Résumé
To keep abreast of the current genetic diversity of the populations of Magnaporthe oryzae, the rice blast disease pathogen, we characterized the genetic diversity of 453 monoconidial isolates recently collected from different geographic areas in the Philippines in experimental plots and in farmer fields. With 13 microsatellite markers, a total of 93 new genotypes were determined when compared to the reference isolates representing the genotypes previously published. Analysis revealed a weak geographic structuring of pathogen genotypes, with some of those collected from Luzon clustering with those from Visayas or Mindanao. Nevertheless, none of the strains from Mindanao clustered with those from the Visayas or vice versa. Significant differences were noted on isolates collected from distinct agroecosystems and isolates collected from the upland tended to be more diverse when compared to isolates from lowland. This structure likely results from limited natural migration and from some events of long distance migration, probably through the transport of infected seeds. A subset of 30 isolates representative of the genotypic diversity were tested for fertility status. Of these, eight induced the production of perithecia, of which six also were only male-fertile and two were also female-fertile. These fertile isolates were mostly found in the upland agroecosystem. In addition, the same isolates and 23 additional isolates were subjected to PCR test for mating type. MAT1.1 isolates represented 36% of our sample whereas it represented only 5% in the collection of reference strains. This result confirms that this new sampling improves our knowledge of the blast population in the Philippines. Pathogenicity test of a subset of 24 isolates inoculated to 16 differential rice varieties demonstrated that most isolates were avirulent to varieties C101Lac and C104Lac which carry the Pi1 resistance gene. The very popular Pi33 gene, integrated into the varieties IR 64, Bala, and IR1529 also conferred resistance to most of the isolates. In addition to Pi1 and Pi33, other possible candidate resistance gene for plant breeding effort to manage the rice blast disease in the Philippines are Pi9, Pikm, and the combination ofPiz and Pish. PCR amplifications of cloned avirulence gene were also carried out. Data analysis is in progress and results will be presented. (Texte intégral)