Agriculture
IDENTIFICATION OF CROPPING AREAS AT RISK
TO SOIL EROSION IN ENGLAND
The risk of erosion to agricultural crops is a serious and long term problem. In order to determine the total hectarage of agricultural crops at risk from soil erosion and their location, the highly cropped areas in England at risk to soil erosion by water were identified. Also the number of holdings affected was calculated so that specific erosion risk management advice could be better targeted in the future.
Soil erosion risk and crop distribution information was combined with the local knowledge of ADAS consultants to evaluate erosion risk in areas where the percentage of arable crop cover is high. Three definitions of risk areas were determined: very high risk, high risk and locally high risk. It was determined that arable cropping is at some level of risk from water erosion in 34 separate areas throughout England. The total hectarage of arable crops considered to be at very high risk to water erosion is 17,990, with 62,170 hectares of crops at high risk and 74,590 hectares of crops at locally high risk.
The risk of soil erosion is a serious and long term problem with considerable environmental and economic implications. Nuisance problems caused by soil being eroded from fields and deposited onto roads, urban areas and into rivers are considerable and subsequently require remedial action. Clearing the build up of sediment in water courses, removing soil from roads, road-side reinforcement and the monitoring of aquatic environments are undertaken at considerable economic cost. This cost may be contributed to by the erosion of soil from agricultural areas.
In addition to these nuisance problems there are more serious concerns as to the effect of continually losing topsoil from agricultural areas. The impact of soil loss on farming systems, crop yield and the long term fertility of the soil is considerable as topsoil is only slowly replenished by natural soil-forming processes. Again the economic implications are considerable both in the short and long term.
In order to gain a better understanding of the economic impact of the loss of soil due to water in agricultural areas and the continual environmental problems subsequently faced by farmers in England the Ministry of Agriculture, Fisheries and Food (MAFF) commissioned ADAS (formerly an executive agency of MAFF) to undertake a study. The aims of this study were to determine the total hectarage of agricultural crops at risk from soil erosion by water and the location of the areas most severely affected. Also to identify the number of holdings affected so that specific erosion risk management advice could be better targeted in the future.
A desk study commenced in August 1996 which involved the combination of soil erosion risk and crop distribution information with the expert local knowledge of ADAS consultants. The study was completed in January 1997.
In order to determine the spatial distribution of agricultural areas at risk from soil erosion by water a number of different spatial datasets needed to be `created' or analysed separately and in combination. These included soil erosion risk assessment classifications for England and agricultural crop distribution information. This information was subsequently combined with expert advice provided by individual ADAS consultants who provided invaluable local knowledge of different areas.
All spatial data was implemented, manipulated and analysed using the ArcInfo geographical information system (GIS). The spatial analysis of this information subsequently resulted in the development of a GIS based map which classified erosion risk according to areas where the percentage of arable crop cover is high and where a `real problem' to agricultural operations was perceived. Three definitions of risk areas were determined: very high risk, high risk and locally high risk. A number of different areas falling within each classification were identified.
Two sources of soil erosion risk assessment information of England currently exist. One of these is the `Risk of Soil Erosion in England and Wales by Water on Land under Winter Cereal Cropping' map as produced by the Soil Survey of England and Wales (SSLRc) (SSLRc, 1993). This is available in hard copy format at 1:650000 or 1:250000 scale. Subsequently, this source of risk assessment information was unavailable for use directly in the study due to copyright issues and limiting time factors preventing the digital capture of the mapped information. However, the ADAS consultants used to provide expert local knowledge of different areas referred to the published maps in order to aid their assessments.
The other source of soil erosion risk assessment information was provided by Evans (1990) who categorises erosion risk by soil associations. Evans categorises the 296 soil associations of the National Soil Map of England and Wales (produced by SSLRc) into five categories of erosion risk taking into account land use, landform, soil properties and upland areas with erosion by water or wind (Evans, 1990). The soil erosion risk definitions for water erosion only by soil association were applied to the National Soil Associations dataset for England and Wales held at a 1km grid resolution in the GIS. This spatial implementation resulted in eight categories of erosion risk being defined. These were the five categories of risk as defined by Evans, wind erosion (essentially occurring in East Angila), areas defined as unsurveyed by SSLRc (mainly urban areas and large water bodies) and upland areas as defined by SSLRc's `Risk of Soil Erosion in England and Wales by Water on Land under Winter Cereal Cropping' map. Table 1 details the definition of each of these eight erosion risk categories.
Table 1: Erosion risk classifications based on Evans (1990) and the National Soil Map of England and Wales
| Erosion Risk | Definition | Source |
| Very High Risk | on average > 10% of fields affected and two years in five 20-25% is affected. Volumes of soil eroded are often greater that in any other category. | Evans (1990) |
| High Risk | > 5% of fields per annum with volumes of soil eroded likely to be greater than those in the smaller risk categories | Evans (1990) |
| Moderate Risk | 1-5% of the land at risk of erosion per annum | Evans (1990) |
| Small Risk | < 1% of the land eroded per annum | Evans (1990) |
| Very Small Risk | erosion rare or not at all | Evans (1990) |
Erosion other than Water |
erosion due to wind | Evans (1990) |
Unsurveyed Soil and Water |
urban areas, water bodies and inaccessible | National Soil Associations dataset |
| Upland | upland areas | SSLRc map |
Three different datasets were used to provide information on the spatial distribution of key agricultural crops and landcover. Six crop groups were considered `at risk' of erosion and three landcover types considered `no risk' (Marks et al, 1997). The `at risk' crops were considered to be those at risk of experiencing soil erosion by water and are most affected by loss of yield and soil fertility. The following at risk crop groups were identified: winter cereals & oilseed rape, spring cereals & oilseed rape, root crops, forage maize, field vegetables (excluding peas and beans), and hops & fruit. The no risk landcover types were considered as aiding the reduction of soil erosion and relate to woodland, permanent and temporary grass (Marks et al, 1997).
Agricultural census information for the years 1994 and 1995 were used to provide information on the spatial distribution for all of the at risk crop groups and the permanent and temporary grass no risk groups. The Landcover Map of Great Britain (produced by the Institute of Terrestrial Ecology) was used to provide information on the spatial distribution of woodland throughout England by definitions of deciduous and coniferous woodland. Although agricultural census returns do provide information on the area of woodland only woodland on farmland and not woodland as a whole is considered. Table 2 contains a full list of the different crop/landcover groups considered, source and resolution of the data used.
Due to the spatial format of the erosion risk assessment data being used in the study it was decided that the use of a similarly structured dataset for the agricultural census information would be most appropriate. The 1995 agricultural census data (sourced from the University of Edinburgh) at a 5km grid resolution was subsequently purchased[1]. This dataset was used to determine the total hectarage of land under each crop group throughout England. Totals were calculated as a percentage of the total land area per 5km grid and not as a percentage of the agricultural area as it was considered that non-agricultural land use could influence erosion risk in some catchments (Marks et al, 1997). The 1995 census dataset was used to determine the percentage cover and location of areas with high percentage cover for the at risk crop groups and the no risk landcover groups. However, information on the hectarage and location of the at risk crop groups of field vegetables and hops & fruit were not available as part of the 1995 spatial census dataset.
The 1994 agricultural census data supplied directly to ADAS by MAFF census division[2] was used to calculate the percentage cover and location of areas with high percentage of cover for the two remaining at risk crop groups. The finest resolution of information available on the location of field vegetables and hops & fruit was at the agricultural district level[3]. The data was released at this spatial resolution to ensure that the information would remain confidential to ensure that the extent of individual holdings could not be determined.
Table 2: Details of the Risk Crop/Landcover Groups
| At Risk Crop Group | Source | Resolution |
| Winter Cereals & Winter Oilseed Rape | 1995 census | 5km grid |
| Spring Cereals & Spring Oilseed Rape | 1995 census | 5km grid |
| Root Crops (potatoes & sugar beet) | 1995 census | 5km grid |
| Forage Maize | 1995 census | 5km grid |
| Field Vegetables (ex. Peas and beans) | 1994 census | agricultural district |
| Hops & Fruit | 1994 census | agricultural district |
| No Risk Landcover Group | Source | Resolution |
| Permanent Grass | 1995 census | 5km grid |
| Temporary Grass | 1995 census | 5km grid |
| Woodland | Landcover map of GB | 1km grid |
ADAS Consulting Limited operates on the basis of having a number consultancy offices and research farms distributed throughout England and Wales. Numerous specialised consultants who are agricultural experts are based at each of these centres. These consultants provide advice to local farmers and environmental bodies on a daily basis which frequently involve site visits. Subsequently the knowledge developed by ADAS consultants of their local area is considerable and is often focused on areas where there are agricultural and environmental problems, such as erosion. A number of these ADAS agricultural consultants were involved in providing local advice and guidance on the actual occurrence of soil erosion by water to agricultural crops in their area for the study. The contribution of this advice proved invaluable in incorporating `real world' occurrences of erosion into the analysis (Marks et al, 1997). In order to aid the incorporation of this information on a spatial basis England was divided into six geographical areas as detailed in table 3.
Table 3: Geographical Areas of Erosion Risk to Cropping Analysis
| Geographical Area | Counties Comprising Area |
| South West | Cornwall, Devon, Dorset, Somerset, Avon, Wiltshire |
| South East | Oxfordshire, Kent, West Sussex, East Sussex, Hampshire, Berkshire, Surrey, Isle of Wight |
| West Midlands | Hereford & Worcester, Gloucestershire, Shropshire, Staffordshire, Cheshire, Lancashire, Warwickshire, Merseyside, West Midlands |
| East Anglia | Cambridgeshire, Norfolk, Suffolk, Essex, Buckinghamshire, Bedfordshire, Hertfordshire |
| East Midlands | Lincolnshire, Northamptonshire, Nottinghamshire, Leicestershire, Humberside (south) |
| North | Cumbria, North Yorkshire, Yorkshire, West Yorkshire, Derbyshire, Cleveland, County Durham, Northumberland, Humberside (north), Tyne and Wear |
Maps of each of the crop/landcover groups and erosion risk assessment were produced for each of the six geographical areas at various scales. These maps were supplemented by additional spatial information[4] as required by the consultants to aid their assessment of the actual occurrence of erosion on the ground. No field visits were made and in a few areas which were not well known some field work would have been advantageous.
Information provided by the ADAS consultants relating to the erosion risk were incorporated into the existing datasets to produce a single spatial dataset at a 1km grid resolution to outline the areas experiencing soil erosion by water. Essentially this involved amending the Evans defined erosion risk dataset by specific areas thus curtailing the original method of blanket categorisation by soil association. Refining the erosion risk classification in this manner enabled the amalgamation of numerous items of information and enabled the appropriate adoption of a more representative risk definition framework (i.e. the introduction of a locally high risk category). A total of nine categories were used in the final erosion risk map; very high risk, high risk, locally high risk, moderate risk, small risk, very small risk, erosion other than water, unsurveyed soil & water and upland. Refer to table 1 for the full definition of each category.
During the assessment of soil erosion risk to agricultural crops some originally defined areas of high risk were down graded to moderate risk on the grounds of insufficient `at risk' cropping or flat topography. In other situations areas of moderate risk were upgraded to high risk because of widespread `at risk' cropping. The `locally high risk' category was introduced for areas where the overall risk was considered to be moderate but there were localised high risk areas. This final spatial dataset is highly specialised in order to meet the objectives of identifying the location of the areas most severely affected by soil erosion due to water and determining the total hectarage of agricultural crops at risk.
In order to determine the number of holdings affected by each of the defined risk areas the 1994 agricultural census dataset was used. The number of holdings affected by each area was statistically calculated according to the size of the risk areas identified in relation to the size of the agricultural parish group for which information on holding numbers was available. The final map showing the spatial distribution of different risks to soil erosion by water in England is shown in figure 1.
It was determined that arable cropping is at some level of risk from soil erosion by water in 34 separate areas throughout England. Each of these separate risk areas is shown in figure 2 and numbered by area. Refer to tables 4, 5 and 6 for each area's unique number reference (risk area code).
Four separate areas were determined as being of very high risk. These areas were defined as having an average of 10% of fields affected by soil erosion due to water each year. The total area affected by each of these four areas was 49,586 hectares with 17,990 hectares estimated to be in `at risk' cropping each year. The total number of holdings estimated to be affected by this level of erosion risk is 339 with over seventy-two percent of the cropping considered at risk and the holdings to which they relate occurring in the West Midland geographical area. A full breakdown of each of these four designated areas and their summary statistics is listed in table 4.
Table 4: Area Statistics for Very High Erosion Risk Designation
Area |
Risk Area Code | Very High Risk Areas | Total Area Affected (Ha) | Estimated Area of `At Risk' cropping (Ha) |
Estimated Number of Holdings |
| 1 | West side of Exe Estuary |
3,086 | 870 | 18 | |
| South West | 2 | Lower Otter Valley, East Devon |
2,900 | 780 | 21 |
| 3 | South Somerset Fine Sands |
6,000 | 2,040 | 55 | |
| West Midlands | 4 | West Midland Sands | 37,600 | 14,300 | 245 |
| Total | 4 | 49,586 | 17,990 | 339 |
A total of 163,615 hectares of land is considered to be at high risk of soil erosion. Of this total 62,170 hectares is considered to be in `at risk' cropping each year and is distributed over a total of 1,181 holdings. Over 50% of the holdings affected were found to occur in the South West geographical area whilst the greatest hectarage of land considered to be in at risk cropping occurred in the East Midlands geographical area. These high risk areas are defined as having an average of 5% of fields affected by soil erosion each year. A full list of statistics for each of the sixteen areas designated as at high risk is listed in table 5.
Table 5: Area Statistics for High Erosion Risk Designation
Area |
Risk Area Code | High Risk Areas | Total Area Affected (Ha) | Estimated Area of `At Risk' cropping (Ha) |
Estimated Number of Holdings |
| South | 5 | Eastern South Downs | 24,800 | 11,400 | 105 |
| East | 6 | Lower Greensand - Kent |
6,100 | 2,400 | 27 |
| 7 | Devon - Torbay to Exeter and Credition |
29,515 | 7,400 | 450 | |
| South West | 8 | East Devon to North Somerset |
11,400 | 2,650 | 101 |
| 9 | East Somerset and South West Dorset |
9,800 | 2,250 | 84 | |
| 10 | Chippenham/Calne | 1,500 | 440 | 9 | |
| West Midlands | 11 | South Hertfordshire | 13,100 | 5,570 | 101 |
| East Anglia | 12 | Central Bedfordshire | 11,400 | 5,700 | 53 |
| East Midlands | 13 | Nottinghamshire Sands | 40,400 | 18,500 | 174 |
| 14 | Northumberland | 800 | 260 | 2 | |
| 15 | Tyne and Wear | 1,300 | 250 | 3 | |
| 16 | Durham | 1,500 | 330 | 6 | |
| North | 17 | North Yorkshire | 4,700 | 2,450 | 26 |
| 18 | East Riding | 1,000 | 590 | 6 | |
| 19 | South Yorkshire | 2,900 | 1,130 | 12 | |
| 20 | South West Derbyshire |
3,400 | 850 | 22 | |
| Total | 16 | 163,615 | 62,170 | 1,181 |
A further fourteen separate areas in England were determined to experience locally high risk of soil erosion due to water. The overall risk of erosion in these areas was considered to be moderate but with locally high risk areas present. The total area defined as this designation is 198,700 hectares with 74,590 hectares of this considered to be in `at risk' cropping each year. A total of 1,218 holdings are estimated to be affected throughout England by this designation. Nearly forty percent of the area considered to in at risk cropping each year was found to occur in the East Anglian geographical area although this area had the lowest proportion of holdings affected by this designation. A full breakdown of each of these fourteen designated areas and their summary statistics is listed in table 5.
Table 6: Area Statistics for Locally High Erosion Risk Designation
Area |
Risk Area Code | Locally High Risk Areas | Total Area Affected (Ha) | Estimated Area of `At Risk' cropping (Ha) |
Estimated Number of Holdings |
| 21 | Lower Greensand - Oxfordshire |
5,100 | 1,780 | 19 | |
| South East | 22 | Lower Greensand - Hampshire, West Sussex and Surrey |
29,900 | 8,372 | 122 |
| 23 | Isle of Wight Greensand |
2,300 | 650 | 12 | |
| 24 | Tunbridge Wells Sands |
5,800 | 700 | 26 | |
| South West | 25 | West and South Cornwall and South Devon |
7,800 | 1,000 | 115 |
| 26 | South East Cheshire | 8,300 | 1,830 | 88 | |
| West Midlands | 27 | North Central Shropshire and North West Staffordshire |
26,500 | 9,280 | 217 |
| 28 | Central and South East Staffordshire | 13,100 | 4,190 | 91 | |
| 29 | Newent Sands | 3,400 | 1,190 | 38 | |
| East | 30 | North Norfolk Coast | 5,400 | 1,840 | 31 |
| Anglia | 31 | North West Norfolk | 26,100 | 14,100 | 86 |
| 32 | Yorkshire Wolds | 45,100 | 27,060 | 267 | |
| North | 33 | Cumbria - Vale of Eden | 16,400 | 1,970 | 84 |
| 34 | West Derbyshire | 3,500 | 630 | 22 | |
| Total | 14 | 198,700 | 74,590 | 1,218 |
Areas of very high, high and locally high risk of actual soil erosion by water were identified by combining soil definitions and cropping information with local knowledge on a spatial basis. 34 distinct areas were identified as being at risk of experiencing soil erosion due to water.
Very high risk areas were defined as having an average of 10% of fields affected by soil erosion by water each year. Erosion events tend to be more severe in this category. Four areas of very high risk were identified. 49,586 hectares were identified as being at very high risk to erosion with 17,990 hectares of this estimated to be in `at risk' cropping each year. 339 holdings were estimated to be affected by this level of erosion risk.
High risk areas were defined as having an average of 5% of fields affected by soil erosion each year. Sixteen areas of high risk were identified. 163,615 hectares were identified as being at high risk to erosion with 62,170 hectares of this estimated to be in `at risk' cropping each year. 1,181 holdings were estimated to be affected by this level of erosion risk.
A category of locally high risk was used to identify areas where the overall risk was considered to be moderate but which contained localised areas at high risk of soil erosion by water. Fourteen areas of locally high risk were identified. 198,700 hectares were identified as being at locally high risk to erosion with 74,590 hectares of this estimated to be in `at risk' cropping each year. 1,218 holdings were estimated to be affected by this level of erosion risk.
The total hectarage of crops actually affected by each of the erosion risk categories is likely to be less than estimated in this study. The spatial variation of erosion risk within each 1km grid was not know. Also the spatial distribution of agricultural crops within each 5km grid was not known. Subsequently the exact occurrence of cropping on any erosion risk classification within this study (national scale for England) is not known.
In addition, the total hectarage of crops actually affected by soil erosion in areas defined as being at locally high risk is likely to be much less than 74,590 hectares. This total accounts for all the at risk cropping grown in the designated areas and not in the areas which are actually `at risk'. The exact location of the areas experiencing high risk as opposed to moderate risk within each 1km grid is not known.
This study enabled the identification of specific agricultural areas experiencing soil erosion at a 1km grid resolution for England. The identification of the spatial location of these areas and the numbers of holdings being affected provided invaluable guidance as to the number and location of holdings affected. Subsequently MAFF was able to more efficiently and effectively target its resources on providing erosion management advice to farmers than would have been possible without the undertaking of this study.
Chambers, B.J., Davis, D.B., and Holmes, S. (1992). Monitoring of Water Erosion on Arable Farms in England and Wales 1989-90. Soil Use and Management 8, 163-170.
Colbourne, G.J.N. and Staines, S.J. (1985). Soil Erosion in South Somerset. Journal of Agricultural Science. Cambridge University Press. 104, 107-112.
Evans R. (1990). Soils at Risk of Accelerated Erosion in England and Wales. Soil Use and Management 6(3) 125-131.
Marks, M.J., Soloman, D.R., Johnson, P.A., Watson, R.L., Royal, S.M., Richardson, S.J., and Goodlass, G. (1997). Soil protection Studies: Identification of Areas of the Country at High or Very High Risk to Soil Erosion by Water. ADAS report to the Ministry of Agriculture, Fisheries and Food.
Soil Survey and Land Research Centre (1993). Risk of Soil Erosion in England and Wales by Water on Land under Winter Cereal Cropping. SSLRc, Cranfield University, Silsoe, Bedford.
Robinson, D.A. and Boardman, J. (1988). Cultivation Practice, Sowing Season and Soil Erosion on the South Downs, England: A Preliminary Study. Journal of Agricultural Science. Cambridge University Press. 110, 169-177.
Author: Deborah Soloman
Position: Research Consultant
Land & Environment Business Development Team
Company: ADAS Consulting Limited
Address: ADAS Gleadthorpe Research Centre, Meden Vale,
Mansfield, Nottinghamshire, NG20 9PF, England.
Telephone: +44 1623 844331
Fax: +44 1623 844472
E-mail: Deborah_Soloman@adas.co.uk
[1] Parish census information is provided to the University of Edinburgh by MAFF who then spatially manipulate the data to ensure that the confidentiality of the data is maintained.
[2] Although the agricultural census data is collected on a parish basis the statistical data is released by agricultural parish groups, agricultural districts or by county.
[3] Amalgamated from agricultural parish groups prior to release by MAFF in order to maintain the confidentiality of the data.
[4] such as specific soil association, crop hectarage, proximity to water course/urban area, etc.