Southern pine beetle (SPB), Dendroctonus frontalis (Zimmerman), is one of the most destructive insect pests of pine trees in southern United States, Mexico and Central America. There is relatively little information on the effect of habitat fragmentation on the connectivity and the spatial population genetics of SPB. This study therefore, adds to previously generated information by assessing how habitat fragmentation affects the spatial population genetic structure of SPB. It also introduces a new approach to the study of bark beetle population dynamics by assessing how landscape variables shape their effective dispersal. To address this issue, a suite of eight highly polymorphic DNA microsatellite markers were used to measure SPB movement over a representative range of SPB habitat and non-habitat (matrix). At the broadest scale, highly significant genetic differentiation suggests that the sampled populations are not panmictic. Loci with higher variability yielded higher resolution for both the infinite allele model based measure of differentiation (FST) and the stepwise mutation based measure (RST) estimates. It is apparent that allelic frequency differences, allelic size ranges and repeat motif played a role in the observed patterns of pair-wise differentiations between the sampled localities. It is supposed that gene flow, wide-range dispersal and recent divergent time could have contributed to the lower level of genetic structure observed in the pair wise estimates. The sampled populations did not show any differentiation attributable to the host species from which they were collected. Mantel test of genetic distance and Euclidean geographic distance revealed no correlation. Mantel tests of the correlation between genetic distance and cost weighted Euclidean distances also suggest that dispersal of SPB across geographic barriers is not significantly reduced. Thus, landscape features and host preference do not appear to have had an impact on population genetic structure of SPB. Since movements of these beetles were not significantly hindered by environmental factors like major rivers, roads, elevation and host type, it is advisable for regional pest management offices to put an effort and coordinate their prevention and management plans in a broader scale to alleviate the problem associated with this native insect pest. Disequilibrium in the observed homogenized pattern of the beetle in this study suggests human contribution in the dispersal of SPB. Therefore, stringent control is deemed necessary in transportation of logs. This could improve our pest management system hence its positive implication in timber industry is obvious.

This PhD thesis studies the evolution of the peculiar mitochondrial DNA (mtDNA) in the mollusc Melarhaphe neritoides. We measured mtDNA diversity and elucidated the evolutionary forces that shape the evolution of the organelle.The mtDNA in M. neritoides harbours a remarkable amount of polymorphism at selectively neutral nucleotide sites (π_syn = 6.8 %), called hyperdiversity when above the threshold of 5 %. We revealed that an elevated mutation rate (µ = 5.8 × 10-5 per site per year at the COI locus), which is 1000 fold higher than in other metazoans, is likely the primary force generating mtDNA hyperdiversity. Such mtDNA hyperdiversity may be more common across other phyla and more frequently linked to high µ values, than currently appreciated.Natural selection is a second force, which shapes mtDNA hyperdiversity. Positive selection influences the overall mtDNA polymorphism in the 16S, COI and Cytb genes, including synonymous sites at which mtDNA hyperdiversity is calculated. Therefore, synonymous sites in M. neritoides are not neutral but possibly positively selected. Strong purifying selection maintains a low non-synonymous polymorphism in the 13 protein-coding genes of the mitogenome, so that a very few changes in nucleotide sequence induce changes in amino acid sequence. The effective population size of this planktonic-dispersing species is surprisingly small in the North East Atlantic (Ne = 1303), likely biased by selection, and for this reason, Ne is a poor indicator of mtDNA hyperdiversity.Migration is a third force, which homogenises the gene pool of the species through high rates of gene flow, predominantly eastward, and results in high connectivity and panmixia over the entire North East Atlantic.Genetic drift, the fourth force, is not sufficient in M. neritoides to lower mtDNA diversity, and populations show no differentiation.This thesis also highlights an important pitfall. The use of hyperdiverse markers may easily lead to erroneous interpretations of differentiation statistics and connectivity pattern, due to the lack of shared haplotypes in datasets induced by a high µ. First, D_EST may reach a maximal value of 1 but is not indicative of differentiation in terms of fixation (D_EST = 1 ≠ φ_ST = 1), and only reflects differentiation in terms of lack of shared haplotypes. Second, the signal of gene flow is concealed in haplotype network bush-like pattern.Rapid evolution of mtDNA results in significant selection pressure for co-adaptation of the nuclear genome encoding mitochondrial proteins. The elevated µ underlying mtDNA hyperdiversity provides an interesting framework for better understanding how mutational dynamics and selection that drive mitonuclear coevolution contribute to speciation. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished