Abstract
In agricultural production systems, the use of varietal resistance is an ecological alternative to the application of phytosanitary products in the control of plant pathogens. However, several examples of breakdown or erosion of varietal resistance due to adaptation in populations of pathogens have been reported. A recent case is the erosion of quantitative resistance in bananas by the fungus Pseudocercospora fijiensis, the causative agent of Black Leaf Streak Disease. The objectives of this thesis aimed to highlight an adaptation of P. fijiensis populations to quantitative resistance with in vitro inoculations, to identify the underlying adaptive architecture and to find out if adapted genotypes were introduced in other countries, in particular in the French West Indies. Cross inoculations of detached leaves of resistant and sensible varieties of banana, suggest the existence of local adaptation to quantitative resistance in populations of P. fijiensis from Cuba and the Dominican Republic. Population genomics and quantitative genetics analyses identified 25 genomic regions potentially involved in adaptation to these resistances. These results suggest a complex adaptive architecture involving a polygenic basis, functional redundancy and low level of convergence between the populations studied. Annotation of genes detected in genomic regions and biological data support the hypothesis that some of them may play a role in pathogenicity. Finally, a genetic analysis of populations in the Latin America - Caribbean region made it possible to propose the hypothesis of an introduction in Guadeloupe of genotypes pre-adapted to resistance from Cuba, which could explain the high levels of disease observed in resistant varieties. This study provides first elements and a methodology to understand the adaptation of fungi to quantitative resistance.