Efficiency-enhancing methods for predicting nitrogen mineralization characteristics in paddy soils using soil properties and rapid soil extraction

Abstract

Soil mineralized nitrogen (N) is a vital component of soil N supply capacity and an important N source for rice growth. Unveiling N mineralization (Nm) process characteristics and developing a simple and effective approach to evaluate soil Nm are imperative to guide N fertilizer application and enhance its efficiency in various paddy soils with different physicochemical properties. Soil properties are important driving factors contributing to soil Nm differences and must be considered to achieve effective N management. Nevertheless, discrepancies in Nm capacity and other key influencing factors remain uncertain. To address this knowledge gap, this study collected 52 paddy soil samples from Taihu Lake Basin, south China, which possesses vastly different physicochemical properties. The samples were subjected to a submerged anaerobic incubation experiment at a constant temperature to obtain the soil Nm characteristics. In addition to other analyses, reaction kinetics models were employed to compare characteristic differences between Nm potential (Nmp) and short-term accumulated mineralized N (Amn) processes in relation to soil physicochemical properties. Based on these relationships, simplified Nmp prediction methods for paddy soils were established. The results revealed that the Nmp values in pH < 6.50, pH range from 6.50 to 7.50, and pH > 7.50 were 145.18 mg kg-1, 88.64 mg kg-1 and 21.03 mg kg-1, respectively. Significantly, after comparing different incubation days, short-term Amn at 14 days showed a good correlation (P < 0.01) with Nmp (R2 = 0.94), indicating that the prevailing short-term incubation experiment is an acceptable marker for Nmp. Moreover, Nmp correlated well with the ultraviolet absorbance obtained at 260 nm using NaHCO3 extraction (Na260), streamlining the Nmp estimation method further. The incorporating easily obtainable soil properties, including pH, total nitrogen (TN), and the ratio of total organic carbon to TN (C/N) alongside Na260 for Nmp evaluation allowed the multiple regression model, Nmp = 58.62 × TN - 23.18 × pH + 13.08 × C/N +86.96 × Na260, to achieve high predictive accuracy (R2 = 0.95). The reliability of this prediction was further validated with published data in paddy soil from the same region and the other rice regions, demonstrating the regional applicability and prospects of this model. This study underscored the roles of soil properties in Nm characteristics and mechanisms and established a site-specific prediction model based on rapid extraction and edaphic properties from the paddy soil, paving the way for developing rapid, precise Nm prediction models.