Analysis of UV–Vis spectral characteristics and content estimation of soil DOM under mulching practices

Dissolved organic matter (DOM), which is generally expressed in terms of dissolved organic carbon (DOC), is a key indicator used to monitor the short-term dynamics of soil quality. However, its content and molecular composition are strongly influenced by agricultural practices, and the determination of DOC content is cumbersome and expensive. A field experiment with three treatments, including (i) no mulching (CK), (ii) gravel mulching (GM) and (iii) film mulching (FM), was selected to study the content and composition characteristics of DOC under different agricultural practices using ultraviolet visible (UV–Vis) technology. The results showed that the GM and FM treatments significantly decreased the DOC contents in topsoil (0–20 cm) by 6.15% and 12.18%, respectively, in comparison with the control. The FM treatment significantly decreased the values of specific UV absorbance at 254 nm (SUVA₂₅₄), 260 nm (SUVA₂₆₀) and Chromophoric DOM (CDOM) in each soil layer, indicating the aromaticity, hydrophobicity and CDOM content in the FM treatment were significantly decreased in each soil layer. To establish an accurate model for rapid determination of soil DOC, Spearman correlation analysis was performed on normally and non-normally distributed datasets, and linear regression analysis was utilized for modeling with normalized data. The results of linear regression analysis showed that the regression models established among CDOM/DOC, CDOM, SUVA₂₅₄, SUVA₂₆₀, and specific UV absorbance at 280 nm (SUVA₂₈₀) had a high degree of fit, with coefficients of determination (R²) all exceeding 0.95. The model validation results showed that the regression model established between CDOM/DOC and CDOM had the highest accuracy with the lowest root mean squared error (RMSE) (3.58 mg kg⁻¹) and the lowest relative error (RE) (2.74%), which was DOC = 58.065 + 184.355 × A(355) mg kg⁻¹. The results showed that the estimation of soil DOC can be achieved with good prediction using UV–Vis technology. This study highlights the potential of UV–Vis spectroscopic technology to rapidly estimate soil DOM in agriculture.