Evaluation of Dollar Spot Predictive Models on Bentgrasses in New Jersey.

The incidence of dollar spot disease (caused by Sclerotinia homoeocarpa F.T. Bennett) varies among bentgrass (Agrostis spp.) species and cultivars. The objective of this field study managed as a fairway turf was to assess the reliability of two weather-based models for predicting dollar spot epidemics on bentgrasses that range in susceptibility. ‘Independence’, ‘Penncross’, ‘Shark’, ‘007’, and ‘Declaration’ creeping bentgrass (A. stolonifera L.), and ‘Capri’ colonial bentgrass (A. capillaris L.) were seeded in a randomized complete block design with 25 blocks in North Brunswick, NJ on 29 Sept. 2014. All plots were inoculated with S. homoeocarpa isolates NJDS003 and NJDS007 on 7 Apr. 2015; plots were not inoculated in 2016 or 2017. Disease incidence was assessed every 2- to 18-d and used to assess the accuracy of a growing degree day (GDD, base 15°C and biofix 1 April) model for predicting the onset of disease symptoms and a logistic regression model for predicting disease progress. An evaluation of the ability of the logistic regression model to predict disease activity during mid- to late-2016 was not feasible due to unintended dollar spot suppression from fludioxonil used to control anthracnose. The onset of disease symptoms in highly susceptible cultivars occurred at 73-, 27-, and 92-GDD during 2015, 2016, and 2017, respectively; whereas, disease onset occurred at 79-, 140-, and 112-GDD for tolerant cultivars. The logistic regression model reached a 20% risk index at 7-, 7-, and 21-d before disease onset in highly susceptible cultivars during 2015, 2016, and 2017, respectively; whereas, a 20% risk index occurred at 11-, 29- and 28-d before symptoms on tolerant cultivars. The logistic regression model accurately forecasted disease progress in susceptible cultivars throughout 2015, early-2016, and 2017 (to date). Interestingly, periods of disease recovery often occurred while the risk index was declining, albeit greater than 20%.