How do draining and re-wetting affect carbon stores and greenhouse gas fluxes in peat soils? (systematic review)

Background

Draining peatland and lowering water tables has traditionally been carried out to prepare land for afforestation and agriculture, or for its extraction and use as fuel or in horticulture. This practice has been implicated in increased greenhouse gas (GHG) emissions from peatland together with a reduction in the total carbon (C) store. Concern about both the loss of wet peatland habitat and the global warming potential of these emissions have led to attempts in recent years at restoration by raising the water table and re-wetting peatland, for example by blocking drainage ditches (oftenreferred to as „grips‟). This practice is intended to restore the function of the peatland as a net sink of carbon dioxide (CO2) and a semi-permanent C store. This review assesses the evidence-base regarding change in peatland C stores and GHG fluxes in response to wetting and drying regimes, as a direct result of environmental management.

Objectives

The primary objective of this review is to retrieve all available evidence relating to the question „How do draining and re-wetting affect C stores and GHG fluxes in peatland soils?‟ and provide a synthesis of evidence on the climate change mitigation effects of re-wetting as a management intervention.

Methods

Search strategy A search for articles and datasets on draining and re-wetting peatland was conducted using a variety of electronic data bases. Website searches and organisational searches were also performed to find grey literature. An extensive consultation was also carried out to retrieve any unpublished information and to improve the search.

Selection criteria

Studies retrieved by the search were included in the review only if they included at least one alternative from each of the following categories:

• Subjects: Carbon in any form, or GHGs, held in, released from, or sequestered by peatland or peat-type soils.

• Interventions: Long-term re-wetting or draining of peatland or peat-type soils, or natural experiments comparing areas of peatland or peat-type soils in the same region with different long-term (not seasonal or sporadic) hydrology. Studies that involved peat cutting, extraction or burning were only included if draining or re-wetting was also clearly involved.

• Comparators: No intervention or before-after comparisons or both (BACI).

• Outcomes: Amount or change in C or GHG stored in, or released from, soils.

Data collection and analysis

Raw data on GHG flux (CO2, methane (CH4) and nitrous oxide (N2O)) as well as fluxes of dissolved organic C (DOC), and C mineralisation were extracted. Data on storage of C as total C, microbial C and yield were also extracted. Random effects meta-analysis was used to generate effects sizes (standardised mean difference) and to examine location-level data on the effectiveness of the interventions. Sub-group analysis and random effects meta-regression were used to investigate variation in effectiveness in relation to methodological and environmental co-variates.

Main results

There is a greater amount of evidence on emissions from intact and drained peatland than from re-wetted peatland. A random effects meta-analysis on the five studies that measured the net efflux of all three main GHGs suggests that drained peatland has a greater net efflux of GHG than intact peatland, but the effect was not statistically significant. No studies measured all three GHG fluxes simultaneously from re-wet peatland.

More studies measured the fluxes of the individual GHGs separately. Drained peatland produces less CH4 emissions by around 8 mg CH4 m-2 d-1 when compared with intact peatland (27 studies). Meta-regression showed this effect to be well correlated with greater water table depth and pH, while sub-group analysis showed a larger effect in fens than in bogs. Conversely, re-wetted peatlands show increased CH4 emissions by around 16 mg CH4 m-2 d-1 when compared to drained peatlands (five studies).

Drained peatland soils show a net increase in N2O emissions of around 133 μg N2O m-2 d-1 over intact peatland. Only one study measured the effect of re-wetting peatland on nitrous oxide flux: it caused an overall reduction in efflux of 7.1 mg N2O m-2 d-1 compared to drained peatland.

The most frequent measure of CO2 flux was total respiration (measured in the dark only). The available evidence suggests that drained peatland has a higher net CO2 efflux as total respiration than intact peatland (1.41g CO2 m-2 d-1, p=0.094) but this effect is not statistically significant. While, only one study measured CO2 in re-wet peatland, as daily respiration, but showed no effect on net efflux.

Conclusions

The evidence-base concerning GHG emissions and C storage in peatland after re- wetting is poor. There are too few studies and those that exist have limitations in their design, often lacking sufficient replication at the location or site level to enable the overall effect to be determined. There is a greater amount of evidence on comparative emissions from intact and drained peatland; however the methodology of most of these studies is similarly limiting. Better studies are required that use greater replication, baseline measurements and improved reporting of the data (showing mean, sample size and variance) and effect modifiers. There is a particular need for studies to address the flux of all GHGs simultaneously in the same locations so that the net global warming potential can be determined. In future studies on re-wetting peatland, researchers, policy makers and managers should ensure that appropriate measurements are put in place to ascertain whether this management intervention ishaving the planned net benefit for climate change mitigation. The available evidence to date is not inconsistent with a re-wetting intervention mitigating climate change, but better evidence of effectiveness is urgently needed.

Background

Peat and peatlands are composed of partly decomposed plant material deposited under saturated soil conditions. Peatland is a generic term including all types of peat- covered terrain and many peatlands are a complex of swamps, bogs, and fens, sometimes called a “mire complex” (NWWG 1988)

Global estimates of peatlands land surface cover vary between 2-5% yet they contain between 30-50% of the world’s soil carbon store; as much carbon as is held in the atmosphere (Gronlund et al. 2006; Lavoie et al. 2005; Treat et al. 2006). While net accumulation of carbon is relatively low (Vitt et al. 2000), they represent a long term store of carbon when compared with mineral soils that have relatively high turnover and oxidation rates (Yu et al. 2001). Moreover, losses of carbon from peatlands tend to be in the form of highly radiatively active methane (CH4), as well as carbon dioxide (CO2) (Moore & Knowles 1989; Moore & Roulet 1993). Therefore, peatlands may become significant sources of atmospheric carbon under a changing climate (Yu et al. 2001).

Although peatland’s contribution to long-term fluctuations in these atmospheric gases has been a matter of considerable debate (MacDonald et al. 2006), there is some evidence to suggest that increasing temperature increases the release rate of carbon from soils (Trumbore & Harden 1997), as well as through soil thaw in northern ecosystems (Tokida et al. 2007). Also, hydrological changes play an important role in regulating peatland dynamics, both in terms of the total flux and in terms of the nature (CH4 vs. CO2) of that flux (Moore & Knowles 1989; Siegel et al. 1995). For example, a lower water table caused by increased evapotranspiration, altered precipitation, and increased frequency of droughts, along with increased atmospheric temperature, may decrease soil CH4 and increase CO2 emissions from the peat surface (Trettin et al. 2006). Also, nitrous oxide emissions in wetlands can be erratic depending on the type of wetland, management and redox potential (Regina et al. 1999; Yu et al. 2008)

Peat soils have traditionally been managed in a number of different ways; such as drainage for forestry production (Gustavsen et al. 1998) and extraction of peat for use as fuel or for use in horticulture (Charman & Warner 2002). While direct measures of changes in peat carbon pool following water table drawdown are rare, both decreases (Sakovets & Germanova 1992) and increases (Minkkinen & Laine 1998) have been reported. Also, drier peat lands increase the risk of peat fires, further altering the carbon balance (Charman & Warner 2002).

Drained peat lands commonly undergo restoration attempts by rewetting, often resulting in reduced CO2 efflux (Komulainen et al. 1999), but alternatively increasing CH4 emissions as a result of longer hydroperiods near the soil surface (Komulainen et al. 1998). This review will synthesise the available literature regarding long term change in carbon stores and greenhouse gas flux in peatlands, in response to changing flooding and drying regimes, either as a direct result of environmental management or in comparisons of areas with naturally different water tables in the same region over long periods of time.