Noninvasive methods for monitoring wildlife species have revolutionized the way population parameters, such as population density and survival and recruitment rates, are estimated while accounting for imperfect detection using capture-recapture models. Reliable estimates of these parameters are vital information required for making sound conservation decisions; however to date, noninvasive sampling methods have been of limited use for a vast number of species which are difficult to identify to the individual level–a general requirement of capture-recapture models. Capture-recapture models that utilize partial identity information have only recently been introduced and have not been extended to most types of noninvasive sampling scenarios in a manner that uses the spatial location where noninvasive samples were collected to further inform complete identity (i.e. spatial partial identity models). Herein, I extend the recently introduced spatial partial identity models to the noninvasive methods of remote cameras for species that are difficult to identify from photographs and DNA from hair or scat samples. The ability of these novel models to improve parameter estimation and extend study design options are investigated and the methods are made accessible to applied ecologists via statistical software. This research has the potential to greatly improve wildlife conservation decisions by improving our knowledge of parameters related to population structure and dynamics that inform those decisions. Unfortunately, many species of conservation concern (e.g., Florida panthers, Andean bears) are managed without having the necessary information on population status or trends, largely a result of the cost and difficulty of studying species in decline and because of the difficulty of applying statistical models to sparse data, which can produce imprecise and biased estimates of population parameters. By leveraging partial identity information in noninvasive samples, the models I developed will improve these parameter estimates and allow noninvasive methods to be used for more species, leading to more informed conservation decisions, and a more efficient allocation of conservation resources across species and populations. / Ph. D. / Noninvasive methods for monitoring wildlife species have revolutionized the way population parameters, such as population density and survival and recruitment rates, are estimated while accounting for imperfect detection using capture-recapture models. Reliable estimates of these parameters are vital information required for making sound conservation decisions; however to date, noninvasive sampling methods have been of limited use for a vast number of species which are difficult to identify to the individual levela general requirement of capture-recapture models. Capture-recapture models that utilize partial identity information have only recently been introduced and have not been extended to most types of noninvasive sampling scenarios in a manner that uses the spatial location where noninvasive samples were collected to further inform complete identity (i.e. spatial partial identity models). Herein, I extend the recently introduced spatial partial identity models to the noninvasive methods of remote cameras for species that are difficult to identify from photographs and DNA from hair or scat samples. The ability of these novel models to improve parameter estimation and extend study design options are investigated and the methods are made accessible to applied ecologists via statistical software. This research has the potential to greatly improve wildlife conservation decisions by improving our knowledge of parameters related to population structure and dynamics that inform those decisions. Unfortunately, many species of conservation concern (e.g., Florida panthers, Andean bears) are managed without having the necessary information on population status or trends, largely a result of the cost and difficulty of studying species in decline and because of the difficulty of applying statistical models to sparse data, which can produce imprecise and biased estimates of population parameters. By leveraging partial identity information in noninvasive samples, the models I developed will improve these parameter estimates and allow noninvasive methods to be used for more species, leading to more informed conservation decisions, and a more efficient allocation of conservation resources across species and populations.

Nesta tese propomos dois modelos espaciais de captura-recaptura para estimação da abundância populacional em população aberta. Os modelos estatísticos propostos ajustam-se a dados obtidos via amostragem de captura-recaptura com marcação individual realizada em diferentes locais dentro do habitat, levando em consideração as taxas de nascimentos e mortes durante o período de estudo e as localizações geográficas das capturas. No primeiro modelo, propomos uma modelagem hierárquica para os tamanhos populacionais locais a fim de obter a distribuição preditiva da abundância populacional para regiões não visitadas pela amostragem. Nesta etapa, uma estrutura para dados zero-inflacionados foi adotada para acomodar situações quando realizam-se amostragens em locais sem a presença da espécie. O segundo modelo proposto leva em consideração o deslocamento dos animais entre os diferentes locais de amostragem, generalizando o primeiro modelo no qual consideramos a permanência dos animais em um mesmo local. Neste caso, tornou-se possível estimar o tamanho da área de vida (movimentação) da espécie além de predizer locais com maiores abundâncias de animais. Em ambos modelos, propomos uma abordagem bayesiana para o processo inferencial e derivamos algoritmos de simples implementação computacional, a partir do uso de técnicas de dados aumentados. As propriedades frequentistas dos estimadores bayesianos foram avaliadas por meio de estudos de simulação e, por fim, estas propostas de modelagem foram aplicadas a três conjuntos de dados reais de aracnídeos. / In this thesis we propose two spatial capture-recapture models for estimation of population abundance in the open population. The proposed statistical models conform to data obtained through individual tag capture-recapture sampling performed in different areas within the habitat, taking into account the rates of births and deaths during the study period and the geographical locations of the catches. In the first model, we propose a hierarchical modeling for local population sizes in order to obtain the predictive distribution of population abundance for regions not visited by sampling. In this step, a structure for zero-inflated data was adopted to accommodate situations when sampling is performed in areas without the presence of the species. The second proposed model takes into account the movement of the animals among the different sampling areas, generalizing the first model in which we consider the permanence of the animals in the same area. In this case, it became possible to estimate the size of the area of movement of the species and to predict areas with higher abundances of animals. In both models, we propose a Bayesian approach to the inferential process and derive algorithms from simple computational implementation, from the use of augmented data techniques. The frequentist properties of the Bayesian estimators were evaluated by simulation studies and, finally, these modeling proposals were applied to three real data sets of arachnids.

The red-backed salamander (RBS; Plethodon cinereus) has a large geographic range and, though it is common throughout, abundance varies spatially. It has been studied extensively for at least a century and as a result; its distribution, habitat use, behavioral ecology, reproductive ecology, and ecological role are well understood in comparison to other cryptic wildlife. Multiple sampling methods have been developed to monitor RBS, and I discuss in detail one common method – artificial cover objects (ACOs). Spatial capture-recapture (SCR) is uniquely suited to estimate population parameters, including true density, and is paired well with spatially referenced sampling methods, such as ACOs, and animals capable of retaining unique marks, such as RBS. In the introduction of this thesis, I review RBS, ACOs, and SCR independently in detail, and go on to discuss their compatibility for monitoring terrestrial salamander populations. Detailed, and often range-wide, monitoring of ecosystems is necessary to gather the information needed to achieve broad multiple-scale conservation objectives. Indicator species are recommended tools for filling in gaps of knowledge where these range-wide data do not exist. The use of indicator species is precluded by the lack of evaluation of candidate indicators and their relationship to indicated processes. In this thesis, I discuss in detail the practicality of indicator species as wildlife management tools, and challenges in their application – primarily their practicality when direct measurement of a variable of interest is possible. I advocate for integrative indicator species applications that make use of relationships to latent variables, review two conceptual models involving latent variables, and propose a modification to these models that makes relationships between variables more explicit. Inference of among-population variation to adaptive capacity, response to large-scale threats, and the condition of ecosystems is limited in part by unstandardized methods. Ecological relationships are made difficult to characterize by gaps in data - and this is especially true of links between indicator species their and related ecosystem processes. Using a candidate indicator species, I tested the congruence of population parameter estimates from study designs that varied. In Wendell State Forest, MA, I manipulated spatial arrangements of artificial cover objects (ACO) arrays and evaluated their use for monitoring terrestrial salamanders. ACOs mimic natural habitat - and attraction of RBS to traps may induce behavioral bias in parameter estimates if not accounted for. I sought to determine if variation in ACO design can be accommodated to make comparable estimates. I found that analyzing data from ACOs using spatial capture-recapture (SCR) modeling produces consistent within-population density estimates regardless of ACO configuration.

Though the American marten (Martes americana) is widely distributed across northern North America, habitat use and population abundance vary widely across the range. Due to its status as a furbearer, the species has been extensively researched, resulting in a large body of knowledge about the species’ ecology, distribution, and abundance, as well as drivers of population structure and dynamics. More recently, marten research has shifted focus to genetics, habitat associations, and estimation of population state variables. The rapid increase in estimation of states such as occupancy, abundance, and density has likely been driven by the increasing accessibility of noninvasive field technology, such as noninvasive genetic sampling and remote camera trapping, and by the statistical development of ecological hierarchical models. This convergence of advances in field and analytical methods is most apparent in the now widespread application of spatial capture-recapture, an approach that produces robust estimates of population densities and abundance that can be compared across time and space. These new models are especially valuable near the edges of marten distribution where populations are often recovering from historic overexploitation, and expanding into areas they have previously been absent from. In these areas, detailed, landscape-scale understanding of marten populations is necessary in order to establish current conditions, effectively monitor changes, and predict what effect management actions may have on marten populations. I utilized these models to study marten populations in New Hampshire where marten are a species of management interest, and recent recovery has led to their removal from the state endangered species list. Through a collaborative effort with New Hampshire Fish and Game Department in the winters of 2017 and 2018, marten were surveyed across northern New Hampshire using a novel camera trap design that allows for the identification of individuals. These data were analyzed using spatial capture-recapture models, allowing me to evaluate habitat associations that explain spatial variation in marten density and provide a population status assessment for the New Hampshire marten population. Marten densities are highest in the White Mountain National Forest, though other protected lands in northern New Hampshire also appear to support larger populations. The greatest population densities coincided with deeper snows, increased canopy closure, and intermediate boreal biomass. These results provide additional support for several hypotheses explaining marten habitat use across their range while also providing novel insight that will inform active management of both marten and the habitat they occur in. In addition to the population status assessment, I evaluated the relationship between estimates of occupancy and density in New Hampshire. Though utility of non-invasive methodology can decrease research costs, the need for individual identification in spatial capture-recapture models represents a cost increase over occupancy models. My results suggest that the two are positively correlated; however, occupancy is a poor predictor of the entire range of density, especially because the variables used to predict each of the state variables are different. Thus, occupancy is likely not a good proxy for density in New Hampshire, however it could be used to track general trends through time so long as density is re-evaluated periodically.

Rares, discrets, fascinants et sanguinaires sont autant d'adjectifs utilisés en Europe pour qualifier les grands carnivores qui arpentent nos forêts. La dégradation de leur habitat et la raréfaction de leurs proies, associées au folklore qui les entoure, ont engendré de nombreux conflits avec l'Homme et les ont menés à disparaître de la plupart de leurs aires historiques de répartition. Depuis, ces espèces ont le statut d'espèces protégées dans la plupart des pays européens. Ce statut est notamment garanti pour la plupart par la Directive Habitat et la convention de Berne. Ces textes législatifs ont permis d'instaurer un contexte favorable pour un retour progressif de ces espèces au sein de nos écosystèmes. Afin de remplir les exigences instaurées par ces différents textes législatifs, il était nécessaire d'améliorer les connaissances scientifiques sur l'écologie de ces espèces et la dynamique de leurs populations. Les pays européens ont ainsi déployé des efforts considérables afin de contribuer à une connaissance globale et à une gestion durable des grands carnivores. Ces efforts ne sont clairement pas homogènes entre l'ours, le loup et le lynx. Le premier enjeu de cette thèse était donc d'évaluer les facteurs pouvant expliquer l'hétérogénéité d'investissement dans la conservation de ces espèces en Europe. Ce chapitre repose à la fois sur des critères écologiques des populations locales et sur des critères économiques des pays considérés. Le premier résultat fort de cette étude montre que l'ours et le lynx auraient un même profil et bénéficieraient du même intérêt pour les scientifiques européens, le loup différant de ces deux espèces. Le second résultat probant révèle que les travaux de recherche seraient davantage orientés vers les populations à forts effectifs plutôt que vers les petites populations. L'investissement scientifique dans ce premier chapitre est en partie quantifié par l'effort investi dans le suivi des populations, qui reste un véritable défi pour les grands carnivores. En effet, le comportement discret de ces espèces, leur faible densité et leur besoin de grands espaces sont autant de contraintes pour leur suivi qui requiert alors d'importants moyens humains et financiers. Le suivi des effectifs du lynx boréal (Lynx lynx), en France, est un exemple révélateur de ces contraintes. Il reposait jusqu'à récemment sur la collecte d'indices de présence indirects. Motivés par la mise en place d'un protocole de suivi non-invasif mais coûteux par piégeage photographique dans le massif jurassien français, nous avons évalué dans un deuxième chapitre une nouvelle méthode d'estimation des effectifs de cette population qui permet d'inclure l'information spatiale dans l'analyse. Cette méthode a permis de fournir la première estimation fiable des effectifs de lynx en France. Cette estimation est fournie néanmoins avec une précision toute relative au vu du peu de données collectées lors de ce suivi. L'écart entre le budget nécessaire pour obtenir un recensement de la population et le budget disponible pour le suivi de l'espèce étant considérable, il a fallu dans un troisième chapitre développer un nouvel outil pour optimiser l'utilisation des données disponibles. La combinaison des données de présence-absence et des données de piégeage photographique a permis d'améliorer considérablement les estimations d'effectifs qui sont, dans le Jura français, plutôt en hausse ces dernières années. La situation n'est pas aussi favorable pour l'espèce dans la région des Vosges. Cette population, issue d'une réintroduction, semble décliner de manière drastique depuis les 5 dernières années. Dans un quatrième chapitre, nous avons donc étudié l'efficacité de deux stratégies de conservation visant d'une part à favoriser la connectivité entre les populations vosgienne et jurassienne et d'autre part à réintroduire des individus dans la forêt Palatine allemande, située en continuité du massif vosgien. / Rare, discrete, fascinating and bloodthirsty are all adjectives used in Europe to describe the large carnivores that roam our forests. Degradation of their habitat and depletion of their prey, combined with these wild and "bloodthirsty" aspects, have led to numerous conflicts with humans and led them to disappear from most of their historical range. Since then, most of them have a protected status in most European countries. This status is guaranteed for most of these countries by the Habitats Directive and the Berne Convention. These laws created a favorable context for a gradual return of these species in our ecosystems. To fulfill the requirements established by these laws, it was necessary to improve scientific knowledge of the ecology of these species and to develop methodological tools to understand the dynamics of their populations. European countries then made considerable efforts to contribute to global knowledge and sustainable management of large carnivores. These efforts, in the case of the bear, wolf and lynx, are clearly not homogenous within Europe. The first challenge of this thesis was to evaluate the factors that might explain the heterogeneity of investment in the conservation of large carnivores in Europe. This chapter is based both on ecological criteria of local species and economic criteria of the countries of interest. The first strong result of this study revealed that the bear and lynx have the same profile and receive the same interest from European scientists, wolf differing from the two other species. The second convincing result revealed that the research would be more oriented towards abundant populations rather than small populations as previously assumed by the scientific community. The scientific investment in this first chapter is partly quantified by the amount of effort invested in monitoring populations. It turns out to be a real challenge for large carnivores. The secretive behavior of these species, their low density and their need for large spaces are all constraints to monitoring requiring substantial human and financial resources. Eurasian lynx (Lynx lynx) monitoring in France is a typical example of these constraints. Until now, monitoring was based on the collection of indirect presence signs. Recently, a non-invasive but expensive camera-trapping protocol has been settled in the French Jura Mountains. We then evaluated in a second chapter a new method for estimating the size of this population. This method has provided the first reliable estimate of the abundance of lynx in France. This estimate, however, came with a relative precision given the limited quantity of data collected during this session. The gap between the needs for a census of the population and the budget available for the monitoring of the species is huge so we had to develop a new modeling tool to achieve our goal. In the third chapter, the objective was to improve these estimates, optimizing the use of the available data. The combination of presence-absence data and camera trapping data has greatly improved the French Jura population estimates that go rising in recent years. The situation is not as favorable for the species on the Vosges part. This population, after reintroduction, appears to be declining drastically over the last 5 years. In a fourth chapter, we therefore investigated the effectiveness of two conservation strategies: the first one was to instate some connectivity between the Vosges and Jura populations and the second one to reintroduce individuals in the German Palatinate Forest, situated in continuity with the Vosges.

Ecological inquiry is rooted fundamentally in understanding population abundance, both to develop theory and improve conservation outcomes. Despite this importance, estimating abundance is difficult due to the imperfect detection of individuals in a sample population. Further, accounting for space can provide more biologically realistic inference, shifting the focus from abundance to density and encouraging the exploration of spatial processes. To address these challenges, Spatial Capture-Recapture (“SCR”) has emerged as the most prominent method for estimating density reliably. The SCR model is conceptually straightforward: it combines a spatial model of detection with a point process model of the spatial distribution of individuals, using data collected on individuals within a spatially referenced sampling design. These data are often coarse in spatial and temporal resolution, though, motivating research into improving the quality of the data available for analysis. Here I explore two related approaches to improve inference from SCR: sampling design and data integration. Chapter 1 describes the context of this thesis in more detail. Chapter 2 presents a framework to improve sampling design for SCR through the development of an algorithmic optimization approach. Compared to pre-existing recommendations, these optimized designs perform just as well but with far more flexibility to account for available resources and challenging sampling scenarios. Chapter 3 presents one of the first methods of integrating an explicit movement model into the SCR model using telemetry data, which provides information at a much finer spatial scale. The integrated model shows significant improvements over the standard model to achieve a specific inferential objective, in this case: the estimation of landscape connectivity. In Chapter 4, I close by providing two broader conclusions about developing statistical methods for ecological inference. First, simulation-based evaluation is integral to this process, but the circularity of its use can, unfortunately, be understated. Second, and often underappreciated: statistical solutions should be as intuitive as possible to facilitate their adoption by a diverse pool of potential users. These novel approaches to sampling design and data integration represent essential steps in advancing SCR and offer intuitive opportunities to advance ecological learning about spatial processes.

Coyote (Canis latrans) range expansion in the Central Appalachian Mountains has stimulated interest in ecology of this predator and potential impacts to prey populations. This is particularly true in the Ridge and Valley Region in western Virginia where white-tailed deer (Odocoileus virginianus) populations are restricted by low nutritional carrying capacity and are subject to two other predators, bobcats (Lynx rufus) and American black bears (Ursus americanus), in addition to an active hunter community. I address two primary objectives of the Virginia Appalachian Coyote Study: to investigate 1) spatial ecology and 2) population dynamics of coyote populations in Bath and Rockingham counties. I deployed 21 GPS satellite collars on 19 coyotes over 32 months. I estimated home range size (mean = 13.46 km², range = 1.23 km² - 38.24 km²) across months using biased-random bridges and second-order habitat selection at four scales using eigenanalysis of selection ratios. I developed a metric to classify social status of individuals as either resident or transient based on stability of home range centers over time. I found evidence for class substructure for selection of territories where adult residents had a higher probability of mortality in high productivity/high risk habitats, compared to subadults and transients that were restricted to less productive habitats. I collected scat samples over five scat surveys across 2.5 years and extracted fecal DNA to identify individual coyotes in a mark-recapture framework. I estimated coyote densities in Bath (5.53 – 9.04 coyotes/100 km²) and Rockingham Counties (2.41 – 8.53 coyotes/100 km²) using a spatial capture-recapture model. Six-month apparent survival was lower in Bath County (Φ_Bath = 0.442, 0.259 – 0.643; Φ_Rockingham = 0.863, 0.269 – 0.991). The Bath County population demonstrated persistence despite high mortality and the Rockingham population demonstrated boundedness with recruitment inverse of changes in density. Findings at both sites suggest density-dependence, and tests of territoriality, presence of transients, and territory turnover demonstrate a capacity for immediate local immigration in response to high mortality in Bath County. I suggest that landscape-level habitat management may be a viable strategy to reduce potential conflicts with coyotes in the region. / Ph. D.