Refractory organic matter in coastal salt marshes-effect on C sequestration calculations

Highlights: • We studied how refractory carbon inputs affect marsh age and C sequestration budgets. • We measured Δ¹⁴C and δ¹³C of total organic carbon (TOC) and refractory carbon (C_RF). • TOC was dominated by autochthonous inputs, C_RF was dominated by allochthonous C. • Allochthonous C delivery was controlled by the size and slope of each watershed. • Steep-gradient rivers delivered Δ¹⁴C-depleted C_RF to their estuarine marshes. The age and ability of salt marshes to accumulate and sequester carbon is often assessed using the carbon isotopic signatures (Δ¹⁴C and δ¹³C) of sedimentary organic matter. However, transfers of allochthonous refractory carbon (C_RF) from the watershed to marshes would not represent new C sequestration. To better understand how refractory carbon (C_RF) inputs affect assessments of marsh age and C sequestration, Δ¹⁴C and δ¹³C of both total organic carbon (TOC), C_RF, and non-C_RF organic matter fractions were measured in salt marshes from four contrasting systems on the North Atlantic coast. To our knowledge, no salt marsh sediment study has considered refractory or allochthonous carbon in carbon budget calculations or the impact on chronologies. Stable and radiogenic isotope data suggest that while TOC was dominated by autochthonous plant inputs, C_RF was dominated by locally recycled or allochthonous C, the delivery of which was controlled by the size and slope of each watershed. Steep-gradient rivers analyzed delivered Δ¹⁴C-depleted C_RF to their estuarine marshes, while the site located in the low-gradient river was associated with larger C_RF content. Finally, the marsh isolated from riverine input contained the least fraction of TOC as C_RF. Laterally transported C_RF caused only a small offset in Δ¹⁴C in relation to TOC in low-gradient systems (average Δ¹⁴C offset was −44.4 and −24.2‰ at each location). However, the presence of allochthonous Δ¹⁴C-depleted C_RF in sediments of steep-gradient rivers led to large overestimates of the time of organic matter deposition (i.e. apparent age was older than the 'true' time of deposition) (Δ¹⁴C offset ranged from −170.6 to −528.9‰). Further, reliance on TOC or loss on ignition analyses to calculate C sequestration by marshes might produce overestimates of at least as much as 10 to 20% since neither account for the lateral transport of allochthonous carbon.