A nature-based solution to enhance aquifer recharge: Combining trees and infiltration basins

Managed aquifer recharge (MAR) is the replenishment of aquifers for environmental recovery or later use of the stored water. Under increasing occurrence of hydrological extremes, it shows great potential to buffer water availability. Nonetheless, these systems are affected by clogging, which decreases the infiltration rates over time. To deal with this issue and evaluate the potential to enhance recharge, we studied the effect of poplars (Populus Nigra predominantly) on two infiltration basins in a MAR system in central Spain, namely Santiuste (ST) and La Laguna del Señor (LS). The hypothesis is that plant roots can increase infiltration rates and counteract the effect of clogging through macropore formation and changes in soil characteristics. To this end, we conducted eight on-site infiltration test campaigns over two years in areas with and without trees, in addition to determining bulk density, organic carbon content and soil textures. We also built a 1-D vadose zone model and compared measured steady-state infiltration rates (SSIR) with theoretical values. When compared to bare soil, areas with trees show higher SSIR (37 % difference on average), higher organic carbon content (113 % difference on average), and lower bulk density (38 % difference on average). In all cases, measured SSIR values were larger than theoretical values, meaning that textures alone cannot explain local infiltration rates. Moreover, the difference between theoretical and measured SSIR is larger in trees than in bare soil by 40 % in ST and 198 % in LS. The 1-D vadose zone model shows that in the area studied, the transpiration by trees is largely offset by the additional recharge due to enhanced soil infiltration rates. These results suggest that trees increase infiltration rates, likely due to macropores and aggregate formation and the rearrangement of the soil structure. This nature-based solution could be extended to other MAR water-spreading methods and in-channel modifications.