Effects of fungicide movement between plots in field experiments

Interactions between plots of spring barley which were either unsprayed, or sprayed with triadi mefon or tridemorph to control powdery mildew ( Erysiphe graminis f.sp. hordei ) were measured in experiments of balanced design in 1978 and 1979. The average amounts of mildew in untreated plots with triadimefon-reated neighbours on the predominaotly up-wind side were usually less than half those in untreated plots with tridemorph-treated neighbours on the up-wind side. Triadiniefon-treated plots had most effect on the adjacent sides of their untreated neighbours but there was evidence that their influence extended across the whole widths of these neighbouring plots. Amounts of mildew in tridemorph-treated plots were increased by untreated neighbours.
Differences in amounts of disease between these plots of spring barley with differently treated neighbours had no significant effects on grain yield. However, in a rando'mized block experiment with spring wheat in 1975 there was evidence that similar inter-plot interactions caused consider able bias in the yield data. Possible causes of the observed effects are discussed.