MULTIVARIATE ANALYSIS ON RESIN PRODUCTION FROM <em>PINUS</em> <em>ROXBURGHII</em> SARGENT: A CASE STUDY OF HIMACHAL PRADESH

Abstract

Multivariate technique is extremely helpful in assisting researchers in making sense of large, complicated and complex datasets that contain many variables measured on a large number of experimental units. Data for the present study was collected for different morphological characters, namely tree diameter, tree height, bole height, number of branches and age of tree from three major resin-producing districts of Himachal Pradesh, viz. Bilaspur, Hamirpur and Sirmour. The primary data on 360 trees of Pinus roxburghii were collected through well planned survey from six ranges randomly selected through a multistage random sampling technique. Two forest ranges were considered from each districts, viz. Swarghat and Bhradi forest ranges from Bilaspur district, Aghar and Hamirpur forest ranges from Hamirpur district, and Rajgarh and Narag forest ranges from Sirmour districts. In the present study, two independent methodologies, viz. discriminant analysis and principal component analysis were discussed. Although resin production may depend on a variety of characteristics, researchers are always interested in identifying those sets of attributes that most significantly affect resin production of the tree. Results from the principal component analysis revealed that based on the average of all forest ranges selected for the sampling procedure, tree diameter played a main role in resin production with an eigenvalue of 2.752, explaining 55.38% of the variation. Results from the discriminant analysis revealed that Rajgarh, Hamirpur and Swarghat forest ranges were found to be high yielders whereas Bhradi, Narag and Aghar forest ranges were found to be low yielders.