Net primary productivity but not its remote‐sensing proxies predict mammal diversity in Andean‐Amazonian rainforests

Tropical forests are disappearing, but we have a limited understanding of the factors driving species coexistence in mammal communities of old-growth forest ecosystems. The total energy that is bound by plants is assumed to be a key factor determining mammalian species richness, but accurately measuring energy flows in complex ecosystems is difficult, and most studies therefore rely on remote-sensing-based surrogates of net primary productivity (NPP). We monitored mammal species richness across three seasons using camera traps on 26 study plots along a forested, elevational gradient from 245 to 3588 m above sea level in southeastern Peru for which a unique dataset on field-measured NPP exists. Using linear-regression models and path analysis, we disentangled the effects of climate and NPP on the diversity of mammals, testing the predictions of the more-individuals hypothesis, stating that energy availability drives the number of individuals and, thus, the number of coexisting species. We compared detailed field measurements of NPP with remote-sensing products (MODIS NPP and MODIS NDVI). Mammal species richness, abundance, and biomass decreased in a negative exponential pattern with elevation. Field-measured data on NPP, which was largely driven by temperature, was a strong predictor of both abundance and species richness, while remotely sensed proxies for NPP failed to accurately predict mammal diversity. Our study underpins the importance of field-based ecosystem data and emphasizes the role of high primary productivity for maintaining diverse mammal communities, which is a particularly pressing issue in light of recent anthropogenic impacts on the Amazonian forest system.