Biogeographic biases in research and their consequences for linking amphibian declines to pollution

The collapse of amphibian populations within pristine reserves worldwide suggests that diffuse, globally distributed factors such as pollution may be a cause of these declines. Nevertheless, cause-effect relationships between pollution and declines have proven difficult to establish at all scales, from local to global. We therefore aimed to quantitatively evaluate the weight of evidence for the role of pollution in global amphibian declines by first quantifying the published research on the effects of pollutants for all amphibian species in the world and then cross-referencing this information with species' biogeographic distribution, range area, and conservation status and with threats to species as summarized in the Global Amphibian Assessment. We found strong biogeographic and related taxonomic research biases, with a few, common, widely distributed generalist species from the northern hemisphere accounting for the majority of studies. Tropical regions, where more species and declines occur, were severely underrepresented in ecotoxicological research; therefore, current knowledge does not permit assessment of the significance of pollution in amphibian declines at a global scale or in regions where most declines occur. Moreover, if broader distributional ranges and occurrence at higher latitudes are correlated with broader environmental tolerances, current knowledge may considerably underestimate the sensitivity of the majority of amphibian species to pollutants. Finally, because species studied represent only a narrow subset of traits that mediate exposure and susceptibility to pollution, the current potential for extrapolation among species, guilds, ontogenetic phases, levels of complexity, habitat types, sites, and regions is weak. Ultimately, to mitigate these research biases and thus improve the basis for risk management, research policy must balance the spatial scale of the problem (global) and the distribution of knowledge (regional) so that biodiversity representation can be increased and regional bioindicators of environmental health can be developed.