QTL Mapping And Analysis Of Fruit And Flower Related Traits In Cucumber (Cucumis Sativus L.)

Cucumber (Cucumis sativus L.) is one of the most important vegetable crops grown wildly in the world. So far most studies on the cucumber mainly focus on the processing type in the west. In China, cucumber is utilized mainly by fresh-eaten, so it is very important to research the commercial characters of fruit-appearance traits. The study would provide much information for the future marker-assisted selection (MAS) breeding.The cucumber monoecious, indeterminate, and dull fruit skin line S94 (Northern China type) was crossed with the gynoecious, indeterminate line S06 (Northern European type) possessing small spines and uniform immature fruit color. The S94 plants normally produce relatively long fruit and stalk, and exhibit low female flower ratios. In contrast, S06 has shorter fruits and stalks, and possess comparatively high female flower ratios. Two hundred and fifty two F2 plants were produced from a single F1 plant. The F2 plants were then self-pollinated to obtain 130 F3 families. A total of 173 markers (116 SRAPs, 33 RAPDs, 11 SSRs, 9 SCARs, 3 ISSRs and 1 STS) was grouped into seven large linkage groups at a LOD≥3.0. The map spanned 1014 cM with an average marker interval of 5.9 cM. The markers were distributed relatively evenly, with the longest linkage group spanning 240.2 cM (LG1), and the shortest spanning 71 cM (LG7). Using the F2 population and its F3 derived families, QTL detection was going for nine fruit-related traits: fruit weight, length, and diameter. Fruit flesh thickness, seed-cavity diameter, fruit-stalk length, fruit pedicel length, length/diameter and length/stalk ratio. The results were as follows: 1. A total of 46 QTLs were detected in F2(16), F3/autumn(18) and F3/spring(28) using WinQTLcart2.5, and each trait had QTLs with more than 10% explanation of phenotypic variation; 2. The QTLs for different fruit traits were detected in the same region (i.e. LG1: ME5SA18 nearby; LG2: F(f) nearby; etc), and the correlations between these traits indeed; 3. Several QTLs for different season/generation combination were detected stably with relatively high r2 (i.e. fw2.1, fl4.1, fd1.1, fft1.3, scd1.1, fsl4.1, fpl1.2, ldr4.1 and ldr5.1); 4. The detection stability of WinQTLcart2.5 is higher than QTLmapper1.6; QTL identification is more precise in F3 families than in F2 individuals.Through self-pollination by single-seed descent (SSD), 224 F2S6 recombinant inbred lines (RILs) and 224 F6:7 families were obtained. More sequence-anchored makers (SSR&SCAR) reported and developed recently were added in the new map, so the linkage map was much more saturated. It contained a total of 254 DNA markers (206 SRAPs, 22 SSRs, 25 SCARs and 1 STS) besides ss, D and u and spanned 1006 cM with an average marker interval of 3.9 cM. Some segregation distortion regions (SDR) were found with clustered skewed markers in favor to a certain parent. A total of 48 anchor markers (SSR, SCAR and STS) were distributed on all seven groups (6 on LG1; 16 on LG2; 5 on LG3; 4 on LG4; 5 on LG5; 7 on LG6; 5 on LG7). By the shared anchor markers, our LG2 should be matched to LG1 of Fazio's map and our LG6 have potential corresponding relationships with LG4 of Fazio's.Based on the new RIL linkage map, the QTL analysis of the fruit and flower traits was done in F7 families between two seasons. The results were as follows: 1.Sixty-two QTLs were detected for nine fruit related traits in F7/autumn and F7/spring. Fruit stalk length had the most QTLs (9). QTL fsl4.1 had the highest r2 values (r2 = 14.3% and 25.9% and LOD = 11.9 and 19.8 in autumn and spring, respectively) among the entire fruit trait QTLs. 2.A total of 16 QTLs for the flower related traits (female flower ratio, first flower node, and first female flower node) were detected in F7/autumn and F7/spring. The QTL sex2.1 neighboring the F locus had the highest variance (r2 = 60.2%, 40.2%; LOD = 72.9 and 84.4 in autumn and spring, respectively). 3. Of fruit-related traits, there were 25 loci being detected in both seasons. For each of these QTLs, its explained phenotypic variances and its additive effect directions were consistent and uniform, respectively, between the two seasons.Using the molecular markers tightly linked to targeted agronomic traits is an efficient breeding approach for crops improvement, especially for the improvement of some complex traits. In this study, several sequence-anchor markers (CSWCT25, CS30, CMBR41, CS08 etc.) were closely linked with the stable QTLs with high phenotypic variation. Because the fresh-eaten cucumber types were chosen as parental materials, which are different from U.S. processing types, so the information of this study would be more helpful to improve precocity and fruit shape traits in cucumber breeding in the future.