Terrestrial Wetlands

13.7.1 Summary of Terrestrial Wetlands Carbon Cycling

North American wetlands constitute a significant proportion (37%) of the global wetland area. The uncertainty in wetland area for North America is relatively low because wetlands in CONUS and Alaska, Mexico, and Canada have relatively recent inventories and assessments. However, more information about soil carbon and vegetation biomass within the wetlands is needed to assess carbon pools and fluxes and reduce uncertainties in the estimates. Wetland soil type varies significantly with latitude, with Alaska and Canada having the majority of the peatland area. Mineral soil wetlands are predominant (79%) in CONUS and contain 38% of its wetland carbon stock. An important consideration regarding the estimate of carbon pools in peatlands, which consist of 58% of the North American wetland area, is that total depth of peat is seldomly reported, while the average depth commonly exceeds the typical assessment depths of 1 to 2 m. Peatlands contain approximately 80% of the North American carbon, a proportion that is likely to increase substantially if the entire peat depth were considered. Nonforested vegetation communities compose 44% of the wetland area in North America, contain approximately 43% of the carbon pool, and accumulate 47% of the net carbon gain.

Historically, the wetland loss in North America has been significant, particularly in CONUS. However, to assess contemporary losses, periodic inventories at the national scale are needed. Currently, only the United States has regular updates to its wetlands inventory. Restoration and creation of new wetlands are major offsets to loss of natural U.S. wetlands. Whether these new wetlands have the same carbon dynamics as natural wetlands is a major uncertainty that will become more important as restored wetlands become a larger proportion of the total wetland area. A global meta-analysis comparing 621 restored and created wetlands to 556 reference wetlands indicated that functions related to biogeochemical cycling (mainly to carbon storage) were 23% lower in the restored and created wetlands (Moreno-Mateos et al., 2012). Successful functioning of those wetlands will be critical to mitigate the long-term losses of carbon from degraded wetlands.

13.7.2 Knowledge Gaps and Associated Uncertainties in the Wetland Carbon Cycle

The following are some major gaps in current knowledge about the North American wetland carbon cycle.

1. Future wetland response to climate change is uncertain. Because temperatures are predicted to increase at greater rates at higher latitudes, northern temperate wetlands, especially peatlands, are expected to be the most affected. More uncertainty exists in the predictions of precipitation, changes in which could either mitigate or exacerbate carbon sequestration rates in terrestrial wetlands. Although contemporary measurements and modeling offer perspective, additional manipulative experiments—such as the U.S. Department of Energy’s Spruce and Peatland Responses Under Changing Environments (SPRUCE) experiment in northern Minnesota (Hanson et al., 2017) and USDA’s former PEATcosm experiment in the Upper Peninsula of Michigan (Potvin et al., 2015)—are critical to test how wetlands will respond to changes in temperature and hydrological regime in the field. Work in mineral soil wetlands is particularly needed because of the paucity of studies and the functional linkages with aquatic systems.

2. Greater understanding is needed of the factors controlling carbon cycling in wetlands. Additional measurements of GHG fluxes and processes regulating the fluxes and carbon storage using improved inventories and methods at multiple spatial scales are required to 1) understand the interactions of soil, vegetation, and climatic factors; 2) provide a basis for quantifying fluxes to reduce significant uncertainties; and 3) evaluate biogeochemical and inverse-atmospheric models. Particularly needed are studies that assess convergence across diverse spatial and temporal scales or lead to a process-based understanding of why convergence does not occur.

3. Dissolved carbon export, including both DIC and DOC, is a major uncertainty in the wetland carbon cycle. Dissolved carbon affects water quality and is an important food source for aquatic systems and estuaries, and dissolved gases may contribute to atmospheric loading. Understanding the mechanisms controlling dissolved carbon production and transformation is a major gap requiring field and watershed-scale assessments.

4. A better understanding is needed of the relationship between the sustainability of stored carbon and the particular chemistry of the carbon compounds that make up the carbon sink. Preliminary research shows that polyphenol content may serve to preserve peats under moderate drought conditions (Wang et al., 2015), but little is known about either the exact types of polyphenols or the plant communities that have the highest sustainability under projected climate and environmental conditions.

5. Data on restored and managed wetlands are sparse and insufficient to support assessment and modeling needs. Measurements to document the carbon balance in these wetlands are needed. Also necessary are standardized measurements and methods for collecting basic data in the field at the same depth and for analyzing parameters such as bulk density and percent of organic carbon. Monitoring of wetland restoration needs to extend through the entire trajectory of the project to gain a functional understanding of the differences in gaseous fluxes and carbon accumulation between natural and restored wetlands.

13.7.3 Tools for Assessing the Wetland Carbon Cycle

Due to the extremely wide variation in wetlands across North America, as well as the certainty that there will never be enough measurements to adequately quantify the wetland carbon stocks and fluxes, models present the means to represent the biophysical processes inherent to wetlands at variable spatial scales. Those tools provide needed capabilities to inform conservation, management, and mitigation strategies to sustain ecosystem services inherently linked to the wetland and global carbon cycle. Models also are useful for addressing the uncertainties within the carbon cycle and, in turn, for focusing field monitoring and experiments to fill critical information gaps. Mechanistic models provide the capabilities for simulating the processes that regulate carbon dynamics in wetlands reflecting the myriad soil, vegetation, and climatic conditions and management influences. Because of the water table’s regulatory function in the wetland carbon cycle, an accurate representation of wetland hydrology is critical to model performance. There are fewer models for wetlands compared to those for uplands. Among biogeochemical models that are widely applicable to terrestrial wetlands and have the broadest capabilities with respect to soil and vegetation types are the Forest DNDC (or DeNitrification DeComposition) model, which was identified by USDA in the development of its carbon accounting framework (Ogle et al., 2014), and the DayCent model (Parton et al., 1998), which is widely used in grassland and agroecosystem simulations. Scaling wetland hydrology within a biogeochemical model is difficult; hence, coupling a biogeochemical model with a hydrological model can provide an effective basis for considering the inherent spatial variability among uplands and wetlands (Dai et al., 2012a). Simulating CH₄ fluxes is particularly difficult because of various interactions among controls of CH₄ production and transport from wetlands, including ebullition, that vary over very short distances such as 10 m or less (Bridgham et al., 2013). Correspondingly, uncertainties associated with plant carbon allocation and organic matter quality and decomposition impair the ability of field-scale biogeochemical models to predict CH₄ flux from the soil surface. These considerations are particularly important for small-scale models that are evaluated with field data.

Another major challenge to modeling carbon dynamics in wetlands is the inherent heterogeneity of conditions within a wetland and the spatial heterogeneity of wetlands across the landscape. Accordingly, new approaches for accommodating high-resolution geospatial data with robust biogeochemical models are needed to provide capabilities to simulate wetland carbon dynamics at large scales. Such capabilities, in turn, would provide a basis for linking wetland biogeochemical models with atmospheric models (Gockede et al., 2010), thereby improving the basis for simulating the effects of climate change on wetland carbon. Large-scale bottom-up and top-down models are providing those capabilities to address CH₄ fluxes at the regional and global scales (Melton et al., 2013; Saunois et al., 2016; Bloom et al. 2017; Zhang et al., 2017a). However, estimates among the CH₄ models can vary considerably (Miller et al., 2016). Correspondingly, there is a real need for tools to assess wetland NEE; unfortunately, the large-scale models for assessing wetland NEE are not available or widely reported. Accordingly, ecosystem models must be upscaled to develop the components to simulate wetland NEE.