Land use changes rank among the highest threats to biodiversity, but assessment of their ecological impact is impaired by data paucity in vast regions of the world. For birds, land use changes may mean habitat loss or fragmentation, changes in resource availability and disruption of biotic interactions or dispersal pathways. As a result, avian population sizes and assemblage diversity decline in areas subjected to urbanization, agricultural intensification and land abandonment worldwide. This threat is especially sensitive in hotspots such as the Mediterranean basin, where avifaunas of several biogeographic origins meet, encompassing numerous endemic taxa and ecological specialists with low resilience to habitat modifications. Here, we correlated several facets of bird taxonomic and functional diversity to a fine-grained land-use change classification, in order to identify priority areas in need for enforced protocoled bird sampling in a conservation prospect. For this, we computed the species richness, functional richness, originality and specificity of 211 bird assemblages based on bird extent-of-occurrence data for 279 species and 10 ecological traits. We used a spatialized regression model to correlate bird diversity patterns with bioclimatic gradients and land use change between 1992 and 2018, assessed from an unsupervised clustering on 2km resolution data. We showed that species-rich bird assemblages are subjected to agricultural intensification, while functionally diverse assemblages are mainly undergoing desertification and land abandonment. Unfortunately, most of these changes occur in areas where protocoled bird surveys with sufficient spatial and temporal resolution are lacking. In light of these results, we urge for the setting of bird monitoring programs targeted mainly on parts of North-Africa and the Levant, in order to allow a region-level evaluation of the threat posed by recent land use changes on the exceptional avifaunistic diversity of the basin. Fostering such regional-scale evaluations of congruences between human threats and centers of diversity is a necessary preliminary step for a pragmatic response to data deficiencies and ultimately setting appropriate responses to avoid the collapse of avian assemblages.

A ReadMe file containing all necessary information on data and processing have been uploaded along with the datasets.