Does reforestation increase or deplete soil water storage? Quantifying the effect of reforestation on the soil water balance: a HYDRUS-1D case study comparing a regenerated plot and an adjacent pasture in the Atlantic Forest, Brazil

Details

Supervisors

Javaux, Mathieu ; De Simone Borma, Laura

Faculty

Faculté des bioingénieurs

Degree label

Master [120] : bioingénieur en sciences et technologies de l'environnement, à finalité spécialisée

Abstract

enThe Atlantic Forest has been heavily reduced by centuries of deforestation, and forest restoration is now a central conservation policy. Yet the consequences of regrowth for water resources remain unsettled, since a recovering forest both restores the soil structure that governs infiltration and raises the evaporative demand returned to the atmosphere, so whether a reforested plot is a net consumer or contributor of water must be quantified rather than assumed. This thesis was carried out in collaboration with the INPE through the co-supervisor of this thesis, Laura Borma. It continues the series of studies conducted at the São Francisco Xavier site (São Paulo State). This present work compares two adjacent plots that differ only in land-use history, a 40-year regenerated forest and a degraded pasture, asking over a complete hydrological year whether the forest gains or loses soil water relative to the pasture and through which fluxes any difference arises. The one-dimensional water balance of each plot was simulated with HYDRUS-1D under the van Genuchten Mualem relations with Lenhard hysteresis, the hydraulic parameters being constrained by Bayesian inverse estimation against in-situ records measured by the INPE. Over three common hydrological years the pasture gains 4.4 cm of soil water while the reforested site loses 2.8 cm, a difference of about 7.2 cm in favor of the pasture, so the regenerated forest is a marginally larger net consumer of soil water. This difference is small and emerges as the residual of much larger, largely offsetting fluxes rather than froma single dominantprocess. Thecontrast clearly attributable to land use is the higher transpiration of the forest, whereas the recovered near-surface infiltration capacity had no effect, and the larger capillary supply that keeps the pasture in surplus is governed by the prescribed lower boundary rather than by an interactive water table. The forested profile is also far more hydrologically dynamic between wet and dry years. The study thus confirms the evapotranspiration side of the infiltration–evapotranspiration trade-off, while the sign and magnitude of the storage difference remain conditioned by the poorly constrained lower boundary and the single plot design, the main avenue identified for future work.