Evaluating deep subsidence in a rapidly-accreting mangrove forest using GPS monitoring of surface-elevation benchmarks and sedimentary records

Ultimately, the fate of mangrove forests depends on substrate-elevation gains keeping pace with relative sea level rise (RSLR). Some of the world's largest mangrove forests occur on tectonically active margins, river deltas and sedimentary basins where sea-level trends are largely controlled by vertical land motion (i.e., subsidence), so that RSLR can be markedly higher than sea-level rise induced by climate warming. The Rod Surface Elevation Table - Marker Horizon (RSET-MH) method has been applied globally to evaluate coastal-wetland resilience to SLR by quantifying net rates of elevation change relative to a benchmark. A limitation of the RSET method is that the stability of the benchmark (i.e., vertical trend) is unknown and RSLR in wetlands is typically inferred from regional tide-gauge records. In the present study, we evaluate RSLR in a rapidly-subsiding Avicenna marina mangrove forest with a large terrigenous sediment supply (Firth of Thames, New Zealand) using independent and complimentary methods: (1) campaign-GPS surveys of the stability of three RSET benchmarks driven 18 m into unconsolidated sediment, a tide gauge some 10 km distant and a reference station located on basement rock (2007-2016) that are tied to a network of satellite-based geodetic sites; and (2) Cs-137-validated Pb-210 sediment accumulation rates (SAR) measured in cores as a proxy for RSLR. The similarity of deep-subsidence rates at the RSET benchmarks located several-hundred metres apart (7.7 +/- 0.5 to 9.4 +/- 0.5 mm yr(-1)) provides confidence in these results. These subsidence rates are two- to five-fold higher than recorded at the nearby tide gauge (3.6 +/- 0.7 mm yr(-1)) and reference station (1.6 +/- 0.5 mm yr(-1)). Weighted-average Pb-210 SAR of 9.9 mm and 93 mm yr(-1) yield similar estimates of deep subsidence (i.e., 8.4 and 6.9 mm yr(-1) + SLR adjusted for vertical land motion), indicating that this geological, process is the major driver of the long-term sea-level trend in the mangrove forest. Our findings suggest that regional tide gauge records may not provide reliable estimates of RSLR for all coastal wetlands due to local subsidence associated with natural processes and human activities. Subsidence rates and RSLR within coastal wetlands can be evaluated using campaign-GPS surveys of RSET benchmarks and independently from longer-term sedimentary records. Finally, evaluations of the resilience of coastal wetlands to rising sea levels over the coming decades must be based on measurements of RSLR from these systems themselves. (C) 2016 Elsevier B.V. All rights reserved.