Information and Advisory Note Number 116
Back to menu
Trends in the natural heritage: land cover 1940S' 1980s
1. Introduction
1.1 Looking back over the latter-half of the twentieth
century, considerable changes are evident in Scotland's urban and rural
environments. Some changes, such as urban expansion and afforestation, have been
striking. More subtle and cumulative changes, such in the structure of farmland
or the extent and condition of moorland, may be less obvious. Yet they are no
less relevant to the visual appearance of the countryside, to its wildlife, and
to opportunities afforded for outdoor recreation and enjoyment.
1.2 Archive air photography has made it possible to look back over this period,
to investigate and quantify change: its magnitude, rate and geographical
variation. A land cover audit, known as the 'National Countryside Monitoring
Scheme' (NCMS), sampled photography from around 1947 (as the imperative for
economic growth in the aftermath of the Second World War promoted the
modernisation and expansion of farming and forestry) to around 1973 (when the UK
joined the European Economic Communities) and then to 1988 (as environmental
balance in rural policies was receiving greater recognition).
1.3 This Information and Advisory Note uses results from the NCMS to summarise
changes in Scotland's land cover between the late 1940s and late 1980s. Forces
responsible for change and relationships between land cover features are
discussed. The note concludes with an outline of the dataset itself and how it
has been utilised.
2. An overview of key findings
2.1 Long established and semi-natural features were displaced by modified or
artificially created features [Table 1, Figure 1].
2.2 In the uplands, mire drainage and access into remote areas were associated
with ditching and track cutting. In the lowlands, simplified cropping patterns
and field amalgamation brought about hedgerow removal [Table 2, Figure 2].
2.3 Trends in afforestation and urban development were fairly constant over the
study period. The decline in blanket mire
increased between the 1970s and 1980s. Loss of coniferous woodland, evident from
the 1940s to the 1970s, was arrested [Figure 3].
2.4 There was an overall reduction of long
established and semi-natural features by 18%,
with change being greatest in the south and
east [Figure 4].
Urban development
2.5 Although the population of Scotland altered little from 1951 to 1991 (around
5 million), pressure on housing increased by 40% (to more than 2 million
households). With a rapid growth of car ownership and mobility, roads became a
major focus of transport policy (between 1950 and 1990, vehicles on British
roads increased from around 4 million to around 25 million). Leisure time and
associated activities increased.
2.6 Key urban changes:
• Built land increased by an estimated 46%, mainly on smooth grassland and
arable farmland.
• Land managed for formal recreation (playing fields, golf courses, etc.)
increased by around 138%, mainly on grassland and arable land.
• The area of transport corridor (roads) increased by about 22%, mainly on mire,
arable, smooth grassland and heather moorland.
2.7 Urban development on lowland farmland
was in-part compensated for by bringing lower
grade land into production. Where conditions
were favourable, moorland was converted to
pasture, grassland was improved, and some
better quality grasslands were converted to
arable use.
Arable
2.8 The Agriculture Act 1947 introduced a
framework of measures for growth. Research
and advice were directed towards labour-
saving and yield-increasing technologies.
Agro-chemicals permitted simplified cropping
rotations, making continuous cereal growing
possible in place of mixed arable and livestock
farming. Grants from the early 1950s
promoted intensification and specialisation,
pasture improvement, field enlargement and
land drainage. Productivity gains were
secured through mechanisation, the
application of fertilisers and pesticides,
selective plant and animal breeding, capital
investment in buildings, plant and machinery.
The Common Agricultural Policy, from 1973 onwards, gave further stimulus to
arable expansion and grassland improvement. With little overall change in the
agricultural land area from 1947 to 1988, agricultural holdings were amalgamated
and the agricultural labour force was substantially reduced.
2.9 Key arable changes:
• Arable land expanded by 11%, and became more dominant in the east
• The increase was mainly at the expense of smooth grassland (improved pasture).
• Hedgerow length was reduced by half, from over 40,000 kilometres in the 1940s
to under 20,000 kilometres in the 1980s.
Grassland
2.10 Grassland improvement (mainly to increase stocking rates of cattle in the
lowlands, sheep in the uplands) was accomplished through land drainage,
re-seeding and fertiliser use. From the 1960s to the 1980s the average rate of
nitrogen on grassland in the UK trebled; silage making increased five-fold
(replacing hay and turnip feeds). On hill land, measures such as drainage,
liming, bracken control and re-seeding continued into the 1970s (when they were
replaced by headage payments). Despite a net reduction in grassland area, output
of British beef and lamb doubled from the 1940s to the mid-1980s; milk increased
more than three-fold.
2.11 Key grassland changes:
• Rough grassland decreased by 10%, mainly due to afforestation and grassland
improvement
• Intermediate grassland increased by 15%, mainly from rough grassland, heather
moorland and drained mire.
• Smooth grassland (permanent pasture in the lowlands) decreased by 11%, mainly
due to conversion either to arable use or urban development.
Forest
2.12 Faced with much depleted timber
reserves at the end of the Second World War,
the Forestry Act 1945 introduced measures for
expanding private and state planting. To avoid
competition with agriculture on the most
productive soils, forest expansion was directed
into upland areas. There, the sale of large
estates provided a supply of land for
afforestation. Improved plough design allowed
blanket peat to be drained for afforestation.
2.13 Afforestation took place mainly on mire (2400km2), heather moorland
(2200km2) and rough grassland (2500km2). Mire was further reduced by drainage
operations (1900km2). Conifer planting changed the character and extent of
former woodland.
2.14 Key woodland changes:
• Forest and woodland (combined) expanded by nearly 200%.
• Former woodland was replaced by coniferous plantations: broadleaved woodland
decreased by 23%; mixed woodland by 37%; and coniferous woodland by 47%.
• Young coniferous plantation increased nine-fold and mature plantation
increased more than four-fold, mainly on heather moorland, rough grassland and
blanket mire.
• The length of ditches, largely associated with mire drainage, doubled.
Moorland and Peatland
2.15 While the uplands were becoming more intensively utilised for farming and
forestry, there was also a marked increase grazing pressure from red deer
(100,000 in 1950 to 300,000 in 1989).
2.16 Key upland changes:
• Heather moorland was reduced by 23%, due mainly to afforestation and
conversion to rough grassland.
• Blanket mire decreased by around 21% and lowland mire decreased by around 44%,
mainly due to afforestation and drainage.
• The length of un-surfaced tracks (mainly in the uplands) increased by around
29%.
3. Utilisation of results
3.1 Results from the NCMS have been utilised at a range of geographical scales
[Table 3], in support of environmental research and education, environmental
profiling, planning and reporting.
Table 1: 1940s-1980s change, areal features
Figures ordered by size of change (with statistically significant change judged
at the 5% level)
Total Area: 77837 km²
Notes
• Significant changes are defined as those for which the likelihood that no
change took place is less than 0.05.
• Non-significant changes should be treated with caution as they may be a
product of 'sampling error'.
• The lower and upper limits define a range of values which contains the true
area or length with 95% confidence.
• Percentages show the increase or reduction of the feature as a percentage of
its 1940s extent.
Table 2: 1940s-1980s change, linear features
Figures ordered by size of change (with statistically significant change judged
at the 5% level)
Notes
• Significant changes are defined as those for which the likelihood that no
change took place is less than 0.05.
• Non-significant changes should be treated with caution as they may be a
product of 'sampling error1.
• The lower and upper limits define a range of values which contains the true
area or length with 95% confidence.
• Percentages show the increase or reduction of the feature as a percentage of
its 1940s extent.
Table 3: The utilisation of results
4. Data Source
4.1 Archive aerial photography exists for the whole of Scotland, around three
dates: c1947, c1973 and c1988. These were utilised as source data for
interpretation.
4.2 It was impracticable to interpret photography for the whole of Scotland.
Instead, a stratified random sample (representing 7.5% of Scotland's land area)
was developed to estimate the extent of extensive land cover types and changes
through time. Stratification was achieved (on a regional basis) through the
classification of Landsat multi-spectral scanner images. Each stratum was then
sampled to select photography for interpretation. The sample size was determined
by a need to detect changes of 10% or more in the major features, with 95%
confidence.
4.3 Structural elements of land cover were classified into 31 areal types and
five linear features. For each sample square (there were 467 in all, 5km x 5km
or 2.5km x 2.5km in size), they were mapped at a scale of 1:10,000. The minimum
mapping resolution was around 0.1 ha for areal features and 30m for linear
features.
4.4 Land cover maps for each sample square were digitised and processed on a
Geographical Information System (GIS). This enabled the land cover in each
sample square to be calculated for each time period. Overlay analyses performed
the computation of land cover change between time periods. Statistical software
allowed GIS outputs to be combined with sampling information to estimate extent,
change and interchange.
4.5 In June 2000, the entire dataset is to become more widely accessible in
CD-ROM form, for use in an ArcView environment at the desktop.
4.6 The study has gained recognition as one of the most thorough and
comprehensive of its kind. Despite rigorous quality control measures, a range of
sampling, interpretation and digitising errors are nevertheless associated with
studies of this kind. Together with the methodology, these have been assessed
and documented. An independent accuracy assessment concluded that
misclassification rates were comparable with similar studies.
5. Further reading
Mackey, E.C., Shewry, M.C. and Tudor, G.J. (1998). Land Cover Change: Scotland
from the 1940s to the 1980s. Edinburgh: The Stationery Office.
Tudor G.J., Shewry M.C., Mackey E.C., Elston DA and Underwood F.M. (1999). Land
Cover Change in Scotland: the Methodology of the National Countryside Monitoring
Scheme. Scottish Natural Heritage Research, Survey and Monitoring Report No 127.
Elston, DA, Gauld, J.H., Miller, J.A., Shewry, M.C. and Underwood, F.M. (1999).
The National Countryside Monitoring Scheme Accuracy Assessment. Scottish Natural
Heritage Research, Survey & Monitoring Report No. 133.
Scottish Natural Heritage (1998). Land Cover Change in Scotland. Advances Series
No. 5. Perth: Scottish Natural Heritage.
A more colourful view can be obtained from the SNH web site: www.snh.org.uk
(under environmental audit).
6. Authors
Ed Mackey,
Head of Environmental Audit Unit
Mike Shewry,
Environmental
Statistician
7. Contact for advice and information
Ed Mackey
Environmental Audit, Chief Scientist's Unit
Scottish Natural Heritage
2 Anderson Place
EDINBURGH
EH6 5NP
Tel: 0131-447 4784
Back to menu