Similar dietary but different numerical responses to nonnative tamarisk (Tamarix spp.) by two native warblers

Abstract

Native species can have a range of responses to nonnative introductions, from negative to positive, and understanding how and why native species respond differently to nonnatives remains an important management challenge. Based on differences and similarities in ecology and behavior, we predicted how abundance and diet of two native warblers, Lucy’s warbler (Oreothlypis luciae) and yellow warbler (Setophaga petechia), would differ in habitats with different amounts of nonnative tamarisk trees and the three nonnative insects obligately dependent on tamarisk (Tamarix spp.). Specifically, we predicted that Lucy’s warblers would have similar densities across sites, yellow warbler densities would be inversely related to tamarisk cover, and both warblers, being generalist insectivores, would incorporate tamarisk biocontrol insects in their diet. Based on point counts and fecal samples at six sites along the Virgin River in the southwestern United States, we found that yellow warblers decreased in abundance with increasing tamarisk cover, while Lucy’s warbler abundance did not and that diet of the two warblers did not differ, with both species exhibiting strong selection for the nonnative tamarisk weevil (Coniatus splendidulus) and weak to no selection for the nonnative tamarisk leafhopper (Opsius stactogalus). Both warblers showed negative selection for the tamarisk beetle (Diorhabda carinulata) and its larvae, even when those insects were 10–100 times more abundant during outbreaks. Although both warblers exploited the novel food resources offered by tamarisk, with those insects contributing half or more of total prey biomass, Lucy’s warblers were better able to maintain densities in tamarisk habitats. We hypothesize this was due to the Lucy’s warbler’s ability to exploit a broader array of habitats surrounding tamarisk sites and its cavity nesting habit that buffers its nests from the higher temperatures and lower humidity of tamarisk-dominated habitat. Our results suggest that predictions based on detailed knowledge of the form and function of native and nonnative species can be used to predict native bird response to nonnatives.