Population genetics along a continuum of butterfly population structures

It is now commonly accepted that we are currently experiencing the sixth biodiversity crisis, the result of rapid and human – induced environmental change. The unrestrained loss of species due to habitat loss, deterioration and fragmentation, is also accompanied by a more subtle but no less important process, the loss of genetic diversity. Genetic diversity is important for at least two biological reasons: 1) the fundamental theorem of natural selection tells us that the rate of evolutionary change in a population is proportional to the amount of genetic diversity available, and 2) heterozygosity, or high genetic variation within populations is positively related to higher fitness. Many organism naturally from sub-populations, partly due to the patchy distribution of their habitat. When there is population subdivision, there is almost inevitably some genetic differentiation, i.e. allele frequencies that differ among the sub-populations. Genetic variation between populations is mostly determined by a combination of gene flow, natural selection and genetic drift. Whereas gene flow reduces the genetic differentiation between populations, a decrease in genetic diversity within populations (due to genetic drift) is often accompanied by an increase in the genetic differentiation between populations. Natural selection also results in a reduction in local genetic variation, but may as well enhance as weaken genetic differentiation between population. Natural populations display a variety of different population structures, from “closed” population structures (well-defined colonies within discrete areas where local birth and death rates are the main factors influencing local population size) to “open” population structures (individuals range widely and regularly migrate from one breeding area to another). In this thesis, I study the genetic population structure of several different Lepidoptera species along such a continuum. “Aglais urticae”, “Melanargia galathae”, “Proclossiana eunomia” and “Boloria aquilonaris” were the selected species. The principal objective of this thesis is the analyses of the effects of dispersal capacity and other ecological characteristics, such as the degree of specialisation of a species, on the genetic population structure. I find that species-specific traits lead to dissimilar genetic population structures. Indeed, population differentiation tends to increase with decreasing ecological constraints and increasing mobility. Isolation by distance is detected at very small local scales in less mobile species, while it is absent in the most mobile species at a very large scale, i.e. Palaeartic. Although ecological variation is common in this species, the wide geographic distribution of both common and rare haplotypes, absence of geographic structure and high genetic diversity suggest that high gene flow is the primary factor shaping the population structure of his very mobile species. The comparison of gene flow in two different landscapes reveals that the matrix matters. Indeed, gene flow within the linearly structured landscape is twice as important as the gene flow within the non-linearly structured landscape. The study of dispersal using both indirect (gene flow) and direct (capture-mark-recapture) data in an endangered species, shows a high congruence between these two methods. These results are particularly relevant to conservation biologists. The use of a “cost-friendly” genetic marker, easily applicable to numerous species and necessitating little DNA, correctly quantified (the number of dispersal events) and qualified (the type of dispersal events, i.e. stepping stone) gene flow in this endangered butterfly species. The reduction in population size, due to habitat loss and deterioration, leads to an increased probability of reproduction by related individuals, a process known as inbreeding. Inbreeding was detected in a vulnerable species using genetic data. The high amount of inbreeding came as a surprise considering the relatively large population sizes of the studied species. Finally, I discuss the choice of genetic markers, as these influence the genetic population structure that is revealed. I conclude that RAPDs appear to be an efficient option for the genetic population studies for fragmented and endangered Lepidoptera species by avoiding excessive investment in the development of particular genetic markers yet producing sufficient polymorphism.Doctorat en sciences (biologie) -- UCL, 200