Understanding the demography of species over recent history (e.g., < 100 years) is critical in studies of ecology and evolution, but records of population history are rarely available. Large single nucleotide polymorphism datasets generated with restriction-site associated DNA sequencing (RADseq), in combination with demographic inference methods, are improving our ability to gain insights into the population history of both model and non-model species. However, to assess the performance of genetic methods it is important to compare their estimates of population history to known demography, in both simulation and empirical settings. Here, I used a simulation approach to examine the potential for RADseq datasets to accurately estimate effective population size (Ne) in Wright-Fisher populations over the course of stable and declining population trends, and distinguish stable from steadily declining populations over a contemporary time scale (20 generations). Overall, my results reveal that demographic inference using genome-wide data can be successfully applied to estimate Ne, and the detection of population-size declines. Next, I assess these methods in an empirical study from a wetland with 37 years of amphibian mark-recapture data to study the utility of genetically-based demographic inference on salamander species with documented population declines (Ambystoma talpoideum) and expansions (A. opacum). For both species, demographic model inference supported population size changes that corroborated mark-recapture data. To further validate these findings, I used individual-based population models of the pond-breeding salamander, Ambystoma opacum, with life-history parameters estimated from a long-term dataset, over a 50 year projection. My results demonstrate that genetically estimated Ne is positively correlated with census size in isolated and subdivided A. opacum populations. Finally, I investigated metapopulation patterns of genomic diversity in A. opacum and A. talpoideum and how migration may impact Ne estimation. I found strong patterns of subpopulation structuring, signatures of migration between subpopulations, and differences in Ne at the subpopulation level in both species. Overall, my findings suggest the ability of genomic data to reconstruct recent demographic changes, which can have important applications to conservation biology, and ultimately can help us elucidate the effects of environmental disturbances in the demography of endangered or declining species.

Cette thèse porte sur la structure spatiale de la diversité génétique. Dans un premier temps, nous étudions un processus à valeurs mesure décrivant l'évolution de la composition génétique d'une population soumise à la sélection naturelle. Nous montrons que ce processus satisfait un théorème de la limite centrale, et que ses fluctuations sont données par la solution d'une équation aux dérivées partielles stochastique. Nous utilisons ce résultat pour donner une estimation du fardeau de dérive au sein d'une population structurée en espace.Dans un deuxième temps, nous nous intéressons à la composition génétique d'une population lorsque les individus se déplacent plus facilement dans une région de l'espace que dans l'autre (on parle alors de dispersion hétérogène). Nous démontrons dans ce cas la convergence des fréquences alléliques via la convergence des lignées ancestrales vers un système de mouvements browniens de Walsh.Nous détaillons également l'impact d'une barrière géographique traversant l'habitat d'une population sur sa diversité génétique. Nous montrons que les lignées ancestrales décrivent dans ce cas des mouvements browniens partiellement réfléchis, dont nous donnons plusieurs constructions.Dans le but d'appliquer ces travaux, nous adaptons une méthode d'inférence démographique au cas de la dispersion hétérogène. Cette méthode utilise les blocs continus de génome hérités d'un même ancêtre entre les paires d'individus dans l'échantillon et permet d'estimer les caractéristiques démographiques d'une population lorsque celles-ci varient dans l'espace. Pour terminer nous démontrons l'efficacité de notre méthode sur des données simulées. / This thesis deals with the spatial structure of genetic diversity. We first study a measure-valued process describing the evolution of the genetic composition of a population subject to natural selection. We show that this process satisfies a central limit theorem and that its fluctuations are given by the solution to a stochastic partial differential equation. We then use this result to obtain an estimate of the drift load in spatially structured populations.Next we investigate the genetic composition of a populations whose individuals move more freely in one part of space than in the other (a situation called dispersal heterogeneity). We show in this case the convergence of allele frequencies via the convergence of ancestral lineages to a system of skew Brownian motions.We then detail the effect of a barrier to gene flow dividing the habitat of a population. We show that ancestral lineages follow partially reflected Brownian motions, of whom we give several constructions.To apply these results, we adapt a method for demographic inference to the setting of dispersal heterogeneity. This method makes use of long blocks of genome along which pairs of individuals share a common ancestry, and allows to estimate several demographic parameters when they vary accross space. To conclude, we demonstrate the accuracy of our method on simulated datasets.

Coalescent theory is a method often used by population geneticists in order to make inferences about evolutionary parameters. The coalescent is a stochastic model that approximates ancestral relationships among genes. An understanding of the coalescent pattern of a sample of sequences, along with some knowledge of the mutations that have occurred, provides information about the evolutionary forces that have acted on the population. Processes such as migration, recombination, variable population size, or natural selection are the forces that affect the genealogies and lead to genetic variability in a sample. Coalescent theory provides a statistical description of the variability in the sample, which in turn leads to inference about evolutionary parameters such as population size, population growth rates, and migration rates. Several methods have been developed that model the coalescent under different sets of evolutionary assumptions. We have examined and compared three computer packages that estimate parameters under the coalescent model: LAMARC, Genetree, and UPBLUE. These are not commercial computer programs, but have been developed by researchers for their own use and made available to others. Their performance in various areas has not been previously well established. We compared the programs in the general areas of model assumptions, availability, usability, and results. No single program is superior to the others in all areas, but each have strengths and weakness. Program selection must be based on the researcher's data, goals and preferences. This comparison should help population geneticists determine which program would be most applicable for their data and research.

The understanding of diversification of intraspecific lineages can shed light on speciation processes and ultimately biogeographic patterns across multiple spatial and temporal scales. In this dissertation I investigated the geographical and ecological factors promoting diversification across the South American dry diagonal biomes (i.e. Cerrado, Chaco, and Seasonally Dry Tropical Forests - SDTFs), through a coupled approach between multilocus phylogeographic and geospatial methods, in the larger context of interpreting the consequences of the resulting patterns for the conservation of biodiversity and evolutionary processes. In Chapter 1 I evaluate biogeographic hypotheses previously proposed and emphasize that the dry diagonal biomes are particularly biodiverse and biogeographically complex, but poorly studied and under protected. I also propose testable predictions for the subsequent chapters and future diversification studies. In the subsequent chapters I adopt a biodiversity prediction approach based on estimating palaeodistributions and habitat stability surfaces to formulate and test spatially explicit diversification hypotheses based on squamate richness and phylogeography. In Chapter 2 I identify historically stable areas of SDTFs and in Chapter 3 I found that the historical climatic stability is a good predictor of Cerrado squamate richness. In Chapter 4 I use a multilocus dataset to estimate the phylogenetic relationships among described species of the lizard genus Phyllopezus (Phyllodactylidae), distributed across the ‘dry diagonal’ biomes. In Chapter 5 I used a dense sampling design focused in the species complex P. pollicaris (more individuals, localities, and markers), and coalescent phylogeographic methods to test the relative influences of Tertiary geomorphological vs. Quaternary climatic events on diversification in this lizard. I found unprecedented levels of cryptic genetic diversity, deep phylogeographic structure, and diversification dating back to at least the Neogene with persistence across Quaternary fluctuations. My dissertation emphasizes that patterns of diversification across the ‘dry diagonal’ biomes are much more complex than previously proposed and reflect the primary influence of geologically old processes. Evidence of allopatric and ecological speciation between lineages that coincide with genetic clusters associated with each of the biomes, contradicts early views that the biomes would have a shared diversification history. These patterns illustrate that low-vagility complexes, characterized by strong structure and pre-Pleistocene divergences, represent ideal radiations to investigate broad biogeography of associated biomes. Future studies should investigate patterns of temporal and spatial congruence across co-distributed taxa, and integrate morphological and further ecological data to refine species limits, taxonomy, and patterns of trait evolution across these radiations.

The Neotropical region is the most biologically diverse region on the planet. The region encompasses a variety of ecosystems and has long been the target of researchers interested in patterns of species diversity and distribution. More recently, molecular data have been incorporated into methods for reconstructing the historical relationships among geographical areas and their biotas. Molecular phylogenetics has provided insights into diversification patterns and the influence of Late Cenozoic events on the evolutionary history of the region. Nevertheless, considering the vast extent and complexity of the region, more studies are needed to fully appreciate the patterns of biogeography and the mechanisms that generate and maintain its biodiversity. Therefore, in Chapter 1 I employed molecular methods to reconstruct the phylogenetic relationships of the subfamily Sigmodontinae, which is the most diverse and widespread radiation of Neotropical rodents. I was able to evaluate controversial hypotheses about the paleogeographic scenarios implicated to explain the biogeography of sigmodontines. Advances in sequencing technology and analytical approaches have revolutionized the role of historical biogeography in elucidating the spatial and temporal context of diversification, and the integrative field of phylogeography was fundamental to the development of biogeography at the intraspecific level. However, the potential of phylogeography to unravel diverse historical scenarios in a tractable statistical framework has been largely unexplored for the Neotropics as a whole. In order to integrate more robust hypothesis testing to elucidate the evolutionary history of Amazonia's biota, I devoted Chapter 2 to a review of Amazonian phylogeography that I anticipate will improve the basis for interpreting the patterns and processes of diversification in Amazonia. Chapter 3 is a thorough species account of spiny rats of the genus Proechimys, which is poorly known taxonomically despite its diversity and widespread distribution in the Neotropics. This taxonomic revision will benefit researchers interested in using such information with coalescent-based methods of species delimitation aimed at an integrative and stable taxonomy. Lastly, Chapter 4 deals with the phylogeography of P. roberti. This species occurs in southeastern Amazonia and the Cerrado of central Brazil. I employed a dense taxon sampling and used coalescent-based methods to demonstrate that rivers and topography have a causal link to the geographic structure of P. roberti populations. In my dissertation, I used a combination of molecular genetics tools to provide a better understanding of the biogeography and evolution of some of the most diverse groups of Neotropical mammals. My dissertation interacts in many levels with my future research interests. These present and future efforts hold promise for unraveling the evolutionary history of the Neotropical region and its biota, and will assist in conservation decisions aiming at preserving its unparalleled biodiversity.

Gene flow and hybridization are pervasive in nature, and can lead to different evolutionary outcomes. They can either accelerate divergence and promote speciation or reverse differentiation. The process of divergence and speciation are strongly influenced by both neutral and selective forces. Disentangling the interplay between these processes in natural systems is important for understanding the general importance of interspecific gene flow in generating novel biodiversity in plants. This thesis first examines the importance of introgressive hybridization in the evolution of the genus Pinus as a whole, and then focusing on specific pine species, investigates the role of geographical, environmental and demographical factors in driving divergence and adaptation. By examining the distribution of cytoplasmic DNA variation across the wide biogeographic range of the genus Pinus, I revealed historical introgression and mtDNA capture events in several groups of different pine species. This finding suggests that introgressive hybridization was common during past species’ range contractions and expansions and thus has played an important role in the evolution of the genus. To understand the cause and process of hybrid speciation, I focused on the significant case of hybrid speciation in Pinus densata. I established the hybridization, colonization and differentiation processes that defined the origin of this species. I found P. densata originated via multiple hybridization events in the late Miocene. The direction and intensity of introgression with two parental species varied among geographic regions of this species. During the colonization on Tibetan Plateau from the ancestral hybrid zone, consecutive bottlenecks and surfing of rare alleles caused a significant reduction in genetic diversity and strong population differentiation. Divergence within P. densata started from the late Pliocene onwards, induced by regional topographic changes and Pleistocene glaciations. To address the role of neutral and selective forces on genetic divergence, I examined the association of ecological and geographical distance with genetic distance in Pinus yunnanensis populations. I found both neutral and selective forces have contributed to population structure and differentiation in P. yunnanensis, but their relative contributions varied across the complex landscape. Finally, I evaluated genetic diversity in the Vietnamese endemic Pinus krempfii. I found extremely low genetic diversity in this species, which is explained by a small ancestral population, short-term population expansion and recent population decline and habitat fragmentation. These findings highlight the role of hybridization in generating novel genetic diversity and the different mechanisms driving divergence and adaptation in the genus Pinus.

The genealogy of a random sample of a population of organisms can be represented as a rooted binary tree. Population dynamics determine a distribution over sample genealogies. For large populations of constant size and in the absence of selection effects, the coalescent process of Kingman determines a suitable distribution. Neuhauser and Krone gave a stochastic model generalising the Kingman coalescent in a natural way to include the effects of selection. The model of Neuhauser and Krone determines a distribution over a class of graphs of randomly variable vertex number, known as ancestral selection graphs. Because vertices have associated scalar ages, realisations of the ancestral selection graph process have randomly variable dimensions. A Markov chain Monte Carlo method is used to simulate the posterior distribution for population parameters of interest. The state of the Markov chain Monte Carlo is a random graph, with random dimension and equilibrium distribution equal to the posterior distribution. The aim of the project is to determine if the data is informative of the selection parameter by fitting the model to synthetic data. / Foundation for Research Science and Technology Top Achiever Doctoral Scolarship

The genealogy of a random sample of a population of organisms can be represented as a rooted binary tree. Population dynamics determine a distribution over sample genealogies. For large populations of constant size and in the absence of selection effects, the coalescent process of Kingman determines a suitable distribution. Neuhauser and Krone gave a stochastic model generalising the Kingman coalescent in a natural way to include the effects of selection. The model of Neuhauser and Krone determines a distribution over a class of graphs of randomly variable vertex number, known as ancestral selection graphs. Because vertices have associated scalar ages, realisations of the ancestral selection graph process have randomly variable dimensions. A Markov chain Monte Carlo method is used to simulate the posterior distribution for population parameters of interest. The state of the Markov chain Monte Carlo is a random graph, with random dimension and equilibrium distribution equal to the posterior distribution. The aim of the project is to determine if the data is informative of the selection parameter by fitting the model to synthetic data. / Foundation for Research Science and Technology Top Achiever Doctoral Scolarship

The genealogy of a random sample of a population of organisms can be represented as a rooted binary tree. Population dynamics determine a distribution over sample genealogies. For large populations of constant size and in the absence of selection effects, the coalescent process of Kingman determines a suitable distribution. Neuhauser and Krone gave a stochastic model generalising the Kingman coalescent in a natural way to include the effects of selection. The model of Neuhauser and Krone determines a distribution over a class of graphs of randomly variable vertex number, known as ancestral selection graphs. Because vertices have associated scalar ages, realisations of the ancestral selection graph process have randomly variable dimensions. A Markov chain Monte Carlo method is used to simulate the posterior distribution for population parameters of interest. The state of the Markov chain Monte Carlo is a random graph, with random dimension and equilibrium distribution equal to the posterior distribution. The aim of the project is to determine if the data is informative of the selection parameter by fitting the model to synthetic data. / Foundation for Research Science and Technology Top Achiever Doctoral Scolarship