Spatial variation of apparent electrical conductivity of saline-sodic paddy soils in the Songnen Plain

The Songnen Plain is one of the largest saline-sodic soil regions in not only China but whole World. The saline-sodic soils in the plain have high spatial variability in soil salinity and sodicity. This heterogeneity has made it very difficult to use the saline-sodic soils for large scale paddy rice production in the region. The objective of this study was to investigate the degree of spatial variability and structural variance of soil apparent electrical conductivity in the plain. The study also determined the relationships between soil apparent electrical conductivity and soil-water extract electrical conductivity (EC_1︰5) and then soil sodicity (ESP) using Pearson regression analysis. Soil salinity and spatial variability were measured using EM38 (electromagnetic induction) and GPS (global position system) systems in saline-sodic paddy soils in the plain. The data were then analyzed using classical statistical and geo-statistical techniques. Classical statistical analysis suggested that EC_h (soil apparent electrical conductivity in the horizontal direction) and ECv (soil apparent electrical conductivity in the vertical direction) had moderate spatial variation with normal distribution curves. However, geo-statistical analysis showed that EC_h and EC_v had strongly spatial auto-correlation due to compound impact of structural factors. The optimal model of empirical semivariograms used to simulate EC_h and EC_v was the exponential model. The spatial distribution maps interpolated with Kriging apparently showed similar spatial distributions of EC_h and EC_v. Based on Pearson correlation, it was noted that soil EC_1︰5 and ESP were significantly correlated with EC_h and EC_v, with coefficients of correlation greater than 0.8. Regression analysis showed exponential function relations between EC_h and EC_v and between EC_1︰5 and ESP, with coefficients of determination greater than 0.76. Comparison of soil apparent electrical conductivity in the vertical and horizontal directions suggested that the correlation coefficients and determination coefficients between ECh, EC_1︰5 and ESP were greater than those between EC_v, EC_1︰5 and ESP. This suggested that EC_h better reflected the degree of soil salinity and was more appropriate for saline-sodic soil studies in the region. The results of the study provided the theoretical basis for the integration of salinity spatial variability theory with soil improvement practices. It also provided data support for homogenous in situ reclamation of saline-sodic soils to reduce soil salinity for rice production.