Partial resistance to downy mildew ( Plasmopara halstedii) and to black stem (Phoma macdonaldii) in sunflower were investigated under natural field infection and a controlled growth chamber respectively. Genetic control for resistance to the diseases was determined in recombinant inbred lines (RILs) and their two parents, 'PAC-2' and 'RHA-266.' The experiments were undertaken in a randomized complete block design with two replications, in a field severely infected by downy mildew and in a controlled growth chamber with plants inoculated with an agressive French isolate of P. macdonaldii. Each replication consisted of three rows, 4.6-m long, giving 48 plants per RIL or parent in the field and 15 plants in the growth chamber. Genetic variability was observed among the RILs for resistance to both diseases. When 10% of the selected RILs were compared with the mean of the two parents genetic gain was significant for partial resistance to the diseases. Four putative QTLs for resistance to downy mildew on linkage groups 1, 9 and 17 were detected using composite interval mapping. The QTLs explained 54.9% of the total phenotypic variance. Major QTLs ( dmr1-1 and dmr1-2) for resistance were found on linkage group 1 with up to 31% of the phenotypic variability explained by two peaks. QTL analysis of resistance to black stem showed seven QTLs on linkage groups 3, 6, 8, 9, 11, 15 and 17. The detected QTLs together explain 92% of the phenotypic variation of the trait. Crosses between RILs contrasted for their resistance to downy mildew and black stem, and exhibiting molecular polymorphism in detected QTLs, will be made in order to focus more-precisely on the genomic region of interest.

Crosses were made between two inbred lines of sunflower. Parents and 118 F(3) families were planted in the field in a randomized complete block design in two replications. Genetic control for some agronomical traits: grain weight by plant (GWP), 1,000-grain weight (TGW), percentage of oil in grain (POG) and sowing to flowering date (STF) was investigated in F(3) families and their parents. Genetic variability was observed among the 118 F(3) families for all the traits studied. Genetic gain was obtained when the best F(3) family, or the mean of 10% of the selected families was compared with the best parent for GWP, TWG and POG. Heritability was 0.23 for GWP, 0.55 for TGW, 0.57 for POG and 0.32 for STF. A set of 244 F(3) families from the same cross, including the above 118 mentioned families and their two parents, were screened with 276 AFLP and microsatellite markers and a linkage map was constructed based on 170 markers. Two putative QTLs for the GWP trait ( gmp), one QTL for TGW ( tgw), six QTLs for POG ( pog) and two for STF ( stf) were detected. The percentage of phenotypic variance explained by each QTL ranged from 2.6% to 70.9%. The percentage of total phenotypic variance explained was 50.7% for GWP, 5.4% for TGW, 90.4% for POG and 89.3% for STF. Although these regions need to be more-precisely mapped, the information obtained should help in marker-assisted selection.

A population of 77 recombinant inbred lines (RILs) were developed through single-seed descent from a cross between 'PAC-2' and 'RHA-266'. Seeds of the above-mentioned RILs and their parents were planted in the field in a randomised complete block design with two replications. Genetic control for some agronomical traits-sowing-to-flowering date, plant height, stem diameter (SD), head diameter (HD), grain weight per plant, 1,000-grain weight (TGW) and the percentage of oil in grains-were measured for RILs and their parents. Genetic variability was observed among 77 RILs for all traits studied. Transgressive segregation occurred for some traits, and the comparison between 10% of selected RILs with the best parent showed significant difference for SD and HD as well as for TGW. A set of 123 RILs from the same cross, including the 77 above-mentioned RILs and their two parents, were screened with 409 AFLP and SSR markers, and a linkage map was constructed based on 367 markers. Several QTLs associated with the studied traits were identified. The effects of each QTL are moderate, ranging from 7% to 37%, but a high percentage of phenotypic variance is explained when considering all the covariants (TR2 mean around 80% in each trait). Although the detected regions need to be more precisely mapped, the information obtained should help in marker-assisted selection.

No abstract