A MULTIVARIATE APPROACH TO ASSESSING VARIABILITY IN SOME SOILS OF THE NIGERIAN SAVANNA

Authors

  • R. I. SOLOMON
  • J. H. ABDULKAREEM
  • D. MAMZING

Keywords:

Principal component analysis (PCA); Nigerian Savanna; soil properties; variability

Abstract

Multivariate analysis is a vital tool for investigative data analysis as it permits grouping of samples based on similarity and at the same time allows the selection of the most important variables that differentiates them. The study was conducted with a multivariate approach to determine the soil properties that account for soil variability in selected soils of the Nigerian Savanna. To achieve this, soil samples were collected from twelve (12) different locations where they were analyzed for various soil physical and chemical properties using standard laboratory procedures. Twenty-one soil properties were subjected to principal component analysis (PCA). Results revealed that only six principal components (PCs) were found to have eigenvalue of >1 out of the twenty-one. Principal component (PC) 1 with an eigenvalue of 6.35 was the most influential and accounted for 28.86 % of the cumulative variance. Similarly, PC 2, 3, 4, 5 and 6 have eigenvalues of 5.17, 2.55, 1.41 and 1.29 respectively and together with PC 1 accounted for a cumulative variation of 83.55%. Clay content, field capacity water (FCW), soil porosity, total nitrogen, available nitrogen, nitrate, exchangeable bases, ECEC, EC, pH in CaCl2, H + Al, organic carbon and sulphate were the main soil properties influencing soil variability in this study and are related to water and nutrients retention as well as soil salinity and reaction. Therefore, management of these properties through the incorporation of organic residues can enhance in minimizing their variability.

Published

2022-04-27