Characterization and predictive modelling of quasi-saturated hydraulic conductivity across land use systems in the Northwest Himalayas.

This study investigated the variation in soil hydrological properties-saturated water-holding capacity (SWHC), field capacity (FC), total porosity (TP), and saturated hydraulic conductivity (Ks)-across different land use systems in the Semalta watershed of the north Himalayan foothills. Soil samples were collected from four land use types (agriculture, barren land, deodar forest, and oak forest) at two depths (0-15 cm and 15-30 cm). Key parameters were measured using mini-disk infiltrometers and standard laboratory techniques. Statistical analyses included redundancy analysis (RDA), variance partitioning analysis (VPA), and multiple linear regression (MLR) for predictive modelling. Results showed that forested lands exhibited superior soil structure and water dynamics: porosity was highest in deodar (TP: 0.49) and oak (TP: 0.48) soils, while barren land had the lowest (TP: 0.41). Similarly, Ks values were higher in deodar (0.15 cm/hr) and oak (0.14 cm/hr) soils, indicating better permeability than agricultural and barren soils. RDA and VPA highlighted the dominant influence of aggregate stability, bulk density, and soil organic carbon on hydrological responses. MLR models achieved adjusted R² values above 0.90, confirming their utility in predicting soil-water dynamics. These findings underscore the importance of vegetation in maintaining soil hydraulic function and provide a framework for land use planning and sustainable watershed management.