Beta diversity of stream fish assemblages: partitioning variation between spatial and environmental factors

Studies that analyse the structure of assemblages across relevant spatial scales can ascertain generalisable patterns and be used to guide efforts that allocate resources meant to conserve regional biodiversity. Beta diversity can shed light on the underlying factors that drive variation in assemblage structure including spatial and environmental influences. The purpose of this study was to address two questions: (1) Which is more important for the structuring of fish assemblages, spatial or environmental factors? (2) What is the dominant pattern underlying species turnover between fish assemblages, species addition (i.e. nestedness) or species replacement (i.e. spatial turnover)? We examined fish beta diversity in southeastern Oklahoma by sampling 65 wadeable stream reaches and measuring 30 environmental factors at each sampling location across the Muddy Boggy River drainage. Variation in fish assemblage structure was partitioned between environmental and spatial predictors using partial redundancy analysis. Overall species turnover was calculated and separated into its two additive components of spatial turnover and nestedness to determine which of these two accounted for the most turnover across the drainage. Spatial and environmental factors combined accounted for 25.5% of fish beta diversity. Environmental factors alone accounted for 20.1%, while spatial factors alone only accounted for 3.5% of the variation among assemblages. Nine environmental factors were significantly related to fish beta diversity: (1) elevation; (2) stream order; (3) stream width; (4) % riffle habitat; (5) water temperature; (6) conductivity; (7) turbidity; (8) % gravel substrate and (9) current velocity. Overall species turnover was driven mostly by spatial turnover rather than nestedness. This pattern was found to be the same across multiple spatial scales (entire drainage, sub-drainages, mainstems) and despite several techniques used to extract turnover coefficients. These results suggest that fish assemblages—particularly in headwater streams—are structured by environmental filtering and that these assemblages tend to be compositionally distinct, rather than being nested derivatives of one another.