Evolution of Parasitism in the Lycaenidae (Lepidoptera)

DSpace/Manakin Repository

Evolution of Parasitism in the Lycaenidae (Lepidoptera)

Citable link to this page

 

 
Title: Evolution of Parasitism in the Lycaenidae (Lepidoptera)
Author: Kaliszewska, Zofia
Citation: Kaliszewska, Zofia. 2015. Evolution of Parasitism in the Lycaenidae (Lepidoptera). Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences.
Full Text & Related Files:
Abstract: Of the four most diverse insect orders, the Lepidoptera contain remarkably few predatory and/or parasitic taxa, and while species with carnivorous life histories have evolved independently numerous times in moths and butterflies, this has rarely led to diversification. As a rule, aphytophagous taxa seem prone to extinction. In this dissertation, I explore the ecological and evolutionary consequences of entomophagy in the butterfly family Lycaenidae using several approaches: natural history observation, phylogenetics, population genetics and stable isotope chemistry.
A striking exception to the lack of radiation and persistence in aphytophagous lineages is the lycaenid subfamily Miletinae, which with 13 genera and 190 species is among the largest and most diverse groups of aphytophagous Lepidoptera. Most miletines eat Hemiptera, although some consume ant brood or are fed by trophallaxis from their host ant. I inferred the higher-level phylogeny of this group using data from one mitochondrial and six nuclear genes sampled from representatives of all genera and nearly half the described species. Biogeographic analyses indicate that Miletinae likely diverged from an African ancestor near the start of the Eocene, and four lineages dispersed between Africa and Asia. Phylogenetic constraint in prey selection is apparent at two levels: related miletine species are more likely to feed on related Hemiptera and are also more likely to associate with closely related ants species, either directly by eating the ants, or indirectly by eating hemipteran prey attended by those ants.
I then examined the influence of diet on the population structure of lycaenid butterflies, and more specifically, I investigated whether particular feeding habits are correlated with traits that might make species vulnerable to extinction. To do this, I compared the phylogeography and population genetics of two endemic lycaenid species of roughly similar age from southern Africa: Chrysoritis chrysaor, whose caterpillars are strictly herbivorous, and Thestor protumnus, whose cuckoo-like caterpillars survive by soliciting regurgitations from their host ants. I sampled both species from populations throughout their entire known ranges, and found that in contrast to C. chrysaor, T. protumnus has exceedingly small effective population sizes and individuals disperse poorly. With its aphytophagous life history, T. protumnus exhibits a high degree of host dependence and specialization. Although these results are correlative and based on only a single comparison, it seems likely that small population sizes and extreme ecological specialization make populations of T. protumnus more susceptible to disturbance and prone to extinction.
Having focused in detail on the population biology of just one species, I then analysed the evolution of Thestor as a whole. This genus is exceptional because all of its 27 described species are thought to be entomophagous, and all are thought to be predators or parasites of a single species of ant, Anoplolepis custodiens. Using representatives sampled from all known species and populations of Thestor as well as 15 outgroup species, I inferred the phylogeny of the genus in two ways: first by using characters from mitochondrial and nuclear genes, and second by analyzing genome-wide SNPs generated for each species using double digest RADseq. I also sequenced the ants associated with each of these taxa using ddRADseq. This investigtion showed that all 24 of the species in the Western Cape utilize Anoplolepis custodiens, while T. protumnusand T. dryburghi (the two species that are found in the north-western part of South Africa) use a closely related, but different species of Anoplolepis, and T. basutus (the species found in the eastern part of South Africa) utilizes yet a third species. Thus factors driving diversity in the genus Thestor may have initially involved ant associations and/or geographic isolation, but other forces are likely to be responsible for generating and maintaining the more recent diversity in the group. Flight time may have separated the “black” and “yellow” groups of Thestor: the black group fly predominantly in the summer months, while the yellow group fly predominantly in the spring. And while species spread across the genus fly in the spring and summer months, only members of the yellow group fly during the winter and fall months. Despite these broad scale differences, species in the genus Thestor show little evidence of niche partitioning, especially those in the Western Cape, and represent an extreme example of the coexistence of 24 species apparently utilizing a single food resource.
While working on the previous three projects, I was surprised by the number of species of South African Lycaenidae with incomplete life histories despite decades of work by avid lepidopterists in the region. For example, in the genus Thestor, although all 27 species are assumed to be aphytophagous, partial life histories have been described for only four species. In part the paucity of data is due to the difficult terrain occupied by these butterflies, and the fact that those whose caterpillars associate with ants often spend significant portions of their lives hidden in ant nests in crevices of rock that are intractable for excavation and observation. To deepen our understanding of South African lycaenid life histories, I used nitrogen and carbon stable isotopic methods to survey a large number of species and their potential food sources. With these methods, I confirmed some known or suspected life histories and showed that in any one area, a species can have a highly variable diet. I also discovered that some of the nitrogen stable isotope values are much higher than expected for land animals, implying longer than average food chains and/or extreme environmental conditions.
Together, these studies shed light on how carnivorous life histories affect the evolution of lycaenid butterflies, and help to explain why entomophagous lineages appear to be an evolutionary “dead end” in contrast to their herbivorous counterparts.
Terms of Use: This article is made available under the terms and conditions applicable to Other Posted Material, as set forth at http://nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA
Citable link to this page: http://nrs.harvard.edu/urn-3:HUL.InstRepos:23845507
Downloads of this work:

Show full Dublin Core record

This item appears in the following Collection(s)

 
 

Search DASH


Advanced Search
 
 

Submitters