Race structure and relationships among ecotypes in cultivated
common bean (Phaseolus vulgaris L.).
Matthew W. Blair, Hector F. Buendía, Lucy Díaz, Juan M. Díaz, Myriam C. Duque, Steve
Kresovich, Sharon Mitchell, Maria J. Peloso, Rosana Brondani, Xiaoyan Zhang, Shumin Wang,
Teresa Avila, Ximena Rojas, Andrea Davila, Sandra Lorigados
Introduction Tables and Figures Results and Discussion
Common bean is the third most important grain legume
Table 1. Genetic diversity values for microsatellites evaluated across the core Allele sizes were estimated by comparing the
in the world produced over an area of 18 million collection. Genomic (n=20) fragment peaks with the internal Genescan-500 LIZ
hectares with large amounts of production in developing No.
AG01 9 0.590 0.186 0.542
Marker alleles Exp. Het. Obs. Het. PIC
BM137
BM139
44
19
0.947
0.749
0.000
0.077
0.945
0.735 size standard in ABI3730 electropherograms and
countries of Latin America and Eastern and Southern BM140 29 0.726 0.126 0.713
Gene-based (n=16)
BMd01 17 0.896 0.386 0.887
BM141
BM143
35
40
0.864
0.929
0.269
0.138
0.852
0.925 calculated with Genemapper v. 3.7 software.
Africa (Broughton et al., 2003). Cultivated common BMd02
BMd08
7
8
0.499
0.460
0.027
0.018
0.387
0.402
BM149
BM156
9
43
0.481
0.856
0.033
0.105
0.434
0.845
bean germplasm is especially diverse due to the
BMd15
BMd16
13
15
0.491
0.560
0.325
0.117
0.424
0.467
BM160
BM172
37
41
0.811
0.813
0.091
0.145
0.804
0.805
Raw allele size calls were binned to assign a whole
BMd17 7 0.630 0.081 0.565
existence of two genepools in the Mesoamerican and
BMd18
BMd20
12
6
0.347
0.590
0.331
0.018
0.338
0.543
BM175
BM183
BM187
24
32
59
0.807
0.809
0.946
0.051
0.113
0.291
0.784
0.790
0.944
integer allele value using the software program
Andean centers of diversity. The two genepools can
BMd46
BMd47
BMd51
3
5
4
0.507
0.511
0.011
0.009
0.016
0.004
0.386
0.393
0.011
BM188
BM200
38
55
0.859
0.872
0.721
0.186
0.846
0.865 AlleloBin. Allele data was used to estimate simple
BM201 15 0.725 0.070 0.693
be morphologically distinguished into various races
PV-ctt001
PV-ag003
PV-at003
14
9
10
0.778
0.556
0.545
0.072
0.030
0.126
0.748
0.459
0.441
BM205
GATs54
15
7
0.820
0.453
0.230
0.061
0.800
0.370
matching dissimilarity matrices and to draw Neighbor
GATs91 28 0.891 0.082 0.885
(Singh et al., 1991), however the association of these
PV-at001
PV-cct001
70
7
0.970
0.087
0.000
0.028
0.969
0.086
BMd56
Total
2
581
0.264 0.018 0.229 Joining (NJ) trees for the genotypes in Darwin software
Total 220
phenotypic divisions with genetic structure has not been
Average 13 0.527 0.099 0.469
Average
Overall Average
29.1
13.8
0.761
0.657
0.150
0.127
0.740
0.620 v. 5.0 and parameters of diversity evaluation were then
clear. In the GCP genotyping project for common bean evaluated with PowerMarker v. 3.25. Finally, the
we addressed this through a large-scale analysis of Figure 1. Dendrogram DJ1 DJ2-A
number of populations (K) was evaluated with the
based on dissimilarity software STRUCTURE assuming an admixture model
international and national germplasm collections index for the reference M2
DJ2-B
representing wide genetic variability from both primary collection of common DJ2-C with K=2 to K=15 and a total of 50,000 iterations each
bean showing the
and secondary centers of diversity using genomic and subdividisions within each M1
DJ2-D for both MCMC repetitions and burn in times.
race. Race abbreviations INT-2
genic microsatellites. For this poster we concentrate on are M: Mesoamerica, DJ: INT-1
All the markers analyzed were polymorphic with
the analysis of the CIAT core collection as it was Durango-Jalisco, P: Peru, from 2 to 59 alleles per marker (Table 1). The total
NG: Nueva Granada,
selected as a broad set of mostly landrace accessions Introg: Introgression
INT-3
number of alleles identified was 801 with an average of
from primary and secondary centers of diversity. All NG-A
P-A 22.3 alleles per locus. The most polymorphic markers
major agro-ecologies were covered in the core P-D
were Pv-at01, BM187, BM200, BM137 and BM156, all
NG-B
collection as this was selected using a geographic P-C
genomic microsatellites, while the gene-based
information systems (GIS) approach taking into account NG-D
NG-C
microsatellites were correspondingly less polymorphic.
P-B
rainfall, temperature patterns, soils and maturity period. The genetic diversity based on Nei’s index for the
Table 2. Genetic diversity parameters for clusters found within the core collection.
entire set of genotypes was 0.657, which was high
Materials and Methods Cluster Sample Size Allele No Availability Gene Observed
compared to previous studies with microsatellites (Blair
Diversity Heterozygosity
DJ1 61 6.81 0.93
(He)
0.45
(Ho)
0.15
et al., 2006; Diaz and Blair, 2006). These values were
Mesoamerican
Genotypes: About half of the CIAT core collection was DJ2
DJ1-2
M1
101
162
52
9.72
10.69
6.14
0.94
0.94
0.96
0.51
0.51
0.40
0.15
0.15
0.10
Genepool even higher for the genomic microsatellites analyzed
analyzed (600 genotypes) and included 293 M2
M1-2
54
106
7.39
8.94
0.95
0.95
0.48
0.46
0.11
0.11
alone (0.761) compared to the gene-based
INT1 24 6.22 0.92 0.54 0.27
Mesoamerican genepool accessions and 307 Andean INT2
INT3
10
19
4.00
6.69
0.91
0.91
0.48
0.61
0.12
0.22
microsatellites analyzed alone (0.527). Observed
genepool accessions. The majority of the accessions NG
P
135
144
11.86
12.00
0.92
0.93
0.45
0.49
0.08
0.13
Andean
Genepool heterozygosity was low, averaging 0.127 across all
Total 600 22 0.93 0.657 0.13
were from the primary center of diversity especially the markers (0.099 for gene based and 0.150 for genomic
countries of Mexico (183 genotypes) and Peru (172 markers). Some observed heterozygosity could be
genotypes), while the remainder were from Argentina explained by out-crossing (which can range from 1 to
(9), Brazil (27), Bolivia (11), Chile (6), Colombia (32), K=2 Figure 2. Structure 5%) while higher Ho may be due to mixed samples or
analysis with K=2 to
Costa Rica (16), Cuba (2), Dominican Republic (2), K=7 populations for
multiple banding patterns observed for some markers.
Ecuador (37), El Salvador (8), Germany (2), Guatemala reference collection Clustering results show a primary division between
of common beans.
(59), Haiti (9), Honduras (8), Nicaragua (6), United K=3
Number codes 1=D1, the Mesoamerican and Andean genepools (Figure 1,
States (3). One accession each were from Australia, 2=D2, 3=INT1,
orange and blue lines). Within Mesoamerican beans,
4=INT2, 5=INT3,
Burundi, France, Jamaica, Japan, Malawi, Rwanda and 6=M1, 7=M2, 8=NG, the division between the Mesoamerica race and the
K=4
Uganda. The accessions had the following phaseolin 9=P as described in
Durango-Jalisco group was very evident while the
Figure 1 and TAble 1
alleles: Mesoamericans (S, Sb, Sd, B and M), Andean genepool there was somewhat less diversity
Andeans (T, C, H and A). Control genotypes run K=5
overall and a continuum between the Nueva Granada
between all the studies were: Andeans (Calima/G4494 and Peru races (Figure 2, clusters).
and Chaucha Chuga/G19833) and Mesoamericans K=6
The Chile race could not be distinguished within the
(ICA Pijao/G5773 and Dorado/DOR364). Seed
Andean genepool but there was some support for a
samples for the core collection are maintained by the K=7 Guatemala race within the Mesoamerican genepool.
bean project at CIAT and were originally from the Introgression between the genepools was evident as
Genetic Resources Unit. DNA extraction were from 10 was probable introgression between cultivated and
seedlings selected at random from each accession with wild common beans. Studies of national collection
the method of Afanador et al. (1993). DNA was diluted References germplasm from Bolivia, Brazil, China and Cuba are
to 10 ng/ml for further experiments. 1. Afanador L, Hadley S, Kelly JD (1993) Adoption of a mini-prep DNA extraction
being used to determine which races are present in
method for RAPD marker analysis in common bean (Phaseolus vulgaris L). Bean
Molecular Analysis: Microsatellite amplification used Improv Coop 36:10-11 these primary and secondary centers of diversity.
2. Blair MW, Pedraza F, Buendia H, Gaitan-Solis E, Beebe S, Gepts P,Tohme J
the fluorescent marker kit that we developed as part of (2003) Development of a genome-wide anchored microsatellite map for common In conclusion, our study has shown that common
this project and were based on microsatellites selected bean (Phaseolus vulgaris L). Theor Appl Genet 107:1362-1374. bean has very significant populations structure that
3. Blair MW, Giraldo MC, Buendia HF, Tovar E, Duque MC, Beebe SE (2006)
from those used by Blair et al. (2006). The kit included Microsatellite marker diversity in common bean (Phaseolus vulgaris L.) Theor could help guide the construction of genetic crosses
a total of nine four-color marker panels for the analysis Appl Genet 113: 100–109. that maximize diversity as well as serving as a basis
4. Broughton WJ, Hernandez G, Blair MW, Beebe SE, Gepts P, Vanderleyden J
of a total of 36 individual microsatellite loci selected (2003) Beans (Phaseolus spp.); Model Food Legumes. Plant & Soil 252: 55-128 for future association studies.
based on their polymorphism information content and 5. Díaz LM, Blair MW (2006) Race structure within the Mesoamerican gene pool of
amplification signal strength and on their even common bean (Phaseolus vulgaris L.) as determined by microsatellite markers.
Theor Appl Genet 114: 143-54.
Acknowledgements
distribution in the genome. 6. Singh. S., Gepts. P. & Debouk. D (1991) Races of common bean (Phaseolus Assistance from Cornell University Biotech. Res. Ctr. And funds
vulgaris, Fabaceae). Econ. Bot. 45(3): 379-396. from the Generation Challenge Program – SP1 and from CIAT.