The Scottish Government

« Previous | Contents | Next »

Listen

Module 6 Options for Mitigating C and N Loss from Organic Soils used for Agriculture

This module summarises the impacts of various aspects of agricultural management on organic soils, to aid in developing policies to minimise the loss of stored carbon and the emission of greenhouse gases. Where management strategies cause trade-offs between different greenhouse gases, Global Warming Potential ( GWP) is used to determine the best practice for minimising climatic effects. This is generally expressed as the amount of carbon as CO ₂ which would have the equivalent heating effect over 100 years, and therefore allows comparison of different GHG emissions in terms of their impact on climate change. Practices covered include the management of natural heathland by grazing and burning, and practices associated with using the soils for improved grassland or cropland, such as drainage, fertilisation, and tillage. Good practice guidance is provided for each management practice within the review and then summarised by land use in section 6.2. Conversion to forestry is covered in module 7 of this report.

6.1 Literature review of impacts of agricultural management on organic soils

6.1.1 Introduction

Reports have revealed that around 30 % of Scotland's greenhouse gas ( GHG) emissions come from the agriculture, forestry and land use section, of which roughly half originate from land use changes on high carbon soils (Milne et al., 2004). In Wales these figures are lower, with agriculture producing 12.1% of total GHG emissions, and land use change sinks and sources being almost in balance (Baggott et al., 2006). These GHG emissions are now recognised as affecting our climate, posing one of the most serious environmental threats to the planet. Therefore, issues surrounding the management of these soils is of immediate concern to both UK and devolved policy makers. Organic soils, when underlying natural vegetation such as heathland, tend to be sources of methane (CH ₄), sinks for carbon dioxide (CO ₂), and neutral as regards nitrous oxide (N ₂O). When managed for agricultural purposes however, activities such as drainage and the application of fertiliser, increase the rate of decomposition and produce significant emissions of both CO ₂ and N ₂O, but also reduce emissions of CH ₄, although not sufficiently to counteract the increase in other gases. Assessing the effect of changes in land management on organic soils is a major area of uncertainty in the estimation of GHG emissions.

Fig 6.1 Area of organic soils in 1 km grid cells from the DEFRA soil carbon database

Fig 6.2 Area of organo-mineral soils in 1 km grid-cells from the DEFRA Soil Carbon Database

Organic soils form when decomposition processes are slower than the input of organic matter from plants, allowing a build up of litter and partially decomposed matter. This occurs when conditions are unfavourable for the microbial populations which break down the litter, typically due to a combination of factors such as low pH, and anaerobic conditions due to water logging. These conditions are also unfavourable for agricultural usage, especially the growing of crops, but 'improving' the soil leads to more rapid turnover and tends to result in the loss of stored carbon. The organic soils considered in here include both truly organic soils, such deep peats, and organo-mineral soils, where a shallower (> 50 cm) but highly organic layer overlays a mineral soil, such as podzols and gleys. While true peats are relatively well understood and conserved rather than converted to forestry or other land uses, organo-mineral soils are widespread and are susceptible to land use change as well as to the impacts of climate change.

In Scotland, organic soils cover 18 000 km ² while organo-mineral are more common than mineral soils, covering 32 000 km ² (38 % of total land cover) (Bradley et al., 2005). Recent revision of estimates of carbon stocks stored within these soils has indicated that carbon in deep peat has been overestimated in the past, but in the 0-30 cm and 30-100 cm soil layers, carbon content has been underestimated by 320 Tg and 261 Tg respectively, representing 22 % and 20 % of the newly calculated total stocks at these depths (Bradley et al., 2005). In Wales, organic and organo-mineral soils are less extensive, covering 3% and 17.3 % of the land surface, and total revised stocks amount to 396 Tg (Bradley et al., 2005, Module 1). Appropriate management of these soils to retain soil carbon levels is becoming increasingly important, as research into carbon levels in soils in England and Wales shows that soils with carbon content greater than 100 g kg ^-1 lost carbon at an average rate of more than 2 % yr ^-1 between 1978 and 2003, and suggests a link to climate change among possible reasons (Bellamy et al., 2005). It is however, important to note that these represent total losses and it is unclear what proportion is lost to water as dissolved or particulate carbon, as opposed to gaseous losses to the atmosphere.

Large areas of organic and organo-mineral soils have been drained for agricultural usage over the years. In the UK, peat drainage was at it's height between the 1940s and 1970s but still continues today on a smaller scale (Holden et al., 2004), although the UKGHG inventory assumes no fenland (lowland) drainage has taken place since 1990 and that emissions due to drainage continue only in England (Milne et al., 2004). Land use change matrixes developed for this project (see Module 8) indicate that land use change has slowed since the 1980s but the most significant change on these soil types has been change to cropland and improved grassland on the East coast of Scotland, change to improved grassland throughout Wales, and change to cropland in South West Wales (see below). Current land use can also be a problem.

Fig. 6.3. Carbon density in 1 km grid cells, highlighting Scotland and Wales, from Bradley et al (2005)

Fig. 6.4 Land use change in the UK a) total LUC 1980s to present on organic and organo-mineral soils (rainbow colours) b) change to cropland 1990s and 2000s (purple) and c) change to pasture (improved grassland) 1990s and 2000s (green). All units ha/decade/20 km square. (see module 8 in this report)

For example, the South West of Scotland has been identified as a potential hotspot of N ₂O ( GHG) emissions due to high concentrations of cattle grazed, fertilised pastures combined with a mild wet climate (Flynn et al., 2005). The presence of large carbon stocks in the local soils mean any mitigation methods should be tailored to more organic soils. In Wales, concern has been raised about overgrazing degrading montane vegetation and reducing soil carbon stocks (Britton et al., 2005), while in Scotland, burning of heathland may be on the increase (W. Towers, pers. obs.) with potentially damaging consequences.

6.1.2 Agricultural Management Practices

Drainage: The rate of organic matter decomposition is dependent on soil water conditions (Goncalves and Carlyle, 1994; Leiros et al., 1999; Rodrigo et al., 1997) and therefore altering the natural water table position in an organic soil has knock-on effects on carbon and nitrogen cycling and associated emissions of greenhouse gases. While natural wet peatlands tend to be sources of CH ₄ emissions, drainage leads to increased loss of carbon and a greater contribution to total GHG emissions and climate change. Decomposition is faster above the water table and lower water tables have been shown to enhance CO ₂ production in a number of studies, including laboratory studies (Moore and Dalva, 1993), drained peatlands (Silvola et al., 1996), natural variability in water table position (Alm et al., 1997; Kim and Verma, 1992) and drier than average years (Bellisario et al., 1998; Carroll and Crill, 1997). In Germany, where a high proportion of peatlands have been drained for agricultural usage, peat mineralization has been reported to be highest where the water table is 0.8-1.2 m below the surface during the growing season (Hoper, 2002).

The position of the water table controls how much of the soil profile is anaerobic, as oxygen is quickly depleted in the lower waterlogged portion. Anaerobic conditions are perhaps the most important factor in reducing the rate of decomposition to allow the formation and maintenance of highly organic soil layers. This reduced rate of decomposition is illustrated by measurements of carbon dioxide emissions from peatland soils, which have been reported to be around 2.5 times lower under anaerobic than aerobic conditions in short-term laboratory incubations (Bridgham and Richardson, 1992; Moore and Dalva, 1997), and much lower, an average of 14 times less, in long-term incubations, suggesting anaerobic CO2 production is inhibited by the accumulation of repressing agents (Magnusson, 1993).

The effect of a drop in water table depth may however, be dependent on the usual conditions of the system. Laiho (2006) notes that the increase in CO2 emission generally tails off when the water table drops below a certain depth (Chimner and Cooper, 2003; Silvola et al., 1996) and argues that deeper layers may lack easily oxidizable labile C (Chimner and Cooper, 2003; Hogg et al., 1992). Peatlands where the water table is generally 20 cm or more below the surface over the summer will already have been exposed to aerobic decomposition for extended periods, leaving only more resistant organic matter (Bridgham and Richardson, 1992) and therefore further drawdown due to drainage may not lead to increased C turnover. In contrast, peatlands such as wet fens which are generally continuously inundated, may exhibit strong responses to a drop in water table as there may be a large pool of labile C which has not been decomposed due to the waterlogged, anaerobic condition.

This clearly demonstrates that drainage, particularly of soils which are generally inundated, increases CO ₂ emissions and leads to a loss of stored carbon. However, oxidizing decomposition processes are not the only processes involved in nutrient cycling and GHG emission production in organic soils. Methane (CH ₄) is produced by anaerobic fermentation processes below the water table and is either emitted from the soil profile, or oxidized into CO ₂ by aerobic methanotrophic bacteria above the water table. Therefore, when the water table is below the soil surface, oxidation becomes a major controlling variable for methane efflux (Christensen et al., 2000) and a lower water table decreases methane emission (Blodau et al., 2004). This decrease in emissions with increasing water table depth has been reported using soil cores incubated in the laboratory by a number of workers (Daulat and Clymo, 1998; MacDonald et al., 1998; Moore and Dalva, 1993) but has also been confirmed using field measurements over a peat wetland in Caithness (Hargreaves and Fowler, 1998). Wetlands are not always a significant source of CH ₄ however; where the water table is above the surface, oxidation can also reduce methane emissions, particularly if light availability allows benthic photosynthetic activity (Le Mer & Roger, 2001). The highest emissions of CH ₄ in temperate systems are associated with wetlands with significant reed or sedge cover as these plants have stems which can transport the gas to the surface and reduce the likelihood of it being oxidised before it reaches the atmosphere. Indeed, sedge cover has been identified as the single most important single variable affecting CH ₄ flux in several studies (Bellisario et al., 1999; Nilsson et al., 2001; Schimel, 1995; Tuittila et al., 2000; van den Pol-van Dasselaar et al., 1999).

Although CH ₄ is a potent greenhouse gas, it makes up a small proportion of total C cycling. CH ₄ flux generally accounts for between 1 and 4.7 % of carbon respired, and <2 to 7 % of net primary production in northern wetlands (Christensen et al.,1996). Other workers have reported that around 2 % of total C cycling in a high arctic fen (Friborg et al., 2000), and 4 % of net annual CO ₂ assimilation and 9 % of net carbon fixation from a reed-dominated wetland in Denmark (Brix et al., 2001), is returned to the atmosphere as CH ₄. Whiting & Chanton (1997) report that for wetlands in general, the ratio of methane released to annual net carbon fixed varies between 0.05 and 0.13 on a molar basis. While draining peats would reduce CH ₄ emissions and may result in producing net sinks for CH ₄ (Blodau & Moore, 2003; Huttunen et al., 2003; Maljanen et al., 2002), these are generally very small, particularly for upland soils (Le Mer and Roger, 2001), and it would not reduce total carbon emissions. Measurements from sites across Europe suggest that ombrotrophic bogs drained for forestry or peat cutting have an average Global Warming Potential ( GWP) of 1253 CO ₂-C equivalents kg ha ^-1 yr ^-1 in comparison with an average GWP of 192 CO ₂-C equivalents kg ha ^-1 yr ^-1 (entirely due to CH ₄ emissions) for undrained sites with natural vegetation (Byrne et al., 2004).

Nitrogen dynamics are also affected by soil water conditions. The greenhouse gas nitrous oxide (N ₂O) is formed through both nitrification and denitrification, and high N ₂O emissions often only occur when soil water content is high or immediately following rainfall (Aulakh et al., 1984; Skiba et al., 1992; Hansen et al., 1993: Smith et al., 1998; Dobbie et al., 1999; Ruser et al., 2001). These peak emissions are generally associated with denitrification which, as an anaerobic process, becomes dominant at higher water contents, generally reported to be above 60-65 % WFPS (water-filled pore space) (Clayton et al., 1997; Lin and Doran, 1984). In a peat soil, total N ₂O emissions have been reported to be nearly constant at 40-80 % WFPS but significantly higher at 100 % (Pihlatie et al., 2004). Peat soils have also been shown to emit more N ₂O in drier conditions than sandy loam soils, but less when wet; in dry soils (40% and 60 % WFPS), N ₂O production was an average of 2.0 % (peat), 0.3 % (loamy sand) and 0.5 % (clay) of that produced in wet soils (80-100% WFPS), where production was highest in the loamy sand and lowest in the clay (Pihlatie et al., 2004). However, N ₂O production is a complex balance of processes with different drivers and when nutrient levels are not limiting, drainage does tend to increase emissions in general, while nutrient-poor organic soils tend to be very small sources of N ₂O even when drained (Byrne et al., 2004; Martikainen et al., 1993).

This is a pertinent issue, particularly in the South West of Scotland where managed grassland is a common land use and organic soils are more prevalent than in the arable areas of the East coast. This area has already been identified as a hotspot for N ₂O emissions which may increase further as the climate is predicted to become wetter and milder, and drainage has been recommended to mitigate this (Smith et al., 2004). For organic soils, since drainage will reduce C storage and increase CO ₂ emissions, N ₂O emissions could be reduced by reducing N inputs instead, since production processes rely on available ammonium or nitrate (see fertilisation section below).

As well as affecting soil biochemical processes, drainage may also have a physical impact on C stocks by increasing vulnerability to erosion. Subsidence rates of 2-4 cm per year have been reported for drained fens in England (Bradley, 2000; French and Pryor, 1993; Hutchinson, 1980). Increases in decomposition may also lead to increased losses of dissolved organic matter ( DOM) (Tipping et al., 1999) which can lead to problems when it colours drainage water used for water supply. Particulate losses of carbon from the development of drainage pipes is also a significant problem, Holden (2006) calculates that the loss from a peat slope drained for 40 years would be sufficient to approximately halve the size of the net carbon sink of a healthy blanket peat catchment. Modelling studies have suggested that although restoring a high water table increases CH4 emissions, it may reduce total GHG emissions from peat soils (van Huissteden et al., 2006). This is supported by measurements from European sites suggesting that restored bogs have a GWP of 517 CO2-C equivalents kg ha-1 yr-1 less than those drained for forestry or cutting, and restored fens have a GWP of 368 CO2-C equivalents kg ha-1 yr-1 less than those drained for forestry (Byrne et al., 2004). Blocking old drains may also be worthwhile as research suggests that subsurface piping increases over time causing particulate carbon loss from drained peat slopes to increase exponentially (Holden, 2006).

Grazing: Grazing is probably the most common land use on organic soils in the UK and its effects are dependent on livestock type and stocking density. Direct impacts on soil are caused by a combination of trampling and nutrient addition via deposition of dung and urine. Trampling effects tend to be concentrated around fence lines and feeding stations, are more pronounced with larger, heavier animals, and greater stocking densities, and are also exacerbated by wet conditions (Shaw et al., 1996). Trampling has been shown to stimulate denitrification due to reduced soil aeration and plant N utilisation (Menneer et al., 2005) and may therefore increase N ₂O emissions, especially under wet conditions. Under dry conditions, trampling disturbs the soil and may act like tillage to increase aeration and stimulate decomposition and associated CO ₂ emissions. Both of these effects may be amplified by the addition of nutrients. It is difficult to quantify soil emissions resulting from grazing however, as field measurements are usually taken using chambers and grazing animals are often excluded from field sites to prevent damage to equipment, and also the effects of nutrient deposition are highly localised. Data collected for this study indicates that N ₂O emissions from organo-mineral soils are highest where grazing pressure is highest. Effects on CO ₂ emissions are even harder to pin down as net ecosystem exchange ( NEE) is usually measured and this includes plant and rhizosphere exchange from respiration and photosynthesis as well as soil respiration from decomposition processes (see for example, Nieveen et al., 2005).

Overgrazing can cause severe degradation, not only damaging vegetation but also reducing soil organic matter and removing upper organic layers. In Wales, there was a 71 % increase in sheep numbers between 1974 and 1998 (Countryside Council for Wales and Forestry Commission, 1999) and an upland area of North Wales has seen an increase in grazing density from around 1.2 sheep ha ^-1 in the 1950s to an average of around 5-6 sheep ha ^-1. This has had a detrimental effect on the ranker and peaty podzol soils, with degraded areas containing significantly less carbon and nitrogen, means of 5 % C and 0.4 % N in comparison with 24-27 % C and 1.1-1.4 % N in intact heathland ecosystems at the same site. Acidity has also significantly reduced (Britton et al., 2005). These effects can be considered typical of sustained heavy grazing pressure (Milne et al., 1998; Rudeforth et al., 1984). In recent years, numbers of sheep in both Scotland and Wales have declined. In Scotland however, red deer numbers have increased sharply, although estimates are subject to debate, and it may become increasingly important to control grazing pressure by taking account of these wild grazers as well as sheep numbers.

Properly managed for site conditions however, grazing can be beneficial. There is some evidence that carefully managed grazing can aid carbon storage in organic soils. Studies of the Moorhouse NNR in the Pennines have indicated that grazed plots tend to accumulate carbon slightly faster than ungrazed areas and certainly light grazing did not cause any reduction in C accumulation in comparison with no grazing (Garnett et al., 2000). Grazing effects can also be minimised by not allowing animals access all year round, for example by moving them to pastures on mineral soils in winter when wetter conditions will exacerbate trampling effects. This has been shown to help maintain heathland vegetation (Grant et al., 1982). It should also be recognised that most heathlands can support higher stocking densities during summer than in winter. Finally, different types of grazing animals may have different preferences for different types and ages of vegetation and can therefore be selected on the basis of the type of plant community land managers wish to promote.

This is discussed in some detail in a report produced for English Nature, which describes dietary preferences and reported effects of cattle, sheep, ponies, goats, deer and other grazers, such as grouse and mountain hares (Shaw et al., 1996). Cattle may have a role to play as they tend to graze less selectively than other grazers, however, this needs to be balanced against the increased risk of trampling from these heavier animals, and they also deposit more nutrients through dung and urine than smaller ruminants.

Burning: Burning is a traditional management technique for maintaining open shrub (usually heather) or grass dominated vegetation in upland areas and preventing succession to scrub and woodland. It is however, a source of CO ₂ emissions to the atmosphere. An investigation into wildfire on peatlands in North Carolina using remote sensing reported that carbon emissions ranged from 0.2-11 kg C m ^-2 and total emissions were between 1 and 3.8 Tg (Poulter et al., 2006). Burning also has long term effects on soil properties; investigations into a drained peatland in Turkey have reported significantly reduced soil organic carbon, even in areas which were burnt as long ago as 1965 (Dikici and Yilmaz, 2005). Organic matter accumulation at the Moorhouse NNR in the Pennines has been reported to be lower for plots which have been burnt every 10 years than for those which have not been burnt since 1954 (Adamson, 2003; Garnett et al., 2000), and data from a Finnish mire also suggests that frequent fires reduce carbon accumulation (Pietikäinen et al., 1999).

Organic matter may be physically broken down by burning but evidence also suggests that further mineralization may occur after the fire due to increased microbial activity; microbial respiration has been reported to be three times higher following burning, in response to higher nutrient and substrate levels in remnant soils and enhanced soil temperature (Kim and Tanaka, 2003). There is also some evidence that burning increases the pH of organic soils, which would also favour increased rates of decomposition (Allen, 1964; Stevenson et al., 1996). Measurements taken in the Everglades to compare soil properties pre- and post-peat fire also confirm that total C and N contents are significantly reduced and not just by physical processes (Smith et al., 2001).

As burning removes vegetation, it also increases vulnerability to physical erosion, especially as charred peat can form loose crusts which are easily washed away, and increases losses of DOC (Tucker, 2003). This erosion is arguably the most important impact of burning on peatlands, and is most severe when all the vegetation cover and root matt is removed, leaving bare soil exposed to sun, wind, water, and in winter, more intense freeze-thaw cycles. A study of post-fire erosion of podzols and peaty gleys in the North Yorkshire Moors concluded that vegetation cover was the main determinant of erosion rates and reported relatively slow rates where the burnt remains of heather made up the ground surface. More severe burning which exposed the peaty or mineral subsoils however, caused much more rapid erosion, and in winter (September to April) erosion rates on bare soil were up to 10 times higher (Imeson, 1971). This is particularly significant as current UK law only allows for moorland burning between October and April to reduce the likelihood of dry peat being consumed along with the heather stand. Vegetation recovery limits erosion (Kinako and Gimingham, 1980) and it has been suggested that heather recovers better from autumn rather than spring burns (Miller and Miles, 1970; Phillips, 1991), although this may not be the case if regeneration is dependent on seedling establishment (Tucker, 2003). Autumn burning is generally considered to be less intense due to greater moisture content in vegetation and litter (Hobbs and Gimingham, 1987) but burning after a wet period, post thawing should limit fire severity and avoid the worst effects of erosion. The best way to avoid detrimental effects from burning however, is not to use it at all, as heather cover in many cool, moist upland European heathlands can probably be maintained without burning (MacDonald et al., 1995) provided grazing is used to prevent succession to woodland (Marrs, 1988).

Fertilisation: Fertilisation is generally considered to aid carbon storage in mineral soils (Gregorich et al., 2005). In organic soils however, this assumption may not hold true. Tundra organic soils which had been fertilised annually for 8 years showed different effects depending on ecosystem type, with moist acidic soils from between sedge tussocks containing significantly less carbon and nitrogen in comparison with unfertilised plots, soils from beneath the tussocks containing significantly more carbon but no significant effect on nitrogen, and the drier heath soils with a shallower organic layer showing no significant effect of fertiliser on carbon or nitrogen stocks (Shaver et al., 2006). In general however, fertiliser would be applied to drier organic soils rather than waterlogged bogs, and added nutrients combined with aerobic conditions accelerate decomposition and increase CO ₂ emissions (Byrne et al., 2004). This effect may be particularly acute when lime is also applied, making conditions more favourable for decomposition as well as supplying extra nutrients.

Emissions of N ₂O have a more straight forward relationship with fertiliser inputs. Like mineral soils, farmed organic soils have been shown to emit high levels of N ₂O temporarily after fertilisation (Augustin et al., 1998), especially in wet conditions and early in the growing season when vegetation competes less for the available N (Kettunen et al., 2005; Nykänen et al., 1995). However, emissions cannot be calculated from fertiliser N input in the same way as for mineral soils, as existing stocks of N in the soil mean that emissions are not directly related to input levels (Mosier et al., 1998). Different types of fertiliser may also have different levels of impact; the application of solid manure has been reported to produce substantially lower N ₂O emissions than mineral fertilisers or liquid manure (Gregorich et al., 2005), while sewage sludge pellets and poultry manure resulted in emissions 12-26 times larger than mineral NPK when applied to a Scottish grassland (Jones et al., 2005).

Aerobic farmed organic soils have been reported to be negligible sources or sinks of CH ₄ with low production and oxidation potentials (Kasimir-Klemedtsson et al., 1997; Kettunen et al., 2005) and are therefore largely neutral in terms of the CH ₄ budget. There is some evidence however, that fertilisation may reduce any small oxidation potential they may have, as Crill et al (1994) report that inorganic NH ₄⁺ can inhibit CH ₄ oxidation in drained peat soil.

Liming: Many organic soils are naturally acidic and this is generally considered to limit the activity of decomposers, which favour a neutral environment, aiding the build up of organic matter. A significant effect of soil pH on decomposition has been observed in many studies (eg. Andersson and Nilsson, 2001; Hall et al., 1997; Situala et al., 1995) and there is a positive correlation between soil biological activity levels and soil pH (Motavalli et al.,1995). Acidic irrigation studies have shown that reduced pH leads to reduced CO ₂ fluxes; Sitaula et al. (1995) reported that pH 3 produced CO ₂ fluxes 20 % lower than those at pH 4 and 5.5, and Persson and Wiren (1989) reported increasing the acidity of forest soil from pH 3.8 to 3.4 reduced CO ₂ production by 83 % and from pH 4.8 to 4 by 78 %. This clearly suggests that increasing pH towards neutral by the addition of lime to a naturally acidic soil will lead to an increase in CO ₂ emissions and a reduction in C stocks. This is supported by a study of carbon flow in an upland grassland, which showed that liming caused more rapid C turnover (Rangel-Castro et al., 2004).

Soil pH may also have a varying impact depending on aeration and water logging. Production rates of CO ₂ have been shown to increase more with increasing pH under anaerobic conditions; Bergman et al (1999) compared CO ₂ production rates at pH 4.3 and 6.2, and found that under anaerobic conditions rates were 21- 29 times greater at the more neutral pH (depending on temperature), while under aerobic conditions rates were 3 times greater at 7 ^oC but pH had no significant effect at 17 ^oC. This suggests that liming will have a greater impact on wet organic soils.

Just as neutralising low pH tends to increase CO ₂ production, as a general rule it also increases rates of organic N mineralization, ammonification, nitrification and denitrification. Cumulative N ₂O emissions have been shown to be positively correlated to pH between 4.14 and 7.53 in arable soils incubated under flooded conditions (Wlodarczyk et al., 2002) and experiments carried out at the Rothamsted Experimental station have indicated that lime treated grassland plots have significantly higher N ₂O emissions; Yamulki et al (1997) reported that emissions from a pH 3.9 plot were around 36 % of those from a pH 7.6 plot. In the short term, there is evidence that neutralising a naturally acidic soil may temporarily reduce N ₂O emissions as the indigenous microbial community is adapted to the acidic conditions. For example, Yamulki et al (1997) found that treating a pH 3.9 soil to increase the pH to 7 reduced N ₂O emission rates by more than 40 % and Brumme and Beese (1992) found that liming from pH 4.5 to 6.5, reduced N ₂O emissions by 74%. However, in the longer term, the microbial population will adapt to the new conditions and organic agricultural soils generally emit more N ₂O than mineral agricultural soils (Duxbury et al., 1982; Martikainen et al., 2002).

More neutral conditions may also favour CH ₄ production as methanogenic bacteria are generally reported to exhibit maximum activity under neutral or slightly higher pH conditions (Garcia et al., 2000). An investigation of CH ₄ production in peat soil samples from temperate and subarctic areas (pH 3.5-6.3) reported an optimum pH of 5.5 to 7.0 (Dunfield et al., 1993), and a positive correlation between CH ₄ production activity and soil pH has been reported for peat soil samples from a temperature wetland with a pH range of 5-7 (Inubushi et al., 2005). Experiments which increase the pH of peat samples towards neutral have also shown an increase in CH ₄ production;

Williams and Crawford (1984) found that a pH increase from 3.2 to 5.8 increased the methane production of an incubated peat from a Minnesota peatland by 1.5-2.2 fold, depending on depth, and Murakami et al (2005) also report that liming acid peats increases CH ₄ production potential. However, this increase in CH ₄ production activity may not necessarily lead to increased CH ₄ emissions, as more neutral conditions may also favour CH ₄ oxidising bacteria (methanotrophs), which have a reported optimum pH of 5.0 to 6.5 in temperate and subarctic peats (Dunfield et al., 1993). Hutsch et al (1994) reported that, in a non-fertilised permanent grassland at the Rothamsted experimental station, a decrease in pH from 6.3 to 5.6 reduced methanotrophy by almost half. In practice, limed soils are generally drained agricultural soils and therefore either slight consumers or low sources of CH ₄ due to their aerobic condition.

A final issue is the impact of liming on the production of DOC (dissolved organic carbon). An increase in the export of DOC in surface waters has been observed across much of Europe and North America over the past twenty years and the causes of this are still a subject of debate, but it has been suggested that declining acid deposition may be a major factor (Evans et al., 2006) because DOC solubility is suppressed by high soil water acidity. Experiments have shown that liming can increase the concentrations of organic matter, DOC and DON in soil water (Andersson et al., 1994, 1999; Curtin and Smillie, 1983). Although loss of organic matter in surface water is an issue nationwide, the management of upland organic soils may play a disproportionately large role in determining the future scale of this C loss. This is because the impact of a change in pH on DOC solubility is greatest in the pH 4 to 5 range (Thurman 1985; Peterson 1990), which is typical of upland organic soils in Britain, and titration experiments have indicated that an increase in soil water pH of 0.5 units could cause a 50% increase in DOC (Tipping and Woof 1990). This has important implications for Wales and Scotland where liming to mitigate historic acidification is underway and may have serious knock-on effects on water quality.

Tillage: The impacts of tillage on soil organic matter ( SOM) vary depending on cropping system, residue management, and climate, but in general, it promotes decomposition through crop residue incorporation into soil, physical breakdown of residues, and disruption of aggregates protecting SOM (Baldock and Skjemstad, 2000; Paustian et al., 2000; Six et al., 2000). Microbial metabolic activity has been shown to increase in response to residue incorporation and improved aeration; Balota et al (2004) reported that, averaged across crop rotations and depth, conventional ploughing increased metabolic quotient (soil respiration per unit of microbial C) by 32 % compared with no ploughing. Impacts are however, generally confined to the surface layers, which are physically disturbed. Several workers have reported that tillage reduces soil organic carbon and nitrogen in the surface layers. Salinas-Garcia et al (2002) found that SOC in the top 5 cm was between 36 and 57 % lower under conventional tillage (disk ploughing to 30 cm) compared to conservation tillage (no ploughing or disking to 10 cm), which retained crop residues at the surface, reducing contact with the decomposer community, and Wright et al (2005) also reported that tillage reduced SOC and SON at the surface, although the effect was not significant under all the cropping regimes.

Impacts are also dependent on the type of ploughing carried out. Moldboard ploughing is generally reported to have the most significant impacts, followed by conventional or disk ploughing, and with conservation ploughing or disking being most similar to no tillage regimes (Wright et al., 2005). Deeper disturbance generally corresponds to greater overall impact, although shallower mixing may result in greater impacts at the surface: Friedal et al (1996) compared the effects of ploughing to a depth of 25 cm with rotary cultivation to a depth of 10-12 cm, and reported that in the upper soil layer (0-10 cm) rotary cultivation caused a greater increase in C and N mineralization potential than ploughing.

Stimulating microbial activity also leads to an increase in N ₂O emissions. Although it generally improves aeration, Jackson et al (2003) found that tillage significantly increased denitrification from day 2 to a week after tillage, and cumulative N2O fluxes have been reported to increase by 5 to almost 12 times, depending on crop type, in response to ploughing to 15-20 cm depth (van der Weerden et al., 1999). Again, shallow tillage practices such as rotovation may lead to lower N ₂O emissions than traditional ploughing, but both stimulate emissions to some extent.

Soils which are under agricultural management and suitable for ploughing are generally net consumers of CH ₄ due to their aerobic surface layers, and therefore tillage does not affect CH ₄ emission levels. Agricultural practices may affect their ability to oxidise atmospheric CH ₄ (Le Mer and Roger, 2001) but the size of this flux is insignificant in terms of total GHG emissions.

As previously discussed (see burning section above), organic soils are more prone to erosion when bare in winter, and freeze-thaw cycles also increase N ₂O emissions (Kaiser et al., 1998; van Bochove et al., 2000), and therefore timing of ploughing is also an important issue, as leaving stubble over winter will partially mitigate these effects.

Conversion to grassland: Conversion to grassland, comprising drainage to a depth of around 0.4-0.8 m below the surface (Joosten et al., 2002), removing natural vegetation and replacing it with productive grass species, and moderate to high levels of fertiliser addition, has been assessed as causing fast loss of stored carbon (Byrne et al., 2004). Conversion of a peat bog to pasture for dairy farming has been reported to reduce carbon storage by an average of 3.7 t ha ^-1 yr ^-1 over 40 years (Schipper and McLeod, 2002). Studies of boreal organic soils used for grassland in Finland have reported annual net carbon losses of 750 g CO ₂-C m ^-2 (Maljanen et al., 2001), and 3.3-4.6 t CO ₂-C ha ^-1 yr ^-1 (Maljanen, 2003), while in the Netherlands, carbon loss of 300 g CO ₂-C m ^-2 has been reported for pasture (Langeveld et al., 1997).

Based on flux measurements from European sites, GHG emissions from grassland on nutrient-poor, ombrotrophic bog soils are estimated to have a GWP of 2367 CO ₂-C equivalents kg ^-1 ha ^-1 yr ^-1, based on CO ₂ emissions of 1.5-3.5 t C ha ^-1 yr ^-1, CH ₄ emissions of 2 kg C ha ^-1 yr ^-1 and N ₂O emissions of 0-0.2 kg N ha ^-1 yr ^-1. For more nutrient-rich, minerotrophic fen soils, the values are GWP of 4794 CO ₂-C equivalents kg ha ^-1 yr ^-1, consisting of CO ₂ emissions of 0.82-6.58 t C ha ^-1 yr ^-1, CH ₄ emissions of

-1.04-105 kg C ha ^-1 yr ^-1, and N ₂O emissions of 5.05 kg N ha ^-1 yr ^-1 (Byrne et al., 2004). Boreal peat soils drained and used for pasture may also have higher nitrate availability than those drained for afforestation, and have been reported to emit 6-10 times more N ₂O at a rate of 8-9 kg N ha ^-1 yr ^-1 (Nykänen et al., 1995), while a range of 1.7-11 kg N ₂O-N ha ^-1 yr ^-1 has been reported for organic soils under grass across Scandinavia (Maljanen, 2003).

Conversion to cropland: Conversion to arable land use is the worst case scenario for organic soils, with the combined effects of drainage, tillage and fertiliser input causing very fast loss of stored carbon (Byrne et al., 2004). Not only does this land use generally involve the greater disturbance of the soil, it also requires the most drainage, as the water table needs to be lowered to around 1.0-1.2 m below the surface (Joosten et al., 2002). Carbon losses of 400-830 g CO ₂-C m ^-2 yr ^-1 have been reported for boreal organic soils used to grow barley in Finland (Maljanen, 2003). These losses are also exaggerated by the presence of bare soil between crops, which may emit more CO ₂ than under crops, for example up to 11 t ha ^-1 yr ^-1 for Finnish organic agricultural soils (Maljanen, 2003), and may also be subject to physical erosion. Therefore, emissions can be partially mitigated by changing to more permanent crops, and also by avoiding root crops such as potatoes and sugar beet, as these involve frequent and intensive soil disturbance (Freibauer et al., 2004).

As well as losing stored carbon rapidly, organic soils used for agriculture can be a significant source of N ₂O. In Finland, such soils may account for as much as 25 % of national N ₂O emissions, despite making up less than 10 % of agricultural soils (Maljanen, 2003), and in general, organic soils emit more N ₂O than mineral agricultural soils (Duxbury et al., 1982; Martikainen et al., 2002). A range of 5.4-24.1 kg N ₂O-N ha ^-1 yr ^-1 has been given for Finnish organic soils used to grow barley (Maljanen, 2003).

Flux measurements from sites across Europe indicate that an average of 4400 CO ₂-C kg ha ^-1 yr ^-1 is emitted from nutrient-poor bog soils converted to arable usage, with no net flux of either CH ₄ or N ₂O, while nutrient-rich fen soils have an average GWP of 5634 kg CO ₂-C equivalents ha ^-1 yr ^-1, of which 1.09-10.6 t C ha ^-1 yr ^-1 is emitted as CO ₂, 4.0-56.4 kg N ha ^-1 yr ^-1 is N ₂O and there is a net sink of 0.2 kg C ha ^-1 yr ^-1 for CH ₄ (Byrne et al., 2004).

6.2 Good practice guidance for the agricultural management of organic soils

6.2.1 Summary

Greater disturbance of natural conditions equates to greater impact in terms of both loss of C stocks and increased greenhouse gas emissions. Therefore, conversion to arable cropland is the worst-case scenario for organic and organo-mineral soils, as this involves drainage, tillage, fertilisation, liming, and bare soil between crops which increases erosion risk. The best-case scenario is restoring natural vegetation and high water table conditions, and managing with light grazing.

6.2.2 Good practice guidance for arable land use on organic soils

• Organic soils converted to arable land use lose stored carbon very rapidly, typically at rates of around 4-8 C t ha ^-1 yr ^-1, emit more N ₂O than mineral agricultural soils, and have average Global Warming Potentials of 4.4-5.6 t CO2-C equivalents ha ^-1 yr ^-1, making this the worst-case scenario for land use change.
• Any further conversion to arable usage on these soils should therefore be strongly discouraged.
• Ideally, existing arable land should be restored to heathland, although conversion to managed permanent grassland is less damaging than continued usage for arable crops, provided any grazing animals are stocked at low density.
• Where arable land use continues, the following measures will partially mitigate carbon losses and greenhouse gas emissions:
  • Fertiliser and lime additions should be kept to a minimum.
  • Fertiliser should be applied when crops are in growth phases and can make best use of the applied nutrients.
  • Fertiliser and lime should not be applied in wet conditions.
  • Application of solid farmyard manure is preferable to sewage sludge or liquid manure/slurry, as it has been shown to produce lower N ₂O emisisons.
  • Conservation or zero-till regimes should be encouraged.
  • Deep ploughing should be discouraged.
  • Winter ploughing should be avoided to reduce erosion risk and effects of freeze-thaw cycles on bare soil, and instead where necessary, ploughing should be carried out as close to new crop sowing as possible to minimise soil exposure.
  • Avoid root crops such as potatoes and sugar beet which require more soil disturbance.
  • Minimise time in which soil is without crop coverage and change to permanent crops if possible.
  • Maintain as shallow a water table as possible.

6.2.3 Good practice guidance for improved grassland on organic soils

• Organic soils converted to managed grassland lose carbon rapidly, typically at rates of around 3-5 C t ha ^-1 yr ^-1, and have average Global Warming Potentials of 2.4-4.8 t CO2-C equivalents ha ^-1 yr ^-1.
• Therefore, further conversion of heathland by draining, fertilising and liming should be strongly discouraged.
• Ideally managed grassland should be restored to a more natural state by blocking drains and stopping applications of fertiliser and lime. Restoring high water tables could reduce their Global Warming Potential to 0.4-0.5 t CO2-C equivalents ha ^-1 yr ^-1.
• Resources for blocking drains should be focused on slopes where the drainage is oldest as this will have the greatest impact for reducing carbon losses through erosion.
• Blocking drains may also help reduce problems with colour in water due to particulate organic matter in catchments where drainage water goes into the water supply, as will stopping liming in these areas.
• Drainage should not be used to mitigate N ₂O emissions on non-mineral soils, instead options such as reducing N inputs and grazing intensity should be explored.
• Where grassland management continues, the following measures will reduce its impact:
  • Maintain as shallow a water table as possible
  • Minimise fertiliser and lime inputs, particularly on wetter sites
  • Permanent grassland is preferable to reduce soil disturbance
  • Where grassland is grazed, low stocking densities will reduce impacts, as will using sheep or other smaller animals, rather than cattle, and reducing numbers in winter and on wetter sites.
  • Nutrients deposited on pastures should be taken into account to reduce fertiliser requirements
  • Fertiliser inputs should be timed so that vegetation is in growth phase and can make maximum use of the nutrients, and to avoid wet conditions.
  • Application of solid manure is preferable to sewage sludge or liquid manure/slurry.

6.2.4 Good practice guidance for rough grazing on organic soils

• Organic soils store most carbon and emit less greenhouse gas when underlying natural vegetation and managed with minimal disturbance and light grazing to prevent succession to scrubland and ultimately woodland.
• Drainage, especially of water logged areas, should be prevented.
• No fertiliser or lime should be applied.
• Overgrazing causes significant damage to soil organic layers, reduces carbon storage and increases greenhouse gas emissions so stocking densities need to be carefully managed to minimise this, while limiting succession to protect vegetation communities.
• Heavier animals such as cattle should only be used in very limited numbers, where their less selective feeding will aid vegetation management, and not on wetter sites where they are more likely to cause significant damage.
• Stocking densities should be reduced in winter or animals removed completely, particularly on wetter sites.
• Stocking densities must take into account grazing pressure from wild animals such as deer and a reduced number of domestic livestock should be used on sites where deer are numerous.
• Deer control measures should be concentrated in areas with organic soils which are at risk of overgrazing.
• Serious consideration should be given to halting the practice of burning on soils with significant stores of carbon, and the use of grazing alone to control community succession and encourage vegetation renewal, although this may require reductions in grazing density.
• Burning should not be allowed in areas with a high risk of erosion such as exposed areas or areas with extensive drainage gullies or ditches.
• Fires should be as small and controlled as possible so that only vegetation is consumed and litter and root mats remain to protect the soil from exposure
• The burning season should be confined to after wet periods in early spring, towards the end of the current legal season, as winter freeze-thaw cycles increase the risk of erosion when soil is no longer protected by vegetation.

« Previous | Contents | Next »