PNNL

Abstract

Soil moisture dynamics strongly influence plant water supply, ecosystem resilience, and land–atmosphere feedbacks in tropical forests, but their vertical and spatial heterogeneity remains poorly quantified. We combined high frequency, in?situ time domain reflectometry (TDR) measurements from 5 to 100?cm depth across plateau, slope, and valley landforms in the Central Amazon’s ZF2 site to (1) quantify soil moisture memory, (2) characterize the timing and strength of rainfall responses, (3) determine dry down time constants following rainfall events, and (4) determine the influence of fine-root distributions in plateaus and valleys (0–40 cm) to these patterns. Landform specific TDR calibration curves were developed to ensure volumetric water content accuracy in these highly weathered soils. During the 2023 dry- to-wet season (August–November), valley soils, despite their sandy texture, showed the longest memory (up to 154?h at 100?cm) and the slowest drying rates (t?>?300?h after =?10?mm rains), due to high water tables and lateral water inputs. Plateau soils showed shorter memory (65–124?h) and rapid drying rates (t?200?h), indicating efficient drainage and little lateral recharge, while slope soils were intermediate. Cross correlation analyses revealed immediate soil moisture responses (0?h lag at 5?cm) in valleys versus delayed responses on plateaus and slopes. Fine root traits suggest potential biological mediation of these patterns. Our results demonstrate that topography plays a key role in the persistence and flow of subsurface water. Accounting for landform and depth dependent dynamics is essential to improve models of drought stress, carbon fluxes, and biosphere atmosphere interactions under increasing hydroclimatic variability.