Community isolation drives lower fish biomass and species richness, but higher functional evenness, in a river metacommunity

Abstract

The flow of individuals among communities and their interactions with local environmental filters are increasingly recognised as determinants of biodiversity patterns in riverine ecosystems. Both incoming dispersers and local conditions are expected to systematically change along connectivity gradients from headwaters to downstream communities. However, the interplay between isolation-centrality gradients and environmental conditions as determinants of biodiversity structure and function has seldom been considered. Here, we represented the dendritic structure of the Negro River basin riverscape (Uruguay) in a directed graph quantifying the isolation-centrality of each river section and evaluated the direct and indirect pathways by which riverscape structure and environmental local drivers determine fish community assembly. Fish communities (n = 58) were sampled following a stratified sampling design that properly represents this isolation-centrality connectivity gradient through the riverscape. In each community, fish abundance, biomass, richness, and functional diversity were estimated, and the direct and indirect hypothesised connections among them were evaluated with structural equation models. We showed that the range of isolation among river sections determines a 2-fold, 5-fold, and 25-fold variation in total fish richness, abundance, and biomass, respectively. Additionally, isolation-centrality was positively associated with local temperature and conductivity, while negatively related to local depth. These variables and taxonomic richness accounted for most of the variation in total fish biomass (81%) herein used as measurement of ecosystem function. Local fish abundance was negatively and positively associated with functional evenness and taxonomic richness, respectively. Furthermore, once the effect of isolation on biomass and richness was accounted for, an effect of diversity on biomass became evident. Our results provide empirical evidence for the role of riverscape structure on taxonomic and functional diversity, biomass, and the relationship between biodiversity and ecosystem function. We emphasise that in the understanding of river biodiversity and its management, local determinants should not be considered without attention to metacommunity processes.

The flow of individuals among communities and their interactions with local environmental filters are increasingly recognised as determinants of biodiversity patterns in riverine ecosystems. Both incoming dispersers and local conditions are expected to systematically change along connectivity gradients from headwaters to downstream communities. However, the interplay between isolation-centrality gradients and environmental conditions as determinants of biodiversity structure and function has seldom been considered. Here, we represented the dendritic structure of the Negro River basin riverscape (Uruguay) in a directed graph quantifying the isolation-centrality of each river section and evaluated the direct and indirect pathways by which riverscape structure and environmental local drivers determine fish community assembly. Fish communities (n = 58) were sampled following a stratified sampling design that properly represents this isolation-centrality connectivity gradient through the riverscape. In each community, fish abundance, biomass, richness, and functional diversity were estimated, and the direct and indirect hypothesised connections among them were evaluated with structural equation models. We showed that the range of isolation among river sections determines a 2-fold, 5-fold, and 25-fold variation in total fish richness, abundance, and biomass, respectively. Additionally, isolation-centrality was positively associated with local temperature and conductivity, while negatively related to local depth. These variables and taxonomic richness accounted for most of the variation in total fish biomass (81%) herein used as measurement of ecosystem function. Local fish abundance was negatively and positively associated with functional evenness and taxonomic richness, respectively. Furthermore, once the effect of isolation on biomass and richness was accounted for, an effect of diversity on biomass became evident. Our results provide empirical evidence for the role of riverscape structure on taxonomic and functional diversity, biomass, and the relationship between biodiversity and ecosystem function. We emphasise that in the understanding of river biodiversity and its management, local determinants should not be considered without attention to metacommunity processes.