Malaysian Journal of Sustainable Agriculture (MJSA)

Genetic diversity is essential to explore novelties in flower and plant architecture to develop new cultivars. 50 genotypes of Chrysanthemum morifolium were systematically characterized for accurate genotype identification based on DUS descriptors in UPOV guidelines. The experiment was conducted for two consecutive years. Significant variation was observed for fourteen quantitative traits. Shannon equitability index was > 0.75 for 6 out of 11 qualitative traits indicating good diversity. The highest GCV (67.52%) and PVC (68.05%) were recorded for number of flowers per plant. High estimates of heritability and genetic advance as percentage of mean were shown by number of flowers per plant, stipule size and ray floret width suggesting effective selection. Principal component analysis (PCA) for both quantitative and qualitative traits performed independently, resulted in 4 principal components accounting for 74.7% and 63.27% of total variability respectively. Factor loading for quantitative traits revealed that leaf characters have major contribution to germplasm variability. Among qualitative characters, ray florets traits have major contribution to germplasm variability. The genotypes were classified into 4 groups at Euclidean distance of 88.94. Cluster 1 has only genotype Gauri which is an outlier for plant height and number of flowers per plants. Pusa Shwet is alone in cluster 2 having unique combination of large sized and semi double type flower head. Cluster 3 and cluster 4 have 28 and 20 genotypes respectively. Grouping based on Jaccard’s similarity coefficient for binary data gave nine clusters at similarity coefficient of 0.33 with genotype Gauri alone constituting cluster 3.

2.MATERIALS AND METHODS

The experiment comprising 50 diverse chrysanthemum genotypes was laid at the experimental farm, Division of Floriculture and Landscaping, IARI, New Delhi in randomized block design with two replications during two consecutive years (2021–2022 & 2022-23). The experimental plot (4.5 m × 1.8 m) comprised of three rows distanced 60 cm apart with plants spaced at 45 cm apart. Standard agronomic practices were followed during crop growth. Data were recorded on 10 randomly selected plants for 25 descriptors including 14 quantitative characters viz., Plant based traits: plant height (PH) & number of primary branches (NPB), Leaf based traits: stipule size (SZ), leaf lamina length (LL), leaf width (LW), ratio of leaf length/width (L/W), petiole length (PL), Terminal lobe length (TLL), length of lower lobe of leaf (LLL), flower based traits: number of flowers per plant (NFP), flower head diameter (FD), peduncle length (Pd. L), ray florets length (RL), ray florets width (RW) and 11 qualitative characters viz., Plant based trait: plant type (PT), leaf based trait: petiole attitude, leaf predominant shape of base, leaf color, flower based traits: flower head type, number of types of ray floret, predominant type of ray floret, cross section of ray floret, rolling of margins of ray floret, longitudinal axis of
majority of ray florets and shape of tip of ray floret.

The picture of flower type of chrysanthemum genotypes is given in Figure1. Plant characters, stem, stipule, leaf and petiole characters were observed when the terminal buds show color, just before they open. Plant height was measured from the tallest point of the canopy to the base of the plant. Plant type was visually observed as bushy (having branches and nonmain single stem) or non bushy (producing single stem). Leaf length was measured from the lamina tip to the intersection of lamina and petiole along the lamina midrib. Leaf width was measured from the widest lamina lobes. Petiole length was measured from the stem margin to the end of lea lamina. The number of flowers was counted during the flowering time of each cultivar. Flower head diameter was measured from the widest point of the flower. Ray floret characteristics were observed on the outermost row of the floret. Both quantitative and qualitative traits were observed according to UPOV (International Union for the Protection of New Varieties of Plants) descriptors. The year-wise quantitative data were pooled and used for statistical Analysis. The analysis of variance for quantitative traits was computed as described (Panse and Sukhatme,1967).

Phenotypic (σ 2p) and genotypic (σ 2 g) variances were calculated using the method suggested by a researcher as σ 2p= σ 2 g + σ 2 e, σ 2 g= MSg–MSe/rand σ 2 e= MSe, where MSg, MSe and r denote mean squares of genotypes, mean squares of error and number of replications respectively (Baye,2002). The PCV and GCV were obtained by using the formulae, PCV(%)=√σ 2 p / x̅ × 100, GCV(%) =√σ 2 g / x̅ × 100,where x̅ is the sample mean(Baye, 2002). GCV and PCV values were categorized as low (< 10%),moderate (10–20% ) and high (> 20%) (Deshmukh et al., 1986). Estimates of broad sense heritability (h2b) were calculated according to the formulae:h2b= σ 2 g / σ 2 p (Allard, 1999). The expected genetic advance (GA) under selection, assuming the selection intensity of 5% was calculated as proposed by a group researchers i.e. GA = k (√σ2p). σ2g / σ2p, K is the standardized selection differential (2.5) (Johanson et al., 1955). Geneticadvance as percent of mean, GAM = (GA/ x̅) x100. The Shannon equitability index (evenness) is simply the Shannon diversity index (H) divided by the maximum diversity (log k). Shannon diversity index (H) is calculated as H= Σ 𝑃𝑖 ∗ 𝐼𝑛 𝑃𝑖 𝑘𝑖=1 , where k denotes the number of group and pi denoting the proportion in group k.

This normalizes the Shannon diversity index to a value between 0 and 1.Evenness closer to 1 indicates more diversity (Shanon and Weaver, 1949).The mean values of each quantitative trait and transformed data (scoresfor the each descriptor state used for transforming qualitative data is given supplementary table S3) of qualitative traits were used to perform PCA and cluster analyses using unweighted pair group method with arithmetic averages (UPGMA) clustering algorithm with dissimilarity between genotypes expressed as Euclidean distance (Kim et al., 2014; Kachare etal., 2016; Khatun et al., 2022). Euclidean distance between two genotypes and j, having observations on morphological characters (p) was denoted by x1, x2, …, xp and y1, y2,…, yp for i and j, respectively, and was calculated with the following formula: d (i, j) = [(x1 – y1)2 + (x2 – y2)2 + (xp- yp)2]1/2.The computation of PCA, Euclidean distance matrices, construction of dendrogram and calculation of cophenetic correlation coefficient was done with the help of XL STAT software (Khatun et al., 2022). Additionally Jaccard’s similarity coefficient was used to perform cluster analysis on binary data based on presence (1) or absence (0) of specific descriptor state (UPOV guidelines) for both qualitative and quantitative trait (Chunget al., 2019). A cluster analysis based on Jaccard’s similarity index wasachieved by using the UPGMA in NTSYS-pc Version 2.1q (Chen et al., 2013).

3. RESULTS

3.1 Morphological Characterization

A total of 25 morphological characteristics (14 quantitative and 11qualitative traits) in 50 genotypes were evaluated. The Mean, range,treatment mean sum of square, standard error of difference (SED) andcoefficient of variation (CV) for quantitative characters are given in Table1. Significant variation was reported among the genotypes for all thequantitative traits. Plant height ranged from 24.64 cm (Basanti) to 92.95cm (Gauri) which was divided into 19 short (20–40 cm), 27 medium (40–60 cm) and 6 tall (> 60 cm) genotypes. Observation on plant type showedthat 44 genotypes were bushy type and 6 were non-bushy type. Numberof primary branches varied from 3.5 (Thaichen queen ) to 23 (Jaya) andthe genotypes were grouped into 4 sparse (<6), 15 medium (6-12) and 31dense (>12) branching.

** means significant at 0.1 level of significance

Figure 2 shows the pictures of leaves. The shape of base of leaves wasclassified into 16 acute, 10 obtuse, 6 rounded, 5 truncate, 5 cordate and 8asymmetric shape. Leaf color was categorized into three types light,medium and dark present in 11, 21 and 18 genotypes respectively. Forpetiole attitude, 18 genotypes showed horizontal petiole attitude, 15 verystrongly upward attitude, 13 moderately upward attitude, 3 moderatelydownward and 1 showed drooping petiole attitude. The leaf lengthsranged from 3.2 cm (Kaul) to 10 cm (Pusa Chitraksha), wherein 12genotypes have short (< 5cm), 22 genotypes medium (5-7 cm) and 16genotypes have long (>7cm) leaves. Leaf width varied from 2.36 cm (Kaul)to 7.9 cm (Star).

Yellow where 13 genotypes bear narrow leaf (<4cm), 21 genotypesmedium leaf (4-5 cm) and 16 genotypes have broad leaf (> 5 cm). Leaf L/Wratio ranged from 1.2 (Ajay orange) to 2.87 (Garden Beauty) andgenotypes were described as 5 low (<1.5), 41 medium (1.5 – 2.5) and 4high (>2.5) L/W. For petiole length the genotypes were grouped as 33short (< 2.5 cm), 13 medium (2.5 cm – 4.5) and 4 long (>4.5 cm). Shortestpetiole length was observed in Kaul (0.83 cm) and longest in Gardenbeauty (6.05cm) which was at par with Sunny (5.85 cm). For terminal lobelength genotypes were was evaluated as 12 short (< 2 cm), 17 medium (2cm – 3 cm) and 21 long (> 3cm). Agni and Kaul had shortest terminal lobeof 1.3 cm and Star Yellow had longest one i.e. 5.02cm and was at par withSunny (4.55cm) and Chitraksha (4.5 cm). The lower lobe length rangedfrom 0.69 cm in Kaul to 4.6 cm in Tata Centenary and genotypes werecategorized as 14 short (< 2 cm), 17 medium (2 cm – 3 cm) and 19 long (>3cm). The stipule size varied from 0.45 cm in Silk Brocade to 2.9 cm in Jayaand genotypes were grouped as 27 small (< 1 cm), 13 medium (1 cm – 2cm) and 10 large (> 2cm).

Evaluation of number of flowers per plant showed that 12 genotypes bearfew flowers (< 50), 17 medium (50 – 100) and 21 genotyes had manyflowers (>100). Gauri was a outlier with highest number of flowers (400)whereas Raja and Garden Beauty had the least (34). Flower diameter wasobserved as small (< 4 cm), medium (4 cm – 8 cm) and large (>8 cm) in 4,36 and 10 genotypes respectively. Gauri had the smallest flower headdiameter of 2.7 cm and Thaichen queen had the widest diameter of 11.4cm followed by Tata Centenary and Star Yellow. Based on peduncle lengthgenotypes were grouped as 4 short (< 5 cm), 33 medium (5 cm – 9 cm) and13 long (> 9 cm). Shortest peduncle length was observed in Anmol (3.4cm) and longest in Himanshu (11.7 cm). Flower head was classified into 4categories viz. single type 5 genotypes, semi double 9, daisy eyed double 6and 30 genotypes were double type. Different genotypes had wide rangeof flower color (supplementary table S1). Ray floret length was observedas short (< 2 cm) , medium (2 cm – 3 cm) and long (> 3 cm) present in 8,18 and 24 genotypes respectively. Longest ray floret was in TataCentenary (6.1 cm) and the smallest in Gauri (1.42 cm).

The ray floret width ranged from 0.27 cm in Winter queen to 1.7 cm in StarYellow. Based on ray floret width genotypes were classified as 11 narrow(<0.5 cm), 34 medium (0.5 cm – 1 cm) and 5 broad (> 1 cm). Out of 50cultivars, 6 cultivar consist of two types of ray floret, among theseAgnireka and Silk Brocade had spatulate and ligulate ray floret, BidhanManasi and Little pink had spatulate and quilled ray floret and Himanshuand Bidhan Bishnupriya had Ligulate and funnel shaped ray floret. Fourgenotypes comprised of three types of ray floret, out of which 3 genotypesnamely Pusa Chitraksha, Ajay orange and Himani contain spatulate,ligulate and quilled ray floret, whereas Haldighati consist of incurved,spatulate and quilled type of ray floret. The remaining genotypesconstitute of only one type of ray floret. Ligulate, spatulate, incurved,funnel shaped and quilled were predominant ray floret in 27, 16, 5, 1 and1 genotype respectively. For the ray floret-cross section, 4 genotypes werewith strongly concave ray floret, 8 had moderately concave, 9 showedweakly concave, 15 had flat and 13 exhibited weakly convex ray floretcrosssection, the variety Winter Queen was quilled type, therefore thischaracter was not applicable in this.

Besides flat margins of ray floret in 34 genotypes, 3 more types:moderately involute in 7, weakly revolute in 7 and weakly involute in 1genotype, this character was also not applicable in cultivar Winter Queen.The longitudinal axis of ray floret was classified into four types, out of 50genotypes studied, 1 genotype had twisted longitudinal axis, 6 genotypesshowed incurving longitudinal axis, 5 cultivars showed reflexinglongitudinal axis and remaining 38 genotypes showed straightlongitudinal axis. Depending on the shape of tip of ray floret, 50 genotypeswere categorized into five types viz. emarginated (10), pointed (13),mammillate (5), dentate (9), rounded (12) and fringed (1). The aboveclassification of quantitative and qualitative traits is based on descriptorstate for different characters given in UPOV DUS guidelines. The meanperformance of various genotypes for 14 quantitative characters andobserved trait for 11 qualitative trait is given in table 2. Figure 3 gives boxplot depicting data dispersion for quantitative traits.

3.2 Shannon Equitability Index for qualitative traits

The Shannon Equitability (EH) was estimated for 11 qualitative charactersto measures the evenness of different types in the population. It variedfrom 0.52 to 0.96. Six characters showed EH greater than 0.75 viz. leafshape of base (0.94), leaf green color of upper surface (0.96), petioleattitude (0.82), flower head type (0.79), ray floret profile in cross section(0.94) and ray floret shape of tip (0.91). Diversity of phenotypic classes forqualitative trait (supplementary table S2).

3.3 Genetic Variability, Heritability and Genetic Advance

Estimates of genotypic coefficient of variance (GCV) and phenotypiccoefficient of variance (PCV) of different traits are given in Table 3. Thehighest GCV and PVC values were found particularly for number of flowersper plant (67.52% and 68.05%), stipule size (54.33% and 54.76%),petiole length (45.8% and 48.00%) and Ray floret width (45.71% and46.21%) respectively. Whereas moderate GCV and PCV were recorded forleaf lower lobe length (40.17% and 41.51%) and ray floret length (36.08%and 36.29%) respectively. Low GCV and PCV viz. 17.28% and 18.98%respectively was recorded for ratio of leaf length to leaf width indicatingexistence of less variability. Most of the traits in this study showed broadsense heritability > 90%. Ray floret length (98.84%) exhibited highestheritability followed by number of flowers per plant (98.43), stipule size(98.42) and ray floret width (97.84%). Leaf length to width ratio showedrelatively low heritability. Genetic advance as percent of mean was highestfor number of flowers per plant (138.01) followed by stipule size (111.03)and ray floret width (93.13).

3.4 PCA Results

PCA identifies variables that are most significant in describing the overallvariability in the dataset. PCA was performed separately on 14quantitative characters and 11 qualitative traits is given in Table 4 & 5respectively showing eigen values, factor loading, proportion ofvariability and cumulative variability.

3.4.1 Quantitative traits

A total of 14 quantitative characteristics of Chrysanthemum wereevaluated for classification in multivariate analysis. Eigen values variedfrom 0.01 to 5.33 (Figure 5 A). 4 principal components were having Eigenvalues more than 1.00 viz. PC 1 (5.33), PC 2 (2.24), PC 3 (1.52) and PC 4(1.37) which together explaining 74.7% of the phenotypic variationpresent in the data. The PC 1 accounted for 38.09% of the total phenotypicvariability with major contribution from 6 characters namely leaf laminalength, leaf width, Terminal lobe length, stipule size, flower diameter andray floret length. The PC 2 covered 16 % of the total variation and wasclosely related with 4 characters viz. petiole length, leaf length to widthratio and number of flowers per plant. The PC 3 constituting10.83% of thetotal phenotypic and is mainly contributed by leaf length to width ratio. PC4 explains 9.83% of variability with major contribution from pedunclelength. The biplot (Figure 4) provides an insight into the direction ofcorrelation between variables. Flower diameter showed highly positivecorrelation with ray floret length and ray floret width. Number of flowersper plant is positively correlated with number of primary branches. Leaflamina length strongly and directly correlated with terminal lobe length.

3.4.2 Qualitative characters

PCA for qualitative traits was conducted using harmonized values(transformed data. Eigen value ranged from 0.22 to 2.66 (Figure 5 B). 4principal components viz. PC 1 (2.66), PC 2 (1.61), PC 3 (1.51) and PC 4(1.2) were having Eigen values more than 1.00 which cumulativelyexplained 63.27% of the total phenotypic variation present in the data. PC1 explains 24.18% of the total variation and was correlated with ray floretrolling of margins, ray floret cross section, plant type and ray floretlongitudinal axis. PC 2 accounted for 14.6518% of total phenotypicvariability and mainly contributed by petiole attitude, leaf color andpredominant type of ray floret. PC 3 contributes 13.71% of total variabilityand is associated with flower head and ray floret: shape of tip. PC 4constitutes 10.73% with major contribution from number of types of rayfloret. The biplot for qualitative trait is given in Figure 6.

3.5 Cluster analysis based on Euclidean distance

Unweighted pair group method with arithmetic average (UPGMA) wasused for cluster analysis and Euclidean distance matrices were alsoconstructed. The harmonized values for qualitative trait and measuredvalue of quantitative traits were used for analysis. The Euclidean distanceranged from 5.37-372.36. The maximum distance of 372.36 was observedbetween Gauri and Raja belonging to cluster 1 and cluster 3 respectively.This was followed by a distance of 372.02 between Gauri (cluster 1) andGarden beauty (cluster 3) and a distance of 369.72 between Gauri andRoyal Princess (cluster 3). The closest related cultivars were BidhanLalima and Bidhan Mallika with a distance of 5.37 followed by BidhanMadhuri and Bidhan Sabita with a distance of 6.7 and all four of them weregrouped together in cluster 4. The UPGMA dendrogram aligned with thedistance matrix as indicated by cophenetic correlation coefficient value of0.92.At a Euclidean distance of 88.94 the genotypes were grouped into fourclusters. The distribution pattern revealed maximum number ofgenotypes i.e., 28 in cluster 3 followed by cluster 4 having 20 genotypesand cluster 1 and cluster 2 have Gauri and Pusa Shwet respectively.Cluster 3 has 28 genotypes viz. Ajay orange, Anemone red, Baggi, Basanti,Bidhan Bishnupriya, Bidhan Chitra, Bidhan manasi, Classic, Discovery,Garden beauty, Jyotsana, Little pink, Magenta , Mother Teresa, Kaul,Neelam, Prevalo, Punjab Shyamal, Pusa Arunodaya, Pusa Centenary, Raja,Royal princess, Sharad, Star yellow, Sunny, Tata centenary, Thaichenqueen and Winter queen. Cluster 4 comprise of 20 genotypes namely Agni,Agnirekha, Aprajita yellow, Bidhan Lalima, Bidhan Madhuri, BidhanMallika, Bidhan Sabita, Chandni, Haldighati, Himani, Himanshu, Jaya, PusaAditya, Pusa Chitraksha, Pusa Guldasta, Ragini, River city, Sensation andSilk Brocade. Figure 7 represents clustering based on Euclidean distance.

3.6 Cluster analysis based on jaccard’s similarity index

UPGMA cluster analysis was also performed based on Binary data toresolve the genetic relationships among the 50 genotypes. The Jaccard’ssimilarity coefficient varied from 0.04 to 0.69. The dendrogram dividedthe genotypes into 9 main clusters at the similarity coefficient of 0.33. Theminimum Jaccard’s similarity coefficient was 0.04 between genotypeBidhan Manasi and Thaichen Queen which belong to cluster 5 & 1respectively, whereas maximum Jaccard’s similarity coefficient of 0.69was same between Agnirekha and Bidhan Lalima, Bidhan Lalima andBidhan Madhuri and Bidhan Sabita and Chandini. Cluster 1 contained 5genotypes namely Thaichen Queen, Star Yellow, Tata Centenary, PusaCentenary and Pusa Arunodaya. Cluster 2, 4 and 5 have two genotypeseach. Cluster 2 consisted of Garden Beauty and winter Queen. BidhanBishnupriya and Jaya made cluster 4. Ajay orange and Bidhan Manasiconstituted Cluster 5. Cluster 3 and 6 had only one genotype each viz. Gauriand Punjab Shyamal respectively. Cluster 7 had 6 genotypes namelyAnmol, Magenta, Mother Teresa, Kaul, Royal Princess and Rivercity.Cluster 8 comprised of 7 genotypes namely Discovery, Haldighati, Himani,Sunny, Raja, Pusa Chitaksha, Little Pink. There were 24 genotypes incluster 9 namely Agni, Pusa Aditya, Sensation, Agnirekha, Bidhan Lalima,Bidhan Madhuri, Neelam, Prevalo, Bidhan Sabita, Chandini, Basanti, SilkBrocade, Bidhan Mallika, Ragini, Pusa Shwet, Baggi, Sharad, Aprajitayellow, Pusa Guldasta, Bidhan Chitra, Anemone Red, Jyotsana, Classic andHimanshu. Cluster 9 showed sub-clustering at Jaccard’s similarity index of0.35 with two sub clusters i.e. cluster 9a and cluster 9b. Cluster 9a included4 genotypes namely Anemone Red, Jyotsana, Classic and Himanshu.Cluster 9b contain rest of the 20 genotypes. Fig 8 represent clustering of50 chrysanthemum genotypes based on Jaccard’s similarity index.

4. DISCUSSION

Assessing the diversity within a crop species is crucial for identifyingunique traits or alleles that can be used to enhance desirablecharacteristics. Crossing genotypes from different clusters with greatergenetic distance can enhance genetic diversity and potentially result inimproved crop varieties with desirable traits.

4.1 Morphological characterization

The 50 accessions showed significant variation with respect to 14quantitative characters. Shannon equitability index showed that 6 out of11 qualitative traits had equitability index of more than 0.75 indicatinghigh level of diversity. Reasonable diversity was exhibited by predominanttype of ray floret (0.67) and ray floret rolling of margins (0.63). Plantattributes like plant height, plant type and number of primary branchesalong with floral characteristics determines the end use ofchrysanthemum as pot plant, loose flower, garden display and cut flower.Basanti, Himanshu, Prevalo, Sharad and Jyotsana could be used for potpurpose based on plant height, flower color and good number of flowering.Based on number of flowers, Gauri, Pusa Shwet, Himani, Jaya, BidhanLalima, Bidham Madhuri, Bidhan Mallika, Bidhan Sabita, and Chandni aresuitable for loose flower purpose. Similar evaluation of genotype was done(Suvija et al., 2016). Leaf characteristics enable early identification ofvarieties which helps in early selection in breeding (Gao et al., 2020). Itwas found that leaf length and petiole length were important parametersin the evaluation of hybrid varieties in breeding studies. Differences wereobserved in terms of both leaf size and leaf shape. Morphological variationfor leaf characteristics in Chrysanthemum have also been reported (Zhenet al., 2013). Variation in flower characteristics among Chrysanthemumgenotypes were reported earlier and might vary depending on climaticconditions (Guo et al., 2008; MacDonald et al., 2017; Wang et al., 2021).Ray floret characteristics determine aesthetic value of cultivar and maybeuseful in crop improvement program. Color variations in ray florets werenoted among cultivars. Pusa Aditya , Pusa Guldasta, Punjab shyamali andRiver city showed presence of secondary color in the inner side of the rayfloret. Large genetic variation for ray floret traits was previously observed(Lim et al., 2014).

4.2 Genetic variation, Heritability and Genetic advance

High values of PCV and GCV value indicates high variability and vice versa. Presence of high variability indicates effective selection for the character. Moderate to low variability indicates the need for improvement of base population (Chauhan et al., 2020). The results indicated that PCV are slightly greater than the GCV for all the traits, this mean that the trait sunder study were less influenced by environment. Similar PCV & GCV values for growth and floral characters were observed (Sarkar et al.,2005). High estimates for heritability and genetic advance as percentage of mean was recorded for number of flowers per plant, stipule size and ray floret width. In chrysanthemum, the high heritability values and genetic advance as per cent of mean for number of flowers per plant was also reported (Henny et al., 2021).

4.3 PCA Analysis

PCA is a valuable tool in multivariate analysis to reduce the dataset’s dimensionality without losing important information about the relationships between variables. PCA was done using a correlation matrix as it helps to ensure that the PCA results are robust, interpretable, and nonbiased by the original measurement units of the variables when dealing with variables measured on different scales. PCA for quantitative traits revealed that leaf characters have major contribution to germplasm variability. For qualitative characters, ray florets traits have major contribution to germplasm variability. A group researcher conducted PCA of 35 morphological characters in 15 taxa of Chrysanthemum species and identified 12 principle components explaining 99.4% of variation (Kim etal., 2014).

4.4 Cluster analysis based on Euclidean distance

Cophenetic correlation coefficient value equal to or greater than 0.85 isconsidered good ensuring the consistency of the dendrogram with thedistance matrices (Stuessy, 1990). Cluster 1 consist of only one genotypei.e. Gauri which bear white colour small sized double type flower, longestplant height of 92.95 cm, petiole length (5.15 cm), smallest flowerdiameter (2.7 cm) and highest number of flowers per plant (400). PusaShwet alone belongs to cluster 2 which has white color semi double flower(2-3 rows of ray floret), long leaf (8.15 cm), wide leaf (6.27 cm), highernumber of flower (231) and large flower diameter (7.75 cm). Cluster 3constitute of 28 genotypes and demonstrated comparatively less meanvalues for plant height (39.88 cm), number of branches per plant (11.62)and number of flowers per plant (54.76). Moderate mean value for flowerdiameter was (6.3 cm) and peduncle length (7.15 cm). With respect toqualitative characters, most genotypes in cluster 3 showed upward petioleattitude, flat cross section of ray floret and round shape of tip of ray floret.

Cluster 4 comprise of 20 genotypes having adequate number of primarybranches per plant with mean value 15.66, good number of flowers perplant with average value 139.72 and long peduncle with mean value 7.93.All the plants in cluster 4 are bushy type, majority of plants have downward petiole attitude and weakly convex cross section of ray floret. Fifteen taxa of Chrysanthemum species were classified into three groups through PCA and cluster analysis based on 35 qualitative and quantitative traits (Kim et al., 2014). Mean value of quantitative traits for different cluster is given in Table 6.

4.5 Cluster analysis based on jaccard’s similarity index

The range of Jaccard’s similarity coefficients (0.04 – 0.69) indicatedsignificant genetic diversity between chrysanthemum cultivars. In cluster1 all 5 genotypes are of non bushy types, large diameter double typeflower, large stipule size, long and wide ray floret and a less number offlowers. All except Pusa Arunodaya exhibited incurving type of ray floret,moderately involute rolling of margins of ray floret and concave crosssection of ray floret. Cluster 2 had single flower type, short plants, fewnumbers of flowers and large stipule size. Genetic relationships withinchrysanthemum were partly indicated by their ray floret type was alsoshown (Chen et al., 2013; Mia et al., 2007). Cluster 3 had only genotypeGauri which showed Jaccard’s similarity index of 0.38 or less with all thegenotypes showing exclusive higher value for plant height (92.95 cm) andnumber of flowers (400). Petiole length, flower diameter, ray floret lengthand ray floret longitudinal axis in Gauri varied from majority of thegenotypes. Cluster 4 showed higher number of primary branches, long andwide leaf, medium terminal lobe length of leaf and long lower lobe length.Clustering based on various leaf characteristics was also performed (Kimet al., 2014). Genotypes in cluster 5 exhibited short height, mediumnumber of primary branches, short leaf length, low leaf length to widthratio and medium stipule size. Punjab Shyamal , the only cultivar in cluster6 showed reflexing longitudinal axis of ray floret, 2 types of color on theinner side of ray floret, short ray floret along with long and narrow length.All the genotypes in cluster 7 showed long and narrow leaves with shortpetiole length and medium length of ray floret. In cluster 8 all cultivarsexhibited bushy plant type, straight longitudinal axis of ray floret, mediumleaf length to width ratio, medium width of ray floret and also allgenotypes had medium to large size flowers. Majority of cultivars incluster 9 have medium sized flowers, many primary branches, manynumber of flowers and flat margins of ray floret. The mean value ofquantitative traits for different cluster is given in Table 7.

5.CONCLUSION

The 50 genotypes of Chrysanthemum varied significantly for 25 Morphological characters. The PCVs were slightly higher than GCVs indicating correspondence between genotype and phenotype. PCA showed that among quantitative traits, leaf characters and ray florets characters among qualitative trait have major contribution to germplasm variability. Clustering of genotypes based on Euclidean distance and Jaccard’s similarity coefficient gave 4 and 9 clusters respectively. Gauri was an outlier for plant height and number of flowers per plants and thus was grouped alone in both type of clustering. The morphological identification of genotypes will enable efficient utilization of genetic resources, selection of parents for hybridization that will help in improvement of desired traits.

ACKNOWLEDGEMENT

I acknowledge the help and support provided by the Director, IARI to carry out the study.

FUNDING DETAILS

No separate funding was available for this study. The study was conducted with the available institute’s funds.

COMPETING INTEREST

No potential conflict of interest was reported by the author(s).

AUTHORS CONTRIBUTION

Conceptualization: Gunjeet Kumar, Vartika Budhlakoti; methodology: Gunjeet Kumar, Vartika Budhlakoti, A.K. Tiwari , V.M. Hiremath; data analysis: Gunjeet Kumar, Vartika Budhalkoti, V.M. Hiremath, Saipriya Panigrahi, Shreekant; writing: original draft preparation, Gunjeet Kumar, Vartika Budhlakoti, A.K. Tiwari, V.M. Hiremath; Review and editing: Gunjeet Kumar, Markandey Singh. All authors read and made suggestions which were incorporated in the final manuscript.

DATA AVAILABILITY

All data pertaining to this research is available in the manuscript or list of supplementary table.

REFERENCES

Allard, R.W., 1999. Principles of Plant Breeding. New York: John Wiley and Sons.
Azad, M.A.K., Biswas, B.K., Alam, N., and Alam, S.K.S., 2012. Genetic Diversity in Maize Zea mays L.. Inbred Lines. The Agriculturists, 10, (1), Pp. 64-70.

Baye, T., 2002. Genotypic and phenotypic variability in Vernonia galamensisgermplasm collected from eastern Ethiopia. The Journal of Agricultural Science, 139, Pp. 161.

Chauhan, S., Mishra, U., and Singh, A.K., 2020. Genetic variability, heretability and genetic advance studies for yield and yield related traits in pearlmillet [Pennisetum glaucum L.. R. Br.]. Journal of pharmacognosy and Phytochemistry, 93, Pp. 1199-1202

Chen, X., Sun, M., Liang, J., Xue, H., and Zhang, Q., 2013. Genetic Diversity of Species of Chrysanthemum and Related Genera and Groundcover Cultivars Assessed by Amplified Fragment Length Polymorphic Marker. HortScience horts, 485., Pp. 539-546. https://doi.org/10.21273/HORTSCI.48.5.539

Chung, N.C., Miasojedow, B., Startek, M., and Gambin, A. 2019. Jaccard/Tanimoto similarity test and estimation methods for biological presence-absence data. BMC Bioinformatics, 2420.: 644. doi: 10.1186/s12859-019- 3118-5. PMID: 31874610; PMCID: PMC6929325

Dai, S.L., Song, X.B., Deng, C.Y., Gao, K., Li, M.L., Ma, C.F., and Zhang, M.M., 2019. Comprehensive approach and molecular tools for breeding and production of ornamental crops. Acta Horticulturae, 1263, Pp. 1–16.

Deshmukh, S.N., Basu, M.S., and Reddy, P.S., 1986. Genetic variability, character association and path coefficients of quantitative traits in Virginia bunch varieties of groundnut. Indian J. Agric. Sci., 56, Pp. 816-821.

Dowrick, G.J., 1953. The chromosomes of Chrysanthemum, II: garden varieties. Heredity 7, Pp. 59-72. https://doi.org/10.1038/hdy.1953.5

Guo, Q.S., Wang, T., Cheng, L.T., Wen, J.J., Wang, T.Y., and Liang, Y.N., 2008. Study on quality offlavone in various cultivars of Chrysanthemum morifolium for medicine. Chin Med J., 337., Pp. 756–779. doi:10.17660/ActaHortic.2019.1263.1

Hair, J.R., Anderson, R.E., Tatham, R.L., and Black, W.C., 1995. Multivariate data analysis with readings. 4th edition, Prentice-Hall, Englewood Cliffs, NJ.

Hammer, O., Harper, D.A.T., and Ryan, P.D., 2001. PAST: Paleontological statistics software package for education and data analysis. Palaeontologia Electronica, 41, Pp. 9. http://palaeo-electronica.org/2001_1/past/issue1_01.htm.

Henny, T., Palai, S.K., and Chongloi, L., 2021. Assessment of genetic variability, heritability and genetic advance in spray chrysanthemum Chrysanthemum morifolium Ramat. Crop Research, 566, Pp. 336−340. doi: http://dx.doi.org/10.31830/2454-1761.2021.054

Jaime, A., Silva, T.D., Shinoyama, H., Aida, R., Matsushita, Y., Raj, S.K., and Chen, F., 2013. Chrysanthemum Biotechnology: Quo vadis?, CRC Crit. Rev. Plant Sci., 32, Pp. 21-52. https://doi.org/10.1080/07352689.2012.696461

Johanson, N., Robinson, H., and Comstok, R., 1955. Estimates of genetic and environmental variability in soybean. Agron. J., 47, Pp. 314–318.

Kachare, S., Tiwari, S., Tripathi, N., and Thakur, V.V., 2016. Assessment of Genetic Diversity of Soybean Glycine max. Genotypes Using Qualitative Traits and Microsatellite Markers. Agric Res., 91., Pp. 23–34.https://doi.org/10.1007/s40003-019-00412-y

Khatun, R., Uddin, M.I., Uddin, M.M., Howlader, M.T.H., and Haque, M.S., 2022. Analysis of qualitative and quantitative morphological traits related to yield in country bean Lablab purpureus L. sweet. genotypes. Heliyon, Pp. 812. https://doi.org/10.1016/j.heliyon.2022.e11631

Kim, S.J., Lee, C.H., Kim, J., and Kim, K.S., 2014. Phylogenetic analysis of Korean native Chrysanthemum species based on morphological characteristics. Scientia Horticulturae, 175, Pp. 278-289. https://doi.org/10.1016/j.scienta.2014.06.018

Lim, J.H., Shim, M.S., Sim, S.C., Oh, K.H., and Seo, J.Y., 2014. Genetic variation of flower characteristics in a population derived from a cross between the chrysanthemum cultivars ‘Falcao’ and ‘Frill Green’. Horticulture, Environment and Biotechnology, 55, Pp. 322–328. https://doi.org/10.1007/s13580-014-0140-4

MacDonald, J., Hackett, M., and Mirmak, B., 2017. Handbook on chrysanthemum classifcation. National Chrysanthemum Society, USA. https:// mums. org/ product/ classification hand book

Miao, H., Chen, F., and Zhao, H., 2007. Genetic relationship of 85 Chrysanthemum [Dendranthema grandiflora Ramat. Kiramura] cultivars revealed by ISSR analysis. Acta Hort.Sin., 34, Pp. 1243–1248. https://www.ahs.ac.cn/EN/

Mohammadi, S.A., and Prasanna, B.M., 2003. Analysis of genetic diversity in crop plants-salient statistical tools and considerations. Crop Science, 43, Pp. 1235-1248. https://doi.org/10.2135/cropsci2003.1235

Panse, V.G., and Sukhatme, P.V., 1967. Statistical Methods for Agricultural Workers. New Delhi: Indian Council of Agricultural Research, Pp. 381.

Roxas, N.J., Tashiro, Y., Miyazaki, S., Isshiki, S., and Takeshita, A., 1995. Meiosis and pollen fertility in Higo chrysanthemum Dendranthema× grandiflorum Ramat. Kitam. J. Jpn Soc. Hortic. Sci., 64, Pp. 161–168. Doi: https://doi.org/10.2503/jjshs.64.161

Sarkar, I., Ghimiray, T.S., and Roy, A., 2005. Evaluation of chrysanthemum varieties under naturally ventilated low cost polyhouse and open field condition. Stress Biology, Pp. 181.

Shannon, C.E., and Weaver, W., 1949. The Mathematical Theory of Communication. Urbana: University of Illinois Press.

Suvija, N.V., Suresh, J., Kumar, R.S., and Kannan, M., 2016. Evaluation of chrysanthemum Chrysanthemum morifolium Ramat.. genotypes for loose flower, cut flower and pot mums. Int. j. innov. Res. Adv. Stud., 34., Pp. 100-104

Wang, Y., Jung, J.A., Kim W.H., Lim, K.B., Hwang, Y.J., 2021. Morphological and rDNA fluorescence in situ hybridization analyses of Chrysanthemum cultivars from Korea. Hortic Environ Biotechnol., 626, Pp. 917–925. Doi: https://doi.org/10.1007/s13580-021-00361-y

Zhen, L.P., Yang, J., and Yu, N.J., 2013. Study on the characteristics of the lower leaf surface of wild chrysanthemum plants in Anhui. Chinese Journal of Plant Science, 31, Pp. 99–106.