Differential Perceptions of Buffelgrass (Pennisetum ... · species, and what factors influence this...
Transcript of Differential Perceptions of Buffelgrass (Pennisetum ... · species, and what factors influence this...
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
Differential Perceptions of Buffelgrass Walker 3
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
Cluster analysis ................................................................................................................. 49
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
Cluster analysis ................................................................................................................. 65
VI. Conclusion ........................................................................................................................... 68
Management implications ................................................................................................. 68
Recommendations for future research .............................................................................. 70
VII. Appendix A – Survey Instrument ........................................................................................ 72
Differential Perceptions of Buffelgrass Walker 4
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
Differential Perceptions of Buffelgrass Walker 5
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.
Differential Perceptions of Buffelgrass Walker 6
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
Differential Perceptions of Buffelgrass Walker 7
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
Differential Perceptions of Buffelgrass Walker 8
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
Differential Perceptions of Buffelgrass Walker 9
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
Differential Perceptions of Buffelgrass Walker 10
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
Differential Perceptions of Buffelgrass Walker 11
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
Differential Perceptions of Buffelgrass Walker 12
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;
Differential Perceptions of Buffelgrass Walker 13
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
Differential Perceptions of Buffelgrass Walker 14
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.
Differential Perceptions of Buffelgrass Walker 15
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.
Differential Perceptions of Buffelgrass Walker 16
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
Differential Perceptions of Buffelgrass Walker 17
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
Differential Perceptions of Buffelgrass Walker 18
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.
Differential Perceptions of Buffelgrass Walker 19
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).
Differential Perceptions of Buffelgrass Walker 20
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
Differential Perceptions of Buffelgrass Walker 21
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).
Differential Perceptions of Buffelgrass Walker 22
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.
Differential Perceptions of Buffelgrass Walker 23
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
Identification of BG on the landscape
SNP usage
Environmental attitudes
Sonoran Desert perceptions
Favored BG invasion response
Risk perception of BG
Conservation organization participation
Reasons for moving to
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
Differential Perceptions of Buffelgrass Walker 24
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).
Differential Perceptions of Buffelgrass Walker 25
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
Differential Perceptions of Buffelgrass Walker 26
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.
Differential Perceptions of Buffelgrass Walker 27
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).
Differential Perceptions of Buffelgrass Walker 28
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
Differential Perceptions of Buffelgrass Walker 29
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
Differential Perceptions of Buffelgrass Walker 30
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
Differential Perceptions of Buffelgrass Walker 31
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
Differential Perceptions of Buffelgrass Walker 32
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
Differential Perceptions of Buffelgrass Walker 33
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
Differential Perceptions of Buffelgrass Walker 34
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
Identification of BG on the landscape
SNP usage
Environmental attitudes
Sonoran Desert perceptions
Favored BG invasion response
Risk perception of BG
Conservation organization participation
Reasons for moving to
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.
Differential Perceptions of Buffelgrass Walker 35
Figure 9a. Figure 9b.
Figure 9c. Figure 9d.
Figure 9e. Figure 9f.
Differential Perceptions of Buffelgrass Walker 36
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.
Differential Perceptions of Buffelgrass Walker 37
Figure 10a. Figure 10b.
Figure 10c. Figure 10d.
Figure 10e Figure 10f.
Differential Perceptions of Buffelgrass Walker 38
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
Differential Perceptions of Buffelgrass Walker 39
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.
Differential Perceptions of Buffelgrass Walker 40
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).
Differential Perceptions of Buffelgrass Walker 41
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.
Differential Perceptions of Buffelgrass Walker 42
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.
Differential Perceptions of Buffelgrass Walker 43
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.
Differential Perceptions of Buffelgrass Walker 44
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
Differential Perceptions of Buffelgrass Walker 45
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.
Differential Perceptions of Buffelgrass Walker 46
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
Differential Perceptions of Buffelgrass Walker 47
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 =
Differential Perceptions of Buffelgrass Walker 48
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.
Differential Perceptions of Buffelgrass Walker 49
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.
Differential Perceptions of Buffelgrass Walker 50
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…
Differential Perceptions of Buffelgrass Walker 51
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.
Differential Perceptions of Buffelgrass Walker 52
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
Differential Perceptions of Buffelgrass Walker 53
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
Differential Perceptions of Buffelgrass Walker 54
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
Differential Perceptions of Buffelgrass Walker 55
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
Differential Perceptions of Buffelgrass Walker 56
(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
Differential Perceptions of Buffelgrass Walker 57
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
Differential Perceptions of Buffelgrass Walker 58
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
Differential Perceptions of Buffelgrass Walker 59
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
Differential Perceptions of Buffelgrass Walker 60
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,
Differential Perceptions of Buffelgrass Walker 61
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
Differential Perceptions of Buffelgrass Walker 62
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
Differential Perceptions of Buffelgrass Walker 63
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
Differential Perceptions of Buffelgrass Walker 64
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,
Differential Perceptions of Buffelgrass Walker 65
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
Differential Perceptions of Buffelgrass Walker 66
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
Differential Perceptions of Buffelgrass Walker 67
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.
Differential Perceptions of Buffelgrass Walker 68
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
Differential Perceptions of Buffelgrass Walker 69
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.
Differential Perceptions of Buffelgrass Walker 70
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
Differential Perceptions of Buffelgrass Walker 71
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.
Differential Perceptions of Buffelgrass Walker 72
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
Differential Perceptions of Buffelgrass Walker 73
If you wish to participate in an interview, please write your contact information here (below). Otherwise, leave this page blank.
Differential Perceptions of Buffelgrass Walker 74
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
Differential Perceptions of Buffelgrass Walker 75
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?
Differential Perceptions of Buffelgrass Walker 76
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.
Differential Perceptions of Buffelgrass Walker 77
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
Differential Perceptions of Buffelgrass Walker 78
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.
Differential Perceptions of Buffelgrass Walker 80
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.
Differential Perceptions of Buffelgrass Walker 81
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.
Differential Perceptions of Buffelgrass Walker 82
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
Differential Perceptions of Buffelgrass Walker 83
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.
Differential Perceptions of Buffelgrass Walker 84
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
Differential Perceptions of Buffelgrass Walker 85
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-
Differential Perceptions of Buffelgrass Walker 86
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
Differential Perceptions of Buffelgrass Walker 87
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.
Differential Perceptions of Buffelgrass Walker 88
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.
Differential Perceptions of Buffelgrass Walker 89
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.
Differential Perceptions of Buffelgrass Walker 90
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
Differential Perceptions of Buffelgrass Walker 91
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%
Differential Perceptions of Buffelgrass Walker 92
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
Differential Perceptions of Buffelgrass Walker 93
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
Differential Perceptions of Buffelgrass Walker 94
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.
Differential Perceptions of Buffelgrass Walker 95
Figure 1. Profile for cluster 1 with average, minimum, and maximum values from all 5 clusters.
Figure 2. Profile for cluster 2 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
Differential Perceptions of Buffelgrass Walker 96
Figure 3. Profile for cluster 3 with average, minimum, and maximum values from all 5 clusters.
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
Differential Perceptions of Buffelgrass Walker 97
Figure 5. Profile for cluster 5 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
Differential Perceptions of Buffelgrass Walker 98
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/
Differential Perceptions of Buffelgrass Walker 99
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
Differential Perceptions of Buffelgrass Walker 100
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.
Differential Perceptions of Buffelgrass Walker 101
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.
Differential Perceptions of Buffelgrass Walker 102
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
and Behavior, 35(3), 347.
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.
Differential Perceptions of Buffelgrass Walker 103
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
Differential Perceptions of Buffelgrass Walker 104
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.
Differential Perceptions of Buffelgrass Walker 105
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,
Differential Perceptions of Buffelgrass Walker 106
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.
Differential Perceptions of Buffelgrass Walker 107
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.
Differential Perceptions of Buffelgrass Walker 108
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
Differential Perceptions of Buffelgrass Walker 109
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.
Differential Perceptions of Buffelgrass Walker 110
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
Differential Perceptions of Buffelgrass Walker 111
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.
Differential Perceptions of Buffelgrass Walker 112
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
Applications, 15(3), 799-805.
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
Differential Perceptions of Buffelgrass Walker 113
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
and Behavior, 26(2), 239.
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.
Differential Perceptions of Buffelgrass Walker 114
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.
Differential Perceptions of Buffelgrass Walker 115
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
Differential Perceptions of Buffelgrass Walker 116
(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.
Differential Perceptions of Buffelgrass Walker 117
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
Applications, 10(3), 685-688.