Anthropogenic Management Dominates the Spatial Pattern of Soil Organic Carbon in Saline Cotton Fields of Xinjiang: A Modeling Investigation Based on the Modified Process-Based Model

Haiyan Han, Jianli Ding, Jinjie Wang, Ping Wang, Shuang Zhao, Zheng Zhang, Xiangyu GeAgronomy 2025 · DOI

TL;DR

This study enhances the RothC soil organic carbon (SOC) model to accurately simulate SOC dynamics in Xinjiang's saline cotton fields, revealing that human management (fertilization and straw return) is the dominant factor shaping SOC patterns, often synergistically interacting with natural elements. This work offers a crucial tool for optimizing carbon sequestration in arid agricultural regions.

What Problem Does This Paper Address?

Accurately quantifying soil organic carbon (SOC) is vital for understanding carbon cycles, ensuring food security, and mitigating climate change. However, regional-scale assessments of SOC dynamics, especially in challenging environments like arid, saline agricultural lands, are scarce. Xinjiang, a major cotton-producing region, faces significant salinity stress, which complicates crop growth and SOC management. Existing models often struggle to account for these specific environmental conditions, leading to uncertainties in SOC estimations and hindering effective land management strategies. This paper aims to bridge this gap by developing and applying a modified process-based model to simulate SOC in Xinjiang's saline cotton fields and identify the key drivers of its spatial and temporal variations.

Methods

The core of this research lies in the modification of the well-established RothC process-based model. The authors introduced two key enhancements to better represent SOC dynamics in saline cotton fields:

  • Salinity Adjustment Factor: Recognizing that soil salinity directly impacts microbial activity and carbon decomposition, they incorporated a salinity adjustment factor (d) into the RothC model's decomposition equations. This factor quantifies the regulatory effect of soil salinity on SOC decomposition, using soil osmotic potential and electrical conductivity as inputs.
  • Vegetation Carbon Decomposition Index: To account for the influence of salinity on plant residue decomposition and subsequent carbon input, they integrated a vegetation carbon decomposition index (PPI). This index adjusts the theoretical plant carbon input based on soil salinity levels.

These modifications were then applied to simulate SOC dynamics from 1980 to 2022 using multi-source data. To understand the drivers of SOC variation, the researchers employed the Geodetector method. This statistical technique quantifies the explanatory power of individual factors and their interactions on spatial patterns. They analyzed nine key factors, including climate variables, management practices (fertilizer application, straw return), and soil properties.

Datasets

The study utilized a comprehensive set of datasets:

  • Meteorological Data: Monthly temperature, precipitation, and evaporation data from the European Centre for Medium-Range Weather Forecasts (ERA5) for 1980-2022, downscaled to a 0.0025° resolution using the MTCLIM model.
  • Soil Salinity Data: Inverted using a "multi-source covariate fusion + machine learning inversion" approach on the Google Earth Engine platform, with a spatial resolution of 0.0025°. Covariates included precipitation, potential evapotranspiration, aridity index, slope, Sentinel-2 remote sensing bands, and NDVI time series characteristics.
  • Plant Carbon Return Data: Calculated based on cotton economic yield, dry matter fraction, carbon proportion, residue-to-economic yield ratios, and root-to-shoot ratios.
  • Other Input Data: Fertilizer application rates and irrigation volumes were obtained from the Xinjiang Statistical Yearbook (1980-2022). Cotton field distribution data was sourced from the Chinese Academy of Sciences (2022 distribution).
  • Validation Data: Field samples collected by the research team in the Weiku-Aksu region of Xinjiang (2016-2024) were used for model validation.

Key Insights and Ideas

  • Human Management Reigns Supreme: The most striking finding is that anthropogenic management, specifically fertilizer application (ORG), is the dominant driver of SOC spatial variation, explaining over 62% of the variation. This highlights the profound impact of human interventions on soil carbon in these arid agricultural systems.
  • Synergistic Power: Interactions between factors are crucial. Management practices like fertilization and straw return exhibit strong synergistic enhancement effects when combined with other factors (e.g., precipitation, temperature, elevation), amplifying their influence on SOC. This suggests that a holistic approach to management is more effective than isolated interventions.
  • Salinity's Subtle but Significant Role: While not the primary driver, the model's modification to account for salinity was essential for accurate simulation, demonstrating that even in the face of dominant human factors, environmental stresses like salinity cannot be ignored.
  • Regional Heterogeneity Matters: SOC distribution is not uniform. High SOC is found south of the Tianshan Mountains, likely due to favorable water and thermal conditions, while southwestern Xinjiang shows lower SOC accumulation, potentially linked to desert margins and lower precipitation.
  • Temporal Trends are Complex: While an overall fluctuating upward trend in SOC was observed across Xinjiang, localized declines in both high- and low-SOC zones indicate that even seemingly stable systems can experience degradation under specific conditions.

Key Results

  • The modified RothC model achieved high accuracy in capturing SOC dynamics in salinized cotton fields, with an R² of 0.47 and an RMSE of 3.5 t C/ha during validation. This RMSE represented only 5.8% of the regional average SOC stock.
  • Fertilizer application (ORG) was identified as the most critical factor influencing SOC spatial variation (Q-value = 0.6260).
  • Straw return (STRAW) also played a significant role (Q-value = 0.2823), demonstrating the importance of residue management.
  • Climate factors like precipitation (PREC) and temperature (TEMP) had moderate explanatory power (Q-values around 0.2610).
  • Interactions between factors, particularly between management practices and climate, showed synergistic enhancement, with some interaction Q-values exceeding 0.77, significantly higher than individual factors.
  • Over the 43-year study period (1980-2022), Xinjiang cotton fields showed an overall fluctuating upward trend in SOC, increasing by 1.42 t C/ha. However, the southwestern low-SOC zone experienced a significant decline.

Limitations and Future Work

The authors acknowledge several limitations:

  • Data Input Uncertainty: The assumption of uniform management practices within administrative regions (prefecture-level cities) may not fully capture field-scale heterogeneity. Detailed agricultural management data for specific regions was also limited due to confidentiality.
  • Model Parameterization: Regional parameters for salinity adjustment and vegetation carbon decomposition were not systematically calibrated with field measurements specific to Xinjiang. The model also didn't fully account for variations in soil salt types or cotton variety-specific salt tolerance.
  • Assumptions: The study assumed stable cotton planting patterns and ignored fallow land dynamics, which could introduce minor deviations. Meteorological data downscaling might also have incurred accuracy loss.

Future research directions suggested include:

  • Comprehensive Model Optimization: Developing a multi-scale nested validation system, establishing refined observation units, and calibrating regional parameters with encrypted validation data.
  • Enhanced Data Acquisition: Cooperating with local departments for partitioned management data and utilizing UAV remote sensing and farmer surveys for accurate input data quantification.
  • Integrating Historical Data: Developing a standardized input database and integrating multi-source historical land use datasets to reconstruct dynamic spatial patterns and explore the impact of historical planting boundary changes.
  • Long-term Field Trials: Calibrating salinity adjustment factors and vegetation carbon decomposition indices through fixed-site trials, incorporating diverse salt types and cotton varieties.

Summary updated

Abstract

Salinity is a key abiotic stress limiting crop growth. Accurate quantification of carbon budgets and their environmental responses is critical for sustainable cotton production, yet regional-scale assessments remain scarce. To clarify the evolutionary patterns and driving mechanisms of soil organic carbon (SOC) in saline cotton fields of arid Central Asia, this study focused on Xinjiang and modified the RothC model by integrating salinity adjustment factors and vegetation carbon decomposition indices, simulating SOC dynamics (1980–2022) with multi-source data. Results showed the improved model achieved high accuracy in capturing SOC dynamics in salinized cotton fields. Spatially, SOC exhibited high levels south of the Tianshan Mountains and low levels in southwestern Xinjiang; temporally, it showed an overall fluctuating upward trend, though both high- and low-value zones displayed localized declines. Geodetector analysis revealed fertilizer application as the primary driver of SOC spatial variation, followed by straw return, precipitation, and temperature, with most factors showing synergistic enhancement effects. Human management (fertilization and straw return) is the core regulator of SOC, and its synergy with natural factors shapes SOC spatiotemporal patterns. The salinization-adapted RothC model provides a novel framework for arid cotton field SOC simulation, offering scientific support for carbon pool optimization and sustainable agriculture in arid regions.
BibTeX
@inproceedings{han2025anthropogenic,
  author = {Haiyan Han and Jianli DING and Jinjie Wang and Ping Wang and Shuang Zhao and Zheng Zhang and Xiangyu Ge},
  title = {{Anthropogenic Management Dominates the Spatial Pattern of Soil Organic Carbon in Saline Cotton Fields of Xinjiang: A Modeling Investigation Based on the Modified Process-Based Model}},
  booktitle = {Agronomy},
  year = {2025},
  doi = {10.3390/agronomy16010017},
}
Figure