Background and aims – Epiphyte bryophytes play a crucial role in forest ecosystems by contributing to water retention, nutrient cycling, and microhabitat creation. However, despite their ecological importance, epiphyte bryophytes in Madagascar remain poorly studied, especially in terms of their functional traits and responses to environmental change. This study addresses this gap by investigating epiphytic liverwort communities along an elevational gradient in northern Madagascar. We aim to assess the relationship between liverwort species' functional diversity and community assembly using a trait-based approach, focusing on how habitat filtering and niche differentiation shape community structure.

Material and methods – Epiphytic bryophytes were collected across 10 elevational belts along a 2230 m elevational transect. We measured 12 morphological traits related to resource use, life history, defense, desiccation resistance, and photosynthetic activity. Functional evenness (FEve) and functional dispersion (FDis) were calculated for each community. To investigate functional responses underlying community assembly, we measured community-weighted means (CWM) and variances (CWV) for each trait along the transect. We analyzed trait-environment relationships using multiple general linear models and evaluated community assembly processes (convergence or divergence) across the gradient.

Key results – Community assembly is influenced by environmental conditions and vegetation structure. At lower elevations, higher temperatures and taller vegetation lead to trait convergence, particularly in size-related traits. In contrast, at higher elevations, increased species richness increased trait divergence. Habitat filtering drives trait convergence at lower elevations, while niche differentiation becomes more important at species-rich higher elevations.

Conclusion – Morphological traits reveal how habitat filtering and niche differentiation jointly influence species distribution. Including physiological traits like water-holding capacity and carbon fixation in future studies will provide deeper insights into ecosystem processes. These findings contribute to understanding tropical forest ecosystems.