Linking regeneration niche to monodominance in biodiverse tropical forest landscapes.

Aims: Although monodominance has attracted the attention of ecologists for many decades, few studies have devoted attention to how abiotic factors could influence the occurrence of monodominant forest patches at the biome scale. Here, we assessed whether the occurrence of monodominant forest patches of Moquiniastrum polymorphum (Less.) G. Sancho (Asteraceae), an early‐successional tree species with wind‐dispersed seeds, could be predicted using optimum germination temperature and past deforestation. We also verified under what edaphic and climatic conditions the species could reach monodominance. Location: The Atlantic Forest, Brazil. Methods: We estimated optimum germination temperature across the species' geographic range as a function of annual mean temperature based on the results of germination tests available in the literature. Past deforestation (a proxy for suitable habitat for the species' dispersal and establishment) around monodominant forest patches was estimated by calculating the forest cover in 1985. We also modeled the upper limit of dominance (relative abundance) as a function of climatic and edaphic variables considered important for the species' establishment. Results: The results showed that the probability of occurrence of monodominant forest patches is statistically null in places where germination can take more than 10 days and the landscape had more than 20% forest cover. Relative density values for the monodominant condition (> 60%) occurred only in warmer regions with infertile soils and median precipitation conditions (ca. 1,075–1,700 mm per year) in the Atlantic Forest. Conclusion: We conclude that only under optimal conditions of germination and dispersal (i.e., regeneration niche) does monodominance occur. This highlights germination traits as an important mechanism for regulating monodominance. In addition, the approach used to predict regions with optimum germination temperature has further implications for understanding species abundance and distribution more generally. [ABSTRACT FROM AUTHOR]