The Usefulness of Multivariate Analysis in Forestry Research: A Case Study of Wild Pomegranate (Punica granatum L)

The present study deals with the Contrast, Cluster, Discriminant and Principal Component Analysis. The data for the present study was collected. Data on different morphological and seedling characters, namely Tree height (m), Tree diameter (cm), Crown spread E-W (m), Crown spread N-S (m), Fruit weight (g), Leaf length(cm), Internodal length(cm), Collar diameter (mm), Number of branches per plant and leaf petiole (cm) were considered from five different districts comprising of ten seed sources. Two seed sources were considered from each district viz; Narag and Neripul from Sirmour district, Waknaghat and Sadhupul from Solan district, Sundernagar and Rewalsar from Mandi district, Mohal and Banjar from Kullu district, Basantpur and Sunni from Shimla district. Different morphological and seedling characteristics of ten seed sources were evaluated by contrast analysis, which varied significantly among the different districts and within the districts. It is concluded that Tree diameter (cm), Crown spread E-W (m) and Crown spread N-S (m)was found Maximum in Solan district and Minimum in Shimla district, whereas Collar diameter (mm), Leaf length (cm) and Number of branches per plant was found Maximum in Mandi district and Minimum in Shimla district. Tree height (m) and Leaf petiole was recorded maximum in Kullu and Sirmour district and minimum in Shimla district whereas Fruit weight (g) and internodal length (cm) was found Maximum in Mandi district and minimum in Solan and Sirmour district.Cluster analysis was performed and the seed sources were grouped into three clusters. Discriminant analysis was carried out to categorize the seed sources into high and low yielders. Tree diameter, Tree height and Crown spread E-W are the most important characters that discriminate the two groups. Six seed sources from Mandi, Kullu and Shimla district were high yielder whereas other four seed sources were low yielder. Three principal components (PCS) were extracted out of ten which explained 34.675, 23.002, and 11.587 per cent of the total variation respectively amounting to 69.26 per cent of total variation.