Population Genetic Variation in the New Zealand Greenshell Mussel Perna Canaliculus

Abstract

This thesis reports that the endemic New Zealand greenshell mussel, Perna canaliculus, constitutes a single panmictic unit on the basis of biochemical genetic variation observed at seven protein-encoding allozyme loci. Levels of gene flow are sufficiently high to counteract genetic drift along the entire habitat range of the species. Two previous allozyme studies reported that a north-south division of this mussel species occurred at ~ 38°S latitude and that mussel populations around the country were isolated by distance, respectively. However, the results of the present allozyme study, which involved extensive geographical sampling of the species entire distribution range, can best be explained by a model of panmixia, which is consistent with regional hydrography, and the history and biology of P. canaliculus. Possible explanations for the discrepancy between the 3 studies include choice of loci, differences in sampling strategy or genuine inter-annual change of allele frequencies due to environmental factors. For the first time, genetic variation at mitochondrial genes was used to assess the population genetic structuring of P. canaliculus. In contrast to the allozyme loci, three mitochondrial DNA (mtDNA) markers revealed significant population genetic heterogeneity among mussel populations, as might be expected from the nature of these markers. However, a clear and pronounced north-south split below Cook Strait at ~ 41°S latitude was identified between all North and northern South Island populations and the remainder of South Island populations and a Stewart Island population. The population genetic differentiation was caused by a significant frequency shift of the most common haplotype (A), the presence of regional haplotypes, (particularly, the occurrence of the unique southern haplotype L at high frequency), and significant differences in haplotypic and nucleotide diversities. This pattern of population genetic structuring is probably in the process of breaking down if the allozyme markers correctly reflect high levels of contemporary gene flow. On the other hand, the phylogeny of the haplotypes appears to be deeper than the population genetic structuring inferred from genetic distances among populations, and the north vs south pattern might therefore be emerging. New Zealand's greenshell mussel industry is based for 80% on a naturally occurring spat source, which is washed up on northern New Zealand beaches near Kaitaia. This so-called Kaitaia spat is collected and transferred to mussel farms around the country, a practice that might add to genetic homogenisation among populations. An attempt was made to evaluate the extent of possible "genetic contamination" of wild stock with genetic material of introduced Kaitaia spat caused by the potential spread of Kaitaia spat and replacement of local alleles. This thesis demonstrates that the mtDNA markers traced back all Kaitaia spat seeded populations to their northern origin, regardless of the location they were sampled from. The mtDNA markers revealed that populations originating from Kaitaia spat were significantly genetically differentiated from and significantly less genetically diverse than any other sampled mussels, suggesting that "genetic contamination" is non-existent. Preliminary research revealed that the gene order of five protein-coding genes of a segment of mtDNA of P. canaliculus was different from that of nineteen other molluscan and non-molluscan invertebrates. Interestingly, the gene order of this particular segment of mtDNA more closely resembled those of gastropods than that of another Mytilid, the blue mussel, Mytilus edulis. This finding highlights that although the use of mtDNA genome organisation as a phylogenetic tool may be warranted to reconstruct phylogenetic relationships among some taxa, it may not be an appropriate character to deduce systematic relationships within the Mytilidae.