Farm Level Soil Properties Variability in an Agriculture Landscape of Southwestern Nigeria

Abstract Spatial variability of soil properties is a major limitation to precision nutrient management in smallholder farming in Nigeria, where fertilizer is applied uniformly without considering spatial variability within the field. This research examined such variability in a 12.5-ha farmland in the Olowogbo village of Akinyele in Ibadan and using reliable geostatistical fertility maps for characterisation and making recommendations for sustainable soil management. Thirty-four soil samples were collected at 0–30 cm depth and analysed for selected soil properties. Descriptive statistics, Pearson correlation, and geostatistical analyses (semivariogram modelling and ordinary kriging) were used. Total nitrogen (0.06–0.17 g kg⁻¹, CV = 23.95%, RMSE = 0.03), organic carbon (6.53–46.43 g kg⁻¹, CV = 33.72%, RMSE = 6.62), calcium (1.45–4.37 cmol kg⁻¹, CV = 26.54%, RMSE = 0.71), magnesium (0.45–1.67 cmol kg⁻¹, CV = 35.19%, RMSE = 0.30), and showed considerable variability, with available phosphorus (1.37–37.03 mg kg⁻¹, CV = 123.49%, RMSE = 8.21), exhibiting extremely high variation and the lowest being exchangeable sodium (0.15–0.34 cmol kg⁻¹, CV = 16.99%, RMSE = 0.04) and potassium (0.16–0.32 cmol kg⁻¹, CV = 18.87%, RMSE = 0.04). Most soil properties had moderate spatial dependence and there was significant potential for site-specific nutrient management, and ordinary kriging was used to produce soil fertility maps which clearly identified areas of nutrient sufficiency and deficiency throughout the landscape. Finally, this study also shows the usefulness of GIS and geostatistical methods as decision support tools in sustainable land management in tropical agricultural landscapes.