Tachinid Parasitoids Affect Host Plant Choice by Caterpillars to Increase Caterpillar Survival

Abstract

Current models of parasite–host interactions generally emphasize considerations of parasite virulence and parasite transmission rather than host responses to attack. We describe a situation in which parasitism causes a host to change its feeding behavior and this change improves the expected fitness of the host and probably the fitness of its parasites. We have found that a tachinid parasitoid (Thelaira americana) often emerges from its caterpillar host (Platyprepia virginalis) without killing the host. Whether the host caterpillar survives its parasites depends in part on what it has been eating. Unparasitized caterpillars were more likely to survive to adulthood when feeding on lupine, whereas parasitized caterpillars were more likely to survive on poison hemlock. Development time and pupal masses of caterpillars (both parasitized and unparasitized) were not found to be affected by the host plants that they fed on. Survival of fly larvae in caterpillars that we determined were parasitized using ultrasound was not affected by host plant. However, fly pupae that emerged from caterpillars that had been reared on hemlock were heavier than those emerging from lupine-fed caterpillars. This was due primarily to the direct effect of diet on the flies and less so to the indirect benefit to flies whose host caterpillars survived their parasites by feeding on hemlock. Parasitized caterpillars were more likely to select hemlock, and unparasitized caterpillars were more likely to select lupine when offered both host plants in field tests. These results were consistent for the two years that the choice experiments were conducted. These results were also consistent with the hypothesis that caterpillars change their food plant choices so as to increase their conditional success. Conditional food choices that increase success depending upon parasite load are well accepted for humans, controversial for other primates, and unknown for insects. If caterpillars alter their host plant choices as a result of their parasite load, then this phenomenon could help to explain the evolution of host plant choices that have defied explanation in the past. Such a suggestion assumes that nonlethal parasitism is a common phenomenon. We believe that this may be the case since we did not detect the nonlethal nature of this interaction until we began the unconventional practice of rearing in the field. Other workers have described nonlethal parasitism for several tachinid–host systems, and many families of flies are similar to tachinids except that their hosts are vertebrates; these interactions are nonlethal. Future empirical work, as well as models of parasite–host interactions, should consider the possibility that hosts alter their plant choices depending upon their parasite loads.