Evolutionary divergence of field and laboratory populations ofManduca sextain response to host-plant quality

Abstract

1. The tobacco hornworm, Manduca sexta, has been an important model system in insect biology for more than 50 years. In nature, M. sexta successfully utilises a range of host plants that vary in quality. The consequences of laboratory domestication and rearing on artificial diet for fitness of phytophagous insects on natural host plants have not been explored. 2. We examine the evolutionary divergence of two domesticated laboratory populations and a field population (separated for more than 40 years, or > 250 laboratory generations) of M. sexta with respect to performance and fitness on two natural host plants: a typical host plant, tobacco (Nicotiana tabacum) and a novel host plant, devil's claw (Proboscidea louisianica). 3. For both field and laboratory populations, rearing on devil's claw resulted in animals with lower survival, smaller final size, longer development time, and reduced size-corrected fecundity than animals reared on tobacco. Reductions in some fitness components (survival and fecundity) were greater for the laboratory population animals than the field population animals. 4. When reared on tobacco, the laboratory population animals had similar or larger pupal masses and slightly shorter development times than when reared on artificial diet, suggesting that laboratory domestication on artificial diet has not greatly affected the ability of M. sexta to perform well on a typical natural host plant. 5. Although field and laboratory populations exhibited qualitatively similar responses to host-plant quality, i.e. reduced performance on devil‘s claw, the magnitude of this reduction differed across populations, with the domesticated laboratory populations having greater reductions in performance than the field population. The use of domesticated populations as models for responses of field populations may therefore be more appropriate for considering environmental conditions that are relatively benign or near-optimal, than when exploring responses to extreme or stressful conditions.