Commentary Ecological mechanisms of extinction

E theory offers predictions, sometimes conflicting, about the ecological characteristics of species that correlate with their risk of extinction. It is generally agreed that risk should be higher for species with small populations, small geographic ranges, and poor dispersal ability than for their ecological counterparts (1–3). How suites of life history characteristics affect risk of extinction is less clear. Species with high variance in the intrinsic rate of population increase (r), which is often associated with high fecundity, moderate to low survival rates, short generation times, and small body size, are predicted to be more susceptible to extinction because they are prone to large stochastic population fluctuations (4). Alternatively, species with a low r (because of low fecundity, high survival, and long generation times) are predicted to be at increased risk, because they would recover slowly from a severe reduction in population size and remain threatened longer by demographic and genetic stochasticity (5, 6). Such species are typically large. Thus, it is unclear whether the ‘‘fast lifestyle’’ associated with small body size and short generation times or the ‘‘slow lifestyle’’ represented by large organisms with long generation times makes species and lineages more or less likely to become extinct. Empirical studies of island fauna yielded contradictory conclusions about the effects of body size and lifestyle on the risk of extinction and produced conflicting explanations to account for the mechanisms underlying the patterns (5, 7–11). Resolution of such issues transcends academic debates, as governments and conservation organizations struggle to apply laws, like the United States Endangered Species Act, and decide how to rank threats and allocate funds among taxa that may differ in risk (12). Now Owens and Bennett (13) present new evidence that the ecological mechanisms underlying extinction may differ for lineages of birds threatened by habitat loss and for lineages threatened by human persecution and introduced predators. The study both supports and challenges current thinking in extinction theory and raises a number of intriguing issues. The authors tested predictions about extinction theory on a database of 95 avian families using phylogenetic comparative methods. Birds are one of the few taxa whose species are well enough described, whose phylogeny has been widely investigated (14), and whose ecology is sufficiently known to permit a global analysis. The outstanding scholarship of Collar et al. (15), who compiled life history accounts of all 1,111 species of birds in the world thought to be at risk, made possible this and similar analyses (16). Multiple factors may interact to threaten species. About one-third of the world’s threatened bird species are at risk from direct mortality because of human persecution, including harvesting, poisoning, egg collecting, and capture for trade, and by predation from introduced predators, which has been especially devastating to fauna and flora on island ecosystems. These factors directly reduce survival andyor reproduction, to result in population declines. Birds are primarily threatened by habitat loss because of habitat destruction and habitat degradation from agricultural practices and water management, which affects over two-thirds of the threatened species. For birds, habitat loss may not result in direct mortality unless the impacts destroy active nests. However, mortality may occur after habitat has been lost through starvation, accidents, and predation caused when birds must disperse in search of unspoiled areas to live in and from crowding into remaining habitats (17). The impacts of habitat loss, however, are likely to be different for less vagile or smaller animals and plants, which may suffer more immediate mortality. Although habitat loss and human persecutionyintroduced predators can occur simultaneously to drive a species toward extinction, Owens and Bennett (13) found that they often acted independently on lineages, as there was no correlation between the percentage of species within a family threatened by one force or the other. Could differences among species in extinction risk be caused by differential vulnerability of lineages to habitat loss vs. persecutionypredation as a result of the differing ecological pathways that these forces affect? Owens and Bennett’s (13) results suggest they could. Extinction risks through human persecution and introduced predators were associated with birds that had large body size and long generation times. This result is entirely expected, because rate of population change (r) of long-lived and slowly reproducing species and lineages is especially sensitive to small perturbations to adult survival (18). For example, factors that lowered the survival of long-lived adult California condors (Gymnogyps californianus) or albatrosses would have a far greater impact on population change than proportional changes in their reproductive success, which is limited by a clutch size of one egg annually. Thus, slow lifestyle species and lineages should be more susceptible to human persecution and introduced predators if the impact primarily affects survival, compared to their fastlifestyle counterparts. This phenomenon is illustrated effectively by the differential vulnerability of parrots and finches harvested as adults for the international pet trade (Table 1). Between 1,600,000 and 3,200,000 birds were taken annually from wild populations for the live bird industry in the 1990s (19). Finches of the families Passeridae and Fringillidae composed 70% of the trade, and parrots (Psittacidae) accounted for 25% of the volume. The body mass of a typical finch is nearly 10 times smaller than the mass of a modal parrot. Despite the large numbers of finches traded, parrots suffer over three times the rate of threat that finches incur from trade (Table 1), making Psittacidae among the most threatened families of birds (6). Many parrots are threatened by a combination of trade and habitat destruction (20), but trade may be more threatening because often species persist in a variety of disturbed habitats (21). Life history differences partly explain why parrots are more susceptible to overharvesting than finches (Table 1). Annual fecundity of parrots is much less than that of finches by virtue of smaller clutch sizes and fewer broods per year. Finches do not require specialized structures for nesting, whereas parrots typically nest in tree cavities, which are often in short supply. This results in large