POTENTIAL OF WILD COMMON BEAN FOR SEED YIELD IMPROVEMENT OF CULTIVARSIN THE TROPICS

Thirty nine wild or weedy common bean (Phaseolus vulgaris L.) accessio ns, representing the two extremes of geographical range of distributio n and domestication of cultigens in the Americas, were crossed to a hi gh-yielding, small-seeded cultivar, ICA Pijao. The resulting F-1, F-2, and F-3 population bulks, along with ICA Pijao, were evaluated for se ed yield, days to maturity, and 100-seed weight under favorable growin g conditions at two locations in Colombia. The F-3 and F-4 bulks were tested separately under soil-fertility and moisture stresses. Thirty r andom Fs-derived Fs lines from the highest yielding population involvi ng both Andean and Middle American wild beans, along with the parents and a control cultivar, were also evaluated for 2 yr, under favorable conditions. No F-1 hybrid, population bulk, or F-5-derived F-8 line si gnificantly outyielded ICA Pijao in any test environment. The mean yie ld of F-1 hybrids, population bulks, and F-5-derived F-8 lines involvi ng wild and weedy beans of Middle America was higher than that involvi ng Andean South American wild beans. Heritablity for seed yield, seed weight, and days to maturity, as well as gains from selection were com parable to those obtained in crosses among cultivars. The small seed o f the progenies and the positive correlation between seed size and see d yield suggest that alternative mating schemes that increase the reco very of progenies with a cultivated phenotype should be investigated. Until this research is done, it would be premature to draw conclusions about the yield improvement potential of wild beans.