The Scottish Government

« Previous | Contents | Next »

Listen

Module 1 The Distribution of Organic Soils in Scotland and Wales and the C Contained in Them

1.1 Types of organic and organo-mineral soils in Scotland and Wales

1.1.1 Scotland

Introduction: In order to allow definitive estimations of the carbon stocks of organic and organo-mineral soils in Scotland to be made, identification of the relevant soils on which to focus attention was required. In addition, modal soil profiles with carbon and bulk density attribute data were needed to estimate carbon stocks. The soil series, i.e. soils with a similar type and arrangement of horizons developed on similar parent material, is the basic soil mapping unit in Scotland and these occur as components of each of the 1:250 000 scale soil map units. Each soil series belongs to a major soil subgroup, which is the reference group adopted in this work.

Methodology: Using expert knowledge and an understanding of the soil classification adopted in Scotland, the major soil subgroups known to occur in Scotland were categorized into:

1) Peats
2) Organo-mineral soils
3) Mineral soils

1) Peats contain more than 60% of organic matter and exceed 50 centimetres in thickness. Peat is an accumulation of partly decomposed plant remains formed under waterlogged conditions where excess moisture has inhibited the normal aerobic process of decomposition. Other conditions conducive to the formation of peat are low average temperature, high acidity and nutrient deficiency, all of which depress microbiological activity. The most important factors that influence peat development are therefore functions of climate, topography and geology. These also control the nature and abundance of vegetation which in turn largely determines the type of peat formed. This grouping includes hill, blanket, and basin peats, all of which are extensive throughout Scotland. In a number of locations, particularly at higher altitudes, the peat is often severely eroded.

2) Organo-mineral soils are classified as soils having organic surface horizons, less than 50 cm thick in Scotland (<40cm in Wales), overlying mineral horizons or rock. This grouping includes humus-iron podzols, peaty podzols, subalpine podzols, alpine podzols, peaty gleys, humic gleys, peaty rankers, podzolic rankers, peaty lithosols and peat alluvium. In most of these soils, the organic surface horizon is a consequence of anaerobic and waterlogged conditions at the surface, due either to slowly permeable subsoil or rock, high ground-water tables in depressions and receiving sites, or to climatic conditions where a combination of low temperatures and high rainfall initiate the formation of the organic surface horizons. The main exceptions to this are humus-iron podzols and subalpine and alpine podzols where it is considered that the shallow organic surface horizon has developed under aerobic conditions. Many humus-iron podzols are cultivated and no longer have an organic surface horizon, having been mixed and incorporated with the underlying mineral horizons through years of cultivation. 3) Mineral soils have no organic horizons in the soil profile and were excluded from this investigation.

Results: Using this general classification, the relevant major soil subgroups occurring in Scotland have been identified and are shown in Table 1.1.

Table 1.1 Major soil subgroups in Scotland relevant to the study

1.1.2 Wales

It is important to define what is meant by peat and organo-mineral soils in the context of this review for Wales. In early soil surveys of Wales, in order to be classified as a peat soil, the layer of accumulation had to contain at least 50% organic matter and extend to a depth of over 12 inches (30 cm) (Roberts 1958) over 15 inches (37.5 cm) (Rudeforth 1970) or 20 inches (50 cm) (Ball 1960). Organic or peat soils, the names appeared inter-changeable, were then further classified according to their mode of formation into Basin Peat and Moss Peat.

With the introduction of a unified classification of soils for England and Wales (Avery 1980), horizons described as peat contain a minimum of between 20 and 50% organic matter, depending on the clay content. The profile classification has a Major Group of Peat Soils in which the thickness of accumulation of organic material exceeds 40 cm depth within the upper 80 cm of the profile. These criteria have been applied in all surveys published since 1978, including the National Soil Map for Wales (Rudeforth et al. 1984).

Organo-mineral soils are classified as soils having organic surface horizons less than 40 cm thick (in Wales), overlying mineral horizons or rock. The extent and organic matter content of the organo-mineral soils are well documented in the various soil maps, bulletins and memoirs of the Soil Survey of England and Wales, for example, Rudeforth et al. 1984, Rudeforth 1970 and Rudeforth and Thomasson 1970. The desk study and fieldwork in Wales carried out as part of ECOSSE concentrated on the extent and depth of peat soils which were less well known or documented.

1.2 The Spatial Distribution Of Peats And Organo-Mineral Soils In Scotland And Wales

1.2.1 Scotland

Introduction: The 1:250 000 scale soil map of Scotland is the only spatial dataset available in digital format for the whole of Scotland. It uses the mapping concept of landscape units with associated soils. Map units can comprise a range of soil types (major soil subgroups) depending on the complexity of the soil pattern on any particular area of land.

Methodology: Each of the 580 mapping units was assigned to the dominant soil type (major soil subgroup) within it and using GIS, the spatial distribution of the relevant major soil subgroups was plotted and the area of each determined.

Results: The areas of the relevant major soil subgroups are shown in Table 1.2. The distribution of blanket peats and peaty gleys (the most prevalent true peat and organo-mineral soil types) in Scotland are shown in Figure 1.1.

Table 1.2 Area of peats and organo-mineral soils of Scotland

Figure 1.1 Distribution of (a) blanket peats and (b) peaty gleys in Scotland

1.2.2 Wales

Peat soils: Taylor (1975) in the first published figures for Wales estimated that there were 1588 km ² of peat soils (9.7% of the Principality based on total area of agricultural land of 16,335 Km ²; Welsh Agricultural Statistics, 2005). Comparison with more recent surveys is difficult because no depth or organic matter content were given to distinguish between organic and non-organic soils. Soil associations dominated by peat were mapped over 706 km ² (3% of the Principality) during the National Soil Map programme (Rudeforth et al., 1984) (Figure 1.2a). Two soil associations dominate the upland areas, Crowdy 1 (222 km ²) and Crowdy 2 (441 km ²). Amorphous raw peat soils of the Crowdy series are the most common soil series in both associations. Stagnopodzols of the Hafren series and stagnohumic gley soils of the Wilcocks series are common in the former and less humified peat soils with moss and Sphagnum remains (Winter Hill series) are in the latter. Crowdy 1 is mapped discontinuously across the main north to south watershed, generally above 350 m O.D., at the heads of streams and on saddles between the major hills. Crowdy 2 occupies wide upland tracts of blanket bog and scattered peat-filled basins from sea-level at Borth and Tregaron to more than 600 m O.D. in east Wales. In the lowlands some 34 km ² of earthy peat soils (Adventurers' and Altcar associations) have been mapped, almost entirely under permanent grassland.

Small areas of less than 1 km ² were not identified during the National Soil Map programme, nor were the specific locations of peat soils within the associations. Other associations include small areas of peat soils, for example Wilcocks 2 (721d).

Figure 1.2 Distribution of (a) peat soils, and (b) organo-mineral soils in Wales

Organo-mineral soils in Wales: The extent of organo-mineral soils is given in Table 1.3 and Figure 1.2b. The thickness of their organic horizons is as described in a Defra report ( UK soil database for modelling soil carbon fluxes and land use for the national carbon dioxide inventory - project SP0511) as summarised in Bradley et al. (2005).

Table 1.3 Extent of organo-mineral soils in Wales

Figure 1.3 Thickness of organic layers in Welsh soils

1.3 Calculation of organic matter depth for organic soils

1.3.1 Scotland

Introduction: Most recent work assessing the carbon stocks present in Scottish and UK soils (Bradley et al. (2005) focused on the top 1 metre of the soil surface, although it was realised that many of the peat deposits were deeper. Assigning depths to the peat polygons occurring on the 1:250 000 scale soil map gives a more realistic estimation of the carbon stocks contained in the peat deposits.

Methodology: The 1:250 000 scale digitized soil map of Scotland formed the framework for the work and the 1328 peat polygons occurring on the map were extracted along with a unique National Grid Reference ( NGR), the area of each polygon, ranging in size from 1.25 ha to 53989.3 ha, and the type of peat deposit mapped. The map units and corresponding peat type are shown in Table 1.4.

Table 1.4 Peat map units and type on 1:250 00 scale soil map of Scotland

All sources of peat data available were examined to ascertain peat depths and comprised the following sources:

1) The peat bogs occurring in the peatland database compiled by Sally Ward (circa 1991) which included bog name, eastings and northings of roughly the centre of the bog and mean depth, calculated from the depthing transects across the bog.

2) Depth information from Department of Agriculture and Fisheries for Scotland reports - Scottish Peat (1962), the second report of the Scottish Peat Committee and Scottish Peat Surveys Vols 1-4 (1964-1968).

3) Reports held at The Macaulay Institute by the Scottish Peat Committee ( SPC) and International Survey Committee ( ISC) Reports of individual bogs were checked and relevant data extracted. Soil Survey of Scotland memoirs were also referred to and data extracted.

4) Peat Survey field notebooks held at the Macaulay Institute were studied and depth information extracted. In most instances this necessitated calculating average depths, sometimes from 300 or more individual recordings.

5) Peat Survey maps were also checked and, although data on most of the bogs had been ascertained from sources mentioned previously, some missing depth data was available, including some additional bogs.

6) More recent reports on survey work, undertaken on the main bogs used for commercial extraction were made available, which allowed more up to date measurements for the estimation of peat depth.

7) Forestry Commission site survey reports.

All bogs with known depth data (278) were entered in a spreadsheet containing bog name, NGR, average depth and source of data. Using GIS, these bogs were matched using the NGR, with peat polygons on the 1:250 000 scale soil map. Remaining polygons without depth data were assigned relevant data using trends from the information gathered above, local knowledge of the landscape and any depth information gathered by the Soil Survey of Scotland during systematic soil mapping.

Results:

1. Spreadsheet of all peat bogs with measured depth data, detailing name, average depth calculated from transect depth data and the source of information.

2. Spreadsheet detailing all peat polygons with NGR, area, estimated depth. The distribution of carbon stocks in soils at depths greater than 1m are show in Figure 1.4.

Figure 1.4 Distribution of carbon stocks at depth greater than 1m in Scotland

1.3.2 Wales

Soil patterns of the upper 100 cm of soil material have been mapped at scales from 1:10,000 to 1:250,000. However, there is a considerable gap in the knowledge and understanding of the thickness of peat deposits across Wales (and England) below 100 cm depth. The primary reason for this is that traditional methods of field survey use 100 cm long hand augers. A secondary reason is that considerable additional effort, time and equipment is needed to investigate below 100 cm depth.

Given the costs in terms of the time and labour required using an auger to measure the peat below 100cm depth, the approach adopted was to carry out a desk study of existing sources of data from Wales. Although some attempts have been made at shallower depths with radar the technique is not suitable for peats as deep as found here. In the desk study, many hundreds of paper-based observations were reviewed to derive the tables for Pwllpeiran, the Dee catchment, National Soil Inventory and Forestry Commission sites. The outcome of these investigations gives the best possible estimates available for the amount of organic material below 100 cm that is now available.

For consistency, the same methodology for material below 1 m depth was used as for material above 1 m in the estimate of soil carbon in the UK (Bradley et al., 2005).The one exception was to reduce the Bulk Densities for some of organic layers in the light of work both within the Project, and for from other studies. Within Wales three main blocks of land were identified where there was evidence suggesting that the thickness of additional organic material below 1 m depth would be similar (Table 1.15). CEH then added the stocks from these deep organic layers to the totals derived in the UK study (Bradley et al., 2005) using the same methodology and spatial database to compile the data presented in Figure 1.16.

The main assumption was that the stock of organic carbon, bulk density and layer thicknesses are the same for a soil series wherever it occurs in Wales. This is consistent with the usual assumptions for estimates of soil properties, used across the soils community in similar exercises. Four field-based studies with measured data on the thickness of peat soils from various parts of Wales have been re-assessed. In addition, as part of the ECOSSE project, transects were made across three areas with landscapes typical of south mid and north Wales.

Taylor (1974) presented some estimates of peat depths (Table 1.5) and suggests that the bulk of peat above 300 m OD is characteristically 1.83 ± 0.61 m in depth but up to 3.05 m in places. In the lowland peat areas, he estimated that Cors Fochno (Borth Bog) attained 7.62 m and Cors Caron (Tregaron Bog) 9.15 m depth.

Table 1.5 Estimated thicknesses of peat deposits in Wales (after Taylor, 1974)

Lowland Peat Inventory: Thirty seven lowland peat areas totalling 50.9 km ² were identified by Burton and Hodgson (1987) in the Lowland Peat Inventory. The thickness of the peat was estimated, with some measurements in the larger areas.

National Soil Inventory: Examination of the National Soil Inventory data for Wales shows that there are 29 sites described as peat soils with less than 90 cm peat (median thickness 80 cm) and 7 sites with more than more than 90 cm (mainly described as "90+", with one observed depth of 300 cm) (Table 1.6).

Table 1.6 Summary of results from surveys of thicknesses of peat soils across Wales

Dee catchment study: Rudeforth and Thomasson (1970) reported on survey work across the upper Dee catchment. Their original survey records have been re-examined and the medians for peat thicknesses calculated (Table 1.6 and Figure 1.4). These were based on 65 sites with peat 40 cm or more thick. Sites with more than 99 cm peat were recorded in the field as "90+ cm".

Figure 1.4 Thickness of peat soils in the Dee catchment

Waun Figlen Felen, Powys: Waun Figlen Felen is a small basin bog at 450 m OD in the Brecon Beacons above the Dan yr Ogof cave system (SN825180). Parts of the Waun are suffering from severe erosion where the surface vegetation has disappeared following uncontrolled burning. Three transects were made from south east to north west across the area, the thickness of peat measured with an extendable gouge auger (Figure 1.5) and the median peat thickness calculated (Table 1.6).

Figure 1.5 Distribution of peat thickness along transects on Waun Figlen Felen

Pwllpeiran, Ceredigion: Pwllpeiran farm extends over 1130 ha south of Plynlimon in the upper Ystwyth valley (SN820780). The four transects were located at 480 to 550 m OD based across areas dominated by peat soils in the grid survey of 1985. The thickness of peat was measured with an extendable gouge auger (Figure 1.6) and the median peat thickness calculated (Table 1.6).

Figure 1.6 Distribution of peat thickness along transects at Pwllpeiran

Meigneint, Ffestiniog, Gwynedd: The moorland below Llyn y Dywarchen west of Mynydd Meigneint (SH760420) rises from 420 to 500 m OD. The thickness of peat was measured with an extendable gouge auger (Figure 1.7) and the median peat thickness calculated (Table 1.6).

Figure 1.7 Distribution of peat thickness along transects at Meigneint, Ffestiniog

1.4 Validation of mapped estimates of carbon stocks in organic soils in two upland catchments

1.4.1 Introduction

This work aims to assess the accuracy of mapped estimates of soil C stocks in two upland catchments. The first site was located in Plynlimon, mid Wales (SN800860) and the second in Glensaugh near Aberdeen, Scotland (NO650800). In both areas three 1km grid squares were sampled intensively (200m grid), with samples taken from the organic and mineral horizons (where present). Measurements included depth of organic horizon, depth of mineral horizons, bulk density (organic horizon only) and % C (Table 1.7). The sampling design enabled an estimate of the carbon stock to be made, which could then be compared to values contained in the National Soils Map (Natmap).

Table 1.7 Soil properties measured at Plynlimon and Glensaugh

1.4.2 Calculating carbon stocks

The stock of carbon, in t C ha ^-1, was calculated for the organic and mineral horizons separately and then added together (Figure 1.8). For the organic horizon, the depth, bulk density for 0-15cm and % carbon for 0-15cm was used in the calculation. The errors associated with using these depths are discussed below (Section 1.4.4). For the mineral soil, the depth of each horizon, % carbon and estimated bulk density was used. Since bulk density was not measured for the mineral soil, representative figures were used for each horizon based on soil type and land use (Source: NSRI and MLURI).

1.4.3 Differences between measured and modelled carbon

The carbon stock calculated for each area was compared to the carbon stock predicted from the Defra funded GHG inventory project (Defra project SP0511). The results for each 1km square and for the total area are given in Table 1.8. Both Plynlimon and Glensaugh show large differences between the measured and modelled carbon for each 1km square. At Plynlimon, some squares are overestimated and others underestimated. For example, in Square 1 at Plynlimon the amount of carbon determined from the measured data is 23 Kt C, compared to 68 Kt C from the model, a difference of 193%. At Glensaugh, the model underestimates the amount of carbon in each square. The main reason for these differences relates to the proportion of organic soil. At Plynlimon, the model data shows considerably more peat for Square 1 then was observed. Whereas at Glensaugh, more peat was found in each square then predicted. This is discussed in the following section.

Figure 1.8 Total carbon (t C ha ^-1), including organic and mineral layers for (a) Plynlimon and (b) Glensaugh.

When the differences from each square are summarised as a total, however, they appear to be less, with an error of 28% for Plynlimon and 24% for Glensaugh. Therefore, although there are large differences within individual squares, overall the differences appear smaller. This is, however, only for six 1 km squares and the same might not be true elsewhere.

Table 1.8 Measured and modelled values of Kt C for Plynlimon and Glensaugh.

1.4.4 Sources of uncertainty when estimating carbon stocks

Two main sources of uncertainty were identified when determining carbon stock, the value used for bulk density and the accuracy of defining the coverage of different soil types due to differences in the sampling density required for different mapping scales (1:250,00 for Wales as a consistently available scale for the whole Principality and 1:25,000 for the squares mapped in this Module). A further factor could be variations in % C, however, since this changed only slightly down the profile it was not considered here.

Bulk density: At both sites, the bulk density of the organic horizon decreased with depth, and varied depending whether it was a deep or shallow peat. For example, at Plynlimon the bulk density for deep peats at 50-65cm (0.12 g cm ³) was significantly less then 0-15cm (0.20 g cm ³) and less then half that for 0-15cm in shallow peats (0.31 g cm ³). To examine the effect that bulk density has on the carbon stock, different bulk density values were used in the calculation: 1. measured bulk density for 0-15cm, 2. average bulk density for 0-15cm, 3. average bulk density for 15-30cm, 4. average bulk density for 50-65cm. This analysis focuses on the organic horizon only, the mineral horizons were excluded for this purpose. Average bulk density values had to be used for 15-30 cm and 50-65 cm depths due to the fewer sample points. The maps for Plynlimon and Glensaugh are given in Figure 1.9 and Figure 1.10. For Plynlimon, using the measured or average bulk density for 0-15cm produces very similar results. There are large differences, however, when using the average value for 15-30cm and 50-65cm. The maps for Glensaugh do not suggest that these changes are as extreme.

To examine this further, the total carbon for each area, and within a 1km square, was calculated using the 4 different approaches. For Plynlimon, using an average bulk density for 0-15cm and 50-65cm produces differences of around 40% compared to values determined using the measured bulk density at 0-15cm. These depths also have the greatest errors at Glensaugh, but here the errors are around 25% (Table 1.9). This analysis suggests that the bulk density figure should be representative, particularly for organic soils, otherwise it could produce misleading results.

Table 1.9 Comparison of Kt C ha ^-1 calculated using four different methods (organic horizon only).

Soil type: A second source of error is the proportion of each soil series within a square. Many models use this to determine the carbon stock, as each soil type has associated with it a typical bulk density and percentage carbon.

Figure 1.11 shows the differences in coverage between the observed soil and NSRI mapped soils for Plynlimon. There were three main soil types observed at this site (a raw oligo-amorphous peat soil, a ferric stagnopodzol, and a cambic stagnohumic gley), with an intergrade between the two mineral soils also present. The classified soil for Plynlimon, however, shows that in some squares the amount of deep peat (raw peat) has been over estimated and in others underestimated. This has a significant effect on the carbon stocks, as discussed earlier. At Glensaugh, nine soil types were observed, ranging from deep peat to surface ground water gleys (Figure 1.12). This complexity is not reflected in the classified soil. It should be noted that the 1:250,000 map was not designed to be used at this scale, nor does it give the proportions of soils within map units. These were made available by NSRI from un-published data. However, it highlights the importance of using reliable data which is suitable for the purpose intended, particularly where organic soils are concerned.

Figure 1.9 Plynlimon t C ha ^-1 for the organic horizon using (a) measured bulk density for 0-15cm, (b) average bulk density for 0-15cm (c) average bulk density for 15-30cm and (d) average bulk density for 50-65cm.

Figure 1.10 Glensaugh t C ha ^-1 for the organic horizon using (a) measured bulk density for 0-15cm, (b) average bulk density for 0-15cm (c) average bulk density for 15-30cm and (d) average bulk density for 50-65cm.

Figure 1.11 Proportions of soil series within each grid square at Plynlimon (a) observed soils (b) Classified soil (Source: NSRI - Natmap data 1:250,000 map).

Figure 1.12 Proportions of soil series within each grid square at Glensaugh (a) observed soils (b) classified soil types (Source: Ronnie Milne).

1.4.5 Summary

This work has produced high-resolution maps of measured soil carbon stocks for two representative upland sites: Plynlimon, Wales and Glensaugh, Scotland. Estimates of soil carbon stocks were compared to model predictions, which gave overall errors of 24-28%. Accurate estimates of soil carbon stocks were found to depend on:

• Bulk density. The sample data showed that bulk density varied for both deep and shallow organic horizons, and also decreased with depth. Selecting an unrepresentative value gave large errors.
• Appropriate soil mapping scale. There were large discrepancies between measured and mapped series. The resultant differences in carbon stock estimation were large at the local scale but somewhat reduced at larger scales. The largest errors occurred where peat soils (relating to the depth of the organic horizon) were misrepresented. It should be recognised that most national soil maps were not designed to be used at such fine scales. Soil maps at more appropriate scales are available for more local areas in Wales, for example Rudeforth (1970).

1.5 Estimation Of The Carbon Within The Peats And Organo-Mineral Soils

1.5.1 Scotland

Introduction: Calculation of carbon stocks using the peat depths determined in Section 1.4, in combination with best estimates of bulk density, enhances the data from Defra work (Defra SP0511 project), giving a better estimate of carbon stocks within the peat lands of Scotland.

Methodology: A spreadsheet was constructed of all peat map units with grid references, peat depth, estimated profile horizons, % carbon of each horizon (from analytical data in the National Soils Database), estimated bulk density of each horizon and estimated carbon mass to profile depth. Bulk density values were predicted using regression equations which are described in Appendix 1. After additional sampling, modifications to the bulk density values used in the Defra project, were incorporated and used in estimations of carbon stocks. As this impacted on the original Defra database, the carbon stocks were re-calculated using the new bulk density values and the revised figures sent to CEH (Bush) for appropriate amendments to the database and national totals.

Results: To allow for consistency with the Defra database and spatial calculations, which assumed peat depths of 1m, weighted average depth values for the 6 types of peat (deep blanket, eroded deep blanket, undifferentiated blanket, eroded undifferentiated blanket, basin and eroded basin) occurring on the 1:250 000 scale soil map of Scotland were used. These were weighted averages taking into account the areas of each polygon and the type of peat (Table 1.10).

Table 1.10 Weighted average peat depth in peat map units

Using these depths, the carbon stocks were calculated for the peat polygons and using GIS, the carbon stocks present at depths greater than 1 metre were determined and plotted. In addition to the peat polygons, peat also occurs as a component in other mapping units and these were also included. A similar procedure was adopted and peat depth values were assigned to mapping units where peat (3 types) was a component.

Table 1.11 Weighted average peat depth in other map units

Weighted average mean depths were determined (Table 1.11) and used to calculate carbon stocks for peat present at depths greater than 1 metre where it occurred as a map unit component. Combining the two datasets the carbon stocks at depths greater than 1 metre were calculated (Table 1.12). The carbon stocks values (>1 metre depth) for the peat soils were calculated for each kilometre square. This data was supplied to CEH (Bush) to amend the total national values as calculated previously and a map showing the carbon stocks >1 metre for Scotland was produced.

Organo-mineral Soils: Using soil profile morphological data in the National Soil Database, which is held at the Macaulay Institute, modal profiles comprising soil horizon type, sequence and thickness for each of the component soil series of the 1:250 000 soil map units had previously been assembled and used in work conducted under Defra project SP 0511. Using analytical data, % carbon values for each soil horizon were assigned and in combination with bulk density values, predicted by regression equations ( Appendix 1), carbon stocks were calculated for the depth ranges 0-30 cm, 30-100 cm (Defra SP 0511). With area figures determined by GIS for each soil series, and the profile carbon stock totals, the carbon stocks for each series with an organic surface horizon was calculated. The soil series were then grouped according to major soil subgroups and the carbon stock totals for these major soil subgroups and the organic soils estimated (Table 1.12).

Table 1.12 Estimated carbon stocks (MtC) for peat and organo-mineral soils of Scotland

* Only those major soil subgroups with an organic surface horizon.
** Excludes mineral soils.

1.5.2 Wales

Bulk density of peat soils: Vernik (2005) visited a number of National Soil Inventory sites across upland Wales previously described as having > 40 cm of peat. Samples of known volume were taken at 10, 30 and 50 cm depth and the bulk densities measured (Figure 1.13). The bulk densities of the samples at 10 cm depth are related to land use in Figure 1.14. This new information on bulk density of peat soils have been used to inform the conversion of weight to volume calculations of carbon stocks (Bradley et al., 2005).

Figure 1.13 Variation in bulk density with depth in six Welsh peat soils

Figure 1.14 Bulk density at 10 cm depth by land use in eight Welsh peat soils

Revised inventory of soil organic carbon: The inventory of soil organic carbon ( SOC) for the UK (Milne et al., 2004 and Bradley et al., 2005) quantified the amount of SOC in the profile to a maximum of 100 cm depth. One of the aims of the ECOSSE project was to quantify the amount of SOC below 100 cm depth to complete the picture of soil carbon stocks in Wales and Scotland. The approach adopted for Welsh soils, given that there is no systematic depth data, was as follows:

1. A "bog-wide" depth was allocated to each lowland area dominated by peat soils. Each is associated with Adventurers or Altcar soils and the results summarised in Table 1.10,

2. A "region-wide" depth was allocated to each association mapped in upland Wales,

3. This added a fixed amount of SOC (as kg x 10 ⁶ per km ²) to each series, regionalised as much as is feasible given the nature of the data,

4. Each series has a known distribution from the spatial database compiled by Bradley et al. (2005) and therefore spatial distribution of the addition SOC can be mapped (Figure 1.16).

Table 1.15 Additional soil organic carbon below 100 cm (kg x 10 ⁶ per km ²)

NA - no peat below 100 cm depth.

Organic soils (peats and organo-mineral soils) cover about 20% of the area of Wales, yet contain 50% of the carbon because peats are carbon rich all the way down the profile and organo-mineral soils have organic rich top soils. In mineral soils, which cover the remaining 80% of Wales, the maximum soil organic carbon content in the top soil is about 4%, over a subsoil with 1 to 2% organic carbon, whereas the organo-mineral top soils (and peats throughout the whole profile) can be up to 50% organic carbon. Total stocks of carbon in peats and organo-mineral soils in Wales are shown in Table 1.16.

Table 1.16 Estimated carbon stocks (MtC) for peats and organo-mineral soils in Wales

Errors of quantifying organic carbon up to 1 metre depth have been assessed previously. These data formed the basis of the UK study (Bradley et al., 2005). Figure 1.15 shows the organic carbon of topsoils of individual soil series where organic carbon is greater than 5% plotted against the Standard Deviations, showing that the variance becomes greater as the carbon content increases.

Figure 1.15 Organic carbon of topsoils of individual soil series where organic carbon is greater than 5% plotted against the Standard Deviations

1.6 Conclusions

The work in Module 1 provides estimates of the carbon stocks for the peat and organo-mineral soils of Scotland and Wales (organic soils collectively), taking cognizance of previously unused peat depth data and a re-classification of the soils to allow estimates at major soil subgroup level. Estimation of carbon stocks for depths > 1 metre is now possible and in conjunction with previous work (Defra project SP0511), will enable more precise national estimates to be made. Estimates for peats and organo-mineral soils are higher than those quoted in previous work (Bradley et al., 2005), which is due to a combination of factors such as the inclusion of peat greater than 1 metre, modifications to the bulk density and the methods adopted in calculating the areas. The combined map of soil organic carbon stocks in Scotland and Wales are shown in Figure 1.16.

Figure 1.16 Distribution of organic carbon in soils (developed from Milne et al., 2004 and Bradley et al., 2005).

« Previous | Contents | Next »