Spatial Variability Of Soil Nutrients And Regional Fertilization Management In A Leased Cane Field

A field experiment was conducted at Laibin in the Guangxi Zhuang Autonomous Region, a key base for the Sugarcane industry located in southwest China. Samples were collected from sixty-eight 50mx50m plots in a small scale field to study the spatial variability and distribution of soil nutrients and to build regional fertilization zones in a red, acidic sugarcane field soil in a low hill subtropical region. Results obtained from statistical analysis indicated that the variable intensity of soil pH was low, while that of soil organic matter and available nutrients was mid-range. The distribution of soil organic matter and macronutrients covered several grades, but that of soil pH and micronutrients covered mainly one grade. The correlation matrix of soil pH and soil nutrients content showed that the correlations between some items were significant.From the results of geostatistical analysis, the semivariograms for soil organic matter, Ln available N, Ln available K, exchangeable Mg, available Cu, and available Zn values fit the Spherical Model; while that for soil pH, Ln available P, exchangeable Ca, available Fe and available K, available Si, fit the Exponential Model; while Ln available Mn fit the Gaussian Model. Geostatistical analysis also revealed that the spatial correlations of soil pH, available P, available K, exchangeable Mg, available Mn and available Si were moderately dependent. Their spatial variability was caused by structural and random factors. Moreover, the spatial correlations of soil organic matter, available N, exchangeable Ca, available Cu, available Zn and available Fe were strongly dependent. Their spatial variability was mainly caused by structural factors. The effective range of macronutrients was relatively small, while that of soil pH and micronutrients was relatively large. The effective range of these items was 117.6 to 3105.0 meters. The maps from the Ordinary Kriging Method showed that each item had its own distribution pattern. The maps revealed the distribution values of the cane field, and could be used to conduct optimal fertilization for sugarcane planting.The map surface layers of available N, available P and available K were overlaid and revealed five fertilization zones. Initial fertilization trials of the zones suggested that sugarcane grown by regional fertilization was better than sugarcane grown by traditional fertilization and the cane yield and sugar yield were higher in most regional fertilization plots as compared to those in traditional fertilization plots, across the five zones.