as a whole, but other clades are likely to have very different—and equally complex—patterns of diversity and extinction risk. The lack of strong surrogacy among groups introduces extra uncertainty into the measured biodiversity value of the regions being considered. In addition, we have focused on conservation benefits rather than costs, but costs vary spatially by several orders of magnitude (Balmford et al., 2003a; Halpern et al., 2006). Cost–benefit models can suggest very different priorities from allocations based solely on perceived biodiversity value (Possingham et al., 2002). So, if we assume that rational decision making must consider both benefits and costs, perhaps the most sensible investment would be in intact but susceptible regions (Balmford et al., 2003a; Cardillo et al., 2006). Public health-care systems may provide a useful analogy: A balanced health-care strategy includes money for preventative medicine as well as for hospital wards and life support.
The primary goal for most conservation management has been to maximize preservation of current diversity. However, by altering the environment, humans also influence future evolution (Smith and Bernatchez, 2008). A previous National Academy of Sciences Colloquium (Cowling and Pressey, 2001; Myers and Knoll, 2001) raised an important question: Should conservation goals be extended to consider the evolutionary future? A range of timescales might be considered. In the short term, species-recovery plans can address the requirements needed for continued adaptive evolution within populations (Mace and Purvis, 2008). But what of the longer term? We can identify clades that have recently diversified and the regions in which they are found (Fig. 14.2b). These lineages or areas might represent engines of speciation: Are they therefore conservation priorities? The distinction between maximizing evolutionary history versus centers of diversification is nontrivial. A network of reserves designed to capture maximal evolutionary history would look very different from one designed to capture rapidly speciating lineages (Table 14.1), because rapid diversification results in low phylogenetic diversity per species (Mace et al., 2003; Roy and Goldberg, 2007). However, past “success” may be a poor indicator of future performance because of the contingent nature of evolution (de Queiroz, 2002; Mace et al., 2003). The geographic pattern of mammalian diversification rates has changed markedly through time (Fig. 14.2a–c), and different lineages have also radiated at different times (Alroy, 2000; Bininda-Emonds et al., 2007).
One secure prediction is that future environmental conditions will almost certainly differ from those in the past. Nonetheless, extrapolating current trends, key environmental changes are likely to include increas-