Using salt marshes for coastal protection: Effective but hard to get where needed most

1. Salt marshes fronting coastal structures, such as seawalls and dikes, may offer important ecosystem-based coastal defence by reducing the wave loading and run-up levels during storms. We question (i) how the long-term salt marsh development in the Dutch Wadden Sea relates to the tidal-flat foreshore bathymetry and (ii) how the wave run-up onto dikes, which enhances the risk of dike failure, depends on foreshore bathymetry, the presence/absence of marshes, marsh vegetation properties, tidal range and wind exposure. 2. We analysed 15 years of vegetation and bathymetry maps along the entire Dutch Wadden Sea coast, in combination with detailed process-based measurements at five locations during 3 years, to understand where salt marshes naturally form and what features determine their contribution to coastal protection. 3. The horizontal extent of marshes along the dikes remained relatively stable over the past decade. The presence of marshes was associated with higher elevations of adjacent tidal flats (above ~0.5 m NAP), while landward-directed marsh retreat was associated with surface erosion of the fronting tidal flats. 4. Wave run-up during storms was lower at sites with wider marshes and higher foreshore elevations. This was attributed to the marsh attenuation effect, which led to a reduction in wave heights at the dike toe. As the tidal range varies across the Dutch Wadden Sea, areas to the East with generally higher water levels experienced higher wave run-up. 5. Synthesis and applications. We found that (i) marshes, where present, effectively protected the dikes from wave loading and (ii) the sites where marshes typically do not develop spontaneously were the most vulnerable to high wave run-up. This catch-22 problem implies that increasing reliance on nature-based coastal defences along soft-bottom coasts may require human interventions to stimulate marsh formation at the locations where it is most needed. Alternatively, ‘hard engineering’ solutions m