The evolution and maintenance of reproductive strategies in bag worm moths (Lepidoptera: Psychidae)

Abstract

Maintenance of sex is one of the central questions of evolutionary ecology. This thesis examines the evolution and maintenance of sexual and asexual reproduction in Dahlica and Siederia bag worm moths (Lepidoptera: Psychidae). Bag worm moths are taxonomically complex and ecologically poorly known species. Dahlica (Sexual and asexual) and Siederia (strictly sexual) are closely related, morphologically and ecologically very similar, and co-occur in same habitats. These species are particularly interesting due to their peculiar life cycle (long larval phase, short adulthood), female winglessness, and the occurrence of parthenogenesis, which is relatively rare among other Lepidoptera. A phylogenetic study with mtDNA markers questioned the current taxonomic classification of genera Dahlica and Siederia and supported multiple origins of parthenogenesis within this group of species. While co-occurrence of several bag worm moth species in the same habitats was common, the proportions of sexual and asexual reproduction varied from O to 100 % between populations. Parasitoid prevalence also showed considerable variation among different moth populations. The prevalence of sexual reproduction correlated positively with parasitoid prevalence across the study populations, suggesting that parasites select for sexual reproduction. An additional cost of sexual reproduction was indicated in two sexual bag worm moth species by a decreased number of fertilised eggs if males had previously copulated. In small, often female biased, populations this decrease may reduce the survival of sexual moths compared to co-existing asexual competitors. In a genetic analysis, populations of asexual Dahlica fennicella had lower genotype diversity but higher heterozygosity levels than their most common sexual competitors Siederia rupicolella and Dahlica charlottae. Clonal diversity was surprisingly high among asexual populations, probably explaining why asexuality has not led to an evolutionary dead end in D. fennicella. Among sexual D. charlottae populations, parasitoid prevalence was higher in the most inbred or genetically less varied populations supporting the importance of high genetic diversity of the host. Parasitoid prevalence in bag worm moth populations is also related to the abundance of ants in their habitats. In ant behaviour experiments, I found that wood ants (Formica rufa group) aggressively attack parasitoids while bag worm moth larvae remain unharmed. Bag worm moths were abundant in habitats where wood ants were abundant and, within habitats, in trees where wood ants were most numerous. These results suggest that wood ants indirectly protect bag worm moth larvae. While apparently ideal prey for wood ants, the larvae escape ant predation through low detectability while their most harmful enemies are readily attacked by the ants. Bag worm moths were not previously known to be ant-associated in any way. These studies offer new empirical evidence of the diversity of factors affecting the occurrence and competitive abilities of closely related sexually and asexually reproducing species.