Introduction
Coastal wetland ecosystems are some of the most efficient ecosystems for carbon sequestration (Mcleod et al., 2011). At the interface of land and ocean, these ecosystems are characterized by mixing of water, nutrients, and sediment from watersheds and nearshore marine waters. Productive plant communities occupy the continuum from upstream to estuary mouths, despite the dynamic gradients in physicochemical conditions. Dense vegetation slows water velocity, allowing sediment accretion that further buries organic matter produced or trapped in coastal wetlands. In addition, soils are often anoxic, slowing the decomposition of buried organic matter and allowing soil organic matter to accumulate. Land surface models (LSMs) are useful for examining carbon cycling at ecosystem scales and are critical for estimating carbon cycling in global scale Earth system model simulations. However, despite the importance of coastal wetlands in carbon storage, and their vulnerability to sea level rise, LSMs have limited representation of coastal ecosystems (O’Meara et al., 2021; Ward et al., 2020).
Plant functional types (PFTs) developed for terrestrial vegetation are inadequate for coastal wetlands because the controls on vegetation productivity differ between these ecosystems (LaFond-Hudson and Sulman, 2023). Wetland plants are sometimes exposed to soil moisture, pH, and nutrient concentrations rarely experienced by terrestrial plants, so wetland vegetation traits sometimes stray from the predictable patterns of resource tradeoffs observed in terrestrial plants (Moor et al. 2017). Models sometimes lack critical for processes in wetlands if the processes are less influential in terrestrial ecosystems. For example, in the Energy Exascale Earth System Model (E3SM)’s land surface model (ELM), vegetation productivity is limited by low soil moisture, but not by saturated soils (Oleson et al. 2013), so the model likely overestimates productivity in coastal wetlands. The model also does not represent the impact of salt on soil water potential and photosynthesis. These missing processes lead to uncertainty in global carbon cycle projections; improving these processes will open opportunities to explore carbon cycle questions within coastal wetland ecosystems using LSMs.
Inundation decreases oxygen availability to roots under saturated soil conditions and by covering stems and leaves under flooded conditions (Colmer and Vosenek 2009). Plant adaptations to flooding can include aerenchymous tissue that transports oxygen to the root zone, increased vertical growth under flooded conditions, and floating tissues for plants adapted to permanently flooded conditions. Despite the abundant water in coastal wetlands, salinity affects plants by lowering the soil water potential outside of roots, creating challenges for water uptake. Plants may respond by limiting water uptake, or by taking up salt along with water and accumulating salt in tissues (Munns and Tester, 2008). Plants that limit water uptake must conserve water; this strategy requires lowering stomatal conductance and carbon uptake. Plants that accumulate salt have adaptations to alleviate ion toxicity within tissues, such as storage in vacuoles and production of osmolytes to maintain osmotic potential of cells. Although these adaptations are better for chronic salt exposure, these adaptations are energetically costly.
The distribution of plant communities is tightly coupled to average salinity concentrations and hydroperiod, but the physicochemical conditions of coastal wetlands are dynamic. Additionally, sea level rise, droughts, and more frequent, intense storms are likely to further vary water levels and salinity in coastal wetlands. Land surface models will be more capable of exploring carbon cycling in coastal wetlands if vegetation dynamics can be accurately simulated at the salinity and hydroperiod for which vegetation is well adapted, as well as the dynamics when vegetation is exposed to conditions for which it is poorly adapted. For example, much of the research on salt marshes in the past two decades concerns the fragility of salt marshes in the face of accelerating sea level rise. Some evidence suggests they are resilient to moderate increases in the duration of inundation via enhanced vegetation productivity that contributes to elevation gain (Kirwan et al., 2010; Morris et al., 2002). However, in some locations, marshes are eroding because local sea level rise rates are outpacing marsh elevation gain (Wasson et al., 2019). At the same time, saltwater intrusion into ecosystems that are not adapted to saline conditions can cause widespread mortality of vegetation and consequent loss of carbon stocks. Incorporating vegetation responses to salinity and inundation into LSMs is important for simulating the impacts of seasonal patterns and long-term trends in water levels and salinity on vegetation productivity in coastal ecosystems. In this paper, we describe a new capability of the DOE’s Energy Exascale Earth system model (E3SM) land model (ELM) to represent vegetation responses to salinity and inundation. We show how the vegetation responses to salinity and inundation improve model accuracy in a salt marsh ecosystem. Several scenarios are then presented to demonstrate how a salt marsh might respond to systemic changes in salinity and water level, as might occur in departures from average annual precipitation.