Demographic and trophic dynamics of fishes in relation to hydrologic variation in channel and floodplain habitats of the Brazos River, Texas.

Large rivers in North America have been subjected to a variety of hydrologic alterations that have negatively impacted aquatic fauna. These impacts have triggered restoration efforts, including management of flows, to restore or maintain ecological integrity. Ecological data relevant to flow management and habitat restoration is scarce, and conceptual models of ecosystem function have been widely applied to large rivers despite a lack of quantitative evaluation of these models. Here, I examine demographic and trophic dynamics of fishes with divergent life histories and trophic guilds in relation to habitat heterogeneity and flow variability in a relatively unaltered floodplain system: the Brazos River, Texas. Reproductive activity of fishes with three divergent life history strategies was positively associated with long-term river hydrology, although species with alternate strategies exploited different portions of the hydrograph (peak flow versus increasing flow). Despite the positive association with hydrology, low-flow periods were favorable for recruitment, and food resources for larvae and juveniles were denser during these periods. Some species used both oxbow and channel habitats during some point in their life cycle, whereas other species appeared to be almost entirely restricted to one habitat type. Terrestrial C3 macrophytes accounted for a significant fraction of the biomass of most consumer species examined. Small-bodied species in oxbow lakes assimilated large fractions of biomass from benthic algae, whereas this pattern was not observed in the river channel. Frequent flow variations in the river channel may reduce algal standing stocks, and significant contributions from autochthonous algal sources may only occur during low-flow periods. Trophic positions of detritivores indicated that terrestrial carbon sources were assimilated, for the most part, via invertebrates rather than by direct consumption. My results indicate that current conceptual models are too vague to provide accurate predictions for restoration strategies in rivers with variable flow regimes. Flow and habitat management strategies that focus on reproducing key features of historical fluvial dynamics are likely to be more successful than strategies that focus on single indicator species or flow dynamics that differ from the historical hydrograph.