THE INTERACTIONS BETWEEN POPULATIONS OF WHEAT CULTIVARS AND WHEAT POWDERY MILDEW RACES Yl LI N ZHOU 1, SHI MAI ZENG2 1 Institute of Plant Protection, CMS, 100094, Beijing, China; 2 Department of Plant Protection, China University of Agriculture, 100094, Beijing, China Background and objecuves Wheat powdery mildew (Blumeria graminis f. sp. tritici) is one of the most important diseases in wheat in China. The most practical, economic and effective means for controlling this disease is to cultivate miIdew resistant cultivars. But when one resistant cultivar is grown extensively the frequency of corresponding vimlence in the pathogen population often increases rapidly because of the selection pressure of the host cultivar. Finally, it leads to the "breakdown" of the resistant gene. The purpose of this study is to establish the model of populations interactions between wheat cultivars and powdery mildew races and to forecast the changes of the races or the virulence of the pathogen. The information of the prediction of pathogen virulence is essential for the breeding of wheat resistance, the deployment and alternation of cultivars and the strategy of the disease management. Materials and methods An albino isolate, WC-9, of wheat powdery mildew, kindly provided by Dr. J. Brown, John Innes Centre, UK, was utilized in the experiments for its white coloured colony is a good morphological marker. A: Two isolates (WC-9 and E19) were inoculated individually onto three susceptible cultivars, as well as the mixed spores of WC-9 and E19 with different ratio onto the same three susceptible cultivars separately. B: The mixed spores of three combinations of WC-9 and E02, E14, E19 separately onto one same cultivar, and one combination of WC-9 and E02 onto three cultivars, respectively. C: Mixed inoculum of WC-9 and E02 onto three cultivars mixed in the different ratios, respectively. The leaf segments of all cultivars tested were laid on water agar containing 60ppm benzimidazole in Petri dishes. The dishes were then placed in growth chamber on a14 hour photoperiod and transferred to produce the next generation 10 days after incubation, successively for 5-7 generations. The colony number of every isolate of population in the different generations was recorded. SAS was used for the statistics and modeling of the data. Results and conclusion The results of experiments of two single races separately on three susceptible cultivars showed that the increase of a single race population between successive generations was obviously limited by the host tissue. The fitting of the Richard model to the increase curve of a single race population was better than that of Logistic model. The author therefore derived a competition model (RBCM) based on the Richard's model. The data obtained from the experiments of two mixed races separately on the three cultivars were fitted with RBCM, showing that competition existed between races. The fitting of RBCM was apparently better than that of Lokta-Voltera competition model[1]. The data of selection of one cultivar to three combinations of races and that of two cultivars to one same combination of races was analyzed. The effect of selection of single cultivar to a mixed population composed of two races was clarified. It is ascertained that the change of race ratio in successive generation populations was determined by two parameters - parasitic fitness (W) of the races on the host cultivars and the carrying capacity (K) of the host cultivars to races. The experiment data of two mixed races on host of three cultivars mixed in different ratios indicates that the change of race ratio in population was also related to the ratio of the different host cultivars apart from W and K. In accordance with the above results and based on RBCM model, a model that describes the interactions between multiple races and multiple cultivars was established. It is named as pathogen-host population interaction model (PHPIM). The model was tested using the data obtained from laboratory experiments of multiple race and multiple cultivar populations, and showed a better than that of Hovmoller's model [2]. References 1. Barrett JA, 1983. Phytopathology 73,510-513. 2. Hovmoller MS, Munk L, Ostergard H, 1993. Phytopathology 83,252-265.