Unlike desert rodents from North America, Allenby’s gerbil (Gerbillus andersoni allenbyi) from the Negev Desert, Israel has evolved with snakes that do not have heat-sensitive sensory pits that enhance night vision. Does this history affect their ability to assess and respond to a snake that has this ability? As a test, we exposed gerbils to risk of predation from various predators, including snakes, owls, and foxes. The snakes included the Saharan horned viper (Cerastes cerastes) and the sidewinder rattlesnake (Crotalus cerastes). The former snake lacks sensory pits and shares a common evolutionary history with the gerbil. The latter snake, while convergent evolutionarily on the horned viper, has sensory pits and no prior history with the gerbil. The gerbils exploited depletable resource patches similarly, regardless of snake species and moon phase. While the gerbils did not respond to the novel snake as a greater threat than their familiar horned viper, the gerbils were cognizant that the novel predator was a threat. In response to both snakes, giving-up densities (GUDs; the amount of food left in a resource patch following exploitation) of the gerbils were higher in the bush than open microhabitat. In response to moonlight, GUDs were higher on full than on the new moon. Based on GUDs, the gerbils responded most to the risk of predation from the red fox, least from the two snake species, and intermediate for the barn owl. Keywords:

Les plantes sont des organismes immobiles qui doivent répondre et s’adapter à des contraintes abiotiques et biotiques. Parmi les stress biotiques, les maladies virales, établies ou émergentes, peuvent être responsables de pertes de rendement majeures aux conséquences économiques importantes. Face aux phytovirus la lutte génétique constitue le moyen de lutte le plus efficace, le plus respectueux de l’environnement et du consommateur. Comprendre l’interaction entre les plantes et les virus reste indispensable pour rechercher de nouvelles sources de résistances. Ce travail de thèse s’intéresse à l’étude du pathosystème naturel Arabidopis thaliana/Turnip mosaic virus (TuMV). Les essais ont été menés majoritairement en conditions extérieures permettant une analyse de l’interaction dans un environnement multistress. La réponse d’A. thaliana a été explorée par l’étude de traits liés à la maladie et par la variation en métabolites primaires et secondaires. Ce travail a permis i) de caractériser de façon fine la réponse d’A. thaliana au TuMV en conditionsmultistress en exploitant la diversité naturelle d’une population mondiale et française ii) de déterminer l’architecture génétique de cette interaction par des approches de génétique d’association et de QTL mapping. Plusieurs nouveaux loci potentiellement impliqués dans la réponse ont été identifiés iii) de montrer l’intérêt du phénotypage métabolique pour discriminer les accessions en fonction de leur sensibilité au TuMV. La multidisciplinarité des approches constitue la richesse de ce travail de thèse qui contribue à une meilleure caractérisation et compréhension de la réponse des plantes lors d’une infection virale. / Plants are immobile organisms which have to adapt to abiotic and biotic constraints. Among bioticstress, established or emerging viral diseases, may be responsible for major yield losses withsignificant consequences. Genetic control is the most effective, environmentally and consumerfriendlyway to control viral infections. Understanding plant/virus interactions remains essential tosearch for new sources of resistance. This work, focuses on the study of the natural pathosystemArabidopsis thaliana/Turnip mosaic virus (TuMV). Most of the trials were conducted in commongarden conditions allowing the analysis of the interaction in a multistress environment. A. thaliana’sresponse was explored through the study of disease-related traits and the variations in primary andsecondary metabolites. This work allows i) the fine characterization of A. thaliana’s response toTuMV in multistress conditions through the exploration of the natural diversity of a world and Frenchpopulation ii) to determine the genetic architecture of this interaction by genome wide associationsand QTL mapping. Several new loci potentially involved in the response have been identified iii) tohighlight the interest of metabolic phenotyping to discriminate accessions according to theirsusceptibility to TuMV. The multidisciplinary approaches contribute to a better characterization andunderstanding of plant-virus interaction.