The potential role of quantitative traits for understanding the genetic diversity in Sorghum (Sorghum bicolor L.)

185 Ullah et al. Int. J. Biosci. 2020 RESEARCH PAPER OPEN ACCESS The potential role of quantitative traits for understanding the genetic diversity in Sorghum ( Sorghum bicolor L.) Muhammad Ihsan Ullah1*, Zaib-un-Nisa1,2, Barkat Ali1, Guljana Nazir1, Muhammad Zeeshan Munir2, Muhammad Amir Siddique3, Arif Hussain4, Dilawar Khan5, Muhammad Atif Muneer6 1Sorghum Research Sub-Station, Dera Ghazi Khan, Punjab, Pakistan 2School of Biological Sciences and Technology, Beijing Forestry University, Beijing 100083, China School of Landscape Architecture and Design, Beijing Forestry University, Beijing 100083, China 4Ghazi University, Dera Ghazi Khan, 32200, Pakistan 5School of Soil and Water Conservation, Beijing Forestry University, Beijing 100083, China 6School of Grassland Science, Beijing Forestry University, Beijing 100083, China Key words: Cluster analysis, Dendrogram, PCA, Genetic diversity, Morphological traits, Sorghum. http://dx.doi.org/10.12692/ijb/16.6.185-195 Article published on June 29, 2020 Abstract Sorghum is a popular cereal crop worldwide, and therefore, the understanding and utilization of sorghum accessions play role crucial for improving crop productivity. The in-depth knowledge of genetic variation between the sorghum accessions will allow the plant breeder for precise breeding. Thus, exploration of sorghum crop genetic diversity is inevitable. In this study, we investigated the genetic diversity among the thirty sorghum accessions by using the quantitative morphological traits. The findings were concluded on the basis of ten quantitative traits, among which 07 diverse quantitative traits which maximized the variation, chosen carefully for genetic diversity analysis. We found a significant positive correlation of stalk yield with days to flowering and no. of leaves per plant. Similarly, for grain yield with no. of leaves per plant and panicle width. The PCA revealed that days to flowering, plant height, panicle width, and leaf area contributed maximum towards the divergence. Sorghum accessions were grouped under 4 clusters by using hierarchical analysis. Cluster-2 contained minimum while cluster4 had maximum number of sorghum accessions. The maximum inter-cluster distance was recorded between cluster-2 and cluster-4. The cluster-4 had the highest mean values for no. of leaves per plant, stalk yield, and days to maturity. Therefore, parent selection must be dependent on best yield related components and wider inter-cluster distance. Therefore, in current study, the quantitative morphological traits showed the broader genetic diversity range among the thirty sorghum accessions, and can be utilized effectively for the improvement of genetic architecture. * Corresponding Author: Muhammad Ihsan Ullah  ih[email protected]om International Journal of Biosciences | IJB | ISSN: 2220-6655 (Print), 2222-5234 (Online) http://www.innspub.net Vol. 16, No. 6, p. 185-195, 2020 186 Ullah et al. Int. J. Biosci. 2020 Introduction Sorghum (Sorghum bicolor L.) is a C4 plant with dynamic growth patterns, and higher biomass production is widely grown cereal crop in the world (Hariprasanna and Patil, 2015; Ullah et al., 2016). It is the world's most important cereal crop and ranks at the fifth position after wheat, maize, rice, and barely (Sinha and Kumaravadivel, 2016). It is extensively cultivated as a major food crop in the regions of South Asia and Africa (Bibi et al., 2010). It is commonly grown in tropical and subtropical areas of the world. It can also be cultivated in the changing agro-climatic conditions, e.g., rainfall, temperature, and soil conditions (Reddy et al., 2008). Therefore, besides the importance of sorghum crop for food, forage, and feed purposes, it also supplies the raw material for fiber processing, alcohol, biofuels, starch, , and many other byproducts (Mumtaz et al., 2018; Assar et al., 2020). In Pakistan, Sorghum is commonly called Jowar, and cultivated for the purpose of fodder and grain. The cultivated area was 274 thousand hectares with annual yield about 161 thousand tons in the year 2015-2016 (Wing, 2016). The shortage of the forage crops in the Pakistan is mostly two times a year, i.e., MayJune, and during October-November (Iqbal et al., 2010). Therefore, improvement of sorghum accessions for the quality and yield traits can do a great deal to overcome the scarcity of forage production during the summer season. Sorghum is known as the fast-growing crop that constitutes a considerable proportion of stalk yield as well as grain yield. It possesses carbohydrates about 70%, fat and protein content 0.3%. Therefore, it can be successfully employed in the feeding programs for dairy cattle and poultry (Khan et al., 2007; Ullah et al., 2016). Therefore, it is imperative to explore the genetic diversity in sorghum accessions and should be utilized in the breeding programs. Genetic diversity is explained as classification or grouping of population or an individual in contrast to the other populations or individuals (Adugna, 2014; Raza et al., 2019). The analysis of genetic diversity is an essential technique for the development of new varieties or cultivars on the basis of genetic distance and similarities. Sorghum was native to the African continent(specifically in the Ethiopia) and then it was introduced globally to the different regions with diverse agro-climatic conditions (Li et al., 2010). Therefore, a wider range of diversity has been reported at both phenotypic and genotypic levels among and within sorghum cultivars (Kong et al., 2000; Hart et al., 2001). The awareness of a crop's genetic diversity helps the breeder in the breeding programs to choose the suitable parents, and the introgression of genes from distant genetic material produces the superior hybrids having great potential against the biotic and abiotic stresses. By knowing genetic diversity richness in the sorghum crop may allow the improvement of genetic architecture of this crop (Jayaramachandran et al., 2011). Genetic variability in the sorghum accessions is the gift of nature, and It is caused by spatial separation or genetic cross-compatibility barriers. The morphological characteristics are used as traditional methods in the breeding programs to research genetic variation. In general, the morphological assays need neither the preparatory procedure nor the specialized equipment. These are usually inexpensive and simple to record the data. These quickly identified morphological characteristics are valuable tools for a preliminary assessment, since they provide a practical and straightforward methodology for determining the level of diversity among the cultivars. Over the years, a variety of studies have used the morphological characteristics to predict the genetic variation in the cultivated sorghum (Zongo et al., 1993; Adugna, Rao et al., 1996; Ayana and Bekele, 1998; Dahlberg et al., 2002; Shehzad et al., 2009; Adugna, 2014) . The most popular method that is used to estimate the associations between the genotypes is the use of morphological traits. The genetic diversity of cultivated varieties/species along with their wild relatives provides a strong base for breeding 187 Ullah et al. Int. J. Biosci. 2020 improved and new varieties of crops. A better knowledge of genetic variation in sorghum can help us to boost up the crop yield and quality. Therefore, it is of the prime importance to assess the genetic diversity in the available accessions of Sorghum. In this study, we have tried to determine the genetic diversity/variation between the thirty accessions of sorghum crops by analyzing the quantitative traits. Materials and methods Plant material Thirty accessions of sorghum including one check variety were received from Maize and Millet Research Institute (MMRI), Yusafwala, Sahiwal, Pakistan. These accessions were planted at Sorghum Research Sub-Station, D. G. Khan, Pakistan, during Kharif2018 (Fig. 1). Fig. 1. The geographical location of Sorghum Research-Substation, D.G. Khan. The real image taken by the google earth pro is showing the original site of field trials (A), and sorghum accessions cultivated in this L-shaped field highlighted by yellow color. These thirty accessions were named as Yusafwala Sorghum Selection (YSS), i.e., YSS-1 (check variety), and YSS-2 to YSS-30 (29 lines) because these were collected form MMRI. Methodology The thirty sorghum accessions were raised in a randomized complete block design (RCBD) with three replications during Kharif-2018 at Sorghum Research Sub-Station, Dera Ghazi Khan. The plot size was consisted of two rows of 5m length by the spacing of 75cm × 15cm (row-to-row, and plant-to-plant, respectively), for each accession. The agronomic practices like fertilizer doses, crop protective measures, and irrigation etc., were ensured as and when required during the crop season. Five plants were selected randomly from each replicate, and the observations for quantitative traits were noted at maturity (except for days to flowering). Data were recorded for various quantitative traits like the 188 Ullah et al. Int. J. Biosci. 2020 number of days to flowering (DFL), plant height (PHT), days taken to maturity (DMA), number of leaves per plant (NPL), Leaf area (LFA), panicle length (PNL), panicle width (PWD), “000” grain weight (TWT), grain yield (GYI), and stalk yield (SYI) tons/ha. For statistical analysis, the mean values were used to determine the genetic diversity of the thirty sorghum accessions. Statistical analyses The experiment was performed by following the CRBD (Complete Randomized Block design). The data were analyzed by using the analysis of variance and descriptive statistics with the help of IBM SPSS 25.0 software package, while the person correlation analysis was performed in R programming software. The factor analysis was conducted to determine what trait contributes to the highest variability. The PCA analysis was performed to analyze the contribution of each quantitative traits to the overall genetic variation. The hierarchical clustering was conducted on “Euclidean distance matric” using the Ward’s linkage method. We performed these analyses by the MINITAB software V.18. Results and discussion Variance analysis using a randomized block method revealed a substantial variance for all the traits (results not shown here), showing the presence of a high degree of genetic variability among the 30 accessions. Table 1. Descriptive statistics of the quantitative traits. DFL DMA PHT NPL LFA PNL PWD TWT GYI SYI Mean 80.08 126.84 243.87 14.44 2.19 25.71 14.33 27.79 515.63 19.63 Standard error 0.49 0.60 5.16 0.16 0.09 0.48 0.19 0.46 10.36 0.73 standard deviation 4.65 5.66 48.98 1.48 0.81 4.52 1.76 4.39 98.28 6.90 Sample variance 21.67 32.02 2399.15 2.20 0.66 20.44 3.11 19.31 9659.18 47.58 Range 20.00 31.00 195.00 6.00 3.06 20.25 7.41 21.00 498.42 27.00 Descriptive statistics The grain yield provided the highest value of mean (515.63), standard deviation (98.28), standard error (10.36), variance (9659.18), and the range (498.42) (Table 1). The descriptive statistics data of ten quantitative traits confirmed the presence of morphological diversity amongst these thirty sorghum accessions, providing the potential for crop improvement via selection and hybridization. The variation coefficients for plant height, and stalk yield seemed to be high, revealing the vulnerability to environmental changes that influence the expression patterns to some extent. Table 2. Rotated factor loadings of the quantitative traits. Traits Factor-1 Factor-2 Factor-3 Factor-4 Days to flowering 0.833 -0.189 -0.172 0.180 No. of leaves per plant 0.812 -0.152 0.232 0.274 Stalk yield 0.762 -0.033 0.048 -0.336 Days to maturity 0.520 0.147 0.230 0.093 Plant height 0.240 0.827 -0.233 0.033 Panicle length 0.204 -0.797 -0.223 0.175 Thousand-seed weight -0.343 0.499 -0.168 0.218 Panicle width 0.118 0.028 0.922 -0.168 Grain yield 0.147 -0.327 0.688 0.482 Leaf area 0.101 0.031 -0.034 0.915 Total variance 2.468 1.758 1.596 1.407 Variance % 24.681 17.581 15.965 14.067 Cumulative variance % 24.681 42.262 58.226 72.294 189 Ullah et al. Int. J. Biosci. 2020 Correlation analysis We found a significant positive correlation for stalk yield (SYI) with days to flowering (DFL) and no. of leaves per plant (NPL) (Fig.2). The grain yield (GYI) was also significantly and positively correlated with NPL, and panicle width (PWD) that could play a significant role in food synthesizing by the plants through early flowering and the process of photosynthesis that have direct effects on stalk yield as well as the grain yield. From these investigations, we concluded that DFL and NPL are associated with SYI, while NPL and PWD are correlated with and GYI. Thus, the selection of these quantitative traits will be of great importance and have a significant impact on stalk yield and grain yield. Table 3. Principal component analysis showing the contribution of 8 traits among sorghum accessions. Traits PC1 PC2 PC3 Days to flowering 0.514 0.131 0.265 No. of leaves per plant 0.526 0.123 0.023 Stalk yield 0.421 0.206 0.433 Days to maturity 0.298 0.253 -0.681 Plant height -0.065 0.703 -0.055 Panicle length 0.298 -0.540 0.105 Thousand-seed weight -0.316 0.275 0.513 Eigenvalue 2.593 1.517 1.019 % Variance 37.05 21.67 14.56 % Cumulative variance 37.05 58.72 73.27 Factor analysis The factor analysis was conducted to lessen the large morphological traits to a lesser but more illustrative set of attributes, and we may interpret which quantitative traits have significant contribution in the maximum variation (Table 2). The first three factors contributed to72.294% of the total variance that we recorded (Table 2). The factor-1 had the highest factor loadings from days to flowering, no. of leaves/plant, stalk yield, and days to maturity contributed 24.681% of total variance. The factor-2 had the highest loading role from the plant height, the panicle length, and 1000-seed weight attributed 42.262% of total variance. Similarly, factor-3 had the high panicle width and grain yield contribution loadings. In the loading plot, the distribution of quantitative traits in the first two factors has been shown (Fig. 3). The loading plot made clear explanation of the quantitative traits, i.e., panicle width (PWD), leaf area (LFA), and grain yield (GYI), contributed low variability towards the genetic variation. Table 4. The inter-cluster distances among the sorghum accessions. Cluster-1 Cluster-2 Cluster-3 Cluster-4 Cluster1 0.00 Cluster2 109.90 0.00 Cluster3 49.99 61.42 0.00 Cluster4 44.77 154.24 94.07 0.00 PCA (Principal component analysis) The principal component analysis (PCA) is interpreted according to the correlations between variables; either it is positive or negative values. Of the ten quantitative traits, a set of seven different quantitative morphological traits was selected, i.e., days to flowering, no. of leaves per plant, stalk yield, days to maturity, plant height, panicle length, and thousand-seed weight used to group the thirty sorghum accessions by PCA analysis. About the contribution made by first three factors was 73.27% of total variance (Table 3). 190 Ullah et al. Int. J. Biosci. 2020 Table 5. Characteristic means of the four similarity cluster groups of the sorghum accessions. Traits Cluster-1 Cluster-2 Cluster-3 Cluster-4 Days to flowering 78.13 78.67 82.11 81.46 No. of leaves per plant 14.54 13.61 13.78 14.92 Stalk yield 18.40 13.94 18.96 23.15 Days to maturity 127.54 122.56 124.67 128.9 Plant height 282.71 173.17 327.11 233.39 Panicle length 23.77 28.89 21.92 26.32 Thousand-seed weight 30.63 28.00 28.56 25.77 The first principal component (PC1) accounting for 37.05% of total variance., and showed maximum contribution of loading factor from the days to flowering, no. of leaves per plant, and stalk yield. The second principal component (PC2) showed the highest contribution of loading factor from the plant height and panicle length, and made contribution to 21.67% of total variation. The third principal component (PC3) showed the highest of factor loading from days to maturity, thousand seed weight, and contributed to 14.56% of the total variation. Fig. 4 shows the score plot of 30 accessions based on the first two main components. The distributed patterns also showed the presence of a significant amount of variability among the sorghum accessions. Fig. 2. Pearson correlation coefficients of quantitative traits. These results indicated that PCA analysis could be instrumental in the evaluation of sorghum accessions for crop production. Makanda et al. (2012), Sinha and Kumaravadivel (2016) also found significant differences among the different quantitative morphological traits in the sorghum crop. Shegro et al. (2013) found the maximum contribution of the first two principal components in the sorghum 191 Ullah et al. Int. J. Biosci. 2020 accessions. Our results are in line with previous findings (Sinha and Kumaravadivel, 2016),who also found that the first three PCs contributed the maximum to the genetic diversity among the sorghum accessions. These previous findings also assessed genetic diversity by the contribution of different morphological quantitative traits. Our results are also consistent with the findings of Mumtaz et al. (2018), they also investigated the genetic diversity among the sorghum accessions. Fig. 3. The loading plot of quantitative traits based on the factor analysis. Fig. 4. The distribution of sorghum accessions for the first two principal components based on quantitative traits. 192 Ullah et al. Int. J. Biosci. 2020 Cluster analysis The hierarchical clustering study was performed using Ward's linkage method based on the Euclidean distance matrix, and the resulting dendrogram is shown in Figure 5. Thirty sorghum accessions comprised four clusters (Fig.5). The size of clusters varied from three to thirteen of the various groups of clusters. The cluster-1 contained 08 accessions, cluster-2 contained the minimum number of accessions, i.e., 03 accessions, cluster-3 comprised of 06 accessions, while, cluster-4 having the maximum number of accessions about 13. The clustering patterns showed the significant amount of variability in the sorghum accessions. Fig. 5. The dendrogram of sorghum accessions based on seven quantitative traits. The highest inter-cluster distance was recorded between the cluster-2 and cluster-4 (154.24), it concluded that if the sorghum accessions are selected from these cluster for the hybridization program, may give the broad spectrum of the variability in the segregating generation (Table 4). While, the lowest inter-cluster was noted between the cluster-1 and cluster-4 (44.77). It has been cited in the previous studies that the clusters contributing the maximum divergence to the variability given the greater emphasis for the type of cluster for the purpose of the selection process, and used as parents in the hybridization program (Rohman et al., 2004). Cluster-1 had the highest mean value for the thousand-seed weight (30.63). Cluster-2 showed the highest mean value for the trait panicle length (28.89). Cluster-3 showed the highest mean values 82.11, and 327.11 for days to flowering, and plant height, respectively. Cluster-4 showed the highest mean values no. of leaves per plant (14.92), stalk yield (23.15), and days to maturity (128.90) (Table 5). Based on mean values, we found the most crucial cluster is the cluster-4 because of having the highest mean values for no. of leaves per plant, stalk yield, and days to maturity. Thus, the sorghum accessions falling this cluster could be used in the hybridization program as parents. (Shegro et al., 2013) found the five clusters, while (Kisua et al., 2015) observed the three clusters for sorghum accessions on the basis of quantitative traits. Similarly, (Sinha and 193 Ullah et al. Int. J. Biosci. 2020 Kumaravadivel, 2016) performed the cluster analysis on forty sorghum accession and found the clusters, but in our study, we found the four clusters for thirty accessions based on quantitative characteristics. Our findings are similar to the results of (Sinha and Kumaravadivel, 2016), because we also performed cluster analysis by using sorghum accessions with their quantitative traits, but there was difference in number of clusters because of difference in number of accessions, and we also found the higher genetic diversity. Conclusion These findings suggested that quantitative traits play a crucial role, as well as a useful tool for preliminary assessment of genetic diversity among the various sorghum accessions. The analysis of correlations reveals that the traits, namely, no. of leaves/plant and panicle width, had positive significant correlation with the grain yield. At the same time, stalk yield had also positive significant correlation with no. of leaves/plant, and days to flowering. The principal component analysis grouped the sorghum accessions into three factors that contributed about 73.27% of the total variance, and the quantitative have great potential in producing the genetic diversity, and selection of sorghum accessions. From cluster analysis, the thirty accessions grouped into four clusters of genetic variability based on selected seven quantitative traits. Thus, the selection of the parents must be on the base of different clusters. Based on the quantitative traits of sorghum, it is concluded that accessions YSS-13 and YSS-20 should be used in future breeding programs for the development of superior verities. Acknowledgments The authors are thankful to the Maize and Millet Research Institute, Yusafwala, Sahiwal Pakistan, for providing the experimental material of sorghum accessions. We are grateful to the field staff of the Sorghum Research Sub-Station, D. G. Khan, for their assistance in conducting the field trials. Conflict of interest The authors declare that they have no conflict of interest. References Adugna A. 2014. Analysis of in situ diversity and population structure in Ethiopian cultivated Sorghum bicolor (L.) landraces using phenotypic traits and SSR markers. Springer Plus 3(1), 212. http://dx.doi.org/10.1186/2193-1801-3-212 Assar AHA, Uptmoor R, Abdelmula AA, Wagner C, Salih M, Ali AM, Ordon F, Friedt W. 2020. Assessment of sorghum genetic resources for genetic diversity and drought tolerance using molecular markers and agro-morphological traits. University of Khartoum Journal of Agricultural Sciences 17(1). Ayana A, Bekele E. 1998. Geographical patterns of morphological variation in sorghum (Sorghum bicolor (L.) Moench) germplasm from Ethiopia and Eritrea: qualitative characters. Hereditas 129, 195205. https://doi.org/10.1111/j.1601-5223.1998.t01-100195.x Bibi A, Sadaqat HA, Akram HM, Mohammed MI. 2010. Physiological markers for screening sorghum (Sorghum bicolor) germplasm under water stress condition. International Journal of Agriculture and Biology 12, 451-455. Dahlberg J, Zhang X, Hart G, Mullet J. 2002. Comparative assessment of variation among sorghum germplasm accessions using seed morphology and RAPD measurements. Crop Science 42, 291-296. https://doi.org/10.2135/cropsci2002.0291 Hariprasanna K, Patil J. 2015. Sorghum: origin, classification, biology and improvement, Sorghum molecular breeding. Springer, p 3-20. https://doi.org/10.1007/978-81-322-2422-8_1 Hart G, Schertz K, Peng Y, Syed N. 2001.