Drought is the second major constraint to common bean (Phaseolus vulga~s L) production after disease. This study examined yield under drought, yield potential, drought susceptibility index, harvest index, and geometric mean as potential indicators of drought resistant genotypes. The performance of two common bean populations, consisting of 78 and 95 recombinant inbred lines, was examined under moisture stress and nonstress regimes. Experiments were conducted at seven locations (1990-1994) in Michigan and Mexico to identify effective selection criteria for drought resistance. Two genotypes from each population yielded in the top 10% under both stress and nonstress conditions. Heritability estimates for yield in the Sierra/AC1028 population, based on 5 yr of data, ranged from 0.55 to 0.59 for stress and nonstress, respectively, and from 0.20 to 0.19 for stress and nonstress, respectively, in the Sierra/Lef-2RB population. Heritability for plant biomass was 0.52 for stress and 0.55 for nonstress in the Sierra/ AC1028 population and 0.15 under stress and 0.05 under nonstress in the Sierra/Lef-2RB population. One-hundred seed weight was the most highly heritable trait in both populations with heritability estimates of 0.80 for the Sierra/AC1028 population and 0.65 for the Sierra/Lef-2RB population. The geometric mean of the two moisture regimes was the single strongest indicator of performance under stress and nonstress, and a breeding strategy that involves selection based first on the geometric mean, followed by selection based on yield under stress, was suggested as the most effective strategy to improve drought resistance in common bean. S IXTY PERCENT of common bean production worldwide is grown under water stress, making drought the second largest contributor to yield reduction after disease (Singh, 1995). One of the largest production areas the world is the Mexican highlands (1800-2200 masl), where more than one million hectares of common bean are planted annually. Ninety-eight percent of this region is subjected to intermittent rainfall, and much of this area does not receive sufficient moisture for optimum performance (annual precipitation 200-400 mm). Drought in this region is categorized as intermittent stress where rainfall and/or drought can occur at any time during the growing season. This type of stress is typical of the semiarid tropics (Ludlow and Muchow, 1990). However, drought stress in the semiarid highlands is not aggravated by high temperatures inherent to the tropics (Acosta-G and White, 1995). Common bean is well

No abstract

Culinary quality in dry bean (Phaseolus vulgaris L.) depends on attributes of seeds prevailing at harvest and is determined by the genetic architecture of cultivars and by unpredictable environmental factors. Variation among genotypes for culinary quality has been shown to be heritable; however, the efficacy of selection depends on a knowledge of the genetic control of the measured traits. A diallel mating design was used to estimate the combining ability of parents and determine the inheritance of nine culinary quality traits important to processors and consumers. Genetic variability among eight parents, 56 F2, and 56 F3 progenies was confirmed by significant mean squares from analyses of variance. Significant variability detected between F2 and F3 progenies for soaked bean weight (SBWT), soaked bean water content (SBWC), and clumps (CLMP) was due to inbreeding effects. General combining ability (GCA) components were highly significant and overshadowed specific combining ability (SCA) components in the F2 and F3 for SBWT, SBWC, split beans (SPLT), and the washed-drained weight coefficient (WDWTR), indicating that additive variance predominated. Ratios of GCA: SCA components were equal to or less than unity for CLMP, washed-drained weight (WDWT), and texture (TEXT), indicating that both additive and nonadditive effects contributed to trait expression. Significant SCA effect variances were noted for `Sanilac', `San Fernando', `Nep-2', and `A-30' for WDWT and TEXT, implying that progeny from crosses of these parents had higher or lower mean values for the traits titan the average expected on the basis of GCA. Graphs of the regression of Vr on Wr showed that genes controlling WDWT and TEXT were completely dominant in most cases. Recurrent selection, which seeks to concentrate favorable alleles with additive effects in populations, may he an effective breeding procedure to improve the culinary quality of dry beans. It is not feasible to breed for TEXT and WDWT simultaneously because of a negative correlation between the traits.

of mineral nutrition on cooking quality of lentils. Can. J. plant sci. 5g: 165-168. A pot experiment was conducted on lentils to study the influence of major and trace elements on the-cooking quality of a hard cooliing lentil. cooking

Tannin in dry bean (Phaseolus vulgaris L.) seeds has been associated with low protein availability and digestibility. Protein losses may occur during soaking and cooking and lessen the potential nutritional benefit from eating beans. The aim of the present study was to evaluate general combining ability (GCA) and specific combining ability (SCA) effects on tannin content and protein characteristics of dry and cooked beans. A fixed‐effect model permitted the evaluation of parents for use in breeding programs to improve bean nutritional quality. Eight parents were crossed in diallel to produce the F2 and F3 generations on which combining ability mean squares and genetic effects were calculated. Tannin was determined by the vanillin‐HCl method and protein of dry and cooked beans was measured using an infrared reflectance instrument. Protein loss during soaking and cooking was determined by subtracting the protein in cooked beans from that in dry beans. Significant differences were detected among parents and progeny for all traits, indicating the presence of genetic variability. Mean squares for both GCA and SCA were significant, indicating the presence of additive and dominance effects in the populations. Ratios of GCA:SCA mean square components were greater than unity for all traits, indicating that GCA predominated. Maternal effects influenced tannin content. The distribution of points representing the parental arrays along the regression line of the Wr, Vr graph revealed that genes with partial dominance controlled tannin content. Genes for protein content of raw and cooked beans showed partial dominance in some cases and complete dominance in other cases. Genes controlling the protein loss trait had the property of complete dominance. Seed protein contents of raw and cooked beans were correlated but raw bean protein and tannin were not. Study results suggested that it is not feasible to simultaneously breed for multiple traits, except for seed protein of raw and cooked beans. Screening for raw bean protein and tannin content should be done independently.