Spatial heterogeneity of dissolved organic matter in vegetation-restored gully systems as shaped by coupled erosional and ecological processes

• Vegetation restoration reshaped the spatial distribution of DOM in gully systems. • Topsoil DOM content was 1.3–1.9 times higher than that in deeper soil layers. • Humic-like DOM dominated top layers and protein-like elements increased with depth. • Abrupt DOM changes occurred at permanent gully heads along the gully system floor. • Erosion-driven landforms altered surface traits, jointly affecting DOM spatial patterns. In drylands, gully systems serve as critical zones where vegetation restoration can shape the dynamics of dissolved organic matter (DOM), thereby regulating carbon cycling and ecosystem functioning. However, few studies have addressed this issue. Therefore, the spatial differentiation patterns of soil DOM in gully systems were clarified by the ultraviolet–visible spectroscopy and fluorescence excitation-emission matrix spectroscopy coupled with parallel factor analysis in a typical semiarid region. Overall, the dissolved organic carbon (DOC) content (76.80 mg kg −1 ) of the top layer (TL, 0–2 cm) was 1.3 and 1.9 times higher than the middle layer (ML, 3–20 cm) and the bottom layer (BL, 21–50 cm) within the gully system floor (GSF). Humic-like components (plant-derived macromolecules) predominated in the TL of the GSF (65.8 %), whereas protein-like components (microbial metabolites) accounted for a higher proportion in the BL (62.1 %). Sharp shifts in DOM concentration and composition were commonly observed at the head of permanent gully situated along the GSF within the gully systems. DOM composition was also varied significantly with landscape position (i.e., catchment area, GSF, and gully system bank (GSB)). The humic-like content in the TL of the catchment area was 13 % and 14 % higher than in the GSF and GSB, respectively. The variation of DOM in the gully systems was driven by soil erosion–induced geomorphic alterations and their impacts on the distribution of surface characteristics (e.g., biocrusts properties and plant litter and root densities). This study provides a spatially optimized and ecologically grounded strategy for enhancing carbon sequestration in semiarid environments.