Abstract
The tight coupling of host and parasite makes parasitism a powerful model for studying ecology and evolution. The null hypotheses of most evolutionary studies of parasites presume the dominance of host factors in mediating parasite evolution, both at the macro- and microevolutionary scales. At the population level, this is reflected in the expectation that the genetic structure of parasite populations should mirror that of their hosts, due to parasites’ reliance on their hosts for dispersal. Additionally, the low effective population sizes, short generation times, and high mutation rates typical of parasites should lead to their populations becoming more differentiated than those of their hosts. In cases in which this expectation is met, the higher levels of parasite population differentiation may provide more detailed information for studying host phylogeography, functioning as a sort of “biological tag”. Unfortunately, most studies comparing the population genetics of hosts and parasites have been restricted to a subset of parasitic taxa, and therefore any conclusions drawn from them are likely not representative of parasites in general.
This thesis set out to test these assumptions in two under-studied host-parasite systems on New Zealand’s South Island: hairworms (Nematomorpha) and their various orthopteran hosts, and the mermithid nematode Thaumamemis zealandica and its sandhopper host, Bellorchestia quoyana.
Identification by scanning electron microscopy of free-living hairworms collected from five mountainous locations in Central Otago revealed five species, including one new species. Sequencing of the mitochondrial gene cytochrome oxidase subunit 1 (CO1) and nuclear rDNA regions suggests cryptic species in two taxa as well as a lack of population structure for the most commonly encountered species. A comparison with previously published data of known and likely hosts suggests the absence of host-parasite co-structure, indicating higher levels of parasite gene flow, potentially mediated by passive dispersal of hairworm eggs or larvae, intermediate host dispersal, and/or lack of definitive host specificity.
Bellorchestia quoyana specimens were collected from thirteen locations along the south-eastern coast and assessed for infection with T. zealandica. The mermithid was restricted to locations south of Long Beach, Dunedin. Sequencing of sandhopper CO1 reveals the existence of regional metapopulations, with breaks occurring at the Otago and Banks Peninsulas. Sequencing of two fragments of CO1, NADH dehydrogenase subunit 4 (NADH4), cytochrome b (CYTB), and the nuclear rDNA internal transcribed spacer region (ITS) of the parasite revealed absolutely no intraspecific variation. This indicates that there may be no parasite population structure, although it is more likely that the sequence invariance is a result of extremely slow rates of nucleotide substitution, which may be characteristic of the family Mermithidae.
The results of this thesis suggest that host traits, while important, are not the only factor influencing parasite microevolution. This in agreement with an emerging, more nuanced framework for conducting host-parasite population genetic studies. By investigating the population genetics of host and parasite, this thesis also provides the first genetic data for most of the study species, new information on their geographical ranges, additional ecological data, and novel primers which may prove useful for future studies.