometricMeanParticleDiam, component of RelativeBedStability

Bank stability and cover Core BankCover, BankStability, BnkCoverStab, BnkCoverBedrock, BnkCoverCobble, BnkCoverLWD, BnkCoverVeg

Floodplain connectivity Core BankfullHeight, IncisionHeight, FloodplainConnectivity

Ocular estimate of instream habitat complexity

Contingent InstreamHabitatComplexity

Bank angle Contingent BankAngle

Thalweg depth profile Contingent ThalwegDepthCV, ThalwegDepthMean, PctDry, component of RelativeBedStability

Cov

aria

tes/

othe

r

Bankfull width Covariate3 BankfullWidth, Entrench, component of RelativeBedStability

Wetted width Covariate WettedWidth, component of RelativeBedStability

Flood-prone width Covariate FloodWidth, Entrench

Slope Covariate Slope, component of RelativeBedStability

Reach length Covariate Sinuosity Photos Covariate NA Human impacts Covariate None currently available

1Core method: measurable ecosystem component applicable across many different ecosystems, management objectives, and agencies. Core aquatic methods are recommended for application wherever the BLM implements monitoring and assessment of wadeable perennial streams. 2Contingent method: measureable ecosystem component having the same characteristics of cross-program utility and consistent definition as core methods, but that are measured only where applicable. Contingent methods are not informative everywhere and, thus, are only measured when there is reason to believe they will be important for management purposes. 3Covariate: measured or derived parameter used to account for natural spatial or temporal variation in a core, contingent, or supplemental method (e.g., gradient); covariates help determine the potential of a given stream reach or other aquatic system.

11

Water Quality

Background To assess water quality, the AIM-NAMF identifies three core field methods (pH, specific conductance, and instantaneous temperature) and four contingent methods (total nitrogen, total phosphorus, continuous temperature, and turbidity). In most cases, the water quality field methods have a single indicator that can be derived from them, as what you measure is what you report; however, a variety of indicators can be derived from continuous temperature. The core and contingent indicators are not meant to be representative of all state water quality standards. Rather, the indicators are meant to help determine the common chemical stressors resulting from land uses, such as irrigation water withdrawals and return flows, grazing, mining, timber harvest, and other activities occurring on or adjacent to public lands. Furthermore, the methods in the AIM-NAMF start with a one-time grab sample collected during base flow conditions. Such sampling is used to identify potential water quality exceedances requiring additional sampling or utilization of existing water quality monitoring networks to determine the temporal persistence of observed exceedances. In addition, one-time grab samples can be used to make correlations with macroinvertebrate biological integrity estimates to identify biologically relevant stressors. In contrast, monitoring to determine the attainment of state water quality standards requires the sampling frequency of each indicator to be consistent with individual state standards. Example Applications (the big picture)

• Identify priority water quality exceedances or stressors requiring additional sampling to assess standard attainment

• Assess attainment of state water quality standards (with increased sample frequency to match state standard operating procedures)

• Relate water quality conditions to observed biological condition as measured by benthic macroinvertebrates. Such correlations help to identify biologically relevant stressors (i.e., those water quality indicators that might be related to degraded biological conditions)

• Relate water quality conditions to land uses or permitted activities in a correlative assessment to inform adaptive management

Indicators pH

A. Description and Applications: pH is a measure of hydrogen ions in a solution. The more hydrogen ions in a solution the more acidic the solution, and the fewer hydrogen ions the more alkaline the solution. The acidification of aquatic systems can be detrimental to both the biota and ecosystem processes. Acidification can act as both a direct (e.g., reduction of survival rates) and indirect (e.g., mobilization of toxic metals) toxin to aquatic organisms, leading to reductions in species richness and density (Allan and Castillo 1995). Leaf litter breakdown rates, one of the primary energy sources of lotic food webs, are also known to decrease as pH decreases (Burton et al. 1985) due mainly to a reduction in microbial respiration and decreased number of invertebrate shedders (Dangles et al. 2004). Anthropogenic activities such as mining, logging, and burning of fossil fuels can decrease the pH of surface waters leading to a reduced quality of aquatic

12

resources (Allan and Castillo 1995). Acidification of aquatic systems is more commonly a concern than high alkalinity. High alkalinity can be caused by elevated concentrations of bicarbonate and/or increased photosynthetic activity, with pH varying diurnally by as much as 2.0 standard units. Therefore, high alkalinity can be indicative of productive systems with high nutrient levels (Allan and Castillo 1995).

B. Indicator Computations: For pH, no indicator computations are required, as what one measures in the field is what is reported (Table 2).

Specific Conductance A. Description and Applications: Conductivity measures the capacity of water to conduct

an electrical charge. Distilled water contains very low levels of ions and is a poor conductor of electricity, which results in a low specific conductance (i.e., conductivity standardized to 25oC). Specific conductance increases with the concentrations of ions in solution (e.g., nitrates, chloride, phosphate, magnesium, calcium, iron) and concentrations can be elevated by anthropogenic activities that increase erosion and/or ion loading (e.g., irrigation water withdrawals, mining, grazing) (Miller et al. 2007, Vander Laan et al. 2013). Excessive conductivity degrades the quality of domestic and/or animal drinking water and can impact freshwater organisms through acute toxicity or less dramatically through disrupting osmoregulation. Altered osmoregulation can decrease organismal fitness and/or change species distribution ranges (Blasius and Merritt 2002, Miller et al. 2007, Vander Laan et al. 2013).

B. Indicator Computations: For specific conductance, no indicator computations are required, as what one measures in the field is what is reported (Table 2).

Water Temperature

A. Description and Applications: Instantaneous temperature can be used to identify priority systems for continuous temperature monitoring and to provide context for other water quality indicators. However, to quantify the thermal regime and its viability for aquatic organisms, continuous temperature monitoring is needed. The thermal regime is among the most important abiotic drivers of biological patterns and processes in river systems (Vannote and Sweeney 1980). Over evolutionary time, organisms have evolved strategies to maximize fitness in response to different thermal regimes (Ward and Stanford 1982), while in contemporary time temperature constrains the distribution, development, and growth of aquatic organisms (Hauer and Benke 1987, Newbold et al. 1994). Consequently, activities such as grazing, dams, and water diversions that alter thermal regimes represent primary threats to lotic ecosystems (Dynesius and Nillson 1994).

B. Indicator Computations: For instantaneous temperature, no indicator computations are required, as what one measures in the field is what is reported (Table 2). For continuous temperature, a variety of indicators can be computed. Some of the more biologically relevant indicators include maximum daily temperature, maximum summer temperature, 7-day average of the maximum daily, and number of days above organismal thresholds such as 28°C for many trout species.

13

Total Nitrogen and Phosphorous A. Description and Applications: Nitrogen and phosphorous are the two major nutrients

that influence rates of primary productivity in stream systems. The major natural sources of nitrogen in streams are derived from N-fixing soil microbes and decomposing vegetation entering the stream through runoff and groundwater inputs (reviewed in Allan and Castillo 2007). In contrast to nitrogen, the predominant phosphorous source is the weathering of soils and rocks, particularly the weathering of sedimentary rock deposits. For both nitrogen and phosphorous, anthropogenic activities such as logging, cattle grazing, accelerated erosion, and agriculture can significantly increased nutrient loading (reviewed in Allan and Castillo 1995), making it among the ubiquitous freshwater stressors in the United States (Paulsen et al. 2008b). Excess nutrient loading can have adverse impacts on other water quality indicators and biological assemblages, as well as changing ecosystem processes such as food web structure and function (Citations). For example, excess nutrient loading can reduce dissolved oxygen concentrations through increased decomposition rates resulting from spikes in primary productivity (Citations)

B. Indicator Computations: All grab samples are sent to the Aquatic Biogeochemistry Lab at Utah State University for analysis. Lab Standard Operating Procedures can be found at http://canoeecology.weebly.com/uploads/2/1/0/0/21002098/abl_analytical_lab_manual.pdf. Detection limits are 25 µg/L and 10 µg/L for total nitrogen and phosphorus, respectively. Besides lab analysis, no indicator computations are required, as what one measures in the lab is what is reported (Table 2).

Turbidity

A. Description and Applications: Turbidity is a measure of light transmission through water and can provide an approximation of suspended sediment loads. High suspended sediment loads and low water clarity can reduce the quality of drinking water for domestic livestock and impair habitat quality for aquatic organisms. For example, high suspended sediment loads can directly influence biota through gill abrasion, smothering of eggs, or a reduction in foraging efficiency (Henley et al. 2000). Indirectly, high suspended sediment loads can reduce photosynthetically active radiation, which in turn reduces rates of primary productivity and can have cascading effects within aquatic food webs (Allan and Castillo 1995).

B. Indicator Computations: Three turbidity measurements are taken in the field and then averaged to get the reported value.

Precision of Water Quality Indicators Content in development

14

Watershed Function – Instream Habitat

Background Assessments of the physical functioning of stream and river systems are a major component of the BLM’s fundamentals of land health (43 CFR 4180.1). The watershed function fundamental calls for assessments of whether channel form and function are characteristic for the region (i.e., proper functioning condition), while the habitat fundamental requires the maintenance or improvement of aquatic habitat for threatened and endangered, proposed or candidate threatened and endangered, and other special status species (i.e., similar to the maintenance of cold- or warm-water fisheries under the Clean Water Act). To assess the condition of instream habitat, the AIM-NAMF identifies five core methods (pool depth, length, frequency; streambed particle sizes; bank stability and cover; floodplain connectivity; and large woody debris) and three contingent methods (bank angle, ocular estimates of instream habitat complexity, and a thalweg depth profile) from which a variety of indicators can be computed (Table 2). Specifically, the indicators characterize five physical habitat attributes:

1. Habitat type and volume o Percent pools o Pool frequency o Residual pool depth o Mean thalweg depth o Percent dry

2. Habitat complexity and cover for aquatic biota o Thalweg depth CV o Instream habitat complexity o Large woody debris (LWD) frequency o LWD volume

3. Bed stability and the type of benthic substrates for aquatic biota o Percent fines o D50, D16, D84 o Geometric mean o Relative bed stability

4. Bank type and condition o Bank cover

Percent cover of bedrock, cobble, LWD, or vegetation o Bank stability o Bank cover and stability o Bank angle

5. Floodplain interaction and channel dimensions o Bankfull height o Incision height o Floodplain connectivity

15

Example Applications (the big picture) • Assess attainment of biological opinion conditions, land health standards, or other policy

standards for physical habitat • Relate physical habitat conditions to observed biological condition as measured by

benthic macroinvertebrates. Such correlations help to identify biologically relevant stressors (i.e., degraded physical habitat conditions that might be related to degraded biological conditions, such as excessive fine sediment loading)

• Relate physical habitat conditions to land uses or permitted activities in a correlative assessment to inform adaptive management

• Assess habitat viability for threatened, endangered, or other species of management concern

Stream Channel Form, Function, and Habitat Indicators Pool depth, length, and frequency

Description and Applications: Stream habitat can generally be classified into “riffles” and “pools” based on the morphological and hydraulic properties of the channel. Riffles are shallow, fast-water habitats, while pools are deep, concave shaped, slow-water habitats (Hawkins et al. 1993). In general, greater frequency of pool-riffle sequences can increase hydraulic and geomorphic complexity and increases biological diversity (Citation). Due to the differences in hydraulic properties, riffles and pools generally have different sediment characteristics, as well as different water retention times and nutrient processing rates (Citations). From a biological perspective, riffles provide ideal habitat for benthic macroinvertebrates and spawning fishes due to the stable nature of the substrate, highly oxygenated water, and increased primary productivity (Citations). Subsequently, riffles are rich in food resources for fish. However, pools can provide important habitat for fishes, as deep water habitats can act as refugia from predators and fish expend less energy to swim in these slower water habitats. Additionally, pools can provide important refugia for invertebrates and fish during low flow periods when sections of streams may dry up (Citations). In many stream systems, anthropogenic activities have decreased the prevalence of the physical mechanisms that create and maintain pool habitat, thereby reducing the amount and quality of pool habitat. Pools are typically created by one of three mechanisms: the vertical force of water falling down over logs and boulders, high flow events that scour out sediment, or beaver activity (Archer et al. 2015). Decreased amounts of large trees and riparian vegetation within a watershed decreases the supply of pool forming woody debris. Various human activities can increase fine sediment loading to streams, which can increase fine sediment deposition in pools, leading to reduced pool habitat quality. Additionally, fine sediment can eventually fill in pools thereby reducing pool habitat availability. Flow reduction and alteration due to drought, water withdrawals, and dams can decrease the scouring ability of streams (Allan and Flecker 1993, Muotka and Syrjänen 2007). Beaver are considered a nuisance for human infrastructure, have been trapped for their pelts, and populations have been drastically reduced across the landscape. Recent research indicates that beaver population reduction has severely altered the physical structure of many streams across the West resulting in reduced pool frequency and quality.

16

From both the core field methods of pools and the contingent field methods of thalweg, we can compute indicators that characterize the amount of pool habitat (e.g., percent pools, residual pool depth, and pool frequency) and thereby also characterize the habitat and hydraulic complexity. Core methods use Archer et al. (2015)’s approach and define pools in the field, while contingent methods use the approach taken by Kaufmann et al. (1999) and use thalweg depth measurements and slope as a covariate to define pools post-hoc. Defining pools post-hoc is very difficult unless specific depth criteria can be easily applied and interpreted. Therefore, we recommend only using the contingent method of thalweg to characterize bed heterogeneity and to compute relative bed stability, rather than to characterize amount of pool habitat. The pool indicator computations below are only for the core method following Archer et al. (2015).

A. Indicator Computations: Pool habitat within the reach can be described by three indicators: percent pools, residual pool depth, and pool frequency. If there are no pools found in the reach, then Residual Pool Depth= NA, Pool Frequency= 0, and Percent Pools= 0.

• Percent Pools Percent of the sample reach (linear extent) classified as pool habitat. The lengths of all the pools in the reach are summed and divided by the reach length that was surveyed for pool habitat. The result is multiplied by 100 to express the value as a percent.

𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 = ∑ 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃ℎ(𝑚𝑚)𝑝𝑝𝑛𝑛𝑝𝑝=1

𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑃𝑃𝑆𝑆𝑃𝑃𝑆𝑆𝑆𝑆𝑃𝑃𝑆𝑆𝑃𝑃ℎ𝑃𝑃𝑃𝑃𝑃𝑃 (𝑚𝑚)∗ 100

Where p= 1….n and n is the total number of pools in the reach, SurveyedReachLen is the total field determined reach length.

If there is interrupted flow at a site, crews are supposed to only survey the flowing section of the reach and record how much of the reach was flowing as the SurveyedReachLen. If the crew did not fill in SurveyedReachLen for sites that had interrupted flow or were partially sampled, PctPools is reported as “NA”.

• Pool frequency Frequency of pools in the reach (# pools/km)

𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑆𝑆𝑃𝑃𝑃𝑃 =𝐶𝐶𝑃𝑃𝑆𝑆𝑃𝑃𝑃𝑃 𝑃𝑃𝑜𝑜 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑖𝑖𝑃𝑃 𝑃𝑃ℎ𝑃𝑃 𝑆𝑆𝑃𝑃𝑆𝑆𝑃𝑃ℎ 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑃𝑃𝑆𝑆𝑃𝑃𝑆𝑆𝑆𝑆𝑃𝑃𝑆𝑆𝑃𝑃ℎ𝑃𝑃𝑃𝑃𝑃𝑃 (𝑚𝑚)

∗ 1000

If there is interrupted flow at a site, crews are supposed to only survey the flowing section of the reach and record how much of the reach was flowing as the SurveyedReachLen. If the crew did not fill in SurveyedReachLen for sites that had interrupted flow or were partially sampled, PoolFreq is reported as “NA”.

17

• Residual Pool Depth Residual pool depth is a flow-independent measure of pool depth. Residual pool depth is calculated for each pool by subtracting the pool tail depth from the max depth.

𝑀𝑀𝑆𝑆𝑀𝑀𝑀𝑀𝑃𝑃𝑀𝑀𝑃𝑃ℎ𝑝𝑝 − 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑆𝑆𝑖𝑖𝑃𝑃𝑀𝑀𝑃𝑃𝑀𝑀𝑃𝑃ℎ𝑝𝑝 = 𝑆𝑆𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑀𝑀𝑃𝑃𝑀𝑀𝑃𝑃ℎ𝑝𝑝 Residual pool depths for each pool are then averaged across all pools in the reach.

𝑆𝑆𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑀𝑀𝑃𝑃𝑀𝑀𝑃𝑃ℎ = 1𝑃𝑃 �𝑆𝑆𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑀𝑀𝑃𝑃𝑀𝑀𝑃𝑃ℎ𝑝𝑝

𝑛𝑛

𝑝𝑝=1

Where p= 1…...n and n is the total number of pools in the reach Thalweg

A. Description and Applications: The thalweg is the deepest part of the stream containing the most flow. Thalweg is a contingent method and indicators computed from thalweg can be used as measures of habitat volume, connectivity, and complexity. Thalweg depths can also be used as a covariate in the computation of other indicators including residual pool depth, length, and, frequency as well as relative bed stability (see above). Mean thalweg depth can be a general descriptor of stream maximum depth at low-flow. Thalweg depths are systematically spaced measurements that also can be used to compute the percent of the reach that is dry. The coefficient of variation of thalweg depths can be used as a measure of stream bed heterogeneity and thereby habitat complexity. Indicators computed from thalweg depths include mean thalweg depth, percent of the reach that is dry, and thalweg depth CV.

B. Indicator Computations: • Mean Thalweg Depth: The mean of all thalweg depths collected. The number of

thalweg depths varies depending on the bankfull width. If thalweg measurements are missing, measurements are graphically interpolated; however, this interpolation has not been completed so at the present only sites with complete dataset have mean thalweg depth reported.

• Percent Dry: The percent of the reach that is dry is computed as the number of dry thalweg measurements divided by the total number of thalweg measurements taken.

• Thalweg Depth CV: The thalweg depth coefficient of variation is computed as

standard deviation of thalweg measurements divided by the mean thalweg depth. Thalweg CV is not computed for sites with any dry thalwegs because zeros could artificially inflate the CV.

Instream Habitat Complexity

A. Description and Applications: The complexity of instream habitat contributes to biodiversity, energy dissipation during high flow events, habitat refugia during extreme conditions (i.e., high and low flow), and stream processes (e.g., nutrient retention) (Solazzi et al. 2000, Ward et al. 2002). Historic and current anthropogenic land use such

18

as logging, splash dams, and water withdraw can decrease habitat complexity and volume by reducing inputs of large woody debris, channelizing streams, and reducing water levels and channel-forming flow (Allan and Flecker 1993, Muotka and Syrjänen 2007). Only one indicator is computed from this contingent method: an aggregate estimate of total instream habitat complexity, with an emphasis on cover provided for fish. While this indicator is only semi-quantitative because it is based on ocular estimates, it assesses a diversity of cover types of and complements more quantitative assessments of habitat complexity such as LWD frequency and volume.

B. Indicator Computations: This indicator is an aggregate measure of average cover provided by boulders, overhanging vegetation, live trees and roots, LWD > 0.3 m diameter, small woody debris < 0.3 m diameter, and stream banks for stream fishes measured at 11 plots. In the field, crews assess cover using five cover categories: 0 = absent 0%, 1 = sparse: <10%, 2 = moderate: 10-40%, 3 = heavy: 40-75%, and 4 = very heavy >75% (Table 3). For analysis, the categories are converted to the mid-point of each category (0%, 5%, 25%, 57.5%, and 87.5%, respectively) and then converted to proportional cover (0.05, 0.25, 0.575, and 0.875 respectively). The proportional cover for each individual cover type (e.g., boulders) is averaged across all transects, and then these averages are summed across the six cover types (Table 4). Note that crews collect proportional cover of macrophytes, filamentous algae, and artificial structures but following Kaufmann et al. (1999) these cover values are not included in the aggregate instream-habitat complexity indicator.

Table 3. Example raw instream habitat data from one site. Values represent cover classes ranging from 0 (absent) to 4 (>75% cover).

TRANSECT BOULDER OVER_VEG TREE_ROOT LWD SWD BANKS A 1 0 2 1 0 0 B 0 0 0 0 0 1 C 3 0 0 0 0 0 D 3 1 0 0 0 0 E 2 0 0 0 0 0 F 1 0 0 0 1 0 G 0 0 0 0 0 0 H 0 0 0 0 0 0 I 0 0 0 1 0 0 J 0 4 0 0 0 0 K 0 0 0 0 0 0

19

Table 4. Example calculation of instream habitat complexity from the raw data in Table 3. Table values are the mid-points of the five cover categories. The final instream habitat complexity for the example sample reach is 0.26, which is the sum of the average among cover categories.

TRANSECT BOULDER OVER_VEG TREE_ROOT LWD SWD BANKS A 0.05 0 0.25 0.05 0 0 B 0 0 0 0 0 0.05 C 0.575 0 0 0 0 0 D 0.575 0.05 0 0 0 0 E 0.25 0 0 0 0 0 F 0.05 0 0 0 0.05 0 G 0 0 0 0 0 0 H 0 0 0 0 0 0 I 0 0 0 0.05 0 0 J 0 0.875 0 0 0 0 K 0 0 0 0 0 0 SUM

AVERAGE 0.14 0.08 0.02 0.01 0.00 0.00 0.26 Large Woody Debris

A. Description and Applications: Large woody debris (LWD) is an important source of cover and velocity break for aquatic organisms such as fishes and amphibians (Whiteway et al. 2010; others). Geomorphically, LWD plays a critical role in the creation and maintenance of complex geomorphic channel units such as pools and in the local storage of bed sediments (Montgomery et al. 1995; others). The relative role of LWD as habitat and in structuring channel morphology varies geographically as a function of climate and the capacity of an ecosystem to support tree growth. Despite such limitations, even small wood, live trees, shrub, or roots can create geomorphic heterogeneity and provide habitat for a diversity of aquatic organisms. Human activities such as timber harvest, grazing, and channelization can limit the production, recruitment, and storage of LWD within a stream or river system.

LWD can serve as both an indicator and a covariate since it can provide critical habitat for aquatic organisms and influence geomorphic conditions. Regardless of the application, LWD computations are the same.

B. Indicator and Covariate Computations: LWD is defined as wood that is greater than 0.1 m in diameter for at least 1.5 m in length. LWD is binned into size categories (see LWD volume below) based on whether it was 1) within the bankfull channel or 2) bridging above bankfull channel. Pieces of LWD that are considered “qualifying” are tallied for each size category and location. Using the LWD tallies and measurement information are compiled for each sample reach, and the number, size, and volume of LWD are computed. Specifically: the following indicators are computed:

• LWD Frequency: The number of pieces of LWD is counted, and this value is divided by the length of the reach surveyed for LWD. All LWD size classes are

20

included, but only LWD within the bankfull channel is included. The length of the reach surveyed for LWD is determined by counting the number of transects for which LWD was assessed at and then multiplying by the distance between transects.

𝑃𝑃𝐿𝐿𝑀𝑀_𝑃𝑃𝑆𝑆𝑃𝑃𝑃𝑃 =∑ 𝐶𝐶𝑃𝑃𝑆𝑆𝑃𝑃𝑃𝑃𝑃𝑃𝐿𝐿𝑀𝑀𝑛𝑛 𝑝𝑝=1 𝑝𝑝

𝐶𝐶𝑃𝑃𝑆𝑆𝑃𝑃𝑃𝑃𝑃𝑃𝑆𝑆𝑆𝑆𝑃𝑃 ∗ 𝑃𝑃𝑃𝑃𝑃𝑃𝑆𝑆𝑃𝑃𝑆𝑆𝑃𝑃ℎ𝑃𝑃𝑃𝑃𝑃𝑃10

Where CountLWDp is the number of pieces of wood of size class “p”, Count Tran is the number of transects LWD was collected at, and TotReachLen is the total reach length specified by the protocol based on average bankfull width for the reach.

• LWD volume: Volume of LWD within the bankfull channel in the reach

expressed as m3/100 m. Where volume of each size class of LWD is computed using the following equation as taken from Robison (1998):

𝑉𝑉𝑃𝑃𝑃𝑃𝑆𝑆𝑚𝑚𝑃𝑃 = 𝜋𝜋 ��0.5 �𝑀𝑀𝑖𝑖𝑃𝑃𝑀𝑀𝑖𝑖𝑆𝑆𝑚𝑚𝑃𝑃𝑃𝑃𝑃𝑃𝑆𝑆 + 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀−𝑀𝑀𝑀𝑀𝑛𝑛𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀3

��2� �𝑀𝑀𝑖𝑖𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃ℎ + 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑛𝑛𝑀𝑀𝑀𝑀ℎ−𝑀𝑀𝑀𝑀𝑛𝑛𝑀𝑀𝑀𝑀𝑛𝑛𝑀𝑀𝑀𝑀ℎ

3�

Where MinDiameter and MaxDiameter are one of the following LWD diameter categories (large end):

MinDiameter (m) MaxDiameter (m) 0.1 0.3 0.3 0.6 0.6 0.8 0.8 2

And MinLength and MaxLength are one of the following LWD length categories (considering the section of the LWD where the diameter is greater than 0.1 m):

MinLength (m) MaxLength (m) 1.5 3 3 5 5 15 15 30

The total volume of large wood is determined by multiplying the volume of each size class by the number of pieces observed within that size class and then summing across all size classes.

𝑃𝑃𝐿𝐿𝑀𝑀_𝑉𝑉𝑃𝑃𝑃𝑃 = �𝐶𝐶𝑃𝑃𝑆𝑆𝑃𝑃𝑃𝑃𝑃𝑃𝐿𝐿𝑀𝑀𝑝𝑝 ∗ 𝑉𝑉𝑃𝑃𝑃𝑃𝑆𝑆𝑚𝑚𝑃𝑃𝑛𝑛

𝑝𝑝=1

Where CountLWDp is the number of pieces of LWD of a particular size class and p= 1…...n and n is the total number of LWD size classes.

21

Streambed Particle Sizes A. Description and applications: The substrate of streams and rivers plays a critical role in

the type, diversity, and abundance of organisms inhabiting a given system (reviewed in Allan and Castillo 1995). For example, benthic macroinvertebrates closely interact with the substrate during the egg, larvae, pupae, and in some cases adult life stages. Streambed substrate provides invertebrates cover from predators and complex hydraulics, food resources in the form of algal growth and organic matter retention, and attachment surfaces for feeding, matting, and egg laying (Citations). These same functions are provided to other aquatic organisms such as aquatic macrophytes, amphibians, and fishes (Citations). Consequently, substrate size, type, and diversity are frequently monitored to determine the habitat suitability for aquatic organisms.

In particular, habitat suitability for aquatic organisms across BLM lands is frequently assessed in terms of the amount of fine sediment, average particle size, and the diversity of particles sizes available. Excessive fine sediment is among the most deleterious stressors to aquatic biota (Wood and Armitage 1997, Paulsen et al. 2008b). Fine sediment can reduce food resource availability for benthic organisms (Henley et al. 2000), decrease benthic egg survival (Bjornn and Reiser 1991), and decrease habitat quality by filling interstical spaces, which are important micro-habitats for macroinvertebrates and smaller fishes (Cunjak and Power 1986, Gries and Juanes 1998). For these reasons, fine sediment is one of the dominant indicators used by the BLM to assess the quality of the streams or rivers. Similarly, we often seek to determine whether a system is in balance with the sediment and hydraulic regime (i.e., are fine sediment levels natural or the result of poor land management). To assess this, we report on relative bed stability (See more detailed description below). Lastly, the average or diversity of available substrate sizes can influence the diversity and abundance of aquatic organisms (Citations).

B. Indicator Computations: Five indicators summarize particle size distribution within a

sample reach (percent fines, D16, D50, D84, geometric mean particle diameter) and are computed from the 210 substrate particle measurements taken from the active channel, defined as scour line on one side of the stream to scour line on the other side (10 per each of 21 transects)(Table 1). These values are computed to characterize substrate conditions within the entire sample reach, and because, in most instances, sample reaches are randomly located and transects are systematically spaced, the descriptive statistics can be interpreted as unbiased representations of substrate conditions.

The measured particle sizes are used to calculate the substrate indicators by ordering the substrate particles from smallest to largest and computing:

• Percent fines = percent of measured particles with a b-axis finer than 2 mm. Percent of measured particles with a b-axis finer than 6 mm can also be computed and provided on request. This indicator is needed to calculate Al-Chokhachy et al. (2010) index of physical habitat condition for streams in the Columbia River basin.

• D16 = particle size corresponding to the 16th percentile of measured particles

22

• D50 = particle size corresponding to the 50th percentile of measured particles

• D84 = particle size corresponding to the 84th percentile of measured particles

• Geometric mean = another measure of central tendency similar to that of the D50, but it is more heavily influenced by fine particle size classes. For example, Faustini and Kaufmann (2007) found that the geometric mean was, on average, half as large as the D50 at a site. While a D50 is a more commonly reported indicator for characterizing geomorphic and sediment transport, geometric means may be more biologically relevant because they are skewed towards fine sediment. For example, Kaufmann and Hughes (2006) found that the geometric mean was a significant covariate for a fish based index of biotic integrity. The geometric mean is computed as:

Geometric mean = exp ∑ log10 𝑋𝑋𝑖𝑖𝑛𝑛𝑖𝑖=1

𝑛𝑛

Where: N = number of measured substrate particles X = individual particle b-axis size

Measured particles used in the above calculations include hardpan and bedrock, which are given values of 4098 mm and 4097 mm respectively; however, organic particles are excluded from the above calculations. Field protocol for data from 2013 differed significantly for stream bed particle sizes and should be compared with following years’ data with caution. Specifically, particles were collected only within the wetted channel, particles were binned into size categories rather than measured, and only 110 particles were collected.

Relative Bed Stability (can only be computed if thalweg was collected)

A. Description and Applications: The size distribution of bed particles naturally varies across the landscape as a function of lithology, climate, watershed size, and slope (Wood and Armitage 1997). However, anthropogenic activities such as roads, dams, mining, logging, and grazing can increase fine sediment loading to streams. Imbalances in the supply and transport of bed sediments can alter channel structure and function, which in turn degrades the physical habitat for stream biota. A measure of stream bed stability is used to assess the degree to which sampled streams are in balance with the sediment and discharge regimes, (Kaufmann et al. 2008, 2009). Relative bed stability measures the capacity of a stream to transport bed sediments given the stream size, slope, discharge and roughness. Values less than zero indicate a streambed consisting of particles that are finer and more mobile than would be expected, whereas values greater than zero indicate coarser or more stable bed particles than would be expected under natural conditions.

B. Indicator Computations: Relative bed stability is calculated as the ratio of the median observed substrate diameter (approximated by the geometric mean diameter) divided by

23

the reach average for the largest particle that is mobile during bankfull flow. The calculation of the reach average for the largest particle that is mobile during bankfull flow indicator requires prior computation of the following indicators and covariates: percent slope, bankfull hydraulic radius as approximated by average bankfull width and height and corresponding thalweg depths, and measures of hydraulic roughness as approximated by pools delineated from the thalweg depth profile and LWD. Due to the computational intensity of this indicator, values are not currently available but will be available summer or fall 2018 for all sites that collected thalweg as a contingent method.

Bank Stability and Cover A. Description and Applications: Bank cover and stability measurements assess the

susceptibility of stream banks to both natural and accelerated erosion rates associated with anthropogenic activities. Streambank erosion overwhelmingly occurs during bankfull discharge events, which is the discharge associated with greatest amount of sediment transport and that has the largest influence on channel morphology (Wolman and Miller 1960). Anthropogenic activities that increase stream power (e.g., flow alteration, improperly sized culverts) or alter the composition and cover of stabilizing vegetation can increase bank erosion rates (Knapp and Matthews 1996, Coles-Ritchie et al. 2007, Herbst et al. 2012). Stream bank erosion is a source of fine sediment loading and channel widening. Elevated fine sediment loading can reduce the viability of the stream bottom environment for aquatic organisms such as amphibians, macroinvertebrates, and fishes (Henley et al. 2000). Bank erosion can also alter channel morphology and subsequent habitat quality (e.g., width:depth ratios) through changing the balance between the sediment and water supply and thus the transport capacity of a stream or river (citations). The composition and cover of vegetation and other stabilizing features such as large woody debris and boulders significantly influences streambank erosion. Therefore, the AIM protocol quantifies both streambank cover and the presence of any erosional features (e.g., fractures, slumps, sloughs).

B. Indicator Computations: Because depositional banks (e.g. point bars) are inherently unstable, we exclude all depositional banks from all analyses and only consider banks that could contain erosional features. The bank cover and stability field measurements from each of the 21 transects (42 plots) within a sample reach are used to compute the following three indicators:

a. Bank cover: the number of plots classified as ‘covered’ are divided by the total number of plots and expressed as a percent. Cover constituents include perennial vegetation, wood greater than 10 cm in diameter, bedrock, and mineral substrate with a b-axis greater than 15 cm. A plot with greater than 50% cover from any one or a combination of these cover categories is considered ‘covered’.

b. Bank stability: the number of plots classified as ‘stable’ are divided by the total number of plots and expressed as a percent. The ‘stable’ designation results from the absence of erosion features: fracture, slump, slough, and eroding.

24

c. Bank stability and cover: the number of plots that met both the ‘covered’ and ‘stable’ criteria, expressed as a percent.

In addition to these three indicators, we also report the observed aerial coverage for each of the four cover constituents. Aerial coverage by constituent is provided for practitioners to understand what contributed to a ‘covered’ or ‘uncovered’ rating.

a. Bedrock: average percent cover by bedrock among the plots. b. Cobble: average percent cover by cobble among the plots. c. Large woody debris: average percent cover by large woody debris among the

plots. d. Perennial vegetation: average percent cover by perennial vegetation among the

plots. Bank Angle

A. Description and Applications: Bank angle is the angle of the bank in degrees, as defined by Archer et al. (2015). Bank angle can be used in conjunction with bank stability information to assess the natural erosional progression of banks from ≥ 90° to undercut ≤ 90°, and back to ≥ 90° (Fig. 3). Outside of sand-bed systems, streams should have some undercut banks, and these undercut banks can provide important fish cover and habitat. In contrast, an increase in bank angle (i.e., more laid back banks) may results from alterations to vegetative composition and cover by livestock trampling (Kauffman and Krueger 1984; Platts 1991), hydrologic alterations associated with roads (Furniss et al. 1991), or the destabilizing effects of forest harvest (Dose and Roper 1994). This contingent indicator is an important component of Al-Chokhachy et al's. (2010) index of

physical habitat condition for streams in the Columbia River basin, and therefore this contingent indicator should be collected if computation of this index is desired. Figure 3. The general shape of obtuse (A) and acute (B) banks.

B. Indicator Computations: All bank angles <45 degrees are changed to 45 degrees

following Archer et al. (2015). Then all bank angles are averaged across all banks and transects.

Floodplain Connectivity

A. Description and Applications: The connectivity or access of a stream channel to its floodplain is critical for the maintenance and recruitment of riparian vegetation, the dissipation of energy during high flow events, and the creation of seasonal habitats during inundation (Naiman and Decamps 1997). Anthropogenic activities that alter the sediment and/or hydrologic regime or directly manipulate the stream channel can decrease the connectivity between streams and their adjacent floodplains (Citations).

25

The floodplain connectivity indicator is used to assess differences between bankfull height and that of the first flat depositional feature at or above bankfull. When the difference between these two heights is minor, the system has access to the floodplain during annual or semi-annual high flow events and floodplain functionality is not expected to deviate from potential natural conditions. However, when the height of the first flat depositional feature significantly exceeds that of the bankfull elevation because of down-cutting or bed degradation, the frequency and duration of inundation are reduced and floodplain functionality deviates from potential natural conditions. It is important to note that naturally confined or high gradient systems are not expected to support floodplains and do not require floodplains for proper functionality. For such systems, the indicator is designed to set measured bankfull and floodplain heights to be one in the same and the system is considered to be properly functioning. Similarly, the flood-prone width covariate is used to assess the degree of valley confinement, the potential of the system to support a floodplain, and the potential extent of the system’s active floodplain.

B. Indicator Computations: • Bankfull height: average of 11 bankfull elevation heights measured from the

water’s surface • Incision height: average of 11 floodplain elevation heights measured from the

water’s surface • Floodplain connectivity: In plain text, floodplain connectivity is the difference

between average floodplain height and average bankfull height. Computationally,

Where: X = floodplain height

Y = bankfull height n = number of sampled bankfull and incision heights

Precision of Watershed Function – Instream Habitat Indicators Content in development

𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭 𝒄𝒄𝑭𝑭𝑭𝑭𝑭𝑭𝒄𝒄𝒄𝒄𝒄𝒄𝑭𝑭𝒄𝒄𝑭𝑭𝒄𝒄𝒄𝒄 = 𝑭𝑭𝑭𝑭𝒍𝒍𝟏𝟏𝟏𝟏 ��∑ 𝑿𝑿𝑭𝑭𝑭𝑭𝑭𝑭=𝟏𝟏

𝑭𝑭� -�

∑ 𝒀𝒀𝑭𝑭𝑭𝑭𝑭𝑭=𝟏𝟏

𝑭𝑭� + 𝟏𝟏.𝟏𝟏�

26

Biodiversity and Riparian Habitat

Background Land health standards associated with the biodiversity and riparian habitat quality standards can be assessed by three core and one contingent indicators (Table 2). Macroinvertebrate biological integrity is used to quantify the intactness of instream aquatic assemblages to directly address the biodiversity fundamental, while also informing assessments of the water quality fundamental. Riparian habitat quality and intactness can be measured by either ocular or quantitative estimates of the riparian vegetative type, cover, and structure, depending on the particular monitoring application. Specifically, ocular estimates are recommended for use in regional-scale assessments (e.g., land use plan or larger spatial-scale monitoring) or where general information is sought regarding the intactness of riparian areas to buffer against anthropogenic stressors (e.g., thermal, sediment, or nutrient loading), to promote properly functioning channel form and function, or to provide wildlife habitat, among other functions. In contrast, quantitative estimates of riparian vegetative cover and composition (contingent indicator) are recommended for local, site-specific estimates of riparian condition and trend, to assess the impacts of a particular land use (e.g., grazing), or when conducting assessments for riparian obligate species. In addition to the ocular estimates of riparian vegetation, AIM-NAMF uses canopy cover as a core indicator. Canopy cover directly measures the capacity of riparian vegetation and other features such as cliff walls to shade the stream and mitigate thermal loading and, thus moderate stream temperatures (Beschta 1997; Johnson and Jones 2000). Canopy cover estimates also provide information regarding the amount of potential leaf litter and other terrestrial organisms that may be available to subsidize aquatic food webs (Cummins 1974; Baxter et al. 2005). Supplemental methods and indicators, such as periphyton and fishes, can incorporate additional lines of evidence regarding the biological integrity of instream assemblages, and should be added if management objectives explicitly address these forms of biological integrity. Example Applications (the big picture)

• Assess attainment of biological opinion terms and conditions, land health standards, or other policy standards for physical habitat

• Relate physical habitat conditions to observed biological condition as measured by benthic macroinvertebrates. Such correlations help to identify biologically relevant stressors (i.e., those degraded physical habitat conditions that might be related to degraded biological conditions, such as excessive fine sediment loading)

• Relate physical habitat conditions to land uses or permitted activities in a correlative assessment to inform adaptive management

• Assess habitat viability for threatened, endangered, or other species of management concern

Biodiversity and Riparian Habitat Indicators Benthic Macroinvertebrates

A. Description and Applications: To address both the biodiversity and water quality fundamentals aquatic macroinvertebrates are sampled. Aquatic macroinvertebrates are ubiquitously used by state and federal agencies as the primary screening tool for

27

assessing chemical, physical, and biological conditions, with all 50 states using macroinvertebrates in biomonitoring programs (USEPA 2002). Macroinvertebrates are used to compliment more traditional chemical and physical monitoring techniques (i.e., provide multiple lines of evidence regarding degradation) during baseline monitoring because they: 1. Are relatively long-lived and thus integrate conditions through time and space; 2. Are ubiquitously found in perennial stream systems; 3. Exhibit a variety of life history strategies, which can be used to discriminate among causes of impairment; and 4. Can be sampled and identified in an efficient and cost-effective manner (Bonada et al. 2006). Furthermore, macroinvertebrate sampling can be used to estimate food resource availability for higher trophic levels such as amphibians and fishes.

B. Indicator Computations: • O/E or MMI indices: Bioassessments of stream and river systems work by

comparing observed biota at sample sites with an estimate of the sample site’s biological potential. Two common tools for conducting bioassessment of lotic systems are observed/expected (O/E) and Multi-Metric (MMI) indices. O/E models compare the macroinvertebrate taxa observed at sample sites of unknown condition to the assemblages predicted to occur in the absence of anthropogenic stressors (Hawkins et al. 2000, Hawkins 2006). In contrast, MMI models aggregate multiple macroinvertebrate metrics (e.g., total richness, proportion of tolerant individuals, combined richness of mayflies, stoneflies, and caddisflies) to assess biological condition (Stoddard et al. 2008). Metrics that differentiate between reference and degraded conditions are selected, rescaled to standardized scale (e.g., 0 to 100), and aggregated into a single measure of biological condition. Metrics are computed for sample sites of unknown condition and compared to metric values predicted to occur in the absence of anthropogenic activities (Hawkins et al. 2010a, Vander Laan and Hawkins 2014). Both types of indices use empirical models built with data from a network of reference sites to predict conditions of sample sites in the absence of anthropogenic impacts. O/E scores range from 0 to approximately 1, with a score of zero indicating that sample sites have no taxa in common with expected reference conditions. In contrast, an O/E score of 1 occurs when macroinvertebrate assemblages at sample sites are equal to those of reference conditions. Biological condition of sample sites is assessed based on the precision of the reference site model used to predict the expected number and type of taxa. Specifically, the standard deviation (SD) of predicted reference site O/E scores, with sample sites scoring less than one SD below the mean of reference sites having ‘minimal departure’ from reference; sites scoring between one SD and two SD having ‘moderate departure’; and sites scoring more than two SD below the mean of reference sites having ‘major departure’ from reference conditions. Final aggregate MMI scores commonly range from 0 to 100, with lower scores indicating that sample sites significantly deviated from reference sites across all metrics. In contrast, higher MMI scores occur when all macroinvertebrate metrics are equal to or greater than those of reference conditions. MMI scores computed

28

for sample sites are compared to reference site scores, with a percentile approach commonly used to assign different degrees of departure or condition categories (e.g., sample sites falling below the 5th percentile of reference MMI distribution are considered to have major departure). Caution should be exercised when interpreting O/E or MMI scores to ensure sample or laboratory error does not have an undue influence on final scores. For example, taxonomic richness of macroinvertebrate assemblages typically increases asymptotically with the number of individuals in a sample (Vinson and Hawkins 1996). Therefore, indices commonly recommend a minimal number of individuals per sample to minimize this sample artifact (e.g., 200 individuals). Low samples counts can result from sampling and/or laboratory processing errors, but can also be a signal of degraded biological condition. For samples with low counts, additional samples should be collected to verify the precision of ‘major’ or ‘moderate’ departures from reference. Additionally, care should be taken to ensure the environmental conditions of sample sites are similar to those of the reference conditions used to developed biological indices. This can be determined from the “ModelApplicability” field in AquADat, which reports whether or not the sample site's environmental conditions are within the range of experience of the model. A “fail” indicates the model had to extrapolate, rather than interpolate, to accommodate one or more of the habitat variables. O/E or MMI scores and condition ratings should be interpreted cautiously if a site failed the test for range of experience of the model. The NOC can currently compute state O/E or MMI bioassessment indices for the following states: UT, NV, CA, CO, and OR (only available for the Marine Western Coastal Forest and Western Cordillera Columbia Plateau ecoregions). If a state model is not available, a BLM Westwide model can be used for all other states. , with the exception of Idaho, which falls within the PIBO regional model (see below). Additionally, the following two regional models are available on request if applicable to the sample locations and if needed to meet policy requirements: for CA, WA, and OR, the Northwest Forest Plan Region (model developed for AREMP program); and for MT, ID, and OR, the Columbia River Basin (model developed for the PIBO program) can be applied. Model specific metadata can be found in Appendix B (in development) and more information about different bioassessment indices can be found at http://www.qcnr.usu.edu/wmc/bioassessments/. All macroinvertebrate samples are sorted and identified by NAMC, with the exception of midges, which are sent to EcoAnalyst if identification to genus or species level is needed (e.g., models for CO and NV). NAMC sorting and taxonomic SOPs can be found at www.usu.edu/buglab/.

• Invasive Species: In addition to computing state bioassessment indices from the macroinvertebrate identification and enumeration data, we also compute the

29

presence of invasive invertebrates. The list of considered invertebrates includes: 1. All individuals within the crayfish family Cambaridae; 2. The Asian clam (Corbicula fluminea); 3. Zebra mussels (Dreissena polymorpha); 4. Quagga mussels (Dreissena rostriformis); 5. New Zealand mud snail (Potamopyrgus antipodarum) and 6. The red-rimmed melania snail (Melanoides tuberculatus).

Percent Canopy Cover A. Description and Applications: Percent canopy cover measures the capacity of riparian

vegetation to mitigate thermal loading (i.e., provide shade) and thus moderate stream temperatures (Beschta 1997, Johnson and Jones 2000). The extent of the riparian canopy also provides information on the amount of potential leaf litter to subsidize aquatic food webs (Cummins 1974). We use a modified convex densiometer to quantify percent canopy cover at the left bank, stream center, and right bank of the 11 main transects following the methods of the USEPA (2009).

B. Indicator Computations: • Percent Overhead Cover: Convex densiometers assess shade as the number of

17 intersections on a grid that are covered. The number covered is divided by 17 to get percent cover. Four densiometer measurements are taken in the center of the stream looking upstream, downstream, left, and right. The percent overhead cover is an average of the percent cover at these four locations.

• Bank Overhead Cover: Bank overhead cover is computed similarly to percent overhead cover except only two densiometer measurements are averaged: those taken at left and right banks.

Riparian Habitat Complexity and Cover A. Description and Applications: Riparian zones are transitional areas between terrestrial

and aquatic ecosystems that provide important habitat for organisms and influence many ecological processes (Naiman and Decamps 1997). Healthy riparian zones provide refugia during periods of stress (e.g., drought or floods), influence ecosystem microclimates, provide energy and nutrients to the stream (e.g., leaf litter), shade surface water, stabilize banks, filter runoff and groundwater inputs, and provide a corridor for plants and animals (Hauer and Lamberti 1998). However, these functions can be altered by land use such as grazing and road development, which is one of the most pervasive land uses in western riparian ecosystems (Fleischner 1994, Beschta et al. 2012). Riparian habitat complexity is estimated using visual estimates of vegetation type and cover at 11 transects.

B. Indicator Computations: • Vegetative Complexity: This indicator is an aggregate measure of the average

vegetative cover provided by three different vegetative height categories: canopy (>5m), understory (0.5-5m), and ground cover (<0.5m). The canopy category only applies to woody vegetation, while the understory and ground cover categories are divided into two vegetation types: woody or non-woody. In the field, crews assess cover for each height category and vegetation type using five cover classes: 0 = absent 0%, 1 = sparse: <10%, 2 = moderate: 10-40%, 3 = heavy: 40-75%, and 4 =

30

very heavy >75% (Table 5). For analysis, the cover classes are assigned mid-points (0%, 5%, 25%, 57.5%, and 87.5% respectively) and then converted to proportional cover (0.05, 0.25, 0.575, and 0.875 respectively). Proportional cover is then summed across the two vegetation types and three heights, and finally averaged across the left and right banks of 11 transects (Table 13).

31

Table 5. Example raw vegetative cover data from one site. Values are an ordinal scale from 0 (absent) to 4 (>75% cover).

Canopy (> 5m) Understory Ground Cover

Transect Bank Big trees Small trees Woody Non-

woody Woody Non-

woody A Left 1 0 3 1 3 2 A Right 0 0 3 0 3 1 B Left 3 0 2 0 3 3 B Right 3 1 3 0 2 2 C Left 2 0 2 4 2 2 C Right 1 0 1 2 2 2 D Left 0 0 0 1 1 1 D Right 0 0 0 0 2 0 E Left 0 0 0 1 4 1 E Right 0 4 4 0 4 1 F Left 0 0 1 1 1 1 F Right 0 0 0 2 3 1 G Left 0 0 0 4 4 1 G Right 0 0 0 4 4 1 H Left 0 0 0 4 4 1 H Right 1 0 3 0 2 2 I Left 0 1 2 4 2 2 I Right 0 0 1 2 2 2 J Left 0 0 1 1 1 1 J Right 0 0 0 2 3 1 K Left 0 0 1 4 2 1 K Right 0 0 0 0 1 4

32

Table 6. Example calculation of vegetative complexity from the raw data in Table 5. Table values are mid-points of the five cover classes. The final vegetative complexity value is the sum of cover at a given transect (left and right bank) averaged across all transects.

Canopy (> 5m) Understory Ground Cover

Transect Bank Big trees Small trees Woody Non-

woody Woody Non-

woody Sum A Left 0.05 0 0.575 0.05 0.575 0.25 1.5 A Right 0 0 0.575 0 0.575 0.05 1.2 B Left 0.575 0 0.25 0 0.575 0.575 1.975 B Right 0.575 0.05 0.575 0 0.25 0.25 1.7 C Left 0.25 0 0.25 0.875 0.25 0.25 1.875 C Right 0.05 0 0.05 0.25 0.25 0.25 0.85 D Left 0 0 0 0.05 0.05 0.05 0.15 D Right 0 0 0 0 0.25 0 0.25 E Left 0 0 0 0.05 0.875 0.05 0.975 E Right 0 0.875 0.875 0 0.875 0.05 2.675 F Left 0 0 0.05 0.05 0.05 0.05 0.2 F Right 0 0 0 0.25 0.575 0.05 0.875 G Left 0 0 0 0.875 0.875 0.05 1.8 G Right 0 0 0 0.875 0.875 0.05 1.8 H Left 0 0 0 0.875 0.875 0.05 1.8 H Right 0.05 0 0.575 0 0.25 0.25 1.125 I Left 0 0.05 0.25 0.875 0.25 0.25 1.675 I Right 0 0 0.05 0.25 0.25 0.25 0.8 J Left 0 0 0.05 0.05 0.05 0.05 0.2 J Right 0 0 0 0.25 0.575 0.05 0.875 K Left 0 0 0.05 0.875 0.25 0.05 1.225 K Right 0 0 0 0 0.05 0.875 0.925 Average 1.20

• Riparian Vegetative Cover o Canopy, Understory, Ground: These three indicators are all computed

similarly to overall cover values. They are measures of the cover provided by riparian species ONLY within each respective layer: canopy (> 5m), understory (0.5-5m), and ground (<0.5m). In the field, crews assess cover for each height category using five cover classes: 0 = absent 0%, 1 = sparse: <10%, 2 = moderate: 10-40%, 3 = heavy: 40-75%, and 4 = very heavy >75%. For analysis, each cover class is assigned a mid-point (0%, 5%, 25%, 57.5%, and 87.5% respectively) and then converted to proportional cover (0.05, 0.25, 0.575, and 0.875 respectively). Proportional cover is then averaged across 11 transects and right and left banks. These calculations are identical to those detailed above for vegetative complexity except cover is not assessed separately for different

33

vegetation types (e.g. woody, non-woody) and each layer (canopy, understory, and ground) is kept separate rather than summing across layers.

• Percent of Native and Nonnative Species o Woody: The presence of nonnative woody vegetation is assessed at 11

transects on left and right banks for a total of 22 vegetation plots. The number of plots with native woody vegetation present is divided by 22 and multiplied by 100 to get the percent of plots with nonnative woody species present. Similarly, the percent of plots with native woody species is calculated for context of whether any natives were present or if it was all nonnative vegetation.

o Herbaceous: The presence of nonnative and native herbaceous vegetation is assessed and calculated the same as detailed above for nonnative and native woody vegetation.

• Percent of Sedges and Rushes: The presence of sedges or rushes is assessed at 11 transects on left and right banks for a total of 22 vegetation plots. The number of plots with sedges or rushes present is divided by 22 and multiplied by 100 to get the percent of plots with sedges and rushes present.

Precision of Biodiversity and Riparian Habitat Indicators Content in development

34

Covariates

Background and Example Applications The chemical, physical, and biological potential of stream and river systems naturally varies across the landscape (reviewed in Allan 2004). Six covariates can be computed from the core methods to begin to account for this natural variability: bankfull width, wetted width, flood-prone width, entrenchment, slope, and sinuosity (Table 4). Covariates such as slope and bankfull width are critical for computing and interpreting several of the watershed function – instream habitat indicators, particularly residual pool depth, length, and frequency and relative bed stability. In addition to field-based covariates, site coordinates and geographic information systems are used to compute a large number of covariates (e.g., watershed area, precipitation, geology, soil types) for use in the computation of O/E type indices and general data interpretation. Photos can be important in reviewing the accuracy of many the computed indicators too, as well as for reporting and communicating results. Qualitative assessment of the extent and type of anthropogenic impacts adjacent to or within the assessed reaches can also be useful in general data interpretation, but this indicator has not been made a priority to compute because much of this information is available from readily available geospatial layers. Covariate Description and Computations Bankfull Width and Wetted Width Bankfull width is used in the calculation of floodplain connectivity, entrenchment ratio, and relative bed stability. It is also provides a coarse estimate of stream size. Stream size is a natural factor that influences many indicators such as canopy cover and bed particle size distributions. Therefore, stream size along with ecoregion are important covariates to consider prior to applying benchmarks (see below). Wetted widths are stage dependent; however, they can provide important context for other indicators such as instream habitat complexity, which is rated as zero if the transect is dry. Additionally wetted width can provide a coarse assessment of available habitat at low flows and can be paired with pool data to determine residual pool volume. Hydraulic retention can also be approximated by the width-depth product, which has been shown to be an important covariate for fish IBIs (Citation). Reported values for bankfull width and wetted width are the average value across all sampled transects (11 and 21 for bankfull and wetted width, respectively). Wetted width for dry transects equals zero.

Flood-prone Width • Flood-prone Width: Average flood-prone width is defined as the valley width at two

times maximum bankfull height. Flood-prone width is an approximation of the size of the floodplain and can be compared to bankfull width to compute entrenchment ratio (see below). Two flood-prone width measurements are taken within riffle habitats per sample reach, one near the top and bottom of the sample reach. Reported flood-prone width is the average of the two measured widths.

• Entrenchment Ratio: Entrenchment ratio equals average flood-prone width divided by average bankfull width. Ratios of 1-1.4 represent entrenched streams; 1.41-2.2 moderately entrenched streams; and ratios greater than 2.2 indicate rivers only slightly entrenched in a well-developed floodplain (Rosgen 1996). This entrenchment value can be used with other covariates such as slope and sinuosity to determine stream type and the potential of the system for floodplain formation (Rosgen 1996).

35

Slope Slope is used in the computation of residual pool depth, length, and frequency when computing these indicators from thalweg measurements. Slope is also used in the computation of stream bed stability and is an important factor in determining stream velocity and power (the ability of the stream to move sediment). For example, slope can be used to help develop context for the potential of a stream system to support a given number of pools (Al-Chokhachy et al. 2010). Slope is measured as the total change in elevation from the top to the bottom of the sample reach divided by the distance of the reach along the thalweg and then multiplied by 100 to convert to a percentage. To ensure precision of this measurement, most crews measure the total elevation change until they get two measurements that are within 10% of one another. The reported value for slope uses the average of two measurements that are within 10% of each other for the total change in elevation.

Sinuosity Sinuosity is a key descriptive characteristic of stream type that is correlated with sediment size and slope. Sinuosity can also be a form of habitat complexity, creating features such as backwaters and oxbows. Sinuosity is computed as the reach length along the thalweg divided by the straight line distance between bottom of reach (BR) and top of reach (TR) coordinates. Reach length along the thalweg is measured by crews, and the straight line distance between the bottom of reach and top of reach coordinates is computed as:

𝑆𝑆𝑃𝑃𝑆𝑆𝑆𝑆𝑖𝑖𝑃𝑃ℎ𝑃𝑃𝑃𝑃𝑖𝑖𝑃𝑃𝑃𝑃𝑀𝑀𝑖𝑖𝑃𝑃𝑃𝑃 = acos �𝑠𝑠𝑀𝑀𝑛𝑛(𝑀𝑀𝐿𝐿𝑇𝑇𝐵𝐵𝐵𝐵)𝜋𝜋

180� ∗ sin(𝑃𝑃𝐿𝐿𝑃𝑃𝑇𝑇𝑇𝑇) 𝜋𝜋

180+ cos �𝑀𝑀𝐿𝐿𝑇𝑇𝐵𝐵𝐵𝐵 𝜋𝜋

180� ∗ cos (𝑀𝑀𝐿𝐿𝑁𝑁𝑇𝑇𝐵𝐵𝜋𝜋

180− 𝑀𝑀𝐿𝐿𝑁𝑁𝐵𝐵𝐵𝐵𝜋𝜋

180) ∗ 6371000

Sinuosity is not computed for sites that were partially sampled because crews were not consistent with recording the partial reach length with matching coordinates. Precision of Covariates Content in development

36

Benchmarks: From Indicator Values to Management Decisions What are Benchmarks and Why are They Needed?

Monitoring objectives are quantitative statements about desired resource conditions that help translate information to action in natural resource management. These quantitative statements are critical to answering questions such as: Was the management goal achieved? Are watersheds, streams, and rivers functioning properly? Are management actions maintaining the health of public lands? A key component of monitoring objectives is benchmarks. Benchmarks are indicator values, or ranges of values used to indicate the need for change or conversely, project success. For example, total nitrogen values characterize ambient nutrient concentrations at a single point in time, but without appropriate benchmarks, such measurements lack context and cannot be used to assess water quality condition and potential eutrophication. Benchmark attainment or exceedance can trigger the need to adjust management practices, to collect additional data, or indicate project success among other applications. The importance of establishing benchmarks, either formal or informal, cannot be understated. In a review of judicial decisions, Fischman and Ruhl (2016) found the failure to establish benchmarks as one of the leading causes for adaptive management to be ruled against in U.S. courts. In the absence of quantifiable benchmarks, management agencies struggle to make objective and decisive decisions as to when current management strategies should be reviewed, amended, or changed all together. Approaches to Setting Benchmarks

There are many different sources of benchmarks and methods for their development (reviewed in Hawkins et al. 2010). However, in aquatic systems benchmarks are generally set relative to some approximation of the environmental conditions expected in the absence of anthropogenic impacts. Because of the paucity of knowledge regarding pre-European stream and river conditions and it being unrealistic to manage for such conditions, least disturbed or minimally impacted sites are commonly used to establish reference conditions (Hughes et al. 1994, Stoddard et al. 2006). This does not mean though that the end goal of all management is the attainment of reference condition. For example, the Federal Land Policy Management Act requires the BLM to manage public lands under a multiple use mandate, which differs from the preservation mandate of the National Park Service. Thus, benchmarks can be used to assess the degree of departure from reference condition and managers must decide whether such departures are sustainable given management objectives. In other words, it is best practice to set benchmarks relative a reference condition and change the degree of allowable departure from that benchmark based on management objectives, as opposed to varying benchmarks based on management objectives. Such an approach is critical to facilitate comparable condition estimates among management units or geographic areas. A central challenge of environmental monitoring is the ability to discriminate between natural environmental gradients and those resulting from anthropogenic activities. Traditionally, regulatory and land management agencies established benchmarks using narrative criteria, professional judgment, or by selecting ‘paired’ reference site(s). Although these approaches are feasible and have been effectively used, they have significant drawbacks capable of limiting the

37

repeatability and inference drawn from monitoring assessments. These include: (1) the subjective interpretation of narrative standards; (2) the considerable expense of identifying paired reference and test reaches; (3) the susceptibility of paired reference sites to pseudo-replication and (4) the difficulty of identifying replicate stream reaches given the multitude of confounding factors that can occur (reviewed in Hawkins et al. 2010). Where available, the AIM-NAMF recommends using benchmarks set by policy (e.g., state water quality standards, biological opinions), as these encompass the legal commitments made by the Bureau. However, in many cases, policy does not include objective, quantifiable benchmarks or policy objectives do not exist for a given indicator. In these instances and where available, the AIM-NAMF recommends utilizing networks of least disturbed sites established by state and federal agencies to define reference conditions and subsequent benchmarks (Hughes et al. 1994, Stoddard et al. 2006, Hawkins et al. 2010). Benchmarks established outside of policy should be considered monitoring benchmarks used to alert practitioners of potential problems requiring additional investigation before changing management or making policy decisions. Networks of least disturbed reference sites consist of stream or river monitoring locations screened for the presence or density of anthropogenic impacts using a mix of field- and/or GIS-based variables (e.g., road density, dams, artificial channel, agricultural land uses)(e.g., Herlihy et al. 2008, Ode et al. 2016). The use of reference site networks is advantageous because natural spatial and temporal gradients are more likely to be adequately represented, thus minimizing the chance of confounding natural and anthropogenic gradients. However, given the natural environmental heterogeneity among streams and rivers, one must ensure that monitoring sites of unknown condition are compared to reference sites of similar potential (Hawkins et al. 2010b). In aquatic monitoring, there are two widely used methods for setting benchmarks based on networks of reference sites: predicted natural conditions and percentiles of regional reference. Predicted natural conditions use empirical models based on geospatial predictors to understand the spatial variability among reference sites for a given indicator. For example, Hill et al. (2013) were able to use nine GIS-derived variables (e.g., air temperature, watershed area, reservoir index) to explain 87% of the spatial variability in mean summer stream temperature (root-mean-square deviation of 1.9oC) among reference sites throughout the conterminous U.S. In addition to stream temperature, models have been developed for macroinvertebrate biological integrity through the use of observed/expected or multimetric indices (e.g., Hawkins 2006; Hargett et al. 2007; Vander Laan et al. 2013), total nitrogen and phosphorous (Olson 2012), conductivity (Olson and Hawkins 2012), and instream habitat complexity (Al-Chokhachy et al. 2010). Such models account for natural environmental gradients and are used to make predictions of chemical, physical, or biological values expected at a site in the absence of anthropogenic impairment. Condition is then determined based on the deviation of the observed indicator value from the site specific predicted value. If this deviation is greater than specified percentiles of model error (e.g., 75th and 95th), the value is assigned a condition of moderate or major departure respectively. Predictive modeling approaches are advantageous because they result in site specific predictions with known levels of accuracy and precision. However, predictive models have not been built for all indicators. For indicators lacking predictive models, the BLM can start by using the natural

38

range of variability among regional reference site networks to set expectations and assign benchmarks (i.e., percentiles of regional reference conditions) (Hughes et al. 1986; Paulsen et al. 2008). Specifically, distributions of reference site indicator values can be used to characterize the natural range of variability from which test sites can be compared and subsequent exceedances identified. Reference site networks are typically grouped by physiographic boundaries (e.g., level III ecoregions; Omernik 1987) to account for differences in reference site distributions resulting from factors such as climate and topography. For example, the 90th and 70th percentiles of reference site fine sediment values for the Colorado Plateau ecoregion can be used as benchmarks to classify the condition of a monitoring site as “major departure,” “moderate departure,” or “minimal departure” from reference conditions, respectively. In other words, a site would be categorized as having major departure from reference conditions if the fine sediment value for a sample site is greater than that observed among 90% of reference sites in the Colorado Plateau ecoregion. Where protocols are concordant and empirical models have not been developed, the BLM can use the percentiles of regional reference condition approach established by the EPA’s National Rivers and Streams Assessment program to make a first pass at setting benchmarks (e.g., Stoddard et al. 2005). Limitations to Benchmark Approaches

All approaches for setting benchmarks are subject to error and the potential to under (type I errors) or over protect (type II errors) natural resources. An important part of developing and applying benchmarks is to be aware of potential limitations. The two main approaches described herein, percentiles of regional reference conditions and predicted natural conditions, are no exception and have important limitations that can influence the interpretation of monitoring data. The National AIM team is currently working on additional content to be provided here that explains the limitations of both of these approaches. Understanding Benchmarks Available in the Benchmark Tool

As stated in the introduction, the AIM-NAMF recommends using benchmarks set by policy (e.g., state water quality standards, biological opinions), as these encompass the legal commitments made by the Bureau. However, in many cases, policy does not include objective, quantifiable benchmarks or policy objectives do not exist for a given indicator. Therefore, the national AIM team has compiled example benchmarks for a majority of indicators that can serve as a starting point for indicator interpretation (Table 7). Additionally, the national AIM team has developed an Excel based tool that allows practitioners to apply these or any other benchmarks to assist with the interpretation of AIM data. Again, the provided benchmarks should only be used after fully understanding their limitations and vetting the benchmark values with a BLM interdisciplinary team. Lastly, the benchmark methods described in Table 7 are only applicable to a subset of computed indicators because not all indicators make sense to assign a condition category or have readily available methods to determine benchmarks.

39

Table 7. Overview of indicator benchmark methods that can be used in the absence of benchmarks established by policy. The provided benchmarks were developed using one of three approaches: predicted natural conditions, percentiles of regional reference, or best professional judgement. All benchmarks should be reviewed by an interdisciplinary team prior to using results to inform management decisions.

Indicator(s) Citation Condition Benchmarks Appropriateness of

benchmarks Predicted natural conditions Total Nitrogen Olson

and Hawkins 2013

Predicted natural condition plus 75th (moderate) and 95th (major) percentiles of model error, 52.1 µg/L and 114.7 µg/L, respectively.

Use these benchmarks if no state water quality standards are available.

Total Phosphorous Olson and Hawkins 2013

Predicted natural condition plus 75th (moderate) and 95th (major) percentiles of model error, 9.9 µg/L and 21.3 µg/L, respectively.

Specific Conductance Olson and Hawkins 2012

Predicted natural condition plus 75th (moderate) and 95th (major) percentiles of model error, 27.1 µg/L and 74.5 µg/L, respectively.

OE_Macroinvertebrate Varies depending on the model (see www.usu.edu/buglab/ for more information). In general, most model's benchmarks are the mean of the reference distribution plus 1SD (moderate) or 2SD (major).

Should always be appropriate to meet state regulations, if a state based model was used.

Percentiles of regional reference conditions VegComplexity, LWD_Freq, LWD_Vol, InstreamHabitatComplexity, PctOverheadCover, BankOverheadCover

Kaufmann et al. 1999; Stoddard et al. 2005

30th (moderate) and 10th (major) percentiles of regional reference conditions defined by 23 groups of EPA hybrid level III ecoregions and a combination of stream size (> or ≤ 10 m bankfull width) and sampling protocol (wadeable vs. boatable). See Tables 10-11 for specific values.

Use these if no other policy benchmarks are available. Benchmarks should be carefully reviewed based on local knowledge of environmental conditions and thus site potential.

PctFines, FloodplainConnectivity

Kaufmann et al. 1999; Stoddard et al. 2005

70th (moderate) and 90th (major) percentiles of regional reference conditions defined by 23 groups of EPA hybrid level III ecoregions and a combination of stream size (> or ≤ 10 m bankfull width) and sampling protocol (wadeable vs. boatable). See Tables 10-11 for specific values.

40

Water Quality Benchmark Development

Under the water quality fundamental, BLM land health standards specify the attainment of state water quality standards as a management objective; therefore, the development of water quality benchmarks for assessing condition is relatively straightforward for those indicators for which state standards have been developed. Practitioners should consult with their state regulatory agency to determine the availability of state standards and the required field sample frequency to assess standard attainment. For those states or indicators for which standards have not been identified, we recommend use of the predicted national conditions approach for setting benchmarks due to the site specific nature of these predictions, as well as known levels of accuracy and precision (see introduction to benchmark section). Specific conductance benchmarks can be established using methods in Olson et al. (2012). This model uses 15 GIS-derived variables (e.g., % calcium carbonate in local geology, air temperature, precipitation) to explain 71% of the spatial variability in base-flow specific conductance concentrations (root-mean-square error 84.2 µS/cm) among reference sites throughout the western U.S. Benchmarks were then established by taking the site-specific predicted natural conditions from the model and adding the 75% and 95% of model error, 27.1 µg/L and 74.5 µg/L respectively, to the prediction. We are currently working to refine this

Table 7 continued.

Indicator(s) Citation Condition Benchmarks

Determining if these benchmarks are appropriate

Best professional judgement InvasiveInvertSp Assumed all sites with invasive species

present had major departure from reference and if no invasive species present had minimal departure from reference.

Consult management objectives to determine if invasive species presence is acceptable.

pH Kaufmann et al. 1999

Acidic (7, 6.5) and alkaline (8.5, 9) for moderate and major departure from reference respectively.

Use these benchmarks if no state standards are available. Otherwise use state standards.

BankCover, BankStability, BnkCoverStab

80% of banks stable and/or covered (moderate) and 69% of banks stable and/or covered (major) for all Hybrid III ecoregions except Plains or the southern xeric and eastern xeric ecoregions. These ecoregions should have naturally lower bank stability so thresholds were 70% of banks stable and/or covered (moderate) and 50% of banks stable and/or covered (major).

Use these benchmarks if no other policy or information available, but carefully examine for your sites and verify condition ratings with local knowledge of physiographic conditions and thus site potential.

41

approach by differentially assigning model error to small and large streams given expectations for naturally higher specific conductance values for larger systems. Total nitrogen and total phosphorus benchmarks can be established using methods in Olson and Hawkins (2013). The total nitrogen model uses 12 GIS-derived variables (e.g., atmospheric nitrogen deposition, air temperature, precipitation) to explain 23% of the spatial variability in base-flow total nitrogen values (root-mean-square error 80.1 µg/L). The total phosphorus model uses 15 GIS-derived variables (e.g., % calcium carbonate in local geology, air temperature, precipitation) to explain 46% of the spatial variability in base-flow total phosphorus values (root-mean-square error 20.5 µg/L) among reference sites throughout the western U.S. Benchmarks for total nitrogen and total phosphorus were established similar to specific conductance. The 75% and 95% of model error are 52.1 µg/L and 114.7 µg/L respectively for total nitrogen and are 9.9. µg/L and 21.3 µg/L respectively for total phosphorus. We are currently working to refine this approach by differentially assigning model error to small and large streams given expectations for naturally higher nutrient concentrations for larger systems. Watershed Function, Instream Habitat, Riparian Habitat, and Biodiversity Benchmark Development Overview We combine descriptions of benchmark methods for all indicators that address the watershed function, instream habitat quality, biodiversity, and riparian habitat quality fundamentals of land health because benchmarks for most of these indicators were developed very similarly. The exception is biodiversity. For our main indicator of biodiversity, OE_Macroinvertebrate, benchmarks are largely determined using state bioassessment indices, which include benchmarks for assigning condition categories. Because of the large number of states and associated models, we do not go into specific methods here. More information about each state’s model and OE and MMI score interpretation see Appendix B (in development) and http://www.qcnr.usu.edu/wmc/bioassessments/. All instream habitat and riparian indicators generally lack predictive models and therefore the provided benchmarks are based on the percentiles of regional reference conditions. Specifically, we used data from 1096 reference sites throughout ten hybrid level II/III ecoregions across the west to characterize the natural range of indicator variability expected to occur in the absence of anthropogenic impairment (Fig. 4; Table 8) (Stoddard et al. 2006). Benchmarks were established at the extremes of reference site distributions to identify significant departures from reference for each of three stream sizes: small wadeable reaches ≤10 m bankfull width, large wadeable reaches >10 m bankfull width, and boatable reaches. For example, the 70th and 90th percentiles of reference site percent fine sediment values (<2 mm) for the small wadeable streams in the Eastern Xeric Basin ecoregion, 44 and 73% respectively, were used to separate minimal, moderate, and major departure from reference conditions, respectively. In other words, sites were categorized as having major departure if fine sediment measurements exceeded levels observed among 90% (73% fine sediment) of reference sites. Subsequent text explains details about reference site selection and screening, as well as detailed methods used to determine

42

ecoregion/stream size groupings of references sites. Lastly, we present benchmark values and number of references sites used for a given ecoregion/stream size.

Figure 4. EPA hybrid level II/III ecoregions used to group 1096 reference sites for determining the natural ranges of variability among indicators and subsequent benchmarks. Figure developed by Stoddard et al. 2005.

43

Table 8. Translation table between EPA hybrid level II/III ecoregions and EPA level III ecoregions. EPA Hybrid Level II/III Ecoregion EPA Level III Ecoregion Northern Xeric Basin Columbia Plateau

Snake River Plain Northern Basin and Range

Pacific Northwest Coast Range Puget Lowland Willamette Valley Cascades Sierra Nevada Eastern Cascades Slopes and Foothills North Cascades Klamath Mountains

Xeric California Southern and Central California Chaparral and Oak Woodlands Central California Valley

Northern Rockies Blue Mountains Northern Rockies Idaho Batholith Middle Rockies Canadian Rockies

Southern Rockies Wasatch and Uinta Mountains Southern Rockies

Eastern Xeric Basin Wyoming Basin Colorado Plateaus Arizona/New Mexico Plateau

Southern Xeric Basin Central Basin and Range Mojave Basin and Range Chihuahuan Deserts Madrean Archipelago Sonoran Basin and Range

Southwest Mountains Southern California Mountains Arizona/New Mexico Mountains

Northern Cultivated Plains High Plains Rangeland Plains Southwestern Tablelands

Northwestern Glaciated Plains Northwestern Great Plains

44

Reference sites The national AIM team used an EPA dataset to identify the range of variability among least disturbed sites (i.e., reference) by hybrid Omernick level II/III ecoregions. The reference dataset was comprised of 1096 sites sampled between 2000 and 2009 as part of EPA’s Wadeable Streams Assessment (WSA), Western Environmental Monitoring and Assessment Program (EMAP-West), and National Rivers and Streams Assessment (NRSA) surveys (Stoddard et al. 2005, Olsen and Peck 2008, Paulsen et al. 2008a, EPA 2016). The EPA screening process for reference site designations differed slightly among surveys, but we sought to use all three datasets to maximize sample sizes within each ecoregion. In general, the EPA used multiple lines of evidence to screen sampled sites for those in least disturbed conditions. At the broadest scale, GIS derived metrics of land use (e.g., row crop and urban land use) and other anthropogenic activities (e.g., dams and impoundments) were used to screen sites. At the site-scale, field observations of the magnitude and proximity of streamside human activities such as roads, agricultural, and urban development and riparian disturbance were used as described by Kaufmann et al. (1999). Instream habitat variables such as habitat complexity and fine sediment levels were also utilized. Lastly, GIS and field-based observations were used in conjunction with water chemistry data, as described by Herlihy et al. (2008), to designate sites as least, moderately, and highly disturbed relative to other sampled sites. To avoid circularity in above process, no measures directly related to a given indicator were used to screen sites. For example, field measurements of riparian vegetation, sediment, or instream habitat complexity were not used to screen sites and determine ranges of variability for any instream indicators. The only exceptions were the use of riparian conditions for designations of instream habitat indicators and visa-versa. We are compiling further detail on reference screening criteria, including values used, and this information is forthcoming. Development of Ecoregion/Size Groupings Given a lack of predictive models of instream habitat and riparian indicators, the national AIM team attempted to minimize natural variability associated with ecoregions and stream size. Similar to the approach taken by the EPA in the Western Environmental Monitoring and Assessment Program (EMAP-West) surveys, we used EPA hybrid level II/III ecoregions to divide reference sites into relatively homogenous physiographic regions. Then within a given ecoregion, we used bankfull width to separate reference sites into small streams (≤ 10 m bankfull width) and large streams (> 10 m). We chose 10 m as an arbitrary cutoff based on balancing sample sizes and maximizing discriminatory efficiency for individual indicators between groups. Boatable AIM data is collected using a slightly different protocol and these sites were more similar to each other than wadeable streams of similar bankfull widths. Therefore, boatable sites were considered their own stream size category. Using this approach, most indicators had a substantial difference between benchmark values for small streams and large streams that made ecological sense (Table 9 & 10), while still providing

45

adequate sample sizes for most indicators for a given ecoregion and stream size (e.g., > 20 sites) (Table 11 & 12). For example, the benchmark for major departure from reference for PctOverheadCover in the Northern Rockies is 20.9% for small streams (which generally support more overhead cover than large streams) but 2.3% for large (Table 9). When subdividing the dataset by both ecoregion and stream size, we attempted to maintain >30 reference sites in each grouping. However, this wasn’t always possible, and we had to lump ecoregions in some cases in an effort to increase sample sizes (see specific bullets below). This approach enabled us to use all available reference site data for a given indicator. For ecoregion and stream size groupings where minimal sample sizes were not achieved, extreme care should be exercised when using the provided benchmarks, especially if sample sizes are <20. Groups or indicators with low sample sizes included:

• Boatable reaches within all ecoregions o Boatable ecoregions were lumped into most similar groups (Table 10 and 12).

• Large streams in the Xeric Basin Ecoregion o These values were different enough from other ecoregions that we didn’t feel

justified in lumping these sites with another ecoregion, but care should be exercised when using the provided benchmarks.

• Wadeable small and large streams within the Xeric California ecoregion o Wadeable AIM sites sampled in the Xeric California to date have been on the

border of Pacific Northwest ecoregion and so Pacific Northwest reference sites have been used in the current benchmark tool to set benchmarks for these sites rather than lumping reference sites from these two diverse ecoregions to obtain benchmarks.

• Floodplain Connectivity for both wadeable and boatable sites o Sample sizes were not egregiously low for most ecoregions, so ecoregions were

not lumped for wadeable sites, but care should be exercised when using the provided benchmarks.

o Sample sizes were much lower for boatable than wadeable sites, and there was little difference among ecoregions so all ecoregions were lumped.

• LWD_Vol, which had inadequate sample sizes across the board. o No benchmarks were developed for LWD_Vol.

The national AIM team is in the progress of refining these benchmarks further and attempting to develop models for physical habitat indicators. Our most immediate attempts at refining these benchmarks are in the Columbia River Basin were models exist for habitat MMIs, percent fines, pool tail fines, bank angle, and percent pools (Al-Chokhachy et al. 2010). We are currently working on assessing the applicability of these models to BLM sampled sites.

46

Table 9. Benchmarks used to assign condition ratings of minimal, moderate, or major departure for wadeable reaches determined by quantiles of regional values by Omernik hybrid level III/IV ecoregions and stream size: small wadeable streams ≤ 10 m or large wadeable streams > 10 m. All benchmarks should be reviewed by an interdisciplinary team prior to using results to inform management decisions.

Northern Xeric

Basin

Pacific Northwest and

Xeric California

Northern Rockies

Southern Rockies

Indicator Response Benchmark Percentile Small Large Small Large Small Large Small Large

PctOverheadCover Decreases with stress

Moderate 30th 47.2 7.6* 65.8 37.6 51.2 15.0 40.7 19.0 Major 10th 12.2 0.0* 38.1 18.3 20.9 2.3 12.2 6.3

BankOverheadCover Decreases with stress

Moderate 30th 69.0 55.1* 84.5 73.8 76.5 61.1 73.1 66.5 Major 10th 32.1 25.2* 67.8 56.8 53.9 38.3 52.7 58.6

VegComplexity Decreases with stress

Moderate 30th 1.03 0.76* 1.14 1.03 1.06 0.84 1.02 0.99 Major 10th 0.60 0.73* 0.83 0.76 0.79 0.57 0.90 0.78

Pctfines Increases with stress

Moderate 70th 45 44* 15 12 29 15 23 22 Major 90th 66 81* 33 26 48 27 37 36

InstreamHabitat Complexity

Deceases with stress

Moderate 30th 0.41 0.19* 0.32 0.29 0.43 0.33 0.64 0.49 Major 10th 0.16 0.11* 0.17 0.16 0.28 0.21 0.34 0.27

LWD_Freq Decreases with stress

Moderate 30th 4.70 0.00* 0.00 0.00 0.00 1.35 0.00 0.00 Major 10th 0.00 0.00* 0.00 0.00 0.00 0.00 0.00 0.00

Floodplain Connectivity

Increases with stress

Moderate 70th -0.09 0.11* -0.31 0.20 -0.26 -0.20 -0.09 -0.13 Major 90th 0.22 0.22* 0.26 0.54 0.00 0.01 0.17 0.09

*Sample size less than 20

47

Table 9 continued. Benchmarks used to assign condition ratings of minimal, moderate, or major departure for wadeable reaches determined by quantiles of regional values by Omernik hybrid level III/IV ecoregions and stream size: small wadeable streams ≤ 10 m or large wadeable streams > 10 m. All benchmarks should be reviewed by an interdisciplinary team prior to using results to inform management decisions.

Eastern Xeric

Basin Southern Xeric

Basin Southwest Mountains

Northern Cultivated

Plains Rangeland

Plains Indicator Response Benchmark Percentile Small Large Small Large Small Large Small Large Small Large

PctOverheadCover Decreases with stress

Moderate 30th 23.9 0.9 70.6 0.9 64.6 11.9 26.9 6.3 5.3 0.1 Major 10th 10.0 0.0 51.9 0.0 33.0 9.5 3.7 0.0 0.6 0.0

BankOverheadCover Decreases with stress

Moderate 30th 70.9 27.1 81.8 28.4 85.8 58.1 76.1 50.8 60.5 34.9 Major 10th 39.8 10.4 65.5 6.2 64.4 32.7 63.3 26.7 31.4 17.6

VegComplexity Decreases with stress

Moderate 30th 0.83 0.56 1.01 0.62 0.89 0.48 0.79 0.68 0.88 0.72 Major 10th 0.71 0.41 0.67 0.26 0.59 0.33 0.49 0.53 0.55 0.50

Pctfines Increases with stress

Moderate 70th 44 46 54 64 26 28 77 84 84 72 Major 90th 73 82 77 84 41 52 96 99 100 93

Instream HabitatComplexity

Deceases with stress

Moderate 30th 0.42 0.12 0.46 0.14 0.36 0.24 0.21 0.16 0.19 0.09 Major 10th 0.12 0.05 0.27 0.08 0.23 0.11 0.08 0.05 0.08 0.03

LWD_Freq Decreases with stress

Moderate 30th 1.80 0.00 0.45 0.50 0.95 5.02 1.60* 8.07 2.92 0.00 Major 10th 0.00 0.00 0.00 0.00 0.00 0.54 0.26* 1.84 0.00 0.00

Floodplain Connectivity

Increases with stress

Moderate 70th 0.11 0.09 0.11 0.23* -1.00 0.26* 0.07* 0.30* 0.30 0.26 Major 90th 0.66 0.50 0.43 0.44* 0.03 0.65* 0.29* 0.37* 0.61 0.50

*Sample size less than 20

48

Table 10. Benchmarks used to assign condition ratings of minimal, moderate, or major departure for boatable reaches determined by quantiles of regional values by groupings of Omernik hybrid level III/IV ecoregions. Note that PctOverheadCover is not collected at boatable sites and is therefore not included in this table. All benchmarks should be reviewed by an interdisciplinary team prior to using results to inform management decisions.

Indicator Response Benchmark Percentile

Eastern and Southern

Xeric Basin

Northern and Southern Rockies

and Northern Xeric Basin

Pacific Northwest

Northern Cultivated

Plains Rangeland

Plains

BankOverheadCover Decreases with stress

Moderate 30th 14.9 10.1 16.6 19.0 7.8 Major 10th 4.3 3.5 6.6 4.9 1.0

VegComplexity Decreases with stress

Moderate 30th 0.72 0.78 1.17 1.13 0.67 Major 10th 0.54 0.63 0.54 0.72 0.49

Pctfines Increases with stress

Moderate 70th 40 3 14 93 69 Major 90th 97 35 98 100 99

InstreamHabitat Complexity

Deceases with stress

Moderate 30th 0.13 0.14 0.14 0.09 0.10 Major 10th 0.08 0.07 0.06 0.06 0.05

LWD_Freq Decreases with stress

Moderate 30th 1.34 0.00 3.58 3.89 1.05 Major 10th 0.00 0.00 0.00 0.13 0.00

Floodplain Connectivity*

Increases with stress

Moderate 70th 0.22 0.22 0.22 0.22 0.22 Major 90th 0.40 0.40 0.40 0.40 0.40

* All ecoregions combined due to low sample sizes

49

Table 11. Number of reference sites for wadeable reach benchmark development and subsequent condition rating assignments of minimal, moderate, or major departure. Samples sizes presented by Omernik hybrid level III/IV ecoregions and stream size: small wadeable streams ≤ 10 m or large wadeable streams > 10 m.

Northern

Xeric Basin

Pacific Northwest and

Xeric California Northern Rockies

Southern Rockies

Indicator Small Large Small Large Small Large Small Large PctOverheadCover 30 11 138 66 143 42 53 33

BankOverheadCover 30 11 138 66 143 42 53 33 RiparianVegComplexity 30 11 138 66 143 43 53 33

Pctfines 30 11 137 65 138 42 53 33 InstreamHabitatComplexity 30 11 137 65 138 42 53 33

LWD_Freq 30 11 136 64 135 41 52 32 FloodplainConnectivity 22 6 117 55 115 38 39 27

Table 11 continued. Number of reference sites for wadeable reach benchmark development and subsequent condition rating assignments of minimal, moderate, or major departure. Samples sizes presented by Omernik hybrid level III/IV ecoregions and stream size: small wadeable streams ≤ 10 m or large wadeable streams > 10 m.

Eastern Xeric

Basin

Northern Cultivated

Plains Rangeland

Plains Southern

Xeric Basin Southwest Mountains

Indicator Small Large Small Large Small Large Small Large Small Large PctOverheadCover 40 29 24 31 75 74 34 29 31 22

BankOverheadCover 40 29 24 31 75 74 34 29 31 22 RiparianVegComplexity 40 29 24 31 75 74 34 29 31 23

Pctfines 40 29 23 28 73 68 34 29 31 23 InstreamHabitatComplexity 40 29 23 28 73 68 34 29 31 23

LWD_Freq 40 28 19 24 70 64 33 29 31 23 FloodplainConnectivity 27 21 10 13 44 25 21 13 27 16

50

* All ecoregions combined due to low sample sizes

Table 12. Number of reference sites for boatable reach benchmark development and subsequent condition rating assignments of minimal, moderate, or major departure determined by quantiles of regional values by Omernik hybrid level III/IV ecoregions. Note that PctOverheadCover is not collected at boatable sites and is therefore not included in this table.

Indicator

Eastern and Southern

Xeric Basin

Northern and Southern

Rockies and Northern

Xeric Basin Pacific

Northwest

Northern Cultivated

Plains Rangeland

Plains BankOverheadCover 29 52 46 36 26

RiparianVegComplexity 29 52 46 36 26 Pctfines 29 49 45 35 26

InstreamHabitatComplexity 29 51 45 36 26 LWD_Freq 29 51 45 36 26

FloodplainConnectivity* 65 across all ecoregions

51

Appendix A. BLM AIM AquADat Local Feature Class Metadata Description Abstract: This feature class includes monitoring data collected nationally to understand the status, condition, and trend of resources on BLM lands. Data are collected in accordance with the BLM Assessment, Inventory, and Monitoring (AIM) Strategy. The AIM Strategy specifies a probabilistic sampling design, standard core indicators and methods, electronic data capture and management, and integration with remote sensing. Attributes include the BLM aquatic core indicators: pH, conductivity, temperature, pool depth, length, frequency, streambed particles sizes, bank stability and cover, floodplain connectivity, large woody debris, macroinvertebrate biological integrity, ocular estimates of vegetative type, cover, and structure and canopy cover. In addition, the contingent indicators of total nitrogen and phosphorous, turbidity, bank angle, thalweg depth profile and quantitative vegetation estimates (see the Data Structure and Attribute Information section for exact details on attributes). Data were collected and managed by BLM Field Offices, BLM Districts, and/or affiliated field crews with support from the BLM National Operations Center. Data are stored in a centralized database (AquADat) at the BLM National Operations Center. Purpose: This dataset was created to monitor the status, condition and trend of national BLM resources in accordance with BLM policies. The methodology used for the collection of these data can be found in TR 1735-2 (AIM National Aquatic Monitoring Framework: Field Protocol for Wadeable Lotic System). These data should not be used for statistical or spatial inferences without knowledge of how the sample design was drawn or without calculating spatial weights for the points based on the sample design. Update frequency: Annually Data Access Constraints Access constraints: NON-PUBLIC. BLM INTERNAL USE ONLY. Unverified Dataset. These data will be restricted to internal BLM staff, contractors and partners directly involved with developing the associated planning documents. These data might contain sensitive information, and may only be accessed by the public by filing a FOIA request, which may or may not be granted depending on the applicable FOIA exemption(s). Use constraints: "NON-PUBLIC, BLM INTERNAL USE ONLY. NOT FOR DISTRIBUTION. NO WARRANTY IS MADE BY BLM AS TO THE ACCURACY, RELIABILITY, OR COMPLETENESS OF THESE DATA FOR INDIVIDUAL USE OR AGGREGATE USE WITH OTHER DATA. The User is cautioned that these data have not been verified, and have not been approved for release. The User should take reasonable measures to ensure that these data are protected from disclosure. Although these data might be available to internal BLM staff, contractors or partners; the quality and fit for use of these data should be considered unknown. The User is advised that the content of the metadata file associated with these data might be incomplete. The User assumes the entire risk associated with its use of these data. The BLM shall not be held liable for unintentional disclosure; nor for any use or misuse of the data described or contained herein. Further, the BLM assumes no liability for the current accuracy, reliability, completeness or utility of these data on any system or for any general or scientific purposes. The User bears all responsibility in determining whether these data are fit for the User's intended use. These data are neither legal documents nor land surveys, and must not be used as such. Official records can be referenced at most BLM offices. Please report any errors in the data to the BLM office from which it was obtained. Any products derived from these data should clearly identify the source as unverified data. They must also include the statement ""REVIEW AND/OR DISPLAY COPY - NOT FOR DISTRIBUTION.” The BLM should be cited as the data source in any products derived from these data. Any Users wishing to modify the data are obligated to describe within the process history section of the metadata the types of modifications they have performed. The User specifically agrees not to misrepresent the data, nor to imply that changes made were approved or endorsed by BLM. This data may be updated by the BLM without notification."

Spatial Domain Boundary Coordinates- Unprojected (geographic) West -157.514027 (longitude) East -102.365554 (longitude) North 70.761066 (latitude) South 32.0226 (latitude)

Point Of Contact Bureau of Land Management Scott Miller Director, National Aquatic Monitoring Center (720) 545-8367 BLM National Operations Center; Denver Federal Center, Building 50 Lakewood, CO 80225

Citation Title: BLM AIM AquADat Local Feature Class Originators: US Dept of Interior, Bureau of Land Management Publication date: 20170101 Data type: vector digital data Dataset credit: US Department of the Interior - Bureau of Land Management Assessment, Inventory, and Monitoring Project Team; NAMC

52

Data Structure and Attribute Information (most commonly used columns are in light gray)

Data Type Indicator or Column Heading Description

Site

Des

crip

tors

SiteCode

Site code assigned during the design process and used by field crews and project leads to track samples. SiteCodes should be used to track design specific information; whereas MasterCode should be used to query information about a site across multiple designs or revisits.

StreamName Stream name based off the USGS National Hydrography Dataset (NHD) layer

MasterCode Code used to identify a unique location. SiteCode may (or may not) change on site revisits. However, MasterCode will remain the same across all site visits.

UID Unique code for an individual site visit. This is the database primary key. Date Sample date (units: m/d/yyyy)

MergeSiteCodes List of existing monitoring sites that fall in the same location as this site. All such sites have been screened for merging using the site scouting protocol

VisitNumber A sequential number indicating the number of times the site has been visited up to the date of this sample. MidLat Latitude of the reach midpoint in NAD 83 (units: decimal degrees) MidLong Longitude of the reach midpoint in NAD 83 (units: decimal degrees) Project Project associated with data collection Protocol Protocol used for collecting the data (wadeable or boatable) State BLM Administrative State where the site is located District BLM District where the site is located FieldOffice BLM Field Office where the site is located Stratum The original design stratum for the site

Targeted Specifies whether the site was a part of a probabilistic random design (Random) or whether it was selected as a targeted site to address a specific management concern (Targeted)

StreamOrder Strahler stream order of the site

StreamSizeOrder

Stream size category as defined in the design by grouping Strahler stream orders together. Generally, SS- Small Streams (Stream Order:1-2) , LS-Large Streams (Stream Order:3-4), RV-Rivers (Stream Order >5), RM-River Major (only streams designated as major rivers).

53

Data Type Indicator or Column Heading Description

Site

Des

crip

tors

StreamSizeBankfull

Stream size category as defined by bankfull width and protocol. SmallWadeable: wadeable streams <10 m bankfull width, LargeWadeable: wadeable stream >10 m bankfull width, Boatable: All boatable streams regardless of bankfull width.

NAMC_Benchmark NAMC assigned category used to determine default benchmarks for making indicator specific condition ratings. This field is a combination of EcoregionHybrid10 and StreamSizeBankfull.

EcoregionHybrid10 EPA hybrid level III ecoregion Climate EPA climatic zone (Mountain, Xeric, Plains) BRLat Bottom of reach latitude in NAD 83 (units: decimal degrees) BRLong Bottom of reach longitude in NAD 83 (units: decimal degrees) TRLat Top of reach latitude in NAD 83 (units: decimal degrees) TRLong Top of reach longitude in NAD 83 (units: decimal degrees)

TotRchlen

Total length of the reach (m) measured along the thalweg as calculated by 20 times average bankfull width (wadeable) or 40 times wetted width (boatable), with a min of 150 m and a max of 4000 m. This field is provided for context for the site but sampled reach lengths may differ from this total reach length for partially sampled sites (FieldStatus= Sampled - Partial). (units: m, min: 150, max: 4000, n=1)

FieldStatus

Whether the reach was fully sampled, partially sampled, or sampled with interrupted flow. Data from partially sampled sites or sites with interrupted flow should be examined carefully to insure crews followed the modified protocols properly.

54

Data Type Indicator or Column Heading Description

Biod

iver

sity

and

Ripa

rian

Habi

tat Q

ualit

y

PctOverheadCover Average % overhead cover provided by stream banks, vegetation, or other objects measured mid-channel (looking 4 directions) across 11 transects (units: %, min: 0, max: 100, n= 44)

BankOverheadCover Average percent overhead cover provided by stream banks (left and right), vegetation or other objects measured at the scour line of the left and right banks across 11 transects (units: %, min: 0, max: 100, n= 22)

VegComplexity

Aggregate measure of the average vegetative cover provided by three different vegetative height category: Canopy (>5m), Understory (0.5-5m), and Ground (<0.5m). Each vegetative height category is then divided into two vegetation types (e.g. woody or nonwoody). Proportional cover was binned into four classes (0.875, 0.575, 0.25, and 0.05) per vegetation type, summed across the three heights, and then averaged across the left and right banks of 11 transects. (units: none, min: 0, max: 2.6, n= 132)

RiparianVegCanopyCover

Measure of the average riparian vegetative cover provided by canopy vegetation (>5m). Proportional cover was binned into four classes (0.875, 0.575, 0.25, and 0.05) and then averaged across the left and right banks of 11 transects. (units: none, min: 0, max: 0.88, n= 22)

RiparianVegUnderstoryCover

Measure of the average riparian vegetative cover provided by understory vegetation (0.5-5m). Proportional cover was binned into four classes (0.875, 0.575, 0.25, and 0.05) and then averaged across the left and right banks of 11 transects. (units: none, min: 0, max: 0.88, n= 22)

RiparianVegGroundCover

Measure of the average riparian vegetative cover provided by the ground cover vegetation (<0.5m). Proportional cover was binned into four classes (0.875, 0.575, 0.25, and 0.05) and then averaged across the left and right banks of 11 transects. (units: none, min: 0, max: 0.88, n= 22)

NonNativeWoody Percent of 22 vegetation plots with invasive woody vegetation present (units: %, min: 0, max: 100, n= 22)

NativeWoody Percent of 22 vegetation plots with native woody vegetation present (units: %, min: 0, max: 100, n= 22)

NonNativeHerb Percent of 22 vegetation plots with invasive herbaceous vegetation present (units: %, min: 0, max: 100, n= 22)

NativeHerb Percent of 22 vegetation plots with native herbaceous vegetation present (units: %, min: 0, max: 100, n= 22)

SedgeRush Percent of 22 vegetation plots with sedges and rushes present (units: %, min: 0, max: 100, n= 22)

55

Data Type Indicator or Column Heading Description

Biod

iver

sity

and

Ripa

rian

Habi

tat Q

ualit

y

InvasiveInvertSp Presence or absence of invasive macroinvertebrates

ObservedInvertRichness Observed macroinvertebrate richness standardized to model specific operational taxonomic units (OTU) (units: # of taxa)

ExpectedInvertRichness Expected macroinvertebrate richness in the absence of anthropogenic impacts from the O/E model (units: # of taxa)

OE_Macroinvertebrate

Biological condition was assessed using an observed/expected (O/E) index. O/E models compare the macroinvertebrate taxa observed at sites of unknown biological condition (i.e., ‘test sites’) to the assemblages expected to be found in the absence of anthropogenic stressors (see Hawkins et al. 2000 for details). The specific model used can be found in the OE_MMI_ModelUsed column and the model specific metadata can be found at www.usu.edu/buglab/. (units: none, min: 0, max: 1.5)

MMI_Macroinvertebrate Biological condition was assessed using the MMI (MultimetricIndex) model specified in the OE_MMI_ModelUsed column.

OE_MMI_ModelUsed

The O/E or MMI model used to determine biological integrity. NAMC currently has the following models available UT, NV, CA, CO, OR, regional models for areas sampled by AREMP or PIBO programs (Northwest Forest Plan or Columbia River Basin), and a West-wide model. Generally, State based models are used if available, otherwise the West-wide model is used.

MacroinvertebrateCount

This field indicates whether or not the site's environmental gradients were within the range of experience of the model. A fail indicates the model potentially had to extrapolate, rather than interpolate, to accommodate one or more of the habitat variables. O/E scores and condition ratings should be interpreted cautiously if a site failed the test of experience.

ModelApplicability

Number of macroinvertebrates identified and resampled to a standardized fixed count (i.e. rarefaction). Samples with counts less than 200 macroinvertebrates can result from sampling and/or laboratory processing errors, but low counts can also be a signal of degraded biological condition. Additional samples should be taken to verify Major or Moderate departure from reference. (units: # of individuals, min: 0, max: 400)

56

Data Type Indicator or Column Heading Description

Wat

er Q

ualit

y TotalNitrogen Measured total nitrogen value (units: µg/L, n=1) PRD_TotalNitrogen Site specific predicted values for reference nitrogen concentrations (Olson and Hawkins 2013) (units: µg/L) TotalPhosphorous Measured total phosphorous value (units: µg/L, n=1)

PRD_TotalPhosphorous Site specific predicted values for reference phosphorus concentrations (Olson and Hawkins 2013) (units: µg/L)

SpecificConductance Measured specific conductance value. The specific conductance is conductivity standardized to 25 degrees C. (units: µS/cm, min: 0, max: 65500, n=1)

PRD_SpecificConductance Site specific predicted values for reference specific conductance values (Olson and Hawkins 2012) (units: µS/cm, min: 0, max: 65500)

pH Measured pH value (units: SU, min: 0, max: 14, n=1) InstantTemp Instantaneous water temperature measurement (units: degrees C, n=1) Turbidity Average water clarity as measured by the suspended solids in the water column (units: NTU, n=3)

Wat

ersh

ed F

unct

ion

and

Inst

ream

Hab

itat Q

ualit

y PctPools Percent of the sample reach (linear extent) classified as pool habitat as assessed using the core pool method (units: %, min: 0, max: 100, n=1)

ResPoolDepth Average residual pool depth as assessed using the core pool method (units: m, n= variable depending on number of pools)

PoolFreq Frequency of pools in the reach as assessed using the core pool method (units: # pools/km, n=1) LWD_Freq Frequency of large woody debris within the bankfull channel of the reach (units: # pieces/ 100 m, n= 1) LWD_Vol Volume of LWD within the bankfull channel of the reach (units: m^3/100 m, n=1) PctFines Percent of 210 particles with a b-axis < 2 mm (units: %, min: 0, max: 100, n=210) PctFines6 Percent of 210 particles with a b-axis < 6 mm (units: %, min: 0, max: 100, n=210)

D16 Particle size corresponding to the 16th percentile of measured particles (units: mm, min: 1, max: 4098, n=210)

D84 Particle size corresponding to the 84th percentile of measured particles (units: mm, min: 1, max: 4098, n=210)

D50 Particle size corresponding to the 50th percentile of measured particles (units: mm, min: 1, max: 4098, n=210)

GeometricMeanParticleDiam

Geometric mean bed particle diameter= exponential function[mean(log(particle diameter)]. This is a less frequently used metric of characterizing central tendency of substrate sizes, but is the main metric used by the EPA for determining relative bed stability. It is less variable than a D50 and more biologically meaningful because it is more influenced by fine sediment. (units: mm, min: 1, max: 4098, n=210)

57

Data Type Indicator or Column Heading Description

Wat

ersh

ed F

unct

ion

and

Inst

ream

Hab

itat Q

ualit

y

PoolTailFines Average percent fine sediment (< 2mm) on the pool tail (units %, min: 0, max: 100, n= 3 per pool) PoolTailFines6 Average percent fine sediment (< 6mm) on the pool tail (units %, min: 0, max: 100, n=3 per pool)

BankCover Percent of 42 erosional banks with greater than 50% cover provided by perennial vegetation, wood or mineral substrate > 15 cm (units: %, min: 0, max: 100, n= 42)

BankStability Percent of 42 banks lacking visible signs of active erosion (e.g., slump, slough, fracture) (units: %, min: 0, max: 100, n= 42)

BnkCover_Stab

Percent of 42 banks both stable (lacking visible signs of active erosions (e.g., slump, slough, fracture)) and covered (greater than 50% cover provided by perennial vegetation, wood or mineral substrate > 15 cm) (units: %, min: 0, max: 100, n= 42)

BnkCoverBedrock Average bank cover composed of bedrock (units: %, min: 0, max: 100, n= 42) BnkCoverCobble Average bank cover composed of cobble > 15 cm (units: %, min: 0, max: 100, n= 42) BnkCoverLWD Average bank cover composed of LWD (units: %, min: 0, max: 100, n= 42) BnkCoverVeg Average bank cover composed of vegetation (units: %, min: 0, max: 100, n= 42) BankfullHeight Average bankfull height measured from water surface across 11 transects (units: m, n = 11) FloodplainHeight Average floodplain height measured from water surface across 11 transects (units: m, n = 11)

FloodplainConnectivity Logarithm of the difference between average bankfull height and average floodplain height= log(FloodplainHeight - BankHeight + 0.1) (units: none, min: -1, max: 2, n=11)

InstreamHabitatComplexity

Aggregate measure of average cover provided by boulders, overhanging vegetation, live trees and roots, LWD, small woody debris, and stream banks for stream fishes measured at 11 plots. Proportional cover was binned into four classes (0.875, 0.575, 0.25, and 0.5), averaged across transects, and then summed across six types of cover. (units: none, min: 0, max: 2.3, n= 66)

BankAngle Measured angle of the stream bank; banks with obtuse angles = >90° and undercut banks with acute angles = <90° (units: degrees, min: 0, max: 180, n= 22)

ThalwegDepthCV Indicator of bed heterogeneity computed as the coefficient of variation of 100-300 thalweg depth measurements (units: none, n= 1)

ThalwegDepthMean Mean thalweg depth. Metric of how deep water was at the site. (units: m, min: 0, max: none, n variable depending on reach length (100 - 300))

PctDry

Percent of the reach that was dry. This is calculated as the number of dry thalweg measurements divided by the total number of thalweg measurements collected and expressed as a percentage. (units: %, min: 0, max: 100, n= variable depending on reach length (100-300))

58

Data Type Indicator or Column Heading Description

Cova

riate

s/ O

ther

BankfullWidth Average bankfull width across 11 transects (units: m, n= 11) WettedWidth Average wetted width across 21 transects (units: m, n= 21)

FloodWidth Average flood prone width as defined as valley width at 2 times bankfull height. The larger the value the larger the floodplain is. (units: m, n= 2)

Entrench

Entrenchment ratio = average floodprone width divided by average bankfull width. Ratios of 1-1.4 represent entrenched streams; 1.41-2.2 represent moderately entrenched streams; and ratios greater than 2.2 indicate rivers only slightly entrenched in a well-developed floodplain (Rosgen 1996). This entrenchment value can then be used with other ancillary site data such as slope and incision to determine stream type (Rosgen 1996) and the potential of the system for floodplain formation. (units: none, min: 1, max: 3, n= 1)

Slope Reach slope measured from the water's surface. In most cases, the reported value is an average of 2 independent measurements that were within 10% of one another. (units: %, min: 0, max: ~45, n= 2)

Sinuosity Reach sinuosity (reach length along the thalweg/straight line distance between BR and TR coordinates) (units: none, min: 1, max: NA, n= 1)

BeaverFlowMod Qualitative visual assessment of extent of beaver flow modifications within the reach (NONE, MINOR, MAJOR)

BeaverSign Qualitative visual assessment of frequency of beaver signs (e.g. chewed logs) within the reach (ABSENT, RARE, COMMON)

WaterWithdrawal Presence or absence of water withdrawals DateChange Date that the site's data was updated and changed (units: m/d/yyyy) ReasonChange Reason for change and field changes; detailing what data was changed and why

59

Appendix B. Macroinvertebrate Model Metadata Content in Development

60

Literature Cited Al-Chokhachy, R., B. B. Roper, and E. K. Archer. 2010. Evaluating the Status and Trends of

Physical Stream Habitat in Headwater Streams within the Interior Columbia River and Upper Missouri River Basins Using an Index Approach. Transactions of the American Fisheries Society 139:1041–1059.

Allan, J. D. 2004. Landscapes and riverscapes: The Influence of Land Use on Stream Ecosystems. Annual Review of Ecology, Evolution, and Systematics 35:257–284.

Allan, J. D., and M. M. Castillo. 1995. Stream Ecology: Structure and function of running waters. Second. Springer.

Allan, J. D., and A. S. Flecker. 1993. Biodiversity Conservation in Running Waters. BioScience 43:32–43.

Baxter, C. V., K. D. Fausch, and W. C. Saunders. 2005. Tangled webs: reciprocal flows of invertebrate prey link streams and riparian zones. Freshwater Biology 50:201–220.

Beschta, R. L. 1997. Riparian Shade and Stream Temperature: An Alternative Perspective. Rangelands 19:25–28.

Beschta, R. L., D. L. Donahue, D. a Dellasala, J. J. Rhodes, J. R. Karr, M. H. O’Brien, T. L. Fleischner, and C. Deacon Williams. 2012. Adapting to Climate Change on Western Public Lands: Addressing the Ecological Effects of Domestic, Wild, and Feral Ungulates. Environmental management.

Bjornn, T., and D. Reiser. 1991. Habitat requirements of salmonids in streams. American Fisheries Society Special Publication 19:83–138.

Blasius, B. J., and R. W. Merritt. 2002. Field and laboratory investigations on the effects of road salt (NaCl) on stream macroinvertebrate communities. Environmental Pollution 120:219–231.

Bonada, N., N. Prat, V. H. Resh, and B. Statzner. 2006. Developments in aquatic insect biomonitoring: a comparative analysis of recent approaches. Annual review of entomology 51:495–523.

Burton, T. M., R. M. Stanford, and J. W. Allan. 1985. Acidification effects on stream biota and organic matter processing. Canadian Journal of Fisheries and Aquatic Sciences 42:669–675.

Coles-Ritchie, M. C., D. W. Roberts, J. L. Kershner, and R. C. Henderson. 2007. Use of a Wetland Index to Evaluate Changes in Riparian Vegetation After Livestock Exclusion. Journal of the American Water Resources Association 43:731–743.

Cummins, K. W. 1974. Structure and Function Stream Ecosystems. BioScience 24:631–641. Cunjak, R. A., and M. E. Power. 1986. Winter habitat utilization by stream resident brook trout

(Salvelinus fontinalis) and brown trout (Salmo trutta). Canadian Journal of Fisheries and Aquatic Sciences 43:1970–1981.

Dangles, O., M. O. Gessner, F. Guerold, and E. Chauvet. 2004. Impacts of stream acidification on litter breakdown: implications for assessing ecosystem functioning. Journal of Applied Ecology 41:365–378.

Dynesius, M., and C. Nillson. 1994. Fragmentation and flow regulation of river systems in the northern third of the world. Science 266:753–762.

Fischman, R. L., and J. B. Ruhl. 2016. Judging adaptive management practices of U.S. agencies. Conservation Biology 30:268–275.

Fleischner, T. L. 1994. Ecological Costs of Livestock Grazing in Western North America. Conservation Biology 8:629–644.

Gries, G., and F. Juanes. 1998. Microhabitat use by juvenile Atlantic salmon (Salmo salar)

61

sheltering during the day in summer. Canadian Journal of Zoology 76:1441–1449. Hargett, E. G., J. R. ZumBerge, C. P. Hawkins, and J. R. Olson. 2007. Development of a

RIVPACS-type predictive model for bioassessment of wadeable streams in Wyoming. Ecological Indicators 7:807–826.

Hauer, F. R., and A. C. Benke. 1987. Influence of temperature and river hydrograph on black fly growth rates in a subtropical blackwater river. Journal of the North American Benthological Society 6:251–61.

Hauer, R., and G. A. Lamberti. 1998. Methods in Stream Ecology. 1st edition. Academic Press. Hawkins, C. P. 2006. Quantifying biological integrity by taxonomic completeness: its utility in

regional and global assessments. Ecological applications 16:1277–94. Hawkins, C. P., Y. Cao, and B. Roper. 2010a. Method of predicting reference condition biota

affects the performance and interpretation of ecological indices. Freshwater Biology 55:1066–1085.

Hawkins, C. P., J. L. Kershner, P. A. Bisson, M. D. Bryant, L. M. Decker, S. V. Gregory, D. A. Mccullough, C. K. Overton, G. H. Reeves, R. J. Steedman, M. K. Young, M. Lynn, S. V. Gregory, D. A. Mccullough, C. K. Overton, G. H. Reeves, R. J. Steedman, and M. K. Young. 1993. A Hierarchical Approach to Classifying Stream Habitat Features. Fisheries 18:3–12.

Hawkins, C. P., R. H. Norris, J. N. Hogue, J. W. Feminella, and W. S. Unit. 2000. Development and Evaluation of Predictive Models for Measuring the Biological Integrity of Streams. Ecological Applications 10:1456–1477.

Hawkins, C. P., J. R. Olson, and R. a Hill. 2010b. The reference condition: predicting benchmarks for ecological and water-quality assessments. Journal of the North American Benthological Society 29:312–343.

Henley, W. F., M. A. Patternson, R. J. Neves, and A. D. Lemly. 2000. Effects of Sedimentation and Turbidity on Lotic Food Webs : A Concise Review for Natural Resource Managers. Reviews in Fisheries Science 8:125–139.

Herbst, D. B., M. T. Bogan, S. K. Roll, and H. D. Safford. 2012. Effects of livestock exclusion on in-stream habitat and benthic invertebrate assemblages in montane streams. Freshwater Biology 57:204–217.

Herlihy, A. T., S. G. Paulsen, J. Van Sickle, J. L. Stoddard, C. P. Hawkins, L. L. Yuan, and J. Van Sickle. 2008. Striving for consistency in a national assessment: the challenges of applying a reference-condition approach at a continental scale. Journal of the North American Benthological Society 27:860–877.

Hill, R. A., C. P. Hawkins, and D. M. Carlisle. 2013. Predicting thermal reference conditions for USA streams and rivers. Freshwater Science 32:39–55.

Hughes, R. M., S. A. Heiskary, W. J. Matthews, and C. O. Yoder. 1994. Use of ecoregions in biological monitoring. Pages 125–151in S. L. Loeb and A. Spacie, editors.Biological monitoring of aquatic systems. Lewis Publishers, Boca Raton, Florida.

Hughes, R. M., D. P. Larsen, and J. M. Omernik. 1986. Regional reference sites: a method for assessing stream potentials. Environmental management 10:629–635.

Hurlbert, S. H. 1984. Pseudoreplication and the Design of Ecological Field Experiments. Ecological Monographs 54:187–211.

Johnson, S. L., and J. A. Jones. 2000. Stream temperature responses to forest harvest and debris flows in western Cascades, Oregon. Canadian Journal of Fisheries and Aquatic Sciences 57:30–39.

62

Kaufmann, P. R., J. M. Faustini, D. P. Larsen, and M. A. Shirazi. 2008. A roughness-corrected index of relative bed stability for regional stream surveys. Geomorphology 99:150–170.

Kaufmann, P. R., D. P. Larsen, and J. M. Faustini. 2009. Bed Stability and Sedimentation Associated With Human Disturbances in Pacific Northwest Streams. Journal of the American Water Resources Association 45:434–459.

Kaufmann, P. R., P. Levine, E. G. Robison, C. Seeliger, and D. V Peck. 1999. Quantifying Physical Habitat in Wadeable Streams. EPA/620/R-99/003. U.S. Environmental Protection Agency, Washington, D.C.:130.

Knapp, R. a., and K. R. Matthews. 1996. Livestock Grazing, Golden Trout, and Streams in the Golden Trout Wilderness, California: Impacts and Management Implications. North American Journal of Fisheries Management 16:805–820.

Vander Laan, J. J., and C. P. Hawkins. 2014. Enhancing the performance and interpretation of freshwater biological indices : An application in arid zone streams. Ecological Indicators 36:470–482.

Vander Laan, J. J., C. P. Hawkins, J. R. Olson, R. A. Hill, and J. J. Vander Laan. 2013. Linking land use , in-stream stressors , and biological condition to infer causes of regional ecological impairment in streams Linking land use , in-stream stressors , and biological condition to infer causes of regional ecological impairment in streams 32:801–820.

Miller, S. W., D. Wooster, and J. Li. 2007. Resistance and resilience of macroinvertebrates to irrigation water withdrawals. Freshwater Biology 52:2494–2510.

Muotka, T., and J. Syrjänen. 2007. Changes in habitat structure, benthic invertebrate diversity, trout populations and ecosystem processes in restored forest streams: A boreal perspective. Freshwater Biology 52:724–737.

Naiman, R. J., and H. Decamps. 1997. The Ecology of Interfaces: Riparian Zones. Annual Review of Ecology and Systematics 28:621–658.

Newbold, J. D., B. W. Sweeney, R. L. Vannote, S. Journal, N. American, B. Society, and N. Mar. 1994. A Model for Seasonal Synchrony in Stream Mayflies. Journal of the North American Benthological Society 13:3–18.

Ode, P. R., C. P. Hawkins, and R. D. Mazor. 2008. Comparability of biological assessments derived from predictive models and multimetric indices of increasing geographic scope. Journal of the North American Benthological Society 27:967–985.

Ode, P. R., A. C. Rehn, R. D. Mazor, K. C. Schiff, E. D. Stein, J. T. May, L. R. Brown, D. B. Herbst, D. Gillett, and K. Lunde. 2016. Evaluating the adequacy of a reference-site pool for ecological assessments in environmentally complex regions 35:237–248.

Olsen, A. R., and D. V. Peck. 2008. Survey design and extent estimates for the Wadeable Streams Assessment. Journal of the North American Benthological Society 27:822–836.

Olson, J. R. 2012. The Influence of Geology and Other Environmental Factors on Stream Water Chemistry and Benthic Invertebrate Assemblages. Utah State University.

Olson, J. R., and C. P. Hawkins. 2012. Predicting natural base-flow stream water chemistry in the western United States. Water Resources Research 48:W02504.

Olson, J. R., and C. P. Hawkins. 2013. Developing site-specific nutrient criteria from empirical models. Freshwater Science 32:719–740.

Paulsen, S. G., C. P. Hawkins, J. Van Sickle, L. L. Yuan, and S. M. Holdsworth. 2008a. An invitation to apply national survey data to ecological research. Journal of the North American Benthological Society 27:1017–1018.

Paulsen, S. G., A. Mayio, D. V. Peck, J. L. Stoddard, E. Tarquinio, S. M. Holdsworth, J. Van

63

Sickle, L. L. Yuan, C. P. Hawkins, A. T. Herlihy, P. R. Kaufmann, M. T. Barbour, D. P. Larsen, A. R. Olsen, and J. Van Sickle. 2008b. Condition of stream ecosystems in the US: an overview of the first national assessment. Journal of the North American Benthological Society 27:812–821.

Stoddard, J. L. J., A. T. A. T. Herlihy, D. D. V. Peck, R. M. R. Hughes, T. R. T. Whittier, and E. Tarquinio. 2008. A process for creating multimetric indices for large-scale aquatic surveys. JOURNAL OF THE NORTH AMERICAN BENTHOLOGICAL SOCIETY 27:878–891.

Stoddard, J. L., D. P. Larsen, C. P. Hawkins, R. K. Johnson, and R. H. Norris. 2006. Setting expectations for the ecological condition of streams: the concept of reference condition. Ecological Applications 16:1267–76.

Stoddard, J. L., D. V. Peck, A. R. Olsen, D. P. Larsen, J. Van Sickle, C. P. Hawkins, R. M. Hughes, T. R. Whittier, G. Lomnicky, A. T. Herlihy, P. R. Kaufmann, S. A. Peterson, P. L. Ringold, S. G. Paulsen, and R. Blair. 2005. Western Streams and Rivers Statistical Summary Environmental Monitoring and Assessment Program (EMAP) Western Streams and Rivers Statistical Summary. EPA 620/R-05/006. US Environmental Protection Agency, Office of Research and Development, Washington, D. C.

Toddard, J. O. H. N. L. S., D. A. P. L. Arsen, C. H. P. H. Awkins, R. I. K. J. Ohnson, J. L. Stoddard, D. P. Larsen, C. P. Hawkins, R. K. Johnson, and R. H. Norris. 2006. Setting expectations for the ecological condition of streams: the concept of reference condition. Ecological Applications 16:1267–76.

USEPA. 2002. Summary of Biological Assessment Programs and Biocriteria Development for States, Tribes, Territories, and Interstate Commissions: Streams and Wadeable Rivers. EPA-822-R-02-048. U.S. Environmental Protection Agency, Washington D.C.

USEPA. 2009. National Rivers and Streams Assessment Field Operations Manual. EPA-841-B-07-009. U.S. Environmental Protection Agency, Washington, D. C.

Vannote, R. L., and B. W. Sweeney. 1980. Geogrpahic analysis of thermal equilibria: a conceptual model for evaluating the effect of natural and modified thermal regimes on aquatic insect communities. The American naturalist 115:667–695.

Vinson, M. R., and C. P. Hawkins. 1996. Effects of Sampling Area and Subsampling Procedure on Comparisons of Taxa Richness among Streams. Journal of the North American Benthological Society 15:392–399.

Ward, J. V., and J. A. Stanford. 1982. Thermal Resonses in the Evolutionary Ecology of Aquatic Insects. Annual review of entomology 27:97–117.

Wolman, M. G., and J. P. Miller. 1960. Magnitude and Frequency of Forces in Geomorphic Processes. The Journal of Geology 68:54–74.

Wood, P., and P. Armitage. 1997. Biological Effects of Fine Sediment in the Lotic Environment. Environmental management 21:203–17.