Comparative analysis of response to phenotypic and marker-assisted selection for multiple lateral branching in cucumber (Cucumis sativus L.)

Yield increase in processing cucumber (Cucumis sativus L.) is positively correlated with an increase in number of fruit-bearing branches. Multiple lateral branching (MLB) is a metric trait controlled by at least five effective factors. Breeding efficacy might be improved through marker-assisted selection (MAS) for MLB. Experiments were designed to independently confirm previously determined linkage of molecular markers (L18-2-H19A SNP, CSWTAAA01 SSR, CSWCT13 SSR, W7-2 RAPD and BC-551 RAPD) to MLB, and to determine their utility in MAS. These markers were present in significantly higher frequency than expected (1, presence:3, absence; p < 0.001) in BC2 plants selected based on a high MLB phenotype (BC2PHE). However, markers that were considered selectively neutral fit the expected segregation of donor parent DNA in BC2 progeny. Markers linked to MLB were used in MAS of BC1 and BC2 plants to produce BC(2)MAS, and BC(3)MAS progeny. Means for MLB in MAS populations were compared with backcross populations developed through phenotypic selection (BC2PHE, BC3PHE) and by random mating where no selection had been applied (BC2RND, BC3RND). Statistical analysis showed no significant differences (p < 0.001) between means of phenotypic (BC2PHE = 3.02, BC3PHE = 3.29) and marker-aided selection (BC(2)MAS = 3.12, BC(3)MAS = 3.11) for MLB. However, both phenotypic and MAS population means were significantly higher than the random control (BC2RND = 2.27, BC3RND = 2.41) for MLB. Thus, given the observed response to selection and the rapid life-cycle of cucumber (4 months), markers linked to MLB when used in MAS will most likely be effective tools in cucumber improvement.