Species interactions underpin the structure and function of food webs. Human alteration of ecosystems leads to changes in species composition and dominance, with cascading effects on the distribution and magnitude of interaction strengths among species, as well as flows of energy and materials in food webs. An increasing number of studies suggest that the patterns of energy flow and interaction strengths affect food web stability, and this information is critical as society is faced with difficult decisions regarding ecosystem management and provision of ecological goods and services. We quantified aquatic food web characteristics at six sites on the Colorado River in Grand Canyon. Sites differed in distance downstream of Glen Canyon Dam and turbidity as a result of tributary inputs of sediment and organic matter. Multi-year measurements of dietary composition and production of invertebrates and fishes allowed us to construct detailed quantitative food webs to examine spatial variation in (a) the degree of autochthony vs. allochthony among consumers and (b) the distribution and magnitudes of interaction strengths.
Results/Conclusions
Our results demonstrate large spatial variation in allochthony vs. autochtony that was strongly related to the longitudinal position of large tributaries. The tailwater food web was predominantly supported by autochthonous algae, while downstream food webs were variably supported by allochthonous carbon derived from tributary inputs. This pattern of increasing allochthony was paralleled by an increase in foodweb complexity and a broader spectrum of interaction strengths. Despite intersite differences in the distribution of interaction strengths, two links between fishes and invertebrates (i.e., consumption of blackflies and midges) were consistently strong and appear to represent ‘keystone’ linkages in the highly-regulated Colorado River. Management actions that alter the magnitude of these linkages may have large cascading effects on the food web. These results, coupled with previously-published responses of the tailwater food web to an experimental flood, suggest that empirically-based estimates of interaction strength may be an effective tool for predicting responses of river food webs to large-scale management actions.