La gestion des maladies infectieuses dans la faune sauvage se heurte à de nombreuses limites, et le développement de stratégies efficaces représente un défi de taille. Pour atteindre cet objectif, une compréhension fine des facteurs influençant la transmission et la persistance de l’infection est nécessaire. Parmi ces facteurs, l’hétérogénéité de transmission est une caractéristique importante des populations sauvages. En effet, la diversité des comportements, des structures sociales et spatiales, ou encore des espèces peut conduire à des contributions très variables au nombre de nouvelles infections. Par conséquent, quantifier l’hétérogénéité de transmission pourrait permettre d’améliorer l’efficacité des mesures de gestion sanitaire dans la faune sauvage, en ciblant les individus ou les unités de population qui sont responsables de la majorité des évènements de transmission. L’objectif de cette thèse était d’améliorer les connaissances sur la gestion des maladies infectieuses dans des populations sauvages hétérogènes, en utilisant la brucellose à Brucella melitensis dans une population de bouquetin des Alpes (Capra ibex) comme modèle d’étude. En effet, la biologie de la brucellose et l’écologie de l’espèce hôte se prêtent bien à l’existence et donc à l’étude d’une hétérogénéité de transmission à différentes échelles. A l’aide de cultures bactériennes, nous avons tout d’abord montré que seulement 58 % des individus séropositifs sont à risque d’excréter la brucellose, et que ce risque diminue avec l’âge. Ensuite, mettant à profit l’existence d’informations détaillées sur la dynamique de population et le comportement du bouquetin, et de données épidémiologiques dans la population d’étude, nous avons développé un modèle individu-centré afin de quantifier l’hétérogénéité individuelle et spatiale de la transmission. Nous avons démontré que la transmission de la brucellose était hétérogène entre individus, les femelles provoquant environ 90% des nouvelles infections, et entre unités spatiales, plus de 80% des cas de transmission ayant lieu dans les trois sous-unités socio-spatiales qui forment la zone cœur du massif. Nous avons également estimé l’évolution temporelle de la séroprévalence et de la force d’infection, en utilisant différents modèles statistiques. Les résultats suggèrent que l’importante opération de capture menée en 2015, avec test systématique et élimination des individus séropositifs, a permis de diminuer la transmission de la brucellose dans la population. Sur la base de l’ensemble de ces résultats, nous avons évalués une série de stratégies de gestion sanitaire qui pourraient être utilisées à l’avenir dans la population. Les résultats, issus du modèle individu-centré, confirment que la stratégie prioritaire devrait être d’éliminer le plus d’individus séropositifs, et que cibler les femelles et/ou la zone cœur permet d’améliorer l’efficacité des mesures. Bien qu’il n’y ait pas de solution évidente pour la gestion de la brucellose dans notre cas d’étude, les stratégies de gestion ciblées sont très prometteuses et permettent de raffiner les mesures sanitaires classiquement utilisées. Il est donc primordial de bien comprendre l’hétérogénéité de transmission dans les populations sauvages infectées, et de rechercher des stratégies ciblées qui peuvent permettre d’améliorer la gestion en termes d’efficacité et d’acceptabilité / The management of infectious diseases in wildlife reservoirs is particularly challenging and faces several limitations. The development of appropriate management strategies requires a detailed understanding of the factors affecting the transmission and persistence of the infectious agent in the population. Among these factors, heterogeneity of transmission is a common characteristic in natural host-pathogen systems. Indeed, wild animals express a broad range of behaviours, are organised in a variety of social and spatial structures, occupy many areas with very different characteristics and belong to a large diversity of species. Such heterogeneities, from between-individuals to between-species, may result in different contributions to the overall number of new cases of infections. Thus, understanding transmission heterogeneity could provide valuable insights on how to effectively manage these systems, by targeting the individuals or areas that are responsible for most transmissions. The aim of this thesis was to provide insights on the monitoring and management of infectious diseases in heterogeneous wild populations, using Brucella melitensis infection in a French population of wild Alpine ibex (Capra ibex) as a case study. The biology of brucellosis and the ecology of Alpine ibex makes this case study a good candidate for transmission heterogeneity at several levels. Using bacterial examinations, we first established that only 58% of seropositive individuals were at risk to excrete Brucella, and that this risk decreased with increasing age. Then, we took advantage of detailed information available on ibex population dynamics, behaviour, and habitat use, and on epidemiological surveys, to build an individual-based model in order to quantify heterogeneity at the individual and spatial levels. The transmission is extremely heterogeneous between individuals, with females generating around 90% of the new cases of brucellosis infection, and between spatial units, three of the five socio-spatial units (the core area) accounting for more than 80% of brucellosis transmission. Using statistical models to estimate the temporal dynamics of the seroprevalence and of the force of infection in the population, we found evidence that the massive captures with test-and-remove operations that were conducted in 2015 managed to reduce brucellosis transmission in the population. Based on these results, we evaluated several predictive disease management strategies in the individual-based model. Our results confirmed that the primary strategy should be to remove as many seropositive individuals as possible, and that strategies targeting females and/or the core area are more effective than untargeted management. Although there is no silver bullet for the management of brucellosis in the population of study, targeted strategies offer a wide range of promising refinements to classical sanitary measures. We therefore encourage to look for heterogeneity in other infection-wildlife systems and to evaluate potential targeted strategies for improving management schemes in terms of efficiency and acceptability

Biological invasions provide a unique opportunity to study the mechanisms that regulate community composition and ecosystem function. Invasive species that are also ecosystem engineers can substantially alter physical features in an environment, and this can lead to cascading effects on the biological community. Aquatic-terrestrial interface ecosystems are excellent systems to study the interactions among invasive ecosystem engineers, physical features, and biological communities, because interactions among vegetation, sediment, and fluids within biophysical feedbacks create and modify distinct physical features. Further, these systems provide important ecosystem services including coastal protection afforded by their natural features. In this dissertation, I investigate the interactions and feedbacks among sand-binding beach grass species (a native, Elymus mollis (Trin.), and two non-natives, Ammophila arenaria (L.) Link and A. breviligulata Fernald), sediment supply, and dune shape along the U.S. Pacific Northwest coast. Dunes dominated by A. arenaria tend to be taller and narrower compared to the shorter, wider dunes dominated by A. breviligulata. These patterns suggest an ecological control on dune shape, and thus, coastal vulnerability to overtopping waves. I investigate the causes and consequences of these patterns with experiments, field observations, and modeling. Specifically, I investigate the relative roles of vegetation and sediment supply in shaping coastal dunes over inter-annual and multi-decadal time scales (Chapter 2), characterize a biophysical feedback between beach grass species growth habit and sediment supply (Chapter 3), uncover the mechanisms leading to beach grass coexistence and whether A. breviligulata can invade and dominate new sections of coastline (Chapter 4), and examine the non-target effects resulting from management actions that remove Ammophila for the recovery of the threatened Western Snowy plover (Charadrius alexandrinus nivosus) (Chapter 5). I found that vegetation and sediment supply play important roles in dune shape changes across inter-annual and multi-decadal time scales (Chapter 2). I determined that a biophysical feedback between the beach grass growth habits and sediment supply results in species-specific differences in sand capture ability, and thus, is a likely explanation for differences in dune shape (Chapter 3). I found that all three beach grass species can coexist across different sediment deposition rates, and that this coexistence is largely mediated by positive direct and indirect species interactions. I further determined that A. breviligulata is capable of invading and dominating the beach grass community in regions where it is currently absent (Chapter 4). Combined, these findings indicate that A. breviligulata is an inferior dune building species as compared to A. arenaria, and suggest that in combination with sediment supply gradients, these species differences ultimately lead to differences in dune shape. Potential further invasions of A. breviligulata into southern regions of the Pacific Northwest may diminish the coastal protection ability of dunes currently dominated by A. arenaria, but this effect could be moderated by the predicted near co-dominance of A. arenaria in these lower sediment supply conditions. Finally, I found that the techniques used to remove Ammophila for plover recovery have unintended consequences for the native and endemic dune plant communities, and disrupt the natural disturbance regime of shifting sand. A whole-ecosystem restoration focus would be an improvement over the target-species approach, as it would promote the return of the natural disturbance regime, which in turn, would help recover the native biological community. The findings from this dissertation research provide a robust knowledge base that can guide further investigations of biological and physical changes to the coastal dunes, can help improve the management of dune ecosystem services and the restoration of native communities, and can help anticipate the impacts of future beach grass invasions and climate change induced changes to the coast. / Graduation date: 2012 / Access restricted to the OSU Community at author's request from Sept. 22, 2011 - March 22, 2012