The need of optimal conservation strategies

Abstract

The biodiversity crisis that the world is facing (Pimm et al., 1995) urges immediate conservation actions to counteract it, and the extreme paucity of the resources available to undertake these actions means that conservation managers are in urgent need of optimal choices. Prioritization among possible actions is therefore of paramount importance to choose strategies that give the best return on conservation investment and thus maximize benefit (Murdoch et al., 2007; Wilson et al., 2007). Yet conservation biology is by definition (Soule, 1985) a crisis discipline, with conservation scientists often asked to make tactical decisions based on scarce information. This is one of the reasons why conservation biology is pervaded by uncertainty (Regan, Colyvan & Burgman, 2002; Rondinini et al., 2006), which in turn paves the road to subjectivity (much more so than in other related disciplines, e.g. ecology and genetics) because incomplete data complicate the task of comparing objectively alternative, competing hypotheses and reject wrong hypotheses. In the last 25 years, quantitative techniques that allow an objective prioritization of conservation strategies have been developed for systematic conservation planning (Margules & Pressey, 2000), the selection of sites to form networks of conservation areas to protect a species assemblage. The first, intuitive but groundbreaking paper on the subject introduced a technique to identify the minimum set of sites that represented a given assemblage of species in an area (Krikpatrick, 1983). Although not explicitly stated in that paper, the method did allow to rank alternative conservation strategies and select the most parsimonious. Thousands of papers on the subject have been published since. The theory has developed in complex algorithms that aid the identification of optimal strategies for large numbers of species and natural processes while taking into account cost, threats, temporal dynamics, zoning and uncertainty. The techniques are now routinely used to inform conservation action (Pressey, Johnson & Wilson, 1994; Ball & Possingham, 2000; Wilson et al., 2006). Optimization methods for conservation strategies other than site selection have been much less explored. Recent investigations include: how to minimize conflicts between humans and large carnivores in a spatially explicit manner (Rondinini & Boitani, 2007); the investigation of the tradeoffs between collecting data on threatened species and taking conservation action based on incomplete data (Chades et al., 2008); a multi-criteria evaluation framework for making conservation decisions that have consequences for multiple species, when each possible decision is more beneficial to some species than others (Drechsler, 2004). Gilioli et al. (2008) extend the scope of optimization to the choice among alternative conservation strategies for a metapopulation of a single species. They do it in a particularly intuitive and appealing manner, by measuring the distance to extinction of the metapopulation, modelled by an incidence function model (Hanski, 1994) under the alternative conservation strategies. As reviewed by Gilioli et al. (2008), a few indices of risk of extinction of metapopulations exist, but the Kullback–Leibler information measure (Kullback & Leibler, 1951) proposed by the authors has advantages over these indices: it can easily be computed irrespective of the number of patches modelled, and allows to evaluate the contribution of each patch to metapopulation persistence under the hypothesis of equilibrium. Gilioli et al. (2008) apply the measure on populations of amphibians, a group of particular conservation interest because one-third of their species are threatened with extinction (Stuart et al., 2004). While the species selected for the analysis (Bufo bufo and Rana temporaria) are still common outside the study area and therefore are not a conservation priority worldwide, the same technique could be readily applied to other, threatened amphibian species. Even though (as Gilioli et al., 2008 point out) it is always difficult to assess if a real population behaves as a metapopulation, their paper may be relevant for a large number of species beyond amphibians. Habitat destruction, fragmentation and degradation, mechanisms that may induce metapopulation dynamics, are for example the highest threat worldwide to mammal persistence (Schipper et al., 2008).