Bird predation and the host-plant shift by the goldenrod stem galler

Escape from natural enemies may favor the incorporation of a novel host plant into the diet of an herbivorous insect. This scenario has been suggested for the recent host-plant shift by the goldenrod stem galler, Eurosta solidaginis Fitch (Diptera: Tephritidae), from the ancestral host Solidago altissima L. (Compositae) to the derived host Solidago gigantea Ait. In this study, we examined the effects of predation from downy woodpeckers, Picoides pubescens L. (Aves: Picidae), and black-capped chickadees, Parus atricapillus L. (Aves: Paridae), on these two host races of insects at the western edge of their zone of sympatry. Based on a field census, bird predation was concentrated near the cover of trees where S. gigantea tends to occur; few attacks occurred in the open where S. altissima is prevalent. We conducted a field experiment to evaluate the preference of these avian predators for galls of the two host races when differences in the microgeographic distribution, size, and height of galls were controlled. In allopatric sites where only S. gigantea occurs, attacks by birds were 58% more frequent on S. gigantea than on S. altissima galls. Similar results were found for sympatric sites, although the difference in attack was only 26% and not significant. We could find no difference in the toughness of galls or the nutritional value of a larva within the gall (in terms of biomass) to explain avian preference for the S. gigantea host race; however, we found that from 1999 to 2000, the S. gigantea race offered a 27–107% higher reward rate (i.e., the probability that a gall harbored a larva of E. solidaginis) than the S. altissima race. Our studies suggest that avian predators can assess a gall’s content prior to pecking it open, preferring galls that are inhabited by both E. solidaginis larvae and the inquiline predator Mordellistena convicta Leconte (Coleoptera: Mordellidae). It is possible that birds have either learned through experience or evolved through natural selection to choose the more profitable S. gigantea galls. Finally, our results suggest that avian predators act against the maintenance of two distinct host races in the midwestern United States.