Genetic distance provides a measure of the degree of relatedness between individuals in a population
(Garcia et al., 2004) and it also plays a key role in genetic improvement through breeding methods
(Liu et al., 2019). Results indicating genetic diversity showed sufficient dissimilarity characteristics and reflected significant genetic diversity among Indian mustard genotypes. Such significant genetic variation has also been reported by
Alie et al., (2009); Singh et al., (2010) on metric traits in
B. juncea. On the basis of Manhattan dissimilarity coefficients, 95 genotypes of Indian mustard were demarcated into five diverse clusters and discriminated these genotypes on the basis of 11 quantitative characters (Fig 1).
Vinu et al., (2013) and
Sheikh et al., (2011) also estimated genetic diversity among 44 genotypes of Indian mustard into four clusters using Manhattan methods.
Mahmud et al., (2008) and
Nath et al., (2003) reported four and five clusters in
Brassica species, respectively. Mean value of diverse clusters for eleven agronomic traits is presented in Table 2.
The first cluster comprised of 16 genotypes developed/maintained mainly by three centre
viz., ICAR-DRMR, Bharatpur (Rajasthan), PAU, Ludhiana (Punjab) and HAU, Hisar (Haryana). The genotypes of this cluster were characterized by tall plants (236.6 cm) higher seed yield per plant (19.87 g), high oil content (39.90%), long main shoots (87.40 cm), more number of siliqua on main shoot (64.6), more number of primary branches per plant (6.27), more number of secondary branches per plant (15.10) and maturity period of 139 days. Main shoot length is considered as the most important fruiting zone in mustard. Hence, its length and number of siliquae on main shoot are desirable traits for increasing seed yield. The genotypes of cluster-I may be used as donor parents for these traits. The second cluster had only 3-genotypes developed/maintained at ICAR-DRMR, Bharatpur and PAU, Ludhiana. These genotypes were poor performer for most of the characters possessing low seed yield, lower test weight and medium plant stature
etc. Eighteen genotypes were grouped in cluster III of which three were having exotic origin. Genotypes of this cluster belong mainly to three centers
viz., DRMR, Bharatpur, IARI, New Delhi and PAU, Ludhiana (except one from HAU, Hisar). In this cluster genotypes were characterized by medium early maturity (137 days), medium tall plant stature (177.6 cm), medium siliqua length (3.78 cm) along with medium number of seeds per siliqua (13.9), medium to high test weight and medium main shoot length. Low plant height is considered as desirable trait due to ease in carrying out agronomic practices; hence genotypes of cluster-III may be used as donor for this trait. Such results are in concurrence with the results of
Singh et al., (2013). The fourth cluster comprised of 23 genotypes from three centers (except two from IARI, New Delhi). The genotypes of this cluster were characterized with oil content (38.90%), moderate to high plant height and seed yield per plant. These genotypes were average performer for most of the characters. Thirty five genotypes were grouped in cluster V which randomly belonged to all four centers. The genotypes of this cluster had high test weight, taller plant stature, high main shoot length and moderate estimates for remaining characters. In earlier study,
Gohel and Mehta (2014);
Anushree and Pandey (2017);
Chandra et al., (2018) reported similar trend of genetic diversity in some oilseed genotypes. None of the cluster/genotypes was found to be most promising collectively for all the quantitative traits. However, some genotypes can be identified as promising for different traits (Table 3). In breeding programmes, these genotypes can play a significant role in achieving specific goals and also be helpful in broadening the genetic base of mustard germplasm.
The Manhattan dissimilarity coefficients ranged from 0.741 to 8.299 indicating the diverse nature of genotypes under study. Based on the genetic dissimilarity matrix, the maximum dissimilarity (8.299) was observed between the genotypes, DRMRIJ-15-133 and M 62. On the other hand, a minimum dissimilarity value of 0.741 was found between genotypes, RC-110 and NPJ-161 which was followed by 1.236, between M 13 and Pusa Jagannath and 1.240, between DRMRIJ-14-30 and Pusa Barani; DRMRIJ-15-251 and Pusa Barani. The genotype, DRMRIJ-15-133 was found to be the most diverse as it showed the highest dissimilarity coefficient values (8.299) with all of the genotypes
viz. M 62; DRMRIJ-14-23 and M 28
etc. These diverse genotypes can be used effectively in the mustard breeding programme to select some desirable recombinants. Thus the obtained results confirmed that the use of diversity analysis is a good tool to determine the phenotypic differences among the genotypes, which agrees with the results of
Crossa and Cornelius (1997);
Marijanovic-Jeromela et al., (2009). Similar results concerning the genetic diversity for yield and its component traits have also been reported by
Singh et al., (2013); Vinu et al., (2013); Chandra et al., (2018).
Principal coordinate analysis (PCoA)
A two dimensional scattered plot of the genotypes was constructed based on two principal axes to visualize the resemblance or divergence between individual genotypes. The genotypes were distantly located from each other. The scatter plot revealed that majority of samples placed at the center of a two-dimensional coordinate plane formed five apparent clusters C-I, C-II, C-III, C-IV and C-V (Fig 2). There is a strong tendency for the PCoA to show the same trends with clustering of lines as in the dendrogram.
Vinu et al., (2013) also used PCoA to delineate and visualize 44 Indian mustard genotypes into four clusters.
