| Most studies of plant disease epidemiology have focused on epidemics in host populations composed of a single genotype. But crop plant diversity can be a useful means of improving the deployment of resistance genes. In a series of experiments with wheat stripe rust and potato late blight, we studied factors influencing the effects of host diversity on epidemics. For wheat stripe rust, we considered the influence of host density and the composition of host mixtures on stripe rust severity. The intermediate planting density from our investigation experienced the greatest reduction in disease severity due to host diversity. We studied host-diversity effects on potato late blight in the USA, Ecuador, and Peru. In Oregon, USA, the level of outside inoculum was low enough that we could manipulate the spatial pattern of inoculum through introducing inoculum in a general pattern throughout an experimental plot or in only one corner of a plot. In single-genotype plots, disease severity was greater for the general inoculum pattern than for the focal pattern. The estimated host-diversity effect was greater for the general inoculum, though the difference between inoculum patterns was not statistically significant. In Ecuador, host-diversity effects were more variable; at one site they were comparable to the results in Oregon, but at the other two sites there was little or no effect of host diversity on the severity of late blight. Smaller host-diversity effects in Ecuador probably result in part due to higher levels of outside inoculum from year-round potato production. In Peru, two areas with different degrees of seasonality were selected for comparison. Sites near Cajamarca were nearer the equator and lower in altitude; sites near Huancayo were further south, higher in altitude, and more seasonal. Host-diversity effects for reduced potato late blight were generally greater near Huancayo. In a meta-analysis of all our studies of late blight, there was generally a greater host-diversity effect for reduced disease for sites where we predicted lower outside inoculum levels and for mixtures with greater differences in disease resistance between components. |
