Farmers’ actions towards sustainability: An evidence-based methodology for native forest ecosystems supporting low- carbon livestock in grazed dairy systems

Knowledge of the spatiotemporal distribution of forest carbon (C) stocks is essential for designing strategies that address global environmental change. Here, we developed a transparent framework to establish a 2016–2021 baseline for quantifying, monitoring, and reporting greenhouse gas (GHG) emission reductions and removal enhancements in native forests embedded within grazed dairy systems in southern Chile. We integrated data from 107 forest inventory plots with Sentinel-2 multispectral imagery and implemented a Random Forest regression model on the Google Earth Engine cloud platform. Our results indicate that native forests across the six study sites store a total of 113,751.4 Mg C, with a mean of 157.7 Mg C ha⁻¹. The mean annual increase in C stock between 2016 and 2021 was 3.2 Mg C ha⁻¹ yr⁻¹, corresponding to a total gain of 12,598.3 Mg C (46,194 Mg CO₂eq). Model performance was high, with R² = 0.93 and RMSE = 22.9 Mg C ha⁻¹, indicating robust agreement between observed and predicted values. The approach proposed in this study provides a robust, transparent, and replicable methodology for measuring and monitoring carbon stocks in Chile's native forests. This methodological framework can be used to derive baseline estimates of additional GHG emissions and removals and may support the development of scientifically sound forest carbon projects in dairy landscapes.