Defining soil organic carbon sequestration potential in Mediterranean agroecosystems for effective carbon credit policies: a multi-method approach

The implementation of carbon credit policies is particularly relevant in Mediterranean agroecosystems, which are highly vulnerable to SOC losses due to climate change. However, to what extent each soil has potentials to increase the SOC content is still debated, making it necessary to determine reliable benchmark values to achieve for maximum SOC accumulation. Here, we implemented and tested different model approaches to determine topsoil SOC reference values in the Euro-Mediterranean region, including the Balkans , with data from the LUCAS soil database.. According to other studies, the reference value was identified as equivalent to those of less disturbed agroecosystems such as managed grasslands.  Three methods were tested to estimate the sequestration potential. The first method was based on the identification of pedoclimatic zones through probabilistic clustering using a Gaussian Mixture Model (GMM), where each homogeneous areas was identified based distribution of  pedological, topological and climatic conditions  (e.g., texture, air temperature , elevation and net photosynthetic production ). The second method employed random forest, a machine learning technique. The model, trained on grassland reference points, was then applied to agricultural land, estimating potential SOC levels as if every point were converted to grassland. Lastly, the RothC  model was applied to each point, simulating grassland management for 100 years, and the obtained SOC value was taken as the reference one. Each point then had three potential SOC reference values, and the median value has been taken as the final sequestration potential.
The harmonization of these three methods provided reliable sequestration potential estimates for each LUCAS point, allowing for point-specific predictions. Simultaneously, this approach enabled the delineation of geographical zones with distinct pedoclimatic properties, producing maps of reference zones for SOC sequestration. These maps will allow carbon credit policies to be tailored to the specific conditions of each region in the Euro-Balkan Mediterranean area, ensuring more effective policy implementation.

This work was co-financed by the Interreg Euro-MED Programme within the project “Capturing and Storing Atmospheric CO₂ for Improvement of Soil Quality - CARBON 4 SOIL QUALITY”.