Local environmental and spatial factors are associated with multiple facets of riverine fish β-diversity across spatial scales and seasons

Disentangling the factors governing community structure across various spatial and temporal scales is a core task in community ecology and biogeography, because it could provide valuable insights into biodiversity conservation and management. The relative role of environmental (a proxy for niche-based mechanism) and spatial (a proxy for dispersal) effects in shaping ecological communities depends on diversity facets, network positions, seasonality and biological characteristics of the focal organisms. However, such patterns are not well understood for fish communities in large river systems, especially if all of these patterns are examined simultaneously. We investigated fish taxonomic, functional and phylogenetic total beta-diversity (and their turnover and nestedness phenomena) at different spatial scales (basin vs. section scales) and seasons (wet vs. dry seasons) in a subtropical river system (the Chishui River basin, China). We also quantified the relative importance of environmental variables and spatial factors in shaping fish beta-diversity by using generalised dissimilarity modelling. At the basin scale, fish taxonomic and phylogenetic beta-diversity across seasons mainly reflected turnover, yet turnover and nestedness contributed likewise to total functional beta-diversity. Taxonomic and phylogenetic beta-diversity were strongly correlated (r from 0.39 to 0.87), whereas functional and both taxonomic and phylogenetic facets were relatively weakly correlated (r from 0.27 to 0.65). Generalised dissimilarity modelling suggested that phylogenetic beta-diversity phenomena were consistently more influenced by environmental variables (wet 16.1%, dry 22.4%) than by spatial factors (wet 6.7%, dry 4.8%). However, for both taxonomic and functional facets, environmental variables (turnover 13.8%, nestedness 6.4%) and spatial factors (turnover 4.3%, nestedness 9.4%) accounted for a greater proportion of variations in turnover and nestedness, respectively. At the section scale, only phylogenetic total beta-diversity and nestedness, as well as taxonomic nestedness significantly differed between the headwater and downstream sections. Furthermore, headwater communities were primarily explained by spatial (15.7%) and environmental (12.7%) variables in the wet season, while environmental factors (25.9%; spatial 7.7%) dominated in the dry season. By contrast, downstream communities were more influenced by environmental factors (17.9%; spatial 8.3%) in the wet season, whereas environmental (11.0%) and spatial variables (11.5%) contributed likewise to beta-diversity in the dry season. Our results highlight that, compared with taxonomic beta-diversity, accounting for turnover and nestedness of functional and phylogenetic beta-diversity helps to reveal distinct biodiversity patterns, and potential environmental and spatial controls. Given that community-environment relationships and the relative contribution of environmental and spatial factors varied between seasons and between the basin and section scales, our study underlines the importance of long-term dynamics and spatial scales in metacommunity studies. Our findings also suggest that maintaining environmental heterogeneity and adequate hydrological connectivity would be essential for the conservation of fish diversity. Overall, our study advances current knowledge of fish biodiversity patterns and their maintenance mechanisms by adding insights from multi-faceted diversity and seasonality, which has rarely been done in large subtropical rivers.