Differential Perceptions of Buffelgrass (Pennisetum ... · species, and what factors influence this perception? The specific realms of scientific literature that can be used to answer

Differential Perceptions of Buffelgrass (Pennisetum

ciliare) in the Sonoran Desert of Tucson, Arizona

Sam Walker

GEOG 499: Honor Studies in Geography

May 2010

ABSTRACT: The invasive plant buffelgrass (Pennisetum ciliare) poses a threat to the

saguaro (Carnegiea gigantea)-palo verde (Parkinsonia florida) desert scrub around

Tucson, Arizona. Buffelgrass outcompetes native grasses and can cause the local

extinction of iconic species such as the saguaro cactus by introducing the threat of

wildfire. While the physical dimensions of this invasion have been studied for 20 years,

the social factors that help and hinder the spread of the plant around Tucson are poorly

understood. This study seeks to understand Tucsonans’ perceptions of the invasive grass

through a novel conceptualization of invasive species perceptions based on an

environmental knowledge, attitudes, and behaviors framework. A mail survey yielded a

sample of 122 randomly selected residents from two subgroups living at the urban-

wildland interface (UWI) and in the central city. These respondents were grouped using

cluster analysis into five significantly different stakeholder groups. Statistical tests and

linear regression show that many demographic and attitudinal variables significantly

affect the public’s perception of buffelgrass. The findings suggest that while

demographics can be used to predict public perceptions of invasive species, a complex

interaction of numerous factors exists. The results have implications for theoretical

understanding of public perceptions of invasive species and their management.

KEYWORDS: Buffelgrass, invasive species, public perceptions, risk perception,

Tucson.

Differential Perceptions of Buffelgrass Walker 2

Table of Contents

I. Acknowledgements ................................................................................................................ 5

II. Introduction ............................................................................................................................ 6

Buffelgrass in the Sonoran Desert ...................................................................................... 6

Literature review ................................................................................................................. 8

Demographics, values, behavior, knowledge, and experience ............................... 9

Identification of invasive species .......................................................................... 10

Risk perception ..................................................................................................... 11

Favored response to invasion ................................................................................ 12

Methods and research questions ....................................................................................... 13

III. Methods................................................................................................................................ 16

Survey design .................................................................................................................... 16

Survey sample ................................................................................................................... 18

Variable relationships and statistical analysis................................................................... 20

Cluster analysis ................................................................................................................. 24

IV. Results .................................................................................................................................. 28

Sample............................................................................................................................... 28

Overall results ................................................................................................................... 30

Variable relationships ....................................................................................................... 33


Variable relationships – Background on intervening ............................................ 34

Variable relationships – Intervening on intervening ............................................. 40

Variable relationships – Intervening on outcome ................................................. 42

Variable relationships – Background on outcome ................................................ 43

Variable relationships – Outcome on outcome ..................................................... 46

Regression analysis ........................................................................................................... 46


V. Discussion ............................................................................................................................ 52

Demographics and sample ................................................................................................ 52

Variable relationships – Background and Intervening ...................................................... 53

Essential species identification ......................................................................................... 55

Variable relationships – Outcome ..................................................................................... 56

Risk perception ..................................................................................................... 56

Preferred management response ........................................................................... 57

Buffelgrass identification skill .............................................................................. 62


VI. Conclusion ........................................................................................................................... 68

Management implications ................................................................................................. 68

Recommendations for future research .............................................................................. 70

VII. Appendix A – Survey Instrument ........................................................................................ 72


VIII. Appendix B – Variable Descriptions ................................................................................... 84

IX. Appendix C – Scale Variable Creation ................................................................................ 85

X. Appendix D – Gower’s Similarity Coefficient from Wishart (2006) .................................. 88

XI. Appendix E – Increase in Sum of Squares or Ward’s Method from Wishart (2006) .......... 89

XII. Appendix F – Sample Demographics .................................................................................. 90

XIII. Appendix G – Regression Analysis Results ........................................................................ 92

XIV. Appendix H – Cluster Analysis Results............................................................................... 94


I. Acknowledgements

This study would not have been possible without the guidance and aid of my

research and thesis advisors Jessica Graybill and Jake Brenner. For additional help on

various sections of my paper I owe a debt to professors Ellen Kraly and Peter Klepeis and

the work of former Colgate student Dara Seidl (‘10). My survey could not have been

made without data generously provided by the Pima County GIS website and the

excellent work of Bob Keats and Christine Scheve from the Colgate Print Shop, and I am

thankful for their aid. Colgate student Drew Colbert (‘11) also assisted in the data entry

for this project. Finally, I would also like to thank the Colgate Geography Department

for providing a stimulating and rewarding undergraduate education and for aiding me in

acquiring a Student Wage Grant and a Summer Undergraduate Research Grant to support

my undergraduate research.


II. Introduction

Invasive alien species (IAS) pose a growing threat to humans and their

environment, causing an estimated $120 billion of damage every year in the United States

alone (Pimentel et al., 2005). While the ecological aspects of IAS invasions are

increasingly well-studied and funded by government agencies, the spread of such

organisms depends largely on human beings and their social relationships and networks

(Robbins, 2004). Furthering understanding of how the public understands and relates to

IAS is critical for developing effective management strategies to reduce their impact

(García-Llorente et al., 2006). This study aims to improve that understanding through a

case study of Pennisetum ciliare, or buffelgrass, in the Sonoran Desert of Tucson,

Arizona.

Buffelgrass in the Sonoran Desert

The non-native, invasive buffelgrass poses a threat that could drastically alter the

Sonoran Desert of the southwestern United States and northern Mexico. Since its

introduction as a ‘miracle grass’ to battle drought and overgrazing problems facing cattle

ranchers in the late 19th and early 20th century, buffelgrass has spread profusely,

outcompeting native plants (Bowers et al., 2006; Brooks et al., 2004; Burgess et al., 1991;

Van Devender et al., 1997; Van Devender & Dimmitt, 2006; Williams & Baruch, 2000).

Buffelgrass poses 2 major ecological problems: competition and the introduction of

wildfire to the Arizona Upland desert scrub environment, a biome within which few

native plants have adaptation to fire ( Van Devender et al., 1997; Van Devender &

Dimmitt, 2006). Thus, buffelgrass invasion may potentially convert desert scrub into

grassland with greatly reduced species richness, eradicating iconic native species such as


saguaro (Carnegiea gigantea) and organ pipe (Stenocereus thurberi) cacti (Bowers et al.,

2006). Competition from buffelgrass has in fact been shown to cause the local extinction

of native cacti (Pachycereus pecten-aboriginum) in the Mexican Sonoran Desert (Lyons,

2009; Morales-Romero, 2008).

The city of Tucson is an appropriate place to study the invasion of buffelgrass and

the social perceptions of the grass because the invasion is being addressed on multiple

levels by various stakeholders (see Frost, 2010; SABCC, 2010), and the invasion has

been studied since the 1990s (Burgess et al., 1991; Yetman & Búrquez, 1994). From an

ecological standpoint, the relationship between invasive plants, fire, and the loss of native

plant cover is widely covered in the literature (Brooks et al., 2004; Rossiter et al., 2003).

Fire is also an environmental hazard to the citizens of Tucson due to both the rapid

expansion of the urban environment and encroachment of buffelgrass into more densely

populated areas.

For these reasons, Tucson also presents a unique opportunity to explore the social

perceptions of an invasive species geographically. The social and physical environments

of the central city and the urban-wildland interface (UWI), where human residential

development encroaches on undeveloped wildlands (Radeloff et al., 2005; Stewart et al.,

2007), are drastically different. Residents of the UWI face much more direct threats from

the buffelgrass invasion, with their homes being threatened by fire and the vast majority

of buffelgrass appearing on disturbed land or invading the iconic desert scrub near their

residences (Frost et al., 2010; Van Devender & Dimmitt, 2006). Comparing these two

subgroups of the Tucsonan population presents a unique opportunity to explore the

nature-society relationship geographically. This aspect of the invasion, where the social


and ecological impacts meet, is only just beginning to be explored in the literature.

Recent work by Brenner (2010, 2011) and Marshall et al. (2010) has focused on

landowner responses to buffelgrass in Mexico and Australia, respectively, and this study

represents the first attempt to understand the social dimensions of the buffelgrass

invasion in Arizona.

Literature review

Before attempting to investigate the social dimensions of the buffelgrass invasion,

a conceptual framework for analyzing public perceptions of invasive species must be

developed. The successful management of invasive species requires further

understanding of the ecological and social aspects at play in biological invasions, and

understanding the public’s perception of invaders and the risks they pose is critical to

developing such a conceptual framework (Andersen et al., 2004). Geographical study can

also inform the study of invasive species perceptions by investigating the spatial

component to this social issue.

To aid in developing a conceptual understanding of this issue, the following

question is asked: what measurable concepts are part of the public perception of invasive

species, and what factors influence this perception? The specific realms of scientific

literature that can be used to answer this question include studies of environmental

knowledge-attitudes-behavior, risk perception, and invasive species management. A

synthesis of existing work drawn from these areas of research is displayed in Figure 1.

The diagram is split into factors influencing perceptions (demographics, values, behavior,

knowledge, and experience) and the concepts that are part of invasive species perceptions

(risk perception, favored invasion response, and identification of invasive species). By


synthesizing the findings of relevant studies, this conceptualization will underpin the

investigation of invasive species perceptions throughout this paper. It is important to

note that this model is generalized and holistic and seeks to address the full spectrum of

factors influencing perceptions of invasive species. To justify the choices made in this

model, relevant findings from the literature will be briefly discussed.

Demographics Behavior

Values

Experience Risk perceptionFavored invasion

response

Identification of invasive species

Knowledge

Social

Am

plificatio

n o

r A

ttenu

ation

FACTORS PERCEPTIONS

Figure 1. Conceptualization of public perceptions of invasive species.

Demographics, values, behavior, knowledge, and experience

The inclusion of demographics, values, behavior, knowledge, and experience in

this model is based on the findings of psychological studies addressing the relationship

between environmental knowledge, attitudes, and behavior. This field of study seeks to

understand how individual’s personal traits can influence their opinions and actions when

addressing environmental issues and has already found application to the study of

invasive species perceptions in Seidl & Klepeis (2011). Although not all of these studies

specifically address invasive species, the assumption is made that perceptions of issues

that pose similar environmental or ecological threats share conceptual similarities. By

measuring such personal attributes, researchers can predict how individuals will view

invasive species based on independent variables such as environmental attitudes or


sociodemographics.

The relationships between these concepts have all been thoroughly explored in the

literature, but several key findings are of particular importance to this conceptualization.

The relationship between demographics and environmental values was explored by

Slimak & Dietz (2006), who found that social demographic influences (such as gender,

age, and ethnicity) combined with social structural influences (including education,

income, political views) to inform individuals’ environmental attitudes and values. The

connection of environmental values and knowledge with behavior was established by

Kaiser et al. (1999) through their theory of reasoned action. This theory states that

individuals’ factual knowledge combined with their social and moral values influences

their behavior; the results of their study support this theory for the environmental issue of

ecologically responsible transportation use. Additional support for the relationship

between environmental values and knowledge with behavior is provided by Fransson &

Gärling (1999). The final relationship in this section of the conceptualization is between

behavior and experience. Carrying out certain behaviors naturally results in experiences

related to the intended behavior, and this relationship works the other way also, with

experiences influencing future behavior. Therefore, this conceptual relationship is based

on logical inference.

Identification of invasive species

As Figure 1 shows, individuals’ success at identifying invasive species is posited

to be a function of their experience and knowledge. It is logical that individuals’ past

encounters with a specific species have the capacity to influence their understanding and

knowledge of the organism, thereby affecting their ability to successfully identify it


visually. Indeed, many campaigns or management programs seeking to combat the

spread of invasive species seek to provide the public with pictures of the plant or animal

in question in an effort to spread awareness. Because no previous attempts to understand

individuals’ success at identifying invasive species were found in the literature,

experience and knowledge are chosen as the determinants of this aspect of perception.

Risk perception

Perhaps the most well studied aspect of invasive species perceptions is public risk

perception. As more invasive species management programs seek to account for the

opinions and attitudes of multiple stakeholders including the public (Maguire, 2004;

McNeely, 2001; Stokes et al., 2006), understanding how individuals make risk judgments

regarding invasive species is crucial. Numerous authors (e.g. (McDaniels et al., 1995;

Slimak & Dietz, 2006)) have shown that individuals’ environmental values can influence

their risk perception of ecological issues. Such studies build on the work of Stern’s

(1999, 2000) Value-Belief-Norm theory, which seeks to create a psychological

explanation of the formation of environmental or environmentalist values. Other

researchers (Halpern-Felsher et al., 2001) have linked knowledge of and experience with

natural disasters to individuals’ risk perception, providing impetus to conceptually link

these three concepts in the model of invasive species perceptions.

The social amplification and attenuation of risk is an important concept from the

risk perception literature which can aid in conceptualizing invasive species perceptions.

Kasperson et al. (1988: 177) first introduced this concept by theorizing that “hazards

interact with psychological, social, institutional, and cultural processes in ways which

amplify or attenuate public responses to the risk or risk event.” Through means such as


the media (Koné & Mullet, 1994; Wåhlberg & Sjöberg, 2000), organizations, personal

relationships, and other entities, individuals’ risk perception can be altered, a process

which must be accounted for in this study’s conceptual model. In order to do so, the

social amplification and attenuation modifies the relationship between individuals’

values, experience, and knowledge and their risk perception, as indicated in Figure 1.

Favored response to invasion

The final aspect of invasive species perceptions addressed by this study is

individuals’ favored management response to the invasion. Although this facet of

perceptions could be seen as separate from risk perception because it implies a certain

course of explicit action, it is included in the model due to its significance for invasive

species management. The relationship between hazard experience, risk perception, and

favored response is perhaps best discussed by the late Gilbert White, a pioneering

geographer in the field of natural hazards. As discussed in Burton et al. (1968), White’s

work in floodplain management revealed that the public’s experience with natural

hazards informs their risk perception and their ‘adjustment’ or response. Often, people

with intimate experience with hazard events are in fact less likely to act, especially if the

hazard occurs at a low frequency. This significant finding could be applied to the

public’s perception and subsequent response to invasive species, whose impacts are

spatially and temporally hard to predict and often even more difficult to manage than

flood events.

In addition to the research of natural hazard risk, some research has already been

conducted in linking the public’s specific risk perception of invasive species with their

preferred management response (Andreu et al., 2009; Fischer & van der Wal, 2007;


García-Llorente et al., 2008). These studies have found that the numerous complicated

factors influencing risk perception can also affect the public’s favored management

response, but that many of the other variables in this study’s conceptualization are

important, including demographics, values, knowledge, and experience.

Methods and research questions

The complex relationships present in the conceptualization presented in Figure 1

will be applied to the case of buffelgrass with two primary goals in mind, the first of

which seeks to improve the theoretical understanding of factors affecting the public’s

perceptions of invasive species. This goal will be accomplished through surveying

Tucsonans to explore the relationships presented in the conceptual model. The second

goal is to inform management schemes by clarifying the relationship between these

perceptions and favored management strategy. An important facet of the social response

to buffelgrass in Tucson is citizen involvement in management, with many Tucsonans

participating in ‘weedwacking’ activities to control the invasion (Arizona Native Plants

Society). Therefore, understanding Tucsonans’ knowledge, attitudes, and beliefs about

the Sonoran Desert and buffelgrass is integral to controlling the invasion because of their

intimate involvement. The successful implementation of a management scheme is aided

by citizens having knowledge of the invasive species and its negative effects and being

willing to support management programs (Hershdorfer, 2007; Tidwell, 2008) this study

provides a unique opportunity to explore how demographics and other factors affect

citizens’ views of an invasive species, an under-explored area of invasive species

management (García-Llorente et al., 2008).

To assess Tucsonans perceptions of buffelgrass, a mail survey was created and


administered in May 2010. It was sent to a stratified random sample of 800 Tucson

residents split into 2 groups: those living in central, urban census tracts, and those living

on the fringe of the city at the urban-wildland interface (UWI). 113 completed surveys

were returned and the responses from this 32 question survey were entered into a data set

containing 195 variables. Of this sample, approximately 40% of respondents live in the

central city and 60% live on the periphery of the city at the UWI, allowing useful

comparisons to be made between these 2 groups.

Based on the literature review conducted to generate the conceptualization of

public invasive species perceptions, the main research questions and hypothesized

answers of this study are:

Question 1.) Do Tucsonans perceive the invasion of buffelgrass as a risk?

Hypothesis 1.) The public will be concerned about the invasion.

Question 2.) How do Tucsonans characterize and value the native desert landscape?

What do they believe the essential elements of the Sonoran Desert ecosystem are?

What do they believe the Sonoran Desert is good for?

Hypothesis 2.) Residents will value the Sonoran Desert beyond its utilitarian or use value.

Question 3.) Do Tucsonans recognize buffelgrass invasion on the landscape?

Hypothesis 3.) Residents will be successful at identifying buffelgrass on the landscape

(>50% of the time) due to media attention and the widespread nature of the invasion.

Question 4.) What do Tucsonans believe are appropriate responses to buffelgrass

invasion?

Hypothesis 4.) Tucsonans will favor strong management options such as total eradication,

because most members of the public are not aware of the practical and budgetary

restraints on invasive plant management.

Question 5.) How do all of these opinions differ among members of the Tucson

community based on demographics (age, income, race, etc.), location within the city

(UWI versus central city residents), and their relationship with the desert (time spent

outdoors, utilitarian or conservation-based view of the desert)?

Hypothesis 5.) Older, more affluent, better educated UWI respondents who are more

‘conservationist’ in attitude towards the Sonoran Desert will be more concerned about the

invasion, favor stronger management responses, and be better at identifying buffelgrass.


These questions will be answered primarily through statistical analysis of the survey data

and their implications will be assessed through the use of scholarly literature.


III. Methods

Survey design

Data were collected using a mail survey sent to a stratified random sample of

residents of Tucson, Arizona in the summer of 2010 (see Appendix A for a copy of the

instrument). This 12 page survey included 32 questions. Dillman (1978) and Fink

(2003) provided useful instruction on question design and survey formatting; the

instrument used by Seidl & Klepeis (2011) also addressed public perceptions of invasive

species and therefore served as a useful model. For a description of the variables

measured by this survey for use in the study, please see Appendix B.

In addition to conventional narrative questions, this survey also made use of

photographs and a map to convey visual information included in the questions. Indeed,

visualization was essential in this survey, and visual identification of buffelgrass, native

vegetation, and other features was a specific capability the study sought to evaluate. The

first question to utilize this unconventional approach (Q13) displayed a matrix of 14

photographs of common Sonoran Desert plants. By providing both the common species

name and a representative photograph it was hoped that the respondents would be better

able to identify and connect with the plant, ensuring accurate responses to this question,

which asked which plants they believed were ‘essential’ to the Sonoran Desert. Similar

surveys with photographs have been employed by economic and conservation literature

(e.g. Damigos, 2011; Home, 2009), but their results have yet to be used in a study of

invasive species. This question was designed to include iconic plants such as the giant

saguaro and the ironwood and palo verde trees, other recognizable common species

including cacti, native grasses, and finally buffelgrass. By including this range of plants


we hoped to gain an understanding of respondents’ perceptions of the different types of

plants in the Sonoran Desert.

Another question (Q19) also made use of photographs in order to judge

respondents’ success at identifying buffelgrass invasion on the Sonoran Desert landscape.

Similar exercises are prevalent in the literature dealing with landscape attachment and

have been shown to produce replicable and theoretically consistent results (Kaplan &

Kaplan, 1989; Petrich, 1984; Ryan, 2005; Swihart & Petrich, 1988). Although the 2

photograph-based questions in this survey have different goals (eliciting respondents’

feelings and opinions in Q13 and testing their capabilities in Q19), both methods are

grounded in previous studies’ approaches to such questions.

In fact, such use of pictures to assess respondents’ skill at identifying invasive

grasses is a novel method and it is hoped that the results from this study will serve to

establish this method as a useful technique for incorporating this important variable in

further research. The photographs in this question come from the collections of

buffelgrass researchers (Brenner, personal communication, May 11, 2010) and were

chosen because they are clear representations of 4 stages of invasion as defined by

ground coverage, pattern, and extent (no buffelgrass, dispersed individuals, distinct

patches, and extensive, continuous buffelgrass) and can be seen in Appendix A.

Buffelgrass identification questions were coded based on respondents’ correct circling of

all buffelgrass; 1 point was assigned if all buffelgrass was correctly circled, 2 points if

more than fifty percent was correctly circled, and 3 points if no buffelgrass was correctly

circled. Respondents who incorrectly circled areas not covered by buffelgrass were given

3 points; Respondents who did not circle any patches on all 4 photographs were marked


as not responding and excluded from the analysis. The mean point score for all 4

landscapes was calculated and used as the scale variable representing success of

buffelgrass invasion on the landscape.

Aside from the photographs in Q13 and Q19, a visual component was also

included in Q24, which displayed a map of Tucson produced by a local realtor to display

the residential zones of the city (The Pepper Group Diversified, 2010). Respondents

were asked to indicate which zones of the city they felt were threatened by the

buffelgrass invasion and also in which zone they live. The data of respondents’ zone of

residence was only used to divide the sample into central city and urban-wildland

interface (UWI) and was not analyzed in the context of any other information. Visual

exercises were used in Q13, Q19, and Q24 with 2 main objectives: first, to ensure

accurate measurement of variables that have a fundamentally visual component (e.g.

identification of buffelgrass on the landscape), and second, to keep respondents interested

and attentive through a relatively long survey. For the latter reason these 3 questions

were spaced out in survey to provide a visual and mental change for the respondents.

Survey sample

The 800 household sample for this survey was taken from tax year 2009 Pima

County tax data obtained from the County Geographic Information Systems (GIS) File

Transfer Protocol (FTP) server (Pima County, 2011). This data was cleaned to remove

commercial entries and joined with a shapefile of all tax parcels in Pima County. It was

then clipped to the study area defined as the extent of residential zones determined by the

real estate map. This study area corresponds closely to the limits of the City of Tucson

but is slightly larger because it covers some suburban areas. By sampling such an area an Comment [j1]: Where is a map of the study area? I can tell you really quickly once I see it.


appropriate scale of analysis is achieved because it contains the majority of the city’s

population and includes the range of central city and urban-wildland interface areas under

consideration in the study. Figure 2 shows the areas chosen to represent the central city

and the urban-wildland interface, respectively. The distinction between zones was then

coded into the tax parcel data, effectively forming 2 groups of residents. From this edited

tax data a stratified random sample was formed by randomly selecting 400 households

each in the central city and urban-wildland interface zones by using Hawth’s Tools

(2003) for ArcGIS (9.3, ESRI, 2009).

Figure 2. Central city and urban-wildland interface sample areas in Tucson, AZ. Basemap ©Google 2011

The use of a stratified random sample focused on the comparison between urban

and suburban/urban-wildland interface residents. Stratification was central to this

research because it sought to generate approximately equal numbers of respondents from

each zone. After generating the sample 800 surveys were sent out in June 2010 with

cover letters elaborating on the study goals (also presented on the first page of the survey)

and prepaid envelopes for the return of the completed surveys. Instructions on the cover

Urban Wildland Interface

Urban Wildland Interface

Central City

Comment [j2]: Once I see the map, I think a little more discussion about how the central city and UWI vary will be in order (b/c there are more types of

neighborhoods than these two categories, it will be

essential to talk about why only 2 categories. Do-able to discuss and necessary).


letter requested that the household member over 18 years old with the most recent

birthday fill out the survey, ensuring a random selection of appropriate individual

respondents from within each household.

Variable relationships and statistical analysis

Each of the 32 questions in the survey corresponds either directly or indirectly to

one of the research questions and seeks to measure all of or a portion of the variables

relevant to said questions. After receiving the survey responses over the summer of

2010, responses to the 32 questions were coded into a dataset consisting of 195 variables

using Predictive Analytics SoftWare (PASW) Statistics (Version 18) created by SPSS

Incorporated. 195 variables were created from only 32 questions because the survey has

many multi-part questions that all require multiple numeric variables to fully code the

responses.

From these 195 variables, a further 28 scale variables were calculated. These 28

scales were created by mathematically combining multiple variables in order to develop a

more comprehensive measure of specific research concepts. For example, a scale was

created to measure the level of respondents’ participation in environmental or

conservation organizations. This calculated variable drew from responses to Q32, which

had respondents indicate their level of participation in 22 organizations that have offices

in Tucson; they could indicate being a donor, member, volunteer, or employee. A point

value was assigned to each level of membership from 1 to 4, and the scale variable was

simply the sum of all these values. Most other scale variables were calculated in a

similar way, usually being either the sum or the mean of a series of questions. For a

detailed explanation of the calculations involved in creating the other scale variables, see


Appendix C.

The variables created through the creation of the survey dataset corresponded

directly to the conceptual organization of the study. Appendix C contains additional

information about how concepts were measured by scale variables and should be referred

to by the reader throughout the text if the meaning of any variables is unclear. The

relationships between variables were specified using the Elgie-Kraly method and

conceptualized based on the initial literature review results (see Figure 1). This method

works by organizing the dataset into background, intervening, and outcome variables.

Such a method has also recently found application in the study of invasive species

perception (Seidl & Klepeis, 2011). The conceptualization of study variables is displayed

in Figure 3, which displays all the key variables obtained from the dataset. The

background variables are independent variables, which are mostly demographic. The

only explicitly spatial variable, central city or urban-wildland interface resident, is also

included in this category. Intervening variables are those that are between background

and outcome variables. They are different than demographic variables but are

conceptualized to also strongly affect the outcome variables. For example, studies have

shown that environmental attitudes tend to be related to education and income levels, but

it is hypothesized by this study that respondents’ relationship with and understanding of

their natural environment is also a mediating factor on their risk perception of buffelgrass

(for similar conceptualization of mediating factors see Barr, 2007; Thapa, 2010).


BACKGROUND INTERVENING OUTCOME

Seasonal residence

Age

Education

Income

Hispanic or Latino

‘Wildland’ recreation frequency

‘Urban’ recreation frequency

Identification of BG on the landscape

SNP usage

Environmental attitudes

Sonoran Desert perceptions

Favored BG invasion response

Risk perception of BG

Most essential SD plant

Conservation organization membership

Race

Gender

Year moved to Tucson

Reasons for moving to

Tucson

CC or UWI Resident

Figure 3. Initial conceptualization of study variables based on the Elgie-Kraly method.

The initial conceptual diagram in Figure 3 shows how variables were theoretically grouped at the onset of the

study. An updated conceptual diagram (see Figure 4) was created to specify the significant relationships

between all variables (all tests for significance in this study used a value of p < 0.05). This new diagram removed

all variables that did not have significant relationships, thereby simplifying the conceptualization of the study.

Lines between concepts show that a significant relationship exists and help clarify the connections between

background, intervening, and outcome variables.


BACKGROUND I N T E R V E N I N G O U T C O M E

Seasonal residence

Age

Income

Outdoor recreation frequency


SNP usage





Conservation organization participation


Tucson

CC or UWI Resident

Education

Test LegendPearson X2 TestANOVASpearman CorrelationIndep. Samples t-test

Figure 4. Updated conceptualization of study variables including only variables with significant relationships.

Statistical tests used to determine significance at the 95% confidence interval are shown in the legend.

The results of this analysis will be discussed later, but generally speaking

independent sample t-tests, Pearson X2 tests, one-way analysis of variance (ANOVA),

and Spearman rank correlations were used to see if these variables of differing

measurement levels had significant relationships. When testing, independent variables

came from the column on the left and dependent on the right; for example, background

variables were independent when testing against dependent outcome variables.

Therefore, intervening variables serve as both independent and dependent variables, but

not for the same statistical test.

After determining the significant relationships among variables, 3 linear

regression models were employed to determine the relative effects of each independent


variable on the dependent outcome variables focused on in the research question: favored

response to buffelgrass invasion, risk perception of buffelgrass, and identification of

buffelgrass on the landscape. These models were built by entering the significantly

related independent variables into the equation and excluding missing cases listwise

(observations with missing values on any of the variables in the analysis are omitted).

The outcomes of these 3 regressions are analyzed in the Results section.

Cluster analysis

In addition to running individual statistical tests and building regression models, a

third statistical technique was used to uncover patterns in the survey dataset and answer

the research question of how different stakeholder groups view the buffelgrass invasion.

Exploratory heuristic cluster analysis was used to group respondents into similar

categories based on all independent variables present in the updated conceptual diagram

that had significant relationships. The use of a Q-mode technique such as cluster analysis

can complement R-mode statistical approaches by discovering different patterns in the

data (Aldenderfer & Blashfield, 1984). Cluster analysis has been criticized as the ‘poor

man’s factor analysis’, but as Tryon (1939) points out, cluster analysis is a more

appropriate choice in certain situations. For example, this study’s variables were clearly

conceptualized and measured through pre-determined survey questions; therefore there

were no hypothesized unknown factors that underlie the issues being examined that could

be measured through factor analysis. In this case cluster analysis is a more appropriate

approach than factor analysis for seeking trends in the data even though it is a simpler

procedure. For further discussion of the benefits and challenges of cluster analysis, see

Aldenderfer & Blashfield (1984) and Lorr (1983).


In this study the initial conceptualization of variables was based heavily off

findings from the literature (see the Introduction) and R-mode statistical techniques were

used to narrow down the variables under consideration to only those with significant

relationships. Grouping variables that are significantly linked strengthens the assumption

necessary for cluster analysis that the factors under consideration are theoretically related.

For this study cluster analysis was performed on the 16 variables that were

significantly related using the computer software ClustanGraphics (Version 8) by Clustan

Limited. While some studies choose to employ cluster analysis to group respondents

based strictly on demographic variables, researchers such as García-Llorente et al. (2008)

have garnered insightful results by clustering attitudinal or behavioral variables such as

environmental attitudes or outdoor recreation. Therefore, the 16 variables used in this

study’s cluster analysis measure concepts beyond demographics.

Due to the fact that these 16 variables measured respondents’ attitudes or

demographics on several different scales at various levels of measurement, several

considerations had to be made for the mixed-mode nature of the dataset (Ichino &

Yaguchi, 2002; Wong & Chiu, 2009). First, the dataset was loaded into ClustanGraphics

and the missing values were coded so they would be recognized by the program and

excluded when appropriate (see Appendices D and E for further discussion of missing

variables). Then, the variables were organized by type within ClustanGraphics, with

level of measurement being assigned as either binary, ordinal, or continuous. After this

process the ordinal and continuous variables were transformed to their z-scores to

standardize the data into the same scale. Proximities were calculated for the data using

the Gower’s General Similarity Coefficient, a method recommended for calculating


proximities among mixed-mode variables (Gower, 1971). For details on the calculation

of Gower’s General Similarity Coefficient, see Appendix D. From these proximities a

hierarchical agglomerative cluster analysis was run using the Increase in Sum of Squares

or Ward’s method, the calculation of which can be found in Appendix E. Increase in

Sum of Squares assumes that the cases can be represented by points in Euclidian space,

and requires a proximity matrix of Squared Euclidian Distances. For this reason,

ClustanGraphics was used to convert the Gower’s General Similarity Coefficients, which

are similarities, to Squared Euclidian Distances by subtracting every value from the

maximum similarity.

After computing proximities and then clustering the resulting similarity matrix a

dendrogram is produced that visually presents the clustering process. For an example,

see Figure 5, in which the result of clustering the milk composition of 25 mammals is

shown through a dendrogram. The vertical axis shows the cases, in this case the mammal

milk samples, and the horizontal axis shows the units used to distinguish between the

cases, in this case Reduction in Error Sum of Squares. Dendrograms are useful tools for

presenting the results of a cluster analysis as they show the level of similarity at which

clusters are formed visually. The results of this study’s clustering are displayed through a

dendrogram in the Results section.


Figure 5. Dendrogram showing clustering of the milk composition of 25 mammals. From Wishart (2005).

In order to determine the best cut or number of clusters to analyze,

ClustanGraphics Best Cut tool was used. This tool uses fusion values (k) to determine

when adding a cluster no longer results in significant differences between all clusters.

Fusion values are simply the numeric value at which various cases merge to form a

cluster. While tests using fusion values to determine the number of clusters are not

perfect, they are one of the best methods to dealing with the difficulty of appropriately

‘cutting’ cluster analyses (Aldenderfer & Blashfield, 1984). The Best Cut tool in

ClustanGraphics calculates the realized deviates and t-statistics for all possible cluster

partitions and then reports those with t-statistics that are significant at the 5% level. For

this study the Moving Average Quality Control Rule was selected as the method for

determining which partition with the largest number of clusters would be selected as the

best cut of the tree. For each fusion value k, this method fits a linear trend to the first k-1

values and then computes the expected values for the kth fusion value from this trend

line. The calculations used by ClustanGraphics are based on the methods detailed by

Mojena (1977) and Mojena & Wishart (1980).


IV. Results

Sample

By October 2010 (approximately 4 months after sending out the survey) 122 of

800 surveys were returned, resulting in a response rate of 15.25%. Appendix F shows the

demographic information of the survey sample split between the 2 main sampled

populations: residents living in the central city and residents living around the urban-

wildland interface (UWI). Despite sending 400 surveys to each group in an effort to

obtain an evenly divided sample, the response rate among UWI residents was greater,

with that group making up 46.7% (n=57) of the sample and central city residents

comprising 30.3% (n=37). This variable was calculated based on respondents’ indication

of which zone of Tucson they reside in; because not every respondent filled out this

question 23% (n=28) of the cases were not assigned to either group.

As is shown in Appendix F the sample demographics seem to have been slightly

skewed by a response bias but still appear to be generally representative of the population

under study. The variables considered ‘demographic’ in nature include seasonal

residence (measured in months per year spent in Tucson), gender, age, education level,

2010 income tax bracket, and race and ethnicity questions following the 2011 census

format. The only variable with a significant difference (p < .05) between the central city

and UWI samples as determined by independent samples t-tests was 2010 income tax

bracket, with the most common bracket in the central city being 15% ($16,750 –

$68,000 for married couples filing jointly and $8,375 – $34,000 for single filers) and

around the UWI 25% ($68,000 – $137,300 and $34,000 – $82,400). The UWI also had

the only respondents in the highest income tax bracket of 35% (over $373,650 for either


joint or single filers). The income disparity between central city and UWI residents is

also apparent in data from the 2000 census: the median income for the UWI was $47,470

compared to only $27,665 for the central city (see Appendix F).

The other demographic variables show differences between the central city and

the UWI also, although they are not statistically significant. Education levels are slightly

higher in the central city, although the most prevalent level for both samples was a 4 year

college degree (30% of the total sample), showing that respondents were generally well

educated. The level of master’s degrees was higher in the central city (29.7% compared

to 21.4%), but the UWI sample had more doctoral degrees (2.7% compared to 14.3%).

Compared to the census data, this sample of the population of Tucson was

disproportionately educated. The most common level of education for both the central

city and UWI was ‘some college’ according to the 2000 census with 25.30% and 24.32%,

respectively. The high levels of 4 year college degrees in this sample (30% of total

sample), master’s degrees (29.7% for the central city and 21.4% for the UWI), and

doctoral degrees (2.7% for the central city and 14.3% for the UWI), are not found in the

census data (8.20% central city and 11.59% UWI for master’s and 2.08% central city and

2.94% UWI for doctoral). Additionally, the balance of master’s and doctoral degrees is

different in the census data, with the UWI having more Master’s and both zones having

similar levels of doctoral degrees.

The sample’s age distribution was heavily skewed towards an older population,

which can be expected in a city with a large retiree population. The most prevalent group

for both samples was older than 60 years old (45.5% of the total sample). Additionally,

the central city was slightly younger with a much greater proportion of the population in


their thirties (13.5% compared to 1.8%). Although Tucson’s overall median age

according to the census is actually under the national average (32.1 versus 35.3), the

sampled areas differ from the city median, with the central city being 34.25 and the UWI

a much higher 42.05. This pattern of older residents living in the UWI was successfully

captured by this sample.

Respondents were 34.2% male and 65.8% female. This imbalance was slightly

greater in the UWI group. This disparity is not reflected in the census data, which show

that the 2 sample zones had similar male to female ratios: around 49% male and 51.5%

female. Therefore, this sample may be skewed towards female respondents. The

overwhelming majority of respondents were full-time residents (11-12 months per year

spent in Tucson) at 92.6%.

The predominant race of respondents in both sample zones was white, comprising

100% of central city and 94.5% of UWI respondents, which is also higher than the rate

for the entire Tucson population (75.44% in the central city and 85.50% in the UWI).

Additionally, there were only 7.4% of respondents who reported being of Hispanic,

Latino, or Spanish origin, a number smaller than the total rate for Tucson (25.08% in the

central city and 15.78% in the UWI). Finally, the median year that most respondents

began living in Tucson was approximately 1980 for both the subgroups and the total

sample.

Overall results

Here, the overall results of the survey independent of the statistical analysis will

be reported in order to provide a general overview of the respondents’ knowledge,

attitudes, and beliefs about the Sonoran Desert and buffelgrass. The 3 outcome variables


of interest (risk perception of and preferred response to the buffelgrass invasion and

buffelgrass identification skill) are also those most interesting in terms of the overall

sample and will therefore be discussed briefly here. Respondents reported uniformly

high risk perception of the buffelgrass invasion, with the mean of the perceived threat

scale being 4.48 out of a maximum possible of 5. The median preferred response to the

invasion was control while the mode preferred response was eradication, indicating that

respondents favored strong management responses. The median for buffelgrass

identification skill was 1.75 and the mode was 1.50. In this scale a value of 1 indicated

all buffelgrass was successfully identified, 2 indicated more than 50%, and 3 indicated no

buffelgrass was successfully identified. These results show that most respondents were

reasonably successful at identifying the plant.

The photograph identification section which asked respondents whether they

believed certain common and iconic Sonoran Desert plants were essential to the desert

did not yield any variables related to demographics or other independent variables, but

the general results are nevertheless interesting and work towards confirming some of the

secondary hypotheses of this study about how Tucsonans view their desert. Respondents

were asked to identify all essential plants and what they thought of as the most essential

plant.

The results for both questions are shown below and strongly suggest that the giant

saguaro (Carnegiea gigantea) cactus is the most essential plant of the Sonoran Desert

according to respondents. An overwhelming 83% of respondents chose the saguaro as

the most essential plant of the Sonoran Desert, with the other individual plants chosen

garnering 6% at most (Desert prickly pear Opuntia engelmannii). The frequencies of


plants chosen as most essential are shown below in Figure 6. Additionally, 96% of

respondents indicated that they felt the saguaro was an essential plant of the desert,

followed by the ocotillo (Fouquieria splendens) with 91% and desert prickly pear

(Opuntia engelmannii) with 90%. The frequencies of plants chosen as essential are also

displayed in Figure 7 below.

The other noteworthy result from this exercise comes from respondents’ views on

the grasses included as choices: the native grasses curly mesquite grass (Hilaria

belangeri (Steud.) Nash) and purple three-awn (Aristida purpurea) and the alien invasive

buffelgrass (Pennisetum ciliare). None of these plants were chosen as the most essential

Sonoran Desert plant, and they were the bottom 3 results for essential plants in the

descending order of curly mesquite grass, purple three-awn, and then buffelgrass. Only 1

respondent identified buffelgrass as an essential plant for the Sonoran Desert.

Figure 6. . Frequency of Sonoran Desert plants selected as most essential.

83

6 4 3 2 1 1 0

10

20

30

40

50

60

70

80

90

Giantsaguaro

Desertpricklypear

Desertironwood

Creosotebush

Bluepaloverde

Desertagave

Arizonapoppy

Fre

qu

en

cy

Plant


Figure 7. Frequency of Sonoran Desert plants selected as essential.

Variable relationships

As explained in the Methods section, an updated conceptual map for all the

variables was created to aid in specifying the relationships between the variables under

consideration and as a precursor to building linear regression models for the outcome

variables. The final conceptual diagram is shown in Figure 8 and shows the results of

each statistical test. In order to find significant (defined here as p < 0.05) relationships

between the variables independent sample t-tests, Pearson X2 tests, one-way analysis of

variance (ANOVA), and Spearman correlations were used.

107 101 100 95 93 88 87 86 64 63 58

36 18 1

0

20

40

60

80

100

120

Fre

qu

en

cy

Plant


BACKGROUND I N T E R V E N I N G O U T C O M E

Seasonal residence

Age

Income

Outdoor recreation frequency


SNP usage





Conservation organization participation


Tucson

CC or UWI Resident

Education

-2.491

-3.466

3.412

3.863

4.995

2.767

30.675

2.298

3.549

2.624

30.457

-0.182

-0.342

0.314

-0.340 -0.469-0.302

-0.420

3.711 0.353

Test LegendPearson X2 TestANOVASpearman CorrelationIndep. Samples t-test

23.868

Figure 8. Final conceptualization of study variables with results of tests for significant relationships.

Variable relationships – Background on intervening

This section will detail the results of the statistical tests using background

variables as independents and intervening variables as dependents. For further

explanation of how variables were calculated or to which questions they respond see

Appendices B and C. A one-way ANOVA between seasonal residence and

environmental attitudes showed that respondents that reported a greater connection to

their surrounding environment tended to spend more time of the year in Tucson with an

F-statistic of 3.412 (p = 0.02). The important exception to this trend is shown in Figure

9a, which shows that residents in the 5-7 months category had the strongest

environmental attitudes of all 4 groups. The outcome of this ANOVA shows that

respondents’ environmental attitudes vary by their seasonal residence.


Figure 9a. Figure 9b.

Figure 9c. Figure 9d.

Figure 9e. Figure 9f.


Figure 9. Means plots displaying results of one-way ANOVAs.

A Spearman correlation between environmental attitudes and reasons for moving

to Tucson resulted in a Spearman’s rho of -0.182 (p = 0.049). The scatterplot in Figure

10a shows that respondents with stronger environmental values are correlated with

having more practical reasons for moving to Tucson. The trendline shown on the graph

is provided to give an idea of the level of correlation present, but note that the Spearman

correlation coefficient does not correspond to such a linear relationship, but rather

measures the monotonicity1 of the data.

1 A monotone sequence has numbers which consistently increase or decrease but do not oscillate in relative

value.

Figure 9g. Figure 9h.

Figure 9i.


Figure 10a. Figure 10b.

Figure 10c. Figure 10d.

Figure 10e Figure 10f.


Figure 10g. Figure 10h.

Figure 10. Scatterplots with linear trendlines showing results of Spearman rank correlations.

The relationship between respondents’ age and their practical- or environmental

amenity-based usage of the Saguaro National Park was also found to be significant. The

ANOVA returned a significant F-statistic of 4.995 (p = 0.001). Additionally, the means

plot (Figure 9b) shows a fairly clear linear trend present in the relationship between the

variables, with respondents’ usage of the park becoming more focused on environmental

amenities (e.g. views, observation of local flora and fauna, etc.) in older age groups.

Income level (measured by 2010 tax bracket) was also significantly related to

conservation organization membership, with the ANOVA returning a significant F-

statistic equal to 3.549 (p = 0.006). The means plot (Figure 9c) shows an interesting

result: it appears that participation in conservation organizations remains relatively and

uniformly low among respondents with lower and middle incomes but becomes much

higher at the highest level of income. While this result is also effectively based on

outliers due to the low n in the highest income bracket of 2 cases, there is a slight upward

trend in the next 2 highest brackets, which suggests that this trend is genuine. In any case


the ANOVA shows there are significant differences in group means for these variables.

An independent samples t-test with central city or UWI residency and recreation

frequency measured as ‘urban’ (e.g. walking, playing tennis, etc.), ‘wildland’ (hiking,

mountain biking, etc.), and total levels of recreation returned a t-statistic of -2.491 with a

2-tailed p-value of 0.015 (equal variances assumed). The analysis of these variables

shows that Tucson residents living around the UWI report higher levels of ‘wildland’,

‘urban’, and total outdoor recreation, with means of 4.73, 4.84, and 4.80 on the respective

scales. Central city residents show lower levels of recreation with means of 4.41, 4.52,

and 4.48. These scales are based on averaged 6-level ordinal responses for frequency of

recreation ranging from a response of 1 being ‘daily and 6 being ‘never’. This result is

displayed in Figure 11 and shows that most respondents were not frequent outdoor

recreators, but that those living near the UWI made use of their natural environment for

recreation significantly more frequently.

Figure 11. Wildland recreation frequency by central city (CC) or urban-wildland interface (UWI) residence.


The final background variable that had a significant relationship with an

intervening variable was education, which was linked to respondents’ perceptions of the

Sonoran Desert through a one-way ANOVA. The F-statistic for this test was significant

at 2.298 ( = 0.05). The means plot (Figure 9d) shows an interesting pattern: respondents

with a 2-year college degree showed the most utilitarian view of the Sonoran Desert,

while those with Master’s degrees reported the most conservation-based view of the

desert. Again, this trend is interesting but because ANOVAs only show differences

between group means the directionality of this relationship may not hold true in the

greater population.

Variable relationships – Intervening on intervening

Another important set of relationships between variables comes from the effect of

intervening on other intervening variables. These relationships are important because

they aid in understanding how these mediating variables relate to one another. It is

acknowledged that due to the abstract versus concrete nature of these variables (e.g.

environmental attitudes versus frequency of outdoor recreation) that their inclusion in the

intervening variables section may appear somewhat arbitrary; however, as discussed in

the Methods section, this conceptualization is appropriate for this study.

Considering that these variables are in the same conceptual group, it would not be

logical to relate them using tests that require categorizing them into independent or

dependent variables. Therefore, given the ordinal nature of these scales, Spearman

correlations were used to explore their relationships. Significance is two-tailed and

missing values were excluded pairwise (if either or both paired values for the 2 variables

were missing, they were excluded from the analysis).


Conservation organization membership was found to be significantly correlated

with 2 intervening variables: outdoor recreation frequency and Sonoran Desert

perceptions. The Spearman correlation between organization membership and recreation

frequency was -0.340 (p < 0.01.) This inverse relationship means that as organization

membership goes up recreation frequency goes down, which can be seen in Figure 10b.

Between organization membership and Sonoran Desert perceptions the correlation

coefficient was -0.302 (p-value < 0.01). This additional inverse relationship shows that

as organization membership goes up respondents tend to have a more conservation-based

view of the Sonoran Desert; this trend is depicted in Figure 10c.

Outdoor recreation frequency was also significantly correlated with respondents’

environmental attitudes and their perceptions of the Sonoran Desert. The correlation

coefficient between recreation frequency and environmental attitudes was -0.342 (p <

0.01). This trend implies that as respondents feel a closer connection to the natural

environment they are less likely to participate in outdoor recreation, as depicted in Figure

10d. Between outdoor recreation and perceptions of the Sonoran Desert the correlation

coefficient was 0.314 and was significant (p < 0.01). According to this result,

respondents who recreated outdoors more frequently had a more utilitarian view of the

Sonoran Desert, as is shown in Figure 10e.

The final significant relationship discovered between intervening variables

involved environmental attitudes and Sonoran Desert perceptions. The correlation

coefficient was -0.469 (p < 0.01). As seen in Figure 10f, this correlation shows that

respondents with stronger environmental attitudes had a more conservation-based view of

the Sonoran Desert.


Variable relationships – Intervening on outcome

Significant relationships were also found between the intervening and outcome

variables. Environmental attitudes were significantly correlated with respondents’ risk

perception of the buffelgrass invasion as measured by the mean of their perceived threat

that the plant poses to both Tucson and the Sonoran Desert in general, with a correlation

coefficient of 0.353 (p < 0.01). As shown in Figure 10g, this result shows that

respondents with stronger environmental attitudes also felt that the buffelgrass invasion

posed more of a threat to the Sonoran Desert and Tucson.

Sonoran Desert perceptions were also correlated with respondents’ risk perception

of buffelgrass and their favored response to buffelgrass invasion. The relationship

between risk perception and Sonoran Desert perception was explored through a Pearson

correlation, which resulted in a significant correlation coefficient of -0.420 (p < 0.01). As

shown in Figure 10h, this statistic shows a correlation suggesting that respondents with a

more conservation-based view of the Sonoran Desert also feel that the buffelgrass

invasion poses a greater threat to the desert and Tucson.

The variable measuring respondents’ favored response to the invasion came from

Q23, which asked respondents how they believed the invasion should be handled, with

responses lying on a continuum from promoting buffelgrass to completely eradicating it.

The relationship between this ordinal variable and Sonoran Desert perceptions was

analyzed through a one-way ANOVA and yielded a significant F-statistic of 3.711 (p =

0.014). The means plot is displayed in Figure 9e and suggests that stronger management

strategies are favored by respondents with a more conservation-based view of the

Sonoran Desert.


Variable relationships – Background on outcome

Another set of important relationships in this study’s dataset are those between

background and outcome variables. Seasonal residence, income level, and central city or

UWI residence were all significantly related to respondents’ skill at identifying

buffelgrass on the landscape. A one-way ANOVA between seasonal residence and

buffelgrass identification skill yielded a significant F-statistic of 3.863 (p = 0.026). The

means plot displayed in Figure 9f does not aid in clarifying the directionality of this

relationship very much; however, it appears that residents spending more of the year in

Tucson are better at identifying buffelgrass than their seasonal counterparts. The

ANOVA comparing the means of different tax bracket groups on the variable of

buffelgrass identification skill yielded a significant F-statistic of 2.624 (p = 0.045). The

means plot (Figure 9g) shows a relatively clear trend that respondents with higher

incomes are better at identifying buffelgrass on the landscape.

The third background variable with a significant relationship to buffelgrass

identification was residency in the central city or around the UWI. A one-way

independent samples t-test with buffelgrass identification skill as the dependent variable

and residency as the independent variable resulted in a significant t-statistic of -3.466 (p

= 0.001). The box and whiskers plot in Figure 12 shows that although both groups have a

large range of skill in identifying buffelgrass, central city residents are significantly better

at identifying the plant on the landscape.

Successfully

Comment [u3]: I’m not seeing this—in fact your BEST IDers have the lowest income. WAIT WAIT WAIT….I SEE—LOW NUMBERS

MEAN BETTER ID SKILLS. THIS SHOULD BE

CLARIFIED ON YOUR FIGURE AXIS LABEL, SINCE IT’S BEEN 35 PAGES SINCE WE HEARD

ABOUT HOW THIS VARIABLE WAS

CONSTRUCTED. THE MESSAGE IS ALSO OBSCURED BY THE FACT THAT WE READ

DOWNWARD TRENDS AS DECLINES.


Figure 12. Success rate at identifying buffelgrass by central city (cc) or urban-wildland interface (UWI)

residency. Lower values on the identification scale mean more successful identification.

In addition to these 3 background variables being related to respondents’ skill at

identifying buffelgrass, the background variables of age and income both had significant

relationships with respondents’ favored buffelgrass invasion response. Due to the ordinal

nature of these variables, Pearson X2 tests were used to determine the strength of these

relationships. The X2 test between age and invasion response resulted in a X

2 of 30.675

and a 2-sided asymptotic p-value of 0.010. Figure 13 shows the respondents grouped by

favored response and age and although there are exceptions, the overall trend suggests

that older respondents are more likely to favor stronger management responses.

Not successfully

identified


Figure 13. Age group by favored response to buffelgrass invasion.

The X2 test between income and invasion response resulted in a X

2 of 30.457 (p =

0.010). Figure 14 shows the respondents grouped by favored response and income level

and although there are exceptions, the overall trend suggests that more middle-income

respondents are more likely to favor stronger management responses.

Figure 14. Income tax bracket groups by favored response to buffelgrass invasion.


The final significant background on outcome variable relationship discovered was

between age and risk perception of the buffelgrass invasion. To measure this connection

a one-way ANOVA was used, which resulted in an F-statistic 2.767 (p = 0.046). The

means plot (Figure 9h) shows that the 30-39 year old age group reported the highest level

of perceived threat and the 40-49 reported the lowest. This group seems to be somewhat

of an outlier as the other 3 age groups with valid data show a linear decline in perceived

threat with increasing age.

Variable relationships – Outcome on outcome

The final set of variable relationships found in the dataset came from the

relationship between favored buffelgrass invasion response and risk perception. This

relationship was measured using a one-way ANOVA, which resulted in an F-statistic of

23.868 (p < 0.01). The means plot (Figure 9i) shows respondents who identified the

invasion of buffelgrass as a greater threat also favored more extreme management

responses.

Regression analysis

As described in the Methods section, linear regression models were built for the 3

outcome variables of this study in order to better understand the relative importance of

their significantly related variables. The results of this regression analysis are displayed

in Tables 1-4 in Appendix G. The model predicting respondents’ favored response to

buffelgrass invasion had the scale for pragmatic- or environmental amenity-based usage

of Saguaro National Park (SNP), the scale for environmental attitudes, the scale for

pragmatic- or conservation-based view of the Sonoran Desert, and the scale for perceived

threat to the SD and Tucson as independent variables. Due to the ordinal nature of the


favored response variable, a PLUM ordinal regression model was used with the

dependent variables entered as covariates. This model had a Cox and Snell pseudo-R2 of

0.205, meaning that it is able to account for around 20.5% of variation in the dependent

outcome variable (see Table 1). The model fit was significant with p < 0.01, which is to

be expected considering the independent variables entered into the model were all

significantly related to the dependent outcome variable (see Table 3).

The parameter significances for this model shed some light on the relative effect

of each factor on the outcome variable: the scale for usage of SNP had a p = 0.047, the

scale for environmental attitudes had a p = 0.618, the scale for perception of the Sonoran

Desert had a p = 0.721, and the perceived threat to the Sonoran Desert and Tucson had a

p < 0.01 (see Table 4). This shows that the risk perception scale and the usage of SNP

scale were the most relatively important influences on the favored response to buffelgrass

in this model and of the 2 factors, risk perception had the greatest influence.

The model for respondents’ risk perception of the buffelgrass invasion used the

factors of age, environmental attitudes, and perception of the Sonoran Desert. Although

favored response to the buffelgrass invasion was also significantly correlated with this

variable, the conceptualization of this study implies that risk perception informs favored

response but that this relationship is not reciprocal. The variable of age was recoded into

dummy variables due to the interval nature of the data. The regression had an R2 of

0.261, suggesting that it is capable of predicting 26.1% of the variance in respondents’

risk perception (see Table 1). This model was also significant with a p < 0.01(see Table

2). As shown in Table 4, the standardized beta coefficients for this model were 0.245 for

environmental attitudes (p = 0.011) and -0.276 for perceptions of the Sonoran Desert (p =


0.005). For the age dummy variables the 30-39 group had a beta of 0.139 (p = 0.185), the

39-40 group had a beta of -0.032 (p = 0.085), the 50-59 group had a beta of 0.157 (p =

0.269), and the 60 and older group had a beta of 0.172 (p = 0.250). These betas suggest

that respondents with stronger environmental attitudes and a more conservation-based

view of the Sonoran Desert are more likely to perceive the buffelgrass invasion as a

serious threat, while age plays a less significant role.

The third model sought to predict respondents’ skill at identifying buffelgrass on

the landscape and used the significantly related variables of seasonal residence, income,

and residence in the central city or UWI as dummy variable factors. This model had a

significant R2 of 0.307 (p < 0.001), as displayed in Table 1. The beta coefficients for this

model were 0.066 for the 5-7 months group (p = 0.675), -0.116 for the 8-10 months

group (p = 0.325), 0.306 for the 10% tax bracket (p = 0.073), 0.314 for 15% (p = 0.108),

0.162 for 25% (p = 0.398), 0.013 for 28% (p = 0.938), 0.032 for 33% (p = 0.806), and -

0.397 for central city residents (p < 0.01). These results show that central city or urban-

wildland interface residency foremost and income to a lesser degree are the relatively

most important factors in predicting respondents’ skill at identifying buffelgrass on the

landscape.

Despite the fact that the models for favored response and risk perception both included

other outcome variables as factors, multicollinearity does not present a problem in this

dataset: a tolerance of less than 0.20 or 0.10 and/or a Variance Inflation Factor of 5 or 10

and above indicates a multicollinearity problem (O’Brien, 2007)

, and as Table 4 shows, none of the factors in any of the 3 regression models meet

these criteria.


Cluster analysis

Using ClustanGraphics software to run a cluster analysis on this dataset yielded

the dendrogram shown in Figure 15, which displays how respondents were grouped into

clusters. In addition to producing this dendrogram, which visually shows the clustering

process, the Best Cut tool in ClustanGraphics was used to determine the largest

significant number of clusters. The results of this analysis yielded 5 clusters with a

deviate of 4.65 and a significant t-statistic of 51.18 (p < 0.05). The results of this analysis

are visible in Appendix H, which shows the mean of each variable and the number of

cases for each cluster. This result is also displayed in the dendrogram below: the blue

area represents parts of the dendrogram included in the 5 significant clusters and the

yellow areas the clusters without significant differences from each other. The exact

location of the ‘cut’ can be seen by following the deviate value of 4.65 from the scale of

Euclidian Sum of Squares along the x axis vertically through the dendrogram, as

indicated by the thick vertical black line.


Figure 15. Dendrogram showing results of cluster analysis. Thick horizontal line shows the location of the 'best

cut' of the analysis into five significantly different clusters.

Visualizing the differences between clusters can be difficult due to the many

variables under consideration and the relatively small differences between clusters;

however, the bar charts in Appendix H Figures 1-5, show the levels of all variables for

each cluster (see Appendix C for more information on variables). Additionally, the

minimum, average, and maximum values for each variable across all clusters are overlaid

on top of the bar chart. This allows the level of variables in each cluster to be compared

to the others to give an idea of how much that particular cluster varies with respect to the

others. Note that some variables, such as seasonal residence, show vary little variation

across the clusters, whereas others, such as conservation organization membership and

participation show a great deal of variation. Also important is that these variable values

Comment [u4]: I like this. Very snappy figure. Add some labels…


are not all directly related to the original values in the data; as explained in the Methods

section, some of these variables were transformed to Z-scores to standardize the scale

used to calculate similarities. An analysis of these cluster analysis results can be found in

the Discussion section.

Comment [SW5]: Appendix? PERHAPS.


V. Discussion

Table 1. Study hypotheses and results.

Hypothesis Confirmed?

1.) The public will be concerned about the invasion. YES: 4.48/5, higher numbers

indicate more concerned

2.) Residents will value the Sonoran Desert beyond its

utilitarian or use value.

YES: 2.2/5, smaller numbers

indicate more conservation-

based view

3.) Residents will be successful at identifying

buffelgrass on the landscape (>50% of the time) due to

media attention and the widespread nature of the

invasion.

YES: average success fell

between identifying greater

than 50% and 100% of the

buffelgrass

4.) Tucsonans will favor strong management options

such as total eradication, because most members of the

public are not aware of the practical and budgetary

restraints on invasive plant management.

YES: mode preferred

response was eradication, the

strongest management

response

5.) Older, more affluent, better educated UWI

respondents who are more ‘conservationist’ in attitude

towards the Sonoran Desert will be more concerned

about the invasion, favor stronger management

responses, and be better at identifying buffelgrass.

Generally YES, with

important exception that

central city residents were

better at identifying

buffelgrass. See Results.

Demographics and sample

Before larger implications and trends are discussed, several factors that could be

responsible for the demographic trends of the study sample will be elucidated. Although

the demographic patterns in this study’s sample were not extremely different from those

in the U.S. Census (see Appendix F), there does appear to be some response bias. At

15.25%, the survey’s rate of return is not abnormal for this kind of study due to budget

constraints that precluded sending out introductory and reminder postcards, which have

been shown to increase response rates, and the relatively long length of the survey (Fink,

2003). The high response rate among full-time residents could be explained by the fact

that they have a greater stake in Tucson and the surrounding Sonoran Desert and

therefore would be more likely to respond to a survey on such topics. This idea is also

supported because the average length of time respondents have lived in the region was

Comment [SW6]: How much should I reference results (#s) here? I don’t want to repeat


almost 30 years. The fact that the sample was more highly educated than the census

average is somewhat to be expected given the tendency for survey respondents to be

more educated than normal (Kaczensky et al., 2004; Mayer & Frantz, 2004). Finally, the

fact that minorities were less likely to respond to the survey also reflects a common

problem with social science surveying methods (see Hillygus, 2006). Despite these

shortcomings, the survey sample obtained a generally reliable look at Tucsonans

perceptions of invasive species.

Variable relationships – Background and Intervening

Several of the relationships between background and intervening variables in this

study confirm findings from other research and provide insight into the social dimension

of the buffelgrass invasion. These results also suggest that the study’s variables

adequately represent their concepts because, as will be shown, they replicate results from

previous studies. One noteworthy trend present in the data comes from the relationship

between income and conservation organization participation. As shown earlier in Figure

9c, organization participation is very low among most income groups except for the

highest tax bracket, which shows significantly higher participation rates. This finding is

in agreement with many studies finding that voluntary organization membership is much

higher among higher income groups due to their increased leisure time and resources

(Coombs, 1972; Devall, 1970; Harry et al., 1969; Hausknecht, 1962; van Liere & Dunlap,

1980; Wikle, 1995) and that such membership is especially higher among those with

high-status, well-paying occupations (Defee et al., 1974; Harry et al., 1969; Milbrath,

1984; Wikle, 1995).

Another important set of findings is related to respondents’ participation in


outdoor recreation. First, a significant difference was found between residents of the

central city and those living near the urban-wildland interface (UWI), with UWI residents

reporting significantly higher levels of ‘wildland’, ‘urban’, and total outdoor recreation.

This result makes sense in light of the literature because the UWI sample tended to be

more affluent, one of many sociodemographic factors found to influence outdoor

recreation (Kelly, 1980; Lee et al., 2001). It also makes logical sense because these areas

are closer to natural recreation opportunities like the Saguaro National Park and the

Coronado National Forest.

Respondents’ level of outdoor recreation was also significantly related to the

correlated variables of Sonoran Desert perceptions and environmental attitudes. The

findings of Theodori et al. (1998) suggest that the location of outdoor recreation is not

important in mediating the effect of environmental attitudes, a result supported by the

findings of this study, which show that all 3 scales for wildland, urban, and total outdoor

recreation have similar significant relationships to environmental attitudes. However, the

relationship discovered in this study is contrary to than that found in others (e.g. Jackson,

1986; Larson et al., 2011; Thapa, 2010) in that respondents who reported a stronger

connection to the natural environment actually recreated less.

It is unclear if this finding stems from inadequate measurement of environmental

attitudes or if there is a genuine trend present in the population; further study would be

required to explore this relationship. This result could stem from the survey design,

which did not distinguish between consumptive or appreciative recreation; participants

with high levels of outdoor recreation may have only been using the desert environment

practically as a place to walk or exercise with no perceived benefit from the natural


surroundings. Such recreation has been found to be linked to weaker environmental

attitudes (Jackson, 1986) and would logically explain why respondents with a more

utilitarian view of the desert recreated more frequently.

Essential species identification

The results from the plant photograph exercise are also hypothesis-confirming. It

was hypothesized that one of the major reasons the public will see the buffelgrass

invasion as a threat is due to its potential destruction of the iconic saguaro cacti

(Carnegiea gigantea) by increasing the risk of fire in the desert. This hypothesis is

partially supported by the finding that the saguaro was overwhelmingly chosen as the

most essential Sonoran Desert plant in the photograph exercise, showing that the public

values this plant over others. Very few respondents identified native grasses as essential

and only 72% of respondents chose the desert ironwood (Olneya tesota). This result

shows that the public may not be aware of or value the ecological function of desert

plants; for example desert ironwood trees have been shown to be extremely valuable to

the Sonoran Desert ecosystem by functioning as nurse plants that provide shade to

Saguaro and other cacti and by fixing nitrogen (Felker & Clark, 1981; Franco & Nobel,

1989; Nabhan & Carr, 1994; Suzán et al., 1996).

An implication of this exercise is that the results from the photograph exercise

suggest that the Tucson public may value Sonoran Desert plants based primarily on their

aesthetics; the saguaro and ocotillo (Fouquieria splendens) were the top plants identified

as essential and both are large plants that are visually interesting. It has been discussed

elsewhere that such so-called ‘charismatic megaflora’ may garner attention from the

public similar to that afforded to ‘charismatic megafauna’ like the Siberian tiger


(Panthera tigris altaica) or the African elephant (Loxodonta sp.) (Hounslow, 2009;

Kanowski & Williams, 2009; Lorimer, 2007; McIntosh, 2003). Other obvious examples

exist within the Tucson area that illustrate the importance charismatic megaflora for the

public’s perceptions of the desert: Saguaro National Park, Organ Pipe National

Monument, and Ironwood Forest National Monument are all named after megaflora. As

suggested by Hounslow (2009), charismatic megaflora can be used to further

conservation goals and therefore the use of the saguaro and/or ocotillo image in public

outreach regarding the buffelgrass invasion could prove useful.

Variable relationships – Outcome

Risk perception

This section examines the implications of findings regarding the outcome

variables of risk perception, preferred management response, and buffelgrass

identification skill. The results regarding risk perception suggest that the scale measuring

environmental attitudes is accurate: respondents with stronger environmental attitudes

perceived the buffelgrass invasion as a significantly higher threat. This finding follows

similar results in the literature that show respondents with stronger environmental

attitudes or a conservation-based view of nature tend to perceive greater ecological risk

(Fischer & van der Wal, 2007; Slimak & Dietz, 2006). In the regression model for risk

perception, the environmental attitudes scale was a significant factor with a beta of 0.245.

An even more powerful predictor of respondents’ risk perception came from their

opinion of the Sonoran Desert, measured by a scale reporting if they took a conservation-

based or utilitarian view of the desert. This scale was a significant factor with a beta of -

0.276, showing that respondents with a more conservation-based view of the Sonoran


Desert reported a higher perceived risk from the buffelgrass invasion. The results of this

regression model and the statistical tests used to analyze respondents’ risk perception

present both confirm one of the study’s hypotheses. Prior to data collection it was

theorized that respondents more ‘environmentalist’ or ‘conservationist’ in attitude

generally and in respect to the Sonoran Desert specifically would perceive a greater risk

from the invasion. These results show that this assumption holds and that each factor is

responsible for predicting a significant amount of the 26.1% of variation in the data

explained by the regression model for risk perception.

Preferred management response

The results of this study also illuminate the factors influencing respondents’

preferred management response to the buffelgrass invasion. According to Fraser &

Zealand (2006), accounting for public opinion in invasive species management is

essential to creating a realistic and effective response, but while many management

schemes attempt to involve public stakeholder groups, it is still not well understood what

factors affect public opinion of invasive species. The results of the PLUM ordinal

regression to predict favored management responses in this study’s sample contribute to

the understanding of this complex issue: risk perception was the most influential factor,

followed by utilitarian or environmental amenity-based usage of the Saguaro National

Park (SNP) and then age.

The fact that respondents who felt that buffelgrass was a greater threat also

favored the stronger management responses of total eradication or strict control confirms

the study’s initial hypothesis that the public will favor unrealistic and/or expensive strong

management options such as eradication. It is understandable that respondents who view


buffelgrass as a pressing issue would favor a stronger response, as other studies have

found risk perception to be a strong predictor of the public’s preferred response to

invasive plants (Norgaard, 2007). However, this result shows a disconnect from reality:

considering its current distribution, rapid expansion, and the difficulty of removal,

completely eradicating buffelgrass in southern Arizona is not a cost- or time-effective

option; experts such as Brigham & Betancourt (2010) favor adopting an adaptive

management strategy as soon as possible to slow the spread of the grass and mitigate the

impacts of invasion. While the total removal of the invasive plant is the most desirable

outcome, even organizations like the Southern Arizona Buffelgrass Coordination Center

(SABCC) acknowledge that the most effective way to reduce the negative impacts of the

invasion is through an integrated management strategy that seeks to:

1. Minimize the spread of buffelgrass

2. Set and implement control priorities

3. Restore treated areas

4. Reduce wildfire risks

5. Motivate legislation and seek federal funding (SABCC, 2010)

As outlined in the Congressional Field Hearing from the House Natural Resources

Committee, Subcommittee on National Parks, Forests and Public Lands, immediate and

large-scale action against the invasion is recommended by numerous organizations,

SABCC included, but the point stressed is that the invasion is continuing essentially

unabated and mitigation and management efforts must precede any long-term complete

eradication goal (Frost et al., 2010). This study suggests that more effort is needed to

inform the public about the integrated management response favored by the SABCC,

because their plan seems far removed from the eradication favored by survey

respondents. This implication is particularly important given the fact that public


participation is key to the successful implementation of integrated management (Brigham

& Betancourt, 2010). This result begs the question: what factors could be influencing

Tucsonans to favor this strong management response?

One possible explanation comes from Kasperson et al. (1988) and their concept of

social amplification of risk. As discussed in the risk perception literature review section,

the influence of powerful institutions such as the media or industry lobbies can modify

the public’s perception of risk, thereby altering their preferred response. By facilitating

cooperation among multiple agencies and stakeholder groups, the SABCC has

contributed to the management of the buffelgrass invasion, but its ambitions have been

exaggerated by the media. For example, a 2010 article from Arizona Public Media

claimed the organization’s “end goal [is] the removal of buffelgrass from the entire

Sonoran Desert region” (McLemore, 2010). Additionally, local events such as Beat Back

Buffelgrass Day (SABCC, 2011) and organizations such as the Arizona-Sonora Desert

Museum have spread awareness of the invasion to the public (Arizona-Sonora Desert

Museum, 2011; Kreutz, 2011). Such media coverage could amplify the public’s risk

perception of the grass, as numerous studies have shown that the public’s view and

favored response to environmental issues is subject to modification by the news media

(DiTomaso, 2000; Koné & Mullet, 1994; Nelkin, 1989; Wåhlberg & Sjöberg, 2000).

A second possible explanation for the public favoring eradication comes from the

second most important factor in the preferred response regression analysis: their

utilitarian or environmental amenity-based usage of the SNP. Respondents who reported

a more environmental amenity-based usage of the park also favored stronger

management responses; although this relationship is captured by the means plot displayed


in Figure 16, the difference in group means was actually not significant according to a

one-way ANOVA (p=0.209) and therefore was not discussed in the Results section. The

usefulness of the regression analysis is obvious in this case, as this trend would have been

missed if ANOVAs were relied upon as the only method of analysis.

Figure 16. Means plot showing respondents’ usage of the Saguaro National Park (SNP) by their favored invasion

response. This figure was not included in the results due to the ANOVA finding no significant difference in

group means, but is shown here to illustrate that the regression model for favored response may have detected a

trend missed by the ANOVA analysis.

Interestingly, in the regression model the SNP scale had a much greater relative

influence than the scale for environmental attitudes or the scale measuring utilitarian or

conservation-based view of the Sonoran Desert. In addition to showing the utility of the

regression analysis to discover relationships missed by ANOVAs, the relative importance

of these 3 scales suggests that respondents’ preferred response to the buffelgrass invasion

may be more affected by personal, everyday experience such as the use of a local

wilderness area like SNP than more abstract concepts of environmental attitudes or views

of the Sonoran Desert. This explanation is supported by the risk perception literature,


which finds that direct personal experience is often a powerful amplifier or attenuator or

risk perception (Halpern-Felsher et al., 2001; Kasperson et al., 1988). However, the fact

that SNP usage has a significant effect on respondents’ preferred management response

but not their risk perception is a key finding of this study: it suggests that the public may

mentally separate abstract concepts like environmental attitudes and Sonoran Desert

perceptions (significantly related to the risk perception scale) from more concrete

experiences including usage of a wilderness area like the SNP and choosing real

management responses.

Studies such as Bremner & Park (2007) also found that experiential concepts such

as having prior knowledge of invasive species and management techniques and being a

member of conservation organizations were most likely to affect respondents’ level of

support for management of invasive species. This result highlights the importance of

hands-on education and experience in affecting the public’s view of invasive species

management responses. It also highlights a deficiency in the study’s initial

conceptualization: it appears that some respondents’ experiences may directly inform

their management choices without the influence of risk perception. This complicated

relationship undoubtedly deserves more study.

The least powerful predictor in the favored management response regression was

age, which showed that older respondents favored stronger management responses.

Although the dummy variable factors were not significant predictors in the regression

model, the results nevertheless correspond to findings in other studies. Many

environmental knowledge-attitudes-behavior studies (e.g. Scott & Willits, 1994; Slimak

& Dietz, 2006) find that sociodemographic factors like age are strong predictors of


environmental attitudes and even more so, behavior (in this case preferred management

response). Additionally, other studies have also found that older members of the public

prefer stronger management responses (Bremner & Park, 2007).

Buffelgrass identification skill

Due to the novel methodology employed in this study, situating the results

regarding respondents’ skill at identifying buffelgrass in the literature is difficult.

However, the findings are extremely valuable to invasive species management, which is

increasingly aiming to involve the public in their programs through both measuring

public opinion and the use of citizen or participatory science. Such programs can aid in

creating better informed and more realistic management strategies and in some cases

reduced costs due to volunteer labor (Bonney et al., 2009; Cohn, 2008; Silvertown,

2009). The results of well-designed citizen science efforts are can also be generally

reliable: Delaney et al. (2008) found that middle/junior high students could, with

instruction, correctly identify species of invasive/native crabs in greater than 80% of

cases; the success rate for students with 2 years of university education was greater than

95%. Citizen science programs are in fact already being used to combat the buffelgrass

invasion in Tucson. A citizen science cyber-infrastructure similar to that outlined by

Graham et al. (2008) is being developed by researchers at the University of Arizona.

This ‘Spatial Decision Support System for Buffelgrass Management’ seeks to combine

citizen science efforts with advanced ecological modeling and spatial analysis in a

Geographic Information System to predict the spread of buffelgrass and recommend

management strategies based on the results (Olsson et al., 2009).

The utility of citizen science programs in combating the spread of invasive alien


species is undeniable. However, participatory projects, especially those requiring the use

of technology by the public, represent a relatively new approach, and as such their

validity and design are often in question (Cohn, 2008). Even if they can produce

internally reliable results, public input into scientific projects is subject to the biases that

come with any social research, especially due to sociodemographic variables (Graham et

al., 2008). Therefore, an understanding of what factors are important in affecting the

public’s skill at identifying invasive alien species is necessary; this study creates a

starting point for such insight.

Three variables were found to have a significant effect on respondents’ skill at

identifying buffelgrass on the landscape: seasonal residence, income, and residence in the

central city or around the urban wildland interface (UWI). In general, year-round,

wealthier, and/or central city residents are better at identifying the grass, and the sample

in general was moderately successful (average between 50-100% success). This finding

confirms the study’s hypotheses with one important exception: central city residents were

significantly better at identifying the grass than UWI residents.

The fact that urban residents were more successful at recognizing buffelgrass

presents a very significant challenge to risk perception science, and specifically this

study’s conceptualization of public invasive species perceptions. As discussed in the

literature review section, the public’s risk perception of environmental hazards is

certainly modified by their personal experiences, but this relationship is difficult to

conceptualize and can manifest itself in seemingly paradoxical ways (Halpern-Felsher et

al., 2001). Gilbert White’s research on flood events found that citizens who had

experience with natural hazards such as flooding are actually be less likely to take steps


to avoid the risk (Burton et al., 1968), while other studies have found that people are

much more willing and able to avoid environmental risks if they have had previous

experience (Slovic, 1987). Indeed, some authors have found that individual variation can

have a huge impact on how experiences are accounted for in risk perception (Barnett &

Breakwell, 2001).

The findings from this study present further issues in the relationship between

experience and risk perception. Because the buffelgrass invasion both poses a greater

risk and is more noticeable on the landscape at the UWI (Frost et al., 2010; Van

Devender & Dimmitt, 2006), the finding that central city residents are significantly better

at identifying the grass is intriguing. This result is even more surprising when combined

with the fact that respondents with higher incomes were better at identifying the grass,

but that the central city respondents have significantly lower incomes.

Unlike the findings of White and others, it appears that respondents direct

experience with observing buffelgrass on the landscape did not result in them being better

at identifying the grass or perceiving a greater risk. Perhaps residents of the UWI have

come to see buffelgrass as a normal part of their surrounding environment and are not

concerned by its presence, which could help explain why they do not perceive it as a

pressing threat and cannot identify it on the landscape. The psychometric paradigm of

risk perception finds that one of the most influential factors determining the public’s risk

perception of hazards is whether the risk is known and observable (Slovic, 1987);

therefore, residents of the UWI could be less concerned and educated about buffelgrass

because they observe it regularly, or are at least aware of its existence near their homes.

Having not experienced any direct hazard from the grass, they pay it no mind. However,


this mindset could be harmful should buffelgrass expand enough to create a continuous

understory in the native Arizona Upland saguaro-palo verde desert, which would create

the possibility of devastating fires in the ecosystem. The lack of concern among the very

population that can directly observe the invasion proves the importance of investigating

the public’s perception of invasive species.

Cluster analysis

Interpretation of the cluster analysis results allows the identification of 5

significantly different major groups of respondents (this method is similar to that used by

García-Llorente et al., 2008). The combination of demographic and other variables

yielded the following clusters: 1. active and concerned, lower class, central city residents

(15.6%), 2. less active and less concerned, lower class, central city residents (15.6%), 3.

Concerned, oldest, mobile, most highly educated, wealthy, UWI residents who are

members of conservation organizations but are not good at identification (25.4%), 4.

youngest, educated, most affluent, not at all concerned, mostly UWI residents (18%), and

5. middle class, less than college education, most concerned, UWI residents who favor

the strongest response but are the worst at ID, (25.4%). Additionally, the sample can be

grouped into 2 clusters to show the largest differences: cluster 1 in this case is older, less

wealthy, less educated, lives in the central city, is more aware of and concerned about

buffelgrass, can identify it better, and favors stronger management responses. Cluster 2

is younger, wealthier, more educated, lives around the UWI, is less aware of and

concerned about buffelgrass, is worse at identifying it, and favors weaker management

responses.

The presence of these clusters of respondents further confirms the study’s


hypotheses while also discovering nuances not found by the statistical tests measuring

variable relationships or linear regression analysis. For example, while it was correctly

hypothesized that older, more affluent, better educated, UWI respondents who are more

‘environmentalist’ or ‘conservationist’ in attitude towards the Sonoran Desert and in

general will be more concerned about the invasion, favor stronger management

responses, and be better at identifying buffelgrass, no assumption was made that there

would be subgroups within the central city and UWI samples. The central city residents

present in clusters 1 and 2 share similar demographics but have polar opposite views of

the invasion. The 3 subgroups of UWI residents (clusters 3 through 5) show that the

younger, more affluent respondents are not at all concerned about the buffelgrass

invasion, while the 2 older groups are both concerned but differ greatly in income and

education. The simple distinction between suburban and urban residents identified in the

initial hypothesis may be somewhat supported by statistical analysis methods such as

ANOVAs and linear regression, but a more holistic and complex picture emerges through

the use of cluster analysis.

The main implication of this study’s cluster analysis is that when the population is

seen as a whole, demographics seem to have unexpected effects in determining

respondents’ understanding and view of the buffelgrass invasion. While variables like

seasonal residence, age, income, and residency in the central city or around the UWI all

have statistically significant effects on the outcome variables of favored response, risk

perception, and buffelgrass identification, when respondents are clustered together it is

seen that subgroups with greatly varying views of the invasion exist within demographic

groups. As mentioned earlier, there are wealthy respondents who are concerned about


the invasion and those who are not; there are highly educated respondents who are

concerned about the invasion and those who are not. Cluster analysis also suggests that

members of the UWI group that were hypothesized to be successful at identifying the

grass due to their proximity to wildlands were in fact worse at identification, even if they

saw the invasion as a hazard and were otherwise concerned about its effects.

In addition to showing that the effects of demographics on the public’s perception

of invasive species can be complex, the use of cluster analysis in this study gives

credence to the Robbins-ian idea of social and physical networks that facilitate the spread

of invasive species. Robbins (2004) claims that “it is not species, but sociobiological

networks that are invasive.” The social organization and power structure of human

communities can help or hinder the expansion of invasives, a fact that becomes apparent

upon examining the results of this cluster analysis. The public’s risk perceptions and

favored management responses vary demographically and attitudinally, showing that

certain parts of society may be more influential in facilitating the spread of buffelgrass.

The social groups present in Tucson who are dealing with the buffelgrass invasion show

that the segments of society that could facilitate invasion through a lack of action or

awareness are not easily pigeon holed, suggesting that the invasive networks identified by

Robbins (2004) may be even more complex than previously thought.


VI. Conclusion

Management implications

Due to the holistic nature of this multi-variable study, many potential insights into

invasive species management can be garnered. The groups identified through cluster

analysis are especially useful for creating management schemes in Tucson specifically or

for any invasive species generally. Understanding what stakeholders could be affected

by a given invasion and what their opinions and values are is often identified as one of

the most important aspects of any invasive species management plan (Maguire, 2004;

McNeely, 2001; Stokes et al., 2006). Knowing how the population is divided into such

stakeholder groups can be difficult, however (Bryson, 2004). Researchers’ or managers’

biases or previous experience can often affect how they view stakeholder groups,

sometimes leading to false assumptions about the population, which in the case of

invasive species management could lead to ineffective communication and education

resulting in a sub-par management result. The use of quantitative techniques like cluster

analysis can overcome this difficulty by finding statistically valid groups within the

sample of population.

The results concerning respondent success at identifying buffelgrass also can

inform management. Due to the unexpected case that central city residents were better at

identifying the grass on the landscape than their urban-wildland interface (UWI)

counterparts, ideas about how the public perceive invasive species must be reexamined.

Because the subgroup of suburbanites who live nearest to invaded areas are actually

worse at identifying the grass and generally feel less threatened by it, managers should

not assume that citizens experiencing risk from invasive species firsthand automatically


perceive the invader as a risk, or even care enough to learn to identify it. Careful

consideration should be taken to properly assess all stakeholders’ perspectives and

compare this result to their assessed risk; finding disparities such as those present in

residents of the UWI should signal managers that education initiatives or other actions

must be taken.

A final management implication that emerges from this study’s results is the

importance of megaflora for public perception of native environments. The wide support

for the saguaro (Carnegiea gigantea) as the most essential Sonoran Desert plant displays

that the public prizes such iconic organisms and will take threats to their wellbeing

seriously. As Hounslow (2009) has stated, such iconic flora can be employed by

conservation organizations or groups seeking to raise awareness of environmental issues

because they capture the attention and feelings of the public. Although the image of the

saguaro is already used widely in the Tucson region, explicitly linking the spread of

buffelgrass with the subsequent destruction of the saguaros by fire must be a key part of

any education campaign. It has been demonstrated that the public cares for saguaros, so

using them to motivate action is a logical next step; indeed many groups already use the

loss of the saguaro as an impetus for combating the spread of buffelgrass (Wing, 2010;

Yetman & Búrquez, 1994). Additionally, because the photographic identification

exercise results show that the public does not find native grasses essential to the

ecosystem, the factors behind citizens’ differential perception of native species requires

further study. Understanding how the public perceives both native and invasive species

will help inform future management endeavors.


Recommendations for future research

By conceptualizing and measuring public perceptions of buffelgrass, this study

can provide several useful lessons for future research. First, a novel method was

presented through quantifiably measuring respondent success at identifying an invasive

species on the landscape and then comparing that result to various attitudinal and

sociodemographic variables. By seeking to incorporate this visual aspect of invasive

species perceptions, this study has demonstrated that researchers can begin to measure

how well the public recognizes elements of the environment, and that this capacity varies

significantly with several factors, including urban or rural residence, income, and

seasonal residence.

One key discovery of this research is that the interplay between demographics and

invasive species perception is complex and multi-faceted. The picture presented of

respondents’ risk perception or favored management response through standard R-mode

statistical methods such as t-tests and ANOVAs was much less nuanced than that given

by further exploration of the dataset through cluster analysis. By clustering respondents

and their attitudes, numerous statistically different subgroups appeared in what could

superficially be seen as clear-cut groups of the population. For example, while lower

class, uneducated residents living in the central city could at first be written off as not

being concerned about the invasion based on statistical tests, the cluster analysis reveals

that there are in fact two demographically similar groups in the city who have very

different risk perceptions of buffelgrass and favor very different responses. The initial

grouping of the sample into central city and UWI subgroups was based on the faulty

assumption that this distinction would be most powerful in affecting public perceptions of


buffelgrass; future studies are challenged not to simplify the effect of demographics on

invasive species perceptions but to fully explore the social dimensions of this issue.

A final implication for continued study of this topic is related to the success of the

cluster analysis for this project: examination of public opinion regarding invasive species

should use multiple methods to question assumptions and produce more robust results.

The combined use of statistical tests, regression models, and cluster analysis allowed the

dataset to be examined from several different perspectives and resulted in a more refined

picture of Tucsonans’ view of buffelgrass. As displayed through the use of multiple

conceptual diagrams, numerous factors are at play in shaping the public’s relationship

with buffelgrass, and the use of multiple methods of analysis helps make sense of these

widely varying dynamics. Future studies should explore public perceptions of invasive

species with similar multi-method designs to continue refining the understanding of this

complicated social and geographical issue.


VII. Appendix A – Survey Instrument

DESERT VIEWS: A STUDY OF TUCSONANS’ KNOWLEDGE, ATTITUDES, AND BELIEFS

ABOUT THE SONORAN DESERT

You are one of a lucky few Tucsonans who have been randomly selected to participate in a survey that will help us

understand prevailing knowledge, attitudes, and beliefs about the Sonoran Desert. This research could help local environmental policy better serve your interests.

The survey should take 10-20 minutes of your time.

Your participation is voluntary, but the success of this project depends on you answering all the following questions as accurately as you can. When you are finished, use the enclosed postage-paid envelope to mail your survey back to us.

_____________________________________________________

You may contact us (see below) or Colgate’s Institutional Review Board Chair ([email protected]) with any questions about this study or your rights as a participant. If you wish to participate in an interview, please write your

contact information on page 2. This information will be kept confidential and used only by us for this project. All your responses are otherwise anonymous. We are happy to provide our research results upon request.

With sincere thanks and best regards,

Jake Brenner, Visiting Assistant Professor ([email protected]) and Sam Walker, Class of 2011 ([email protected])

Colgate University Geography Department 13 Oak Drive, Hamilton NY 13346

Phone: 520-664-5757


If you wish to participate in an interview, please write your contact information here (below). Otherwise, leave this page blank.


Section I: How and Why You Live in Tucson

We would first like to know about your residence in Tucson. 1.) How many months per year do you spend in Tucson? Check one box.

Less than 1 month 1-4 months 5-7 months 8-10 months 11-12 months

2.) In what year did you begin living in Tucson? Write your answer or check the box.

_________ I was born here 3.) What factors influenced your decision to live in Tucson? Check all appropriate factors and circle the most important.

Employment College or university education Good schools for children Family roots in the region Family currently living in the region Scenic views The natural desert environment Vibrant community Good place to start or raise a family Warm climate Dry climate

Sunshine Safe community to live in Cultural events or attractions Interesting regional culture or history Retirement Health or wellness “Urban” outdoor recreation

(e.g. golf, tennis, swimming, city walking) “Wildland” outdoor recreation

(e.g. hiking, mountain biking, camping)

4.) Do you rent or own your Tucson residence? Check one box.

Rent Own 5.) How would you describe your Tucson residence? Check one box.

Single-family Duplex Apartment Mobile home


Section II: Involvement with the Outdoors

We are also interested in your participation in outdoor activities. 6.) How often do you participate in the following activities? Check one box in each row to indicate frequency.

Activity Daily Weekly Monthly Yearly Rarely Never “Urban” outdoor recreation (e.g. golf, tennis, swimming, walking)

“Wildland” outdoor recreation (e.g. hiking, mountain biking, camping)

7.) How often do you visit the following sites? Check one box in each row to indicate frequency.

Site Daily Weekly Monthly Yearly Rarely Never Saguaro National Park (East or West) Ironwood Forest National Monument Catalina State Park Mt. Lemmon / Coronado National Forest

Tucson Mountain Park Reid Park Rillito River Walk Santa Cruz River Walk Reddington Pass Local swimming pools or splash parks Local golf courses Arizona-Sonora Desert Museum Tohono O’Chul Park Tucson Botanical Garden

8.) If you visit Saguaro National Park (East or West), Tucson Mountain Park, or Ironwood Forest National Monument, what are your main reasons for doing so? Check all appropriate boxes.

I do not visit these sites. Scenic beauty Exercise Recreation Environmental/ecological education

Nature observation (including birds) Relaxation or solitude Activity with family or friends “Weedwacking” or other natural resource

management activities

Comment [SW7]: L.F. seems to think these instructions won’t be read and people will write in a number, I’m not sure I agree. Even if they do it

could be converted to the correct response.

M.H. thinks we should change them completely, I

think maybe either a free response of how many

times per month or an ordinal scale for how many times per month. We can extrapolate from there.

Comment [jcb8]: I’m reviewing LF’s and MH’s rationales before commenting. I’m inclined to agree

with LF, just because she’s got more training on this methodology. I also recall some kind of quasi-

interval bin system where you put an X on a

graduated line. I’ll get back to you on this and the photos later.

Comment [SW9]: And maybe a scale for how

many times per year here?

Comment [SW10]: M.H. wants this to capture weedwacking- how would we do that?

Conservation-related activities?

Comment [jcb11]: Recently heard from a senior

seminar student who surveyed a bunch of

weedwackers that a principal motivator for weedwacking was access to remote (sometimes

permit-only) areas of these very reserves.

Interesting….

Comment [SW12]: M.H. wants this to capture weedwacking- how would we do that?

Conservation-related activities?


Section III: Views of the Sonoran Desert

We are also interested in how you view yourself in relation to the surrounding Sonoran Desert environment. Check the most appropriate box to indicate your beliefs about the following statements. 9.) People’s wellbeing depends on their surrounding ecosystem.

Strongly Disagree Disagree Neutral Agree Strongly Agree 10.) Ecosystems need to be preserved, even if doing so requires social or economic compromises.

Strongly Disagree Disagree Neutral Agree Strongly Agree 11.) How would you finish this sentence? Check one box in each row.

“I believe the Sonoran Desert is…” Strongly Disagree

Disagree Neutral Agree Strongly

Agree

…beautiful. …unique. …barren. …harsh. …dangerous. …prime real estate. …a tourist attraction. …important for agriculture. …important for livestock. …biologically diverse. …culturally significant. …historically significant. …in need of preservation.

12.) How do you consider the following in relation to the Sonoran Desert environment? Check one box in each row.

Severe threat

Moderate threat

Neutral Moderate benefit

Strong benefit

Industrial production

Agriculture

Livestock production

Residential development

Invasive alien species

Climate change

Recreational activities

Comment [SW13]: This is good but only some measure if they value this aspect of the S.D.- some are just measures of their perception, i.e. they may

think the desert is biologically diverse but not care

about that.

Comment [SW14]: What about something like this for 12? I am still worried about putting ideas in

people’s heads; I think it is worth erring on the side

of caution in this regard, because while it may be unlikely that people will view residential dev as

helpful, we can’t rule that out.


13.) What are the essential plants of the Sonoran Desert? Check all appropriate photos. Then circle one photo showing the Sonoran Desert’s most important plant.

Desert ironwood Desert marigold

Chain fruit cholla Creosote bush Giant saguaro

Desert agave Banana yucca Desert prickly pear

Ocotillo Curly mesquite grass Buffelgrass

Purple three-awn Arizona poppy Blue paloverde


Section IV. Buffelgrass

Perhaps you have heard of buffelgrass (Pennisetum ciliare), an exotic species in the Sonoran Desert. 14.) How, if ever, did you first learn about buffelgrass? Check one box.

Word of mouth Printed news or magazines Scientific publications Television Radio Internet Direct observation on my property or in my neighborhood Direct observation in a public place Direct observation on a roadside or median strip Direct observation in an empty lot Other - Please explain: _______________________________________________________________. or This is the first time I have heard of buffelgrass.

Indicate your level of agreement with the following statements. Check one box for each question. 15.) Buffelgrass invasion poses a problem in the Sonoran Desert at large.

Strongly Disagree Disagree Neutral Agree Strongly Agree 16.) Buffelgrass invasion poses a problem in the Tucson region.

Strongly Disagree Disagree Neutral Agree Strongly Agree 17.) Buffelgrass invasion poses a problem in my neighborhood.

Strongly Disagree Disagree Neutral Agree Strongly Agree 18.) Buffelgrass invasion poses a problem on my property.

Strongly Disagree Disagree Neutral Agree Strongly Agree

19.) Where do you see buffelgrass in these local Sonoran Desert landscapes? Carefully circle all buffelgrass patches.


20.) What is your level of concern about the following statements? Check one box in each row.

Serious Moderate Slight None Unsure Buffelgrass will outcompete native perennial plants (e.g. cactus, shrubs, and trees)

Buffelgrass will outcompete native grasses (annual or perennial) Buffelgrass will outcompete other native plants (e.g. wildflowers)

Buffelgrass will deplete soil nutrients Buffelgrass will deplete soil moisture Buffelgrass will fuel wildfires Buffelgrass fires will kill native plants Buffelgrass fires will kill native animals Buffelgrass fires will damage people’s homes and property Buffelgrass fires will hurt or kill me Buffelgrass invasion will depreciate the value of my home Buffelgrass invasion will interfere with my job Buffelgrass invasion will interfere with my favorite outdoor activities Buffelgrass invasion will degrade my desert views Buffelgrass invasion will damage protected areas (e.g. Saguaro National Park)

Buffelgrass invasion will interfere with conservation activities (e.g. the Sonoran Desert Conservation Plan)

Buffelgrass invasion will hurt Tucson’s real estate market Buffelgrass invasion will strain limited public funds and/or resources (e.g. by increasing fire-fighting demands)

Buffelgrass invasion will spread quickly Buffelgrass invasion will be uncontrollable

Now review your answers above and circle your most important concern. 21.) What, if any, are appropriate responses to buffelgrass invasion in and around Tucson? Check all appropriate responses and circle the most appropriate response.

Manual removal and/or herbicide spraying on roadsides Manual removal from protected areas (e.g. Saguaro National Park) Manual herbicide spraying in protected areas Aerial herbicide spraying in protected areas Paid staff time for buffelgrass control in protected areas Volunteer efforts for buffelgrass control in protected areas (e.g. the Sonoran Desert Weedwackers) Public education about buffelgrass Regulation stipulating buffelgrass removal from private land (e.g. Pima County Ordinance #2008-117) Regulation of buffelgrass sale, transport, and cultivation

(e.g. Arizona’s 2005 designation of buffelgrass as a “Noxious Weed”) Increased support for Tucson’s fire departments Biological control (e.g. an insect that eats buffelgrass) Native ecosystem restoration

Comment [jcb16]: On further review, these

MUST be definitive statements, or else they’re almost impossible to disagree with.

Comment [SW15]: Perhaps we should change

this to a Likert scale plus an unsure option- that way

we can capture people who think the opposite and it will make them easier to disagree with/form an

opinion. Maybe even switch the wording of some

like we have in other sections to keep the respondents on their toes.


22.) Do you volunteer or otherwise participate in buffelgrass-related activities? Check all appropriate boxes.

Buffelgrass Summit 2007 Arizona-Sonora Desert Museum’s Invaders citizen science program Beat Back Buffelgrass Day: 2008 2009 2010 Sonoran Desert Weedwackers or other Weedwacker group Involvement with another buffelgrass-related organization (e.g. the Buffelgrass Working Group)

23.) Which statement best matches your attitude about the buffelgrass invasion? Check one box.

No response is needed. Buffelgrass should be promoted. Buffelgrass should be eradicated. Buffelgrass should be controlled. Buffelgrass invasion cannot be controlled, but should be managed as part of the Sonoran Desert ecosystem.

24.) Please circle on the map the zone where you live (e.g. N, SE, SW, etc.).

Then mark with Xs all of the locations where you think buffelgrass poses a threat.

Comment [SW17]: M.H. suggested this to include all Weedwacker groups.

Comment [SW18]: Thought this should be wider in scope

Comment [jcb19]: Agreed-both.


Section V. Demographics and Socioeconomics

In order to analyze the results of this survey, we have to collect some demographic and socioeconomic data. Rest assured that this information is anonymous. 25.) Are you male or female?

Male Female 26.) What is your age?

18-22 23-29 30-39 40-49 50-59 60+ 27.) What is the highest level of education you have completed? Check one.

Less than high school High school/GED Some college 2-year college degree 4-year college degree Master’s degree Doctoral degree (e.g. Ph. D., M. D., J. D., etc.)

28.) What is your 2010 IRS income tax bracket?

Tax bracket Married couples filing jointly Most single filers

10% Not over $16,750 Not over $8,375 15% $16,750 – $68,000 $8,375 – $34,000 25% $68,000 – $137,300 $34,000 – $82,400 28% $137,300 – $209,250 $82,400 – $171,850 33% $209,250 – $373,650 $171,850 – $373,650 35% Over $373,650 Over $373,650

29.) Are you of Hispanic, Latino, or Spanish origin?

No, not of Hispanic, Latino, or Spanish origin Yes, Mexican, Mexican American, Chicano Yes, Puerto Rican Yes, Cuban Yes, another Hispanic, Latino, or Spanish origin

30.) What is your race?

White Black, African-American, or Negro American Indian or Alaska Native Asian Indian Japanese Native Hawaiian Chinese Korean Guamanian or Chamorro Filipino Vietnamese Samoan

Other Asian Other Pacific Islander Another race


31.) Under which category does your occupation fall, according to the 2010 Standard Occupational Classifications? If you are retired or unemployed, under what category did your occupation fall? Check the most appropriate box.

Management Business and financial operations Computer and mathematical Architecture and engineering Life, physical, and social science Community and social services Legal Education, training, and library Healthcare practitioners and technical Healthcare support Arts, design, entertainment, sports, media

Personal care and service Sales and related Office and administrative support Farming, fishing, and forestry Construction and extraction Installation, maintenance, and repair Production Transportation and material moving Military specific Protective service Building and grounds cleaning and maintenance

32.) How, if at all, are you involved with the following organizations? Check all appropriate boxes for each row.

Organization Donor Member Volunteer Employee

Tucson Audubon Society

National Audubon Society

The Nature Conservancy

League of Conservation Voters

Friends of Sabino Canyon

Coalition for Sonoran Desert Protection

Arizona Open Land Trust

Safari Club International

Sierra Club

Sky Island Alliance

The Sonoran Institute

Tucson Wildlife Center

Rincon Institute

International Dark Sky Association

Wildlands Network

Tucson Clean & Beautiful

Friends of Saguaro National Park

Defenders of Wildlife

Arizona Wilderness Coalition

The Arizona-Sonoran Desert Museum

Arizona Native Plants Society

Center for Biological Diversity

This is the end of the survey. We thank you for your participation. If you are interested in an interview, please write your contact information on page 2.

Now please enclose your completed survey in the provided envelope and drop it in the mail.


VIII. Appendix B – Variable Descriptions

Survey Question

BA

CK

GR

OU

ND

Ca

teg

ori

cal

CC or UWI 24

Gender 25

Hispanic or Latino 29

Race 30

Most essential SD plant 13

Ord

ina

l

Seasonal residence 1

Age 26

Education 27

Income 28

Continuous Year moved to Tucson 2

Ord

ina

l/C

on

tin

uo

us

Reasons for moving to Tucson Scale from 3

INT

ER

VE

NIN

G

Wildland recreation frequency Scale from 6 & 7

Urban recreation frequency Scale from 6 & 7

Saguaro National Park (SNP) usage Scale from 8

Environmental attitudes Scale from 9 &

10

Sonoran Desert perceptions Scale from 11

Conservation organization participation Scale from 32

Conservation organization membership Scale from 32

OU

TC

OM

E

Risk perception of BG Scale from 15 &

16

Identification of BG on the landscape Scale from 19

Ordinal Favored BG invasion response 23


IX. Appendix C – Scale Variable Creation

Reasons for moving to Tucson - FACTORS3_SCALE = FACTORS3_SCALE = -1*(

FACTORS3i + FACTORS3ii + FACTORS3iii + FACTORS3iv + FACTORSv +

FACTORS3viii + FACTORS3ix + FACTORS3xiii + FACTORS3xiv + FACTORS3xv +

FACTORS3xvi + FACTORS3xvii) + (FACTORS3vi + FACTORS3vii + FACTORS3x +

FACTORS3xi + FACTORS3xii + FACTORS3xviii + FACTORS3xix)

This scale was created based on responses to Q3, which asked what reasons respondents had for

moving to Tucson. This list was divided into utilitarian (school, work, family, etc.) and

environmental amenity-based (outdoor recreation, climate, desert environment, etc.) reasons,

with a checkmark for each reason coded with a value of 1. The calculation of the scale took

utilitarian reasons for moving to Tucson, added them together, and multiplied them by -1. This

number was then added to the environmental amenity-based reasons for moving to Tucson to

result in a scale with low numbers corresponding to more utilitarian reasons and high numbers

more environmental amenity-based reasons. For a discussion of utilitarian- versus

environmental amenity-based reasons for migration, see Dearien et al. (2005) and Rudzitis &

Johnson (2000).

Wildland recreation frequency - OUTD_REC_W_FQ = MEAN(REC_FREQ6A,

SITE_FREQ7A,SITE_FREQ7B,SITE_FREQ7C,SITE_FREQ7D,SITE_FREQ7E,

REC_FREQ6A)

This scale was created from Q6 & Q7 and reports the mean of respondents’ frequency of visiting

5 ‘wildland’ recreation sites (Saguaro National Park [East or West], Ironwood Forest National

Monument, Catalina State Park, Mt. Lemmon / Coronado National Forest, Tucson Mountain

Park) and their self-reported ‘wildland’ recreation frequency.

Urban recreation frequency - OUTD_REC_U_FQ = MEAN(SITE_FREQ7F,SITE_FREQ7G,

SITE_FREQ7H,SITE_FREQ7I,SITE_FREQ7J,SITE_FREQ7K,SITE_FREQ7L,SITE_FREQ7M

,SITE_FREQ7N, REC_FREQ6B)

This scale was created from Q6 & Q7 and reports the mean of respondents’ frequency of visiting

9 ‘urban’ recreation sites (Reid Park, Rillito River Walk, Santa Cruz River Walk, Reddington

Pass, local swimming pools or splash parks, local golf courses, Arizona‐Sonora Desert Museum,

Tohono O’Chul Park, Tucson Botanical Garden) and their self-reported ‘urban’ recreation

frequency.

Saguaro National Park (SNP) usage - SNP_REASONS8_SCALE = -1*( SNP_REASONS8iii +

SNP_REASONS8iv + SNP_REASONS8vii + SNP_REASONS8viii) + (SNP_REASONS8ii,

SNP_REASONS8v, SNP_REASONS8vi, SNP_REASONS8ix)

Q8 asked respondents to indicate their reasons for visiting the SNP; this list of reasons included

both utilitarian and environmental amenity-based choices. To calculate the scale, the utilitarian

reasons were added together and multiplied by -1 and the environmental amenity-based reasons

were added to this sum. The resulting scale ranges from -4 to +4, with negative numbers

meaning more utilitarian usage and positive numbers meaning more environmental amenity-


based usage. The conceptual division of reasons into utilitarian and environmental amenity-

based categories follows ideas similar to those presented in outdoor recreation literature

(Jackson, 1986).

Environmental attitudes - ENV_ATT_SCALE = MEAN(WELLBEING_ECO9,

PRESERVE_ECO10)

This scale is the mean of Q9 and Q10, which asked how respondents relate to their natural

environment. This variable serves as an indicator of respondents’ overall environmental

attitudes because it combines 2 key aspects of humans’ relations to nature: Q9 measures how

strongly respondents believe their wellbeing is dependent on their surrounding ecosystem and

Q10 measures how strongly they wish to preserve ecosystems, even if doing so requires

compromises. This scale is a simplified, concise version of similar scales such as the New

Ecological Paradigm scale developed by Dunlap et al. (2000).

Sonoran Desert perceptions - SD_PER_SCALE = MEAN(MEAN(SD_IS11C, SD_IS11F,

SD_IS11H, SD_IS11I), MEAN(SD_CHANGE12A, SD_CHANGE12B, SD_CHANGE12C,

SD_CHANGE12D, SD_CHANGE12E, SD_CHANGE12F, SD_CHANGE12G),

MEAN(SD_IS11A_R, SD_IS11B_R, SD_IS11J_R, SD_IS11K_R, SD_IS11L_R,

SD_IS11M_R)

This scale reports the mean of several other calculated means: the mean of respondents’

responses on a Likert-type scale to parts of Q11 corresponding to a ‘conservation-based’ view

the Sonoran Desert; the mean of respondents’ responses to Q12, which asked if they saw certain

kinds of development in the desert as a threat or a benefit; and recoded responses from Q11 that

reversed the scale for parts corresponding to a ‘utilitarian-based’ view of the desert. This

calculation resulted in a scale with low numbers corresponding to a ‘conservation-based’ view

and high numbers to a ‘utilitarian-based’ view of the desert. Similar quantitative ‘conservation-

based’ and ‘utilitarian-based’ scales have been used to measure attitudes in studies of water

consumption (Corral-Verdugo, Bechtel, & Fraijo-Sing, 2003) and fisheries management (Olver,

Shuter, & Minns, 1995).

Conservation organization participation - ORGS32_SCALE= ORGS32A + ORGS32B +

ORGS32C + ORGS32D + ORGS32E + ORGS32F + ORGS32G + ORGS32H + ORGS32I +

ORGS32J + ORGS32K + ORGS32L + ORGS32M + ORGS32N + ORGS32O + ORGS32P +

ORGS32Q + ORGS32R + ORGS32S + ORGS32T + ORGS32U + ORGS32V

This scale measured total level of participation in all conservation organizations, with an

additional point of weight given to responses as they increased in level of participation from

donor to member to volunteer to employee.

Conservation organization membership - ORGS32_MEMB = ORGS32A_R + ORGS32B_R +

ORGS32C_R + ORGS32D_R + ORGS32E_R + ORGS32F_R + ORGS32G_R + ORGS32H_R

+ ORGS32I_R + ORGS32J_R + ORGS32K_R + ORGS32L_R + ORGS32M_R + ORGS32N_R

+ ORGS32O_R + ORGS32P_R + ORGS32Q_R + ORGS32R_R + ORGS32S_R + ORGS32T_R

+ ORGS32U_R + ORGS32V_R


This scale is similar to the scale for conservation organization participation except that it

measured simple yes or no participation in each organization from recoded responses, resulting

in a scale showing how many organizations each respondent was involved in at any level.

Risk perception of BG - THRT_SDT = MEAN(PROB_SD15,PROB_TUC16)

This scale reported the mean of 2 Likert-type scale responses for perceived threat level from

buffelgrass to the Sonoran Desert (Q15) and Tucson (Q16). These combined responses served as

an indicator primarily of respondents’ view of the invasion’s general impact on their region.

While many other methods exist for measuring risk perception of ecological hazards (McDaniels

et al., 1995; Slimak & Dietz, 2006), this method was chosen to be a concise measure of

respondents’ general risk perception.

Identification of BG on the landscape - BG_ID = MEAN(BG_ID19A,BG_ID19B,BG_ID19C,

BG_ID19D)

Buffelgrass identification questions were coded based on respondents’ correct circling of

buffelgrass in the 4 landscapes presented in Q19; 1 point was assigned if all buffelgrass was

correctly circled, 2 points if more than fifty percent was correctly circled, and 3 points if no

buffelgrass was correctly circled. Respondents who incorrectly circled areas not covered by

buffelgrass were given 3 points; Respondents who did not circle any patches on all 4

photographs were marked as not responding and excluded from the analysis. The mean point

score for all 4 landscapes was calculated and used as the scale variable representing success of

buffelgrass invasion on the landscape. Values range from 1 to 3, with a value of 1 indicating

successful identification and a value of 3 indicating failure.


X. Appendix D – Gower’s Similarity Coefficient from Wishart (2006)

Gower’s Similarity Coefficient compares two cases i and h as follows:

𝑠𝑖ℎ =∑ (𝑤𝑖ℎ𝑗 − 𝑠𝑖ℎ𝑗)𝑗

∑ 𝑤𝑖ℎ𝑗𝑗

where sihj denotes the contribution provided by the jth

variable, and wihj is usually 1 or 0

depending upon whether or not the comparison is valid for the jth

variable. The effect of the

denominator ∑ 𝑤𝑖ℎ𝑗𝑗 is to divide the sum of the similarity scores by the number of variables.

Gower (1971) defines the value of sihj for ordinal and continuous variables as follows:

s𝑖ℎ𝑗 = 1 − |𝑥𝑖𝑗 − 𝑥ℎ𝑗|

𝑟𝑗

where rj is the range of values for the jth

variable. For continuous variables sihj ranges between 1,

for identical values xij = xhj, and 0, for the two extreme values xjmax – xjmin. For a binary variable

(or dichotomous attribute), Gower defines the component of similarity and the weight according

to the table (below), where + denotes that attribute j is ‘present’ and – denotes the attribute j is

‘absent’.

Value of attribute j

Case i + + - -

Case h + - + -

sihj 1 0 0 0

wihj 1 1 1 0

Thus sihj = 1 if cases i and j both have attribute j ‘present’ or 0 otherwise, and the weight wihj

causes negative matches to be ignored.

The value of sihj for nominal variables is 1 if xij = xhj, or 0 if xij ≠ xhj. Thus sihj = 1 if cases

i and h both have the same ‘state’ for attribute j, or 0 if they have different ‘states’, and wihj = 1 if

both cases have observed states for attribute j, or zero if either value is missing.


XI. Appendix E – Increase in Sum of Squares or Ward’s Method from Wishart (2006)

The Euclidian Sum of Squares Ep for a cluster p is the sum of the Squared Euclidian

Distances between all the members of the cluster p and its mean:

𝐸𝑝 = ∑ ∑ (𝑥𝑖𝑗 − 𝜇𝑝𝑗)

2𝑗𝑖𝜀𝑝

𝑣

where for each variable j, 𝑥𝑖𝑗 is the value in case i, 𝜇𝑝𝑗 is the mean in cluster p, and v is the

number of variables. Any missing proximities in this study that occurred due to missing values

in the data are ignored. Thus two cases or clusters p and q whose similarity spq is missing cannot

be directly combined. However, they can subsequently join the same cluster by merging with

another case or cluster r for which spr and sqr are both valid. Although the presence of missing

values can lead to unpredictable results if the proximity matrix contains a large number of

missing entries, or where the proximities between two clusters are all missing, the number of

missing values in this data set is relatively low (around 9%) and so these issues do not present a

major problem.

The total Euclidian Sum of Squares over all clusters, for a given classification, is

therefore 𝐸 = ∑ 𝐸𝑝𝑝 . Increase in Sum of Squares combines two clusters p and q which result in

the least increase 𝐼𝑝∪𝑞 in E; that is, for which 𝐼𝑝∪𝑞 = 𝐸𝑝∪𝑞 − 𝐸𝑝 − 𝐸𝑞 is a minimum.


XII. Appendix F – Sample Demographics

Table 1. Demographics of sample (n) by central city (CC) and urban-wildland interface (UWI) residence.

Months per year

spent in Tucson

Gender Age Education Level 2010 Income

Tax Bracket*

Hispanic, Latino, or

Spanish origin

Race

Response Percent Response Percent Response Percent Response Percent Response Percent Response Percent Response Percent

CC UWI Total CC UWI Total CC UWI Total CC UWI Total CC UWI Total CC UWI Total CC UWI Total

1-4 months

0 0 0.8 Male 40.5 32.1 34.2 23-29

0 1.8 0.9 High school/GED

8.1 5.4 5.5 10%

5.7 3.9 4.0 No

91.7 96.4 92.6 White

100 94.5 97.2

5-7

months

0 1.8 1.7 Female 59.5 67.9 65.8 30-39

13.5 1.8 7.3 Some

college

18.9 21.4 17.3 15%

54.3 29.4 36.6 Yes,

Mexican

5.6 1.8 4.6 American

Indian or Alaska

Native

0 1.8 0.9

8-10

months

2.8 1.8 5.0 40-49

16.2 19.6 17.3 2-year

college degree

5.4 10.7 7.3 25%

25.7 37.3 33.7 Yes,

another origin

2.8 1.8 2.8 Chinese 0 1.8 0.9

11-12

months

97.2 96.5 92.6 50-59

24.3 32.1 29.1 4-year

college

degree

35.1 26.8 30.0 28%

8.6 19.6 15.8 Another

race

0 1.8 0.9

60+ 45.9 44.6 45.5 Master's degree

29.7 21.4 29.1 33%

5.7 5.9 7.9

Doctoral

degree

2.7 14.3 10.9 35% 0 3.9 2.0

Table 2. Sample respondents' median year began living in Tucson.

Table 3. Number and percent of respondents by sub-sample.

Total CC Dweller UWI Dweller Not recorded

n 122 (100%) 37 (30.3%) 57 (46.7%) 28 (23%)

CC: 1979.5

UWI: 1980.5

Total: 1980

*significant (p<.05) difference in means

between CC and UWI groups according to

independent samples t-test


Table 1. Demographics of sample population (N) from US Census by central city (CC) and urban-wildland interface

(UWI) residence.

CC UWI

Median income 27665 47470

Median age 34.25 42.05

Percent male 48.98% 48.46%

Percent female 51.02% 51.54%

Percent over 65 12.79% 16.76%

Percent white 75.44% 85.50%

Percent black 3.98% 2.67%

Percent Hispanic/Latino 25.08% 15.78%

Less than high school 14.28% 6.28%

High school 23.96% 18.99%

Some college 25.30% 24.32%

2 year college 6.56% 8.50%

4 year college 17.67% 23.79%

Master's 8.20% 11.59%

Doctoral 2.08% 2.94%


XIII. Appendix G – Regression Analysis Results

Table 1. Variation explained by regression models

Dependent variable R R2

Adjusted R2 Std. Error of the Estimate

Favored response to

buffelgrass invasion

0.205a

Risk perception of

buffelgrass invasion

0.511 0.261 0.217 0.68246

Buffelgrass identification

skill

0.554 0.307 0.206 0.41990

a Cox and Snell pseudo-R

2

Table 2. ANOVA results for regression models

Dependent variable Sum of Squares df Mean Square F Sig.

Risk perception of buffelgrass

invasion

30.590 10 3.059 8.979 0.000

Buffelgrass identification skill 4.287 8 0.536 3.039 0.007

Table 3. Model fit for favored response to buffelgrass invasion PLUM regression model

Model -2 Log

Likelihood

Chi-

Square

d

f

Sig.

Intercept

Only

210.715

Final 189.664 21.051 4 .00

0

Table 4. Coefficients and collinearity measures

Standardized PLUM Coefficients - Favored response to buffelgrass

invasion

No Collinearity

Factor Estima

te

Std.

Error

Wald Sig.

Threshold Response = ‘no response’ 1.632 2.247 .527 .468

Response = ‘manage’ 4.518 2.300 3.858 .050

Response = ‘control’ 5.633 2.326 5.866 .015

Location Scale for pragmatic- or environmental

amenities-based reasons for using SNP

.413 .208 3.948 .047

Scale for environmental attitudes .158 .318 .248 .618

Scale for pragmatic- or conservation-

based view of the Sonoran Desert

-.153 .428 .127 .721

Threat to the SD and Tucson 1.196 .345 12.000 .001

Standardized Coefficients - Risk perception of buffelgrass invasion Collinearity

Factor Beta t Sig. Toleranc

e

VIF

(Constant) 6.000 0.000

Dummy for 30-39 0.139 1.336 0.185 0.672 1.488

Dummy for 40-49 -0.032 -0.247 0.805 0.437 2.288


Dummy for 50-59 0.157 1.111 0.269 0.366 2.733

Dummy for 60+ 0.172 1.158 0.250 0.331 3.019

Scale for environmental attitudes 0.245 2.604 0.011 0.826 1.211

Scale for pragmatic- or conservation-based

view of the Sonoran Desert

-0.276 -2.886 0.005 0.801 1.249

Standardized Coefficients – Buffelgrass identification skill Collinearity

Factor Beta t Sig. Toleranc

e

VIF

(Constant) 10.576 .000 .508 1.969

Dummy for 5-7 months .066 .421 .675 .930 1.076

Dummy for 8-10 months -.116 -.994 .325 .452 2.214

Dummy for 10% .306 1.830 .073 .340 2.942

Dummy for 15% .314 1.632 .108 .348 2.873

Dummy for 25% .162 .852 .398 .472 2.117

Dummy for 28% .013 .079 .938 .734 1.362

Dummy for 33% .032 .247 .806 .815 1.228

Dummy for CC resident -.397 -3.190 .002 .508 1.969


XIV. Appendix H – Cluster Analysis Results

Table 2. Cluster profiles with average variable or z-score value.

Cluster 1 2 3 4 5

Weight (number of cases) 19 19 31 22 31

Percentage 15.6 15.6 25.4 18.0 25.4

Seasonal residence 4.74 5 4.9 4.82 4.97

Age 5.33 5.5 5.27 4.7 4.86

Education 5.33 5.33 4.12 5.75 4.57

Income 2.59 3.33 2.48 3.5 2.96

CC or UWI 0 1 0.17 0.81 0.96

Buffelgrass identification 1.36 2.1 1.58 1.77 1.8

Wildland' recreation 4.02 5.14 4.73 4.75 4.49

Urban' recreation 4.23 5.12 4.76 4.75 4.68

Risk perception 4.6 4.14 4.67 4.26 4.57

Reasons for moving to Tucson 0 -0.26 0 0.82 0.84

SNP useage -0.37 0.17 -0.31 -0.55 -0.83

Sonoran Desert perceptions 0 0 0 0.05 0

Enviromental attitudes 4.33 3.92 4.3 4.39 4.22

Organization participation 3.63 3 1.1 0.95 0.77

Organization membership 2.05 1.74 0.74 0.5 0.45

Favored response 4.69 4 4.35 4 3.85

Comment [j20]: This can be made to look more accessible for your readers.

Comment [u21]: A LOT more accessible—in fact, while I haven’t read much cluster analysis

research, my inclination is not to report raw-ish data

such as these. What’s interesting is a dendrogram with labeling that makes it clear what the groupings

are and what they mean.


Figure 1. Profile for cluster 1 with average, minimum, and maximum values from all 5 clusters.


-1

0

1

2

3

4

5

6

7

Va

ria

ble

/Z

-sco

re v

alu

e

Variable

Average

Minimum

Maximum

-1

0

1

2

3

4

5

6

7

Va

ria

ble

/Z-s

core v

alu

e

Variable

Average

Minimum

Maximum



Figure 4. Profile for cluster 4 with average, minimum, and maximum values from all 5 clusters

-1

0

1

2

3

4

5

6

7

Va

ria

ble

/Z-s

core v

alu

e

Variable

Average

Minimum

Maximum

-1

0

1

2

3

4

5

6

7

Va

ria

ble

/Z-s

core v

alu

e

Variable

Average

Minimum

Maximum



-1

0

1

2

3

4

5

6

7

Va

ria

ble

/Z-s

core v

alu

e

Variable

Average

Minimum

Maximum


References

Agarwal, C., Green, G. M., Grove, J. M., Evans, T. P., & Schweik, C. M. (2000). A

review and assessment of land-use change models: dynamics of space, time, and

human choice. Fourth International Conference on Integrating GIS and

Environmental Modeling (GIS/EM4), 2–8.

Albaum, G. (1997). The Likert scale revisited: an alternate version. Journal of the Market

Research Society, 39(2), 331-348.

Alberti, M., Marzluff, J. M., Shulenberger, E., & Bradley, G. (2003). Integrating Humans

into Ecology: Opportunities and Challenges for Studying Urban

Ecosystems. Bioscience, 53(12), 1169.

Aldenderfer, M. S., & Blashfield, R. K. (1984). Cluster analysis. Quantitative

Applications in the social sciences. Beverly Hills: Sage Publication,

Allen, T. C., & Thomas W. Hoekstra. (1992). Toward a unified ecology . New York:

Columbia University Press.

Alston, K. P., & Richardson, D. M. (2006). The roles of habitat features, disturbance, and

distance from putative source populations in structuring alien plant invasions at the

urban/wildland interface on the Cape Peninsula, South Africa. Biological

Conservation,132(2), 183-198. doi:DOI: 10.1016/j.biocon.2006.03.023

Andersen, M. C., Adams, H., Hope, B., & Powell, M. (2004). Risk assessment for

invasive species. Risk Analysis, 24(4), 787-793.

Andreu, J., Vila, M., & Hulme, P. E. (2009). An Assessment of Stakeholder Perceptions

and Management of Noxious Alien Plants in Spain. Environmental

Management, 43(6), 1244-1255.

Arizona Department of Agriculture. (2010). Prohibited, Regulated and Restricted

Noxious Weeds. Retrieved 4/28/2010, 2010,

from http://www.azda.gov/PSD/quarantine5.htm

Arizona Native Plants Society.Arizona Native Plant Society. Retrieved 2/25/2010, 2010,

from http://www.aznps.com/invasives/weedwackers.html

Arizona-Sonora Desert Museum. (2011). Arizona-Sonora Desert Museum. Retrieved

3/23/2011, 2011, from http://www.desertmuseum.org/

http://www.azda.gov/PSD/quarantine5.htm

http://www.aznps.com/invasives/weedwackers.html

http://www.desertmuseum.org/


Arriaga, L., Castellanos, A. E. V., Moreno, E., & Alarcon, J. (2004). Potential Ecological

Distribution of Alien Invasive Species and Risk Assessment: a Case Study of Buffel

Grass in Arid Regions of Mexico. Conservation Biology, 18(6), 1504.

Arrow, K., Solow, R., Portney, P. R., Leamer, E. E., Radner, R., & Schuman, H.Report of

the NOAA Panel on Contingent Valuation.

Bakken, S. (1995). Conflicts between natural resources and structural protection. DR

Weise and RE Martin (Technical Coordinators). the Biswell Symposium: Fire Issues

and Solutions in Urban Interface and Wildland Ecosystems. General Technical

Report PSW-GTR-158, US Department of Agriculture, Forest Service, Pacific

Southwest Research Station, Albany, CA, 105-107.

Barnett, J., & Breakwell, G. M. (2001). Risk perception and experience: hazard

personality profiles and individual differences. Risk Analysis, 21(1), 171-178.

Barr, S. (2007). Factors influencing environmental attitudes and behaviors. Environment

and Behavior, 39(4), 435.

Batty, M. (2008). The Size, Scale, and Shape of Cities. Science, 319(5864), 769-771.

doi:10.1126/science.1151419

Best, S. J., Krueger, B., Hubbard, C., & Smith, A. (2001). An assessment of the

generalizability of Internet surveys. Social Science Computer Review, 19(2), 131.

Ordinance No. 2008-117, Ordinance U.S.C. 7.33.010 (2008).

Bond, W. J., & Keeley, J. E. (2005). Fire as a global ‘herbivore’: the ecology and

evolution of flammable ecosystems. Trends in Ecology & Evolution, 20(7), 387-394.

doi:DOI: 10.1016/j.tree.2005.04.025

Bonney, R., Cooper, C. B., Dickinson, J., Kelling, S., Phillips, T., Rosenberg, K. V., &

Shirk, J. (2009). Citizen science: a developing tool for expanding science knowledge

and scientific literacy. Bioscience, 59(11), 977-984.

Bowers, J. E., Bean, T. M., & Turner, R. M. (2006). Two Decades of Change in

Distribution of Exotic Plants at the Desert Laboratory, Tucson,

Arizona. Madrono, 53(3), 252-263.

Bremner, A., & Park, K. (2007). Public attitudes to the management of invasive non-

native species in Scotland. Biological Conservation, 139(3-4), 306-314. doi:DOI:

10.1016/j.biocon.2007.07.005


Brenner, J. (2011). Pasture Conversion, Private Ranchers, and the Invasive Exotic

Buffelgrass (Pennisetum ciliare) in Mexico's Sonoran Desert. Annals of the

Association of American Geographers, 101(1), 84-106.

Brenner, J. C. (2010). What Drives the Conversion of Native Rangeland to Buffelgrass

(Pennisetum ciliare) Pasture in Mexico’s Sonoran Desert?: The Social Dimensions of

a Biological Invasion. Human Ecology, 38(4), 495-505. Retrieved from

http://www.springerlink.com/content/022408v3225g7778/

Brigham, L., & Betancourt, J. (2010). Collaboration in Adaptive Management of

Buffelgrass Invasion and Its Impacts Retrieved from

http://www.adaptivemanagement.net/sites/default/files/Southern%20Arizona%20Buff

elgrass%20Coordination%20Center_Lindy%20Brigham_Julio%20Betancourt.pdf

Brigham, L. (2010). Mapping Buffelgrass for Strategic Planning and Funding. Tucson,

Arizona: Southern Arizona Buffelgrass Coordination Center. Retrieved

from http://www.governor.state.az.us/fhc/documents/PPT/2010/MappingBuffelgrass_

FHC011410.pdf

Brooks, M. L., D'Antonio, C. M., Richardson, D. M., Grace, J. B., Keeley, J. E.,

DiTomaso, J. M., . . . Pyke, D. (2004). Effects of invasive alien plants on fire

regimes. Bioscience, 54(7), 677-688.

Bryson, J. M. (2004). What to do when stakeholders matter: A guide to stakeholder

identification and analysis techniques. Public Management Review, 6(1), 21-53.

Buffelgrass Information Center. (2007). Buffelgrass Summit. Retrieved 04/10, 2010,

from http://www.buffelgrass.org/summit.php

Buffelgrass Information Center. (2009). In the News: Stories and Interviews. Retrieved

4/20/2010, 2010, from http://www.buffelgrass.org/stories.php

Buijs, A. E., Pedroli, B., & Luginbühl, Y. (2006). From hiking through farmland to

farming in a leisure landscape: changing social perceptions of the European

landscape. Landscape Ecology, 21(3), 375-389.

Bureau of Economic Analysis. (2010). Regional Economic Accounts. Retrieved

4/30/2010, 2010, from http://www.bea.gov/regional/index.htm

Burgess, T. L., Bowers, J. C., & Turner, R. M. (1991). Exotic Plants at the Desert

Laboratory, Tucson, Arizona. Madrono, 38(2), 96-114.

http://www.springerlink.com/content/022408v3225g7778/

http://www.adaptivemanagement.net/sites/default/files/Southern%20Arizona%20Buffelgrass%20Coordination%20Center_Lindy%20Brigham_Julio%20Betancourt.pdf

http://www.adaptivemanagement.net/sites/default/files/Southern%20Arizona%20Buffelgrass%20Coordination%20Center_Lindy%20Brigham_Julio%20Betancourt.pdf

http://www.governor.state.az.us/fhc/documents/PPT/2010/MappingBuffelgrass_FHC011410.pdf

http://www.governor.state.az.us/fhc/documents/PPT/2010/MappingBuffelgrass_FHC011410.pdf

http://www.buffelgrass.org/summit.php

http://www.buffelgrass.org/stories.php

http://www.bea.gov/regional/index.htm


Burton, I., Kates, R. W., & White, G. F. (1968). The human ecology of extreme

geophysical events. (White Paper Toronto: Dept. of Geography, University of

Toronto. Retrieved

from http://www.colorado.edu/ibs/hazards/publications/wp/wp1.pdf

Carson, R. T., & Mitchell, R. C. (1989). Using surveys to value public goods: The

contingent valuation method. Resources for the Future, Washington DC, 82

Carson, R. T., Mitchell, R. C., Hanemann, W. M., Kopp, R. J., Presser, S., & Ruud, P. A.

(1992). A contingent valuation study of lost passive use values resulting from the

Exxon Valdez oil spill. MPRA Paper,

Carson, R. T., Mitchell, R. C., Hanemann, M., Kopp, R. J., Presser, S., & Ruud, P. A.

(2003). Contingent Valuation and Lost Passive Use: Damages from the Exxon Valdez

Oil Spill. Environmental and Resource Economics, 25(3), 257-286.

doi:10.1023/A:1024486702104

Chambers, N., & Hawkins, T. O. (2002). Invasive plants of the Sonoran Desert, a field

guide. Tucson, Arizona: Sonoran Institute, Environmental Education Exchange,

National Fish and Wildlife Foundation. Retrieved from

http://archive.sonoran.org/programs/pdfs/invasive_plants_en.pdf

Cihacek, L. J., & Macha, D.Carbon Sequestration and Storage in Select Grass

Monocultures. Fargo: North Dakota State University Department of Natural Resource

Sciences.

Clark, W. C., & Dickson, N. W. (2003). Sustainability Science: The Emerging Research

Program. Proceedings of the National Academy of Sciences, 100(14), 8059-8061.

Clenton E. Owensby, Coyne, P. I., Ham, J. M., Auen, L. M., & Knapp, A. K. (1993).

Biomass Production in a Tallgrass Prairie Ecosystem Exposed to Ambient and

Elevated CO"2. Ecological Applications, 3(4), 644-653.

Clout, M. N., & Williams, P. A. (2009). Invasive species management: a handbook of

principles and techniques Oxford University Press, USA.

Cohn, J. P. (2008). Citizen science: Can volunteers do real research? Bioscience, 58(3),

192-197.

Collins, J. P., & Kinzig, A. (2000). A New Urban Ecology. American Scientist, 88(5),

416.

http://www.colorado.edu/ibs/hazards/publications/wp/wp1.pdf

http://archive.sonoran.org/programs/pdfs/invasive_plants_en.pdf


Coombs, D. (1972). The Club Looks at Itself. Sierra Club Bulletin, 57(7), 35-39.

Corral-Verdugo, V., Bechtel, R. B., & Fraijo-Sing, B. (2003). Environmental beliefs and

water conservation: An empirical study. Journal of Environmental Psychology, 23(3),

247-257.

Costanza, R., d'Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., . . . van, d. B.

(1997). The value of the world's ecosystem services and natural

capital. Nature, 387(6630), 253-260.

Cottrell, S. P. (2003). Influence of sociodemographics and environmental attitudes on

general responsible environmental behavior among recreational boaters. Environment


Cox, J. R. (1992). Maria Wilman: An Outstanding Contributor to Rangeland

Improvement. Rangelands, 14(5), 276-278.

Cummings, R. G., Brookshire, D. S., Schulze, W. D., Bishop, R. C., & Arrow, K. J.

(1986). Valuing environmental goods: an assessment of the contingent valuation

method Rowman & Littlefield Pub Inc.

Damigos, D., & Anyfantis, F. (2011). The value of view through the eyes of real estate

experts: A Fuzzy Delphi Approach. Landscape and Urban Planning, 101(2), 171-

178.

Davis, M. (1999). Ecology of Fear: Los Angeles and the Imagination of Disaster. New

York: Vintage Books.

Davis, M. (1995). The Case for Letting Malibu Burn. Environmental History

Review, 19(2), 1-36.

Dearien, C., Rudzitis, G., & Hintz, J. (2005). Chapter 8: The role of wilderness and

public land amenities in explaining migration and rural development in the American

Northwest. In G. Green (Ed.), Amenities and rural development: theory, methods, and

public policy: based on a conference held in Madison, Wis. in the summer of

2004 (pp. 113-129). Madison, WI: Elgar.

Defee, J. F., Schultz, J. H., & Pasewark, R. A. (1974). Occupational level and

organizational membership. Journal of Leisure Research, 6(1), 20-26.


Delaney, D. G., Sperling, C. D., Adams, C. S., & Leung, B. (2008). Marine invasive

species: validation of citizen science and implications for national monitoring

networks. Biological Invasions, 10(1), 117-128.

Devall, W. B. (1970). Conservation: An Upper-middle Class Social Movement: A

Replication. Journal of Leisure Research,

Dillman, D. A. (1978). Mail and telephone surveys :the total design method. New York: J

Wiley & Sons.

DiTomaso, J. M. (2000). Invasive weeds in rangelands: species, impacts, and

management. Weed Science, 48(2), 255-265.

Downing, R. (2006). Bad Grass: What could be a more serious threat to the Sonoran

Desert than development? Retrieved 4/29/2010, 2010,

from http://www.tucsonweekly.com/tucson/bad-grass/Content?oid=1083667

Duany, A., Plater-Zyberk, E., & Speck, J. (2001). Suburban nation: The rise of sprawl

and the decline of the American dream. New York: North Point Pr.

Dunlap, R. E., Van Liere, K. D., Mertig, A. G., & Jones, R. E. (2000). New trends in

measuring environmental attitudes: measuring endorsement of the new ecological

paradigm: a revised NEP scale. Journal of Social Issues, 56(3), 425-442.

Editors of the American Heritage Dictionaries. (2006). The American Heritage

Dictionary of the English language (4th ed.) Houghton Mifflin. Retrieved

from http://books.google.com/books?id=uPCFIQAACAAJ

Felker, P., & Clark, P. R. (1981). Nodulation and nitrogen fixation (acetylene reduction)

in desert ironwood (Olneya tesota). Oecologia, 48(2), 292-293.

Fink, A. (2003). The survey kit (2nd ed.). Thousand Oaks, Calif.: Sage Publications.

Fischer, A., & van der Wal, R. (2007). Invasive plant suppresses charismatic seabird-the

construction of attitudes towards biodiversity management options. Biological

Conservation, 135(2), 256-267.

Fischer, A., & Young, J. C. (2007). Understanding mental constructs of biodiversity:

Implications for biodiversity management and conservation. Biological

Conservation, 136(2), 271-282.

Foresman, T. W., Pickett, S. T. A., & Zipperer, W. C. (1997). Methods for spatial and

temporal land use and land cover assessment for urban ecosystems and application in

http://www.tucsonweekly.com/tucson/bad-grass/Content?oid=1083667

http://books.google.com/books?id=uPCFIQAACAAJ


the greater Baltimore-Chesapeake region. Urban Ecosystems, 1(4), 201.

doi:10.1023/A:1018583729727

Franco, A. C., & Nobel, P. S. (1989). Effect of nurse plants on the microhabitat and

growth of cacti. Journal of Ecology, 77(3), 870-886.

Franklin, K. A., Lyons, K., Nagler, P. L., Lampkin, D., Glenn, E. P., Molina-Freaner, F., .

. . Huete, A. R. (2006). Buffelgrass (Pennisetum ciliare) land conversion and

productivity in the plains of Sonora, Mexico. Biological Conservation, 127(1), 62-71.

Franklin, K. A., Lyons, K., Nagler, P. L., Lampkin, D., Glenn, E. P., Molina-Freaner, F.,

Huete, A. R. (2006). Buffelgrass (Pennisetum ciliare) land conversion and

productivity in the plains of Sonora, Mexico. Biological Conservation, 127(1), 62-71.

Fransson, N., & Gärling, T. (1999). Environmental concern: conceptual definitions,

measurement methods, and research findings. Journal of Environmental

Psychology, 19(4), 369-382. doi:DOI: 10.1006/jevp.1999.0141

Fraser, A., & Zealand, N. (2006). Public attitudes to pest control: a literature

review Science & Technical Pub., Dept. of Conservation.

Freudenburg, W. R., Frickel, S., & Gramling, R. (1995). Beyond the Nature/Society

Divide: Learning to Think about a Mountain. Sociological Forum, 10(3), 361-392.

Friedel, M., Grice, T., Marshall, N., & van Klinken, R. D. (2008). The costs and benefits

of buffel grass and its management: how are they valued? A Climate of Change in the

Rangelands. 15th Biennial Conference, 132–135.

Losing Ground: The War On Buffelgrass In The Sonoran Desert: Subcommittee on

National Parks, Forests and Public Lands, Arizona House Natural Resources

Committee, (2010). Retrieved from

http://www.buffelgrass.org/sites/default/files/Cong_Field_Hearings_041010_Written

_V2.pdf

García-Llorente, M., Martín-López, B., González, J. A., Alcorlo, P., & Montes, C.

(2008). Social perceptions of the impacts and benefits of invasive alien species:

Implications for management. Biological Conservation, 141(12), 2969-2983.

Gaston, K. J., & Fuller, R. A. (2008). Commonness, population depletion and

conservation biology. Trends in Ecology & Evolution, 23(1), 14-19.

http://www.buffelgrass.org/sites/default/files/Cong_Field_Hearings_041010_Written_V2.pdf

http://www.buffelgrass.org/sites/default/files/Cong_Field_Hearings_041010_Written_V2.pdf


Gower, J. C. (1971). A General Coefficient of Similarity and Some of Its

Properties. Biometrics, 27(4), pp. 857-871. Retrieved

from http://www.jstor.org/stable/2528823

Grace, J. B., Smith, M. D., Grace, S. L., Collins, S. L., & Stohlgren, T. J. (2000).

Interactions between fire and invasive plants in temperate grasslands of North

America. Proceedings of the Invasive Species Workshop: The Role of Fire in the

Control and Spread of Invasive Species. Fire Conference, 40–65.

Graham, J., Simpson, A., Crall, A., Jarnevich, C., Newman, G., & Stohlgren, T. J. (2008).

Vision of a cyberinfrastructure for nonnative, invasive species

management. Bioscience, 58(3), 263-268.

Grimm, N. B., Faeth, S. H., Golubiewski, N. E., Redman, C. L., Wu, J., Bai, X., &

Briggs, J. M. (2008). Global change and the ecology of cities. Science, 319(5864),

756.

Grimm, N. B., Grove, J. M., Pickett, S. T. A., & Redman, C. L. (2000). Integrated

approaches to long-term studies of urban ecological systems. Bioscience, 50(7), 571.

Grove, J. M., & Burch, W. R. (1997). A social ecology approach and applications of

urban ecosystem and landscape analyses: a case study of Baltimore, Maryland. Urban

Ecosystems, 1(4), 259-275.

Halpern-Felsher, B. L., Millstein, S. G., Ellen, J. M., Adler, N. E., Tschann, J. M., &

Biehl, M. (2001). The role of behavioral experience in judging risks. Health

Psychology: Official Journal of the Division of Health Psychology, American

Psychological Association, 20(2), 120.

Hanemann, W. M. (1994). Valuing the Environment Through Contingent Valuation. The

Journal of Economic Perspectives, 8(4), 19-43.

Hanselka, C. W. (1988). Buffelgrass: South Texas Wonder Grass. Rangelands, 10(6),

279-281.

Harry, J., Gale, R., & Hendee, J. (1969). Conservation: an upper-middle class social

movement. Journal of Leisure Research, 1(2), 255-261.

Hauser, A. S. (2008). Pennisetum ciliare. Retrieved 4/30/2010, 2010,

from http://www.fs.fed.us/database/feis/plants/graminoid/pencil/all.html

Hausknecht, M. (1962). The joiners. New York: Bedminster,

http://www.jstor.org/stable/2528823

http://www.fs.fed.us/database/feis/plants/graminoid/pencil/all.html


Hayden, D. (2003). Building Suburbia: Green Fields and Urban Growth, 1820-2000.

New York: Pantheon Books.

Head, L. (2007). Cultural ecology: the problematic human and the terms of

engagement. Progress in Human Geography, 31(6), 837-846.

Head, L., & Muir, P. (2004). Nativeness, invasiveness, and nation in Australian

plants. Geographical Review, 94(2), 199-217.

Heberlein, T. A. (1981). Environmental attitudes. Zeitschrift Fur Umweltpolitik, 2(3),

241-270.

Hershdorfer, M. E., Fernandez-Gimenez, M. E., & Howery, L. D. (2007). Key attributes

influence the performance of local weed management programs in the southwest

United States. Rangeland Ecology & Management, 60(3), 225-234.

Hillygus, D. S. (2006). The hard count: The political and social challenges of census

mobilization Russell Sage Foundation Publications.

Hobbie, J. E., Carpenter, S. R., Grimm, N. B., Gosz, J. R., & Seastedt, T. R. (2003). The

US Long Term Ecological Research program. Bioscience, 53(1), 21.

Home, R., Keller, C., Nagel, P., Bauer, N., & Hunziker, M. (2009). Selection criteria for

flagship species by conservation organizations. Environmental Conservation, 36(02),

139-148.

Hounslow, E. (2009). What is a charismatic plant?. Sheffield: University of Sheffield.

Retrieved from http://www.shef.ac.uk/aps/mbiolsci/l4-students-09/hounslow-

emily/what_%20is_a_charismatic_plant.doc

Hughes, H., Aakre, D. & Johnson, L. (1997). Leasing Beef Cows for a Profit. Retrieved

4/25/2010, 2010, from http://www.ag.ndsu.edu/pubs/agecon/farmmgt/ec1086w.htm

Hulme, P. E. (2006). Beyond control: wider implications for the management of

biological invasions. Journal of Applied Ecology, 43(5), 835-847.

Huntsinger, L., & Hopkinson, P. (1996). Viewpoint: sustaining rangeland landscapes: a

social and ecological process. Journal of Range Management, , 167-173.

Ichino, M., & Yaguchi, H. (2002). Generalized Minkowski metrics for mixed feature-

type data analysis. Systems, Man and Cybernetics, IEEE Transactions on, 24(4), 698-

708.

http://www.shef.ac.uk/aps/mbiolsci/l4-students-09/hounslow-emily/what_%20is_a_charismatic_plant.doc

http://www.shef.ac.uk/aps/mbiolsci/l4-students-09/hounslow-emily/what_%20is_a_charismatic_plant.doc

http://www.ag.ndsu.edu/pubs/agecon/farmmgt/ec1086w.htm


Ikeda, S. (2006). Risk analysis, the precautionary approach and stakeholder participation

in decision making in the context of emerging risks from invasive alien species. Paper

presented at the Assessment and Control of Biological Invasion Risks: International

Conference on Assessment and Control of Biological Invasion Risks, Yokohama

National University. 15-26. Retrieved from http://data.iucn.org/dbtw-

wpd/edocs/2006-061.pdf

Jackson, E. (1986). Outdoor recreation participation and attitudes to the

environment. Leisure Studies, 5(1), 1-23.

Jackson, K. T. (1987). Crabgrass frontier: The suburbanization of the United

States Oxford University Press, USA.

Jenerette, G. D., & Wu, J. (2001). Analysis and simulation of land-use change in the

central Arizona – Phoenix region, USA. Landscape Ecology, 16(7), 611.

doi:10.1023/A:1013170528551

Kaczensky, P., Blazic, M., & Gossow, H. (2004). Public attitudes towards brown bears

(Ursus arctos) in Slovenia. Biological Conservation, 118(5), 661-674. doi:DOI:

10.1016/j.biocon.2003.10.015

Kaiser, F. G., Wölfing, S., & Fuhrer, U. (1999). Environmental attitude and ecological

behaviour. Journal of Environmental Psychology, 19(1), 1-19.

Kanowski, P. J., & Williams, K. J. H. (2009). The reality of imagination: Integrating the

material and cultural values of old forests. Forest Ecology and Management, 258(4),

341-346. Retrieved from http://www.sciencedirect.com/science/article/B6T6X-

4VKMW5W-2/2/afdd8eb14f12e7648c430c571c643d3e

Kaplan, R., & Kaplan, S. (1989). The experience of nature: A psychological

perspective Cambridge Univ Pr.

Kasperson, R. E., Renn, O., Slovic, P., Brown, H. S., Emel, J., Goble, R., . . . Ratick, S.

(1988). The social amplification of risk: A conceptual framework. Risk Analysis, 8(2),

177-187.

Kearney, A. R., Bradley, G. A., Petrich, C. H., Kaplan, R., Kaplan, S., & Simpson-

Colebank, D. (2008). Public perception as support for scenic quality regulation in a

nationally treasured landscape. Landscape and Urban Planning, 87(2), 117-128.

http://data.iucn.org/dbtw-wpd/edocs/2006-061.pdf

http://data.iucn.org/dbtw-wpd/edocs/2006-061.pdf

http://www.sciencedirect.com/science/article/B6T6X-4VKMW5W-2/2/afdd8eb14f12e7648c430c571c643d3e

http://www.sciencedirect.com/science/article/B6T6X-4VKMW5W-2/2/afdd8eb14f12e7648c430c571c643d3e


Keller, R. P., Lodge, D. M., & Finnoff, D. C. (2007). Risk assessment for invasive

species produces net bioeconomic benefits. Proceedings of the National Academy of

Sciences, 104(1), 203.

Kelly, J. R. (1980). Outdoor recreation participation: a comparative analysis. Leisure

Sciences, 3(2), 129-154.

Kennedy, A. M., & Wilson, P. N. (2009). Reduced tillage systems as economical dust

mitigation strategies. Journal of Soil and Water Conservation, 64(1), 61-69.

doi:10.2489/jswc.64.1.61

Koné, D., & Mullet, E. (1994). Societal risk perception and media coverage. Risk

Analysis, 14(1), 21-24.

Kreutz, D.01/). It's a threat, so help to buffalo buffelgrass. Retrieved

from http://azstarnet.com/news/local/article_e3ea27d5-90c4-5e33-a940-

9e1c254c2d1c.html

Larson, L., Whiting, J., & Green, G. (2011). Exploring the influence of outdoor

recreation participation on pro-environmental behaviour in a demographically diverse

population. Local Environment, 16(1), 67-86.

doi:doi:10.1080/13549839.2010.548373"

Lee, J. H., Scott, D., & Floyd, M. F. (2001). Structural inequalities in outdoor recreation

participation: a multiple hierarchy stratification perspective. Journal of Leisure

Research, 33(4), 427-449.

Lorimer, J. (2007). Nonhuman charisma. Environment and Planning D, 25(5), 911.

Lorr, M. (1983). Cluster analysis for social scientists Jossey-Bass San Francisco.

Luck, M. A., Jenerette, G. D., Wu, J., & Grimm, N. B. (2001). The urban funnel model

and the spatially heterogeneous ecological footprint. Ecosystems, 4(8), 782-796.

Luck, M., & Wu, J. (2002). A gradient analysis of urban landscape pattern: a case study

from the Phoenix metropolitan region, Arizona, USA. Landscape Ecology, 17(4),

327. doi:10.1023/A:1020512723753

Lyons, K. G., Maldonado-Leal, B. G., Owen, G., Espinosa-García, F., Hubbard, T., Van

Devender, T. R., & Harper-Lore, B. (2009). Community and ecosystem impacts of

the non-indigenous C4 grass Pennisetum ciliare (Buffelgrass) in the Plains of the

http://azstarnet.com/news/local/article_e3ea27d5-90c4-5e33-a940-9e1c254c2d1c.html

http://azstarnet.com/news/local/article_e3ea27d5-90c4-5e33-a940-9e1c254c2d1c.html


Sonoran desert, Sonora, Mexico. Invasive Plants on the Move: Controlling them in

North America (). Tucson, AZ: Arizona-Sonora Desert Museum Press.

Maguire, L. A. (2004). What can decision analysis do for invasive species

management? Risk Analysis, 24(4), 859-868.

Marshall, N. A., Friedel, M., van Klinken, R. D., & Grice, A. C. (2010). Considering the

social dimension of invasive species: the case of buffel grass. Environmental Science

& Policy,

Mayer, F. S., & Frantz, C. M. P. (2004). The connectedness to nature scale: A measure of

individuals' feeling in community with nature. Journal of Environmental

Psychology, 24(4), 503-515.

McDaniels, T., Axelrod, L. J., & Slovic, P. (1995). Characterizing perception of

ecological risk. Risk Analysis, 15(5), 575-588.

McDonnell, M. J., & Pickett, S. T. A. (1990). Ecosystem Structure and Function along

Urban-Rural Gradients: An Unexploited Opportunity for Ecology. Ecology, 71(4),

1232-1237.

McDonnell, M. J., Pickett, S. T. A., Groffman, P., Bohlen, P., Pouyat, R. V., Zipperer, W.

C., . . . Medley, K. (1997). Ecosystem processes along an urban-to-rural

gradient. Urban Ecosystems, 1(1), 21. doi:10.1023/A:1014359024275

McIntosh, C. R., Shogren, J. F., & Finnoff, D. C. (2010). Invasive species and delaying

the inevitable: Valuation evidence from a national survey. Ecological

Economics, 69(3), 632-640. doi:10.1016/j.ecolecon.2009.09.014

McIntosh, M. E. (2003). Review: Giants of the Desert: And the Thorn Scrub, and the

Semi-Arid Woodlands, and the Tropical Dry Forest, and. . . Ecology, 84(7), pp. 1937-

1939. Retrieved from http://www.jstor.org/stable/3450011

McIntyre, N. E., Knowles-Yanez, K., & Hope, D. (2000). Urban ecology as an

interdisciplinary field: differences in the use of “urban” between the social and

natural sciences. Urban Ecosystems, 4(1), 5-24.

McIntyre, S., & Hobbs, R. (1999). A Framework for Conceptualizing Human Effects on

Landscapes and Its Relevance to Management and Research Models. Conservation

Biology, 13(6), 1282-1292.

http://www.jstor.org/stable/3450011


McKinney, M. L. (2002). Urbanization, Biodiversity, and

Conservation Bioscience, 52(10), 883. doi:10.1641/0006-

3568(2002)052[0883:UBAC]2.0.CO;2

McKinney, M. L. (2006). Urbanization as a major cause of biotic

homogenization. Biological Conservation, 127(3), 247-260. doi:DOI:

10.1016/j.biocon.2005.09.005

McLemore, M. (2010). The Battle Against Buffelgrass. Tucson, AZ: Arizona Public

Media. Retrieved from http://radio.azpm.org/azspotlight/podcasts/2010/8/27/830-the-

battle-against-buffelgrass/

McNeely, J. A. (2001). The great reshuffling: human dimensions of invasive alien

species World Conservation Union.

Meinhold, J. L., & Malkus, A. J. (2005). Adolescent environmental

behaviors. Environment and Behavior, 37(4), 511.

Metropolitan Tucson Convention & Visitors Bureau. (2009). Tucson Facts. Retrieved

4/30/2010, 2010, from http://www.visittucson.org/media/research/facts/

Michalak, J., & Lerner, J. (2007). Linking conservation and land use planning. Defenders

of Wildlife,

Milbrath, L. W. (1984). Environmentalists, vanguard for a new society State Univ of

New York Pr.

Mojena, R. (1977). Hierarchical grouping methods and stopping rules: An

evaluation. The Computer Journal, 20(4), 359.

Mojena, R., & Wishart, D. (1980). Stopping rules for Ward's clustering method. Paper

presented at the COMPSTAT 1980: Proceedings in Computational Statistics, 4th

Symposium Held at Edinburgh 1980, 426.

Morales-Romero, D., & Molina-Freaner, F. (2008). Influence of buffelgrass pasture

conversion on the regeneration and reproduction of the columnar cactus, Pachycereus

pecten-aboriginum, in northwestern Mexico. Journal of Arid Environments, 72(3),

228-237.

Nabhan, G. P., & Carr, J. L. (1994). Ironwood: an ecological and cultural keystone of the

Sonoran Desert. Washington DC: Conservation International 92p.ISBN, 1881173070

http://radio.azpm.org/azspotlight/podcasts/2010/8/27/830-the-battle-against-buffelgrass/

http://radio.azpm.org/azspotlight/podcasts/2010/8/27/830-the-battle-against-buffelgrass/

http://www.visittucson.org/media/research/facts/


Nelkin, D. (1989). Communicating technological risk: The social construction of risk

perception. Annual Review of Public Health, 10(1), 95-113.

Norgaard, K. M. (2007). The Politics of Invasive Weed Management: Gender, Race, and

Risk Perception in Rural California. Rural Sociology, 72(3), 450-477.

Nunes, P. A. L. D., & van den Bergh, J. C. J. M. (2004). Can People Value Protection

against Invasive Marine Species? Evidence from a Joint TC–CV Survey in the

Netherlands Environmental & Resource Economics, 28(4), 517 <last_page> 532.

doi:10.1023/B:EARE.0000036777.83060.b6

O’Brien, R. M. (2007). A caution regarding rules of thumb for variance inflation

factors. Quality and Quantity, 41(5), 673-690.

Olsson, A., Orr, B., Marsh, S., & Guertin, D. P. (2009). Development of a Spatial

Decision Support System for Buffelgrass Management for the Tucson Mountain

Sector of Pima County. Tucson: Arizona Remote Sensing Center, Office of Arid

Lands Studies, University of Arizona.

Olver, C. H., Shuter, B. J., & Minns, C. K. (1995). Toward a definition of conservation

principles for fisheries management. Canadian Journal of Fisheries and Aquatic

Sciences, 52(7), 1584-1594.

Perrings, C., Williamson, M., Barbier, E. B., Delfino, D., Dalmazzone, S., Shogren, J., . .

. Watkinson, A. (2002). Biological invasion risks and the public good: an economic

perspective. Conservation Ecology, 6(1), 1.

Peterson, M. N., Hull, V., Mertig, A., & Liu, J. (2008). Evaluating household-level

relationships between environmental views and outdoor recreation: The Teton Valley

Case. Leisure Sciences, 30(4), 293-305.

Petrich, C. (1984). EIA scoping for aesthetics: Hindsight from the Green County nuclear

power plant EIS. Improving Impact Assessment: Increasing the Relevance and

Utilization of Scientific and Technical Information, Hart, SL, GA Enk, and WF

Horrick (Eds.).Westview Press, Boulder, CO, , 57-92.

Pickett, S. T. A., Burch, W. R., Dalton, S. E., Foresman, T. W., Grove, J. M., &

Rowntree, R. (1997). A conceptual framework for the study of human ecosystems in

urban areas. Urban Ecosystems, 1(4), 185-199.


Pickett, S., Cadenasso, M., Grove, J., Groffman, P., Band, L., Boone, C., . . . Wilson, M.

(2008). Beyond Urban Legends: An Emerging Framework of Urban Ecology, as

Illustrated by the Baltimore Ecosystem Study. Bioscience, 58(2), 139.

Pickett, S. T. A., Cadenasso, M. L., Grove, J. M., Nilon, C. H., Pouyat, R. V., Zipperer,

W. C., & Costanza, R. (2001). Urban Ecological Systems: Linking Terrestrial

Ecological, Physical, and Socioeconomic Components of Metropolitan Areas. Annual

Review of Ecology and Systematics, 32(1), 127-157.

Pima County. (2011). Pima County GIS Data FTP Server., 2010, from ftpgis.pima.gov

Pima County Administrator. (2009). Pima County Administrator's Recommended Budget

09/10. Retrieved 4/12/2010, 2010,

from http://www.pima.gov/finance/HTML/RecBdgt/2009-2010/Index.htm

Pima County Assessor. Tax Year 2011 Data. Retrieved 4/30/2010, 2010,

from http://www.asr.co.pima.az.us/prevweb/ftp/ty2011.aspx

Pimentel, D., Zuniga, R., & Morrison, D. (2005). Update on the environmental and

economic costs associated with alien-invasive species in the United States. Ecological

Economics, 52(3), 273-288.

Pitt, J. L., Virtue, J. G., Feuerherdt, L. J., Preston, C., Watts, J. H., & Crossman, N. D.

(2006). Onion weed: pest or perception? Paper presented at the 15th Australian

Weeds Conference, Papers and Proceedings, Adelaide, South Australia, 24-28

September 2006: Managing Weeds in a Changing Climate. 454-457.

Radeloff, V. C., Hammer, R. B., Stewart, S. I., Fried, J. S., Holcomb, S. S., & McKeefry,

J. F. (2005). The wildland–urban interface in the United States. Ecological

Applications, 15(3), 799-805.

Radeloff, V. C., Hammer, R. B., Stewart, S. I., Fried, J. S., Holcomb, S. S., & McKeefry,

J. F. (2005). The Wildland–Urban Interface in the United States. Ecological


Redman, C. L. (1999). Human Dimensions of Ecosystem Studies Ecosystems, 2(4), 296.

doi:10.1007/s100219900079

REGAN, T. (2004). Capturing expert knowledge for threatened species assessments: a

case study using NatureServe conservation status ranks. Acta Oecologica, 26(2), 95

<last_page> 107. doi:10.1016/j.actao.2004.03.013

http://www.pima.gov/finance/HTML/RecBdgt/2009-2010/Index.htm

http://www.asr.co.pima.az.us/prevweb/ftp/ty2011.aspx


Robbins, P. (2004). Comparing invasive networks: cultural and political biographies of

invasive species. Geographical Review, 94(2), 139-156.

Rossiter, N. A., Setterfield, S. A., Douglas, M. M., & Hutley, L. B. (2003). Testing the

grass-fire cycle: alien grass invasion in the tropical savannas of northern

Australia. Diversity & Distributions, 9(3), 169-176.

Rudzitis, G., & Johnson, R. (2000). The impact of wilderness and other wildlands on

local economies and regional development trends. Paper presented at the Wilderness

Science in a Time of Change Conference, , 2 14-26.

Ryan, R. L. (2005). Exploring the effects of environmental experience on attachment to

urban natural areas. Environment and Behavior, 37(1), 3.

SABCC. (2010). SABCC: Promoting a regionally integrated approach to control

buffelgrass. Tucson, AZ: Southern Arizona Buffelgrass Coordination Center.

Retrieved from http://buffelgrass.org/sites/default/files/SABCCbrochure2011-sm.pdf

SABCC. (2011). Beat Back Buffelgrass Day 2011. Tucson, AZ: Southern Arizona

Buffelgrass Coordination Center. Retrieved

from http://www.buffelgrass.org/content/beat-back-buffelgrass-day-2011

Sands, J. P., Brennan, L. A., Hernández, F., Kuvlesky Jr, W. P., Gallagher, J. F., Ruthven

III, D. C., & Pittman III, J. E. (2009). Impacts of Buffelgrass (Pennisetum ciliare) on

a Forb Community in South Texas.

Scott, D., & Willits, F. K. (1994). Environmental attitudes and behavior. Environment


Seidl, D., & Klepeis, P. (2011). Human Dimensions of Earthworm Invasion in the

Adirondack State Park. Unpublished manuscript.

Shen, W. (2008). Effects of Urbanization-Induced Environmental Changes on Ecosystem

Functioning in the Phoenix Metropolitan Region, USA Ecosystems, 11(1), 138.

doi:10.1007/s10021-007-9085-0

Sikder, I. U., Mal‐Sarkar, S., & Mal, T. K. (2006). Knowledge‐Based Risk Assessment

Under Uncertainty for Species Invasion. Risk Analysis, 26(1), 239-252. Retrieved

from http://onlinelibrary.wiley.com/doi/10.1111/j.1539-6924.2006.00714.x/full#b63

Silvertown, J. (2009). A new dawn for citizen science. Trends in Ecology &

Evolution, 24(9), 467-471.

http://buffelgrass.org/sites/default/files/SABCCbrochure2011-sm.pdf

http://www.buffelgrass.org/content/beat-back-buffelgrass-day-2011

http://onlinelibrary.wiley.com/doi/10.1111/j.1539-6924.2006.00714.x/full#b63


Simberloff, D. (2005). The politics of assessing risk for biological invasions: the USA as

a case study. Trends in Ecology & Evolution, 20(5), 216-222.

Skogen, K. (2001). Who's Afraid of the Big, Bad Wolf? Young People's Responses to the

Conflicts Over Large Carnivores in Eastern Norway*. Rural Sociology, 66(2), 203-

226.

Slimak, M. W., & Dietz, T. (2006). Personal values, beliefs, and ecological risk

perception. Risk Analysis, 26(6), 1689-1705.

Slovic, P. (1987). Perception of risk. Science, 236(4799), 280.

Smith, S. D., Huxman, T. E., Zitzer, S. F., Charlet, T. N., Housman, D. C., Coleman, J.

S., Nowak, R. S. (2000). Elevated CO2 increases productivity and invasive species

success in an arid ecosystem. Nature, 408(6808), 79. doi:10.1038/35040544

Sonoran Desert Conservation Plan. (2009). SDCP Ranch Conservation. Retrieved

4/30/2010, 2010, from http://www.pima.gov/CMO/SDCP/Ranch.html

Starrs, P. F. (1998). Let the cowboy ride: cattle ranching in the American West .

Baltimore: Johns Hopkins University Press.

Stern, P. C. (2000). New environmental theories: Toward a coherent theory of

environmentally significant behavior. Journal of Social Issues, 56(3), 407-424.

Stern, P. C., Dietz, T., Abel, T., Guagnano, G. A., & Kalof, L. (1999). A value-belief-

norm theory of support for social movements: The case of environmentalism. Human

Ecology Review, 6(2), 81-98.

Stevens, J., & Falk, D. A. (2009). Can Buffelgrass Invasions Be Controlled in the

American Southwest? Using Invasion Ecology Theory to Understand Buffelgrass

Success and Develop Comprehensive Restoration and Management. Ecological

Restoration, 27(4), 417.

Stewart, S. I., Radeloff, V. C., Hammer, R. B., Hawbaker, T. J., & USDA, F. (2007).

Defining the wildland-urban interface. Journal of Forestry-Washington-, 105(4), 201.

Stokes, K. E., O’Neill, K. P., Montgomery, W. I., Dick, J. T. A., Maggs, C. A., &

McDonald, R. A. (2006). The importance of stakeholder engagement in invasive

species management: a cross-jurisdictional perspective in Ireland. Human

Exploitation and Biodiversity Conservation, 489-512.

http://www.pima.gov/CMO/SDCP/Ranch.html


Suzán, H., Nabhan, G. P., & Patten, D. T. (1996). The importance of Olneya tesota as a

nurse plant in the Sonoran Desert. Journal of Vegetation Science, 7(5), 635-644.

Swihart, M., & Petrich, C. (1988). Assessing the aesthetic impacts of small hydropower

development. Environmental Professional, 10, 198-210.

Thapa, B. (2010). The Mediation Effect of Outdoor Recreation Participation on

Environmental Attitude-Behavior Correspondence. The Journal of Environmental

Education, 41(3), 133-150.

The Pepper Group Diversified. (2010). The Pepper Group. Retrieved 01/01, 2009,

from http://thepepper.com/

Theobald, D. M. (2004). Placing exurban land-use change in a human modification

framework. Frontiers in Ecology and the Environment, 2(3), 139-144.

Theodori, G. L., Luloff, A. E., & Willits, F. K. (1998). The Association of Outdoor

Recreation and Environmental Concern: Reexamining the Dunlap-Heffernan

Thesis1. Rural Sociology, 63(1), 94-108. doi:10.1111/j.1549-0831.1998.tb00666.x

Tidwell, L. S., & Brunson, M. W. (2008). Volunteering to Manage Rangeland

Weeds. Rangelands, 30(4), 19-24.

Trousdale, W., & Gregory, R. (2004). Property evaluation and biodiversity conservation:

Decision support for making hard choices. Ecological Economics, 48(3), 279-291.

doi:DOI: 10.1016/j.ecolecon.2003.09.011

Tryon, R. C. (1939). Cluster analysis: correlation profile and orthometric (factor)

analysis for the isolation of unities in mind and personality Edwards brother, inc.,

lithoprinters and publishers.

US Census Bureau. (2010). Pima County QuickFacts from the US Census

Bureau. Retrieved 4/30/2010, 2010,

from http://quickfacts.census.gov/qfd/states/04/04019.html

Van Devender, T. R., Felger, R. S., & Búrquez, A. (1997). Exotic Plants in the Sonoran

Desert Region: Arizona and Sonora. California Exotic Pest Plant Council Annual

Symposium,

Van Devender, T. R., & Dimmitt, M. A. (2006). Final Report on “Conservation of

Arizona Upland Sonoran Desert Habitat. Status and Threats of Buffelgrass

http://thepepper.com/

http://quickfacts.census.gov/qfd/states/04/04019.html


(Pennisetum ciliare) in Arizona and Sonora. Project #2004-0013-003.”. No. #2004-

0013-003).Arizona-Sonora Desert Museum.

van Liere, K. D., & Dunlap, R. E. (1980). The social bases of environmental concern: A

review of hypotheses, explanations and empirical evidence. Public Opinion

Quarterly, 44(2), 181.

Vasquez-Leon, M., & Liverman, D. (2004). The political ecology of land-use change:

Affluent ranchers and destitute farmers in the Mexican municipio of Alamos. Human

Organization, 63(1), 21-33.

Wåhlberg, A., & Sjöberg, L. (2000). Risk perception and the media. Journal of Risk

Research, 3(1), 31-50.

Wikle, T. A. (1995). Geographical patterns of membership in U.S. environmental

organizations. Professional Geographer, 47(1), 41. Retrieved

from http://search.ebscohost.com/login.aspx?direct=true&db=aph&AN=9505150252

&site=ehost-live

Williams, D. G., & Baruch, Z. (2000). African grass invasion in the Americas: ecosystem

consequences and the role of ecophysiology. Biological Invasions, 2(2), 123-140.

Willis, H. H., DeKay, M. L., Morgan, M. G., Florig, H. K., & Fischbeck, P. S. (2004).

Ecological risk ranking: Development and evaluation of a method for improving

public participation in environmental decision making. Risk Analysis, 24(2), 363-378.

Retrieved from http://onlinelibrary.wiley.com/doi/10.1111/j.0272-

4332.2004.00438.x/full

Wing, T. (2010). Troop 007 "licensed to kill" buffelgrass. Retrieved 3/7/2010, 2010,

from http://www.kvoa.com/news/troop-007-licensed-to-kill-buffelgrass/

Wishart, D. (2005). Mammals.cls Dendrogram Retrieved

from http://www.clustan.com/cluster_keys.html

Wishart, D. (2006). ClustanGraphics Primer: A Guide to Cluster Analysis (4th ed.).

Edinburgh: Clustan Ltd.

Wong, A. K. C., & Chiu, D. K. Y. (2009). Synthesizing statistical knowledge from

incomplete mixed-mode data. Pattern Analysis and Machine Intelligence, IEEE

Transactions on, (6), 796-805.

http://search.ebscohost.com/login.aspx?direct=true&db=aph&AN=9505150252&site=ehost-live

http://search.ebscohost.com/login.aspx?direct=true&db=aph&AN=9505150252&site=ehost-live

http://onlinelibrary.wiley.com/doi/10.1111/j.0272-4332.2004.00438.x/full

http://onlinelibrary.wiley.com/doi/10.1111/j.0272-4332.2004.00438.x/full

http://www.kvoa.com/news/troop-007-licensed-to-kill-buffelgrass/

http://www.clustan.com/cluster_keys.html


Yetman, D., & Búrquez, A. (1994). Buffelgrass--Sonoran Desert Nightmare. Arizona

Riparian Council News, 7(1), 8-10.

Zipperer, W. C., Wu, J., Pouyat, R. V., & Pickett, S. T. A. (2000). The application of

ecological principles to urban and urbanizing landscapes. Ecological


Differential Perceptions of Buffelgrass (Pennisetum ... · species, and what factors influence this perception? The specific realms of scientific literature that can be used to answer

Documents