Current scenarios for climate changes over the 21st century suggest major changes in temperature, precipitation and some other climate phenomena that are likely to result in substantial changes to peatlands. Any assessment of future impacts of climate change on ecosystems must be based on some assumptions concerning the most likely trajectory for future climate change. Here, we use the summarised results of the IPCC Fourth Assessment Report (IPCC 2007) as the basis for an evaluation of future changes to global peatland systems, focusing on scenarios for climate change to the end of the 21st century. Based on IPCC data, the magnitude of climate change predicted suggests substantial impacts on the distribution, functioning and biodiversity of peatlands throughout the world. The most recent projections for climate change now have sufficient spatial resolution and adequate confidence levels to begin to suggest likely trajectories for peatlands at the regional scale.

Climate change will have far-reaching consequences for peatlands. There will be a wide range of impacts on biodiversity and carbon cycling. Indirect effects will follow, specifically in terms of feedbacks to the climate system through greenhouse gas exchange, and impacts on other functional values of peatlands. Peatlands are fundamentally linked to climate variability and conditions. Climate is a key determinant of their distribution, condition and typology. Their biogeography and biodiversity are also linked to climate, and variability in weather and climate is a primary control on many aspects of their functioning in relation to the carbon cycle and greenhouse gas exchange. Peatlands can therefore be affected in many ways by projected future climate changes. Some of the reactions to climate change will result in feedbacks through changes in greenhouse gas fluxes and storage.

Effects of increasing temperatures

Increasing temperatures will increase peatland primary productivity in some regions due to lengthened growing seasons and increased average temperatures during the growing season. In other areas increased temperatures will reduce primary productivity due to enhanced evapotranspiration and increased moisture stress. In future, it is expected that the growing season length will increase by 20-50 days by 2100 as measured by the number of days above 5oC. Whilst increased growing seasons will be experienced across many biozones, the productivity response may be greater in northern regions where peatland extent is greatest, because there will be much less drought stress to limit growth. Increasing summer temperatures in the Arctic and mountain areas will result in the northern tree line migrating to higher latitudes and altitudes in both peatland and non-peatland areas. This will decrease the albedo (reflectivity) of the surface, further enhancing warming of the atmosphere. Increased temperatures will generally result in enhanced decay rates and loss of carbon. However, the response of decay rates in the surface layers of peatlands to increased temperatures is complex and depends on hydrological and other conditions as well as temperature.
		Factors affecting the sensitivity of organic matter to decay. The intrinsic temperature sensitivity is affected by the temperature and the substrate. The environmental constraints dampen the sensitivity to decay because of water availability, soil structure, freezing. Although the response is shown as linear, the relationship between these factors and decay is often non-linear. Reproduced from Davidson and Janssens (2006).

Effects of precipitation changes

		Responses to precipitation changes will be highly variable depending on local and regional factors. Projections of future precipitation show a much more varied response than those for temperature. There are large regional variations in both the sign and magnitude of change, as well as in seasonality. In regions of increased year-round precipitation, peatlands may become wetter, but only if the increase in precipitation is not counterbalanced by increased evapotranspiration losses due to increased temperatures. In northern high latitudes where precipitation is projected to increase year-round, peatland water tables may rise, particularly in spring, when combined with the effects of snowmelt. However, increases in summer precipitation may not compensate for higher temperatures and greater numbers of very hot days, so there will be periods of increased dryness and drought during the summer.

In other regions such as the northern mid-latitudes, a decrease in summer precipitation will cause a fall in water tables during the summer, exacerbated by increased temperatures. Increased winter precipitation in the mid-latitude regions will not provide any compensation for summer drought, as excess moisture will be lost through runoff. In tropical peatland regions, projections suggest decreases (e.g. Amazon basin), little change (e.g. tropical west Africa) or increases (e.g. Indonesia) in precipitation. At any particular location, the hydrological impacts will be mediated through topography and mire type. For example, ombrotrophic bogs may be more susceptible to reduced precipitation than fens, where groundwater flow can buffer some of the temporary drought periods.

Increased frequency of heavy rainfall and greater intensity of precipitation may cause enhanced erosion of susceptible peatlands. Peatlands that are susceptible to erosion through exposure of bare peat surfaces lose large amounts of peat during heavy rainfall events and also through aeolian processes (Evans and Warburton 2007). In temperate and high latitude peatland regions, potential threats are wetter winters and increased drought frequency in summer. In particular, the predicted increased intensity of rainfall events may exacerbate erosion of peat in areas already disrupted by the effects of grazing, fire and pollution, and also in regions subject to other future climate-related damage. Increased severity and frequency of droughts would cause desiccation of the peat surface, resulting in structural damage such as cracking and increased susceptibility to erosion during wetter periods (Evans et al. 2006). In extreme cases, very wet conditions have led to mass movement of peat in ‘peat-slides’ both during intense rainfall events in summer and more prolonged wet conditions during winter (Warburton et al. 2004). More extreme hydrological conditions in future may increase the likelihood of such events.

Where increased flooding of peatlands occurs due to higher frequency and magnitude of heavy precipitation events, there may be increased CH4¬ release. The relationship between water-table depth and methane emissions is well known, with higher water tables causing increased methane emissions.

Hydrological Changes

The hydrological regime is the principal factor controlling ecosystem processes in peatlands. Any changes in water balance should be expected to have far-reaching effects on biogeochemical cycles, productivity and community composition. Increased evapotranspiration, altered precipitation, and increased frequency of extreme events (e.g. droughts, intense rainfall) are likely to result in generally lower water-table depths during the growing season (Roulet et al. 1993, Waddington et al. 1998).

A combination of increased temperatures and changes in precipitation regime will determine the future hydrological status of peatlands. The direction and magnitude of surface wetness (moisture, water table) changes will vary between regions. Changes in precipitation will be particularly important in determining the local and regional changes in hydrology. The regional complexity of projected future climate change (see Section 8.1) will lead to a regionally and seasonally variable hydrological response in peatlands.

Changes in water-table levels will affect the carbon dioxide and methane exchange of peatlands. The global net effect of these changes on ‘global warming potential’ of peatlands is not known with any accuracy, but it is likely that the carbon sink function of existing peatlands will be reduced overall. This may be compensated for in part by expansion of some peatland areas. In general, falling water tables in many peatlands will result in lower CH₄ emissions but higher CO₂ emissions. In areas where peat accumulation is already marginal, this will result in a gradual loss of the peat by oxidation. Conversely, in some regions (notably thawing permafrost areas), increased flooding will increase CH₄ release but may also increase CO₂ drawdown from increased productivity.

Lower water tables will increase the rate of release of N₂O. Increased aerobic decay in the surface layers of peatlands following water-table draw-down is accompanied by increased mineralization of organic compounds including nitrogen (N). Although levels of N are often low in ombrotrophic systems, they are much higher in minerotrophic peatlands such as floodplain fens. The N is released in the form of nitrous oxide (N₂O), another powerful greenhouse gas (see Chapter 7). Although N2O emissions from natural peatlands are generally low, peatland drainage is known to increase N₂O release (Minkkinen et al. 2002, Alm et al. 2007). This suggests that N₂O emissions may increase with water-table draw-down and increased temperatures in the future.

Hydrological change will result in major changes in microbial, plant and animal communities. Hydrology is the most important influence on peatland communities and many species occur in restricted hydrological niches. Changes in hydrology will inevitably result in changes to the distribution of individual species and communities. This will happen at various scales. Small-scale micro-topographical patterning provides a range of microhabitats for peatland taxa. Changes in the balance between pools, lawns and hummocks on individual sites will lead to a change in the plant and animal communities present. Specialist taxa tend to be more dependent upon the wetter microhabitats in shallow pools and hollows and loss of these habitats may lead to local extinctions of aquatic and semi-aquatic taxa. The ability of these taxa to migrate to more suitable locations may be limited in regions where peatlands have become fragmented.

Hydrological changes will result in the expansion of forest cover on drying peatlands and a reduction in tree cover on peatlands with increased surface wetness. Tree cover on peatlands is usually less dense than on surrounding mineral soils. Typically, trees become smaller and less dense towards the wetter, central area of a peatland. Tree establishment and growth is limited by high water tables that keep the rooting zone saturated. Thus when water tables fall, tree cover becomes more extensive and existing trees may be able to grow (Choi et al. 2007). Once trees become established on drier areas of peatlands they may further enhance transpiration, providing a positive feedback and further drying of the peatland surface beneath the trees and adjacent areas. Increases in tree cover may also add to the decrease in surface albedo arising from the more general changes in tree lines, and reduced snow extent. The situation may be more complex in tropical peatlands; although some trees will increase growth if water tables fall, others may suffer increased mortality in drought years (Nishimua et al. 2007).

Fire frequency may increase on peatlands that are subject to greater extremes of drought. Natural fires occur throughout many parts of the world, including areas where peatlands are common. Although peatlands are some of the wettest locations in the landscape, in dry summer seasons, surface layers of peat and plant litter provide a highly combustible mixture for fire to start in or to spread to from surrounding upland. More prolonged, more severe or more frequent periods of drought may occur in some peatland regions. Even in some areas where rainfall is projected to increase, there may be more severe short periods of drought if rainfall variability increases. There are many causes of peatland fires, principally related to human activities such as clearance for agriculture, improvement of grazing conditions or accidental causes. Lightning strikes may be a cause of natural fires in some circumstances but the occurrence of natural fires is limited when compared with those started by people. It is unknown if the frequency of natural fires will increase significantly in the future, because lightning strike frequency cannot be predicted with any certainty. However, any increase in droughts will make uncontrolled natural or anthropogenic fires more likely and will lead to increased intensity of burn. In particular, lowered water tables may make the loss of upper peat layers more likely during peatland fires and the areas burnt may be more extensive.

In areas where peatlands become drier, extreme events such as drought and storm rainfall will increase erosion by water and wind. Peatlands in some parts of the world, such as Central Asia, are likely to become much drier in future. Desiccation of peat surfaces is likely to make them more susceptible to erosion. Exposure of bare peat surfaces and disturbance from activities such as grazing will result in further destabilization of the peat structure. During droughts, the upper peat layers can be eroded by wind and the surface becomes deflated. During intense rainfall peat can also be moved down slope and lost in runoff.

In regions strongly affected by drying during ENSO years, the frequency of drought is likely to increase due to the background increase in temperature and changes in precipitation. Several peatland regions are currently affected by ENSO. During these events, peatlands in southeast Asia and to a lesser extent, the Amazon basin, experience much drier than normal conditions. Although ENSO may not increase in frequency or intensity, the frequency and impacts of drought will increase in ENSO years because of the background climatic changes against which they occur. This will enhance fire frequency and intensity especially in drained areas and where people light fires for clearance and agriculture.

Changes in river runoff regime may change flood periods and other characteristics of the hydrological regime of riverine and lacustrine peatlands. The changing balance between precipitation and evapotranspiration will result in alterations to river flows and their seasonal variability. These changes will affect peatlands in floodplains and lake margins through more dynamic flood regimes from increased intensity of rainfall and droughts. Increases in precipitation and earlier snowmelt in the northern high latitudes will lead to earlier and increased runoff of rivers discharging into the Arctic Ocean. The total increase in runoff is expected to be in the region of 10-20% depending on the climate scenario used, with much greater percentage increases during winter (Walsh et al. 2005). This will affect the salinity, ice cover and circulation of the Arctic Ocean.

Changes in permafrost and snow cover

		Melting of permafrost may stimulate CH4 emissions from wet peatland sites and peat decomposition and CO2 release from dry peatland sites such as palsas. Rapid permafrost melting has been reported to increase formation of thermokarst ponds and fen-bog vegetation with rapid peat accumulation through natural successional processes of terrestrialization (Payette et al. 2004). Both increased temperatures and increased snow depth (and therefore insulation) may be a cause of permafrost degradation. Contrary to current expectations, the melting of permafrost caused by recent climate change did not transform the peatland to a carbon-source ecosystem, as rapid terrestrialization exacerbated carbon-sink conditions (Payette et al. 2004).

However, the developing fen vegetation may develop into an efficient producer of methane emissions (Wickland et al. 2006), which may counteract the climatic impact of CO₂ sequestration.

Reduction in permafrost extent in peatlands may lead to greater release of dissolved organic carbon in river runoff. Permafrost prevents or reduces decay of organic matter in northern high latitude peatlands. Where decay occurs, carbon is lost as CO₂ and also in dissolved organic carbon (DOC) in runoff. Much higher DOC concentrations have been found in catchments with permafrost-free peatlands than in those dominated by permafrost. Furthermore, there is a correlation between the extent of peatland cover and DOC concentrations in runoff in permafrost free catchments. On this basis it is predicted that losses of DOC in rivers draining the West Siberian region may increase by 29-46% by the year 2100 (Frey and Smith 2005).

Increasing temperatures are likely to lead to a large reduction in the occurrence of palsa mires. Palsa mires occupy a zone in the high northern latitudes. They are tightly constrained climatically, suggesting they are particularly sensitive to climate change (Parviainen and Luoto 2007). Recent observations suggest that palsas are already declining in some areas (Zuidhoff 2002). In Fennoscandia, an increase in mean annual temperature of 4^oC would result in the loss of all palsa mires from this region.

Increased snow depth during winter may affect the thermal regime of northern peatlands, exacerbating the effect of increased temperatures on permafrost melt. Snow cover is important in insulating the peat surface during winter. Deeper snow leads to less intense freezing and reduced permafrost formation. For example, at the southern margins of palsa distribution, palsas only occur in particular areas where wind action is sufficient to reduce snow cover.

Sea level rise

		Inundation of coastal peatlands may result in biodiversity and habitat losses with conversion of freshwater peatlands to saline marshes. A rise in base level may allow the spread of new peatlands inland if land is made available for this. Gradual sea-level rise will result in a shoreward displacement of the boundary between saltmarsh and freshwater wetland as the height of highest astronomical tide rises. Complete inundation of saltmarsh and coastal peatlands may even result in destabilisation of the peat matrix and the release of CH4. However, due to the rise in base level, areas further inland may be increasingly susceptible to flooding and freshwater peatlands could spread inland if the topography is suitable and land use policy allows for this.

In low-lying peatland areas, intrusion of saline water into aquifers may give rise to increased salinity and changes in the ecology and functioning of the system. Floodplain fens and some other peatlands close to the coast are partly dependent on groundwater for their water supply. Aquifers close to the coast are often linked to the sea and it is possible that over time their groundwater will become more saline. This will change the growing conditions for plants and the geochemical conditions at the surface.

There is an increased risk of erosion of coastal peatlands in areas experiencing sea-level rise. In some locations where peatlands meet the coast, rising sea-levels may result in incremental erosion of peat, with loss of particulate organic carbon (POC) directly into the ocean. The extent of physical erosion depends upon the wave conditions and rate of sea-level rise. Gradually increasing sea-level in a relatively sheltered wave environment is more likely to result in succession to saltmarsh and the accumulation of marine sediments rather than the loss of peat.

Other impacts of climate change on peatlands

Non-linear responses to climate change may lead to ‘surprise’ changes in peatland systems. Some of these changes will be irreversible. Past climate change has led to sudden changes in peatlands. Sometimes peat accumulation is interrupted for several hundred years by events such as fires, floods or long-term droughts. Even if climate change proceeds as a gradual process, it is unlikely that the peatland response will be smooth and monotonic. Many of the processes expected to change in response to climate change are likely to have thresholds past which changes are sudden and perhaps irreversible. For example, drying of peat surfaces can lead to cracking and desiccation to the point where rewetting the peat is extremely difficult. Once this initial severe drying has occurred, slow oxidation and peat removal by physical erosion occurs and is largely irreversible until the layer of altered peat has been lost.

There will be longer term (100-1000 years) impacts as a result of climate changes taking place this century, due to altered successional processes, and the slow response of some processes. Many peatland ecosystems have a high level of ‘self-determination’ due to successional processes. This means that once an initial change has taken place, it may lead to other (different) changes in the future. For example, in the case of permafrost collapse, the initial formation of pools may be succeeded by a phase of pool-infill and renewed peat growth, which leads to a very different peatland system over hundreds to thousands of years.

Climate change will interact with anthropogenic disturbance. Some of the impacts of climate change will exacerbate the impacts of activities such as drainage, grazing, burning and logging. Human disturbance to peatlands often makes them much more vulnerable to climate change impacts. Climate change is only one of many factors affecting peatlands. In most cases, human impacts on peatlands will increase their vulnerability to climate change. In a number of situations, climate change will simply act to accelerate existing degradation of damaged peatlands. For example, the occurrence of fires in Indonesian peatlands is largely due to drainage, logging and fire-setting (Page et al. 2002), but the frequency and severity of fires is increased by changes in the length and severity of droughts. Likewise, over-grazing of vegetation in central Asia has already led to erosion and loss of peat. These impacts are likely to be accelerated by future climate change.

The species that are most vulnerable to climate change are those specifically adapted to peatland conditions, endemics, species at their geographical limits (latitudinal and altitudinal), and those with a disjunct distribution. Some peatland species have clear relationships with climatic variables. Those species are expected to disappear in areas where significant climate shifts occur. Changes will occur first at the geographical limits of these species. Where taxa have a disjunct distribution, perhaps due to fragmentation of the peatland habitat, these taxa are more likely to become locally extinct rather than migrating in concert with climate changes. Endemic and rare peatland taxa are the most threatened species. The Siberian Crane (Grus leucogeranus) is a threatened species dependent upon suitable nesting areas in tundra regions of Yakutia. This and other migratory birds dependent on open breeding areas may suffer reduced populations as the area north of the treeline becomes smaller, with expansion of northern forest areas.

The microclimate of peatlands may present opportunities for ecological adaptation to climate change. Peatlands may provide non-peatland taxa with temporary refugia as they retreat to higher latitudes and altitudes. Peatlands are often isolated within anthropogenically modified landscapes. They provide some of the last ‘wild’ landscapes in such areas and thus are a refuge for specialist peatland taxa which may be more widespread elsewhere, but also for other wildlife which can survive in a peatland or around its margins. Peatlands may therefore act as ‘stepping stones’ for the migration of more adaptable species that could not survive in agricultural or other strongly modified landscapes. In particular, they will be refuges for birds and mammals sensitive to disturbance.