Abstract
Aim The underlying processes driving the origin and spatial variation of species richness stand as one of the most fundamental and enigmatic questions in ecology. Here, we assess the roles of water stress, energy availability, climatic seasonality, physiological tolerance to cold extremes, historical climate stability, and topographic heterogeneity in shaping spatial richness variation among various plants of the Atlantic Forest. Location Atlantic Forest Hotspot, South America. Taxon 14,042 species of vascular plants (angiosperms and gymnosperms), including trees, lianas, shrubs, subshrubs, terrestrial herbs, and epiphytic herbs. Methods We employed stacked ecological niche modelling to estimate species richness across different plant life forms. We then applied spatial models and hierarchical partitioning to assess the role and significance of each hypothesis in explaining species diversity variation. Results Spatial richness variation in the Atlantic Forest results from the interplay of multiple drivers, with effects differing among plant life forms. Overall, regions with higher energy availability, lower temperature seasonality, higher temperature stability but lower precipitation stability over the last 21,000 years, and more complex topography harboured more species. However, the effects of water stress and cold sensitivity contrast among life forms. Main Conclusions Observed species richness variation among life forms is primarily rooted in their life‐history strategies. Trees and lianas exhibited lower sensitivity to water stress but were less tolerant of cold extremes, whereas subshrubs and terrestrial herbs showed the opposite pattern, with lower tolerance to water stress but greater cold tolerance. Shrubs and epiphytic herbs displayed an intermediate pattern, combining the reduced sensitivity to water stress of trees and lianas with the higher cold tolerance of subshrubs and terrestrial herbs. Our study deepens the understanding of the factors shaping Atlantic Forest diversity and opens new pathways for predicting and managing large‐scale impacts of human activities.