Abstract
The effect of host [Avena sativa] genotype unit area (the ground area occupied by an individual, independent unit of the host population that is genetically homogeneous) on the effectiveness of oat mixtures in reducing epidemic development of crown rust [Puccinia coronata] was studied in two types of epidemics. The initial inoculum was confined to a single focus in the first, and the same amount of initial inoculum was distributed uniformly in the second. Plots with 100% pure-line susceptible plants were compared with those having two plant population mixtures, each consisting of 25% susceptible and 75% immune plants. The positions of the two genotypes were completely random in one mixture. In the other mixture, host genotype unit area was increased by aggregating 200 seeds of like genotype into randomly positioned blocks within plots. When the plots were artificially inoculated in a single focus in 1984, both mixtures reduced epidemic development relative to the pure-line susceptible check and there was a relatively small difference in the effectiveness of the two mixtures. When the same amount of initial inoculum was distributed uniformly over the plots in 1984, the completely random mixture greatly reduced rust development relative to the pure-line check, but the mixture of large genotype units did not. A similar result was obtained in 1983 when a natural spore shower resulted in an apparently random initial distribution of disease in the plots.