Effects of fish and crayfish on ecosystem structure and function during stream drying

Central stonerollers (Campostoma anomalum) and Meek's crayfish (Orconectes meeki meeki) are important benthic consumers that can have strong effects on ecosystem structure and function in Arkansas streams. Summer drying in these systems greatly reduces flow and habitat volume and might alter interactions between these consumers and stream benthic communities. Consumer effects on algal and invertebrate abundance and rates of primary production and detrital decomposition were examined in a series of field and greenhouse mesocosm experiments. Electrical exclusions were used to manipulate consumer densities in Little Mulberry Creek, AR during early and late drying. The relationship between consumer density and the size of consumer effects on benthic periphyton and invertebrate abundance was examined with multi-trophic dynamic models and a mesocosm experiment (crayfish density). In a second mesocosm experiment (mesocosm disturbance) I determined whether flow-related disturbance altered the effects of stonerollers and crayfish on benthic ecosystem structure and function. Through these experiments, I tested the following hypotheses; (1) benthic consumers will reduce primary production and periphyton, sediment, and invertebrate abundance on exposed tiles compared with exclusions, but increase leaf litter decomposition, (2) disturbance will alter the strength of these consumer effects, (3) the size of consumer effects will be related to biotic and abiotic factors, and (4) effects of stonerollers and O. meeki will differ. In field exclusions, consumers strongly reduced algae, sediment, and invertebrate abundance in 2006, but effects varied among pools and intensified during stream drying. Stoneroller density was positively related to the size of consumer effects on periphyton, but only in the absence of fish predators. In 2007 tile sediment and invertebrate densities did not differ among treatments, and chlorophyll a increased in exposed treatments in August, though not in June. Exclusion did not affect net primary production or biomass specific primary production, but significantly reduced leafpack decomposition rates. Consumer density was positively related to the size of consumer effects on periphyton and invertebrate abundance, and low crayfish densities were related to increased algae on exposed tiles. Differences in consumer effects between years may be related to greater flow-related disturbance in 2007. In the crayfish density study, model results indicated that crayfish effects would benefit algal abundance only for low levels of omnivory. In the mesocosms, crayfish effects on lower trophic levels were similar across the density gradient as strong crayfish omnivory reduced the abundance of both periphyton and chironomids even in low crayfish density treatments. In the mesocosm disturbance experiment, disturbance and interactions between disturbance and consumer identity did not significantly affect any response variable. Leafpack decomposition, periphyton, and invertebrate abundance were influenced by consumer identity, and effects differed between stonerollers and crayfish and with time. These experiments emphasize that predicting when consumer effects will influence ecosystem processes is a formidable challenge. In both mesocosms and natural pools fish and crayfish had strong but variable effects on benthic structure and function and effects were influenced by disturbance, biotic and abiotic environmental factors, consumer identity, and experimental venue. Future studies should explicitly consider the influence of spatial and temporal variability on ecological processes when assessing the effects of organisms on ecosystem function.