Males of the noctuid moths, Heliothis virescens and H. subflexa locate mates based on species‐specific responses to female‐emitted pheromones that are composed of distinct blends of volatile compounds. We conducted genetic crosses between these two species and used AFLP marker‐based mapping of backcross families (H. subflexa direction) to determine which of the 30 autosomes in these moths contained quantitative trait loci (QTL) controlling the proportion of specific chemical components in the pheromone blends. Presence/absence of single H. virescens chromosomes accounted for 7–34% of the phenotypic variation among backcross females in seven pheromone components. For a set of three similar 16‐carbon acetates, two H. virescens chromosomes interacted in determining their relative amounts within the pheromone gland and together accounted for 53% of the phenotypic variance. Our results are discussed relative to theories about population genetic processes and biochemical mechanisms involved in the evolution of new sexual communication systems.

Larval feeding was assayed in a generalist caterpillar (Heliothis virescens (F.)), a specialist caterpillar (Heliothis subflexa (Gn.)), their F1 hybrids and a backcross with H.s. to gain a preliminary understanding of the genetic basis of host use in H.v. Plants used in these experiments were tobacco (Nicotiana tabaccum), cotton (Gossypium hirsutum), soybean (Glycine max) (hosts of H.v.) and ground cherry (Physalis pubescens) (host of H.s.). A feeding study of H.v., H.s. and their reciprocal hybrids showed that, after the first eight days of feeding, H.v. had its highest survival and weight gain on soybean and had lowest survival and weight gain on its non‐host, Physalis. H.s. had its highest survival and weight gain on its host plant, Physalis, and performed very poorly on the three non‐host plants. The two F1 hybrids were SV0 (offspring of H.s. female and H.v. male) and VS0 (offspring of H.v. female and H.s. male). The hybrids did not differ from each other, indicating no sex linkage or maternal effects, except that VS0 had greater weight gain on tobacco than did SV0. The hybrids, unlike their parents, survived well on all four host plants and their weight gain was intermediate on all four host plants. In a separate experiment the VS0 hybrid was mated to the specialist to produce the VS1 backcross. In contrast to the F1, the backcross had significantly lower survival than the generalist on soybean, cotton, tobacco and weight gain was lower on soybean, cotton and tobacco but higher on Physalis. Survival and weight gain on cotton and tobacco were inherited as partially dominant traits; on Physalis there was overdominance for survival and complete dominance for weight gain; on soybean both survival and weight gain were additive. Survival on cotton, Physalis, soybean and tobacco and weight gain on Physalis could not be completely explained by a model that included only dominance and additive effects. These traits may be influenced by epistatic and/or environmental effects.

No abstract

Abstract.-Two species of moths and their hybrids were studied to determine the genetic basis of host range. One species, Heliothis virescens (HV), has a broad host range and is an agricultural pest on cotton (Malvaceae), soybean (Fabaceae), and tobacco (Solanaceae). The other species, Heliothis subjiexa (HS), has a narrow host range, feeding on plants in the genus Physalis (Solanaceae). Experiments were done to determine whether the abilities to feed on cotton, soybean, and tobacco were each under separate genetic control in HV (genetically uncorrelated) or whether feeding on all three hosts was genetically correlated. By repeatedly backcrossing hybrids to HS while selecting for high larval survival and weight on soybean, loci conferring the ability to feed on soybean were moved from HV into the genetic background of the specialist, HS. After six generations of selection on soybean we tested the selected line on soybean, cotton, Physalis, and tobacco to determine if ability to feed on soybean was correlated with feeding ability on these other hosts. We found no evidence to support the hypothesis that feeding on all three hosts was correlated and controlled by a single locus. Compared to HS, the survival and weight of the selected line were significantly higher on soybean, similar on tobacco and Physalis, and slightly, but not significantly, higher on cotton. We also conducted a behavioral choice test to determine if larval preference for soybean was correlated with preference for other hosts. The selected line larvae had higher preference for soybean and cotton than HS. Preference for tobacco was equal in the selected line and in HS. These results support the trends seen in survival and growth tests where performances on soybean and cotton appeared partially correlated and performances on soybean and tobacco were uncorrelated. In conclusion, the generalist, HV, did not have a single set of loci that governed feeding on soybean, cotton and tobacco. It is likely that some of the loci governing performance and preference for soybean also contributed to performance and preference for cotton and that the soybean loci examined did not interact negatively to lower performance and preference for Physalis, the specialist's host. This approach can be applied more widely to address evolutionary questions about host range and other ecological traits.Key words.-Diet breadth, host range, monophagous, performance, polyphagous, preference.Received July 11, 1994. Accepted March 21, 1995 In 1920, Brues observed that insects were instinctively polyphagous, oligophagous or monophagous and began to address many of the questions still central to the study of plant-herbivore interactions. Aside from intimating the beginnings of chemical ecology and coevolution, he initiated the study of host range evolution by examining some phylogenetic patterns in diet breadth and by asking whether host shifts were caused by major mutations or by accumulation of many small changes. Although Brues' work was conducted prior to the Modern Synthesis (Hu...