Mate choice, predation and chemical defense in two species of alpine leaf-beetles

Evidence for the use of defensive compounds for sexual purpose is scarce. However sexual selection might have some importance in the evolution of defense-related traits. The present study reports a parallel analysis of defense-related traits and mate choice in two sister species of leafbeetles differing in their type of chemical defense. Oreina gloriosa produces autogenous cardenolides, whereas O. cacaliae sequesters pyrrolizidine alkaloids (PAs) from its food plant. We analysed order of mating, assortative mating and number of matings as measures of individual sexual success. The relationship between these measures and body weight and toxicity (i.e. concentration of defensive secretions) was investigated. We report mating pattern in relation to toxicity in O. gloriosa but not in O. cacaliae, which is most likely due to both male and female choice. These results support our hypothesis that sexual selection is expected to occur only in O. gloriosa in which toxicity is genetically determined, and not in O. cacaliae which defense depends mostly on host plant content in PAs. Reciprocal male and female choice of sexual partners may be responsible for the observed pattern. We also analyzed the mating pattern for these traits in field-sampled beetles. In both species, body weight and volume of the defensive secretion produced were important factors in mate selection, and in O. gloriosa age also played a role. The concentration of defensive secretion did not influence the mating pattern. However in this study our results did not support the hypothesis of a sexual selection for toxicity trait only in the species were toxicity is genetically determined. We discuss the ability and the need of these beetles to evaluate the defensive capacity of their mates. The harvestman Mitopus morio is a major predator of the leafbeetles Oreina gloriosa and O. cacaliae at the larval stage. We investigated both learning and local adaptation of M. morio towards these two preys, by performing choice experiments. We found that to some extent O. cacaliae was better defended than O. gloriosa. We propose that the pyrrolizidine alkaloids (PAs) contained in O. cacaliae are more deterrent to this predator than O. gloriosa's cardenolides. Deterrence against PAs did not occur at first, but after prior experience which we interpret as a dose-dependent response to avoid intoxication. Thus there would be some post-ingestion deterrence, caused by the toxic effect of PAs. Moreover, O. gloriosa did not gain protection from its gut content with toxic plant material. This suggests that O. gloriosa has specialised on its only food plant Peucedanum ostruthium for other reasons than gaining better protection against predators. We also tested the preference of M. morio for the PAs and cardenolide-defended O. elongata, vs. Cassida sanguinolenta which has a faecal shield defense. M. morio appeared to be a predator to both species. However it always largely preferred O. elongata despite the cardenolides and PAs contained by this prey. Since O. elongata larvae are very abundant and more vulnerable, C. sanguinolenta larvae may benefit from living in the same habitat as O. elongata. This example might illustrate that in some cases, counter-adaptations to chemical defense are more likely to evolve in predators than traits for by-pathing mechanical defense.

Thèse de doctorat : Université de Neuchâtel, 2000 ; 1568