Municipal wastewater treatment plants in coastal areas are facing numerous challenges, including the need to provide cost-effective approaches for removing nutrients from wastewater, as well as adapting to the effects of climate change. Retrofitting conventional wastewater treatment plants to remove nutrients can be expensive and is technically challenging. Moreover, wastewater treatment facilities are often situated along coasts, leaving them uniquely susceptible to sea level rise. For these reasons, innovative wastewater utilities have begun to consider more seriously the use of nature-based approaches, such as horizontal levees, to manage these simultaneous pressures. Horizontal levees consist of sloped subsurface wetlands that are built between storm control levees and tidal marshes. When combined with tidal marshes, these systems can help attenuate storm surges and provide space for wetland transgression to higher elevations as sea levels rise. At the same time, removal of residual wastewater-derived contaminants (i.e., nutrients and trace organic contaminants) from treated wastewater effluent can be achieved in the subsurface of these systems.In the research detailed within this dissertation, we evaluated the ability of horizontal levees to improve water quality. Specifically, we focused on the ability of these systems to remove nutrients (i.e., nitrogen and phosphorus) from nitrified secondary wastewater effluent. To identify optimal operating conditions, we monitored water quality improvements at a 0.7-ha experimental horizontal levee system. First, we assessed the impact of design and operational parameters on contaminant removal, as detailed in Chapter 2. The removal of nitrogen and trace organic contaminants was particularly sensitive to hydrology in this system: rapid and near complete removal (>97%) was observed in water flowing through the subsurface of this system, whereas water flowing over the surface did not receive treatment. When overland flow was eliminated, removal of F+ coliphage (>99%) and phosphate (>83%) was also significant. However, phosphate removal was not as sensitive to hydrology as was removal of other contaminants.Using porewater sampling, isotope measurements and mass balances on nitrogen and other redox active species, we investigated the mechanisms responsible for nitrogen removal in this system. As detailed in Chapter 3, we identified that microbial metabolic processes (i.e., denitrification and anammox) were responsible for the majority (approximately 80%) of nitrogen removal in this system. The addition of labile organic carbon in the form of wood chips to this system stimulated heterotrophic microorganisms, leading to a progression of reduction of dissolved oxygen, nitrate, Mn-oxides, Fe-oxides and sulfate within the first 15 m of the subsurface. This progression was especially rapid in the summer. Fe(II)(aq) and sulfide produced from these processes precipitated to form a reservoir of sulfide minerals in the wetland sediments. During cooler winter months, autotrophic denitrifiers paired oxidation of those Fe(II)-sulfides to nitrate reduction, consuming as much as 30% of the nitrate removed in the wetland during the winter. To project long-term removal of nitrogen in horizontal levees, we developed an electron transfer model, described in Chapter 3, to account for production and consumption of electron donors (e.g., organic carbon) that are required to fuel denitrification. Results indicated that horizontal levees could remove nitrate from wastewater effluent for at least 50 years before the carbon amended to the system (e.g., as wood chips) would be depleted. After the wood chips are depleted, sulfide minerals, decaying vegetation and root exudates may provide enough electrons to fuel long-term nitrogen removal.Plant uptake can also be a significant removal pathway for nitrogen in nature-based treatment systems. However, past methods for quantifying plant uptake have often relied on harvesting plants and assuming that all nitrogen stored in plant biomass is derived from wastewater. This assumption is inappropriate in pilot- or full-scale systems where other sources of nitrogen are available. Moreover, harvesting methods can be laborious and inaccurate when extrapolated to large wetland areas due to heterogeneous distributions of plant biomass. To improve our understanding of this removal mechanism, we developed a new method for quantifying plant uptake, detailed in Chapter 4, in which we used a stable isotope mixing model to distinguish between nitrogen sources. We applied this new method at the field site and found that 14% of nitrogen in plants was derived from wastewater with the remaining nitrogen obtained from the soil. By combining these results with remote-sensing derived biomass measurements, it was determined that 8% of nitrogen removal in this system was due to plant uptake. There were large variations in plant uptake along the wetland slope, both seasonally and with plant maturity. Plant uptake also varied significantly based on design parameters, suggesting that design decisions can have an important impact on this removal pathway. We present this new method as a useful way to inform our understanding of nitrogen cycling and optimization of nature-based nutrient control systems.We also assessed the cycling of phosphorus in the horizontal levee test facility (Chapter 5). Despite observing significant phosphate removal, removal of phosphate was largely offset by export of dissolved organic phosphorus from the pilot system. This suggested that phosphate may be consumed by microorganisms or assimilated into plant biomass and then exported in other forms. However, preliminary experiments were conducted to investigate the possible use of aerated ponds to convert Fe(II)(aq) in the effluent from these systems to Fe(III)-oxide flocs that can settle out of the water column and have a high capacity for adsorption of phosphorus. This presents a relatively simple method for increasing phosphorus removal and recovery from these systems.Overall, our work identifies horizontal levees as a promising alternative to conventional treatment systems for the removal of nitrogen and other contaminants from wastewater effluent. Our findings provide us with a better understanding of the impact of design and operational parameters on contaminant removal in these systems and can be used to inform the design of future systems. Further work is needed to test the ability of these systems to treat a variety of additional contaminants (e.g., trace metals), to characterize removal mechanisms for many contaminants (e.g., trace organic contaminants), and to understand the impact of alternative water matrices (i.e., reverse osmosis concentrate) on contaminant removal in these systems.