1Danish Institute of Agricultural Sciences, Department of Plant Protection, Flakkebjerg , DK-4200 Slagelse, Denmark. 2Department of Plant Biology and Biogeochemistry, Risø National Laboratory, DK-4000 Roskilde, Denmark

Background and objectives
In Denmark over the last 15 to 20 years, the most commonly used fungicides for controlling powdery mildew have been propiconazole and fenpropimorph. The rates at which these have been applied have been reduced over the last ten years. There has been much debate about the effects of using reduced rates on evolution of fungicide resistance. A project was started to investigate the effect of reduced doses of fungicide on population structure and the rate of development of fungicide resistance. In order to speed up the process of evolution and contain the pathogen population, the experiment was carried out in the laboratory.

Materials and methods
The initial barley mildew population was made up of samples collected using stationary spore traps, before fungicides had been applied to spring cereals. The initial population was split into smaller populations, which were propagated for 31 generations in boxes (1m x 1m x 1m) in a glasshouse (150C). The populations were grown on barley treated with a spray of Corbel (750 mg/l fenpropimorph) with two dosage rates (0.3 mg/l and 5 mg/l, the doses which killed 50% and 95% of sampled mildew populations). An untreated control was also included in the experiment. Single colony isolates were taken from the start population, after 15 generations and at the end of the experiment (31 generations). These were tested for their fenpropimorph sensitivity and virulence spectrum.

Results and conclusions
The results show that using different doses of fenpropimorph have different effects on mildew populations. Fungicide resistance increased in both treatments, but this increase was greater in the populations treated with the higher dose. Differences could also be seen in the population structures. In the high dose treatment, the population became dominated by a single pathotype, whereas the other populations were more diverse.
The results suggest that fenpropimorph dose only has a small effect on resistance in a population so any effect of a strategy of applying reduced doses of fungicides like fenpropimorph would only be seen after many years. This correlates well with what is seen in the fields. Monitoring in Denmark 1992-1996 (and also many other European countries) has shown that there has only been a small change in fenpropimorph sensitivity in barley powdery mildew, in spite of intensive sprayings.
Optimising dose rates to control mildew resistance should be a balance between keeping the doses high enough to keep the amount of mildew at an economically acceptable level, and minimising the rate at which resistance builds up in the pathogen population.