Morpho–Molecular Characterization of Brazilian Common Bean Landraces Varieties and Commercial Cultivars

Abstract

This study aimed to characterize forty genotypes (31 landrace varieties and nine commercial cultivars) of common bean (Phaseolus vulgaris L.) morphologically and molecularly. Morphological descriptors were evaluated during seedling, flowering, physiological maturation and post-harvest stages. Ten microsatellite markers were used for molecular screening. The markers were analyzed according to the number of alleles per locus, the allele frequency per locus and the polymorphism in content (PIC). Genetic distances and cluster analysis were performed using Bayesian inference and the UPGMA method. All black beans evaluated have anthocyanin in the cotyledons, hypocotyls and stems, and their flowers are purple. For the growth habit, 50% of the evaluated genotypes belong to type III, and at the physiological maturation stage, yellow color is predominant in the pods (85%). Through the genetic dissimilarity, three groups were observed for the likelihood reason, and five groups were observed through the UPGMA method, a strong indication of the wide genetic diversity among the evaluated genotypes. All genotypes from the Andean center of origin were grouped into the same cluster.

1. Introduction

Cultivated and consumed around the world due to wide soil and climate adaptation, high protein [1], carbohydrates, dietary fiber and minerals contents [2], common beans (Phaseolus vulgaris L.) are one of the most important foods in the human diet [3]. In addition to the benefits to human health, the consumer’s profile must be considered, given the various ways of preparing dishes, color, sizes and shape diversity, reflecting the flavor variety [4].

In Brazil, much of the genetic diversity has been traditionally maintained by family farming as a source of food, income and culture. Seeds that display greater adaptability to adverse climatic conditions and resistance to pathogens, among other characteristics of interest, are selected by the farmer for the next season. These landraces possess great allelic variability, which is important for choosing new parent genotypes with desirable characteristics for commercial production and the development of segregating populations [5]. The use of landrace varieties is very popular with small farmers in Brazil, and this germplasm is a very important source of variability for common beans. Studies about the genetic variability in common beans show the importance of the landrace varieties as a source of variability in Brazil [5,6,7,8].

There is a wide genetic and morphological diversity observed in bean plants; studies based on these parameters are complementary and help identify characteristics of agronomic interest and select superior genotypes for breeding programs [3]. The genetic variability derives from two main centers of origin (gene pools), the Mesoamerican covering the Durango, Jalisco, and Mesoamerica original races, and the Andean covering the New Granada, Peru, and Chile original races [9,10]. The Mesoamerican, composed of small (<25 g.100 grains⁻¹) and medium-sized beans (25 to 40 g.100 grains⁻¹), is the most consumed and cultivated in Brazil and presents several groups with emphasis on carioca and black, representing 85% of Brazilian consumption. Some grains from the Andean group are also cultivated, which are red and white in color and have medium to large grains (>40 g.100 grains⁻¹) [4].

The recent research on vegetable beans, particularly in relation to molecular marker technology, has shown significant advancements in breeding and genetic improvement. Molecular markers such as SSRs (simple sequence repeats) and SNPs (single nucleotide polymorphisms) are increasingly employed to enhance traits such as disease resistance, yield, and nutritional quality. Studies have demonstrated the efficacy of these markers in constructing high-density genetic maps and facilitating marker-assisted selection (MAS) [10,11]. Additionally, the integration of genomic tools with traditional breeding methods is enabling more precise and efficient development of new vegetable bean varieties [12]. This synergy between molecular marker technology and breeding practices promises to accelerate improvements in vegetable bean production, addressing both consumer preferences and environmental challenges.

Parenting selection can be assisted by different data analyses based on plant behavior, crossbreeding, or those of a predictive nature based on differences between the parents (morphological, physiological, molecular), using a measure of similarity or dissimilarity in order to evaluate the diversity between the genotypes [13]. Thus, the aim of this work was to evaluate the genetic diversity of thirty-one landraces and nine commercial cultivars of common beans by morphological and molecular characterization, using different methodologies to determine their genetic dissimilarity.

2. Materials and Methods

The field experiment was conducted in the experimental area of the Universidade Tecnológica Federal do Paraná (Dois Vizinhos campus) (25°41′ S, 53°05′ W, 526 m above sea level) in the Southwest region of the state of Paraná, Brazil, during dry crop season (second season, January to April) of 2018. The local soil is described as Red Latosol [14] with a clay texture (773 g.kg⁻¹ of clay). The climate of the region, according to the Köppen classification, is Cfa type, humid subtropical, with no defined dry season [15]. Annual precipitation for the region varies from 2200 to 2400 mm. year⁻¹ [16] with a total of 453.6 mm (113.4 mm. month⁻¹) in the experiment period, with an average temperature of 19.75 °C.

Forty genotypes of common bean were evaluated, divided into thirty-one landraces and nine commercial cultivars (Figure 1). The plants were grown under field conditions, with seeds homogeneously distributed in the furrows at a depth of approximately 2.5 cm. Fertilization was carried out in the furrow with 350 kg. ha⁻¹ of the formula NPK (05-20-20) following the manual planting of the seeds. Agricultural practices and plant protection management were carried out to control weeds and pests, as well as nitrogen complementary fertilization (100 kg. ha⁻¹ of N as Urea applied in V3), according to the technical recommendations for the crop [17].

The experimental design was complete randomized blocks, with three replications, with individual plants being evaluated within each replication (block). The size of the experimental plot was 3.6 m², composed of two rows of 4 m with 0.45 m between rows, with 20 plants per row. The final plant density was 350,000 plants.ha⁻¹.

Morphological characteristics were evaluated in the seedling, flowering, physiological maturity, and post-harvest stages. Evaluations of minimum descriptors were performed according to the Brazilian legislation of plant varieties protection described by Silva [18] as follows: (i) commercial group; (ii) presence or absence of anthocyanin in cotyledons, hypocotyl and stem; (iii) growth habit; (iv) color of wings and standard of flowers; (v) primary pod color; (vi) number of days until flowering; (vii) cycle (number of days until physiological maturation); (ix) primary and secondary grain colors; (x) presence or absence of forehead venations; (xi) grain shape (spherical, elliptical or oblong/kidney shape); (xii) degree of flatness; (xiii) brightness; (xiv) presence or absence of halo; (xv) halo color; and (xvi) mass of 100 seeds.

Microsatellite markers were analyzed according to the number of alleles per locus, the allele frequency per locus and the polymorphism in content (PIC) given by the equation $PIC=1-\sum _{i=1}^x{fi}^2$, where fi is the frequency of the i-th allele for a given band, added along the n alleles (

Table 1. List of the ten microsatellite markers used for DNA amplification of 40 common bean genotypes (Phaseolus vulgaris L.) with their respective annealing temperature (AT °C).

Nº	SSR Nomenclature	Reason	AT °C	Authors
01	SSR-IAC67	(GT)7	56	[19]
02	SSR-IAC183	(AG)18 A (AC)4	56	[19]
03	SSR-IAC268	(TC)9	56	[20]
04	SSR-IAC245	(AT)2 (GT)3	56	[20]
05	SSR-IAC284	(CT)13	56	[20]
06	SSR-IAC262	(GT)9	56	[20]
07	SSR-IAC276	(GAA)4(GT)7	56	[20]
08	SSR-IAC282	(CAA)2T(CA)3	56	[20]
09	SSR-IAC65	(TG)5	56	[19]
10	SSR-IAC239	(AG)15	56	[19]

).

Reading of polymorphic bands was performed according to the molecular weight of each allele and with the aid of the ladder used (100 bp Plus Opti-DNA Marker). Distances were calculated using the method described by Bruvo et al. [21], and cluster analysis was performed using the UPGMA (Unweighted Pair Group Mean Average) method, using the DARwin Program version 6.0 [22]. Graphical representation of the forty genotypes was performed through Bayesian inference with Structure software version 2.3.4 [23,24] with values ranging from K = 2 to K = 10. Five repetitions were performed for each K value, using the No Admixture Model with 200,000 burn-in periods and 500,000 Markov Chain Monte Carlo simulations (MCMC). To verify which K was the most appropriate to infer the clusters, the likelihood ratio (LnPD) was calculated, according to Evanno et al. [24]. Descriptive analysis was used for morphological parameters, with data presentation according to the minimum descriptors described by Silva [18]. The dendrogram represents the UPGMA grouping based on both morphological and molecular data, with the morphological descriptors presented on a scale and standardized as the molecular data. The data analysis was performed using the R software version 4.4.1 [25].

3. Results

Based on the color groups, 13 genotypes were classified as black (32.5%), 10 as Carioca (25%), 1 as Rosinha (2.5%), 1 as Pinto (2.5%), 1 as Mulatinho (2.5%), 2 as red (5%), 3 as cranberry (7.5%), and 10 as others (25%). Regarding the growth habit, 5 genotypes were classified under type I (determined growth and shrub size) (12.5%), 15 under type II (erect and shrubby growth, presence of guides with more than 12 knots) (37.5%), 20 under type III (prostrate or semi-creeper growth, with shrub bearing in favorable environments and longer guides than type II) (50%), and none under type IV (climber and undetermined) (

Table 2. Minimum descriptors of common bean (Phaseolus vulgaris L.): commercial group (CG), growth habit (GH), presence of anthocyanin (ANT), flower color (wing and standard—FC), primary pod color (PPC), number of days to flowering (NDF) and number of days from emergence to harvest maturity (CYCLE).

Genotype	Minimum Descriptors
	CG	GH	ANT	FC	PPC	NDF	CYCLE
1. Serrana Vagem Roxa	Black	II	Present	Purple	Purple	48 days	90 days
2. 90 Dias Preto	Black	II	Present	Purple	Yellow	47 days	94 days
3. Rosinha	Rosinha	II	Absent	White	Yellow	48 days	95 days
4. Vagem Branca Lustroso	Black	III	Present	Purple	Yellow	35 days	75 days
5. Carioca Siriri	Carioca	III	Absent	White	Yellow	43 days	85 days
6. Carioca Um Rajado	Others	II	Present	Purple	Purple	43 days	85 days
7. Cavalo BR UM	Cranberry	I	Absent	Pink	Yellow	35 days	74 days
8. Chumbinho Preto	Black	III	Present	Purple	Yellow	44 days	87 days
9. Gralha Coop	Black	III	Present	Purple	Yellow	43 days	95 days
10. Pardinho Mineiro	Others	III	Absent	White	Yellow	44 days	95 days
11. Argentino	Red	I	Absent	White	Yellow	36 days	79 days
12. Mourinho	Others	II	Absent	Purple	Yellow	44 days	95 days
13. Taquara	Black	II	Present	Purple	Yellow	42 days	95 days
14. Serrana Vagem Branca	Black	II	Present	Purple	Yellow	41 days	89 days
15. Iapar 40	Carioca	III	Absent	White	Yellow	43 days	95 days
16. Marronze	Others	III	Absent	White	Yellow	43 days	95 days
17. Carioca Iapar 16	Carioca	II	Absent	White	Yellow	41 days	95 days
18. Chumb. Preto Lus. troso	Black	III	Present	Purple	Purple	41 days	95 days
19. Pardinho	Others	III	Absent	White	Yellow	41 days	88 days
20. Carioca Vermelho	Others	II	Present	Purple	Purple	42 days	87 days
21. Gralha MST	Black	II	Present	Purple	Yellow	44 days	88 days
22. Vinagrinho	Others	II	Absent	Purple	Yellow	44 days	95 days
23. Cavalo UM PR	Cranberry	I	Absent	Pink	Yellow	39 days	77 days
24. IPR Rajado	Cranberry	I	Absent	Pink	Yellow	37 days	79 days
25. Chumbinho	Black	III	Present	Purple	Yellow	47 days	95 days
26. Vermelho	Red	I	Absent	Pink	Yellow	36 days	79 days
27. Pombinho	Others	III	Absent	White	Red	43 days	102 days
28. Vagem Roxa Seca	Black	III	Present	Purple	Purple	41 days	87 days
29. Mulatinho	Mulatinho	II	Present	Purple	Yellow	42 days	88 days
30. Carioca Rosa	Others	III	Absent	Purple	Yellow	46 days	100 days
31. Carijó	Pinto	III	Absent	White	Yellow	44 days	95 days
32. Bem-te-vi	Carioca	III	Absent	White	Yellow	44 days	90 days
33. Campos Gerais	Carioca	III	Absent	White	Yellow	46 days	88 days
34. ANFC 9	Carioca	III	Absent	White	Yellow	46 days	95 days
35. Nhambu	Black	II	Present	Purple	Yellow	39 days	90 days
36. Bola Cheia	Carioca	III	Absent	White	Yellow	44 days	90 days
37. Quero-quero	Carioca	III	Absent	White	Yellow	43 days	90 days
38. BRS Esteio	Carioca	II	Absent	White	Yellow	44 days	90 days
39. BRS Dama	Carioca	III	Absent	White	Yellow	43 days	95 days
40. BRS Expedito	Black	II	Present	Purple	Yellow	41 days	88 days

I: determined growth and shrub size; II: erect and shrubby growth, presence of guides with more than 12 knots; III: prostrate or semi-creeper growth, with shrub bearing in favorable environments and longer guides than type II.

). Regarding the flower color, 19 genotypes had purple flowers (47.5%), 17 had white flowers (42.5%), and 4 had pink flowers (10%). For primary pod color, 5 genotypes were purple (12.5%), 34 were yellow (85%), and 1 was red (2.5%). The number of days until flowering (NDF) varied from 35 to 48 days, with 1 genotype classified as an early cycle (<75 days from emergence to maturation) (2.5%), 6 as a semi-early cycle (75 to 85 days) (15%), 31 as a normal cycle (86 to 95 days) (77.5%), and 2 as a late cycle (>95 days) (5%) (Table 2).

In this study, varieties with shorter cycles and NDF were also classified under type I growth habit, with red or cranberry grain types reflecting characteristics of the Andean gene pool.

Morphological descriptors resulted in 17 genotypes with primary and secondary grain colors (varying from cream to reddish brown for primary color and pink to black as secondary) (42.5%), while 23 genotypes were uniform in their grain color (varying from cream to black) (57.5%) (

Table 3. Morphological descriptors for common bean (Phaseolus vulgaris L.), primary (%PC) and secondary (%SC) colors, presence of forehead venations (PFV), shape (S), degree of flatness (DF), brightness (B), halo (H) and mass of one hundred grains (MHG).

Genotpype	(%PC and %SC)	PFV	S	DF	B	H	MHG (g)
1. Serrana Vagem Roxa	100% Black	P	El	F	O	P	21.47
2. 90 Dias Preto	100% Black	P	El	F	O	P	21.10
3. Rosinha	100% Pink	P	So	Hf	G	P	18.07
4. Vagem Branca Lustroso	100% Black	A	Sf	F	G	P	19.23
5. Carioca Siriri	60% Cream and 40% Brown	P	El	F	O	P	21.90
6. Carioca Um Rajado	80% Reddish brown and 20% Black	P	El	Hf	O	P	18.47
7. Cavalo BR UM	60% Cream and 40% Pink	P	Ob	Hf	I	P	29.37
8. Chumbinho Preto	100% Black	A	Sf	Hf	G	P	21.33
9. Gralha Coop	100% Black	A	So	F	O	P	20.23
10. Pardinho Mineiro	100% Greenish brown	P	El	Hf	O	P	19.10
11. Argentino	100% Red	P	So	Hf	I	P	24.63
12. Mourinho	100% Grayish	A	El	Hf	I	P	22.80
13. Taquara	100% Black	P	El	F	O	P	21.63
14. Serrana Vagem Branca	100% Black	P	El	F	O	P	21.50
15. Iapar 40	80% Cream and 20% Brown	P	El	Hf	O	P	25.47
16. Marronze	100% Brown	P	El	Hf	O	P	19.10
17. Carioca Iapar 16	70% Cream and 30% Brown	P	El	Hf	O	P	25.53
18. Chumb. Preto Lustroso	100% Black	A	Sf	Hf	I	P	22.03
19. Pardinho	100% Light Brown	P	El	F	I	P	23.83
20. Carioca Vermelho	80% Reddish brown and 20% Black	P	El	Hf	I	P	22.23
21. Gralha MST	100% Black	P	El	F	O	P	21.10
22. Vinagrinho	100% Brown	P	Sf	F	G	P	20.80
23. Cavalo UM PR	80% Red and 20% Black	P	Ob	Hf	O	P	27.17
24. IPR Rajado	80% Cream and 20% Red	P	So	Hf	O	P	26.77
25. Chumbinho	100% Black	P	Sf	Hf	O	P	23.57
26. Vermelho	100% Red	P	So	Hf	I	P	20.77
27. Pombinho	100% Brown	P	Sf	Hf	O	P	18.77
28. Vagem Roxa Seca	100% Black	P	El	Hf	O	P	20.03
29. Mulatinho	100% Cream	P	El	F	O	P	23.77
30. Carioca Rosa	90% Pinkish e 10% Brown	P	Sf	F	G	P	17.37
31. Carijó	70% Cream and 30% Brown	P	El	Hf	O	P	21.37
32. Bem-te-vi	80% Cream and 20% Brown	P	El	F	O	P	20.43
33. Campos Gerais	80% Cream and 20% Brown	P	El	Hf	O	P	21.07
34. ANFC 9	80% Cream and 20% Light Brown	P	El	Hf	O	P	21.90
35. Nhambu	100% Black	P	El	Hf	O	P	22.23
36. Bola Cheia	80% Cream and 20% Brown	P	El	Hf	O	P	25.00
37. Quero-quero	80% Cream and 20% Brown	P	El	Hf	O	P	23.37
38. BRS Esteio	80% Cream and 20% Brown	P	El	F	O	P	22.87
39. BRS Dama	80% Cream and 20% Light Brown	P	El	Hf	O	P	20.17
40. BRS Expedito	100% Black	P	El	Hf	O	P	21.40

P: present; A: absent; Sf: spherical; El: elliptical; So: short oblong; Ob: oblong; F: flat; Hf: semi-flat; O: opaque; I: intermediate; G: glossy.

). Most of the genotypes (87.5%) presented forehead venations, elliptical (65%), half-full (65%) and opaque (70%) grains with the presence of halo (100%) and mass of 100 grains variating between 18.07 g and 29.37 g. Within the 35% of genotypes with flat grains, half belong to the black group and the other half to the Carioca and other groups (Table 3).

Molecular analysis resulted in all 10 microsatellites with polymorphic patterns with 43 polymorphic bands generated, with 2 to 6 alleles and an average of 4.2 alleles per locus (

Table 4. Molecular data for the size of the amplified fragments, number of alleles detected and polymorphism in content (PIC) of 10 microsatellites.

Nº	SSR Nomenclature	Size of Amplified Fragments	Number of Alleles	PIC
01	SSR-IAC67	102–108	4	0.63
02	SSR-IAC183	182–200	5	0.66
03	SSR-IAC268	186–200	3	0.33
04	SSR-IAC245	186–204	5	0.55
05	SSR-IAC284	202–208	4	0.58
06	SSR-IAC262	202–210	5	0.60
07	SSR-IAC276	186–206	6	0.65
08	SSR-IAC282	292–298	4	0.61
09	SSR-IAC65	296–298	2	0.37
10	SSR-IAC239	202–208	4	0.63

). According to Botstein et al. [26], markers were considered moderately to very informative, with the marker SSR-IAC183 being the most informative (0.66) and the least informative being SSR-IAC168 (0.33), with an average value of 0.56 (Table 4).

Genotypes were divided into three subpopulations (K = 3) according to the Bayesian inference, suggesting a robust and consistent genetic structuring pattern within the genotypes in this study (Figure 2). Subpopulation 1 (red) has 11 genotypes, of which 7 are commercial cultivars (ANFC 9, Bola-Cheia, Quero-Quero, BRS Esteio, BRS Dama, Nhambu and BRS Expedito) and 4 are landraces (Cavalo BR Um, Argentino, Mourinho and Vermelho). Subpopulation 2 (green) has 19 genotypes, all landraces). Subpopulation 3 (blue) has 10 genotypes, of which 2 are the remaining commercial cultivars (Bem-Te-Vi and Campos Gerais) and 8 the remaining landraces (Cavalo Um PR, IPR Rajado, Chumbinho, Pombinho, Vagem Roxa Seca, Mulatinho, Carioca Rosa and Carijó).

The unrooted dendrogram based on the UPGMA method revealed high variability among genotypes forming five clusters (Figure 3). The first cluster (purple) has eight genotypes, six commercial cultivars (ANFC9, Nhambu, Bola-Cheia, Quero-Quero, BRS Esteio and BRS Dama) and two landraces (Mourinho and Taquara). The second cluster (blue) has six genotypes, all landraces (Vermelho, IPR Rajado, Cavalo Um PR, Argentino and Cavalo BR Um). The third cluster (pink) has 13 genotypes, 10 are landraces (Gralha Coop, Pardinho Mineiro, Carioca Vermelho, Gralha MST, Chumbinho, Pombinho, Vagem Roxa Seca, Mulatinho, Carioca Rosa and Carijó), and 3 are the remaining commercial cultivars (Bem-Te-Vi, Campos Gerais and BRS Expedito). The fourth cluster (green) has 12 genotypes, all landraces (Serrana Vagem Roxa, 90 Dias Preto, Vagem Branca Lustroso, Carioca Siriri, Carioca Um Rajado, Chumbinho Preto, Serrana Vagem Branca, IAPAR 40, Marronze, Carioca IAPAR 16, Chumbinho Preto Lustroso and Pardinho). The fifth cluster (orange) has the remaining landrace genotype, Rosinha.

The purple cluster from the dendrogram has predominantly Mesoamerican characteristics with all genotypes with types II and III growth habits and normal cycle (between 86 and 95 days) (Table 2). The predominance in this cluster is of Carioca grains (62.5%); absence of anthocyanin (87.5%); white flowers (62.5%); yellow as the predominant pod color (100%) (Table 2); the presence of forehead venations (87.5%); grains that are elliptical (100%), semi-flat (75%), oblong (87.5%), opaque (87.5%) and with halo (100%) (Table 3); and average mass of 100 grains of 22.5 g. The blue cluster from the dendrogram has predominantly Andean characteristics with genotypes with type I growth habit and semi-early life cycle (between 75 and 79 days) and one genotype with Mesoamerican characteristics, such as type II growth habit and normal cycle (95 days) (Table 2).

The predominance in this cluster is of cranberry grains (50%); absence of anthocyanin (100%); pink flowers (66.7%); yellow as the predominant pod color (100%) (Table 2); the presence of forehead venations (100%); grains that are oblong (50%), semi-flat (83.3%), opaque (87.5%) and with halo (100%) (Table 3); and average mass of 100 grains of 24.92 g. Two genotypes of this cluster consisted of landrace varieties with normal cycle and type I growth habit (IPR Rajado) and semi-early cycle and type I growth habit (Cavalo BR UM), both with pink flowers (Table 2). Both presented large, oblong, opaque and semi-flattened grains (Table 3). Similar characteristics to cycle, growth habit and grain characteristics are present in the genotypes of the Andean center of origin. (Figure 3).

The pink cluster from the dendrogram also has predominantly Mesoamerican characteristics with genotypes with type II or III growth habits and normal life cycles (between 87 to 102 days) (Table 2). The predominance in this cluster is of black grains (38.7%); the presence of anthocyanin (53.9%); purple flowers (61.5%); yellow as the predominant pod color (76.9%) (Table 2); the presence of forehead venations (92.3%); grains that are elliptical (69.2%), semi-flat (61.5%), opaque (84.6%) and with halo (100%) (Table 3); and average mass of 100 grains of 20.80 g. The orange cluster also presents Mesoamerican characteristics, type II growth habit and normal life cycle (95 days) (Table 2). The only genotype in this cluster has Rosinha grains; the absence of anthocyanin; white flowers; yellow as the predominant pod color (Table 2); the presence of forehead venations; grains that are short-oblong, semi-flat, glossy and with halo; and mass of 100 grains of 18.07 g (Table 3).

The genotypes Pardinho Mineiro and Gralha Coop have a normal cycle, type III growth habit, yellow pods (Table 2), opaque grains and a similar mass of 100 grains (Table 3). These two genotypes were also together under Bayesian analysis (group 2), showing how genetically similar these genotypes were (Figure 2 and Figure 3). The genotypes Campos Gerais and IPR Bem-Te-Vi that are in the cluster together with the landrace varieties consisted of two commercial cultivars from IAPAR, which belong to the Carioca group, with normal cycle and type III growth habit (Table 2), with opaque and elliptical grains colored with similar shades of cream and brown (Table 3). Both Bayesian (Figure 2) and UPGMA (Figure 3) analysis revealed the diversity that exists between the studied genotypes and the elevated similarity among them, with some minor exceptions. However, UPGMA analysis (Figure 3) provided better discrimination of genotypes within the Bayesian groups (Figure 2), where genotypes were primarily grouped by their centers of origin.

4. Discussion

The morphological and molecular differences detected in this study corroborate the results found by other studies [20,27,28,29,30] that detected great genetic variability for local common beans through molecular tools. For breeding programs, the greater the diversity, the better, reflecting the possibility of greater gains. Thus, similar studies can be used by researchers and breeders to plan new crossings to generate elite cultivars and to meet farmers’ and consumers’ needs [31]. The life cycle can vary from 65 to 120 days, depending on the genotype, edaphoclimatic conditions and crop season, so it is possible to have three crop seasons during the year, depending on the environmental conditions [32,33,34]. In this study, genotypes with normal cycles prevailed, with only a few extremes (five varieties of early and two with late life cycles) (Table 2).

There are regional preferences in different locations in Brazil [34], particularly where low and medium-technological rural producers and family farming are more concentrated, with a predominance of landrace genotypes with grains classified under black and Carioca groups. Black grains have a nutraceutical appeal because dark-colored beans have higher anthocyanin content than light-colored ones [35,36]. The presence of anthocyanin in parts of the plant is already well studied [37,38,39]. This characteristic is desired in new varieties given the antioxidant activity of these compounds and easily identified in bean cultivars that present dark-colored grains (Figure 1).

Besides the nutraceutical appeal, consumers also rely on the visual appeal of grains. Differences in the preferences among all regions of Brazil lead to a large variety of morphological traits such as color (one or two), form, shine and the presence or absence of other traits such as streaks or spots [40]. This variability was observed within evaluated genotypes. In Brazil, black and Carioca groups are the most accepted ones, tending to have elliptical and semi-flattened grains [4,40]. The Carioca group with beige grains with brown streaks is the most consumed in Brazil, as lighter-colored grains were correlated to the preference for consumption and the commercial value of the product [34]. This was partially reflected in this study, with the Carioca beans being the second most predominant group. The predominant morphological characteristics that most impacted the separation of the groups were morphological grain traits such as its shape, degree of flattening, color, brightness, color of the flower, of the pods, growth habit and cycle. These results are in agreement with the grouping performed by Kloster et al. [41] and Rana et al. [30] for common bean cultivars, and it also relates their grouping to the two centers of origin of the common bean (Andean and Mesoamerican).

When characterizing grains of the black and Carioca groups as the most important types in Brazil and belonging to Mesoamerican origin center [42], it emphasizes that the genetic basis of bean breeding programs in Brazil comes from Mesoamerican accessions. This also occurs in the results found in the present study, as grains of black and carioca type were not grouped in the center of Andean origin, separated into genotypes with larger grains which were grouped separately from the genotypes with characteristics from the Andean Pole, as also observed by Catarcione et al. [27].

Many studies have been carried out in recent years to characterize bean genotypes in relation to their domestication centers [43,44,45], evaluated the diversity of bean genotypes using microsatellite markers and managed to group the genotypes according to the domestication centers. Researching the implications of univariate and multivariate analysis on the dissimilarity of bean accessions [42,46] has also been successful in grouping bean genotypes according to their centers of origin.

In terms of molecular characteristics, the degree of polymorphism between possible parents can be used to indicate/select recombinant genotypes based on the studied markers [47]. The results found for the microsatellites SSR-IAC65, SSR-IAC67, SSR-IAC183 and SSR-IAC239 are superior to the results found by other authors [19] for Carioca beans, where the authors detected the absence of polymorphism for the SSR-IAC67 marker. However, this marker was polymorphic and had the second-highest PIC of the evaluated SSRs, presenting four alleles and showing a greater diversity among evaluated genotypes in this study, agreeing with other studies [30,48].

UPGMA separated two commercial cultivars from the Rural Development Institute (IDR) (Bem-Te-Vi e Campos Gerais) from the others (Figure 2) that belong to the Carioca commercial group. Each breeding institution has different objectives and explores different sources of germplasm. We assume that similar parents were used in the creation of these cultivars and, therefore, genetic similarity is high, as observed by Pereira et al. [34] analyzing cultivars from Embrapa when cultivars from the Carioca group grouped together. In this study, clusters were formed by similar genotypes, corresponding to their morphologic similarity, types of varieties (landrace or commercial cultivars) and from the same breeding program.

5. Conclusions

It was possible to detect dissimilarity between the genotypes studied, evidencing genetic variability for the landraces and commercial cultivars used in Brazil.

Andean group landraces and the commercial cultivars were grouped in the same groups, evidencing the accuracy of the clustering performed.

The methodologies employed were complementary, indicating that they can be used together in studies on genetic variability.