Throughfall heterogeneity in tropical forested landscapes as a focal mechanism for deep percolation

Forest structure can both reduce and intensify precipitation inputs to the ground surface at fine spatial scales. Areas of localized input at the soil surface may have large effects on deep drainage because of the strongly nonlinear relationship between soil water content (SWC) and unsaturated hydraulic conductivity. We therefore explored the following questions: Does forest structure that creates high spatial heterogeneity in canopy throughfall also create associated deep percolation pathways capable of quickly moving water beyond the rooting zone? Or alternatively, do soil properties resulting from biological activity (e.g. root networks) reduce SWC heterogeneity created by the focused inputs from the canopy and eliminate the potential for these deep percolation pathways? We explored these questions by measuring spatial variation in both throughfall and SWC within 8 forested plots of the Sarapiquí region, Costa Rica where soil texture is relatively homogeneous within deep, clay-rich soils. A novel method that combined soil augering and frequency domain reflectometry was used to assess SWC profiles below the most extreme wet and dry throughfall locations within each plot. Findings revealed relatively homogeneous soil moisture within the surface root zone (0-90. cm depth) with SWC values of roughly 45%. Below the root zone, SWC heterogeneity increased, with the wettest throughfall sites having significantly (α= 0.05) higher SWC than their paired driest throughfall end-members (by 2-15%). Below approximately 130. cm depth, SWC homogeneity was observed again. Physically-based modeling in HYDRUS-3D supports these findings and suggests processes that may explain these changes in SWC patterns observed with increasing depth, such as redistribution through macropores, focused deep-percolation, and lateral downslope flow, respectively. This is the first field-based study that explores the linkage between throughfall heterogeneity and focused deep-percolation, and therefore advances the integrated understanding of how the structure, diversity, and spatial heterogeneity of forests influence their hydrologic outputs. © 2014 Elsevier B.V