Mexican land birds reveal complexity in fine-scale patterns of endemism

Abstract

Aim: Understanding patterns of endemism is a key to deciphering the history of biotas and setting conservation priorities, but resolving the complexity of distributional patterns quantitatively into areas of endemism is often a difficult task. We report here an analysis of a comprehensive biodiversity dataset for the study of endemism, including virtually all vouchered records available for resident land birds of Mexico (> 100,000 georeferenced data points for all 780 species). Location: Mexico. Methods: The dataset was analysed with methods that recover areas without assuming prior endemic status for any species. This grid-based method for detecting areas of endemism considers co-occurrence and exclusiveness of species in alternative sets of geographic cells at different spatial resolutions, and finds optimal sets using heuristic, computationally intensive searches. Results: We provide the most detailed study of endemism in Mexico to date. Our analysis recovered 17 of 18 previously recognized areas of endemism for Mexican birds, plus many additional areas clearly supported by distributional data totalling 33 areas of endemism at different spatial scales. These areas cover 70% of the country's surface and form a network of nested and partially overlapping regions, some of which are also disjoint. Main conclusions: This picture contrasts strongly with previous conceptions of areas of endemism as non-overlapping and spatially simple in terms of scale. Our results reveal that endemism may be spatially complex and shed new light on its role as a key manifestation of biodiversity. Species identified as endemic to these areas comprise > 30% of the land birds of Mexico, with a disproportionately large fraction endangered according to IUCN or SEMARNAT.