A primary goal of biodiversity research is to uncover the processes acting in space and time to create the global distribution of species richness. However, we currently lack an understanding of how recent versus ancient biodiversity dynamics shape patterns of diversity for most groups. Here, we introduce a method to partition lineage turnover into recent and ancestral components, and use it to identify hotspots of turnover at the global scale for 8,296 bird species. Counter to the tropical niche conservatism hypothesis, we find extra-tropical regions such as Greenland and the Sahara are hotspots of ancestral turnover, while areas with high climatic variation such as (sub)tropical mountains and biome transitions are recent turnover hotspots. We can now quantify the relative contribution of contemporary and ancient lineage dynamics to assemblage structure, which enables future research to explore the processes generating earth’s diversity in a more temporally-explicit framework.