Top predators and biodiversity: much debate, few data

Abstract

Summary In two recent papers, we found support for the hypothesis of an association between the occurrence of top predators and biodiversity. Such results were recently criticized. Cabeza, Arponen & Van Teeffelen (2008) claim that the proper way to test the investigated relationship was to conduct a regional‐level complementarity analysis that pools all predators across all habitat‐types. However, such an approach may not help in assessing the performance of predators as surrogates for wider biodiversity, because: (a) management programmes on top predators almost invariably focus on a single species; (b) conservation planning and management often occur within habitats, not only across them; and (c) the required surveys and analyses may often be logistically unfeasible. Re‐analysis of our data suggests that a different methodology would not change the results. Kéry, Royle & Schmid (2008) used data from a Swiss volunteers’ monitoring scheme to claim that we missed a large number of species in our estimates of avian richness. However, we list 10 reasons why their calculations were inflated, given major differences in surface‐coverage, habitat‐sampling, field effort and in the way that secretive species were treated. Furthermore, our results for birds were replicated in analyses of butterflies and trees, which could not suffer from the same detectability biases. Roth & Weber (2008) analysed the same Swiss database and found that a group of lower‐trophic‐level species (tits) out‐performed raptors as indicators of butterfly richness (but not of bird richness). However, their analyses did not control for ecosystem identity, and data on raptor locations were of dubious quality. Furthermore, because tits are major caterpillar predators, the reported findings could have been expected on other grounds. Synthesis and applications. We disagree with Cabeza and colleagues that local studies are of limited importance. Given the paucity of available information, many more local studies are needed to check the generality of the ecological pattern we found. Until further analyses become available, polarized positions for or against top predators as biodiversity indicators are premature. Meanwhile, like any tools, top predators should be employed in conservation decision‐making with appropriate caution.