Biodiversity and land-use change: understanding the complex responses of an endemic-rich bird assemblage

Aim: Anthropogenic land-use change is a major driver of the current extinction crisis, but the processes through which it acts on biodiversity are complex and still poorly understood. Here, we use several biodiversity metrics to make a comprehensive assessment of the response of an endemic-rich bird assemblage to land-use change. Location: São Tomé Island (São Tomé and Príncipe, Africa). Methods: We sampled bird assemblages in 220 point counts, stratified across three regions and four land-use types. Species richness was examined using accumulation curves and generalized linear mixed-effect models. Compositional and structural changes were analysed with ordination techniques. We used correlations and model selection to identify species traits and environmental variables associated with such changes. Results: At the point level, there was a slight decrease of species richness in more intensive land uses. However, higher dissimilarities within these land uses led them to have higher overall species richness. Endemics dominated forest environments and, although most preferred less-disturbed land uses, many persisted across all land uses. Non-endemics were virtually absent from forests, but became very abundant in the more intensive land uses. Canopy cover was the best predictor of the shift between endemic and non-endemic species, and allowed identifying regional differences in the local responses of bird assemblages to land-use change. Main conclusions: To better understand biodiversity, it is crucial to use several metrics simultaneously. In São Tomé, simply using species richness, the single most used biodiversity metric, could have been misleading as it concealed major shifts in the bird assemblage structure towards an endemism impoverished avifauna. Despite scarce data on land-use cover, we demonstrate how landscape context can be vital to understand biodiversity patterns and that highly forested surroundings might overestimate the strict value of more intensive land uses. Our results raise concerns about the long-term persistence of endemic species restricted to islands where forested land uses are being lost.