Past orchard performance as a guide to decisions on use of fungicides to control sooty blotch and flyspeck

Abstract

We investigate the usefulness of past apple orchard performance as a guide for making decisions concerning treatment for the diseases sooty blotch (caused by Gloeodes pomigena (Schw.) Colby) and flyspeck (caused by Zygophiala jamaicensis Mason), in the light of a previously reported decision analytic model. Applicability of three decision rules is investigated, one based directly on the model of Gold and Sutton, one based on a simple binomial model, and a third which combines features of the first two. The decision-theoretic value of information for each of the rules was evaluated using the method of cross-validation. Five years of performance data were available for each of 29 orchards in western North Carolina, and four-year training sets were drawn from these data. The combined rule gave the best overall results. The value of past information averaged over all orchards was US $32·62/ha. This is reduced to $20·84/ha, if growers will apply two sprays of fungicide to protect against late season rot diseases, independent of the decision to protect against sooty blotch and flyspeck. The value of the past information was zero for orchards producing consistently high-quality fruit, and averaged $180/ha for orchards producing consistently low-quality fruit. We also discuss the usefulness of data for more or for less than 4 years, inferences concerning risk attitudes of orchardists, and implications for general orchard management strategy.