Statistical Techniques to Understand Soils

Thirteen soils were identified in 518 ha of the farm area of the Indian Grassland and Fodder Research Institute. They were also described in terms of effective soil volume (ESV), occupied by fine earth and pore space, because of the spatial variability in the landscape attributes, which are found to control ESV. Two important properties, which determine ESV, depth and coarse fragment contents varied from 0. 14 to 1. 65 m and 0 to 55%, respectively. Sand, silt and clay contents varied from 37 to 72, 16 to 43 and 9 to 32%, respectively. Multivariate statistical tool like factor analysis extracted two factors namely, ‘surface area factor’ and ‘exploitable soil volume factor’, which jointly described 83% of variability. Sand (−0. 825), silt (0. 830) and clay (0. 825) had larger loading on surface area factor (SA factor) (−0. 825, 0. 830 and 0. 825, respectively) and described 51% of variability in the data. Similarly, ESV had the largest loading (0. 946) on exploitable soil volume factor (ESV factor), which described 32% variability of the data. Scatter pattern based on factor scores, which indicated the nature of soils, placed three soils in Group I with positive scores on both the factors while two soils were put in Group II with positive and negative scores on SA and ESV factor, respectively. Three soils were classified in Group III, characterized by negative scores on both the factors, while five soils were placed in Group IV with negative and positive scores on factor 1 and 2, respectively.