Gene flow between divergent cereal- and grass-specific lineages of the rice blast fungus Magnaporthe oryzae

Abstract

Delineating species and epidemic lineages in fungal plant pathogens is critical to our understanding of disease emergence and the structure of fungal biodiversity, and also informs international regulatory decisions. Pyricularia oryzae (syn. Magnaporthe oryzae) is a multi-host pathogen that infects multiple grasses and cereals, is responsible for the most damaging rice disease (rice blast), and of growing concern due to the recent introduction of wheat blast to Bangladesh from South America. However, the genetic structure and evolutionary history of M. oryzae, including the possible existence of cryptic phylogenetic species, remain poorly defined. Here, we use whole-genome sequence information for 76 M. oryzae isolates sampled from 12 grass and cereal genera to infer the population structure of M. oryzae, and to reassess the species status of wheat-infecting populations of the fungus. Species recognition based on genealogical concordance, using published data or extracting previously-used loci from genome assemblies, failed to confirm a prior assignment of wheat blast isolates to a new species (Pyricularia graminis tritici). Inference of population subdivisions revealed multiple divergent lineages within M. oryzae, each preferentially associated with one host genus, suggesting incipient speciation following host shift or host range expansion. Analyses of gene flow, taking into account the possibility of incomplete lineage sorting, revealed that genetic exchanges have contributed to the makeup of multiple lineages within M. oryzae. These findings provide greater understanding of the eco-evolutionary factors that underlie the diversification of M. oryzae and highlight the practicality of genomic data for epidemiological surveillance in this important multi-host pathogen.