Multivariate Analysis for Soil Fertility Characterization of Harve Sub-watershed, Chamarajanagar District, Karnataka Sathish A.*, Kumar Vinay, Parama Ramakrishna Department of Soil Science and Agricultural Chemistry, University of Agricultural Sciences, Gandhi Krishi Vigyan Kendra, Bangalore, 560065, Karnataka *Corresponding author Email: soilsathish@gmail.com
Online published on 21 August, 2020. Abstract Multivariate statistical methods such as principal component analysis (PCA) and cluster analysis (CA), coupled with correlation coefficient analysis, were used to analyze various soil parameters in Harve subwatershed, Chamarajanagar District, Karnataka. A total of 844 soil samples from a depth of 0–15 cm at an interval of 250 m were collected from the sub-watershed. Soil properties showed large variability with greatest variation in DTPA-Fe (80%), whereas the smallest variation was in pH and EC (9.49%). The PCA applied on the investigated soil properties identified three components with eigen values greater than 1, which explained 58% variability and same grouping was also obtained from CA. Cluster 1 has highest positive loadings for K2O, Ca and S and moderate positive loadings for Mg and Zn in PC-1, and also showed strong significant relationship. In cluster 2, pH and Mg showed highest positive loadings and EC, N and Zn showed moderated positive loadings whereas Fe and Mn showed negative loadings in PC-2 and also significant positive relationship with each other. Cluster 3 contains N and P2O5 which has highest positive loadings and DTPA-Zn, DTPA-Mn and DTPA-Fe has moderate positive loadings. Correlation study showed that available N had highly significant positive correlation with available P. Top Keywords Correlation, cluster analysis, multivariate analysis, principal component analysis, soil fertility. Top |