Development and Evaluation of Two Pecan Scab Prediction Models

Abstract

Fusicladium effusum (syn. Cladosporium caryigenum), causal agent of pecan scab, is the most economically important pathogen of pecan (Carya illinoinensis). A weather-based advisory is currently used in Oklahoma to assess the need for fungicide application and requires the accumulation of scab hours. A scab hour is defined as an hour with average temperature ≥21.1°C and relative humidity ≥90%. To assess the validity of the thresholds in the advisory, repeated ratings of disease severity were taken on fruit at five locations during the 1994–96 and 2009–10 growing seasons, resulting in a total of eight site years. Hourly weather variables were also examined, including temperature, relative humidity, dew point, dew point depression, total solar radiation, and total rainfall. Rain and disease severity were converted to binomial variables where a rain event (≥2.5 mm) and disease severity (≥25%) were coded as 1 and all other events as 0. Logistic regression models adjusted for correlated data were developed using generalized estimating equations. Two models were developed: a temperature/relative humidity model and a dew point/dew point depression model. For the temperature/relative humidity model, the best fitting model included all main effects. Using this model, validation exercises assuming no rain and total solar radiation of 22.5 MJ m–2 resulted in a 0.45 probability of pecan scab development when the temperature was 21°C and relative humidity was 90%. Findings of this model were further validated during field studies that evaluated different combinations of temperature and relative humidity thresholds for scheduling fungicide applications. These analyses indicated that the current thresholds of temperature and relative humidity are viable, but a modification of the relative humidity component should be considered. For the dew point/dew point depression model, a reduced model, including dew point, dew point depression, and the binomial rain variable, was considered adequate for explaining scab events, which suggests that future model building to describe pecan scab epidemics should include dew point, dew point depression, rain, and total solar radiation as independent variables. This article originally appeared in the January issue, Volume 96, pages 117-123. It was changed to correct errors in a measurement conversion that appeared throughout.