Local environmental, geo-climatic and spatial factors interact to drive community distributions and diversity patterns of stream benthic algae, macroinvertebrates and fishes in a large basin, Northeast China

Understanding processes and mechanisms driving patterns of species distribution and diversity is a vital theme in community ecology and conservation biology. There has been a continuous increase in studies focusing on diversity patterns in freshwater ecosystems during the last few decades. However, comparative studies of species distributions, diversity patterns and underlying processes across trophic levels remain limited. The unique characteristics of the study area (i.e. Hun-Tai River in Northeast China) generates a wide range of environmental conditions to advance our understanding of what drives community assembly and diversity pattern of three stream organism groups. We applied variance partitioning (VPA) to correlate community compositions with environmental and spatial factors to infer stochastic and deterministic assembly processes, respectively. Linear regression (LR) models were used to identify the main drivers of species richness and local contributions to beta diversity (LCBD) as a function of different factors, including local environmental (e.g., in situ parameters, hydrology, nutrients), geo-climatic variables (e.g., land use, topography, climate) and spatial factors. Results indicated that species compositions of stream biota showed significant correlations with local environmental, geo-climatic and spatial factors. VPA demonstrated that both paradigms (i.e. deterministic and stochastic processes) interact to influence the biota distributions with stochastic process contributing more than deterministic process. In addition, the strongest stochastic process was found in fishes (10%), followed by benthic algae and macroinvertebrates with the same effects (8%). Notably, geo-climatic factors explained a substantial fraction of species composition, richness and beta diversity, although their effects were partially manifested via local and spatial variables. We demonstrate the relative importance of both stochastic and deterministic processes in shaping community composition and biodiversity of three stream organism groups in a large basin. This emphasizes the need to move beyond observed patterns and consider metacommunity theory into river management and conservation practices.