Assessing carbon storage and soil erosion patterns in different habitat types: A case study of Belihuloya, Sri Lanka

Abstract

Environmental degradation in Sri Lanka has accelerated since the 1870s, as commercial agriculture expanded, driving widespread deforestation that resulted in severe land degradation through soil erosion and carbon loss. This study evaluates carbon storage and soil erosion across five land use types in the Belihuloya region: natural forests, forest patches, plantation forests, home gardens, and tea plantations. InVEST Sediment Delivery Ratio (SDR) and Carbon Storage models (version 3.12.1) were employed to generate spatially distributed predictions, evaluating how land-use changes impact ecosystem services. Land Use Land Cover (LULC) classification identified 16 land cover types, with forests and forest patches achieving the highest mean carbon storage (129.30 t ha-1), while home gardens and tea plantations showed reduced storage due to human management practices. Soil erosion modelling revealed a mean annual potential of 315.81 t ha-1 based on RKLS (R × K × LS) values, though actual Universal Soil Loss Equation (USLE)-based erosion rates were significantly lower (9.16 t ha-1), exceeding the national soil erosion tolerance threshold of 5 t ha-1. Avoided erosion and sediment export values further demonstrated the protective capacity of forested landscapes. Results identified erosion hotspots concentrated in tea plantations on steep slopes, where cultivation practices and vegetation removal increase vulnerability. Conversely, forested areas provided superior carbon retention while maintaining effective erosion control. These findings emphasise the critical need for soil cover management and conservation strategies to prevent further degradation. This research demonstrates the effectiveness of GIS-based modelling tools for assessing environmental impacts of land use change and supports integrating conservation practices, such as agroforestry and reforestation, into land-management planning.