Korean Journal of Soil Science and Fertilizer ISSN:0367-6315(Print) 2288-2162(Online) 한국토양비료학회 학회지

One of the most important practices for nitrogen (N) management in arable lands is to minimize N loss using soil-customized N supply. Despite that N mineralization of soils is an useful parameter to manage N fertilization, it requires a lot of analysis. To predict N mineralization rate of upland soils, we performed regression analysis between N mineralization parameters (Mineralized N, N₀, k) and soil properties. Experimental soils were randomly sampled from 40 pepper cultivation fields before fertilization. The rate of soil N mineralization was analyzed through incubation for 16 weeks at 27°C. N mineralization parameters of soils were estimated from first-order kinetic model, and a stepwise regression was conducted to predict the mineralized N, N₀ and k using soil properties. Mineralized N was linearly correlated with EC (R²=0.745) which meant a good indicator to predict N mineralization rate. N₀ as potentially mineralizable N was positively correlated with soil organic matter (SOM) and total N, and those represented a linear relationship (y=0.0682x, R²=0.729). SOM, sand, clay and C/N ratio were selected as parameters for prediction of N₀ (R²=0.448). In addition, a regression equation for estimation of k was composed of EC, CEC, sand and clay (R²=0.508). In conclusion, we suggest that N mineralization rate by combination with soil properties could be determined with a regression equation. However, model equation represented in this study should be validated and improved through field experiments with different levels of N application and different types of soil property.

Stepwise multiple regression equations for estimating nitrogen mineralization factors in relation to soil properties.