Tidal Wetlands and Estuaries

The CCARS synthesis report (Benway et al., 2016) is the most comprehensive attempt to develop a science plan for carbon cycle research of North American coastal systems. While clarifying key regional differences in processes and projections, this synthesis effort also exposed major knowledge gaps and disconnects between measurement and modeling scales. These knowledge gaps are currently being explored by multiple synthesis efforts, and below is a review of some of the major gaps being investigated.

15.7.1 Lateral Exchanges Between Tidal Wetlands and Estuaries

Estimates of lateral fluxes of carbon between tidal wetlands and estuaries are mostly based on discrete sampling events at monthly to seasonal intervals, with sampling resolution from hourly to one ­half of a tidal cycle, leaving the majority of time unsampled and thus requiring large interpolation between sampling events and producing substantial uncertainty in export fluxes (Downing et al., 2009; Ganju et al., 2012). A recent estimate of the DIC lateral flux from a pristine intertidal wetland marsh on Cape Cod, Massachusetts, with minute-scale resolution revealed that previous estimates of marsh DIC export—such as those summarized by Najjar et al. (2018) and used here—may be severalfold too low (Wang et al., 2016). Previous studies generally show a positive carbon export from tidal wetlands to estuaries but may not fully resolve the export magnitude and temporal heterogeneity, which, in turn, are controlled by variability in water flux and constituent concentration across timescales from minutes to tidal cycles to years. Such observational gaps extend beyond DIC to include DOC and particulate organic carbon (POC) as well. In particular, the fate of exported POC from eroding marshes, though virtually unknown, is important for carbon accounting. Future studies should be directed to capture appropriate temporal scales of variability of carbon exports from marshes to accurately constrain lateral exchanges.

15.7.2 Coastal Subhabitat Boundaries

The definition of estuarine subhabitat within the coastal ocean is fluid, primarily associated with bottom depth and mixing processes. This boundary may not be mappable, but the absence of a robust definition inhibits future monitoring efforts and projections. Progress has been made in defining estuaries and quantifying their fundamental characteristics (such as residence time) in CONUS via NOAA’s Coastal Assessment Framework (NOAA 2017). Such a framework has been essential for scaling up carbon and nitrogen fluxes from limited data (Herrmann et al., 2015; Najjar et al., 2018) and is greatly needed for all of North America. The global estuarine delineation based on MARCATS (Project Geocarbon 2017) has been very helpful, but the coarse resolution (i.e., 0.5 degrees) is a concern. For coastal wetland boundaries, multiple criteria have been used by different entities: political boundaries, salinity gradients, elevation thresholds, and tidal criteria. This variability has led to great confusion in the literature (e.g., Lu et al., 2017), in agency policies, and in market-based carbon accounting protocols. A strong gap is the lack of a boundary mapped for head of tide. Tidal wetlands, by definition, cross a wide range of salinities (i.e., saline, brackish, and freshwater), with the singular distinction of having a hydroperiod influenced by ocean tides (paraphrased from web link; U.S. EPA 2016). Networks of available data may be useful in monitoring this boundary, as it is a key distinction of carbon dynamics in coastal habitats. These networks include, for example, a NOAA repository of coastal LIDAR; NOAA tide gauge networks; USFWS wetland mapping efforts; and USGS Land Change Monitoring, Assessment, and Prevention (LCMAP; USGS 2017). In the absence of a mapped boundary, spatial accounting of tidal and estuarine extent—current, past, and future—is fraught with uncertainty, with a likely underestimate of at least 50% for freshwater tidal wetlands alone.

15.7.3 Spatial Variability in Burial Rates and in Air-Water Flux

Because of ocean influences and similar processes along coastlines, spatial variability can be much greater within an estuarine and tidal wetland complex than among regions. Tracking the drivers of spatial variability in ecosystem properties—sea level, bathymetry, river flow, elevation, soil properties, and vegetation types—can greatly improve the use of remotely sensed data to validate carbon flux models and their variability between years. Accounting processes generally rely on spatial data, and mapping stocks and fluxes in these spatially dynamic habitats will require improved use of geospatial datasets and, thus, improved attribution of location information with observations. Relative sea level rise is particularly variable in its magnitude and influence. Geomorphic models (e.g., Kirwan and Megonigal 2013; Morris et al., 2016) are improving understanding of the sustainability of wetland carbon storage, showing enhanced carbon sequestration under modest increases in sea level but rapid carbon emissions after wetland accretion reaches its conditional “tipping point.” Empirically, many GMx wetlands undergoing land subsidence appear to have crossed their threshold of sustainability and are being rapidly eroded or drowned (Couvillion et al., 2017).

15.7.4 Other Greenhouse Gases: CH₄ and N₂O

The bulk of data on CH₄ and N₂O fluxes in tidal wetlands is modeled from pore-water ­measurements in profile or from atmospheric chamber measurements under static conditions. However, these methods generate an incomplete picture of these dynamic environments and fluid boundaries. The growing network of eddy covariance and other continuous data-rich approaches (“movies” instead of “snapshots”) is improving the understanding of the episodic nature of these processes and emergent thresholds of concern. Nitrous oxide fluxes likely are heightened under enhanced nitrate runoff (i.e., “nitrate saturation”; Firestone and Davidson 1989), but documentation is poor. Further, CH₄ production is likely low when sulfate is available (Poffenbarger et al., 2011), but it is enhanced by increased carbon fixation, such as through global changes that include rising atmospheric CO₂ concentrations or invasions of more productive species (e.g., Phragmites australis; Martin and Moseman-Valtierra 2015; Mueller et al., 2016).

Estuarine CH₄ emissions currently appear to be a small fraction of global emissions (i.e., <1%; Borges and Abril 2011), but they may be poised to increase with enhanced rates of methanogenesis in response to organic matter inputs and hypoxia expansion under future conditions (Gelesh et al., 2016). A seaward decrease in near-surface porewater concentrations of CH₄ is observed often, likely due to both increasing sulfate availability and in situ water column oxidation. Water column CH₄ and pCO₂ are positively correlated in well-mixed estuaries, suggesting in situ production from organic matter transferred from surface waters to ­methane-producing bottom waters (Borges and Abril 2011). Like tidal wetlands, many estimates of emission rates are modeled from profiles of surface and porewater concentrations of CH₄, but continuous sampling and eddy covariance data likely will reduce uncertainty in emissions and allow better characterization of the physical and biogeochemical processes associated with atmospheric CH₄ emissions.

15.7.5 Regional Gaps

Much assessment has been focused on estuaries along different regions of the Atlantic Coast (e.g., GOM, MAB, and SAB), but modeled carbon fluxes for large estuaries still remain poorly constrained. For example, few measurements of air-water CO₂ flux are available for upscaling within the Chesapeake Bay, the largest East Coast estuary (e.g., Cai et al., 2017).

The Gulf of Mexico also is well studied, but it has surprisingly few gas flux measurements in its tidal wetlands and estuaries (see, however, Holm et al., 2016). One of the most extensive regional monitoring programs, Louisiana’s Coastwide Reference Monitoring System (CRMS 2017), supports GMx soil and vegetation stock change assessments and predictive models through annual records of tidal wetland conditions. These data also help illustrate the wide within-watershed variability in conditions, such as land subsidence (Jankowski et al., 2017), that drive organic carbon accretion, erosion, and mineralization processes. In addition, the Texas Commission on Environmental Quality (TCEQ) has been maintaining quarterly measurements of total alkalinity and pH in all coastal estuaries across the state in the northwestern GMx since 1969 (TCEQ 2017). This dataset may offer insight on multidecadal changes in CO₂ flux that await further investigation.

In contrast, Pacific Coast estuaries lack published carbon cycle measurements with sufficient resolution and duration to afford insight into short- or long-term changes associated with climate or human-caused forcing. Observation and modeling gaps are notably large in the Gulf of Alaska and Central American isthmus regions. For instance, very few studies have addressed CO₂ cycling and air-sea exchange in lagoons (Ávila-López et al., 2017), a dominant habitat type in the tropical Pacific and the Gulf of California in Mexico. Estimates of air-sea exchange of climate-reactive gases (e.g., CO₂, CH₄, and N₂O) in open waters of Pacific Coast estuaries, along with estimates of primary production and carbon burial, are insufficient for a systematic analysis.

Finally, high-latitude estuaries are experiencing rapid shifts in salinity and seasonality, making relationships between climatic drivers difficult to assess. Some clear data needs for a monitoring framework in Arctic systems include depths of coastal peats along rivers, the sensitivity of productivity to rising temperatures and longer growing seasons, terrestrial carbon fluxes (including DOC and DIC), and the long-term prognosis for coastal erosion rates due to relative sea level rise.

Carbon stock and flux data from Pacific Islands, Puerto Rico, and Hawai‘i are not included in this chapter because of their limited datasets (Fagan and MacKenzie 2007; MacKenzie et al., 2012) and the inability to extrapolate their data in space and time. Emerging carbon assessments may be useful for upscaling (Selmants et al., 2017), but the necessary measurements are lacking to estimate carbon fluxes of similar confidence as reported herein for continental coastlines. Hence, there is a clear need for studies of carbon cycling in the coastal environments of Pacific Islands, Puerto Rico, and Hawai‘i.

15.7.6 Outlook and Conclusion

Current outlooks and understanding of tidal wetland and estuarine carbon cycling are represented herein, recognizing that synthetic and novel research activities are ongoing. The current state of knowledge represented is sufficient to identify predictable processes and responses, but uncertainty in modeling is higher when applied at continental scales and across datasets of varied confidence. Whereas coastal habitats have distinct responses to myriad global changes, regional and temporal drivers of carbon exchanges and internal processing remain critical knowledge gaps. Monitoring advances, such as high-frequency field data, remotely sensed imagery, and data integration platforms, may shed light on the carbon dynamics at the land-ocean margin and provide the clarity needed to close continental-scale carbon budgets. Improved confidence in projected changes of coastal carbon storage and processing is needed for contributing to more effective policy and management decisions in coastal communities and nationally within North America.