Criteria and Indicators of Sustainable Forest Management in a Changing Climate : An Evaluation of Canada ’ s National Framework

Sustainability is now woven throughout forest management and policy. Criteria and indicators (C&I) provide a means of defining the concepts of sustainability in the context of forest management and establishing goals to gauge progress. There have been no major research initiatives to determine the implications of climate change for C&I. We evaluated the 46 indicators of the 2003 Canadian Council of Forest Ministers framework. Indicators were evaluated for their relationships with climate, relationships with other indicators, robustness and utility under climate change, and future prospects, including abandonment, improvement, or continued use. An evaluation framework was developed to analyze indicator linkages, direct and indirect climate-change influence, and potential modifications. 12 indicators were considered unaffected by climate change. The remaining 34 indicators were considered to be influenced by climate change. No modification seemed warranted for 23 of these indicators, while modifications for the remaining 11 indicators were recommended. Six new indicators were identified for monitoring forests sustainably under climate change. The difference between action and state indicators had implications for the influence of climate change on indicator effectiveness. State indicators were more prone to declines in their tracking ability, while action indicators were often unaffected, or even improved under climate change, as measured by several key traits of indicator effectiveness. The most prevalent theme in the evaluations was a decline in indicator predictability. We suggest moving from predominantly retrospective analysis to a balance of retrospective and prospective analysis, given that monitoring is inherently backward-looking and the threats and uncertainties of climate change are impending.


Introduction
The theme of sustainability is now globally woven throughout forest management and policy.Commitments to sustainability and the sustainable development of forests became internationally accepted after the United Nations Conference on Environment and Development (UNCED) in 1992, also known as the Rio Summit, and the statement of Forest Principles and Chapter 11 of Agenda 21 of the conference's action plan (UNCED, 1992).
A key initiative to emerge from the Forest Principles commitments to sustainable development in forestry was the development of criteria and indicators (C&I; UNCED, 1992).C&I provide a means of defining the broad and ambitious concepts of sustainability in the context of forest management and establishing measurable goals to gauge progress towards sustainable forest management (SFM; Wijewardana, 2008).Several international agreements and initiatives for C&I-SFM exist, which combined have involved almost 150 countries.Notable examples include the Montréal Process, the Helsinki Process (now Forest Europe), and the International Tropical Timber Organization Process (Duinker, 2011;Hall, 2001).The use of C&I-SFM to define and measure SFM progress has also been refined and implemented at the local, forest-management-unit level (Duinker, 2001).Local-level applications of C&I include industrial forest management planning, model-forest initiatives, and forest certification standards, such as those provided by the Canadian Standards Association (CSA), the Sustainable Forestry Initiative (SFI), and the Forest Stewardship Council (FSC; Duinker, 2011).There have been two significant national/international-level C&I-SFM initiatives in Canada: the Canadian Council of Forest Minister's (CCFM) national framework of C&I-SFM (CCFM, 1995;2003) and the Montréal Process and Santiago Declaration (Montreal Process, 1995).
Prior to the formation of the Montréal Process (Montreal Process, 1995), the CCFM initiated a C&I Task Force to develop a national framework of C&I-SFM and meet Canadian commitments to SFM made at the Rio Summit and in the 1992 national forest strategy (CCFM, 1992).In 1995, the CCFM released a national framework of C&I-SFM with six criteria and 83 indicators (CCFM, 1995).The C&I-SFM were first reported on in the 1997 technical report (CCFM, 1997) and then the 2000 national assessment report (CCFM, 2000), and a public review process was initiated soon after in 2001.The revised C&I-SFM, consisting of six criteria and 46 indicators, were released in 2003and reported on in 2005(CCFM, 2003;;2006).C&I-SFM are widely recognized as a valuable science-based tool both to define the full spectrum of values associated with forests and their management and to measure and gauge the degree to which forests are being managed sustainably (CCFM, 2003).There have been many recent studies on the potential improvement of C&I and sustainability indicators (Gough et al., 2008;Rametsteiner et al., 2011;Wijewardana, 2008;Wolfslehner & Vacik, 2011).However, there have been no major research initiatives to determine the implications of climate change for existing national/international frameworks of C&I, and certainly no research on how climate change might influence the utility of C&I-SFM.Canada's national framework of C&I-SFM was developed to have the quality and utility to track SFM progress with confidence.Because significant climate change in Canada is expected during the next century (Intergovernmental Panel on Climate Change [IPCC], 2007), and because climate is a strong driver and determinant of forests and the forest sector (Williamson et al., 2009), it is paramount that C&I-SFM have sufficient robustness to be both meaningful and valid in a changing climate.Consequently, a comprehensive and systematic assessment of the national C&I framework was undertaken to explore its effectiveness and validity under a changing climate (Steenberg et al., 2011b).This paper summarizes an evaluation of each of the CCFM indicators, as defined within the revised 2003 C&I framework (CCFM, 2003), for their suitability in defining and monitoring SFM in a changing climate.

Key Term Definitions
We define C&I herein as they are used in the Canadian forest sector, and specifically by the CCFM.A criterion is a collection or homogeneous category of values, by which SFM is assessed (CCFM, 1995;Montréal Process, 1995).An indicator is some identified system component that can be objectively and empirically measured to assess the status of a criterion or progress towards a goal associated with the SFM values of a criterion (CCFM, 1995;Duinker, 2001;Prahbu et al., 1999).The CCFM national framework also includes elements, which are simple categories used to organize related SFM values within several of the criteria.For example the biological diversity criterion includes ecosystem, species, and genetic diversity elements.For this analysis it is important to differentiate between indicators and variables.We define variables as alternative ways to measure an indicator.An indicator may have one or many variables, as was evident in the 2006 C&I-SFM assessment.For example, Indicator 1.2.2 (population levels of selected forest-associated species) is tracked using 41 variables (i.e., population levels of 41 species), while the related Indicator 1.2.3 (distribution of selected forest associated species) is represented by only four variables.
There are several types of SFM indicators, in terms of what they measure and track.It is, however, important to delineate the difference between two key groups, which are the action indicators and state indicators.Action indicators track both the quality and quantity of management actions, such as the rate of compliance with soil disturbance standards.State indicators measure the condition or state of a system that is being monitored.One type of state indicator is the condition/response indicator, which tracks the response of phenomena to management actions, such as the population levels of forest-associated species.Another state indicator is the context indicator, tracking variables that cannot be directly influenced by management, such as the contribution of timber products to the gross domestic product (Duinker, 2001).
A key mechanism for analyzing the effects of climate change on the indicators was a set of indicator traits, which have been established to define the characteristics of a valid, relevant, and effective indicator (Duinker, 2001).These traits include measurability, feasibility, responsiveness, relevance, understandability, validity, and predictability.Measurability and feasibility refer to the degree to which a phenomenon can be objectively and empirically measured on a continual basis and how practical, expensive, and obtainable data are to measure an indicator, respectively.More often than not, biophysical indicators are more easily measured than socioeconomic and socio-political indicators.Responsiveness, also called sensitivity, is the degree to which a phenomenon responds to management actions in known ways, and is a vital factor for indicator effectiveness.Relevance is an indicator's relationship to a defined SFM value within a criterion and the insight it provides into the sustainability of that value.Understandability is an obvious trait, as the SFM signal generated by an indicator must be clear and approachable to decision-makers and forest stakeholders.Validity refers to the overall soundness of the science behind an indicator.Finally, predictability is the degree to which the future value or range of an indicator can be confidently forecast.The implications of climate change for these indicator traits are discussed in Section 4.1.Our approach was to review and evaluate each indicator in the national set for a) its relationships with climate, b) its systemic relationships with other indicators in the set, c) its robustness and utility in the face of climate change, and d) future prospects for the indicator including possible abandonment, improvement, or continued use unchanged.This analysis is a form of applied policy research, which brings with it specific challenges and methodological needs.A particular challenge is the lack of access or repeatability of the analysis (Ritchie & Spencer, 1994), which is certainly true of this study.Because there is a strong reliance on the conceptual abilities of the analysts in applied policy research and in this study, there is a need for explicitness in description of the research methods (Ritchie & Spencer, 1994).Consequently, we developed a three-stage, qualitative evaluation framework for the analysis of the indicators (Figure 1; Appendix A), which was applied to each of the 46 indicators in the 2003 C&I-SFM framework (Table 1;CCFM, 2003).In the evaluation, we attempted to decipher the often complex, uncertain, or ambiguous effects of climate change on indicator functioning and ability to gauge SFM progress.The development of C&I-SFM has often been seriously impeded in past initiatives because they failed to address relationships or linkages between indicators and among disciplines (Wolfslehner & Vacik, 2011;Yamasaki et al., 2002).Indicators developed in the confines of one discipline, whether in the biophysical, social, or economic realm, will be insufficient to address adequately the integrated nature of SFM that they are meant to define and measure.It stands to reason that an evaluation of the potential influence of climate change on C&I-SFM should avoid these same pitfalls.We therefore began the indicator evaluations with a linkages assessment to explore the linkages among all indicators prior to the assessment of climate change.

Evaluation Framework
The current CCFM C&I-SFM do not thoroughly address the complex linkages that exist among indicators within the framework.Thus, our goal in the linkages assessment was to ascertain the extensive network of relationships among the indicators of SFM so that these linkages could be incorporated into our climate-change assessments.This was done through critical review and comparison of indicator pairs to gauge the nature of the relationship between them, and whether one was influential over the other.The basis of comparison was both the fundamental indicator design and intent (CCFM, 2003) and the specific variables or processes for which data were collected in the 2005 national assessment (CCFM, 2005), as they were found occasionally to differ.Climate change was not addressed in this assessment, but rather this was a preliminary stage of evaluation prior to the independent and integrated climate-change assessments.
An assessment of the effects of climate change on each individual indicator, independent of its linkages with other indicators, was done to ensure a detailed and uncluttered examination of all possible avenues of effects, prior to the integrated assessment.While the relationships with other indicators may open further possibilities for climate change to influence an indicator, we assumed that these relationships would never reduce or reverse the potential influence of climate change.Thus, the independent climate-change assessment was the logical second stage of evaluation.The assessment consisted mainly of a review of the literature in relevant fields of study.
The final stage of evaluation was the integrated climate-change assessment.Here, the efforts of the two previous assessments were combined to explore both the direct and indirect effects of climate change on the indicators in order to assign indicators to one of the four final post-evaluation categories or create a new indicator.A key focus of the integrated climate-change assessment was the indirect effects of climate change on the indicators through the complex network of interactions among them.

Results
In total, there were found to be 12 uninfluenced indicators, 23 unmodified indicators, 11 modified indicators, and no abandoned indicators.We also recommended creation of six new indicators (Table 2).All six of the new indicators were created within the biophysical criteria.The indicator modifications were more evenly distributed, with six of the 20 biophysical indicators modified within Criteria 1 through 4, and five of the 26 social/economic indicators modified within Criteria 5 and 6.It was initially expected that the effects of climate change would have resulted in the abandonment of some of the indicators.However, our project was restricted to the analysis of the influence of climate change alone on indicator effectiveness, and did not address any existing deficiencies in indicator strength.There could conceivably be instances where the combined influence of climate change and existing indicator effectiveness would warrant abandonment.
Given the text length of the unabbreviated results and usual confines of journal publication, it was not possible here to present the full findings of the three-stage evaluation for each of the 46 indicators.Instead, we present a brief summary of the implications of climate change for each of the indicators, organized by criterion.A more-detailed presentation of the findings can be found in Appendix A or in the two original reports (Steenberg et al., 2011a;2011b).

Criterion 1 -Biological Diversity
The biophysical indicators within Criterion 1, biological diversity, were highly susceptible to the changing climate, as biological diversity at all scales is intrinsically linked to the surrounding climate.Indicators 1.1.1 and 1.1.2represent the ecosystem diversity element, tracking forest type, age, and area, and wetlands in and out of protected areas, and were both assigned to the unmodified category.Both of these indicators were broad and encompassing in what was measured in the 2005 assessment, with a total of 20 measured variables (CCFM, 2005).The anticipated effects of climate change on species distribution, natural disturbance regimes, and hydrological processes, as well as the sheer breadth of what was measured by the indicators, suggests that Indicator 1.1.1,and to a lesser degree Indicator 1.1.2,were among the indicators most afflicted by climate change, with a decline in the predictability, responsiveness, and relevance of the Indicators.However, these indicators are at the foundation of SFM monitoring, so developing alternative approaches to measurement to mitigate the deterioration of these indicator traits would be unwise.
Indicator 1.1.2is highly related to 1.1.1 and vulnerable to all the aforementioned climate-change impacts.However, because its function in signalling progress towards SFM is to yield insight into the ability to protect representative forest areas (CCFM, 2006), we found it to be further compromised by climate change due to the core assumption of biogeographic stability in the protected areas system.Because of this loss of biogeographic stability and the shifting ranges and relative abundance of species, we recommended the creation of an additional indicator to track the connectivity of protected areas.Given the degree to which Indicators 1.1.1 and 1.1.2are interrelated with other indicators in the framework and the fact that they track physical forest-ecosystem conditions within Canada and its terrestrial ecozones, they represent arguably one of the most fundamental sets of information pertaining to biodiversity.Therefore, we saw no possible justification for abandoning these indicators in light of climate change.
The species diversity indicators were also found to be affected quite extensively by climate change, resulting in the modification of Indicators 1.2.2 and 1.2.4.Indicator 1.2.1 was somewhat anomalous, as the status of forest-associated species that it tracks refers to the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) risk status (CCFM, 2006).Therefore, the indicator's capacity under climate change is dependent on the COSEWIC status assessment and designation process.Fortunately, climate change is being more frequently considered in this process and Indicator 1.2.1 was subsequently assigned to the unmodified category.
Indicators 1.2.2 and 1.2.3 are similar in concept but quite different in terms of what is measured and tracked.Indicator 1.2.2 currently tracks population levels of ten mammal species and 31 bird species, while Indicator 1.2.3 tracks four species with detailed case studies (CCFM, 2006).Consequently, we feel that the decline in predictability, responsiveness, and relevance of these indicators was unacceptable and warranted modification to Indicator 1.2.2, but not to Indicator 1.2.3.Since climate change may cause both increases and decreases in the population levels of some forest-associated species, both directly and indirectly and often in an unpredictable manner, their ability to signal SFM is compromised.Species monitored by Indicator 1.2.2 in the future will need to be either relatively uninfluenced by climate change, or have a climate-change response that is adequately researched and understood.This will certainly result in the abandonment of some of the 41 species currently monitored.The more-focused and in-depth approach to the measurement of Indicator 1.2.3 suggested a smaller decline in the strength of its SFM signal, and it was left unmodified.
Invasive, alien, forest-associated species will likely become more abundant and detrimental to forest ecosystems and the sustainable flow of their resources under climate change.Therefore, Indicator 1.2.4 not only suffered a considerable decline in its predictability, but also an increase in its relevance.The impetus behind the switch to measuring area disturbed rather than the number of species was that it could be more easily aligned with a directional goal statement; the minimization of forest area disturbed by invasive species.Moreover, the mountain pine beetle is an example of how climate change may destabilize the population dynamics of native forest-associated biological disturbance agents.As such, we also modified the indicator to track both native and alien invasive species.
Indicator 1.3.1 offers a relatively simple measure of genetic diversity by assessing the number of parents used for collecting seed of regeneration seed-lots, and as an action indicator, it was largely uninfluenced by climate change and was assigned to the uninfluenced category.However, the rate and magnitude of climate change in Canada may lead to a decline in the genetic diversity of tree species.One potential adaptation to this is adjusting jurisdictional seed-transfer guidelines to promote assisted migration of more southerly genetic resources.This issue has serious implications for the relevance of Indicator 1.3.1,yet the current structure of the indicator does not reflect this.We therefore recommended a new indicator for the Genetic Diversity Element to track seed-transfer updates on public land.
Indicator 1.3.2,another action indicator, has two key components, in situ and ex situ conservation efforts for native tree species, which were afflicted by climate change in differing ways.The ex situ component of the Indicator was relatively uninfluenced by climate change.The in situ component shares the same assumption of biogeographic stability as Indicator 1.1.2,and thus was reduced in its ability to monitor SFM progress.We saw no possible modifications that would mitigate the decline in this indicator's effectiveness.However, it is reasonable to assume that many of the in situ and especially the ex situ conservation efforts will adapt to incorporate the additional threats of climate change.In this case, the current design of Indicator 1.3.2 is sufficient to monitor these potential adaptations.

Criterion 2 -Ecosystem Condition and Productivity
Indicators within Criterion 2, ecosystem condition and productivity, were the most directly affected by climate change, resulting in recommendations for two modifications and two new indicators.These five biophysical state indicators are also all relatively well researched in the climate-change literature and far more tractable than most of the socioeconomic indicators within Criteria 5 and 6.Therefore, the evaluation of, and recommendations for, these indicators were more in-depth and developed than many of the more conceptual and ambiguous socioeconomic and socio-political indicators.
Indicator 2.1 tracks merchantable volume and was found to increase in relevance from observed and anticipated effects of climate change, but was also expected to decline in predictability and measurability.The SFM signal generated by Indicator 2.1 was slightly diminished, not because of how we interpret growing stock and its relation to SFM in a changing climate (i.e. its relevance), but because of how it is often measured and the predictability of these measurements.The measurement has a reliance on pre-established relationships and processes in forest ecosystems that are based on historically stable climates.Thus, inventory and modelling methods will need to be modified to account for a variable climate.The creation of a new indicator to track temperatures was also recommended because of its importance for phenology, tree growth performance, and forest productivity.
Indicator 2.2, though having numerous interactions with climate change, was not found to be significantly diminished in its effectiveness.Climate change may directly influence forest area through species migration and dieback and indirectly through new afforestation and carbon sequestration policies, which increased the relevance of Indicator 2.2 in a changing climate.Moreover, the proliferation of remote-sensing technologies may improve the feasibility of monitoring the indicator in the future.The decline in predictability was found for nearly every biophysical state indicator, and in this case we felt it did not overly diminish the Indicator's SFM signal and we assigned it to the unmodified category.
Indicator 2.3 greatly increased in its relevance to SFM because of the more frequent and severe natural disturbance events, as is already being observed across Canada.The decline in predictability of Indicator 2.3 was concerning, given the major influence of Indicator 2.3 on other indicators in the C&I framework, though we saw no possible modifications that could mitigate this decline in predictability.However, given the increased relevance of natural disturbance in a changing climate, we did recommend modifying the Indicator to incorporate other major forms of natural disturbance.
Indicator 2.4 tracks forest area disturbed by ozone and acid rain, and was far more linked with anthropogenic emissions from fossil-fuel combustion than the resulting climatic change.However, there was some decline in the predictability of the indicator due to the influence of changing precipitation rates and temperatures with acid rain and tropospheric ozone, respectively.Despite this drop in predictability, we felt that Indicator 2.4 would remain largely unchanged in its ability to signal SFM progress and we assigned it to the unmodified category.
Indicator 2.5 was also left unmodified, but resulted in the recommendation of a new indicator.We surmised that the Indicator experienced major declines in predictability and responsiveness because of the anticipated impacts of climate change.The relevance of Indicator 2.5 to SFM also increased given the vulnerability of forest regeneration to climate change.We felt that the best approach to address the introduced uncertainty and relevance from climate change was to create a new indicator as opposed to modifying the original Indicator 2.5, which was assigned to the unmodified category.The newly created indicator tracks assisted migration initiatives on public land, since this is an adaptation to climate-change impacts on natural regeneration that is growing in relevance.However, assisted migration is still a new and developing science and practice, with many uncertainties (Ste-Marie et al., 2011).Consequently, further research into its effectiveness and appropriateness is needed.

Criterion 3 -Soil and Water
Criterion 3, soil and water, contains three indicators, the least of the six criteria.Indicators 3.1 and 3.2 are action indicators tracking compliance rates with various standards, and were found to be uninfluenced by climate change.Indicators 3.1 and 3.2, along with Indicator 1.3.1 (another action indicator), were the only biophysical indicators to be uninfluenced by climate change.Some observed and potential climate-change impacts are relevant to Criterion 3, such as changes in decomposition rates of dead organic matter, soil disturbance arising from less frozen-ground conditions, and changes in hydrological processes, but they were not captured by Indicators 3.1 and 3.2.
Indicator 3.3 has considerable overlap with Indicator 2.3, and naturally was influenced to some degree by the effects of climate change on natural disturbance regimes.The most notable effects of climate change on the indicator were a decline in predictability.However, the SFM values associated with Criterion 3 and Indicator 3.3 are the maintenance of soil and water resources, not ecosystem condition and productivity.Therefore, we chose to leave the indicator unmodified and created a new indicator in response to the likely impacts of climate change on hydrological processes.This new indicator may also satisfy some of the current misalignment of the indicator with its described goal of monitoring major changes in water yield, timing, and peak flow (CCFM, 2003).Indicator 4.1.3tracks carbon stored in forest products and was much more relevant in a changing climate, though we did not modify the Indicator.Instead, given the growing recognition of forest products carbon and the carbon emissions avoided through product substitution to mitigate climate change, we recommended the creation of an additional indicator to track emissions avoided through product substitution.However, it will be important to examine what we classify as forest products in the future measurement of this indicator, considering the growing relevance of bioproducts, biofuels, and other forest-derived products.

Criterion
Indicator 4.1.4also increased in relevance because of the contribution of forest-sector emissions to climate forcing by anthropogenic greenhouse gas emissions.Future uncertainty around the magnitude and rate of climate change and society's response, namely the employment of policy tools for emissions reductions, also confounded the predictability of Indicator 4.1.4.However, unlike more complex and inclusive indicators like Indicators 1.1.1 and 2.3, Indicator 4.1.4is relatively simple and transparent, so we felt no modifications were necessary to maintain its SFM signal.

Criterion 5 -Economic and Social Benefits
Indicators within Criterion 5, economic and social benefits, track the social and economic benefits derived from forests, and were far more difficult to evaluate in light of climate change.While on average the indicators within Criteria 5 and 6 were less influenced by climate change, it was far more difficult to expose and quantify the full suite of direct and indirect effects of climate change on social systems and the indicators that track them, in comparison to those that track ecological systems (Beckley, 2000).The economic indicators of Element 5.1 were influenced by climate change to varying degrees.Indicator 5.1.1 and 5.1.4are simple economic measures similar in design that track the contribution of timber products, non-timber forest products, and forest-based services to the gross domestic product (GDP).Indicator 5.1.4was modified to include the contribution of biotechnologies, bioproducts, and carbon credits to the GDP.Indicator 5.1.1 was similarly modified to include the contribution of bioenergy generated from forest biomass to the GDP.This subsector is increasing in relevance as a low-emissions substitute for fossil fuels in light of climate change.
Indicator 5.1.2suffered a decline in predictability due to the indirect effects of climate change on the supply of wood fibre.The linear dependence of secondary manufacturing on the supply of primary manufactured products make fluctuations in this indicator easily understandable, despite the likelihood that climate change will influence the flow of primary timber products.Therefore, Indicator 5.1.2was assigned to the unmodified category.
Indicator 5.1.3is highly dependent on external factors, especially ones external to the C&I framework and to Canada, which run the gamut from the price of fossil fuels to consumer preferences and trade patterns.These relationships obscured the influence of climate change on quality, quantity, and type of wood available, and subsequently the production, consumption, and trade of timber products, thereby diminishing the SFM signal generated by Indicator 5.1.3.However, Canada has long been the world's biggest exporter of forest products (CCFM, 2006), so monitoring international trade is critical to SFM.Furthermore, Indicator 5.1.3was found to be highly influential on other socioeconomic indicators within Criterion 5.So while no modifications were deemed appropriate, we were not ready to abandon the indicator and it was assigned to the unmodified category.Indicator 5.1.5tracks the value of unmarketed forest-based services and products.This indicator has existing challenges in its measurability and feasibility.However, the influence of climate change alone does not warrant any action, and the indicator was left unmodified.
Indicator 5.2.1 tracks the type and distribution of forest tenures on public land, and is an important indicator for the distribution of benefits from the forest sector.The ability of Indicator 5.2.1 to signal SFM progress may be slightly diminished if it does not incorporate emergent societal demand for climate-change mitigation and adaptation on public land.This was largely addressed by the creation of new indicators under Criteria 1 and 2. We did, however, recommend modifying the indicator so that the proportion of tenured area with agreements that specifically address climate change would also be measured.Indicator 5.2.2 tracks the distribution of financial benefits specifically, and was assigned to the unmodified category, as most of the economic impacts of climate change were far removed from the indicator and only marginally reduced its predictability.It also increased in relevance in light of emerging subsectors like carbon markets and bioenergy, as well as job creation from assisted migration and research to discover new forest practices and products.Indicator 5.3.5 is also a widely accepted and used indicator of economic and social welfare, so modifications to the indicator would be both unfavourable and unrealistic, despite the influence of climate change on its effectiveness.

Criterion 6 -Society's Responsibility
Criterion-6 indicators of social responsibility were the most uninfluenced by climate change of all the criteria of SFM.As previously mentioned, these social indicators were far more difficult to evaluate for their interaction with climate change, as all or any effects of climate change were indirect and frequently ambiguous.However, many of these indicators were genuinely independent of climate and climate change and eight of the 13 indicators are action indicators.Several of the indicators were evaluated in groups, which was done not because they were considered less important but because they were closely related and largely independent of climate change.
Indicators 6.1.1,6.1.2,and 6.2.1 are action indicators that track Aboriginal rights, land use, and traditional knowledge, and were all left unmodified.Indicator 6.1.1 was deemed to be uninfluenced by climate change.Indicators 6.1.2and 6.2.1 both had some decline in their predictability due to the effects of climate change, most notably tree-line advance in Canada's northern regions and the threats to traditional knowledge arising from environmental change.However, no modifications were considered to be necessary for these two indicators to remain effective SFM signals.
The four socioeconomic indicators used to track forest-dependent community well-being in Element 6.3 were also evaluated jointly.Indicators 6.3.1, 6.3.3, and 6.3.4 were all influenced by climate change in a near-identical fashion.The short-and long-term ecological, economic, and social impacts of climate change will reduce the predictability of these three socioeconomic state indicators.However, these three indicators are well established measures of community well-being and resilience (Beckley, 2000), so it is more than likely that these will remain a vital tool in monitoring SFM progress in relation to society's responsibility in managing the economic and social benefits of forests in Canada as the climate changes.They were therefore all assigned to the unmodified category.Indicator 6.3.2 tracks education attainment levels, and was determined to be uninfluenced by climate change, as it tracks a climatically insensitive phenomenon.
Indicator 6.4.1 tracks satisfaction with the public engagement process in forest management in Canada.The Indicator was almost entirely unrelated to climate change and had no decline in its ability to signal SFM progress.However, the growing public awareness of climate-change issues, especially those that pertain to forest management, may surface more frequently in advisory committee processes and increase the relevance of the Indicator.Indicator 6.4.2 is another action indicator, and tracks compliance rates with SFM laws and regulations, which makes it closely related to Indicators 3.1 and 3.2.There was a very low potential for the indicator to decline under climate change, and it was determined to be uninfluenced.Indicators 6.5.1 and 6.5.2, evaluated jointly, are action indicators that track the quality and availability of forest inventory data, and did not sustain any declines in their ability to signal progress towards SFM.It seems likely that the indicators will increase in relevance due entirely to their role in the measurement of other major climate-change impacts, especially given recent improvements to inventory capacities with new remote-sensing technologies.However, we still regarded both these indicators as fundamentally uninfluenced by climate change.
Indicators 6.5.3 and 6.5.4 were the only indicators within Criterion 6 where some form of change was recommended.Indicator 6.5.3 is a socioeconomic action indicator tracking investment in the forest sector.The Indicator had a considerable increase in its relevance, largely because of the likelihood for new major research initiatives into climate-change adaptation and mitigation, like Natural Resources Canada's Regional Adaptation Collaborative initiatives (Natural Resources Canada, 2011).We therefore modified the Indicator to also measure investment into climate-change adaptation and mitigation.Indicator 6.5.4 is another broad and encompassing indicator that because of its breadth and relatedness with other indicators was influenced by climate change.Specifically, this action indicator increased in relevance because of the myriad of ecological issues expected and currently arising from climate change.In response to this, we modified the indicator so that the development of new standards and guidelines pertaining to ecological issues would also address those ecological issues caused by the changing climate.

Interpreting the Effects of Climate Change on Indicator Effectiveness
Assessing the influence of climate change on an indicator is far more complex than simply determining how the phenomenon tracked by an indicator will be affected.Moreover, assuming that simply measuring and monitoring a phenomenon that is heavily influenced by climate change will continue to be adequate to gauge SFM progress is misleading.Measurability is just one of many traits that are instrumental to indicator effectiveness.To understand and characterize the complex interactions of climate change with the ability of an indicator to track SFM progress, we analyzed the effects of climate change through a set of seven traits of indicator effectiveness (Duinker, 2001).These traits included measurability, feasibility, responsiveness, relevance, understandability, validity, and predictability.
A simple example of the effects of climate change on measurability is the expected decrease in American marten (Martes americana) populations (Steventon & Daust, 2009).The smaller marten population, as an indicator, would become physically more difficult to count and less measurable.Conversely, climate change is predicted to lead to more invasive alien species (Dukes et al., 2009), and those more-abundant species would become easier to count, and thus more measurable.Feasibility is inextricably linked to measurability, and is more of a limiting factor for indicator effectiveness, especially at the national level.Any change in the measurability of an indicator caused by climate change will subsequently correspond to a change in feasibility for the same reasons.
Responsiveness may also be used to interpret the effects of climate change on indicators.Moose (Alces alces) populations might become far less sensitive to forest management actions due to increasing competition with deer, more parasites, physiological stress, and habitat alteration from more-frequent natural disturbances (Thompson et al., 1998).The cumulative effects of these climatically introduced stressors would mean that the moose population is now far less proportionally affected by forest management activities in its habitat and therefore a poor signal of SFM.
The relevance of an indicator may increase or decrease with climate change.For example, forest-management initiatives for maximizing carbon sequestration and storage are becoming more relevant due to climate change and major mitigation initiatives (Hall, 2002).Moreover, an existing indicator that will be heavily influenced by climate change, such as area of forest disturbed by natural disturbance, will also increase in relevance.A decline in relevance corresponds to a diminishing ability to interpret how changes in an indicator correspond to a given SFM value.Any decline in relevance due to climate change would only arise from a decline in responsiveness.
We have made the assumption that the understandability of an indicator is independent of climate change.
Conversely, validity is such a broad and overarching trait of indicator effectiveness that we assumed that most fluctuations in other indicator traits would correspond to a fluctuation in validity.It is a less tangible indicator trait that was found to have less utility in assessing the effects of climate change on individual indicators, but is far more important in initial indicator design.
Predictability is fundamental in the context of SFM, climate change, and adaptive management because it is the expected future range of an indicator by which we derive management goals and indicator targets (Duinker, 2001).Monitoring is inherently focused on the present and past (Duinker, 1989), so the ability to predict, forecast, or model an indicator into the future is vital to gauging whether indicators will be within acceptable ranges or progressing towards desired targets.Climate change is not only a significant driver of change in forest ecosystems and SFM, but is tremendously uncertain.Consequently, the predictability of indicators is highly sensitive to climate change.
Each of the 46 indicators in the C&I framework has existing strengths and weaknesses in these different traits.In our assessment, we focused exclusively on the effects of climate change on these indicator traits, though the current effectiveness of the SFM indicators, independent of climate change, was often a determinant of how an indicator fared in the evaluation.

Common Patterns in Climate-change Influence on Indicators
As can be expected, there are some major differences for the implications of climate change on biophysical indicators versus social indicators.First, it is important to recognize that developing ecological and economic indicators tends to be a less challenging task than developing social indicators, as they are longer-standing disciplines with many existing systems of data collection and monitoring in place (Bridge et al., 2001).Moreover, the inherent normative nature and political implications of social sustainability indicators further complicate their development (Rametsteiner et al., 2011).Social indicators from Criterion 6, and to a lesser degree Criterion 5, are therefore faced with existing challenges in indicator development and effectiveness (Gough et al., 2008), which will in all likelihood be exacerbated by climate change.The biophysical indicators of Criteria 1 through 4 were heavily affected by predicted and observed climate change, while the influence of climate change on the socioeconomic and socio-political indicators in Criteria 5 and 6 were difficult to assess and, in our assessment, much less pronounced.Also, the vast majority of climate-change influence on the social indicators was indirect, and dependent on linkages with biophysical indicators.
Another key division in the climate-change evaluations was between action indicators, which track the quality and quantity of management actions, and state indicators, which track how forest and forest-related systems react to management action and external drivers.The state indicators were more prone to a decline in their ability to track SFM progress, while the utility of action indicators were often unaffected, or even improved in the face of climate change.It is useful to examine this division through the indicator traits.
Every state indicator declined in predictability to some degree because of climate change.Many of them also had substantive declines in responsiveness and relevance, which often resulted in the modification of the indicator.The action indicators most often had an increase in relevance under climate change, frequently because of the capacity of the actions they track to influence climate-change mitigation or adaptation.However, action indicators were also found to be insufficient in capturing the threats of climate change to SFM values identified by the criterion that contains them.For example, Indicators 3.1 and 3.2 are action indicators designed to track phenomena associated with the SFM values pertaining to water and soil.We know that climate change will affect soil and water to varying degrees (Johnston et al., 2010;Jones et al., 2009), yet both of these indicators were uninfluenced by climate change.This division may emphasize the advanced utility of state indicators in place of action indicators, if possible, despite the propensity for state indicators to decline in effectiveness under climate change.
A general theme of these evaluations was that the C&I most often responded to climate change with a decline in their predictability.This response, though not beneficial for the overall effectiveness of the C&I framework, is far more desirable than a complete lack of response of the indicators.The inherit dependence of forest ecosystems and their management on climate predisposes them to vulnerability to climatic fluctuation.If many of the SFM indicators, especially the biophysical ones, were largely unaffected, it would suggest that the indicators do not fully respond to system variation or are not entirely aligned with the SFM values defined for their respective criteria.This was largely the case for Criterion 3, consisting of two action indicators and a state indicator with low measurability/feasibility. It could also be stated that climate change certainly will make defining and measuring SFM more difficult, but we have no comparable or adequate replacement for C&I-SFM, which are already integrated into many scales and levels of forest management and policy development.
A decline in indicator predictability was in fact the most common outcome for state indicators.Many of these indicators, the biophysical ones in particular, were classified as unmodified in the evaluation.However, it is no surprise to us that the majority of indicators were not modified.They represent fundamental components of forests and SFM that are entirely embedded in the surrounding climate and have no surrogates to replace them as indicators.For example, the growing stock of trees (Indicator 2.1) will be affected by changes in temperature and precipitation, reducing its predictability.However a good surrogate indicator does not exist nor would it be favourable.These declines in predictability occurred because of the uncertainty and variability of future climate change, and rather than necessitating modification to individual indicators, they instead strongly reinforce the need for prospective insight and analysis in applications of C&I-SFM.Current approaches to C&I implementation and monitoring are intrinsically backward-looking.Through measurement of a suite of indicators, we take a snapshot of present conditions in order to monitor trends over time.A key implication of climate change is future uncertainty.Effective C&I implementation will need to draw from forecast modelling and scenario analysis toolsets to address future uncertainty with prospective insight (Duinker, 2011).
A major challenge to the indicator evaluations and especially the indicator modifications pertained to how these national indicators are currently defined and constructed, which we believe also is largely related to issues of spatial scale.Some of the indicators consist of a single objective and measurable variables that are easily aligned with a desired management direction or target -an important attribute of SFM indicators (Duinker, 2001).A prime example of this is Indicator 2.5 (proportion of timber harvest area successfully regenerated).Conversely, Indicator 1.2.2 (population levels of forest-associated species), tracks 41 species of birds and mammals in the national assessment, and is in essence a cluster of indicators.Indicator 6.5.4 (status of new or updated forest management guidelines and standards related to ecological issues), is an exceedingly broad and inclusive indicator that is not easily aligned with a management target.
The above distinctions are important as they heavily influence the type of modifications made to an indicator due to climate change.More specifically, we suggested two types of modifications to the indicators.The first arose when we needed to alter the foundational concepts of an indicator, by making contributions or omissions to the phenomenon tracked by an indicator because of the changing climate.An example of this is the modification to Indicator 6.5.3 to include investment in climate-change adaptation and mitigation in addition to scientific research, industry research and development, and education.The second type of modification arose when we had no need to change what phenomenon an indicator tracks, but instead had to change how the phenomenon is measured.An example of this is again Indicator 2.1, which is currently measured and subsequently modelled under the assumption of a stable climate.

Indicator Adjustments and Improvement for a Changing Climate
Often the case for indicator modification in this project was founded on existing deficiencies in one or more of the seven indicator traits of effectiveness.The most frequent deficiency appeared to be low understandability, resulting from the broad definition of indicators and 'indicator clusters', as defined in Section 4.2.Again, this is at least partly due to scale, as explicit indicator definition is difficult at the national level given the diversity of Canada's jurisdictions, forests, and SFM values (Hickey & Innes, 2005).Clearly defined, understandable indicators with adequate measurability and feasibility are critical, though of course a challenge, as indicators of SFM strive to monitor incredibly complex systems while at the same time being as simple as possible (Wijewardana, 2008).Another recommended course of action for future C&I framework revision would be to adopt state indicators over action indicators, when feasible.The analysis of action-indicator performance under climate change as well as their interaction dynamics in the linkages assessment suggests that action indicators are not as effective as the state indicators within the national framework of C&I-SFM.
One concept relating to indicator-based monitoring of SFM is that of an early warning system for climate change.
For example, climate change, notably warmer winter temperatures, was the catalyst for the mountain pine beetle (Dendroctonus ponderosae) outbreak in western Canada (Carroll et al., 2004).However, the abundance of over-mature lodgepole pine (Pinus contorta) in the region, a major host species, was an early indicator of vulnerability that was not entirely heeded.This represents a situation where the C&I-SFM, in this case Indicators 1.1.1 and 2.3, have the capacity to serve as an early warning system for climate-change vulnerability and impacts, but only if they are so interpreted.Certainly this will not always be possible because of the tremendous uncertainty associated with future climatic change.However, known climate-change impacts should be incorporated into future national assessment initiatives so that early warnings can be heeded.
An ongoing limitation of the indicators is the lack of, or inability to be linked with, a direction or target.This is a vital attribute of indicators that will likely become more critical under climate change and echoes the need for more prospective insight in C&I-SFM.A considerable threat to the utility of indicators to gauge SFM progress under climate change will be an introduced uncertainty and variability in indicator targets.Furthermore, there may be emergent ecological thresholds that are unanticipated because of the interaction of climate change and management actions (Millar et al., 2007), which further complicates the setting of desirable targets.For example, if reforestation initiatives are focused on a given species mix that is no longer climatically favoured, eventually some key thresholds may be exceeded, resulting in time-lags and potential ecosystem collapse.Established targets may be too close to or beyond these thresholds introduced by climate change.In addition to clear indicator definition, the latter point reinforces the use of prospective analysis in C&I implementation.

Conclusions
C&I-SFM are, without question, a key component in the pursuit of SFM.Canadian and international forest and forest-sector stakeholders have embraced them and, using them, tried to determine the degree to which SFM is becoming a reality.Climate change adds considerable complexity and uncertainty to this enterprise.With concerted and coordinated efforts, we can address these complexities and uncertainties in useful ways to improve the capacity to gauge progress in SFM.This study points to a range of improvements that can be made to further the C&I cause.Many of the identified implications of climate change for C&I and subsequent recommendations may also have relevance in C&I applications at the local level (Duinker, 2001).This warrants an independent investigation, not least because of the continuing importance of forest certification and the central position of C&I in certification schemes.A greater degree of uncertainty around potential climate-change impacts at the local level and their implications for management are issues, among others, that may make our recommendations for national-level indicators less applicable at the local level.
It has been established that a failure to address linkages among indicators can lead to deficiencies in the monitoring and assessment of SFM (Wolfslehner & Vacik, 2011).Our evaluation of climate change and C&I-SFM emphasises this point further, and suggests that a failure to address indicator linkages may also reduce the capacity to track the influence of climate change on SFM values.It will also be important in future C&I reviews for the adoption of climate change to acknowledge the participatory nature of indicator development as a process of political negotiation (Rametsteiner et al., 2011), as lack of political commitment has been found to be one of the key challenge in C&I-SFM implementation (Wijewardana, 2008).Stronger links into management and policy decision-making processes will help to guide C&I-SFM forward under the changing climate.Finally, prospective analysis in C&I-SFM implementation will be a critical addition in light of climate change.

Figure
Figure 1.Evaluation framework Indicators 5.3.1 and 5.3.2 are two of the most influential and important indicators in the framework as they track the flow of timber and non-timber forest products from forest ecosystems in Canada.Both indicators were influenced by climate change but were left unmodified.The anticipated impacts of climate change and the already observed alterations to annual allowable cut (AAC) in response to the mountain pine beetle outbreak meant a considerable decline in the predictability of these indicators.However, these two indicators, Indicator 5.3.1 especially, are at the core of SFM in Canada, and we saw no possible modifications or alternative indicators to alleviate the decline in their predictability.Moreover, because they interact so heavily with the biophysical indicators and are therefore influenced by climate change indirectly, it was hoped that much of the decline in the effectiveness of these two indicators would be addressed by the new and modified indicators within Criteria 1 through 4. Indicators 5.3.3 and 5.3.4 are two traditional economic indicators used to gauge the economic welfare of an industry.Both indicators are largely unrelated to other indicators in the C&I framework and were both determined to be uninfluenced by climate change.Indicators 5.3.5 and 5.3.6 track employment variables in the forest sector.Indicator 5.3.6 tracks the distribution of income in employment categories and was deemed uninfluenced by climate change.Indicator 5.3.5 sustained a decline in predictability due to short-and long-term fluctuations in employment deriving from climate change.

Table 1 .
The CCFM 2003 national framework of C&I-SFM Area of forest by type and age class, and wetlands in each ecozone Indicator 1.1.2.Area of forest by type and age class, wetlands, soil types and geomorphological feature types in protected areas in each ecozone Status of in situ and ex situ conservation efforts for native tree species within each ecozone Contribution of timber products to the gross domestic product Indicator 5.1.2.Value of secondary manufacturing of timber products per volume harvested Indicator 5.1.3.Production, consumption, imports and exports of timber products Indicator 5.1.4.Contribution of non-timber forest products and forest-based services to the gross domestic product Indicator 5.1.5.Value of unmarketed non-timber forest products and forest-based services Annual harvest of timber relative to the level of harvest deemed to be sustainable Indicator 5.3.2.Annual harvest of non-timber forest products relative to the level of harvest deemed to be sustainable Extent of consultation with Aboriginals in forest management planning and in the development of policies and legislation related to forest management Indicator 6.1.2.Area of forest owned by Aboriginal peoples Element 6.2.Aboriginal traditional land use and forest-based ecological knowledge

Table 2 .
New indicators recommended for inclusion into the C&I-SFM framework (CCFM, 2003)lobal Ecological CyclesAll four indicators in Criterion 4, role in global ecological cycles are within the carbon cycle element, tracking the Canadian forest's contribution in global carbon cycling.This criterion is unique in the framework in that it expressly addresses climate change(CCFM, 2003), and its indicators were all found to track phenomena that can potentially mitigate climate change and were all therefore more relevant in the face of global environmental change.Two out of four indicators were modified, a new indicator was created, and all were influenced by climate change to some degree.Indicators 4.1.1 and 4.1.2both track the carbon stored in forest ecosystems and are similar in design.As such, they both were subject to the same recommendations for modification.These indicators suffered from the same decline in measurability, and arguably in validity, as Indicator 2.1.Climate change is predicted to affect the physiological growth rate of trees and the productivity of forests, as well as decomposition rates of dead organic matter, which invalidates some of the core assumptions of how we measure net changes in forest carbon.Consequently, Indicators 4.1.1 and 4.1.2were both assigned to the modified category.

Table 2 .
Abbreviated evaluation results, including indicator number, total number of linkages with other indicators in the framework, keywords illustrating the interaction of the indicator with climate change, main references used during the indicator evaluation, and final evaluation category, respectively