Answers are grounded in this paper and cite their sources.
Global biodiversity is rapidly declining under the influence of human‐induced environmental change, but our understanding of what is driving change at the local scale is often incomplete. Global datasets such as the Living Planet Database (LPD) provide invaluable site‐based time‐series data on wildlife populations but often lack detailed information on the local threats simpacting them. Here, we test whether global broad‐scale spatial threat data based on the probability of threat impact or threat occurrence data can predict the presence of threats recorded in the LPD for 1668 terrestrial bird and mammal populations. We test six generalised additive model structures, comparing the predictive performance of models using global threat impact probabilities and threat occurrence data to an intercept only and Red List species threats null model. Predictive performance was assessed using 100 repetitions of five‐fold spatial blocked cross‐validation. Across taxa and threat types, the overall ability of models to detect true threat presences was low. Models based on probabilities of threat impact or threat occurrence rarely showed significant gains in predictive performance over the null models. Our findings demonstrate that current broad‐scale threat datasets act as imperfect proxies for on‐the‐ground population impacts, highlighting the essential need for higher‐resolution environmental predictors and increased local reporting to strengthen threat‐biodiversity response analyses.
No takes yet. Share an insight, caveat, or question.
Oppenheimer et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: