Relationship between frequencies of resistance gene in the host and corresponding virulence gene in the pathogen in rice blast disease.

Abstract

The relationship between resistance gene frequency in the host population and virulence gene frequency in the pathogen population on rice blast disease was analyzed by a multiple regression method to develop a method for predicting the virulence gene frequency. The virulence gene frequency was computed frorn the data on race distribution obtained in 1976 and 1980 by YAMADA and his co-workers on virulence genes, Av-a+, Av-i+, Av-k+ and Av-ta+ corresponding to the four resistance genes. Pi-a. Pi-i. Pi-k and (Pi-ta+ Pi-ta2). The resistance gene frequency for the four resistance genes was calculated from growing areas of leading varieties in each prefecture during ten years before the examination of virulence gene frequencies. By a multiple regression analysis of the pooled data on four gene pairs for resistance and virulence, the following equation was obtained, as an equation to predict frequencies of four virulence genes (F). F=0.081+ O.24M1-0. 028Dy+0.71Da+2.38M2Dt+1.26M2Dk-5.82M22Di-4.07DyS1Dk. Here M1 and S1 are the mean and slope (increasing rate) of resistance gene frequencies during the period from the sixth to tenth year (the former five years) before prediction year. M2 is the mean of the five years from the first to fifth year (the latter five years) before prediction year. Dy is the dummy variable for years. Da, Di and Dk are dummy variables for Pi-a, Pi-i and Pi-k gene frequencies, respectively. Incorporation of virulence gene frequency of a past year in the regression analysis increases the accuracy of prediction as evaluated by contribution ratio.