森林认证

ly ity o t and Forest Ecology and Management 256 (2008) 1311–1319 Keywords: Forest certification Adaptive risk management a st n on est eable future. We find that current forest certification schemes are largely deficient because they fail to demand: (i) measurable management objectives for biodiversity, (ii) formal risk assessment of competing management options that integrate impacts on biodiversity, (iii) Contents lists availab Forest Ecology an .e l 1. Introduction Forests are among themost species-rich environments on earth and the way they are managed has a substantial impact on global biodiversity loss (Millennium Ecosystem Assessment, 2005). Much of the focus on conserving forest biodiversity has centered on setting aside large reserves (Soule´ and Sanjayan, 1998; Mitter- meier et al., 2005) and wilderness areas (Donlan et al., 2005). Reserves undoubtedly play a key role (Mittermeier et al., 2005), but it is increasingly clear that off-reserve conservation is critical (Lindenmayer et al., 2006), especially asmost of theworld’s biota is presently not in reserves or wilderness areas (Daily, 2001). Approximately 92% of the world’s forests (and associated biota) occur in unreserved areas used for the production of wood, paper and other forest products (Lindenmayer and Franklin, 2002). Biodiversity conservation is now widely acknowledged around the world as a fundamental part of ecologically sustainable forest management (Hunter, 1999; Lindenmayer et al., 2006). Policy documents note that the conservation of biodiversity requires ‘‘conserving species throughout their known ranges’’, maintaining the ‘‘evolutionary potential’’ of populations, and maintaining species interactions and ‘‘ecological processes’’ (e.g. Common- wealth of Australia, 1992, 1996; Haynes et al., 2006). Workable interpretations of these policy statements must be developed through cooperation between managers, the community, and ecologists to provide specific goals and performance measures as a basis for forest management. Market-based instruments such as certification are rapidly gaining popularity as effective motivators for improved forest management. Certification schemes have developed in the fishing industry (Marine Stewardship and Council, 2002) and some areas of agriculture (USDA, 2000). As ofmid-2005,more than 214million ha of forest worldwide had been certified under various standards withmore than 50% of European forests and 30% of North American forests managed under certification schemes (UNECE/FAO, 2005). The area of forest certified under the Forest Stewardship Council (FSC) has increased approximately linearly since 1998 (Fig. 1) and the total area of forest certified under the Pan European Forest Certification Scheme alone is now greater than 200 million hectares. Forest certification is considered a potentially important measure to counter the current ecological problems being created by globalization of the wood products industry (Viana et al., 1996; Due diligence Monitoring Population viability analysis (PVA) Multi-model inference monitoring that directly addresses management performance requirements and a clear plan for how monitoring information will be used to make better management decisions, and (iv) ongoing research targeted toward practices that enhance biodiversity inmanaged landscapes.We argue that the credibility of certification schemes hinges on their ability to dictate scientifically defensible management systems for biodiversity conservation. We present a framework for adaptive risk management (ARM) of biodiversity that is both responsibly proactive and diligently reactive and recommend its incorporation in all certification schemes. We highlight the need for substantial government and agency investment in fostering ARM. � 2008 Elsevier B.V. All rights reserved. * Corresponding author. Tel.: +61 3 83443306. E-mail address: brendanw@unimelb.edu.au (B.A. Wintle). 0378-1127/$ – see front matter � 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.foreco.2008.06.042 Adaptive risk management for certifiab B.A. Wintle a,*, D.B. Lindenmayer b aCommonwealth Environment Research Facility (AEDA), School of Botany, The Univers bCommonwealth Environment Research Facility (AEDA), Fenner School of Environmen A R T I C L E I N F O Article history: Received 22 November 2007 Received in revised form 19 June 2008 Accepted 23 June 2008 A B S T R A C T The past decade has seen hectares of theworld’s fore species-rich environments protected area systems, for biodiversity for the forese journal homepage: www sustainable forestry f Melbourne, Parkville, Victoria 3010, Australia Society, The Australian National University, Canberra, ACT 0200, Australia global surge in forest management certification, with over 200 million ow certified as sustainably harvested. Because forests are some of themost earth and more than 90% of the world’s forests occur outside formal management certification will be one of the pervasive influences on global le at ScienceDirect d Management sev ier .com/ locate / foreco Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 m ag im 19 ogy13 Gullison, 2003). Thus, it is likely that forest management, practiced according to certification standards, will be one of, if not the major influence on forest biodiversity the foreseeable future. Other competing influences on forest biodiversity include forest con- version in the tropics, development in third world economies, and climate change. Under forest certification schemes, standards of conduct are prescribed for forest operations. Some certification schemes defer to existing institutional arrangements in the jurisdiction under which the forest is managed, such as codes of practice and forest management plans. Successful certification rests largely on the existence and adherence to such processes (AFS, 2007). Other schemes are more prescriptive about what constitutes sustainable forest management (FSC, 1996). Common to all certification processes are periodic, third party assessments of adherence to the certification standard. The overall goal in certification is the adoption of standards that will ensure forest management is environmentally sensitive, socially aware, and economically viable (Upton and Base, 1996). The focus of conservation biologists on reserve design as the pre-eminent tool for biodiversity conservation has meant that significantly less effort has been allocated to the development of ecologically sustainable management practices in forests outside reserves (Lindenmayer and Franklin, 2002). A convincing working Fig. 1. Rate of growth in forest areas certified under the Forest Stewardship Council certification scheme since 1998. de fo m (Li su m de m go be ou de ar 1. B.A. Wintle, D.B. Lindenmayer / Forest Ecol12 finition is yet to be developed of what ecologically sustainable rest management actually means in terms of off-reserve forest anagement, making demonstration of sustainability difficult ndenmayer and Franklin, 2003). Noss (1993) concludes that stainable forestry is a ‘‘multifaceted and relative concept’’. A ore realistic approach to demonstrating sustainability may be to fine it in terms of well measurable local and regional manage- ent goals, and attempt to demonstrate progress toward those als (Lindenmayer and Franklin, 2003). Such an approach would consistent with the principles of adaptive risk management tlined below. We believe that six key factors underpin the failure to monstrate ecologically sustainable forest management. These e: A failure to clearly specify biodiversity management objectives and constraints in terms of measurable attributes at the management, landscape and regional level. This hinders transparent evaluation of management performance through monitoring and renders managers largely unaccountable for their management performance (Bunnell et al., 2003). Managers to su ag AF co m in lar m m un m 19 re te plemented as a whole package from goal-setting and system odeling to monitoring and model-updating (e.g. Johnson et al., 97). Despite frequent claims to the contrary, forest management lies more on trial-and-error management (sometimes augmen- 200 im gument in favor of this position may be defended, it ignores the ge body of work that has developed the theory of adaptive anagement to a high degree of sophistication. Adaptive manage- ent provides a coherent approach to decision-making under certainty and a philosophy for learning (Nichols and Williams, 6). However, this is only the case when it is properly syst ar ntext in which the expression ‘‘adaptive management’’ is mmonly used in existing standards indicates a pervasive isconception that any decision to change a management action light of an observed (usually unexpected) change in the state of a em is, by definition adaptive management. While a semantic stan co rious commitment to adaptive management (sensu Walters, 86), linked to a systematic risk assessment protocol is necessary provide a sound basis on which to assert ‘ecologically stainable forest management’. The expression ‘‘adaptive man- ement’’ can be found in standards documents (e.g. FSCC, 2005; S, 2007) although the exact meaning seems to vary from dard to standard and definitions are largely absent. The vers se substantially undermine the decision-making are not being resolved and many research projects are addressing questions that have only a minor influence on decision-making. If forest management were not subject to uncertainty, then the ajor challenge facing managers would be to set goals that were reeable to stakeholders. If agreeable goals could be set, plementation of management would proceed without contro- y. However, because uncertainty is pervasive, we argue that a q have largely failed to setmeasurable performance thresholds for biodiversity or to specify remedial actions that would be conducted if thresholds are breached. 2. Management options (e.g., silvicultural systems) are typically uniform throughout a forest type (e.g. wet schlerophyll eucalypt forest in Australia is almost always clear-felled Lutze et al., 1999), with no attempt to undertake management experiments to test competing theories about best practice and competing social preferences. 3. A failure to formalize competing views about the impacts of forest management (or relative impacts of competing manage- ment options) as transparent models. This makes it difficult for outside observers to identify the expected outcomes of manage- ment and how those expectations were determined. 4. A failure to embrace prospective biodiversity risk analysis (but see FEMAT, 1993). We could find no published peer-reviewed examples of biodiversity risk analyses being used to support the assertion that forest management practices are sustainable. However, there have been several cases where risk assessments demonstrate the opposite (Burnham et al., 1996; Noon and Blakesley, 2006). 5. A failure to design and implement monitoring (sensu Nichols and Williams, 2006) to assess the performance of management strategies for biodiversity conservation. There is commonly a mismatch between the amount ofmoney required to implement successful monitoring and the amount of money managers and policymakers are prepared to invest inmonitoring. A reluctance to set measurable biodiversity management objectives and thresholds (Point #1 above) also makes designing effective monitoring strategies very difficult. 6. A failure to take a systematic approach to setting research priorities based on the uncertainties that most impact on the uality ofmanagement decisions.Many of the uncertainties that and Management 256 (2008) 1311–1319 d by the results of definitive experiments) than formal adaptive Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 Administrator 铅笔 g p ogy management (sensu Walters, 1986; Johnson et al., 1997; Nichols and Williams, 2006, see below). In the remainder of this paper, we outline the key ingredients of an adaptive forest management strategy that would better meet the aims of ecologically sustainable forest management. We argue that scientifically defensible certification schemes should embrace true adaptive management as an overarching framework and philosophy for management and as a minimum standard for certification. 2. Adaptive risk management Formal approaches to adaptive management (Walters, 1986; Walters and Holling, 1990) integrate information from research, monitoring and management to test and improve management practices. Experimentation is central to understanding the system under management, enabling learning from both successes and mistakes under a systematic, replicated experimental design (Taylor et al., 1998). Adaptive management is not management by ‘trial and error’ (Linkov et al., 2006). This is because trial-and- error management: (1) is not underpinned by a formal model (or models) for the system being managed, (2) does not formally identify and select between competingmanagement options using competing systemmodels, (3) does not involve a plan for learning, and (4) is usually not replicated and statistically rigorous. Adaptive management is not a new concept (Walters, 1986), but successful applications are rare in natural resource manage- ment (Stankey et al., 2003, 2005). Several barriers have been identified, including difficulties in modeling ecosystem responses to management, risk avoidance, lack of institutional flexibility, the cost of monitoring, and a lack of community involvement (Stankey et al., 2003). To date, there has been no attempt (that we could find) to reconcile the adaptive management literature with the equally prolific literature on formal risk analysis methods (Burgman, 2005). Risk analysis may be defined as ‘‘the consideration of the sources of risk, their consequences and the likelihood that those consequences may occur’’ (AS/NZS 4360–1999). Risk assessment has become an integral part of conservation science, providing a basis for comparing the value of alternative management options (Akc¸akaya et al., 2004; Wintle et al., 2005), prioritizing conserva- tion effort between species (IUCN, 2001), and setting research priorities (Lindenmayer and Possingham, 1996). Surprisingly, references to formal risk assessment methods and literature are largely absent in the adaptive management literature. The integration of formal risk analysis methods with adaptive management will help overcome some of the major impediments to successful adaptive management, including dealing with risk- averse stakeholders (Gray, 2000; Stankey et al., 2003). It will improve approaches to characterizing uncertainty about manage- ment outcomes and developing robust management strategies. By integrating risk assessment and adaptive management, we envisage a forest management system that is both responsibly proactive and diligently reactive. Given this, we argue that certifiably sustainable forest management systems must be underpinned by adaptive management principles and formal risk assessment methods (Fig. 2). In the following sections, we detail key components of an adaptive risk management (ARM) system needed to underpin ecologically sustainable forest management and, in turn, underpin credible forest certification schemes. 2.1. Statement of goals, constraints and performance measures The first step in the development of an adaptive management program is to clearly define management goals and constraints as B.A. Wintle, D.B. Lindenmayer / Forest Ecol well as measures by which management performance may be limited value unless there is an identified action if that threshold is breached. In our example, one such action might include the cessation of logging until it can be proven (with sufficient confidence) that the population in question has recovered to the required level. The inherent unpredictability of natural systems means that unforeseen population declines may occur that were not predicted by rigorous risk assessment. This should not reflect badly on a manager. Rather, a manager should be judged by how quickly the decline was detected (i.e. how robust was their monitoring strategy) and the speed of implementation of remedial actions. 2.2. Specification of management options Specification of management options is partly a social and partly a scientific process. Management options are usually generated by opinions of stakeholders and scientists about the best means to achieve management objectives. The need for multiple management options arises from uncertainty about the outcomes of particular management options. For example, a manager may predict that the implementation of clearfell harvesting with scattered tree retention will maintain sufficient habitat for large forest owls while ensuring the minimum acceptable economic return. Alternative opinions about the best of t h opulations of forest-dependent species would not fall below 80% he current estimated population size. Setting a threshold has ide oals defined above, a biodiversity performance threshold is ntifiable; the manager must ensure, with 90% confidence, that ma assessed (Possingham, 2001). Without clearly stated goals and performance measures, assertions of sustainability are essentially baseless. Appropriately constructed statements of goals and constraints convert broad (but often opaque) policy objectives such as ‘‘maintain species throughout their range’’ into operational and measurable goals. A possible example would be: ‘‘achieve at least a 7% internal rate of return on investment in timbermanagementwithin the region, subject to the constraint of maintaining (with at least 90% confidence) priority species in populations no less than 80% of their estimated population size within the management region for the next 100 years’’. This statement is characterized by measurable performance criteria (in units of dollars and population size) and both an explicit spatial context (management region) and an explicit temporal context (100 years). It also explicitly states acceptable levels of uncertainty (>90% confidence). The goals are social preferences that must be elicited throughout the management planning process via community engagement. Clear statements of goals make trade-offs explicit; here, some loss of population size may be tolerated for some gain in net economic benefit. Management performance can then be assessed against goals and constraints. We reinforce the key points that: � Goals and constraints must be measurable and clearly define the spatial and temporal scale. � Required confidence (tolerable uncertainty) is explicitly speci- fied. � Specificat