Terrestrial land use signals on groundwater fauna beyond current protection buffers

Terrestrial and aquatic ecosystems are tightly linked, with direct implications for applied resource management and conservation. It is well known that human land use change and intensification of terrestrial systems can have large impacts on surface freshwater ecosystems. Contrastingly, the study and understanding of such land use impacts on groundwater communities is lagging behind. Both the impact strength of land use on groundwater communities and the spatial extents at which such interlinkages are operating are largely unknown, despite our reliance on groundwater for drinking water extraction as a key ecosystem service. Here, we analyzed groundwater amphipod occurrence from several hundred shallow groundwater aquifers used for drinking water extraction across a region of varying agricultural intensity and human population density in Switzerland. Despite drinking water extraction sites being generally built at locations with expected minimal aboveground impacts on water quality, we found a direct correlation between land use type and intensity within the surrounding catchment area and the locally measured nitrate concentrations, which is a direct proxy for drinking water quality. Furthermore, groundwater amphipods were more likely to be found at sites with higher forest coverage than at sites with higher crop and intensive pasture coverages, clearly indicating a tight connection between aboveground land use and groundwater biodiversity. Our results indicate that land use type effects on groundwater communities are most relevant and pronounced to spatial scales of about 400-1000 m around the groundwater sampling site. Importantly, the here identified spatial scale is 1.2- to 3-fold exceeding the average extent of currently defined groundwater protection zones. We postulate that incorporating an ecosystem perspective into groundwater management strategies is needed for effective protection of groundwater quality and biodiversity.