Land Use-Dependent Preferential Flow Paths Affect Hydrological Response of Steep Tropical Lowland Catchments With Saprolitic Soils

Forested catchments in Central Panama can produce more base flow during the dry season compared to pasture catchments-the so-called forest sponge effect. During rainfall events, peak storm runoff rates and storm runoff coefficients can be lower for forested catchments than pasture catchments, even when they have similar topographic characteristics, underlying geology, and soil texture. The internal mechanism of these differences in hydrological response due to land use is yet to be fully understood. A distributed model explicitly simulating preferential flow paths (PFPs), which is referred to as PFPMod, is used to explain the hydrological response caused by land use using data from three catchments with distinct land covers in Central Panama. Input parameters of forest and pasture land covers were identified using field observations and literature values. Multiple satisfactory objective criteria demonstrate that the two end-member land cover parameter sets are adequate to explain the observed difference in dry-season base flow and storm runoff coefficients. Field measurements of matrix infiltrability using soil cores and plot-scale infiltration capacity enabled estimating the number of vertical macropores that fully penetrate the root zone. Model simulation results demonstrate that fast drainage through lateral PFPs in the early wet season and high flow in vertical PFPs to recharge deep groundwater in the late wet season contribute to the observed differences in peak storm runoff and the forest sponge effect during the dry season. This study provided insights to the mechanism by which reforestation may help to restore ecosystem services and water security in tropical settings.