lUniversity of Reading, Department of Agricultural Botany, 2 Earley Gate, Reading RG6 2AU, UK; 2 AgrEvo UK Limited, Chesterford Park, Saffron Waiden, Essex CB10 1XL, UK

Background and Objectives
There is a lack of understanding of the dynamics of resistance development to fungicides and the many interacting factors that could influence the short and long term survival of resistant fungal isolates in field populations. Field resistance to DMI fungicides is known to have occurred in Erysiphe graminis populations and is a potential risk with Septoria tritici. Studies have attempted to measure shifts in sensitivity in naturally developing field populations of S. tritici [1], but the natural variability in fungicide sensitivity of these populations has prevented the measurement of significant shifts.
The objectives of this work were to determine whether significant shifts in fungicide sensitivity could be measured in inoculated populations of S. tritici and how the strength of selection is influenced by dose and physio-chemical properties of DMI fungicides.

Materials and Methods
In 1996 and 1997 field plots of spring wheat of the cultivar Baidus were spray inoculated at GS 13 with S. tritici isolates of known fungicide sensitivity. Four isolates of similar pathogenicity were used, two from each end of the flutriafol sensitivity spectrum. Uninoculated plots were included to determine if a naturally occurring S. tritici population developed. In 1996 single spray applications of one of four DMI fungicide formulations chosen for their physio-chemical properties were compared for effects on selection for resistance. Flutriafol and fiuquinconazole were chosen for their systemic activity. Two experimental formulations of fiuquinconazole, produced to provide differing rates of fungicide uptake, were used. Prochloraz, a non-transiocated compound, was chosen to contrast with flutriafol and fluquinconazole. All fungicides were applied at half the recommended field dose at GS 39-47. In 1997 two fungicide doses and a split application were included in the treatment programme. In both years plots sprayed with water plus Tween 80 served as standard for comparison of treatments. A visual assessment of S. tritici severity was made just before fungicide application and at weekly intervals thereafter. Selection for resistance was measured using a discriminating dose assay to determine the difference between the ratio of susceptible to resistant isolates in each field plot before and after fungicide application. The strength of selection was compared for each treatment on each of the top three leaves of the crop canopy.

Results and Conclusions
In 1996 and 1997 all fungicide treatments significantly reduced S. tritici infection. The amount of reduction varied with fungicide. Before fungicide application the proportion of susceptible to resistant isolates was almost equal. After fungicide application in 1996 significant shifts in the populations' sensitivity were detected with all four fungicides. The strength of selection for resistance varied with fungicide and position of the infection in the crop canopy. On leaf three, flutriafol and both fluquinconazole formulations caused significant shifts towards a more resistant pathogen population, but shifts in plots treated with prochloraz did not differ significantly from those in water treated plots. On the upper two leaves of the crop canopy the strongest selection for resistance occurred in plots treated with tlutriafol. In contrast to the results on leaf three prochloraz caused significant selection for resistance. These results can be explained by the physio-chemical properties of each fungicide and spray distribution in the crop canopy. The changes in resistance levels found in both years and a discussion of the possible causes of the effects will be presented in the full poster.

1. Pijls CFN, Shaw MW, 1997. Plant Pathology 46, 247-263