Geographical Information System Approach to Suitability Assessment of Cassava Production in Southeastern Nigeria

This study applied geographical information system in the study of suitability of cassava production in soils of Southeastern Nigeria. The study was conducted in three states of Southeastern Nigeria namely Anambra, Enugu and Ebonyi. Geology map aided sampling and six parent material was considered which include Ogwashi-Asaba Formation, Imo Clay Shale, Bende-Ameki formation, Ajali Sandstone, Asu River Group, and Afikpo Sandstone. Three profile pits were dug at two selected parent materials from each of the selected states. Hand held GPS receiver was used to geo-referenced all sites. 18 profile pits were dug in total and sampled for laboratory analysis. The rainfall of the study areas ranged from 1200 mm to 3000 mm, temperature ranged from 220C to 370C and relative humidity ranged from 25% to 85% which were optimum for cassava production. Soil texture varied from sandy, clayey and loamy. The soils were generally acidic (pH: 5.47 to 7.14). The total nitrogen (0.3 g/kg to 1.46 g/kg), available phosphorus (6.84 mg/kg to 12.43 mg/kg), exchangeable potassium (0.04 cmol/kg to 0.54 cmol/kg) and cation exchangeable capacity (4.03 cmol/kg to 11.23 cmol/kg) were rated low. For the aggregate suitability for cassava production, the soils studied were currently not suitable (N1) for cassava production due to fertility constraint. The potential suitability score indicated that most of the soils studied were optimum for cassava production when all the fertility constraints were removed. For the descriptive statistics, the log normal transformation was done to the parametric suitability score for cassava (potential) base on Shapiro–Wilk test. The results were skewed to the left with cassava (potential) having the higher outlier (-3.232) and lower coefficient of variation (4.871). The semivariogram results indicated that Exponential model was best fitted for cassava (Current) having none log normal transformation while Spherical Model was best fitted for cassava (Potential) having log normal transformation. Both falls within strong spatial class with cassava (Potential) having the lowest nugget effect (0.011). The result of the cross validation showed that the model made a moderate prediction in the variability of the suitability of the cassava. From the Kriged map for suitability of cassava (Current), all the study areas were not suitable for cassava production with areas in Ebonyi State showing the lowest suitability score. When the fertility constraints were removed (potential) the study area became highly suitable for cassava production with suitability score between 83.99 to 95 covering large area of the study area and very few places indicating moderately suitable.