Evaluating ecosystem goods and services after restoration of marginal upland peatlands in South-West England

Authors


Correspondence author. E-mail: e.grand-clement@exeter.ac.uk

Summary

  1. In the UK, drainage for agricultural reclamation during the 19th and 20th centuries is responsible for an alteration of the ecological and hydrological functioning of peatlands, in turn, affecting a whole suite of ecosystem services. Today, initiatives are in place throughout the UK to reinstate this eco-hydrological functioning by blocking drainage ditches. Effects on ecosystem services remain unclear, as does the overlapping impact of climate change on peatland recovery.
  2. This article uses a conceptual model to present the effects of restoration on ecosystem services, that is, water provision and quality, carbon storage, biodiversity, food and fibre provision and cultural services, both immediately after ditch blocking and in the few years post-restoration. The model is then applied in the context of Exmoor National Park, in South West England and used to perform a cost–benefit analysis of the restoration and monitoring programme, as these shallow peatlands are located in geographically marginal areas, and therefore more sensitive to climate change.
  3. Past research indicates that some processes tend to return progressively to their predisturbance state, but whether the complete recovery of peatlands to functioning mires occurs after restoration remains unclear, partly due to the difference between the temporal and spatial scale at which processes occur (i.e. up to decadal) and are monitored (typically a few years).
  4. Overall, on Exmoor, the long-term benefit of peatland restoration to some ecosystem services, such as a reduction in carbon losses and improvement of water storage and quality, has the potential to balance high financial investment.
  5. Synthesis and applications. Gaining a better understanding of the effects of peatland restoration on ecosystem services provided is essential to assess the potential value of restoration projects. Using the case of the shallow peatlands of Exmoor National Park, located in geographically marginal areas in the UK and therefore more vulnerable to the effects of climate change, we find that there is potential for both the value of carbon storage and water provision to offset the costs of restoration in the long-term. Our results from Exmoor can provide ecological analogues of impending change further north.

Introduction

Functioning mires are dependent on the fragile connection between water table levels and vegetation communities. Natural and anthropogenic-driven changes acting over large temporal and spatial extents are generally affecting the environmental conditions for peat formation (Bonnett et al. 2009). Management practices are known to affect some of the critical ecosystem services (ES) provided by peatlands, namely, the storage of carbon (C), support of specific and rare habitats and supply of drinking water (Hubacek et al. 2009). In the UK, focus on food production in the uplands meant that drainage was undertaken extensively in the 19th and 20th centuries to lower the water table in order to improve the quality of the vegetation for grazing and game management, and ensure the safety of stock (Ratcliffe & Ostwald 1988). From 1900, drainage was also implemented widely as a flood alleviation measure (Holden, Chapman & Labadz 2004). Between the 1970s and 1990s, EU subsidies for sheep farming intensified this practice and resulted in a 30% increase in sheep numbers on UK moorlands (Holden et al. 2007). In the south west (SW) of England, large areas of Dartmoor and Exmoor were extensively drained and additionally used for peat cutting for fuel.

There is little evidence that drainage fulfilled its role (Stewart & Lance 1983; Wilson, Wilson & Johnstone 2011), as peatlands cannot sustain high grazing densities (Holden, Chapman & Labadz 2004). Instead, historical and recent agricultural use of peatlands is now perceived to have contributed to degradation through drainage, peat cutting, burning, grazing or afforestation (Bonnett et al. 2009). Concerns over ES provided by peatlands have now largely shifted towards environmental factors: drainage ditches and areas of peat cutting for fuel are major threats to peatland sustainability, as they have long-lasting consequences on the water table, which in turn regulates the balance between peat accumulation and decomposition, and impacts on biodiversity (Holden, Chapman & Labadz 2004; Holden et al. 2006). Consequently, numerous conservation projects are now aiming to re-establish active peat-forming mires through the restoration of eco-hydrological function by blocking drainage features and cuttings (Lunt et al. 2010). A key component of these restoration activities is to understand and quantify the impacts on ES provision.

This article first presents a conceptual model to describe the observed effects of ditch blocking and trade-offs between various ES. The model is then applied to the marginal peatlands of Exmoor and used to conduct a cost–benefit analysis of the current restoration work. Focus is given to the shallow peatlands in SW England, as conditions here are different from other northerly peatlands. For example, they are shallower than peatlands in northern areas (Smith 2009), as highlighted in Table 1, and consequently are likely to be more sensitive to drainage and peat cutting. Their location on the southernmost margin of the UK peatlands' geographical extent also makes them extremely vulnerable to climate change (Clark et al. 2010). Whilst more information is needed about the behaviour of such mires in a changing climate, our working hypothesis is that SW peatlands may be early indicators of climate change in damaged peatland due to the warm maritime climate that they experience.

Table 1. Contextual information for the main peatland areas in England
AreaLocationUpland peat area (ha)Precipitation (mm)Average winter temperature (°C)Average summer temperature (°C)Peat depth (m)Mires vegetation communitiesHistorical land use and degradation patterna
  1. a

    MetOffice (2012b).

  2. b

    Heal & Smith (1978).

  3. c

    Armstrong et al. (2010).

  4. d

    Wishart & Warburton (2001).

  5. e

    Rothwell et al. (2007).

  6. f

    Charman et al. (1999).

  7. g

    Maltby, Legg & Proctor (1990).

  8. h

    Smith (2009).

  9. Precipitation and temperature data were estimated from regional climatic data for 1971 to 2000 (MetOffice 2012a), unless stated; Peatland area and land-use data originate from Natural England (2012), with PC for peat cutting, D for drainage, E for erosion and B for burning. Mires vegetation communities data were sourced from Drewitt & Manley (1997) using the following abbreviations: M6, rushes, sedges and Sphagnum; M17 and M20, Eriphorum vaginatum; M19, Cotton grass/Sphagnum; M21, Narthecium ossifragum, Sphagnum papillosum; M23, rushes; M25, Molinia caerulea.

North Pennines54°42′N; 2°29′W107 4801700–23001–45.5–9.50.4–5bM6, M19, M20D, B, E
Yorkshire Dales54°13′N; 2°23′W49 7071700–23004–59.5–112cM6, M20, M23PC, D, E, B
Northumberland55°28′N; 2°8′W31 0841200–15001.5–3.56–9.51.5–2.5dM6, M19, M20PC, D, E, B
Peak District53°22′N; 1°52′W30 2621050–16003.4–4.88.8–11.83 eM6, M20D, E, B
Dartmoor50°40′N; 4°0′W18 6941974a5.4a11.2a>6fM6, M17, M21PC, D, E, B
North York Moors54°23′N; 1°3′W5 514900–12003.5–4.59.5–113.5gM6, M20PC, D, E, B
Exmoor51°9′N; 3°34W4 2791800–26004.5–5.510–120.33hM25PC, D

The restoration of ecosystem services

Practically, restoration seeks to create storage of water behind drain blocks, and thus to increase the water table levels in the surrounding area to a predrained level. The intricate suite of changes in ES provided by peatlands after blocking is shown within the conceptual model in Fig. 1. These impacts will be examined and discussed in more detail in the following sections. This conceptual model will then be applied to Exmoor National Park.

Figure 1.

Conceptual model of the impact of drain blocking on ES provided by peatlands; circle sizes (small, medium or large) were determined by qualitative scaling of the magnitude of the expected impact. Light and dark circles indicate overall positively and negatively affected ES, respectively.

Water storage potential

Studies have reported an increase in water table level following the blocking of drainage ditches within 6 years after restoration (e.g. Worrall, Armstrong & Holden 2007; O'Brien et al. 2008; Holden et al. 2011). The amplitude of the increase is, however, variable; going from 25 cm after 7 months of restoration in the North Pennines (Jonczyk et al. 2009), to only 2 cm in Wales, remaining 8 cm below the peat surface (Wilson et al. 2010). The water table tends to exhibit behaviour similar to that of undisturbed conditions after 1 year, due to the increased water storage potential in areas of the surrounding peat where it would have previously flushed out of the system more rapidly because of drainage connectivity. The reduction in hydraulic conductivity from blocking ditches therefore allows for the water table to remain high between rainfall events and stable during the summer (Wilson et al. 2010). Wilson et al. (2010) also observed that the rise in water table primarily occurs on the downslope side of drains, and that the buffering and swelling ability of the peat can be reversed, which is a good sign of recovery.

The direct effect of ditch blocking is to alter the response of the drainage network to rainfall, slowing down the flow of water in the ditches and receiving channels, and modifying the connectivity of the network (Jonczyk et al. 2009; Ramchunder, Brown & Holden 2009). This attenuation of flow velocities in ditches increases the likelihood of overland flow, which is significantly slower than drainage through the channels (Holden et al. 2008a), and ultimately leads to a decline in the magnitude of peak flows in streams (Wilson et al. 2010). However, acrotelm storm flow, subsurface pipeflow and macropore flow remain important pathways of flow generation (Holden & Burt 2003), which are often unquantified. On the other hand, the saturation of the peat following restoration can sometimes cause a greater discharge due to the lack of space for holding more water (Evans et al. 1999). As a result, the potential for decreased flooding in restored areas remains unclear but is still used to justify restoration (e.g. Bain et al. 2011). Studies of flow paths at the scale of the drain, but also their evolution during storm events, are therefore still needed to understand catchment-wide responses post-restoration.

Carbon storage

Carbon storage has to be quantified based on the balance between C losses from the system through gaseous emissions, that is, carbon dioxide (CO2) and methane (CH4), aquatic pathways (Dissolved and Particulate Organic Carbon, DOC and POC, respectively) (Limpens et al. 2008), as well as C accumulation processes related to vegetation. Current understanding and areas of uncertainty on the effect of restoration on these pathways is reviewed in the following section and illustrated in Fig. 2.

Figure 2.

Hypothesised effects of drainage and restoration through time on CO2 and CH4 (top), and DOC concentrations (bottom), based on findings from Komulainen et al. (1999), Wallage, Holden & McDonald (2006), Worrall, Armstrong & Holden (2007), Höll et al. (2009), Gibson et al. (2009), Armstrong et al. (2010), and Lindsay (2010). Dashed lines and grey areas represent epistemic (knowledge) uncertainties; amplitude of concentrations and time are therefore schematic, and not to scale.

Gaseous fluxes

There have been very few studies looking at CO2 emissions in rewetted peatlands, but it is generally thought that a rise in water table will decrease respiration. Results by Komulainen et al. (1999), 2 years after the restoration in a drained ombrotrophic bog in Finland, show that respiration is reduced after restoration, leading to a significant increase in C storage compared with that of the drained site. Differences in the peat structure and behaviour make quantification and comparisons of CO2 emissions difficult (Lindsay 2010). Figure 2 highlights the opposite behaviour of CO2 and CH4 following both drainage and restoration. Potential counteracting effects of restoration through methane emissions are a key concern: despite being several orders of magnitude lower than CO2, CH4 has a higher global warming potential. The creation of open water pools behind ditch blocks is likely to increase CH4 emissions in the short-term (Baird, Holden & Chapman 2009), but this increase is only a temporary stage of recolonisation that will be mitigated by Sphagnum spp. within 5–10 years (Lindsay 2010). Overall, it is thought that methane emissions remain largely compensated by the long-term benefits of restoration, that is, decreased CO2 emissions and increased C accumulation.

Aquatic fluxes

A general increase in DOC concentrations in lakes, rivers and reservoirs has been observed during the past 15 years in areas largely covered by peat (e.g. Driscoll et al. 2003; Hejzlar et al. 2003). This process suggests a systematic response of DOC releases to a combination of external drivers acting over large areas (Evans, Monteith & Cooper 2005), such as an increase in atmospheric CO2 (Freeman et al. 2004), a decrease in acidic deposition (Monteith et al. 2007), changes in pH (Krug & Frink 1983), changes in the nature of flow (Tranvik & Jansson 2002), occurrence of severe drought (Worrall, Burt & Adamson 2004) or eutrophication (Harriman, Curtis & Edwards 1998). However, small-scale factors (i.e. land use) can have an additional effect on the general trend. For instance, moorland burning seems to exert a stronger influence on DOC losses than global and regional drivers such as climate or acid deposition (Clutterbuck & Yallop 2010), whereas ditch blocking could help to mitigate DOC export in the short-term (Worrall, Gibson & Burt 2007).

Within the aquatic losses, DOC typically represents the main component. For example, in Welsh unblocked peatlands, the ratios of dissolved organic carbon to total organic carbon (DOC/TOC) range between 15 : 1 and 50 : 1, in conditions of low and high rainfall, respectively (Wilson et al. 2011). Certain studies working on short timescales demonstrate that, up to 3 years after restoration, DOC losses can increase (O'Brien et al. 2008; Gibson et al. 2009), as illustrated in Fig. 2. Such behaviour is explained by different processes, including: (1) a flush of DOC that built up in the peat during aerobic conditions (Kalbitz, Rupp & Meissner 2002; Worrall, Armstrong & Holden 2007); (2) an enzymatic latch where the return to anaerobic conditions does not immediately stop the degradation of recalcitrant materials by key enzymes (phenol-oxidase), initiated in aerobic conditions (Freeman, Ostle & Kang 2001) and (3) a change in ionic strength and pH as water table recovers from a severe drought that increases DOC solubility and mobility (Clark et al. 2005). In the longer term (more than 4 years after restoration), most studies show a decrease of DOC concentrations (e.g. Höll et al. 2009; Armstrong et al. 2010), possibly related to a store flushing and exhaustion process in response to higher water table levels (Wallage, Holden & McDonald 2006). However, studies are usually based on DOC concentrations but rarely consider change in flow after restoration and therefore might not give a reliable picture of the extent of change (Wilson et al. 2011): if DOC production remains upon rewetting due to the processes explained above, a decrease in the flow after blocking would still cause an overall decrease in DOC yield from the system (Gibson et al. 2009). In some cases, the run-off decline is smaller than the increase in DOC concentration, causing an increased DOC export after ditch blocking (Worrall, Gibson & Burt 2007).

In addition to C losses, DOC exports also affect water quality, causing water acidity, nutrient and pollutant transport (Thurman 1985), with concomitant ecological implications downstream (Urban, Bayley & Eisenreich 1989). Furthermore, discoloration gives the water a low aesthetic value, and complicates the costly water treatment process. Consequently, peatland restoration also has the potential to improve provision of water quality as an ES, as shown in Fig. 1.

Current knowledge seems to indicate that reverting DOC and CO2 losses to levels similar to that of the general evolution of undisturbed peatlands is possible in the long-term (Fig. 2). However, the speed of recovery and whether the magnitude of DOC and CO2 losses can reach undisturbed levels is currently unknown. In addition, the amplitude and duration of the CH4 increase until recolonisation by mire species remains uncertain, as shown by the grey area in Fig. 2. Overall, in the long-term, restoration has the potential to lead to a positive C balance, although DOC production and gaseous emissions would clearly benefit from understanding the behaviour of drains in various conditions (for example, varying peat depth, width and depth of the drain), as well as during rainfall run-off events of varying magnitude.

Vegetation change and biodiversity

One of the goals of restoration is to re-initiate the establishment of Sphagnum and other bog species, leading to a C-accumulating system. This has been observed 10 years after the restoration of ombrotrophic bogs in Finland (Haapalehto et al. 2010). Nevertheless, some common vascular species were absent, and the overall species composition remained different to that of a pristine bog. This is thought to be linked to different factors, such as (1) the poor dispersal potential of these species (Haapalehto et al. 2010), (2) the connectivity between drains and gullies and the surrounding peat (Ramchunder, Brown & Holden 2009) and (3) a lack of a sustainable natural seed bank (Ramchunder, Brown & Holden 2009).

The return of mire-forming species is an important goal in restoration, not only for its implications for biodiversity and habitat condition, but also because it allows an increased C accumulation, which, together with decreased C losses, could lead the system to a positive C balance. Subtle vegetation changes can, however, be difficult to observe. Remote-sensed data are now increasingly used as proxies for monitoring changes in near surface wetness (Anderson, Bennie & Wetherelt 2010), and thus, species composition (Couwenberg & Joosten 2005).

Food and fibre provision

High financial incentives in the 1970s, linked to national efforts to increase food productivity, promoted drainage to improve vegetation quality for grazing. As a result, stocking densities are thought to have increased to unsustainable levels (Holden, Chapman & Labadz 2004). Overall, there is little evidence that drainage actually fulfilled its initial purpose (Stewart & Lance 1983; Wilson, Wilson & Johnstone 2011), but instead caused long-lasting and extensive ecological and hydrological damage in upland peatlands. Recent work indicates that restoration will make little difference to the agricultural productivity of those areas (Wilson, Wilson & Johnstone 2011), although in some cases, the potential changes to vegetation from restoration will inevitably lead to a decrease in stocking densities. In addition, agri-environment schemes, including Higher Level Stewardships agreements, currently allow the compensation of potential loss of income to farmers, and provide payments for environmental management and have already led to a decrease in grazing densities (Condliffe 2009).

Cultural services

Cultural services provided by peatlands are notably difficult to evaluate and quantify because of their subjectivity and because they are perceived differently by different sectors in society (Suckall, Fraser & Quinn 2009). Changes in accessibility will clearly affect the amenity value of rewetted peatlands, but whether the impact of restoration is negative or positive is open to debate. For example, some of the recreational activities that are common on Exmoor, such as hill walking, deer hunting or horse riding, will become difficult due to the land becoming saturated for longer periods of time. However, ditch blocks will make movement across ditches and heavily degraded peatlands easier for both animals and humans, which can balance this negative aspect. Conservation of the historical environment is another crucial service of functioning peatlands: anaerobic conditions in restored peatlands will preserve rare archaeological assets and palaeo-environmental records better than degraded peatlands (Bonn et al. 2009). If the educational value of peatlands from archaeological and aesthetical points of view is commonly considered, there is also an educational value from the restoration itself: this process is not only important for scientists and professionals, but can also improve the general publics' understanding of environmental issues, such as water usage and peatland functions. In this case, communication has an important role to play in restoration programmes and can be achieved through stakeholder engagement.

An approach for evaluating benefits of peatland restoration on Exmoor

The specific context of Exmoor

Most of the current understanding of the effects of peatland restoration is based on sites located in the north of England, but comparatively little is known about the situation in the SW. Within this region, Exmoor is of particular interest, due to its specific characteristics (Table 1). For example, based on long-term regional meteorological data, the climate in the SW appears to be wetter and warmer compared with other peatland areas. In terms of peat depth, although extensive data sets seem to be lacking, and most of the values presented in Table 1 are based on few localized studies, peatlands on Exmoor appear to be shallower (0.33 m on average; Smith 2009) compared with other areas that usually exhibit several metres of peat. Erosion and bare peat are common problems in deep peat areas, such as the Peak District, but are rarely found on Exmoor. Instead, the degradation mostly comes from historical drainage and peat cutting, leading to a poor diversity in terms of vegetation communities on Exmoor, mostly dominated by Molinia caerulea.

The marginal location of peatlands on Exmoor and their characteristics make them very vulnerable to climate change (Clark et al. 2010); little is also known on the effects of restoration in such heavily drained and shallow peatlands. The Exmoor Mires project is currently using an ES approach to address such issues. To present this approach, Table 2 applies our conceptual model (Fig. 1) within the specific context of Exmoor. It highlights the methods employed for monitoring ES, the range of stakeholders benefiting from potential changes, and further helps to establish a cost–benefit analysis for this area. Overall, the measurement of physical processes such as flow dynamics, DOC, gaseous fluxes or vegetation cover, through the combination of remote sensing and event-based sampling techniques, will allow the quantification of the C storage potential (Table 2). For other processes, such as the potential release from reservoirs, pre- and post-restoration situations will be compared.

Table 2. Application of the conceptual model for peatland restoration to Exmoor National Park: methods of ES quantification, and beneficiaries in SW England; with HLS, High Level Stewardship; EU WFD, the European Union Water Framework Directive; WRT, the Westcountry Rivers Trust; and ENPA, the Exmoor National Park Authority
Ecosystem goodsEcosystem ServicesProcess to quantifyCurrent monitoring approach on ExmoorBeneficiaries on Exmoor
Water quantityFuture water securityFlow dynamicsMapping of hydrological proxies using LiDAR and thermal imaging

SWW (decreased operating costs)

Devon and Cornwall water customers (reduced water bills)

Devon and Cornwall Councils (reduced flood prevention measures)

Environment Agency (reduced release from reservoir)

Land owners (HLS agreements)

National government (EU WFD)

 Capital investmentsComparison of current projected water quality with post restoration situation
 Improved predictability of water provisionRelease from reservoirs to maintain river base flowComparison of pre- and post- restoration situations 
Water qualityReduced water treatment downstreamDOC losses, colourSuspended sediment dynamics Operational costs of treatment Operational costs of reservoir dredgingRain event based monitoring using pump samplers and laboratory analyses Post restoration modelling and monitoring Comparison of pre- and post- restoration water treatment costsMeasurement of rate of sediment accumulation

SWW (reduced operating costs)

Devon and Cornwall water customers (reduced water bills)

Land owners (HLS agreements)

National government (EU WFD)

C storageClimate change mitigationC reduction and accumulation in peatMonitoring vegetation change by remote sensing techniques and field-based vegetation surveys

Landowners of Exmoor (HLS agreements; potential C offsetting market)

SWW (financial value of climate change impact)

ENPA (habitat condition)

Gaseous CO2 and CH4 emissions Monitoring of gaseous fluxes and spatial remote sensing proxies
  DOC losses, erosion ratesRain event based monitoring using pump samplers and laboratory analyses 
BiodiversityEcosystem resilienceKey indicator species cover, condition (SSSI areas)Vegetation surveys

ENPA

National government (EU Habitat directive and Natura 2000)

Environmental stakeholders, i.e. Exmoor Society, RSPB, WRT and National Trust.

Food and fibre provisionAppropriate stock densities Grazing densitiesEngagement with land owners and ENPALandowner (EU CAP, HLS agreements, enhanced value of stock through food supply chain, easier drinking water provision for stock)
Pools of drinking water for stocksRemote sensing techniques Study of impact on agricultural and economic aspects
  Vegetation communitiesRemote sensing techniques and vegetation surveys 
Cultural services DistinctivenessBare/damaged peat coverRemote sensing techniques

Public

ENPA (increased record and understanding of archaeological features)

Public, SWW customers (education on water provision and environmental processes)

Tourists

Tourism industry

Vegetation conditionVegetation monitoring and surveys
 Archaeological features Archaeological surveys and palaeo-ecological studies
Education Attendance to public events and field visitsField visits, public talks and other events, promotion of “Upstream Thinking”
RecreationMonitoring of access and public useSurveys of access and public use; use of national statistics
AestheticsPublic opinionLandscape impact studies

Cost–benefit of peatland restoration on Exmoor

A holistic cost–benefit analysis of peatland restoration is essential, first because it is likely that payments for ensuring continued provision of ES will contribute to future agricultural incomes (Dunn 2011), and second because of the increasing involvement of water companies and land owners in upland restoration schemes who need to justify their financial contribution in such activities to shareholders and customers. However, such an approach is challenging because of the temporal and spatial variability of certain ES, the subjectivity of some others and their trade-offs.

The cost of peatland restoration varies according to the aim of restoration, the level of degradation of the peat, whether land has to be purchased or not, and the techniques used on site. The choice of technique used is dependent upon local conditions, namely peat depth, drain size, slope, vegetation, erosion status and access to the site, and is therefore made on a case-by-case basis. For example, the cost of recent work by the Exmoor Mires project varied between just under £1 per m of ditch blocked (peat dams) to £16.50 per m (wooden dams), translating to an average of £490 per ha. This represents a third of the median restoration costs estimated overall at other sites (i.e. £1500 ha-1; Holden et al. 2008b). The next step in the project involves the restoration of an additional 2000 ha of peatlands over the next 5 years, and the detailed monitoring of processes to allow a holistic quantification of ES benefits (Table 2), which has a high impact on the overall cost (£2.4 m).

The benefits provided by restoration in terms of C storage and water provision have been investigated (e.g. Billett et al. 2010) to understand whether restoration can mitigate climate change, but also because they are perceived to be easier to observe. However, the time frame over which some of them occur (i.e. C accumulation) means that significant levels of uncertainty are involved, and changes to C accumulation rates as a product of restoration have high approximations. Moxey (2011) estimates that degraded peatlands could be emitting 2.9 tons of CO2 equivalent per year (tCO2e year-1), translating into 5800 tCO2 ha-1 year-1 for the 2000 ha to be restored on Exmoor (Table 3), whereas the restoration of this same area could allow a reduction of 104 000 tCO2e over 20 years, highlighting the great potential of restoration on Exmoor. The increase in mitigation costs, from £9 to £24 per tCO2e, reflects here the financial impact of monitoring, and nearly doubles the typical restoration costs estimated at £13 per tCO2e by Moxey (2011). In 2011, projections of the value of climate change impact through the shadow price of carbon for sectors covered by non-Emission Trading Scheme, such as water industries, were increasing from £56 per tCO2e in 2012 to £81 per tCO2e 20 years later (DECC 2011). For the current 2000 ha considered on Exmoor, and assuming a restoration rate of 400 ha year-1 for 5 years and stable thereafter, the value of the emission reduction is £311 700 after 5 years, and £421 800 if cumulated after 20 years (Table 4). More recent developments in Kyoto agreements now allow quantified reductions in greenhouse gases emissions from restoration to be included in both voluntary and international C markets. The potential of such income for creating new funding opportunities for restoration programmes needs to be considered in the future.

Table 3. C reduction and mitigation costs for current and future restoration areas on Exmoor, based on emission rate in degraded peatlands of 2.9 tCO2e year-1, and actual C reduction achieved through restoration of 2.6 tCO2e ha-1 yr-1 from Moxey (2011)
 MonitoringArea restored (ha)Overall cost (£)Cost/ha (£ha-1)Emissions avoided (tCO2e year-1)Total reduction over 20 years (tCO2e)Mitigation cost (£ tCO2e)
Recent restored areaNo390191k490113120 2809.4
Future areaYes20002400k12005800104 00023.8
Table 4. Potential saving from C emissions reduction due to peatland restoration (C shadow price from DECC 2011)
TimescaleCumulated restored area (ha)Shadow C price (£ tCO2e)Cumulated shadow C price (£)
Year 1 (2012)4005658 731
Year 5 (2016)200060311 672
Year 20 (2031)200081421 767

Additionally, substantial financial income can come from benefits to water provision and increased water quality. This involves quantifying benefits from, for example, more reliable base flows in dry weather, a possible improvement in peak flow control to prevent downstream flooding, improved water quality with consequent water treatment savings, but also a reduced demand for reservoir releases and cost savings from a reduced need for pumped storage. An estimate of the cost of restoration and operating water treatment works in SW England suggests that delaying the upgrade of major treatment works by 10 years would offer a benefit to cost ratio of 65 : 1 over 30 years (M. Ross, unpublished data). To address funding uncertainty, South West Water is working with Defra, the local National Park Authorities and universities to value the ‘farming of clean water’ by land managers on the uplands to yield more reliable and better quality water downstream. Alongside the value of storing C, this approach could enhance the return from restoration to the landowner, and should allow the development of a possible payment scheme that recognizes the value of upland water management to the company and its customers on top of the value of CO2 from the C market.

This example from the Exmoor Mires project shows that, altogether, there is potential for both the value of C storage and water provision to offset the costs of restoration in the long-term. If the substantial costs of monitoring often constrain the level of detail or duration at which research is carried out, a multiple ES approach should be used to estimate overall benefits, as illustrated in Table 2. Benefits also often occur at timescales going beyond those anticipated by funding bodies (Holden et al. 2011), and can also be absorbed by long-term benefits. For example, the benefit of enhancing biodiversity of moorlands may also provide a future income stream alongside C storage and water provision. Holistic monitoring of ES benefits should therefore not be overlooked, as it is essential to provide proof of concept and improve future forecasts of the effects of restoration, ideally over medium-long timescales. Ultimately, the results of this study suggest that costs of restoration remain lower than costs to society of leaving peatlands damaged (Bain et al. 2011).

Conclusion

Drainage never quite fulfilled its initial purpose of improving vegetation for grazing but caused long-lasting and extensive ecological and hydrological damage in upland peatlands. Today, the importance of peatlands goes beyond their sole ability to provide land for agriculture, grazing or fuel. Carbon pricing and new policies for rewarding landowners for practices that positively reinforce ES also mean that the future of peatlands and the way they are valued is changing.

The various results discussed here highlight the complex interaction between the eco-hydrological functioning of mires, vegetation and C storage and loss. Studies of the recovery of peatlands give some indication of the potential for improvement in terms of water quantity, and this clearly is a key to further restore other ES. However, the interaction and speed of recovery of these processes remain unclear. Monitoring the consequences of rewetting at the scale of the catchment or the region requires the detailed study of eco-hydrological processes occurring at fine spatial and temporal scales, such as individual drains over discrete rainfall events. Linking hydrological and ecological changes, such as water quality, gaseous emissions or biodiversity, in turn to build hydrological and carbon budgets, can also give a better picture of the changes occurring post-restoration. Finally, involvement of social science research, in the form of paid ES assessment, is also required to gain a complete overview of the benefits of restoration and their trade-offs. Such an approach has been adopted in the shallow peatlands of Exmoor.

The location and specific characteristics of Exmoor mean that it is essential that we gain a better understanding of the recovery in these areas: the SW is significantly warmer and influenced by a more maritime climate than the north of the UK, making the drying out of these peatlands arguably more advanced than their counterparts in the north. Other future challenges, such as those brought by climate change, might jeopardize the recovery of peatlands whether restoration occurs or not. Recent modelling work has shown that marginal peatlands in SW England are highly vulnerable to change due to temperature and precipitation changes (Clark et al. 2010), and that, depending on various climatic scenarios, they would be the first to disappear from as early as 2050 (Gallego-Sala et al. 2010). Therefore, monitoring responses to ditch blocking in SW England can tell us whether peatland restoration is feasible in a marginal peat-forming landscape. Subsequently, it is critical that we understand the way in which marginal peatlands respond to restoration, as they may provide us with ecological analogues of impending change further north.

Acknowledgements

This research received financial support from South West Water, The University of Exeter and the Knowledge Transfer Partnership programme (KTP). KTP is funded by the Technology Strategy Board along with Government Funding Organisations. We thank all project partners (ENPA, NE, EA, The Exmoor Society, Exmoor Farmers and English Heritage) and are grateful to two anonymous referees for comments, which improved the manuscript.

Ancillary