Evaluating vegetation–hydrology interactions in Mediterranean catchments using the distributed ecohydrological model MEDFATELAND

Abstract. Mediterranean forest ecosystems strongly regulate catchment hydrology through vegetation-driven processes such as rainfall interception, transpiration, and soil water uptake. In recent decades, increasing forest density following land abandonment has altered water balances across many Mediterranean basins, raising concerns about declining streamflow and increasing drought vulnerability. Process-based ecohydrological models provide a valuable framework for investigating these coupled vegetation–water dynamics and exploring potential management strategies. A spatially distributed implementation is particularly valuable because it allows representing within-catchment heterogeneity, evaluating predicted vegetation dynamics against remotely sensed products, and—crucially—enabling spatially explicit forest management scenarios in which species-specific treatments are applied heterogeneously across space.

In this study, the distributed ecohydrological model MEDFATELAND is presented and applied at watershed scale in Mediterranean catchments (Aigua d'Ora and Siurana) in northeastern Spain. The objectives were (i) to evaluate the ability of the model to reproduce observed hydrological and vegetation dynamics, (ii) to analyze the relationships between vegetation structure and catchment water balance under current conditions, and (iii) to assess the potential effects of forest management on water availability, plant water stress, fuel moisture, and fire behavior.

Model calibration using daily streamflow resulted in moderate performance at the daily scale (NSE = 0.49 in Aigua d'Ora and 0.15 in Siurana) and improved skill at the monthly scale (NSE = 0.62 and 0.46, respectively). Independent evaluation against remotely sensed data showed that simulated gross primary production (GPP) reproduced observed temporal variability reasonably well, while simulated leaf area index (LAI) showed greater spatial variability. Historical trend analyses over the 2001–2023 observation period indicated increasing LAI without significant trends in precipitation or temperature, accompanied by a decline in river discharge, suggesting that vegetation structural changes may be contributing to increased catchment water consumption.

Scenario simulations indicated that forest management reduced canopy structure and interception while increasing blue water fluxes (runoff and deep drainage) in both catchments. The hydrological response was stronger in the wetter Aigua d'Ora basin and more modest in the drier Siurana basin, highlighting the interplay between climate, vegetation composition, and the spatial extent of treatments on catchment-scale outcomes.

Overall, the results demonstrate that MEDFATELAND can realistically represent coupled vegetation–hydrology dynamics at watershed scale and provides a useful tool for evaluating forest management strategies aimed at improving water availability and reducing wildfire risk in Mediterranean environments.