Using a Logistic Regression Model to Advise Fungicide Sprays for the Control of Dollar Spot of Creeping Bentgrass Putting Greens.

Dollar spot, caused by Sclerotinia homoeocarpa, is the most damaging disease of closely mown creeping bentgrass and annual bluegrass swards.  Repeated fungicide applications are required to suppress dollar spot to maintain acceptable playing conditions.  A reliable dollar spot prediction model would be useful for managing dollar spot on these turfgrasses.  Logistic regression was used to develop a model that input weather variables to predict probability of dollar spot development [P(Y)] on creeping bentgrass putting greens and fairways at sites in Oklahoma and Wisconsin.  The model inputs included 5-day moving averages of minimum air temperature [MNT], average relative humidity [RH], and a class variable for fungicide [FUNG] use [1 = fungicide used; 0 = no fungicide used] to describe P(Y). According to the best model, average 5-day temperatures above 57°F were conducive for the development of dollar spot.  As temperature increased above that point, the probability of disease development slowly decreased.  However, dollar spot development was still possible during periods when temperatures were as high as 85°F.  Conversely the probability of dollar spot increased with increasing average relative humidity.  When temperatures were between 57°F and 85°F, 5-day average relative humidity values of 70% or above were considered sufficient for dollar spot development. Independent validation indicated that an action threshold of 30% probability of dollar spot development was required to provide adequate fungicide protection when using the model as a spray advisory in OK and WI.  Using this threshold, the model correctly identified warm/hot, dry periods, which were considered of low risk for the development of dollar spot and no fungicide sprays were advised at either location.  In 2009, the advisory would have saved three fungicide sprays in Oklahoma and two fungicide sprays in Wisconsin if it were used to advise actual fungicide applications.