Effects of natural enemies on relative fitness of Heliothis virescens genotypes adapted and not adapted to resistant host plants

We investigated the potential of two natural enemies of Heliothis virescens (F.) (Noctuidae) to affect its rate of adaptation to tobacco containing Bacillus thuringiensis Berliner toxin. Larval fitness of two laboratory strains of H. virescens, one adapted to B. thuringiensis toxin and one not adapted, was compared on toxic and nontoxic plants, in the presence of the parasitoid Campoletis sonorensis (Cameron) (Ichneumonidae) or the entomopathogenic fungus Nomuraea rileyi (Farlow) Samson. By exposing larvae to plants and enemies for no more than 24 h, we focussed on the behavioral rather than physiological component of their interaction with toxic plants and natural enemies. Parasitism of H. virescens larvae by C. sonorensis during exposure periods of 1–4 h was lower on toxic plants than nontoxic plants and was lower for nonadapted larvae than for toxin-adapted larvae. Decreased larval feeding damage on toxic versus nontoxic plants, and by nonadapted versus adapted larvae, may explain differences in parasitism, because C. sonorensis locates host larvae using cues from damaged plants. Effects of plant toxicity and larval strain on H. virescens survival were numerically consistent with effects on parasitism, but they were not statistically significant. When mean larval survival is used to estimate fitness of the nonadapted genotype relative to the toxin-adapted genotype, we find that C. sonorensis is expected to delay adaptation to toxic plants. Percent infection by N. rileyi of H. virescens larvae exposed to fungus-treated plants for 24 h was greater when plants were toxic, and was greater for nonadapted larvae than toxin-adapted larvae. There were corresponding decreases in larval survival on toxic compared to nontoxic plants, and of nonadapted compared to adapted larvae. Interaction of effects of plant line and larval strain on survival was significant in the presence of fungus, but not in the absence of fungus, which indicates that the effect of toxic plants on the relative fitness of toxin-adapted and nonadapted larvae was mediated by fungus. As in the interaction with C. sonorensis, behavior of larvae on plants may explain differences in susceptibility to N. rileyi. Because nonadapted larvae moved more than toxin-adapted larvae on toxic plants, nonadapted larvae may have been more likely to encounter a lethal dose of conidia. In contrast with C. sonorensis, N. rileyi, which decreased the fitness of the nonadapted genotype relative to the adapted genotype, is expected to accelerate adaptation to toxic plants.