Plant Biology and Biogeochemistry Department, PBK-301, PO Box 49, DK-4000 Roskilde, Denmark

Background and objectives

Changes over time in multilocus virulence genotype frequencies of an aerial spore pool in an isolated geographical area were modelled for an airborne pathogen [1]. The basic model assumes that cultivars with different resistance genes are sown at the same time, over specified areas, and that spores are randomly dispersed on the host genotypes according to area. Selection on the virulence genotypes takes place when a new crop is emerging and on a specific host only spores with the corresponding virulence genes are able to grow. The pathogen reproduces asexually with no mutations. The frequency of multilocus virulence genotypes in the aerial spore pool at the end of a growing season can then be expressed as a function of the frequencies at the beginning of the growing season and the area of different host genotypes. The total amount of disease can be estimated from the number of spores in the initial aciial spore, pool and their average fitness on the crop. This basic model has been extended in different ways to include sexual reproduction [2], to consider cultivar mixtures, to include overlapping growing seasons (spring and autumn sown crops) and to consider quantitative host-pathogen interactions [1].

Results and conclusions
The modelling has given an overall understanding of the population dynamics of pathogens for which selection due to host plant resistance genes is the most important evolutionary force. The model has predicted the generation of gametic disequilibria among virulence genes, positive or negative, depending on the use of different resistance genes in different cultivars. This prediction was largely in accordance with observations in Danish populations of barley powdery mildew [l]. The genetic disequilibria generated results in changes in frequencies of unnecessary virulence genes owing to 'hitch hiking', in differences in virulence gene frequencies between spores in the aerial pool and in specific host plant genotypes, and they influence the interpretation of complexity of multilocus virulence genotypes. For virulence surveys, the modelling has stressed the importance of assaying multilocus virulence genotypes and of sampling at the right time in relation to crop emergence. Further, the model has enabled different resistance gene diversification schemes to be compared, e.g., different uses of resistance genes in autumn- and spring-sown cultivars.

1. Hovmoller MS, Ostergard H, Munk L, 1997. In: The Gene-for-Gene Relationship in Plant-Parasite Interactions (Eds Crute IR, Holub EB, Burdon JJ). Wallingford, UK: CAB International, pp. 173-190.
2. Ostergdrd H, Hovmoller MS, 1991. Plant Pathology 40, 166-177.