Mahajan, Manoj K Srivastava, Vineet Kumar, N V P R Ganga Rao, S Nagarajan and R
Directorate of Wheat Research, P.O. Box-158, Karnal-132001,
Field Crop Research
Author: Dr Vinay Mahajan, email: firstname.lastname@example.org
hundred and fifty two genotypes of national and international origin used in
hybrid wheat program since 1995 were analysed for genetic diversity and found
clustered in six groups. Cluster IV was superior for flag leaf area while
cluster V had members with more spikelets\spike and protein content.
High yield and high hectolitre weight were observed in cluster VI but
were tall. Hybrids coming out of cross combinations obtained between clusters V
and VI may be more useful and promising.
Key words: Diversity, wheat, hybrid, genetic diversity, clusters and D2
Genetic divergence is dependent on geographical diversity as well as phenotypic components of varieties, and its quantitative assessment could provide a rational basis for the selection of parents for any breeding programme. The hybrid wheat program initiated in 1995 laid an emphasis on diversity of genotypes used to identify harnessable commercial heterosis. Since then a number of genotypes were selected from various national and international sources. These were evaluated for various morphological characters, which can be useful in hybrid wheat programme for hybrid seed production and standard heterosis. Non- hierarchical Euclidean cluster analysis helps in separating out genotypes in clusters. This assists in identifying probable male and female lines\groups that will result in to standard heterosis. The aim of the present investigation is to develop clusters of genotypes for various yield-contributing characters, which may be used in hybrid wheat programme.
hundred and fifty two genotypes of spring wheat (Triticum aestivum L.) of national and international origin were
evaluated at Directorate of Wheat Research, Karnal during winter 1999-2000. The
genotypes were grown in paired rows of two-meter length. The distance between
rows was 0.23m while distance between plants was 0.02m. Five competitive plants
from each plot was randomly taken for recording observations on flag leaf area,
plant height, spike length and spike lets/spike. The flag leaf area was computed
by using the product of flag leaf length, breadth and a factor 0.80. The
observation on days to 50% heading was taken on plot basis. A random sample of
seeds from plot was used for measurement on protein content (%) and hectolitre
weight (kg\hl). The protein content (%) was estimated using Near Infrared
Reflectance Analyser (NIR) Model Infratec-1255 Food and Feed Analyser. The
hectolitre weight was estimated using “new DWR hectolitre instrument”. The
genetic divergence and related statistics were estimated by using D2 statistics
as suggested by Mahalanobis (1936) and Rao (1952).
hundred fifty two genotypes collected from various national and international
sources were grouped into different clusters based upon the method as applied in
non–hierarchical Euclidean cluster analysis (Beale 1969, Spark 1973). The
method is used to form a range of clusters from 2-15 so as to see if the
genotypes could be grouped into lesser number of clusters without any loss of
information. No major differences in variance of genotypes in each cluster was
observed between 6-clusters and 15-clusters, as 6 clusters were as effective as
15 clusters in providing useful information on diversity (Table
1). However, as we drop one more cluster from 6-clusters the value of
variance become just more than double which suggests inconsistency in grouping
the genotypes in to 5- clusters or below. Hence, it is judicious to group 152
genotypes in to 6-clusters without loosing any information. Suri and Sharma
(1999) had divided 200 genotypes of wheat into 10 clusters and found that grain
yield and tiller number were major contributors towards genetic divergence with
moderate contribution from 1000-grain weight, grains/ear and harvest index.
The intra-and inter-cluster distance for
6-clusters have been discussed in Table 2. It is
desirable to have high inter-cluster distance and low intra-cluster distance.
The inter-cluster distances range from 1.933 to 3.346. The maximum inter-cluster
distance (3.346) was observed between clusters II and I. The intra-cluster
distance was lowest in cluster I (1.942) and highest in cluster II (2.270).
Hence, desirable segregates as well as single cross hybrids can be obtained by
crossing between 28 and 18 genotypes spread in cluster I and cluster II,
respectively in various characters. The cluster superior in different characters
like yield (Cluster VI), protein (Cluster V), spikelets\spike (Cluster V), spike
length (Cluster II), flag leaf area (Cluster IV) and dwarf plants (Cluster III).
The members of cluster VI have highest value for yield and hectolitre weight but
were tall (Table 3). On the other hand members of
cluster V had highest value for protein % and spike lets\spike but it was poor
in hectolitre weight. A combination of both yield and protein can be attempted
in hybrids by selecting parents from cluster V and VI. The genotypes of cluster
II had long spike and more spikelets/spike coupled with long vegetative phase
expressed as delayed heading. Longer flag leaf was a characteristic of members
of cluster IV. Members of cluster III exhibited early heading with dwarf plant
type but poorer in protein, spike length and flag leaf size.
at the strengths of each cluster depending upon the out performance of a cluster
for a character or group of characters indicate that cluster I, II and III are
close to each other while cluster V and VI are close and complimented for yield
and protein characters (Fig.1). The overlapping
between clusters in Fig.1 indicates commonality
in superiority for specific characters under study. Hence, the cumulative effect
of desirable characters in wheat hybrids can be felt by attempting a number of
cross combinations between members of the clusters like V and VI with desirable
characteristics. Cross combinations between members of cluster V and VI may
combine high yield as well high protein content. It will be desirable to use
members of cluster VI as male which may be superior for yield and yield
components, hectolitre weight and tall while members of cluster V were superior
for protein content, yield components and dwarf as females. Singh and Chatrath
(1993) opined that grain yield, ear-bearing tillers and plant height were the
potent variables which could be used as parameters in selecting genetically
divergent parents in crossing programme for breeding high yielding wheat
varieties for salt-affected soils. The inter- and intra- cluster distances among
cluster III and I was lowest and had genotypes with early heading. On the other
hand on the basis of overall diversity among clusters, members of cluster I and
II are diverse.
To search some commonality in genetic background of genotypes in each cluster it was not possible to find any commonness in pedigree of genotypes especially in cluster V and VI (Table 4). Our observation was in accordance with Singh and Chatrath (1993), Suri and Sharma (1999) who observed that genetic divergence was independent of pedigree as well as place/origin.
authors are grateful to PI (CI) and PI (Quality) for providing necessary
facility for the conduct of the experiment. We are also grateful to Sh Subhash
Chander for assisting in improving the manuscript.
E.M.L., 1969. Euclidean Cluster analysis. A paper contributed to 37th
session of the International Statistical Institute.
P.C., 1936. On the generalised distance in statistics. Proc. Natl. Inst. Sci.
India. 2, 49-55.
C.R., 1952. Advanced Statistical Methods in Biometrical Research. John Wiley and
Sons, New York.
K.N. and Chatrath, R. 1993. Genetic divergence in bread wheat (Triticum
aestivum L. em Thell) under sodic soil conditions. 76, 35-38.
D.N., 1973. Euclidean cluster analysis. Algorithm As.58. Applied Statistics. 22,
V., Sharma, S.C., 1999. Genetic diversity in relation to number of clusters in
wheat (Triticum aestivum). Crop Improv.