Particulate organic carbon reveals regional‐specific patterns of soil structural degradation in two contrasting cropland systems

Abstract

Soil degradation threatens agricultural productivity and environmental sustainability, a concern often manifested through declining levels of soil organic carbon (SOC). However, the reliability of SOC as a degradation indicator can be confounded by region‐specific edaphic conditions and land use history, suggesting that soil degradation assessments should move beyond simplistic relations with SOC quantity. This study investigates whether particulate organic carbon (POC), a labile SOC fraction, could serve as a more sensitive indicator of soil structural degradation. This question was explored by comparing two intensively managed agricultural regions, northeast China (NEC) and the Belgian Loam Belt (BEL), which share similar loess‐derived soils but differ in their degradation trajectories. Two regional‐scale datasets revealed contrasting relationships between SOC and soil aggregate stability (SAS), despite their similar SOC ranges and textural compositions. In BEL, SAS strongly correlated with SOC ( r  = 0.84), driven by a POC‐SOC synergy in which elevated POC supplied the labile binding agents needed to enhance macroaggregate formation. Conversely, NEC exhibited a weak SOC‐SAS relationship ( r  = 0.37), with persistently low POC content aligning with highly unstable soil structure across SOC gradients, indicative of widespread degradation from intensive tillage and erosion that disrupted the POC‐mediated aggregation pathway sustaining soil structure. Cross‐scale investigation into the impact of soil erosion on POC characteristics across four test cases, including large‐scale POC datasets in Europe and North America, showed that severely eroded areas consistently coincided with low POC levels. These findings underscore the potential of POC in capturing regional patterns of soil structural change, thereby supporting its potential as an indicator for cropland degradation monitoring.