1L Daehnfeldt A/S, Breeding Station Danefeld, 5290 Marsiev, Denmark; 2Department of Plant Biology, and 3Department of Agricultural Sciences, Royal Veterinary and Agricultural University, 1871 Frederiksberg C, Denmark

Background and objectives
The biotrophic oomycete Peronospora parasitica (Pers. ex. Fr.) Fr. causes downy mildew of brassicas. Seedlings of horticultural brassicas raised in seedbeds for later transplanting are frequently lost due to infection. Seedling resistance to P. parasitica is a means to secure plant survival. Partial, quantitative disease resistance by definition results in reduced host invasion without complete protection [1], and is generally anticipated to be based on several genes. It is considered to be a possible source of more durable resistance, which is less isolate-specific than resistance based on single genes [1]. Partial resistance to P. parasitica has previously been reported at the seedling stage in gene bank accessions of broccoli [2]. This study describes characteristics of partial resistance to P. parasitica within 20 double haploid broccoli lines from a wide range of breeding material, and presents results from a half diallel cross study of the inheritance of this resistance.

Materials and methods
8-day-old seedlings of 20 dihaploid broccoli lines were screened for resistance to three isolates of P. parasitica by drop inoculation of the cotyledons. The cotyledons were assessed for resistance by a visual sporulation score, and by determination of conidial production. Two lines, exhibiting high levels of resistance, were screened with 13 isolates to further evaluate the isolate-specificity of the resistance. Six of the broccoli lines, representing different levels of resistance, were crossed in a half diallel. The offspring and parents were evaluated for resistance, and the results tested for additive and non-additive genetic effects by Hayman's analysis.

Results and conclusions
All broccoli lines supported sporulation of the pathogen to a varying extent and intensity. Analysis of variance of sporulation score as well as conidia production showed highly significant differences between the genotypes. Two lines exhibited the highest level of resistance in both assessments, but quantitative differences were also present in the remaining material. Partial resistance of the most resistant lines resulted in 50-70% reduction of conidia production in comparison to the most susceptible lines. This may be crucial for the survivial of young seedlings under critical conditions. All lines exhibited a low and non-significant degree of isolate specificity towards three isolates from Denmark, England and Mozambique, respectively. Inoculation of the two lines exhibiting the highest level of resistance with 13 isolates of different geographical origins showed that one line exhibited some isolate specificity, while the other line exhibited a rather uniform level of resistance to all isolates. A low degree of isolate specificity is a highly desirable feature for the use of partial resistance as a supplement to more isolate-specific types of resistance, because it may be built up in the basic breeding germplasm for subsequent use in many different growing environments. Analysis of the diallel revealed that additive genetic effects were dominating the inheritance of the resistance. This strongly indicates that recurrent selection for resistance in early generation inbreds or in populations of well adapted broccoli will be efficient. Seedling screening with P. parasitica can be performed routinely during most breeding operations as a simple and inexpensive means for elimination of the most susceptible genetic combinations.

1. Parleviet JE, 1992. In Stalker HT, Murphy JP, eds, Plant Breeding in the 1990s. CAB International, Wallingford, UK, pp. 281-302.
2. Thomas CE, Jourdain EL, 1990. HortScience 25, 1429-1431.