Genetic diversity for yield and grain quality of fourth-cycle recurrent-selection progenies of black common bean

Alves, Eduardo Almeida; Melo, Patrícia Guimarães Santos; Coelho, Alexandre Siqueira Guedes; Pereira, Helton Santos; Melo, Leonardo Cunha

doi:10.1590/1984-70332026v26n3a34

Abstract

Recurrent selection methods are most suitable for improving quantitative traits in plant breeding. This study estimated genetic parameters and assessed the genetic diversity of black common bean progenies from the fourth cycle of recurrent selection. Thirty-five S_0:4 progenies and four cultivars were evaluated in a randomized complete block design with three replications across four environments. Grain yield (YIELD), 100-seed weight (100SW), and sieve yield (SY) were evaluated. Molecular characterization was performed using microsatellite markers. Statistical analyses included analysis of variance, genetic parameters, genetic divergence, and population structure. Different strategies were used to select progenies. The analysis of variance showed significant effects for genotypes and the genotype × environment interaction. Expected gains within the program’s potential were 6.01% for YIELD, 3.47% for 100SW, and 2.77% for SY. Molecular analysis indicated inconsistent clustering and lack of genetic structure. For progeny selection, the Mulamba and Mock index with economic weights achieved better gain.

Keywords:
Phaseolus vulgaris L.; molecular markers; population breeding methods; genetic parameters; multivariate analyses

INTRODUCTION

Increasing grain yield remains the primary objective of common bean breeding programs. However, the polygenic nature of yield and the influence of environment make development of lines for higher yield a laborious process (Ramalho et al. 2024). Other traits must also be considered, such as disease resistance, plant architecture, yield stability, and commercial grain quality. Growers typically seek higher yield, whereas the packaging industry and consumers demand quality grain, with varying regional requirements (Pereira et al. 2017). For carioca-type beans, 100-seed weight from 25 to 27 grams, a lighter seed coat color, and lighter colored seed coat streaks are priority traits (Pereira et al. 2017, Dias et al. 2021). For commercialization of black-type beans, visual quality is also important, such as the absence of bluish or purplish discoloration.

The genetic base used to develop populations is narrow, constrained by commercial requirements for grain quality. Chiorato et al. (2010) evaluated genetic gains in common bean from 1989 to 2007 and found that after the inclusion of grain-quality assessments, there were no significant genetic gains for yield. Furthermore, the advancement of segregating populations without recombination does not effectively maintain genetic variability. Recurrent selection emerges as a method with greater capacity to develop high-yielding genotypes (Melo et al. 2019).

Recurrent selection involves successive cycles of developing, selecting, and recombining genotypes, which promotes genetic progress while maintaining genetic variability (Ramalho et al. 2024). Several studies have investigated recurrent selection in common bean (Lopes et al. 2019, Costa et al. 2024, Lima et al. 2026). Costa et al. (2024) applied recurrent selection to improve anthracnose resistance in common bean and reported heritability estimates ranging from 86.09% to 96.85%, with gains observed over five selection cycles. Similarly, Lima et al. (2026), working with recurrent selection in common bean for white mold resistance, reported heritability of 72.8% after thirteen cycles and a 3.96% to 4.27% reduction in the disease severity mean. For quantitative traits, recurrent selection tends to provide better performance in the long term because it progressively accumulates favorable alleles (Ramalho et al. 2024).

Molecular marker information and diversity metrics effectively show stratification and the proportion of parental alleles present in the progenies (Melo et al. 2019). Maintaining genetic variability is one of the main requirements for sustaining progress in a recurrent selection program; without this variability, success in exploiting genetic variability may be reduced by selection pressure. This study sought to evaluate agronomic performance while assessing genetic divergence using molecular markers. In this regard, genomic information helps guide parental selection and determine parental potential, ensuring efficient exploitation of genetic variability in future combinations.

This study aimed to estimate genetic parameters and assess the genetic diversity of common bean progenies from a recurrent selection program in the black bean group, selecting genotypes that combine high yield and optimal commercial grain-quality traits.

MATERIAL AND METHODS

Progenies were evaluated originating from the black bean recurrent selection program for grain yield at Embrapa Arroz e Feijão. The base population was established by crossing 13 parental lines (A525, A806, LM95204101, A797, A429, AN9022180, VAX1, IAPAR14, Aporé, A774, ARC1, MAN48, and Pérola) from the black and carioca commercial groups. These parents showed improved performance for plant architecture, response to low soil fertility, biological nitrogen fixation, grain yield and quality, disease resistance, and drought tolerance. The C₀S₀ generation was sown with 5500 seeds at a single site with high anthracnose incidence. From this population, 500 individual anthracnose-resistant plants were selected; and after selection for commercial grain quality, 400 plants were retained. The 400 C₀S_0:1 progenies were grown under water-stress conditions to select genotypes for drought tolerance; from these, 100 C₀S_0:1 progenies were selected and advanced for multiplication (C₀S_0:2) and evaluated at two locations. Subsequently, 60 C₀S_0:3 progenies were selected for evaluation at five locations. Among the 60 C₀S_0:3 progenies, 20 C₀S_0:4 progenies were selected and evaluated at four locations (Table S1). From the 20 evaluated C₀S_0:4 progenies, the 10 best progenies were recombined in a partial diallel design. For cycle zero, a partial diallel was used; for subsequent cycles, recombination was carried out through a complete diallel design.

In cycles I, II, and III, 1500 seeds were sown in Santo Antônio de Goiás-GO (STA-GO). These 1500 seeds resulted in 1000 progenies to establish the S_0:1 generation at the same location. For the S_0:2 generation, 400 progenies were sown in an environment without an experimental design in Ponta Grossa-PR (PG-PR), with progenies represented by three 3-meter rows. The S_0:3 generation comprised 200 progenies, also grown without an experimental design, with three 3-meter rows per progeny in PG-PR and a check cultivar allocated to every nine progenies. In the present study, progenies from the fourth recurrent selection cycle were evaluated.

The experiments comprised 35 C₃S_0:4 progenies and four cultivars: BRS Esteio, BRS FP403, BRS FP417, and BRS Esplendor. The cultivars were chosen due to their high performance for specific traits: BRS FP403 for technological grain quality (Souza et al. 2019), BRS Esplendor (Costa et al. 2009) and BRS FP417 (Aguiar et al. 2023) for upright architecture and yield potential, and BRS Esteio for yield stability (Pereira et al. 2014). A randomized complete block design was used, with three replications. Plots consisted of three 3-meter rows, with a density of 12 plants per linear meter and 0.45 meters between rows. Experiments were conducted in PG-PR and STA-GO across three sowing seasons (Table S1). The evaluated agronomic and commercial traits were grain yield (YIELD - kg ha^-1), 100-seed weight (100SW - g), and sieve yield (SY - %).

All genotypes were sampled for molecular genetic analysis, selecting 10 plants from each treatment in the STA-GO trial. This analysis was performed using microsatellite markers (Simple Sequence Repeat - SSR). Four multiplex panels containing six markers each were employed, for a total of 24 molecular markers with adequate amplification, informativeness, and genome-wide distribution (Valdisser et al. 2013, Morais et al. 2016, Melo et al. 2019). As described in Table S2, the panels comprise 12 markers from the BM series (Gaitán-Solís et al. 2002), eight from the PvBR series (Buso et al. 2006), and four from the Pv-ESTBR series (Garcia et al. 2011).

DNA was extracted from two leaf discs per sample according to Xin et al. (2003). PCR amplification was performed using the Qiagen® PCR Multiplex kit, and amplified fragments were analyzed on an ABI3500xL automatic fragment analyzer and genotyped using GeneMapper 4.1 (Applied Biosystems).

Individual and combined analyses of variance were performed, considering all effects as fixed. Based on the analysis of variance results and expected mean squares, the genetic, phenotypic, environmental, and genotype × environment interaction quadratic components were estimated at the mean level, as well as the broad-sense coefficient of determination, according to Vencovsky and Barriga (1992). Confidence intervals for the quadratic components were calculated according to Burdick and Graybill (1992) and Barbin (2019), and heritability was estimated according to Knapp et al. (1985) and Burdick and Graybill (1992). Genetic and experimental coefficients of variation were also estimated. Selective Accuracy (SA), which indicates the accuracy of genotypic values, was estimated according to Resende and Duarte (2007): $S A = {(1 - \frac{1}{F c})}^{0.5}$ where Fc is the ratio for the treatment effects (genotypes), associated with the analysis of variance.

Based on genomic data, the distance matrix and estimates of genetic diversity among the progenies were calculated using Rogers’ distance modified by Wright (Wright 1978). The following diversity measures were estimated: number of alleles per locus (An), expected heterozygosity under Hardy-Weinberg equilibrium ( $H_{E}$ ), observed heterozygosity ( $H_{O}$ ), and the within-population fixation index (f). The distance matrix was represented using a dendrogram generated through the Unweighted Pair Group Method with Arithmetic Mean (UPGMA) algorithm and by principal coordinate analysis (PCoA). Node support was assessed by bootstrap resampling with 30,000 resamplings. Correlations between the dendrogram and the dissimilarity matrix and between the principal coordinates and the distance matrix were tested using the Mantel test (Mantel 1967). The error probability associated with the cophenetic correlation was determined using 10,000 resamplings.

To evaluate crossing efficiency and exploitation of genetic variability, the degree of genetic structure was quantified based on molecular markers and allele frequencies. A Markov Chain Monte Carlo (MCMC) procedure was performed using the Structure software (Pritchard et al. 2000). A burn-in of 6,000 iterations was used, followed by an iterative process with 60,000 randomizations. For simulations, the number of groups or clusters (K) ranged from one, assuming no genetic structure, to 10, assuming genetic structure with 10 different populations, with 30 iterations for each K. The optimal ∆K number of groups was defined using the method of Evanno et al. (2005).

Selection pressure of 28% was applied for choosing progenies. Multiple strategies were used for progeny selection. First, simultaneous selection (SS) was used, based on the breeder’s empirical knowledge of the progenies. Subsequently, the Kennard-Stone (KS) algorithm (Kennard and Stone 1969) was applied to select more divergent samples based only on molecular information. Then, an association between SS and the KS algorithm was evaluated to weight genetic diversity (SKS). In addition, the Mulamba and Mock (1978) selection index (MMI) was used. The MMI index was applied under the following scenarios: without economic weights (MMI-1); with economic weights (MMI-2), assigning a weight of 2 for YIELD and 1 for 100SW and SY; incorporating diversity information from marker data, based on ranking using the genetic distance matrix (MMI-3); and combining both economic weights and diversity data (MMI-4). Statistical analyses were performed using Structure (Pritchard et al. 2000), GENES (Cruz 2016), and R (R Core Team 2024) softwares.

RESULTS AND DISCUSSION

Significant differences were observed for both yield and technological grain quality (Table S3). Quadratic components and genetic parameters are presented in Tables S4, S5, and S6. Most of the estimated quadratic components differed significantly from zero (α = 0.05), showing low-magnitude ranges.

The combined analysis of variance for yield showed significant differences for all sources of variation except for the check cultivars. For 100-seed weight, only the contrast effect was non-significant; whereas for sieve yield, non-significance was observed for both the check × environment interaction contrast and the contrast across environments (Table S7 and Figure S8). Significant results indicate genetic variability and environmental effects. Furthermore, the genotype × environment interaction was significant, showing that genotypes respond differently to environmental stimuli.

Mean values of the progenies were 2883 kg ha^-1 for yield, 21.37 grams for 100-seed weight, and 86.63% for sieve yield (Table S7). The coefficients of variation ranged from 3.36% to 12.87% for the three traits, indicating high precision. Selective accuracy ranged from 0.87 to 0.97, indicating precision of the estimated genotypic values (Resende and Duarte 2007).

Mean quadratic components and genetic parameters estimated from the combined analysis of variance are shown in Table S9. The estimates were consistent and significant, thus different from zero (α = 0.05). The coefficient of determination reached high magnitudes across all traits: 67.97% for yield, 90.67% for 100-seed weight, and 74.29% for sieve yield. As these traits are quantitative and polygenic, they are sensitive to environmental influences (Ramalho et al. 2024). Estimated heritability was high, predicting success under selection.

Variance components are estimates that allow assessment of the genetic nature of traits. Traits with greater genetic variance can be improved using less laborious and simpler breeding methods. However, releasing genotypes on the market requires the association of traits, which depends on hybridizations. When parents with superior performance are crossed and when their genetic variance is of high magnitude, tangible genetic gains can be expected.

Genotypic variation does not change markedly at intermediate gene frequencies; however, partitioning it into additive and dominance components varies according to the degree of dominance and overdominance (Falconer 1960). Consequently, evaluating S_0:2 progenies may be more efficient for observing existing genetic variability and improving selection efficiency, depending on the adopted selection intensity. Thus, Menezes Júnior et al. (2008) and Lopes et al. (2019) used S_0:2 progenies from a common bean recurrent selection program. However, in early generations, the amount of seed available to perform evaluations at more than one location is limited. Therefore, using more advanced generations with more seed stock allows evaluation across more locations, improving genetic and environmental estimates.

The environment does not have an equal influence on all genotypes, even when environmental conditions are similar, and genotypes with low inbreeding show greater responsiveness to environmental variance (Falconer 1960). Moreover, S_0:4 progenies have a higher proportion of additive genetic variance, which enhances selection efficiency, provided that the selection intensity is consistent with the available genetic variability. Selection intensity determines the rate of change in allele frequencies. Even under conditions of high genetic variability, excessively high selection intensity can be detrimental to future selection cycles and progeny recombination. Both effective population size and selection intensity permanently influence the potential for exploiting genetic variability (Pereira and Vencovsky 1988). To mitigate limitations related to genetic variability, new genotypes can be introduced into the recombination phase.

Evaluation of the S_0:4 generation provides advantages because multi-environment trials become feasible due to seed availability. Phenotypic data used to assess diversity in a limited number of environments are often biased by the genotype × environment interaction (Pereira et al. 2019). Therefore, using S_0:4 progenies enables evaluation across multiple locations and more precise estimates of genetic effects.

For common bean, yield and disease resistance traits must be accompanied by commercial grain-quality traits for cultivar adoption. Traits such as grain shape (assessed by sieve yield), size, and density (assessed by 100-seed weight) are criteria associated with higher commercial returns desired by growers (Pereira et al. 2017, Pereira et al. 2021).

Mean values, selection differentials, and selection gains under the different selection criteria are shown in Table 1. Based on individual selection, the predicted genetic gains for the progenies in this study were 6.01% for yield, 3.47% for 100-seed weight, and 2.77% for sieve yield (Table S10).

Thumbnail

Table 1
Estimates of population means, selection differentials (SD), and expected gains from selection (GS) (%) for the grain yield (kg ha^-1) (YIELD), 100-seed weight (g) (100SW), and sieve yield (%) (SY) grain traits under the selection strategies [Kennard-Stone algorithm (KS), and the Mulamba and Mock (1978) selection index (MMI)] based on mean selection [direct and simultaneous selection (SS)]

The gains observed correspond to only a single selection and recombination cycle. Evaluating different methods for advancing segregating populations in common bean, Pontes Júnior et al. (2016) reported selection gains for yield of 1.86% using Bulk within F₂, 4.45% using Bulk, and 2.97% using Single Seed Descent (SSD). Similarly, Silva et al. (2013) found yield gains of 10.18% for Bulk, 4.09% for Bulk F_2:7, and 8.77% for SSD. Although these methods are widely used, their long-term gains are lower than those achieved through recurrent selection. The common bean breeding program at Embrapa Arroz e Feijão is an example of this difference. While conventional methods for advancing segregating populations require six to eight years per selection cycle to achieve gains, recurrent selection can yield gains ranging from 1% to as much as 9% per progeny evaluation cycle, with an average duration of one to two years per cycle. In addition, common bean can be sown in up to three seasons during the year, which can accelerate gains.

In common bean breeding, selection aims to identify high-yielding genotypes that also meet commercial quality requirements, as consumers prefer specific grain size and shape (Pereira et al. 2017). Among the eight selection strategies evaluated in the present study, the MMI-4 index achieved the greatest gain for yield, differing by only 0.14% from the maximum potential gain of the progenies. The MMI-1 index was the best strategy for 100-seed weight, differing by 1.62% from individual selection. For total potential for sieve yield, simultaneous selection (SS) was the best strategy, differing by 1.07% from the maximum potential.

The main objective of the recurrent selection program evaluated here is yield gain. The strategy that enabled adequate and positive gains in both yield and commercial grain-quality traits, considering lower variation, was SKS, followed by SS. However, considering the program objective of maximizing yield gain, MMI-2 is predicted to be the best strategy.

Studies on the genetic structure of natural populations are quite common (Zucchi et al. 2003) and frequently show structural effects due to mating systems or geographic barriers. In addition, genetic structure analyses are often conducted in germplasm bank studies (Delfini et al. 2021). Analysis using the Structure software (Pritchard et al. 2000), based on an MCMC algorithm, combined with the method of Evanno et al. (2005) for selecting the optimal number of groups, offers advantages in using iterative processes.

In natural populations or germplasm bank studies, the $∆ K$ range of values typically varies widely (Zucchi et al. 2003, Delfini et al. 2021). In this study, low $∆ K$ values indicated the absence of structure (Figure S11). This result confirms the efficiency of recombination, the balanced distribution of parental alleles in the population, and the maintenance of genetic variability.

The selected molecular marker panel is designed to assess genetic diversity among common bean genotypes (Valdisser et al. 2013). The SSR markers provided multi-allelic information at low cost and could increase the speed and accuracy of analyzing genomic data (Cardoso et al. 2014). Multiplex systems containing the 24 markers have already been applied to common bean by Morais et al. (2016), Melo et al. (2019), and Pereira et al. (2019). These markers are routinely used in breeding activities at Embrapa Arroz e Feijão (Pereira et al. 2019) and have been effectively employed for diversity analyses (Cardoso et al. 2014, Pereira et al. 2019), parental characterization (Batista et al. 2014), genotype identification, and seed genetic purity assessment (Morais et al. 2016).

A total of 83 alleles were detected across the 24 loci, with a mean of 3.46 alleles per polymorphic locus, ranging from one allele at locus BM164 to eight alleles at locus BM143 (Table S12). Pereira et al. (2019) reported a similar mean number of alleles (4.15). Cardoso et al. (2014) found a higher mean number of alleles (8.29); however, in that study, diversity was assessed across distinct grain classes and cultivars from different countries. Valdisser et al. (2013) reported a mean of 10 alleles when evaluating accessions in an active germplasm bank.

Expected heterozygosity ( $H_{E}$ ) averaged 0.357, ranging from 0.051 at locus BM181 to 0.684 at BM143. Observed heterozygosity ( $H_{O}$ ) averaged 0.203, with values ranging from 0.051 at locus PV11 to 0.436 for BM143. The within-population fixation index ranged from 0.031 (locus PV35) to 1.00 (loci BM138, BM155, BM181, BM201 and PV251), with an overall mean of 0.436.

The mean expected and observed heterozygosity values in the present study differed from those reported in previous studies with common bean cultivars and accessions, with mean values of $H_{E}$ = 0.793 and $H_{O}$ = 0.039 reported by Valdisser et al. (2013); $H_{E}$ = 0.646 and $H_{O}$ = 0.031 by Cardoso et al. (2014); and $H_{E}$ = 0.55 and $H_{O}$ = 0.05 reported by Pereira et al. (2019). In addition, the positive mean value of the within-population fixation index reflects the selective effect imposed on the progenies, indicating that the applied phenotypic selection favored homozygous individuals.

Various classes of molecular markers are used in genetic diversity studies; and measures such as simple matching, Jaccard, Nei (1972), Nei and Li (Cruz et al. 2020), Rogers (Cruz et al. 2020), Wright, and Rogers-modified by Wright (Wright 1978) can be applied. Visualizing dissimilarity information among groups from the matrices is not a straightforward process. Thus, clustering analyses, which may or may not include dendrogram construction, are useful multivariate tools for dissimilarity measures. Clustering procedures include the Tocher method, the nearest- and farthest-neighbor methods, UPGMA, and WPGMA (Cruz et al. 2020).

Node consistency assessed through iterative procedures, such as bootstrapping, has increasingly been used because of its conservative nature and broad applicability across species and data types (Cruz et al. 2020). Node support values did not exceed the 70% threshold, which is an adequate reliability threshold for concluding that groups are not formed. This result was observed for both genotypes (Figure 1, Figure S13) and progenies (Figure S14). Therefore, these finding confirm that there is no apparent group formation among the progenies and checks (Figures 1 and S13).

Figure 1
Assessment of node support in the dendrogram including common bean progenies and cultivars. Cophenetic correlation = 0.73 (p < 0.01). 1: p-057; 2: p-103; 3: p-127; 4: p-197; 5: p-213; 6: p-235; 7: p-242; 8: p-264; 9: p-268; 10: p-308; 11: p-310; 12: p-312; 13: p-369; 14: p-403; 15: p-446; 16: p-471; 17: p-474; 18: p-481; 19: p-526; 20: p-579; 21: p-580; 22: p-617; 23: p-645; 24: p-652; 25: p-675; 26: p-698; 27: p-744; 28: p-745; 29: p-751; 30: p-759; 31: p-789; 32: p-803; 33: p-834; 34: p-852; 35: p-873; 36: BRS Esteio; 37: BRS FP403; 38: BRS FP417; 39: BRS Esplendor.

Characterizing genetic diversity assists in choosing parental lines, exploiting the combination of genetic variability and complementarity to increase the mean of traits (Pereira et al. 2019). Consequently, the normal recommendation is to select genotypes from distinct and distant groups based on genetic diversity analysis. In the present study, however, as no groups were detected, selection of any family can be recommended for generation advancement and recombination, for all provide similar representation of genetic variation in terms of allele frequencies.

A tendency toward group formation was not observed for the SS, MMI-2, and MMI-4 selection strategies (Figure S15). Selection based on these principles sampled progenies that were more randomly distributed across the total variation. The KS and MMI-3 strategies tended to select progenies with greater similarity to each other.

Comparisons among values obtained from different strategies are shown in Figures S15 and S16. Without considering individual selection and focusing on yield and 100-seed weight, the MMI-4 index, weighted by both economic weights and diversity information, showed the highest gain from selection. In contrast, for sieve yield, simultaneous selection achieved the highest gain. Selection using the Kennard-Stone algorithm produced the lowest genetic gains across all traits. The mean values for all progenies and checks are shown in Table S17. Progeny p-481 had the best yield (3230 kg ha^-1). For 100-seed weight and sieve yield, the best progenies were p-745 (22.56 g) and p-242 (90.09%), respectively.

CONCLUSION

Genetic variability among the progenies for the evaluated traits was confirmed, ensuring potential for gains through recurrent selection for grain yield.

The best strategy for progeny selection was the Mulamba and Mock (1978) index incorporating economic weights.

Population structure analysis revealed no genetic structure among the progenies. Diversity measures indicated broad genetic variability and the absence of cluster formation.

Ten progenies were selected to compose the next recurrent selection cycle aiming at improved grain yield and the development of elite lines.

Data Availability Statement

The datasets generated and/or analyzed during the current research are available from the corresponding author upon reasonable request.

REFERENCES

Aguiar MS, Pereira HS, Souza TLPO, Faria LC, Costa JGC, Torga PP, Guimarães CM, Souza Filho BF, Almeida VM, Melo LC2023 BRS FP417: black common bean cultivar with high yield, high commercial-quality grain, and disease resistanceFunctional Plant Breeding Journal 5:1-7
Barbin D2019 Componentes de variância: teoria e aplicações. Fealq, Piracicaba, 144p.
Batista CEA, Bueno LG, Resende TN, Lima VR, Wendland A, Pereira HS, Faria L C, Melo LC, Souza TLPO2014 Parental genetic representativeness in black seeded common bean progenies from the Embrapa recurrent selection pro- gram for tolerance to BGMVAnnual report of the Bean Improvement Cooperative 57:221-222
Burdick RK, Graybill FA1992 Confidence intervals on variance components. CRC Press, New York, 230p.
Buso GSC, Amaral ZPS, Brondani RPV, Ferreira ME2006 Microsatellite markers for the common bean Phaseolus vulgarisMolecular Ecology Notes 6:252-254
Cardoso PCB, Brondani C, Menezes IPP, Valdisser PAMR, Borba TCO, Del Peloso MJ, Vianello RP2014 Discrimination of common bean cultivars using multiplexed microsatellite markersGenetics and Molecular Research 13:1964-1978
Chiorato AF, Carbonell SAM, Vencovsky R, Fonseca Júnior NS, Pinheiro JB2010 Genetic gain in the breeding program of common beans at IAC from 1989 to 2007Crop Breeding and Applied Biotechnology 10:329-336
Costa JGC, Melo LC, Pereira HS, Del Peloso MJ, Faria LC, Díaz JLC, Wendland A, Rava CA, Carvalho HWL, Costa AF, Almeida VM, Melo CLP, Vieira Júnior JR, Faria JC, Souza JF, Marangon MA, Cargnin A, Abreu AFB, Moreira JAA, Pereira Filho IA, Guimarães CM, Bassinello PZ, Brondani V, Magaldi MCS2009 BRS Esplendor: cultivar de feijoeiro comum de grão tipo comercial preto, com arquitetura de planta ereta, alto potencial produtivo e tolerância a doenças. Embrapa Arroz e Feijão, Santo Antônio de Goiás, 4p. (Boletim Informativo, 185).
Costa LC, Ramalho MAP, Abreu ÂFB, Souza EA2024 Recurrent selection for broad-spectrum resistance to anthracnose in common beanCrop Science 64:3073-3085
Cruz CD2016 Genes software - extended and integrated with the R, Matlab and Selegen. Acta ScientiarumAgronomy 38:547-552
Cruz CD, Ferreira FM, Pessoni LA2020 Biometria aplicada ao estudo da diversidade genética. Editora UFV, Viçosa, 620p.
Delfini J, Moda-Cirino V, Santos Neto J, Ruas PM, Sant’Ana GC, Gepts P, Gonçalves LSA2021 Population structure, genetic diversity and genomic selection signatures among a Brazilian common bean germplasmScientific Reports 11:2964
Dias PAS, Almeida DV, Melo PGS, Pereira HS, Melo LC2021 Effectiveness of breeding selection for grain quality in common beanCrop Science 61:1127-1140
Evanno G, Regnaut S, Goudet J2005 Detecting the number of clusters of individuals using the software STRUCTURE: a simulation studyMolecular Ecology 14:2611-2620
Falconer DS1960 Introduction to quantitative genetics. Longman Group Limited, Edinburgh, 109p.
Gaitán-Solís E, Duque MC, Edwards KJ, Tohme J2002 Microsatellite repeats in common bean (Phaseolus vulgaris) isolation, characterization, and cross-species amplification in Phaseolus sspCrop Science 42:2128-2136
Garcia RAV, Rangel PN, Brondani C, Martins WS, Melo LC, Carneiro MS, Borba TCO, Brondani RPV2011 The characterization of a new set of EST-derived simple sequence repeat (SSR) markers as a resource for the genetic analysis of Phaseolus vulgarisBMC Genetics 12:1-14
Kennard RW, Stone LA1969 Computer aided design of experimentsTechnometrics 11:137-148
Knapp SJ, Stroup WW, Ross WM1985 Exact confidence intervals for heritability on a progeny mean basisCrop Science 25:192-194
Lima GAS, Pereira FAC, Ferreira AN, Carneiro VQ, Souza EA2026 Accumulating alleles of resistance to white mold in common beans over 13 recurrent selection cyclesCrop Science 66:e70240
Lopes FS, Leite ME, Porto ACM, Miguel LA, Reis VOL, Santos JB2019 Recurrent selection in common bean aiming at resistance to white mold in a greenhouseCrop Breeding and Applied Biotechnology 19:95-101
Mantel N1967 The detection of disease clustering and a generalized regression approachCancer Research 27:209-220
Melo LC, Silva ACF, Pereira HS, Melo PGS, Souza TLPO2019 Genetic diversity as an additional selection criterion in a carioca common bean recurrent selection program for resistance to common bacterial blightGenetics and Molecular Research 18:1-15
Menezes Júnior JÂN, Ramalho MAP, Abreu AFB2008 Seleção recorrente para três caracteres do feijoeiroBragantia 67:833-838
Morais SRP, Vieira AF, Almeida LCS, Rodrigues LA, Melo PGS, Faria LC, Melo LC, Pereira HS, Souza TLPO2016 Application of microsatellite markers to confirm controlled crosses and assess genetic identity in common beanCrop Breeding and Applied Biotechnology 16:234-239
Mulamba NN, Mock JJ1978 Improvement of yield potential of the Eto Blanco maize (Zea mays L.) population by breeding for plant traitsEgyptian Journal of Genetics and Cytology 7:40-51
Nei M1972 Genetic distance between populationsThe American Naturalist 106:283-292
Pereira HS, Alvares RC, Silva FC, Faria LC, Melo LC2017 Genetic, environmental and genotype x environment interaction effects on the common bean grain yield and commercial qualitySemina: Ciências Agrárias 38:1241-1250
Pereira HS, Melo LC, Wendland A, Del Peloso MJ, Faria LC, Nascente AS, Díaz JLC, Costa JGC, Carvalho HWL, Costa AF, Almeida VM, Melo CLP, Magaldi MCS, Abreu AFB, Moreira JAA, Pereira Filho IA, Cargnin A, Posse SCP, Souza Filho BF, Moura Neto FP, Guimarães CM, Braz AJBP, Ferreira SB, Marangon MA, Souza NP, Oliveira JP, Faria JC, Bassinello PZ2014 BRS Esteio - cultivar de feijoeiro comum com grãos pretos, alto potencial produtivo e resistência à antracnose. Embrapa Arroz e Feijão, Santo Antônio de Goiás , 4p. (Boletim Informativo, 213).
Pereira HS, Mendonça FR, Rodrigues LL, Melo LC, Melo PGS, Faria LC, Costa AF, Carvalho HWL, Pereira Filho IA, Almeida VM2021 Selection of carioca common bean lines with slow darkeningPesquisa Agropecuária Brasileira 56:e02471
Pereira HS, Mota APS, Rodrigues LA, Souza TLPO, Melo LC2019 Genetic diversity among common bean cultivars based on agronomic traits and molecular markers and application to recommendation of parent linesEuphytica 215:38
Pereira MB, Vencovsky R1988 Limites da seleção recorrente. Fatores que afetam o acréscimo das frequências alélicasPesquisa Agropecuária Brasileira 23:769-780
Pontes Júnior VA, Melo PGS, Pereira HS, Melo LC2016 Genetic potential of common bean progenies obtained by different breeding methods evaluated in various environments. Genetics and Molecular Research 15: gmr.15038622.
Pritchard JK, Stephens M, Donnelly P2000 Inference of population structure using multilocus genotype dataGenetics 155:945-959
R Core Team2024 R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna.
Ramalho MAP, Nunes JAR, Pádua JMV, Toledo FHB2024 Aplicações da genética quantitativa no melhoramento de plantas autógamas. UFLA, Lavras, 456p.
Resende MDV, Duarte JB2007 Precisão e controle de qualidade em experimentos de avaliação de cultivaresPesquisa Agropecuária Tropical 37:182-194
Silva ACF, Melo PGS, Melo LC, Bassinello PZ, Pereira HS2013 Eficiência de métodos de melhoramento para teor de fibra e produtividade de grãos em progênies de feijoeiro comumBragantia 72:326-331
Souza TLPO, Pereira HS, Del Peloso MJ, Faria LC, Costa JGC, Wendland A, Díaz JLC, Magaldi MCS, Aguiar MS, Carvalho HWL, Souza Filho BF, Melo CLP, Costa AF, Almeida VM, Posse SCP, Melo LC2019 BRS FP403: high-yielding black-seeded common bean cultivar with superior grain quality and moderate resistance to fusarium wiltCrop Breeding and Applied Biotechnology 19:240-244
Valdisser PAMR, Mota APS, Bueno LG, Menezes IPP, Coelho GRC, Magalhães FOC, Vianello RP2013 Protocolo de extração de DNA e genotipagem de SSRs em larga escala para uso no melhoramento do feijoeiro comum (Phaseolus vulgaris L.). Embrapa Arroz e Feijão, Santo Antônio de Goiás , 6p. (Boletim Informativo, 208).
Vencovsky R, Barriga P1992 Genética biométrica no fitomelhoramento. Sociedade Brasileira de Genética, Ribeirão Preto, 496p.
Wright S1978 Evolution and the genetics of populations. University of Chicago Press, Chicago, 580p.
Xin Z, Velten JP, Oliver MJ, Burke JJ2003 High-throughput DNA extraction method suitable for PCRBioTechniques 34:820-826
Zucchi MI, Pereira R, Brondani V, Coelho ASG, Vencovsky R2003 Genetic structure and gene flow in Eugenia dysenterica DC in the Brazilian Cerrado utilizing SSR markersGenetics and Molecular Biology 26:449-457

Correspondence

Corresponding author: eduardoalmeidaalves10@gmail.com

SCIENTIFIC EDITOR:

Luiz Antônio dos Santos Dias

Publication Dates

Publication in this collection
25 Sept 2026
Date of issue
2026

History

Received
09 Feb 2026
Accepted
09 May 2026
Published
26 June 2026

This is an open-access article distributed under the terms of the Creative Commons Attribution License

[1] Aguiar MS, Pereira HS, Souza TLPO, Faria LC, Costa JGC, Torga PP, Guimarães CM, Souza Filho BF, Almeida VM, Melo LC2023 BRS FP417: black common bean cultivar with high yield, high commercial-quality grain, and disease resistanceFunctional Plant Breeding Journal 5:1-7

[2] Barbin D2019 Componentes de variância: teoria e aplicações. Fealq, Piracicaba, 144p.

[3] Batista CEA, Bueno LG, Resende TN, Lima VR, Wendland A, Pereira HS, Faria L C, Melo LC, Souza TLPO2014 Parental genetic representativeness in black seeded common bean progenies from the Embrapa recurrent selection pro- gram for tolerance to BGMVAnnual report of the Bean Improvement Cooperative 57:221-222

[4] Burdick RK, Graybill FA1992 Confidence intervals on variance components. CRC Press, New York, 230p.

[5] Buso GSC, Amaral ZPS, Brondani RPV, Ferreira ME2006 Microsatellite markers for the common bean Phaseolus vulgarisMolecular Ecology Notes 6:252-254

[6] Cardoso PCB, Brondani C, Menezes IPP, Valdisser PAMR, Borba TCO, Del Peloso MJ, Vianello RP2014 Discrimination of common bean cultivars using multiplexed microsatellite markersGenetics and Molecular Research 13:1964-1978

[7] Chiorato AF, Carbonell SAM, Vencovsky R, Fonseca Júnior NS, Pinheiro JB2010 Genetic gain in the breeding program of common beans at IAC from 1989 to 2007Crop Breeding and Applied Biotechnology 10:329-336

[8] Costa JGC, Melo LC, Pereira HS, Del Peloso MJ, Faria LC, Díaz JLC, Wendland A, Rava CA, Carvalho HWL, Costa AF, Almeida VM, Melo CLP, Vieira Júnior JR, Faria JC, Souza JF, Marangon MA, Cargnin A, Abreu AFB, Moreira JAA, Pereira Filho IA, Guimarães CM, Bassinello PZ, Brondani V, Magaldi MCS2009 BRS Esplendor: cultivar de feijoeiro comum de grão tipo comercial preto, com arquitetura de planta ereta, alto potencial produtivo e tolerância a doenças. Embrapa Arroz e Feijão, Santo Antônio de Goiás, 4p. (Boletim Informativo, 185).

[9] Costa LC, Ramalho MAP, Abreu ÂFB, Souza EA2024 Recurrent selection for broad-spectrum resistance to anthracnose in common beanCrop Science 64:3073-3085

[10] Cruz CD2016 Genes software - extended and integrated with the R, Matlab and Selegen. Acta ScientiarumAgronomy 38:547-552

[11] Cruz CD, Ferreira FM, Pessoni LA2020 Biometria aplicada ao estudo da diversidade genética. Editora UFV, Viçosa, 620p.

[12] Delfini J, Moda-Cirino V, Santos Neto J, Ruas PM, Sant’Ana GC, Gepts P, Gonçalves LSA2021 Population structure, genetic diversity and genomic selection signatures among a Brazilian common bean germplasmScientific Reports 11:2964

[13] Dias PAS, Almeida DV, Melo PGS, Pereira HS, Melo LC2021 Effectiveness of breeding selection for grain quality in common beanCrop Science 61:1127-1140

[14] Evanno G, Regnaut S, Goudet J2005 Detecting the number of clusters of individuals using the software STRUCTURE: a simulation studyMolecular Ecology 14:2611-2620

[15] Falconer DS1960 Introduction to quantitative genetics. Longman Group Limited, Edinburgh, 109p.

[16] Gaitán-Solís E, Duque MC, Edwards KJ, Tohme J2002 Microsatellite repeats in common bean (Phaseolus vulgaris) isolation, characterization, and cross-species amplification in Phaseolus sspCrop Science 42:2128-2136

[17] Garcia RAV, Rangel PN, Brondani C, Martins WS, Melo LC, Carneiro MS, Borba TCO, Brondani RPV2011 The characterization of a new set of EST-derived simple sequence repeat (SSR) markers as a resource for the genetic analysis of Phaseolus vulgarisBMC Genetics 12:1-14

[18] Kennard RW, Stone LA1969 Computer aided design of experimentsTechnometrics 11:137-148

[19] Knapp SJ, Stroup WW, Ross WM1985 Exact confidence intervals for heritability on a progeny mean basisCrop Science 25:192-194

[20] Lima GAS, Pereira FAC, Ferreira AN, Carneiro VQ, Souza EA2026 Accumulating alleles of resistance to white mold in common beans over 13 recurrent selection cyclesCrop Science 66:e70240

[21] Lopes FS, Leite ME, Porto ACM, Miguel LA, Reis VOL, Santos JB2019 Recurrent selection in common bean aiming at resistance to white mold in a greenhouseCrop Breeding and Applied Biotechnology 19:95-101

[22] Mantel N1967 The detection of disease clustering and a generalized regression approachCancer Research 27:209-220

[23] Melo LC, Silva ACF, Pereira HS, Melo PGS, Souza TLPO2019 Genetic diversity as an additional selection criterion in a carioca common bean recurrent selection program for resistance to common bacterial blightGenetics and Molecular Research 18:1-15

[24] Menezes Júnior JÂN, Ramalho MAP, Abreu AFB2008 Seleção recorrente para três caracteres do feijoeiroBragantia 67:833-838

[25] Morais SRP, Vieira AF, Almeida LCS, Rodrigues LA, Melo PGS, Faria LC, Melo LC, Pereira HS, Souza TLPO2016 Application of microsatellite markers to confirm controlled crosses and assess genetic identity in common beanCrop Breeding and Applied Biotechnology 16:234-239

[26] Mulamba NN, Mock JJ1978 Improvement of yield potential of the Eto Blanco maize (Zea mays L.) population by breeding for plant traitsEgyptian Journal of Genetics and Cytology 7:40-51

[27] Nei M1972 Genetic distance between populationsThe American Naturalist 106:283-292

[28] Pereira HS, Alvares RC, Silva FC, Faria LC, Melo LC2017 Genetic, environmental and genotype x environment interaction effects on the common bean grain yield and commercial qualitySemina: Ciências Agrárias 38:1241-1250

[29] Pereira HS, Melo LC, Wendland A, Del Peloso MJ, Faria LC, Nascente AS, Díaz JLC, Costa JGC, Carvalho HWL, Costa AF, Almeida VM, Melo CLP, Magaldi MCS, Abreu AFB, Moreira JAA, Pereira Filho IA, Cargnin A, Posse SCP, Souza Filho BF, Moura Neto FP, Guimarães CM, Braz AJBP, Ferreira SB, Marangon MA, Souza NP, Oliveira JP, Faria JC, Bassinello PZ2014 BRS Esteio - cultivar de feijoeiro comum com grãos pretos, alto potencial produtivo e resistência à antracnose. Embrapa Arroz e Feijão, Santo Antônio de Goiás , 4p. (Boletim Informativo, 213).

[30] Pereira HS, Mendonça FR, Rodrigues LL, Melo LC, Melo PGS, Faria LC, Costa AF, Carvalho HWL, Pereira Filho IA, Almeida VM2021 Selection of carioca common bean lines with slow darkeningPesquisa Agropecuária Brasileira 56:e02471

[31] Pereira HS, Mota APS, Rodrigues LA, Souza TLPO, Melo LC2019 Genetic diversity among common bean cultivars based on agronomic traits and molecular markers and application to recommendation of parent linesEuphytica 215:38

[32] Pereira MB, Vencovsky R1988 Limites da seleção recorrente. Fatores que afetam o acréscimo das frequências alélicasPesquisa Agropecuária Brasileira 23:769-780

[33] Pontes Júnior VA, Melo PGS, Pereira HS, Melo LC2016 Genetic potential of common bean progenies obtained by different breeding methods evaluated in various environments. Genetics and Molecular Research 15: gmr.15038622.

[34] Pritchard JK, Stephens M, Donnelly P2000 Inference of population structure using multilocus genotype dataGenetics 155:945-959

[35] R Core Team2024 R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna.

[36] Ramalho MAP, Nunes JAR, Pádua JMV, Toledo FHB2024 Aplicações da genética quantitativa no melhoramento de plantas autógamas. UFLA, Lavras, 456p.

[37] Resende MDV, Duarte JB2007 Precisão e controle de qualidade em experimentos de avaliação de cultivaresPesquisa Agropecuária Tropical 37:182-194

[38] Silva ACF, Melo PGS, Melo LC, Bassinello PZ, Pereira HS2013 Eficiência de métodos de melhoramento para teor de fibra e produtividade de grãos em progênies de feijoeiro comumBragantia 72:326-331

[39] Souza TLPO, Pereira HS, Del Peloso MJ, Faria LC, Costa JGC, Wendland A, Díaz JLC, Magaldi MCS, Aguiar MS, Carvalho HWL, Souza Filho BF, Melo CLP, Costa AF, Almeida VM, Posse SCP, Melo LC2019 BRS FP403: high-yielding black-seeded common bean cultivar with superior grain quality and moderate resistance to fusarium wiltCrop Breeding and Applied Biotechnology 19:240-244

[40] Valdisser PAMR, Mota APS, Bueno LG, Menezes IPP, Coelho GRC, Magalhães FOC, Vianello RP2013 Protocolo de extração de DNA e genotipagem de SSRs em larga escala para uso no melhoramento do feijoeiro comum (Phaseolus vulgaris L.). Embrapa Arroz e Feijão, Santo Antônio de Goiás , 6p. (Boletim Informativo, 208).

[41] Vencovsky R, Barriga P1992 Genética biométrica no fitomelhoramento. Sociedade Brasileira de Genética, Ribeirão Preto, 496p.

[42] Wright S1978 Evolution and the genetics of populations. University of Chicago Press, Chicago, 580p.

[43] Xin Z, Velten JP, Oliver MJ, Burke JJ2003 High-throughput DNA extraction method suitable for PCRBioTechniques 34:820-826

[44] Zucchi MI, Pereira R, Brondani V, Coelho ASG, Vencovsky R2003 Genetic structure and gene flow in Eugenia dysenterica DC in the Brazilian Cerrado utilizing SSR markersGenetics and Molecular Biology 26:449-457

Estimates	YIELD	100SW	SY
$\bar{x}$	2846	21.36	86.70
${\bar{x}}_{s e l e c t e d}$	3097.50	22.18	89.94
$S D$	251.46	0.82	3.23
$G S$ (%)	6.01	3.47	2.77
$G S$ (%) - individual - YIELD	6.01	1.35	0.82
$G S$ (%) - individual - 100SW	0.90	3.47	-1.66
$G S$ (%) - individual - SY	0.65	-0.32	2.77
$G S$ (%) - SS	4.14	1.29	1.76
$G S$ (%) - KS	1.68	0.03	-0.13
$G S$ (%) - SKS	3.53	1.32	1.24
$G S$ (%) - MMI-1	5.55	1.85	0.87
$G S$ (%) - MMI-2	5.65	1.29	1.23
$G S$ (%) - MMI-3	4.21	0.73	1.38
$G S$ (%) - MMI-4	5.87	1.57	0.64