Improving groundwater dynamics in restored tidal marshes: an explorative field and modelling study

Along estuaries and coasts, tidal marsh restoration projects are increasingly being executed on formerly embanked agricultural land to regain the ecosystem services provided by tidal wetlands. There are, however, more and more indications that restored tidal marshes do not deliver these ecosystem services to the same extent as natural tidal marshes. In particular, we found that marsh restoration on a compacted agricultural soil (which has a very low porosity and hydraulic conductivity) leads to reduced groundwater fluxes and soil aeration, which may imply decreased soil-water interactions, reduced biogeochemical cycling and impaired vegetation development.We studied the subsurface hydrology in the restored marsh Lippenbroek (Scheldt estuary, Belgium). To investigate spatial and temporal variation of groundwater fluxes in the restored tidal marsh, we developed a real-time groundwater flux sensor (iFLUX sensor) that enables us to measure both the groundwater flow velocity and flow direction in real-time. With these instruments installed at multiple locations and depths in the marsh soil, we were able to capture the effects of the tidal regime and soil stratigraphy on groundwater flow in high detail.Furthermore, we set up a 2D vertical model in HYDRUS with a domain representing a creek and marsh cross-section. The model enables variably saturated flow calculations for groundwater flow and solute transport in dual porosity soils. Input parameters for the model were obtained by soil sampling and laboratory measurements of saturated hydraulic conductivity and soil water retention curves. Simulated results are in good agreement with in situ measured groundwater levels in monitoring wells.With a scenario analysis, we showed that in a scenario in which the compact subsoil is absent, 6 times more water passes through the marsh soil during a spring tide – neap tide cycle compared to the reference scenario in which the compact soil starts at a depth of 60 cm. In the compact layer, which is always saturated, flow rates are so low that this layer is expected not to contribute to nutrient cycling.We then simulated the effect of (i) creek excavation (by varying the number of creeks in the domain) and (ii) soil amendments (by varying the depth to the compact layer) on groundwater flow in newly restored tidal marshes. We found that increasing the creek density from 1 creek to 2 creeks per 50 m marsh, or changing the depth to the compact layer from 20 cm to 40 cm, both more than doubles the volume of water processed by the marsh soil. As such , our study demonstrates that groundwater modelling is a useful tool in support of designing marsh restoration measures aiming to optimize groundwater fluxes and related ecosystem services.