Contents of this file: Table S1. List of CMIP6 GCMs and the availability of runoff, salinity, and SST in historical simulations and projections four SSPs, covering the period of 1850-2100. -- Figure S1. Results from the diagnostics of the overimputation in the PCA. The overimputed values for the different measured biogeochemical properties show that they are in acceptable range (i.e., values are close to the diagnostic slope). The data included in the PCA were organic carbon (oc), C:N-ratio (cn) and stable isotopes of C (d13c) and N (d15n), XRF-elements (Cl, Br, Zr, Sr, Ca, Rb, Fe, K, Mn, and Ti), n-alkanes (average chain length, MCL), mean sediment grain-size distribution (MEAN) and sorting (SORTING), and magnetic susceptibility (MS_x). -- Figure S2. Sediment age models based on cumulative mass vs. combined 210Pb and 14C ages obtained with Bacon for (A) Gåsö S and (B) Gåsö N. The solid lines show the best “fitted” model based on 210Pb-derived ages and calibrated 14C-dates and the dashed lines are 2 σ range (equal to 0.95). Note the difference in scale on the axes. In the above panels, iteration is the number of Markov Chain Monte Carlo (MCMC) runs that was performed to assess the accumulation rate and produce the age model, acc. rate is the gamma distribution of the accumulation rate and memory shows the variability of the accumulation rate before (green curve) and after (grey box) the model was run. -- Table S2. Radiocarbon dates used for estimating sediment age models in Gåsö S and N. AMS = Accelerator Mass Spectrometry. BP = Before present. -- Figure S3. Mean (±SD) n-alkane (µg g dry weight sediment-1) signature of Zostera marina above- and belowground biomass based on analysis of three replicates for each of the above- and belowground plant parts.

Peer reviewed