Assessment of Genetic Divergence by Using Multivariate Analysis for Physico Chemical Characters of Mango Table and Juicy Cultivars Grown in Telangana Region

Mango (Mnagifera indica L.) is a major fruit crop which is widely grown in tropical and subtropical countries of the world. It’s related to Anacardiacea family in order Sapindales. India ranks first in production and consumption of mango in the world. In India, mango is cultivated in an area of 2.31 million ha with production of 22.35 million tonnes and 7.3 MT/ha productivity (NHB Data Base 2019-20). Major mango growing states are Uttar Pradesh, Andhra Pradesh, Odisha, Karnataka and Telangana.
 
In India, 1000 cultivars are well known for its commercial importance, Telangana state also has richest source of mango germplasm. Mango is having the chromosome number 2n=40 and n=20 and the genome size is 4.39 × 108 bp which is small in size but, morphologically distinguished among the cultivars of mango (Arumuganathan and Earle, 1991).
 
Genetic diversity analysis is generally used to identify diverse genotypes for hybridization purpose. Analysis of divergence can be performed by many statistical tools such as D² Mahalanobis distance, Euclidean distance, average distance, etc. (Shirkhorshidi et al., 2015). Among the various methods identified to assess the genetic divergence in the varieties, the Mahalanobis D² (Mahalanobis, 1936) is reliable and most frequently used technique. D² analysis is a useful tool in quantifying the degree of divergence between biological population at genotypic level and to assess relative contribution of different components to the total divergence, both at the inter- and intra-cluster levels by using the concept of statistical distance employing multivariate measurements and Principal Component analysis is used to identify traits that explain the phenotype variability best.
 
There is a huge genetic resource exists in Telangana state. 477 mango cultivars established at Fruit Research Station (FRS), Sangareedy. Based on the strength of mango germplasm an attempt has been made to estimate genetic diversity among the mango table and juicy cultivars based on post-harvest physico chemical traits by using multivariate analysis. The present experiment was carried out with fifty mango table and juicy cultivars which are commercially importance of the state and an experiment entitled “Assessment of genetic divergence by using multivariate analysis for physico chemical characters of mango table and juicy cultivars grown in Telangana region”.

The experimental material comprised of fifty mango table and juicy cultivars which were collected from Fruit Research Station, Sangareddy by random sampling method. Present experiment was conducted on pre-established 15 years old and uniformly maintained mango orchard at FRS during the year 2018-2019 and 2019-2020 further analysis was carried out at College of Horticulture, Rajendranagar. Both table (36) and juicy (14) cultivars of mango used in the experiment.
 
Selection and preparation of fruits
 
Selected varieties were harvested based on their maturity standards and these fruits immediately transported to the laboratory and subjected to ripening with enripe and kept in low-cost ripening chamber. Fifty mango cultivars were analysed for all quantitative and qualitative traits and all these characters were subjected to completely randomized design with three replications, the level of significance was tested at 5% using F test (Panse and Sukhatme, 1989). The genetic divergence among the mango cultivars was worked out by using Mahalanobis’s D² statistics (1936) and principal component analysis was done to identify traits that explain the phenotype variability best in SPSS version v.27.

The results obtained from the principal component analysis was given in Table 1 and Fig 1. D² statistics (Toucher’s method (Rao, 1952) based on D² values and distribution of genotypes in each cluster, Intra and inter cluster distances, cluster means and per cent of contribution towards genetic diversity was accounted in Table 2 to 4 and Fig 2 and 3.
 

 

 

 

 

 

 

 
Principal component analysis (PCA)
 
PCA analysis based on correlation matrix for physico chemical characters studied in fifty mango cultivars includes principal components, eigen values, per cent of variability and cumulative per cent of variability and also factor loading values of different traits for the respective principal components are furnished in Table 1 and Fig 1. The principal components with eigen values above one was considered as significant and less than one was considered as non-significant as per the procedure.
               
As per the PCA analysis, the first seven principal components with eigen values more than one are explained 74.71% of the total variance among the fifty mango cultivars. The first principal component (PCI) accounted for 27.73% of total variation, included fruit length (cm), fruit width (cm), fruit thickness (cm), fruit weight (g), pulp per cent, pulp to peel ratio, pulp to stone ratio, shelf life, brix acid ratio, total sugars (%), reducing sugars (%) showed positive loadings. The second principal component was explained 14.28% total variation and was positively associated with fruit length (cm), fruit thickness (cm), stone per cent, firmness (kg/cm²), DA reading, fibre content (g/100g). The third principal component accounted for 8.99% variability and showed high positive correlation for acidity (%), pulp (%), non reducing sugars (%), total sugars (%), fruit weight (g). The fourth principal component explained the variability 8.59% and positively associated with firmness (kg/cm2), total sugars (%), non reducing sugars, total flavonoid content (mg QE/100 g).

The fifth component was showed 5.52% of the total variation and was correlated with pulp per cent, TSS (°Brix). Sixth component was accounted 5.31% variation and associated with TSS (°Brix), total phenolics content (mg of GA/100 g), total flavonoid content (mg QE/100g) and antioxidant activity (µg/100 g) and fibre content (g/100 g). Seventh principal component was explained 4.27% of total variation and correlated with vitamin C content (mg/100 g) and total phenolics content (mg of GA/100 g). The eighth principal component was accounted 3.78% of total variation and associated with fruit width (cm), reducing sugars (%) and total phenolics content (mg of GA/100 g).
       
However, the first two principal components for fifty cultivars were showed maximum variation and widely distributed along the axis which also showed in Fig 1. The similar results were also noticed by Krishnapillai and Wijeratnam (2016) and Majumdar et al., (2013), Tewodros Bezu Neguse et al., (2018) and Himabindu et al., (2017) in mango. Hence, it is indicating that, to give emphasis on traits which had a significant contribution to the observed variation for future breeding program.
       
The fifty mango cultivars were grouped into and eight clusters by using D² analysis was illustrated in Table 2 Fig 2. The results showed that cluster I comprising of 26 genotypes followed by cluster II and cluster III each with nine cultivars. Cluster 4 consisting of two cultivars viz., Pulihora and Yerra Mulgoa. Cluster V, VI, VII and VIII uniform clusters (Vanraj, Shendriya, Kaju and Sora respectively). All popular table cultivars except Zardalu, Chinnarasam, Kothapalli Kobbarii, Panchavarnam and Cherukurasam were grouped in cluster I. Almost all juicy cultivars except Mahamooda Uppal were formed as single group (cluster II). Large sized table cultivars except one juicy cultivar (Nagulapalli Iraslu) formed as cluster III. Genotypes with high fruit weight can be utilized in crossing programme to realize broad spectrum of the genetic variability in segregating generations to affect the selection for fruit weight improvement. This clustering pattern clearly reflects the presence of considerable extent of genetic diversity among the genotypes under study. Similar results in relation to formation of large sized table cultivars in a cluster were reposted by Kumar et al., (2006); Rathod (2007) and Raina et al., (2015), Himabindu et al., (2015) in mango.
       
Dinesh et al., (2015) attempted to study the genetic diversity in some indigenous mango varieties of seedling origin and carried out evaluation of morphological traits in the Chittoor area of Andhra Pradesh in India.
       
The average intra and inter cluster distances for fifty genotypes are furnished in the Table 3. Inter cluster distance ranged from 218.93 between cluster V and VII to 1475.21 between cluster IV and VIII and it was maximum between cluster IV and VIII (1475.21) followed by clusters VI and VIII (1269.55) and clusters VII and VIII. Cluster VIII showed maximum inter cluster distance with other clusters indicating wide genetic diversity between the genotypes. Selection of parents from such clusters for hybridization programme would help to evolve novel hybrids in mango. Intra cluster distance ranged from 0.00 in cluster V, VI, VII and VIII to 280.69 in cluster III. Cluster III contained 9 cultivars showing maximum intra cluster distance (280.69) thus, these cultivars were most heterogenous and followed by cluster II (212.22), cluster I (189.56) and cluster IV (106.10). Hence, genotypes from these clusters may be utilized in the hybridization programme to produce wide variability and transgressive segregants from diverse parents. Similar studies were conducted by Rajan et al., (2007) in guava; Rai and Misra (2005) in Bael, Kalia et al., (2001), Govanakoppa et al., (2002), Ramaprasad et al., (2006), Sharma et al., (2013) in apple, Barhate et al., (2012); Barholia and Sangeeta (2014); Indian et al., (2019), Himabindu et al., (2017), Shazia et al., (2017), Indian et al., (2019); Manchekar et al., (2011) and Rajan et al., (2009) in mango.
       
The genetic diversity was also corroborated with cluster means of fifty genotypes for different physico chemical traits under study revealed that considerable differences between the groups was given in Table 4. From the present data, it is evident that cluster I was characterized with minimum fruit thickness (5.9), total sugars (8.88) and reducing sugars (4.19). Cluster II was found to have genotypes with maximum mean values for DA reading (1.52), antioxidants (229.65) and fibre content (5.88) with minimum shelf life. The highest mean value for fruit width (9.08), fruit weight (460.03), pulp to stone ratio (5.12), brix acid ratio (56.05) with lowest mean value for peel per cent (13.88), stone per cent (12.39) and total phenols content (86.02) was observed in cluster III, indicating that genotypes having wide genetic base and desirable characters could be utilized in selection of parents in mango breeding. TSS, non-reducing sugars, flavonoids (19.59, 5.58 and 297.07) were recorded maximum under cluster IV. Cluster VI mixed up with desirable characters like TSS, total sugars, total phenols, antioxidants and flavonoids. The maximum mean value for physiological loss in weight (8.58), fruit length (12.11), fruit weight (770.64) pulp per cent (94.26), pulp to stone ratio (6.37), acidity (0.61) and minimum mean value for firmness (1.11), flavonoids (47.32) and beta carotene content (1.54) was found in cluster VIII.
 

       
The mean obtained for various characters from different genotypes in each cluster gives an idea about diversity among the clusters compared. It also helps to group the clusters according to their average performance. These results are in line with the reports of Manchekar et al., (2011), Shazia et al., (2017), Majumdar et al., (2013), Himabindu et al., (2017), Rajan et al., (2009), Sandra et al., (2013), Barholia and Sangeetha (2014), Rathod (2007), Indian et al., (2019) in mango; Ismail (2008) in case of lemon.
       
It is very essential to know the characters whose contribution is the most for the total genetic diversity so as to improve that character in the further breeding programs. It was observed from the per cent contribution data (Table 4 and Fig 3), among physical traits fruit weight (g) ranked first with a maximum contribution towards genetic divergence of 21.14 per cent followed by DA reading (11.51%), fruit thickness (4.24%), fruit length (4.65%). Characters such as stone per cent, pulp to peel ratio, pulp to stone ratio, shelf life had no contribution towards total divergence. Among the chemical traits total phenols content (17.22%), flavonoids (11.67%), fibre content (10.12%), beta carotene content (8.49%) antioxidants (7.10%), reducing sugars (1.14%), brix acid ratio (0.98%), total sugars (0.16%) contributed towards the genetic diversity in decreasing order. Acidity, vitamin C and non reducing sugars did not contributed towards diversity.
       
The experimental results further revealed that the mango genotypes selected for the present study are most divergent for total phenolics content, total flavonoid content, DA reading, fibre content, beta carotene content, antioxidants. Therefore, these characters should be given greater importance for the improvement of quality in further selection of segregants and choice of parents during hybridization programmes in mango. Similar studies were also carried out by Singh (2005), Rajan et al., (2009), Rufifni et al., (2011), Barhate et al., (2012), Majumder et al., (2013), Barholia and Sangeetha (2014) and Sandra et al., (2013) in mango; Clemilton et al., (2017) in papaya and Singh et al., (2003) in pomegranate.

In the present investigation Fifty mango genotypes were grouped into eight distant clusters by performing Tocher’s clustering method using Mahalanobis D² statistical analysis. Cluster I was the largest and comprised of 26 genotypes, cluster II and III each with 9 genotypes. Cluster IV with two cultivars and cluster V, VI, VII and VIII formed as solitary cluster. Inter cluster distance was maximum between cluster IV and VIII (1475.21). The cluster mean for most of the desirable characters was found in cluster II and III. Fruit weight contributed maximum towards genetic diversity followed by total phenolic content, total flavonoid content, DA reading, fibre content, beta carotene, antioxidants, fruit thickness and fruit length. Selection of genotypes from clusters that are far genetically and had the maximum association of physico chemical characters to obtain superior segregants in advanced generation to improve the quality of breeding in future.

The authors thank the SKLTS Horticulture University, Rajendranagar for providing financial support and Fruit Research Station, Sangareddy for providing fruits and also Institute of Biotechnology and Quality Control Laboratory (PJTSAU, Rajendranagar) for technical support.