Terrestrial Wetlands

13.6.1 Global and Continental Perspectives

Observational studies suggest that wetlands cover an estimated 8.2 million km² globally (Lehner and Döll 2004). However, based on recent studies that use both observations and models, the mean global area may be 12.3 million km² (Melton et al., 2013). The largest concentrations of wetlands generally are found between 50° and 70°N latitude, with substantial concentrations also found between 0° to 10°S latitude (Lehner and Döll 2004). North of 70°N latitude, continuous permafrost ecosystems also contain considerable soil carbon (see Ch. 11: Arctic and Boreal Carbon). Wetlands are estimated to cover approximately 2.2 million km² in North America (see Table 13.1), or about 9% of the continental land area. Although approximate global and regional extents of wetlands are generally known, there are significant challenges that hinder estimating wetland coverage with a high degree of confidence. These challenges include, but are not limited to, lack of detailed inventories, nonuniform definitions of wetlands, limitations of remotely sensed data and models, and continuing drainage and conversion of wetlands worldwide.

Positioning the North American wetland carbon stock in a global context is difficult due to the broad range (300 to 530 Pg C) reported (Mitra et al., 2005). Accordingly, the North American wetlands (161 Pg C) compose a significant but uncertain proportion (30% to 54%) of the global wetland carbon stock.

Natural wetlands are the largest natural source of CH₄ fluxes to the atmosphere (Kirschke et al., 2013) and thus are an important consideration of large-scale modeling assessments. Saunois et al. (2016) conducted a comprehensive assessment of the global atmospheric CH₄ budget using “top-down” and “bottom-up” approaches, which respectively are based on inversions of atmospheric CH₄ data and process-based wetland biogeochemical models. Twenty top-down and 11 bottom-up estimates were provided for North American wetland fluxes averaged from 2003 to 2012. The multimodel mean (±1 standard deviation) was 16 ± 4 Tg CH₄-C emitted per year for the top-down estimates, and 35 ± 11 Tg CH₄-C per year for the bottom-up estimates. Boreal North America (i.e., Alaska and Canada) account for most of the difference between these two estimates, with the bottom-up approaches exceeding the top-down approaches by 19 Tg CH₄-C per year. Estimating the CH₄ flux from North American wetlands between 1979 and 2008, Tian et al. (2010) estimated an average of 17.8 Tg CH₄-C per year. Those simulation approaches are less than the estimate of North American wetland fluxes reported in this chapter, 44.8 Tg CH₄-C per year (see Table 13.1). Both approaches have relatively large uncertainty levels associated with the CH₄ flux. Extrapolation of measurement data across the wetland area presumes a uniform response that belies the considerable differences among wetlands across the landscape. The large-scale model assessments suffer from the same issue of not having the capacity to consider variation among wetlands, but they have the ability to accommodate some aspects of spatial variability. The relative correspondence of the wetland CH₄ flux attests to the merits of both the large-scale process-based models and the need for additional empirical studies, particularly on mineral soil wetlands, to provide a broad base for model validation.

13.6.2 Regional Perspectives­—United States, Canada, and Mexico

Within North America, Canada has the greatest wetland coverage, with estimates ranging from 1.27 to 1.60 million km², followed by Alaska with an estimated 0.18 to 0.71 million km² of wetlands (Lehner and Döll 2004; Zhu and McGuire 2016). Estimates of terrestrial wetlands for CONUS from the USFWS National Wetlands Inventory (0.39 million km²) and Mexico (~0.05 million km²) are smaller than the total wetland area suggested by Lehner and Döll (2004), 0.45 and 0.16 million km², respectively. The reported soil carbon stock for CONUS terrestrial wetlands (12.6 Pg C) approximates the estimate (10.6 Pg C) provided through the U.S. EPA’s National Wetland Condition Assessment (NWCA; Nahlik and Fennessy 2016). The relatively small difference in soil carbon stock is attributable to less wetland area as reported in the NWCA (a difference of about 11,000 km²) and a shallower reporting depth (120 cm). Wetlands in Canada are dominated by peatlands, which harbor large carbon stocks estimated at 115 Pg C for this assessment (see Table 13.1) and 150 Pg C by Tarnocai et al. (2005). The greatest concentration of wetlands is in the provinces of Manitoba and Ontario, which contain about 41% of Canada’s wetlands (Mitsch and Hernandez 2013).

The recent cartographic assessment of Mexico’s wetlands provides important new information about the distribution of wetlands and context for assessing their loss (Landgrave and Moreno-Casasola 2012). Inland marshes are found in deltaic regions of the southeastern states of Veracruz, Tabasco, and Campeche, where the floodplains have deep organic soils (Smardon 2006). Marshes also are found in mountain ranges of central Mexico and in localized areas in the Sonoran and Chihuhuan deserts where springs feed shallow swamps (Mitsch and Hernandez 2013). However, little is known about their carbon stock or CO₂ and CH₄ fluxes.

The U.S. Geological Survey’s LandCarbon Program developed ecoregion estimates of current and future projections of carbon storage, net CO₂ exchange and CH₄ fluxes, and net carbon balance of U.S. wetlands (Zhu and McGuire 2010), providing context for the current assessment. Wetland area, carbon stocks, and fluxes were estimated using process-based models and land-use and land-cover maps. These estimates, originally reported by level II ecoregion in a series of reports, are summarized by region in Table 13.2. The LandCarbon assessment provides a basis for regional comparisons using a common methodology. However, the reported pools and fluxes are substantially different than those included in Table 13.1 which uses the National Wetlands Inventory as the basis for wetland area, summarizes geospatial databases for the pools, and synthesizes observational studies as the basis for the pools and fluxes.