Aims: The objective of this paper was to test the hypothesis that epiphytic lichen metacommunities in the Brazilian Atlantic Forest are hierarchically structured and not only correspond to forest formations and their floristic subdivisions, but also reflect the main geopolitical regions in the study area (Northeast, Southeast, South) and particularly the division of the Atlantic Forest, along a latitudinal gradient, into a tropical and a subtropicaltemperate zone north and south of the Tropic of Capricorn. Study area: Atlantic Forest in Brazil. Methods: We employed niche distribution modeling for 245 species of four large and representative families of epiphytic lichens in the study area, using a stacked species approach to model metacommunity structure via quantitative methods of community analysis (clustering and ordination techniques). To account for sampling bias, we performed background manipulation via a bias layer. Results: We detected a hierarchical structure with three main clusters (A, B, C) subdivided into ten subclusters. Both the main clusters and each of the subclusters correlated strongly with environmental variables and latitude and featured statistically significant, characteristic indicator species, and the subclusters matched floristic subdivisions of the main forest formations recognized in the study area. There was a significant correlation of the three main clusters with the three geopolitical regions in the study area which after all also reflect ecogeographical features. The main separation between the two northern (A, B) and the southern cluster (C) almost precisely followed the division of the Atlantic Forest by the Tropic of Capricorn, supports the notion that the latter is an important delineator in biome formation, reflected by epiphytic lichen metacommunities. Conclusion: We conclude that modeling metacommunities via a stacked species approach is a powerful tool, also to elaborate baseline data for future monitoring of metacommunities in relation to climate change. [ABSTRACT FROM AUTHOR]