Velocity of spread of wheat stripe rust epidemics

Controversy has long existed over whether plant disease epidemics spread with constant or with increasing velocity. We conducted large-scale field experiments with wheat stripe rust at Madras and Hermiston, Oregon, where natural stripe rust epidemics were rare, to test these competing models. Data from three location-years were available for analysis. A susceptible winter wheat cultivar was planted in pure stand and also in a 1:4 or 1: 1 mixture with a cultivar immune to the stripe rust race utilized in the experiments. Plots were 6.1 in wide and varied from 73 to 171 in in length. A 1.5 by 1.5-m focus was inoculated in either the center (2001) or upwind of the center (2002 and 2003) of each plot. Disease severity was evaluated weekly throughout the epidemics in each plot at the same points along a transect running upwind and downwind from the focus. Velocity of spread was calculated from the severity data and regressed separately on time and on distance from the focus. In all location-years and treatments, and at all levels of disease severity, velocity consistently increased linearly with distance, at an average rate of 0.59 m/week per m, and exponentially with time. Further, across epidemics there was a significant positive relationship between the apparent infection rate, r, and the rate of velocity increase in both space and time. These findings have important implications for plant diseases with a focal or partially focal character, and in particular for the effectiveness of rate-reducing disease management strategies at different spatial scales.