Temporal patterns of airborne conidia of Podosphaera leucotricha, causal agent of apple powdery mildew

In each of the five years 1969 and 1971-1974 inclusive a volumetric spore trap was used in an apple orchard to monitor changes in the number of airborne conidia of Podosphaera leucotricha , the causal agent of apple powdery mildew. The number of trapped conidia varied greatly between years. Time-series analyses, using autoregressive integrated moving average (ARIMA) models, revealed that the temporal pattern of the number of airborne conidia was similar in all years, generally following a diurnal pattern with an afternoon peak. A strong correlation between consecutive hourly counts indicated that the number of trapped conidia depended on the strength of sporulating sources. Using the time-series transfer function (TF) method, it was shown that in each year the most important weather variables influencing the number of airborne conidia were vapour pressure deficit (VPD) and rainfall. Variation between years in the dynamic effects of these variables on conidium numbers was detected, and may reflect weather differences between years. Stepwise regression analysis was applied to the combined daily data for 1973 and 1974 using a subset of weather variables as independent variables, chosen on the basis of TF analysis. A resulting regression model accurately predicted the temporal pattern of conidium numbers (expressed as a percentage of the maximum daily number trapped in the same year) in both years. When this model was applied to the other three years there was good agreement between predicted and observed temporal patterns. Application of this regression model for practical disease control is discussed.