Weather-driven infection events as indicators of severe wheat stripe rust

Plant pathologists, agronomists and crop consultants are often called on to help growers evaluate in-season disease risk and determine if management actions are needed to suppress disease development. Tools that help disease specialists and other practitioners better evaluate the influence of environment on disease development could help guide these recommendations. The objectives of this research were to identify weather conditions influencing the infection of wheat by stripe rust ( Puccinia striiformis f.sp. tritici) in field environments, to develop models estimating the probability of infection based on environment, and to evaluate if the predicted infection events were an indicator that weather conditions were favorable for severe stripe rust development in commercial wheat fields. The influence of environment on infection was studied using a bioassay approach that involved exposing inoculated, susceptible wheat seedlings within a wheat field environment. Following the exposure, the plants were incubated with a diurnal temperature cycle of 12-20°C and stripe rust incidence and severity visually evaluated after 14 days. The bioassay was conducted 116 times over three wheat growing seasons. Infection was coded as a binary variable with ≥1% incidence considered an infection event and logistic regression used to model the probability of infection based on weather during the exposure period. Multiple predictive models of stripe rust infection in field environments were developed based on the bioassay results. Several of the identified models correctly classified days as favorable for infection with more than 80% accuracy. The infection models confirmed the importance of extended periods of high relative humidity and favorable temperature on the infection processes of P. striiformis f.sp. t ritici and provided guidance on specific thresholds that were conducive for infection in field environments (e.g. > 8 h of RH ≥ 87% and a temperature between 5-25°C). The experimental approach also helped evaluate the influence of wind speed and rainfall events on the infection process that are difficult to accurately simulate in controlled environments. The resulting infection models were combined with simulations of wheat growth and used to estimate the risk of severe stripe rust developing at locations throughout Kansas (n=104). This analysis identified that the accumulation of infection events between jointing and flowering stages of growth were influential on the development of severe disease in research plots and commercial wheat fields. Models estimating the risk of severe disease correctly classified 79% of disease observations collected over 17 years. The timing of these predictions relative to crop growth and disease development could help disease specialists evaluate and communicate the risk of stripe rust in the future.