Modeling net ecosystem carbon balance and loss in coastal wetlands exposed to sea-level rise and saltwater intrusion

Abstract

Coastal wetlands are globally important stores of carbon (C). However, accelerated sea-level rise (SLR), increased saltwater intrusion, and modified freshwater discharge can contribute to the collapse of peat marshes, converting coastal peatlands into open water. Applying results from multiple experiments from sawgrass (Cladium jamaicense)-dominated freshwater and brackish water marshes in the Florida Coastal Everglades, we developed a system-level mechanistic peat elevation model (EvPEM). We applied the model to simulate net ecosystem C balance (NECB) and peat elevation in response to elevated salinity under inundation and drought exposure. Using a mass C balance approach, we estimated net gain in C and corresponding export of aquatic fluxes ( $F_{\mathrm{AQ}}$ ) in the freshwater marsh under ambient conditions (NECB = 1119 ± 229 gC m^-2 year^-1 ; F_AQ = 317 ± 186 gC m^-2 year^-1 ). In contrast, the brackish water marsh exhibited substantial peat loss and aquatic C export with ambient (NECB = -366 ± 15 gC m^-2 year^-1 ; F_AQ = 311 ± 30 gC m^-2 year^-1 ) and elevated salinity (NECB = -594 ± 94 gC m^-2 year^-1 ; F_AQ = 729 ± 142 gC m^-2 year^-1 ) under extended exposed conditions. Further, mass balance suggests a considerable decline in soil C and corresponding elevation loss with elevated salinity and seasonal dry-down. Applying EvPEM, we developed critical marsh net primary productivity (NPP) thresholds as a function of salinity to simulate accumulating, steady-state, and collapsing peat elevations. The optimization showed that ~150-1070 gC m^-2 year^-1 NPP could support a stable peat elevation (elevation change ≈ SLR), with the corresponding salinity ranging from 1 to 20 ppt under increasing inundation levels. The C budgeting and modeling illustrate the impacts of saltwater intrusion, inundation, and seasonal dry-down and reduce uncertainties in understanding the fate of coastal peat wetlands with SLR and freshwater restoration. The modeling results provide management targets for hydrologic restoration based on the ecological conditions needed to reduce the vulnerability of the Everglades' peat marshes to collapse. The approach can be extended to other coastal peatlands to quantify C loss and improve understanding of the influence of the biological controls on wetland C storage changes for coastal management.

Keywords: elevation change; net ecosystem C balance; peat collapse; saltwater intrusion; sea-level rise; wetland vulnerability.

FIGURE 1

Work flow diagram illustrating how the collected experimental data at varying levels of salinity and inundations were utilized to compute net ecosystem carbon balance and develop mechanistic modeling framework for parameterization and simulations of peat stock and peat elevation change.

FIGURE 2

Conceptual schematic showing balancing of ecosystem carbon (C) budget. The net residual C flux (F _R ) is estimated by equating (a) land–water–atmospheric flux and (b) ecosystem C pool–based C balance approaches. F _R represents remaining components of C budget such as the fraction of F _AQ (e.g., dissolved inorganic carbon, dissolved organic carbon, particulate organic carbon), and F _PC that are not incorporated in approach (b).

FIGURE 3

Diagram of system dynamics Everglades peat elevation model (EvPEM) used to simulate change in peat stock ($M_S$) and peat elevation ($\mathrm{PE}$). Adj, AG, ANPP, BG, BNPP, NECB, and SLR refer to adjusted, aboveground, aboveground net primary productivity, belowground, belowground net primary productivity, net ecosystem carbon balance, and sea‐level rise, respectively.

FIGURE 4

Decline in sawgrass (Cladium jamaicense) aboveground net primary productivity (ANPP) and belowground net primary productivity (BNPP) with increasing porewater salinity for submerged (SUB), exposed (EXP), and extended depth of exposure of peat surface (EXTEXP) hydrologic treatments.

FIGURE 5

Everglades peat elevation model calibration plot showing comparisons between mean ± 1 standard error experimental observed and simulated changes in (a) peat elevation and (b) net ecosystem carbon balance (NECB) over 1‐year period across the seven treatments representing freshwater (FW) and brackish water (BW) marshes. SALT = elevated porewater salinity of 20 ppt, AMB = 0.5 ppt for FW and 10 ppt for BW. SUB, EXP, and EXTEXP respectively refer to submerged, exposed, and extended exposed hydrologic treatments for different salinity levels. Minus sign refers to loss of elevation and NECB.

FIGURE 6

Simulated peat elevations (PEs) in sawgrass marsh under mean global 3 mm year⁻¹ sea‐level rise showing examples of accumulating, steady‐state (no change), and collapsing peats in response to annual net primary productivity (NPP), porewater salinity, and hydrology. The inset shows different threshold levels of NPP as a function of salinity used to simulate the corresponding PEs. Ratios of above/below primary productivity, turnover rates, and soil bulk density remained constant during the simulation period. NPP represents the sum of aboveground and belowground net primary productivity. Water level (WL) at the beginning of the simulation (year 1) includes seasonal variability with 8 months of peat soil surface submergence, 3 months of moderately exposed peat soil surface, and 1 month of highly exposed peat soil surface. The WLs of the remaining simulated years were subject to seasonal variability in addition to the overall 3‐mm rise in each year.