Maping the Soil Fertility of Bisankhel Catchment of Chitlang Vdc and Comparison of Different Geo-Spatial Interpolation Techniques

Abstract

The aim of this research work was to map the status of soil nutrients in the Bisankhel catchment of Chitlang VDC, Makawanpur, Nepal. The study area covered 1023.25 hectares of land, extending from 85°8'8.433"E to 85°10'10.198"E longitude and 27°37'24.251"N to 27°40'21.560"N latitude. Total Nitrogen (N), available Phosphorus(P), extractable Potassium (K), soil organic matter (SOM) and soil pH were measured for 50 soil samples collected using random sampling representing different land uses in the study area. Most of the samples were found very acidic where uplands (4.879±0.119) were more acidic than lowlands (5.036±0.0973). Most samples upon analysis were found to be medium in SOM and total N. SOM was slightly higher in lowlands (3.385±0.256%) than in uplands (3.05±0.206%). Total N was also slightly higher in lowlands (0.145±0.00634%) than in uplands (0.127±0.00639%). Similarly, Available P and K were found very high in majority of the soil samples. Available P was slightly higher in lowlands (92.34±16.96ppm) than in uplands (71.58±15.47ppm). However, extractable K was slightly more in uplands (269±31.32ppm) than in lowlands (240.4±23ppm). Spatial prediction using various interpolation methods was performed in ArcGIS 10.5 software platform using ‘Geostatistical Analyst’ extension. The predicted values in raster data structure were used for mapping the soil fertility status of the catchment. Different interpolation methods were evaluated following cross-validation approach. Comparison of prediction errors was carried out to select the optimum prediction methods for the interpolation of soil nutrient values. Completely Regularized Spline and Ordinary Kriging methods were found to yield better prediction of soil nutrient status among the Deterministic and Kriging methods respectively. Soil pH being the limiting factor, pH control was necessary.