A theoretical analysis of transport in a controlled hydrologic volume, inclusive of two willow trees and forced by erratic water inputs, is carried out contrasting the experimental data described in a companion paper. The data refer to the hydrologic transport in a large lysimeter of different fluorobenzoic acids seen as tracers. Export of solute is modeled through a recently developed framework which accounts for nonstationary travel time distributions where we parameterize how output fluxes (namely, discharge and evapotranspiration) sample the available water ages in storage. The relevance of this work lies in the study of hydrologic drivers of the nonstationary character of residence and travel time distributions, whose definition and computation shape this theoretical transport study. Our results show that a large fraction of the different behaviors exhibited by the tracers may be charged to the variability of the hydrologic forcings experienced after the injection. Moreover, the results highlight the crucial, and often overlooked, role of evapotranspiration and plant uptake in determining the transport of water and solutes. This application also suggests that the ways evapotranspiration selects water with different ages in storage can be inferred through model calibration contrasting only tracer concentrations in the discharge. A view on upscaled transport volumes like hillslopes or catchments is maintained throughout the paper.

This paper reports about the experimental evidence collected on the transport of five fluorobenzoate tracers injected under controlled conditions in a vegetated hydrologic volume, a large lysimeter (fitted with load cells, sampling ports, and an underground chamber) where two willows prompting large evapotranspiration fluxes had been grown. The relevance of the study lies in the direct and indirect measures of the ways in which hydrologic fluxes, in this case, evapotranspiration from the upper surface and discharge from the bottom drainage, sample water and solutes in storage at different times under variable hydrologic forcings. Methods involve the accurate control of hydrologic inputs and outputs and a large number of suitable chemical analyses of water samples in discharge waters. Mass extraction from biomass has also been performed ex post. The results of the 2 year long experiment established that our initial premises on the tracers' behavior, known to be sorption-free under saturated conditions which we verified in column leaching tests, were unsuitable as large differences in mass recovery appeared. Issues on reactivity thus arose and were addressed in the paper, in this case attributed to microbial degradation and solute plant uptake. Our results suggest previously unknown features of fluorobenzoate compounds as hydrologic tracers, potentially interesting for catchment studies owing to their suitability for distinguishable multiple injections, and an outlook on direct experimental closures of mass balance in hydrologic transport volumes involving fluxes that are likely to sample differently stored water and solutes.

Flow velocities, residence times, and tracer breakthroughs at the lysimeter scale are affected by matrix properties and preferential flow. Despite their relevance to transport processes, however, the relative timing of preferential flow, and its link to transit times through the soil block, is still poorly described. Here we present and analyze tracer data from a 2.5 m3 vegetated lysimeter experiment where 18 mm of isotopically labeled water was added as a pulse and then followed with a series of tracer‐free controlled rainfall events for 5 months. A solution of two fluorobenzoid acid tracers was also injected and tracked. Time series of soil water samples at three different depths and bottom drainage samples were collected and analyzed. Unlike past lysimeter experiments, a willow tree grown within the lysimeter exerted strong evapotranspiration fluxes. By comparative analysis of soil water and bottom drainage samples, we show the presence of both translatory and preferential flow features reflecting the interplay of slow vertical percolation and fast recharge through macropores. We found that water ponding and evaporating from the top of the lysimeter after irrigation prompted samples to be highly and irregularly fractionated. Comparative analyses of multitracer breakthroughs (adjusted by removing fractionation effects) showed that fluorobenzoid acid tracers reached the bottom of the lysimeter earlier than the isotopes, likely due to the effect of plant uptake. Our results underscore the essential role of models to interpret tracer behavior and, critically, the importance of future experiments aimed at measuring the ages of the water abstracted by vegetation.

No abstract

No abstract