STUDIES ON GENETIC DIVERSITY IN RICE (Oryza sativa L.)
yield (68.72). Number of filled grains per panicle, 1000-grain weight, grain length and breadth exhibited
less environmental effect and high heritability coupled with moderate to high genetic advance.
Zen and Bahar (2001) an experiment was carried out to study on genetic variability of plant
characters and yield of eleven promising lines of highland rice and Batang Sumani at nine location 800-
900 m. a.s.i., the results showed that heritability of all plant characters and yield were high (76.98 percent
– 97.96 percent) heritability of filled spikelets per panicle was medium (37.29 percent). Plant height,
productive tillers, spikelets per ear, and yield had a wide genetic variability. Selection for those characters
could be done at early generation.
Mohammad etal.(2002) reported that phenotypic and genotypic coefficient of variation (PCV and
GCV) were of comparable magnitudes except for grain yield./plant and grains per panicle where the
environmental coefficient of variation (ECV) contributed more to the PCV than GCV and also observed
high heritability along with high relative expected genetic advance for 1000 grain weight, primary
branches per panicle and productive tillers in an experiment comprising ten rice varieties or pure lines.
Nayak et al. (2002) studied genetic variability for grain yield and nine yield contributing
characters in 200 scented rice genotypes in Cuttack, Orissa. They obtained high estimates of genetic and
phenotypic coefficients of variation for number of spikelets per panicle, number of panicles per plant,
number of grains per panicle and grain yield per plant. Further they also observed high heritability and
high genetic advance for number of spikelets per panicle, number of grains per panicle and grain yield per
Singh etal.(2002) evaluated 52 genotypes of (low land) rice for 15 characters and reported high
genotypic and phenotypic variances for grain yield per plant, panicle weight, number of grains per
panicle and number of branches per panicle, medium for panicle length, 1000-seed weight and low for
panicle length and milling percent and found that heritability in broad sense ranged from 3.61 for
number of effective tillers per plant to 99.55 for grain length. High heritability with high genetic advance
was recorded for number of grains per panicle followed by panicle weight and grain yield per plant in an
experiment conducted at Meghalaya.
Yadav et al. (2002) genetic variability heritability and expected genetic advance which were
estimated for length of grain, breadth of grain, L/B ratio, kernel elongation, gel consistency, amylase
content and gelatinization in rice. The experimental material included 150 rice germplasm categorized
into three groups. i.e. long slender, medium slender and short slender. Maximum variability was recorded
for test weight in long slender and medium slender groups and for amylase content in medium and short
slender groups. Water updake, volume expansion ratio, kernel elongation and gel consistency appeared to
be the useful traits in all the groups because high heritability and high genetic gain were recorded for
Chaudhary and Motiramant (2003) fifty four traditional aromatic rice accessions were evaluated
for 19 descriptors to obtain information on genetic variability and character association of grain quality
and yield attributes. A wide range of variation as recorded for most of the characters. Heritability in broad
sense was very high for all the characters exhibited high heritability coupled with high genetic advance
except harvest index. Grain yield per plant showed significant positive correlation with effective tillers per
plant, spikelet density and biological yield per plant, path analysis indicated a greater contribution of
effective tillers per plant, spikelet density and biological yield per plant towards grain yield.
al.(2003) genetic variability, heritability and genetic advance were studied in 200 scented
rice genotypes including one non-scented check, Ratna for grain yield and its nine attributing characters.
High GCV and PCV values spikelets/panicle, number of grains/ panicle and grain yield/plant. High
heritability along with high genetic advance were observed for number of spikelets/panicle, number of
grains/panicle, grain yield/plant followed by other characters. Emphasis should be given on these
characters while selecting scented rice varieties to improve grain yield.
Singhara et al. (2003) genetic variability and association of different panicle components with
number of grains panicle were determined in 36 genotypes of rice grown under Kashmir conditions wide
range of variability was observed for all characters studied. The PCV and GCV values were larger for
number of secondary branches, panicle and grains borne primary branch and low for days to 50%
flowering, panicle length, kernel length and kernel breadth. High heritability coupled with high genetic
advance were recorded for 1000-seed weight, seeds panicle and primary branch length indicating greater
scope for yield improvement through selection.
Sinha et al. (2004) variability, heritability and genetic advance estimates and their correlation
coefficient were determined in Ambikapur, Chattisgarh, during the 1998-2000 kharif seasons from 19
local midland rice landraces, with IR-36 as the standard control, for yield and its attributing characters
(plant height, tillers per hill, panicles per hill, panicle length, days to 50% flowering, days to maturity, test
weight and grain yield). A high genotypic coefficient of variation was observed in grain yield followed by
test weight and panicles per plant. High heritability with high genetic advance was found for grain yield
followed by test weight and panicles per plant. Significant positive genetic correlation were observed for
panicles per plant, test weight and grain yield.
Vivek et al. (2004) estimated genetic variability for 12 characters in 39 “new plant” tropical
Japonica lines from International Rice Research Institute, Philippines along with three control cultivars.
The genotypes showed a wide range of variation for all the characters. Grain yield per plant, biological
yield per plant, number of tillers and panicles per plant had high values of genetic coefficient of variation
and phenotypic coefficient of variation. They reported high heritability with high genetic advance for
grain yield per plant, biological yield per plant, panicle in an experiment.
Elayaraja etal.(2005) observed high heritability associated with moderate to high genetic advance
as a percent of mean for number of productive tillers, panicle length, number of grains per panicle, 100grain
weight and grain yield per plant in M2 generation.
Panwar (2005) observed high heritability and genetic advance for grain yield per panicle, chaffy
grains per panicle, grain yield per plant, filled grains per panicle and secondary branch number per
panicle, indicating the effectiveness of selection for these characters.
The nature and magnitude of genetic variability was assessed by Singh et al. (2005) for seven
characters in 20 rice genotypes including a local check during kharif 2004. The analysis of variance
revealed significant differences among the genotypes for all the characters. High estimates of genotypic
coefficient of variation (GCV) and phenotypic coefficient of variation (PCV) were recorded for grain yield
(t/ha) and biological yield (t/ha), particle density/sq.m. and harvest index (%).
Vivek etal. (2005) observed 39 “new plant” tropical Japonica lines during the wet season of 2000
in Pantnagar. The genotypes showed a wide range of all characters. High genotypic and phenotypic
coefficient of variation was observed for grain yield, followed by harvest index and biological yield. They
observed high heritability coupled with high genetic advance which were also observed for grain yield,
followed by harvest index and biological yield.
Amudha et al.(2006) reported that high genetic variability for the number of days to flowering,
plant height, number of productive tillers per plant, panicle length, spikelet fertility, number of grains per
panicle, 100-grain weight, grain yield per plant, dry matter production and harvest index in biparental
progenies of the cross MP x Norungans, for all the characters examined except 100-grain weight, panicle
length and root volume in the biparental progenies of cross PM x Mattaikar, and for all the characters
examined except 100-grain weight, root volume, number of grain per panicle and grain yield in the
biparental progenies of the cross PM x Poonagar.
Singh etal.(2006) genetic variability and correlation studies in rice were conducted during kharif
2002 in Khudwani, Anantnag district, Jammu and Kashmir. Thirty-two genotypes were evaluated for days
to 50% heading, days to maturity, plant height, panicles per plant, biological yield, grain yield and harvest
index. A wide range of variation was recorded for all traits. The highest genotypic and phenotypic
coefficients of variation were recorded for grain yield. High heritability and high genetic advance were
recorded for plant height, indicating the predominance of additive gene action for this trait. Genotypic and
phenotypic correlation studies indicated that biological yield per plant and harvest index were
significantly and positively correlated with yield. Thus, selection for these two traits might be helpful in
enhancing rice grain yield.
Kumar et al. (2007) genotypic and phenotypic coefficient of variation, heritability and genetic
advance as percent of mean were estimated in the F2 and F3 segregating populations of six crosses of rice
for six yield and yield component characters. The F2 populations of the cross P1 P3 showed high PCV, GCV
coupled with high heritability estimates and high genetic advance as percentage of mean for number of
filled grains per panicle, 100-grain weight, biomass per plant and grain yield per plant. Similarly, the F3
population of the cross P2 P1 exhibited high genetic parameters for number of productive tillers per plant
and grain yield per plant. These populations could be subjected to simple pure line selection to improve
grain yield per plant.
Ishwar etal.(2007) the estimates of variability and genetic divergence were carried out in rice for
17 characters at RRS, Kaul during kharif 2002-03. Substantial amount of genotypic variability was
observed for all the traits under study.
2.2 Genetic Divergence:
The expression “divergence in character” was used by Darwin (1859) for the variation in genera
and species. Another term “morphism” was used by Huxley (1955) for genetic diversity implying “genetic
polymorphism” which means the coexistence of distant genetic forms in population.
Early workers regarded the geographical isolation as a reasonable index of genetic diversity
(Vavilov, 1926; Joshi and Dhawan, 1966). The varieties, which come from different localities are usually
presumed to diverse and are utilized in hybridization programme. However, several workers in different
crop species have emphasized that there is no parallelism in geographical distribution and genetic
diversity (Murthy and Anand, 1966 in linseed; Bhatt, 1970 in greegram; Maurya and Singh, 1977 and De
et al., 1992 in rice) advocating that varieties with the same geographical origin could have undergone
changes under selection pressure. Thus, for estimation of variation within the germplasm divergence study
in the form of classification into different homogenous groups is an important practice. Multivariate
analysis based on Mahalanobis-D2 statistics and canonical variant analysis has been considered as an
important tool in quantifying the genetic divergence in different crops (Rao, 1952).
A number of scientists (Griffing and Lindstrom, 1954; Mall etal., 1962; Arunachalam, 1981) have
emphasized the importance of genetic diversity in plant breeding for obtaining broad spectrum of
desirable variability in segregating generations. Some of the earlier reports on genetic diversity in rice
have been reviewed below:
Pradhan and Roy (1990) multivariate analysis of data on 13 yield components in 25 breeding lines
grown in shallow and intermediate depth water showed 6 and 7 clusters, respectively. The composition of
the clusters differed under the 2 regimes due to pronounced genotype – environment interactions.
Roy and Ponwar (1993) genetic divergence for grain yield and 9 yield-related traits were studied
in 99 diverse genotypes. Total genotypes were grouped into 16 clusters. Genetic divergence was controlled
mainly by panicles/plant, grains/panicle, grain yield/plant, spikelets/panicle. The results indicated a
significant divergence for the traits measured.
Sharma and Hore (1993) seventeen diverse genotypes of upland rice raised during 1989 and 1990
were grouped into 5 clusters on the basis of D2 analysis of data on 11 characters. Inter - and – Intracluster
distances are given and their use in the selection of parents for breeding programmes is mentioned.
Sharma and Richaria (1995) estimated genetic diversity among thirty-nine rice genotypes and
grouped them into 8 clusters. The distribution of highest and lowest mean values in distant clusters,
indicated the importance of traits contributing to the divergence (days to 50% flowering, days to maturity,
secondary branches per panicle and spikelets per panicle).
Sawant etal.(1996) grouped 75 genotypes of rice grown in kharif, 1991, on the basis of data on 8
yield components into 10 clusters. Clustering pattern revealed that geographic diversity is not a reasonable
index of genetic diversity.
Singh et al. (1996) assessed the nature and magnitude of genetic divergence in 40 genotypes of
scented and fine using Mahalanobis D2-statistic for 10 characters. The genotypes were grouped into six
clusters. Grain yield contributed 40.6% and plant height contributed 16.5% to genetic divergence.
Kumari and Rangasamy (1997) estimated genetic divergence using Mahalanobis’s D2 statistics in
62 early rice genotypes obtained from sixteen countries. Based on eight important yield contributing
characters, these genotypes were grouped into six clusters. They found that there was no relationship
between geographical distribution and genetic diversity. Characters like grain yield per plant, panicle
exertion and plant height made largest contribution to total divergence.
Mishra and Dash (1997) an experiment was conducted during the rainy season of 1991-1994 to
study genetic diversity in 10 genotypes of aromatic rice (Oryza sativa L.). Pooled data on 9 quantitative
characters, viz. days to 50% flowering, plant height, tillers/hill, panicles/hill, panicle length, grain/panicle,
chaffs/panicle, 100-grain weight and grain yield, were analysed for 4 environments, and 4 clusters of
genotypes were formed on the basis of D2 statistics.
Hanamaratti etal.(1998) grouped 50 rice genotypes for 10 yield components in low and upland
environments using cluster analysis into 18 and 17 clusters under low and upland conditions,
respectively. Genotypes were found independent of their geographic origin.
(1998) twenty-five genotypes of rice (Oryza sativa L.) were grouped into five clusters
on the basis of yield components data. The maximum inter cluster divergence was observed between
cluster B and components (63.04); followed by components and D (51.90) and cluster B and E (48.30)
indicating that these groups of genotypes were highly divergent from each other. The genotypes in above
clusters revealed substantial differences in the means for important yield contributing characters,
suggesting that the genotypes belonging to these clusters forms ideal pairs for planning a hybridization
Ahmad and Borah (1999) estimated genetic divergence among 85 indigenous glutinous rice varieties
from Assam using D2 analysis. The genotypes were grouped into 12 clusters based on 13 agronomic
characters. Tillers number, panicles per hill, grains per panicle, grain fertility and grain yield accounted
for the major portion of divergence.
Bansal et al. (1999) assessed genetic diversity in 34 rice stocks collected from seven countries
which were grouped into 15 clusters using D2 analysis for 10 economic traits. The pattern of distribution
of genotypes within various clusters was independent of geographical distribution.
Kandhola and panwar (1999) genetic diversity among 52 indigenous and exotic genotypes of rice
was studied using Mahalanobis D2 statistics in kharif 1996 under 2 sowing dates and 2 nitrogen fertilizer
levels. Based on 16 agromorphological and quality characters, these genotypes were grouped into 11
clusters. Cluster 1 with 26 genotypes was largest, while clusters VII, VIII, IX, X and XI were
monogenotypic. There was no association between genetic and geographic diversity. The maximum
intercluster distance was observed between genotypes of clusters V and IX (18984.4). It was concluded
that hybridization among genotypes drawn from widely divergent clusters with high yield potential were
likely to produce heterotic combinations and wide variability in segregating generation.
Pandey etal.(1999) 50 genotypes were grown in iron-toxic soil at Barapani in Meghalaya during
rainy seasons of 1992 and 1993 and evaluated for 12 yield-related and morphological traits. On the basis
of D2 analysis of the data collected, the 50 genotypes were grouped into 6 clusters. The characters
contributing most total divergence were days to 50% flowering, plant height, primary branches per
panicle and 100-seed weight (27.9, 24.7, 16.4 and 10.4, respectively). Genetic diversity was not correlated
with geographical diversity.
Singh et al. (1999) studied genetic divergence in 42 genotypes of boro rice. Multivariate analysis
revealed considerable genetic diversity in the material and led to their grouping in four clusters. Harvest
index, total number of grains per panicle, number of fertile grains per panicle and stability accounted
90.6% of the total divergence.
Hegde and Patil (2000) assessed genetic divergence in 40 genotypes of rainfed rice using D2
statistics. The cultivars fell into 7 clusters. The highest contributing characters to D2 values were spikelet
number per panicle, photosynthetic rate, unit and 1000-grain weight.
Rather etal.(2001) studied genetic divergence in 56 rice cultivars for 12 characters and grouped
them into six clusters. The grouping of cultivars from various regions into the same cluster indicated that
the geographical distribution did not necessarily suggest genetic divergence.
Shanthi, P. and Singh, J. (2001) studied genetic divergence for yield and its components in 17
induced mutants (induced by gamma rays, EMS and their combinations) including one non-mutant
mother variety of Mahsuri rice, for six qualitative characters (plant height, number of tillers/plant, panicle
length, number of grains/panicle, 1000-seed weight, and grain yield/plant).The genotypes differed
significantly for six characters considered collectively and were grouped into four clusters.
Arun etal.(2002) assessed genetic diversity for 28 yield and morphological traits in 100 aromatic
rice genotypes at Hyderabad, Andhra Pradesh. The observed that the pattern of distribution of genotypes
within various clusters was random and independent of geographical isolation.
Bhave et al. (2002) studied association analysis of seed and seedling characters with adult plant
characters, genetic divergence with hybrid performance and genetic distance with hybrid performance
helps for prediction of hybrid performance at early stages of crop growth and can become a very good tool
in the hands of breeders. Genetic divergence showed significant positive association with mean
performance of hybrid for tillers/plant and productive tillers/plant while genetic distance showed
significant negative association with plant height, panicle length and grain/panicle.
Chaudhary and Sarawgi.(2002) assessed genetic diversity in 54 genotypes for 19 morphological
and quality traits. The analysis of variance indicated that the genotypes differed for almost all the traits
under study. The genotypes were grouped into five clusters.
Reddy etal.(2002) applied D
2 statistics to group 36 genotypes of low land rice grown in Cuttack,.
Significant varietal differences were observed for all 13 characters studied. The genotypes were grouped
into 12 clusters. Among the different characters, 1000-grain weight, grain length, number of grains per
panicle and plant height played a major role in the formation of clusters. The diversity was not related to
Babu etal.(2003) worked out genetic diversity among 33 rice cultivar planted in Madurai, Tamil
Nadu and grouped them into 10 clusters using Mahalanobis D2 statistics based on genetic distance. Among
the eight different characters, the trait days to 50% flowering contributed maximum to words genetic
diversity (51.89%), followed by plant height (22.92%) and panicle length (9.47%).
Chauhan, J.S. and Singh, K.H. (2003) studied genetic divergence in forty-five elite rainfed upland
rice cultivars. The D2 values among the genotypes ranged from 1.69 (between CR 544-1-1 and CR 544-
1-6) to 257.8 (between CR 636-7 and Sattari). Based on divergence, 45 genotypes were grouped into
eleven clusters. The average intra-cluster distance this indicates the divergence among the genotypes of
the same cluster.
Manna, M; Hossain,A.M. and Sasmal, B.G.(2003) worked out genetic divergence for yield and yield
components .The clustering pattern based on genetic diversity did not correlate with the grouping of
cultivars based on growing conditions. The cultivars categorized into clusters II and IV, which exhibited
the greatest inter-cluster distance and superiority for most of the traits, may be used for the development
of superior genotypes..
Mishra etal.(2003) studied the nature and magnitude of the genetic diversity for 20 quantitative
and qualitative characters were determined for 16 rice cultivars and their 72 F1 hybrids. The genotypes
were grouped in 12 clusters based on the relative magnitude of multivariate D2 values. Based on the
cluster means, plant height, flag leaf width, ear bearing tillers per plant, 100 seed weight, panicle length,
biological yield, harvest index. Analysis of variance indicated highly significant differences for the most of
the characters studied.
Shiv, Datt etal.(2003) studied the degree and nature of genetic divergence among a set of 61 Elite
Basmati rice genotypes collected from different parts of India and abroad and grouped them into 4
clusters. Plant height contributed maximum to the genetic divergence (52.24%), followed by days to 50%
flowering (22.56%) and grain yield per plant (8.63%).
Cheema etal.(2004) the genetic variation among 17 mutants and their respective parents (from
Indonesia, Malaysia, Korea, India, Vietnam, Thailand, Philippines, Bangladesh, Pakistan, China and IRRI)
was evaluated in Faisalabad, Pakistant, during 2002-2003. Deviations observed by metroglyph method
regarding the number of clusters formed, number of genotypes in the cluster, and super imposition of the
genotypes within the cluster indicated the possibility of genetic improvement for yield and yield
components. Metroglyph scatter diagram classified the genotypes into 11 groups. Based on this grouping,
hybridization between group-Individuals and group-II is expected to yield superior rice cultivars.
Das etal.(2004) studied genetic divergence in fifty land race collections of rice. The genotypes
were grouped in 10 clusters. Days to 50% flowering, grain yield per plant, grain length, kernel breadth
and 100-kernel weight were identified as potential characters that can be used as parameters while
selecting diverse parents in the hybridization programme for yield and quality improvement.
Nayak etal.(2004) studied nature and magnitude of genetic divergence among 200 genotypes of
scented rice including one non-scented check using Mahalanobis D
2 statistics for 10 quantitative
characters. On the basis of D2 values, the genotypes were grouped into 10 clusters. Grain length and days
to 50% flowering played important role in the formation of clusters. Among the different characters,
panicle length contributed minimum (0.7%) to total divergence.
Nandini etal.(2004) 60 rice cultivars were evaluated for their organoleptic qualities. Divergence
of sample ws measured by Mahalanobis D2 statistics and clustering done by Tocher’s method. For rice, the
cultivars formed 6 clusters, while for the parboiled samples, 10 clusters could be recognized. Results of the
D2 analysis revealed that among the 60 rice cultivars, as much as 35 were homogenous with respect to
quality attributes such as appearance, colour, flavour, texture and taste for the preparation of boiled rice
either in the row of parboiled forms.
Roy etal. ( 2004) thirty-five aman rice cultivars were evaluated for 10 traits (number of panicles
per plant, panicle length, number of primary branches per panicle, number of secondary branches per
panicle, number of filled grains per panicle, number of unfilled grains per panicle, 1000-grain weight,
panicle weight, grain yield per plant and sterility) over 2 environments. The cultivars were grouped into 5
clusters. The greatest genetic divergence was observed between clusters II and IV. As clusters II (Nagra,
Khayersali, CRM-30 and Langulmutha) and IV (Randhunipagal) showed the greatest divergence and
higher mean values for characters contributing to genetic divergence, the cultivars from both clusters may
be used in hybridization programmes to obtain good recombinants.
Xie Rong et al. (2004) assessed Principal component analysis as well as cluster analysis was
performed on 53 somatic lines from regenerated plants of early maturing restorer rice line 402. The 53
somatic lines might be divided into 6 groups. There were larger variation for shelled seed length, shelled
seed length/width, brown rice rate, milled rice rate, head rice rate, number of spikelets per panicle, grain
weight per plant, effective panicles per plant and plant height among 21 quantitative characters studied.
There were significant differences among the quantitative characters in various groups of somatic lines.
The first and sixth group of somatic lines, which had better comprehensive characters, may be employed
as key crossing lines in breeding programme.
Awasthi etal.(2005) a field experiment were conducted to determine the genetic divergence of 21
Indian aromatic rice genotypes. The genotypes were grouped into 6 clusters for different characters. The
inter-cluster distance was observed to be highest between clusters II and III, indicating that the genotypes
of these 2 clusters were genetically more diverse. The number of grains per panicle, grain yield per plant,
days to 50% flowering, leaf length and leaf width showed high percent contribution towards total genetic
Bhutia,K . (2005) assessment of genetic divergence using Mahalanobis D2 statistics was carried out
on 41 high yielding and local genotypes of rice (Oryza sativa). The genotypes were grouped into six
clusters. Cluster IV showed the maximum genetic distance from cluster VI followed by its distance from
cluster V. The desirable yield and quality characteristics were distributed mainly in clusters III and IV and
cluster V. The genotypes included in clusters III and IV may be used as parents in hybridization
programme to improve yield.
Bose and Pradhan (2005) assessed the nature and magnitude of genetic divergence among 35 deep
water rice genotypes from India using Mahalanobis D2 statistics. The genotypes wre grouped into 10
clusters showing fair degree of relationship between geographic distribution and genetic divergence.
Traits such as plant yield, days to 50% flowering, and plant height were the major contributors to genetic
Chand et al. (2005) 19 genotypes of Aman rice were studied for their genetic divergence using
Mahalanobis D2 statistics for 12 characters. Based on D2 values, the genotypes were grouped into six
clusters. Cluster Individuals was the largest with eight genotypes followed by cluster II with four
genotypes. Intra-cluster IV. The maximum inter-cluster distance was found diversity between these
groups. The major part of total divergence was imported by single trait i.e. 1000-grain weight and panicle
length, grain length and plant height were also very important in this regard.