摘要

The introduction of invasive Spartina grasses (C-4 plant) into previously C-3-dominated saltmarshes of western Europe complicates the reconstruction of past relative sea-level (RSL) using sediment geochemistry-based approaches. We apply a Bayesian mixing model (SIAR - Stable Isotope Analysis in R) to identify organic carbon (C-org) source contributions to saltmarsh sediment and correct delta C-13 values for the contaminating influence of C-4 vegetation. We first assess the performance of SIAR by simulation, producing synthetic saltmarsh sediment samples with variable contributions from three organic matter sources (C-3 plants, C-4 plants and particulate organic matter (POM)) and then comparing these known contributions to those determined by SIAR. For each source type, SIAR 68% credible intervals include true source contribution in 70% of calculations (n = 60) and all three sources were incorporated within credible intervals for 45% of sediment samples (n = 20). SIAR improves the mean absolute error of source contribution predictions by up to 26% compared to a conventional ternary mixing model approach. We then apply SIAR to real sediment samples recovered along transects spanning the intertidal zones at two contrasting saltmarshes in the inner and outer Shannon estuary, western Ireland. SIAR identified decreasing Corg contributions from POM with increasing elevation in the intertidal zone, reflecting differences in tidal inundation with height. Sediment delta C-13 values across the intertidal zones varied by 4.4 parts per thousand in the inner estuary and 9.4 parts per thousand in the outer estuary. Removal of the invasive C-4 signatures reduced these differences to 0.5 parts per thousand and to 2.4 parts per thousand in the inner and outer estuary, respectively, with the larger range due to increased deposition of marine POM. Although subtle, when used in combination with additional proxies, these geochemical gradients can assist in RSL reconstruction. This will be most effective when POM from marine sources is a significant carbon source.

  • 出版日期2017-10