Comparative Analyses Of Mammalian Extinction Risk

Perhaps the most obvious proposed risk factor for extinction is large body size. The end-Pleistocene mass extinction of mammals removed mostly large species (Barnosky, Chapter 12, this volume), and declining mammals are an order of magnitude heavier, on average, than are non-threatened species (Cardillo et al., 2005). There are several possible reasons: Large-bodied species are more tempting targets than small ones for hunters; they are, on average, less abundant; and they take longer to reach sexual maturity, have smaller litters of larger offspring, and have larger individual home ranges. Narrow ecological tolerances are also a plausible risk factor—habitat specialists may be more at risk than generalists from

272 / T. Jonathan Davies et al.

habitat loss (Fisher and Owens, 2004). A small geographic range size may reflect narrow tolerances and increase the risk that the whole of the species' range is in the firing line. Which of these features matter most for extinction risk, and are any associations consistent across all mammals?

Cardillo et al. (2005) carried out the most comprehensive investigation to date. Threatened species were included only if they were on the IUCN Red List because of observed decline, to avoid autocorrelation with predictor variables. Red List status, on a 0-5 scale, was used as the response variable. Many facets of geography (including human population density), ecology, and life history were tested as predictors of extinction risk, by using phylogenetically independent contrasts. A phylogenetic approach is needed because, although extinction risk and some of the possible predictors listed above (e.g., geographic range size) do not evolve along the phylogeny's branches like, say, body size does, they nonetheless tend to show phylogenetic signal [i.e., they tend to take more similar values in close relatives than in species chosen at random; (McKinney, 1997; Purvis et al., 2000b; Fisher and Owens, 2004)]. Minimum adequate models were derived from a large initial set of predictors. This approach helps exclude variables that correlate only indirectly with extinction risk, for example, because another variable shapes both them and risk.

The predictors of risk were significantly different for smaller and larger species, with the importance of many predictors changing markedly at a body size of =3 kg. Species smaller than this fit the firing-line model: They are more likely to be threatened if they have small geographic ranges, live in temperate areas, face high human population densities, and live where a high proportion of the other mammal species are also threatened. Larger species, however, face multiple jeopardy: Biology matters as well as geography, with high-abundance, small neonates, and many litters per year all independently helping to bullet-proof species. High abundance is predicted to bullet-proof species if the field-of-bullets model operates at the level of individuals rather than species (Erwin, 2006a), but such a model also predicts that no other biological traits would independently predict risk.

For both large and small mammals, the most important single risk factor is small geographic range size. The firing-line model predicts that small-ranged species will be most at risk because a single localized threat can impact their entire distribution. However, range size itself varies systematically among clades [although it shows weaker phylogenetic signal than, e.g., body size (Gaston, 2003; KE Jones et al., 2005)], suggesting that it is shaped, at least in part, by organismal traits such as dispersal ability (Bohning-Gaese et al., 2006) or niche breadth as well as by circumstances of geography. For example, small-ranged species are more common at low latitudes and within climatically stable regions. Any traits, including geographic location, that confer large species ranges also help make species bulletproof (although geographical variation in species' range sizes again complicates separation of geographical variability in threat intensity from intrinsic biological vulnerability).

Large-scale analyses can find general predictors of extinction risk but can miss interesting variation among regions or clades, which more narrowly focused models might pick up (Fisher and Owens, 2004). Order-specific models typically have higher explanatory power than the large-scale models. These models have some common features, such as the importance of geographic range size, but also differ considerably (Cardillo et al., 2008). For example, body size is a predictor in bats but not in rodents, whereas different life history traits predict risk in carnivores (gestation length) and ungulates (weaning age, interlitter interval). Likewise, different environmental factors and measures of human impact are implicated in different taxa. The models also vary regionally, with life history mattering less in north temperate regions than elsewhere (Cardillo et al., 2008).

One likely source of variation in models is that different drivers may select against different characteristics and show spatial variation in intensity. Broad-scale analyses may therefore lump competing signals together (Owens and Bennett, 2000). Within artiodactyls, predictors of extinction risk differ between hunted and nonhunted species: Late weaning age was the sole risk factor among the former, whereas low income levels among local people and small range size predicted risk among the latter (Price and Gittleman, 2007). More generally, low reproductive rates and large size are likely risk factors for overexploitation, but a specialized habitat may matter more under habitat loss (Owens and Bennett, 2000). Analyses focused more tightly on driver-specific responses often tend to consider far fewer species, in which case far fewer predictor variables can be considered simultaneously without overfitting, and statistical power may be lower. On the plus side, the tighter focus can reduce the chance of mixed signals [although interaction terms can also do this (Price and Gittleman, 2007)], and more precise measures of driver intensity and extinction risk might be available than can be had globally (Fisher et al., 2003; Isaac and Cowlishaw, 2004). Broad- and narrow-scale analyses each give part of an obviously very complex picture. Furthermore, we have focused on phy-logenetic nonindependence, but to fully consider the interaction between biology and geography, the development of methods that also deal with spatial nonindependence in comparative data will be critical.

Analyses modeling risk as a function of intrinsic biology (i.e., not including driver information) can highlight species at lower risk than expected from their geography, ecology, and life history (Cardillo et al., 2006). Such species may be particularly likely to decline rapidly if drivers intensify, because their attributes are repeatedly found in rapidly declining

274 / T. Jonathan Davies et al.

taxa. Cardillo et al. (2006) termed this ''latent risk'' and proposed 20 regions with largely intact, but intrinsically susceptible, mammalian faunas. These include the Nearctic boreal forests and the island arc of Southeast Asia, and are mostly not exceptionally high in numbers of total, endemic, or threatened species. Many have much less than 10% of their land within reserves, and some (especially in Southeast Asia) face very rapid human population growth. As such, latent-risk hotspots might represent cost-effective options for long-term conservation. However, these analyses do not yet consider realistic scenarios of future driver patterns; rather, they implicitly assume that places with low intensity will experience an increase to more typical levels (Cardillo et al., 2006). The next section discusses how more policy-relevant predictions could be obtained by projecting future driver patterns based on explicit scenarios.

0 0

Post a comment