Wudpecker Journal of Agricultural Research Vol. 1(6), pp. 223 - 232, July 2012 Available online at http://www.wudpeckerresearchjournals.org 2012 Wudpecker Research Journals Full Length Research Paper

Greengram germplasm for constituting of core collection

Pandiyan M., N.Senthil, D. Packiaraj, S. Jagadeesh

National Pulses Research Centre, Vamban colony – 622 303, Pudukkottai District, Tamil Nadu, South India.

Accepted 13 June 2012

With the objectives to constitute a core collection the available 646 germplasm accessions of greengram were characterized for 18 quantitative traits and 37 qualitative traits listed in NBPGR minimal descriptor. Two multivariate statistical approaches viz., Hierarchical cluster analysis and Principal component analysis were utilized for forming a core consisting of 65 accessions (approximately 10% of the base collection). For selection of parental lines with maximum genetic diversity, the 65 accessions of the core collection were subjected to Hierarchical cluster analysis and Principal component analysis. The 65 accessions of the core were grouped into nine clusters indicating existence of enormous genetic diversity among the accessions of the core. The cluster mean performance for 18 quantitative characters in the core revealed high mean performance for the characters namely plant height, number of branches per plant, length of branch, number of clusters per branch, number of clusters per plant and number of pods per plant registered by cluster VI among the nine clusters. The cluster IV exhibited highest mean value for the characters, length of pod, seed length and seed breadth. Among the nine clusters, the cluster VII registered the highest mean performance for many traits namely number of seeds per pod, dry matter production, protein content, days to initial flowering, days to 50 per cent flowering, days to initial maturity and days to full maturity. Clusters VIII and I registered highest mean value for hundred seed weight and single plant yield respectively. Hence the accessions of these clusters can be utilized for hybridization programme. Among the 65 accessions AC 285, LM 196, AC 198, EC 396124, EC 396125, PLS 287, PLS 315, VBN1 and MDU 3405 were found promising for high per se performance for multiple quantitative traits. The accessions CO 4, VRM (Gg) 1, K1, Pusa bold, ML 267, ML 682, VBN (Gg) 2 and Vellore local were found to perform moderately for various traits. Key words: Greengram, germplasm, genetic diversity

INTRODUCTION Vigna radiata (L.) Wilczek, commonly known as greengram or mungbean is the most widely distributed species among the six Asiatic Vigna species. It is one of the predominant sources of protein and certain essential amino acids like lysine and tryptophan in vegetarian diets. It possessed certain added features compared to other pulses. It is relatively drought tolerant and well adapted to a range of soil conditions including light soils and can thrive even under limited irrigation, more over, it is suited for crop rotation and crop mixtures Baldev (1988) and Sadaphal (1988). However, this crop is *Corresponding author Email: mpandiyan8@yahoo.co.in.

suffering from the yield advantage as realized in case of C4 cereals.

Presently, the yield level of green gram as like other pulses is well below the optimum level. The average yield of mungbean is very low not only in India (425 kg/ha) but in entire tropical and subtropical Asia. In Tamil Nadu it is cultivated in an area of 0.163million hectare with a production of 0.78 lakhs tonnes. Besides management factors the prime cause for the low productivity can be ascribed to the inherently low yielding potential of the cultivars coupled with susceptibility to diseases. Stepwise utilization of primary gene pools of this crop can result in tremendous improvement in yield when we used distinct variability available germplasm lines. In order to utilize the variability available in the primary gene pool, it

is imperative to critically characterize and evaluate the available germplasm collections, establishing core collections and employing divergent parents to obtain useful segregants. The core collection concept was suggested by Frankel and Brown (1984), who defined it as a representative sample of a collection where as much as possible of the diversity of the collection, is retained with minimum redundancy.

The core collection is a logical first step in the screening for desirable alleles Holbrook and Anderson (1995). The economy of size makes it possible to increase the number of characteristics evaluated and or to promote the use of other techniques for screening such as molecular markers. Selection of parents from the core collection for hybridization program is likely to generate superior transgressive segregants With a view to evaluate the available germplasm using the descriptors and form the core collection, this study was taken up keeping the following objectives in mind. 1. To evaluate and characterize available germplasm of green gram using NBPGR descriptors. 2. To perform hierarchical cluster analysis and ordinate analysis of available germplasm and to establish core collection following specific scientific principles. 3. To assess genetic components and to perform association analysis among yield and yield attributes within the core collection. MATERIALS AND METHODS Seeds of 646 accessions of greengram were collected from the Department of Pulses and utilized for the study. The field experiments were carried out at Department of Pulses, Centre for Plant Breeding and Genetics (CPBG), Tamil Nadu Agricultural University, Coimbatore during 2002 to 2004 Rabi and Kharif seasons. Each genotype was sown in two single row plot each of four metre length, plants were raised at a spacing of 30 x 10 cm and replicated twice.

The package of practices recommended in the crop production guide was followed. Eighteen quantitative traits viz., plant height (cm), number of branches per plant ,length of branch (cm), days to initial flowering, days to 50% flowering, number of clusters per branch, number of clusters per plant, number of pods per plant, pod length (cm), number of seeds per pod, seed length (cm), seed breadth (cm), hundred seed weight (g), single plant yield (g), dry matter production (g), protein content (%), days to initial maturity and days to full maturity and 37 qualitative characters like days to emergence, growth habit, growth pattern, hypocotyl colour, stem colour, primary leaf shape, terminal leaflet shape, terminal leaflet length, leaf pubescence, leaf colour, petiole colour, petiole length, leaf senescence, peduncle length, raceme position, calyx colour, corolla colour, pod colour at immature pod, colour of ventral suture of immature pod, pod colour at mature stage, attachment of mature pod to peduncle, pod pubescence, pod peak shape, cross-section of the mature pods, seed colour, luster on seed surface, seed shape, twinning tendency, branching pattern, leafiness, flowering period, shattering in the field, reaction to cercospora leaf spot, reaction to powdery mildew, reaction to bacterial leaf spot, reaction to mungbean yellow mosaic virus and reaction to bruchids infestation listed in the NBPGR minimal descriptors The scaling adopted was as per the NBPGR minimal descriptors for greengram

Pandiyan et al. 224 (Kawalkar et al. (1996) ) were recorded in five randomly selected plants in each of the accession per replication.

For hierarchical cluster analysis, the computer software NTSYS pc was used by Rohlf (1992). A phenetic tree was constructed using the TREEPLOT programme of NTSYS pc after standardization of the data. The PCA was performed as adopted by Bisht (1998) in order to confirm the diversity patterns brought out by cluster analysis. The cumulative percentage contribution to variation of different traits to Eigen vector was also computed. The selection of the core was done based on the combination of logarithmic strategy (L) and Relative Diversity (RD) method as described by Diwan et al. (1995) which considered the variation inside the group prior to the selection of the accession. The steps involved in building the core were as follows: 1. Fixing the size of core collection at 10 per cent level as suggested by Brown (1989a) 2. For fixing the number of accessions for each cluster resulting from cluster analysis, logarithmic strategy suggested by Brown (1989 b) was adopted. The logarithmic strategy adopted is as follow:

3. Performing second level of cluster analysis (cluster within cluster) for further assortment of accessions. 4. Selection of accessions as fixed in step 2 from each second level cluster at random. 5. These accession put together constitute core. RESULTS The result of HCA utilizing 18 quantitative traits on 646 germplasm accessions is presented as a dendrogram (Figure 1). Eight distinct clusters could be recognized from the tree. The composition of clusters is presented in Table 1. The cluster II, V and I were large with 195, 144, and 133 accessions respectively while the cluster III, VI and IV were medium sized with 86, 49 and 27 accessions respectively. The cluster VII and VIII were the smallest with 6 accessions each.

The PCA revealed that the first three principal components accounted for 33.61, 15.69 and 11.16 per cent of the total variability respectively, cumulatively accounting for 60.47% (Table 2). Principal component analysis scatter plot of first three PCA axes is shown in Figure 2. The distribution of accession along this axes confirmed the grouping obtained through hierarchical cluster analysis.

To constitute the core and then selection of one accession from each cluster at random. For fixing the

clusternumbertotalmeicollectionbasecoreforselectedbetoaccessionsofNumbern

clusteriofproportionPclustertheofsizep

characterainselectedaccessionsofnumberSwhere

np

pS

thi

t

m

i

)65..%10(

log

log1

225 Wudpecker J. Agric. Res.

Figure 1. Phenetic tree depicting the clustering pattern of 646 greengram base collection accessions.

number of accessions the logarithmic strategy was adopted. The results of the cluster analysis performed within the first level clusters along with the accessions selected adopting logarithmic strategy is furnished in Table 3. Ten accessions selected each from cluster I and III ,11 accessions each from cluster II and V and four accessions each from cluster VII and VIII were selected a apiece. The remaining fifteen accessions were selected from cluster IV and VI, and this constituted a total of 65 accessions in the core collection. The result of HCA utilizing 18 quantitative traits on 65 accessions from the core collection accessions is presented as a dendrogram (Figure 3). Nine distinct clusters could be recognized from the tree. The composition of clusters is presented in Table 4. The cluster I and II were large with 12 accessions each while the cluster III comprised of 11 accessions. Cluster IV and V were medium sized with 7 accessions each. Cluster VI

contained 6 accessions. The clusters VIII and IX were small with 4 and 5 accessions respectively while cluster VII was the smallest with only one accession. The cluster mean performance for 18 quantitative characters for 65 core collection accessions is furnished in Table 5. Cluster VI revealed highest mean performance for many morphological characters viz., plant height (82.45 cm), number of branches per plant (2.80), length of branches (53.80 cm), number of clusters per branch (8.95), number of clusters per plant (15.35) and number of pods per plant (41.80). The cluster IV exhibited the highest mean performance for the traits length of pod (7.92), seed length (0.42 cm) and seed breadth (0.35 cm). The highest mean performance was observed for the characters viz., number of seeds per pod (15.75), dry matter production (174.47), protein content (22.25) and earliness for maturity traits viz., days to initial flowering (36.00) , days to 50 per cent flowering (41.00) , days to initial maturity (54.00) and days to full

Pandiyan et al. 226

Table1. The composition of clusters.

Cluster No. of accession Accession Number

I 133

1,564,24,341,386,511,21,524,25,351,530,172,354,324,334,3,497,456,500,150,153,507, 312,336,399,29,290,174,184,358,408,175,343,385,640,498,187,361,582,369,333,335,368,402,583,512,2,631,137,628,120,322,390,494,163,340,419,84,128,227,140,633,180,431,364,585,565,611,26,178,183,209,328,103,323,533,392,421,30,296,303,307,586,32,119,382,233,302,91,319,96,186,367,243,38,79,448,105,225,215,542,301,393,434,240,253,460,355,372,379,136,255,377,4,164,210,213,501,627,562,383,489,514,645,510,529,156,315,531,23,19, 46,403

II 195

203,239,223,374,366,320,332,20,159,220,237,550,625,48,196,459,234,569,622,95,346,115,289,363,144,201,316,130,297,398,375,405,33,360,477,106,182,45,587,247,362,413,304,99,236,216,147,329,580,339,291,61,493,257,345,526,581,139,321,389,151,211,305,350,86,87,88,342,629,451,444,50,160,162,344,35,195,551,418,121,443,188,254,561,55,485,574,480,420,185,238,59,85,544,577,97,296,54,63,472,337,70,71,5,18,27,194,414,25,349,404,464,94,559,537,357,454,7,10,82,198,547,56,424,57,126,108,425,423,463,566,478,646,11,436,34,432,435,468,433,563,356,299,505,541,482,539,76,83,536,560,567,430,6,124,310,427,13,338,40,74,14,473,78,232,326,540,44,51,170,318,387,47,441,179,438,245,409,571,327,401,380,437,36,447,412,442,426,453,538,553,58,117, 546, 249

III 86

578,407,42,110,330,570,53,104,113,395,90,522,446,161,400,548,568,219,579,572,17,65,67,9,231,294,557,347,428,197,248,222,226,417,12,406,77,235,410,73,308,92,242,306,359,370,317,49,429,623,31,260,348,277,449,15,101,43,141,41,66,93,112,397,440,116,252,64,123,131,39,556,462,80,470,545,250,543,584,37,552,75,16, 122,230,411

IV 27 132,8,206,176,626,193,177,415,624,558,60,102,158,644,143,246,309,488,373,554,166,157,259,169,589,276, 261

V 144

266,62,129,146,461,69,72,125,465,450,452,535,28,313,311,515,314,643,516,481,142,528,457,154,171,325,590,165,635,384,518,466,394,492,487,591,612,532,149,479,504,156,168,513,467,509,98,228,388,549,597,608,641,214,523,109,189,474,300,598,221,458,639,471,200,439,602,642,293,609,576,118,603,615,52,207,422,114,376,555,81,107,445,217,490,573,378,506,575,476,22,416,190,371,138,181,204,630,241,396,508,525,588,111,212,224,331,455,148,191,496,202,258,288,475,527,495,68,519,520,521,517,192,292,391,502,134,365,208,135,167,127,173,133,353, 155,483,486,499,607,610,491,503,613

VI 49 595,643,601,268,534,218,484,617,618,632,619,244,592,638,636,599,637,593,594,600,606,605,620,621,614,596,616,89.100,145,298,469,256,273,275,283,269,279,265,285,281,282,278,280,286,604,287,199,205

VII 6 262,263,267,229,271,264 VIII 6 284,270,272,274,381,352

Table 2. Extraction of principal component for base collection.

PC Axes Eigen value Variance (%) Cumulative eigen value

Cumulative variance (%)

I 6.05 33.61 6.05 33.61 II 2.82 15.69 8.87 49.31 III 2.00 11.16 10.88 60.47

maturity (66.00) in the cluster VII. The highest mean value for the hundred seed weight (3.84g) was registered by the cluster VIII, while cluster I registered the highest mean value of 7.31 g for single plant yield.

The PCA revealed that the first three principal components accounted for 27.10, 23.52 and 13.16 per cent of the total variability respectively, cumulatively accounting for 63.79% (Table 6). The first eight principal components accounted for more than 90 per cent variability. Principal component analysis scatter plot of first three PCA axes is shown in Figure 4.

The distribution of accessions along these axes confirmed the grouping obtained through Hierarchical Cluster Analysis.

The contribution of various traits to the first three principal components is presented in Figure 5. Traits affecting the first principal axis were the branching characters viz., length of branch and number of branches per plant. Traits that had maximum influence on the second principal components axis were the pod characters like length of pod and number of pods per plant. Traits that separated accession along the third

227 Wudpecker J. Agric. Res.

Figure 2. Three dimensional ordination of 646 greengram germplasm accessions.

Table 3. Original cluster size and log value of each cluster for selection of core collection.

Cluster Original size

Log value

Proportions of log value

No. of accessions to be selected

No. of second level cluster formed

No. of accession selected from each cluster

Accession number

I 133 2.12 0.16 10 10 1 21,23,32,38,128,164,240,377,582,645

II 195 2.29 0.17 11 11 1 130,162,198,374,412,418,427,436,505,526,560,

III 86 1.93 0.14 10 10 1 16,67,93,131,330,348, 406,411,543,548

IV 27 1.43 0.10 7 7 1 8,60,132,259,276,488, 558

V 144 2.15 0.16 11 11 1 28,62,68,72,133,138,207,224,266,293,523,

VI 49 1.69 0.12 8 4 2 89,205,256,280,287,285, 619,620,

VII 6 0.77 0.05 4 2 2 229,262,264,267 VIII 6 0.77 0.05 4 2 2 270,272,352,381 Total 646 13.18 65 65

Pandiyan et al. 228

Figure 3. Phenolic tree depicting the clustering pattern of 65 greengram core collection accessions.

Table 4. Distribution of 65 genotypes in different clusters in core collection.

Cluster No of accessions Accession number I 12 8,505,558,418,488,162,436,23,138,526,164,645 II 12 38,330,548,130,374,582,377,406,412,259,266,276 III 11 89,348,21,224,523,28,293,68,32,128,240 IV 7 207, 267, 133, 619, 620, 280, 60 V 7 287,16,131,93,427,67,411 VI 6 62, 72,198,543,560,132 VII 1 205 VIII 4 262,229,256,264 IX 5 270,272,381,352,285 Total 65

principal component axis were the seed dimension related characters like seed length, seed breadth and hundred seed weight. DISCUSSION The success of hybridization programme depends upon

selection of suitable parents which are genetically diverse from the existing germplasm. Hence for selection of genetically diverse parents from the germplasm it is essential that the collection should be systematically characterized using descriptors encompassing both quantitative and qualitative traits. In addition to this systematic characterization, the development of a core collection or working germplasm representing maximum

229 Wudpecker J. Agric. Res.

Table 5. Cluster mean performance for core collection. Clusters PHT NOB BRL DIF DFF NCB NOC NPP POL NSP SDL SDB HSW SPY DMP PTC DIM DFM I 46.54 2.16 31.22 36.36 41.09 5.52 11.59 36.68 7.35 11.84 0.34 0.28 3.26 7.31 11.61 20.66 54.36 66.54 II 51.84 2.24 35.95 38.90 44.00 5.00 10.40 30.34 7.15 11.65 0.36 0.29 3.45 6.02 19.51 20.08 56.90 69.00 III 45.57 1.27 17.33 37.00 42.40 1.95 7.67 21.77 7.07 11.37 0.35 0.28 3.15 5.02 16.09 21.22 55.00 67.40 IV 37.87 1.12 11.25 38.66 44.00 1.54 6.50 14.66 7.92 11.12 0.42 0.35 3.74 3.45 28.36 20.68 56.66 69.00 V 70.32 1.91 39.25 42.50 48.33 4.95 11.23 27.12 7.33 11.54 0.34 0.28 3.33 5.55 25.14 19.51 60.50 73.33 VI 82.45 2.80 53.80 38.40 43.80 8.95 15.35 41.8 7.44 11.61 0.36 0.28 3.62 6.87 16.49 21.15 56.40 68.80 VII 31.50 1.75 21.25 36.00 41.00 3.50 8.25 18.25 6.88 15.75 0.31 0.27 3.42 2.94 174.47 22.25 54.00 66.00 VIII 50.83 2.16 28.25 39.00 43.33 3.75 8.41 24.66 6.81 11.00 0.40 0.31 3.84 5.43 75.69 20.01 57.00 68.33 IX 47.18 1.93 2.03 36.50 41.75 3.87 8.18 20.56 6.98 11.12 0.33 0.28 3.26 4.52 58.34 20.58 54.50 66.75

Table 6. Extraction of principal components for core collection.

S/No Eigen value Variance (%) Cumulative Eigen value

Cumulative variance (%)

1 4.87 27.10 4.87 27.10 2 4.23 23.52 9.10 50.63 3 2.36 13.16 11.46 63.79 4 1.27 7.05 12.73 70.85 5 1.11 6.19 13.84 77.05 6 0.97 5.42 14.81 82.47 7 0.77 4.31 15.58 86.79 8 0.68 3.80 15.26 90.59 9 0.43 2.39 15.69 92.99 10 0.35 1.98 16.04 94.98 11 0.32 1.78 16.36 96.76 12 0.19 1.05 16.55 97.82 13 0.16 0.89 16.71 98.72 14 0.08 0.46 16.79 99.18 15 0.07 0.43 16.76 99.61 16 0.05 0.32 16.81 99.94 17 0.00 0.03 16.81 99.98 18 0.00 0.01 16.81 100.00

diversity in the base collection will be of immense practical utility for mungbean improvement. Hence a core collection was established in this investigation, following recommended scientific

principles. The core collection concept was first suggested by Frankel and Brown (1984) who defined it as a representative sample of the base collection where as much as possible of the

diversity in the base collection is retained with minimum redundancy. For developing a core collection from the existing germplasm a number of methods has been proposed by many workers.

Pandiyan et al. 230

Figure 4. Three dimensional ordination of 65 greengram core collection accessions.

Among them, the method suggested by Diwan et al. (1995) which combine the logarithmic method of stratification and relative diversity principle of selection that takes into account the morphological diversity of different strata has been recommended as the best approach for developing better core with maximum diversity by Malosetti and Abadie (2001).

By applying this strategy in the existing 646 germplasm accessions a core collection consisting of 65 (approximately 10%) diverse accessions were constituted in this investigation. Constituted core collection was tested for proportion of variability of the base collection, represented by two statistical parameters viz., average range retention and Shannon diversity index studied by Shannon and Weaver, (1963). Computed values of average range retention for quantitative traits suggested

by Diwan et al. (1995) as well as Shannon diversity indices for qualitative traits were high indicating that the established core collection represented majority of genetic variation present in base collection.

It is envisaged by Marshall (1989), that core collection concept will receive priority in evaluation and characterization, so that in a short time, a large number of traits would be evaluated at different locations. This could lead to a better understanding of genetic structure of the crop and promote the exchange of information and material, thus facilitate the use of the core collection. Macky (1995), has discussed the practical application of core collection and the possible inclusion of core collection principles into gene bank information management systems. Even though the established core collection represented majority of the variation of the

231 Wudpecker J. Agric. Res.

Figure 5. Contribution of various traits to three principal components of the core collection.

base collection, it is essential to estimate the diversity represented by the core and partition the accessions based on diversity for selection of diverse parents for hybridization. For this, hierarchical cluster analysis using quantitative traits was performed.

The hierarchical cluster analysis of the quantitative traits with 65 accessions of core collection formed nine clusters which clearly indicated the existence of high level of genetic diversity in the core collection; similar findings were reported by Bisht et al. (1998) who obtained six clusters from 111 accessions in green gram. Hierarchical cluster analysis has been widely utilized for stratification of germplasm accession and selection of diverse parents by many workers in many crops like urdbean, Ghafoor and Ahmed (2002), Bacopa monnieri, Shalini et al. (2003), Sesbania Veasey et al., (2001) common bean Escribano et al., (1998) red colver Kouame and Quesenberry (1993), Ethiopian cereal tef cereal Assefa et al. (1999), sorghum Ayana and Bekele (1999) and uruguayan maize Malosetti and Abadie (2001).

Principal component analysis of the quantitative traits in the core collection supported the cluster constellation formed by hierarchical clustering procedure. Principal component has been utilized to confirm the hierarchical clustering patterns in green gram by Bisht et al. (1998) and Ghafoor and Ahmed (2002) in urd bean. The principal component analysis has thrown light on the key

characters that contribute to diversity. Among the 18 quantitative traits viz., length of branch, number of branches per plant, length of pod, number of pods per plant, seed length, seed breadth and hundred seed weight contributed maximum to the genetic diversity. Similar results were also reported by Bhist et al. (1998) in mungbean. Hence these traits can be considered as key characters for estimating genetic diversity in green gram. In present study the first three principal components accounting for 63.79 per cent of variation indicated that the graphical representation of the accessions is statistically valid. Similar high value for the first three principal components was reported by Bhist et al. (1998) in greengram and Ghafoor and Ahmad (2002) in urd bean.

The cluster mean performance for 18 quantitative characters for 65 core collection accessions is furnished in Table 1. Cluster VI revealed highest mean performance for many morphological characters viz., plant height, and number of branches per plant, length of branch, number of clusters per branch, number of clusters per plant and number of pods per plant. The cluster IV exhibited the highest mean performance for the traits, length of pod, seed length and seed breadth. The highest mean performance was observed for the characters viz., number of seeds per pod, dry matter production , protein content and earliness for maturity

traits viz., days to initial flowering, days to 50 per cent flowering, days to initial maturity and days to full maturity in the cluster VII. The highest mean value for the hundred seed weight was registered by the cluster VIII, while cluster I registered the highest mean value for single plant yield. These clusters are suggested for planning a hybridization programme in order to generate more heterotic F1s and better segregants in F2 generation. In the core, the accession AC 285 was found to exhibit maximum mean values for characters viz., plant height, number of branches per plant, length of branch and number of pods per plant. Accession LM 196 recorded minimum mean value for days to initial flowering, 50 per cent flowering, initial maturity and full maturity. AC 198 exhibited maximum mean value for the characters viz., number of clusters per branch and number of clusters per plant. Maximum values for traits length of pod and hundred seed weight were observed in EC. 396124. The accession EC 396125 recorded highest mean value for the characters seed length and seed breadth. The highest mean value for seed yield was observed in PLS 287 and highest number of seeds per pod was recorded by PLS 315 A. The high dry matter production was exhibited by the accession VBN1 and the highest protein content was registered in the accession MDU 3405. Accessions CO4, VRM (Gg) 1, K1, Pusa bold, ML 267, ML 682, VBN (Gg) 2 and Vellore local were found to perform moderately for various traits. REFERENCE Ayana A, Bekele E (1999). Multivariate analysis of morphological

variation in sorghum (Sorghum bicolor (L.) Moench) germplasm from Ethiopia and Eritrea. Genetic Resourc. and Crop Evolution. 46 (3): 273-284.

Baldev B (1988). Cropping patterns: In Pulse Crop (Grain Legumes), Eds B Baldev. S.Ramanujam and H K Jain. Oxford & IBH Publishing Co. Pvt Ltd. pp. 513-557.

Bhist IS, Mahajan, RK, Kawalkar TG (1998). Diversity in greengram (Vigna radiata (L.) Wilczek) germplasm collection and its potential use in crop improvement. Ann. Appl. Biol., 132: 301-312.

Pandiyan et al. 232 Brown AHD (1989a). The case for core collections. In: Brown AHD,

Frankel OH, Marshall DR. Williams JR (eds.) The use of plant genetic resources. Cambridge University Press, Cambridge, UK. pp. 136-156.

Brown AHD (1989b). Core collections: a practical approach to genetic resources management. Genome 31: 818-824.

Diwan N, McIntosh MS, Bauchan GR (1995). Methods of developing a core collection of annual Medicago species. Theor. Appl. Genet., 90: 755-761.

Escribano MR, Santalla PA, Ron DE (1998). Parrern of genetic diversity in land races of common bean (Phaseolus vulgaris L.) from Galicia. Pl. Breeding, 117: 49-56.

Frankel OH, Brown AHD (1984). Current plant genetic resources- a critical appraisal. In Genetics: new frontiers. Oxford & IBH. Publishing Co., New Delhi, 4: 1-11.

Ghafoor A, Ahamed Z, Qureshi AS, Basheer M (2002). Genetic relationship in V. mungo (L.) Hepper and V. radiata (L.) Wilczek based on morphological trait and SDS-PAGE. Euphytica., 123: 367-378.

Holbrook CC, Anderson WF (1995). Evaluation of a core collection to identify resistance to late leafspot in peanut. Crop Sci., 35: 1700-1702.

Kawalkar TG, Bhist IS, Mahajan RK, Patel DP, Gupta PN, Chandel.KPS (1996). Catalogue on greengram (Vigna radiata L. Wilczek) germplasm. NBPGR. New Delhi. pp 130.

Kouame CN, Quesenberry KH (1993). Cluster analysis of a world collection of red clover germplasm. Genetic Resources and Crop Evolution 40: 39-47.

Macky MC (1995). One core collection or many. In: Hodgkin T, Brown AHD, Hintum JL, Morales AAV (Eds) Core collection of Plant Genetic Res. John Wiley and Sons. pp199-210.

Malosetti M, Abadie T (2001). Sampling strategy to develop a core collection of Uruguayan maize landraces based on morphological traits. Genetic Res. and Crop Evol. 48: 381-390.

Marshall DR (1989). Limitations to the use of germplasm collection. In : Brown AHD, Frankel OH, Marshall DR, Williams (Eds), The use of Plant Genetic Res. Combridge University Press.

Rohlf FJ (1992). NTsys-PC; Numerical taxonomy and multivariate analysis system verson 1.70. State University of New York, Brook, New York.

Sadaphal MN (1988). Agronomy of pulses. In Pulse Crops ( Grain Legumes), Eds Baldev B, Ramanujam S, Jain HK. Oxford & IBH Publishing Co. Pvt Ltd. pp. 456-512.

Shalini M, Srikant S, Gupta, Sushil K (2003). Evaluation of an Indian germplasm collection of the medicinal plant Bacopa monnieri (L.) Pennell by use of multivariate approaches. Euphytica, 133: 255-265.

Shannon CE, Weaver W (1963). The mathematical Theory of Communication. Urbana illinors.USA: University of illinors press.