Genetic Variability and Path Analysis of Chickpea (Cicer arientinum L.) Landraces and Lines
Identifying suitable parental materials is an important
phase in the development of hybrid seeds. Thus, a study was conducted
to determine the genetic variability among 360 chickpea land races and
lines. Appropriate parents selected from a pool of genes of 360 chickpeas
which was carried out during the year 2000-2001 in DARI, Sararoud of Iran.
The experiment included 12 blocks, each contains 20 plots of 2 m rows.
The traits studied were growth type, number of leaflet per leaf, leaflet
size, plant height, days taken for 50% flowering, flower color, flowering
period, days to maturity, pod size, pod per plant, seed numbers per pod,
seed color, seed shape and 100 seed weight. Data based on morphological
and phonological traits were analyzed using SPSS software and the statistical
procedures: correlation coefficient, cluster analysis, principal component
analysis and path analysis. Among the morphological characters, numbers
of branches, pod numbers with CV: 41.77 and 37.25% had higher variation,
respectively while leaflet with CV: 10.49% had minimum variation. Among
the phonological traits the flowering period with CV: 22.02% had highest
and flowering time had the least variability. The seed yield per plant
ranged from 4.27 to 0.41 g and CV: 51.43% reflected highest variation.
The highest correlation coefficient (r = 0.78) was between seed yield
per plant and pod numbers. Chickpeas genotypes could be classified into
four clusters and 63% of the variance were explained by five PCAs. Path
analysis revealed that the pod numbers with 0.745, seed numbers with 0.386,
100 seed weight with 0.268 and single seed with 0.267 had highest direct
effect on seed yield.
Plant genetic resources are the basis of global food security. They comprise
diversity of genetic material contained in traditional varieties, modern
cultivars, crop wild relatives and other wild species. To meet the need
for more food, it would be necessary to make better use of a broader range
of the world`s plant genetic diversity (Karaöz and Zencirci, 2005;
Farshadfar and Farshadfar, 2008).
Chickpea (Cicer arientinum L.) with 17-24% protein, 41-50.8% carbohydrates
and high percentage of other mineral nutrients and unsaturated linloeic
and oleic acid is one of the most important crops for human consumption.
Chickpea with low production cost, wide climate adaptation, use in crop
rotation and atmospheric nitrogen fixation ability is one of the most
important legume plants in sustainable agriculture system (Anonymous,
2002; Singh, 1997). Gene pool of chickpea with 39 species included 31
perennial and 8 annual land races, mutants, cultivars and wild types of
Cicer. Chickpea has high variation for different quality and quantity
traits, included ideal plant type (tall type), shape and color grain,
flower color, podding, color of seed coat, earliness, resistance to disease
and pests, which helps breeders to release improved and advanced lines
and varieties (Dasgupta et al., 1987; Singh, 1997).
First systematic attempt to establish, gene pool of chickpea was done
in 1977 by Indian IKWIST and 11195 lines were registered by 1978. In order
to avoid genetic erosion and releasing new varieties, cultivars and recombinant
lines the genetic variation for desirable traits should be broadened.
To study the genetic diversity among population, varieties and species,
the multivariate statistical procedures viz. cluster analysis, principal
components, factor analysis, discriminant function and path analysis have
been widely applied (Romesburg, 1984; Singh and Bejiga, 1991).
Traditionally, diversity is assessed by measuring variation in phenotypic
and morphological characters traits, which are of direct interest to users.
In the 1960s, biochemical methods based on seed protein and enzyme. Electrophoresis
were introduced, which proved particularly useful in analysis of genetic
diversity as they reveal differences between seed storage proteins or
enzymes encoded by different alleles at one (allozymes) or more gene loci
(isozymes). Molecular methods such as RAPDs, AFLPs, SSRs and microsatellites
used for detecting DNA sequence variation are being used as complementary
strategies to traditional approaches for assessment of genetic diversity
(Karp et al., 1997; Karp, 2002).
Earlier studies had established that evaluation of genetic variability
is one the important breeding objectives. Singh (1973) studied the variability
of 75 chickpea cultivars (DC and Kabuli type) by D2 statistic
and cluster analysis. Each DC and Kabuli type was grouped in separate
cluster. The 1000 grain weight, flowering time and pod size had highest
variation. Variability of chickpea has been studied using multivariate
statistical methods by different scientists (Bahl et al., 1976;
Singh et al., 1990). Grain yield of chickpea depends on many related
traits. Correlation coefficient could show linear relationship between
them, however, path analysis would elucidate direct and indirect relationship
among these traits, hence on the basis of that the breeder could select
the most effective traits to release varieties (Ulukan et al.,
2003; Yucel et al., 2006). Saleem et al. (2002) concluded
that pod number and 100 grain weight were the most important traits in
chickpea breeding programs. Yucel et al. (2006) with the path analysis
results showed that grain number and pod number were the most desirable
traits for chickpea improvement. Similar results were reported in many
studies for chickpea and Vicia (Yücel, 2004; Ulukan et
al., 2003; Ciftci et al., 2004; Güler et al.,
2001). Padi (2003) studied relationship between different characters of
chickpeas and found that harvest index and pod number had the greatest
direct effect on yield. Similar finding was reported by Toker (2004).
Noor et al. (2003) found that pod number and 100 grain weight had
most important traits to improve chickpea.
The purpose of the study is to find out the genetic variability which
is bared on different characters.
MATERIALS AND METHODS
To study genetic variability of 360 chickpea line and land races and
lines, the experiment was carried out at Dry Land Agriculture Research
Institute (DARI), Sararoud, Kermanshah, Iran. The experiment included
12 blocks, each contains 20 plots of 2 m rows. The traits studied were
growth type, number of leaflet per leaf, leaflet size, plant height, days
taken for 50% flowering, flower color, flowering period, days to maturity,
pod size, pod per plant, seed numbers per pod, seed color, seed shape
and 100 seed weight. The Bivanij (famous local chickpea) was the control.
Data based on morphological and phonological traits were analyzed using
SPSS software and the statistical procedures: correlation coefficient,
cluster analysis (UPGMA), principal component analysis and path analysis.
The statistical parameters viz. means, standard deviation, minimum, maximum,
Coefficient of variation (CV), skewness and kurtosis for different traits were
shown in Table 1. Among the morphological characters, stem
numbers, pod numbers with CV = 41.77 and 37.25%, respectively; had higher variation
while the leaflet with CV = 10.49% had minimum variation. Among the phonological
traits the flowering period with CV = 22.02% had the highest and flowering time
had the least variability. The seed yield per plant ranged from 4.27 to 0.41
g and CV = 51.43% revealed the highest variation. The height character also
showed high variation; consequently it can be rich gene pool for breeder.
To find out relationships between different traits and yield, correlation
coefficient between different quantitative and qualitative characters
were shown in Table 2 and 3, respectively.
|| The statistical parameters of chickpea genotypes
highest correlation coefficient r = 0.78 was between seed yield per plant
and pod numbers. Single seed weight, canopy height, seed numbers, pod
size, stem numbers and leaflet size were relatively higher than others,
respectively. Significant correlation was observed between plant color
and flower color, seed color and seed shape with r = 0.48% and 0.35%,
respectively. Spearman rank correlation among quality traits is shown
in Table 3. There was negative correlation between seed
shape and plant color, it means if the stem and leaf contain antocyanine,
the seed will be smaller and angled (DC type).
Path and regression analysis: The results of stepwise multiple
regression analysis is shown in Table 4. Grain yield
per plant was the dependent trait and other traits were independent variables.
Pod numbers was entered in the model at first and explained 62% of variation,
followed by 100 grain weight and grain numbers entered in the model, respectively.
In order to study the direct and indirect effect of traits entered into
the step wise regression analyses on the yield, path analysis was carried
out (Fig. 1). According to Table 4,
the pod numbers, 100 grain weight, canopy and seed numbers per pod had
high correlation coefficient values. Path analysis revealed that the pod
numbers with 0.745, seed numbers with 0.386, 100 seed weight with 0.268
and single seed with 0.267 had highest direct effect on seed yield, respectively.
Indirect effect via other characters displayed different values.
Cluster analysis: Cluster analysis using UPGMA method was carried
out in two steps. First, for all genotypes regardless of geographical
regions and second, for quality and quantity traits separately.
All genotypes were classified into four clusters. Cluster 1 contains
338 genotypes, cluster 2 and 3 had one and cluster 4 includes eight genotypes.
Genotypes 41-3550 in cluster No. 2 due to highest plant height, the highest
pod size, grain numbers per pod with higher average greater than grand
mean of genotypes were located in separate cluster.
|| Stepwise regression analysis on different characters
|*, ** are significant on 5 and 1% probability level,
|| Five principal components on quantity traits
|| Principal component analysis on different traits
with high 100 grain weight and good performance of other yield components
was in separate cluster as well.
Members of cluster 4 had desirable situation for pod per plant, leaflet
numbers, days to 50% flowering and maturity. The results of principal
components analysis for 360 chickpea genotypes are shown in Table
5. About 63% of variance was demonstrated by five PCAs. The 1st PCA
covered 22% of variance, 2nd PCA; 15%, 3rd PCA; 10%, 4th PCA; 8% and 5th
PCA with 7% of variance. Contribution of each character in PCAs is given
in Table 6. According to these results the seed yield
per plant had highest value and 100 grain weight was high, respectively.
There was positive significant correlation among these traits with grain
yield per plant (Table 2). Stepwise regression analysis
was significant and some of the traits were entered into the model for
grain yield per plant. Therefore in breeding program grain yield could
be focused on. The first PCA is named yield component PCA (with 15.42%
variance) and the second one is called phonological PCA (15.42% of variance),
the third would be called as morphometrical PCA (10/18% of variance) and
fourth PCA is generation phase and fifth PCA is called canopy PCA.
Determination of genetic variation is the main step in breeding programs.
Different statistical parameters of accessions were estimated and according
to the results applied breeding programs would designed. Such procedures
have been done by many authors and similar results have been reported.
Singh and Bejiga (1991) studied the 38 pea cultivars and concluded that
ripening date with CV = 3% was minimum and flowering period with CV =
26% had largest coefficient of variation. These values in this experiment
for same traits ranged between 4.57 and 22.02%. They showed that Phenotypic
Coefficient of Variation (PCV) for all traits were higher than Genotypic
Coefficient of Variation (GCV) indicating the highly environmental effects.
Therefore breeders should be careful to select the best parent based on
Correlation coefficient and regression analysis help the breeders to
select an efficient trait, that`s why this parameter was estimated. Sandhu
and Singh (1972) reported high correlation between seed yield and pod
per plant, seed per plant; seed weight per plant and stem number per plant.
Also the positive significant correlation among yield and pod number per
plant, branch number, plant height were reported by Singh et al.
(1990). Joshi (1972) reported that if the seed will be circle and angle
less, the leaf color will be whiter. Path analysis identified the character
having direct and indirect effects on the yield. Breeders would try to
find out relationships between different characteristics obtaining new
varieties and lines. Pod numbers had highest direct effect on yield in
this experiment. Therefore, promising lines could be selected on the basis
this trait. Phadnis et al. (1970) studied 45 chickpea lines and
revealed that seed weight had largest direct effect on seed yield. Bahl
et al. (1976) also studied yield components by path analysis in
21 chickpeas and similar results were reported. Paliwal (1987), Chaudhary
et al. (1988) and Dasgupta et al. (1992) reported the higher
effect of pod numbers and seed numbers on economical yield of chick pea.
These results help breeders to select new parents and lines based on pod
numbers, seed numbers and 100 seed weight.
Multivariate statistical procedures such as cluster analysis, principal
component analysis etc. were used to determine genetic variability based
on different characters. These methods have done by many scientists and
a lot of results have been reported. Singh (1973) analyzed 75 lines of
DC and Kabuli peas by D2 statistics and classified into 9 clusters
and each DC and Kabuli type were located in separate cluster. The 100
grain weight and flowering time had the highest effect on variation.Ansawa
(1981) reported such results as well. Govil et al. (1980) classified
Iranian and Indian pea samples into 13 clusters. In the most studies pod
numbers per plant had the most variation (Singh, 1997; Toker, 2004; Noor
et al., 2003).
Anonymous, 2002. FAO Web Site. WWW. FAO. Org.
Ansawa, B.M., 1981. Factor analysis in chickpea. J. Agric. Sci., 5: 156-159.
Bahl, P.N. R.B. Mehra and D.B. Rajue, 1976. Path analysis and its implications for chickpea breeding. Plant Breed., 77: 67-71.
Direct Link |
Chaudhary, B.D., P. Singh, A. Kumar and R. Singh, 1988. Correlation and multiple Regression analysis in Chickpea. Crop Improve., 15: 180-186.
Ciftci, V., N. Toay, Y. Toay and Y. Doan, 2004. Determining relationships among yield and some yield component using path coefficient analysis in chickpea (Cicer arientinum L.). Asian J. Plant Sci., 3: 632-635.
CrossRef | Direct Link |
Dasgupta, T., M.O. Islam and P. Gayenm, 1992. Genetic variability and analysis of yield components in chickpea. Annu. Agric. Res., 132: 157-160.
Dasgupta, T., M.O. Islam, P. Gayen and K.K. Sarkak, 1987. Genetic divergence in Chickpea (Cicer arientinum L.). Exp. Genet., 3: 15-21.
Farshadfar, M. and E. Farshadfar, 2008. Genetic variability among lucerne cultivatrs based on biochemical (SDS-PAGE) and morphological markers. J. Applied Sci., 8: 1867-1874.
CrossRef | Direct Link |
Govil, J.N., B.S. Rana and B.K. Murty, 1980. Factor diversity and yield component in chickpea Egypt. J. Genet. Cytol., 90: 20-25.
Guler, M., M.S. Adak and H. Ulukan, 2001. Determining relationships among yield and some yield components using path coefficient analysis in chickpea (Cicer arietinum L.). Eur. J. Agron., 14: 161-166.
CrossRef | Direct Link |
Joshi, S.N., 1972. Variability and association of some yield components in chickpea. Indian J. Agric. Sci., 49: 35-38.
Karagoz, A. and N. Zencirci, 2005. Variation in wheat (Triticum spp.) landraces from different altitudes of three regions of Turkey. Genet. Resour. Crop Evol., 52: 775-785.
CrossRef | Direct Link |
Karp, A., 2002. The New Genetic Era: Will it Help Us in Managing Genetic Diversity?. In: Managing Plant Genetic Diversity, Engls, J.M.M., R.V. Ramanatha, A.H.D. Brown and M.T. Jackson (Eds.). CAB International and IPGRI, Wallingford and Rome, pp: 43-56.
Karp, A., S. Kresovich, K.V. Bhat, W.G. Ayad and T. Hodgkin, 1997. Molecular tools in plant resources conservation. A Guide to the Technologies (IPGRI) Technical Bulletin No. 2). International Plant Genetic Resources Institute, Rome.
Noor, F., M. Ashaf and A. Ghafoor, 2003. Path analysis and relationships among Quantitative traits in chickpea (Cicer arientinum L.). Pak. J. Biol. Sci., 6: 551-555.
Direct Link |
Padi, F.K., 2003. Correlation and path coefficient analysis of yield and yield components in pigeonpea. Pak. J. Biol. Sci., 6: 1689-1694.
Direct Link |
Paliwal, K.K., 1987. Correlation and path analysis in Chickpea (Cicer arientinum L.) legume research. Indian J. Agric. Sci., 10: 47-48.
Phadnis, B.A., A.P. Ebote and S.S. Anichwar, 1970. Path coefficient analysis in gram. Indian J. Agric. Sci., 40: 1013-1016.
Romesburg, H.C., 1984. Cluster Analysis for Researchers. 1st Edn., Life Time Learning Publications, Belmount, California, pp: 168-195.
Saleem, M., M.H.N. Tahir, R. Kabir, M. Javid and K. Shahzad, 2002. Interrelationship and path analysis of yield attributes in chickpea (Cicer arientinum L.). Int. J. Agric. Biol., 3: 404-406.
Direct Link |
Sandhu, T.S. and N.B. Singh, 1972. Correlation, Path coefficient analysis and discriminant function selection in gram. J. Res. Punjab. Agric. Uni. Ludhiana, 9: 417-421.
Singh, K.B. and G. Bejiga, 1991. Evaluation of world collection of Chickpea (Cicer arientinum L.). Field Crop Res., 28: 12-15.
Singh, K.B., 1997. Chickpea (Cicer Arientinum L.). Field Crop Res., 53: 161-170.
Singh, K.B., G. Bejiga and R.S. Malhotra, 1990. Association of some characters with seed yield in chickpea collection. Euphytica, 49: 83-88.
Direct Link |
Singh, K.P., 1973. Genetic divergence and quantitative inheritance studies in gram. Indian. J. Agric. Sci., 6: 392-396.
Toker, C., 2004. Evaluation of yield criteria with phenotypic correlations and factor analysis in chickpea. Plant Soil Sci., 54: 45-48.
CrossRef | Direct Link |
Ulukan, H., M. Guler and S. Keskin, 2003. A path coefficient analysis some yield and yield components in faba bean (Vicia faba L.) genotypes. Pak. J. Biol. Sci., 6: 1951-1955.
CrossRef | Direct Link |
Yocel, D.O., A.E. Anlarsal and C. Yocel, 2006. Genetic variability, correlation and path analysis of yield and yield components in chickpea (Cicer arietinum L.). Turk. J. Agric. For., 30: 183-188.
Direct Link |
Yucel, C., 2004. Correlation and path coefficient analysis of seed yield components in the narbob bean (Vicia narbonensis L.). Turk. J. Agric. For., 28: 371-376.