Impacts of Long‐Term Vegetation Restoration on Soil Water Field Capacity in China's Loess Plateau

ABSTRACT

Vegetation restoration is one of the most effective measures to mitigate or prevent land degradation. Long‐term vegetation restoration will inevitably have an impact on soil properties. However, studies on the prediction of soil hydraulic properties such as field capacity (FC) under long‐term revegetation on a regional scale are still limited. This study investigated 243 sampling sites across China's Loess Plateau and measured FC values at depths of 0–10, 10–20 and 20–40 cm. MODIS raw images and four algorithms were used to develop predictive models for predicting FC and mapping FC distributions during 2001–2020. Structural equation modelling was used to analyse the factors driving FC changes. The results showed that Random Forest achieved the highest prediction accuracy for FC at depths of 0–10 cm ( R ²  = 0.43, RMSE = 0.050 cm ³  cm ⁻³ , IQR = 0.053 cm ³  cm ⁻³ ) and 10–20 cm ( R ²  = 0.56, RMSE = 0.042 cm ³  cm ⁻³ , IQR = 0.075 cm ³  cm ⁻³ ), while Backpropagation Neural Network provided the best results at a depth of 20–40 cm ( R ²  = 0.54, RMSE = 0.048 cm ³  cm ⁻³ , IQR = 0.072 cm ³  cm ⁻³ ). From 2001 to 2020, the FC generally increased across the region, with an average annual growth rate of 0.002 cm ³  cm ⁻³ ( p  < 0.01). Mean FC of 0–40 cm increased from 0.223 cm ³  cm ⁻³ in 2001 to 0.251 cm ³  cm ⁻³ in 2020. The most significant increase occurred in the southern and southeastern regions. The Leaf Area Index was identified as the primary factor influencing FC changes, with a direct impact path coefficient of 0.65 in the 0–40 cm. These findings indicate the effectiveness of vegetation restoration in enhancing the soil hydraulic properties and provide a scientific basis for sustainable land management and water‐resource allocation in the Loess Plateau.