Aim  To examine the roles of local and regional environmental variables and biotic interactions in determining the structure of local stream fish assemblages, and to compare results derived from analyses based on taxonomic and functional groups. Location  Texas, USA. Methods  Species abundance data were compiled for 157 stream fish assemblages in several river basins across Texas. Species were condensed into functional groups based on trophic and life‐history characteristics. Local and regional environmental variables were either measured at each location or determined from scale maps and public‐access data bases. The original taxonomic and functional group data sets were analysed using similarity indices, null models of co‐occurrence, and direct and indirect ordination techniques. Results derived from taxonomic and functional group data sets are compared. Results  Inferences regarding the relative roles of local and larger‐scale factors in determining stream fish assemblage structure differ dramatically between analyses of taxonomic and functional groups. Taxonomic analyses suggest a prominent role of regional‐scale environmental factors, and local assemblages sorted according to a biogeographic pattern. Functional group analyses suggest almost equal roles of factors representative of local and larger scales, and assemblages were distinguished by a habitat template irrespective of geographic region. Main conclusions  The structure of local stream fish assemblages is determined ultimately by factors representing multiple scales, with the relative importance of each depending on the biological unit employed (species or functional groups). We suggest that analyses using functional groups can more directly infer ecological responses to environmental variation, and therefore may provide a more fruitful avenue for developing and testing ecological theory of community organization across biogeographic scales.

Intermittent headwater streams serve important functions for humans and wildlife in semi-arid rangelands. Increases in ashe juniper coverage in central Texas over recent decades are believed to have negatively impacted stream flows. Few studies have examined relationships between aquatic species and environmental factors in these systems as well as the influence of juniper coverage on assemblage structure. During summer 2003 and spring 2004, we examined species-environment relationships to infer potential effects of juniper cover on aquatic ecology relative to local-scale and watershed-scale environmental variables. Fish and crayfish species assemblages and physicochemical variables were investigated in spring-fed headwater tributaries of the Pedernales River, Texas. Fish abundance was much higher in summer 2003, whereas crayfish abundance was higher in spring 2004. Fish species richness was lower during spring 2004, possibly due to below average precipitation during spring 2004 that reduced deep-water refugia. Higher abundance of crayfish in spring 2004 samples was probably due to their ability to survive low-flow conditions, and a release from fish predation pressure. Fish assemblage structure was more strongly associated with local abiotic factors during spring 2004 when flow was reduced, whereas structure during summer 2003 samples suggested a relatively greater influence of predation. In general, juniper cover was weakly associated with fish and crayfish assemblages, although it tended to be positively associated with relatively high-quality habitat for sensitive taxa (flowing runs with coarse substrate; deep, connected pools). We suggest that intermediate levels of juniper cover in the region provide indirect benefits to aquatic organisms. However, short-term, local environmental factors appear to have a much greater influence than watershed vegetation on fish and crayfish assemblages in these intermittent streams.

No abstract