. I
z 1.11.:
. in.

. .2112-
Iwu~aq
v

.3; ..
1. 31’...)

L...
11: 1... . 1....

Nd.‘

 

. 1...
1, .1 .-v
it... v 1...
l1....¢v§ t...
.qV... !

 

0 .l1 1 1
.nJuuu .11 «43v .19 1.
. . . .1 .. 1. ..

 

‘13 [7. I1. .
H. 0.... a! rgflihz§
1 . 1 . .gxéfiéagéam. .
. . .113p....11..m..au...,&.m«an...1a

THESIS

This is to certify that the
dissertation entitled

A HOUSEHOLD PERSPECTIVE FOR WILDLIFE
CONSERVATION IN COUPLED HUMAN AND NATURAL

Q
SCI)?

«9 I
co
>-- :
Luca-*5
agar
99-23
4:523;

2 l

 

Ph.D

 

SYSTEMS

presented by

M. Nils Peterson

has been accepted towards fulfillment
of the requirements for the

degree in Fisheries and Wildlife

 

 

Major Professor 3 /Sign7re0?

5/0!

Date

MSU is an affirmative-action, equal-opportunity employer

o1-u-n-.—.-.-.—1—1—-.-.-.—-----.-1-.—1--»-1-1----.—-—A4

 

' PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRCIDateDue.mdd-p.1

A HOUSEHOLD PERSPECTIVE FOR WILDLIFE CONSERVATION IN
COUPLED HUMAN AND NATURAL SYSTEMS

By

MARKUS NILS PETERSON

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Department of Fisheries and Wildlife

2007

ABSTRACT

A HOUSEHOLD PERSPECTIVE FOR WILDLIFE CONSERVATION IN COUPLED
HUMAN AND NATURAL SYSTEMS

By
M. Nils Peterson

Identifying human society and the natural environment as separate systems is the
most fundamental challenge facing wildlife conservation. Human society cannot A
conserve wildlife without understanding the relationship between natural and social
processes, and then applying that understanding in the policy arena. This dissertation
focuses on assessing the role of households as a nexus in coupled human and natural
systems (CHANS). Under this goal I address 4 objectives: 1) evaluate the relationship
between global wildlife endangerment and household density in biodiversity hotspot
nations, 2) explore how households mediate the relationship between social and natural
systems using the outdoor recreation and environmental view relationship, 3) identify
socio-demographic variables explaining variation in household location decisions with
direct impacts on natural systems and wildlife conservation and consider potential
feedbacks in the model of environmental behavior, and 4) use a systems modeling
approach to predict land use change in Teton Valley (Idaho and Wyoming) by integrating
immigration, home construction, household location decisions, and different policy
scenarios.

I use linear multiple regression models to demonstrate that household density
provides a viable alternative statistical hypothesis to human population density for
explaining species endangerment at the global level (household model, r2 = 0.85;

population model, r2 = 0.84). I then suggest adopting a household perspective for

biodiversity conservation because social norms and practices render a household
approach to conservation more pragmatic than a human population perspective. I
addressed the final objectives using data from a survey of Teton Valley residents.
Results suggested environmentally oriented views relate positively to appreciative
outdoor recreation participation and negatively to non-appreciative outdoor recreation
participation for participants and their non-participating household members. Older and
highly educated immigrants with the most environmentally oriented worldviews chose to
live in natural areas (e.g., riparian zones, wetlands, critical winter range for wildlife) in
disproportionately high numbers, and required significantly more homes per person than
other groups. Length of residency was negatively related to more environmentally
oriented worldviews. Simulation results from the systems model suggest cluster
development was the most important strategy for protecting open space. Growth slowing
restrictions, or lack thereof, were the most important policies for regulating home
construction. Finally composite rankings favored high levels of cluster development and
not implementing growth slowing restrictions or a vacation home ban. This simulation
model may also be a simple and intuitive tool for other regions where conservation goals
appear to clash with the economic well being of people.

Aside from establishing the importance of household dynamics for wildlife
conservation, this research helps establish a new interdisciplinary approach for CHANS
research that focuses on homes as expressions of environmental views and behavior. The
methods may be applied in many other areas where communities struggle to meet the

needs of wildlife and humans in CHANS.

COPYI‘ight by
M. Nils Peterson
2007

ACKNOWLEDGMENTS

First, I thank the Teton Valley community for sharing their insights and life
experiences. Without their generous assistance this effort would have been impossible. I
also would like to thank my committee Lawrence Busch, Shawn Riley, and Paul
Thompson for providing guidance and support throughout my tenure at Michigan State
University.

I thank Larry and Paul in particular for filling my head with intriguing ideas that
most Fisheries and Wildlife students never experience, and I thank Shawn for his
continuing efforts to help me make those ideas accessible to applied conservationists.
I’m grateful for Angela Mertig’s guidance while developing the primary survey used in
this dissertation. My doctoral experience would not have been successful without these
advisors.

.I thank Jack Liu for providing me more opportunities than I could possibly take
advantage of. Jack brought the leading figures from several academic fields into our lab
to discuss each student’s research. He set an academic standard far higher than I
preferred or thought possible and then helped me achieve it. Jack’s ability to balance
competitive, time consuming, and onerous tasks with fun and devoted mentorship is a
testament to what academia should be.

Finally, I thank my family. Thank you, Mom and Dad, for giving me a passion for
scholarship and the social side of conservation. Special thanks to you, Shannon, for your
infectious love of wildlife, your patience with my eccentricities, your encouragement, and
your willingness to leave Texas. Thanks for making the summer in a trailer full of

spiders and encounters with wild dogs and drug runners fun. Thanks to my brothers, in

laws, and all the rest who I subject to “exciting” research stories whenever I can. I must
thank Kate Vickery in the family category because she listened to Wayne’s snoring for
two months. Thanks for your field work and wonderful insights about Teton Valley
culture.

I thank Michigan State University, the Budweiser Conservation Fellowship
program, the NASA Earth Systems Science Fellowship program, and the National
Science Foundation for financial support. I also thank the Center for Systems Integration
and Sustainability and Department of Fisheries and Wildlife Sciences at Michigan State
University for their support. I particular I thank Sherrie Lenneman, Julie Traver, and
Mary Witchell for last minute financial shuffles that kept food on the table and for

helping me graduate.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. ix

LIST OF FIGURES ........................................................................................................... xi

CHAPTER 1

INTRODUCTION ............................................................................................................... 1
Background ..................................................................................................................... 2
Objectives ..................................................................................................................... 14
Teton Valley Study Area .............................................................................................. 14

CHAPTER 2

A HOUSEHOLD PERSPECTIVE FOR BIODIVERSITY CONSERVATION .............. 19
Abstract ......................................................................................................................... 20
Introduction ................................................................................................................... 20
Methods ......................................................................................................................... 23
Results ........................................................................................................................... 25
Discussion ..................................................................................................................... 30
Conservation Implications ............................................................................................ 33

CHAPTER 3

EVALUATING HOUSEHOLD LEVEL RELATIONSHIPS BETWEEN

ENVIRONMENTAL VIEWS AND OUTDOOR RECREATION ................................... 36
Abstract ......................................................................................................................... 37
Introduction ................................................................................................................... 37
Methods ......................................................................................................................... 41
Results ........................................................................................................................... 45
Discussion ..................................................................................................................... 51
Conservation Implications ............................................................................................ 54

CHAPTER 4

HOUSEHOLD LOCATION CHOICES: IMPLICATIONS FOR BIODIVERSITY

CONSERVATION ............................................................................................................ 59
Abstract ......................................................................................................................... 60
Introduction ................................................................................................................... 61
Methods ......................................................................................................................... 63
Results ........................................................................................................................... 68
Discussion ..................................................................................................................... 71
Conservation Implications ............................................................................................ 74

vii

CHAPTER 5
A SYSTEMS MODEL FOR INTEGRATING CONSERVATION, LAND USE

CHANGE, AND HOUSEHOLD PROLIFERATION ...................................................... 81
Abstract ......................................................................................................................... 82
Introduction ................................................................................................................... 83
Basic structure of the model .......................................................................................... 88
Quantitative description of the model ........................................................................... 91
Model evaluation .......................................................................................................... 97
Simulation experiments ................................................................................................ 97
Results ........................................................................................................................... 98
Discussion ................................................................................................................... 100
Conservation Implications .......................................................................................... 102

CHAPTER 6

CONCLUSIONS ............................................................................................................. 112

LITERATURE CITED .................................................................................................... 120

APPENDIX

CODE FOR THE SYSTEMS MODEL ........................................................................... I35

viii

LIST OF TABLES

Table 2.1. Coefficient values (Coeff) for independent variables in multiple linear
regression models predicting species endangerment in hotspot countries ........................ 35

Table 3.1.Principal Components Analysis with Varimax rotation factor loadings for
outdoor recreation activities performed by: respondents who lived with others in the same
household, their associated household members, and respondents who lived alone ......... 56

Table 3.2. Zero-order and partial correlations between the New Ecological Paradigm

(N EP) score of respondents and outdoor recreation activities performed by: respondents
who lived with others in the same household, their associated household members, and
respondents who lived alone. Zero-order correlations consider only the outdoor
recreation variable and partial correlations account for: education, income, political
affiliation, gender, and age ................................................................................................ 57

Table 3.3. Comparison of mean NEP scores between respondents who share each type of
outdoor recreation participation, or non-participation, with another household member
and respondents that do not share each type of outdoor recreation participation, or non-
participation, with another household member .................................................................. 58

Table 4.1. Primary reason for home location choices of natives and immigrants moving to
natural areas, agricultural areas, and residential areas in Teton Valley ............................. 76

Table 5.1. Relationship used for the growth slowing index. Relationship between ratio of
total population to long term resident population and home construction index (home
construction index * home demand = actual homes built) .............................................. 104

Table 5.2. Comparison of model predictions of household numbers in 2000 to census
estimates of household numbers in 2000. Land use specific estimates were derived by
multiplying the number of households in Teton County by the percent of households in
that land use type estimated in the 2004 social survey (see Chapter 3) ........................... 105

Table 5.3. Differences in total open space after 30 and 100 years respectively based on
growth slowing restrictions, vacation home ban, and 5 levels of cluster development
implementation in Teton County, Idaho. Ranks reflect significantly different groups
identified using the Duncan’s post hoc comparison of means (p < 0.05). A rank of 1 is
the strategy leaving the most open space ......................................................................... 106

Table 5.4. Differences in total development allowed (i.e., number of homes) after 30 and
100 years respectively based on growth slowing restrictions, vacation home ban, and 5
levels of cluster development implementation in Teton County, Idaho. Ranks reflect
significantly different groups identified using the Duncan’s post hoc comparison of
means (p < 0.05). A rank of 1 is the strategy allowing the most home building ............ 107

ix

Table 5.5. Scores and ranks for development scenarios in Teton County over 30 and 100
year time scales. Scores represent the sum of total open space preserved ranks (derived
from Table 5.3) and total development ranks (derived from Table 5.4). A rank of 1 is the
best, and worse development scenarios receive progressively higher rankings .............. 108

LIST OF FIGURES

Figure 1.1. The conceptual causal model of environmental concern and behavior
(modified from Stern et al. 1995) ...................................................................................... 17

Figure 1.2. Map of Teton Valley Study area ..................................................................... 18

Figure 4.1. Relationship between age (9 equal intervals) and the weighted percent of
immigrants within that age level choosing to live in natural areas (hillsides, 100m riparian
buffer, and wetlands) ......................................................................................................... 77

Figure 4.2. Relationship between education level and the weighted percent of immigrants
within that education level choosing to live in natural areas (hillsides, 100m riparian
buffer, and wetlands) ......................................................................................................... 78

Figure 4.3. Relationship between NEP percentile (in 10% increments) and the weighted
percent of immigrants within that percentile range choosing to live in natural areas
(hillsides, 100m riparian buffer, and wetlands). ................................................................ 79

Figure 4.4. Relationship between household size (average number of people per
household) and the percent of respondents choosing to live in natural areas. ................... 80

Figure 5.1. Conceptual model of land use change in Teton Valley of Idaho and Wyoming
where a. is the population sub model, b. is the home construction sub model, c. is the land
conversion sub model, and d. is the policy module. Dashed lines indicate system
components that are not explicitly simulated. Plus and minus signs indicate positive and
negative relationships respectively. ................................................................................. 109

Figure 5.2. Annual population estimated for Teton County, Idaho, with pivot point
indicated by line between 1991 and 1992 ........................................................................ 110

Figure 5.3. Time series of total open space in Teton County showing the decreasing
importance of growth slowing restrictions (magnitude indicated by brackets) as cluster
development rates increase. ............................................................................................. 111

xi

CHAPTER 1

INTRODUCTION

BACKGROUND

In 2007 the Earth Day celebrations were joined by over 1,000 grassroots rallies
for climate change action. Step It Up 2007 organizers (Bill McKibben and graduate
students at the University of Vermont) declared April 14th the national day for climate
action. In many ways this expanding public awareness of global warming and climate
change issues represents a public awakening to degradation of coupled human and natural
systems (CHANS) just as the original Earth Day in 1970 represented a public awakening
to degradation in natural systems. Coupled human and natural systems are systems in
which humans and entities from their environment interact. Because humans now
dominate most malfunctioning ecosystems, the new ecology for understanding
environmental problems focuses on coupled human-natural systems (Botkin 1990,
Vitousek et al. 1997, Redman 1999, Liu et al. 2007). Humans have interacted with their
environment since the beginning of humanity, but the breadth and intensity of these
interactions exploded after the Industrial Revolution. When the catastrophic impacts of
humans on natural systems became painfully evident, the environmental movement
emerged in the 19605 and 19705 embracing ideals of preservation and biocentrism
(Mertig et al. 2002). Since that time, society has attempted to buffer critical natural
systems from human impacts by designating more than 100,000 nature reserves (IUCN
2003), but the reserves face increasing pressure from internal and cross-boundary human
activities (Dompka 1996, Kramer et al. 1997).

Climate change has demonstrated the futility of efforts to isolate natural and
human systems. Failure to understand couplings between human and natural systems has

led to many other serious conservation failures besides global warming (Liu 2001).

Expulsion of millions of “conservation refugees” from nature reserves not only gave
international conservation organizations a bad reputation for perpetrating an inexcusable
social injustice, it magnified pressure on non-protected natural lands when the refugees
were forced to abandon sustainable lifestyles (Dowie 2005). In China efforts to lure
locals out of Wolong Nature Reserve (established in 1975 to protect giant pandas

[A iluropoda melanoleucaD with free housing backfired because 2 social issues were
ignored: 1) older residents did not want to relocate, even to improved housing, for
cultural reasons, and 2) there was little open land near the free housing for the farmers
(Liu et al. 2001).

Human ability to address environmental problems stands at a crossroads. Society
can either continue down the dangerous path of “protecting” natural systems in reserves
and sacrificing the remaining land required to support CHANS, or society can recognize
the inescapable coupling of natural and social systems and work to make them more
inhabitable for humans and non-humans. The exponentially increasing impacts of
humans on natural systems and vice versa make this decision crucial (Liu et al. 2007).
Humans have changed ecosystems more in the last few decades than in any other period
of human history (Millennium Ecosystem Assessment 2005). Human activities,
including those contributing to global warming, are homogenizing our environment and
reducing biodiversity on a global scale (Chapin et al. 2000, Root et al. 2003, Root et al.
2005).

These changes in natural systems have fed back into impacts on human systems.
The incidence of natural disasters increased at an almost exponential rate due mostly to

human population growth in disaster prone areas (Liu et al. 2007), although advances in

technology probably contributed to greater reporting of those disasters. Anthropogenic
sea surface warming over the last 3 decades has contributed to increased intensity of
hurricanes (Emanuel 2005, Webster et al. 2005). The social costs associated with these
natural disasters have created huge demands for more natural resources further
exacerbating the problem (Liu et a1. 2007). Addressing the challenges associated with
anthropogenically induced extinctions, global warming, human starvation in developing
countries, wetland 1055, 1055 of entire hydrological cycles, soil loss, melting polar ice
caps, holes in the ozone layer, acid rain, desertification, deforestation, and unprecedented
global pollution requires understanding processes linking human and natural systems.
While conservation biology needs the insight of CHANS research, the movement
faces a strong legacy of research pitting humans against nature. The human population as
the problem paradigm dates back to the origins of environmental science and Malthus’s
An Essay on the Principle of Population (Malthus [1798] I970). The essay is known
today for suggesting starvation is likely in natural populations because food supplies tend
to grow arithmetically while the populations consuming them grow geometrically. The
essay is less known for justifying the starvation of poor classes on grounds it was a divine
way of curing social ills (Young 1985, Murphy 1999). In the next century the population
paradigm emerged in arguments justifying the atrocities of Social Darwinism and
eugenics (Hitler 1943, Ferry 1995). When the modern environmental movement was
born in the late 19605 and early 19705 the population as the problem paradigm was
central to arguments for forced sterilization of all Indian men with 3 or more children and
ending food aid to poor nations without population control measures in The Population

Bomb (Ehrlich 1971:151). This perspective even emerged in ethical arguments made by

Garrett Hardin (1993). Hardin’s lifeboat ethics suggested beating back the drowning
(i.e., starving) poor masses from first—world lifeboats was essential to prevent capsizing
the boat and drowning everyone. Eric Pianka’s recent series of lectures provide a
contemporary example of this paradigm (Austin 2006). Pianka called human population
growth the fundamental ecological disaster of our time and mused about the potential for
mutated strains of Ebolavirus to save biodiversity. Pianka defended himself by arguing
he does not want humans to die, but a death threat, angry email, and floods of negative
press (Austin 2006) reflect some inescapable logic: if population is the problem, then de-

population is the solution.

The landmark paper “Principles for the conservation of wild living resources”
reflects the population as the problem paradigm (Mangel et al. 1996). The first action
prescribed by 42 of conservation biology’s leaders was recognizing and acting on Ehrlich
and Holdren’s I = PAT model (I = environmental impact, P = population, A = affluence,
and T = technology: Dietz and Rosa. 1994, Chertow 2001). The I = PAT model has been

interpreted as:

Environmental impact = Population x GDP per capita x Environmental impact per unit of

per capita GDP (Graedel and Allenby 1995)

Amazingly leaders in the conservation biology field assumed acting on I = PAT meant:
Thus, the underlying and most critical aspect of any effort to conserve wild
living resources is to slow sown and eventually decrease human per capita
demand for resources. . .It is almost certain that the only practicable way to
reduced human per capita resource demand is to stabilize and then decrease

the human population (Mangel et al. 1996:340).

One is tempted to assume the population paradigm blinded leading conservation
biologists to the flawed logic in this “most critical aspect” of wildlife conservation. How
else would so many world renowned scholars assume the solution to reducing per capita
consumption was reducing population? If we have learned anything from demographic
transitions to stable or declining human populations, per capita resource demand goes up.
Indeed stable or declining natural population growth characterizes the nations with
highest per capita resource consumption (e. g., European nations, Japan, United States).

Efforts to promote a more productive focus for CHANS research must do more
than pummel the population as the problem paradigm; they must provide a viable
alternative. Households provide one such possibility. They are a compelling unit for
CHANS research because they are products of social systems both physically and
symbolically, and they are physical objects literally rooted in natural systems. In social
systems households are both physical edifices created by human society, and fundamental
social units. Households are the basic socio-economic and consumption unit globally
(Wheelock and Oughton 1996). In natural systems households are physical objects
whose number, size, location, and constitution include the impacts of human density,
affluence, consumption, and technology (York et al. 2002, Liu et al. 2003). Unlike other
forms of consumption, households are tied to one physical location, and their dynamics
are recorded by censuses in most nations. Liu et al. (2003) suggested household numbers
as a quantifiable variable indicative of per-capita consumption that should be tied to

biodiversity loss.

The first step for applying a household perspective to wildlife conservation in

CHANS is demonstrating that households are a viable alternative to the population as the

problem paradigm. Evaluating household density as an alternative explanation for
wildlife endangerment seems prudent for several reasons. First, humans have contributed
to species extinction since at least the Pliocene-Holocene megafaunal extinctions (McKee
2001), when the global human population had yet to reach half a billion. Further, urban
development is the second leading cause of endangerment in the United States, while
human numbers have no direct linkage with endangerment (Czech et al. 2000). Finally,
human population growth rates are falling globally while household growth rates are
much higher than population growth rates (Liu et al. 2003). Focusing on the growing

threat seems prudent.

Once households are established as a viable perspective for explaining
global wildlife endangerment, researchers must address how households mediate the
relationship between social and natural systems. As previously mentioned, however,
households can be viewed as a social unit and as a physical object. Research addressing
households as a social unit must address how social relationships at the household level
influence the environmental views emerging from CHANS.

Outdoor recreation provides a compelling place to start, because the relationship
between environmental views and various types of outdoor recreation has been studied
repeatedly using similar measures at the individual level (Dunlap and Heffeman 1975,
Geisler et al. 1977, Pinhey and Grimes 1979, Van Liere and Noe 1981, Jackson 1986,
Nord et al. 1998, Tarrant and Green 1999). Outdoor recreation, particularly those forms
considered to be more appreciative in content, provides a form of consumption that may
ameliorate environmental problems by promoting pro-environmental views and behaviors

(Dunlap and Heffeman 1975). Accordingly, a5 a central, growing. and dynamic part of

American culture (Cordell et al. 2002) outdoor recreation has potential to change
destructive relationships between society and the environment (Diamond 2005).

Scholars have responded to this possibility with detailed studies on the
relationship between participation in outdoor recreation and environmental beliefs,
values, attitudes, and behaviors. Most studies hypothesized a positive relationship
between environmental concern and participation in outdoor recreation activities due to
experiences with the natural environment (Dunlap and Heffeman 1975, Tarrant and
Green 1999). None of these studies, however, addressed the potential for household-
level effects. Household-level effects refer to how outdoor recreation participation of one
household member relates to the environmental views of other, potentially non-
participating, household members. For instance an avid birder may influence the
environmental views of non-birder household members by expressing his or her views,
telling stories, and bringing environmental literature and media into the home.

Research addressing household level effects of outdoor recreation on
environmental views would fall short of the most important metric, human behavior.
Environmentally oriented worldviews, attitudes, intentions, and education do not
guarantee more environmentally sensitive behavior (Stern 2000). Few studies have
addressed the ontogeny of environmental behavior. What little empirical research has
occurred only demonstrates weak correlations between environmental behavior and other
factors (Kanagy and Willits 1993, Nooney et al. 2003, Johnson et al. 2004). Past studies
also rely on a causal chain model without feedback (Figure 1.1: Stern et al. 1995, Dietz et
al. 1998, Johnson et al. 2004). The unidirectional model posits cultural context defines

broad values and worldviews, and worldviews lead, via attitudes, and intentions, to

behavior. Stern et al. (1995), however, suggest important feedbacks probably exist and
non-adjacent variables may affect each other directly.

The traditional model starts by acknowledging humans are embedded in a social
structure that influences the emergence of values and general beliefs or worldviews
(Figure 1.1: Stern et al. 1995). This social structure embodies the opportunities and
constraints associated with behavior and the perceived response to a particular behavior
(Guagnano et al. 1995). Favorable evaluations of abstract concepts (e. g., freedom and
equality) are known as values (Ajzen 2001), while worldviews reflect the set of narrative
symbols humans use to explain the nature of reality (Greeley 1993). If activated, abstract
values influence evaluations of specific entities. Research shows values regarding
religiosity, personal restraint, and liberty are related to liberal versus conservative
attitudes (Braithwaite 1998, Feather 2002). Attitudes represent evaluations of
psychological objects captured in dimensions such as good-bad, dangerous-safe, harmful-
helpful, and likable-dislikable (Eagly and Chaiken 1993, Ajzen 2001). Humans,
however, can hold multiple attitudes toward a given object (Wilson et al. 2000). hnplicit,
habitual, attitudes tend to remain even after people explicitly change an attitude. The
attitudes people endorse with specific preferences depend on whether the person is
motivated to override the implicit attitude and if the person has the cognitive capacity to
retrieve the explicit attitude. This suggests poor correlations between attitudes and
behavior may reflect multiple context dependent attitudes toward the same psychological
object (McConnell et a1. 1997).

The expectancy-value model represents the most popular conceptualization of

attitude formation and activation (F ishbein 1963, Fishbein and Ajzen 1975). This model

suggests attitudes about an object are determined by valuation of the object’s attributes
(each belief associates the object with an attribute) and the strength of the associations
between attributes and the object. Experience with the attitude object increases
accessibility (F azio 1990), but attitude strength does not necessarily affect automatic
attitude activation (Bargh et al. 1992). Automatic attitude activation can occur without
an explicit goal to make an evaluative judgment (Bargh et al. 1996). This suggests that
even if environmental attitude objects are important, infrequent interaction with them
may limit their activation in decision making. Affective (how one feels about the object)
judgments are usually more accessible than cognitive (how one thinks about the object)
judgments (Verplanken et al. 1998). So, when beliefs and feelings regarding an object
conflict, feelings oflen predominate (Lavine et al. 1998).

Most attitude research focuses on how attitudes predict behavioral intentions and
actual behavior. The theory of planned behavior (Ajzen 1991) suggests behavior reflects
intentions and perceived control over the behavior. Intentions reflect attitudes, norms,
and perceptions of control. The difference between norms and attitudes, however, often
blurs as in the case of individualistic cultures (a norm) and positive associations with
individual freedom (an attitude).

Environmental ethicists further complicate this model by suggesting a feedback
loop between behavior and worldviews. Specifically behaviors directly affecting the
environment (e. g., shooting a deer, chopping down and burning an oak tree, plowing a
field) play a role in the ontogeny of environmental worldviews, and those worldviews in
turn influence future behaviors relating to conservation (Leopold 1949, Thompson 2003).

Empirical research addressing this hypothesized feedback loop has proven difficult

10

because activities directly impacting nature (e. g., hunting) often have ambiguous impacts
and small numbers of participants (Peterson 2004).

Previous studies of environmental behavior relied primarily on large scale social
surveys that rarely measured actual behavior or behaviors with direct impact on the
environment. Accordingly, most past research relies on self reported behaviors with
indirect linkages to biodiversity conservation (e.g., signing petitions, voting, consumer
choices, donations, entertainment choices, membership in clubs: Dietz et al. 1998,
Johnson et al. 2004). Further, what people say they do often differs significantly from
what they actually do (Argyris and Schon 1978, Argyris 1992). Finally, many
environmental behaviors directly impacting biodiversity conservation (e.g., hunting,
farming) have ambiguous conservation effects and relatively few participants (Peterson
2004).

Households as physical objects on the land can be used to measure the
consequences of human behavior. Specifically decisions about household location have
direct impacts on wildlife conservation and the function of natural systems. Recent
research suggests choosing a household location represents a directly observable and
almost universal human behavior with direct impacts on biodiversity conservation (Liu et
al. 2003, Peterson et al. 2007). Choosing the type and location of one’s home is
potentially the most pervasive and direct link between human attitudes and intentions and
their physical impacts on natural systems. Further, because people tend to live in
households for extended periods of time, household location decisions provide an

opportunity to evaluate the potential for feedback from an environmental behavior to

11

environmental views. Despite these critical CHANS linkages, little, if any, research has
addressed the ontogeny of household location decisions.

Because C HANS are so complex, any knowledge gained about the role of
households in linking human and natural systems will be more useful and palatable when
presented in a systems framework. The complexity involved in even the simplest
CHANS usually surpasses the limited information processing ability of humans (Grant et
al. 1997). Humans use biases (e. g., anchoring and adjustment, representativeness
heuristic, availability heuristic) to reduce the mental effort required to understand
complex systems (Kahneman et al. 1982, Hogarth 1987), tend to think linearly, rather
than in causal nets (Domer 1980) and have difficulty recognizing imbalanced paths (e.g.,
thresholds) and feedback loops (Axelrod 1976). At the group level these biases against
systems thinking are oflen exacerbated by the forces of group think (Janis 1972, Janis and
Mann 1977) and selective memory related to multiple perspectives (Peterson and Horton
1995).

Implementing a systems thinking paradigm may seem as simple as looking at the
“big picture,” but systems theory presents an approach counterintuitive to a culture
indoctrinated with deductive reasoning, reductionism, and solving problems by breaking
them down and isolating their parts (Senge 1990, Ackoff 1999, Vennix 1999).
Accordingly, the tool of systems modeling and simulation, has been used successfully to
explore and explain complex CHANS (Liu 2001). Scholars suggest the systems
modeling approach can help encourage managers, stakeholders, and the public address
complex CHANS issues without resorting to counterproductive reductionstic approaches

to problem solving (Checkland 1981, Senge 1990, Vennix 1999, van den Belt 2004).

12

Systems modeling also carries the added benefit of facilitating generalization of case
study results (either through scaling up or modifying models to apply results in different
contexts).

The lack of information regarding the role of households in predicting species
endangerment, mediating human nature relationships, and measuring environmental
behavior represents a critical gap in wildlife conservation research. In this dissertation I
attempt to address this gap at the local and global levels. In Chapter 2 I use a global
analysis of wildlife endangerment in biodiversity hotspot nations to evaluate whether a
household perspective provides a viable alternative to the population as the problem
paradigm for wildlife conservation. Chapters 3—5 rely on a case study in Teton Valley
of Idaho and Wyoming, USA (Figure 1.2). In Chapter 3 I assess how households, as
social units, mediate the relationship between natural and human systems systems. I
evaluate how outdoor recreation of one household member influences both their
environmental views and the environmental views of non-participating household
members. In Chapter 4 I identify socio-demographic correlates for household location
choices, a behavior with direct impacts on the environment. I also evaluate the potential
for household location to promote changes in environmental views over time. Finally,
Chapter 4 measures how household location behaviors influence critical wildlife habitat.
In Chapter 5 I integrate information from the previous sections in a systems model that
simulates human immigration, house construction, and land use change under various
policy scenarios. In Chapter 6 I summarize the results of previous chapters and discuss
their implications for wildlife conservation and for today’s dominant conservation

strategies.

13

OBJECTIVES
1. Evaluate the relationship between global wildlife endangerment and household
density in biodiversity hotspot nations (Chapter 2)
2. Explore how households mediate the relationship between social and natural
systems using the outdoor recreation and environmental view relationship
(Chapter 3)
3. Identify socio-demographic variables explaining variation in household location
decisions with direct impacts on natural systems and wildlife conservation
(Chapter 4) and consider potential feedbacks in the model of environmental
behavior (Chapter 4)
4. Develop a systems model to predict land use change in Teton Valley by
integrating immigration, home construction, and household location decisions
(Chapter 5)
TETON VALLEY STUDY AREA

Chapters 3—5 utilize a case study approach in Teton Valley (Figure 1.2). I
identify the “case” in two ways: 1) as a geographic area (i.e., the Valley bounded by
mountains), and 2) as a theoretical construct used to conduct research (Ragin and Becker
1992). As a geographic area, Teton Valley includes the communities of Driggs, Victor,
Tetonia, Felt, and Alta; unincorporated areas throughout Teton County, Idaho; and the
portion of Teton County, Wyoming, west of the Teton Mountain Range (Figure 1.2).
Teton valley is in the Southern Rocky Mountain Steppe—Open Woodland—Coniferous
Forest—Alpine Meadow Province (Bailey 1995). This province is characterized by the

Rocky Mountains and intermountain depressions or “parks” with floors less than 1,800 m

14

(e. g., Teton Valley). Climate is influenced by prevailing west winds and the north-south
orientation of mountains. South and west facing slopes are usually warmer and provide
critical winter habitat for ungulates. Altitude and slope exposure play important roles in
the pronounced vegetation zonation in this region.

As a theoretical construct for research Teton Valley represents the Intermountain
West (Riebsame 1997) and the “Mormon Culture Region” (Hunter and Toney 2005).
Beaver Dick became the first permanent settler in Teton Valley in 1854 after a falling out
with Brigham Young (the leader of the Mormon Church; Teton Valley Historical
Society). Beaver Dick’s surname (Lee) and his wife’s name, Jenny, grace lakes in nearby
Teton National Park where Lee eventually worked as a guide for wealthy tourists
including Teddy Roosevelt. Beaver Dick was followed to Teton Valley by several
Mormon families who left Utah during the late 18805 when federal pressure against
polygamy was building (Teton Valley Historical Society). In 1882, the US. government
passed the Edmunds Act, officially outlawing the practice of plural marriage, and five
years later, the Edmunds-Tucker Act called for the termination of the Mormon Church
and disposal of its property by the Secretary of the Interior Department (Ostling and
Ostling 1999). The Mormon Church officially denounced polygamy with its 1890
manifesto, but immigration to Teton Valley continued. Teton Valley had almost 4,000
residents by 1920. Most settlers lived in small communities spaced eight miles apart, the
distance you could go to town to sell milk and get home for the next milking (Teton
Valley Historical Society personal communication). These Mormon communities shared
a powerful group identity rooted in a common heritage of persecution and pioneer

hardships (Ostling and Ostling 1999).

15

Teton Valley’s population decreased steadily between the Great Depression
(3,921 in 1920) and 1979 (2,351; US. Census). Expansion of a local ski resort,
immigration of laborers from Jackson, Wyoming seeking affordable housing, and
immigration of retirees, second home owners, and telecommuting professionals made
Teton County the fastest growing county in Idaho, the fourth fastest growing state, during
the 19905 (US. Census). During the 19905 population grew from 3,439 to 5,999 (74%
increase) and the number of households grew from 1,123 to 2,078 (85% increase).
During fieldwork for this dissertation population was approximately 7,200.

The post 1990 immigrants, who now make up more than half the community, are
generally more “urban”, more educated, and more secular than their predecessors (Smith
and Krannich 2000, Peterson et al. 2006b). Development accompanying the immigration
has been haphazard, and lacked a general plan. The location of individual homes, golf
courses, and sub-divisions reflects the whim of land speculators and developers rather
than community planning. Development threatens water quality, wetlands, migration
corridors from the greater Yellowstone ecosystem (75% of which are already closed), and
habitat for mule deer and elk, native trout, and waterfowl (see http://www.tetonwater.org:

Berger 2004).

16

 

Position in social structure,
Institutional constraints.
Incentive structure

I

Values

I

Worldview,
General beliefs,
Folk ecological theory

I

Specific beliefs and attitudes

1

Behavioral commitments and intentions

I

Behavior

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1.1. The conceptual causal model of environmental concern and behavior
(modified from Stern et al. 1995).

17

 

Figure 1.2. Map of Teton Valley Study area.

CHAPTER 2

A HOUSEHOLD PERSPECTIVE FOR BIODIVERSITY CONSERVATION

In collaboration with

Markus J. Peterson, Tarla R. Peterson, and Jianguo Liu

l9

ABSTRACT

Many researchers have implicated human population density in species
endangerment, but these correlative studies do not demonstrate causality. We propose
that hypotheses implicating human population density in wildlife endangerment at global
and national scales owe their public and academic currency as thoroughly to inductive
reasoning and repetition as to scientific experimentation. It follows that alternative
research hypotheses generated from the same facts should provide equally tenable results.
Household density provides such an alternative hypothesis and is growing faster than
human population density. We used linear multiple regression models to demonstrate
that household density provides a viable alternative statistical hypothesis to human
population density for explaining species endangerment (household model, r2 = 0.85;
population model, r2 = 0.84). We then suggest adopting a household perspective for
biodiversity conservation because: social norms and practices render a household
approach to conservation more pragmatic than a human population perspective, and (2)
shifting the focus toward households could facilitate movement from a human-versus-
nature ethic to a humans-situated-within-nature ethic (i.e., the land ethic). Wildlife
managers and researchers concerned about the negative influence humans have on
biodiversity should consider grounding research, theory, and policy decisions in the
dynamics of human households as an alternative to human population.
INTRODUCTION

Chapin et al. (2000) argued that human alteration of the environment has
triggered the sixth major extinction event in the history of life, and some conservation

biologists attribute every continental extinction in recorded history to anthropogenic

20

factors (Soule’ 1983, Diamond 1986, Kerr and Currie 1995). Further research on
biodiversity 1055 supports conflating human existence with human impacts. Researchers
have consistently sought and found a correlation between human population density and
species endangerment threats on national and global scales (Kerr and Currie 1995,
Forester and Machlis 1996, Kirkland and Ostfeld 1999, Czech et al. 2000, McKee et al.
2004).

Some may view these studies as bravely disregarding the taboo that prevented
serious consideration of human population growth as the central cause of environmental
problems (Hardin 199324), but the population-as-the-problem paradigm has along history
dating back at least to Malthus’ An Essay on the Principle of Population (Malthus [1798]
1970). This paradigm can be traced through Malthus’ justification of starvation among
the poor (Young 1985, Murphy 1999), the atrocities of Social Darwinism and eugenics
(Hitler 1943, Ferry 1995), arguments for forced sterilization of all Indian men with 3 or
more children in The Population Bomb (Ehrlich 19712151), and Hardin’s (1993) lifeboat
ethics. For Hardin, beating back the poor from first-world lifeboats was essential to
prevent capsizing the boat and drowning everyone. Eric Pianka’s recent series of lectures
focusing on human population growth as the fundamental ecological disaster of our time
provide a contemporary example of this paradigm (Austin 2006). While Pianka was
merely reiterating the population-as-the-problem perspective, talk radio, blogs, and
several newspapers suggested he advocated death for most humans as the only means for
saving the environment. Pianka defended himself by arguing he does not want humans to
die, but a death threat, angry email, and floods of negative press (Austin 2006) reflect

some inescapable logic: if population is the problem, then de-population is the solution.

21

Although research on relationships between human numbers and biodiversity 1055
represents a welcome departure from occasional periods of ignoring population-related
problems, it does not demonstrate causal relationships. More immediately significant to
wildlife managers, it inadvertently builds on a morally repugnant foundation that
promotes defensive responses among the general public. Assuming human population
density causes species endangerment relies solely on induction (Benton and Craib 2001).
The academic currency of this hypothesis owes more to cultural preferences in first world
countries, and repetition, than to the hypothetico—deductive (HD) scientific method
(Romesburg 1981). Without the replication and experimentation implicit to HD science,
correlations can be as meaningless as increasing intelligence tracking shoe size. Because
replicated manipulative experimentation (Sinclair 1991, Krebs 2000) on this issue would
require socially unacceptable activities such as causing species extinctions and
manipulating the number of humans living at a given time, this form of HD science
necessarily is unavailable to biologists interested in protecting biodiversity. Instead, the
strength of population claims must lie in a lack of viable alternative hypotheses.

Per-capita consumption has long been one alternative to the population
hypothesis, but its complex multi-faceted nature made statistical correlations difficult to
demonstrate. Recently, however, Liu et al. (2003) suggested household numbers as a
quantifiable variable indicative of per-capita consumption that should be tied to
biodiversity loss. Households are the basic socio-economic and consumption unit
globally (Wheelock and Oughton 1996), and not only include the impacts of human
density, but also affluence, consumption, and technology (York et al. 2002, Liu et al.

2003).

22

Evaluating household density as an alternative explanation for species
endangerment seems prudent for several reasons. Humans have contributed to species
extinction since at least the Pliocene-Holocene megafaunal extinctions (McKee 2001 ),
when the global human population had yet to reach half a billion. Evidence regarding
direct mechanisms of extinction is clear: urban development is the second leading cause
of endangerment in the United States, while human numbers have no direct linkage with
endangerment (Czech et al. 2000). Finally, Liu et al. (2003) found that annual rate of
growth in the number of households in biodiversity hotspots (3.2%), most of which were
in developing nations, was nearly double the rate in non-hotspot nations (1.7%).

If the household perspective yields a viable alternative to the population
hypothesis, then we should conduct ethical and social evaluations to determine which
perspective would most effectively ground public policy designed to ameliorate loss of
biodiversity. To demonstrate the existence and potential effectiveness of such an
alternative, we use linear multiple regression models to demonstrate that household
density—an index to consumption—provides a viable alternative statistical hypothesis to
human population density for explaining species endangerment, (2) demonstrate that
ethical and social norms render the household perspective more pragmatic than the
population perspective for conserving biodiversity, and (3) argue that shifting
conservation’s focus towards households might facilitate movement from human—versus—
nature ethics to humans-situated-in-nature ethics (e. g., a land ethic: Leopold 1949).
METHODS

We developed a linear multiple regression model to determine whether

households and human population density could be interchangeable predictors of species

23

endangerment. We replicated the model and methods described by McKee et al. (2004)
to facilitate comparisons. The model used species richness, household density, and
population density as independent variables for predicting species endangerment in
biodiversity hotspot countries (Table 2.1). We based our analysis on mammals and birds,
whose status is best known, and we used the International Union for the Conservation of
Nature and Natural Resources (IUCN) Red List data (Hilton-Taylor 2000) for
endangerment (critically endangered, endangered, and vulnerable) and World Resources
Institute (2003) data for species richness. McKee et al. (2004) also used World
Resources Institute data for species richness, although their paper refers to the United
Nations Environment Programme's World Conservation Monitoring Centre (UNEP-
WCMC) Animals of the World Database as the source (J. K. Mckee, Ohio State
University, personal communication). We used 2003 data for species richness because
breeding bird diversity estimates were improved in the more recent version. We included
68 nations containing hotspot areas, eliminating only the smallest island nations (Liu et
al. 2003). We compiled household data from the United Nations Center for Human
Settlements (2001), and population data from the US. Census Bureau. We divided all
frequency data by each nation’s area (in 106 km?) to account for size differences among
nations and log-transformed it (base 10) to meet normality assumptions (McKee (McKee
et al. 2004). We removed independent variables from the multiple regression equation if
t-values for their coefficients were not significant at P < 0.05. We chose not to use
stepwise procedures because probable collinearity between household and population

density could yield multiple regression models with highly significant F-values, but

24

insignificant t-values for the independent variables (Ott and Longnecker 2001). We
calculated all statistics using Statistica 6.1 (StatSoft, Tulsa, Oklahoma, USA).

Our remaining results rely on logical comparison of the household and population
perspectives of biodiversity conservation. All logical arguments require premises, and
we rely on 3: pragmatic approaches for biodiversity conservation must be capable of
implementation within current socio-political contexts because changing dominant social
norms is supremely difficult and beyond the scope of wildlife conservation, (2) pragmatic
approaches to future biodiversity conservation must reflect dominant social norms, and
(3) respect for negative and positive human rights is a dominant social norm. Negative
rights only require restraint (e. g., not shooting people, not burning down someone’s
house: Hospers 2005). Positive-rights require action (e.g., education, welfare: Halper
2003)

RESULTS

As is typically found, species richness predicted numbers of threatened species
best (Table 2.1: McKee et al. 2004). The t values for human household density and
population density were insignificant because neither had additional predictive power
over the other. The correlates of threatened bird and mammal species per unit area and
species richness per unit area, household density, and population density were 0.91 , 0.56,
and 0.58, respectively. Higher collinearity between population density and species
richness, as compared to household density and species richness (r = 0.50 versus 0.45),
explained why household density had a higher t value than population density. By
themselves, both variables were significant predictors of species endangerment (Table

2.1). We found no significant correlation between species endangerment and household

25

or population growth rates for 1985—2000. Household density (coefficient: 0.20) and
population density (coefficient: 0.18) contribute equally to projected increases in species
endangerment in our models.

Since households and human population density are statistically equally powerful
predictors of species endangerment, ecologists and environmental managers could
legitimately ground policy decisions in either (or both). Without experimentation, neither
households nor population should be given preferential treatment in environmental
decision making based on HD science. Although the statistical uncertainty poses a
challenge, it also offers an opportunity to identify a biodiversity conservation approach
consistent with cultural values and social identities.

These values and identities make a household approach to biodiversity
conservation less problematic than a human population approach. Dominant social
norms suggest humans—and humans alone—have intrinsic value and should not be used
as means to an end (Kant 1873). Therefore, parenthood is often considered an inalienable
human right on par with life, liberty, and the pursuit of happiness (O'Neill and Ruddick
1979, Holmes et al. 1980, Philip and Thomas 1986, Moskowitz and Jennings 1996).
Houses, however, clearly are means to an end (shelter for humans). American
jurisprudence has recognized this fact by shifting from viewing property as an inalienable
right to viewing property as a political construct (Horwitz 1992). Although shelter may
also be a basic human right, houses, second homes, and specific size, expense, or
locations for homes certainly are not.

Within an ideal libertarian state, where maximal liberty is constrained only by

interference in the liberties of others, several scenarios dictate regulation of households

26

(e.g., property), but not procreation. Libertarian defenses of property (e. g., households)
are based on respect for negative rights—those requiring restraint (e. g., not shooting
people) rather than acting (e. g., providing universal health care)—and derive from
individual freedom (Horwitz 1992, Hospers 2005). Without property rights, someone
could take the fruits of our labor, depriving us of liberty and thus enslaving us (Hospers
2005). Corporations make similar takings claims, rooted in profit loss, when restrictions
are placed on their properties (Helvarg 1994, Gunningham et al. 2003). Negative rights
perspectives, however, do not preclude regulating households. For example, the products
of one’s labor can be protected in forms of property other than houses or land, and when
houses are used, value can be preserved even if locations and sizes of houses are
regulated. Dwelling rights are part of property rights, which are culturally less
fundamental than life rights. This places fewer negative right restrictions on regulation of
households than on regulation of birth rates.

Many rights based perspectives, however, consider positive-rights (those
requiring action; e.g., education, welfare). Most industrialized societies subsidize
healthcare intended to increase human survival and quality of life. Although public
healthcare systems in these nations vary from the comprehensive care provided to
residents of Scandinavian countries to Medicare/Medicaid in the United States, all
suggest human life is a positive right (i.e., one that imposes an obligation on others or on
the state) valued by humanity. Housing programs also are socially subsidized in most
industrialized nations, indicating that both the shelter and autonomy provided by a house

are valuable positive rights. Although the difference between life and housing is less

27

clear from a positive rights perspective than from a negative rights perspective, life still
outranks housing.

Regulating net household proliferation does ultimately end in questions about
rights (e. g., does each individual have a right to a house? a large house? an opulent large
house?), but not inalienable rights (e. g., life). For example, in both Western Europe and
the United States, the rule of law protects private property, but, when necessary to
provide a public use, allows environmental statutes and regulations that remove rights
associated with ownership (Horwitz 1992, Vamer 1994). The Fifth Amendment of the
US. Constitution states that private property may be taken for “public use,” and a recent
Supreme Court ruling (Susette Kelo et al. v. City of New London, Connecticut et al., 23
June 2005) defined “use” broadly enough to include economic development plans (e. g.,
river walks, restaurants, new hotels). While this last step is controversial, it demonstrates
the flexibility of household regulation as a policy tool relative to life regulation. The
bitter divisions over when life begins in debates over abortion demonstrate the fragility of
a conservation policy rooted in the human population perspective.

Pragrnatically, household regulation fits more comfortably within currently
dominant ethical perspectives, thereby facilitating more immediate implementation of
biodiversity conservation initiatives. Zoning laws protect watersheds, beaches, and parks
throughout the world. In the United States, environmental regulations (e. g., the
Endangered Species Act of 1973) have limited household development in critical areas
and have dictated development location in many communities (Vamer 1994, Peterson et
al. 2002, Peterson et al. 2004). Although many governments are required to pay for

private property takings, they have the authority to destroy homes to create refugia for

28

imperiled species. Both national and local governments limit the number of houses per
hectare, specify who may or may not build houses, and sometimes destroy houses for the
good of the community. Governments restrict the size, shape, location, and efficiency of
houses. In addition, governments indirectly reduce household development by increasing
costs. For example, the US. Congress enacted the Coastal Barrier Resources Act of 1982
at least in part to increase the cost of building and maintaining houses in the coastal
margins of the United States. The Act simply eliminated perverse incentives for
development in the coastal margins (e. g., federally subsidized mortgages, loans, and
flood insurance).

Western law and policy currently addresses the creation, destruction, and type of
houses, but not the creation, destruction, or type of human beings. Governments rarely
have the authority to prevent humans from giving birth (nations characterized by extreme
poverty and human rights abuses (United Nations 1948) are notable exceptions), and,
indeed, have some responsibility for ensuring that such births result in a living child.
Governments may not destroy infants after they are born or dictate the size, shape, or
color of individual humans (United Nations 1948). Further, with few exceptions (e.g.,
military bases, wilderness areas), laws in developed countries do not restrict where
people can go. Rather than restricting human access to relatively pristine areas, the US.
Congress attempted to promote experience with wildlife with the National Wildlife
Refuge System Improvement Act of 1997. It acknowledged that wildlife-dependent
recreation fosters appreciation for fish and wildlife conservation, and made wildlife-
dependent recreation that was compatible with wildlife conservation the priority for

general public use of the system (Public Law 105—57—Oct. 9. 1997, See. 5).

29

Finally, household perspectives toward biodiversity conservation have greater
spatial and temporal immediacy than population perspectives. Decisions regarding
households impact wildlife conservation here and now. A population perspective places
the major challenge for biodiversity conservation spatially distant from the centers of
western science (e. g., wildlife management) and money in Europe and North America
because natural population growth is mostly occurring in developing nations. The
population perspective also makes biodiversity conservation temporally distant in both
developing and developed nations because the impacts of family planning take a
generation to impose themselves.

DISCUSSION

Hypothetico-deductive science lends equal credence to household and population
perspectives for biodiversity conservation. Given this starting point, logic suggests a
household perspective has more to offer because it respects both negative and positive
rights, and can operate within current political and legal constraints. We have painted
ourselves into an ethical comer with the population—as-the-problem paradigm. The
personal attacks (e. g., Dr. Death, Dr. Doom, an advocate of eugenics/ genocide, etc.),
derogatory newspaper articles, and public outcry resulting from Pianka’s 2006
population-as-the-problem lectures demonstrate this fact. The human population
perspective makes biodiversity-conservation advocates the enemies of humanity in a
world where conservation relies on the generosity and empathy of humans. Policy
makers and conservationists contribute to psychological isolation between humans and

other species when they frame human existence as inimical to biodiversity. Considering

30

the very existence of humans inimical to conservation pits value for human life against
value for nonhuman nature: a lose—lose scenario.

Of course, conceptualizing household dynamics as a nexus with biodiversity
conservation will not directly reduce species endangerment. The household perspective
must work indirectly by changing the way society conceptualizes its interaction with the
environment, and may be uniquely situated to do 50. Traditional environmental ethics
find ,value either in nature (e.g., deep ecology, some versions of the land ethic) or in
specific entities deemed worthy of moral standing (e. g., Kantianism, coo-feminism,
animal rights) (Light and Rolston 2003). These perspectives make humans master
consumers of utility, rights, or value provided by nature (Thompson 2003). In order for
biodiversity conservation to resonate with the general public, society requires a moral
foundation rooted in human relationships with the environment rather than simply the
consumption of utility and rights provided by the environment (Leopold 1949, Thompson
2003). Beyond overcoming the pragmatic constraints of a population perspective toward
biodiversity conservation, a household perspective may facilitate movement from this
human-versus-nature ethic to a humans-in-relation-to-nature ethic.

Because community is constructed through social relationships (Hegel 1977,
Peterson et al. 2005), relationship-based environmental ethics provide a necessary context
for creating the inclusive land community advocated by Leopold (1949). Thus, an
environmental ethic that recognizes nonhuman beings as members of the community
requires cultivation of reciprocal relationships between humans and nonhumans (i.e.,
most of biodiversity). Agrarianism represents the dominant relationship-based

environmental ethic, and means of cultivating such reciprocal relationships between

31

humans and the environment, in western culture (Montmarquet 1989, Peterson 1990,
Mariola 2005). In agrarian philosophy the relationship with nature formed through
subsistence activities (e. g., horticulture, animal husbandry, farming, and forestry)
originates value. Human articulation with the land rather than humans “contemplating
natural landscapes as if they were cuts of meat or paintings in a museum” creates value
(Thompson 2003178).

Households provide the most tangible contact with non-human dimensions of the
land community for people in modern industrialized nations. Limiting houses to protect
the habitat of endangered species places humans in a reciprocal relationship with the
environment because more houses in less space reflects crowding (more houses per
hectare) or loss of portions of the land community that no longer fit (the endangered
species). Thus, regulating household proliferation, size, density, and location legitimizes
human experience within environmental limits, and focuses on reciprocal relationships
people actually have with the land, as well as other species via their home. A household
perspective can enrich studies of extinction risk by focusing on human articulations with
nature.

Oikos, or household, is the Greek root of ecology, economics, and related terms,
and could serve as the root of a relationship based environmental ethic. Within such an
ethic, the land—house nexus enables tangible interactions with a place, thus promoting
trust, reciprocity, and connectedness with that environment. Trust, reciprocity, and
connectedness motivate social action and define community in discussions of social
capital as an alternative to privatization and command and control solutions to the

tragedy of the commons (Ostrom 1990, Pretty 2003, Peterson et al. 2006c). The

32

household approach provides a socially acceptable and politically practical perspective
for biodiversity conservation capable of fostering the land ethic. A household-level focus
for conservation also enables deconstruction of the modern environmentalist’s paradox—
people are bad for the environment, yet conservation is about people being in tight
feedback loops with nature.
CONSERVATION IMPLICATIONS

Given that HD science indicates no reason why natural resource managers should
prefer population versus household perspectives toward biodiversity conservation, ethics
and practicality should guide management decisions. Our analysis suggests that a
household perspective to biodiversity conservation will be more effective than a human
population perspective. Research assessing household dynamics over temporal and
spatial scales that match the scale of population data is needed to tease apart the influence
of household density and population density on wildlife extinction. In the meantime,
wildlife and environmental managers concerned about the influence humans have on
biodiversity conservation should ground research and policy in household dynamics as an
alternative to human population. Research should evaluate how household dynamics
(e.g., changes in multi-generational households, family size, and land tenure systems)
influence species endangerment, and clarify socio-structural determinants of those
dynamics. Household regulation via zoning already occurs at local scales in many parts
of the world, and this approach could be effectively scaled up and used in adaptive
management strategies at national and international levels. Of course, cultural and
geographic differences mandate different land tenure strategies (e.g., zoning) for different

contexts. Policy advocacy should stress reciprocal relationships between humans and

33

species endangerment (e. g., how home building influences wildlife survival) instead of
inimical relationships (e. g., human existence versus wildlife existence). This means
managers should think and talk about household dynamics as both a threat and solution to
wildlife conservation. It also means abandoning the misanthropic tendency to pit human
life against biodiversity conservation in public forums. A household approach to
biodiversity conservation positions wildlife managers as advocates for both humanity and

wildlife.

34

Table 2.1. Coefficient values (Coeff) for independent variables in multiple linear regressior
models predicting species endangerment in hotspot countries.

 

Independent variables

 

 

 

 

 

 

Species Household Population Intercept
richness density density
Coeff P Coeff P Coeff P Coeff P r23
Population versus
household”
Full model 0.79 0.00 0.47 0.06 -0.30 0.26 -0.36 0.09 0.82
Finalmodel 0.85 0.00 NA NA NA NA _0.57 000
Household density
Model 0.77 0.00 0.20 0.00 NA NA -0.53 0.00 0.85
Population density
Model 0.77 0.00 NA NA 0.18 0.00 -0.63 0.00 0.84

 

a r2 values reflect final models after we removed independent variables with P values >

0.05.

b In the population and household model, I removed household density and population

density from the final model by stepwise regression due to high autocorrelation.

35

CHAPTER 3

EVALUATING HOUSEHOLD LEVEL RELATIONSHIPS BETWEEN

ENVIRONMENTAL VIEWS AND OUTDOOR RECREATION

In collaboration with

Vanessa Hull, Angela G. Mertig, and J ianguo Liu

36

ABSTRACT

Outdoor recreation may foster positive environmental views among
participants and their household members, but no research has addressed this hypothesis
at the household-level. We address this gap with a case study evaluating both the
individual- and household-level relationship between outdoor recreation and
environmental views. Our results suggest NEP relates positively to appreciative outdoor
recreation participation and negatively to non-appreciative outdoor recreation
participation for participants and their household members. Future research should focus
on how household dynamics mediate the relationship between environmental views and
outdoor recreation.
INTRODUCTION

Current levels of human consumption of natural resources threaten the function of

most ecological systems and the biodiversity they support (Mathews and Hammond
1999). Pragmatic solutions to this problem are incredibly rare because consumption is
central to Western culture (Schnaiberg and Gould 1994)). Outdoor recreation,
particularly those forms considered to be more appreciative in content, provides a form of
consumption that may ameliorate environmental problems by promoting pro-
environmental views and behaviors (Dunlap and Heffeman 1975). Accordingly, as a
central, growing, and dynamic part of American culture (Cordell et al. 2002) outdoor
recreation has potential to change destructive relationships between society and the
environment.

Scholars have responded to this possibility with detailed studies on the

relationship between participation in outdoor recreation and environmental beliefs,

37

values, attitudes, and behaviors. They generally divide outdoor recreation activities into
three types: appreciative (e. g., hiking, camping, and bird watching), consumptive (e.g.,
fishing and hunting), and motorized (e.g., riding all terrain vehicles [ATVs]). Most
studies hypothesized a positive relationship between environmental concern and
participation in outdoor recreation activities (of any kind) due to direct and personal
experiences with the natural environment (Dunlap and Heffeman 1975, Tarrant and
Green 1999). These studies also hypothesized a stronger positive relationship between
environmental concern and appreciative outdoor recreation activities, than between
environmental concern and motorized or consumptive activities. This hypothesized
relationship assumed appreciative activities which leave the environment relatively
untouched correspond to a “preservationist” ideology (Dunlap and Heffeman 1975,
Tarrant and Green 1999).

Some studies provided evidence to support these hypotheses with weak
correlations for motorized and consumptive activities (correlation coefficients < 0.1) and
stronger correlations for appreciative activities (0.15-0.3: Dunlap and Heffeman 1975,
Van Liere and Noe 1981, Jackson 1986). Other studies did not find strong correlations
between environmental concern and any type of outdoor recreation activity (Geisler et al.
1977, Pinhey and Grimes 1979, Nord et al. 1998). Finally, Bright and Porter’s (2001)
findings suggest the meaning of wildlife-related recreation fully mediates the
participation/environmental concern relationship.

The differences among these findings could reflect different measures of
environmental concern. Geisler et al. (1977) used support for environmentally related

public action as a measure of environmental concern; Pinhey and Grimes (1979) based

38

their determination of environmental concemon two questions about land use; Nord et al.
(1998) used the problem of quality of the environment (PQENV) scale; Van Liere and
Noe (1981), Jackson (1986), and Bright and Porter (2001) used the New Environmental
Paradigm scale (a previous version of the New Ecological Paradigm scale); and Dunlap
and Heffeman (1975) used a series of questions regarding concern for specific
environmental entities. Future research using a common measure such as the New
Ecological Paradigm scale (NEP: Dunlap et al. 2000) rather than one devised for the
specific study would facilitate comparisons with both past and future studies. We use the
NEP scale in this study, not only because it is currently one of the most widely used
measures of environmental worldview, but also in the hope that our research can be more
readily replicated and that future results can be compared to ours using a consistent
measure of environmental views.

Household-level effects are probably the biggest gap in scholarship
addressing linkages between outdoor recreation and environmental views. Household-
level effects refer to how outdoor recreation participation of one household member
relates to the environmental views of other, potentially non-participating, household
members. For instance an avid birder may influence the environmental views of non-
birder household members by expressing his or her views, telling stories, and bringing
environmental literature and media into the home. Little if any previous research on the
relationship between environmental views and outdoor recreation has addressed
household-level effects. Household-level effects, however, could play a major role in the
influence of outdoor recreation on the environment because households represent a

fundamental unit in economics (Wheelock and Oughton 1996) and natural resource use

39

(Liu et al. 2003). We hypothesize environmental views may be influenced by outdoor
recreation at the household-level in addition to the individual-level (i.e., the person
participating). This hypothesis suggests outdoor recreation’s impact on environmental
views extends beyond the participant to non-participating household members. For
instance, hiking could promote pro-environmental views for both the hiker, and the
hiker’s non-hiking household members. While this hypothesis has not been directly
addressed, studies finding household-level impacts of individual behaviors (e. g., parental
participation in work programs) on beliefs, attitudes, and behaviors of non-participating
household members are common (Huston et al. 2001). Further, worldviews (the set of
narrative symbols humans use to explain the nature of their environment) evolve largely
from interactions with parents and other family members (Greeley 1993).

In this paper we use interview data from a study conducted in
Teton Valley to test four interrelated specific hypotheses addressing individual and
household-level interactions between outdoor recreation and environmental views
(measured with the NEP): 1) environmentally oriented views are positively related to
participation in appreciative outdoor recreation, 2) environmentally oriented views are
negatively related to participation in non-appreciative outdoor recreation, 3)
environmentally oriented views of non-recreating respondents are positively related to
participation in appreciative outdoor recreation by other household members, and 4)
environmentally oriented views of non-recreating respondents are negatively related to

participation in non-appreciative outdoor recreation by other household members.

40

METHODS

We obtained data for this study from an interview survey conducted in Teton
Valley. The study area included the portion of Teton County, Wyoming, west of the
Teton Mountain Range and Teton County, Idaho. Immigration motivated largely by
outdoor recreation opportunities led to a 74% jump in population (3,439 to 5,999) and
85% (1,123 to 2,078) jump in household numbers during the 1990’s (Smith and Krannich
2000, Peterson et al. 2006a). The population reached approximately 7,200 in 2004 when
this study was conducted. The centrality of outdoor recreation and environmental
amenities, particularly as drivers of immigration, make the relationship between outdoor
recreation and environmental views of particular relevance in this study area.

We used an in-person interview protocol to assess relationships between
household and individual-level participation in outdoor recreation and environmental
views. We chose personal interviews because they promised higher response rates
(Dillman 2000). We purchased a representative sample (n = 550) of telephone listings
(which included physical addresses) from Survey Sampling, Incorporated (F airfield,
Connecticut). Logistical constraints dictated sample size. We pre-tested the
questionnaire with residents of Victor, Idaho (within the study area; n = 23), and Lansing,
Michigan (n = 18). During July—August, 2004, we visited each respondent during 4 time
intervals, morning and evening on a weekend day and on a week day. We made initial
contact via telephone when visits failed or we could not locate a physical address. An
interpreter was enlisted for Spanish interviews. Interviewers defined acronyms, but
answered other questionnaire related queries by reading directly from the questionnaire,

explaining questionnaire format, or stating “whatever it means to you” (Groves 1989).

41

We measured environmental views with the NEP scale (Dunlap et al. 2000). The
scale was designed to address five theoretical dimensions with three questions for each:
endorsement of limits to growth, anti-anthropocentrism, belief in future ecocrisis, belief
in fragile and balanced nature, and rejection of human exemptionalism (i.e., the notion
that humans are free to do as they please because they are exempt from the laws of
nature). Each item utilizes a 5 category Likert response format, ranging from “strongly
disagree” to “strongly agree.” The NEP taps a lay person’s view of human relationships
with the environment (Johnson et al. 2004: 159). Respondents embracing the views of
modern environmentalist groups, consistently score higher than other groups (Dunlap and
Van Liere 1978, Dunlap et al. 2000, Dunlap and Michelson 2002, Mertig et al. 2002).

We assessed outdoor recreation participation of individuals by asking respondents
“about how often in a typical year” do you participate in: l) bird watching, 2) hiking, 3)
camping, 4) boating 5) fishing, 6) hunting, and 7) riding off-road vehicles. We assessed
outdoor recreation participation of other household members by asking the same
question, but replacing “you” with “someone in your household (other than yourself)”.
Possible responses ranged from frequently to never (4 = frequently, 3 = sometimes, 2 =
rarely, l = never). Asking one member of a household to judge outdoor recreation
participation of other household members could create biases if participation in some
activities was systematically over or under estimated by respondents. Interviewing all
household members could identify any biases associated with asking one respondent to
report on outdoor recreation activities of other household members, but was not

logistically possible in this study.

42

We also measured several important demographic variables: we used standard
survey questions to collect data for education (1 = less than high school to 7 = graduate or
professional degree), previous year’s annual income (1 = < 14,999 to 9 = 2200,000), age,
and gender (Dillman 2000). Using an open-ended question, we asked respondents for
their political affiliation and received 6 answers: Conservative, Republican, Independent,
Democrat, Liberal, and non-voting (excluded from analysis). Follow up questions
indicated that all of the Conservatives considered themselves Republicans and all of the
Liberals considered themselves Democrats so we grouped Conservative and Republican
and Liberal and Democrat during coding of political affiliation (l = Republican, 2 =
Independent, and 3 = Democrat).

We explored the relationship between environmental views and participation in
outdoor recreation by calculating Pearson correlation coefficients for the relationship
between NEP score and frequency of participation in each of the outdoor recreation
activities. We grouped participants into individuals within multi-person households and
individuals who lived alone. The first group was analyzed with respect to both activities
of the individuals and activities of their household members. For those individuals who
identified themselves as sharing a household with others, we computed correlations
between the frequency that respondents participated in each of the activities and the
frequency that their household members participated in the activity.

As was performed by Jackson (1986) and Theodori et al (1998), we conducted a
principal components analysis in order to ascertain larger groupings for participation in
outdoor recreation activities, such as appreciative and non-appreciative outdoor

recreation activities. We performed principal components analysis with varimax rotation

43

in order to obtain orthogonal factors that accounted for the greatest proportion of the
variance. We retained all factors with eigenvalues greater than 1 for analysis. We
calculated Cronbach’s alpha to measure the consistency of the NEP scale.

To control for education, income, political affiliation, age, and gender, as
suggested by Dunlap and Heffeman (1975), Van Liere and Noe (1981), Jackson (1986),
and Theodori et al. (1998) we calculated partial correlation coefficients for the
relationship between NEP score and each of the outdoor recreation activities. Partial
correlation coefficients were obtained from regressing NEP score against each of the
outdoor recreation activities, with all demographic variables included in each of the
regression models as controls (Cohen and Cohen 1983).

We evaluated differences in: 1) the correlations between a respondent’s NEP
score and their own participation in outdoor recreation activities; and 2) the correlations
between a respondent’s NEP score and participation in outdoor recreation activities by
other household members. For this we used a modified t-test for comparing partial
correlations among dependent variables from the same sample (Cohen and Cohen 1983).
We used a Fisher’s r—to-z transformation when comparing the partial correlations
between NEP score and outdoor recreation activities across the two independent samples
of individuals who identified themselves as living alone and those belonging to multi—
person households (Cohen and Cohen 1983).

To evaluate the relationship between an individual’s environmental views and the
outdoor recreation activities of other household members, we divided the dataset into two
groups, those who did not participate in the activity and those who did participate. We

then divided each of these groups into sub-groups according to whether the activities of

44

the respondent’s household members matched or did not match their own participation in
the activity. We performed two-sample t-tests on the NEP scores of individuals who had
household members that matched their participation in the activity and individuals
without household members sharing activity participation.

In order to analyze the relationship between an individual’s environmental views
and the outdoor recreation activities of other household members while controlling for
demographic parameters, we created a dummy variable to represent whether the activities
of the respondent’s household members matched (i.e., 0) or did not match (i.e., 1) their
own activities (a value of l reflected any level of participation). We computed the partial
correlation coefficients between NEP scale and the dummy variable for each type of
participation and non-participation, while controlling for the aforementioned
demographic parameters. All statistical analyses were performed using the R package (R
Development Core Team, 2005). With the exception of NEP scale items (see below),
respondents who failed to answer a relevant question were excluded from the analysis.
RESULTS

Only 484 of the initial 550 household listings were usable (several of the
individuals were no longer at the listed address). Of the remaining households, we were
able to contact 436, 20 of which refirsed to participate in the study. Ten additional cases
were excluded due to incomplete questionnaires. The final response rate was 84% (406
of 484; sampling error :t 4.8%). Most of the respondents (n = 312) shared a household
with at least one other person, and 94 lived alone. Item non-response was 31% for age,
gender, and education, as well as for participation in the outdoor recreation activities.

Item non-response was 7% for income and 14% for political affiliation. There were 15

45

individuals who did not answer either 1 or 2 of the 15 NEP questions; therefore we
substituted mean values for these item non-responses.

Our sample matched US. Census data for the study area, with 46% of
respondents being female, 90% Anglo, and 6% Hispanic. Because 90% of respondents
were Anglo, ethnicity was not used in other analyses. The median yearly household
income fell within the $3 5,000—849,999 range, the majority of the population (90%) had
annual family incomes below $100,000 and only 6.5% of respondents had annual family
incomes below $15,000. Nearly 40% of respondents had 4-year college degrees or
higher, 30% had some form of vocational training, 25% completed high school, and 5%
did not complete high school. Mean age of respondents was 46. With regard to political
affiliation, 41% of the respondents identified themselves as being Independent, while
34% of respondents were Republicans and 25% were Democrats.

Cronbach’s alpha for the NEP scale items was 0.87, reflecting a high degree of
consistency. As was the case for Dunlap et al. (2000), principal components analysis
revealed more than one dimension (Table 3.1). Despite this, the scale’s authors (Dunlap
et al., 2000) strongly suggest using the NEP as a single measure; hence we used the NEP
scale as a single measure to represent environmental views. Political affiliation was a
significant predictor of NEP for both individuals belonging to multi-person households

(MPH; p < 0.001) and those living alone (LA; p < 0.001 ). Democrats had higher NEP
scores (R MPH = 61.27, R LA: 60.97) than both Independents (K MPH = 51.56, R LA =

52.06, p < 0.01) and Republicans (32 MPH = 43.51, 32 LA = 45.5; p <0.01), and the NEP
scores of Independents were significantly higher than those of Republicans (p < 0.01 ).

NEP scores were significantly positively related to education level for individuals living

46

alone (p < 0.05), but not for individuals living in multi-person households (p = 0.06).
Surprisingly, given past research on the NEP (Jones and Dunlap 1992), neither gender
nor age contributed significantly to explaining NEP scores. NEP scores were also
unrelated to income.

Camping and hiking were the most common outdoor recreation activities
performed by individuals participating in the study (73% of people answered they
sometimes or frequently participated in both). A majority of respondents participated in
bird watching (62%) and fishing (57%) sometimes or frequently. Boating and riding all-
terrain vehicles (ATVs) were less common (52% and 56% respectively never or rarely
participated in each). Respondents participated in hunting the least often (51% never
hunted and 15% rarely hunted).

Most respondents had household members who sometimes or frequently hiked
(75%) and camped (77%). The majority of respondents also indicated bird watching
(56%) and fishing (61%) were sometimes or frequently performed by household
members. Slightly more than half of the respondents indicated household members
sometimes or frequently boated (51%) or rode ATVs (53%). Hunting was least often
performed by household members (47% never hunted and 13% rarely hunted).
Household members of respondents participated in outdoor recreation activities at similar
rates to respondents. The strongest correlations between respondents’ participation in
outdoor recreation and participation by other household members were for bird watching
(r = 0.84) and boating (r = 0.82), followed by riding ATVs (r = 0.78), camping (r =
0.78), and hiking (r = 0.74). The weakest correlations were for fishing (r = 0.66) and

hunting (r = 0.55).

47

Principal component analyeses on outdoor recreation activities differed depending
on the group (i.e., multi-person household respondent, other household member, or
single-person householder; Table 3.1). For respondents living in multi-person
households, the appreciative activities of camping, hiking, and boating (most “boating” in
Teton Valley is non-motorized) loaded heavily on the first factor (Cronbach’s alpha =
0.67). The non-appreciative activities of fishing, hunting, and riding ATVs loaded
positively on the second factor (Cronbach’s alpha = 0.69). For activities performed by
household members other than the respondent, boating grouped with the non-appreciative
activities on the first factor (Cronbach’s alpha = 0.71), while camping and hiking loaded
on the second factor (Cronbach’s alpha = 0.67). For respondents who lived alone,
camping, boating, fishing, hunting, and riding ATVs grouped together on a single factor
(Cronbach’s alpha = 0.79). Considering the lack of consistency present, particularly in
the group of appreciative activities, and the fact that some activities fell into different
groupings across the three categories of individuals, we analyzed each outdoor recreation
activity separately, rather than combining them to form a composite measure.

We found a positive relationship between NEP scale and frequency of
participating in appreciative outdoor recreation activities, including bird watching,
hiking, and camping (Table 3.2). This relationship carried across all three categories of
participators: respondents who lived in multi-person households, their associated
household members, and individuals who lived alone. NEP score had the strongest zero-
order correlation with'participation in hiking for respondents living in multi-person
households when this activity was conducted either by the respondent or another member

of the household. For respondents living in single-person households, NEP score had the

48

strongest zero-order correlation with participation in bird watching. NEP score had the
weakest zero-order correlation with participation in boating for respondents of both
single-person and multi-person households. For respondents in multi-person households,
respondents’ NEP score had the weakest correlation with participation in camping by
other household members. We found appreciable differences among the zero-order and
partial correlations between NEP score and outdoor recreation activities (Table 3.2). The
largest declines in partial correlation coefficients occurred for hiking performed by
household members of respondents from multi-person households (0.12) and bird
watching performed by individuals who lived alone (0.19); all other changes were < 0.1.

A significant negative relationship existed between NEP scores and the frequency
of participating in all non-appreciative outdoor recreation activities for respondents in
multi-person households and their household members, but not for respondents living
alone (Table 3.2). The strongest zero-order correlation was between NEP score and
frequency of riding ATVs and the weakest for frequency of participation in fishing. As
in the case of the appreciative activities, there were differences between the zero-order
correlations and partial correlations. The largest declines between zero-order and partial
correlation coefficients occurred for ATV riding respondents in multi-person households
(0.10), ATV riding household members of respondents (0.14) and hunting respondents in
multi-person households (0.10). Declines in correlation coefficients when including
control variables were < 0.1 for all other activities.

With the exception of hunting, the correlation between NEP score and outdoor
recreation participation did not differ when the activity was performed by the respondents

themselves or non-respondents within the household (Table 3.2). For hunting, NEP score

49

was significantly more related to participation of the respondent’s household members
than their own participation (1 = 5.62, p < 0.001). The relationship between NEP score
and outdoor recreation activities also depended on whether activities were performed by
individuals who lived with others or individuals who lived alone. Individuals who lived
alone had stronger correlations between NEP scores and the appreciative outdoor
recreation activities and weaker correlations between NEP scores and the non-
appreciative outdoor recreation activities, than individuals who lived with others (Table
3.2).

For birding, hunting, and ATV use, respondent’s NEP scores were correlated with
whether or not their outdoor recreation activities matched those of their household
members (Table 3.3). Respondents who did not hunt had significantly higher NEP scores
if their household members also did not hunt than non-hunters with household members
who did hunt. The same was true for ATV users. In addition, respondents who rode
ATVs had significantly lower NEP scores if their household members also rode ATVs
than if their household members did not. Respondents who did not bird watch had
significantly higher NEP scores if their household members bird watched than if their
household members did not bird watch. There were no significant household effects for
outdoor recreation activities except bird watching, ATV use, and hunting. These results
should be interpreted with caution given the small sample sizes of some sub-groups in
which the participation of the individual did not match the participation of household
members. Furthermore, after controlling for demographic variables the only significant

effect was for non-ATV users (r = -0.22, p < 0.05). As in the case of the two-sample t

50

test, non-ATV users had significantly higher NEP scores if their household members did
not ride ATVs than if their household members did.
DISCUSSION

Our results support the hypothesized positive relationship between
environmentally oriented views and appreciative outdoor recreation and negative
relationship between environmentally oriented views and non-appreciative outdoor
recreation. The latter relationship, however, was only evident for people living in multi-
person households. Fishing, ATV use, and hunting did not negatively influence
environmental views of individuals who lived alone. While larger sample sizes might
detect a significant negative relationship for these activities, the correlations were among
the lowest we found (Table 3.2). This finding suggests it is not the activity so much as
the structure of social relationships that promul gates human exemptionalism among
participants in non-appreciative outdoor recreation. The activity has a different impact
on environmental views in different social structures (e.g., 2 participants, 2 non-
participants, or a mixture).

Unlike earlier studies (Dunlap and Heffeman 1975, Van Liere and Noe 1981,
Jackson 1986, Theodori et al. 1998), we found appreciable differences among the zero-
order and partial correlations between environmental views and outdoor recreation
activities (Tables 3.2 and 3.3). The larger differences probably reflect higher zero-order
correlation coefficients than earlier studies. Previous studies could not find larger
declines in partial correlation coefficients relative to zero-order coefficients because zero-
order coefficients rarely surpassed 0.1. For instance, Jackson (1986) reported higher

coefficients than usual, but only 12% of zero order coefficients were > 0.2 compared to

51

48% in this study. The stronger than average linkages (i.e., collinearity) between
environmental views and culturally important variables (e.g., education, political
affiliation) seems intuitive in communities where culture is defined and divided by
environmental issues (Peterson et al. 2002). In such areas (e.g., ski towns, fishing towns,
beach towns, or mountain biking towns) the environment may have stronger ties to
cultural identity (Smith and Krannich 2000, Peterson et al. 2006a).

This may also explain the relatively strong negative correlations between
environmentally oriented views and non-appreciative outdoor recreation (Table 3.2).
Hunting had high negative correlations with non-hunting household members’
environmental views compared to other activities (Table 3.3) but had the lowest
correlation between respondent and household member participation (r = 0.55). Even
though an individual was not highly likely to hunt when their other household members
hunted, respondents were likely to have more negative environmental views if their
household members hunted. Hunting also provided a notable exception to respondent’s
and household members’ participation in outdoor recreation being interchangeable with
regard to environmental views. Hunting participation of non-respondent householders
predicted respondent environmental views better than the respondent’s participation.
While this result is difficult to interpret, it may suggest social interactions associated with
hunting (e. g., discussions, story telling, hunting related media) may have stronger
negative ties to environmentally oriented views than participation in hunting itself. The
hunter recruitment problems suggested by low household member participation and
decreasing trends in hunter recruitment numbers, implies any negative impact of hunting

on environmentally oriented views will decline along with one of the largest sources of

52

income for many state wildlife and parks departments (Enck et al. 2000, United States
Department of the Interior and United States Department of Commerce 2002, Peterson
2004).

Due to high correlation between respondent and household member participation
(r = 0.78) and apparently strong transmission of the human exemptionalism views within
households (Table 3.3), ATV use presents a serious conservation challenge. While some
research suggested social evolution towards pro-environmental views and economic
constraints would lead to appreciable declines in ATV use (Jackson 1986), the opposite
trend has occurred. Between 1999 and 2003 the proportion of people above 16
participating in ATV outdoor recreation in the US. increased from 16.8% to 23.8%, and
participation rates doubled within the fastest growing demographic group in the US,
Hispanics (Cordell et al. 2005). Individuals living alone, however, did not demonstrate
negative correlations between environmentally oriented views and ATV use. Many
explanations for this phenomenon are possible including ATV users who lived alone:
participated in non-environmentally damaging versions of the activity, were unaware of
the damage they caused, or consciously made ATV use an exception to their
environmental views. In any case human exemptionalism and ATV use do not represent
an unbreakable positive feedback loop.

The unusually strong negative correlations and negative household effects of non-
appreciative outdoor recreation on environmental views may relate to the political
polarization of recent decades (Layman and Carsey 2002). Political affiliation predicted
environmental views better than any other demographic variable in this study. If cultural

groups representing various outdoor recreation activities follow the political trend of

53

polarization environmental value based divisions may grow larger in the future. This
potential problem can be addressed in part by deconstructing stereotypes of political and
recreational linkages (e.g., not all ATV users and hunters are Republican) and promoting
ideological diversity within recreation groups. While this suggestion may seem far
fetched, the successful campaigns of pro-gun Democratic candidates in the 2006 midterm
elections (e.g., Senators Jon Tester [Montana] and Jim Webb [Virginia]) demonstrate
such decoupling is both possible and a politically successful strategy.
CONSERVATION IMPLICATIONS

Our results suggest outdoor recreation participation has a larger impact on
environmental views than previously thought both because we found larger correlation
coefficients than previous studies and because some correlations permeated to non-
participating household members. These findings support Bright and Porter’s (2001) call
for research addressing the social factors influencing and mediating the outdoor
recreation and environmental concern relationship. As the first social unit beyond the
individual, households provide a logical place to begin this effort. Future research should
address how household dynamics (e. g., changing household size and family structure)
mediate the relationship between outdoor recreation and environmental views. Our
results suggest a specific need to understand why the relationship between environmental
views and outdoor recreation differs in multi-person and single-person households.

Since single-person households are increasing in prevalence and household size is
decreasing globally (Liu et al. 2003), the differences between multi-person and single-
person households will become more important for conservation efforts. For instance,

answering why ATV use and hunting in multi-person households correlated with less

54

environmentally oriented views, but the same activities had no effect on environmental
views in single-person households would represent the beginning of efforts to make these
sectors of outdoor recreation more environmentally oriented. Ideological changes will
not make all forms of outdoor recreation environmentally benign, but increasing the
number of environmentally oriented participants should decrease the prevalence of
environmentally damaging forms of recreation.

Future qualitative research could help illuminate how social dynamics in
households mediate the relationship between environmental worldviews and outdoor
recreation. Specifically participant observation and in depth interviews could document
how recreation activities of one household member influence other household members,
and the extent non-participating household members feel their own environmental views
are influenced by outdoor recreation of other household members. For example, this
approach could evaluate the role of media, cultural stereotypes, and story telling in
explaining why non-hunters in hunting households held less environmentally oriented

views than the hunters in their own homes or respondents from non-hunting households.

55

Table 3.1. Principal Components Analysis with Varimax rotation factor loadings for
outdoor recreation activities performed by: respondents who lived with others in the
same household, their associated household members, and respondents who lived
alone.

 

 

 

 

 

Activity Respondents with Respondents
household members Household members living alone
Factor 1 Factor 2 Factor 1 Factor 2 Factor 1
bird watching 0.120 -0.310 -0.088 0.292 0.024
hiking 0.602 -0.318 0.042 0.849 0.329
camping 0.751 0.080 0.459 0.565 0.665
boating 0.633 0.237 0.560 0.354 0.768
fishing 0.491 0.537 0.724 0.088 0.760
hunting 0.355 0.616 0.745 -0. 128 0.606
riding ATVs 0.067 0.613 0.522 -0.074 0.594
Eigenvalue 1.713 1.304 1.885 1.280 2.437
% variance
explained 24.5 18.6 26.9 18.3 34.8

 

 

56

HmEm mi... Nmnooame 8a 08.3m— nofimscoa condoms Gm 294 mooEmSa 08.99.18: 9H3 £on om 4309525 8a 9:909.
$9.038: 035.33 vmfiozpma 3.“ emmvoeamea £5 :30 ES 0353 5 9m £53 #55355. Ema 359080 005385 92:69.? 85
2.3095me 15 :43 £00m. N30102:; 8:05:03 8353 9:0. :5 9.303 333:0: 53.020 0:0 003:: 8:05:05 285:
men 35820:. 583%. 00:38: manage manna. 85 0mm.

 

 

 

 

Wmmcoeamefi 5 55:03.8: 205153 59:5? 0». 309520? Wmmvoeame? :35 30:... A. s H 0:
vocwmroam A: n 0 CV 9 u .38
>343.
NEPQAE 093.2 N9.o-9.03 03d“: N2.o-oafi. 093m:
_. 0 A 0 a. 0 a m .. 0 _. 0

90 0.15.0 0.000 0.00.”. 0000 0pm.“ 0.000 0. 5.» 0-00; 0.1.: 0 0,000 ouuw 0.000
29850.0.
01:5:er 0.03 0.000 0.15: 0.000 0001* 0.000 0. E.» 0 00.. 0 .000 0 000 0.30 0.03
08:05.0 0.74,: 003 0.3L 0000 0.000 0.03 0.0010 0.000 0.03 0000 0.1.5 0000
woBEm 0. H _0 0.0%,.» 0.03 0 03 0. 0 3 00.3 0.000 0.30 0. H.110 0.1.60 0.1100 0.000
3950 -0 0.3 0.000 -0000 0.00:5 10.30 0-00.” -0900 0.000 0.0.5.0 0.an 10.05 0.0:...»
35:50. 0.0.». 0.000 10.000 0. H 0.» 0.0.3 0.000 -0. 5.» 0.00.10 -0. _ 5 0.1% u 0.00... 0.000
5.08% -040.“ 0.000 -0000 0.000 -0120 0.000 0.me 0.000 -0. 00m 0.0.0 -003 0:30

>j.m

 

57

Table 3.3. Comparison of mean NEP scores between respondents who share each type of
outdoor recreation participation, or non-participation, with another household member
and respondents that do not share each type of outdoor recreation participation, or non-
participation, with another household member.

 

 

Type of outdoor

recreation participation _ _

or non-participation n3 N” X a X b t p
Bird watchers 246 12 52.872 51.083 0.578 0.573
Non-bird watchers 46 7 45.657 52.857 -2.342 0.042
Hikers 268 14 52.517 50.357 1.014 0.325
Non-hikers 19 1 1 44.327 46.636 -0.655 0.519
Campers 275 9 51.992 49.333 1.110 0.294
Non-campers 17 10 51.235 47.585 1.265 0.219
Boaters 216 13 53.203 48.846 1.331 0.205
Non-boaters 66 17 47.924 49.706 -0.586 0.564
Fishers 217 22 50.456 51.318 0324 0.749
Non-fishers 42 31 56.451 54.382 0.73 0.468
Hunters 119 35 49.163 53.114 -1.728 0.090
Non-hunters 1 12 45 54.932 56.451 2.892 0.005
ATV-users 166 1 1 47.492 52.091 -2.185 0.046
Non-ATV-users 111 23 57.875 54.932 2.657 0.011

 

 

3 cases where recreation activities of respondents and non-respondent household members match

b . . . . .
cases where recreation activmes of respondents and non-respondent household members do not match

58

CHAI’I‘ER 4

HOUSEHOLD LOCATION CHOICES: IMPLICATIONS FOR BIODIVERSITY

CONSERVATION

In collaboration with

Xiaodong Chen and J ianguo Liu

59

ABSTRACT

Successful biodiversity conservation efforts require understanding the ontogeny
of environmentally conscientious behavior. Household location choices represent a
directly observable behavior that predicts species endangerment as well as any social
variable, but no studies have addressed the socio-cultural determinants of those choices.
In this paper we address this gap with a case study in Teton Valley Idaho and Wyoming.
We explore the spatio-temporal relationship between socio-demographic variables,
environmental worldviews, and the household location choices of immigrants. We
collected socio-demographic data, spatial coordinates, and land cover (residential,
agricultural, and natural) information with 416 household surveys. Older and highly
educated immigrants with the most environmentally oriented worldviews chose to live in
natural areas (e. g., riparian zones, wetlands, critical winter range for wildlife) in
disproportionately high numbers while immigrants with the lowest education and least
environmentally oriented world views chose to live in previously established residential
areas in disproportionately high numbers. Length of residency was negatively related to
more environmentally oriented worldviews, suggesting pro-environmental worldviews
influenced home location choice more than natural home locations influenced
worldviews. Older respondents with more education and more environmentally oriented
worldviews made household location decisions posing the greatest threat to biodiversity
conservation. Further, the smaller household size (i.e., number of people per home) for
this group magnified their per-capita environmental impact. In these contexts

conservation requires providing people who have learned to love nature through the

60

wisdom of years, education, or environmentalism ways to experience nature besides
building a house in it.
INTRODUCTION

Conservation biology faces one of the most difficult and important challenges of
the 21St century, helping society care enough about the environment to behave in an
environmentally conscientious manner (Ehrlich 2003, Freyfogle 2003, Orr 2003).
Successful, if small scale, efforts to achieve this goal revolve around personalizing nature
(Thompson 2003, Thompson 2004, Schwartz 2006). Physical interactions with nature,
have overcome the common tendency to view the environmental impacts of personal
actions (e. g., where one chooses to live) as inconsequential (Thompson 2004). To fully
capitalize on the opportunities associated with public engagement in biodiversity
conservation, however, conservationists must understand the ontogeny of environmental
behavior (Freyfogle and Newton 2002, Peterson et a1. 2005).

Successful conservation initiatives must consider how dominant social lenses
shape public perception (Peterson et al. 2002, Peterson et al. 2006c) and how those lenses
shape the development of specific environmental behaviors impacting biodiversity
conservation (Dietz et al. 1998, Johnson et al. 2004). Research on the ontogeny of
environmental behavior relies heavily on a model wherein cultural context defines
worldviews, and worldviews lead, via attitudes and intentions, to behavior (see: Dietz et
al. 1998, Johnson et al. 2004). Worldviews, the set of narrative symbols humans use to
explain the nature of reality (Greeley 1993), provide the lens filtering human perception
of the environment (Catton and Dunlap 1978, Dunlap and Van Liere 1978).

Environmental ethicists posit a feedback loop between behavior and worldviews.

61

Specifically behaviors directly affecting the environment (e. g., shooting a deer, chopping
down and burning an oak tree, plowing a field) personalize the environment leading to
more environmentally oriented worldviews (Leopold 1949, Thompson 2003, Thompson
2004).

Scholars studying the ontogeny of environmental behavior face several
challenges. First large scale social surveys rarely measure actual behavior or behaviors
with direct impact on the environment. Accordingly most past research relies on self
reported behaviors with indirect linkages to biodiversity conservation (e. g., signing
petitions, voting, consumer choices, donations, entertainment choices, membership in
clubs: Dietz et al. 1998, Johnson et al. 2004). What people say they do, however, often
differs significantly from what they actually do (Argyris and Schon 1978, Argyris 1992).
Further, environmentally oriented worldviews, attitudes, intentions, and education do not
guarantee more environmentally sensitive behavior (Stern 2000). Finally, many
environmental behaviors directly impacting biodiversity conservation (e. g., hunting,
farming) have ambiguous conservation effects and relatively few participants (Peterson
2004).

Recent research, however, suggests choosing a household location represents a
directly observable and almost universal human behavior with direct impacts on
biodiversity conservation (Liu et al. 2003, Peterson et a1. 2007). Household density
predicts species endangerment as well as human population density, and is growing faster
than population density globally (Liu et al. 2003, Peterson et al. 2007). Choosing the
type and location of one’s home is potentially the most pervasive and direct link between

human attitudes and intentions and their physical impacts on the land. Despite this

62

critical linkage, little, if any, research has addressed the ontogeny of household location
decisions. We used a case study in Teton Valley of Idaho and Wyoming, USA to address
this gap by evaluating the spatio-temporal relationship between socio-demographic
variables, environmental worldviews, and the environmental impact of homes immigrants
chose to live in. We tested three hypotheses: 1) older respondents, those with the most
education, and those with more environmentally oriented worldviews preferentially
choose household locations in natural areas (i.e., the “loving nature to death” hypothesis),
2) the ecological impacts of household location decisions are magnified by smaller
household size (i.e., fewer people per house) of people choosing to live in natural areas,
and 3) living in natural areas leads to more environmental worldviews via feedback from
the environment.
METHODS

Teton Valley includes Teton County, Idaho; and the portion of Teton County,
Wyoming, west of the Teton Mountain Range (Figure 1.2). The Valley is within the
largest contiguous ecosystem (Greater Yellowstone) in the continental United States.
The local economy and population remained stable between the great depression and
1970, but immigration related to natural amenities and outdoor recreation led to 74%
jump in population (3,439 to 5,999) and 85% (1,123 to 2,078) jump in household
numbers during the 1990’s (Smith and Krannich 2000, Peterson et al. 2006b). This
pattern reflects the global phenomenon of household numbers growing faster than
population size (Liu et al. 2003).

Post 1990 immigrants were generally more urban, educated, and secular than their

predecessors (Smith and Krannich 2000, Peterson et al. 2006b). Development

63

accompanying the immigration lacked a general plan. The post-1990 development
coincides with the first E. coli outbreaks and near extirpation of native cutthroat trout
(Oncorhynchus clarki) in the Teton River (see http://www.tetonwater.org: Idaho Fish &
Game 2005). Development threatens water quality, wetlands, migration corridors in the
greater Yellowstone ecosystem (75% of which are already closed: Berger 2004), and
winter habitat for mule deer (Odocoileus hemionus) and elk (Cervus elaphus).

We used an in-person interview protocol to assess demographic characteristics,
environmental worldviews, and home location information for respondents. We chose
personal interviews because they promised higher response rates (Dillman 2000),
qualitative insight regarding conclusions made from our analyses, ability to identify exact
locations of homes, and opportunity to verify landscape attributes identified through
remote sensing. We purchased a representative sample of Teton Valley residents (n =
550) from Survey Sampling, Incorporated (F airfield, Connecticut; logistic constraints
dictated sample size). Arbitrarily selected residents of Lansing Michigan (n = 18), and
Victor, Idaho (within the study area; n = 23) pre-tested the questionnaire.

We attempted to visit each respondent during 4 time intervals (morning and
evening on a weekday and on a weekend day) during J uly—August, 2004. If visits failed
and we could not locate a physical address with the aide of local informants we made
initial contact via telephone. Interviewers defined acronyms, but answered other
questionnaire related queries by explaining questionnaire format, reading directly from
the questionnaire, or stating “whatever it means to you” (Groves 1989:451).

We used the New Ecological Paradigm (NEP) scale (Dunlap et al. 2000) to

measure environmental worldviews. The NEP contrasts ecological (i.e., humans and

64

nature share a common fate) and human exemptionalist (i.e., humans are exempt from
environmental constraints) environmental views (Catton and Dunlap 1978). The scale
taps a “folk ecology” or lay person’s view of human and nature relationships (Stern et al.
1995, Johnson et al. 20042159). Those embracing the conservation goals of the modern
environmental movement, consistently score higher on the NEP than other groups
(Dunlap and Van Liere 1978, Dunlap et al. 2000, Dunlap and Michelson 2002, Mertig et
al. 2002).

We used standard survey design to collect data for education (1 = less than high
school to 7 = graduate or professional degree), income (1 = < 14,999 to 9 = 2200,000),
age, household size, and gender (Dillman 2000). We divided households into large (>2)
and small (1 and 2) households based on average household sizes of the US (2.60), Idaho
(2.64), and Wyoming (2.43; US Census). We coded political affiliation as a 3 category
variable (1 = Republican, 2 = Independent, and 3 = Democrat). We coded residency
(native versus immigrant) as a binary variable from response to the following question:
“have you lived all your life in Teton County?”

We identified general household location using address geocoding in ArcView
3.2 (Environmental Systems Research Institute, Redlands, California). We then used
landmarks (e.g., tree lines, creek beds, notable buildings) and homestead attributes (e. g.,
lawn shape, roof type, house shape, topography, driveway shape and type) to locate and
mark the exact location of each house on a digital aerial photograph in ArcView 3.2.. We
classified land cover using categories related to environmental impact of a household,
rather than using typical vegetation classes. Immigrants to Teton Valley could choose to

live in: 1) residential areas, 2) agricultural areas, or 3) natural areas. Homes in residential

65

areas required little new infrastructure (e.g., roads, sewer lines, power lines), and caused
minimal fragmentation of natural land cover. Homes in agricultural areas, however,
required road construction, power line construction, and either extension of sewer lines
or, more likely, installation of septic systems. Finally, homes in natural land cover
required new infrastructure construction, replaced natural land cover, and magnified
environmental damage by either immediate adjacency to wetlands and streams, or
destruction and fragmentation of hillside habitats critical for elk and mule deer winter
range (Skovlin 1982, Kie and Czech 2000).

We used 2004 zoning maps from each municipality to delineate boundaries for
residential areas (i.e., city limits). All households within city limits were categorized as
residential unless they also occurred in a riparian zone. Natural land cover was limited to
wetlands and riparian zones on the valley floor and forest or rangeland areas on hillsides
bordering the valley. We classified riparian zones using a 100m buffer around streams
and rivers identified using 2000 US. Census Bureau's TIGER\Line® datasets. We used
the US. Fish and Wildlife Service’s (2005) National Wetland Inventory to identify
wetlands. The aforementioned pronounced vegetation zonation in this region (Bailey
1995) made visual identification of forest and rangeland on aerial photographs possible.
Because forest and rangeland was limited to hillsides bounding the valley, these land
cover types did not overlap with agricultural or residential areas. We categorized homes
surrounded by crop fields as within agricultural areas, unless they also occurred in a
wetland or riparian zone.

All descriptive and inferential statistics were calculated using SPSS (Release

15.0.0 [6 September 2006]). While we generated descriptive statistics for the entire

66

sample, we focused analysis of household location on immigrants. We made this
decision based on the results of preliminary qualitative research which suggested natives
had fundamentally different household location decisions (i.e., live in the home you were
born in, build a home on parents’ property), and lived in the area for different reasons

(e. g., family versus natural amenities). The latter finding was corroborated by this study.

We used t-tests to compare natives and immigrants, and one-way analysis of
variance (ANOVA) for comparisons between immigrants choosing to live in each of the
three areas (residential, agricultural, natural; 0.05 level of significance). If ANOVA was
significant, we used Duncan’s range test to evaluate differences among means (0.05 level
of significance). When data failed to meet ANOVA assumptions (i.e., normality and
equality of variance) we used chi-square tests of independence for comparisons. We used
binary logistic regression to control for correlation between education, age, gender,
income, and environmental worldviews, and to select variables significantly related to
choosing households in natural areas.

To test our first hypothesis, we evaluated household location preferences with an
adapted habitat-selection ratio (Thomas and Taylor 1990, Lopez et al. 2004) by variables
significantly related to household location decisions (age, education, and NEP). We
calculated habitat-selection ratios by dividing observed use of natural areas by
availability (Thomas and Taylor 1990, Lopez et al. 2004). For example, a selection ratio
(5) for respondent group X would be calculated as: S = ([UNX/UAAXJ/[UNT/UAATJ),
where UNX = the number of group X households in natural areas, UAAX = the total
number of group X households in all areas, UN T= total number of households in natural

areas, and UAA T = total number of households in all areas. To test the second hypothesis

67

we determined if large and small households had different likelihoods of occurring in
natural areas using a chi-square test. Finally we evaluated the potential for feedback
from living in natural areas by calculating Pearson’s correlation coefficients between
years of residency in each area and NEP.
RESULTS

The final compliance rate was 95% (n = 416; sampling error i 4.8%). Of
the 550 households in our original sample, 66 contacts were incorrect (e. g., resident
deceased or moved), 20 refused to provide an interview, and we could not contact
respondents at 48 households. Our sample aligned with 2000 census data in terms of sex
(46% female) and ethnicity (90% Anglo, 6% Hispanic). Median annual family income
was $35,000—$49,999, 6.5% of respondents had annual family incomes below $15,000,
and 90% were below $100,000. Only 5% of the respondents had less than high school

education and 40% held at least a 4 year college degree. Mean respondent age was 46.
Most (74.4%) respondents were immigrants. Immigrants were younger (X = 44.4) than

natives (i = 51.7; p < 0.001). Immigrants had higher education levels than natives (x2 =
23.32, 6 (if, p < 0.001) with more than twice the percentage of college graduates (45.2%
immigrant versus 22.9% native), but income distributions were not significantly different.
Natives were more likely to be Republican (47.7%) than immigrants (30.0%; x2 = l 1.04,
1 df, p = 0.001).

Immigrants and natives chose the location of their household for different reasons
(x2 = 62.93, 2 df, p < 0.001). Immigrants chose their household location based primarily
on natural amenities, and economic considerations (e. g., jobs, cost of living; Table 4.1).

Few immigrants cited home or family as primary considerations in their household

68

location decision. Conversely natives cited home or family as their primary

consideration for household location nearly twice as often as natural amenities, and rarely
cited economic considerations (Table 4.1). Natives scored significantly lower (i = 45.5)

on the NEP scale than immigrants (i = 51.7; p < 0.001). Immigrant groups also chose
the location of their household for different reasons ([7 = 13.36, 4 df, p < 0.01).
Immigrants moving to natural areas chose their household location based on natural
amenities more often than other immigrants (Table 4.1). Residential area immigrants
were most likely to cite economic reasons. Few respondents from any immigrant group
cited home or family as primary considerations in their household location decision
(Table 4.1).

When considered simultaneously via binary logistic regression, age (Wald =
13.42, p < 0.001), education (Wald = 8.15, p = 0.004), and environmental views (Wald =
5.67, p = 0.017) predicted whether immigrants chose to live in natural areas, but neither
income (p = 0.078) nor gender (p = 0.62) were significant. Immigrants moving to each

area had significantly different mean ages (F = 15.53, p < 0.001, p < 0.05 for all post-hoe
tests). Those moving to natural areas were the oldest (n = 80; X = 49.7) followed by
those moving to agricultural areas (n = 159; i = 44.8) and those moving into residential
areas (n = 67; i = 37.0). Older immigrants moved to natural areas at rates greater than
their availability (Figure 4.1).

Immigrants moving to natural areas had higher education levels than immigrants
moving to agricultural or residential areas (x2 = 42.81, 12 df, p < 0.001). Almost a third
(29.1%) of immigrants choosing to live in natural areas had graduate or professional

degrees; only half as many immigrants moving to agricultural areas (14.5%) and 3% of

69

those moving to residential areas did. This education gap was also pronounced at the
college graduate level (natural area = 63.3%, agricultural area = 40.9%, residential area =
34.3%). Immigrants with at least some college preferentially chose to live in natural
areas, and immigrants with Associates degrees or less preferentially chose not to live in
natural areas (Figure 4.2).

Immigrants choosing to live in different areas had significantly different NEP

scores (F = 5.27, p = 0.006). Immigrants moving to natural areas exhibited significantly
higher NEP scores (35 = 57.46, p < 0.05, SE = 1.29, n = 79) than those moving to

agricultural (i = 53.43, SE = 0.91, n = 160) and residential areas (3(- = 51.59, SE = 1.40,
n = 67) which were not significantly different. While the relationship between NEP
percentile and household location appeared random at intermediate NEP levels, the
immigrants with lowest NEP scores moved to natural areas at half the average rate and
the immigrants with highest NEP scores moved to natural areas at double the average rate
(Figure 4.3).

Respondents with small households (1 and 2 persons) were significantly more
likely to live in natural areas than respondents with larger households (Chi-square =
16.63, p < 0.001; Figure 4.4). Years of residency in natural areas was negatively related
to NEP (r = -0.24, p < 0.05), and controlling for age, education, political affiliation, and
income made little difference in the relationship (partial correlation = -0.22, p = 0.052).
We observed a similar relationship in agricultural areas (partial correlation = -0.20, p =
0.011), but the relationship was insignificant in residential areas (partial correlation = -

0.06, p = 0.698).

70

DISCUSSION

These results generally support our first hypothesis (older, more educated, and
more environmentally oriented respondents chose homes in natural areas) and second
hypothesis (ecological impacts of people choosing to live in natural areas were magnified
by their smaller than average household size), but fail to support our third hypothesis
(living in natural areas actually led to less environmentally oriented worldviews). These
findings do not suggest social laws making older, more educated, and more
environmentally oriented people more likely to build homes in environmentally sensitive
areas. Rather they demonstrate the need for explicit consideration of household impacts
on resource use and biodiversity conservation (Liu et al. 2003), and development of ways
to experience pristine environments besides living in them.

Measuring environmental behavior in terms of household location rather than
recycling, watching nature related television, or donating money to environmental
organizations produced different results than previous studies which found either no
relationship or a weak positive relationship between education, environmentally oriented
worldviews, and pro-environmental behavior (Dietz et al. 1998, Nord et a1. 1998,
Johnson et al. 2004). Household location, as an indicator of environmental behavior,
yielded a strong negative relationship between age, education, NEP, and pro-
environmental behavior. Older highly educated immigrants with environmentally
oriented worldviews chose to live in natural areas (e. g., riparian zones, wetlands, critical
winter range for wildlife) in disproportionately high numbers while immigrants with the
lowest education and least ecologically oriented world views chose to live in previously

established residential areas in disproportionately high numbers. In Teton Valley, natural

71

areas were the most difficult to develop due to both permitting and land prices, but older
persons, the highly educated and most environmentally oriented immigrants choose to
live in them even though they did not have higher than average incomes. For these
respondents household location was an explicit choice with higher economic costs than
other options.

These results suggest a serious problem for conservation biology. Those
demonstrating the greatest financial, political, and ideological support for conservation
can become biodiversity’s greatest threat. Moreover, the number of households for older
persons should boom on global, continental, and national scales in the future
(unpublished data). The contribution of older persons to household proliferation will be
significant due to the increasing number of older persons, the decreasing average number
of people in their households, and the increasing spatial size of their homes (unpublished
data). Human over-population has often been identified as the greatest human threat to
biodiversity conservation, but as population growth rates and average household size
decline globally, household proliferation will progressively become a more serious
problem (Liu et al. 2003, Peterson et al. 2007). Further, household density and
population density are scientifically indistinguishable as predictors of species
endangerment on a global scale (Peterson et al. 2007). Our findings should not be
construed as contradicting traditional threats to biodiversity conservation, but they do
suggest looking beyond over-population, poverty, poor education, and neo-conservatives
eager to exploit the environment for economic gain. In this case older highly educated
individuals with environmentally oriented worldviews posed the greatest threat to

biodiversity conservation.

72

The negative relationship between NEP and time as a resident in natural areas
takes the silver lining out of our results by suggesting environmental worldviews
motivated home location choice rather than location creating environmental worldviews.
Environmental worldviews were place based (Norton and Harmon 1997), but place was
defined by social context (e. g., the local values immigrants absorbed after settling) not
living in natural areas (Smith and Krannich 2000). Natural features brought
environmentally oriented immigrants to Teton Valley and disproportionately to Teton
Valley’s natural areas. Living in a social context with less environmental worldviews
eventually led to less environmentally oriented worldviews for immigrants.

A conservation psychology perspective (Clayton and Brook 2005, Saunders et al.
2006) provides several interpretations of our results: 1) people do not link what may be
their most important conservation related behavior, choosing where they live, to
environmental goals, 2) personal motives for living in natural areas trump the desire to
act in accordance with environmental worldviews, and 3) situational context prevents
people from acting in accordance with their environmental worldviews. Each
interpretation seems reasonable since thinking about household impacts on biodiversity is
a relatively new idea (Liu et al. 2003, Peterson et al. 2007), social, psychological and
physical well being are all linked to homes in natural areas (Frumkin 2001), and the
situational context in our study area reflected declining financial viability of farming and
ranching, migration related to natural amenities making land valuable, and a history of
strong property rights.

Education provides the most obvious strategy for helping people align their

environmental views and behaviors (Berkowitz et al. 1999). Our results, however,

73

support previous research suggesting education does not necessarily lead to more
environmentally conscientious behavior (Stern 2000). Ecological education, as opposed
to other types, might reduce the tendency to choose living in ecologically sensitive areas,
but our study suggests problems with this approach as well. Type of education was not
explicitly included in the interview, but we were able determine education type for 44.3%
(n = 31) of our respondents with advanced degrees using follow up contacts and notes
from interviews, and 25.8% of them were educated in environmental fields (e. g., ecology,
forestry, wildlife biology, botany, zoology). Few of these immigrants needed information
about the ecological costs of building homes in riparian areas or on hillsides. One
biologist, a recent retiree from Idaho Fish and Game, pointed out this issue long before it
was supported by survey results. He spoke loudly so his voice would carry over the nail
guns that were tacking his new home together saying: “the biggest problem is the loss of
winter range (for mule deer and elk), and I've now become part of it because my wife
won’t live in town.” This comment followed two stories about other Idaho communities
where Fish and Game was pressured to feed elk after winter range was gobbled up by
development.
CONSERVATION IMPLICATIONS

Linking household location decisions to environmental worldviews represents a
critical step for biodiversity conservation. Environmentally conscious decisions about
home appliances, food consumed, family planning, voting, donations, activism, and
transportation will not protect biodiversity unless people make the environmentally
conscious decision regarding their home. By framing biodiversity conservation in terms

of household impacts one gains a metric with direct impacts on biodiversity conservation,

74

promotes social justice, and reduces the validity of hypocrisy allegations levied against
the environmental community. A household perspective shifts the burden of
conservation from the backs of the uneducated and poor to the highly educated and
wealthy, and expects educated enviromnentalists to sacrifice what they want (e. g., a home
on a river, on a mountain side, or on fragile desert soils) before expecting poor or
uneducated individuals to sacrifice what they need for basic living (e. g., heating, health
care, college education for their children) in the name of conservation. Even considering
the multiplicative effect of small household size, homes of older well educated
environmentalists have relatively small impacts on biodiversity conservation compared to
global household proliferation. Changing global household proliferation, however,
requires leaders with moral authority. To build that authority, local conservation leaders

must make household decisions reflecting what they advocate.

75

Table 4.1. Primary reason for home location choices of natives and immigrants moving to
natural areas, agricultural areas, and residential areas in Teton Valley.

 

 

 

Group Primary Reason for Household Location
Natural Economic Home Place
Amenities Constraints

Natives 34% 16% 56%

Natural Area 72% 14% 1 4%

Immigrants

Agricultural Area 58% 23% 19%

Immigrants

Residential Area 47% 39% 14%

Immigrants

 

76

Selection
Ratio

 

 

 

93 39 J) ‘73 J ‘1 ‘9 ’2 6153‘

Age

Figure 4.1. Relationship between age (9 equal intervals) and the weighted percent of
immigrants within that age level choosing to live in natural areas (hillsides, 100m riparian
buffer, and wetlands).

77

1.5~ g
1.
Selection
Prefer
Ava/d
0.5 J

 

 

 

 

 

 

 

 

 

Figure 4.2. Relationship between education level and the weighted percent of immigrants
within that education level choosing to live in natural areas (hillsides, 100m riparian
buffer, and wetlands).

78

1.5
Selection
Ratio 1

0.5

Prefer

A void

 

 

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 l

NEP Percentile in 10% Increments

Figure 4.3. Relationship between NEP percentile (in 10% increments) and the weighted
percent of immigrants within that percentile range choosing to live in natural areas
(hillsides, 100m riparian buffer, and wetlands).

79

Percent of
Households
in Natural
Areas

 

 

 

l 2 3 4 5 6 7
Household Size

Figure 4.4. Relationship between household size (average number of people per
household) and the percent of respondents choosing to live in natural areas.

80

CHAPTER 5

A SYSTEMS MODEL FOR INTEGRATING CONSERVATION, LAND USE

CHANGE, AND HOUSEHOLD PROLIFERATION

In collaboration with

Jianguo Liu

8]

ABSTRACT

Humans are rapidly becoming a dominant source of disturbance and change in
ecosystems. In many human dominated ecosystems land use change threatens wildlife
conservation. Integration of ecological and socio-demographic data is critical for wildlife
conservation in such contexts. Within the Intermountain West, conflicts between land
rich and cash poor locals wanting to develop their economically unviable farms and more
cash wealthy immigrants wanting to protect natural amenities from development make
the integration of social and environmental information essential for conservation. We
developed a systems model to simulate how socio-demographic change and various land
use policies will influence future land use change in Teton County, Idaho. We developed
20 development scenarios to evaluate strategies for meeting the needs of wildlife by
protecting open space and meeting the needs of long term residents by allowing them to
profit from developing their farmland property. The development scenarios emerged
from a combination of a vacation home ban, gangplank policy, and 5 levels of cluster
development. The vacation home ban reflects the possibility that open space could be
protected by banning the construction of secondary or seasonal homes that lefi empty
rather than used as rental properties. The gangplank policy reflects the tendency of
immigrants to use their progressively growing political power to pull up the gangplank
after themselves with policies slowing new home construction. We then ranked the
strategies based on the amount of open space they preserved, the amount of development
they allowed, and a combination of those factors. Model evaluation suggested the model
provided rigorous estimates of household numbers (i.e., no significant differences

between model predictions of and US Census estimates). The model was most sensitive

82

to changes in immigration rates and household size of immigrants in each land use area.
Cluster development was the most important strategy for protecting open space. The top
2 development strategies in open space rankings always included 100% cluster
development. Gangplank policy or lack thereof, was the most important policy for
regulating new home construction. Finally composite rankings favored high levels of
cluster development and not implementing gangplank policy or a vacation home ban.
These results suggest gangplank policies are not necessary to protect open space, and
may prove problematic if pressure to develop increases in the future or if they cause
conflicts preventing smart growth strategies such as cluster development. The cluster
development strategy presents a win-win scenario for Teton County residents by allowing
them to achieve high levels of protected open space and profit for current landowners.
This simulation model is also a simple and intuitive tool for other regions where
conservation goals appear to clash with the economic well being of people.
INTRODUCTION

Human activities, notably household proliferation (Liu et a1. 2003) are causing
major changes in global ecosystems (Soulé 1983;1986, Vitousek et al. 1997). These
changes can pose a serious threat to biodiversity conservation (Harcourt et a1. 2001, Liu
et al. 2001). Environmental problems and decision making to address the problems are
becoming more complex. Because humans now dominate most ecosystems, the new
ecology for understanding environmental problems focuses on coupled human-natural
systems (Botkin 1990, Vitousek et al. 1997, Redman 1999, Liu et al. 2007).

Conserving biodiversity in coupled human and natural systems

(CHANS) requires understanding the linkages between human systems and

83

ecosystems. Within the context of CHANS, failure to integrate bio-geographic
features with sociology has led to many past conservation failures (Liu 2001).
Within the intermountain west, clashes between residents hoping to cash in on
development and those hoping to conserve wildlife and other natural amenities
have divided many communities (Smith and Krannich 2000), but those clashes
emerge from an unproven assumption that allowing landowners to cash in on

development and protecting open space are incompatible goals.

Several scholars suggest the systems approach can help address similar
challenges associated with conservation in complex CHANS (Checkland 1981,
Senge 1990, Vennix 1999, van den Belt 2004). Systems modeling and
simulation, a tool developed by systems theorists, has been effectively used to
address biodiversity conservation in coupled human-natural systems (Liu et al.
1995, An et al. 2001, Liu 2001). Therefore, we decided to use a systems
modeling approach to integrate ecological and socio-demographic data for
developing land use planning strategies that both protect wildlife and allow
landowners to develop their property in Teton County. Our model uses
households to link social processes and ecological processes. Only one
simulation model (An et al. 2001, An et al. 2002) has used this approach for
wildlife conservation purposes, and that model focused on household level fuel
wood collection decisions instead of household location decisions. Our model
has broad implications for wildlife conservation since it represents the first
attempt to assess the impacts of home location decisions for wildlife

conservation.

84

Dynamics simulated in this study also have broad implications for wildlife
conservation because of Teton Valley’s location. Teton Valley is a natural amenity rich
community inside the Greater Yellowstone Ecosystem, and inside a region of the United
States with high species endemicity and endangerment rates and high ex-urban
immigration rates (Gutmann et al. 1998, Rutledge et a1. 2001). Exurban development
represents the fastest growing development style in the United States today (Crump
2003). Since 1997 over 80% of housing development occurred in rural areas (Heimlich
and Anderson 2001). Reverse migration (urban to rural) to nonmetropolitan natural
amenity rich areas has created a new cultural phenomenon, with serious implications for
conservation biology in the United States. Prior to the 19703 most immigrants followed
economic opportunity to urban centers (Zelinsky 1971, Beyers and Nelson 2000, Carr
2004). This trend was reversed temporarily in the 19703 and most recently in the 1990s
when rural population growth began consistently outpacing urban population growth
(Shumway and Davis 1996, Johnson and Fuguitt 2000, Shumway and Otterstrom 2001).
During the 19903 reverse migration caused the greatest population increases in rural areas
with high natural amenity value to potential residents (Shumway and Davis 1996, Smith
and Krannich 2000, Johnson 2003).

These rural natural amenity rich communities represent a rapidly growing
(double digit population growth over the last 15 years: Johnson 2003, Jones et al. 2003,
Radeloff et al. 2005), but rarely considered, component of the wildland-urban interface.
Typically the wildland-urban interface is defined as the area where houses on the border
of urban areas intermingle with undeveloped wildland vegetation (Radeloff et al. 2005).

This interface created by ex-urban immigrants spilling past current suburbs of urban areas

85

(e. g., Atlanta, Chicago, and Denver) simply represents new suburbs. An alternative
perspective, however, suggests large numbers of ex-urban immigrants can create a
wildland-urban interface in wildlands (areas with previously undeveloped wildland
vegetation) of any distance from the urban center. We consider the latter type of
wildland-urban interface a social interface because the urban component reflects the
cultural background of human residents rather than the built environment. This social
wildland-urban interface can emerge in areas reflecting more traditional connotations of
“wildness” (e. g., within the Greater Yellowstone Ecosystem (Rasker and Hansen 2000)).
Understanding how this new form of wildland-urban interface grows is critical for
progressive wildlife conservation initiatives. Systems modeling, however, can do more
than quantify the effects of household proliferation on land use and population growth. A
systems modeling approach facilitates exploring land use policy scenarios, before
irreversible damage to natural or social systems occurs (Conway and Lathrop 2005).
Many exurban communities have struggled to coordinate and plan land use in the face of
high development pressure. Within the Intermountain West 3 land use planning
strategies have been implemented: 1) market control (developers and land speculators
drive planning), 2) gangplank policy, and 3) cluster development. Gangplank policy
refers to immigrants using progressively growing political power to shift the costs of
conservation to large landholders with growth slowing restrictions. Cluster development
has been used for more than 30 years in the US. Midwest and Southeast to combat
sprawl, but is a relatively novel approach in the property rights-obsessed Intermountain

West (Arendt 1992, Jacobs 1998, Smith and Krannich 2000).

86

In most cases developers and land speculators use the drive development with
limited public oversight in poorly attended planning and zoning meetings. In the few
cases where communities attempt to slow growth (e. g., building moratoriums, slower
permitting processes, larger areas required per home) communities tend to erupt into land
use conflict. Local conflicts over land use in such areas spawned the culture clash (Price
and Clay 1980) and gangplank (see: Smith and Krannich 2000) explanations for land use
conflict in rural areas experiencing natural amenity related immigration. The culture
clash explanation suggests deep-rooted cultural differences led to conflicts between
longer-term residents and recent ex-urban migrants. The gangplank hypothesis suggests
after ex-urban immigrants gain access to their rural refuge they are eager to pull up the
gangplank and prevent further development with growth slowing restrictions than longer-
terrn residents.

The information needed by managers to improve wildlife conservation in these
contexts is diverse but includes how to maximize open space protection without violating
perceived property rights of local residents. In this chapter we start addressing this need
with a simulation model that evaluates how exurban immigration impacts household
proliferation and land use change and how that relationship might change in the future
under various combinations of four policy scenarios: market control, vacation home ban,
growth slowing restrictions, and cluster development. We ranked 20 development
scenarios (Table 5.3) based on the amount of open space they preserved, the amount of

development they allowed (i.e., number of homes), and a combination of those factors.

87

BASIC STRUCTURE OF THE MODEL

The systems model addresses household proliferation and land use change at the
county level because land use decisions and policy are made at the county level in the
study area. We removed the Wyoming portion of the Teton Valley study area (Figure
1.2) because insufficient data were collected in for inference to Teton County Wyoming.
All data were collected using the survey described in Chapters 3 and 4 or collected from
the US. Census. We realized landowners on the property rights side of land use conflicts
were probably concerned about being able to develop land and about the profit per acre
of land developed. Accordingly home numbers could be a poor proxy for meeting the
needs of these stakeholders. To determine if this was a problem we regressed the acreage
and listing price of all houses in the Multiple Listing Service (MLS) in Teton County for
April of 2007. There was no relationship between acreage and asking price (n = 153, R2
= 0.0009). Of the 28 homes listed for > $1,000,000, only 2 were on lots larger than 1.2
ha.

We used three strategies suggested by Xie et al. (1999) to improve the utility of
simulation models: design a simple model to facilitate use by other people, use as much
publicly available data as possible, and design a user friendly interface. The basic
structure of the model consists of 3 sub models representing (1) population growth from
immigration, (2) house construction, and (3) land use change, and a module for
controlling policy scenarios (Figure 5.1). In systems dynamics models, sub models must
have state variables (Grant et a1. 1997). State variables are points of accumulation in the
system, and usually store “material” (e. g., people, plants, Kg’s of a material, m3 of water,

energy, knowledge). The policy module only contains driving (components of the system

88

that affect, but are not affected by, other components), and auxiliary (converted value of
another model component) variables.

The population sub model simulates population growth from immigration. This
relies on the assumption that emigration will prevent natural population grth as it did
between 1900 and 2004 (see Chapters 3 and 4). Immigration, however, receives
feedback from the number of open homes and available land. If development slows due
to gangplank policy or the depletion of undeveloped private land, the reduced number of
new housing units constrains immigration (Figure 5.1). Residents who lived in Teton
County before a major immigration event started had different household sizes, made
different home location decisions (see Chapter 4), and chose their home location for
different reasons (see Chapter 4), so we divided the population into a static number of
longer term residents and a growing immigrant population. We divided the population
using responses to the following question: 1) have you lived all your life in Teton County
(if respondents answered “no” we asked: “how many years have you lived in Teton
Valley”). Those living in Teton Valley prior to 1990 were coded as longer term
residents; otherwise we coded them as newcomers. Previous studies comparing longer
term residents and newcomers in the Intermountain West defined the local group using
either a residency requirement based on theoretical time required to integrate into a new
culture (Graber 1974, Fortmann and Kusel 1990), or the year a major in-migration began
(Smith and Krannich 2000). A major in-migration event began in Teton County, Idaho in
the early 19905, so we chose the latter approach. We used least-squares nonlinear
piecewise regression of annual population data (U.S. decennial census and annual

estimates) to quantitatively estimate the threshold related to the immigration boom

89

(StatSoft 2003). The regression model based on a pivot point between 1991 and 1992
accounted for 99.5% of observed variance in population change. We, however, used
1990 instead of 1991 to divide the population so the cut off date would match decennial
census data. This change still reflects the major inflection point in population related to
immigration (Figure 5.2).

The house construction sub model simulates the number of homes built in natural,
agricultural, and residential land use areas (see Chapter 4 for detailed descriptions of how
these land use areas were defined). The number of houses built depends on demand
created by immigration, but can be constrained by polices restricting development or by
depletion of undeveloped private property (Figure 5.1). The number of houses built in
each land use type (i.e., natural, agricultural, residential) depends on the number of
people moving to that area and the average household size (i.e., number of people in a
household) for new homes built in that land use type.

The land use change sub model simulates how total hectares of land transition
between private open land and protected areas, clustered residential, and low density
residential sprawl as a function of the number of homes in each area and the number of
hectares required by zoning for homes in each land use type (Figure 5.1). The dynamics
within the land use change sub model have direct impacts on water quality, critical
ungulate winter range, and landscape fragmentation.

Finally the policy module allows users to control various combinations and levels
of development. First it allows the user experiment with the gangplank hypotheses.
Specifically the user can use a “gangplank policy” switch to control if immigrants can

slow down development as they gain political power over longer term residents. A “ban

90

vacation homes” (those not lived in by owners) switch allows the user to evaluate the
effect of vacation homes on public open space over time. Finally the model allows the
user to assess the impacts of various percentages of development being cluster
development using a “percent cluster development required” slider ranging from 0% to
100%. Cluster development refers to developing the same number of homes on a given
amount of land, but instead of developing one home on ten 10 ha parcels, cluster
development would place 10 homes on 10 ha and protect a strategically chosen 90 ha plot
for green space. The user can control the percent of development using cluster
approaches with a slider in the user interface.

QUANTITATIVE DESCRIPTION OF THE MODEL

The model is represented mathematically as a stochastic compartment
model based on difference equations with a l-year time step. Simulations are run
on a personal computer using STELLA® 8.1 (High Performance Systems, Inc.,
2004). For all sub models we present the general form of both state variable and
material transfer equations below. In these equations state variables are denoted
by upper case X’s, rates of material transfers (e. g., immigration, home
construction, hectares of land converted from agricultural use to residential use) by
lower case letters, and model parameters (e. g., percent of immigrants demanding
homes in natural areas, size of land parcels converted) by upper case letters.
Superscripts refer to the type of material in a state variable (people [p], houses [h],

hectares [ha]), and subscripts refer to the specified time.

91

The population sub model contains one state variable representing the total
number of residents in Teton County (initial population = 3,439). The state
variable equation is:

X17,” = X”, + (input,)

Where X”, is population at time 1, input, is the sum of material transfers into X” during the
time interval t to t+1 and represents immigration from outside Teton County. Material

transfer equations are as follows:

fPA,ifoz*X p, > PA
mi,=<

 

10‘,“ X ”1, otherwise
where mi, is the number of immigrants migrating to Teton County from time t to H],
a is the number of immigrants moving to Teton County from time t to t+1 per
resident at time t (0.0583), and PA = the population accommodated by new
housing built from time t to t+1. The equation for the PA auxiliary variable is as
follows:

PA = lrbn, * HHSN + lzba, * HHSA + lzbr, * HHSR
where lzbn,, hba,, and hbr,, are material transfers in the house construction sub model
representing new homes (excluding vacation homes which do not actually accommodate
residents) built between time t and (+1. The equation has 3 stochastic driving variables
(with normal distributions) representing average household sizes in each land use type:
HHSN = household size in natural areas (mean = 2.31, SD = 1.39), HHSA = household

size in agricultural areas (mean = 2.85, SD = 1.53), and HHRA = household size in

residential areas (mean = 3.56, SD = 1.80).

92

The house construction sub model contains 6 state variables representing the total
number of occupied and unoccupied “vacation” homes in natural, agricultural, and
residential land use areas. State variable equations for the six state variables are of the
general form:

Xi’,_,+/ = X”,,, + (input,,,)

Where X ',;,+1 is number of homes in land use area i at time t, input” is the sum of material
transfers into X,- during the time interval t to H] and represents new home construction

in land use area 1'. Material transfer equations for each of the state variables are as

follows:
HD, *LAI,,1fGSR = 0
libLt =
HDI. *LAI, * GSRI,,0!IIem'ise
HD, *O/oVHi *LAI, *VB,ifGSR = 0
Vhbm =

HDI. * %VH,. * LAI, * GSR1,, otherwise

where lib“ = the number of occupied homes built in land use area i in Teton County from
time t to t+1, GSR = a binary “switch” for grth slowing restrictions (1 = restrictions),
HD, = the number of houses demanded in land use area i, , LAI, = a home construction
slow down related to land scarcity (LAI, = 1 if > 2,000 ha of private open land exists
otherwise LA], = 0.0005* ha of private open land), GSR], = a home construction slow
down related to political action of immigrants (GSR1,= an exponentially declining
graphical function, Table 5.1), 1211b“: the number of unoccupied vacation homes built in
land use area i in Teton County from time t to t+1, %VH,- = the percent of vacation homes

built in land use area 1' respectively, and VB = a binary “switch” for vacation home ban

93

policy that only applies in material transfers for unoccupied homes (1 = no ban, 0 = ban
and 0 vacation homes built). No data existed for quantifying the LA], or GSRI,, so we
relied on logic to parameterize the variables. The ability of systems dynamics models to
incorporate this type of information is one reason they are useful for modeling complex
systems where empirical data are rarely available for all system components and
relationships (Grant et al. 1997). New home construction should slow down after some
threshold of open land scarcity is passed. Without compelling reasons to use more
complex functions we relied on a linear relationship where LA], declined to zero when no
land was available. Logic suggests the GSR], represents an exponentially declining
function. The GSR], relates to the gangplank hypotheses, and assumes that once exurban
immigrants gain political power (i.e., a majority) they will make rapid gains in slowing
development, but once immigrants gain a majority further population growth will lead to
progressively smaller gains in political power and policy control (i.e., after gaining a
voting majority increases in political power slow down). While these equations were not
based on empirical data, model evaluation assesses their reliability. The equation for the
HD, auxiliary variable (homes demanded) follows:

HD, = (a *X”,) * %IH,/HHS,
where %]H,- = the percent of immigrant houses in land use area i (natural = 0.2073,
agricultural = 0.5207, and residential = 0.2396), and HHS,- = the average household size
of immigrants in land use area i. The initial value for houses in each land use type was
calculated as:

[H.- = LP * %H,- /HSL *( ”/00, or %V)

94

where 1H,- : initial houses in land use area 1', LP = longer term resident (pre-l 990)
population (3439), %H, = percent of homes in land use area i matural = 0.1619,
Agricultural = 0.5143, Residential = 0.3238), HHSL = pre-1990 household size of
residents (3.03), 000, = the percent of occupied homes in each land use area (Natural =
0.9677, Agricultural = 0.9297, Residential = 0.9676), and %V= the percent of vacation
homes built for every occupied home (9%).

The land use change sub model contains 4 state variables representing the total
number of ha in private open space, residential sprawl, cluster development, and
protected open space. State variable equations for the latter three variables are of the

general form:

X"",-,,+, = “2,, + (input,,)
where Xhu,_,+ I is number of ha in land use area i at time t, input“ is the sum of material
transfers into X“,- during the time interval 1 to (+1 and represents conversion of private
open space to residential sprawl, cluster development, or protected open space. The state
variable equation for private open space is:

XML,” = X'm,_, - (output,_,)
where output“ is the sum of material transfers out of X} "’,- during the time interval t to t+1
and represents conversion of private open space to residential sprawl, cluster
development, or protected open space. The material transfer equations for the state

variables are as follows:

95

[(hbm + hbw ) * RHA, if %CD = 0

 

t . =4

C” (1219,], + hbW) + %CD *RHA *
(11190., + hbw', + 12an + hb,.,,,),othemrise
“(121),, + hb,,a.,)*AHA+ (121),, + hbW) * %H *HHA+
(hbm + hbW) * %W*AHA,if%CD=O

is” =<

(th, + hbm) *AHA + (hbn', + hbW) * %H *HHA+
\(hbn', + hbm, ) * %W *AHA*(1-%CD), otherwise

 

r0, if %CD = 0

tpr‘m =<((hba,, + hbW) *AHA —(hba’, + hbW) *RHA) +
((hb,” + hbm,) * %H *HHA—(hbn', + hbw) * %H *
\RHA) +((hb,,', + hb,,,,_,) * %W *AHA) *( l-%CD), otherwise

 

where ((3,, = ha converted to cluster development at time t, %CD = the percent of cluster
development required, hb,.,, = the number of homes built in residential areas at time t,
hbm, = the number of vacation homes built in residential areas at time t, RHA = the
number of ha per home built in a residential area (0.4047), 12b0,, = the number of homes
built in agricultural areas at time t, lzbw, = the number of vacation homes built in
agricultural areas at time t, 12b", = the number of homes built in natural areas at time t,
hb,.,,,, = the number of vacation homes built in natural areas at time t, ts,;, = ha converted to
sprawl development at time t, AHA = the number of ha per home built in agricultural or
riparian areas (8.0937), ”0H = the percent of natural area homes built on hillsides (0.425),
HHA = the number of ha per home built on a hillside (1.01 17), %W= the percent of

natural area homes built in wetlands or riparian areas (0.575), 1pm,, = ha converted to

96

protected areas at time t, and 0pm,, = ha of private open land lost to cluster and sprawl
development and protected areas at time t. The initial values for hectares in each land use
type was extracted from city zoning maps (cluster development = 148.92 ha) and land
ownership maps created by the USGS (protected areas = 39197.56 ha, sprawl
development = 5443.02, private open land = 71815.22 ha).
MODEL EVALUATION

We evaluated the model by comparing model predictions of total house numbers
and homes in each land use type to empirical data. The most recent data are from the
2000 US. Census, so we ran 10 year simulations (1990-2000), and compared model
predictions for 2000 to the Census data. Data for total house numbers came fi'om a
census and was not replicated, so we used a modified t-test to compare the mean of
model predictions to the single real-system datum (Grant et al. 1997). We evaluated our
simulation of house numbers in the same way. We were also able to validate the model
by comparing the year it predicted grth slowing development restrictions to the year
those restrictions were implemented. Finally we conducted sensitivity analysis to
evaluate how sensitive model predictions of total open space were to small changes (5%)
in model parameters (Haefner 1997). The sensitivity index equation is as follows:
Sx = relative change (AX/X0) of component X divided by a relative change (AP/P0) in
component P.
SIMULATION EXPERIMENTS

We simulated 20 different development scenarios for estimated total open space.
The baseline scenario was with no growth slowing restrictions, no vacation home ban,

and no cluster development. The remaining 19 scenarios were combinations of grth

97

slowing restrictions or lack thereof, vacation home ban or lack thereof, and 5
implementation rates of cluster development (0%, 25%, 50%, 75%, 100%). We used 30
replicates of 30 year simulations for all simulation experiments. While error propagation
over longer periods can make statistically significant results biologically and socially
meaningless (Grant et al. 1997, Haefner 1997), we did run 100 year simulations (2090) to
determine if the built out scenario for Teton County followed the same pattern observed
after 30 years.

We tested for significant differences in area of open space, and number of homes
built under each scenario using ANOVA (Ott and Longnecker 2001). If ANOVA was
significant we identified groupings and ranked them using the Duncan’s post hoc test.
Finally we combined open space and development scores into a composite score for
ranking development scenarios by their ability to meet both conservation and social
needs. To create the composite score we ranked development scenarios by the mean
open space remaining (1—20) and mean number of homes built (1—20), and then
summed the ranks. An ideal score would equal 40. The composite ranking gave the
number 1 rank to the development scenario/s with the highest composite score and the
worst ranking to the development scenario/s with the lowest composite score.
RESULTS

Model results were not significantly different from census data for total houses or
houses in any of the land use areas (Table 5.2). Political developments occurring in
Teton Valley provide benchmarks for evaluating the gangplank policy portion of the
model. The model predicts immigrants will gain a majority in 2003. The first County

Commissioner supporting growth slowing restrictions, a newcomer, was elected in 2004,

98

and in 2006 residents elected 2 (of 3) County Commissioners who supported growth
slowing restrictions. The next year a 6 month building moratorium was enacted,
effectively stopping all new building permits. These findings support both the growth
slowing restrictions portion of the model and the hypothesized relationship between
immigrant majority and development rate (GSRI,: exponential decline). Total open space
was robust to small changes in most parameters. It was most sensitive to household size
in natural areas (Sx = 0.28), household size in agricultural areas (Sx = 0.67), household
size in residential areas (Sx = 0.14), and immigration rate (Sx = -0.78). These sensitivity
indexes are reasonable because most houses were built in agricultural areas and 20 acre
zoning laws give each house in agricultural areas greater impact on total open space at
the end of the simulation. While similar numbers of natural and residential homes were
built, natural area homes required more land per home than residential homes (1 acre
zoning). Natural area homes occurred in 20 acre zoning areas (wetlands and riparian
areas) and 2.5 acre zoning areas (low elevation hillsides). As one might expect the model
was most sensitive to immigration rate. This variable controlled total population which
in turn drove development and land use change.

Total open space varied by simulation scenario for both the 30 year and
100 year simulations (F = 14086.01 , p < 0.001). Similar rankings of development
scenarios characterized both simulation lengths as well (Tables 5.3, 5.4, 5.5).
Development scenarios with the highest percentage of cluster development were ranked
highest for protecting open space. Development scenarios with 100% cluster
development were ranked 1 and 2 irrespective of gangplank policy or the vacation home

ban. At lower levels of cluster development gangplank policy had a small impact on

99

open space, but the vacation home ban had no impact on rankings (Table 5.3). Time
series results, however, show the impacts of growth slowing restrictions on open space
become progressively less important as the percent of cluster development increases
(Figure 5.3). Finally, cluster development preserves open space in protected areas, but
growth slowing restrictions preserves open space by pushing back development dates. A
baseline simulation with 50% cluster development and a growth slowing restrictions
simulation with 0% cluster development have almost identical total open space
predictions for 2020, 94,410 ha and 93,409 respectively (Figure 5.3). The cluster
development scenario, however, minimizes open space loss by creating 35,593 ha of
protected land. The growth slowing restrictions save an equal amount of open space, but
by keeping it in the developable privately owned open space category.

The amount of development allowed varied by simulation scenario for the 30 year
(F = 422.96, p < 0.001) and 100 year simulations (F = 21.66, p < 0.001). As in the case
of open space preservation, the vacation home ban had little impact on house numbers
(Table 5.4). Growth slowing restrictions created significant reductions in the number of
homes built over both 30 and 100 year time scales (Table 5.4). Cluster development had
little impact on total houses (Table 5.4). Composite rankings generally favored high
levels of cluster development and not implementing a vacation home ban or growth
slowing restrictions (Table 5.5). Scenarios with growth slowing restrictions only ranked
within the top 4 in the 100 year simulations, but even then they were never ranked 1.
DISCUSSION

These results quantify how much cluster development strategies can

improve conservation development in Teton County, and highlight the risk

100

 

 

associated with using majority rule politics to temporarily take development
rights from land owners. Not only did cluster development leave as much open
space as growth slowing restrictions, it protected that open space. Growth
slowing restrictions maintained open space in private developable lands, a
risky strategy.

One might assume that as long as the immigrant majority exists, they
will be able to severely restrict development. That assumption, however,
ignores on critical component of systems theory, the constraints imposed by
higher levels within a system. In the case of Teton County, a development
moratorium was instated only one month after anti-grth immigrants gained a
2 to l majority of the Teton County Commissioners. That moratorium was
repealed by a ruling in a state level court only one month later (Robinson v.
Board of Teton County Commissioners, Case No. CV-07-107). This suggests
the assumed relationship between immigrant majority and grth slowing
restrictions was correct at the level we modeled, but could not operate based on
constraints from the state level. Because conservation development strategies
rely on enlightened and cooperative developers and landowners (Pejchar et al.
2007), immigrants hurt themselves by using a heavy handed political solution
to stop development.

Residents hoping to protect open space with conservation planning
strategies such as cluster development could gain cooperation from developers
and landowners by supporting development strategies that allow for the

greatest profits (Table 5.4). Supporting development in cluster designs would

101

remove 1224 ha of open space over 30 years from the ideal open space
development strategy (Table 5.3, Table 5.4). Without cooperation from
landowners and developers, continuation of status quo development would
results in similar profits for landowners but 37,493 ha less open space. The
cooperative strategy could help land rich but cash poor farmers receive fair
value for their property without threatening open space and wildlife
conservation.

Over the long term, the cooperative strategy would probably be even
more beneficial from an environmental perspective because open space would
be in protected areas. Both status quo development and growth slowing
restrictions leave open space in private hands, so future changes in zoning
(e. g., changing riparian area zoning from 20 acres to 0.5 acres) could allow
incredible increases in house numbers and loss of open space. In fact
composite rankings for growth slowing restrictions development scenarios
(Table 5.5) in the 100 year simulations reflects continued development after
land was either protected or built out in the other scenarios. The only people
who stand to benefit from status quo development are developers and land
speculators who could hope for firture reductions in zoning areas and
subdivision of properties.

CONSERVATION IMPLICATIONS

Our findings highlight the importance of scale in land use planning. Ignoring

state level constraints caused the growth slowing restrictions of immigrants to backfire.

Barring cooperative efforts capable of transcending the division caused by the building

102

moratorium, nearly one half of Teton County would be converted to low density sprawl
in the next 30 years. The results also suggest some form of cluster development provides
means for meeting the needs of those advocating development and those advocating
sustainable wildlife populations. Cluster development, however, is not necessarily
conservation development (Pejchar et al. 2007). Conservation development strategies
must be designed specifically to protect and restore wildlife populations or ecological
services, and be based on property level assessments. Future models incorporating
spatial data with GIS technology could facilitate such efforts, but may prove less intuitive
to stakeholders. When stakeholders do not understand models, acceptance of model
results drops drastically (Grant et al. 1997).

Our model could be used in Teton County or other areas attempting to balance
sustainability and development in two ways. First the model could be used to help
residents cope with the complexity inherent to C HANS. The complexity involved in
systems thinking usually surpasses the limited information processing ability of humans
(Grant et al. 1997). Using and building systems models can help people understand
feedbacks, interconnectedness, nonlinearity, and hierarchy in complex systems (van den
Belt 2004). Second our model could be used as a building block to start mediated
modeling efforts (van den Belt 2004). It could be modified by stakeholders addressing
similar but different issues in similar but different areas as they struggle to meet the needs

of human and non-human C HANS residents.

103

Table 5.1. Relationship used for the growth slowing index. Relationship between
ratio of total population to long term resident population and home construction index
(home construction index * home demand = actual homes built).

 

Total population/long term Home construction indexa
resident population

2 l
2.8 0.44
3.6 0.29
4.4 0.2
5.2 0.12
6 0.05
6.8 0.01
7.6 0.01
8.4 0.01
9.2 0.01
10 0.01

 

 

a The home construction index relies on the logical relationship between voting majority and political
power. Once recent immigrants gain a voting majority (i.e., total population/longer tem resident
population >2) recent immigrants rapidly exert their political will by slowing new home construction.

104

Table 5.2. Comparison of model predictions of house numbers in 2000 to census
estimates of house numbers in 2000. Land use specific estimates were derived by
multiplying the number of houses in Teton County by the percent of houses in that
land use type estimated in the 2004 social survey (see Chapter 4).

 

 

Model Mean (SD) Census Estimate t p

Total houses 2638 (148) 2632 0.21 > 0.8
Natural area 622 (48) 631 0.94 > 0.2
houses

Agricultural 1388 (115) 1370 0.84 > 0.2
area houses

Residential area 627 (52) 631 0.41 > 0.5
houses

 

105

Table 5.3. Differences in total open space after 30 and 100 years respectively based
on growth slowing restrictions, vacation home ban, and 5 levels of cluster
development implementation in Teton County, Idaho. Ranks reflect significantly
different groups identified using the Duncan’s post hoc comparison of means (p <
0.05). A rank of 1 is the strategy leaving the most open space.

 

Development Scenarios Mean 30 Rank Mean 100 Rank
year year

 

No growth slowing
restrictions, no vacation home

ban
0% Cluster 71890 11 39198 10
25% Cluster 80638 10 77819 8
50% Cluster 90047 6 84207 6
75% Cluster 99329 4 92958 4
100% Cluster 108159 2 105911 2

No growth slowing

restrictions, vacation home

ban
0% Cluster 72014 1 1 39198 10
25% Cluster 82060 9 77752 8
50% Cluster 89510 6,7 84243 6
75% Cluster 99303 4 92930 4
100% Cluster 108078 2 105846 2

Growth slowing restrictions,

no vacation home ban

0% Cluster 88340 8 64287 9
25% Cluster 939999 5 78532 7
50% Cluster 98931 4 85768 5
75% Cluster 104030 3 96632 3
100% Cluster 109350 1 107536 1
Growth slowing restrictions,

vacation home ban
0% Cluster 89030 7,8 64746 9
25% Cluster 94137 5 78548 7
50% Cluster 98929 4 86015 5
75% Cluster 104337 3 96885 3
100% Cluster 109382 1 107586 1

 

106

Table 5.4. Differences in total development allowed (i.e., number of homes) after 30 and
100 years respectively based on growth slowing restrictions, vacation home ban, and 5
levels of cluster development implementation in Teton County, Idaho. Ranks reflect
significantly different groups identified using the Duncan’s post hoc comparison of
means (p < 0.05). A rank of l is the strategy allowing the most home building.

 

Development Scenarios

Mean 30 Rank

Mean 100 Rank

 

year
No growth slowing
restrictions, no vacation home
ban
0% Cluster 8532 2 12063 2
25% Cluster 8511 2 10042 5,6
50% Cluster 8564 1, 10528 4,5,6
75% Cluster 8602 1, 14526 1
100% Cluster 8792 1 11975 2
No growth slowing
restrictions, vacation home
ban
0% Cluster 8555 1,2 11490 2,3
25% Cluster 8411 9868 5,6
50% Cluster 8376 11212 2,3,4
75% Cluster 8586 1,2 14560 1
100% Cluster 8541 11605 2,3
Growth slowing restrictions,
no vacation home ban
0% Cluster 5316 3 9918 5,6
25% Cluster 5438 3 9831 6
50% Cluster 5527 3 10737 3,4,5,6
75% Cluster 5430 3 10093 5,6
100% Cluster 5521 3 10714 3,4,5,6
Growth slowing restrictions,
vacation home ban
0% Cluster 5514 .3 10291 4,5,6
25% Cluster 5410 3 9743 6
50% Cluster 5462 3 10404 4,5,6
75% Cluster 5422 3 10390 4,5,6
100% Cluster 5478 3 10839 3,4,5

 

107

Table 5.5. Scores and ranks for development scenarios in Teton County over 30 and 100
year time scales. Scores represent the sum of total open space preserved ranks (derived

from Table 5.3) and total development ranks (derived from Table 5.4). A rank of l is the
best, and worse deve10pment scenarios receive progressively higher rankings.

 

Development Scenarios 30 year 30 100 100
score year year year
rank score rank

 

No growth slowing
restrictions, no vacation home

ban
0% Cluster 15 l3 l9 7
25% Cluster 16 12 ll 9
50% Cluster 25 7 l9 7
75% Cluster 33 2 33 2
100% Cluster 38 1 35 1

No growth slowing

restrictions, vacation home

ban

0% Cluster 18 10 17 8
25% Cluster 16 12 8 11
50% Cluster 18 10 24 4
75% Cluster 31 4 33 2
100% Cluster 32 3 33 2

Growth slowing restrictions,

no vacation home ban

0% Cluster 6 l6 7 12
25% Cluster 14 14 9 10
50% Cluster 22 8 23 5
75% Cluster l9 9 21 6
100% Cluster 28 5 30 3
Growth slowing restrictions,

vacation home ban

0% Cluster l4 14 ll 9
25% Cluster 12 15 9 10
50% Cluster 17 l 1 21 6
75% Cluster l9 9 24 4
100% Cluster 27 6 33 2

 

108

 

Potential Immigrants

 

 

 

 

 

 

 

Developable Land

 

+

 

 

 

 

 

 

 

New lm migrants ‘ +

+a. + {—— _

Development + Land Conversion

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

 

Restrictions to Residential Use

   

 

 

 

 

 

 

 

 

 

 

 

Natives l

 

 

4 Quality of Life 5 _________ j _________

EWinter Range 5
E and Water 5

 

 

Figure 5.1. Conceptual model of land use change in Teton Valley of Idaho and
Wyoming where a. is the population sub model, b. is the home construction sub
model, c. is the land conversion sub model, and d. is the policy module. Dashed
borders indicate system components that are not explicitly simulated. Plus and minus
signs indicate positive and negative relationships respectively.

109

8000 -

7000 -

6000 r

5000 ~

4000 r

3000 ~

2000 ~

1000 ‘

 

 

0 I I I I I I I I I I T I I I I I I I I I I I I I I I r T I I I I I I

«900849%Qe‘le‘te‘be‘bQQQ'thqQ’OPC96}
(9 <9 {5 45 £5 {5 {5 {a <9 (5 <3 (5 <5 (P S” ‘83 ‘§5

Figure 5.2. Annual population estimated for Teton County, Idaho, with pivot point
indicated by line between 1991 and 1992 (US Census).

110

39000

3980
Mai
m 38.25
6
m
P 80.08
S
m 8 80
o. .
o
8.08
3.80
8.08

III.

 

 

 

 

 

.. 1 II II I 393
0.5.2
.. I
I
I I l
.. l
I I 3.x.
/ 0.53-
L I
I
r I I 5:3 @353... «.953 33.2.2.3 co\
r I $23.65 @353 «32.3 «332.03» Quenc-
q—qfiqq_q____q_qqdq_-dquu_dqqu
o m 8 3 mo mm mo

<34 2 932520:

Emfim mu. H35 818 o». 88— 0?: £58 3 H805 095? macaw 9o nonaafim 560358 0». «823 fleece—m
433303 Agmuwflam 59033 3 63285 3.. 2584 awed—0682: 38m 333mm.

111

CHAPTER 6

CONCLUSIONS

112

This dissertation suggests that hypothetico-deductive science (Romesburg
1981) provides no support for the population as the problem perspective for wildlife
conservation (Chapter 2). Research assessing household dynamics over temporal and
spatial scales that matches the scale of population data is needed to tease apart the
influence of household density and population density on wildlife extinction. Until
that research provides scientific support for one alternative, ethics and practicality
should guide management perspectives. A household perspective toward biodiversity
conservation is both more ethical and more practical than a population perspective for
conserving wildlife because it matches social values for positive and negative human
rights and manipulations of home locations and sizes are facilitated by institutions
and laws in most western governments.

Future household research should evaluate how household dynamics (e. g.,
changes in household size, number, and location) influence species endangerment,
and clarify socio-structural determinants of those dynamics. Local scale household
regulation approaches (e. g., zoning) should be scaled up and used in adaptive
management strategies at national and international levels. Cultural and geographic
differences will require different land tenure and development strategies for different
contexts, but that does not preclude larger scale projects. Indeed regional scale
development planning is essential for biodiversity conservation (Pejchar et al. 2007).

Household dynamics regulation carries the added benefit of advocacy that
stresses reciprocal relationships between humans and species endangerment (e. g.,
how home building influences wildlife survival) instead of inimical relationships

(e. g., human existence versus wildlife existence). This means managers should think

113

and talk about household dynamics as both a threat and solution to wildlife
conservation, and abandon the misanthropic tendency to pit human life against
biodiversity conservation in public forums. Using this strategy, wildlife managers
can be advocates for both humans and wildlife in coupled human and natural systems
(CHANS).

The case study research provided several insights about how households
mediate human and nature relationships in CHANS (Chapters 3-5). First they suggest
households act as a social structure mediating the relationship between humans and
nature (Chapter 3). Specifically, the results suggest outdoor recreation participation
has a larger impact on environmental views than previously thought because
households spread the influence of one person’s participation to other people. Within
a given household, correlations between outdoor recreation and environmentally
oriented views tended to permeate to non-recreating household members. If one
person’s activities promote environmentally oriented views, that person’s household
members adopt more environmentally oriented views through social interactions in
the household and without actual participation in the activity. The household effect
appeared to apply for both activities promoting positive environmental views at the
individual level (bird watching) and for activities promoting negative environmental
views at the individual level (ATV use). The results suggested different relationships
between outdoor recreation and environmentally oriented views based on whether
someone lives in multi-person or single-person households.

These findings support Bright and Porter’s (2001) call for research addressing

the social factors mediating the outdoor recreation and environmental concern

114

relationship. As one of the most basic social units, households provide a logical place
to begin this effort. Future research should address how household dynamics mediate
the relationship between outdoor recreation and environmental views. Specific
emphasis should be placed on single-person households because they are increasing
in prevalence and household size is decreasing globally (Liu et al. 2003). The
differences between multi-person and single-person households will become more
important for future conservation efforts as the number of single person households
continues to grow. In our case, answering why ATV use and hunting in multi-person
households correlated with less environmentally oriented views, but the same
activities had no effect on environmental views in single-person households would
represent the beginning of efforts to make these sectors of outdoor recreation more
environmentally oriented. Ideological changes will not make all forms outdoor
recreation environmentally benign, but increasing the number of environmentally
oriented participants should decrease the prevalence of environmentally damaging
forms of recreation and environmentally irresponsible behavior in general.

Future qualitative research will help illuminate how social dynamics in
households mediate the relationship between environmental worldviews and outdoor
recreation. This type of research is essential for addressing unanswered questions
about process and context (Lincoln and Guba 1985). Participant observation and in
depth interviews would allow researchers to identify how recreation activities of one
household member influence other household members. This type of research
strategy would also allow non-participating household members to describe the extent

they feel their own environmental views are influenced by outdoor recreation of other

115

household members. For example, this approach could illuminate the role of media,
cultural stereotypes, and story telling in explaining why non-hunters in hunting
households held less environmentally oriented views than the hunters in their own
homes.

The case studies also demonstrate how households operate as physical objects
mediating the relationship between social and natural entities within CHANS
(Chapter 4). Social relationships dictate where people choose to place a new home,
but the home is a physical object with direct impacts on the natural system. This
direct impact makes home location decisions an appealing behavior for studying
CHANS. Environmentally conscious decisions in many other realms (e. g., home
appliances, food consumed, voting, donations, family planning, activism, and
transportation) do not have direct impacts on wildlife conservation. Our findings
suggest well educated, older, and more environmentally oriented people were more
likely to choose household locations within environmentally sensitive areas (e. g.,
wetlands, riparian areas, hillsides). Further, since those immigrants had the smallest
household sizes of any group, their impacts were magnified by having more
households per person.

Based on these results, framing wildlife conservation in terms of household
impacts promotes social justice, and reduces the validity of hypocrisy allegations
levied against the environmental community. A household perspective shifts the
burden of conservation from the backs of the uneducated and poor to the highly
educated and powerful and expects educated environmentalists to sacrifice what they

want (e. g., a home on a river, on a mountain side, or on fragile desert soils) before

116

expecting poor or uneducated individuals to sacrifice what they need for basic living
(e. g., heating, health care, college education for their children) in the name of
conservation. Global wildlife conservation, however, requires changes in household
locations and size on a global level. If or when older well educated environmentalists
change their behaviors, they will gain moral authority to lead global efforts to control
household impacts on wildlife conservation. To build that authority, conservation
leaders must make household decisions reflecting what they advocate.

Even with leadership and adequate information, wildlife conservation in
CHANS presents a formidable challenge. Part of the problem stems from the
complexity inherent to CHANS. The complexity involved in even the simplest
systems usually surpasses the limited information processing ability of humans (Grant
et al. 1997). Humans also use biases (e.g., anchoring and adjustment,
representativeness heuristic, availability heuristic) to reduce mental effort (Kahneman
et al. 1982, Hogarth 1987). People tend to think linearly, rather than in causal nets
(Domer 1980) and have difficulty recognizing imbalance paths (e. g., thresholds) and
feedback loops (Axelrod 1976). At the group level these biases against systems
thinking are often exacerbated by the forces of group think (Janis 1972, Janis and
Mann 1977) and selective memory related to multiple perspectives (Peterson and
Horton 1995). Systems theory presents an approach counterintuitive to a culture
indoctrinated with deductive reasoning, reductionism, and solving problems by
breaking them down and isolating their parts (Senge 1990, Ackoff 1999, Vennix

1999).

117

Using and building systems models can help people understand feedbacks,
interconnectedness, nonlinearity, and hierarchy in complex systems (van den Belt
2004). The systems model developed in this dissertation (Chapter 5) quantifies how
much cluster development strategies can improve conservation development in Teton
County, and highlights the risk associated with using majority rule politics to
temporarily take development rights from land owners. Cluster development left as
much open space as growth slowing restrictions, and protected the open space.
Growth slowing restrictions did reduce the number of homes built, while cluster
development did not significantly alter the number from status quo simulations.

Recent developments in Teton County, however, demonstrate the risk
associated with trying to have your cake and eat it too. A building moratorium
implemented in 2007 by recent immigrants who also supported conservation planning
(e. g., cluster development) only lasted one month before a state judge repealed it
(Robinson v. Board of Teton County Commissioners, Case No. CV-07-107). The
building moratorium did, however, help foster alliances between landowners and land
speculators (who paid for the lawyers in the case), and isolated landowners from
recent immigrants hoping to protect open space (personal communication, Wayne
Peterson, Teton County Resident).

This outcome highlights the importance of scale in land use planning.
Ignoring state level constraints caused the growth slowing restrictions imposed by
recent immigrants to backfire. Unless those immigrants can break the new alliance
between landowners and land speculators and forge cooperative land use planning,

nearly one half of Teton County will be converted to low density sprawl in the next

118

30 years (Chapter 5). Cluster development provides means for meeting the goals of
both landowners and those promoting conservation goals (Chapter 5). Our model
could be used in Teton County or other areas attempting to balance development and
conservation as a building block for mediated modeling efforts (van den Belt 2004).
Stakeholders addressing similar but different issues in similar but different areas
could modify the model to project land use under their own set of policy scenarios
and contextual constraints.

This dissertation was grounded in the assumption that the tradition of
identifying human society and the natural environment as fundamentally separate
systems is the most fiindamental challenge facing wildlife conservation (Leopold
1949, World Commission on Environment and Development 1987, Busch 1996,
Latour 2004). Human society cannot conserve wildlife without understanding the
relationship between material processes in nature and socio-political practices, and
then applying that understanding in the policy arena. Integrating society and nature
within CHANS, however, can prove ethically and socially problematic (Chapter 2).
This and future research focusing on households as the nexus between human and
natural systems provides an avenue for CHANS research that helps fulfill the needs

of humans and wildlife.

119

LITERATURE CITED

Ackoff, R. L. 1999. Ackoffs best : his classic writings on management. Wiley, New
York, USA.

Ajzen, 1. 1991. The Theory of Planned Behavior. Organizational Behavior and
Human Decision Processes 50:179-21 1.

. 2001. Nature and operation of attitudes. Annual Review of Psychology 52:27-
58.

An, L., J. G. Liu, Z. Y. Ouyang, M. Linderman, S. Q. Zhou, and H. M. Zhang. 2001.
Simulating demographic and socioeconomic processes on household level and
implications for giant panda habitats. Ecological Modelling 140:31-49.

An, L., F. Lupi, J. G. Liu, M. A. Linderman, and J. Y. Huang. 2002. Modeling the
choice to switch from fuelwood to electricity Implications for giant panda
habitat conservation. Ecological Economics 42:445-457.

Arendt, R. G. 1992. "Open space" zoning: what it is and why it works. Flaming
Commisioners Journal 5: l -9.

Argyris, C. 1992. On organizational learning. Blackwell, Cambridge, MA, USA.

Argyris, C., and D. A. Schon. 1978. Organizational learning: a theory of action
perspective. Addison-Wesley, Reading, MA, USA.

Austin, L. 2006. Prof. criticized for overpopulation view. The Associated Press
Tuesday, April 4, 2006; 2 PM GMT.

Axelrod, R. 1976. Structure of deision: the cognitive maps of political elites. Princton
University Press, Princeton, NJ, USA.

Bailey, R. G. 1995. Description of ecoregions of the United States. USDA Forest
Service, Washington DC, USA.

Bargh, J. A., S. Chaiken, R. Govender, and F. Pratto. 1992. The generality of the
automatic attitude activation effect. Journal of Personality and Social
Psychology 62:893-912.

Bargh, J. A., S. Chaiken, P. Raymond, and C. Hymes. 1996. The automatic evaluation
effect: Unconditional automatic attitude activation with a pronunciation task.
Journal of Experimental Social Psychology 32: 1 04-128.

120

Benton, T., and I. Craib. 2001. Philosopy of social science: the philisophical
foundations of social thought. Palgrave, New York, USA.

Berger, J. 2004. The last mile: How to sustain long-distance migration in mammals.
Conservation Biology 18:320-331.

Berkowitz, A. R., C. H. Nilon, and K. S. Hollweg, editors. 1999. Understanding
urban ecosystems: a new frontier for science and education. Springer-Verlag,
New York, USA.

Beyers, W. B., and P. B. Nelson. 2000. Contemporary development forces in the
nonmetropolitan west: new insights from rapidly growing communities.
Journal of Rural Studies 16:459-474.

Botkin, D. 1990. Discordant harmonies: a new ecology for the 21st century. Oxford,
New York, USA.

Braithwaite, V. 1998. The value balance model of political evaluations. British
Journal of Psychology 89:223-247.

Bright, A. D., and R. Porter. 2001. Wildlife-related recreation, meaning, and
environmental concern. Human Dimensions of Wildlife 6:259-276.

Busch, L. 1996. Bringing nature back in: principles for a new social science of nature.
Centennial Review 40:491-501.

Carr, D. L. 2004. Proximate population factors and deforestation in tropical
agricultural frontiers. Population and Environment 25:585-612.

Catton, W. R., and R. E. Dunlap. 1978. Paradigms, Theories, and Primacy of Hep-
Nep Distinction. American Sociologist 13:256-259.

Chapin, F. S., E. S. Zaveleta, V. T. Eviner, R. L. Naylor, P. M. Vitousek, H. L.
Reynolds, D. U. Hooper, S. Lavorel, O. E. Sala, S. E. Hobbie, M. C. Mack,

and S. Diaz. 2000. Consequences of changing biodiversity. Nature 4052234-
242.

Checkland, P. 1981. Systems thinking, Systems practice. Wiley, New York.

Chertow, M. R. 2001. The IPAT equation and its variants: Changing views of
technology and environmental impact. Journal of Industrial Ecology 4:13-29.

Clayton, S., and A. Brook. 2005. Can psychology help save the world? A model for

conservation psychology. Analyses of Social Issues and Public Policy 5:87-
102.

121

Cohen, J ., and P. Cohen. 1983. Applied Multiple Regression/Correlation for the
Behavioral Sciences 2nd ed. Lawrence Erlbaum Associates, London, England.

Conway, T. M., and R. G. Lathrop. 2005. Modeling the ecological consequences of
land-use policies in an urbanizing region. Environmental Management
35:278-291.

Cordell, H. K., C. J. Betz, G. Green, and M. Owens. 2005. Off-highway vehicle
recreation in the United States: a national report from the National Survey on
Recreation and the Environment. US. Forest Service Southern Research
Station, Asheville, NC, USA.

Cordell, H. K., C. J. Betz, and G. T. Green. 2002. Recreation and the environment as

Cultural dimensions in contemporary American society. Leisure Sciences
24:13-41.

Crump, J. R. 2003. Finding a place in the country - Exurban and suburban
development in Sonoma County, California. Environment and Behavior
35: 187-202.

Czech, B., P. R. Krausman, and P. K. Devers. 2000. Economic associations among
causes of species endangerment in the United States. Bioscience 50:593-601.

Diamond, J. 2005. Collapse: how societies choose to fail or succeed. Penguin, New
York, USA.

Diamond, J. M. 1986. The design of a nature reserve system for Indonesian New
Guinea. Pages 485-503 in S. M. E., editor. Conservation biology: the science
of scarcity and diversity. Sinauer, Sunderland, MA, USA.

Dietz, T., and E. A. Rosa. 1994. Rethinking the environmental impacts of population,
affluence and technology. Human Ecology Review 12277-300.

Dietz, T., P. C. Stern, and G. A. Guagrrano. 1998. Social structural and social
psychological bases of environmental concern. Environment and Behavior
30:450-471.

Dillman, D. A. 2000. Mail and Internet Surveys: The Tailored Design Method. Wiley,
New York.

Dompka, V., editor. 1996. Human population, biodiversity and protected areas:
Science and policy issues. American Association for the Advancement of

Science, Washington, DC.

Domer, D. 1980. On the difficulties people have in dealing with complexity.
Simulation and Games 11:87-106.

122

Dowie, M. 2005. Conservation Refugees. Orion November/December:16-27.

Dunlap, R. E., and R. B. Heffeman. 1975. Outdoor recreation and environmental
concern: an empirical examination. Rural Sociology 40: 18-30.

Dunlap, R. E., and W. Michelson, editors. 2002. Handbook of environmental
sociology. Greenwood, Westport, CT.

Dunlap, R. E., and K. D. Van Liere. 1978. The "New Environmental Paradigm": a
proposed measuring instrument and preliminary results. Journal of
Environmental Education 9: 10-1 9.

Dunlap, R. E., K. D. Van Liere, A. G. Mertig, and R. E. Jones. 2000. Measuring

endorsement of the new ecological paradigm: A revised NEP scale. Journal of
Social Issues 56:425-442.

Eagly, A. H., and S. Chaiken. 1993. The psychology of attitudes. Harcourt Brace,
Fort Worth, TX, USA.

Ehrlich, P. R. 1971. The population bomb. Ballantine Books, New York, USA.
. 2003. Get off the train and walk. Conservation Biology 17:352-353.

Emanuel, K. 2005. Increasing destructiveness of tropical cyclones over the past 30
years Nature 436:686-688.

Enck, J. W., D. J. Decker, and T. L. Brown. 2000. Status of hunter recruitment and
retention in the United States. Wildlife Society Bulletin 28:817-824.

Fazio, R. H. 1990. Multiple processes by which attitudes guide behavior - the mode
model as an integrative framework. Advances in Experimental Social
Psychology 23:75-109.

Feather, N. T. 2002. Values and value dilemmas in relation to judgments concerning
outcomes of an industrial conflict. Personality and Social Psychology Bulletin
28:446-459.

Ferry, L. 1995. The new ecological order. University of Chicago Press, Chicago, IL,
USA.

Fishbein, M. 1963. An investigation of the relationships between beliefs about an
object and the attitude toward that object. Human Relations 16:233-240.

F ishbein, M., and I. Ajzen. 1975. Belief, attitude, intention, and behavior: and
introduction to theory and research. Addison-Wesley, Reading, MA, USA.

123

Forester, D. J ., and G. E. Machlis. 1996. Modeling human factors that affect the loss
of biodiversity. Conservation Biology 10:1253-1263.

Fortmann, L., and J. Kusel. 1990. New voices, old beliefs - forest environmentalism
among new and long—standing rural residents. Rural Sociology 55:214-232.

F reyfogle, E. T. 2003. Conservation and the culture war. Conservation Biology
17:354-355.

Freyfogle, E. T., and J. L. Newton. 2002. Putting science in its place. Conservation
Biology 16:863-873.

Frumkin, P. 2001. Beyond toxicity - Human health and the natural environment.
American Journal of Preventive Medicine 20:234-240.

Geisler, C. C., O. B. Martinson, and E. A. Wilkening. 1977. Outdoor Recreation and
Environmental Concern - Restudy. Rural Sociology 42:241-249.

Graber, E. E. 1974. Newcomers and Oldtimers - Growth and Change in a Mountain
Town. Rural Sociology 39:504-513.

Graedel, T., and B. Allenby. 1995. Industrial ecology. Prentice Hall, Englewood
Cliffs, NJ, USA.

Grant, W. E., E. K. Pederson, and S. L. Marin. 1997. Ecology and natural resource
management: systems analysis and simulation. Wiley and Sons, New York,
USA.

Greeley, A. 1993. Religion and Attitudes toward the Environment. Journal for the
Scientific Study of Religion 32:19-28.

Groves, R. M. 1989. Survey errors and survey costs. Wiley, New York, USA.

Guagnano, G. A., P. C. Stern, and T. Dietz. 1995. Influences on attitude-behavior

relationships - a natrual experiment with curbside recycling. Environment and
Behavior 27:699-718.

Gunningham, N., R. A. Kagan, and D. Thornton. 2003. Shades of green: business,
regulation, and enviromnent. Stanford University Press, Stanford, CA, USA.

Gutmann, M. P., S. Pullum, G. A. Cunfer, and D. Hagen. 1998. The Great Plains
Population and Environment Database. Available online at
http://enceladus.icpsr.umich.edu/gpdb/begin.asp (accessed 9 December 2003).

124

Haefirer, J. W. 1997. Modeling biological systems: principles and applications.
Chapman & Hall, New York, USA.

Halper, T. 2003. Positive rights in a republic of talk: a survey and a critique. Kluwer,
Norwell, MA, USA.

Harcourt, A. H., S. A. Parks, and R. Woodroffe. 2001. Human density as an influence
on species/area relationships: double jeopardy for small Afiican reserves?
Biodiversity and Conservation 10:1011-1026.

Hardin, G. 1993. Living Within Limits: ecology, economics, and population taboos.
Oxford University Press, New York, USA.

Hegel, G. W. F. 1977. Phenomenology of Spirit. Oxford University Press, New York,
USA.

Heimlich, R. E., and W. D. Anderson. 2001. Development at the urban fringe and
beyond: impacts on agriculture and rural land. Agricultural economic report
803. Economic Research Service, US. Department of Agriculture,
Washington, DC, USA.

Helvarg, D. 1994. Title The war against the Greens : the "Wise-Use" Movement, the
new right and anti-environmental violence. Sierra Club Books, San Francisco,

CA, USA.

Hilton-Taylor, C. 2000. 2000 IUCN red list of threatened species International Union
for the Conservation of Nature and Natural Resources, Gland, Switzerland.

Hitler, A. 1943. Mein kampf. Houghton Mifflin company, Boston, MA, USA.
Hogarth, R. 1987. Judgement and choice. Second edition. Wiley, New York, USA.

Holmes, H. B., B. B. Hoskins, and M. Gross. 1980. Birth control and controlling
birth: women-centered perspectives. Hurnana Press, Clifton, NJ, USA.

Horwitz, M. J. 1992. The transformation of American law: 1870-1960 the crisis of
legal orthodoxy. Oxford University Press, New York, USA.

Hospers, J. 2005. What libertarianism is. Pages 321-389 in J. E. White, editor.
Contemporary moral problems. Thomson Wadsworth,Belmont, CA, USA.

Hunter, L. M., and M. B. Toney. 2005. Religion and attitudes toward the

environment: a comparison of Mormons and the general US population.
Social Science Journal 42:25-38.

125

Huston, A. C., G. J. Duncan, R. Granger, J. Bos, V. McLoyd, R. Mistry, D. Crosby,
C. Gibson, K. Magnuson, J. Romich, and A. Ventura. 2001. Work-based
antipoverty programs for parents can enhance the school performance and

social behavior of children. Child Development 72:318-336.

Idaho Fish & Game. 2005. Teton River trout population depressed, but impoving.
Upper Snake Region annual fisheries newsletter 1:1-6.

IUCN. 2003. United Nations List of Protected Areas. UNEP World Conservation
Monitoring Centre.

Jackson, E. L. 1986. Outdoor recreation participation and attitudes towards the
environment. Leisure Studies 5:1-23.

Jacobs, H. M., editor. 1998. Who owns America?: social conflict over property rights.
University of Wisconsin Press, Madison, WI, USA.

Janis, I. L. 1972. Victims of groupthink, a psychological study of foreign-policy
decisions and fiascoes. Houghton Mifflin, Boston, MA, USA.

Janis, 1. L., and L. Mann. 1977. Decision making: a psychological analysis of
conflict, choice, and commitment. The Free Press, New York, USA.

Johnson, C. Y., J. M. Bowker, and H. K. Cordell. 2004. Ethnic variation in
environmental belief and behavior: An examination of the new ecological
paradigm in a social psychological context. Environment and Behavior
36: 1 57-1 86.

Johnson, K. M. 2003. Unpredictable Directions of Population Growth and Migration.
Pages 19-31 in D. L. Brown, and L. Swanson, editor. Challenges for Rural
America in the let Century. Pennsylvania State University Press,University
Park, PA, USA.

Johnson, K. M., and G. V. Fuguitt. 2000. Continuity and change in rural migration
patterns, 1950-1995. Rural Sociology 65:27-49.

Jones, R. E., and R. E. Dunlap. 1992. The Social Bases of Environmental Concern -
Have They Changed over Time. Rural Sociology 57:28-47.

Jones, R. E., J. M. Fly, J. Talley, and H. K. Cordell. 2003. Green migration into rural
America: The new frontier of environmentalism? Society & Natural
Resources 16:221-238.

Kahneman, D., P. Slovic, and A. Tversky. 1982. Judgement under uncertainty:
heuristics and biases. Cambridge University Press, Cambridge.

126

Kanagy, C. L., and F. K. Willits. 1993. A Greening of Religion - Some Evidence
from a Pennsylvania Sample. Social Science Quarterly 74:674-683.

Kant, 1., editor. 1873. Foundation of the metaphysics of morals. T. K. Abbott,
translator, 1949. Bobbs-Merrill Co. New York, New York, USA.

Kerr, J. T., and D. J. Currie. 1995. Effects of human activity on global extinction risk.
Conservation Biology 9:1528-1538.

Kie, J. G., and B. Czech. 2000. Mule and black-tailed deer. in S. Demarais, andP. R.
Krousman, editors. Ecology and management of large mammals in North
America. Prentice Hall,Upper Saddle River, NJ, USA.

Kirkland, G. L., and R. S. Ostfeld. 1999. Factors influencing variation among states
in the number of federally listed mammals in the United States. Journal of
Mammalogy 80:71 1-719.

Kramer, R. C., C. van Schaik, and J. Johnson, editors. 1997. Last Stand: Protected
Areas and the Defense of Tropical Biodiversity. Oxford University Press,
New York, USA.

Krebs, C. J. 2000. Hypothesis testing in ecology. Pages 1-14 in a. T. F. F. L. Boitani,

editor. Research techniques in animal ecology. Columbia University Press,
New York, USA.

Latour, B. 2004. Politics of nature: how to bring the sciences into democracy.
Harvard University Press, Cambridge, MA, USA.

Lavine, H., C. J. Thomsen, M. P. Zanna, and E. Borgida. 1998. On the primacy of
affect in the determination of attitudes and behavior: the moderating role of

affective-cogrritive ambivalence. Journal of Experimental Social Psychology
34:398-421 .

Layman, G. C., and T. M. Carsey. 2002. Party polarization and "conflict extension" in
the American electorate. American Journal of Political Science 46:786-802.

Leopold, A. 1949. A sand county almanac and sketches here and there. Oxford
University Press, London, UK.

Light, A., and H. Rolston, III, editors. 2003. Environmental ethics: an anthology.
Blackwell Publishing, Malden, MA, USA.

Lincoln, Y. S., and E. G. Guba. 1985. Naturalistic inquiry. Sage, Thousand Oaks, CA,
USA.

127

Liu, J ., T. Dietz, S. R. Carpenter, C. Folke, M. Alberti, C. L. Redman, S. H.
Schneider, E. Ostrom, A. N. Pell, J. Lubchenco, W. W. Taylor, Z. Ouyang, P.
Deadman, T. Kratz, and W. Provencher. 2007. Coupled human and natural
systems. Ambio 36: In Press.

Liu, J. G. 2001. Integrating ecology with human demography, behavior, and
socioeconomics: Needs and approaches. Ecological Modelling 140:1-8.

Liu, J. G., G. C. Daily, P. R. Ehrlich, and G. W. Luck. 2003. Effects of household
dynamics on resource consumption and biodiversity. Nature 421 :530-533.

Liu, J. G., J. B. Dunning, and H. R. Pulliam. 1995. Potential Effects of a Forest
Management Plan on Bachman Sparrows (Aimophila-Aestivalis) - Linking a
Spatially Explicit Model with GIS. Conservation Biology 9:62-75.

Liu, J. G., M. Linderman, Z. Y. Ouyang, L. An, J. Yang, and H. M. Zhang. 2001.
Ecological degradation in protected areas: The case of Wolong Nature
Reserve for giant pandas. Science 292:98-101.

Lopez, R. R., N. J. Silvy, R. N. Wilkins, P. A. Frank, M. J. Peterson, and M. N.
Peterson. 2004. Habitat use patterns of Florida Key deer: Implications of
urban development. Journal of Wildlife Management 68:900-908.

Malthus, T. R. [1798] 1970. An essay on the principle of population and a summary
view of the principle of population. Penguin, Harmondsworth, Middlesex,
UK.

Mange], M., L. M. Talbot, G. K. Meffe, M. T. Agardy, D. L. Alverson, J. Barlow, D.
B. Botkin, G. Budowski, T. Clark, J. Cooke, R. H. Crozier, P. K. Dayton, D.
L. Elder, C. W. Fowler, S. Funtowicz, J. Giske, R. J. Hofinan, S. J. Holt, S. R.
Kellert, L. A. Kimball, D. Ludwig, K. Magnusson, B. S. Malayang, C. Mann,
E. A. Norse, S. P. Northridge, W. F. Perrin, C. Perrings, R. M. Peterrnan, G.
B. Rabb, H. A. Regier, J. E. Reynolds, K. Sherman, M. P. Sissenwine, T. D.
Smith, A. Starfield, R. J. Taylor, M. F. Tillman, C. Tofi, J. R. Twiss, J. Wilen,
and T. P. Young. 1996. Principles for the conservation of wild living
resources. Ecological Applications 6:338-362.

Mariola, M. J. 2005. Losing ground: farmland preservation, economic utilitarianism,
and the erosion of the agrarian ideal. Agriculture and Human Values 22:209-
223.

Mathews, E., and A. Hammond. 1999. Critical consumption trends and implications:
degrading earth's ecosystems. World Resources Institute, Washington, DC,
USA.

128

McConnell, A. R., J. M. Leibold, and S. J. Sherman. 1997. Within-target illusory
correlations and the formation of context-dependent attitudes. Journal of
Personality and Social Psychology 73:675-686.

McKee, J. K. 2001. Faunal turnover rates and mammalian biodiversity of the late
Pliocene and Pleistocene of eastern Afiica. Paleobiology 27:500-511.

McKee, J. K., P. W. Sciulli, C. D. Fooce, and T. A. Waite. 2004. Forecasting global
biodiversity threats associated with human population growth. Biological
Conservation 1 15:161-164.

Mertig, A. G., R. E. Dunlap, and D. E. Morrison. 2002. The environmental movement
in the United States. Pages 448-481 in R. E. Dunlap, andW. Michelson,
editors. Handbook of environmental sociology. Greenwood,Westport, CT,
USA.

Millennium Ecosystem Assessment. 2005. Synthesis Report. Island Press,
Washington, DC, USA..

Montmarquet, J. A. 1989. The idea of agrarianism: from hunter-gatherer to agrarian
radical in Western culture. University of Idaho Press, Moscow, ID, USA.

Moskowitz, E. H., and B. Jennings. 1996. Coerced contraception?: moral and policy
challenges of long-acting birth control. Georgetown University Press,
Washington, DC, USA.

Murphy, N. 1999. Darwin, social theory, and the sociology of scientific knowledge.
Zygon 34:573-600.

Nooney, J. G., E. Woodrum, T. J. Hoban, and W. B. Clifford. 2003. Environmental
worldview and behavior - Consequences of dimensionality in a survey of

North Carolinians. Environment and Behavior 35:763-783.

Nord, M., A. E. Luloff, and J. C. Bridger. 1998. The association of forest recreation
with environmentalism. Environment and Behavior 30:235-246.

Norton, B. G., and B. Harmon. 1997. Environmental values: a place based theory.
Environmental Ethics 19:227-245.

O'Neill, O., and W. Ruddick. 1979. Having children: philosophical and legal
reflections on parenthood. Oxford University Press, New York, USA.

Orr, D. W. 2003. Walking north on a southbound train. Conservation Biology 17:348-
351.

129

Ostling, R. N., and J. K. Ostling. 1999. Mormon America: The Power and the
Promise. Harper, New York, USA.

Ostrom, E. 1990. Governing the commons: the evolution of institutions for collective
action. Cambridge University Press, New York, USA.

Ott, R. L., and M. Longnecker. 2001. An Introduction to Statistical Methods and Data
Analysis 5th edition. Duxbury, Belmont, California, USA.

Pejchar, L., P. M. Morgan, M. R. Caldwell, C. Palmer, and G. C. Daily. 2007.
Evaluating the potential for conservation development: Biophysical,
economic, and institutional perspectives. Conservation Biology 21 :69—78.

Peterson, M. N. 2004. An approach for demonstrating the social legitimacy of
hunting. Wildlife Society Bulletin 32:310-321.

Peterson, M. N., S. A. Allison, M. J. Peterson, T. R. Peterson, and R. R. Lopez. 2004.
A tale of two species: Habitat conservation plans as bounded conflict. Journal
of Wildlife Management 68:743-761.

Peterson, M. N., A. G. Mertig, and J. G. Liu. 2006a. Effects of zoonotic disease
attributes on public attitudes towards wildlife management. Journal of
Wildlife Management 70: 1746-1753.

. 2006b. Effects of zoonotic disease attributes on public attitudes towards
wildlife management. Journal of Wildlife Management 70: 1746-1 753.

Peterson, M. N., M. J. Peterson, and T. R. Peterson. 2005. Conservation and the myth
of consensus. Conservation Biology 19:762-767.

Peterson, M. N., M. J. Peterson, T. R. Peterson, and J. G. Liu. 2007. A household
perspective for biodiversity conservation. Journal of Wildlife Management
71:1243-1248.

Peterson, M. N., T. R. Peterson, M. J. Peterson, R. R. Lopez, and N. J. Silvy. 2002.
Cultural conflict and the endangered Florida Key deer. Journal of Wildlife
Management 66:947-968.

Peterson, T. R. 1990. Jefferson's yeoman farmer as frontier hero: a self defeating
mythic structure. Agriculture and Human Values 7:9-19.

Peterson, T. R., and C. C. Horton. 1995. Rooted in the soil: How understanding the

perspectives of landowners can enhance the management of environmental
disputes. The Quarterly Journal of Speech 81:139-156.

I30

Peterson, T. R., M. N. Peterson, M. J. Peterson, S. A. Allison, and D. Gore. 2006c. To
Play the Fool: Can Environmental Conservation and Democracy Survive

Social Capital? Communication and Critical/Cultural Studies 32116-140.

Philip, M., and L. Thomas. 1986. Planned parenthood in Europe: a human rights
perspective. Croom Helm in association with International Planned
Parenthood Federation, Europe Region, London, UK.

Pinhey, T. K., and M. D. Grimes. 1979. Outdoor recreation and environmental
concern: a reexamination of the Dunlap-Heffeman thesis. Leisure Sciences
2:1-11.

Pretty, J. 2003. Social capital and the collective management of resources. Science
302:1912-1914.

Price, M. L., and D. C. Clay. 1980. Structural Disturbances in Rural Communities -
Some Repercussions of the Migration Turnaround in Michigan. Rural
Sociology 45:591-607.

Radeloff, V. C., R. B. Hammer, S. 1. Stewart, J. S. Fried, S. S. Holcomb, and J. F.
McKeefry. 2005. The wildland-urban interface in the United States.
Ecological Applications 15:799-805.

Ragin, C. C., and H. S. Becker, editors. 1992. What is a case? exploring the
foundations of social inquiry. Cambridge University Press, New York.

Rasker, R., and A. J. Hansen. 2000. Natural ammenities and population growth in the
Greater Yellowstone region. Human Ecology Review 7:30-40.

Redman, C. L. 1999. Human dimensions of ecosystem studies. Ecosystems 2:296-
298.

Riebsame, W. E. 1997. Atlas of the new West: portrait of a changing region. W. W.
Norton, New York, New York, USA.

Romesburg, H. C. 1981. Wildlife Science - Gaining Reliable Knowledge. Journal of
Wildlife Management 45:293-313.

Root, T., D. MacMynowski, M. Mastrandrea, and et a1. 2005. Human-modified
temperatures induce species changes: Joint attribution. Proc. Natl. Acad. of
Sci. USA 10227465-7469.

Root, T., J. Price, K. Hall, S. Schneider, C. Rosenzweig, and J. Pounds. 2003.
Fingerprints of global warming on wild animals and plants. Nature 421:57-60.

131

Rutledge, D. T., C. A. Lepczyk, J. L. Xie, and J. G. Liu. 2001. Spatiotemporal
dynamics of endangered species hotspots in the United States. Conservation
Biology 15:475-487.

Saunders, C. D., A. T. Brook, and O. E. Myers. 2006. Using psychology to save
biodiversity and human well-being. Conservation Biology 20:702-705.

Schnaiberg, A., and K. A. Gould. 1994. Environment and society: the enduring
conflict. St. Martin's Press, New York, USA.

Schwartz, M. W. 2006. How conservation scientists can help develop social capital
for biodiversity. Conservation Biology 20: 1 550-1 552.

Senge, P. 1990. The fifth discipline: the art and practice of the learning organization.
Doubleday, New York, USA.

Shumway, J. M., and J. A. Davis. 1996. Nonmetropolitan population change in the
Mountain West: 1910-1995. Rural Sociology 61:513-529.

Shumway, J. M., and S. M. Otterstrom. 2001. Spatial patterns of migration and
income change in the mountain west: The dominance of service-based,
amenity-rich counties. Professional Geographer 53:492-502.

Sinclair, A. R. E. 1991. Science and the Practice of Wildlife Management. Journal of
Wildlife Management 55:767-773.

Skovlin, J. M. 1982. Habitat requirement and evaluations. Pages 369-413 in J. W.
Thomas, andD. E. Toweill, editors. Elk of North America: ecology and
mangement. Wildlife Management Institute, Harrisburg, PA, USA.

Smith, M. D., and R. S. Krannich. 2000. "Culture clash" revisited: newcomer and
longer-terrn residents' attitudes towards land use, development, and

environmental issues in rural communities in the Rocky Mountain west. Rural
Sociology 65:396-421.

Soulé, M. E. 1983. What do we really know about extinction? Pages 111-124 in C.
Schonewald-Cox, editor. Genetics and conservation. Benjamin Cummings,
Reading, MA, USA.

. 1986. Conservation biology and the "real world". Pages 1-12 in M. E. Soulé,
editor. Conservation biology: the science of scarcity and diversity. Sinauer

Associates, Sunderland, MA, USA.

Stern, P. C. 2000. Psychology and the science of human-environment interactions.
American Psychologist 55:523-530.

132

Stern, P. C., T. Dietz, and G. A. Guagnano. 1995. The New Ecological Paradigm in
Social-Psychological Context. Environment and Behavior 27:723-743.

Tarrant, M. A., and G. T. Green. 1999. Outdoor recreation and the predictive validity
of environmental attitudes. Leisure Sciences 21 :17-30.

Theodori, G. L., A. E. Luloff, and F. K. Willits. 1998. The association of outdoor
recreation and environmental concern: Reexamining the Dunlap-Heffeman
thesis. Rural Sociology 63:94-108.

Thomas, D. L., and E. J. Taylor. 1990. Study Designs and Tests for Comparing
Resource Use and Availability. Journal of Wildlife Management 54:322-330.

Thompson, P. B. 2003. Expanding the conservation tradition: the agrarian vision.
Pages 77-92 in B. A. Minteer, and R. E. Manning, editor. Reconstructing
conservation: finding common ground. Island Press, Washington D. C., USA.

Thompson, R. H. 2004. Overcoming barriers to ecologically sensitive land
management - Conservation subdivisions, green developments, and the
development of a land ethic. Journal of Planning Education and Research
24:141-153.

US. Fish and Wildlife Service. 2005. National Wetlands Inventory Data 1979-1994.
St. Petersburg, FL, USA.

United Nations. 1948. Universal declaration of human right. General assembly
resolution 217 A (III). http://www.unhchr.ch/udhr/index.htm.

United Nations Centre for Human Settlements. 2001. Citites in a globalizing world:
global report on human settlements. Earthscan, London, UK.

United States Department of the Interior and United States Department of Commerce.
2002. 2001 national survey of fishing, hunting, and wildlife-associated
recreation. United States Government Printing Office, Washington, DC,
USA.

van den Belt, M. 2004. Mediated modeling: a system dynamics approach to
environmental consensus building. Island Press, Washington, DC, USA.

Van Liere, K. D., and F. P. Noe. 1981. Outdoor Recreation and Environmental
Attitudes - Further Examination of the Dunlap-Heffeman Thesis. Rural
Sociology 46:505-513.

Vamer, G. E. 1994. Environmental law and the eclipse of land as private property.
Pages 142-160 in F. Frederick, editor. Ethics and Environmental Policy.
University of Georgia Press,Athens, GA, USA.

133

Vennix, J. A. M. 1999. Group model-building: tackling messy problems. System
Dynamics Review 15:379-401.

Verplanken, B., G. Hofstee, and H. J. W. Hanssen. 1998. Accessibility of affective
versus cognitive components of attitudes. European Journal of Social
Psychology 28:23-35.

Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human
domination of Earth's ecosystems. Science 277:494-499.

Webster, P. J ., G. J. Holland, J. A. Curry, and H.-R. Chang. 2005. Changes in tropical
cyclone number, duration, and intensity in a warming environment. Science
309:1844-1846.

Wheelock, J., and E. Oughton. 1996. The household as a focus for research. Journal
of Economic Issues 30:143-159.

Wilson, T. D., S. Lindsey, and T. Y. Schooler. 2000. A model of dual attitudes.
Psychological Review 107:101-126.

World Commission on Environment and Development. 1987. Our common future.
Oxford University Press, New York, USA.

World Resources Institute. 2003. Earth trends data tables: biodiversity and protected
areas. [Available online at:
http://earthtrends.wri.org/datatables/index.php?theme=7] Last Accessed 17
April 2007.

Xie, J. L., H. R. Hill, S. R. Winterstein, H. Campa, R. V. Doepker, T. R. Van Deelen,
and J. G. Liu. 1999. White-tailed deer management options model

(DeerMOM): design, quantification, and application. Ecological Modelling
124:121-130.

York, R., E. A. Rosa, and T. Dietz. 2002. Bridging environmental science with
environmental policy: Plasticity of population, affluence, and technology.

Social Science Quarterly 83:18-34.

Young, R. M. 1985. Darwin's metaphor: nature's place in Victorian culture.
Cambridge University Press, Cambridge, Massachusetts, USA.

Zelinsky, W. 1971. Hypothesis of Mobility Transition. Geographical Review 61 :219-
249.

134

APPENDIX

CODE FOR THE SYSTEMS MODELa

Population sector

TotalPop(t) = TotalPop(t - dt) + (Immigration) * dt

INIT TotalPop = 3439

/Total population at any year equals the total population in the previous year plus the

number of immigrants between years. Initial population reflects the number of longer

term residents residing in the area prior to 1990./

IN FLOWS:

Immigration = if Base_lmmigration > New_People_Accomodated then
New_People_Accomodated else Base_Immigration

Base_Immigration = TotalPop*0.059

HH_Size_Ag = normal(2.85,l .53)

HH_Size_Nat = normal(2.3 1 , 1.39)

HH_Size_Res = normal(3.56, 1.80)

Local_Pop = 3439

New_People_Accomodated =
ConstructionAPH*HSFilter+ConstructionNPH*HSFilter
+ConstructionRPH*HSFilter

/ Immigration rate is a function of total population (0.059) unless insufficient housing

is available. In the latter case immigration rate reflects the number of people that can

use new construction. That number depends on the number of new homes

 

a All text bounded by forward slashes (e.g., /hello/) is explanatory and not model code

135

constructed in each land cover area (APH = agricultural area permanent resident

houses, NPH = natural area permanent resident houses, RPH = residential area

permanent resident houses) and average household size in homes within each land

cover type. Household sizes (HH_Size) in each land cover type (Ag = agricultural,

Nat = natural, Res = residential) are random normally distributed variables based on

the mean and standard deviations of household size generated from survey data. The

HSfilter variable prevents household sizes below 1 from being generated]

House construction sector

APH(t) = APH(t - dt) + (ConstructionAPH) * dt

INIT APH = LAH

/Total APH at any year equals the total APH in the previous year plus the number of

new APH built between years. Initial APH reflects the number of houses in

agricultural areas with permanent residents prior to 1990 (LAH)./

INF LOWS:

ConstructionAPH = if GSRFactor=0 then immAPHdemand*OpenLandConstraint
else immAPHdemand*GSRFactor*OpenLandConstraint

/Construction of new houses for permanent residents in agricultural areas is a function

of growth slowing restrictions (GSRF actor) limited open land for development

(OpenLandConstraint) and homes demanded by immigrants (immAPHdemand)./

AVH(t) = AVH(t - dt) + (ConstructionAVH) * dt

INIT AVH = LAH*0.264

/Total agricultural area vacation houses (AVH) at any year equals the total AVH in

the previous year plus the number of new AVH built between years. Initial AVH

136

reflects the ratio of vacation houses to permanent resident houses in agricultural areas

(0.264) in 1990 times the number of permanent resident houses in agricultural areas

in 1990 (LAH)./

INFLOWS:

ConstructionAVH = if Ban_vac_Homes=l then 0 else if GSRFactor=0 then
immAPHdemand*0.09*OpenLandConstraint else
immAPHdemand*0.09*OpenLandConstraint*GSRF actor

/Construction of vacation houses in agricultural areas is a function of policy banning

their construction (Ban_vac_Homes), growth slowing restrictions (GSRF actor),

limited open land for development (OpenLandConstraint) and homes demanded by
immigrants (immAPHdemand) times the percent of those homes that are used as
vacation homes./

NPH(t) = NPH(t - dt) + (ConstructionNPH) * dt

INIT NPH = LNH

/Total NPH at any year equals the total NPH in the previous year plus the number of

new NPH built between years. Initial NPH reflects the number of houses in natural

areas with permanent residents prior to 1990 (LNH)./

INFLOWS:

ConstructionNAH = if GSR=0 then immNPHdemand*OpenLandConstraint else
immNPHdemand*OpenLandConstraint*GSRF actor

/Construction of new houses for permanent residents in natural areas is a function of

growth slowing restrictions (GSRFactor) limited open land for development

(OpenLandConstraint) and homes demanded by immigrants (immNPHdemand)./

137

NVH(t) = NVH(t - dt) + (ConstructionNVH) * dt

INIT NVH = LNH*0.264

/Total natural area vacation houses (NVH) at any year equals the total NVH in the

previous year plus the number of new NVH built between years. Initial NVH reflects

the ratio of vacation houses to permanent resident houses in natural areas (0.264) in

1990 times the number of permanent resident houses in natural areas in 1990 (LNH)/

INF LOWS:

ConstructionNVH = if Ban_vac_Homes=l then 0 else if GangplankFactor=0 then
immNPHdemand*0.09*OpenLandConstraint else
immNPHdemand*0.09*OpenLandConstraint*GangplankF actor

/Construction of vacation houses in natural areas is a fimction of policy banning their

construction (Ban_vac_Homes), grth slowing restrictions (GSRF actor), limited

open land for development (OpenLandConstraint) and homes demanded by
immigrants (immNPHdemand) times the percent of those homes that are used as
vacation homes./

RPH(t) = RPH(t - dt) + (ConstructionResHH) * dt

INIT RPH = LRH

/Total RPH at any year equals the total RPH in the previous year plus the number of

new RPH built between years. Initial RPH reflects the number of houses in

residential areas with permanent residents prior to 1990 (LRH)./

IN FLOWS:

ConstructionRPH = if GSRFactor=0 then immRPHdemand*OpenLandConstraint else

immRPHdemand*GSRFactor*OpenLandConstraint

138

/Construction of new houses for permanent residents in residential areas is a function
of growth slowing restrictions (GSRFactor) limited open land for development
(OpenLandConstraint) and homes demanded by immigrants (immNPHdemand)./
RVH(t) = RVH(t - dt) + (ConstructionRVH) * dt

INIT RVH = LRH*0.264

/Total residential area vacation houses (RVH) at any year equals the total RVH in the

previous year plus the number of new RVH built between years. Initial RVH reflects

the ratio of vacation houses to permanent resident houses in residential areas (0.264)

in 1990 times the number of permanent resident houses in natural areas in 1990

(LRH)/

INFLOWS:

ConstructionRVH = if Ban_vac_Homes=1 then 0 else if GSRFactor=0 then
immRPHdemand*0.09*OpenLandConstraint else
immRPHdemand*0.09*OpenLandConstraint*GSRFactor

/Construction of vacation houses in natural areas is a function of policy banning their

construction (Ban_vac_Homes), growth slowing restrictions (GSRFactor), limited

open land for development (OpenLandConstraint) and homes demanded by
immigrants (immRPHdemand) times the percent of those homes that are used as
vacation homes./

Ban_vac_Homes = 0

/0 = no vacation home ban and l = vacation home ban/

GSR = if GSR_Policy = 0 then 0 else Total_Pop/Local_Pop

139

/1f growth slowing restriction policy is enacted the GSR calculates the ratio of total
population to population of pre-l990 residents./

GSR_Policy = 0

HH_Size_Locals = 3.03

/Average number of residents per house for pre-l 990 residents./
Imm%_Agricultural_Area_H = 0.521

/Percent of immigrants choosing home locations in agricultural areas./
immAgI—lHdemand = (Base_1mmigration*imm%_Agricultural__Area_H)/HSFilter
/Housing demand based on number of immigrants and percent preferring land cover
type. The HSFilter variable represents household size, but prevents values < 1./
immNatHHdemand = (Base_Immigration*imm_%Natural_Area_H)/ HSFilter
immResHHdemand = (Base_lmmigration*imm_%_Residential_Area_H)/ HSFilter
imm_%Natural_Area_HH = 0.207

/Percent of immigrants choosing home locations in natural areas./
imm_%_Residential__Area_HH = 0.240

/Percent of immigrants choosing home locations in residential areas./

LAH = Local%_Agricultural_Area_H*Local_HHs

LNH = Local_%_Natural_Area_H*Local_HHs

Local%_Agricultural_Area_H = 0.514

Local_%_Natural_Area_H = 0.222

local_%_Residential_Area_H = 0.262

Local_HHs = Local_Pop/HH_size_Locals

140

/The number of households occupied by pre-1990 residents equals their population
divided by number of residents per house./

LRH = local_%_Residential_Area_HH*Local_HHs

Percent_CIuster_Development_Required = 0

/The percent cluster development required can be changed from 0—100./

totagh = APH+AVH

/Calculates total number of houses in agricultural areas./

totanath = NPH+NVH

totresh = RPH+RVH

GSRF actor = GRAPH(GSR)

(2.00, 1.00), (2.80, 0.44), (3.60, 0.30), (4.40, 0.21), (5.20, 0.12), (6.00, 0.05), (6.80,
0.01), (7.60, 0.00), (8.40, 0.00), (9.20, 0.00), (10.0, 0.00)

/Growth slowing restrictions do not occur until GSR is > 2 (i.e., post-1990

immigrants have a voting majority. After that point the GSRF actor demonstrates

exponential decay (initially rapid but progressively slower)./

OpenLandConstraint = GRAPH(Private_Open_Hectares)

(0.00, 0.00), (200, 0.09), (400, 0.20), (600, 0.30), (800, 0.400), (1000, 0.50), (1200,
0.60), (1400, 0.70), (1600, 0.80), (1800, 0.90), (2000, 1.00)

/The open land constraint graph calculates growth limitations based on the number of

non-developed private hectares. When that number drops below 2000 development

rate begins a linear decline to zero development with zero hectares available./

141

Land Cover Area

Private_Open_Hectares(t) = Private_Open_Hectares(t - dt) + (- ToSprawlDev -
ToReserve - ToClusterDev) * dt

INIT Private_Open_Hectares = 71815.22

/Tota1 private undeveloped hectares at any year equals the total in the previous year
minus the hectares developed in cluster and sprawl patterns and the number of
hectares protected in reserves./

OUTF LOWS:

ToSprawlDev = if Percent_Cluster_Development_Required=0 then
(ConstructionAVH*AgriculturalHHectares)+(ConstructionNVH*0.575*HillsideHHe
ctares)+(ConstructionNVH*0.425
*RiparianHHectares)+(ConstructionAPH*AgriculturalHHectares)+(ConstructionN P
H*0.575*HillsideHHectares)+(ConstructionNPH*0.425 *RiparianHHectares) else
((ConstructionAVH*AgriculturalHHectares)+(ConstructionNVH*0.575*HillsideHHe
ctares)+(ConstructionNVH*0.425
*RiparianHHectares)+(ConstructionAPH*AgriculturalHHectares)+(ConstructionNP
H*0.575*HillsideHHectares)+(ConstructionNPH*0.425 *RiparianHHectares))*(l -
Percent_Cluster_Development_Required)

/ If cluster development is not required land moves into sprawl development at the rate
per house stipulated in zoning regulations (e.g., HillsideHectares = hectares per house
built on a hillside). The 0.425 and 0.575 values represent the proportion of natural

area homes built in wetland and hillside area respectively. If cluster development is

142

required the area is reduced by multiplying area by l-the percent of cluster
development required/

ToReserve = if Percent_Cluster_Development_Required=0 then 0 else
((ConstructionAVH*AgriculturalHHectares-
ConstructionAVH*ResidentialHHHectares)+(ConstructionAPH*AgriculturalHHectar
es-
ResidentialHHHectares*ConstructionAPH)+(ConstructionNPH*0.575*HillsideHHec
tares-ResidentialHHHectares*ConstructionNPH*0.575)+(ConstructionNPH*0.425
*RiparianHHectares-
ResidentialHHHectares*ConstructionNPH*0.425)+(ConstructionNVH*0.575*Hillsid
eHHectares-
ResidentialHHHectares*ConstructionNVH*0.575)+(ConstructionNVH*0.425
*RiparianHHectares-Residentia1HHHectares*ConstructionNVH*0.425))

/No land is set aside in reserves unless cluster development occurs. If cluster
development is used reserve area equals the difference between land requiredunder
traditional zoning and land required under residential zoning./

ToClusterDev = if Percent_Cluster_Development_Required=0 then
(ConstructionRVH*ResidentialHHHectares+ConstructionRPH*ResidentialHHHectar
es) else
(ConstructionRVH*ResidentialHHHectares+ConstructionRPH*ResidentialHHHectar
es)
+(Percent_Cluster_Development_Required*ResidentialHHHectares*(ConstructionA

VH+ConstructionNVH+ConstructionAPH+ConstructionNPH))

143

/Without cluster development policy only residential homes ad to hectares in cluster
development. With the policy the percent of agricultural and natural area homes in
cluster patterns contribute to hectares of cluster development as well./
Protected_Area(t) = Protected_Area(t - dt) + (ToReserve) * dt

INIT Protected_Area = 39197.56

INFLOWS:

ToReserve = if Percent_Cluster_Development_Required=0 then 0 else
((ConstructionAVH*AgriculturalHHectares-
ConstructionAVH*ResidentialHHHectares)+(ConstructionAPH*AgriculturalHHectar
es-
ResidentialHHHectares*ConstructionAPH)+(ConstructionNPH*0.575*HillsideHHec
tares-ResidentialHHHectares*ConstructionNPH*0.575)+(ConstructionNPH*0.425
*RiparianHHectares-
ResidentialHHHectares*ConstructionNPH*0.425)+(ConstructionNVH*0.575*Hillsid
eHHectares-
ResidentialHHHectares*ConstructionNVH*0.575)+(ConstructionNVH*0.425
*RiparianHHectares-ResidentialHHHectares*ConstructionNVH*0.425))
ResClusterArea(t) = ResClusterArea(t - dt) + (ToClusterDev) * dt

INIT ResClusterArea = 148.92

INFLOWS:

ToClusterDev = if Percent_Cluster_Development_Required=0 then
(ConstructionRVH*ResidentialHHHectares+ConstructionRPH*ResidentialHHHectar

es) else

144

(ConstructionRVH*ResidentialHHHectares+ConstructionRPH*ResidentialHHHectar
es)
+(Percent_Cluster_Development_Required*ResidentialHHHectares*(ConstructionA
VH+ConstructionNVH+ConstructionAPH+ConstructionNPH))

Res_Sprawl_Area(t) = Res_Sprawl_Area(t - dt) + (ToSprawlDev) * dt

INIT Res_Sprawl_Area = 5443.02

INFLOWS:

ToSprawlDev = if Percent_Cluster_Development_Required=0 then
(ConstructionAVH*AgriculturalHHectares)+(ConstructionNVH*0.575*HillsideHHe
ctares)+(ConstructionNVH*0.425
"‘RiparianHHectares)+(ConstructionAPH*AgriculturalHHectares)+(ConstructionNP
H*0.575*Hi1lsideHHectares)+(ConstructionNPH*0.425 *RiparianHHectares) else
((ConstructionAVH*AgriculturalHHectares)+(ConstructionNVH*0.575*HillsideHHe
ctares)+(ConstructionNVH*0.425
*RiparianHHectares)+(ConstructionAPH*AgriculturalHHectares)+(ConstructionNP
H*0.575*Hi11sideHHectares)+(ConstructionNPH*0.425 *RiparianHHectares))*(1-
Percent_Cluster_Development_Required)

AgriculturalHHectares = 8.09

Developable_Area = total_area_of_va11ey-Developed_Area-Protected_Area
Developed_Area = ResClusterArea+Res_Sprawl_Area

HillsideHHectares = 1.01

ResidentialHHHectares = 0.41

RiparianHHectares = 8.09

145

total_area_of_valley = 1 16605

Total_Open_Space = Private_Open_Hectares+Protected_Area
Not in a sector

HHSizeSim = TotalPop/Total_Homes

HSFilter = if HH_Size_Ag > 1 then HH_Size_Ag else 1
HSFiltem = if HH_Size_Nat > 1 then HH_Size_Nat else 1
HSFilterr = if HH_Size_Res > 1 then HH_Size_Res else 1
InitialAgHomes = 635

InitialNatHomes = 268

InitialResHomes = 233

percent_vacation = Totalt_vacation_Homes/Total_Homes
Totalt_vacation_Homes = AVH+NVH+RVH

Total_Homes = APH+AVH+NVH+NPH+RPH+RVH

146

   

llili]]][lli]][i[]ll[[1111]]