Modelling soil organic carbon at multiple depths in woody encroached grasslands using integrated remotely sensed data

Abstract

Woody plants encroachment into grasslands has considerable hydrological and biogeochemical consequences to grassland soils that include altering the Soil Organic Carbon (SOC) pool. Consequently, continuous SOC stock assessment and evaluation at deeper soil depths of woody encroached grasslands is essential for informed management and monitoring of the phenomenon. Due to high litter biomass and deep root structures, woody encroached landscapes have been suggested to alter the accumulation of SOC at deeper soil layers; however, the extent at which woody plants sequester SOC within localized protected grasslands is still poorly understood. Remote sensing methods and techniques have recently been popular in SOC analysis due to better spatial and spectral data properties as well as the availability of affordable and eco-friendly data. In this regard, this study sought to quantify the accumulation of SOC at various depths (30 cm, 60 cm, and 100 cm) in a woody-encroached grassland by integrating Sentinel-1 (S1), Sentinel-2 (S2), PlanetScope (PS) satellite imagery, and topographic variables. SOC was quantified from 360 field-collected soil samples using the loss-On-Ignition (LOI) method and spatial distribution of SOC across the Bisley Nature Reserve modelled by employing the Random Forest (RF) algorithm. The study’s results demonstrate that the integration of topographic variables, Synthetic Aperture Radar (SAR), and PlanetScope data effectively modelled SOC stocks at all investigated soil depths, with high R² values of 0.79 and RMSE of 0.254 t/ha. Interestingly, SOC stocks were higher at 30 cm compared to 60 cm and 100 cm depths. The horizontal reception (VH), Slope, Topographic Weightiness Index (TWI), Band 11 and vertical reception (VV) were optimal predictors of SOC in woody encroached landscapes. These results highlight the significance of integrating RF model with spectral data and topographic variables for accurate SOC modelling in woody encroached ecosystems. The findings of this study are pivotal for developing a cost-effective and labour-efficient assessment and monitoring system for the appropriate management of SOC in woody encroached habitats.