Does degradation affect diet composition of stream predators?

Background/Question/Methods

In the western United States, anthropogenic habitat degradation has severely impacted lotic ecosystems. This degradation has many sources, including beaver dam removal, irrigation, and livestock grazing, yet the impact of such disturbances on aquatic communities is not well understood. In a disturbed western Colorado stream, Kimball Creek, larval Arizona tiger salamanders (Ambystoma tigrinum nebulosum) are the top predators in beaver ponds across a range of degradation intensity. Although A. t. nebulosum larvae have been shown to induce trophic cascades in lentic ecosystems, little is known about the strength of their top-down control in either degraded or lotic systems. Assessing larval salamander diet across the degradation gradient present at Kimball Creek is the first step in understanding how degradation may alter trophic interactions. We predicted that larval diets would shift from more degradation-intolerant species in the upstream reaches to more tolerant species in the more degraded, downstream reaches. Second-year larval salamanders were collected from ponds across the degradation gradient, dietary samples were collected using gastric lavage, and samples were enumerated in the laboratory. Prey diversity, size and abundance were compared across ponds using MANOVA.

Results/Conclusions

Ephemeropterans (mayflies) were the primary prey type across all levels of degradation, in terms of both proportional biomass and abundance. Although the biomass of ephemeropterans did not change across the range of degradation, proportional abundance did, suggesting that differences in prey size across the gradient offset abundance differences. Surprisingly, as degradation increased a smaller proportion of the diet (both biomass and abundance) was made up of degradation-tolerant chironomids, refuting our hypothesis. However, consumption of another tolerant taxon, corixids (water boatmen), increased in both biomass and abundance with degradation. None of the other prey showed a consistent change in proportional biomass or abundance across the degradation gradient. For these prey items, biomass and abundance of prey in the two upper sites was more similar than comparisons of these sites with the most degraded, downstream site, suggesting that areas of intermediate levels of disturbance in the stream have communities more similar to undisturbed areas. An important next step to understanding the community dynamics in this system is to quantify prey availability from benthic samples. This will provide information on how salamander feeding behavior, and potentially trophic interactions, change with degradation.