Diet composition of two larval headwater stream salamanders and spatial distribution of prey

Abstract

Summary Interspecific overlap in resource use can determine the degree to which different presumed guild members have distinct or similar effects on ecosystem processes or responses to environmental change. Many headwater streams of North America support multiple species of larval salamanders commonly defined as a single guild that can influence macroinvertebrate communities and nutrient dynamics and are sensitive to stream alteration. We explored macroinvertebrate distributions in conjunction with stable isotope and gut content analyses of salamanders to examine similarities in diet between two sympatric larval salamander species (Desmognathus quadramaculatus and Eurycea cirrigera) in four headwater streams. We determined the degree to which larval salamanders used similar prey functional feeding groups (FFGs) and taxa and determined the primary source habitat (pools versus riffles) of prey. Stable isotopes of carbon (δ13C) and nitrogen (δ15N) suggest the two salamander species occupied similar trophic positions and individual‐based stable isotope mixing models indicated similar use of macroinvertebrate predators and filterers by both species. Diet analyses were generally consistent with stable isotope results in identifying prey FFGs that composed the largest biomass of salamander diets. However, despite similarities in diet at the resolution of FFGs, there was little overlap in the specific taxa consumed by the two salamander species: 52 prey taxa were consumed over all samples, with only 16 taxa in common. Further, only five prey taxa were common in dominating diet biomass of both species; there was more overlap in taxa in terms of diet abundance. We assessed patterns in benthic macroinvertebrate biomass and compared the biomass of taxa that were in pools versus riffles to the biomass and abundance of taxa in salamander diets. Total macroinvertebrate biomass was generally higher in pools; however, the majority of salamander prey biomass was from riffle habitats, a trend that was stronger for D. quadramaculatus than for E. cirrigera. There was greater similarity in taxa comprising diet by abundance with the majority of prey items originating from pools. The larval salamander species used similar prey FFGs but differed significantly in specific prey taxa. We hypothesise that species differences in diets were most likely a function of differences in larval size and microhabitat use. Consequently, the treatment of larval salamanders as a guild is probably inadequate for predicting the effect of larval salamander diversity on some stream processes, and species may differ in how they respond to factors affecting prey assemblages.