Organic matter flow in stream food webs with reduced detrital resource base

Food webs based on flows of organic matter were developed for two small str eams to examine food web response to a large reduction in detrital inputs. At the study site, Coweeta Hydrologic Laboratory in the southern Appalachia ns, leaf litter inputs and associated microbial assemblages are the main en ergy source for food webs in headwater streams. We eliminated leaf litter i nputs to one stream using a net placed over the first 180 m of stream from its origin. Food webs based on Row of organic matter were developed for a r eference stream and the litter-excluded stream for two months. July and Dec ember of year 1 of the litter exclusion, to examine effects of leaf litter exclusion on the trophic base of the food web, size distribution of flows, predator-prey interactions, and trophic structure. Flows (mg AFDM.m(-2).d(- 1); AFDM = ash-free dry mass) were estimated using gut content analyses for detritus and prey items, coupled with secondary production estimates. We u sed a whole-stream C-13 tracer method to estimate assimilation of bacteria by invertebrates. The food webs encompassed most (84-91%) of invertebrate s econdary production. but <30% of the estimated total links. The primary sou rces of organic matter for the food web in the reference stream were leaf t issue, bacterial carbon, and animal prey, with <similar to>25-30% of total secondary production derived from each. In-stream primary production led to <1% of invertebrate secondary production. A higher fraction of food web pr oduction in the litter-excluded stream was derived from wood. Magnitudes of detrital flows were lower in the litter-excluded stream, and some taxa wer e missing compared to the reference stream. The fraction of predator ingest ion approached 100% of total secondary production for both streams, but thi s predation was distributed diffusely among several tare. Flows to predator s were fewer and smaller in the litter-excluded stream, yet these flows had higher per-biomass consumption coefficients, suggesting stronger interacti ons among the remaining common taxa. These food webs enabled us to examine interactions among taxa in the streams; hence, we found responses of the st ream ecosystem to litter-exclusion that we would not have considered had we only measured changes in invertebrate population sizes or system-level cha nges in organic matter flow.