Spatial prediction of soil nutrients from soil, topography and environmental attributes in the northern part of Ekiti State, Nigeria

Abstract

Knowledge of soil nutrient's spatial distribution and variability is useful for determining the balanced use of fertilizer inputs and for decisions making on the acquisition of land for any activity that will have its base on the lithosphere portion of the earth. This study presents statistical model for the prediction of top soil nutrients (Soil Organic Matter (SOM), Potassium (K), Phosphorus (P) and Nitrogen (N)) from soil and topographic attributes in Ekiti north, Ekiti state, Nigeria using geospatial techniques. Three topographic attributes (elevation, slope and curvature), one environmental variable (Land surface temperature, LST), one vegetation attribute (Normalized difference vegetation index (NDVI)) and four soil attributes (Water holding capacity, and bulk density) were employed to predict the spatial distribution of soil nutrient (N, P, K, SOM) in the study area. 66 samples of top soil nutrient of were collected within the study area. Topographic attributes were extracted from the obtained Digital Elevation Model (DEM) and Land use data while the soil attributes were derived from existing methodology by soil characteristics and Land use data. The derived attributes were statistically correlated with the sampled soil nutrients (N, P, K, SOM) to establish their relationships. Using Multiple Linear Regression Analysis, the contributions of individual factors in the prediction of soil nutrient (N, P, K, SOM) were established and soil nutrient model developed. The model was used to spatially predict the distribution of top soil nutrient in the study area. Results show both negative and positives correlation between the attributes and the soil nutrient (N, P, K, SOM). Spatially, the predicted top soil nutrient compared considerably well with the observed top soil nutrients.