Background/Question/Methods   Riparian forests serve as important allochthonous energy sources for stream food webs, especially along small headwater streams in the eastern United States. Due to relatively high levels of shading and organic matter input, aquatic communities in headwater streams are predicted to be particularly dependent on riparian inputs. In the context of agricultural watersheds, riparian forests are often fragmented and isolated, and their importance as sources of energy for aquatic food webs is uncertain. We studied an agricultural watershed in north-central Ohio to investigate the role of riparian forests as allochthonous energy sources for aquatic food webs. We sampled macroinvertebrates and fish to assess whether there were differences in the composition and structure of aquatic animal communities among headwater streams with forested and non-forested riparian areas. Forested sites were dominated by black cherry and green ash, while non-forested sites were dominated by herbaceous species such as white clover and orchardgrass. We also used stable isotopes to determine the degree to which aquatic food webs were supported by resources derived from terrestrial or aquatic systems.

Results/Conclusions   In both forested and non-forested sites the rank order of abundance for macroinvertebrate functional feeding groups was the same with gatherers being the most abundant followed by filterers, predators, grazers, and finally shredders. The mean total fish density and fish species richness were higher in non-forested streams relative to forested streams. We found that d13C values were generally the lowest (most depleted in 13C) for algae, followed by riparian vegetation (trees and herbs) and CPOM, and finally epilithon and FPOM were the most enriched in 13C. Our analyses of aquatic consumer data revealed that one of the fish species (creek chub) out of eight possible consumer groups had lower d13C values (P<0.05) for forested relative to non-forested reaches; however, the effect size for creek chub was small relative to variation in d13C of basal resources precluding us from making inferences about shifts in energy sources. These results suggest that in this watershed the incorporation of terrestrial energy inputs into aquatic food webs may be largely independent of whether a stream is forested or not.