Ishiguro, Kiyoshi*. A strategy for preventing the spread of super races in the blast fungus population with multiline varieties of rice. In: International Meeting on "Population and Evolutionary Biology of Fungal Symbionts", Ascona, Switzerland, 2007. AB-22. Click here to download the presentation on pdf format.

The blast disease of rice (Oryza sativa) is caused by a fungal pathogen, Magnaporthe oryzae. Several genes conferring high resistance to the disease have been identified. These resistance genes and corresponding avirulence genes of the fungus comply with the gene-for-gene relationship. Using a blast-resistant multiline (ML) variety (a mixture of near-isogenic lines (NILs) each of which possesses a different resistance gene) has been believed to be a classic idea to use these resistance genes more durably. More than 10 years has passed since ML varieties were released in Japan. While the planted area of the varieties has been limited, no disease-resistance breakdown has yet been observed. However, risks of the spread of super races of the pathogen are undeniable. Meanwhile, the management strategy for preventing super race from the spread is still uncertain. Consequently, the policy to determine the number of mixed NILs and the mixture ratio of NILs has been concluded based on anecdotal evidence. Preventing dominance of the super race in the pathogen population is equivalent to maintaining the virulence gene polymorphism in the population. A traditional explanation of the polymorphism in the population is that the virulence genes have their fitness costs. However, experimental evidence to demonstrate the cost has been yet poor. If the cost is slight (if not significant), durability of ML varieties relies not on the costs but on the other driving forces. Random genetic drift and migration may be the plausible forces. Our recent findings suggest that the blast fungus population in the current Japanese rice production area has a very narrow bottle neck between crop seasons due to introduction of advanced field sanitation and qualified seed production systems. The presence of bottle neck usually increases the effect of random genetic drift. Also, our field observations showed that the pathogen has exponential spore dispersal gradients and that most of the spores land within 1 km distance from the inocula in each airborne infection cycle. The migration of pathogen may contribute to temporal change of the genotype frequency in the population. Accordingly, sensitivity analyses of the effect of bottle neck in the pathogen population, dispersal gradient of the pathogen, and the number of mixed NILs in ML variety upon the spread of super races in the population using a spatially explicit lattice-structured model. In the model, we assumed the pathogen with multilocus avirulence genes. Although some parameters in the model lack the reliable values, the sensitivity analyses showed: (1) the narrower bottle neck inhibits the spread of super races (2) the steeper dispersal gradient makes the population structure unstable, and (3) ML varieties that consist of more mixed NILs delays the development of super races. Consequently, mixing maximum-available NILs in a ML variety and efforts to narrow the bottle neck should be essential for preventing the breakdown.

Keywords: random genetic drift, virulence polymorphism, bottle-neck.

*Institution: MAFF AFFRC, Japan

E-mail: ishiguro@affrc.go.jp