|
|
 |
|||||||||||
|
|
 |
|
|||||||||
|
|
 |
 |
Within this section:
|
|
|
Home A Biodiversity Conservation Plan for Papua New Guinea Based on Biodiversity Trade-offs AnalysisConclusions, comparisons and discussion A set of biodiversity priority areas which together represent 608 environmental domains, 564 vegetation types, 10 species bioclimatic profile clusters and 11 rare and threatened species, has been identified for Papua New Guinea. This set of areas also includes all existing protected areas and samples all CNA priority 1 areas (with the possible exception of one in north-west Sepik; Figure 2d). In addition, the set minimizes foregone opportunities for timber production, avoids areas of high agricultural potential, avoids areas of high existing land use intensity and gives preference to areas of low human population density. A total of 16.8% of PNG was required when 1) existing protected areas were included, 2) areas were excluded that were judged unsuitable as biodiversity priority area candidates because of past land use, and 3) areas offering other land use opportunities were deliberately avoided. We reached a biodiversity target in a way that provides a potential benchmark for comparison with other countries. The total number of environmental domains and vegetation types was determined by finding the number that could be represented in 10% of PNG, assuming that there were no people and no opportunity costs. It was possible to do this because the classifications are hierarchical, (in the case of the vegetation types, made partly hierarchical by overlaying types onto broad physico-climatic zones) with finer distinctions between classes expressed at lower levels in the hierarchy and broader distinctions at higher levels. This highlights the fact that more classes, and in theory more biodiversity, could be sampled if the target was, say, 15% (see Faith et al. 2001a). Such a target should be considered in future planning. The approach to biodiversity planning described here recognizes that such planning is an ongoing iterative process. The data sets and the computer software supporting the current set of biodiversity priority areas, and the skills needed to use them, have been delivered by this project to relevant PNG government officers. The current set of biodiversity priority areas can be expected to change as knowledge accumulates and as social and economic conditions change. Faith et al. (1999 and 2001b) outline approaches to linking these maps to implementation issues, such as environmental levies and carbon offsets. This study represents the first ever whole-country study based on systematic biodiversity trade-offs methods. Some comparisons with other approaches are interesting. We have tabulated some alternative analyses of the PNG data (Table 4). First, we examine some differences in total cost. If we had interpreted "efficiency" as minimum number of areas as in conventional minimum sets approaches (e.g., Pressey et al. 1993), the actual cost in timber volume units for the proposed set would have gone up by more than 20% (last column, Table 4). This result reinforces the comparisons from simple case studies (Faith et al. 1994, 1996). It is interesting that taking area as the cost in our trade-offs analyses would approximate the results, calculated in terms of timber costs, that are found when using timber volume costs directly (5th column, Table 4). We do not, however, take this to mean that area might be recommended more generally as a stand-in for opportunity costs (contra Balmford et al. 2000); in other regions and for other opportunities, opportunity costs could be unrelated to area. Our current best set achieves the 10%-based target with a cost of 93,218 timber volume units, but that cost would have been only 71,280 units if the existing protected areas were not included up front as commitments. The cost from the existing protected areas alone is 34,771 units (Table 4). The total area of the current priority set follows the same pattern. Without committing existing protected areas, the total area needed to achieve the 10%-base target would not be 16.8%, but just 12.9% of total PNG area. All these reported analyses ignored, for the purpose of calculating representativeness, attribute occurrences in an RMU of less than 1 kilometer squared, based on a viability argument. Our initial explorations when these occurrences were counted suggested that this is another factor, in addition to existing reserves, that can dramatically influence total area required to reach a target. Faith et al. (2001b) discuss these viability/persistence issues further. Achieving the 10%-based target required 16.8% of total area of PNG. Would we find a higher or lower total area in other regions/countries? It would be interesting to determine the total area needed elsewhere to achieve a 10%-based target, and interesting to determine as well which costs and constraints account for any extra area needed over 10%. Failure to consider biodiversity targets and trade-offs early in the planning process can limit the subsequent capacity of a region/country to achieve effective trade-offs. Our biodiversity target in PNG cost 93,218 timber volume units, with 34,771 units alone contributed by the commitment to an existing set of protected areas that are not particularly representative. That cost is much more than an estimated cost of 58,000 units or less suggested by the baseline analysis (column 2, Table 4). We presented a similar hypothetical example (Faith 1995) in a theoretical study on how past land use could limit the achievable degree of balance - the best-possible "regional sustainability" level. Early consideration of targets and costs can avoid reducing the capacity for compromise and balance (see also Pressey 1994). A long period of time has passed between our initial case study exploring the utility of trade-offs approaches for biodiversity planning (Faith et al. 1994, 1996), their incorporation into the BioRap toolbox, and this real-world application of the methods in PNG. The development of these tools occurred against the backdrop of recent biodiversity planning processes in Australia (Commonwealth of Australia 1997). While not taken up at the time, these trade-offs approaches have now directly influenced the Federal Government environment department's new planning system (Commonwealth of Australia 1999). The role of science in influencing biodiversity planning in Australia is also discussed in Pressey (1998) and Ferrier et al. (2000). Prendergast et al. (1999) recently argued that what they perceive as gaps between theory and practice in selecting nature reserves might be overcome if the research was published in management journals. But their paper unintentionally highlights another problem, given that they advocate publishing new methods in management journals while remaining unaware themselves of the new methods published in management journals (e.g., Faith et al. 1996). If there is no real synthesis by scientists of the scientific developments, the wheel keeps getting re-invented, and methods incorporating cost, for example, remain novelties (as in Prendergast et al. 1999 and Balmford et al. 2000). Applications are needed, rather than more studies that again show that algorithms that achieve a target by minimizing costs will indeed do the best at minimising costs. Re-discoveries of the established links between complementarity and costs present new confusions. For example, Balmford and Gaston (1999) argue that anticipated complementarity-based cost savings (arising from a smaller number of areas used) justify new surveys to obtain the species lists they believe are necessary for conservation. Certainly, the application of complementarity does lead to cheaper representation than any selection of areas that ignores complementarity. But we don't necessarily need "species lists for each candidate site" (Balmford and Gaston 1999) in order to make savings. Thus, the major premise for their argument for new surveys is incorrect. In our PNG study, as in the earlier case studies, existing data provide the basis for using complementarity and reaching biodiversity targets at low cost. Thus, while Balmford and Gaston would call for new surveys, we advocate the kind of rapid assessment carried out in this project using surrogates, which at the end of the day are all that is available if results are to be obtained in time for effective conservation action in countries such as PNG. That assessment not only can determine which areas should be protected now, but also point to the most urgent information needs for ongoing assessments and planning. Cost trade-offs may not be equally applicable at all levels of planning. We have joined others in arguing for consideration of a suitable costs framework in international priority setting (Mace et al. 2000), but we doubt the utility of a recently proposed use of costs for prioritising among countries (Balmford et al. 2000). In that scheme, based on finding a representative set of countries having low cost, a country having unique biodiversity components and low estimated opportunity costs of conservation would end up as a high priority. This might be one form of guidance for international conservation efforts. However, given that their priority set never represents all biodiversity, it presumes that protecting the biodiversity of countries with high opportunity costs is a lost cause. We would argue that if anything the opposite might pertain. Any country in which the opportunity costs of achieving a target (say, a 10%-based target) were estimated to be high, would be one deserving a high, not low, priority for conservation investment. High opportunity costs imply high potential conflicts. Investment could be used to facilitate within-country planning based on trade-offs so as to urgently achieve a balance between biodiversity protection and production, before sustainable options were lost (see also Faith in press). The problem of prioritizing at different geographic scales is relevant also to PNG, where there will be a need for within province biodiversity planning. At present, we have no formal link to propose between whole-country and within province planning. It may be useful to carry out area substitution within provinces, starting with the priority set members, and also identifying those areas that require urgent decisions - to be made (as discussed earlier). The need for within-province planning highlights again that the set of priority areas identified in this study is only the "current best set". Planning is an ongoing iterative process and any set of priority areas can be expected to change as a result of factors such as new decisions on land uses, changes in economic, social and political conditions, and changes in ecological and biological knowledge. It is almost inevitable that some such changes will occur. Even if it were decided in PNG to implement the current best set of priority areas, it would take time to negotiate with all stakeholders. It is likely that only a few priority areas could be accorded some form of protection within, say, a year, which means it would take many years to implement the entire plan. Thus, areas not selected in the current best set may assume higher priority in the future. However, there is always a sub-set of areas which are irreplaceable (Pressey et al. 1993) if the biodiversity target is to be achieved and it will always be necessary to include these in any set of biodiversity priority areas. It is not feasible for any set of priority areas alone to represent all of biodiversity (unless they cover the whole of the country). Areas not selected in the current best set still contain biodiversity and some will contain components of biodiversity not represented in the current best set of priority areas. The priority areas were selected in a way that minimized the number that were attractive/vulnerable for other land uses (specifically timber and agriculture) - in other words, such areas often were given low priority for protection. This approach contrasts with others (see Margules and Pressey 2000) that give high priority for protection to areas that are attractive for/vulnerable to other land uses. Our view is that regional sustainability and corresponding net benefits for society depend on balancing biodiversity and other land uses, as attempted in this study. It is notable that the biodiversity target was achieved here at low cost - the scope for balance and compromise was great. Indeed, opportunity costs were among the least important factors determining the increase in total area needed over that in the baseline 10% analysis. Inclusion of existing protected areas and exclusion of very small samples of the target biodiversity features (less than 1km2 occurrences) were most important. These results suggest that the major demand on effective protected area systems may not be the competing land uses so much as other design issues relating to persistence. Many of the RMUs selected are small, probably too small to form viable protected areas either from the point of view of ecological persistence or management. If the smaller members of this set were to be chosen as potential protected areas it would be rational and necessary to proceed with clusters of them forming the basis of negotiated Wildlife Management Areas. Wildlife Management Areas (WMAs) form the mechanism currently used in PNG to negotiate biodiversity protection. Many of the existing protected areas shown in Figure 1c are WMAs, especially the larger ones. These reappear in Figure 2a as members of the current best set of biodiversity priority areas because existing protected areas were committed to selection. Another approach to dealing with this problem would be to incorporate an adjacency option into the area selection software. This would allow for the possibility of choosing an area adjacent to one already chosen. Choosing adjacent areas implies a cost that may have to be traded-off with other demands. In the workshops accompanying the PNG study, TARGET was used to restrict new selected areas to be adjacent to existing proposed areas. The degree of restriction depended on an "importance weighting". Adjacency options are being programmed into the TARGET software (Walker et al. unpublished) as part of a current project to identify biodiversity priority areas in tropical Queensland, minimising lost opportunities for tourism. We have described how the trade-offs procedure used here represents a departure from traditional methods. In conventional systematic conservation planning, "efficiency" is all about how well biodiversity representation targets are achieved relative to number or area of reserves (e.g., Margules et al. 1988; 1994; Pressey and Nicholls 1989a,b; Pressey et al. 1993). We link efficiency instead to "regional sustainability" (Faith 1995) - a balancing act that encompasses general opportunity costs, in a framework that may or may not have biodiversity targets. The departure from traditional approaches includes more than abandoning "minimum sets". We see "complementarity" itself, which conventionally is defined as amounts of biodiversity, as incorporating both representativeness and persistence (Faith and Walker (1996c). That more general formulation plays an important role in trade-offs analyses, allowing partial protection, design issues and other aspects to be taken into account. Faith et al. (2001b) treat some of these issues in their paper on future planning in PNG.
|
|
|
|||
|
|
|||||||||||
