Understanding the impacts of both natural and anthropogenic landscape features on genetic diversity, population structure and connectivity has important implications for conservation of species living in fragmented environments. Here, I combine population genetic data, detailed land cover information, and computer simulations to explore how landscape shapes genetic structure across two regional populations of the threatened eastern massasauga rattlesnake (Sistrurus catenatus catenatus) in Ontario, Canada: one along the eastern shores of Georgian Bay and the other largely confined to the northern half of the Bruce Peninsula. First I used spatial Bayesian assignment to quantify the genetic population structure within each regional population. I found marked subpopulation structure within eastern Georgian Bay with differentiation of island and mainland snakes, a north-south split within the mainland coinciding with the town of Parry Sound, and evidence of further subdivision within the cluster of snakes north of Parry Sound. In contrast I found no population subdivision within the mainland of the Bruce Peninsula, but genetic distinction of mainland and island snakes. Next, I identified the landscape features that shape spatial genetic structure within regional populations. In eastern Georgian Bay I found local variation in the effect of landscape on populations. North of Parry Sound I found no effect of landscape on inter-individual genetic differentiation, but a strong pattern of isolation-by-distance. In contrast I found that both open water and roads restrict gene flow of snakes south of Parry Sound. I found no evidence of isolation-by-distance or that landscape shape genetic structure within the Bruce Peninsula. Finally I used individual-based, spatially explicit simulations to identify the lag-time associated with the detection of contemporary landscape feature effects on genetic structure of massasaugas, and explore the consequences of using spatially correlated land cover elements in landscape genetic analyses. I found that the genetic consequences of roads could be detected within 2-12 generations when population sizes were small or juvenile dispersal was low. However, I also found that roads could be spuriously identified as impediments to gene flow when spatially correlated features such as water are included in genetic models. / Thesis (Master, Biology) -- Queen's University, 2011-10-14 15:06:35.956

Assessments of landscape connectivity are increasingly required in natural resource management. Understanding how landscape structure affects the movement and dispersal of animals may be essential for ensuring the long-term persistence of species of conservation concern. Functional connectivity models describing how features on the landscape influence animal movement behaviour have been produced in two different ways. The resistance surface models landscape connectivity as its inverse, the resistance to movement and dispersal, while the landscape graph represents landscape connectivity by describing the relationships among resource patches. Both methods have limitations that make them less effective for modelling highly-mobile and wide-ranging species such as ungulates and carnivores. This thesis develops a method called grains of connectivity that combines the continuous representation of landscape connectivity provided by resistance surfaces and the scalability provided by landscape graphs to create a flexible modelling framework for these species. The first half of the thesis reviews the conceptual origins of the grains of connectivity method and examines its properties using simulated landscapes. In the second half, empirical evidence of movement and dispersal in a boreal woodland caribou (Rangifer tarandus caribou) population is used to validate functional connectivity hypotheses generated using the method. Connectivity for caribou at the temporal scale of generations is examined using a landscape genetics approach, while connectivity at the seasonal scale is assessed using the distribution of caribou telemetry locations. Grains of connectivity may be most useful for study systems where animals are not found exclusively in well-defined resource patches and there is uncertainty in the behavioural parameters influencing movement and dispersal. Additionally, the scalability of the analysis can be used to selectively remove spatial heterogeneity that may be uncorrelated with movement and dispersal giving an improved description of the pattern affecting the landscape connectivity process.

Assessments of landscape connectivity are increasingly required in natural resource management. Understanding how landscape structure affects the movement and dispersal of animals may be essential for ensuring the long-term persistence of species of conservation concern. Functional connectivity models describing how features on the landscape influence animal movement behaviour have been produced in two different ways. The resistance surface models landscape connectivity as its inverse, the resistance to movement and dispersal, while the landscape graph represents landscape connectivity by describing the relationships among resource patches. Both methods have limitations that make them less effective for modelling highly-mobile and wide-ranging species such as ungulates and carnivores. This thesis develops a method called grains of connectivity that combines the continuous representation of landscape connectivity provided by resistance surfaces and the scalability provided by landscape graphs to create a flexible modelling framework for these species. The first half of the thesis reviews the conceptual origins of the grains of connectivity method and examines its properties using simulated landscapes. In the second half, empirical evidence of movement and dispersal in a boreal woodland caribou (Rangifer tarandus caribou) population is used to validate functional connectivity hypotheses generated using the method. Connectivity for caribou at the temporal scale of generations is examined using a landscape genetics approach, while connectivity at the seasonal scale is assessed using the distribution of caribou telemetry locations. Grains of connectivity may be most useful for study systems where animals are not found exclusively in well-defined resource patches and there is uncertainty in the behavioural parameters influencing movement and dispersal. Additionally, the scalability of the analysis can be used to selectively remove spatial heterogeneity that may be uncorrelated with movement and dispersal giving an improved description of the pattern affecting the landscape connectivity process.

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第21602号 / 理博第4509号 / 新制||理||1647(附属図書館) / 京都大学大学院理学研究科生物科学専攻 / (主査)教授 曽田 貞滋, 准教授 渡辺 勝敏, 教授 中務 真人 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

Understanding the factors that make non-native species successful invaders is an important step towards mitigating spread. At the same time, species invasions can serve as natural experiments to test range-limit theory. Range-limit theory postulates declines in local abundance (abundant center model) and genetic diversity (central-peripheral hypothesis) towards range edges because of underlying environmental gradients. Such declines constrain adaptation to marginal habitats via gene swamping. However, broader evolutionary theory predicts intermediate rates of immigration into range-edge populations can relieve genetic drift and improve adaptive potential. I tested hypotheses generated from theory while illuminating aspects affecting of the invasion of the Asian tiger mosquito (Aedes albopictus Skuse) into the US. Using reciprocal distribution modeling, I found US populations occupied significantly different climate and habitat than in their native range (SE Asia). Most inconsistencies were found in the northern US range, where Ae. albopictus has recently crept northward, providing an opportunity to test range-limit theory as the range reaches its limit. Because of its limited natural dispersal ability, rapid spread after the 1985 US introduction pointed to human-aided dispersal. I tested the current role of human-aided versus natural dispersal using a landscape genetics framework, and found that natural dispersal dominated current patterns. Some distant localities were highly genetically similar, indicating potential human-aided transport in limited cases. Asymmetric gene flow from core to edge localities supported the abundant center model, but uniformly high genetic diversity contrasted with the central-marginal hypothesis. I detected a significant signature of local adaptation by overwintering diapause-induced eggs in multiple field sites using reciprocal transplants. Surprisingly, most genotypes from throughout the range produced large offspring when overwintered at the range edge. Relative offspring mass between home and away winters peaked at an intermediate immigration rate. These results show that rapid adaptation has occurred in US populations of Ae. albopictus and highlight the potential for further spread. Genetic admixture from multiple introductions may explain high genetic diversity throughout the US range and contribute to high offspring size for all genotypes overwintered at the range edge. Finally, my work highlights the need for a better understanding of contemporary ecological and evolutionary processes leading to range-limits (or expansion) to more accurately reflect processes occurring in a human-dominated world.

Doctor of Philosophy / Department of Biology / Brett K. Sandercock / Samantha Wisely / Anthropogenic activities and climate change have dramatically altered landscapes worldwide. The ability of species to cope and adapt to ongoing changes is likely a function of their behavior, movements, and sensitivity to fragmentation. Greater Prairie-Chickens (GPC) are a lek mating grouse native to the Great Plains Landscape Conservation Cooperative (GPLCC), for which inbreeding depression and anthropogenic avoidance are a concern. The goals of my dissertation were to: 1) identify genetic correlates of male performance which may influence population viability under current land use practices, 2) identify GPC habitat characteristics and delineate areas of critical GPC habitat necessary for GPC conservation, and 3) identify the relative importance of distance and habitat quality for maintaining genetic connectivity among spatially structured populations. First, I found male reproductive success and survival to be positively associated with genetic diversity. Using multistate modeling in Program Mark, male survival across the observed range of variation in number of alleles (15-22) increased more than fourfold from 0.17 to 0.77. Second, I found 35-40% of Kansas, and 1.5 % (11,000 Km squared) of the GPLCC, were considered high-quality lek habitats. Top performing logistic models predicting lek presence (wi=0.95) included strong effects of grassland cover and avoidance of anthropogenic disturbance. When this model was applied to putative future landscapes based on climate change and current land use trends over a 70-year period, I found a 27-40% reduction in habitat area and a 137 Km southeast shift in habitat distribution. Under equilibrium conditions we expect isolation by distance (IBD) to explain the distribution of genetic diversity. However, if the landscape restricts dispersal, then we might observe isolation by resistance (IBR). I used model selection procedures to choose among competing IBR or IBD models to explain the distribution of genetic diversity among GPC populations across Kansas and the GPLCC. IBD was never supported (R-square<0.02, P>0.09). The best models for Kansas (R2=0.69, P<0.02) and for the GPLCC (R-square=0.46, P<0.02) indicated that human-mediated landscape changes have influenced landscape permeability for dispersal. The integration of behavioral, landscape, and genetic data provided new insights on prairie-chicken ecology, and is a powerful approach for developing conservation strategies for sensitive species.

This study examined the impact intervening environment has on gene flow in the insect pollinated understory tree, Cornus florida L., by combining GIS and landscape genetic techniques (Least Cost Path Analysis, Circuit Theory, and Conditional Genetic Distance). Traditional population genetic analysis indicated pair-wise relatedness was significantly correlated to distance (Pearson; r = -0.312, P < 0.001) suggesting a spatial component to offspring relatedness. Dispersal throughout the study site was non-random, exhibiting a high degree of pollen pool structure due to restricted gene flow (Two-Generation Analysis; Φft = 0.161, P = 0.001). Forest structure was quantified in GIS layers representing coniferous canopies, mixed hardwood canopies, C. florida canopies, open understory (roads), and open understory/canopy due to tree removal. Of these layers, landscape isolation for the roads layer provided the best-fit model for describing genetic differentiation among sampled pollen pools (Mantel; r = 0.542, P = 0.001). These data also suggest that improved biological inferences can be gained by examining a range of landscape isolation models.

Understanding worldwide declines in reptiles due to factors such as habitat loss and emerging infectious disease has become an increasingly important focus in conservation biology. Here, I use novel approaches from the field of landscape genetics to combine spatial genetic data with landscape data at both regional and local spatial scales to explore natural and anthropogenic landscape features that shape population structure and gene flow in a federally threatened reptile, Gopherus polyphemus. I also utilize approaches from the field of spatial epidemiology to examine the extent to which environmental variables can be used to predict the seroprevalence of an associated pathogen Mycoplasma agassizzi in gopher tortoise populations. Using mitochondrial data, I find evidence of a historical barrier to gene flow that appears to coincide with the Apalachicola River. I also discover low genetic diversity and evidence of population bottlenecks in the western portion of the range. My evaluation at the regional scale shows that dispersal is limited by geographic distance, areas of low elevation and major roads ways. A finescale study reveals no evidence of spatial genetic structure within a 14 x 35 km area. However, soil type is significantly correlated with pairwise genetic distances between individuals, suggesting that this variable influences fine-scale population structure in the gopher tortoise. In addition to soil, high density canopy cover is an important factor impeding gene flow at the local level for females, while land cover type explains some of the genetic variance between males. Finally, temperature and precipitation appear to be important predictors of the seroprevalence of the pathogen Mycoplasma agassizii in gopher tortoises. The probability of an individual testing seropositive for exposure to this disease increased with high temperature and low precipitation values. The methods presented in this dissertation evaluate novel approaches for assessing the influence of environmental variables on population structure, dispersal and disease occurrence and could be applied in future studies of other threatened and endangered taxa.

En utilisant comme cas d’étude le mouflon Méditerranéen (Ovis gmelini musimon × Ovis sp.), les objectifs de cette thèse étaient de mieux comprendre comment sont liés diversité génétique, comportement des individus, flux de gènes et dynamique parasitaire. Au travers d’approches de génétique des populations et de génétique du paysage, nous avons pu mettre en évidence que la structure génétique spatiale de la population étudiée était impactée par son histoire d’introduction, sa structure socio-spatiale et le paysage dans lequel elle évolue. Etant donné l’impact de ces divers éléments sur les flux de gènes des mouflons, nous nous attendions à ce qu’ils déterminent également les flux de parasites dans la population. Nos résultats ont, au contraire, révélé que les parasites circulent mieux que les gènes de mouflons dans la population. Enfin, nous avons montré que les capacités de résistance des hôtes face à leurs parasites dépendaient de la diversité génétique neutre et adaptative, notamment de l’hétérozygotie d’un gène lié à l’immunité. Les résultats de ce travail décrivent avec précision la distribution de la variabilité génétique et son lien avec les risques sanitaires dans la population d’étude, apportant ainsi des informations cruciales pour la mise en place de stratégies de gestion et de conservation des populations de mouflons dans le contexte actuel de changements globaux et de réémergences de maladies. / Using as a case study the Mediterranean mouflon (Ovis gmelini musimon × Ovis sp.), we aimed at better understanding how are linked genetic diversity, individual behaviour, gene flows and parasitic dynamic. Using population and landscape genetics approaches, we showed that the spatial genetic structure of the studied population was determined by its introduction history, its socio-spatial structure and the landscape in which it evolves. Given the impact of these elements on mouflon gene flow, we expected them to also determine parasite transmission in the population. Our results nevertheless evidenced that parasite are better dispersed than mouflon genes. Finally, we showed that host resistance to parasites depends on neutral and adaptive genetic diversity, and more specifically on heterozygosity at a immunity-linked locus. Our results precisely describe genetic variability spatial distribution and its link with sanitary risks in the studied population, bringing crucial information for wild sheep population management and conservation in the current context of global changes and disease reemergence.

acase@tulane.edu