1Department of Crop Protection, The University of Adelaide, Waite Campus, Glen Osmond, SA 5064, Australia; sup>2South Australian Research and Development Institute, Waite Precinct, GPO Box 397, Adelaide, SA 5001, Australia

Background and objectives
Powdery mildew of grapevine, caused by the obligately biotrophic fungus Uncinula necator, is the most economically important disease of grapevines worldwide. U. necator reproduces both sexually and asexually. Ascospores are the main source of primary inoculum for epidemics in New York State, USA, but the origin of primary inoculum in most regions has not been documented. The sexual stage was first recorded in Australia in 1984 and the relative importance of sexual reproduction in initiating infection and in generating new variants is unknown.

Our aim is to characterize the fungus using molecular and phenotypic markers to identify haplotypes (clonal lines) and to assess genetic diversity and relatedness among individuals. This information will further our understanding of factors affecting the genetic structure of U. necator populations.

Materials and methods
Samples were obtained from nine viticultural regions across southern Australia and established in vitro, as described previously [1,2]. Isolates were characterized using U. necator specific clones (one single-locus and four multi-locus) as RFLP probes. Four primers (mini and microsatellite and intron-splice junction) were used to amplify conidial DNA in PCR experiments. Genetic similarities were calculated between all pairs of isolates. A total of 59 RFLP and 27 PCR bands were scored. Phenograms of genetic similarity for the combined RFLP and PCR data were generated using UPGMA. The Shannon index (D1=[-Spi.lnpi]/lnN) was used to describe diversity for RFLP and PCR haplotypes.

Results and conclusions
Mating types are present in this collection at a ratio of 1:1 (Mat+:Mat-). The RFLP and PCR markers detected DNA polymorphisms among our collection of isolates. Phenetic analyses of banding pattern similarity coefficients revealed two broad groups sharing an average overall similarity of 40%. Most of the variation was detected within Group 1 isolates, comprising both mating types. In comparison, few differences were detected among Group 2 isolates, all of which are Mat+. Genetic variation is evident between and within viticultural regions. Individual haplotypes occur multiple times, are not restricted to any one viticultural region and, therefore, may have been introduced on infected grapevine material or via wind dispersal. Among a subset of 44 isolates, RFLP markers detected 34 different haplotypes and a corresponding diversity index of D'=0.85. In comparison, 15 phenotypes were resolved with the PCR system, resulting in a much lower diversity index of D'=0.59. The overall diversity of combined RFLP and PCR haplotypes was D'=0.88, distinguishing 37 different isolates. These results demonstrate that the RFLP probes detected higher levels of diversity than the PCR primers and highlight the need for using a range of markers for diversity studies. Future population studies will provide information on factors which affect the distribution and evolution of pathogen populations, and contribute to the development of strategies for sustainable disease control.

1. Evans KJ, Whisson DL, Scott ES, 1996. Mycological Research 100, 675-80.
2. Evans KJ, Whisson DL, Stummer BE, Scott ES. 1997. Mycological Research 101, 923-32.