Effects of wheat cultivar mixtures on epidemic progression of Septoria tritici blotch and pathogenicity of Mycosphaerella graminicola

The effects of host genotype mixtures on disease progression and pathogen evolution are not well understood in pathosystems that vary quantitatively for resistance and pathogenicity, We used four mixtures of moderately resistant and susceptible winter wheat cultivars naturally inoculated with Mycosphaerella graminicola to investigate impacts on disease progression in the field, and effects on pathogenicity as assayed by testing isolate populations sampled from the field on greenhouse-grown seedlings. Over 3 years, there was a correspondence between the mixtures' disease response and the pathogenicity of isolates sampled from them. In 1998, with a severe epidemic, mixtures were 9.4% less diseased than were their component pure stands (P = 0.0045), and pathogen populations from mixtures caused 27% less disease (P = 0.085) in greenhouse assays than did populations from component pure stands. In 1999, the epidemic was mild, mixtures did not reduce disease severity (P = 0.39), and pathogen populations from mixtures and pure stands did not differ in pathogenicity (P = 0.42). In 2000, epidemic intensity was intermediate, mixture plots were 15.2% more diseased than the mean of component pure stands (P = 0.053), and populations from two of four mixtures were 152 and 156% more pathogenic than the mean of populations from component pure stands (P = 0.043 and 0.059, respectively). Mixture yields were on average 2.4 and 6.2% higher than mean component pure-stand yields in 1999 and 2000, respectively, but the differences were not statistically significant. The ability of mixtures challenged with M. graminicola to suppress disease appears to be inconsistent. In this system, host genotype mixtures evidently do not consistently confer either fitness benefits or liabilities on pathogen populations.