Community composition exceeds area as a predictor of long-term conservation value

Abstract

Conserving biodiversity often requires deciding which sites to prioritise for protection. Predicting the impact of habitat loss is a major challenge, however, since impacts can be distant from the perturbation in both space and time. Here we study the long-term impacts of habitat loss in a mechanistic metacommunity model in terms of both immediate extinctions and secondary species losses. We find that biomass-at-site, closely related to site area, is a poor predictor of long-term regional species losses following site removal. Knowledge of the compositional distinctness (average between-site Bray-Curtis dissimilarity) of the removed community can markedly improve the prediction of impacts at the regional scale, even when biotic responses play out at substantial spatial or temporal distance from the removed site. Fitting our model to empirical species-by-site tables describing Andean diatoms and Brazilian lichen-fungi, we show that compositional distinctness surpasses area as a predictor of long-term species losses in the empirically relevant parameter range. Our results robustly demonstrate that site area alone is not sufficient to gauge conservation priorities; analysis of compositional distinctness permits improved prioritisation at low cost.