FOOD-WEB INFLUENCES ON FISH POPULATION RE SPONSES TO INSTREAM FLOW

Fish population responses to changes in streamflow have seldom been co rrectly forecasted from physical habitat predictions. Yet, practical i nstream flow assessments rely on analyses of physical habitat dynamics , which are presumed to influence fish populations. This leads to the instream flow dilemma : how many species and life stages to analyze an d how to weight the importance of each. In some streams, PHABSIM (Phys ical Habitat Simulation) analyses for common warmwater fishes indicate that weighted usable area is either insensitive or maximized at low f lows. However, average to high flows are undoubtedly important for a v ariety of reasons, such as recruitment of floodplain tree species, str eambed scouring, organic matter inputs, and benthic invertebrate produ ction. Density, diversity, and production of aquatic insects in a vari ety of streams show considerable annual and spatial variation, often d irectly attributable to streamflow. Similarly, growth and foraging suc cess of fishes varies due to prey abundance. A food web model was deve loped to investigate the influence of the food base (aquatic insects, small prey fishes, and crayfish) on production of key predatory fishes (smallmouth bass Micropterus dolomieu, rock bass Ambloplites rupestri s, and flathead catfish Pylodictis olivaris) in a large warmwater stre am. Analysis of the model suggested that the production and yield of t he fishes were strongly dependent on the bottom-up influence of prey, primarily aquatic insects and crayfish. From model analyses and review of studies on insect production and foraging success of fish, hypothe size that decreases in flow during the growing season may decrease ins ect production and, thereby, indirectly depress sport fish production. Practical instream flow assessments and research should address (1) b roader array of taxa, especially suspected sensitive species, (2) appr opriateness of meso-scale habitat analyses, (3) indirect effects of fl ow on the trophic structure, (4) rigorous testing and refinement of ha bitat suitability criteria, and (5) efficacy of individual-based model s for evaluating alternative flow regimes.