The wheat blast pathogen Pyricularia graminis-tritici has complex origins and a disease cycle spanning multiple grass hosts

The wheat blast disease has been a serious constraint for wheat production in Latin America since the late 1980s. We used a population genomics analysis including 95 genome sequences of the wheat blast pathogen Pyricularia graminis-tritici (Pygt) and other Pyricularia species to show that Pygt is a distinct, highly diverse pathogen species with a broad host range. We assayed 11 neutral SSR loci in 526 Pygt isolates sampled from wheat and other grasses distributed across the wheat-growing region of Brazil to estimate gene flow, assess the importance of sexual reproduction, and compare the genetic structures of Pygt populations infecting wheat and nearby grasses. Our results suggest a mixed reproductive system that includes sexual recombination as well as high levels of gene flow among regions, including evidence for higher gene flow from grass-infecting populations and into wheat-infecting populations than vice versa. The most common virulence groups were shared between the grass- and wheat-infecting Pygt populations, providing additional evidence for movement of Pygt between wheat fields and nearby grasses. Analyses of fruiting body formation found that proto-perithecia and perithecia developed on senescing stems of wheat and other grass hosts, suggesting that sexual reproduction occurs mainly during the saprotrophic phase of the disease cycle on dead residues. Phalaris canariensis (canarygrass) supported the fullest development of perithecia, suggesting it is a promising candidate for identifying the teleomorph in the field. Based on these findings, we formulated a more detailed disease cycle for wheat blast that includes an important role for grasses growing near wheat fields. Our findings strongly suggest that widely grown pasture grasses function as a major reservoir of wheat blast inoculum and provide a temporal and spatial bridge that connects wheat fields across Brazil.Author summaryAfter the first wheat blast epidemic occurred in 1985 in Paraná, Brazil, the disease spread to Bolivia, Argentina, and Paraguay, and was introduced into Bangladesh in 2016 followed by India in 2017. Wheat blast is caused by Pyricularia graminis-tritici (Pygt), a highly diverse pathogen species related to the rice blast fungus P. oryzae, but with an independent origin and a broader host range. We conducted a large scale contemporary sampling of Pygt from symptomatic wheat and other grass species across Brazil and analyzed the genetic structure of Pygt populations. Pygt populations on both wheat and other grasses had high genotypic and virulence diversity, a genetic structure consistent with a mixed reproductive system that includes regular cycles of recombination. The pathogen formed sexual fruiting structures (perithecia) on senescing stems of wheat and other grasses. Historical migration analyses indicated that the majority of gene flow has been from Pygt populations on other grasses and into the Pygt population infecting wheat, consistent with the hypothesis that Pygt originated on other grasses before becoming a wheat pathogen. We found that the Pygt populations infecting wheat were indistinguishable from the Pygt populations infecting other grass species, including signal grass (Urochloa brizantha). Because U. brizantha is a widely grown grass pasture often found next to wheat fields, we propose that it functions as reservoir of Pygt inoculum that provides a temporal and spatial bridge that connects wheat fields in Brazil.