Evolution and ecology of an amphibian emerging infectious disease: a context-dependant approach of ranavirus virulence in Lithobates (Rana) pipiens

Abstract

Host-pathogen investigations have conceptually evolved during the last two decades, from a basic and descriptive approach to a current hypothesis-driven and a more theoretical discipline shaped by evolutionary biology. Our deeper understanding of the elements influencing the mutual selective pressures that the host and the pathogens exert on each other, together with recent conceptual advances, currently position this field of research at the frontier between ecology and evolution. Recent theoretical considerations define hostpathogens systems as an evo-eco mosaic comprised of evolutionary and ecological attributes in turn underlying the context-dependent nature of the system dynamic. Therefore, investigations of host-pathogen interactions should integrate the diversity of the systems drivers by using an integrative approach in order to elucidate both coevolutionary trajectory and epidemiological dynamic of the system. In this thesis, such a framework is used to investigate Amphibian/ranavirus interactions. Ranaviruses are emerging pathogens known to have caused amphibian die-offs on five continents with the greatest number of reported mortality events documented in North America and Europe. Despite an increasing understanding of ranaviral disease properties, ranavirus disease dynamics in the environment remain poorly understood. For instance, the influence of potential abiotic and biotic mechanisms including temperature, local landscape features, larval developmental stages, host density and genetic variability as well as genotypic interactions between the host and the pathogen has on the prevalence and virulence of the virus remains to be elucidated. In order to improve our knowledge regarding these specific determinants of ranaviral disease, I designed a combination of manipulative laboratory experiments and a field mensurative survey using the ranid amphibian Lithobates (Rana) pipiens as the host model for this system. I observed that populations of amphibian hosts inhabiting urbanized landscapes suffered from significant decline in genetic diversity in turn promoting the accrued infection by the ranavirus (manuscript 1). Complementary analysis using two amphibian host species, L.pipiens and L.sylvaticus, and three ranavirus strains revealed significant variation among hosts for their susceptibility to ranavirus, and significant variation among ranavirus strains for infectivity. I also showed that specific amphibian/ranavirus interactions might have a tighter coevolutionary history than other combinations, resulting in sharper mutual coadaptations and the potential for frequency-dependent selection to operate in this system. However, the coevolutionary trajectories in this host-pathogen system are dependent on the temperature conditions in which the interaction takes place. Amphibian/ranavirus interactions outcomes iv are therefore temperature, host, and pathogen genotype-dependent suggesting that the range of infection outcomes in this system is potentially large (manuscript 2). Further, I observed that increasing animal holding density is detrimental for host fitness as mortality rate is higher, day of death earlier, development longer, and growth rate significantly lower when tadpoles are experimentally exposed to ranavirus in high holding density situations. These results paralleled a linear increase of detrimental effects when ranavirus doses increased in low density conditions, with control tadpoles having a significantly higher overall relative fitness. However, this pattern was not observed in high density conditions, where the effects of increasing ranavirus dose were limited, revealing non-trivial density-dependence of virulence expression (manuscript 3). Finally, ranavirus infection rate varied with the host developmental stage as the host immune system clears the infection over the course of individual host development. However the intensity of the clearing depends on both the timing and number of ranavirus exposures (manuscript 4). Overall the results described in my thesis suggest that ranavirus virulence depends on a diversity of ecological, epidemiological, and evolutionary determinants. The underlying complexity of ranavirus