The Response of Foraminifera to the Tidal Restoration of the Bandon Marsh (Oregon) – Implications for Subsidence Estimates During Past Earthquakes
Stratigraphic sequences beneath U.S. Pacific Northwest salt marshes preserve a 7000-year record of plate-boundary earthquakes along the Cascadia subduction zone. The salt marshes record a rapid relative sea-level (RSL) rise
Stratigraphic sequences beneath U.S. Pacific Northwest salt marshes preserve a 7000-year record of plate-boundary earthquakes along the Cascadia subduction zone. The salt marshes record a rapid relative sea-level (RSL) rise during regional coseismic subsidence followed by gradual RSL fall during interseismic uplift. These changes in RSL have been quantified using foraminiferal transfer functions assuming that foraminifera rapidly recolonize salt marshes and adjacent tidal flats following the earthquake.
The tidal restoration of the Bandon marsh (Oregon) that started in 2011 provides a natural experiment to study the re-colonization of foraminifera during rapid “coseismic” (simulated by dike removal that allowed sudden tidal flooding) and “interseismic” (stabilization of the marsh following flooding) RSL change. The tidal restoration also allows us to evaluate the accuracy of our foraminiferal transfer function. We analyzed surface samples from eight stations at the restoration site prior to and during the first five years of tidal restoration and collected a 50-cm-long core from station (St.) 1 in 2016. The tide gauge, salinity and grain-size data show rapid tidal restoration within days but first high numbers of foraminifera of >3500 per 10 cm3 sediment volume were observed ~ten months after tidal restoration at stations with elevation lower than ~0.60 m mean tide level (MTL). The measured elevations agree with transfer function estimations at some stations with established stable foraminiferal assemblages (e.g., for St. 1 predicted elevation is 0.33 m (± 0.14 m) MTL and measured elevation is 0.30 m MTL), but at some low elevation stations with stable assemblages, the transfer function under-predict (e.g., for St. 3 predicted elevation is 0.31 m (± 0.14 m) MTL but measured elevation is 0.57 m MTL). The application of the transfer function to the core estimated a RSL rise of 0.93 m (± 0.21 m) from the pre-restoration peat to the post-restoration mud, which matches the measured increase in tide gauge RSL of 0.97 m. Our results provide further evidence that our transfer function provides accurate predictions except for the lower end of the elevational gradient due to the dominance of the single species Miliammina fusca.