Field and glasshouse observations of Lolium spp. grasses indicated that the lower, abaxial, leaf surface was rarely infected by powdery mildew (Erysiphe graminis) even when the upper, adaxial, surface was densely colonized. Experiments showed that conidia of two strains of E. graminis, one from Lolium and one from Avena, germinated equally well on both surfaces of Lolium and Avena leaves, but that the subsequent growth and development of germlings was impaired on the lower surface of Lolium leaves, so that most formed only multiple short germ tubes or an abnormal long tube, and only c. 25% or fewer formed infection structures. This contributes to the apparent resistance of the lower Lolium leaf surface to powdery mildew and may help to explain why the disease is relatively unimportant in UK ryegrass crops, since infection structures develop at a high frequency on only 50% of the leaf area, i.e. the upper surface. Scanning electron microscopy showed that the epicuticular waxes on the lower Lolium leaf surface form amorphous sheets. This contrasts with the crystalline plate waxes seen on the upper surface of Lolium leaves and on both surfaces of oat leaves. However, when the lower Lolium leaf surface was washed with chloroform to remove epicuticular wax, normal germling and infection structure development was obtained on the wax‐free surface. This suggests that the sheet waxes prevent the pathogen gaining access to features of the cuticular membrane which trigger normal germling development.

SummaryThe evolution of virulence in UK oat powdery mildew (Blumeria graminis f.sp. avenae) populations is presented along with comparative information on the deployment of resistant cultivars. Virulence frequencies have followed classical gene‐for‐gene principles, and there are no effective resistance genes currently deployed in cultivars grown in the UK. The incidence of powdery mildew in continental Europe and pathogen variation is reviewed as well as other strategies for the control of this disease. New resistant sources have been identified and are being used in breeding programmes throughout Europe.

Resistance was found in the meadow fescue (Festuca pratensis) to crown rust (Puccinia coronata), originating from ryegrasses (Lolium spp). A backcrossing programme successfully transferred this resistance into diploid Italian ryegrass (Lolium multiflorum) and genomic in situ hybridisation (GISH) was used to identify the introgressed fescue chromosome segment. The resistant (R) plants in two BC 3 lines all carried an introgressed segment on a single chromosome, which in one of the lines was confined to the short arm of the chromosome. Susceptible (S) plants either contained no introgressed chromosome segment or a segment which was physically smaller than the segments in resistant plants. Using GISH the resistance locus could be physically mapped to the midpoint of a short arm. Segregation ratios of the progeny of BC 3 plants, when crossed as R Â S and R Â R, were in agreement with the hypothesis that the resistance was controlled by a single gene or very closely linked genes. No R plants were produced by crossing S Â S plants.

The causative organism of crown rust in ryegrasses (Puccinia coronata f.sp. lolii) is an obligate biotroph that causes significant economic losses within the temperate grazing industries of dairy, meat, and wool production. This study reports on the development, transferability, and utility of gene-associated simple sequence repeat (SSR) molecular markers for crown rust. Analysis of 1,100 expressed sequence tag (EST) sequences from a urediniospore-derived cDNA library detected 55 SSR loci. The majority of EST-SSR arrays contained perfect trinucleotide repeats with consistently low repeat numbers, and the motifs (ACC)n and (CAT)n were most commonly represented. DNA extraction from single pustules, in conjunction with multiple displacement amplification, provided the basis for PCR-based screening to evaluate genetic marker performance. An example of the identification of intraspecific genetic diversity was obtained from the analysis of 16 P. coronata isolates originating from the United Kingdom, Australia, New Zealand, and Japan. A subset of 12 robust EST-SSR markers was informative for determination of pathogen diversity within and between these localities. It was also demonstrated that crown rust EST-SSR markers were capable of cross-amplification in closely related fungal taxa (Puccinia spp.) and filamentous fungi within the Ascomycota.

RESULTS * * NF = No fungicide; ST = Seed treatment only; ST + F = Seed treatment + foliar sprays. Means within columns followed by the same letter(s) are not significantly different (P < 0.05) according to Duncan's Multiple Range Test.* and *** indicate variance ratios for fungicide x cultivar interactions significant at P Q 0.05 and 0.001 respectively.