Evaluation of Variability in Proximate Compositions Among Accessions of Sword Bean ( Canavalia gladiata Jacq . DC ) and Jack Bean ( Canavalia ensiformis L . DC )

Exploitation of available germplasms particularly in the developing countries is still limited despite the availability of rich gene pools. Canavalia species are underutilized legumes of wide importance for humans and livestock feed. Field and laboratory studies were carried out on evaluation of variability in proximate compositions within and between fifteen genotypes of Canivalia species collected from Genetic Resources Centre, International Institute of Tropical Agriculture (IITA), Ibadan. The study was a randomized complete block design replicated three thrice. Harvested mature seeds were analysed for proximate composition. The results revealed that moisture content ranged from 26.97 to 30.05, crude protein 15.45-21.85%, lipid 7.05-29.70%, ash 10.18-17.43%, sugar 0.75-1.37%, starch 7.40-0.26% and energy levels 1016.56-1825.82 kJ 100 g-1 DM. Significant (p<0.05) variability was observed in sugar, lipid and energy contents among the accessions based on analysis of variance (ANOVA). The first principal component analysis (PCA) had the highest contribution of 31.4% to the total variation. Based on proximate composition, the 15 accessions clustered into six main groups. Lipid varied widely and significantly (p<0.05) correlated to energy. Aggregation of accessions within the same species into a cluster revealed high similarity within a biological and taxonomic unit. The similarities among these 15 accessions could allow for inter specific hybridization.


Introduction
Number of plants that feed the world is becoming smaller as the genetic diversity of many crop species are reducing too.These trends are coming at a time when human population is expanding rapidly and all available options will be needed to meet world food demand (Smith, 1998).
The natural gene pool for most crop species still remains very rich but their exploitation is limited (Vance et al., 2000).A revitalized investigation of all plant resources available worldwide is most needed.Collection and the assemblage of crop accessions for in-situ conservation are primary for genetic improvement in order to preserve the plant diversities from extinction is needful.Many plant genetic resources may have been adversely affected due to neglect and under-utilization.For instance, the most commonly cultivated legumes among peasant farmers in humid tropics are Soy bean (Glycine max), cowpea (Vigna unguiculata), groundnut (Arachis hypogaea) and lima bean (Phaseolus lunatus).However, there are many other pulses that could help meet human and livestock dietary needs; most of which are cultivated only in localized areas and are less used (Nwokolo, 1987, Vance et al., 2000).Some of these underexploited legumes include African yam bean (Sphenostylis stenocarpa), Bambara groundnut (Vigna subterranea), Pigeon pea (Cajanus cajan), Green gram (Phaseolus aureus), Winged bean (Psophocarpus tetragonolobus), Sword bean (Canavalia gladiata) Jack bean (Canavalia ensiformis), Lablab beans (Dolichos lablab L.) (Adewale, 2010;Sørensen et al., 1997).
Canavalia species ranks well among the neglected crop species (Dada et al., 2012;Ekanayake et al., 2000).The nutritional, pharmaceutical usefulness of this legume both to man and his livestock are outstanding; the seeds are used as food for man while foliage is important as meal for animals (Graham and Vance, 2003).The immature pod of most legumes is rich in protein, minerals and vitamins while the seeds are the most nutritious part (Nwokolo, 1987).The fruits of Canavalia spp.have been reported as potential sources of nutritional (Ekanayake et al., 2000), nutraceutical, (Nwokolo, 1987) and pharmaceutical benefits for humans and livestock, (Grubben and Denton, 2004).Both species of Sword bean (Canavalia gladiata) Jack bean (Canavalia ensiformis) are used in Nigeria as ornamental plants and in some places are believed to be "snake repellents".Sword bean contains gibberellin A-15, a growth-promoting hormone (Tokoya and Takama, 1981) nutritional composition.Ground samples were taken for proximate analysis.

Analyses of proximate composition and gross energy
The proximate constituents of the oven-dried materials were determined by the methods of Association of Official Analytical Chemist (AOAC, 1998).All analyses were carried out in duplicate.Crude protein was determined by Kjeldahl method using Kjeltec TM model 2300 as described in Foss Analytical Manual (2003).Gross energy of the dry material was determined against thermocouple grade benzoic acid using a Gallenkamp ballistic bomb calorimeter (Model CBB -330-0104L).

Statistical analysis
The means of the proximate compositions in each replicate were generated and subjected to ANOVA using PROC GLM procedure of Statistical Analysis Software (SAS, 2002).The means of the accessions for the various traits were separated by the least significant difference (LSD) at p< 0.05.Correlation analysis was carried out to determine the relationship among the proximate compositions of the two species.The 15 × 7 data matrix was then subjected to PCA, clustering and Fastclus analysis using SAS (version 9.1).

Results
Proximate composition of 15 accessions of Canavalia spp is presented in Tab. 1. Results revealed significant dif-and its leaf meal has been shown to be a good source of pigment for egg yolks but information on its nutritive value is quite sketchy (Vadivel and Janardhanan, 2001).Generally information on nutritive value of Canavalia species is scanty.This study was carried out to unravel the nutritive compositions as well as variation in proximate constituents of seven accessions of Canavalia gladiata and eight accessions of Canavalia ensiformis.

Materials and methods
Seeds of fifteen ascensions of Canavalia species comprising of eight accessions of sword bean (C.ensiformis) and seven accessions of jack bean (C.gladiata) were obtained from the Genetic Resources Centre, International Institute of Tropical Agriculture (IITA), Ibadan.The experiment was laid out in a randomized complete block design (RCBD) with three replicates in a field at IITA, Ibadan (Lat.7°30 ' , Long.3°54 ' ) during the early planting season ( July to December) of 2009.Planting was done at the spacing of 1 m × 1m.Two seeds were planted per hill; which was later thinned to one after seedling emergence.Field management was carried out as required.

Data collection and analysis
At maturity, matured pods were air dried and further oven dried at 70°C to constant weight.Samples were collected from oven dried seeds for proximate analysis.Twenty dried seeds each of the fifteen accessions of Canavalia species were ground to powder in order to determine their Among the seven proximate variables tested in the 15 accessions moisture content had the least variation (CV = 2.74%) among the samples, while lipid (fat) had the highest variation (CV = 44.08%).
The eigen values of the PC for seven proximate variables is presented in Tab. 2. The first five PC -axes explained the total variation within the 15 accessions.The first three axes had eigen values ≥ 1.0 and explained 75% of the total variation within the studied accessions.The first PC-axis made the highest contribution (31.4%) to explaining the total variation in the proximate compositions.
The seven variables had significant contribution to the variances within each PC-axis.Energy, fat, moisture content and sugar were the four variables which significantly contributed to the quantity of variance in the PC1 axis.In PC2, moisture content, sugar, ash and protein were most significant; while sugar, starch and ash were significant for PC3.Sugar, starch, lipid and protein were the four significant variables, in PC4.In the last PC-axis, moisture content, sugar, ash and protein were significant in their impact in determining the variance within the PC-axis, Sugar consistently featured in the five PC-axes.
The clustering of proximate compositions of the 15 accessions of Canavalia spp is presented in Fig. 1  The intra-cluster variability of the fifteen accessions within the five clusters are presented in Tab. 4. Moisture content was highest in cluster I but lowest in II.Sugar content was highest in II but lowest in IV and V. Least starch content was obtained in cluster I with a mean of 7.76.TCg1C (cluster V) had the highest starch content of 10.25.The least fat and energy obtained in this study was in cluster II while highest was observed in cluster VI.However, ash and protein content was highest in cluster IV.
Tab. 5 presented the Pearson correlations among the seven the proximate variables.The was no significant relationship among most of the seven proximate factors.Nonetheless, there was a significant relationship between fat and ash components of the 15 Canavalia accessions.

Discussion
All the 15 studied accessions belonging to two species of Canavalia differed significantly from one another for all the proximate variables.The range of values for most traits was narrow with the various accessions having superior performance for each variable.The 15 genotypes varied widely in fat content.This study agrees with the con-fication was based on similarities among the 15 accessions for the seven proximate variables.The 15 accessions were unique; differing from each other based on the proximate descriptors.No pairing of accessions occurred at a similarity distance of zero.Six main clustering patterns were possible as shown in Fig. 1.The cluster with the highest membership was I; clusters III and IV had three members each, clusters IV and VI had two accessions each while cluster V had a single accession.Only accessions within gladiata species featured in clusters III and V, while accessions within the ensiformis species featured only in cluster VI.In clusters I, II and IV accessions within the two species (ensiformis and gladiata) freely clustered together based on some proximate similarities (Fig. 1).
The similarity distances within which the fifteen accessions joined or tied are presented in Tab. 3. Based on the proximate variables considered in this study, TCg1B and TCg3B were the closest; they tied at the similarity distance of 0.01 (Tab.4).The next accessions with closest  (2001).This depicts a very high similarity between accessions within the same species.There seems to be a very high level of similarities between C. ensifromis and C. gladiata.The clusters generated using proximate data revealed accessions from the two species uniting in groups based on some similarities.TCe2B (C.ensiformis) and TCg4A (C.gladiata) were the closest accessions within the two species in the genus Canavalia.This is a type of inter-specific similarity within a genus.On the other hand some accessions among the 15 clustered in the same group.The aggregation of accessions within the same species into a cluster further revealed that there are high similarities within a biological and taxonomic unit.Intra and inter-specific hybridization may not be a big problem in any proposed breeding program for the improvement within each species or the genus.This is in line with observations of Nassir et al. (2008) and Vadivel and Janardhanan (2001).
The variability which existed within each clusters suggests the possibility of making selection for introgression of superior accessions with outstanding performance for various quantitative traits.Amongst the proximate variables, percentage fat content highly correlated with the energy constituent.Selection for any of these traits means a simultaneous selection of the other (Nassir et al., 2008).This study provides a template for further improvement through hybridization that could lead to production of elite hybrids.Similar view had been reported by Bandyopadhyay and Santra (2009).
Ash, moisture content, fat and energy were highest in C. ensiformis while starch and protein constituents were highest in the C. gladiata.The accessions from Canivalia species are good sources of quality nutrients for both humans and livestock according to Bamikole et al. (2000).Selection of accession with desirable proximate characters is highly recommended for legume breeding programme (Dada et al., 2012).

Fig. 1 .
Fig. 1.The dendrogram representation of 15 accessions of Canavalia spp using the proximate variables Tab. 1. Assessment of fifteen accessions of Canavalia spp.by seven quantitative proximate descriptors . The classi-Tab. 2. Eigenvectors and variance proportion of seven proximate variables in five PC-axes Significant at eigenvectors ≥ 0.20 Pugalenthi et al. (2004))ry of the fifteen accessions of Canavalia spp determined by the proximate variables Significant clusion ofDoss et al. (2011), that Canavalia species has compositional variability of proximate contents.The high coefficient of variability in lipid, sugar and energy is an indication of availability of opportunity for further selection of these traits for improvement of this species.This is inline with the submission ofNassir and Adewusi (2012),Asch et al. (2005)andPrice et al. (2002).Exploitation of the proximate constituents of Canavalia species with respect to crude protein, carbohydrates, and ash contents becomes necessary since the percentage composition of these compounds are comparable to other legumes whose nutritional demands are highly competitive between humans and livestock.This is in consonance with the view ofAdebowale et al. (2005),Pugalenthi et al. (2004),Siddhuraju and Bressani (2002)and Becker ***