2012 Annual Report
In western Oregon, daily mean temperature measurements in hop yards, grass fields, and vineyards were little affected by time resolution (15 or 60 min collection intervals). However, sensor placement at standard height compared to in-canopy had larger impacts on temperature estimates in grass seed canopies than in hop yards and vineyards. While virtual station estimates of standard height daily mean temperatures recorded hourly were well correlated with observed values, virtual station estimates of rain and relative humidity had weaker correlations with respective observed values. These errors caused differences in the disease risk indexes calculated from observed and virtual weather data, especially if the index rules involved rain. However, overall management recommendations were nearly identical for powdery mildew indexes for hop or grape when calculated with observed verses virtual weather data. As a result, overestimation of rain led to increased predicted risk levels for hop downy mildew and consequently a tendency for increased recommended treatments. For grass seed, estimation errors for leaf wetness reduced infection risk calculations in a stem rust simulation model, leading to an underestimation of stem rust hazard. Field tests were conducted during 2 years that happened to have relatively unfavorable weather for stem rust development; in these cases there was little effect on management decisions or disease outcomes of the error from the virtual weather inputs to the epidemic model. Because the rust decision aid is based on a simulation model, it is additionally possible to estimate the effects of input errors by simulation. A beta group of producers and/or their crop advisers was assembled to trial a web interface. Individualized training and focus groups were conducted during beta-testing in 2011 and early 2012. For hops, beta group users also made presentations at field events to other growers and crop consultants to aid in dissemination and uptake of the technology. Individual training sessions were conducted with beta users at least three times during the season to assist the users in interpreting disease model outputs and to obtain feedback on system improvements. Model outputs were disseminated weekly via social media to producers, crop advisers, and other stakeholders. Testing materials and scenario analyses were developed and given to participants over time to document changes in awareness of factors favoring downy mildew and powdery mildew diseases, disease management decision making, and grower confidence resulting from these educational activities. Five commercial hop yard and vineyard experimental plots were monitored in 2010 and 2011 for disease development along with collection of site specific weather data. Additional field experiments to validate powdery mildew control efficacy with on-site vs. virtual weather data inputs to the hop and grape powdery mildew risk indexes and stem rust simulation model are underway in 2012. Grass seed crop consultants have provided feedback on use of the stem rust model website operated with actual and virtual weather data. This research was conducted in support of objective 2C of the parent project.