Assignment 2

Using analyses of public value orientations, attitudes and preferences to inform national forest planning in Colorado and Wyoming Jessica M. Clement *, Antony S. Cheng Colorado Forest Restoration Institute, Department of Forest and Rangeland Stewardship, Colorado State University, Forestry Building, Colorado State University, Fort Collins, CO 80523, USA Keywords:

Social values and attitudes National forests Forest planning Forest Service Social research Survey methods abstract Understanding public value orientations, attitudes and preferences towards national forests is a critical task for the USDA Forest Service (USFS) during the development of their forest plans. Social surveys are efﬁcient and effective ways to generate this information from a representative sample of the larger public who may care about national forests but do not attend participatory events ethe so-called silent majority. Survey results can be used to complement input generated from participatory and collaborative processes. This paper presents results and discusses implications from social surveys conducted on three national forests in Colorado and Wyoming. The results indicate that although respondents identi ﬁed aesthetic, biodiversity, future and recreation value orientations as most important, there are also surprising linkages between value orientations, attitudes and preferences towards forest uses and policy options associated with speci ﬁc geographic and socio-economic contexts and conditions. The results also suggest some “hotspots ”where value orientations, attitudes and preferences display some apparent contradictions. Such hotspots indicate potential con ﬂicts and suggest opportunities to focus participa- tory, collaborative methods. The relevance of these results to national forest planning in particular and public resource management is explored in the discussion and conclusion, especially as the USFS eand other public resource management agencies eface increased pressure to make room for place-based, collaborative planning while also taking into account broader public sentiments and preferences.

2010 Elsevier Ltd. All rights reserved. Introduction The National Forest and Grasslands System of the USA total 191 million acres (77.3 million ha) and are managed by the USDA Forest Service (USFS) to encompass multiple public values and uses, from commercial activities like ski areas, timber harvesting, and mining to more biocentric values, such as protecting threatened and endangered species and maintaining ecosystems in their wild, undeveloped state. As such, the national forests are geographic features integral to American history, livelihoods, and lifestyles ( Hays, 2009 ). Understanding and taking into account public value orientations, attitudes and preferences towards national forest goals, uses, and management activities are a matter of critical importance for the USFS, especially during the development of a national forest ’s Land and Resource Management Plan emore commonly called a forest plan ( Allen et al., 2009). Required by the National Forest Management Act of 1976 (NFMA), a forest plan effectively sets policy about a national forest ’s goals, priorities, special area designations (i.e., recommendations for Wilderness and Wild & Scenic River designations) and management strategies for 10 e15 years ( Wilkinson & Anderson, 1987 ). Theﬁrst national forest plans were completed in the 1980s and are required to be revised every 10 e15 years.

The USFS has struggled with understanding and integrating social value orientations, attitudes, and preferences into its forest planning processes, which have traditionally emphasized technical analyses of a national forest ’s biophysical conditions and attributes ( Committee of Scientists, 1999; Cortner & Moote, 1999; Larsen et al., 1990 ). The USFS employs public participation processes in attempts to elicit and incorporate public sentiments, such open public meetings, solicitation of written or oral comments, or small-group interactions ( Gericke, Sullivan, & Wellman, 1992 ). Collaborative approaches and processes in particular have rapidly gained favor in national forest planning and management decision-making ( Burns & Cheng, 2005; Selin, Schuett, & Carr, 1997; Wondolleck & Yaffee, 2000 ). However, even if the USFS thoughtfully develops highly inclusive, transparent and well-structured public involvement and * Corresponding author.

E-mail addresses: [email protected] (J.M. Clement),tony.cheng@ colostate.edu (A.S. Cheng). Contents lists available at ScienceDirect Applied Geography journal homepage: www.elsevier.com/locate/apgeog 0143-6228/$esee front matter 2010 Elsevier Ltd. All rights reserved.

doi: 10.1016/j.apgeog.2010.10.001 Applied Geography 31 (2011) 393e 400 collaborative processes, participants may reﬂect only segments of the broader population. What of the so-called “silent majority ”who may not participate in letter-writing campaigns, attend public meetings or commit to an intensive collaborative process, but nevertheless may care deeply about what happens to national forests? A second elicitation approach available to the USFS is the administration of large-N, random sample social surveys contain- ing questions about value orientations, attitudes, beliefs and pref- erences for national forest management. Social surveys are efﬁ cient ways to collect information from a representative sampling of the silent majority ( Allen et al., 2009 ). Social surveys relating to national forests have been conducted at the national ( Shields et al., 2002a, 2002b ), regional (Tarrant & Cordell, 2002 ), and state levels ( Manning, Valliere, & Minteer, 1999; Vaske & Donnelly, 1999 ), as well as comparing national and state populations ( Shindler, List, & Steel, 1993 ). These and other studies examining public values and attitudes towards a variety of natural resource topics suggest that, in general, the U.S. public is generally more favorable towards biocentric values and prefer to protect ecosystems over using resources for material gain.

This general ﬁnding is useful to USFS planners, as it indicates a general public sentiment. But as sense of-place research suggest, people ’s connection to and values for national forest “places ”are complex and multi-layered, encompassing highly personal, inti- mate connections as well as instrumental and symbolic connec- tions ( Williams & S tewart, 1998). Additionally, many national forest planning decisions are not so easily dichotomous choices between ecosystem protection and resource use. For example, tens of millions acres of national forest lands in the Western US are in need of active management in the form of mechanical tree removal (i.e., logging), prescribed burning, or both, in order to reduce the risk of ecologically uncharacteristic and/or socially undesirable wild ﬁres ( Sampson & Adams, 1994 ). By emphasizing forest wild ﬁre risk reduction in a national forest plan, value-added forest products and other wood biomass companies eand the people they employ and do business with emay bene ﬁt economically, people living and recreating in the forests face reduced wild ﬁre risks, and the resilie nce of the ecosystem in the face of wild ﬁre events may likely be improved. While it may be obvious that national forest planning choices affect different people in different ways, USFS planners and decision-makers often do not have empirical data from which to draw in order to assess how different forest management goals and strategies are regarded by the public. As a result, the social impacts of plan decisions epositive and negative emay be a matter of guesswork, thereby heightening public dissatisfaction with the planning process and the ﬁnal plan decision, and fostering con ﬂ ict.

As Allen et al. (2009) note, there is a need and opportunity to re ﬁne and apply social research methodologies to speci ﬁc national forest geographic and socio-economic contexts in order to more fully inform forest planning and decision-making. Even as the USFS employs more participatory, collaborative public involvement strategies, forest planners will bene ﬁt from cross-checking the values, attitudes and preferences of active attendees of participa- tory processes with those of the silent majority. Having social research ﬁndings in hand also allows planners and the public to more precisely identify and closely examine potential con ﬂicts and take appropriate steps to address them. Lastly, federal adminis- trative and environmental laws mandate agency decision-makers to make informed choices supported by evidence ( Haas, 2003); possessing results from valid social research methodologies can give decision-makers the con ﬁdence that their public land management decisions are grounded in empirical data and avoid accusations of making arbitrary and capricious decisions. In this light, we present results and implications from a social survey methodology that was replicated across three national forests undergoing revision of their forest plans from 2004-2008:

the Bridger-Teton (BTNF) in Wyoming, Pike-San Isabel (PSI) in Colorado), and Shoshone (SNF) in Wyoming. The methodology was derived from the work Greg Brown and Pat Reed conducted pur- suant to the Chugach National Forest plan revision process ( Brown, Reed, & Harris, 2002 ) and has since been applied across various resource management settings globally ( Brown & Reed, 2009). The methodology was applied to these three forests at the request of USFS planners and decision-makers, as well as state policy-makers who desired to ensure that social data and analyses were being considered in national forest plans alongside biophysical analyses.

A highly participatory process involving USFS planners, decision- makers, and public stakeholders was employed to tailor the survey to speci ﬁc geographic and socio-economic contexts. Additionally, each national forest was utilizing collaborative stakeholder processes and there was interest in assessing how stakeholder values and preferences compared with those of a large, represen- tative sample of the public. The results of statistical analyses of survey data are presented in the context of key national forest plan decisions for each forest and across all three forests. We present the results related to values for all three forests, and then focus on one critical subject for each forest, explore the role of values in relation to each subject, and describe the implications regarding these results for each forest ’s planning efforts. Our intent is two-fold: 1) to demonstrate how this methodology, based on a speci ﬁc set of constructs of social values and behavior, can generate information useful to national forest planning, and 2) contribute to the evolving scholarship of social values-related research in landscape planning contexts.

Study contexts and concepts Bridger-Teton National Forest (BTNF) is in northwest Wyoming, encapsulating ﬁve counties (Teton, Fremont, Sublette, Park and Lincoln) with a total population of 107,287 as of 2007. The largest sources of income to this area were recreation, tourism, natural gas development and agriculture ( Taylor, Coupal, Foulke, Rashford, & Olsen, 2008 ). The Shoshone National Forest (SNF) includes Park, Teton, Fremont and Hot Springs counties which collectively have a population of 87,159 in 2007 relying on recreation and tourism, livestock grazing, and timber as primary economic activities ( Taylor, Coupal, et al., 2008; Taylor, Foulke, et al., 2008 ). Both national forests contain populations of elk, sage grouse, grizzly bear and wolves. Vegetation on both national forests start with sage- brush and grasslands at lower elevations and proceed to lodgepole pine, mixed-conifer forest types, spruce- ﬁr forests, and alpine tundra as elevation rises. Insect infestations and wild ﬁres are recent signi ﬁcant ecological disturbances on these national forests.

The lands and communities surrounding the BTNF and SNF have experienced considerable in-migration (12.5% population growth between 2000 and 2008), resulting in an increase in land devel- opment and second-home-ownership ( Taylor, Coupal, et al., 2008; Taylor, Foulke, et al., 2008 ).

The Pike and San Isabel National Forests (PSI) in central Colo- rado are adjacent to large population centers totaling approxi- mately 3.2 million people, including Denver, Colorado Springs, Pueblo, and communities of Summit County, one of the fastest growing counties in the U.S. ( USDA Forest Service 2006). The PSI contain the tallest mountains peaks in Colorado, include nine wilderness areas, and support the third highest visitation rate of any forest in the National Forest System ( USDA Forest Service 2006 ). The communities adjacent to the PSI have experienced high population growth rates in both rural and urban areas and J.M. Clement, A.S. Cheng / Applied Geography 31 (2011) 393 e400 394 changes in demographic proﬁles, predominantly by people who enjoy larger incomes and higher education levels than the tradi- tional rural Colorado population ( McGranahan, 1999; Stein, 2005; USDA Economic Research Service, 2005 ).

In structuring the social analysis for the three national forests, we de ﬁne “value ”as “an enduring belief that a speci ﬁc mode of conduct or end-state of existence is personally or socially preferable to an opposite or converse mode of conduct or end-state of exis- tence” (Rokeach, 1973 ). Values are relatively stable and are unlikely to change unless under extreme duress ( Allen et al., 2009; Rokeach, 1973 ). Since values are generally predictive of attitudes and pref- erences towards speci ﬁc topics or actions, they can provide an important and reliable indicator whether proposed goals and activities in a national forest plan will be considered acceptable to the public ( Allen et al., 2009; Fulton, Manfredo, & Lipscomb, 1996; Homer & Kahle, 1988; Machlis, Kaplan, & Tuler, 2002; Vaske & Donnelly, 1999 ).

Reed and Brown ’s social values research methodology was chosen not only because it marked a re ﬁnement in and new application of social values research, but also because it employed three attributes relevant to national forest planning. First, the method draws on Zube ’s “transactional concept of human-land- scape relationships ”to connect individuals’ held values with a landscape ( Brown, 1984; Zube, 1987 ). In these relationships humans are considered participative actors in a landscape since they work and live in it, and therefore value a landscape from this interactive perspective, attributing this contextual meaning to a landscape and to speci ﬁc places as a result. The methodology seeks to capture these meanings through values, attitudes and preferences, both as survey and Geographical Information Systems (GIS) data. This paper reports on the survey results only, see Sherrouse, Clement and Semmens (2011) for GIS survey results.

Second, the methodology operationalized a typology of 13 value de ﬁnitions based on the writings of Holmes Rolston ( Rolston, 1981; Rolston & Coufal, 1991 ). This typology comprises the most inclusive range of social values for natural, undeveloped settings or so-called wildlands, of which national forests can be categorized. Each value de ﬁnition is distinct and has precise meaning, which lends to the typology ’s reliability and validity in survey research.

Third, the method is grounded in the cognitive hierarchy concept stemming from the Theory of Reasoned Action and the subsequent Theory of Planned Behavior which articulate cognitive hierarchy constructs such as beliefs and attitudes ( Ajzen, 1991; Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975 ). The cognitive hierarchy provides a logical and structured format to survey people about their beliefs and attitudes about speci ﬁc national forest goals, conditions, uses and management activities ( Brown & Reed, 2000).

Within the cognitive hierarchy, a belief is a person ’s judgment about what they consider true or false. The aggregation of several basic beliefs regarding a speci ﬁc topic forms a value orientation towards that topic. Thus, the actual value is not measured; rather, it is an individual ’s value orientation to a particular issue that is measured through a set of basic beliefs ( Fulton et al., 1996; Rokeach, 1973 ). In this article we will use the terms “value ”and “value orientations ”to indicate those sets of basic beliefs variables used to measure respondents value orientations. Basic beliefs in turn form the basis for attitudes, which are strongly predictive of behavior and are an individual ’s consistent tendency to respond favorably or unfavorably towards an object, concept or individual. In this study forest values relationships are explored with attitudes, beliefs and preferences, which are a combination of attitudes and beliefs about speci ﬁc forest plan allocations, predominantly related to forest uses ( Brown & Reed, 2000 ). The relationship between values, beliefs and attitudes has been tested in many studies, among them studies speci ﬁc to natural resource contexts ( Bright, Manfredo, & Fulton, 2000; Fulton et al., 1996, 2004; Grilliot & Armstrong, 2005; Lubell & Vedlitz, 2006; Vaske & Donnelly, 1999 ). Using the methodology in this study, values, attitudes and preferences have also been used in Australia ( Brown, 2005, 2006; Raymond et al., 2009 ), Alaska ( Alessa, Kliskey, & Brown, 2008; Brown, Smith, Alessa, & Kliskey, 2004 ), and Idaho ( Nielsen-Pincus, 2006 ).

Methods A mail survey was developed containing ﬁve sections: 1) questions regarding familiarity and use on the forest in question; 2) questions that measured attitudes to 19 forest uses with a 5-point Likert scale, where 1 ¼Strongly Oppose and 5 ¼Strongly Favor; 3) a series of policy preference variables speci ﬁc to each Forest ’s plan revision, using 5-point Likert scales where 1 ¼Strongly Disagree and 5 ¼Strongly Agree; 4) a values typology with forest values ( Table 1 ), requiring respondents to divide a hypothetical $100 between the values listed and mark a map with places that are valued accordingly; and 5) demographic information. On the BTNF and SNF, all values were included in the typology, however on the PSI “subsistence” value was dropped since hunting was judged by managers to be comparatively more recreation- than subsistence- oriented on this Forest. To ensure that the survey addressed national forest-speci ﬁc geographic conditions and forest uses, the authors chose to engage USFS managers, planners and decision-makers in crafting questions pertaining to use and policy variables. On the PSI, the variables were designed based on a questionnaire submitted to the six district rangers to ask them about issues on their districts that were likely to continue for some time in the future, and therefore provided good discussion points during the forest planning process.

On the BTNF and SNF, use and policy variables were drafted with assistance from USFS staff, and then posted on the USFS websites for comments and suggestions by members of the public. Using these various inputs, each survey draft was amended and discussed with the BTNF ’s and SNF ’s respective Cooperator Group composed of county and soil conservation district representatives with whom each national forest consulted throughout the plan revision process. The survey drafts were amended according to suggestions made by the Cooperators, other stakeholders and members of the public while bearing in mind survey size limitations, response rates, and the need for concise, relevant information for planning purposes.

A mail survey was sent to a random sample of home-owner residents in the counties surrounding the forests ethe de ﬁned population of interest as stated by the agency planners and policy- makers in conjunction with the authors when the project strategies were being negotiated. A random sample was generated by Survey Sampling International. The sample was focused on primary home- owners to reach residents who are most likely to be affected by long-term forest planning efforts. The survey mailings were con- ducted using Dillman ’s total design method ( Dillman, 1978, Dillman, Smyth, & Christian, 2007 ). Theﬁrst two and the fourth mailings consisted of the survey, a map of the forest, and a self- addressed return envelope with a letter. The third mailing consisted of a reminder postcard. In each case, the random sample contained 2000 addresses of permanent residents. We identi ﬁed one key national forest plan decision for each forest to explore the link between values and attitude and prefer- ence policy variables for each forest. At the time the studies were being conducted, the Bridger-Teton National Forest was engaged with its public in an oil and gas discourse which involved the trade- offs between economic beneﬁ ts and habitat impairment on the Wyoming Range. On the Pike-San Isabel National Forests the question of amount, timing and location of motorized recreation J.M. Clement, A.S. Cheng / Applied Geography 31 (2011) 393 e400 395 was of special interest. And on the SNF National Forest there had been ongoing discussions regarding forest health and the related role of the private sector to conduct forest treatments and create local economic beneﬁts.

A series of statistical analyses were performed on survey responses to explore these issues similar to analyses used by Brown and Reed using SPSS 17.0 ( SPSS, 2004). Our desire was to look at general commonalities and differences between the three forests in terms of values, attitudes and preferences and to use this data to explore speci ﬁc forest plan decisions in relation to each forest. As illustrated in Fig. 1, several statistical procedures were used to analyze response data. In the ﬁrst step, response frequencies were run and the differences between values ( Table 1) and attitude ( Table 2 ) responses by Forest were explored using analysis of variance (ANOVA) tests.

Second, in order to explore the relationship between values and speci ﬁc plan decisions, we used discriminant function analysis ( Table 3 ). Discriminant function analysis models relationships between a categorical dependent variable (speci ﬁc plan decision) and one or more scale independent variables (value orientations and attitudinal rankings). The three objectives of discriminant analysis are to classify respondents into groups, assess the extent to which independent variables successfully classify the dependent variable and, lastly, determine the extent to which the independent variables explain the variance in the dependent variable ( Stevens, 2002).

Discriminant function analysis attempts to ﬁnd linear combinations between a set of independent variables, i.e. the twelve forest values, that optimally separate the groups of respondents, and a national Step I Step 2 Step 3Step 4 Forest-Specific Planning Issues Use Discriminant Function Analysis to find explanatory values for particular issues.

Table 3 Beliefs Attitudes Preferences Table 2 Values Table 1 Use Demographics to explore relationship between respondents and values to discover possible stakeholders in collaborative process Explore relationship between values and other variables in relation to specific planning issues.

Table 4 Fig. 1. Relationship between values and other variables, and steps to explore values in relation to particular issues, and potential stakeholders. Table 2 Differences in attitudes and preferences across forest (From 1 ¼Strongly Oppose/ Disagree to 5 ¼Strongly Favor/Agree).

PSI BTNF SNF F-value p-value Forest Uses Sport Fishing 4.42 4.62 4.69 38.67 .000 Non-motorized Recreation 4.58 4.48 4.55 2.84 .059 Sport Hunting 3.72 4.52 4.42 103.253 .000 Wildlife Viewing/Observation 4.64 4.76 4.82 16.863 .000 Motorized Recreation 2.69 3.21 3.17 27.995 .000 Oil/Gas Drilling 2.21 2.38 2.48 5.45 .004 Commercial Out ﬁtting 3.47 3.71 3.81 15.53 .000 Wilderness 4.31 3.79 4.05 26.94 .000 Fish and Willdife Habitat 4.65 4.66 4.70 .937 .392 Vegetation management Logging for commercial pro ﬁt only 2.19 3.85 3.62 338.942 .000 Logging to protect life and property 4.30 4.28 4.43 4.419 .012 Logging to remove dead or insect-infested trees 4.48 4.43 4.43 .717 .488 Logging to create or improve wildlife habitat 4.28 4.48 4.46 11.372 .000 No support for any logging 1.94 1.82 1.73 5.305 .005 Tehniques to reduce the wild ﬁre risk close to communities Clearcutting 2.56 2.81 2.88 8.519 .000 forest thinning 4.07 4.18 4.21 3.595 .028 Prescribed ﬁres 3.54 3.46 3.53 .659 .518 No action 2.02 2.07 1.9 2.416 .090 Table 1 Percent of respondents per forest and signiﬁ cant differences between responses per forest.

Value De ﬁnitions PSI % BTNF % SNF% Total Studies % ANOVA F-Value Aesthetic: I value these Forests because I enjoy the scenery, sights, sounds, smells, etc 75 76 76 76 2.72 Biodiversity: I value these Forests because they provide a variety of ﬁsh, wildlife, plant life, etc. 68 75* 72 71 4.77 Cultural: I value these Forests because they are a place for me to continue and pass down the wisdom and knowledge, traditions, and way of life of my ancestors. 32 37 28 34 1.19 Economic: I value these Forests because they provide timber, ﬁsheries, minerals, and/or tourism opportunities such as out ﬁtting and guiding 34 49** 45** 41 13.45 Future: I value these Forests because they allow future generations to know and experience the Forests as they are now. 70* 65 67* 68 5.01 Historic: I value these Forests because they have places and things of natural and human history that matter to me, others, or the nation. 42 47 34 42 1.66 Intrinsic: I value these Forests in and of themselves, whether people are present or not 43 41 38 42 1.88 Learning: I value these Forests because we can learn about the environment through scienti ﬁc observation or experimentation. 40** 37 33 37 8.93 Life Sustaining: I value these Forests because they help produce, preserve, clean, and renew air, soil, and water. 71** 67 58 66 11.72 Recreation: I value these Forests because they provide a place for my favorite outdoor recreation activities 66 77* 74 72 3.28 Spiritual: I value these Forests because they are a sacred, religious, or spiritually special place to me or because I feel reverence and respect for nature there. 33 30 28 31 1.68 Subsistence: I value this Forest because it provides necessary food and supplies to sustain my life. 0 29** 23 26 10.83 Therapeutic: I value these Forests because they make me feel better, physically and/or mentally. 50* 46 47 48 3.54 Coef ﬁcients indicate % of respondents who allocated any number of hypothetical 100 dollars to a value.

ANOVA result: Forest with signi ﬁcantly higher mean frequency: * at p< .05 ** at p< .001. J.M. Clement, A.S. Cheng / Applied Geography 31 (2011) 393 e400 396 forest management and policy variable as the dependent variable.

FollowingBrown and Reed (2000) , the discriminant function analysis was conducted in two ways: 1) with only the forest values and 2) with forest values and forest use attitudes as predictor variables, entering all predictor variables simultaneously in each case. The discriminant function analyses allowed us to get to the last step by informing us what values were most critical in classifying respondents into different groups. The characteristics and nature of those groups can be very important to informing a collaborative process regarding who should be involved. In order to learn more about the groups of respondents who had allocated dollars to those values, we computed Pearson correlations between the values that emerged from the discriminant function analyses with relevant attitude variables which could tell us more about those respon- dents ( Table 4 ). Such correlations can also be done with demo- graphic variables to provide a more cohesive understanding of values, attitudes, preferences and demographic information resulting from a survey to craft more targeted public processes and to understand the context of particular issues. When using tools such as the Progress Triangle as used by Daniels and Walker (2001) to design a collaborative process, this information can directly inform planning staff regarding substance and relationship aspects.

Results The response rate for each of the surveys was approximately 34%.

The results from the frequency analysis indicate that there is a great deal of agreement across all three Forests survey respondents regarding which values have the most bearing, i.e. aesthetic, recrea- tion, biodiversity and future values. On the other hand, there were also some signi ﬁcant differences between the Forests: the importance of economic value is greater on the two more rural Wyoming forests than on the more urban PSI. Values that rated higher on the PSI than in Wyoming forests were learning, life sustaining and therapeutic. The comparison of means for attitudes and preferences across the three Forests generates some interesting insights ( Table 2). Most national forest use attitudinal variables had favorable rankings, except for oil and gas drilling, which was ranked fairly low in terms of favorability; residents near the SNF were signi ﬁcantly more favor- able towards oil or gas drilling than on the BTNF or the PSI but still rated the use low on the Likert scale. PSI respondents weakly oppose motorized recreation, whereas BTNF and SNF respondents saw the use slightly more favorably. Generally, while Wilderness designa- tion was viewed generally favorably, PSI respondents were more favorable than BTNF and SNF respondents. Where wildlife consid- erations were concerned the BTNF and SNF respondents showed signi ﬁcantly more favor, e.g. wildlife viewing, sport hunting, sport ﬁ shing and logging to create or improve wildlife habitat. Hence, it appears that BTNF and SNF respondents epossibly due to living in smaller, more rural municipalities ehave a more favorable attitude towards human use and interaction with the national forests than PSI respondents, who tend to be more urban.

In terms of preferences for vegetation management prescrip- tions and policies, mechanical thinning operations were viewed Table 3 Discriminant Analysis between values and forest policy preference variables for each forest.

PSI Off-Highway Use BTNF Oil and Gas Drilling SNF Wood Products Logging Predictive variables used Values only Values & Att Values only Values & Att Values only Values & Att Eigenvalue .241.627 .3261.78 .251 .616 % variation explained 8995 8697 6077 Classi ﬁcation by probability 20% 20%20%20% 20%20% % grouped cases correctly classi ﬁed 41% 48% 40%60% 4154 Most important predictors Recreation Economic Life sustaining LearningFavor or Oppose OHV Use Recreation Economic Life-sustaining Learning Economic Learning Future Biodiversity Recreation Life sustainingFavor or Oppose Oil/Gas Drilling Economic Learning Future Biodiversity Recreation life sustaining Economic Recreation Intrinsic Logging for Wood Products Learning Subsistence Recreation Spiritual Wilks ’Lambda .783,p< .000 . 594, p< .000 .717,p< .000 .338, p< .000 .680,p< .000 .517 p< .000 Italics indicates attitude variable used.

Table 4 Pearson correlations between values and Forest use attitudes or preferences.

Forest issue A B E F I L LS R S Sb T PSI eOff-highway vehicles on forest Motorized Recreation .102* .227** .080* .119** .156** .386** .084* Non-motorized recreation .236** BTNF eOil and gas drilling on the Forest Sport ﬁshing .116* .157** Sport hunting .100* .099* .172** .183** .121** Wildlife Viewing .104*.098* .223** .096* Oil/Gas .117* .426** .132** .106* .165** .145** .144** .177* Wilderness .117* .111* .208** .102* .180** .122** .255** .130* .129*** Fish/Wildife Habitat .148** .112* .098* .095** .162** SNF - Logging for wood products Commercial Pro ﬁt .373** .149** .157** .101* .170** Protect life/property .198**.112* .212** Remove dead/insect-infested trees .208**.150** .112* .153** Create or improve wildlife habitat .158** .146** No logging .180** .121* .175** .121* No signi ﬁcant correlations resulted for Cultural or Historical Values.

Top Row Values: A ¼Aesthetic, B eBiological Diversity, E ¼Economic, F ¼Future, I ¼Intrinsic, L¼ Learning, LS ¼Life Sustaining, R ¼Recreation, S ¼Spiritual, T ¼Therapeutic.

** Correlation is signi ﬁcant at the .01 level (2-tailed). * Correlation is signﬁ cant at the .05 level (2-tailed). J.M. Clement, A.S. Cheng / Applied Geography 31 (2011) 393 e400 397 with support when such operations are directed at protecting life and property, removing dead or insect-infested trees, and creating and/or improving wildlife habitat. Logging solely for commercial proﬁt, clearcutting, and doing nothing ( “No Action ”) were viewed negatively. However, respondents on the two Wyoming national forests were more favorable towards logging for a variety of purposes, including commercial pro ﬁt, and were more ﬂexible regarding silivicultural techniques to reduce wild ﬁre risk. Respon- dents across all three Forests were generally ambivalent to prescribed ﬁres.

The value orientation, attitudinal and preference data, while providing a snapshot of the respondents, have limited utility in and of themselves. However, discriminant function analyses can employ value orientations to classify respondents according to their management and policy preferences. Discriminant function analyses were performed with values alone, and with values and one relevant attitude variable together. In each case, we found, as expected, that attitudes have an important predictive role to play ( Vaske & Donnelly, 1999 ). However, we were more interested in the role of value orientations in explaining the variance among residents ’ preferences. The discriminant function analysis results suggest that there were some values that were more prevalent in terms of helping to correctly classify groups of respondents for each issue, namely, recreation, economic, life sustaining and biodiversity values ( Ta bl e 3 ). In each case, the value orientations increased the accuracy to classify respondents ’preferences over probability alone. All resulting Wilks ’lambda statistics were signi ﬁcant at the .000 level.

This allowed us to discover notable contradictions, or “hot- spots ”, that are helpful to understand in forest planning or collab- oration. Using the BTNF oil and gas drilling issue as an example, we found that the values that have the most important role to play in classifying our respondents into groups that favor or disfavor oil and gas leasing on the forest include economic, learning, future, biodiversity, life sustaining and recreation. We then explored those values in more detail in relation to other related attitude and preference variables to get a better sense of the relationship between our residents ’values, attitudes and preferences. It appears that the two values that correlate positively with oil and gas leasing, recreation and economic, also correlate positively with sport hunting and ﬁshing. However, those who correlated posi- tively with oil and gas, and sport hunting and ﬁshing also correlated negatively with wilderness and ﬁsh and wildlife habitat. It is understandable that respondents who rank the economic value of a forest highly would oppose wilderness designation, which would eliminate the use of that Forest for mining and gas leasing. But there does seem to be a contradiction in that the value orientations of the respondents who favor sport hunting and ﬁshing would not correlate with ﬁsh and wildlife habitat designation.

Why might certain sets of respondents classi ﬁed by value orientations hold certain policy preferences or exhibit seemingly contradictory preferences? Examining value orientations in light of attitudes towards forest uses using Pearson correlations can provide answers. For each Forest, we identi ﬁed a critical issue that faces dif ﬁcult resolution in its respective forest planning process e off-highway motorized vehicle recreation on the PSI, oil and gas drilling on the BTNF, and commercial logging on the SNF ( Table 4).

Across all three Forests and across all of the issues, economic and recreation values are strong, consistent indicators of positive and negative attitudes or preferences. For example, regarding the PSI ’s motorized recreation use issue, recreation value was most strongly positively associated with motorized recreation (along with economic value) and negatively associated with non-motorized use. When looking at the frequency data in Table 1, neither recre- ation nor economic value ranked high in their average allocations of the hypothetical $100. This suggests that, as an issue, off-highway recreation use and management impacts a relatively small number of people who prefer this use quite intensely, con ﬁrming what researchers have observed with “unmanaged recreation ”(Brooks & Champ, 2006 ). Engaging these users in a pro-active, collaborative approach has been successful in other cases when approached with care and professional mediation ( Forester, 2006).

The oil and gas drilling issue on the BTNF also presents a possible “ hotspot ”, a more complicated, wicked problem, as multiple values are associated with the various resource uses and management options associated with the issue. Values that positively correspond to, for example, Wilderness designation attitudes are predictable, such as aesthetic, biodiversity and therapeutic, but some are unex- pected, such as learning. Interestingly, recreation is negatively correlated with Wilderness, likely due to the prevalence of users who wish to preserve motorized vehicle access for various uses, including off-highway recreation and sport hunting. Regarding commercial logging on the SNF, values corresponding to negative attitudes included future, intrinsic, and learning; however, intrinsic value is positively associated with performing logging to remove dead or insect-infested trees. Spiritual value becomes a predictor of negative attitudes towards logging for purposes other than pro ﬁt.

Implications Generally the value orientations that emerged as most impor- tant to respondents on these three forests overlap with other studies ( Bright & Manfredo, 1995; Brown & Reed, 2000; Teel & Manfredo, 2009; Vaske & Donnelly, 1999 ). Aesthetic, future and biodiversity values lean more towards more biocentric values in the sense that while there is a component that bene ﬁts humans, these values can also beneﬁ t nature for its own sake. Recreation value scored high with respondents, which is directly a more anthropo- centric value. The argument can be made that a human desire to recreate can lead indirectly to bene ﬁts on a landscape, directly it does not provide beneﬁ t nature for its own sake.

One limitation is that in all three studies respondents were overwhelmingly male, only 24% of them were women for the three studies combined. This will have bearing on the results since these studies also support previous research that showed that generally women ’s value orientations are more biocentric ( Tarrant, Cordell, & Green, 2003; Vaske, Donnelly, Williams, & Jonker, 2001 ). Although there were Forest-speciﬁ c differences (e.g. women on the PSI correlated signi ﬁcantly higher with spiritual value than men), women on all three forests signi ﬁcantly differed from men by allocating more dollars to aesthetic, intrinsic, and learning value.

On the other hand in all three studies and combined there was a high correlation between men and recreational value ( .238, p < .001). As shown in the discriminant function analyses, recrea- tion value was a critical variable in classifying respondents. If our respondents had been more evenly divided between men and women, which would also more accurately re ﬂect their distribu- tions among taxpayers in these states and nationally, the results might have been different. We suggest that future research stratify studies for gender. While the random sample design of the survey allows these results to be generalized to the broader population, the population de ﬁned in these three studies was bounded geographically to counties directly adjacent to each of the national forests. Being national resources, however, the relevant population for assessing public value orientations, attitudes and preferences regarding the three national forests would be every tax-paying citizen of the U.S.

Based on the consultation received during the participatory plan- ning of the surveys, bounding the population geographically seemed to make the most sense to planners and decision-makers familiar with the proportion of visitation and uses on their J.M. Clement, A.S. Cheng / Applied Geography 31 (2011) 393 e400 398 respective Forests. Additionally,ﬁnancial, time, and operational resources constrained our ability to administer a national survey akin to Shields et al. (2002a, 2002b) . Further analysis between our dataset and the data unearthed in the Shields et al. survey to test for signi ﬁcant differences may be possible if the measures are equivalent.

Despite the limitations in the population of interest, by using the process illustrated in Fig. 1, we uncovered information that could be useful to both the USFS and public stakeholder involved in national forest planning as suggested by Allen et al. (2009). Similar to Brown and Reed ’s research, we provide quanti ﬁable relationships between people ’s value orientations and attitudes towards various forest uses and management strategies, capturing sentiments from a sample of the silent majority who might not otherwise attend public meetings or participate in an intensive collaborative process. Using value orien- tations as a starting point, our survey results can shine light on the respondents ’contexts behind forest management and policy issues.

For example, when looking at this type of data for all three forests in relation to forest vegetation management it was clear that respondents who rank recreation, economic, historic and cultural values positively have a higher comfort level with all the reasons and silvicultural techniques to reduce the risk of wild ﬁre (results not shown). Respondents who rank aesthetic, biodiversity, future, intrinsic, learning, spiritual and therapeutic values highly were more opposed to these solutions. The greatest discomfort in association with these values appears to be logging for commercial pro ﬁt, removal of dead or insect-infested trees and clear cutting as a silvi- cultural technique. The rub here is that around 90% of respondents favored removal of trees because they are dead or infested with insects, while the majority of respondents also ranked aesthetic, biodiversity and future values highest. However these value orien- tations and the removal of dead and insect-infested trees correlated negatively. Thus it appears that although respondents feel that treatments are warranted when stands are affected by insects, there are contextual factors such as location, silvicultural treatment and management purpose which also matter. For instance, an insect- infected tree removal project might ﬁnd more support when also aiming to increase wildlife habitat rather than to reap economic beneﬁ ts.

Generally, the value orientations results can be interpreted to indicate that the health of these forests is most important to respondents. The majority of respondents also want insect impacted forests to be treated. However, the methods presented in these surveys only ﬁnd whole-hearted support with those respondents who also rank recreation and economic values highly.

This too indicates the presence of a “hotspot ”, where there may be discrepancies between respondents ’forest value orientations and the management options that are generally discussed, including in these surveys. It also indicates that one-size ﬁt all policies and management solutions will not corresponded with constituents ’ attitudes and preferences.

Armed with this awareness of people ’s value orientations towards the issue of logging, managers can demonstrate and communicate a deeper level of understanding about people ’s values and more directly address issues on the public ’s terms, not the USFS ’s terms. The results also demonstrate that, underneath the super ﬁ cial bl ack-and-white portrayals of national forest management issues (i.e., jobs vs. the environment), a mix of social value orientations, attitudes and preferences esome of which appear to be contradic- tory eunderlie why people favor or oppose certain forest policies and management strategies. By accounting for the range of social value orientations and attitudes associated with forest plan deci- sions, the public may gain a more nuanced understanding of why certain decisions were made, and can appreciate trade-offs and contradictions between values and attitudes/preferences. Conclusion Value orientation studies using this 12 or 13-scale typology can help considerably to describe a Forest or a place, on a local Forest- scale, a local place-scale or a regional, multi-state scale. These three studies combined can be used to describe the value orientations, attitudes and preferences that respondents ’have in common across a regional scale, and can be used to compare between Forests.

To add a note regarding the involved survey logistics, these surveys took each about nine months to a year from initiation of the process, through public participative survey design, data collection, data entry and analysis. One person oversaw the whole process and was deeply involved in communications with the USFS and its stakeholders to design the survey, and one person conducted the printings, mailings and data entry on a part-time basis. Altogether the costs were around $ 10,000.00 for printing, mailings and data entry, including the part-time logistical support. The cost related to the expertise needed to oversee the project, design the survey and write the results into various publications may vary. This meth- odology can therefore provide a considerable resource to planning and collaborative efforts at relatively little cost. For more consid- erations, please see further Brown and Reed (2000)andReed and Brown (2003) .

Studies such as these can also help make planning efforts more ef ﬁcient. Understanding that respondents generally value forests for their aesthetic and biodiversity characteristics, for the recrea- tion opportunities they present and respondents ’desire for future generations to enjoy the same provides a solid foundation from which to build. Understanding that intrinsic and economic value orientations, for example, are not as widespread is equally impor- tant. This knowledge can narrow down that policies and projects that con ﬂict with this value orientation pro ﬁle will require more time and deliberation.

Additionally, the results from these studies provide explanatory information, but they also indicate where there are “hot spots ”,areas where values, attitudes and preferences do not line up, and appear to indicate contradictions. This may mean that more information is required to discover the source of the hot spot, or it may provide an opportunity to deliberate new solutions. Either way, more mean- ingful forms of public participation, and more deliberative approaches with stakeholders, can use the information from these studies, and the hotspots they indicate, as a starting point. This too can provide greater ef ﬁciency to planning processes at any scale.

A limitation of surveys is that they capture a point in time, not a trend of any kind. Another limitation is that although a great deal of information can be gathered across many people, they cannot in depth explain the reasons for the responses gathered. We have endeavored to capture that deeper understanding through Q- methodology studies on both the PSI and the BTNF and would recommend this for future studies that explore complex issues and the value orientations that underlie them. A last limitation is the limited amount of women respondents. Although in human dimensions in natural resources related studies it is often the case that the majority of respondents are men, the distinct differences in value orientations across many values studies indicates a need to stratify samples for gender, to ensure that populations are accu- rately represented in these studies.

Acknowledgements We thank the staffs of the Bridger-Teton National Forest, Pike- San Isabel National Forests and Shoshone National Forest for their tremendous cooperation and the Governor of Wyoming ’sOf ﬁce, in particular Temple Stevenson, for funding and support. We also extend great thanks to Julie Schaefers with Region 2 of the USDA J.M. Clement, A.S. Cheng / Applied Geography 31 (2011) 393 e400 399 Forest Service for all her support and advice in conducting these studies. We also thank Kathie Mattor for her assistance in con- ducting the SNF and BTNF studies. Finally, we thank the many interested forest collaboration participants in Wyoming who provided suggestions and comments along the way.

References Ajzen, I. (1991). The theory of planned behavior.Organizational Behavior and Human Decision Processes, 50 ,179e211.

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior . New Jersey: Prentice-Hall, Inc.

Alessa, L., Kliskey, A., & Brown, G. (2008). Social-ecological hotspots mapping: a spatial approach for identifying coupled social-ecological space. Landscape and Urban Planning, 85 ,27e39.

Allen, S. D., Wickwar, D. A., Clark, F. P., Dow, R. R., Potts, R., & Snyder, S. A. (2009).

Values, attitudes, and beliefs technical guide for Forest Service land and resource management, planning, and decision-making . Portland, OR: US Department of Agriculture, Forest Service, Paci ﬁc Northwest Research Station. Gen. Tech. Rep.

PNW-GTR-788.

Bright, A., & Manfredo, M. (1995). The quality of attitudinal information regarding natural resource issues: the role of attitude-strength, importance and infor- mation. Society and Natural Recources, 8 , 399e414.

Bright, A., Manfredo, M., & Fulton, D. (2000). Segmenting the public: an application of value orientations to wildlife planning in Colorado. Wildlife Society Bulletin, 28 (1), 218 e226.

Brooks, J. J., & Champ, P. A. (2006). Understanding the wicked nature of “unman- aged recreation ”in Colorado ’s front range. Environmental Management, 38 (5), 784 e798.

Brown, G. (2005). Mapping spatial attributes in survey research for natural resource management: methods and applications. Society and Natural Resources, 18,17e39.

Brown, G. (2006). Mapping landscape values and development preferences: a method for tourism and residential development planning. International Journal of Tourism, 8 (2), 101e113 .

Brown, G., & Reed, P. (2000). Validation of a forest values typology for use in national forest planning. Forest Science, 46(2), 240e247.

Brown, G., & Reed, P. (2009). Public participation GIS: a new method for use in national forest planning. Forest Sciences, 55,166e182.

Brown, G., Reed, P., & Harris, C. (2002). Testing a place-based theory of environ- mental evaluation: an Alaska case study. Applied Geography, 22,49e76.

Brown, G., Smith, C., Alessa, L., & Kliskey, A. (2004). A comparison of perceptions of biological value with scienti ﬁc assessment of biological importance. Applied Geography, 24 (2), 161e180.

Brown, T. (1984). The concept of value in resource allocation. Land Economics, 60(3), 231 e246.

Burns, S., & Cheng, A. S. (2005). The utilization of collaborative processes in forest planning. Report to the USDA forest service . Durango, CO: Ofﬁ ce of Community Services, Fort Lewis College.

Committee of Scientists. (1999). Sustaining the people’s lands: Recommendations for stewardship of the national forests and grasslands into the next century .

Washington, D.C.: U.S. Department of Agriculture.

Cortner, H. J., & Moote, M. A. (1999). The politics of ecosystem management.

Washington, D.C.: Island Press.

Daniels, S. E., & Walker, G. B. (2001). Working through environmental conﬂict: The collaborative learning approach . Westport, Conn: Praeger Publishers.

Dillman, D. (1978). Mail and telephone surveys: The total design method . New York, NY: Wiley.

Dillman, D., Smyth, J., & Christian, L. (2007). Internet, mail and mixed-mode surveys:

The tailored design method (3rd ed.). Hoboken, N.J.: John Wiley & Sons, Inc.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An intro- duction to theory and research . Reading, MA: Addison-Wesley.

Forester, J. (2006). Making participation work when interests conﬂ ict: moving from facilitating dialogue and moderating debate to mediating negotiations. Journal of the American Planning Association, 72(4), 447 e456.

Fulton, D., Manfredo, M., & Lipscomb, J. (1996). Wildlife value orientations: a conceptual and measurement approach. Human Dimensions of Wildlife, 1 (2), 24e47.

Fulton, David C., Skerl, Kevin, Shank, Erin M., & Lime, David W. (2004). Beliefs and attitudes toward lethal management of deer in cuyahoga valley national park.

Wildlife Society Bulletin, 32 (4), 1166e117 6 .

Gericke, K. L., Sullivan, J., & Wellman, J. D. (1992). Public participation in national forest planning: perspectives, procedures, and costs. Journal of Forestry, 90(2), 35e38.

Grilliot, Amy L., & Armstrong, James B. (2005). A comparison of deer hunters with disabilities and nondisabled hunters in Alabama: motivations and satisfactions in deer hunting. Wildlife Society Bulletin, 33 (1), 243e250.

Haas, G. E. (2003). Restoring dignity to sound professional judgment. Journal of Forestry, 101 (9), 38e43.

Hays, S. P. (2009). The American people and the national forests: The ﬁrst century of the U.S. forest service (1st ed.). Pittsburgh: University of Pittsburgh Press.

Homer, P., & Kahle, L. (1988). A structural equation test of the value-attitude- behavior hierarchy. Journal of Personality and Social Psychology, 54, 638 e646.

Larsen, G., Holden, A., Kapaldo, D., Leasure, J., Mason, J., Salwasser, H., et al. (1990).

Synthesis of the critique of land management planning . Washington, DC: U.S.

Department of Agriculture, Forest Service. Lubell, Mark, & Vedlitz, Arnold (2006). Collective action, environmental activism, and air quality policy. Political Research Quarterly, 59 (1), 149e160.

Machlis, Gary E., Kaplan, Amanda B., Tuler, Seth P., Bagby, Kathleen A., & McKendry, Jean E. (2002). Burning questions: A social science research plan for federal wildland ﬁre management . Moscow: University of Idaho.

Manning, R., Valliere, W., & Minteer, B. (1999). Values, ethics, and attitudes towards national forest management: an empirical study. Society and Natural Resources, 12 (5), 421 e436.

McGranahan, D. A. (1999). Natural amenities drive rural population change .

Washington, D.C.: U.S. Department of Agriculture, Economic Research Service.

Nielsen-Pincus, M. (2006). The nuts and bolts of attachment to landscap-scale places: three case studies of social values in a landscape context. In Interna- tional Association for Society and Natural Resources. (Ed.), 12th International symposium on society and resource management . Vancouver, B.C: International Association for Society and Natural Resources.

Raymond, C. M., Bryan, B. A., MacDonald, D. H., Cast, A., Strathaern, S., Grandgirard, A., et al. (2009). Mapping community values for natural capital and ecosystem services. Ecological Economics, 68, 1301e1315 .

Reed, P., & Brown, G. (2003). Values suitability analysis: a methodology for iden- tifying and integrating public perceptions of ecosystem values in forest plan- ning. Journal of Environmental Planning and Management, 46(5), 653 e658.

Rokeach, M. (1973). The nature of human values. New York, NY: Free Press.

Rolston, H. (1981). Values in nature. Environmental Ethics, 3,113e128.

Rolston, H., & Coufal, J. (1991). A forest ethic and multivalue forest management: the integrity of forests and of foresters are bound together. Journal of Forestry 35 e40.

Sampson, R. N., & Adams, D. L. (1994). Assessing forest ecosystem health in the inland west . Binghampton, NY: Food Products Press.

Selin, S. W., Schuett, M. A., & Carr, D. S. (1997). Has collaborative planning taken root in the national forests? Journal of Forestry, 95(7), 17e21.

Sherrouse, B. S., Clement, J. M., & Semmens, D. J. (2011). A GIS application for assessing, mapping and quantifying the social values of ecosystem services.

Applied Geography, 31 (2), 748e760.

Shields, D., Martin, I., Martin, W., & Haefele, M. (2002a). Survey results of the american public ’s values, objectives, beliefs and attitudes regarding forests and grasslands. In Gen. Tech. Rep. RMRS-GTR-95 . Fort Collins, CO: USDA Forest Service, Rocky Mountain Research Station.

Shields, D. J., Martin, I. M., Martin, W. E., & Haefele, M. A. (2002b). Survey results of the American public ’s values, objectives, beliefs, and attitudes regarding forests and grasslands: A technical document supporting the 2000 USDA Forest Service RPA Assessment . Gen. Tech. Rep. RMRS-GTR-95. Fort Collins, CO: US Department of Agriculture, Forest Service, Rocky Mountain Research Station.

Shindler, B., List, P., & Steel, B. S. (1993). Managing federal forests: public attitudes in Oregon and nationwide. Journal of Forestry, 91(7), 36e42.

SPSS 13.0 Base System . (2004). Chicago. IL: SPSS Inc.

Stein, S. (2005). Poll: Colorado voters say future population pressures will destroy quality of life [cited 10-10-2005 2005]. NPG, Inc. Available from: http://www.

npg.org/pressreleases/npg_03_28_01.html .

Stevens, J. (2002). Applied multivariate statistics for the social sciences (4th ed.).

Mahwah, NJ: Lawrence Erlbaum Associates.

Tarrant, M. A., Cordell, H. K., & Green, G. T. (2003). PVF: a scale to measure public values of forests. Journal of Forestry .

Tarrant, M. A., & Cordell, H. K. (2002). Amenity values of public and private forests:

ex amining the value-attitude relationship. Environmental Management, 30(5), 692 e703.

Taylor, D. T., Coupal, R. H., Foulke, T., Rashford, B., & Olsen, D. (2008). An economic pro ﬁle of the Bridger-Teton National Forest, phase II ﬁnal report . Laramie, WY:

University of Wyoming, Department of Agricultural and Applied Economics.

Taylor, D. T., Foulke, T., & Coupal, R. H. (2008). An economic proﬁle of the Shoshone national Forest: Prepared in support of the Shoshone national forest plan revision process . Laramie, WY: University of Wyoming, Department of Agricultural and Applied Economics.

Teel, T., & Manfredo, M. (2009). Understanding the diversity of public interests in wildlife conservation. Conservation Biology, 24 (1), 128e139.

USDA Economic Research Service. (2005). Colorado fact sheet[cited 10-10-2005 2005]. US Department of Agriculture. Available from: http://www.ers.usda.gov/ StateFacts/CO.htm .

USDA Forest Service. (2006). USDA Forest Service pike & san isabel national forests, Cimarron & Comanche national grasslands . US Department of Agriculture, Forest Service, Pike-San Isabel National Forest. [cited May 7, 2006].

Vaske, J. J., Donnelly, M. P., Williams, D. R., & Jonker, S. (2001). Demographic inﬂuences on environmental value orientations and normative beliefs about national forest management. Society & Natural Resources, 14 ,761e776.

Vaske, J. J., & Donnelly, M. (1999). A value-attitude- behavior model predicting wildland preservation voting intentions. Society and Natural Recources, 12(6), 523 e537.

Wilkinson, C. F., & Anderson, H. M. (1987). Land and resource planning in the national forests . Covelo, CA and Washington, D.C.: Island Press.

Williams, D. R., & Stewart, S. I. (1 998). Sense of place: an elusive concept that is ﬁnding a home in ecosystem management. Journal of Forestry, 96(5), 18 e23.

Wondolleck, J. M., & Yaffee, S. L. (2000). Making collaboration work: Lessons from innovation in natural resource management . Covelo, CA: Island Press.

Zube, E. (1987). Perceived land use patterns and landscape values. Landscape Ecology, 1 (1), 37e45. J.M. Clement, A.S. Cheng / Applied Geography 31 (2011) 393 e400 400