Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers

FORPOL-01295; No of Pages 11 Forest Policy and Economics xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Forest Policy and Economics journal homepage: www.elsevier.com/locate/forpol

Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers Reem Hajjar ⁎, Robert A. Kozak Faculty of Forestry, University of British Columbia, 2900-2424 Main Mall, Vancouver, BC V6T 1Z4, Canada

a r t i c l e

i n f o

Article history: Received 21 October 2014 Received in revised form 29 June 2015 Accepted 7 August 2015 Available online xxxx Keywords: Climate change Forest management strategies Forest adaptation policy Public perceptions Assisted migration

a b s t r a c t Various reforestation strategies that could potentially help forests adapt to a changing climate are currently being debated. We sought to gauge the public's acceptance levels of different reforestation strategies, and explore which factors seem to be associated with people's willingness to accept different forms of human and technological intervention in forest management. To do so, a public survey was administered in British Columbia and Alberta, Canada to assess acceptance of different forest adaptation strategies that could be employed to adapt to a changing climate, and explore variables identified from the literature as potentially affecting public decisions on biotechnology and assisted migration. A logistic regression was used to determine the degree to which variables identified in the literature are associated with levels of acceptance of different forest management strategies. What emerged was an explanatory model that can be used as a starting point to further engage the public in a discussion over appropriate and acceptable technologies and policies to help forests adapt to a changing climate. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Forests have become increasingly vulnerable to mortality due to the direct and indirect effects of climate change (Fettig et al., 2013). In Western Canada, recent increases in the frequency and severity of natural disturbances in forests, such as wildfires, pest outbreaks, and droughts, have been attributed to a changing climate (Aubin et al., 2011; Parkins and Mackendrick, 2007, and citations within). Reforestation practices that source seeds locally as a principal means of local adaptation no longer ensure forest health and productivity, as local climates are expected to change over the coming decades (Wang et al., 2012) and individual trees and populations may be maladapted when exposed to conditions outside of their climate niche (Fettig et al., 2013). For these reasons, scientists have been considering different strategies to inform policy on helping forests adapt to changing climates, such as assisted migration – moving a population or a species outside its historic range to habitats that would become more suitable than present habitats in the future climate (Hewitt et al., 2011; Ste-Marie et al., 2011). However, such interventions are not without controversy, and scientists continue to debate the efficacy of adaptation strategies like assisted migration (Hewitt et al., 2011). Much of this discourse revolves around the ecological risks and benefits of such interventions (Aubin ⁎ Corresponding author at: International Forestry Resources and Institutions, School of Natural Resources and Environment, University of Michigan, 440 Church St., Ann Arbor, MI 48109, USA. E-mail address: [email protected] (R. Hajjar).

et al., 2011), but public opinions regarding these sorts of adaptation strategies are also an issue. Given the substantial uncertainty associated with assisted migration, and the complex scientific, social, policy, and economic implications of the topic, Aubin et al. (2011) argue that the debate on assisted migration is based on more than just an objective risk assessment that weighs the ecological risks and benefits; the debate should also include a discussion of beliefs, values, visions of the future, and subjective perceptions of risk and desirable outcomes (Minteer and Collins, 2010; Aubin et al., 2011). While studies on public perceptions of biotechnology (Durant and Legge, 2005; Bonny, 2003; Costa-Font et al., 2008; Fischhoff and Fischhoff, 2001) and climate change (Kellstedt et al., 2008; Slimak and Dietz, 2006; Maibach et al., 2009) abound, such studies situated within the forest sector are comparatively rare. In this paper, we sought to gauge the public's acceptance levels of different reforestation strategies, including assisted migration, natural regeneration, seed selection for future climates, and use of genetically engineered seeds. We further sought to explore which factors, such as risk perception, knowledge, values, and attitudes, are associated with people's willingness to accept different forms of human and technological interventions in forest adaptation strategies, with a view to informing the policy-making process for forest adaptation. To do so, we tested the association between acceptance of different forest adaptation strategies and variables identified by various studies on attitudes surrounding biotechnology, as well as thought pieces on climate change and assisted migration technologies. These variables formed the basis of a questionnaire administered to the general public in British Columbia and Alberta which also assessed the public's acceptance of various

http://dx.doi.org/10.1016/j.forpol.2015.08.004 1389-9341/© 2015 Elsevier B.V. All rights reserved.

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004

2

R. Hajjar, R.A. Kozak / Forest Policy and Economics xxx (2015) xxx–xxx

forest management strategies that could be used to adapt forests in Western Canada to projected future climate conditions. 2. Background In order to contextualize this study, background information on current climate change adaptation strategies in the forestry sector of Western Canada is provided, followed by a review of the variables explored in this research as being associated with acceptance of forest adaptation strategies. 2.1. Climate change and forest adaptation in Western Canada Current action on adaptation to climate change in Western Canada's forests could best be described as nascent and evolving. While policy surrounding assisted migration has yet to be enacted, the provincial governments of both Alberta and British Columbia seemed poised and willing to adopt this strategy as a means of adapting to climate change. For instance, tree improvement and adaptation programs in Alberta acknowledge that climate change will bring about challenges for the forest sector and Alberta Environment and Parks recommends that trees be planted to better suit future conditions that are expected to be warmer and drier. To that end, they have funded a number of research projects and pilot programs related to tree breeding and assisted migration strategies (Alberta Environment and Parks, 2015). In British Columbia, the Ministry of Forests, Lands and Natural Resource Operations has released a number of relevant reports, including a Climate Change Strategy, 2013–2018 (British Columbia Ministry of Forests, Lands and Natural Resource Operations, 2013a), a Climate Action Roadmap (British Columbia Ministry of Forests, Lands and Natural Resource Operations, 2013b), and a Forest Stewardship Action Plan for Climate Change Adaptation, 2012–2017 (British Columbia Ministry of Forests, Lands and Natural Resource Operations, 2012). The latter document outlines three goals for adapting forests to future climate conditions: fostering resilient forests; maintaining future options and benefits; and building adaptive capacity. Under each, a series of practicable actions are proffered, for example, to prepare a charter and implementation plan for climate-based seed transfer program (British Columbia Ministry of Forests, Lands and Natural Resource Operations, 2012). Like the Alberta government, the British Columbia Ministry of Forests, Lands and Natural Resource Operations has also been very active in supporting research and pilot studies pertaining to the adoption of assisted migration as a viable climate adaptation strategy. 2.2. Variables affecting public acceptance of new forest adaptation technologies A number of factors can affect public acceptance of new technologies. For GMO technology, Costa-Font et al. (2008) found that personal attitudes are formed by complex decision-making processes based on values, knowledge, trust, and risk perception, among other factors. These and other factors, such as desired outcomes, past hardships, and beliefs, have been found to affect public opinions of other such technologies (e.g. Aubin et al., 2011; Marx et al., 2007). We chose to test the importance of these variables in this present study on public acceptance of forest adaptation technologies. While it is not our intent to provide a comprehensive literature review on this ongoing debate (there are many works related to this, e.g. Adger et al., 2008; Aubin et al., 2011; Minteer and Collins, 2010; Slimak and Dietz, 2006), in order to explain our choices of variables in this study, we offer the following brief review of the above-mentioned factors and how they might affect a person's decision to accept or reject forest adaptation strategies. All the variables discussed were selected for exploration given their prominence in the literature on public acceptance of new technologies in the field of biotechnology in agriculture and the limited available literature related to climate change and forestry. While a systematic review of the literature was not conducted, these variables repeatedly emerged

in both empirical studies and synthetic reviews, which were found using various combinations and variations of search terms such as public acceptance/perceptions, biotechnology, climate change, GMO acceptance, assisted migration, and adaptation. 2.2.1. Environmental values and beliefs Values, in the context of forest management, can be defined as, “… culturally and emotionally informed beliefs about desirable and appropriate standards for judging appropriate actions and goals,” influenced by the information received through social ties and other means (Harshaw and Tindall, 2005, p.435). Where the environmental ethics debate is concerned, the value-based concepts of anthropocentrism, ecocentrism, and anthropocentric conservationism are very much at the fore. These values and beliefs can shape attitudes towards forest adaptation strategies. Anthropocentrists, those who see the primary purpose of nature as a resource for human use, might argue that the risks of climate change can be mitigated using human knowledge and technological fixes (Hulme 2009, as cited in Aubin et al. (2011)) and, thus, might be more willing to accept higher degrees of human intervention to help forests adapt. Conversely, ecocentrists, those who emphasize the intrinsic value of nature in a pure and pristine state that should be protected independent of human use and benefits (Osanken in Aubin et al. (2011)), might argue that human and technological interventions for forest adaptation are diverting attention from the need for humans to reduce their impact on nature (Aubin et al., 2011). On the other hand, ecocentrists might also be interested in using such technologies to prevent species extinction, or to preserve vital ecosystem processes that are at risk due to climate change (Aubin et al., 2011). This so-called anthropocentric conservationist view capitulates to the fact that species are in trouble because of human-induced climate change, and that the more pragmatic alternative is to manage populations and species to help them survive changing conditions (Minteer and Collins, 2010, p.1801). Similarly, Donlan et al. (2005) point to an ethical responsibility to save species that we have put at risk through our own actions. Thus, individuals with more ecocentric values may hesitate when considering where to place the threshold of human manipulation of nature; when exactly is a ‘technofix’ acceptable? Steg and Sievers (2000) approach the issue of values using the lens of cultural theory and its effects on risk perception and preference for risk management strategies. In discussing mitigation strategies, they categorize the following: individualists believe that technology will fix the problem so that there is no need to change human behavior for mitigation; hierarchists will put trust in the government or experts to tell them when the limits to environmental damage are reached and when to control resource use; egalitarians have a strong sense of responsibility for the environmental problem and prefer changing human behaviors and reducing needs rather than controlling resources; and fatalists feel that there is no point in learning about and managing nature since nothing is controllable and all is left up to chance (Steg and Sievers, 2000). In the context of forest management, individualists and hierarchists may be more willing to accept adaptation strategies due to their belief in the ‘technofix’ and their trust in expert decisions, while egalitarians may be more hesitant as they would be more inclined to change human behavior rather than changing nature. The choices of fatalists would be unpredictable. 2.2.2. Desired outcomes Acceptance of adaptation strategies may depend on desired outcomes (Aubin et al., 2011) – what society prioritizes in forest management. Desired outcomes can be a manifestation of values and beliefs, and can include non-environmental values such as social, cultural, and economic values. For example, if the desired outcome is to maintain or enhance the productive capacity of forests and, thus, sustain a sector of great importance in many locations, perhaps one would be more accepting of strategies that promise to do just that. If biodiversity

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004

R. Hajjar, R.A. Kozak / Forest Policy and Economics xxx (2015) xxx–xxx

conservation is a priority, then one may either be for or against adaptation strategies: against, if one thinks nature should be left as is and without human intervention for adaptation; or for, if one aligns with anthropocentric conservationism. 2.2.3. Perceptions of risk While Wachinger et al. (2013) state that the connection between risk perception and willingness to act is still not clear, Parkins and Mackendrick (2007) argue that individuals who perceive a risk are more likely to act to mitigate that risk. For this study, we are concerned with perceptions of risk associated with two issues: climate change and the degree to which it will affect our forest ecosystems and, thus, our daily lives; and, the technologies used to help forests adapt to climate change. For the former, if one perceives that the risk of climate change is high, the likelihood to mitigate that risk may increase. Perhaps one would be more willing to undertake or accept actions that would help our ecosystems to adapt to changes, thus reducing the impacts of climate change.1 For the latter, one would expect that those who viewed such technologies as risky would be less inclined to accept them. Debate on the ecological risks associated with various biotechnologies or species introductions is not limited to the scientific community – public perceptions of these risks will also likely affect their willingness to accept them. Many factors can contribute to risk perception, although a review by Wachinger et al. (2013) shows diverging views on these. That said, Slimak and Dietz (2006) found that values and beliefs were good predictors of ecological risk perception. Counter to studies showing that risk perception was influenced by demographic factors, such as age, schooling, income, and gender (e.g. Kellstedt et al., 2008; Savage, 1993), Slimak and Dietz (2006) found that neither demographics nor social structural influences were as important as values and beliefs (Slimak and Dietz, 2006). In addition, risk perception can be influenced by past hardships experienced, knowledge and awareness, and institutional trust (Kellstedt et al., 2008), each reviewed below. 2.2.3.1. Past hardships. Perception of risk can be influenced by past hardships that one has experienced. Marx et al. (2007) argue that experiential processing dominates our analytic processing; personal experience or exposure to others' experiences (vicarious experience) will motivate action more than an analytical understanding of climate risk. Weber (2006) echoes this conclusion, stating that recent personal experiences greatly influence risk management actions, more so than receiving abstract information provided by experts. 2.2.3.2. Knowledge and awareness. How well-informed one is, and how aware one is of their knowledge of the subject, may affect risk perception and even influence values and beliefs. However, using knowledge-deficit theory to explain risk perception has recently come under scrutiny (reviewed in McFarlane et al. (2012)). In some cases, knowledge, or lack thereof, has been shown to be a poor predictor of risk perception or response to risk (Slovic, 1999) as other subjective and contextual factors may also play important roles (Whitfield et al., 2009). In the global warming debate, Kellstedt et al. (2008) argue that the media's presentation of the issue as an unsettled controversy diminishes the effect that increasing information levels might have on risk perception. 2.2.3.3. Trust in science and decision-makers. Risk perception can be either positively or negatively influenced by trust in scientists. For instance, one may place trust in science or scientists to provide correct and objective information on risks; thus, if scientists are worried about climate change, then one should be worried too. On the other hand, trust in 1

One might also undertake actions to mitigate climate change as well, rather than just adapt to it. Alternatively, one may have high perceived risk, but do nothing about it for various reasons (feeling ineffectual, competing interests, etc.).

3

scientists might also lead to the belief that they will devise the best solutions to mitigate risks, thus potentially lowering risk perception. Kellstedt et al. (2008) found this to be the case in their survey on climate change awareness and risk perception in the United States. Taking this further, people's acceptance of adaptation strategies to mitigate risks might be influenced by the confidence that they have in others who can make more informed decisions on those strategies. It has been shown that trust in experts and management agencies is important both to risk perceptions and to the acceptance of risk management options in the forest sector (as reviewed in McFarlane et al. (2012)). Thus, in this case, even though trust in scientists may translate to less perceived risk of climate change, it may also mean that adaptation interventions are more likely to be accepted, not less, as trust in scientists to know what is best is increased. Having confidence in decision-makers to make the right decisions for society might also increase willingness to accept management strategies for mitigation. Durant and Legge (2005) found that support for genetically modified foods was linked to citizen's trust in government. Trust in experts seems to be more important in situations where knowledge is lacking; however, when people are more informed about the risk, they tend to rely on their own judgment (McFarlane et al., 2012). 3. Methods Based on the literature reviewed above, we focused this study on testing the association between, and directionality of, the variables listed in Table 1 with acceptance of different forest management strategies as they relate to climate change adaptation. An online survey instrument was created using FluidSurveys to gather data on the dependent and independent variables described below. The survey, containing 39 questions (detailed in Sections 3.1 and 3.2), was distributed to the public in British Columbia and Alberta through Research Now, a research and marketing survey company. We made use of their pre-existing online panels, which were roughly representative of populations in both provinces in terms of age and gender. In order to avoid an over-representation of major metropolitan areas (MMAs) within the panel samples, a quota was programmed into the online survey to limit the number of respondents living in Vancouver and Victoria (54%), and Calgary and Edmonton (52%); these percentages are proportional to actual populations in those urban centers relative to the total population of each province. After an initial period of pre-testing, the survey was administered between 6 and 25 July 2012, which was how long it took until the desired sample size of 1500 complete responses – 750 from each province – was reached. By statistically comparing survey response patterns to known demographics in the two provinces, nonresponse bias in this sample was determined to be negligible (see Hajjar et al. (2014) for further details). Survey data were analyzed with SPSS Statistics v.20 and Microsoft Excel v.14.3.4 using three multivariate techniques. First, K-means clustering (after conducting hierarchical clustering using the Ward method to get a sense of the possible number of clusters) was used to Table 1 Directionality (based on the literature) for acceptance of scenarios with greater levels of human intervention. Question marks for the directionality indicate that the literature states that the relationship can either be positive or negative, depending on further explanatory factors. Variable

Directionality

Attitudes (characterizing values and beliefs) Perceptions of risk of climate change Perceptions of risk of reforestation technologies Prioritization of future outcomes for forests and forest-dependent communities Knowledge of climate change and reforestation technologies Past hardships experienced related to climate change Confidence in decision-makers Confidence in scientists

? + − ? ? + + +

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004

4

R. Hajjar, R.A. Kozak / Forest Policy and Economics xxx (2015) xxx–xxx

characterize the environmental values of respondents based on a series of Likert-scale statements. Second, forward stepwise binary logistic regressions were used to examine the associations between all the listed variables in the model and the likelihood of accepting each of six presented forest management intervention strategies and to determine the most robust model.2 Finally, Pearson's chi-square tests of independence (alpha = 0.05) were used to provide further clarity on some of the results.

Table 2 Six different forest management strategies for which respondents rated their levels of acceptance. Strategy 1

Strategy 2 Strategy 3

3.1. Operationalization of the dependent variable The questionnaire elicited respondents' levels of acceptance of different forest management strategies that can be used to help forests adapt to climate change. After explaining to respondents that climate change is likely to affect forests in Western Canada and that, with current forest management strategies, tree growth and survival may not be optimal given changing climate conditions, respondents were presented with six different management strategies related to replanting trees after forests have been logged. They were asked to rate their levels of acceptance for each strategy on a four-point scale (completely reject, tentatively reject, tentatively accept, and completely accept, with an additional “don't know” option). The strategies were meant to represent a continuum of options ranging from a “do nothing” approach with no human or technological interference (Strategy 1) to a technologyheavy intervention with genetic engineering of tree seedlings (Strategy 6). The six forest management strategies, as described to the respondents, are listed in Table 2. In order to run the binary logistic regressions, the dependent variable was transformed from a four-point scale of acceptance to a binary acceptance or rejection of the strategy (the more nuanced four-point scale is discussed in Hajjar et al., 2014). We ran forward logistic regressions for each strategy individually, with acceptance of the strategy as the dependent variable against the independent variables listed in Table 1 (the operationalization of these variables is discussed in the next section). We opted to run separate regressions for each strategy in turn and not combine the six strategies into one variable. The reason for this is that there is debate on the order of the scenarios in terms of which is more “intervening” than the others, or alternatively, which is less “natural” (i.e. is it less natural to move seeds or to selectively breed trees?).

Strategy 4

Strategy 5

Strategy 6

Forests are left to re-grow naturally, with no intervention by forest managers. This means climate change is allowed to run its course without any management to help forests adapt. Forest managers plant seedlings grown from seeds that are collected from forests close to the planting area. Forest managers plant seedlings grown from seeds that are collected from forests close to the planting area, but breeding takes place to select local seeds that are expected to grow well in the climate conditions expected to occur in the near future. Forest managers plant seedlings grown from seeds that come from non-local forests (i.e. forests that are distant from the planting area). Breeding takes place to select seeds that are expected to grow well in the climate conditions expected to occur in the near future. Forest managers plant seedlings from seeds of tree species that are different species than what had been there previously. As in the last strategy, the seeds are collected from another region within the province, and breeding takes place to select seeds that are expected to grow well in the climate conditions expected to occur in the near future. Forest managers plant seedlings grown from seeds that are genetically engineered (genetically modified organisms, GMOs) to grow well in the climate conditions expected to occur in the near future.

asked which of the four statements most closely matches their own views of environmental problems (parenthetic statements were not included in the survey, but represent categories of environmental views according to Steg and Sievers (2000)): • We cannot control environmental problems without changing human behavior (egalitarian); • Environmental problems can be controlled if the government dictates clear rules based on expert opinion (hierarchist); • We do not need to worry about environmental problems because they can be resolved by technological solutions (individualists); and • We cannot know whether environmental problems will become an issue or not because nature is unpredictable, so there is no point in coming up with solutions (fatalist).

3.2. Operationalization of independent variables 3.2.1. Characterizing values and beliefs Respondents were asked to state their level of agreement – on a Likert scale – with several statements used to characterize their values and beliefs related to climate change and adaptation. As the focus of the study was on forest adaptation strategies, the statements revolved around the issue of adaptation. In spite of the fact that this may provide for a limited characterization of the respondents, we opted not to characterize their environmental values in full (as the New Ecological Paradigm or other scales would have done). Instead, the statements, shown in Table 3, were chosen to get a sense of respondents' views on science and uncertainty, risks of inaction and incorrect actions, their sense of responsibility in determining actions (anthropocentric conservationism) and humankind's role in the ecological crisis, and their belief in technological fixes and human ingenuity for problem solving. That said, three of the ten statements used were based on the New Ecological Paradigm scale, first devised by Dunlap et al. (2000). We also used an alternative means of characterization of environmental values using Steg and Sievers' (2000) cultural theory analysis. Per their methodology (Steg and Sievers, 2000), respondents were 2 As this was an exploratory study, all the variables extracted from the literature review were entered stepwise into the model, despite potential interactions of independent variables (for example, knowledge affecting risk perception and environmental values). Indirect relationships are explored in Section 4.5.

3.2.2. Knowledge We assessed the respondents' level of knowledge by asking them to rate3 – on a continuous scale from 1 (not at all informed) to 5 (very well informed) – how informed they thought they were about various topics related to climate change (causes, consequences, and ways to lower carbon emissions) and reforestation technologies (strategies to adapt forests to changing local climates, current reforestation strategies in their province, assisted migration of species to new locations, use of genetic tools in tree breeding, and differences between genetically modified organisms (GMOs) and selectively bred crops). Two aggregated knowledge variables were created from these by adding up scores of the climate change questions (scale from 5 to 15) and the reforestation technology questions (scale from 5 to 25). 3.2.3. Perceived risk Respondents were asked to rate on a continuous scale of 1 to 5, with 1 being no threat and 5 being a great threat, how much of a threat they thought the following posed in terms of its impact on the well-being of forest-dependent communities: climate change; pest and disease outbreak; gathering tree seeds in one part of the province and planting 3 Self-reported knowledge was used for simplicity, as getting an accurate assessment of these topics using exam-style questions would have increased the length and complexity of an already long survey.

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004

R. Hajjar, R.A. Kozak / Forest Policy and Economics xxx (2015) xxx–xxx

5

Table 3 Likert-scale statements used to cluster respondents' into three groups and cluster means (final cluster centers after 10 iterations) for each statement, where 1 = strongly disagree and 5 = strongly agree. * indicates questions adapted from the New Ecological Paradigm (Dunlap et al., 2000). Cluster

The world's climate is getting warmer. We know enough science to have a good idea of what will happen under a changing climate. There is so much uncertainty around climate change that there is no point in trying to slow its impacts until we know more. The risks of doing nothing to adapt to climate change outweigh the risks of possibly acting incorrectly. By adapting our ecosystems in order to reduce the impact of climate change, we risk damaging the environment even more than we already have. Since humans are responsible for the loss of many species, it is our duty to use our best science to try to save what we have left. Human ingenuity will ensure that society will be able to adapt to climate change. Technological solutions are some of the best ways forward in dealing with climate change.* Humans are severely abusing the environment.* The so-called “ecological crisis” facing humankind has been greatly exaggerated.*

them in different parts of the province; weather extremes; logging/ timber harvesting; planting tree species in locations where they did not occur before; genetically modified trees; and naturally occurring fires. These data were analyzed in two ways. First, logistic regressions were run with each threat as an individual independent variable. We also aggregated the scores of all threats related to natural disasters (climate change, pest outbreaks, fires, weather extremes) and all threats related to reforestation technologies (moving seeds, moving species, and GM trees) to create two aggregate variables that were then run in the logistic regressions. 3.2.4. Past hardships linked to climate change Respondents were asked if certain negative occurrences had caused them, or someone close to them, hardship in the past ten years. In a separate question, they were also asked if they thought any of the listed negative occurrences had been caused by climate change. The occurrences included in the survey were insect outbreaks, wildland fires, job losses in the natural resources sector, floods, and droughts. As our goal was to see if people had directly or indirectly experienced a hardship which they also linked to climate change, we created a single, binary variable that combined the response of these two questions: if a respondent had experienced a specific hardship, directly or indirectly, which they also linked to climate change, this was scored as “yes”. In addition, we created an aggregated hardship variable, where each hardship linked to climate change was scored as one, and each respondent was given a summated score for this variable. Both the aggregated and individual hardship variables were run in the logistic regression. 3.2.5. Confidence in various actors involved Respondents were asked to rate how much confidence they had in different actors involved in the climate change adaptation debate to make the right choices for managing risks related to climate change. The actors in question were local leaders, scientists, the private sector/ industry, provincial and federal governments, and environmental groups. A 4-point continuous interval scale was used, with 1 being very little confidence, 2 being little confidence, 3 being some confidence, and 4 being lots of confidence. A “don't know” option was also included. Measures of confidence of each of the actors were used in the logistic regression. 3.2.6. Desired outcomes The questionnaire presented the respondents with five possible future outcomes for forests and forest-dependent communities in Western Canada and they were asked to rank the top three in terms of what they thought would be the most important for the province in the next 10 years. These future outcomes were: employment in forestrelated businesses is increased; local sources of income in forestdependent communities are diversified; forest plants and animals are

1

2

3

4.23 3.9 3.55 1.59 2.95 4.55 3.11 3.54 4.52 1.79

3.63 2.42 2.61 1.93 2.96 4.12 2.43 3.05 3.99 2.31

3.18 3.1 3.15 3.12 3.2 3.72 3.47 3.37 3.36 3.72

conserved; forest-based recreation opportunities are enhanced; and the local forest sector is maintained in its current state. The variable used for the logistic regression was their number one ranked future priority. 3.2.7. Additional exit questions After being asked to rate the six forest management strategies presented (see Section 3.1), respondents were then asked to consider four statements that could potentially explain their choices of answers. We asked them specifically to think of the forest management strategies that involved non-local or selectively bred seedlings and to rate their level of agreement on a 5-point Likert scale with the following statements: I think that manipulating nature with such interventions is ethically wrong; it is risky to manipulate nature in these ways because we are never completely sure of the outcomes; I am skeptical that such interventions will be useful for adapting our forests to climate change; and, I trust decision-makers to choose the interventions that would work in the best interest of Western Canadians. These four statements provided additional measures of values (ecocentrism), risk perception of biotechnologies, lack of trust in science, and trust in decision-makers, respectively, and were used as a form of triangulation and a means to strengthen our results. Put another way, respondents were asked to directly link the independent and dependent variables in this analysis, thus potentially helping to explain correlations by hinting at causal mechanisms. Responses to these exit questions were included in the logistic regression. Finally, the questionnaire also solicited basic demographic information, such as location, age, gender, education, and occupation/sector. These were all incorporated into the logistic regressions. 4. Results Prior to presenting the results of the logistic regression, we first present the survey results describing the acceptance of forest management strategies by respondents (the dependent variable in our logistic regression), followed by summaries of the independent variables used in the logistic regression. 4.1. Acceptance of forest management strategies Respondents were asked to rate each of the forest management reforestation strategies (seen in Table 2) based on their levels of acceptance for each strategy. These results formed the basis of our logistic regressions. Fig. 1 shows the general response patterns for all survey participants considered in aggregate. In general, respondents showed low acceptance of a “do nothing” strategy that will let climate change run its course without any human intervention, high acceptance of helping the forest to regenerate by re-planting with local seeds, a decreasing acceptance of breeding strategies that involve increasingly

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004

6

R. Hajjar, R.A. Kozak / Forest Policy and Economics xxx (2015) xxx–xxx

Fig. 1. Proportion of respondents that accepted each of the six forest management strategies (described in Table 2).

more manipulation (notably, there is a large dip in acceptance when it comes to breeding by moving seeds around the province), and low acceptance of GMOs. 4.2. Values and beliefs The values and beliefs of survey respondents pertaining to climate change and environmental issues were summarized in two ways, with the intent of using this reduced data set as predictive variables in the logistic regressions. First, Likert-scale data on the statements in Table 3 were used to derive a K-means cluster solution, with the aim of characterizing the views of segments within our sample. The final cluster solution yielded three groups of respondents, based on their general views towards climate change and the environment. Cluster means for each attitudinal statement are given in Table 3. The three clusters can be roughly characterized as follows: • Cluster 1 (728 respondents) were more inclined to believe that climate change and species loss are problems that should be acted upon, that we (humankind) will be able to figure out the solutions, but that we should proceed cautiously; • Cluster 2 (335 respondents) believed that there is a problem (although viewed less negatively than Cluster 1 respondents) that needs to be acted upon, but were more skeptical as to the extent of knowledge we have about it and to our ability to know what to do about the problem; and

• Cluster 3 (481 respondents) were somewhat indifferent, not convinced that there is a problem, and rather neutral on humankind's ability to deal with the problem if it does exist.

The acceptance of each of the six forest management reforestation strategies was plotted for each cluster (Fig. 2). While there were significant differences among the clusters for strategies 1, 4, 5, and 6, low Cramer's V values indicate that the association is indeed small, limiting the predictive power based on clusters alone. Cluster 3, the cluster that was characterized by neutrality on environmental problems and solutions, was more accepting than the other two clusters of strategies involving heightened human and technological interventions. Steg and Sievers' (2000) cultural theory analysis was also used to characterize the environmental values of respondents. Most respondents identified as egalitarians (69.6%), with 22.9% identifying as hierarchists, 1.8% as individualists, and 5.7% as fatalists. However, we found that these results were not incredibly meaningful. First, minor inconsistencies with responses to other questions meant that these four options oversimplified respondents' opinions. For example, a very small percentage identified as individualists, characterized by Steg and Sievers (2000) as those most likely to think that technology will fix environmental problems. However, this result is inconsistent with the relatively high mean score given to the statement on technological solutions (Table 3) across all three clusters. This inconsistency can, of course, be explained by the fact that, while most respondents thought that technological fixes were

Fig. 2. Differences among the three clusters' levels of acceptance for the six forest management strategies. * indicates a significant difference between clusters at p b 0.05 using a Pearson's chi-square test (Cramer's V: Strategy 1 = 0.152; Strategy 4 = 0.098; Strategy 5 = 0.069; Strategy 6 = 0.094).

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004

R. Hajjar, R.A. Kozak / Forest Policy and Economics xxx (2015) xxx–xxx

important in dealing with climate change, this statement was not the one that best characterized their views, generally speaking. We opine that scoring the degree of agreement with a value-based statement (as in the cluster analysis) was more informative than a categorical choice designation. Additionally, the skew of the majority of the respondents in one category (egalitarians) lowers its power of discrimination. As such, this “cultural theory” variable was excluded from the logistic regressions.4 4.3. Descriptive results A number of other descriptive variables used in the logistic regressions are summarized here. Fig. 3 shows how knowledgeable respondents are on a number of issues related to climate change and reforestation strategies. In general, respondents considered themselves more informed on climate change issues than reforestation technology issues. Respondents were also asked about perceived risks and past hardships as they pertain to a variety of issues that could occur in forestdependent communities. While respondents scored pest and disease outbreaks highly as a threat to the well-being of forest-dependent communities (Fig. 4), the fewest number of respondents had actually experienced, or knew someone who had experienced, this hardship (Fig. 5). That said, a high proportion of those who had experienced a pest and disease outbreak, directly or indirectly, linked it to climate change (81.0%).5 Even though this survey was administered prior to the massive flooding that took place in Alberta in June 2013, flooding also ranked highly among experienced hardships, again with a high proportion of respondents linking this with climate change (80.7%). Finally, respondents were asked some forward-looking questions, specifically about whom they entrust to make decisions and manage risk on matters related to climate change, and what their ideal future outcomes would be for forests and forest-dependent communities in Western Canada. Although scores were generally low for all actors (mean scores mostly below 3 out of 5), respondents seemed to have more confidence in scientists and environmental groups on issues related to climate change than in government or industry (Fig. 6). As Fig. 7 illustrates, conservation of plants and animals featured prominently as the outcome that respondents most often chose as their number one priority for Western Canada's forests and forest-dependent communities. Diversification of local income in forest-dependent communities was the second most prioritized option. 4.4. Logistic regression Forward binary logistic regressions were used to examine the association between the above listed variables and the likelihood to accept each of the six forest adaptation strategies. Several combinations of variables were used to run the regressions, and here we present the most robust model. Several variables emerged as having an influence in our model, with some having a stronger effect in increasing the odds of acceptance or rejection for different strategies (Table 4). Each of the models was significant, and the diagnostics indicate a decent fit of the data, as indicated by the pseudo R-squared values, with the exceptions of the models derived for Strategies 1 and 2. That said, the classification tables show relatively high numbers of correctly predicted responses for each of the models, especially those derived for Strategies 2 and 3, meaning that the models seem to have good predictive power. It is of

4 Initially, the “cultural value” variable was included in the logistic regressions, but was insignificant in every model run. 5 Parkins and MacKendrick (2007) also showed that many residents in forestdependent communities make a direct link between climate change and mountain pine beetle outbreaks, an insect outbreak that has caused great devastation to the forests of Western Canada.

7

interest that diagnostics seem to improve from Strategy 1 to Strategy 6. In other words, there seems to be more consistency in the responses of individuals considering the more interventionist forest management approaches. Four variables were more prominent in explaining outcomes across most of the six strategies. Agreement with the statement that manipulating nature is ethically wrong had a strong positive effect on the “do nothing” strategy, and consistently negative effects on the acceptance of the other strategies (i.e. respondents that believed that it is ethically wrong to manipulate nature were more likely to reject Strategies 3 through 6). Higher risk perception of natural disasters increased the odds of rejecting the do nothing strategy, and increased the odds of accepting Strategies 3 through 5, while the reverse was true for higher risk perception of reforestation technologies. Interestingly, the threat of climate change did not factor into the model. The more one agreed with the statement that it is risky to manipulate nature with such interventions, the more likely one was to reject Strategy 6. Age also seemed to be an important predicting factor for Strategies 2, 4, 5, and 6. Membership in Clusters 1 and 2 increased the odds of rejecting Strategy 1, and membership in Cluster 1 also increased the odds of rejecting Strategy 4. Respondents who had prioritized increased employment in the forest sector or increased recreation opportunities as future outcomes had very high odds of accepting Strategies 4 and 5. Trust in decision-makers to make the right choices related to adaptation and confidence in local leaders also increased the odds of accepting strategies that involved human manipulation of nature, while skepticism of the efficacy of such interventions increased the odds of accepting them. Gender only affected the odds of acceptance of Strategy 6; males were more likely to accept the use of genetically engineered seedlings in reforestation. Very few variables were important in strengthening the model for Strategy 2 (reforestation with local seeds and no selective breeding), likely due to the fact that the majority of respondents accepted this Strategy. 4.5. Exploring possible indirect relationships Knowledge, past hardships, and trust in science did not enter into the model, but previous literature indicates that these variables can affect perceptions of risk. That being the case, we also ran Pearson's correlations to see if there were any associations between these variables and risk perception. We found that respondents' perceptions of risk of reforestation technologies were positively correlated with their knowledge of such technologies (r = 0.157, p b 0.001) and negatively correlated with their confidence in scientists (r = − 0.052, p = 0.026) and trust in decision-makers (r = − 0.145, p b 0.001) to make the right choices for managing risks related to climate change; the more confidence they have in scientists and decision-makers, the less their perceived risk of such technologies. We also found that respondents' perceptions of natural disaster threat were positively correlated with confidence in scientists (r = 0.188, p b 0.001) and with hardships that they had experienced (personally or someone close to them) that they linked to climate change (r = 0.227, p b 0.001). Although respondents' perceptions of the threat of climate change as an individual variable (not the aggregate variable of natural disasters' threat) did not factor into the model, it was interesting to note that this variable was positively correlated with their knowledge of climate change (r = 0.165, p b 0.001), hardships they have experienced that they linked with climate change (r = 0.269, p b 0.001), and their confidence in scientists to make the right choices for managing risks related to climate change (r = 0.254, p b 0.001). 4.6. Summary of results Results of this study indicate that there are a multitude of factors – experiential, attitudinal, demographic, and perceptionally-based – that

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004

8

R. Hajjar, R.A. Kozak / Forest Policy and Economics xxx (2015) xxx–xxx

Fig. 3. Mean scores representing the degree to which respondents are informed about various topics related to climate change and reforestation strategies, with 1 = not at all informed and 5 = very well informed.

can affect people's decision-making processes in accepting or rejecting forest adaptation strategies to climate change. Fig. 8 illustrates a conceptual framework of the predictive variables, as tested in the logistic regression analysis. In addressing the hypotheses laid out at the outset of the study, the important factors in determining acceptance of several of the interventions presented to respondents in this survey seem to be attitudes towards climate change and nature, perceptions of risk of both natural disasters and reforestation technologies, skepticism about the efficacy of those technologies, and trust in decision-makers (Fig. 8). Age has a minor, but consistent, effect on acceptance levels of the various intervention strategies, with older respondents being more inclined to accept Strategies 2, 4, 5, and 6. In addition, prioritizing employment in the forest sector and recreation opportunities as future outcomes increased the odds of acceptance substantially for certain interventions. While statistical differences are observed between the levels of acceptance of the three attitudinally-clustered groupings of respondents, for most of the forest management strategies presented, these were not the most important factors explaining acceptance levels, generally speaking (only entering into the regressions for two strategies). Specific attitudes measured with the exit questions seem to be more telling of reasons underlying acceptance levels; belief that manipulating nature is ethically wrong and skepticism in the efficacy of such interventions both came out as important variables in the regression. It is

interesting to note that those who believe most strongly that climate change is a problem (Cluster 1) are also those who are least likely to accept an assisted migration strategy. This might be explained by the fact that this group was characterized as cautious and believed, more than the other two clusters, that manipulating nature was risky.

5. Discussion and implications for policy-makers Forests continue to play a vital role in Western Canada and around the world, providing countless products and ecosystem services, generating revenues and meaningful forest-based employment, and sustaining rural communities. Unfortunately, climate change threatens to disturb forest ecosystems and affect our reliance upon them. As forests become more vulnerable to changing climatic conditions, forest policy must look to different reforestation strategies and selective breeding technologies to ensure the long-term health and productivity of forests. As we continue to build on the science of how best to adapt our forests to climate change, we should also be seeking public ‘buy-in’ before implementing these strategies. Public participation in forest management and policy decision-making has increased in importance over the last few decades, and forest managers are having to adapt to an inclusive and socially responsive model of decision-making instead of an expert-driven and regulatory one (Beckley et al., 2006).

Fig. 4. Mean scores representing the degree to which respondents feel that various topics are a threat in terms of their impact on the well-being of forest-dependent communities, with 1 = no threat and 5 = great threat.

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004

R. Hajjar, R.A. Kozak / Forest Policy and Economics xxx (2015) xxx–xxx

9

Fig. 5. Number of respondents who have or know of someone close to them who has experienced one of the listed hardships and the portion of respondents who link these hardships to climate change.

As Fig. 8 shows, a wide variety of factors seemed to be important in predicting public acceptance of the forest adaptation strategies proposed here. Of these, attitudes towards nature, skepticism towards the efficacy of reforestation technologies, and trust in decision-makers most strongly impacted acceptance of different strategies, with risk perception of both natural disasters and reforestation technologies playing a weaker, but consistent, role in acceptance. The finding that gender had a large impact on the likelihood of accepting GM trees (specifically, being male greatly increased the odds of accepting Strategy 6) is consistent with previous work on gender differences in acceptance of GM technology in foods (reviewed in Moerbeek and Casimir (2005); Costa-Font et al. (2008)). Otherwise, the lack of gender differences in acceptance of the other strategies, is interesting in the context of forest adaptation strategies, as other studies have shown that women are more likely to have higher risk perception and are more likely to say they would voluntarily act to mitigate climate change, but are less likely to vote for government policies that address climate change (O'Connor et al. (1999), and references within). While not directly affecting acceptance levels, experiencing hardships and confidence in science seem to be related to risk perceptions of natural disasters, while confidence in science and in decisionmakers and knowledge of reforestation technologies are, respectively, negatively and positively correlated with risk perception of those technologies. Experienced hardships did not enter the regression, perhaps due to the small percentages of people who had actually experienced these hardships. Even fewer associated hardships with climate change. As Weber (2006) notes, personal experience with the consequences of climate change is still limited, so this may not adequately explain how people act in response to this risk. While abstract future consequences may not produce a visceral reaction in people yet – at least enough to accept more drastic adaptation measures – experienced hardships

were, in fact, correlated with perceptions of risk of climate change and of natural disasters. It is worth reiterating that while the perception of risk of climate change, on its own, did not enter the logistic regression as hypothesized, the threat of natural disasters that are often caused by climate change seems to be an important variable in determining levels of acceptance. This finding has implications for how the issue of climate change is framed to the public – people worry about natural disasters, particularly those that they have experienced personally or closely. As such, in arguing for the utility of various forest adaptation approaches, it is perhaps best to frame the problem as mitigating a potential series of recurring natural disasters, rather than addressing climate change in general. This study also reaffirms that trust in decision-makers and local leaders can play a vital role in both accepting adaptation strategies and reducing the risk perception of reforestation technologies. This corroborates findings in other contexts (for examples, see McFarlane et al. (2012) on forest management, and Durant and Legge (2005) on GMOs). This is of particular importance in this study, given the finding that respondents did not have much confidence in local, provincial or federal governments to make the right choices for managing risks related to climate change. Governments will need to demonstrate a willingness to work with the public on addressing their concerns in order to build more trusting relationships, prior to or while operationalizing new policies (Beierle, 1999). While knowledge of climate change or reforestation technologies did not enter the regression model as hypothesized, these variables were both positively correlated with risk perception. This would agree with the knowledge-deficit hypothesis – that increased levels of information lead to heightened risk perception (Kellstedt et al., 2008), a common finding in studies on public perceptions of biotechnology (Shaw, 2002). However, given the use of self-reported knowledge in

Fig. 6. Mean scores representing how much confidence participants had in relevant actors to make the right choices for managing risks related to climate change, with 1 = very little confidence and 5 = lots of confidence.

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004

10

R. Hajjar, R.A. Kozak / Forest Policy and Economics xxx (2015) xxx–xxx

Fig. 7. Proportion of respondents who ranked each future outcome as their number one priority for forests and forest-dependent communities in Western Canada.

this survey, we acknowledge the limitations in our ability to test this hypothesis; Durant and Legge (2005) found that self-reported knowledge, specifically of genetic science and biotechnology, tended to be higher than actual knowledge that was objectively measured among their respondents, and that those with high self-reported knowledge and those with high actual knowledge of biotechnology had different attitudes towards GM foods. Regardless, the prominence of factors other than knowledge in better explaining outcomes reinforces another important consideration for policy-makers. It is not just a matter of educating the public with science and facts so that they are able to make better informed decisions or convincing them that climate change is occurring; policy-makers must appeal to the public's beliefs, values, and priorities, and they must gain their trust. Studies have undermined the notion that disseminating environmental knowledge will alone enhance people's commitment to pro-environmental behavior (reviewed in Eden (1998)). Rather than focusing policy efforts on merely “educating” the public, a dialogue needs to be created, incorporating a discussion of the values and services that the public seeks in managing western Canada's forests. Given the complex interplay of these various factors affecting public acceptance, there are many implications for policy-makers who wish to make forest adaptation strategies to climate change more palatable to the public. As Maibach et al. (2009) state, the first would be to “know thy audience”; their knowledge base, risk perceptions, values and

beliefs, levels of trust, desired future outcomes, and past hardships are all essential considerations when initiating a dialogue (Fig. 8 can be used as a starting point for this). A genuine, inclusive, and deliberative process of public engagement can increase acceptance of new policies related to forest management strategies. Processes should ensure a highly interactive, symmetrical, and critical relationship between citizens and specialists, where emphasis is placed on exposing differences, dissent, uncertainties and underlying assumptions (Chilvers, 2008). Models for and analyses of citizen involvement in environmental decision-making abound in the literature, with analytic–deliberative processes being well described for participatory decision-making in the science–policy interface (Petts, 2006; Renn, 2006; Chilvers, 2008; Beckley et al., 2006; Burgess et al., 2007). Such deliberative processes can enhance procedural legitimacy through building trust, increase understanding through social learning, and promote ownership of the decision-making process.

Acknowledgments This research, as part of the AdapTree Project, was funded by GenomeCanada, GenomeBC and the Forest Genetics Council of British Columbia. The authors would also like to thank the rest of the AdapTree team for their input throughout the course of this study.

Table 4 The most robust of models that emerged from the forward logistic regression for each of the six forest management strategies seen in Table 2. The following combination of variables was entered into the regression: all exit questions; all confidence questions; aggregated hardship; aggregated hardship linked to climate change; aggregated knowledge of climate change; aggregated knowledge of reforestation technologies; aggregated natural disasters threat; threat of climate change; aggregated reforestation technologies threat; number one future priority; K-means cluster; age; gender; level of education; and location (major metropolitan area and non-MMA). Only the variables that enhanced the predictive power of the model (i.e. were significant) are listed in this table, with the exception of variables denoted with * which indicate non-significant categorical variables that were analyzed in groups.

Manipulating nature ethically wrong Aggregated threat of natural disasters Aggregated threat of reforestation technologies Cluster 3 (baseline) Cluster 1 Cluster 2 No. 1 priority (status quo) (baseline) No. 1 priority (employment) No. 1 priority (economic diversification) No. 1 priority (biodiversity) No. 1 priority (recreation) Age Trust decision-makers to make right choice Skeptical intervention will work Confidence in local leaders Risky to manipulate nature Gender (male) Nagelkerke R-squared Classification table: percentage correct

Strategy 1

Strategy 2

Strategy 3

Strategy 4

Strategy 5

Strategy 6

Do nothing

Local seeds

Local + breeding

Non-local + breeding

Non-local different sp

GMOs

1.354 0.94 1.05

1.129

0.516 1.187 0.854

0.584 0.661

0.78 1.071 0.873

0.544

1.027 1.51 0.731

1.013 1.531 0.793

0.81

0.732 1.602 0.363 72.0

0.905

0.592 0.855* 0.538* 2.612* 2.829* 3.187* 1.043 1.608

0.078 67.5

0.644 1.116 0.835

0.102 97.7

0.233 91.9

2.775 1.777* 1.716 3.201 1.016 1.458 0.719 1.405

0.35 73.4

0.295 71.0

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004

R. Hajjar, R.A. Kozak / Forest Policy and Economics xxx (2015) xxx–xxx

11

Fig. 8. A conceptual framework of factors affecting willingness to accept various levels of intervention strategies for forest adaptation. Dotted boxes and dotted lines indicate variables that did not factor into logistic regression, but were correlated with variables that were. Shaded boxes indicate that the independent variable factored into the regression for three or more strategies. Thickness of box outline for independent variables is an indication of the strength of the odds in determining outcome and influencing the dependent variable. Note: Belief that manipulating nature is ethically wrong is grouped together with the clusters variable in this diagram as a generalized “attitudes towards climate change/nature” variable.

References Adger, W.N., Dessai, S., Goulden, M., Hulme, M., Lorenzoni, I., Nelson, D.R., Naess, L.O., Wolf, J., Wreford, A., 2008. Are there social limits to adaptation to climate change? Clim. Chang. 93 (3–4), 335–354. http://dx.doi.org/10.1007/s10584-008-9520-z. Alberta Environment and Parks, 2015. Tree Improvement and Adaptation. Retrieved from http://aep.alberta.ca/lands-forests/forest-health/tree-improvement-and-adaptation. aspx, June 23, 2015. Aubin, I., Garbe, C.M., Colombo, S., Drever, C.R., Mckenney, D.W., Messier, C., Pedlar, J., Saner, M.A., 2011. Why We Disagree about Assisted Migration 1: Ethical Implications of a Key Debate Regarding the Future of Canada's Forests. pp. 755–765. Beckley, T.M., Parkins, J.R., Sheppard, S.R.J., 2006. Public Participation in Sustainable Forest Management: A Reference Guide. Management, Edmonton, Canada. Beierle, T.C., 1999. Using social goals to evaluate public participation in environmental decisions. Rev. Policy Res. 16 (3/4), 76–103. Bonny, S., 2003. Why Are Most Europeans Opposed to GMOs? Factors Explaining Rejection in France and Europe. 6 (1). British Columbia Ministry of Forests, Lands and Natural Resource Operations, 2012. Forest Stewardship Action Plan for Climate Change Adaptation, 2012-2017. Retrieved from https://www.for.gov.bc.ca/ftp/HFP/external/!publish/ClimateChange/Adaptation/ MFLNR_CCAdaptation_Action_Plan_2012_final.pdf, June 23, 2015. British Columbia Ministry of Forests, Lands and Natural Resource Operations, 2013a. Climate Change Strategy (2013-2018). Retrieved from https://www.for.gov.bc.ca/het/ climate/strategy/ClimateChangeStrategy_2013-2018.pdf, June 23, 2015. British Columbia Ministry of Forests, Lands and Natural Resource Operations, 2013b. Climate Action Roadmap Retrieved from https://www.for.gov.bc.ca/het/climate/strategy/FLNRClimateActionRoadmap.pdf, June 23, 2015. Burgess, J., Stirling, A., Clack, J., Davies, G., Eames, M., Staley, K., Williamson, S., 2007. Deliberative mapping: a novel analytic–deliberative methodology to support contested science-policy decisions. Public Underst. Sci. 16 (3), 299–322. http://dx. doi.org/10.1177/0963662507077510. Chilvers, J., 2008. Deliberating competence: theoretical and practitioner perspectives on effective participatory appraisal practice. Sci. Technol. Hum. Values 33. http://dx. doi.org/10.1177/01622439073075941. Costa-Font, M., Gil, J., Traill, W., 2008. Consumer acceptance, valuation of and attitudes towards genetically modified food: review and implications for food policy. Food Policy 33 (2). Elsevier, pp. 99–111. Donlan, J., et al., 2005. Re-wilding North America. Nature 436 (7053), 913–914. http://dx. doi.org/10.1038/436913a. Dunlap, R.E., Liere, K.D., Van Mertig, A.G., Jones, R.E., 2000. Measuring Endorsement of the New Ecological Paradigm : A Revised NEP Scale. J. Soc. Issues 56, 425–442. Durant, R.F., Legge, J.S., 2005. Public opinion, risk perceptions, and genetically modified food regulatory policy: reassessing the calculus of dissent among European Citizens. Eur. Union Polit. 6 (2), 181–200. http://dx.doi.org/10.1177/1465116505051982. Eden, S., 1998. Environmental issues: knowledge, uncertainty and the environment. Prog. Hum. Geogr. 22 (3), 425–432. http://dx.doi.org/10.1191/030913298676818153. Fettig, C.J., Reid, M.L., Bentz, B.J., Sevanto, S., Spittlehouse, D.L., Wang, T., 2013. Changing climates, changing forests: a Western North American perspective. J. For. 111 (May), 214–228. Fischhoff, B., Fischhoff, I., 2001. Public opinions about biotechnologies. AgBioforum 4 (3&4), 155–162. Hajjar, R., McGuigan, E., Moshofsky, M., Kozak, R., 2014. Opinions on strategies for forest adaptation to future climate conditions in Western Canada: surveys of the general public and leaders of forest-dependent communities. Can. J. For. Res. 44, 1525–1533. http://dx.doi.org/10.1139/cjfr-2014-0142. Harshaw, H.W., Tindall, D.B., 2005. Social structure, identities, and values: a network approach to understanding people's relationships to forests. Journal of Leisure Research 37 (4). National Recreation and Park Association, p. 426. Hewitt, N., Klenk, N., Smith, A.L., Bazely, D.R., Yan, N., Wood, S., MacLellan, J.I., LipsigMumme, C., Henriques, I., 2011. Taking stock of the assisted migration debate.

Biological Conservation 144 (11). Elsevier Ltd, pp. 2560–2572. http://dx.doi.org/10. 1016/j.biocon.2011.04.031. Kellstedt, P.M., Zahran, S., Vedlitz, A., 2008. Personal efficacy, the information environment, and attitudes toward global warming and climate change in the United States. Risk Anal. 28 (1), 113–126. http://dx.doi.org/10.1111/j.1539-6924.2008.01010.x. Maibach, E., Roser-Renouf, C., Leiserowitz, A., 2009. Global Warming's Six Americas 2009: An Audience Segmentation Analysis. Yale University, New Haven (http://www. climateaccess.org/sites/default/files/Leiserowitz_Six Americas June 2010_0.pdf). Marx, S.M., Weber, E.U., Orlove, B.S., Leiserowitz, A., Krantz, D.H., Roncoli, C., Phillips, J., 2007. Communication and mental processes: experiential and analytic processing of uncertain climate information. Glob. Environ. Chang. 17 (1), 47–58. http://dx.doi.org/10.1016/j. gloenvcha.2006.10.004. McFarlane, B.L., Parkins, J.R., Watson, D.O.T., 2012. Risk, knowledge, and trust in managing forest insect disturbance. Can. J. For. Res. 42, 710–719. http://dx.doi.org/10.1139/ X2012-030. Minteer, B.A., Collins, J.P., 2010. Move it or lose it? The ecological ethics of relocating species under climate change. Ecological Applications 20 (7). Eco Soc America, pp. 1801–1804. Moerbeek, H., Casimir, G., 2005. Gender differences in consumers' acceptance of genetically modified foods. Int. J. Consum. Stud. 29 (4), 308–318. http://dx.doi.org/10. 1111/j.1470-6431.2005.00441.x. O'Connor, R.E., Bord, R.J., Fisher, A., 1999. Risk perceptions, general environmental beliefs, and willingness to address climate change. Risk Anal. 19 (3), 461–471. http://dx.doi. org/10.1023/A:1007004813446. Parkins, J., Mackendrick, Norah A., 2007. Assessing community vulnerability: a study of the mountain pine beetle outbreak in British Columbia, Canada. Glob. Environ. Chang. 17, 460–471. http://dx.doi.org/10.1016/j.gloenvcha.2007.01.003. Petts, J., 2006. Managing public engagement to optimize learning: reflections from urban river restoration. Hum. Ecol. Rev. 13 (2), 172–181. Renn, O., 2006. Participatory processes for designing environmental policies. Land Use Policy 23 (1), 34–43. http://dx.doi.org/10.1016/j.landusepol.2004.08.005. Savage, I., 1993. Demographic influences on risk perceptions. Risk Anal. 13 (4), 413–420 (http://www.ncbi.nlm.nih.gov/pubmed/8234949). Shaw, A., 2002. ‘It just goes against the grain’. Public understandings of genetically modified (GM) food in the UK. Public Underst. Sci. 11 (3), 273–291. http://dx.doi.org/10. 1088/0963-6625/11/3/305. Slimak, M.W., Dietz, T., 2006. Personal values, beliefs, and ecological risk perception. Risk Anal. 26 (6), 1689–1705. http://dx.doi.org/10.1111/j.1539-6924.2006.00832.x. Slovic, P., 1999. Trust, emotion, sex, politics, and science: surveying the risk-assessment battlefield. Risk Anal. 19 (4), 689–701 (http://www.ncbi.nlm.nih.gov/pubmed/10765431). Steg, L., Sievers, I., 2000. Cultural theory and individual perceptions of environmental risks. Environ. Behav. 32 (2), 250–269. http://dx.doi.org/10.1177/00139160021972513. Ste-Marie, C., Nelson, E.A., Dabros, A., Bonneau, M.E., 2011. Assisted migration: introduction to a multi-faceted concept. For. Chron. 724–730. Wachinger, G., Renn, O., Begg, C., Kuhlicke, C., 2013. The risk perception paradox–implications for governance and communication of natural hazards. Risk Anal. 33 (6), 1049–1065. http://dx.doi.org/10.1111/j.1539-6924.2012.01942.x. Wang, T., Campbell, E.M., O'Neill, G.A., Aitken, S.N., 2012. Projecting future distributions of ecosystem climate niches: uncertainties and management applications. Forest Ecology and Management 279 (September). Elsevier B.V., pp. 128–140. http://dx.doi.org/ 10.1016/j.foreco.2012.05.034. Weber, E.U., 2006. Experience-based and description-based perceptions of long-term risk: why global warming does not scare us (yet). Clim. Chang. 77 (1–2), 103–120. http://dx.doi.org/10.1007/s10584-006-9060-3. Whitfield, S.C., Rosa, E.A., Dan, A., Dietz, T., 2009. The future of nuclear power: value orientations and risk perception. Risk Anal. 29 (3), 425–437. http://dx.doi.org/10. 1111/j.1539-6924.2008.01155.x.

Please cite this article as: Hajjar, R., Kozak, R.A., Exploring public perceptions of forest adaptation strategies in Western Canada: Implications for policy-makers, Forest Policy and Economics (2015), http://dx.doi.org/10.1016/j.forpol.2015.08.004