Analysis of the effects of selected gronomic factors on the dynamics of the take‐ all disease of wheat in field plots

Linear models were used to investigate the effects and interactions of treatments for the control of disease in multifactorial experiments on the dynamics of the take‐ al disease of wheat. Analyses were based on repeated measures of disease on second successive winter wheat crops in three successive se sons. The treatments were date of sowing, presence or absence of triadimenol‐containing seed treatment, autumn nitrogen and chloride applications, timing of early spring nitrogen and total spring nitrogen applications. The effects of treatments were investigated on line and quadratic trends of disease increase and on the integral amount of disease over time. Different disease variables were used to express disease relative to the total and diseased plant populations. The effects of treatments on absolute and relative numbers of diseased roots per plant were also analysed. During three se sons when disease was never severe, the effects of the treatments were shown to be variable. Late sowing consistently reduced disease during the early stages of the epidemic (in the winter months). The effect of late sowing in reducing disease, however, persisted through to the end of the growing se son in only one of the 3 years of the study. autumn nitrogen application and Baytan seed treatment (containing triadimenol) also reduced disease in some seasons. The presence of an effect for either was dependent on sowing date and was most evident on the proportion of diseased roots relative to the diseased plant rather than the total plant population.