et al., 2003; Duarte et al., 2013), and combined with high variability in local burial rates, this has resulted in a relatively large range of global burial estimates in salt marsh (4.8 to 190 TgC/yr, Table 1.1). With the most recent estimate of mean carbon sequestration rates (245 g C/m2/yr, Ouyang & Lee, 2014) and the recently updated global salt marsh area of 55,000 km2 by Mcowen et al. (2017) we estimate mean global carbon burial in salt marsh to be 13.5 TgC/yr. Uncertainties of global carbon burial and stocks are due to the small number of measurements from few study sites that are extrapolated to large regions. There are significant data gaps, in particular in South America and South Asia (Ouyang & Lee, 2014).
Important controls on local carbon sequestration rates, decomposition, and stocks include climate and latitude, tidal range, and halophyte genera (Kirwan & Mudd, 2012; Ouyang & Lee, 2014; Holmquist et al., 2018), as well as marsh elevation, i.e., the available accommodation space created by sea‐level rise (Kirwan & Megonigal, 2013; Rogers et al., 2019). Loss of salt marsh habitats and therefore the ability of marsh to capture carbon is predominantly caused by anthropogenic disturbance and land‐use change such as dredging, filling, draining, and construction (annual loss rate of 1–2% between 1980–2000, Duarte et al., 2008). Hurricanes can destroy large areas of marsh and trigger instantaneous losses of sequestered soil carbon (DeLaune & White, 2012). For total salt marsh carbon stock, we only include soil carbon, with an estimate of 0.4 to 6.5 PgC (Duarte et al., 2013).
1.4.3. Tropical Peatlands
In the tropics, areas with year‐round high rainfall, warm temperatures, low nutrient levels (mostly derived only from rainfall), and a topography that favors retention of water, host a unique ecosystem where plant biomass accumulates to form a thick layer of partially decomposed peat. Tropical peatlands cover 243,000–300,000 km2 in Southeast Asia (with the greatest extent in coastal Indonesia), 209,000 km2 in Africa (especially in the Congo Basin), and 131,000–587,000 km2 in the Americas (the large range reflecting uncertainty on what share of these extensive wetlands are peat‐forming systems) (Page et al., 2011; Lähteenoja et al., 2012; Gumbricht et al., 2017; Leifeld & Menichetti, 2018; Xu et al., 2018). Most contemporary tropical peatlands began formation during the Holocene, with rates of C accumulation in the range 30–270 g C/m2/yr for SE Asian sites (Page et al., 2010; Dommain et al., 2011). Depths of up to 20 m are known for some SE Asian peatlands, but most are in the range 5–7 m. Extensive peatlands on other continents are shallower: average thickness in the Peruvian Amazon is 2–3 m (max. 7.5 m; Lähteenoja et al., 2012; Draper et al., 2014) and 2 m in the Congo Basin (max. 6 m; Dargie et al., 2017). Belowground C density values average 2,775 t C/ha for SE Asia (Page et al., 2011); 2,000 t C/ha for the Congo Basin (Dargie et al., 2017), and 800 t C/ha for Peru (Draper et al., 2014). The largest store of tropical peat C is in SE Asia (69 PgC), followed by 34 PgC in Africa and 13 PgC in the Americas (Page et al., 2011; Dargie et al., 2017). Tropical peatland vegetation is dominated by either hardwood trees, palms, or a combination, and displays some level of zonation based on flooding and nutrient gradients (Anderson, 1961; Lähteenoja et al., 2012; Dargie et al., 2017; Draper et al., 2014). Prior to anthropic disturbance, peatland vegetation stored an estimated 7.1 and 1.4–2.5 PgC in SE Asia and Africa (based on 27–275 t C/ha and 67–124 t C/ha of biomass C, respectively) (Dargie et al., 2017; Wijedasa, 2020). South American peatland extent is uncertain, thereby limiting estimation of aboveground C stocks, but Peruvian peatland biomass stores 80–90 t C/ha (Draper et al., 2014). Compared to belowground C density values, aboveground C density is relatively low, but the vegetation plays an essential role in the formation, maintenance, and protection of below ground C pools.
For conversion to agricultural land uses, waterlogged swamp conditions are artificially drained, and accompanied by nutrient addition and pH control (Wijedasa et al., 2017). Drainage shifts conditions from those favoring slow peat accumulation to ones facilitating rapid decomposition, resulting in C emissions in the range 11–20 t C/ha/yr for peatlands under agriculture (Drösler et al., 2014). Due to rapid land use change, the area of peat swamp forest in SE Asia declined between 1990 and 2015 from 76% to 29% of the area (a loss of 4.6 Mha), with a corresponding increase of 11% to 50% in agricultural area (Miettinen et al., 2017), comprising both small‐scale farms and large‐scale industrial plantations (Wijedasa et al., 2018). In addition, drained peatlands are susceptible to fires that result in further, rapid C release to the atmosphere and life‐threatening air pollution. In Africa, the extensive peatlands of the Congo Basin are relatively intact, although the potential exists for future anthropic disturbance due to a lack of protection (Dargie et al., 2017); elsewhere on the continent, peatland has been exploited for agriculture (e.g., Uganda) and as a fuel source for energy generation (Rwanda). In South America, most Peruvian peatlands remain undisturbed, but as with the Congo peatlands, there is the potential for future disturbance from the development of agriculture and transport infrastructure (Roucoux et al., 2017).
The key uncertainties on tropical peatland C stocks and fluxes relate to extent, peat depth, rate of peat decomposition, and ecosystem functioning (Lawson et al., 2015; FAO, 2020). Even in the best studied tropical peatland area of SE Asia the actual extent and depth are uncertain (Vernimmen et al., 2020). This uncertainty is even more acute in Africa, where the recent study of Congo peatlands resulted in a 29% increase in known global tropical area (Dargie et al., 2017). Remote‐sensing techniques combined with ground‐truthing and field measurement campaigns are needed to improve mapping and monitoring of tropical peatlands and will also provide the basis for assessment of peatland condition and likely improvement of understanding of their role in the global C cycle (FAO, 2020). We estimate that tropical peatlands contain 8.5–9.6 PgC in aboveground biomass, 69–129 PgC in organic soil carbon (Dargie et al., 2017), and 77.5–138.6 PgC for total carbon stocks.
1.4.4. High‐Latitude Wetlands
The largest extent of peatlands and wetlands globally are in Boreal and Arctic biomes (here defined as >50°N latitude) where waterlogging and low soil temperatures promotes the accumulation of organic matter (Gorham, 1991). Peatlands >50°N latitude cover ca. 3.3 M km2, which is >70% of the global peatland area (Hugelius et al., 2020) These northern peatlands have been expanding throughout the Holocene, sequestering C into peat deposits and acting as a sink for atmospheric carbon (Smith, 2004; MacDonald et al., 2006; Brovkin et al., 2016). Peatlands are a type of wetland defined by an accumulation of peat (organic soil) at the surface, usually to a thickness of at least 30 or 40 cm (Joosten & Clarke, 2002). Gorham (1991) provided an extensive review of northern peatlands and made a benchmark carbon stock estimate of 455 PgC in the northern high‐latitude region of the world. Over broad spatial scales, peatland initiation is favored by relatively low temperature and high precipitation (Kivinen & Pakarinen, 1981). At local scales, topography and substrate permeability are the main controls on soil water retention (Loisel et al., 2017).
The two main processes of peatland formation (and expansion) are paludification or terrestrialization. Paludification is the most common form of northern peatland formation and involves initiation and expansion of peatlands into other established terrestrial ecosystems (such as poorly drained forests) and terrestrialization is peatland formation due to gradual in‐filling of water bodies (Kuhry & Turunen, 2006). While peatland initiation is favored by wet and cold conditions, the vertical growth and carbon accumulation rate of peat in established peatlands is favored by long growing seasons and high incoming photosynthetically active radiation, so that ecosystem productivity seems more important than decomposition rates for the rate of long‐term carbon accumulation (Charman et al., 2013). Over the past seven millennia, the mean net carbon accumulation rate in northern peatlands has been slightly above 10 g C/m2/yr, with present day flux measurements indicating a mean sink strength of around 30 g C/m2/yr (Yu et al., 2012). Northern peatlands and wetlands are a sink for atmospheric carbon, and this sink is expected to remain or grow in coming decades (Frolking et al., 2011; Chaudhary et al., 2020). However, it has been shown that the sink strength may saturate (Gallego‐Sala et al., 2018) or that peat growth may partly be offset by disturbances, including higher fire frequency (Turetsky et al., 2015), drought (Fenner and Freeman, 2011), and thawing of permafrost