Spatial variability of soil biological and chemical indicators in selected soils in South Africa

Abstract

Understanding the spatial variability of selected soil fertility indicators could make it possible to optimize the use of agricultural inputs with the reduction of economic and environmental risks. The aim of this study was to determine the degree of spatial variability of soil fertility indicators in selected soils in South Africa. There is a paucity of information on the spatial variability of soil fertility indicators in many part of South Africa. Soil samples were collected from two sites viz. University of Venda and Roodeplaat Experimental Farms. The fields were demarcated into 20 m × 20 m grid cells in approximately1 ha using a measuring tape. Hence, the field consisted of 12 grids. Therefore, soil samples were collected from 0 - 0.2 m depth at both sites and the coordinates of each sampling points were recorded with GPS. Then, soil respiration, cellulolytic, catalase, urease, acid phosphatase, organic carbon and total nitrogen, cation exchange capacity, soil pH and soil texture were determined using standard methods. Descriptive and geostatistical analyses were performed using ArcMap® version 9.0. The results showed that most parameters were analysed by exponential model except for acid phosphatase and organic carbon that fitted into Gaussian models in University of Venda Experimental Farm. Spatial dependence of the soil respiration, cellulases and organic carbon have strong spatial dependence with nugget to sill ratios of less than 25% in the field of University of Venda Experimental Farm. Thus, catalase, acid phosphatase and organic phosphorus exhibited moderate spatial dependence with nugget to sill ratios between 25 and 75%. All parameters were analysed by exponential model except cellulases in Roodeplaat Experimental Farm. Hence, all parameters exhibited strong spatial dependence (nugget/sill ratio ˂ 25%) except acid phosphatase and cellulases that were exhibited moderate (nugget/sill ratio 25 and 75%) and weak (nugget/sill ratio >75%) spatial dependence, respectively. The measured parameters were spatial dependent at Roodeplaat Experimental Farm more than in University of Venda Experimental Farm. Hence, spatial dependence of measured parameters at Roodeplaat Experimental Farm mainly controlled by extrinsic factors than intrinsic factors. The study showed that geostatistics is a useful tool to map spatial variabilities of soil fertility indicators under arable lands. Heterogeneity and variation of soil fertility indicators in a field due to intrinsic and extrinsic factors should be taken into consideration for a successful agricultural management.