Search

[Objective] The aims of this study are to investigate the soil infiltration patterns under different land uses in karst trough valley and to propose a suitable soil infiltration model for karst trough valley. [Methods] Four typical land uses (forest, grassland, cropland and orchard) in the karst trough valley were taken as the research objects to determine the differences in soil infiltration characteristics and influencing factors as affected by different land uses, and to evaluate the adaptability of three common infiltration models: Horton, Kostiakov and Philip, the in this area. [Results] (1) Soil bulk density, water content and electrical conductivity were markedly different among different land uses (p＜0.05). Soil bulk density (1.1 g/cm³), water content (26.9%) and electrical conductivity (75.43 mS/cm) in cropland were lower than those in other land uses. Soil bulk density (1.32 g/cm³) in orchard, water content in forest (35.02%) and electrical conductivity in grassland (164.6 mS/cm) were higher than those in other land uses, respectively. (2) The infiltration process of different land use types showed rapid decline, slow decline and stabilization. (3) There were differences in the fitting results of Horton model, Kostiakov model and Philip model for different land use modes in karst trough valley. Horton model had the best results, with the fit R² reaching 0.922～0.951, and Philip model had the worst results, with the fit R² ranging from 0.754 to 0.908. (4) The pH value was highly significantly positively correlated with initial infiltration rate and average infiltration rate (p＜0.01) and significantly positively correlated with stable infiltration rate (p＜0.05); 0.05～0.25 mm and ＜0.05 mm fractions of micro-aggregates were significantly positively and negatively correlated with stable infiltration rate and average infiltration rate, respectively (p＜0.05). The multiple linear regression model showed that soil bulk density, water content and ＜0.05 mm fraction of micro-aggregates were the main factors influencing infiltration rate, which had an inhibitory effect on infiltration. [Conclusion] Cropland has better infiltration performance among different land uses, and the Horton model is more suitable for modeling soil infiltration in karstic troughs and valleys. The results of the study can provide a scientific basis for the assessment of soil infiltration performance and vegetation allocation in karst trough valley areas. [ABSTRACT FROM AUTHOR]