Quantitative trait loci mapping for pathotype-specific powdery mildew resistance in a diploid rose population

Rose is the most important ornamental plant worldwide. Powdery mildew (Podosphaera pannosa) is one of the most important diseases of greenhouse and field grown roses. Use of chemical fungicides to control this disease increases the product cost considerably and contributes to environmental pollution. Breeding of cultivars resistant to powdery mildew would be an efficient, economical and environmentally safe way to address the problems caused by powdery mildew.
The objective of this research was to study the genetic segregation of disease resistance towards powdery mildew in a diploid rose population. To do so, we used two monoconidial pathotypes of powdery mildew, R-E and R-P, to get more insight into the race-specific disease resistance. The diploid rose cultivar ‘Yesterday’ was crossed with a diploid wild species; Rosa
wichurana. Both parent plants vary in susceptibility to the powdery mildew isolates. The 90 progenies of this crossing were planted in the field.
Disease inoculation was done artificially with an inoculation tower under standardised conditions for the parent plants and all progeny plants. Disease symptoms were scored 10 days after inoculation for both fungal pathotypes. The rate of powdery mildew development was used to describe the variation in resistance within the population. A disease index was calculated for both pathotypes and for all plants individually. Besides, microscopy was used to evaluate plant responses upon inoculation with the powdery mildew pathotypes based on germination of the conidia, development of mycelium and plant cell reactions in the leaves. Both tests detected a wide and significant variation among genotypes for resistance. Especially for pathotype R-E on parent ‘Yesterday’ immunity was found based on a failure of fungal growth after germination of the conidia. Analysis of the data indicated that the two pathotypes differed in pathogenicity. The population was molecularly characterized to enable genetic dissection of the variation of the traits. In this study we used AFLP, SSR and morphological markers to construct a genetic linkage map using JoinMap program version 4. Two parental genetic linkage maps were generated. Seven linkage groups, corresponding to seven monoploid homologous chromosomes of roses, were found and the linkage groups were named based on anchor SSR
and morphological markers of previously published research of diploid roses. The total length of the map of the female parent (‘Yesterday’) and the male parent (R. wichurana) spanned 536 and 526 cM, respectively. The map of ‘Yesterday’ was constructed with 278 AFLP, 40 SSR and two morphological markers. The map of R. wichurana was generated with 232 AFLPs, 39 SSR
markers and one morphological marker. An integrated map was constructed consisting of 366 markers, spanning 540 cM. The maps generated in this study were very dense and the markers were evenly distributed. These maps also shared more than 20 anchor SSRs and morphological markers with previously published maps of diploid roses. Morphological markers involved two
flower characteristics: colour and petal number. This could lead to a good consensus map for diploid roses in the future and could be useful in map alignment studies. These new maps were used to identify QTLs controlling disease resistance based on the calculated disease indexes. The QTL analysis combined the Kruskal-Wallis (KW) test and Simple Interval Mapping (SIM) (MapQTL version 5.0). The results indicated that several genomic regions were involved with each of the specific monoconidial pathotypes studied. Major
pathotype-specific QTLs could also be identified. Five QTL loci were found by pathotype R-P; two QTLs on parent ‘Yesterday’ (on Linkage groups 2 and 3) and three QTLs were detected on parent R. wichurana (two QTL loci on linkage group 2 and one QTL on linkage group 5). Four QTLs were found for pathotype R-E; two QTLs on parent ‘Yesterday’ (on linkage groups 3 and 6) and two QTLs on parent R. wichurana (on linkage groups 5 and 6). All QTLs explained, in
total, 15-73% of the phenotypic variance for disease response. Finally, 77% of the QTLs were confirmed by restricted Multiple QTL Mapping (rMQM).
The QTL maps developed here are useful for future rose breeding, resistance research and development of a consensus map for roses.