Aim: Uncertainty remains in revealing the determinants of plant diversity over regional and larger scales. Both plant diversity and the effects of environmental factors on it are scale dependent. This study aims to integrate and analyse the variations in the ecological mechanisms by which environmental factors drive plant diversity. Location: Montane forests, Northeast China Taxon: Tree Methods: In this study, we used field survey data from 3,078 sample plots in temperate forest ecosystems. First, we partitioned the species–area curve into different area scales with segmented models to explain changes in the dominant factors affecting tree richness. Then, we quantified the contribution of environmental factors to tree richness with gradient boosted model (GBM) and fitted their correlations at different spatial scales. Finally, we partitioned the contribution of climate, topography and both to tree richness with redundancy analysis (RDA) and analysed how the driving mechanisms varied with spatial scale. Results: Two area scales (small scale A ₁ and large scale A ₂) were segmented in regional temperate forests. The correlations between environmental contributions and spatial scale at the A ₂ scale were better than those at the A ₁ scale. The effects of environmental factors driving tree richness varied spatially nonlinearly. Tree richness at the A ₁ scale was determined by both climate and topography and climate alone, while it was determined by climate alone at the A ₂ scale. Main conclusion: Spatial heterogeneity affected the effects of environmental factors on tree diversity, especially at the A ₁ scale. Environmental factors exerted a dominant control at different spatial scales. Environmental filtering determined tree diversity at the A ₁ scale, and dispersal limitation determined tree diversity at the A ₂ scale for temperate forests. The spatial heterogeneity and collinearity of climate and topography prevented us from drawing consistent conclusions.