SIGNIFICANCE OF METABOLIC LOAD IN THE EVOLUTION OF HOST SPECIFICITY OF MANDUCA-SEXTA

The existence of elevated metabolic loads associated with the detoxification of plant allelochemicals has been proposed to be an important selective force in the evolution of dietary specialization of herbivorous insects. In this study we have examined the effects of one host plant toxin (nicotine) and three nonhost plant toxins (xanthotoxin, precocene II, and canavanine) on the growth and energy metabolism of the tobacco hornworm, Manduca sexta (Lepidoptera: Sphingidae). Although M. sexta has three mechanisms for reducing the toxic effects of nicotine, nicotine has a dose-dependent growth-inhibiting effect on third-instar larvae at dietary concentrations ranging from 0.25 to 8.0%. Thus, herbivory by adapted species, as well as nonadapted species, can potentially select for higher levels of defensive chemicals in plants. At higher concentrations, where the effects of nicotine are most pronounced, growth reduction is associated with reduced consumption (which reduces the amount of food assimilated) and an increase in the duration of the instar (which increases the amount of the assimilate pool that must be allocated to maintenance metabolism), thus leaving less assimilate for allocation to growth. The respiration rates of larvae on diets containing up to 1% nicotine are the same as the respiration rates of larvae on nicotine-free diets, while the respiration rates of larvae on diets containing 4 to 8% nicotine are lower than those of larvae on the nicotine-free diet. Thus, the processing of nicotine does not impose a significant metabolic cost on M. sexta larvae, and it is invalid to infer increased metabolic costs of detoxification from the observation of concomitant decreases in growth rate (GR or RGR) and the efficiency of conversion of digested food (ECD) of larvae on toxin-containing diets. In a manner similar to nicotine, the growth-reducing effects of the nonhost plant toxins (xanthotoxin, precocene II, and canavanine) can be explained on the basis of their effects on consumption and the duration of the third instar. In no case is growth reduction a consequence of an increase in metabolic rate. Instead, reductions in growth result from the diversion of assimilate from growth to energy metabolism mandated by a longer instar, and from decreases in the size of the assimilate pool resulting from decreases in consumption and/or assimilation. Thus, the effects of allelochemicals on the energy budget of M. sexta result from their activity as metabolic toxins and feeding deterrents reducing rates of growth and consumption, and not from their diversion of energy from growth to allelochemical processing. We conclude that (1) the processing of host plant allelochemicals does not impose a significant energy demand on Manduca sexta and (2) the energy costs of allelochemical processing are unlikely either to constrain the expansion or to drive the contraction of the host plant range of M. sexta. Therefore, we suggest that the concept of metabolic load is not a useful one in understanding the effects of plant allelochemicals on the growth and efficiency of food utilization or in explaining the evolution of dietary specialization of lepidopteran herbivores.