Scale and scope matter when explaining varying patterns of community diversity in riverine metacommunities

Large-scale species and genetic metacommunity patterns are influenced by variation in environmental factors and distance between communities, according to previous studies. However, these studies often used different measures to assess patterns of metacommunity diversity, distances between communities and grain sizes at which environmental variables arc measured. This hinders interpretations and generalizations of the underlying process that drive metacommunity diversity. We applied a synthetic and multi -analytical approach to identify general factors structuring the diversity of a large riverine metacommunity. Using complementing approaches we analyzed how distance, measured as Euclidean or topological distance, and environmental factors, assessed at different grain sizes, influenced different measures of metacommunity diversity (species richness, functional richness and phylogenetic diversity) of mayfly, stonefly and caddisfly species in a large river network (river Rhine, Switzerland). We found the amount of explained variation in species diversity was generally unaffected by grain size, but improved with the use of topological distance, compared to Euclidean distance. Variation in functional diversity was best explained by environmental factors at small grain sizes and topological distance. Variation in phylogenetic diversity was best explained when environmental variables were assessed at larger grain sizes and Euclidean distance was used. Overall, our results indicate that processes structuring metacommunity diversity may differ at the species, functional or phylogenetic level of the community, as recently postulated in the metacommunity phylogenetics approach. While such differences may hinder comparisons across studies using different methodologies, it offers opportunities to disentangle the structuring factors within metacommunities by applying multiple analytical approaches to the same dataset.