Printer Friendly

Marker-assisted selection to improve drought resistance in common bean.

The recent trend in genetic mapping has led to the elucidation of the chromosomal position of many quantitative trait loci (QTL) in several important agronomic crops. Among others (Lee, 1995), Edwards et al. (1987) identified significant associations between isozyme loci and 82 quantitative traits in two F2 populations of maize (Zea mays L.). Using the same hybrid crosses, Stuber et al. (1987) further identified isozyme loci associated with QTL influencing 25 yield-related traits. Edwards et al. (1992) identified restriction fragment length polymorphism (RFLP) markers from a highly saturated genomic map that were associated with the same quantitative traits evaluated by Stuber et al. (1987). They concluded that associated RFLP loci corresponded to those locations already identified by the isozyme markers. The RFLP, however, are more informative than isozymes because they are more abundant. Paterson et al. (1991), with an RFLP map of tomato (Lycopersicon esculentum L.), identified six QTL influencing fruit mass, four for soluble solids concentration and five for fruit pH in an interspecific back-cross. The QTL controlling gray leaf spot (Cercospora zea-maydis Tehon & E. Y. Daniels) resistance in maize also have been mapped (Bubeck et al., 1993) and Beavis et al. (1994) have identified QTL influencing 24 other agronomic traits in maize. Freyre and Douches (1994) have mapped QTL controlling specific gravity and tuber dormancy in potato (Solanum tuberosum L.) with RFLP, isozymes, and RAPD. Martin et al. (1989) have identified three RFLP loci associated with C isotope discrimination in tomato, and Nodari et al. (1992) identified RFLP loci associated with common blight [Xanthomonas campestris pv. phaseoli (Smith) Dye] resistance in common bean.

Since discrete regions of the genome can be identified that contribute to a significant portion of quantitative trait variation, selecting for these regions based on marker genotype may prove effective in improving a particular quantitative trait. Stuber and Edwards (1986) demonstrated that MAS was as effective as phenotypic selection to improve quantitative traits such as grain yield, ear height, and ear number, despite the fact that the markers used to identify QTL represented only 40% of the genome. This suggests that, with a more saturated map, MAS could be more effective than conventional, phenotypic selection (Stuber and Edwards, 1986). Although similar responses were observed between phenotypic and genotypic selection, a threefold increase in marker frequency occurred in the populations selected on the basis of marker genotype. This result suggests that there are unmapped regions of the genome for which marker-based selection would be ineffective (Stuber and Edwards, 1986). If, however, a breeder were to combine genotypic selection with phenotypic selection, breeding efforts to improve quantitative traits could be accelerated (Bubeck et al., 1993).

Soller and Beckmann (1990) stated that MAS for quantitative traits with high heritability would not be as efficient as conventional breeding. The predictive value of MAS is inversely proportional to the heritability of that trait (Paterson et al., 1991). The heritability for yield performance under stress in common bean is low to moderate ([h.sup.2] = 0.19 to 0.59; Schneider et al., 1997), indicating that if markers were identified that explained a proportion of the variation in a population, MAS would facilitate selection. For a quantitative trait with high heritability, MAS could still be effective after the major QTL are fixed and the heritability is reduced (Paterson et al., 1991). Although markers associated with QTL controlling quantitative traits have been reported (Lee, 1995), limited research has been conducted regarding the effectiveness of MAS to improve quantitative traits (Stromberg et al., 1994).

It was our objective to (i) identify significant associations between the polymorphic RAPD markers and yield performance under moisture stress and nonstress conditions in the absence of an effective linkage map; (ii) compare the effectiveness of MAS to conventional phenotypic selection for drought resistance based on actual data in the same population; and (iii) determine if the identified markers are useful in populations other than the one from which the markers were identified.


Genetic Stocks

Two drought resistant Mexican breeding lines (AC1028 and Lef-2RB) and one adapted Michigan pinto cultivar, Sierra, exhibiting drought resistance, were selected as parents to create two recombinant inbred populations. Sierra pinto was crossed to both AC1028 and Lef-2RB in 1986. Two F3 plants from each of 50 [F.sub.2]-derived families from both crosses were advanced by single seed descent to the F5 generation from 1987 to 1990 using nurseries in both Michigan and Puerto Rico. Seed from individual [F.sub.5] plants was bulked to create 78 [F.sub.5]-derived recombinant inbred lines (RIL) in the Sierra/AC1028 population (S/ A) and 95 [F.sub.5]-derived RIL in the Sierra/Lef-2RB population (S/L).

Field Study

Field experiments designed to study drought resistance were planted in seven environments across five years. Both populations were grown at the following locations: Montcalm County, Michigan, 1990; Ingham County, Michigan, 1991; Madero, Durango, Mexico, 1992, 1993, and 1994; and Calera, Zacatecas, Mexico, 1993 and 1994. Genotypes from each population were grown as separate experiments under two treatments (stress and nonstress) with two replications per treatment. The S/A population was grown in a 9 X 9 lattice design, whereas the S/L population was grown as a 10 X 10 lattice (Schneider et al., 1997). At the Montcalm County location, stress was induced by planting along a 2 to 6% slope. Additionally, Seafarer, a navy bean cultivar, was planted between rows to compete for moisture. The nonstress treatment at this location lacked the interrow competition of Seafarer. For the 1991 Ingham County experiment, stress treatments were grown under a movable rain shelter that covered plots during periods of precipitation and at night. The nonstress treatment received normal rainfall. At all Mexican locations, the stress treatment received normal rainfall, whereas the nonstress treatment was irrigated as needed. A detailed description of these experiments has been reported previously (Schneider et al., 1997). A combination of pre-emergence herbicide and mechanical and hand cultivation was used to control weeds. Agronomic practices were applied to ensure normal crop growth and development.

An additional field experiment was conducted in 1994 under a rain shelter at the Kellogg Biological Station, Kalamazoo County, Michigan. Nineteen selected genotypes from S/A and all three parents were planted and analyzed as a split-plot design with four replications. Stress treatments (stress and nonstress) were considered main plots with genotypes as subplots. The genotypes within each treatment were separated, and stress and nonstress treatments were analyzed as randomized complete-block designs, separately, with four replications. The 19 genotypes were selected based on molecular marker data associated with yielding ability under water stress (explained in the following section). Plots were planted by hand in a Spinks sand (sandy, mixed, mesic psammentic hapludalf) in single rows, 3 m long and 0.5 m wide, on 15 June with a density of 13 plants [m.sup.-1]. The shelter remained open and plots received normal rainfall until 50% of the plants had at least one flower, after which the rain shelter closed automatically after sensing 15 mm of continuous rain. Nonstress plots were irrigated by overhead sprinklers once weekly with 13 mm of water for 14 wk (for a total of 178 mm). Plots were maintained weed free. During the first week of October, plants were hand pulled and number of pods per five plants and days to maturity were recorded. Biomass was measured, plants threshed, and yield per plot and 100 seed weight were recorded.

Marker Protocol

DNA Extraction

Several young trifoliolates from each F4,5 plant and parent genotypes were harvested, lyophilized, and ground. Ground tissue was stored at -80 [degrees] C. The DNA was extracted following the protocol described by Saghai-Maroof et al. (1984) and modified by Miklas et al. (1993). The protocol was altered slightly as follows: chloroform/isoamyl alcohol (24:1) was used instead of chloroform/octanol, RNase A was added prior to the second chloroform/isoamyl alcohol extraction, and DNA was resuspended in 0.1 M Tris Acetate ethylenediaminetetra-acetic acid (EDTA) buffer, pH 8.0.

When extra DNA was required, a second mini-extraction was performed following the procedure outlined by Edwards et al. (1991) and modified according to Haley et al. (1994b). All samples were quantified by DNA fluorometry (Hoefer TKO100, Hoefer Scientific, San Francisco, CA). A sub-sample was then diluted to a standard concentration of 10 ng [Micro][L.sup.-1] for amplification by the polymerase chain reaction (PCR).

Polymerase Chain Reaction Protocol

The PCR procedure followed that of Miklas et al. (1993) with slight modifications (Haley et al., 1994a). Random primers were used from selected Operon kits (Operon Technology, CA) to amplify random regions of the genome. The DNA was amplified using a Perkin Elmer Cetus DNA Thermal Cycler 480 (Perkin Elmer, Cetus, Norwalk, CT) using the following cycles: 3 cycles of I min at 94 [degrees] C, 1 min at 35 [degrees] C, and 2 min at 72 [degrees] C; 34 cycles of 1 min at 94 [degrees] C, 1 min at 40 [degrees] C, and 2 min at 72 [degrees] C (final step extended by 1 s for each of the 34 cycles); and a final extension cycle of 5 min at 72 [degrees] C (Haley et al., 1994a).


Approximately 20 [Micro]L, of amplified DNA from each sample were run on a 1.4% agarose gel containing ethidium bromide (0,5 [Mu]g m[L.sup.-1]), 40 mM Tris-acetate, and 1 mM EDTA. The DNA was viewed under ultraviolet light and photographed for permanent record.

Marker Scoring and Nomenclature

Six-hundred random decamer primers were screened against the parents of each population to detect polymorphisms. Seventy polymorphic primers were selected to use for RAPD analysis on the RIL from each population. The resulting marker data were scored as a 1 for RIL that lacked the RAPD band and a 2 for RIL that possessed the band. Nomenclature for RAPD followed that of Miklas et al. (1993). For those markers that were not sized and exhibited more than one polymorphism, lower case letters were used in alphabetical order starting with the longest RAPD from that primer.

Marker Identification

One-way analyses of variance were performed for yield data treating marker genotype as a classification variable for each population using the GLM procedure in SAS (SAS Institute, 1988). Performance data consisted of mean yield under stress (Yd), mean yield under nonstress (Yp), and geometric mean (GM) of yield under the two stress treatments for each location from 1990 to 1993 as well as mean Yd, Yp, and GM averaged across all locations. Associations were determined by F-tests with significance at P [is less than] 0.05.

MAPMAKER/EXP (Lincoln et al., 1987) "group" command was used to identify linkage groups among markers. Default linkage criteria were set at a distance of 10 centimorgans (cM) and a likelihood of the odds ratio (LOD) score of 4.0 so that marker groups would not generally exceed 25 cM. All RAPD from each resulting linkage group were treated as independent variables in a multiple regression analysis with Yd, Yp, and GM as dependent variables. These analyses were performed on data from each environment and the combined analysis using the GLM procedure in SAS (SAS Institute, 1988). The following F-test was performed to confirm the utility of the entire set of markers for predicting yield ([y.sub.i]) values:


where [Beta]'X'Y is the sum of squares for the fitted parameters, n is the number of individuals, p is the number of parameters or markers used, y. is the mean yield for the experiment, and [ss.sub.e] is the sum of squares for the error term (Gill, 1993, p. 298-319). Significance was set at P [is less than] 0.10.

Markers in Sierra/AC1028

Based on these data, four markers were identified that appeared to be associated with QTL controlling drought resistance in S/A. The RAPD markers from linkage group nine ([OH19.sub.690] and [OAB18.sub.650]; Fig. 1) combined with unlinked markers, [OF01.sub.520] and [OH18.sub.710] were chosen as indirect selection criteria to improve performance under drought in S/A. The RAPD [OH19.sub.690] is associated with genotypes performing below average, whereas RAPD [OAB18.sub.650], [OF01.sub.520], and [OH18.sub.710] are all associated with genotypes performing above average. Therefore, RIL that lacked the [OH19.sub.690] band but possessed the [OAB18.sub.650], [OF01.sub.520], and [OH18.sub.710] bands were chosen as potentially above average genotypes (PAA). Those RIL with the [OH19.sub.690] band present but lacking the [OAB18.sub.650], [OF01.sub.520], and [OH18.sub.710] bands were designated as potentially below average genotypes (PBA). Ten RIL (12.8% of the population) were identified in the PAA group, whereas nine RIL (11.5% of the population) were classified into the PBA group. These markers were subsequently used to select genotypes within S/A that potentially differed in their performance under water stress. The 19 selected genotypes were planted with the parents, Sierra, AC1028, and Lef-2RB, At Kalamazoo County in 1994 (described above) to determine if significant yield differences existed between genotypic groups selected based on marker genotype. The means of the two genotypic groups were contrasted (1 df) using the data from Kalamazoo County. In addition, the PAA and PBA were contrasted using yield data under stress and nonstress conditions from the combined 1994 analysis for S/A experiments conducted in Madero and Calera, Mexico. Significance was set at P [is less than] 0.05 and environments considered as random effects.

Markers in Sierra/Lef-2RB

Sixty-one RIL were indirectly selected based on their marker genotype in S/L. Five markers were used to select PAA, which was compared to PBA. The RAPD from linkage group one in S/L (Fig. 2) were used to select 35 RIL (36.8% of the population) as PAA and 27 RIL (29.5% of the population) as PBA. Linkage Group 1 contained five markers: [OA08.sub.780], [OZ08.sub.750], [OA04.sub.560], [OX11.sub.680], and [OX18.sub.980]. The RAPD markers [OA04.sub.560], [OX11.sub.680], and [OX18.sub.980] were all associated with below average performance, whereas RAPD markers [OZ08.sub.750] and [OA08.sub.780] were associated with above average performance. The RIL that lacked the [OA04.sub.560], [OX11.sub.680], and [OX18.sub.980] bands but had the [OZ08.sub.750] and [OA08.sub.780] bands were included in the PAA group. Contrasts were performed to determine if significant differences existed between the two groups. The groups were contrasted in stress and nonstress environments for yield, combined across two locations in Mexico (Madero and Calera) in 1994. Significance was set at P [is less than] 0.05, and environments were considered random effects.

Efficiency of Markers Detected in Sierra/Lef-2RB Compared with Conventional Selection

To compare the efficiency of conventional selection to MAS at the same selection intensity, additional contrasting groups were selected based on previous performance for Yd, Yp, and GM. An above average performing group (AA) of RIL, corresponding to the same number of genotypes as PAA, and a below average group (BA) of RIL, with the same number of genotypes as PBA, were selected based on Yd data combined across four years (1990-1993) and five locations in Michigan and Mexico. Similar contrasting groups were identified and selected based on Yp and GM. With the groups identified based on conventional selection, the means of AA and BA were contrasted (1 df) within each selection criterion (Yd, Yp, and GM), with 1994 yield data from stress and nonstress experiments combined across two locations, Madero and Calera. Thus, the three AA groups selected based on Yd, Yp, and GM data sets were contrasted against the three BA groups in stressed and nonstressed environments, independently. Significance was set at P [is less than] 0.05.

Markers Detected in Sierra/Lef-2RB and Tested in Sierra/AC1028

The five RAPD markers used for MAS in S/L were further analyzed in S/A. Four of the five markers were polymorphic between Sierra and AC1028, and these four RAPD were used to identify extreme genotypic groups, PAA and PBA in S/A. Comparisons between PAA and PBA were performed on yield data combined across the two Mexican locations in 1994. The same selection intensity that was used to create PAA and PBA groups was used to compare contrasting groups, AA and BA selected for Yd, Yp, and GM based on yield data at five locations (1990-1993). Comparisons were performed to determine if AA significantly out-yielded BA with data combined across two 1994 Mexican locations. Significance was set at P [is less than] 0.05.


Marker Analysis

No single marker was associated consistently with any of the yield traits across all locations and years. In fact, no RAPD appeared significant in more than three out of seven environments (data not shown). This can be explained by the strong environmental influence typical of a quantitative trait such as drought resistance. Relative yield differences due to a significant genotype X environment interaction vary dramatically between environments (Schneider et al., 1997). Thus, a marker may explain a significant proportion of the variation for yield with data from one location but not explain a significant proportion in another. Data from a single experimental season are not adequate to determine associated markers that will be useful predictors in future growing seasons. It follows that a single marker representing one locus also is not sufficient to identify stable associations. QTL identification is limited by genetic resolution relative to markers, suggesting that QTL may be placed with an acceptable degree of certainty only within 15 to 20 cM (Lee, 1995). To account for the lack of resolution, a region of the genome must be analyzed against yield data from several locations. As observed in interval mapping, a curve representing LOD scores will be distributed across a region of the chromosome with its maximum as the most likely position for the QTL on that map (Lander and Botstein, 1989).

Markers from each population were analyzed using MAPMAKER/EXP (Lincoln et al., 1987) to identify linkage groups within each population (Fig. 1 and 2). Markers within each group were analyzed as a multiple regression against data from individual years and the combined data from all years to determine if a region of the genome could better explain the variation for yield performance (Tables 1 and 2). Of the nine linkage groups in S/A, all showed significant associations with at least one yield trait in at least one environment and the coefficients of determination ranged from 0.08 to 0.22 (Table 1). Linkage Group 9 was significantly associated in more environments than any other linkage group. In S/L, all linkage groups were significantly associated with at least one yield trait in at least one environment and the coefficients of determination ranged from 0.04 to 0. 18 (Table 2). Linkage Group 1 was significantly associated in nine out of the 18 (6 environments X 3 yield traits; Yd, Yp, and GM) possible analyses. To account for the strong environmental influence on yield performance and the difficulty in ascertaining the exact location of QTL within a range of [is less than] 15 to 20 cM, a region of the genome marked by several markers explained a higher degree of variation compared with single regression analysis.


Multiple regression is used when more than one factor is thought to influence a particular trait (Gill, 1993, p. 298-319). Since the distance for accurately identifying the location of a QTL ranges from 15 to 20 cM, the markers within each linkage group spanning from 4.8 to 50 cM (Fig. 1 and 2) were incorporated into a multiple regression equation to determine if, when added together, a better predictive model could be developed. F-tests were analyzed to confirm the utility of the set of markers. Additionally, analysis was performed to determine which RAPD within the set of markers were useful. However, of the markers used for MAS in S/L and S/A, all were found to be effective in different environments. For example, in Linkage Group 1 of S/L, markers [OZ08.sub.750] and [OA08.sub.780] were useful in predicting Yd in 1990. However, in the 1993 Madero location, [OA04.sub.560] and [OX11.sub.680] were the most useful markers. Thus, the entire region encompassed by the five markers from Linkage Group 1 was used for selection in S/L. For these same reasons, analyses of variance with single RAPD markers did not result in consistent associations across locations. A multiple regression analysis with all RAPD markers within a linkage group better explained the variation for this trait in all locations and represented a larger region of the genome that is required to compensate for QTL positioning problems (Lee, 1995).


Marker Assisted Selection in Sierra/AC1028

Multiple regression analysis detected specific areas of the genome that were consistently significant across locations. For S/A, Linkage Group 9 (Fig. 1) was significantly associated with Yp, Yd, and GM in both 1990 and 1991 Michigan locations, 1992 Madero location, and in the analysis combined across all locations. In all cases, these markers could explain a moderate amount of the genetic variation ([R.sup.2] = 0.08 to 0. 14; Table 1). When markers were analyzed independently in one-way analyses of variance, however, each marker in Linkage Group 9 was significantly associated in only one environment. Markers in this group were chosen as a selection criteria for drought resistance along with two other unlinked markers ([OF01.sub.520] and [OH18.sub.710]).

To test the effectiveness of the MAS method in S/A, RIL selected based on markers [OAB18.sub.650], [OH19.sub.690], ([OF01.sub.520] and [OH18.sub.710] were grown in stress and nonstress treatments under a rain shelter in Kalamazoo County, Michigan in 1994. As demonstrated in previous marker studies (Stuber and Edwards, 1986; Stromberg et al., 1994), differences between genotypes identified as PAA and those that are PBA were significant except for biomass and number of pods per plant under stress conditions (Table 3). Mean yield of PAA was improved compared with the mean of PBA by 10 g [m.sup.-2] under stress and 33 g [m.sup.-2] under nonstress conditions (Table 3). A significant improvement was observed for PAA compared with PBA supporting our hypothesis that selection can be performed with associated markers.


All genotypes from S/A were planted in Madero and Calera in 1994. These data were used to test the effectiveness of MAS with experiments conducted in Mexico. The same genotypes were contrasted with combined data across the two Mexican locations. Significant differences were not detected between PAA and PBA, indicating that MAS was not effective for selection of drought resistance in this region (Table 3). Based on previously calculated heritability estimates for this population, yield should have improved, based on phenotypic selection, by 16% under stress and 10% under nonstress with the same selection intensity of 12.8% (Schneider et al., 1997). The gain from selection based on MAS was 3% for stress and 0% for nonstress. Clearly, MAS was ineffective in this population. Since the markers chosen were associated more consistently in Michigan locations, we may have identified regionally specific markers that were ineffective for improvement of performance under drought in Mexico. Alternatively, these markers may be better indicators of increased performance under severe stress since the drought intensity index for the Kalamazoo County experiment was high (0.76) compared with the experiments conducted in Madero and Calera (0.37).

Two genotypes were selected within the PAA group that, based on GM, ranked fourth and fifth for the S/A population. The PAA group, however, did not include two other genotypes previously identified based on yield data as ranking first and second for GM (Schneider et al., 1997). Genotypes selected among PBA included two genotypes that ranked 70th and 73rd for GM based on yield data for 80 genotypes. The PBA did not include three genotypes identified in the bottom 5% for GM. Genotypes identified by their performance data also should have been identified by their marker genotypes for MAS to be effective. However, the markers used for selection identified only 10 RIL for PAA and 9 RIL for PBA. Thus, the selection intensity was too high although the expected trend of PAA out-yielding PBA was still observed.

Marker Assisted Selection in Sierra/]Lef-2RB

In S/L, Linkage Group 1 was significantly associated with at least one of the three variables (Yd, Yp, GM) for all environments and the combined analysis, except for 1992 Madero and 1993 Calera locations. (Fig. 2; Table 2). This linkage group explained up to 16% of the variation for yield performance. Markers within this linkage group spanned an area of 26.8 cM, which exceeds the limit for minimum resolution (Lee, 1995). All five RAPD associated with increased performance under moisture stress were derived from the Sierra parent (Fig. 2).

Recombinant inbred lines from S/L were grown in Madero and Calera in 1994 to test the effectiveness of MAS on selection for drought resistance. Thirty-five RIL were selected for PAA and 27 RIL were selected for PBA, in this population. Genotypic groups were contrasted with combined 1994 yield data from Mexico under stress and nonstress (Table 4). In the combined analyses, all contrasts were significant. The PAA yielded 15 and 17 g [m.sup.-2] more than PBA under stress and nonstress, respectively. This supports our hypothesis that MAS would be effective in identifying a group of superior performing genotypes.


Within the group of 35 genotypes, included in PAA for S/L, 20 overlapped with the top 35 genotypes identified by conventional selection based on high GM. The MAS identified three genotypes that ranked first, second, and third for GM and within the top 10% for Yd and Yp (Schneider et al., 1997). Included in PAA were two other genotypes that ranked second and third for Yp. Of the 28 genotypes included in PBA for S/L, eight overlapped with the bottom 28 genotypes identified by conventional selection based on low GM.

Marker-Assisted Selection Compared with Conventional Selection

Markers identified from Linkage Group 1 of S/L accounted for a similar level of variation in S/A as in S/L (14%, S/A; 16%, S/L). However, MAS was effective in S/L and not in S/A (Table 4). As quantitative genetic theory suggests, the effectiveness of MAS is inversely proportional to the [h.sup.2] of a given trait (Lande and Thompson, 1990; Paterson et al., 1991). The results from this study are consistent with this statement. The [h.sup.2] for yield under both stress and nonstress treatments in S/A was three times greater than the [h.sup.2] for yield in S/L (Schneider et al., 1997). Additionally, the genetic variation was much greater in S/A than S/L for yield under both stress and nonstress conditions. Thus, for populations where improvement of a particular trait by conventional means is limited due to a lack of adequate genetic variation, MAS becomes an appealing option. In S/A, where [h.sup.2] for yield ranged from 0.55 to 0.59, considerable improvement can be made by conventional selection. However, for S/L where [h.sup.2] 2 for yield ranged from 0. 19 to 0.20 (Schneider et al., 1997), MAS proved more effective than conventional selection (Table 4). If the effectiveness of MAS is consistent with quantitative genetic theory, we would expect that improvement in S/A based on MAS, using the RAPD from Linkage Group 1 of S/L, will not realize the same gains in selection as were observed in S/L.

With 16% of the variation explained by the markers used in this study for selection in S/L, MAS resulted in an increase of 11 and 8% over the experimental mean for stress and nonstress, respectively. Although this is not a large improvement, it was sufficient considering the lack of genetic variation for yield in S/L (Schneider et al., 1997). The gain in selection, based on previously calculated [h.sup.2] 2 (Schneider et al., 1997) for phenotypic selection, at a 30% selection intensity, was 4 and 2.6% for stress and nonstress, respectively. Conventional selection did not improve the mean by the expected amount in either stress or nonstress treatments in S/L (Table 4). Expected gains, calculated based on [h.sup.2], were not achieved with conventional selection, which has also been observed in a previous study with common bean under stress (White et al., 1994). Marker-assisted selection exceeded the expected gain per cycle of selection by 6 and 5.4% for stress and nonstress, respectively (Table 4). This is a substantial improvement, which suggests that MAS is more effective than conventional selection for drought resistance in this common bean population.

Another consideration when developing an effective MAS method is selection intensity. In this study, two unlinked markers and one linkage group with two markers were chosen as selection criteria for S/A resulting in a 12.8% selection intensity for PAA and 11.5% for PBA. Independent assortment will occur among the two unlinked markers and the linkage group resulting in the identification of fewer individuals with the specified parental genotype. Thus, in S/A, 12.8% of the 78 genotypes would have been retained after MAS or alternatively 11.5% would have been discarded. In either case, too few individuals from a population of 78 were selected. In S/L, 36.8% of the population would have been selected, based on marker genotype, or alternatively 29.5% would have been discarded. This appears to be a more practical selection intensity for MAS breeding, especially when selection is targeted at earlier generations. Marker-assisted selection attempts to improve a quantitative trait through selection based on qualitative criteria. Thus, depending on the markers used, selection may be too lax or too rigid. To avoid this problem, Stromberg et al. (1994) used a system of weighted scoring; assigning different weighted values to individuals depending on the genotype at certain marker loci. This type of scoring produces a quantitative value for each individual. In this case, a selection intensity specified by the breeder can be maintained and, for example, the top 30% of genotypes based on their overall scores can be selected.

Marker-Assisted Selection in Sierra/AC1028 with Markers Identified in Sierra/Lef-2RB

The seventy RAPD used to screen each population for associations were not the same markers for each population. Although four of the five markers in Linkage Group 1 (Fig. 2) from S/L (Fig. 2) showed polymorphisms between Sierra and AC1028, they were not initially screened against S/A. Therefore, in the preliminary analyses, these markers were not identified as a linkage group associated with drought performance in S/A. The four polymorphic markers used for MAS in S/L were mapped into S/A with MAPMAKER/EXP. These four RAPD mapped into Linkage Group 3 of S/A (Fig. 1). These markers were then used as selection criteria to identify contrasting genotypic groups in S/A. Twenty-six RIL were identified in the PAA group, and 22 RIL were identified in PBA. Comparisons demonstrated no significant difference between genotypic groups in this population. Both groups, PAA and PBA, yielded equal to the experimental mean under stress and nonstress treatments (Table 4).

Selection based on Yd, Yp, and GM in S/A demonstrated significant differences between AA and BA for both stress and nonstress analyses (Table 4). Selection based on Yd showed a significant difference between AA and BA of 15 and 36 g [m.sup.-2] for stress and nonstress, respectively. Selection based on Yp resulted in a significant difference of 14 and 30 g [m.sup.-2] for stress and nonstress, respectively. Selection based on GM showed a significant difference between AA and BA of 15 and 44 g [m.sup.-2] under stress and nonstress, respectively. Markers identified in S/L and subsequently used for effective selection in S/L were not effective in S/A, although the RAPD explained a similar proportion of the variation in S/A (Tables 1 and 2). The four polymorphic markers from Linkage Group 1 in S/L were significantly associated in only 5 of the 18 possible combinations of analyses in S/A as opposed to 9 out of 18 in S/L (Tables 1 and 2). Compared with conventional selection, MAS was not effective in improving yield above the experimental mean in S/A under stress nor nonstress.

It would be of value to select within S/A until genetic variation for yield is limited, then examine the effectiveness of MAS. However, in S/A, of the 27 RIL selected in PAA, only one out of the top five yielding genotypes ranked by GM would have been selected by the markers identified in S/L. This precludes the use of these markers for early generation selection in this population. Additionally, of the four markers used for MAS in S/L, all alleles associated with improved performance in S/A originated from AC 1028, as opposed to originating from Sierra in S/L. This observation emphasizes the complexity of quantitative traits such as yield performance under moisture stress and the influence of epistasis and the genetic background in which the QTL reside. For markers to be useful when breeding for drought resistance, early generation marker-based selection must be tested. Stromberg et al. (1994) attempted this in maize, but MAS performed on [F.sub.2] genotypes failed to increase performance of the inbred lines above the population mean. However, conventional selection also proved ineffective. A selection index that explained 25% of the phenotypic variation, the influence of environmental factors, and the limited number of loci used to identify associations all contributed to this outcome (Stromberg et al., 1994). To verify the efficacy of the markers associated with drought resistance in this study, new populations must be developed and MAS performed on [F.sub.3:4] RIL. Testing of the predictive value of MAS should be conducted across a broader range of environments. If markers prove to be population specific, Sierra could be used as one parent making MAS useful, presuming that polymorphism for these markers exists between the chosen parents.


Beavis, W.D., O.S. Smith, D. Grant, and R. Fincher. 1994. Identification of quantitative trait loci using a small sample of topcrossed and [F.sub.4] progeny from maize. Crop Sci. 34:882-896.

Bubeck, D.M., M.M. Goodman, W.D. Beavis, and D. Grant. 1993. Quantitative trait loci controlling resistance to gray leaf spot in maize. Crop Sci. 33:838-847.

Edwards, K.C., C. Johnstone, and C. Thompson. 1991. A simple and rapid method for the preparation of plant genomic DNA for PCR analysis. Nucleic Acids Res. 19:1349.

Edwards, M.D., T. Helentjaris, S. Wright, and C.W. Stuber. 1992. Molecular-marker-facilitated investigations of quantitative trait loci in maize. Theor. Appl. Genet. 83:765-774.

Edwards, M.D., C.W. Stuber, and J.F. Wendel. 1987. Molecular-marker-facilitated investigations of quantitative-trait loci in maize. I. Numbers, genomic distribution and types of gene action. Genetics 116:113-125.

Freyre, R., and D.S. Douches. 1994. Development of a model for marker-assisted selection of specific gravity in diploid potato across environments. Crop Sci. 34:1361-1368.

Gill, J.L. 1993. Design and analysis of experiments in the animal and medical sciences: Volume 1. Iowa State University Press, Ames.

Haley S.D., L. K. Afanador, and J. D. Kelly. 1994a. Identification and application of RAPD marker for the I gene (Potyvirus resistance) in common bean. Phytopathology 84:157-160.

Haley, S.D., P.N. Miklas, L. Afanador, and J.D. Kelly. 1994b. Random amplified polymorphic DNA marker variability between and within gene pools of common bean. J. Am. Soc. Hortic. Sci. 119:122-125.

Lande, R., and R. Thompson. 1990. Efficiency of marker-assisted selection in the improvement of quantitative traits. Genetics 124: 743-756.

Lander, E.S., and D. Botstein. 1989. Mapping Mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121:185-199.

Lee, M. 1995. DNA markers and plant breeding programs. Adv. Agron. 55:265-344.

Lincoln, S., M. Daly, and E. Lander. 1987. Constructing genetic maps with MAPMAKER/EXP 3.0. 3rd ed. Tech. Rep., Whitehead Inst., Cambridge, MA.

Martin, B., J. Nienhuis, G. King, and A. Schaefer. 1989. Restriction fragment length polymorphisms associated with water use efficiency in tomato. Science (Washington, DC) 243:1725-1728.

Miklas, P.N., J.R. Stavely, and J.D. Kelly. 1993. Identification and potential use of a molecular marker for rust resistance in common bean. Theor. Appl. Genet. 85:745-749.

Nodari, R.O, S.M. Tsai, R.L. Gilbertson, and P. Gepts. 1992. Towards an integrated linkage map of common bean. H. Development of an RFLP-based linkage map. Theor. Appl. Genet. 85: 513-520.

Paterson, A.H., S. Damon, J.D. Hewitt, D. Zamir, H.D. Rabinwitch, S.E. Lincoln, E.S. Lander, and S.D. Tanksley. 1991. Mendelian factors underlying quantitative traits in tomato: Comparison across species, generations and environments. Genetics 127:181-197.

Saghai-Maroof, M.A., K.M. Soliman, R.A. Jorgensen, and R.W. Allard. 1994. Ribosomal DNA spacer-length polymorphisms in barley: Mendelian inheritance, chromosomal location, and population dynamics. Proc. Natl. Acad. Sci. USA 81:8014-8018.

SAS Institute. 1988. SAS/STAT user's guide, release 6.03 ed. SAS Inst., Inc., Cary, NC.

Schneider, K.A., R. Rosales-Serna, F. Ibarra-Perez, B. Cazares-Enriquez, J.A. Acosta-Gallegos, P. Ramirez-Vallejo, N. Wassimi, and J. D. Kelly. 1997. Improving common bean performance under drought stress. Crop Sci. 37:43-50 (this issue).

Soller, M., and J.S. Beckmann. 1990. Marker-based mapping of quantitative trait loci using replicated progenies. Theor. Appl. Genet. 80:205-208.

Stromberg, L.D., J.W. Dudley, and G.K. Rufener. 1994. Comparing conventional early generation selection with molecular marker assisted selection in maize. Crop Sci. 34:1221-1225.

Stuber, C.W., and M.D. Edwards. 1986. Genotypic selection for improvement of quantitative traits in corn using molecular marker loci. p. 70-83. In Proc. 41st Annual Corn and Sorghum Industry Res. Conf. Chicago, IL. 10-11 Dec. 1985. American Seed Trade Association, Washington, DC.

Stuber, C.W., M.D. Edwards, and J.F. Wendel. 1987. Molecular marker-facilitated investigations of quantitative trait loci in maize. II. Factors influencing yield and its component traits. Crop Sci. 27:639-648.

White, J.W., R. Ochoa M., F. Ibarra P., and S.P. Singh. 1994. Inheritance of seed yield, maturity and seed weight of common bean (Phaseolus vulgaris) under semi-arid rainfed conditions. J. Agric. Sci. (Cambridge) 122:265-273.

Abbreviations: AA, above average genotypic group; BA, below average genotypic group; GM, geometric mean; LOD, likelihood of the odds ratio; MAS, marker-assisted selection; PAA, potentially above average genotypic group; PBA, potentially below average genotypic group; QTL, quantitative trait loci; RAPD, random amplified polymorphic DNA; RIL, recombinant inbred lines; S/A, Sierra/AC1028 population; S/L, Sierra/ Lef-2RB population; Yd, yield under stress; Yp, yield under nonstress; RFLP, restriction fragment length polymorphism; EDTA, ethylenediaminetetraacetic acid; PCR, polymerase chain reaction; cM, centimorgan.

Kristin A. Scheneider, Mary E. Brothers, and James D. Kelly(*)

K.A. Schneider and J.D. Kelly, Crop and Soil Sciences Dep., Michigan State Univ., East Lansing, MI 48824; and M.E. Brothers, USDA-ARS, North Central Regional Plant Introduction Station, Iowa State Univ., Ames, IA 50011. This research was partially supported by Michigan Agric. Exp. Sm. and by the grant DAN1310-G-SS-6008-00 from the USAID Bean/Cowpea Collaborative Research Support Program. Received 2 Jan. 1996. (*) Corresponding author (
COPYRIGHT 1997 Crop Science Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1997 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Scheider, Kristin A.; Brothers, Mary E.; Kelly, James D.
Publication:Crop Science
Date:Jan 1, 1997
Previous Article:Improving common bean performance under drought stress.
Next Article:Estimation of outcrossing rate in common bean.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters