The vegetable Moinordica charontja L., CucurbitaCeae, is known variously as bitter gourd, balsam pear, bitter melon, bitter cucumber, and African cucumber (Heiser 1979). Although it has many culinary uses, especially in south, southeast and east Asia, it is also grown as an ornamental and is used extensively in folk medicine (Heiser 1979). The fruits are cooked with other vegetables, stuffed, stir-fried, or added in small quantities to beans and soups to provide a slightly bitter flavor. However, for most food preparation, fruits are blanched, parboiled, or soaked in salt water before cooking to reduce the bitter taste. In addition to frying or cooking (e.g., for curries), the fruits can be dehydrated, c

Cucumber (Cucumis sativus L.) is a major cucurbit vegetable species whose genetic base has been drastically reduced during its domestication. The crop's narrow genetic base (3-12% DNA polymorphism) has resulted from the use of limited genetic material and intense selection during plant improvement. Recently, however, interspecific hybridization has been successful in Cucumis via mating of C. hystrix Chakr. and C. sativus, which resulted in the amphidiploid C. hytivus. We report herein a marker-assisted strategy for increasing genetic diversity in cucumber through introgression backcrossing employing C. hytivus. The comparatively late-flowering but high-yielding, indeterminate, monoecious line WI 7012A (P 1 ; donor parent) derived from a C. hytivus 9 C. sativus-derived line (long-fruited Chinese C. sativus cv. Beijingjietou) was initially crossed to the determinate, gynoecious C. sativus line WI 7023A (P 2 ; recurrent parent 1), and then advanced backcross generation progeny (BC 2 ) were crossed with the gynoecious indeterminate line WI 9-6A (P 3 ; recurrent parent 2). More specifically, a single F 1 individual (P 1 9 P 2 ) was backcrossed to P 2 , and then BC progeny were crossed to P 2 and P 3 , where markerassisted selection (MAS) for genetic diversity (8 mapped and 16 unmapped markers; designated Sel) or no selection (designated NSel) was applied to produce BC 3 P 2 (Sel) and BC 3 P 3 (Sel), and BC 2 P 2 (NSel) and BC 2 P 2 S 1 (NSel) progeny. Relative vegetative growth, number of lateral branches (LB), days to flowering (DF), yield (fruit number), and fruit quality [as measured by length:diameter (L:D) and endocarp:total diameter (E:T) ratios] were assessed in parents and cross-progeny. DF varied from *20 (BC 3 P 2 Sel) to *25 days (BC 2 P 3 Sel) among the populations examined, where progeny derived from P 2 possessed the shortest DF. Differences in cumulative yield among the populations over six harvests were detected, varying from *8 fruits per plant in BC 3 P 2 (Sel) to *39 fruits per plant in BC 2 P 3 (Sel). Although the vigorous vegetative growth of line P 1 was observed in its backcross progeny, highly heterozygous and polymorphic backcross progeny derived

Two cucumber recombinant inbred lines (RILs) differing in plant habit were crossed and progeny self-pollinated to produce F 3 individuals upon which phenotypic selection was practiced to identify a base population which in turn underwent either two cycles of MAS or random mating without selection (RAN). MAS and RAN were practiced to produce F 4 and F 5 progeny sets. RIL, crossing parents, and F 3 -F 5 progeny sets were then evaluated under replicated field conditions for fruit yield and quality (L:D and E:T) to evaluate gain from selection (DG). The broad-sense heritability (h 2 B) over cycles (C) of selection ranged 0.22-0.45, 0.09-0.20, and 0.11-0.15 for yield, L:D, and E:T, respectively. Although one cycle of PHE selection followed by MAS was effective in conserving the performance of the traits examined during inbreeding, progeny performance during RAN fluctuated (F 4 -F 5 generation; C 2 ). Lack of DG during advanced generations (F 4 -F 5 ) of MAS was likely due to allelic fixation and/or optimized epistatic complementation.

No abstract