Animal bodies house trillions of bacteria, which can influence host behavior in ways that have far-reaching implications for host ecology and evolution. Recent studies have revealed surprising roles for bacteria in shaping behaviors across many animal taxa. But questions remain and recent perspective papers have thus emphasized the need of studying the interaction between non-pathogenic bacteria and host behavior. In many species, individuals preferentially mate with MHC (Major Histocompatibility Complex) unrelated partners which they discriminate using odor cues. The mechanism by which MHC genes influence odor is still unclear, but one hypothesis suggests that MHC genes may influence body odor indirectly by shaping bacterial communities in scent integument. Bacteria are well-known to produce odorants, but whether they shape the odor cues signaling MHC genotype remains unknown. The main objective of the BactOdo project is to inquire whether bacteria in scent integument may be responsible for the production of MHC-related odors in birds. We will adopt a step by step empirical and experimental approach, bringing together the fields of genetic, behavioral, microbial and chemical ecology. We will use a bird model, the blue petrel Halobaena caerulea nesting on Kerguelen Archipelago. The blue petrel is particularly well adapted for such a project as it preferentially mates with MHC unrelated partners, its musky odor carries information on genetic relatedness and it possesses a very well developed olfactory sense. The BactOdo project will be organized around 4 scientific tasks, plus a coordination task and a communication task. More specifically, we will determine (1) whether microbial communities in feathers and preen gland correlate with body odor and MHC traits. Then we will determine whether an experimental change in bacterial community disrupts (2) the MHC signal in odors and (3) the perception by mice of odor similarity between related birds. Finally we will determine (4) whether social relationships between pair mates make their bacterial community more similar and therefore reduce the MHC signal in odor. Our interdisciplinary BactOdo project aims therefore at filling a knowledge gap within one of the most attractive field of research in evolutionary ecology in the past decade, i.e., the role of odor cues in MHC-related mate choice. This very innovative work will include cutting edge ultra-deep sequencing methods, and fells into efforts being made worldwide to describe the factors associated with variations in host-associated microbiota (Human Microbiome Project, Earth Microbiome Project). This project will be carried out at the "Centre d’Ecologie Fonctionelle et Evolutive" (CEFE, UMR 5175), Montpellier, in the Behavioural Ecology group, which has decade-long experience on olfactory communication in petrels and fieldwork at Kerguelen Archipelago. In addition, the CEFE hosts several research groups on chemistry and microbiology and the candidate will be associated to a network of scientists from complementary backgrounds, creating a supremely appropriate environment for research in this area.