Pathogenic and molecular variability in a population of Mycosphaerella graminicola, cause of septoria tritici leaf blotch of wheat

Mycosphaerella graminicola, cause of Septoria tritici leaf blotch of wheat, is an important disease throughout the world. Information about the population structure of the pathogen is useful to improve control strategies. Molecular studies have shown that there was a high level of genetic variability within populations of M. graminicola, but no attempt was made to relate this variability to pathogenicity. The objectives of this research were to study a population of M. graminicola for both variability at the molecular level and pathogenicity and to determine whether any relationship between these two types of variability exists. Ninety isolates of M. graminicola were collected using hierarchical sampling of leaves and lesions from 10 locations within a wheat field near Saskatoon. The isolates were tested for the degree of pathogenicity (aggressiveness) on a single susceptible cultivar and the components of pathogenicity, incubation period, latent period and disease severity, were evaluated. There were significant differences among isolates for all components at the lesion sampling level only but not at the leaf and location level. A subsample of 40 isolates was tested for variability for virulence on a set of six differential cultivars. A significant isolate x cultivar interaction was detected, but since the magnitude of the variability was low no attempt was made to classify the isolates into races. Using 15 RAPD primers the percentage of polymorphic loci, number of molecular phenotypes (haplotypes) and gene diversity of 90 isolates were estimated. A high level of genetic variability was found within the population. Partitioning this variability into different components showed that most of the variability was distributed within locations. This type of distribution suggested that air-borne ascospores were the primary source of inoculum in the field. A similar type of distribution of genetic variability was detected using eight microsatellite markers. Little relationship between molecular and pathogenic variability was found, suggesting that DNA fingerprinting has little value for monitoring the development of new virulent genotypes of the pathogen.