|
|
 |
|||||||||||
|
|
 |
|
|||||||||
|
|
 |
 |
Within this section:
|
|
|
Home A Biodiversity Conservation Plan for Papua New Guinea Based on Biodiversity Trade-offs AnalysisBiodiversity priority setting methods include a range of methods developed originally in Australia (e.g., Kirkpatrick 1983; Margules 1989; Margules et al. 1988; 1994; Pressey and Nicholls 1989a,b; Pressey et al. 1993) and now applied elsewhere (e.g., Rebelo and Siegfried 1992; Kershaw et al. 1994; Lombard 1995). These earlier methods were designed to find sets of areas, which fully represent biodiversity features, while minimizing the area required to do so. The approach adopted here departs from those methods by incorporating opportunity costs (for example, forgone timber production) using a trade-offs approach (Faith et al. 1994; Faith et al. 1996) which incorporates the concept of complementarity established in the earlier methods as the basis for biodiversity values of areas. This form of multi-criteria analysis (see also Faith and Walker 1996a,b) searches for a balance between (often) conflicting objectives, and is linked to a form of "regional sustainability" (Faith 1995). Sustainability is often referred to in the context of individual areas, but a balance may also be sought regionally through the allocation of different land uses to different areas. Attempts to achieve such a balance raise several issues. Reaching a biodiversity target does not by itself imply a high degree of sustainability. Costs and constraints must be taken into account so that solutions can be found along a realistic trade-offs curve, providing high net benefits (Faith 1995). Constraints, such as those implied by land lost through degradation, can mean that the available trade-offs curve no longer provides high net benefits (as in the trade-offs curves in figure 1 of Faith et al. (2001a), where the darker curve implies lower regional sustainability levels). This study addresses explicit trade-offs involved in achieving a biodiversity target in a set of priority areas. Biodiversity priority areas were found by establishing the level of heterogeneity achievable in 10% of PNG (Faith et al. 2001a), then finding that set of areas which together reach this goal efficiently, while minimizing foregone forestry opportunities, avoiding areas of agricultural potential, incorporating existing protected areas, avoiding areas of high land use intensity and high human population density and preferring, where possible, that they coincide with areas chosen previously by experts as high biodiversity priority areas. The TARGET software (Walker and Faith 1998) was used for these analyses. TARGET (or 'TD' for targets and diversity) is one module of the DIVERSITY software package (Faith and Walker 1994, 1996a) which forms part of the BioRap toolbox. TARGET assumes that the areas in a region are described as containing one or more different biodiversity "attributes", which are to be the surrogates for all biodiversity. Within each area, each surrogate also has some quantitative value associated with it - this value might, for example, correspond to the total number of hectares of that attribute within that area. Each attribute is assigned a target for representation. This might be constant over all attributes (e.g., 10% of total area) or vary to reflect the degree of threat or persistence of different attributes (see Faith and Walker 1996c, 1997; Faith et al. 2001b). In the PNG study, the target level of representation was simply a single representative of each attribute from the set of attributes determined by the 10%-based target. TARGET implements the multicriteria approach based on biodiversity complementarity values, described in Faith et al. (1994, 1996). When costs are taken into account, the relative "importance" or weight given to these costs, relative to biodiversity representation, will influence the outcome of the allocation procedure. An area is justified for protection if and only if its "complementarity" value (that is, its marginal contribution to overall biodiversity representation) exceeds its weighted cost. This marginal contribution of a given area simply reflects how much additional contribution it makes to the overall regional achievement of the biodiversity target. For any given area in PNG, the software calculates the number of so-far-under-represented attributes that the area could contribute to the list of selected areas. The software iteratively adds and deletes areas from a list of nominated areas (the "select list") so as to approach the target levels of representation. When cost trade-offs are used, the area which is added to the "select list", at any stage, is the one which has the greatest difference between complementarity and (weighted) cost. TARGET allows a range of search strategies to be implemented. One can start "from scratch" or with all but a set of preferred areas masked out. Alternatively, it is possible to use a set of randomly selected areas as a starting point - the method adds and deletes areas in searching for a set whose members collectively achieve a nominated biodiversity goal and also all have complementarity values exceeding their (weighted) costs. The simple search provided by the basic algorithm can be extended and modified. For example, the search can begin with a high weighting on costs, such that targets are not met, and this partial result read in to a subsequent analysis with lower weight on costs. This strategy, which best minimizes costs, can be applied iteratively until the biodiversity target is met. Similar iterative approaches might initially mask out some areas, giving preference to others until later iterations. Both of these strategies were used to derive the set of biodiversity priority areas shown in Figure 2a below. Of course, it is also possible to find the cost to biodiversity of making the resources available for meeting a production target, such as a certain timber volume, for example. TARGET allows the search for a set of areas for a given biodiversity level, budget level, or where dictated by weightings. When several "costs" are involved, the approach uses weightings applied to each (as in Faith et al 1996).
|
|
|
|||
|
|
|||||||||||
