Biogeographical patterns and compositional turnover of African vascular plants

Abstract Exploring biogeographical patterns of species richness and compositional turnover, and their responses to environmental factors is essential for predicting biodiversity responses to future global change. However, our understanding at continental scales remains limited, especially in Africa, due to data gaps and uneven sampling. Using a model-based distribution dataset of >23 000 vascular plant species in sub-Saharan Africa, we mapped species richness, estimated as species counts at site-level and explained using generalised additive modelling (GAM). Compositional turnover was mapped according to the nonmetric multidimensional scaling of between-site Jaccard dissimilarity and explained using multi-site generalised dissimilarity modelling (MS-GDM) of zeta diversity. We used deviance partitioning to identify potential underlying mechanisms of these biogeographical patterns. The results revealed species-rich pockets and compositional clusters of vascular plants across sub-Saharan Africa. The GAM fits the richness variation well (89.7% deviance explained) and identified mean annual precipitation (contributing 8.52% alone), mean annual temperature, fire frequency, human footprint, soil clay content, and topographic roughness as significant predictors. The MS-GDM explained 63% and 38.49% of the deviance in compositional turnover of narrow-range and widespread species, respectively. Geographical distance between sites contributed the most to the turnover of narrow-range species (9% deviance explained), whereas turnover of widespread species was affected by between-site differences in soil pH (7.55% deviance explained). Turnover of narrow-range and increasingly widespread species was driven mostly by environmental heterogeneity (14% and 10% deviance explained, respectively). These insights enhance our understanding of the processes shaping African plant biogeography and provide a foundation for predicting continental-scale biodiversity responses to future environmental change.