AbstractsBiology & Animal Science

Gene flow in an arctic wetland: modelling landscape effects on fine-scale genetic variation in an isolated muskrat «Ondatra zibethicus» population

by Xavier Giroux-Bougard




Institution: McGill University
Department: Department of Natural Resource Sciences
Degree: MS
Year: 2015
Keywords: Biology - Ecology
Record ID: 2057964
Full text PDF: http://digitool.library.mcgill.ca/thesisfile130625.pdf


Abstract

This thesis evaluates the genetic diversity, the population genetic structure as well as the functional connectivity of an isolated muskrat (Ondatra zibethicus) population at the northernmost limit of its geographic distribution in North America. In the first chapter, nine microsatellite markers were used to assess the genetic characteristics of muskrats sampled from 21 different lakes spread across the Old Crow Flats (OCF), an arctic wetland complex in Northern Yukon. Genetic diversity was relatively low, while patterns of genetic variation were structured into two population clusters within the OCF, suggesting reduced levels of gene flow in arctic habitats. In chapter 2, the relationship between genetic differentiation and the connectivity of wetland features in the OCF was investigated using circuit theory to model isolation by resistance (IBR) based on landscape cover. Resistance surfaces were parameterized using a machine learning algorithm to optimize the fit between an IBR model and the observed genetic distance between lakes. The optimized IBR model was subsequently compared to the alternative models of isolation by distance (IBD) and isolation by barrier (IBB) using two complementary approaches: (1) causal modelling, and (2) partitioning of genetic variation explained by spatial eigenfunctions (Moran's eigenvector maps). The optimized IBR model revealed that hydrological features facilitated gene flow in the landscape. Gene flow through most terrestrial habitats was relatively unimpeded, though woodlands were severely obstructive. Causal modelling showed greater support for the optimized IBR model then for IBD or IBB models. Partitioning of variation suggested that both IBR and IBB models contributed unique explanatory spatial structures, while the IBD model was redundant. This thesis highlights the first muskrat population genetic structures detected at fine spatial scales and identifies the landscape variables that drive this spatial pattern using new modelling approaches. Cette thèse évalue la diversité génétique, la structure génétique ainsi the la connectivité fonctionnelle au sein d'une population isolée de rats musqués (Ondatra zibethicus) située à la limite septentrionale de leur distribution géographique en Amérique du Nord. Dans le premier chapitre, neufs marqueurs microsatellites ont été utilisés pour caractériser la variation génétique de cette population en échantillonnant 21 lacs éparpillés dans le paysage d'Old Crow Flats (OCF), un complexe de terres humides arctiques situé dans le nord du Yukon. Les résultats démontrent une diversité génétique relativement basse, en plus de révéler une structure génétique suggérant deux groupes distincts au sein de la population d'OCF. Le deuxième chapitre évalue la relation entre la différentiation génétique et la connectivité des caractéristiques hydrologiques d'OCF en utilisant la théorie des circuits pour produire un modèle d'isolement par résistance (IPR) depuis la couverture et l'hydrologie du paysage. À cet effet, les surfaces de résistances ont…