Burning management and carbon sequestration of upland heather moorland in the UK.
Abstract: The use of fire in land management has come under increasing scrutiny with regard to its potential effects on sustainability and climate change. Moorlands in the United Kingdom have traditionally used rotational burning of the heather (Calluna vulgaris) to improve the grazing and habitat, especially for grouse (Lagopus lagopus scoticus). However, these ecosystems overlie carbon-rich soils and concerns have been raised about the merits of this practice. In order to assess the impact of rotational burning on carbon balance, an investigation was undertaken on a grouse moor in the Yorkshire Dales, UK. This showed that the quantity of carbon stored above ground in heather biomass ranged from 600 to 1325 g C/[m.sup.2] (typical for UK upland heaths). However, the national UK carbon inventory assumes 200 g C/[m.sup.2], thereby appearing to underestimate considerably the importance of these habitats for carbon storage above ground.

Analysis of 2 bums in subsequent years showed that 16 [+ or -] 4% and 24 [+ or -] 5% ([+ or -] s.e.) of the above- ground material was consumed in the fires, resulting in the direct release of 103 [+ or -] 22 and 201 [+ or -] 62 g/[m.sup.2] of carbon, respectively. Indirect carbon losses, which other studies have shown to be primarily due to erosion, were estimated to release another 5-21 g C/[m.sup.2].year. The significance of other major greenhouse gas fluxes was assessed for the whole system using published parameters and models. We show that, over the burning cycle of 15-20 years, losses of carbon from burning are <10% of the total losses of carbon from the system, implying that careful burning management at this site does not have a major detrimental effect on the carbon budget, which for this moor lies within the range of an annual net loss of 34 g C/[m.sup.2].year to a net uptake of 146 g C/[m.sup.2].year.

Additional keywords: Calluna vulgaris, rotational burning, upland heath.
Article Type: Report
Subject: Carbon sequestration (Research)
Moors and heaths (Environmental aspects)
Authors: Farage, Peter
Ball, Andrew
McGenity, Terry J.
Whitby, Corinne
Pretty, Jules
Pub Date: 07/01/2009
Publication: Name: Australian Journal of Soil Research Publisher: CSIRO Publishing Audience: Academic Format: Magazine/Journal Subject: Agricultural industry; Earth sciences Copyright: COPYRIGHT 2009 CSIRO Publishing ISSN: 0004-9573
Issue: Date: July, 2009 Source Volume: 47 Source Issue: 4
Topic: Event Code: 310 Science & research
Geographic: Geographic Scope: United Kingdom Geographic Code: 4EUUK United Kingdom
Accession Number: 205745895
Full Text: Introduction

Land management in the 21st Century has become multifunctional. The primary roles that land once had for mankind, to provide food and resources, now have to compete with environmental considerations. As human demands intensify, agriculture and land-based systems must increasingly become more sustainable if environmental degradation is to be avoided, or at least lessened. Increasingly, this now includes the impacts that land management has on greenhouse gas emissions and ultimately climate change. The UN Framework Convention on Climate Change has led member countries to assess their natural carbon stocks and inventory all human-induced emissions. The aim is to encourage land management practices which protect existing carbon stocks, promote carbon sequestration, and mitigate the release of other greenhouse gasses.

Fire is one of the oldest land-management tools available, and it is still used extensively today. Wildfires, together with irresponsible and illegal use of fire, can destroy huge swathes of forest with catastrophic environmental consequences (Pearce 1998; FAO 2000). Conversely, managed or 'prescribed' burning is frequently used to prevent the spread of wildfires and is also used as a tool in a variety of land-management regimes. The many issues, aims, and potential impacts of fire are summarised by Lindenmayer and Burgman (2005). All aspects of using fire have become controversial, and its impact on biodiversity and the wider environment is often questioned (e.g. WA Forest Alliance 2003). Concerns have now been raised about the wisdom of using a process that directly releases stored carbon and other trace gases into the atmosphere and which may ultimately lead to additional soil carbon losses in the future (Pearce 2006). However, much more specific research into how fire interacts with the environment is required (Lindenmayer and Burgman 2005), in particular its impacts on the carbon balance.

The upland heaths of the UK are a long-standing and well-documented system that has been subject to burning for centuries, both natural and through human management (Tucker 2003). Moorlands are found in upland temperate zones across the world including New Zealand and Tasmania (Holden et al. 2007). They have a finely balanced ecology that makes them especially susceptible to shifts in the environment through either natural or man-made disturbance. In the UK these upland dwarf shrub heaths have long been affected by human activity, which in the last 2 centuries has included the routine burning of the dominant vegetation, heather (Calluna vulgaris). This encourages the plants to produce nutritious new shoots to improve the grazing for sheep, deer, and, in particular, grouse (Lagopus lagopus scoticus). The burning is on a small scale (<1 ha; Tucker 2003) and is designed to produce a 'patchwork' of differently aged heather stands that provide the grouse with shelter and cover for nesting. This resulting vegetative mosaic has been identified as contributing to the rich biodiversity of these habitats (Thompson et al. 1995). These upland heaths are particularly rich in invertebrate species and are home to 11% of British breeding birds (Usher and Thompson 1993). The importance of upland heaths is further emphasised by the fact that they account for 16% of all English Sites of Special Scientific Interest (SSSIs) (English Nature 2003). Dwarf shrub heaths are now recognised as a resource of international importance (Ratcliffe and Thompson 1988) and include 5 plant communities that are almost entirely confined to Great Britain [EC Habitats Directive 92/43/EEC (Thompson et al. 1995)].

Many heather moorlands in the UK uplands are endangered as a direct result of poor management, land-use change, and, now, the effects of climate change (reviewed by Holden et al. 2007). The coverage of heather moorland has declined by ~40% in England and Wales during the latter half of the 20th Century (Usher and Thompson 1993), and of the heather that remains, a substantial proportion is in poor condition. Natural England (formerly English Nature) has reported that two-thirds of upland heath SSSIs are in unfavourable condition; by far the greatest proportion of damaged SSSIs for any habitat (English Nature 2003). Although controlled burning can be a major contributor to enhancing biodiversity of this semi-natural system, paradoxically, inappropriate burning, together with overgrazing, is considered the biggest contributor to heathland damage (English Nature 2003).

Reversing the decline in upland dwarf shrub heath is vital for conserving an internationally important habitat and also preserving areas that are now valued by many for their recreational benefits. Additionally, if degradation of this habitat were to occur, it would put at risk the carbon that is stored in these systems, particularly below ground. The carbon stored in the vegetation and soils of Great Britain has been estimated at nearly 10 Gt (Milne and Brown 1997), and while the greatest proportion of soil carbon in the UK has accumulated in Scottish blanket peats (Milne and Brown 1997; Chapman et al. 2001), other upland organic soils such as humus-iron podzols, peaty podzols, and peaty gleys also hold a significant proportion of the nations's carbon store (Milne and Brown 1997). These soils are considered to be at greater risk from changes in land management than many of the blanket peats that are now protected (Chapman et al. 2001).

Upland heath vegetation in the UK is still estimated to cover 2-3 million ha (UK Biodiversity Group 1999) but little is known about how traditional moorland management, in particular routine burning, affects the carbon balance. As with all issues relating to burning world-wide, the arguments for and against the use of fire to manage land are often based on a paucity of fundamental research (Lindenmayer and Burgman 2005). Rotational burning of upland heaths provides a well-documented system to investigate the effects of managed fire on carbon dynamics. In this study, only the direct losses of carbon from vegetation burning were measured. In order to put these into context and consider the potential impact of burning on the carbon cycle as a whole, consideration was given to all the other likely greenhouse gas fluxes for the system that might be affected using best estimations from published values obtained from similar environments.

Methodology

Site

Investigation was undertaken on Mossdale Moor in upper Wensleydale, N. Yorkshire, UK (SD 810 910). Sampling was conducted at ~400 m, where annual rainfall is 1670 mm/year. The land had suffered from poor husbandry in the past, but during the last 20 years has been carefully managed for grouse and sheep. Sheep are grazed at ~0.65/ha, which is half the normal stocking density for this area. Depending on aspect, the heather is burnt on an ~15-20-year burning cycle, which is slightly longer than at lower elevation sites in the region. This management regime has resulted in the active regeneration of heather across the moor.

Sampling of biomass

Heather sampling was conducted in mid-April to early May during 2002-2004. Sampling locations were selected to represent different stages in the heather management cycle. These were: areas that had been recently burnt (within the last month); areas with 12-15 years re-growth (i.e. at peak to late building phase; Gimingham 1972); and heather that had grown for 25 years without burning (mature plants, but not yet degenerating, the oldest heather at this site). Sampling was undertaken using 0.25-[m.sup.2] quadrats (n = 12 or 16) to collect above-ground biomass of all species, live and dead, following procedures recommended by the International Biology Programme (IBP 1968). Stems were cut at ground level using secateurs. Although all species were harvested, the material was almost exclusively heather; the only others were low-growing grasses and mosses.

Losses from burning were assessed by selecting new sites where a recent burn had cut through a uniform heather stand dividing the area into burnt and unburnt sectors. Locations for sampling above-ground biomass were consequently selected randomly from areas of ~1600 [m.sup.2] and at least 10m either side of the fire boundary in 2003 and 2004. On these 2 occasions the heather was ~15 years old. The advised burning cycle for heather depends on its growth rate, which will be dependent upon location and generally lies between 8 and 20 years (MAFF 1996). At Mossdale burning is usually undertaken on plants 15-20 years old. For calculating carbon dynamics of the system, a burning cycle of 17.5 years is therefore assumed.

Below-ground biomass was collected by cutting a 0.4-[m.sup.2] by 0.3-m-deep block of soil located in the centre of the quadrat previously used for collecting above-ground material. All samples were stored at 6[degrees]C until further analyses were undertaken.

Soil sampling

A soil core was taken with a Dutch auger (5 cm diameter) to a depth of 0.3 m from each quadrat location. Mean soil bulk density for the top 0.3 m of soil was determined by carefully removing a block of soil to 0.3 m depth at each quadrat site. Four soil cores (5 [cm.sup.3] volume) were taken between ~0.10 and 0.25 m depth. The organic layer was rarely less than 0.30 m and so seldom was there any mineral horizon included in the samples.

Soil respiration

Gaseous efflux of C[O.sub.2] from the soil was measured using a portable infrared gas analysis monitor (EGM-1, PP Systems, Hitchin, UK) linked to a respiration chamber (diameter 10 cm; SRC, PP Systems, Hitchin, UK) as described previously (Sowerby et al. 2000). Soil temperature was measured at 0.05 m depth by thermocouple.

Soil respiration is subject to variation due to seasonal influences (Raich and Schlesinger 1992; Raich and Potter 1995) and is particularly sensitive to temperature (Chapman 1979; Chapman and Thurlow 1996). Several relationships between soil respiration and temperature have been established, including one for heather-covered heathland, although this was at low elevation sites (Chapman 1979):

[log.sub.e] Y = a + bT (1)

where Y is rate of respiration; T is soil temperature; and a and b are constants, -1.9573 and 0.0906, respectively.

This relationship can be extended to partition soil respiration between that produced by plant roots and that which originates from the root-zone humus (Chapman 1979):

Y = [e.sup.0.09000T] (0.0551 [square root]R + 0.0045H) (2)

where: Y is rate of respiration g C[O.sub.2]/[m.sup.2]/h; T is soil temperature [degrees]C; R is root biomass kg/[m.sup.2]; H is root-zone humus (SOM) kg/[m.sup.2].

In order to calculate annual rates of soil respiration at Mossdale, temperature readings were obtained from a meteorological recording station 20km from the sampling site. These were corrected for altitude assuming a lapse rate of 0.061[degrees]C/100m. Daily maximum and minimum air temperatures for the past 3 years were averaged to give the daily mean.

Determination of biomass dry weight

Wet weight of above-ground biomass (and below-ground biomass after washing over a sieve) was measured before the material was chopped with a commercial vegetation shredder. Any remaining long stems were cut with secateurs so that the material could be thoroughly mixed. Three subsamples were taken, weighed, and then oven-dried at 80[degrees]C overnight. The material was then re-weighed until a constant dry weight was obtained. The proportionate decrease in weight of the subsamples was used to estimate the total dry weight of each field sample. Statistical analysis was conducted using 1-way analysis of variance between treatments for the different age classes sampled.

Measurement of plant and soil carbon content

Mixed, representative samples of dried plant biomass were ground to a powder using a hammer mill and then analysed by gas chromatography using an elemental analyser (PE 2400 Series II CHNS/O Analyser, Perkin Elmer, Norwalk, CT, USA). The instrument was first calibrated with acetanilide standards (Perkin Elmer). Nitrogen content was determined so that [N.sub.2]O emissions-associated biomass burning could be assessed. The carbon content of soil obtained from the cores, each of which was first well mixed and dried at 80[degrees]C, were analysed in the same way.

Trace gas emissions and global warming potential

Trace gases emitted when vegetation was burnt in the field were estimated from the quantity of material combusted and the carbon and nitrogen content using the IPCC methodology for vegetation burning (IPCC 1996). Fluxes of C[O.sub.2e] (carbon dioxide equivalent) were obtained from carbon fluxes or from trace gases applying the global warming potentials of 23 for methane and 296 for nitrous oxide (IPCC 2001).

Results

Biomass and carbon storage

The total quantity of above-ground biomass ranged from 1262 [+ or -] 122g/[m.sup.2] at sites which had been recently burnt through 1575 [+ or -] 101 g/[m.sup.2] for 12-15-year-old heather to 2757 [+ or -] 104 g/[m.sup.2] for the oldest heather sampled. The quantity of biomass for this oldest 25-year age class was significantly greater than that recorded at the other 2 sites (F = 45.25, P < 0.01; n = 12-16). Immediately after burning, most of the material was dead (90% of the 2003 sample, Table 1), and that living comprised mainly close-growing ground flora (mosses and grasses). The proportion of dead material was significantly different for each age class (F = 91.041, P<0.01) and was lowest in the 12-15-year age class (45%) but increased to 58% in the 25-year-old heather. The amount of dead material of the 25-year age class was >2.5 times the amount present in the 12-15-year class; conversely, the quantity of live material increased with age (Table 1). Estimations of below-ground root material show that this was ~2000 [+ or -] 350g/[m.sup.2] for the recently burnt sites, significantly higher at 5000 [+ or -] 700 g/[m.sup.2] for the 12-15-year-old heather (F=9.913, P<0.01), and was again ~2000 [+ or -] 420 g/[m.sup.2] for the 25-year site.

Measured carbon content of the vegetation showed little variation, being 48.1 [+ or -] 0.6% across all age classes for aboveground biomass and 48.3 [+ or -] 0.8% for below-ground material. Such consistency is not unexpected and compares with the value of 47% generally taken as the average carbon concentration of vegetation (Allen 1989). Above-ground carbon storage is consequently dependent directly upon biomass and ranged from 606g C/[m.sup.2] to >1324g C/[m.sup.2] at the oldest sampling site (Table 2). This was significantly higher than at the 2 younger sites (F = 45.259, P<0.01). The carbon stored below ground in roots is again determined by their biomass and consequently peaked for the intermediate age class (Table 2).

Soil carbon storage

The amount of carbon stored (expressed on a w/w basis) in a soil is determined by its carbon content and the soil bulk density. The soil sampled at Mossdale Moor was an organic soil (peaty gley), with the thickness of the peat/organic layer generally 0.3-0.5 m but occasionally as little as 0.1 m and as thick as 0.8 m in a few localised areas. Consequently, the area is not typical of blanket peat, which is defined as having a peat thickness >0.5 m (UK Biodiversity Group 1999). The quantity of carbon stored across the site is therefore subject to the local topography. Mean bulk density for the top 0.3 m of soil at Mossdale varied across the sampling sites, ranging from 100kg/[m.sup.3], which is typical of peat for the 12-year age class, up to 180 kg/[m.sup.3] for the 4-year age class (F= 17.8, P<0.01; n = 16). The carbon content of the top 0.3m of soil was primarily dependent upon the bulk density and averaged 9951 [+ or -] 1204gC/[m.sup.2] across all the sites. This is at least 7 times greater than the total amount of carbon stored above ground, and only includes carbon to a depth of 0.3 m.

Burning losses

The quantity of above-ground biomass lost following burning was calculated to be 16 [+ or -] 4% after burning in 2003 and 24 [+ or -] 5% following burning in 2004 (Fig. 1). This corresponds to 103 [+ or -] 22 and 201 [+ or -] 62g C/[m.sup.2] of carbon removed from the system, respectively, on these 2 occasions. Although these emissions are neutral in terms of their effect on atmospheric C[O.sub.2] concentration, the trace gases methane and nitrous oxide are also produced and these will have an enhanced global warming potential. Applying the IPCC methodology for plant residue burning (IPCC 1996), 0.69 and 1.38g C[H.sub.4]/[m.sup.2] together with 0.014 and 0.027g [N.sub.2]O/[m.sup.2] will have been produced in 2003 and 2004, respectively. This corresponds to combined totals of 20 and 39g C[O.sub.2] equivalent/[m.sup.2] being emitted for these 2 years, respectively.

These carbon losses directly associated with burning will only occur periodically; in this location, approximately every 15-20 years. To enable comparisons with other carbon losses that occur continuously, the burning losses have been averaged assuming a 17.5-year burning cycle and are presented in Table 3.

Soil respiration

To estimate the annual amount of carbon lost to soil respiration, the model developed by Chapman (1979) for lowland heath was tested against field measurements on Mossdale Moor. Soil respiration was measured in mid April 2004 when average soil temperature was 7.6[degrees]C. Efflux of C[O.sub.2] from the soil did not differ significantly between sites (F=0.052, P > 0.05; n = 16) and averaged 0.281 [+ or -] 0.063 g C[O.sub.2]/[m.sup.2] . h. For this temperature, Eqn 1 (Chapman 1979) predicts a soil respiration rate of 0.282g C[O.sub.2]/[m.sup.2] . h, which is very close to the measured value. Although Chapman developed this model for soil respiration at a different type of site, the fact that it provides a very similar result to the measured value implies that it may provide a reasonable estimate of the annual carbon efflux at Mossdale. Using daily mean air temperatures for Mossdale, Eqn 1 predicts that the annual carbon efflux would be 799 g C/[m.sup.2] . year (2930 g C[O.sub.2]/[m.sup.2] . year). However, using air temper ature to calculate soil respiration will probably give an overestimate of this process because air has a lower thermal capacity than soil and hence shows greater extremes of temperature, particularly during the day.

[FIGURE 1 OMITTED]

These rates of soil respiration represent the gross flux of C[O.sub.2] from the soil; that is, they include active root respiration together with microbial breakdown of litter together with any other C[O.sub.2] releases. Equation 2 (Chapman (1979), which partitions soil respiration between that produced by the roots and that originating from the root-zone humus, was applied to the Mossdale data. Using this relationship the annual carbon efflux at Mossdale was calculated to be 650g C/[m.sup.2] . year, of which 470 g C/[m.sup.2] . year (72%) is attributable to root respiration. Assuming that the loss of carbon from soil processes other than roots is 28% at Mossdale, then these processes would account for 180 and 224 g C/[m.sup.2] . year using the 2 equations of Chapman (1979), respectively.

Comparison of burning associated carbon losses with carbon fluxes for the whole system

In order to assess the significance of the carbon losses connected with heather burning, the magnitude of other losses and inputs of carbon to the system need consideration. Major losses of carbon or carbon dioxide equivalent (C[O.sub.2e]) from the system include those associated with erosion, and gaseous and fluvial fluxes, while the principle input will be from net primary production. Data for these carbon fluxes have been sourced from other investigations, and are outlined in Appendix 1. The likely range of these fluxes for Mossdale Moor is summarised in Table 4 together with the parameters of this investigation.

Comparison of all the carbon losses shows that those associated with burning are estimated to represent 5-10% of the total losses of carbon or C[O.sub.2e] from the system. The biggest contributors to greenhouse gas emissions are most likely to be those that result from the direct release of carbon dioxide and methane from the soil (68-85% of the total carbon or C[O.sub.2e] losses). However, the interaction between these various factors is complex and data have been obtained from systems not identical to Mossdale. Consequently, it is only possible to give a broad range for the total net carbon budget for Mossdale Moor (-34 to 146g C/[m.sup.2] . year), with a mid-point value corresponding to a carbon sequestration rate of 56 g C/[m.sup.2] . year.

Discussion

Carbon storage

The quantity of above-ground biomass at this upland heath site, dominated by Calluna vulgaris, was typical of that recorded for similar ecosystems in Britain (reviewed by Gimingham 1972). Chapman and Clarke (1980) demonstrated consistency of above-ground standing crop with age of Calluna across a variety of sites in the British Isles, and the results from Mossdale (1200-2700 g/[m.sup.2]) fit within this general pattern. These quantities of biomass represent above-ground carbon storage in the range of 600->1300gC/[m.sup.2] . This compares with a standardised value of 200 g C/[m.sup.2] used for heathland vegetation in the UK national carbon inventory (Milne and Brown 1997; Milne and Smith 2001). Similarly, higher figures than the national inventory have also been found; Wessel et al. (2004) measured 1790g C/[m.sup.2] for a Calluna moor in North Wales and Garnett et al. (2001), working at a higher elevation site where the moorland vegetation was less dense, still obtained a higher carbon mass than that assumed in the national inventory, particularly in the areas dominated by heather.

Although the carbon density of moorland is much lower than that of woodland (e.g. 2000-6000 g C/[m.sup.2]; Milne and Brown 1997) the national inventory value (Milne and Smith 2001) seriously underestimates the quantity of carbon stored on Calluna-covered moorlands. Cruickshank et al. (1998) acknowledge this general problem with the use of generalised vegetation carbon densities and highlight the need for more reliable and specific data. The precise coverage of heather moorland in the UK is difficult to establish accurately, and some of this heather is in poor condition. Nevertheless, the national inventory calculations assume above-ground carbon storage of only 200 g C/[m.sup.2]. Such low values are normally only recorded for the first few years after burning; thereafter, biomass quantities will be significantly higher (reviewed by Gimingham 1972). Consequently, there could be 3-5-fold more carbon stored above ground on both upland and lowland heather moors across large parts of these ecosystems within UK.

The quantity of carbon stored below ground at Mossdale (averaging 9950g C/[m.sup.2]) is 7-16 times that present above ground, and this accounts only for the top 0.3 m of soil. Garnett et al. (2001) measured the soil carbon content for Moor House to be 4000-17600g C/[m.sup.2] for the top 0.3m, the highest values representing areas of blanket bog. No attempt has been made to estimate the total amount of carbon stored within the soil at Mossdale.

Burning management

Outside of natural influences, land management is the primary factor that can affect the quantity of carbon stored within an ecosystem. The act of routinely burning heather releases back into the atmosphere carbon that has been stored above ground over the previous 15-20 years. However, the results from this investigation show that with well-managed burns the proportion of carbon lost is not large because, in this case, the fires did not consume more than a quarter of above-ground biomass. This was ~100-200g C/[m.sup.2] for the 2 burns investigated, which is between one-third and up to the total estimated annual losses for the moor. However, averaged over the burning cycle, it represented 3-10% of the total system losses. The significance of estimated emissions from trace gasses produced during burning was small, even after allowing for their enhanced global warming potential.

These results show the importance of a carefully managed burning regime. This is because the quantity of material combusted during routine burning depends primarily on the intensity of the fire and this will be governed by the age of the heather, aspect, and moisture content (Tucker 2003). For example in north-east Scotland, Kayll (1966) measured above-ground losses of >93% when burning took place in the autumn and <30% for a spring burn. An additional factor with regard to the carbon budget is that some of the material will not be completely combusted, particularly the thicker stems, which will be converted to charcoal. This is extremely resistant to decomposition, having a mean residence time of 10000 years (Swift 2001). The significance of refractory charcoal as a carbon sink is illustrated by the estimation that Australian bush fires produce charcoal that sequesters >8% of the country's baseline C[O.sub.2] emission rate (Graetz and Skjemstad 2003).

The stimulation of soil erosion following burning before vegetation cover is re-established (Imeson 1971; Kinako and Gimingham 1980) has the potential to create carbon losses similar to, or in excess of, those resulting from combustion processes alone (Table 3). However, this value for erosion, providing that the fire is not intense, will represent the maximum loss. This is because burning heather in small patches results in the eroded material being moved to neighbouring unburnt areas where it can then be stabilised, so preventing its loss from the system (Kinako and Gimingham 1980). Ultimately some of this eroded material will be removed from the site, primarily as particulate organic carbon (POC). However, not all POC will necessarily be released to the atmosphere; 25% is oxidised within the riverine system, 25% is stored in the sediment as POC and 50% is transported to the ocean, where not all will be lost to the atmosphere (Allan and Castillo 2008).

Burning-induced erosion is potentially the most serious carbon loss for the system, especially if uncontrolled or accidental fires were to burn the soil surface. This will not only directly release carbon as the peat is combusted, but it is quite likely that significant erosion will follow during the proceeding months and even years (Maltby 1980; Maltby et al. 1990). Drying of organic soils increases their susceptibility to erosion and oxidation, but a recent investigation by Worrall et al. (2007) has shown that the depth to the water table is less in areas that are subject to rotational burning. Cessation of burning would halt all carbon losses attributed directly to scheduled fires and associated erosion. However, the effect on the overall carbon balance for the system will depend on how the primary carbon input to the system, i.e. heather productivity, responds in the longer term. The purpose of burning heather is to produce an abundance of nutritious shoots from young vegetation (Tucker 2003). Individual plants at this growth stage can be expected to be particularly productive but overall biomass accumulation at the pioneer stage will be limited until complete canopy closure occurs (e.g. Miller 1979). Forrest (1971) established that the most productive age was 10-18 years, which coincided with peak green shoot biomass. However, there is some contradiction concerning how much production declines with old age. Miller (1979) found little decline with age although others have recorded results to the contrary (e.g. Chapman 1967; Barclay-Estrup 1970; Gimingham 1972; Chapman et al. 1975). Milne et al. (2002) also found that production increased up to the mature phase and then declined. As occurs with the pioneer stage, productivity will decline during the degenerate phase simply because the canopy becomes more open, leading to a loss of productivity per unit ground area. Crucial to the aspect of burning is whether combustion-related carbon losses, and the reduction in productivity that will occur before canopy closure can be attained, is offset by reduction in heather productivity associated with keeping increasingly degenerate plants. Simply delaying the time interval between burns is unlikely to benefit the carbon balance because old heather, unless markedly degenerated, generally burns at higher temperatures, as there is more woody material present (Kenworthy 1963; Hobbs and Gimingham 1984). Consequently, more material will be combusted, there is a greater likelihood that the soil surface will be damaged, vegetative regeneration and root propagation are reduced (Webb 1986), and many more seeds destroyed (Tucker 2003). However, at present there is insufficient research data to fully estimate the role that burning has on the local heather community (Holden et al. 2007).

Overall the effects of burning on the carbon cycle for heather moorland systems will depend on a complex interaction of many site-specific factors, but our results suggest that if well-managed burns are used, the net effect on the carbon balance can be relatively small. There are no studies that have included analyses of how heather management might affect the carbon balance. Forrest (1971) did report that of 7 floristically varied sites at Moor House, those with the highest production were the ones which had been burnt recently. Garnett et al. (2000), also working at Moor House albeit on blanket bog, estimated from peat profiles that unburnt areas accumulated carbon at a rate of 169 g C/[m.sup.2] . year, while sites burnt every 10 years still sequestered carbon, but at a reduced rate of 97g C/[m.sup.2] . year. This, however, is atypical because burning on blanket bog is not normally practised (Tucker 2003). Further investigations on the responses of heather productivity to management practices are required.

Other fluxes of carbon

Based upon data obtained from similar systems, it is estimated that the loss of carbon attributed to rotational burning accounts for <10% of the total system carbon losses when considered over the burning cycle. The greatest contribution to greenhouse gas emissions will almost certainly be derived from gaseous emissions originating from the soil. Of these, soil respiration and methane flux, the latter because of its enhanced warming potential, can be expected to be the most significant.

Most published measurements of soil respiration from uplands have been made on sites overlain by blanket bog, peatlands, or from systems that have a near-tundra environment and so do not provide for ideal comparison with our results. Soil respiration in peatlands is enhanced by drainage, and consequently, C[O.sub.2] efflux varies not only with site but also with previous land management. Published results include those of Clymo and Pearce (1995) who measured an efflux (based on gas concentrations and averaged for differing topography) of 21 g C/[m.sup.2] . year for a raised bog in south-west Scotland with a maximum peak equivalent to 105 g C/[m.sup.2] . year. Giblin et al. (1991) measured 98 g C/[m.sup.2] . year but this was for hill-top/hill-slope heath in Alaska overlying permafrost. More recently, the technique of eddy covariance has been used to establish net gas budgets. Using this methodology Beverland et al. (1996) measured a net sink (40.6gC/[m.sup.2] . year) for a Calluna blanket bog system in Scotland.

With regard to the overall carbon balance, we provide tentative values for Mossdale Moor ranging from a net source releasing 34g C/[m.sup.2] . year (153g C[O.sub.23/[m.sup.2] . year) to a net sink sequestering 146 g C/[m.sup.2] . year (298 g C[O.sub.2e]/[m.sup.2] . year). There are few detailed carbon inventories of upland systems; again most attention has been directed towards blanket bog communities, although many of these have Calluna populations associated with them, e.g. Moor House National Nature Reserve (Garnett et al. 2001). However, deep peat systems have historically accumulated carbon in much greater quantities than upland heaths and so do not make for ideal comparisons. Evidence is contradictory as to whether blanket bogs still sequester carbon, or if they are now net carbon emitters. Carbon accumulation has been estimated to be in the range 20-50gC/[m.sup.2] . year (Beverland et al. 1996; Clymo et al. 1998; Cannell et al. 1999; Worrall et al. 2003). Other investigators have used a variety of techniques to estimate rates of carbon addition in peatlands. Gorham (1991) calculated that northern peatlands accumulated 29 g C/[m.sup.2] . year, values that are very similar to the 21 gC/[m.sup.2] . year for the same region calculated by Clymo et al. (1998). Cannell et al. (1993) estimated that Scottish peatlands were a carbon sink of 40-70g C/[m.sup.2] . year and Garnett (1998) calculated that Moor House accumulated 27g C/[m.sup.2] . year. These results suggest that in spite of the effluxes of carbon measured from organic upland soils, small rates of carbon sequestration do occur, although this issue is contentious.

The contribution of methane emissions may result in bogs and peatlands being either neutral or net carbon emitters (Cannell et al. 1993; Crill et al. 2000). It is becoming increasingly recognised that the fluxes of greenhouse gases associated with peatlands is highly complex (Sirin and Laine 2007). Methane fluxes depend primarily on soil moisture, waterlogged soils emitting methane while drier mineral soils oxidise this gas. The soil at Mossdale, although reasonably drained was anaerobic in places and on balance is likely to be a net methane emitter, although measurement would be needed to confirm this. It is possible that burning may affect the methane budget but there is little experimental evidence available; that which exits comes from other ecosystems. Castaldi and Fierro (2005) found no effect of experimental fire on dry Mediterranean shrubland, whereas Burke et al. (1997) established that burning slightly increased the methane sink in Canadian boreal forest and Nakano et al. (2006) found that fire in a boreal peat swamp forest turned the system into an emitter because of increased water content.

Conclusion

Heather moorland represents an important component of the UK's carbon store. The soils underlying these upland ecosystems have been acknowledged as important carbon reservoirs, but this investigation has additionally highlighted the underestimate of the carbon stored above ground. Significantly, our findings also show that traditional moorland management incorporating rotational heather burning need not have a major detrimental effect on the carbon balance of this system. Furthermore, this study demonstrates the importance of verifying assumptions about land management practices; the intuitive belief that fire will inevitably diminish carbon stocks was found to be incorrect, providing that rotational burning is carefully planned. Each land management system is unique and appropriate fire regimes must be tailored to the characteristics of the system (Lindenmayer and Burgman 2005). Nevertheless, the use of fire to maintain any system is not without risk, particularly on fragile organic soils. Fires that burn out of control put at jeopardy carbon that has accumulated over several centuries together with the consequential destruction of natural communities.

It is perhaps not surprising that traditional land management that has been practised for many generations has created a sustainable system that not only supports the rural economy but is also beneficial for biodiversity and maintains an internationally important ecosystem in addition to conserving carbon stocks. However, further research is urgently required. The range of potential carbon fluxes highlights the need for more investigation of non-blanket bog moorland systems, particularly in relation to heather productivity and its response to management practices, as well as seasonal and regional measurements of greenhouse gas fluxes.

Appendix 1. Losses and inputs of carbon or carbon equivalent for upland heaths and moors

Erosion

Upland areas are particularly prone to erosion. General erosion is included in the fluvial losses, particularly POC. Only potential erosion caused by specific land management at the site is dealt with in this section. For Mossdale, the most significant factor likely to cause increased erosion is burning because vegetation cover is regarded as crucial in protecting against soil loss (Imeson 1971; Kinako and Gimingham 1980). Severe ecological damage and erosion can be expected if the peat surface becomes burnt (Maltby 1980; Maltby et al. 1990). There is no evidence that this occurs at Mossdale. Other contributors to erosion (overgrazing) are not considered to be a problem for this site.

Kinako and Gimingham (1980) found that with normal routine burning in NE Scotland stability returned to the soil surface within 15-20 months and they found no relationship with slope; total erosion following burning averaged 0.4 cm. Imeson (1971) (N York Moors) found that where Calluna is >20 cm high, erosion is absent and litter accumulation occurs. Vegetation cover prevented erosion within 2-3 years of burning and total erosion was c. 1.6 cm following burning.

The mean loss of soil reported by Imeson (1971) and Kinako and Gimingham (1980) is therefore 1 cm following a burning episode. As burning at Mossdale is not usually intense, and considerable litter remains, excessive erosion would not be expected: a 1-cm loss is therefore a reasonable estimate. Taking the average soil carbon content measured at Mossdale Moor, a 1-cm soil removal relates to a carbon loss from the system of 230 g C/[m.sup.2] per burn. This will be 13 g C/[m.sup.2] . year averaged over the burning cycle.

Gaseous fluxes

Methane

Efflux of methane is dependent upon the height of the water table. The contribution to carbon loss is relatively small with most measured rates commonly ranging between 0 and 11 g C[H.sub.4]/[m.sup.2] . year (Daulat and Clymo 1989; MacDonald et al. 1998; Watson and Nedwell 1998), although diurnal peaks equivalent to 3 times this can be recorded during high summer temperatures (Hargreaves and Fowler 1998; MacDonald et al. 1998). Although the amount of carbon released is relatively small the enhanced greenhouse warming potential of methane increases its potency such that, although a methane flux of 5 g C[H.sub.4]/[m.sup.2] . year represents a carbon flux of just 3.75 g C/[m.sup.2] . year, this is equivalent to the emission of 115 g C[O.sub.2]/[m.sup.2] . year. As these measurements are for peatlands, those for Mossdale with its peaty gley soils are likely to be lower (MacDonald et al. 1996). The average annual methane flux from Mossdale Moor is consequently most likely to be within the range 1-11 g C[H.sub.4]/[m.sup.2] . year.

Nitrous oxide

[N.sub.2]O is produced by moorlands because the acid peaty soils favour the release of nitrogen through denitrification as [N.sub.2]O rather than [N.sub.2] (Dutch and Ineson 1990). Emissions from upland moors are of the order 0.03-0.08 g [N.sub.2]O/[m.sup.2] . year (Chapman et al. 2001) which has a greenhouse warming potential equivalent to 8.88-23.68 g C[O.sub.2]/[m.sup.2] . year.

Isoprene

Biogenic emissions of volatile organic compounds, isoprene and monoterpenes, will also occur. Fewer data are available, and reactive carbon fluxes are very sensitive to land cover and climate (Grote and Niinemets 2008). Isoprene emissions for gorse heathland in early summer have been measured at 43 g/[km.sup.2] . h (Owen et al. 2003). This would represent a carbon loss of 0.3 g C/[m.sup.2] . year if the efflux of isoprene was constant. However, isoprene emissions generally correlate with rising temperature during the day and are highest in the summer. Consequently, the annual loss of carbon to isoprene emissions in the study will be very small.

Fluvial losses

Losses of carbon in water can be divided into several categories: particulate organic carbon (POC), dissolved organic carbon (DOC), dissolved inorganic carbon (DIC), dissolved C[O.sub.2], and dissolved C[H.sub.4].

POC

Published values for POC lost through transport by streams and rivers in upland areas are particularly variable as this reflects the fact that up to 80% of organic matter is transported during brief, high flow events, which in the north Pennines occur for <5% of the year (Crisp and Robson 1979). Consequently, sampling that misses these spates may seriously underestimate fluvial losses. One of the highest values recorded in the north Pennines is 56 g C/[m.sup.2] . year in the form of peat transport (Crisp and Robson 1979). Many lower values have been published but these generally do not include high flow events. Hope et al. (1994) give a range for European temperate moorland and grassland of 0.2-50 g C/[m.sup.2] . year, with a typical value of 12.0 g C/[m.sup.2] . year where precipitation is >1000mm/year. The value for Mossdale Moor will most likely reflect these values and lie somewhere within the broad range 12-59 g C/[m.sup.2] . year.

DOC

As with POC, recorded values for DOC vary. Hope et al. (1994) provide a value of 4.3g C/[m.sup.2] . year for rivers draining European temperate moorland and grassland and specifically 2-3 g C/[m.sup.2] . year for Yorkshire, although this may not account for high flow events. Tipping et al. (1997) give a value of 3.6 g C/[m.sup.2] . year for the Ure catchment as a whole, which includes Mossdale Moor, although the figure will undoubtedly be higher in the head waters where Mossdale Moor is located. Grieve (1984) measured 8.4 g C/[m.sup.2] . year for a site in the Ochil Hills draining blanket bog and peaty soils. This value included storm events and was for moorland with precipitation similar to Mossdale and so is more likely to reflect the situation for this investigation. Interestingly, Worrall et al. (2007) found that DOC content showed significant decrease in areas where rotational burning occurred.

DIC

There are fewer published values for DIC as it is less commonly measured. Worrall et al. (2003) used a value of 5.9 g C/[m.sup.2] . year for Moor House, although the value for Mossdale is likely to be lower because Moor House overlies calcareous rock. Additionally, as DIC represents inorganic carbon it is ignored in this study.

Dissolved C[O.sub.2]

This will be influenced by the local geology. Dawson et al. (2002) measured dissolved C[O.sub.2] in streams draining heather moorland and obtained values <1.0 g C/[m.sup.2] . year, whereas Worrall et al. (2003) used 3.8 g C/[m.sup.2] . year at Moor House.

A summary of the estimate fluvial fluxes of carbon from Mossdale Moor is given below:

Dissolved C[H.sub.4]

Very few data are available and since the quantity of carbon involved is small (<0.001 g C/[m.sup.2] . year; Dawson et al. 2002) it will be excluded in this analysis.

Inputs

Input to the system will be from net primary production, although a small quantity of carbon may enter from rainfall: e.g. Worrall et al. (2003) took this to be 4.2 g C/[m.sup.2] . year at Moor House.

Heather productivity has been extensively studied, although Chapman and Clarke (1980) note that this work has been concentrated at a relatively limited number of sites. Above-ground dry matter production rates taken from the literature are normally within the range 175-300 g/[m.sup.2] . year, although a recent investigation by Milne et al. (2002) obtained values ranging between 350 and 500 g/[m.sup.2] . year, with northern England averaging 425 g/[m.sup.2] . year for the building and mature phases of heather. Milne et al. (2002) point out that most studies have been focused on northern Scotland and that this may explain why the figures they obtained for many English locations were higher. Milne et al. (2002) also reported that production increased up to the mature phase and then declined.

Many production studies include only annual green shoot production because of the difficulties of including wood increment and flower production. Wood production may account for an additional 29% of dry matter increase relative to green shoot production (Forrest 1971), although Chapman et al. (1975) estimate an extra 8%. Miller (1979) calculated flower production to be 23% of shoot production, but Forrest (1971), Grant (1971), and Chapman et al. (1975) all give values in the range 3-7%.

Estimation of root production is problematic because of the difficulties encountered with separating root material from organic soils. A few studies have investigated this aspect; in particular, Forrest (1971) and Forrest and Smith (1975) estimated annual belowground productivity to be 183 g/[m.sup.2] . year at Moor House, which compares with an above-ground production rate of 168 g/[m.sup.2] . year. The other major detailed investigation of below-ground production is that conducted by Chapman, working on lowland heath in Dorset. Using respiration measurements, root production is estimated at 400 g/[m.sup.2] . year while above-ground production is 300 g/[m.sup.2] . year (Chapman et al. 1975). Aerts et al. (1992) calculated Calluna root productivity to be 160 g/[m.sup.2] . year, 25% of stand productivity on dry heathland in The Netherlands.

For estimating heather production at Mossdale Moor, the older literature would imply that above-ground heather production should lie between 175 and 300 g/[m.sup.2] . year but the more recent work of Milne et al. (2002) suggests a higher range with a figure of 425 g/[m.sup.2] . year for the building and mature phases in northern England. Consequently, the value of dry matter production for Mossdale almost certainly lies within the very broad range of 175-500 g/[m.sup.2] . year (84-241 g C/[m.sup.2] . year) and is most likely to be near the upper value, especially if wood and flower production are included. With regard to root production, the 2 British studies (measuring at very different sites) both found that this exceeded above-ground production (9% and 33% greater), although the Dutch study gave contrary results. A conservative assumption for Mossdale is that root production is similar to shoot production (84-241 g C/[m.sup.2] . year).

http://www.publish.csiro.au/journals/ajsr

Acknowledgments

We express our appreciation to Mr Hugh van Cutsem for his enthusiastic support for this work, without whom the field investigation would not have been possible. We are also indebted to Mr G. Roberts and Mr N. Parker for their local knowledge and assistance with the sampling trips. Technical support for analysing the samples was provided by Elisabeta Torok, Daniela Belici, and Cristina Jelescu from the University of Brasov, Romania. Acknowledgment is also due to Paul Igboji for his soil respiration measurements.

Manuscript received 24 April 2008, accepted 6 March 2009

References

Aerts R, Bakker C, de Caluwe H (1992) Root turnover as determinant of C, N, and P in a dry heathland ecosytem. Biogeochemistry 15, 175-190. doi: 10.1007/BF00002935

Allan JD, Castillo MM (2008) Structure and function of running waters. In 'Ecosytem metabolism'. 2nd edn (Springer: Dordrecht)

Allen SE (1989) 'Chemical analysis of ecological materials.' (Blackwell Scientific Publications: Oxford, UK)

Barclay-Estrup P (1970) The description and interpretation of cyclical processes in a heath community. II Changes in biomass and shoot production during the Calluna cycle. Journal of Ecology 58, 243-249. doi: 10.2307/2258179

Beverland IJ, Moncrieff JB, Oneill DH, Hargreaves KJ, Milne R (1996) Measurement of methane and carbon dioxide from peatland ecosystems by the conditional sampling technique. Quarterly Journal of the Royal Meteorological Society 122, 819-838. doi: 10.1002/qj.49712253203

Burke RA, Zepp RG, Tarr MA, Miller WL, Stocks BJ (1997) Effect of fire on soil-atmosphere exchange of methane and carbon dioxide in Canadian boreal forest sites. Journal of Geophysical Research 102, 29289-29300. doi: 10.1029/97JD01331

Cannell MGR, Dewar RC, Pyatt DG (1993) Conifer plantations on drained peatlands in Britain: net gain or loss of carbon. Forestry 66, 353-369. doi: 10.1093/forestry/66.4.353

Cannell MGR, Milne R, Hargreaves KJ, Brown TAW, Cruickshank MM, et al. (1999) National inventories of terrestrial carbon sources and sinks: the U.K. experience. Climatic Change 42, 505-530. doi: 10.1023/A:1005425807434

Castaldi S, Fierro A (2005) Soil-atmosphere methane exchange in undisturbed and burned Mediterranean shrubland of southern Italy. Ecosystems 8, 182-190. doi: 10.1007/s 10021-004-0093-z

Chapman SB (1967) Nutrient budgets for a dry heath ecosystem in the south of England. Journal of Ecology 55, 677-689. doi: 10.2307/2258418

Chapman SB (1979) Some interrelationships between soil and root respiration in lowland Calluna heathland in southern England. Journal of Ecology 67, 1-20. doi: 10.2307/2259333

Chapman SB, Clarke RT (1980) Some relationships between soil, climate, standing crop and organic matter accumulation within a range of Calluna heathlands in Britain. Bulletin d'Ecologie 11, 221-232.

Chapman SB, Hibble J, Rafarel CR (1975) Net aerial production by Calluna vulgaris on lowland heath in Britain. Journal of Ecology 63, 233-258. doi: 10.2307/2258852

Chapman SJ, Thurlow M (1996) The influence of climate on C[O.sub.2] and C[H.sub.4] emissions from organic soils. Agricultural and Forest Meteorology 79, 205-217. doi: 10.1016/0168-1923(95)02283-X

Chapman SJ, Towers W, Williams BL, Coull MC, Paterson E (2001) Review of the contribution to climate change of organic soils under different land uses. Scottish Executive Central Research Unit, Edinburgh.

Clymo RS, Pearce ME (1995) Methane and carbon dioxide production in, transport through, and efflux from a peatland. Philosophical Transactions of the Royal Society of London Series A 351, 249-259. doi: 10.1098/rsta.1995.0032

Clymo RS, Turunen J, Tolonen K (1998) Carbon accumulation in peatland. Oikos 81, 368-388. doi: 10.2307/3547057

Crill P, Hargreaves K, Korhola A (2000) The role of peat in Finnish greenhouse gas balances. 10/2000 Ministry of Trade and Industry, Helsinki, Finland.

Crisp DT, Robson S (1979) Some effects of discharge upon the transport of animals and peat in a north Pennine headstream. Journal of Applied Ecology 16, 721-736. doi: 10.2307/2402848

Cruickshank MM, Tomlinson RW, Devine PM, Milne R (1998) Carbon in the vegetation and soils of northern Ireland. Biology and Environment: Proceedings of the Royal Irish Academy 98b, 9-21.

Daulat WE, Clymo RS (1989) Effects of temperature and watertable on the efflux of methane from peatland surface cores. Atmospheric Environment 32, 3207-3218. doi: 10.1016/S1352-2310(98)00078-8

Dawson JJC, Billett MF, Neal C, Hill S (2002) A comparison of particulate, dissolved and gaseous carbon in two contrasting upland streams in the UK. Journal of Hydrology 257, 226-246. doi: 10.1016/S0022-1694(01)00545-5

Dutch J, Ineson P (1990) Denitrification of an upland forest site. Forestry 63, 363-377. doi: 10.1093/forestry/63.4.363

English Nature (2003) 'England's best wildlife and geological sites. The Condition of Sites of Scientific Special Interest in England in 2003.' (English Nature: Peterborough, UK)

FAO (2000) FAO global forest fire assessment 1990-2000. Working Paper No. 55 (495 p.). (JG Goldammer, RW Mutch, compilers). Published online: www.fire.unifreiburg.de/programmes/un/fao/Wp55_eng.pdf

Forest Alliance WA (2003) Submission to the COAG Inquiry into Bushfire Mitigation and Management: The myths of prescribed burning. Published online: www.coagbushfireenquiry.gov.au/subs_word/1_robertson_wa_fire_alliance.doc

Forrest GI (1971) Structure and production of north Pennine blanket bog vegetation. Journal of Ecology 59, 453-479. doi: 10.2307/2258325

Forrest GI, Smith RAH (1975) The productivity of a range of blanket bog types in the northern Pennines. Journal of Ecology 63, 173-202. doi: 10.2307/2258850

Garnett MH (1998) Carbon storage in Pennine moorland and response to change. PhD Thesis, University of Newcastle upon Tyne, UK.

Garnett MH, Ineson P, Stevenson AC (2000) Effects of burning and grazing on carbon sequestration in a Pennine blanket bog, UK. The Holocene 10, 729-736. doi: 10.1191/09596830094971

Garnett MH, Ineson P, Stevenson AC, Howard DC (2001) Terrestrial organic carbon storage in a British moorland. Global Change Biology 7, 375-388. doi: 10.1046/j.1365-2486.2001.00382.x

Giblin AE, Nadelhoffer KJ, Shaver GR, Laundre JA, McKerrow AJ (1991) Biogeochemical diversity along a riverside toposequence in arctic Alaska. Ecological Monographs 61, 415-435. doi: 10.2307/2937049

Gimingham CH (1972) 'Ecology of heathlands.' (Chapman and Hall: London)

Gorham E (1991) Northern peatlands: role in the carbon cycle and probable responses to climate warming. Ecological Applications 1, 182-195. doi: 10.2307/1941811

Graetz RD, Skjemstad JO (2003) The charcoal sink of biomass burning on the Australian continent. CSIRO Atmospheric Research Technical paper No. 64, Victoria, Australia.

Grant SA (1971) Interactions of grazing and burning on heather moors. 2. Effects on primary production and level of utilization. Grass and Forage Science 26, 173-182. doi: 10.1111/j.1365-2494.1971. tb00660.x

Grieve IC (1984) Concentrations and annual loading of dissolved organic matter in a small moorland stream. Freshwater Biology 14, 533-537. doi: 10.1111/j.1365-2427.1984.tb00173.x

Grote R, Niinemets U (2008) Modeling volatile isoprenoid emissions--a story with split ends. Plant Biology 10, 8-28.

Hargreaves KJ, Fowler D (1998) Quantifying the effects of water table and soil temperature on the emission of methane from peat wetland at the field scale. Atmospheric Environment 32, 3275-3282. doi: 10.1016/S1352-2310(98)00082-X

Hobbs RJ, Gimingham CH (1984) Studies on fire in Scottish heathland communities: I. Fire characteristics. Journal of Ecology 72, 223-240. doi: 10.2307/2260015

Holden J, Shotbolt L, Bonn A, Burt TP, Chapman PJ, et al. (2007) Environmental change in moorland landscapes. Earth-Science Reviews 82, 75-100. doi: 10.1016/j.earscirev.2007.01.003

Hope D, Billett MF, Cresser MS (1994) A review of the water export of carbon in river water: fluxes and processes. Environmental Pollution 84, 301-324. doi: 10.1016/0269-7491(94)90142-2

IBP (1968) 'Methods for the measurement of the primary production of srassland.' C Milner, RE Hughes with contributions on Measurement of the primary production of dwarf shrub heaths. International Biological Programme. (Blackwell Scientific Publications: Oxford and Edinburgh)

Imeson AC (1971) Heather burning and soil erosion on the North Yorkshire moors. Journal of Applied Ecology 8, 537-542. doi: 10.2307/2402889

IPCC (1996) 'Climate change 1995: The science of climate change.' Contribution of Working Group I to the Second Assessment Report of the Intergovernmental Panel on Climate Change. (Cambridge University Press: Cambridge, UK)

IPCC (2001) 'Climate change 2001: The scientific basis.' Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change. (Cambridge University Press: Cambridge, UK)

Kayll AJ (1966) Some characteristics of heath fires in north-east Scotland. Journal of Applied Ecology 3, 29-40. doi: 10.2307/2401664

Kenworthy JB (1963) Temperatures in heather burning. Nature 200, 1226. doi: 10.1038/2001226a0

Kinako PDS, Gimingham CH (1980) Heather burning and soil erosion on upland heaths in Scotland. Journal of Environmental Management 10, 277-284.

Lindenmayer D, Burgman M (2005) 'Practical conservation biology.' (CSIRO Publishing: Melbourne, Vic.)

MacDonald JA, Fowler D, Hargreaves KJ, Skiba U, Leith ID, Murray MB (1998) Methane emission rates from a northern wetland; response to temperature, water table and transport. Atmospheric Environment 32, 3219-3227. doi: 10.1016/S1352-2310(97)00464-0

MacDonald JA, Skiba U, Sheppard LJ, Hargreaves KJ, Smith KA, Fowler D (1996) Soil environmental variables affecting the flux of methane from a range of forest, moorland and agricultural soils. Biogeochemistry 34, 113-132. doi: 10.1007/BF00000898

MAFF (1996) 'The heather and grass burning code.' (MAFF: London)

Maltby E (1980) The impact of severe fire on Calluna moorland in the North York moors. Bulletin d'Ecologie 11, 683-708.

Maltby E, Legg CJ, Proctor MCF (1990) The ecology of severe moorland fire on the North York moors: effects of the 1976 fires, and subsequent surface vegetation development. Journal of Ecology 78, 490-518. doi: 10.2307/2261126

Miller GR (1979) Quantity and quality of the annual production of shoots and flowers by Calluna vulgaris in north-east Scotland. Journal of Ecology 67, 109-129. doi: 10.2307/2259340

Milne JA, Pakeman RJ, Kirkham FW, Jones IP, Hossell JE (2002) Biomass production of upland vegetation types in England and Wales. Grass and Forage Science 57, 373-388. doi: 10.1046/j.1365-2494.2002.00339.x

Milne R, Brown TA (1997) Carbon in the vegetation and soils of Great Britain. Journal of Environmental Management 49, 413-433. doi: 10.1006/jema.1995.0118

Milne R, Smith P (2001) Country specific data for the UK relating to Articles 3.3 and 3.4 of the Kyoto protocol: An update. In 'UK emissions by sources and removals by sinks due to land use, land use change and forestry activities'. Annual report for DETR, Contract EPG1/1/160. (Ed. R Milne) pp. 61-70. (DETR: London)

Nakano T, Takeuchi W, Inoue G, Fukuda M, Yasuoka Y (2006) Temporal variations in soil-atmosphere methane exchange after fire in a peat swamp forest in West Siberia. Soil Science and Plant Nutrition 52, 77-88.

Owen SM, MacKenzie AR, Stewart H, Donovan R, Hewitt CN (2003) Biogenic volatile organic compound (VOC) emission estimates from an urban tree canopy. Ecological Applications 13, 927-938. doi: 10.1890/01-5177

Pearce F (1998) Playing with fire. New Scientist 2126, 36.

Pearce F (2006) Grouse-shooting popularity boosts global warming. New Scientist 191, 2564. doi: 10.1016/S0262-4079(06)60204-2

Raich JW, Potter CS (1995) Global patterns of carbon dioxide emissions from soils. Global Biogeochemical Cycles 9, 23-36. doi: 10.1029/94GB02723

Raich JW, Schlesinger WH (1992) The global carbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus 44, B81-99.

Ratcliffe DA, Thompson DBA (1988) The British uplands: their ecological character and international significance. In 'Ecological change in the uplands'. (Eds MB Usher, DBA Thompson) pp. 9-36. (Blackwell Scientific Publications: Oxford, UK)

Sirin A, Laine J (2007) Peatlands and Green house gasses. In 'Assessment on peatlands, biodiversity and climate change'. (Eds F Parish, A Sirin, D Charman, H Joosten, T Minaeva, M Silvius) pp. 7.1-7.21. (Global Environment Centre: Kuala Lumpur, and Wetlands International: Wageningen)

Sowerby A, Blum H, Gray TRG, Ball AS (2000) The decomposition of Lolium perenne in soils exposed to elevated C[O.sub.2]: comparisons of mass loss of litter with soil respiration and soil microbial biomass. Soil Biology & Biochemistry 32, 1359-1366. doi: 10.1016/S0038-0717 (00)00045-6

Swift RS (2001) Sequestration of carbon by soil. Soil Science 166, 858-871. doi: 10.1097/00010694-200111000-00010

Thompson DBA, Macdonald AJ, Marsden JH, Galbraith CG (1995) Upland heather moorland in Great Britain: a review of international importance, vegetation change and some objectives for nature conservation. Biological Conservation 71, 163-178. doi: 10.1016/0006-3207(94) 00043-P

Tipping E, Marker AFH, Butterwick C, Collett GD, Cranwell PA, et al. (1997) Organic carbon in the Humber rivers. The Science of the Total Environment 194-195, 345-355. doi: 10.1016/S0048-9697(96) 05374-0

Tucker G (2003) 'Review of the impacts of heather and grassland burning in the uplands on soils, hydrology and biodiversity.' English Nature Research Reports, Report No. 550. (English Nature: Peterborough, UK)

UK Biodiversity Group (1999) 'Tranche 2 Action Plans--Volume VI: Terrestrial and freshwater species and habitats.' (HMSO: London)

Usher MB, Thompson DBA (1993) Variation in the upland heathlands of Great Britain: conservation importance. Biological Conservation 66, 69-81. doi: 10.1016/0006-3207(93)90136-0

Watson A, Nedwell DB (1998) Methane production and emission from peat: the influence of anions (sulphate, nitrate) from acid rain. Atmospheric Environment 32, 3239-3245. doi: 10.1016/S1352-2310 (97)00501-3

Webb NR (1986) 'Heathlands.' (Collins: London)

Wessel WW, Tietema A, Beier C, Emmett BA, Penuelas J, Riis-Nielsen T (2004) A qualitative ecosystem assessment for different shrublands in Western Europe under impact of climate change. Ecosystems 7, 662-571. doi: 10.1007/s10021-004-0219-3

Worrall F, Armstrong A, Adamson JK (2007) The effects of burning and sheep-grazing on water table depth and soil water quality in a upland peat. Journal of Hydrology 339, 1-14. doi: 10.1016/j.jhydrol. 2006.12.025

Worrall F, Reed M, Warburton J, Burt TP (2003) Carbon budget for a British upland peat catchment. The Science of the Total Environment 312, 133-146. doi: 10.1016/S0048-9697(03)00226-2

Peter Farage (A), Andrew Ball (B,C), Terry J. McGenity (A), Corinne Whitby (A), and Jules Pretty (A)

(A) Department of Biological Sciences, University of Essex, Colchester, UK.

(B) School of Biological Sciences, Flinders University, Adelaide, Australia.

(C) Corresponding author. Email: andy.ball@flinders.edu.au
gC[O.sub.2e]/
Parameter                    gC/[m.sup.2] x year   [m.sup.2] x year

Particulate organic carbon          12-59               44-216
Dissolved organic carbon             2-9                 7-33
Dissolved C[O.sub.2]                 1-4                 4-15


Table 1. Partitioning total above-ground biomass between live and
dead components ([+ or -] s.e., n = 8)

Samples were collected on Mossdale Moor, north Yorkshire, 2003. The
age class is the time since the sampling location was last subject
to rotational burning

Age class   Live vegetation     Dead vegetation
(years)
                        g/[m.sup.2])

0            110 [+ or -] 42   1039 [+ or -] 102
12-15        756 [+ or -] 67    608 [+ or -] 114
25          1159 [+ or -] 69   1623 [+ or -] 105

Table 2. Quantity of carbon stored in all vegetation above and below
ground ([+ or -] s.e., n = 12-16)

The age class is the time since the sampling location was last subject
to rotational burning, Mossdale Moor, north Yorkshire, 2003

Age class     Above ground       Below ground
(years)
                            (gC/[m.sup.2])

0           606 [+ or -] 58    959 [+ or -] 169
12-15       756 [+ or -] 48    2404 [+ or -] 336
25          1324 [+ or -] 50   1194 [+ or -] 201

Table 3. Carbon fluxes directly associated with heather burning
on Mossdale Moor, north Yorkshire (n=8)

The age class is the time since the sampling location was last
subject to rotational burning

                                                   gC[O.sub.2e]/
                     gC/[m.sup.2]   gC/[m.sup.2]    [m.sup.2] x
Parameter               x burn       x year (A)      year (A)

Combusted
  vegetation (B)        88-210          5-12           18-44
Burning associated     0.5-1.1       0.03-0.06          1-2
  trace gasses (B)
Burning associated
  erosion               88-368          5-21           18-77

(A) Average annual values based on a 17.5-year burning cycle.

(B) Ranges based on measurements made in 2003 and 2004, as
described in text.

(C) Best estimated range based on data from the literature, see
Appendix 1.

Table 4. Measured and estimated carbon fluxes for
Mossdale Moor under a well-managed burning regime

                                                gC[O.sup.2]/
Parameter               gC/[m.sup.2] x year   [m.sup.2] x year

                     Losses

Burning (A)                    10-33               37-63
Soil gaseous effluxes         181-235             692-1098
Fluvial                        15-72               55-264
Total losses                  216-340             784-1485

                     Inputs

Rainfall                        4-4                15-15
Above/below-ground
  production                  168-482             616-1768
Total inputs                  172-486             631-1783
Net totals                  -34 to 146          -153 to 298

(A) Parameter directly measured and averaged for a 17.5-year
burning cycle. All other values have been given a range
reflecting the best estimates as available in the literature,
see Appendix 1.

(B) Losses minus inputs.
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.