Читать онлайн книгу "Scotland"

Scotland
Peter Friend


Harnessing recent developments in computer technology, the latest New Naturalist volume uses the most up-to-date and accurate maps, diagrams and photographs to analyse the diverse landscapes of Scotland.Most people share an enthusiasm for beautiful and breathtaking scenery, explored variously through the physical challenge of climbing to the top of the tallest mountains or the joy of viewing the work of a painter; but while easy to admire from a distance, such landscapes are usually difficult to explain in words.Peter Friend highlights the many famous and much loved natural landscapes of Scotland, ranging from the rolling, agricultural lowlands of the east to the wild and rugged mountains of the west, from the whitewashed villages of Galloway to the traditional fishing ports of the east. He provides detailed explanations for the wide variety of natural events and processes that have caused such an exciting range of surroundings.Setting apart the topography that has resulted from natural rather than man-made occurrences, Friend focuses on each region individually, from the windswept islands that fringe the Atlantic to the sheltered straths of Perthshire, and explains the history and development of their land structures through detailed descriptions and colourful diagrams.Illustrated with beautifully detailed photographs throughout, Scotland comprehensively explores the formation of these wonderful landscapes that are so universally admired.On some devices, certain links to a figure (or references to a page) before or after or very close to the link itself may not work every time. Thanks for your patience. We hope you enjoy the ebook editions of the Collins New Naturalists series.







EDITORS

SARAH A. CORBET, ScD

PROF. RICHARD WEST, ScD, FRS, FGS

DAVID STREETER, MBE, FIBiol

JIM FLEGG, OBE, FIHort

PROF. JONATHAN SILVERTOWN

*

The aim of this series is to interest the general

reader in the wildlife of Britain by recapturing

the enquiring spirit of the old naturalists.

The editors believe that the natural pride of

the British public in the native flora and fauna,

to which must be added concern for their

conservation, is best fostered by maintaining

a high standard of accuracy combined with

clarity of exposition in presenting the results

of modern scientific research.


THE NEW NATURALIST LIBRARY

SCOTLAND

Looking at the Natural Landscapes

PETER FRIEND

with

LEAH JACKSON-BLAKE

and

JAMES SAMPLE












Contents

Cover (#ucb68bbe9-aaff-5d97-8ed0-a64f7bb5a3b2)

Title Page (#u4d07ffb3-d0af-5974-befa-9e627cf004a1)



Editors’ Preface

Authors’ Foreword and Acknowledgements



CHAPTER 1 - Looking at Scotland’s Landscapes

CHAPTER 2 - Surface Modifications

CHAPTER 3 - Movements of the Earth from Within

CHAPTER 4 - Episodes in the Bedrock History of Scotland

CHAPTER 5 - Later Surface Modifications

CHAPTER 6 - Area 1: Galloway

CHAPTER 7 - Area 2: Southern Borders

CHAPTER 8 - Area 3: Jura to Arran

CHAPTER 9 - Area 4: Glasgow

CHAPTER 10 - Area 5: Edinburgh

CHAPTER 11 - Area 6: Mull

CHAPTER 12 - Area 7: Rannoch

CHAPTER 13 - Area 8: Dundee

CHAPTER 14 - Area 9: Uists and Barra

CHAPTER 15 - Area 10: Skye

CHAPTER 16 - Area 11: Affric

CHAPTER 17 - Area 12: Cairngorm

CHAPTER 18 - Area 13: Aberdeen

CHAPTER 19 - Area 14: Lewis and Harris

CHAPTER 20 - Area 15: Cape Wrath

CHAPTER 21 - Area 16: Inverness

CHAPTER 22 - Area 17: Caithness

CHAPTER 23 - Area 18: Orkney

CHAPTER 24 - Area 19: Shetland

CHAPTER 25 - Overview



Further Reading

List of Searchable Terms

The New Naturalist Library

Copyright

About the Publisher (#litres_trial_promo)


Editors’ Preface

IN HIS EARLIER NEW NATURALIST VOLUME on the natural landscapes of Southern England, the author, Peter Friend, presented a new vision of landscape, providing a geological background for our understanding of the distribution and variation of flora and fauna in the lowland parts of Britain.

A division of Britain into lowland and highland regions has often been made in descriptions of our flora and fauna, for example by Arthur Tansley in his classic book on Types of British Vegetation (1911). Now Peter Friend has turned his attention to Scotland, the major highland part of Britain. In contrast to the sedimentary origin of the �soft’ rocks of the lowlands, the �hard’ rocks of Scotland arise from a series of events in the Earth’s crust dating back to the earliest years of the planet, which were far less understood in the days when Dudley Stamp’s New Naturalist volume on Britain’s Structure and Scenery was published in 1946. The resulting structures, now much better understood, underlie Scotland’s great variations in rock type and altitudes. Allied to this is the effect of the northern climate on the distribution of plants and animals, making the Highlands an area of particular interest from the biogeographical point of view, a mountainous region in the far west of Europe, adjacent to the Atlantic.

The illustrations featured in this book take full account of the possibilities of aerial and satellite photography in analysing topography, showing the relation between the geology, the soils, and the directions and angles of sloping features – all factors which must affect flora and fauna. The arrangement into areas, each with a similar treatment and analysis of the landscape, makes the subject very accessible to those interested in the geology or visiting the areas, and to those studying the fauna and flora and wishing to understand the physical background of the natural history. This book is a welcome addition to the New Naturalist Library, and will strengthen our understanding of the important and basic relationships between geology and natural history.


Authors’ Foreword and Acknowledgements

THE PLEASURE OF ENJOYING A LANDSCAPE is greatly increased and deepened by developing some feeling for the events in the history of the Earth that may have caused it. This approach was followed in 2008, when Southern England, by Peter Friend, appeared as New Naturalist 108. The object was to provide a systematic general review of the landscapes visible in the countryside extending from Land’s End in the southwest to East Anglia in the east. Peter has now been joined by two others, Leah Jackson-Blake and James Sample, to apply a similar approach to Scotland.

Peter and his two brothers were brought up in Edinburgh, in the Midland Valley of Scotland, moving with the family for a few years to Peebles in the Southern Uplands, during part of the Second World War. Many of the family activities involved visits to the countryside, and the pleasures and interests of these visits have continued into new generations of the family. The other two authors of this book have recently moved from southern England, where the book has been written, and now enjoy the landscapes of northern Scotland where they live and work.

Landscapes are easy to look at, given reasonable weather conditions, but difficult to describe in words. But developments in computer technology now offer many ways of analysing landscapes using different mapping methods, and these, along with diagrams and photos, form the framework of this book. Working on this imagery has been the main contribution of a succession of enthusiastic helpers. Lucinda Edes, Emilie Galley and Liesbeth Renders, and the second and third authors of this book, have all contributed great skill and enthusiastic innovation to this work, and made the project enjoyable as well as successful.

The home of this project has been the Department of Earth Sciences in the University of Cambridge. Peter Friend walked into the department as a first-year student some 57 years ago, to meet his supervisor, W. Brian Harland, for the first time. Apart from a period in the Scott Polar Research Institute, he has been based in Cambridge Department of Earth Sciences ever since, teaching and exploring the scenery and geology of many parts of the world. This work has included many visits to Spitsbergen (under the guidance of W. Brian Harland), Greenland, Spain, the Arabian Gulf, India and Pakistan. This has been an exciting period to be working in geology in Cambridge, because many key advances in the subject have been made by people working in Cambridge. CASP, originally the Cambridge Arctic Shelf Programme, made a valuable donation in support of the aerial photography used in this book.

All three authors would like to acknowledge their debts to the Cambridge college system. In the case of Peter Friend, his college, Darwin, has provided him with the congenial friendship of many people from diverse backgrounds, and their skills have helped him to remain a generalist in his interests.

Any work of this sort on the British Isles owes a fundamental debt to the British Geological Survey (BGS), now based at Keyworth near Nottingham. The numerous Survey maps and reports provide a remarkable source of carefully observed and objective information. The BGS has readily provided advice and help for this project, and helped to determine the sort of coverage and level that would be best.

The photographs that form such an important part of this book have come from many sources, and we are grateful to the following organisations and individuals for allowing us to use the results of their work (individually credited in the figure captions): Aerographica (Patricia and Angus Macdonald), Nicholas Branson, British Geological Survey, Lorne Gill, Last Refuge Limited (the late Adrian Warren, and Dae Sasitorn), David Law, Planetary Visions/Science Photo Library, Scottish Natural Heritage, Nigel Trewin.

Many other people have made important contributions by providing ideas and materials. These include John R. L. Allen, Wendy Annan, Phillip Gibbard, Alan Smith, Nigel Trewin, Nigel Woodcock and Richard West.

As with the New Naturalist volume on Southern England, this volume is dedicated to the Dr John C. Taylor Foundation, which has provided the financial support essential for the production of the imagery that is such a key part of the book’s presentation. Some 50 years ago, John spent two summers exploring the geology of Spitsbergen with Peter Friend, and the support of his foundation has made both books possible.

We wish to thank HarperCollins Publishers for their support of the New Naturalist series, and particularly Myles Archibald, and then Julia Koppitz, for enthusiasm and help throughout. Hugh Brazier, Martin Brown and Robert Gillmor have brought great talents to different aspects of preparing this book.

The cover shows a view at Siccar Point, Berwickshire (Area 5). When James Hutton and some friends visited in 1788, they recognised what is now known to be an unusally good example of an unconformity. They made geological history by seeing it as evidence of the folding, erosion and deposition of strata over an incredibly long series of episodes in the Earth’s past history (see also picture (#litres_trial_promo)).


CHAPTER 1

Looking at Scotland’s Landscapes



LANDSCAPES AND LANDFORMS

FIGURE 1 (#ulink_2ae4ccda-59c6-5bef-8d97-19db76bbc5b4) IS A SCOTTISH VIEW, showing landscapes that are typical of the confections of topographic shapes, sea, light, colour and atmosphere that are enjoyed by all. The object of this book is to contribute further to that enjoyment by surveying the varied landscapes of Scotland, and to help the reader to discover the stories that lie behind the rich variety.

At least two landscapes are present in this photograph: (1) in the middle distance, terraced hills in autumn gold vegetation extend down to the shores of Loch Slapin in the foreground, about 2 km in coastal length, and (2) behind this landscape rises the dark mass of the Cuillin mountains, providing one of the most famous and distinctive of Scottish landscapes, covering an area some 10 km across and giving a skyline to this photograph that is about 15 km from the photographer.

Many enthusiasts have written about the scenery of specific parts of Scotland. In this book the aim has been to cover the whole country relatively uniformly, because the variations from place to place are interesting in themselves. But this uniformity of approach has made it necessary to adopt a rather broad-brush treatment, whilst establishing the linear scale of features by the use of maps and aerial photographs in which the scale is clear in general terms. It is useful to follow earlier authors who regarded a landscape as an area of land that can be seen from one vantage point. In the case of Figure 1 (#ulink_2ae4ccda-59c6-5bef-8d97-19db76bbc5b4), the oblique aerial view covers two landscapes that are kilometres to tens of kilometres across.

Figure 1 (#ulink_2ae4ccda-59c6-5bef-8d97-19db76bbc5b4) provides a fine example of Scottish scenery that not only allows some questions of scale to be considered, but also illustrates the sorts of features that can be used to investigate the stories behind landscapes. Numerous small cliffs and bays of the coastal cliffs are visible in the foreground, where resistant bedrock provides information about the early history of events in this landscape area. The Cuillin mountains themselves are very special in the amount of bedrock that is visible in their slopes, and in the roughness that this bedrock has given to the peaks and ridges. As we shall see, the bedrock history of the Cuillin provides an explanation of the size and surface style of these remarkable hills. In the middle distance, various smaller landscape features, landforms, are visible, particularly some clearly developed terraces and cliffs. There are also smaller ridges, cross-cutting the terraces and cliffs, and often occupied by small stream valleys. These landforms directly reflect erosion of different features of the bedrock. The middle distance also illustrates the way that the gentler slopes tend to have a covering of surface blanket, often made of peat, soil or relatively weak and young sediment.






FIG 1. Aerial oblique photograph looking northwestwards towards the Cuillin mountains, Skye. (В© Patricia & Angus Macdonald/Aerographica)

PEOPLE

Much has been written about landscape history, and by many people the phrase tends to have been used for the way that mankind has modified landscapes. This approach is not the main focus of this book, which deliberately concentrates on natural landscape features. However, all the landscapes described contain roads and settlements. In some cases, such as in Figure 1 (#ulink_2ae4ccda-59c6-5bef-8d97-19db76bbc5b4), these are visible but have had little impact on the landscape as a whole. In other Scottish cases, landscapes have been changed profoundly by the building works of man, and the city and town landscapes of the Midland Valley are obvious examples (Fig. 2 (#ulink_56cd522e-0f1e-5f3a-80b2-d2e54f4556d3)). In other cases, subtle changes of landscape vegetation across Scotland may well be the result of man’s arrival and growing influence.






FIG 2. Satellite image covering the British Isles, showing population centres (including the Midland Valley of Scotland) picked out by man-made lighting at night. Note also the lighting on oil and gas platforms in the North Sea. (В© Planetary Visions Ltd/Science Photo Library)

AREAS AND MAPS

The systematic survey of Scotland is based on a division into a grid of 19 arbitrary Areas (Fig. 3 (#ulink_7749aab7-4c7a-5bea-88df-a6fb784b0ad5)). Each Area is based loosely on the pattern of double-page areas used in the larger road atlases available, in particular the Collins Road Atlas, Britain. The object is to provide total coverage of the land areas and islands of Scotland, allowing the reader to navigate easily from place to place. At the beginning of each Area chapter, a location map explains the relationship between the Area and its neighbours. Ordnance Survey (OS) National Grid references are provided for the edges of the Area, in kilometres east and north of the arbitrary National Grid origin some 80 km west of the Isles of Scilly, southwest England.






FIG 3. Division of Scotland into 19 Areas.

The sizes and shapes of the Areas have been adjusted to fit the shape of the land areas concerned: these Areas range from 50 to 100 km wide (from west to east) and 70 to 130 km high (from south to north), covering the shape and form of the mainland and islands of Scotland. On average each Area is about 100 Г— 100 km. All Areas are defined by National Grid south to north and west to east lines, and except for a few oblique view maps, all our maps use the same boundary orientation so that Grid North is parallel to the up-and-down margins.

Shaded, colour-coded maps are used to convey the height and approximate shape of the land surface in each Area. These maps have been produced from data collected by the space shuttle Endeavour in February 2000 as part of NASA’s Shuttle Radar Topographic Mission (SRTM). This 11-day mission used stereo pairs of radar images to build up a Digital Elevation Model (DEM) covering nearly 80 per cent of the Earth’s surface. The original dataset used in this book is publicly available via the SRTM website (www2.jpl.nasa.gov/srtm) and consists of pixels, approximately 90 × 90 m, each of which has an associated elevation value. The absolute vertical accuracy of these data is estimated to be ± 16 m, whilst the absolute horizontal accuracy is ± 20 m.

Data on roads, railways, coastlines, town boundaries, rivers etc., suitable for reproduction at a scale of 1 : 200,000, have been made available by the Collins Bartholomew mapping agency. For further detail it is recommended that the Ordnance Survey Landranger (1 : 50,000) maps are used.

We have used ESRI ARC Geographic Information System (GIS) software in the processing and manipulating of the map data. This software makes it possible to present maps with artificial hill-shading, so that topography becomes easier to visualise. Maps presenting the directions and slope angles of sloping features are also very useful in some situations.

Many of the maps make use of a standard colour scheme, ranging from greens for the lowest ground through yellows to browns and greys for the highest ground. In general, the full range of colours has been used for each map, no matter what numerical range of heights is involved. This makes it possible to convey the fine detail of slopes and other features, whether the map covers flat ground or valleys and high peaks. To make it possible to compare between maps using this colour sequence, we have quoted the maximum elevation reached in each Area.


CHAPTER 2

Surface Modifications

THE LANDSCAPE CYCLE

IN CHAPTER 1 (#u1db6a4ef-dcf8-5c72-b3d0-429532ccc75c), WE ILLUSTRATED our use of the word landscape to indicate an area a few kilometres to many kilometres across that is distinctive in appearance and origin. We have also found the word landform useful for smaller features of landscapes formed by distinctive surface processes during the modification of the landscape surface. In this chapter, we shall be examining further some aspects of certain of these landforms.

Our developing understanding of the larger workings of the Earth has shown us that although surface modifications are almost always apparent, the Earth’s crust and its surface have been subject to continual movements generated within the Earth. Earthquakes and volcanoes are obvious signs of these internal movements. Any landscape is the result of the interplay between these contrasting internal and external systems, as illustrated using the cycle diagrams (Figs 4 (#ulink_3ae460b5-a180-5838-b9db-bfc32bb32d7d), 5 (#ulink_a9946da3-7e6d-5b36-9a34-9ef28142c928)). These illustrate the two systems in usefully different ways.

DIGITAL MAPS, SLOPES AND DOWNSLOPE MOVEMENT

We explained in Chapter 1 (#u1db6a4ef-dcf8-5c72-b3d0-429532ccc75c) that our primary information about the shapes and patterns of Scottish landscapes comes from the use of the digital elevation datasets that are now available. Most people are familiar with the representation of elevation information on maps, using colour shading, or contours representing lines of specified elevation on the surfaces. The GIS software that we have used is a powerful tool for presenting topography in these ways. The same software makes it possible to represent topography using a hill-shade approach, which portrays topography using a shadowing effect, as estimated by an artificial light source with a specific orientation and elevation angle. The effects can appear similar to those produced by hachuring, as used in early Ordnance Survey maps, although hachure shading owed much to the eye of the individual draughtsman.






FIG 4. Diagram illustrating the processes of movement occurring within the outer layers of the Earth’s crust, and how these relate to the processes and features of the Earth’s surface and atmosphere.

As outlined in Chapter 1 (#u1db6a4ef-dcf8-5c72-b3d0-429532ccc75c), our maps of Scotland are based on digital elevation data where areas are divided into large numbers of small square unit areas (pixels), arranged in a rectangular grid. The elevation above sea level of each of the pixels is recorded in the database, and much of our data are based on a pixel size of 90 Г— 90 m. Although this resolution is adequate to provide information on larger landforms, we have to accept that many smaller landforms will be invisible if the pixel size is similar in area to, or larger than, the landform.

Digital elevation data can be directly represented on a map using colour shading or contours. It is also possible to define slopes by measuring changes of elevation within clusters of neighbouring pixels, allowing each pixel to be assigned a local slope value and converting the simple grid of elevation measurements into a grid of differences in elevation, or slopes. These maps are sometimes referred to as �first derivative’ maps of the topography, because they represent changes of topography (local slopes) rather than the elevations themselves. Whatever the limitations of scale, there is no question that examining patterns of slope variation is a powerful way of studying the shapes of landscapes, and the Area chapters that follow make frequent use of maps of this sort.






FIG 5. Landscapes are changed by surface modifications (Chapter 2 (#uc0e147ab-0db8-5cd3-a46c-c03cd0831740)) and solid Earth movements (Chapter 3 (#u5ec5bf09-5acf-58e3-8f03-7029db4715a5)).

We now consider the sorts of processes that are likely to give rise to various different features and patterns of slopes through time.

Slopes are likely to have a direct and profound influence on the way topography evolves through time, because any slope surface has the potential for downslope movement under gravity (Fig. 6 (#ulink_5bd8a2ae-a3f8-599c-96fa-9cceeed2fd2a)). Movement will often require triggering, for example by earthquakes, freeze–thaw ice changes or even heavy rainfall.

Even more important in determining the amount of movement and the angle of slope that can occur is the nature of the materials making the slope. Bedrock of igneous or metamorphic origin, consisting mainly of coarse crystals of interlocking minerals, formed at high temperatures in the Earth, such as quartz, feldspar and other silicate minerals, is likely to produce a strong material in terms of its surface weathering behaviour. At the other extreme, certain sedimentary rocks, consisting of small particles of clay minerals separated from their neighbours by films of water, will be weak and strongly liable to downslope mass flow, tending to produce a distinctly lower slope angle.

RIVER CATCHMENTS AND VALLEY PROCESSES

Looking back at the landscapes from Skye (Fig. 1 (#ulink_2ae4ccda-59c6-5bef-8d97-19db76bbc5b4)), and excluding coastline considerations for the moment, the first step in analysing landscape shape or morphology, as just discussed, is to realise that most of the detailed features visible can be considered as combinations of different scales and combinations of slopes. Our survey of Scotland has confirmed for us that, under present-day climate conditions, rivers and streams are the fundamental agents forming and changing valleys and slopes. This is why we have designed our computer-based maps to display clearly the locations and shapes of these landforms.






FIG 6. Models of downslope movement: (a) uniform material and movement; (b) non-uniform material and non-uniform movement (greatest from upper slope).

We can demonstrate the widespread importance of these river-generated landforms by plotting the pattern of large river catchments across present-day Scotland (Fig. 7 (#ulink_7473398e-e912-5841-b9c0-27bb75749122)). This map further supports the claim that these river and stream agents are easily the most important agents in the modification by erosion of Scotland and its landscapes. This map has been compiled using data from the Scottish National River Flow Archive collected by the Environment Protection Agency. The areas of the larger river catchments are superimposed on an elevation-shaded map of most of Scotland. In the interests of clarity, only the larger catchments are shown, leaving clear a zone of land, up to some 40 km wide, around the coast and including the islands, where the catchments are smaller.






FIG 7. The larger river catchments of Scotland, based on the Scottish National River Flow Archive.

One great contrast between western and eastern Scotland is shown very clearly by the rainfall data averaged over the year and over the catchments, varying from more than 2000 mm to less than 1000 mm per year. The other variable plotted in Figure 7 (#ulink_7473398e-e912-5841-b9c0-27bb75749122) is river flow rate (in m3/s), averaged over the year from the daily flow rates measured at the river gauge furthest downstream on each principal river. These flow rates give a first feel for the normal size of the river, but, of course, much of the work of rivers in eroding and transporting material is carried out during major floods, so our data give only limited grounds for comparison. Even with these data, it is interesting to see that the Tay (167.9 m3/s) is easily the largest river under most conditions, reflecting its large catchment area and its mid-range rainfall. It is interesting to note that the Thames and the Severn, in England and Wales, which both have large estuaries, have flow rates upstream from the tidal estuaries that are distinctly smaller (66 m3/s and 106 m3/s, respectively).

Some consideration can be given in our Area treatments to the shape and location of the different catchments, and to the bedrock materials and movements that may have been involved. Here the important general point that emerges is that the surface modification processes occurring in these river and valley catchments in Scotland bear a unique responsibility for the changes taking place now in the landscapes that we want to understand better.






FIG 8. Comparison of valley cross-sections: (a) a simple valley created by channel incision in balance with the valley slope evolution; (b) a valley created by glacial incision and now occupied by a stream channel of similar water discharge to the earlier glacier, but now incising a much smaller channel.

Valleys are bounded by boundary slopes that are inclined downwards towards the stream or river that flows along them. They can therefore be picked out on slope maps as pairs of areas of sloping pixels, providing clear evidence of the action of valley erosion, and the role of channel processes in the erosion of the catchment. Whereas downslope mass movement is a key component in removing material from the catchment, the incision and lowering of the river channel must also exert a control on the extent to which slope materials can be removed and transported down-channel and out of the catchment.

V-shaped cross-valley profiles in upland areas are commonly interpreted as the result of river or stream erosion by the valley channel. In contrast, U-shaped profiles are frequently interpreted as the products of glacial erosion (Fig. 8 (#ulink_2429fe1f-257d-54c2-b31a-a5bdf6804749)). It is generally accepted that most upland river valleys have a V-shaped profile, whereas upland valleys that have been occupied by glaciers tend to have U-shaped profiles. This can be readily understood in terms of the small eroding perimeter of a water-filled channel compared with the larger eroding perimeter of the much larger glacier channel, even though the discharge (in m3/s) of rapidly flowing water or slowly moving ice down the valleys may have been similar under the two climatic regimes.

Another climate-related factor should be mentioned at this point. Under periglacial (near-glacial) conditions, freeze–thaw processes, particularly within the surface blanket, will often keep the blanket materials in a state of frequent downslope movement. The degree to which this happens is likely to have a profound influence on the evolution of valley-slope and channel systems.






FIG 9. Dendritic channel pattern growth. This shows three phases in the development of a computer-based erosional model, in which valley erosion from the southern edge of a study area proceeds by headward erosion of each valley which is randomly free to choose its erosion direction. The map model was provided by Dimitri Lague and was based on the work of A. Crave and P. Davy.






FIG 10. Other landforms typical of river catchments, including channels of different geometries, flood plains and river terraces.

The catchment map (Fig. 7 (#ulink_7473398e-e912-5841-b9c0-27bb75749122)) shows clearly the dendritic patterns developed by all river and stream drainage systems, and it is these patterns (Fig. 9 (#ulink_b57ff4e1-a88e-52e9-b34b-4acd1e7f18df)) that make the recognition of valleys so easy on our slope maps.

In the lower reaches of channel systems, other landforms develop that are typical of deposition of sediment being carried down the system (Fig. 10 (#ulink_79fc7768-e25e-5c04-bc30-50279381b393)). This may influence the sinuosity of the channels, so that only slightly sinuous channels, often carrying rather coarse gravel or sandy sediment, may be replaced downstream by meandering (highly sinuous) channels that flow between banks of muddy sediment. Flood plains of relatively fine-grained sediment are increasingly the obvious landforms low down in river systems, and these are typically very flat extensive plains. Terraces often represent fragments of former flood plains left at a higher level as the active channel has cut down to a lower elevation.

SEA COASTS

A special slope consideration is raised by the landscapes of Figure 1 (#ulink_2ae4ccda-59c6-5bef-8d97-19db76bbc5b4), and frequently in other parts of Scotland also. This occurs where a change of sea level has initiated the formation of new coastal landforms, sometimes largely erosional and sometimes depositional (Fig. 11 (#ulink_4c1873dd-1f55-5a39-b446-b5b4be0b2f34)).

Erosion of a coastal slope has similarities to the erosion of river or stream valley slopes, except that coastal slopes are caused by mass downslope movement towards sites of sea-cliff erosion. A very different situation occurs where sea-level changes initiate the deposition of sediment to form coastal flats, with very flat overall slopes, and patterns of coastal sediment bars, sediment flats and salt-marshes. Sediment supply is obviously a key factor in the growth of potentially extensive landforms and landscapes, often with distinctive landforms and very flat surfaces or zero-angle overall slopes.






FIG 11. Coastal slopes: erosion and deposition; a) �vertical profile’ and b) �plan view’.


CHAPTER 3

Movements of the Earth from Within

EARTH-SURFACE MOVEMENTS DUE TO PLATE TECTONICS

TO UNDERSTAND THE CHANGES AND MOVEMENTS affecting the appearance of the landscape on large scales we need to review current understanding of some geological systems, especially plate tectonics. Many of the widespread changes that have created landscapes over long periods of time can now be understood using this discovery.

Knowledge of the processes causing the movement of large areas of the Earth’s surface (10–1000 km length scale) has been revolutionised by scientific advances made over the last 50 years. During this time, scientists have become convinced that the whole of the Earth’s surface consists of an outer shell of interlocking tectonic plates (Fig. 12 (#ulink_a84d1eeb-47a4-570b-916e-5e05b2035a6c)). The word tectonic refers to processes that have built features of the Earth’s crust (Greek: tekt, a builder). The worldwide plate pattern is confusingly irregular – particularly when seen on a flat map – and it is easier to visualise the plates in terms of an interlocking arrangement of panels on the Earth’s spherical surface, broadly like the panels forming the skin of a traditional leather football.

Tectonic plates are features of the lithosphere, the name given to the ~125 km- thick outer shell of the Earth, distinguished from the material below by the strength of its materials (Greek: lithos, stone). The strength depends upon the composition of the material and also upon its temperature and pressure, both of which tend to increase with depth below the Earth’s surface. In contrast to the mechanically strong lithosphere, the underlying material is weaker and known as the asthenosphere (Greek: asthenos, no-strength). Note that in Figure 13 (#ulink_54110bfc-f411-5583-80ce-e856a9f091fe) the crustal and outer mantle layers are shown with exaggerated thickness, so that they are visible.






FIG 12. World map showing the present pattern of the largest lithosphere plates.






FIG 13. Diagram of the internal structure of the Earth.

Much of the strength difference between the lithosphere and the asthenosphere depends on the temperature difference between them. The lithosphere plates are cooler than the underlying material, so they behave in a more rigid way when forces are generated within the Earth. The asthenosphere is hotter and behaves in a more plastic way, capable of deforming without fracturing and, to some extent, of �flowing’. Because of this difference in mechanical properties and the complex internal forces present, the lithosphere plates can move relative to the material below. To visualise the motion of the plates, we can use the idea of lithospheric plates floating on top of the asthenosphere.

The pattern of earthquake activity and actively unstable mountain belts corresponds very well with the pattern of the tectonic plates now recognised. The largest plates (Fig. 12 (#ulink_a84d1eeb-47a4-570b-916e-5e05b2035a6c)) clearly mark relatively rigid and stable areas of the lithosphere, with interiors that do not experience as much disturbance as their edges. Plates move relative to each other along plate boundaries, in various ways that will be described below. The plate patterns have been located by investigating distinctive markers within the plates and at their edges, allowing the relative rates of movement between neighbouring plates to be calculated. These rates are very slow, rarely exceeding a few centimetres per year, but over the millions of years of geological time they can account for thousands of kilometres of relative movement.

It has proved much easier to measure plate movements than to work out what has been causing them. However, the general belief today is that the plates move in response to a number of different forces. Circulation (convection) within the mantle is driven by temperature and density differences, but other forces are also at play. Where plates diverge, warm material rises from within the Earth to fill the surface gap, and, being warmer, it may also be elevated above the rest of the plate, providing a pushing force to move the plate across the surface of the Earth. At convergent boundaries, cold, older material sinks into the asthenosphere, providing a pulling force that drags the rest of the plate along behind it. Deep within the Earth, the sinking material melts and is ultimately recycled and brought back to the surface to continue the process.

Knowledge of how tectonic plates interact provides the key to understanding the movement history of the Earth’s crust. However, most people are much more familiar with the geographical patterns of land and sea, which do not coincide with the distribution of tectonic plates (Fig. 12 (#ulink_a84d1eeb-47a4-570b-916e-5e05b2035a6c)). From the point of view of landscapes and scenery, coastlines are always going to be key features because they define the limits of the land; we make no attempt in this book to consider submarine scenery in detail.

The upper part of the lithosphere is called the crust (Fig. 13 (#ulink_54110bfc-f411-5583-80ce-e856a9f091fe)). Whereas the distinction between the lithosphere and the asthenosphere is based upon mechanical properties related to temperature and pressure, the distinction between the crust and the lower part of the lithosphere is based upon composition. Broadly speaking, there are two types of crust that can form the upper part of the lithosphere: continental and oceanic. An individual tectonic plate may include just one or both kinds of crust.

Continental crust underlies land areas and also many of the areas covered by shallow seas. Geophysical work shows that this crust is typically about 30 km thick, but may be 80–90 km thick below some high plateaus and mountain ranges. The highest mountains in Britain are barely noticeable on a scale diagram comparing crustal thicknesses (Fig. 14 (#ulink_dbcbd410-6d7d-59fa-9afd-503fa3f26c7b)). Continental crust is made of rather less dense materials than the oceanic crust, or the mantle, and this lightness is the reason why land surfaces and shallow sea floors are elevated compared to the deep oceans. Much of the continental crust is very old (up to 3–4 billion years), having formed early in the Earth’s life when lighter material separated from denser materials within the Earth and rose to the surface.

Oceanic crust forms the floors of the deep oceans, typically 4 or 5 km below sea level. It is generally 5–10 km thick and is distinctly more dense than continental crust. Oceanic crust only forms land where volcanic material has been supplied to it in great quantity (as in the case of Iceland), or where other important local forces in the crust have caused it to rise (as is the case in parts of Cyprus). Oceanic crust is generally relatively young (only 0–200 million years old), because its greater density and lower elevation ensures that it is generally subducted and destroyed at plate boundaries that are convergent.






FIG 14. Scale diagram comparing average thicknesses of oceanic and continental crust and lithosphere.

Figure 12 (#ulink_a84d1eeb-47a4-570b-916e-5e05b2035a6c) shows the major pattern of tectonic plates on the Earth today. The Mercator projection of this map distorts shapes, particularly in polar regions, but we can see that there are seven very large plates, identified by the main areas located on their surfaces. The Pacific plate lacks continental crust entirely, whereas the other six main plates each contain a large continent (Eurasia, North America, Australia, South America, Africa and Antarctica) as well as oceanic crust. There are a number of other middle-sized plates (e.g. Arabia and India) and large numbers of micro-plates, not shown on the world map.

Figures 12 (#ulink_a84d1eeb-47a4-570b-916e-5e05b2035a6c) and 15 (#ulink_d6668883-b8ab-5bd7-bb06-bd220b07a25f) also identify the different types of plate boundary, which are distinguished according to the relative motion between the two plates. Convergent plate boundaries involve movement of the plates from each side towards the suture (or central zone) of the boundary. Because the plates are moving towards each other, they become squashed together in the boundary zone. Sometimes one plate moves below the other in a process called subduction, which often results in a deep ocean trench and a zone of mountains and/or volcanoes, as well as earthquake activity (Fig. 15 (#ulink_d6668883-b8ab-5bd7-bb06-bd220b07a25f)). The earthquakes that happened off Indonesia in December 2004 and off Japan in March 2011 were two of the strongest known since records began. Both seized world attention because of the horrifying loss of life cause by the tsunami waves they generated. Both were the result of sudden lithosphere movements of several metres on faults in the convergent subduction zones where the Australian and Pacific plates have been moving under the Eurasian plate (Fig. 12 (#ulink_a84d1eeb-47a4-570b-916e-5e05b2035a6c)).






FIG 15. Diagram (not to scale) illustrating the movement processes of plates.

In other cases the plate boundary is divergent, where the neighbouring plates move apart and new material from deeper within the Earth rises to fill the space created. New oceanic crust is created by the arrival and cooling of hot volcanic material from below. The Mid-Atlantic Ridge running through Iceland is one of the examples nearest to Britain of this sort of plate boundary, and volcanic ash-cloud activity there caused widespread disruption to air transport during 2010.

Other plate boundaries, sometimes called transform boundaries, mainly involve movement parallel to the plate edges. The Californian coast zone is the classic example but there are many others, such as the transform boundary between the African and Antarctic plates. In some areas, plate movement is at an oblique angle to the suture and there are components of divergence or convergence as well as movement parallel to the boundary.

Britain today sits in the stable interior of the western Eurasian plate, almost equidistant from the divergent Mid-Atlantic Ridge boundary to the west and the complex convergent boundary to the south where Spain and northwest Africa are colliding. In its earlier history the crust of Britain has been subjected to very direct plate boundary activity. The results of convergent activity in Devonian and Carboniferous times (between 416 and 299 million years ago) are visible at the surface in southwest England, and in Ordovician to Devonian times (between 490 and 360 million years ago) in Wales, northwest England and Scotland (see Chapter 4).

Present-day plate boundaries are often picked out by the location of earthquakes, as described above. Mention should also be made at this point of the importance of volcanoes and igneous bedrock in providing information about movements within the upper levels of the Earth. Highly sophisticated analytical work has illuminated the whole subject of the chemical and mineral evolution of igneous material as it evolves and moves in the crust. For the purposes of this book, a very simple twofold division of igneous rocks into felsic and mafic will be sufficient.

Felsic igneous rocks tend to be light-coloured and of relatively low density, containing the minerals quartz and feldspar. Typical types are granite, syenite (coarsely crystalline) and rhyolite (finely crystalline). Continental crust consists of felsic and mafic igneous rocks, as well as sedimentary and metamorphic rocks.

Mafic igneous rocks tend to be darker-coloured and of relatively high density, containing feldspar and dark minerals rich in magnesium and/or iron, such as augite or hornblende. Typical types are gabbros (coarsely crystalline), andesites and basalts (finely crystalline). Oceanic crust is dominated by mafic igneous materials.

MAKING LOCAL MEASUREMENTS OF EARTH SURFACE MOVEMENTS

We have been considering the large movement systems that originate within the Earth. There are also more local movement systems operating on the Earth’s surface, which are linked to a very variable degree to the large-scale movements of plate tectonics. To explore this complex linkage further, it will be helpful to look now at different processes that may combine to cause particular local movements.

Tectonic plates are defined by their rigidity, so there is relatively little horizontal movement between points within the same plate, compared to the deformation seen in plate boundary zones. This extreme deformation may involve folding and fracturing of the rock materials, addition of new material from below, or absorption of material into the interior during subduction.

Nonetheless, deformation is not restricted solely to plate boundaries and does occur within the plates, although to a lesser extent. In some cases, major structures that originally formed along a plate boundary can become incorporated into the interior of a plate when prolonged collision causes two plates to join. The Caledonian convergent boundary that extended across Scotland (see Chapter 4) provides an excellent example of movements that occurred hundreds of millions of years ago, but also contains many examples of structures formed in later movements. These structures have often been reactivated long after they first formed in order to accommodate forces along the new plate boundary via deformation within the plate. Conversely, changes of internal stress patterns can sometimes lead to the splitting of a plate into two, forming a new, initially divergent plate boundary. Many of the oil- and gas-containing features of the North Sea floor (Fig. 2 (#ulink_56cd522e-0f1e-5f3a-80b2-d2e54f4556d3)) originated when a belt of divergent rift faults formed across a previously intact plate.

It needs to be stressed that the patterns of deformation (fracturing and folding) due to these plate motions occur at a wide range of different scales, from centimetres to thousands of kilometres. Sometimes they are visible at the scale of an entire plate boundary, such as the enormous Himalayan mountain chain that marks the collision of India with Asia.

The effects of features as large as plate boundaries on landscapes persist over hundreds of millions of years, long after the most active movement has ceased. For example, parts of southwestern England, Wales and the Scottish Highlands are underlain by bedrocks that were formed in convergent boundary zones of the past. The tin and lead mines of Cornwall owe their existence to a 300-million-year-old convergent plate boundary, where an ocean was destroyed as two plates converged and continents collided. The convergence released molten rock that rose in the crust and gradually cooled to form granite, whilst metals were precipitated in the surrounding crust as �lodes’ containing tin and lead (see Chapter 4).

Mapping the patterns of bedrock exposed at the surface often reveals folds and faults that provide key information about the movements that have taken place during the past (Fig. 16 (#ulink_9bcdd054-5b60-5190-8c20-a8e932b61672)). Figure 17 (#ulink_aee633f4-0b1c-53bf-ad26-bfada4da029c) provides a key to some of the terms commonly used to classify these structures, as a step towards understanding the sorts of movement patterns that they represent. In broad terms, folds tend to indicate some form of local convergent movement, though they may be the result of larger movement patterns of a different kind. Normal faults tend to indicate divergent or stretching movements, at least locally, whereas reverse and strike-slip faults tend to indicate convergence. Two broad types of fold are distinguished: synclines are U-shaped downfolds, while anticlines are the opposite – A-shaped upfolds.

Further mapping of folds and faults often reveals complex patterns of changing movements. A complex example is shown in Figure 18 (#ulink_d852fad3-6ec7-5f0a-b742-2a8ea78db4b1). Divergent movements in an area of crust produced plastic deformation in the warmer lower crust, and faulting into a number of discrete blocks in the colder, more brittle upper crust. This was then followed by an episode of convergent movement that resulted in closing up the upper crustal blocks and further flow in the plastic lower crust, causing crustal thickening and mountain building at the surface.






FIG 16. Outcrop in the Atacama Desert, Chile, showing a very regularly bedded succession of mudstones, formed originally as horizontally layered deposits in a lake. Since their deposition the mudstones have been tilted. They have also been fractured during an earthquake, resulting in a step, or normal fault (see Fig. 17 (#ulink_aee633f4-0b1c-53bf-ad26-bfada4da029c)), that is particularly clear because it has cut through a white layer in the deposits. An outcrop such as this makes it possible to measure the local movements that have taken place in this material after it was deposited. (В© Nicholas Branson)

VERTICAL CRUSTAL MOVEMENTS

The movement of lithospheric plates, as described above, is the main cause of horizontal convergent and divergent movements affecting thousands of kilometres of the Earth’s surface. As shown in Figures 16 (#ulink_9bcdd054-5b60-5190-8c20-a8e932b61672) to 18 (#ulink_d852fad3-6ec7-5f0a-b742-2a8ea78db4b1), horizontal movements are generally accompanied by vertical movements of local crustal surfaces. Some of these could have produced very large scenic features, such as a mountain belt or a rift valley. In this book we are primarily concerned with scenic features at a more local scale, so we now consider various other processes that may contribute to the creation of vertical crustal movements.






FIG 17. The most important types of folds and faults, and the local patterns of forces responsible.

Vertical crustal movement linked to erosion or deposition

Addition or subtraction of material to the surface of the Earth is happening all the time as sediment is deposited or solid material is eroded. The discipline of sedimentology is concerned with the wide range of different processes that are involved in the erosion, transport and deposition of material, whether the primary agent of movement is water, ice, mud or wind. An important point is that few of these sedimentary processes relate directly to the large tectonic movements of the Earth’s crust that we have discussed above. Landscape is often produced by erosion of thick sedimentary deposits that formed in sedimentary basins where material eroded from the surrounding uplands accumulated. One of the characteristic features of these thick deposits is their layered appearance – as, for example, in the Torridonian Sandstones of northwestern Scotland (see Chapter 4). Layering varies from millimetre-scale laminations produced by very small fluctuations in depositional processes, to sheets hundreds of metres thick that extend across an entire sedimentary basin. These thicker sheets are often so distinctive that they are named and mapped as separate geological units representing significant changes in the local environment at the time they were deposited.






FIG 18. Example of a cross-section through the crust, showing how a divergent movement pattern (A) may be modified by later convergent movements (B and C).

Vertical crustal movements due to loading or unloading

In addition to the direct raising or lowering of the surface by erosion or deposition, there is a secondary effect due to the unloading or loading of the crust that may take some thousands of years to produce significant effects. As mentioned above, we can visualise the lithosphere as �floating’ on the asthenosphere like a boat floating in water. Loading or unloading the surface of the Earth by deposition or erosion will therefore lower or raise the scenery, just as a boat will sit lower or higher in the water depending on its load.

An example of such loading has been the build-up of ice sheets during the Ice Age. The weight of these build-ups depressed the Earth’s surface in the areas involved, and when the ice melted the Earth’s surface rose again. Western Scotland provides an example of an area that has been rising because of �rebound’ since the ice of the Ice Age disappeared on melting.

A second example of this is the lowering of the area around the Mississippi Delta, loaded by sediment eroded from the more central and northern parts of North America. The Delta region, including New Orleans, is doomed to sink continually as the Mississippi River deposits sediment around its mouth, increasing the crustal load there.

Conversely, unloading of the Earth’s surface will cause it to rise. Recent theoretical work on the River Severn suggests that unloading of the crust by erosion may have played a role in raising the Cotswold Hills to the east and an equivalent range of hills in the Welsh Borders.

Vertical movements due to thermal expansion or contraction

Changes in the temperature of the crust and lithosphere are an inevitable result of many of the processes active within the Earth, because they often involve the transfer of heat. In particular, rising plumes of hot material in the Earth’s mantle, often independent of the plate boundaries, are now widely recognised as an explanation for various areas of intense volcanic activity (for example beneath Iceland today). These plumes are often referred to as �hot spots’ (Fig. 15 (#ulink_d6668883-b8ab-5bd7-bb06-bd220b07a25f)). Heating and cooling leads to expansion or contraction of the lithosphere and can cause the surface to rise or sink, at least locally.

An example of this is the way that Britain was tilted downwards to the east about 60 million years ago. At about this time, eastern North America moved away from western Europe as the North American and Eurasian plates diverged. The divergence resulted in large volumes of hot material from deep within the Earth being brought to the surface and added to the crust of western Britain. It is believed that the heating and expansion of the crustal rocks in the west has elevated them above the rocks to the east, giving an eastward tilt to the rock layers and exposing the oldest rocks in the west and the youngest ones in the east.

THE CHALLENGE OF MEASURING CRUSTAL MOVEMENTS

Having just reviewed some of the processes that may cause movements of the Earth’s surface, it is useful to consider the practical difficulties of how such movements are measured.

For present-day applications, it seems natural to regard sea level as a datum against which vertical landscape movements can be measured, as long as we remember to allow for tidal and storm variations. However, much work has demonstrated that global sea level has changed rapidly and frequently through time, due to climate fluctuations affecting the size of the polar icecaps and changing the total amount of liquid water present in the oceans and seas (see Chapter 5). It has also been shown that plate tectonic movements can have an important effect on global sea level by changing the size and shape of ocean basins.

Attempts have been made to develop charts showing how sea level, generalised for the whole world, has varied through time. However, it has proved very difficult to distinguish a worldwide signal from local variations, and the dating of the changes is often too uncertain to allow confident correlation between areas.

In sedimentary basins, estimates of vertical movements have been made using the thicknesses of sediment layers accumulating over different time intervals in different depths of water. In areas of mountain building, amounts of vertical uplift have been estimated using certain indicator minerals that show the rates of cooling that rocks have experienced as they were brought up to the surface. However, both these approaches are only really possible in areas that have been subjected to movements of the Earth’s crust that are large and continuous enough to dominate completely other possible sources of error.

Local movements are also difficult to estimate, although fold and/or fault patterns may allow a simple measure in some cases. Over short present-day periods of time it has proved to be possible to detect vertical movement patterns using satellite imagery. Movement of sediment across the Earth’s surface by rivers or sea currents can be estimated if mineral grains in the sediment can be tracked back to the areas from which they have come. In the detailed consideration of landscapes in this book, we have to rely on using the widest possible range of types of evidence, carefully distinguishing the times and scales involved. Even then, we are often left with probable movement suggestions rather than certainties.


CHAPTER 4

Episodes in the Bedrock History of Scotland

CHAPTERS 2 (#uc0e147ab-0db8-5cd3-a46c-c03cd0831740) AND 3 (#u5ec5bf09-5acf-58e3-8f03-7029db4715a5) HAVE INTRODUCED the idea that natural landscapes are the results of combinations of surface modification (Chapter 2 (#uc0e147ab-0db8-5cd3-a46c-c03cd0831740)) and internal movements (Chapter 3 (#u5ec5bf09-5acf-58e3-8f03-7029db4715a5)). The responses of the Earth’s surface to these combinations have depended on the bedrock present locally on the surface at each stage. It is now time, therefore, to turn specifically to Scotland, to summarise the distribution and history of its bedrock.

The mapping of the bedrock of Scotland has been a heroic task that started over 200 years ago. Most of the systematic work has been carried out by the British Geological Survey, and is now available on different scales, forming a monument to the efforts of many remarkable people and the Survey itself (Fig. 19 (#ulink_14158c9a-c802-5caf-a8ea-fcd8d794c594)). For the generalising approach of this book much of this work has had to be simplified.

Scotland’s geological history is unusually long and varied for a country of its size. One reason for this is that present-day Scotland is the result of the convergence or movement together of at least five different areas of crust, often referred to as terranes (Fig. 20 (#ulink_92e7c4dd-e904-5c91-a5eb-a77514394804)). These terranes are fragments of continental crust that have been carried together by plate tectonic movements that resulted eventually in the construction of crustal Scotland, as we find it now.

Although most of the surface modifications and internal movements have overlapped in time and space, it helps to pick out discrete episodes in summarising aspects of Scotland’s history. The first nine of these episodes are represented in the bedrock record and are outlined in the rest of this chapter. Episodes 10–12 are mainly represented in the record of recent surface modifications, and they are described in Chapter 5 (#u661596f2-c933-517a-abc9-8c82715c3114). All 12 episodes have been placed in chronological order in Figure 21 (#ulink_5ef00752-9d7a-5a35-83e6-cb77e10a1de8) (where Episode 1 is the oldest and 12 the most recent) using the International Stratigraphy Chart 2009, which provides an accepted standard for the names used in dividing and describing geological time (see www.stratigraphy.org).






FIG 19. Simplified geological map of Scotland. (Redrawn from British Geological Survey 1 : 625,000 map)

We provide a �Timeline’ as part of the description of the geology of each of our Areas. These timelines are designed to summarise the time sequence of events that is represented in or near each Area, using the standard International Stratigraphic divisions. Standard colours are used for the divisions and ages. If part of the stratigraphic record is absent, the division is not coloured.

The bedrock episodes can be grouped as follows:

(1) Pre-Caledonian Greenland-margin episodes (Episodes 1–3)

(2) Caledonian mountain-building episodes (Episodes 4–6)

(3) Post-Caledonian episodes (Episodes 7–9)

The distribution of these groups of rocks is shown in Figure 22 (#ulink_974e4b5e-a91a-5d0c-be9f-f997dba62da2), and the episodes involved in their formation are described below.

PRE-CALEDONIAN GREENLAND-MARGIN EPISODES

Episode 1: formation of the Lewisian Complex

Rocks of the Lewisian Complex are very largely restricted to the Hebridean terrane, where they make up almost all of the bedrock of the Outer Hebrides, and much of the bedrock of the mainland. They also occur occasionally in the neighbouring part of the Northern Highland terrane, where they became involved in the much younger Caledonian movement history. The Lewisian Complex takes its name from the largest and northernmost of the islands of the Outer Hebrides.

The Lewisian Complex consists of metamorphic rocks (typically coarsely crystalline gneisses) that formed by alteration of earlier rocks when high temperatures and/or pressures peaked during movements at deep levels within the Earth’s crust. The great interest of these metamorphic rocks is that they can provide information about the conditions deep within the crust when these movements were taking place. Unlike igneous rocks that formed by crystallisation from completely melted rock material, metamorphic rocks have involved changes in rocks that were at least partly solid, so they preserve information about features present before, as well as conditions during, the metamorphism. In most cases, the minerals now present are stable at present-day surface temperatures and pressures, are large in crystal size, and interlock with neighbouring crystals, so the rocks are resistant to surface weathering compared with many other rock types.






FIG 20. The five terranes of Scotland. (After Trewin 2002)






FIG 21. Episodes in Scotland’s geological history. The age scale is not linear and has been deliberately chosen so that younger episodes are given greater space than older ones, because they are usually known in greater detail. The chart indicates the ages covered by the 12 episodes, and the dominant processes represented by them. (Redrawn from International Stratigraphy Chart 2009, www.stratigraphy.org)

One of the most important research tools applied to the Lewisian rocks has been the dating of the various mineral components, using the fact that some of them contain radioactive materials that have been steadily changing since they were first trapped when the minerals formed. The amount of change gives a measure of the time over which it has been taking place. New analytical methods have led to increasingly accurate and reliable figures. As this work has continued, it has become clear that the Lewisian is truly a �complex’, made up of many distinct volumes of crust, each preserving certain episodes of movement and rock alteration. Many of the folds or fractures mapped in the Lewisian have a northwest/southeast trend, almost at right angles to the Moine Thrust Zone and the associated folds and fractures that form the margin of the Hebridean terrane. However, mapping of these structures has shown that the movement and alteration of the Lewisian occurred in a number of phases with different compression and shearing directions. The evidence is too fragmentary to allow identification of the boundaries of tectonic plates similar to those that can be identified in younger bedrock areas. This is hardly surprising, because these are some of the earliest movement events recognised anywhere on the surface of the whole Earth, representing glimpses of early crustal activity that has escaped reworking or obliteration in more recent episodes.

Important phases of activity and mineral alteration have been recognised in the Lewisian Complex, some in the Archaean (3.2 and 2.8 billion years ago), generally named Scourian and Inverian. Other rocks were formed and/or altered in the Proterozoic (2.4, 1.7 and 1.1 billion years ago) and are named Laxfordian. The Archaean phases are older than any other for which there is evidence in Britain. Most of the rocks altered in these phases were originally igneous but some were sedimentary, and all had actually been formed as rocks even earlier. It is clear from the minerals present that some phases involved crust being moved downwards to considerable depths – several tens of kilometres below the surface – although before the next episode (described below) the rocks had been moved back upwards and were exposed at the surface.

Surface modification of the Lewisian during the Tertiary and the Ice Age has carved it into typical �knock-and-lochan’ topography, in which the land surface consists of hillocks of exposed rock tens to hundreds of metres across (called knocks, from the Gaelic cnoc, a small, rocky hill), separated by water and bog-filled hollows (lochans) which often pick out folds and linear fractures in the bedrock (Fig. 23 (#ulink_e58116da-0136-51ae-aa28-cb36de3d86d0)). This wild knock-and-lochan landscape was once thought to represent the first formed surface of the Earth, but it is now realised that the surface shapes of the landscape are very much younger, and that the metamorphic alterations and movements, although very old, were preceded by even earlier episodes.






FIG 22. The distribution of the Pre-, Syn- and Post-Caledonian rocks in Scotland.






FIG 23. Aerial oblique view of Suilven (731 m), carved from Neoproterozoic Torridonian Sandstones resting unconformably on the knock-and-lochan topography of the Lewisian Complex. (В© Adrian Warren/lastrefuge.co.uk)

Episode 2: formation of the Torridonian Sandstones

Mountains and slopes made of uniform but well-layered successions of Torridonian Sandstones, often tens to hundreds of metres thick, provide some highly characteristic elements of the Hebridean terrane (Fig. 23 (#ulink_e58116da-0136-51ae-aa28-cb36de3d86d0)). The layers are generally rather flat-lying, except in the folded and faulted bedrock of the Moine Thrust Zone where the Torridonian has been deformed along the eastern terrane boundary. Careful examination of the layering shows that it generally reflects episodes in the life of Torridonian rivers, which occasionally flooded and deposited sandy or gravelly river bars, often with muddy tops that have weathered to pick out the layers . Although much of the Torridonian was deposited by rivers, some of it accumulated in lakes or in the sea.

At least two different episodes of deposition are represented in the Torridonian Sandstones, and these have been dated, using radiometric methods, as Mesoproterozoic and Neoproterozoic (1.2 and 1.0 – 0.95 billion years ago respectively: Fig. 21 (#ulink_5ef00752-9d7a-5a35-83e6-cb77e10a1de8)). These two episodes left thick successions of sediment in the crustal record: up to 2 km in the Mesoproterozoic and up to 5 km in the Neoproterozoic. Thicknesses as great as these imply the downward movement of certain areas (basins) of the Earth’s surface close to areas where erosion due to upward movement was producing large quantities of sediment. In other words, although the Torridonian Sandstones are the result of modification by surface processes, these processes must have been linked to important vertical movements of the crust, due to forces acting within the Earth.






FIG 24. Cross-section through the Ord window of Southern Skye, showing the folded and fractured structure of the bedrock below the Moine Thrust.

The contact between the flat-lying layering of the Torridonian and the underlying Lewisian Complex is an unconformity that formed when the Torridonian sediments were deposited onto topography of valleys and hills (often hundreds of metres in relief) that had been carved in the Lewisian Complex. This unconformity, although preserving local Proterozoic hills and valleys, is relatively flat-lying overall, showing that widespread vertical movements – rather than significant folding or tilting – must have been involved.

The western and eastern margins of the present-day Atlantic Ocean were close to one another within a �super’-continent when the Torridonian sediments were accumulating. It seems likely that much of the sediment was derived from upland areas whose crust is now in Greenland or eastern Canada.

Episode 3: Cambrian and Ordovician sedimentation

A rather uniform succession of sediments of Cambrian and Ordovician age occurs more or less continuously along a strip from Skye, in the south, to the north coast of the Scottish mainland (Figs 19 (#ulink_14158c9a-c802-5caf-a8ea-fcd8d794c594), 20 (#ulink_92e7c4dd-e904-5c91-a5eb-a77514394804), 22 (#ulink_974e4b5e-a91a-5d0c-be9f-f997dba62da2): our �Greenland edge’). This succession was deposited unconformably on the eroded rocks of the Lewisian Complex and the Torridonian Sandstones. These Cambrian and Ordovician sediments only occur in the Hebridean terrane, and their usually gentle tilt is evidence of the lack of later movements, although they have been folded and fractured in the Moine Thrust Zone along the terrane’s edge (Fig. 24 (#ulink_19f49dad-6bb2-521a-9811-4a6da08ecd8c)).

Where most fully developed, this sedimentary succession is about 1 km in thickness and consists of a lower unit of quartzites (up to 100 m thick) that forms a greyish-white cap on some mountains and weathers to produce distinctive angular scree and boulder fields. Above this is a thin unit of mudstones, fractured and folded by subsequent movements on the Moine Thrust, overlain by a thick succession of limestones which, according to the fossil evidence contained within them, span the time interval from early Cambrian to early Ordovician (Fig. 21 (#ulink_5ef00752-9d7a-5a35-83e6-cb77e10a1de8)). In the present-day landscape, these limestones often produce swallow holes and caves formed by solution of the limestone’s fracture joints, as well as unusually lush grass wherever there is significant soil development.

The whole of this succession appears to have formed on the edge of a sea, and the occurrence of similar sediments on both sides of the present-day North Atlantic, a much more recent feature of the Earth’s surface, suggests a widespread uniformity in the coastal environments at this time. It is believed that these deposits formed due to surface modification involving global sea-level change, with no clear evidence of local movements due to processes deeper within the Earth.

CALEDONIAN MOUNTAIN-BUILDING EPISODES

The Latin adjective Caledonian is widely used to indicate Scottish-ness, and is used in geology for the important phase of mountain building that dominated earth movements and surface modification in Scotland between Ordovician and Devonian times. Evidence of similar movements and modifications during the same time periods is found along the east coast of the USA, Canada and Greenland, and through Ireland, Norway, Sweden and Spitsbergen. The terranes now recognised in Scotland have been mentioned above and shown in Figure 20 (#ulink_92e7c4dd-e904-5c91-a5eb-a77514394804). Distinct areas of continental crust, some thousands of kilometres across, others much smaller, rode on plates (Chapter 3 (#u5ec5bf09-5acf-58e3-8f03-7029db4715a5)) that moved independently and came together at different stages over Ordovician, Silurian and Devonian times to create the final assemblage of crustal fragments now present in Scotland. The main crustal fragments and their plates and intervening oceans are tracked in summary in Figure 25 (#ulink_b6cf889e-f4b9-5df3-9253-ea72e9fe5903).

Episode 4: making the core of the Caledonian mountains

The core of the Caledonian mountain belt is represented by the metamorphic bedrock that forms most of the Northern Highland and Grampian Highland terranes (Fig. 20 (#ulink_92e7c4dd-e904-5c91-a5eb-a77514394804)). The metamorphism of the originally largely sedimentary rocks occurred under the high pressures and temperatures that reflect their deep burial when compressive movements caused thickening of the crust and mountain uplift at the surface.

For many years a distinction has been drawn between the Moine and Dalradian supergroups in the mapping of the metamorphic core of the Caledonian mountain belt (Figs 19 (#ulink_14158c9a-c802-5caf-a8ea-fcd8d794c594) and 20 (#ulink_92e7c4dd-e904-5c91-a5eb-a77514394804)). The Moine Supergroup was named after a stretch of moorland on the north coast. It forms most of the Northern Highland terrane and may be present also in part of the Grampian Highland Terrane (Fig. 19 (#ulink_14158c9a-c802-5caf-a8ea-fcd8d794c594)). In contrast, the Dalradian Supergroup contains a greater variety of metamorphic rock types that have made it possible to trace distinctive subdivisions across most of the rest of the Grampian Highland terrane and even into Shetland. The name Dalradian has many historic roots and, in a geological sense it simply indicates association with the Scottish Highlands and parts of Ireland. There is general agreement that the original (pre-metamorphism) sediments of the Moine are older than those of the Dalradian, but the mapping of any boundary between them is still very arbitrary, and is not important in our review of landscapes across Scotland.






FIG 25. Diagram showing major plate-scale ocean closings and openings, with compressive events on the plate margins that generated events during the Caledonian and later Variscan mountain building. Ma (mega-annum) = million years ago.

The dominant bedrock of both these supergroups is metamorphic. In other words, the bedrock has been altered but not melted, during the growth of new minerals under the high temperatures and/or pressures generated by compressive movements and thickening of the crust. The original rocks of the Moine and Dalradian were mostly formed as sediments, mainly muds and sands but also occasionally lime-rich sediments. These sediments have now been transformed into schists (also called pelites; originally mudstones) and psammites (originally sandstones).

Knowledge of the age of the original rocks and the age of their alteration depends on sophisticated analysis of the decay of radioactive mineral components. The Moine Supergroup appears to have been deposited in the Neoproterozoic (about 1000 – 900 million years ago), so it was being formed at the same time as part of the Torridonian succession, although horizontal movements have brought them closer since they formed. Today, the Moine contains evidence of at least three different episodes of mineral alteration, the first around 850 million years ago (Knoydartian), the second 470 million years ago (Grampian; mid-Ordovician) and the last roughly 430 million years ago (Scandian; mid-Silurian), each resulting from phases of movement in the Earth’s crust where the rocks were moved, folded and fractured (Fig. 21 (#ulink_5ef00752-9d7a-5a35-83e6-cb77e10a1de8)). The Grampian and Scandian episodes are usefully distinguished as important phases in building the core of the Caledonian mountain belt. A further phase, the Acadian (mid-Devonian, 400 million years ago), is more clearly seen in other areas, showing that the movement pattern along the mountain belt involved many distinct continental fragments with different movement histories (Fig. 25 (#ulink_b6cf889e-f4b9-5df3-9253-ea72e9fe5903)). Much later, in the Mesozoic and Cenozoic, this belt was split by the plate divergence that formed the Atlantic Ocean, explaining why today there are other fragments of the Caledonian belt in Canada, Greenland and Scandinavia.

The Dalradian Supergroup was originally a succession of sediments more varied in type than the Moine. This has allowed the mapping of distinctive rock types across the country, revealing a complex pattern of folds (some upright, others over-folded) and fracture surfaces, themselves often folded after their original formation. These were formed by complex, multi-phase movements which occurred during a general convergence of the crust in a northwest/southeast direction. Radioactive dating indicates that much of this movement took place 470 million years ago, in the same Grampian episode that also deformed the Moine. It is estimated that the crustal rocks of the northern part of the Grampian Highland terrane were uplifted by some 25–35 km during this event, creating a major mountain range. Note that, despite such large amounts of uplift being indicated by research on the pressures that cause the metamorphism, mountains themselves never reach heights above sea level of this magnitude. The present height of Mount Everest is about 9 km, and this is thought to be some indication of the maximum height to which mountains can be lifted, given the powers of erosion that can be generated in present-day steep and high mountain belts. The mountains being measured in planets and moons may be bigger because of the different gravitational forces present.

Igneous intrusions were also formed during the Caledonian episodes, as heat from the compression produced molten magma that rose in the deforming crust, cooled and solidified, most commonly forming granites. These igneous volumes were emplaced both during and after the various phases of Caledonian movement. Where they have been exposed by erosion, they have given rise to differences in the material properties of the bedrock that have locally influenced the present-day landscapes.

The Great Glen Fault is one of the most obvious features of the landscape when Scotland is viewed from a satellite in space. Unlike the complex forms of the coastline and the river valleys, it represents a simple, straight or perhaps very slightly curved, vertical fracture cutting the crust (Figs 19 (#ulink_14158c9a-c802-5caf-a8ea-fcd8d794c594), 20 (#ulink_92e7c4dd-e904-5c91-a5eb-a77514394804), 22 (#ulink_974e4b5e-a91a-5d0c-be9f-f997dba62da2)). This major feature separating the Northern Highland and Grampian Highland terranes, and bisecting the Caledonian core, is now thought to have been part of a system of fractures that formed first in the Scandian phase (mid-Silurian, 430 million years ago) due to compressive continental movements that involved a strong enough oblique component to produce sliding parallel to the bedrock fabric of folds and faults generated by the general compression. A recent estimate of the amount of strike-slip sliding between Laurentia and Baltica (Fig. 25 (#ulink_b6cf889e-f4b9-5df3-9253-ea72e9fe5903)) during this phase is that it was about 1200 km, although this total movement was distributed between numerous faults. In the simple analysis of fault mechanics in Chapter 3 (#u5ec5bf09-5acf-58e3-8f03-7029db4715a5) (Fig. 17 (#ulink_aee633f4-0b1c-53bf-ad26-bfada4da029c)), a clear distinction was drawn between reverse faulting, resulting from convergence or compression, and strike-slip faulting, resulting from shearing. The present belief is that the Great Glen, and other similar faults, formed as a result of a combination of compression and shearing, sometimes referred to as oblique-slip, or transpression.

Episode 5: formation of the Lower Palaeozoic of the Southern Uplands terrane

Strongly folded, fractured and altered Ordovician and Silurian bedrock predominates in the Southern Uplands terrane. The commonest material is mudstone, often altered to slate. Altered sandstones are also common, with lesser amounts of altered limestone and volcanic material (Fig. 19 (#ulink_14158c9a-c802-5caf-a8ea-fcd8d794c594)). In the present landscapes, much of this material has been weathered and covered to some degree with Ice Age deposits, so good exposures of the sediments are rare and the hills of the Southern Uplands are generally more rounded and less rocky than those of the Highlands.






FIG 26. Sketch sections of an accretionary prism forming during subduction: (a–c) the development of thrusting (reverse faulting) within an accretionary prism; (d) age relations within the thrust stack. Thrust sheets get younger towards the southeast (i.e. 1 is older than 3) but, within each sheet, beds get younger towards the northwest. The beds are often very tightly folded and dip steeply.

It is thought that these sediments first formed as an accretionary prism, created when ocean crust in the southeast was subducted (see Chapter 3) beneath the deforming continent to the northwest, now represented by the Highlands. As subduction continued, the newly deposited sediments were folded and scraped up into a number of slices that were made of younger and younger ocean floor sediment as the movement continued (Fig. 26 (#ulink_441031df-1412-5973-85bc-d346458c101b)). How much of the Southern Uplands formed as one of these accretionary prisms is uncertain, but it is clear that the setting was marginal to the main Caledonian mountains that lay to the north. The oceanic crust was subducted along a line (locally called the Iapetus Suture: see Fig. 20 (#ulink_92e7c4dd-e904-5c91-a5eb-a77514394804)) that lay to the southeast of the Southern Uplands, roughly along the present Scotland–England border.

Episode 6: formation of the Lower Old Red Sandstone

Old Red Sandstone is the name commonly given to the red sandstones, mudstones and conglomerates that underlie rocks of Carboniferous age. The Old Red rests unconformably on older rocks in all of the Scottish terranes except the Hebridean, where it is absent (Figs 19 (#ulink_14158c9a-c802-5caf-a8ea-fcd8d794c594), 20 (#ulink_14158c9a-c802-5caf-a8ea-fcd8d794c594)). Successions of this bedrock have been classified as Lower, Middle and Upper Old Red Sandstone, depending on their fossil content and spatial relationships. Episode 6 concerns only the deposition of the Lower Old Red Sandstone.

Although fossil evidence for dating the Lower Old Red Sandstone is not common, the primitive fish and plant fossils that do occur indicate that it was deposited during the late Silurian and early Devonian, about 420 – 400 million years ago (Fig. 21 (#ulink_5ef00752-9d7a-5a35-83e6-cb77e10a1de8)). The weathering properties of these rocks are such that, in their present-day erosional landscapes, the conglomerates (with their associated lavas) have generally resisted erosion, tending to produce distinct ridges and steep slopes.

The processes of surface modification that deposited the Lower Old Red Sandstone took place largely on land, in rivers and lakes, with small amounts of sediment transported locally by the wind. Great thicknesses of lava are also important, particularly in the Midland Valley, Grampian Highlands and the Cheviot area of the Southern Uplands. The andesitic composition of these lavas suggests they were formed by internal Earth movements related to the plate subduction associated with Episode 5, and they are the earliest Scottish rocks to have yielded reliable measurements of their magnetism at the time of their formation. This information has been used to show that Scotland was located roughly 20 degrees south of the equator at this time, and it is believed that the Scottish terranes had moved into approximately their present-day positions, relative to one another, by the end of this episode (Fig. 25 (#ulink_b6cf889e-f4b9-5df3-9253-ea72e9fe5903)).






FIG 27.Geography of Scotland during deposition of the Lower Old Red Sandstone. (After Trewin 2002)

It seems likely that many of the late Silurian and early Devonian sediments and igneous rocks accumulated in distinct subsiding basins, separated by a series of northeast/southwest-trending uplifting areas that formed during the later phases of the Caledonian mountain building. Although much of the sediment in these basins was derived locally from these actively moving uplands, there is evidence that some of it was transported here by large rivers flowing from other areas of active movement in Scandinavia. The fact that the Lower Old Red sediments are predominantly non-marine in nature shows that most of the crustal surface of Scotland had been raised above sea level by this time (Fig. 27 (#ulink_c03c6115-1272-59b9-bcb6-0b85ac965b39)).

POST-CALEDONIAN EPISODES

Episode 7: formation of the middle to late Devonian, Carboniferous and Permian

It is convenient to group together as one episode the deposition of the Middle and Upper Old Red Sandstone (Devonian), the rocks of the Carboniferous and those of the Permian. The total time period represented by these units extends from about 395 to 290 million years ago, by which time Scotland had moved north to equatorial latitudes. The rocks of this episode consist largely of mudstones and sandstones, deposited by rivers in lakes, on coasts and in shallow seas. They vary considerably in age and extent, lying on the eroded top of the deformed Caledonian bedrock and often reaching thicknesses of many kilometres.

Although there is plenty of evidence of internal earth movements during this episode, their intensity and regional geography indicates a change from the strongly compressive regime associated with the Caledonian mountain building and the closing of the Iapetus Ocean (Episodes 4 to 6). By the mid-Devonian, extension had begun through much of Scotland, resulting in the formation of subsiding basins. The Middle Old Red Sandstone formed in a particularly large basin often referred to as the Orcadian Lake Basin (Fig. 28 (#ulink_97ce30e1-e360-557b-93ae-e0a8135abc77)). This extensional tectonic regime continued to characterise Scotland during much of the Carboniferous.

During the Devonian and Permian, sandy, wind-blown dune fields and evaporating groundwater conditions existed at times when local deserts developed under arid climatic conditions. The Carboniferous by contrast lacks evidence of such arid climates: river mouths were often deltaic, and the regular movement of river channels deposited distinctive cycles in the sedimentary succession, consisting of vertical changes in sediment type – most obviously between sheets of sandstone and mudstone. Limestones are also sometimes dominant where sources of sand and mud were absent. Coal-forming conditions developed repeatedly during the Carboniferous, particularly in parts of what is now the Midland Valley, and hydrocarbon-bearing mudstones were briefly but vigorously exploited west of Edinburgh. Both these had an important influence on economic and social development both locally and nationally. Carboniferous limestones, ironstones and certain sandstones have been economically important as well, at least in local terms.






FIG 28. Geography of Scotland during Middle Old Red Sandstone times. (After Trewin 2002)

Because of their economic significance, many of the Carboniferous deposits formed in this episode have been studied in great detail: tracing individual marker beds and attempting to date them by painstaking analysis of the fossil fauna and flora contained within them. This work has revealed that the Carboniferous sediments were deposited in large numbers of subsiding basins, usually only a few kilometres or tens of kilometres across (Fig. 29 (#ulink_33cd75d4-3832-57ea-8e98-9390fddd9ad7)). These basins formed due to vertical movements of the Earth’s crust along faults, the continued activity of which caused thickening and thinning of the sediments as they accumulated.






FIG 29. Geography of Scotland during early Carboniferous times. (After Trewin 2002)

As well as sedimentation, this episode also involved considerable Carboniferous igneous activity, creating volcanoes and extensive lava fields and injecting large bodies of molten rock into the crust. This igneous bedrock has had a profound effect on the present-day landscape of the Midland Valley, and also on parts of the Southern Uplands. The weathering and erosion of the landscape has preferentially picked out the igneous bedrock because it is generally more resistant than the neighbouring sediments.

The Variscan mountain building (Fig. 25 (#ulink_b6cf889e-f4b9-5df3-9253-ea72e9fe5903)) is clearly represented in southwestern England and southern Ireland. In Scotland, it appears to be represented only by a change from Carboniferous deltaic sedimentation to undoubtedly freshwater or aeolian sedimentation in New Red Sandstone times, ushering in the Mesozoic.






FIG 30. Geography of Scotland and its surroundings during the Jurassic.

Episode 8: Mesozoic sedimentation

There are only relatively small volumes of Mesozoic sediment preserved as bedrock within the land area of modern-day Scotland, but large offshore areas of the sea bed are underlain by sediment of this age. The simple explanation for this is that the approximate map-shape of present-day Scotland was already becoming established by the beginning of the Mesozoic, resulting in extensive erosion of much of today’s landmass, followed by deposition in areas that are still offshore. Reconstructions of the geography of Jurassic times, say 175 million years ago, show an upland area roughly the shape of present-day eastern and northern Scotland. This area was surrounded by basins along the Hebridean and Atlantic margins to the west and by the North Sea to the east, into which sediments accumulated (Fig. 30 (#ulink_bfa30b4d-8430-59c2-86d2-9d5ebbe6a678)). Conditions varied between the areas of accumulation, but this broad pattern continued from the Triassic, through Jurassic and Cretaceous times.

The sandstones and mudstones of the Triassic are often red due to oxidisation of their iron minerals, indicating a dry, desert-like climate. Conditions at this time were influenced partly by the global climate, but also by the general pattern of plate movement which, by the end of the Triassic, saw Scotland at about 30 degrees north – equivalent to the present-day latitude of the Canary Islands.

In Jurassic times, where river deltas fed into shallow seas, a wide variety of rock types was deposited: mudstones, sandstones and limestones, along with rare ironstones and coals. Organic material – largely algal – formed locally in some of the muddy seas and was particularly abundant in the case of the Late Jurassic Kimmeridge Clay. This unit has been the main �source rock’ for the North Sea hydrocarbons that have had such a critical influence on the British economy over the last 40 years. Key points in the trapping and preservation of the hydrocarbons are the presence of sandstone with a suitable porosity, and earth movements that have subsequently stretched the crust, faulting it to seal the hydrocarbon reservoirs. Meanwhile, fault-related Jurassic landslide deposits are a spectacular feature of outcrops on one stretch of the east coast of the northern Highlands (see Areas 16 and 17), while in some parts of the Hebrides Jurassic sandstones have provided resistant bedrock that has influenced the development of the landscape.

Cretaceous bedrock is very rare on land in Scotland and is generally only preserved as isolated fragments in areas of Tertiary volcanism, where sheets of lava have protected the Cretaceous rocks from the erosion that has removed them elsewhere. Small amounts of sandstone and chalk (the Late Cretaceous algal limestone that is such a dominant feature of the landscape of southern England and northern France) are preserved in some of the volcanic centres, but do not tend to influence landscapes on a scale that can be considered in this book. On the other hand, the offshore record of the Late Cretaceous around Scotland is much more complete, and the lack of mud and sand (derived from the erosion of land-based bedrock) in these deposits suggests that Scotland had been eroded down to a largely flat landscape by this time.

Episode 9: Tertiary volcanism

About 60 million years ago, in the earliest Tertiary, a dramatic episode of igneous activity took place along the western seaboard of mainland Britain. The resulting bedrock has played a major role in forming features of the landscape of the western Hebridean, Northern Highland and Midland Valley terranes. Successions of lava flows formed volcanic lava fields tens to hundreds of metres thick in many areas of the Inner Hebrides and northern Ireland. Distinct fields have been dated around Eigg and Muck at 60.5 million years old, around Skye and Canna at 58 million years old and around Mull and Morvern at between 58.5 and 55 million years old. The layered (�stepped’) landscapes eroded in the bedrock of these lava fields are striking, and are due primarily to differences in erosional resistance between the lower and upper parts of each lava flow.






FIG 31. General pattern of processes thought to underlie a typical igneous centre.

Even more striking are the centres of volcanic activity and igneous intrusion that developed in a scatter of localities shortly after the lava fields formed (Fig. 31 (#ulink_4df6b40a-4419-57b5-8cc0-51492f475684)). The coarsely crystalline intrusive rocks of these centres dominate the landscapes of their surroundings, because of the resistance of this material to erosion. The eroded remains of these ancient igneous centres now form the remarkable Cuillin and the Red Hills of Skye, the mountains of Rum, the hills of the Ardnamurchan peninsula and the main mountains of Mull and Arran, not to mention the islands of St Kilda and Ailsa Craig.

In wider geographical terms, these Tertiary igneous activities, along with the associated uplift and erosion, were responses to the tectonic plate divergence movements that created the Atlantic Ocean, with additional igneous input related to �hot-spot’ activity in east Greenland, Iceland, the Faroes, western Scotland and northern Ireland.


CHAPTER 5

Later Surface Modifications

THE PREVIOUS CHAPTER dealt with nine episodes recorded in the bedrock of Scotland. This chapter deals with three more recent episodes (Episodes 10–12; Fig. 21 (#ulink_5ef00752-9d7a-5a35-83e6-cb77e10a1de8)) which have modified the surface, removing bedrock and adding soft material to the surface blanket.

SURFACE-MODIFICATION EPISODES

Episode 10: Tertiary landscape erosion

Dating of the lavas extruded in Episode 9 suggests that Tertiary igneous activity in Scotland lasted for only about 5 million years and finished about 55 million years ago. This was followed by more than 50 million years of Tertiary and Quaternary landscape erosion (Fig. 21 (#ulink_5ef00752-9d7a-5a35-83e6-cb77e10a1de8)), during which time the main valleys of present-day Scotland increasingly approached their present shape and size.

Sedimentary bedrock of Tertiary age (Palaeogene and Neogene) is very largely absent on land in Scotland, even where volcanic and other igneous bedrock is present. This suggests that the crust below the present land area of Scotland was moving upwards and was subjected to net erosion during most of the Tertiary. Part of the evidence for this is the large thickness of Tertiary sandstones and mudstones that are found offshore to the east, north and west of Scotland, as shown by extensive oil exploration.

The valleys and mountains of Scotland, along with the lochs, sea lochs and offshore rock basins, have all been shaped by this erosion, principally by Tertiary rivers but also by more recent glacial ice (Episode 11). The present-day drainage pattern in Scotland (see Chapter 2) represents the latest phase in the evolution of this erosional system, and provides clues to the way it may have developed over the past 55 million years.

Episode 11: the Ice Age

During the nineteenth century, it became generally accepted that much of Britain had been subjected to glaciation by ice sheets and valley glaciers. Since then, this distinctive episode in the history of the British landscape has been referred to as the Ice Age, broadly equivalent to the Quaternary period of the internationally accepted series of time divisions (Fig. 21 (#ulink_5ef00752-9d7a-5a35-83e6-cb77e10a1de8)).

Over the last few years of geological research, one of the most far-reaching developments has been the establishment of the detailed record of fluctuating climate changes that have occurred during the Ice Age. A key step in this advance was the realisation that various indicators (often called proxies) of climate change can be measured at very high time resolution in successions of sediment or ice. The first of these successions to be tackled covered only the last few thousand years, but further work has now provided estimates of global temperature extending back several million years.

One of the best climate indicators has turned out to be variations in the ratios of oxygen isotopes (oxygen-16 versus oxygen-18), as recorded by microfossils that have been deposited over time on deep ocean floors. When alive, these organisms floated in the surface waters, where their skeletons incorporated the chemistry of the ocean water – including the relative amounts of oxygen-16 and oxygen-18. During cold climatic periods (glacials) water evaporating from the oceans may fall as snow on land and may be incorporated within ice sheets. Because oxygen-16 is lighter than oxygen-18 it evaporates more easily, so during cold periods the newly formed ice sheets tend to be rich in oxygen-16, relative to the oceans. The ratio of oxygen isotopes in the world’s oceans, as recorded by microfossils, can therefore be used to distinguish glacial and interglacial periods. Other useful indicators of ancient climate have come from measuring the chemical properties of ice cores, which preserve a record of the atmospheric oxygen composition, to complement the oceanic data from sediment cores.

Ratios of the isotopes of oxygen have turned out to provide one of the most important indicators of climate change, because they depend principally on ocean temperature and the amount of water locked up in the world’s ice sheets. There are, however, numerous other factors that can affect the ratios in ice and sediment cores, so interpretation of the data is rarely straightforward.

Figure 32 (#ulink_aec88710-8551-5019-b030-be6ddd3d7bd4) shows corrected oxygen isotope ratios as an indicator of temperature over the last 3.3 million years. The numbers on the vertical axis are expressed as δ18O values (pronounced �delta 18 O’), which compare the oxygen-18/oxygen-16 ratios in a given sample to those in an internationally accepted standard. The greater the proportion of heavy oxygen-18 in a sample the larger the δ18O value and, as described above, the lower the corresponding ocean temperature. For this reason, the vertical axis on Figure 32 (#ulink_aec88710-8551-5019-b030-be6ddd3d7bd4) is plotted with the numbers decreasing upwards, so that warmer temperatures are at the top of the figure and cooler ones at the bottom. The pattern shown in Figure 32 (#ulink_aec88710-8551-5019-b030-be6ddd3d7bd4) is of an overall cooling trend with, in detail, a remarkable series of over 100 warm and cool periods or oscillations. These alternations have been numbered, for ease of communication by the scientific community, with even numbers for the cold periods and odd numbers for the warm periods.






FIG 32.Oxygen isotope ratios track the more than 100 climate fluctuations over the last 3.3 million years. Warm episodes (red lines above the curve) alternate with cold episodes (blue lines below the curve). These have been used as the basis for numbering the global oxygen isotope stages, as shown.

Our next step involves looking in greater detail over roughly the last 400,000 years (Fig. 33 (#ulink_787309c8-3647-59bf-868b-c3329bbc0f2e)). Over this period, there has been a distinctive pattern of increasingly highly developed 100,000-year-long cold stages, separated by 10,000-year-long warmer stages. This temperature curve (also calculated from isotope ratios) is saw-toothed in shape, representing long periods of cooling followed by rapid warming events. The most recent of the four glacial episodes covered in this diagram (the Devensian) has left abundant fresh evidence on the landscapes of Scotland and obliterated most of the evidence of the earlier ones. In this important respect, the Scottish evidence differs strongly from that of southern England, where the much earlier Anglian glacial episode has left abundant evidence of ice as far south as London. This is because later glaciations, such as the Devensian, did not reach so far south. Not surprisingly, the older evidence in southern England is not as fresh as that of the younger glaciation in Scotland.

An even closer look at the last of these cold-to-warm changes (Fig. 34 (#ulink_645d655f-0818-5449-8f51-98fedb3f9682), black line) allows us to appreciate better the glaciation which has been responsible for much of the recent modification of Scottish landscapes. Starting with the Ipswichian interglacial, the Greenland curve shows fluctuations in the oxygen isotope ratios that were frequent and short-lived, though generally implying increasingly cool conditions. This part of the record is helping to define the Devensian glaciation and shows clearly the Late Glacial Maximum (LGM) at between about 30,000 and 20,000 years ago. Following this, the beginning of the Holocene warm period (about 10,000 years ago) is also clear.






FIG 33. Isotopic temperature of the atmosphere changing through the last 400,000 years, measured from ice cores taken from Vostok, Antarctica.

The link between oxygen isotopes, temperature and sea level becomes clear if we compare oxygen isotope ratios from the Greenland ice (Fig. 34 (#ulink_645d655f-0818-5449-8f51-98fedb3f9682), black line) with sea-level data from tropical reefs in Papua New Guinea (Fig. 34 (#ulink_645d655f-0818-5449-8f51-98fedb3f9682), red line). The data show how colder climates are generally associated with lower sea levels, reflecting the locking up of oxygen-16-rich water in land-based ice sheets during these colder times.






FIG 34. Black line: oxygen isotope ratios sampled from cores taken in the Greenland ice sheet. Red line: sea-level determinations from tropical reefs in Papua New Guinea.

At its maximum extent the Devensian ice sheet covered the whole of Scotland, including the western and northern islands. It also covered most of Wales and northern England and extended as far south as the Midlands, the Bristol Channel and the Wash. Maintaining a thickness of many hundreds of metres, it joined Norwegian ice on the Norwegian side of the northern North Sea (Figs 35 (#ulink_7e6b468b-fba8-5cf0-915e-b3347ed0d30e), 36 (#ulink_214ad505-b82c-5eec-9f25-23bc1af6c8e5)).






FIG 35. One estimate of the maximum extent of the Devensian ice sheet, with generalised ice-flow directions. At a later stage the Scottish and Norwegian ice became separated.






FIG 36. West-to-east generalised cross-section at the maximum extent of the Devensian ice sheet.






FIG 37. The larger rock basins are the result of erosion by Quaternary ice streams.

There is abundant local evidence in Scotland of the modification of valleys by glaciers and ice streams, which deepened and opened out the valley profiles, removing spurs and side ridges, to produce classic U-shaped glacial troughs. These troughs are very different from the V-shaped cross-sections and sinuous forms typical of river erosion (see Fig. 8 (#ulink_2429fe1f-257d-54c2-b31a-a5bdf6804749), Chapter 2 (#uc0e147ab-0db8-5cd3-a46c-c03cd0831740)). This modification work is likely to have taken place in every one of the Ice Age glacial stages that occurred in Scotland, and the same processes have also been responsible for the elongate rock basins now recognised in many offshore areas (Fig. 37 (#ulink_84f91e1e-e530-58b9-8d5f-abe42f81268b)).






FIG 38. Shrinking of the main Scottish ice sheet over the last 18,000 years.

Episode 12: since the Devensian Late Glacial Maximum

The period of rather more than 20,000 years since the Late Glacial Maximum represents one of the most recent phases of intense landscape evolution (Fig. 38 (#ulink_b8e5fada-2208-5fa5-b5ee-8243ebdd5187)). Because this was a period when ice cover was generally decreasing, local evidence is often preserved that would have been destroyed during a major phase of advancing ice. The last 10,000 years is often referred to as either the Holocene or the Flandrian Interglacial, the latter name emphasising that the ice may well return.






FIG 39. Oxygen isotope variation from (a) Greenland ice cores and (b) Northeast Atlantic sea-surface temperatures, both over the time period from 15,000 to 10,000 years BP (before present).

The record of climate change since the Late Glacial Maximum has been greatly illuminated by the same use of oxygen isotopes as described above for Episode 11. One important advantage in working on these recent times is that it is possible to seek additional, independent information for the ages of samples. Some of this dating may be based on comparison of plant remains, particularly pollen from cores extracted by drilling into lake beds or peat-rich wetlands. Other dates come from the analysis of radioactive carbon, whose rapid decay rate makes it a powerful tool in dating material that is so relatively young.

Although the dominant feature of global climate change over the past 20,000 years has been the general warming trend, detailed research has established a complex pattern of climatic fluctuations. In Scotland, the most important of these fluctuations is the Younger Dryas cold phase, also known as the Loch Lomond Stadial (Fig. 39 (#ulink_fb953f5d-69f7-5726-b236-2e0f344fd41f)). During this time, between about 13,000 and 11,500 years ago, the generally retreating ice re-advanced to form an icecap covering much of the western Highlands (Fig. 38 (#ulink_b8e5fada-2208-5fa5-b5ee-8243ebdd5187), red line). The local effects of this Loch Lomond Advance are particularly clear within the area of western Scotland where moraines were pushed forward.

SEA-LEVEL CHANGE

In Areas with coastlines, some of the freshest features of the landscape have formed since the Late Glacial Maximum as a result of changes in sea level. Two different mechanisms have combined to produce these changes:

(1) Worldwide ocean-volume changes of the water occupying the world’s ocean basins. These have been the direct result of the locking-up or releasing of water from land-based ice sheets as they grow or shrink due to climate fluctuations. The water itself may also have expanded or contracted as its temperature changed. These worldwide processes are often grouped together as eustatic.

(2) Solid Earth local movements which have resulted in the local raising or lowering of the ground surface relative to the level of the sea. These movements were responses to changes in the local temperature or stress pattern within the Earth. Ice-sheet melting unloaded the crust of the Earth locally, resulting in uplift, while ice-sheet growth loaded the crust, resulting in subsidence. These effects are often referred to as isostatic adjustments of local sea level (see Chapters 2 and 3).

Some parts of the world, for example many tropical areas, have been free of ice since before the Late Glacial Maximum and so have avoided any solid Earth movements associated with loading and unloading by ice. Records of changing sea level from these areas can therefore be used to estimate worldwide (eustatic) changes in the volume of the world’s oceans since the Late Glacial Maximum. Figure 40 (#ulink_98c98f73-a241-5a85-9d85-77ecebb0d443) shows that eustatic sea level has risen by about 120 m over the past 18,000 years, beginning with a slow, steady rise until about 12,000 years ago, followed by a rapid increase until about 6,000 years ago, and then another slow, steady phase up to the present day.

Curves of local sea-level change for any area can be estimated (relative to the present) by recognising and dating various features that indicate elevations in ancient coastal profiles. These features, preserved in the rocks either above or below the present sea level, include former erosional cliff lines, wave-cut platforms and ancient tidal, estuarine or freshwater deposits. The similarity or otherwise of such curves to the eustatic curve (Fig. 40 (#ulink_98c98f73-a241-5a85-9d85-77ecebb0d443)) depends on whether the areas in question have been subjected to any localised solid Earth movements, such as ice loading or unloading.

Two examples of British sea-level curves, relative to the present, illustrate how the local uplift and subsidence history varies for different coastal areas around Britain. In the Thames Estuary, local evidence shows that a rise of some 40 m has taken place through time over the last 10,000 years, at first very rapidly but then more slowly between about 6,000 years ago and the present (Fig. 41 (#ulink_98352187-f11b-538f-90c7-d1eb7b85a983), red circles). Modelling of the processes involved, incorporating estimates of eustatic (global) sea-level change and local solid Earth movements, gives a fairly good match to the observational data (Fig. 41 (#ulink_98352187-f11b-538f-90c7-d1eb7b85a983), black line).






FIG 40. Generalised change of worldwide (eustatic) sea level over the past 18,000 years. (After Van Andel 1994, Fig. 4.11)






FIG 41.Relative sea-level curve for the Thames Estuary.






FIG 42. Relative sea-level curve for the upper River Forth at Arnprior.

Our second example of relative sea-level change comes from the upper River Forth and is quite different. It shows that there has been a fall of relative sea level of about 50 m over the past 15,000 years, so that former coastline features are now visible well above the present-day coast (Fig. 42 (#ulink_61b601bf-5753-5504-937e-c927f5e20c19), red circles). This type of curve is common in Scotland and, given the ~120 m worldwide rise in sea level shown in Figure 40 (#ulink_98c98f73-a241-5a85-9d85-77ecebb0d443), it is clear that the crust of the Forth region must have been subjected to significant uplift (~170 m) in order to produce the curve shown in Figure 42 (#ulink_61b601bf-5753-5504-937e-c927f5e20c19). This uplift is largely the result of isostatic rebound due to unloading of the crust as the ice retreated.

In some curves, as with the upper Forth, oscillations in the curve represent changes in the rates at which the two mechanisms of change were operating. Such changes may leave characteristic coastline features in the landscape, which will be considered in the Area descriptions.

At larger scales, it is useful to consider average rates of crustal movement over a given time period, which can then be plotted as contour maps. The contours shown on Figure 43 (#ulink_c7d3716f-a14d-5537-9ca0-f022ad054044) are based on estimates of local elevation changes averaged over roughly the last 5,000 years, attempting also to allow for the effects of eustatic sea-level variations. Additional support for this approach comes from data from tide-gauge studies, collected over the last 200 years, which show some consistency with this pattern.

Although our two local studies in the Thames region and the upper Forth (Figs 41 (#ulink_98352187-f11b-538f-90c7-d1eb7b85a983), 42 (#ulink_61b601bf-5753-5504-937e-c927f5e20c19)) involve a rather longer timescale than the regional analysis (Fig. 43 (#ulink_c7d3716f-a14d-5537-9ca0-f022ad054044)), all three studies highlight the clear contrast between crustal movements in southeastern England and those in western and central Scotland. These variations have been produced by differences in ice-sheet thickness and extent during the last (Devensian) glacial.

The distinctive rise of the land of Scotland relative to sea level lends itself to another approach to the study of sea-level change. Figure 44 (#ulink_8e1f8a88-c56c-588e-8bdb-bd05a4a3d355) shows a plot of the elevations and gradients of various old shoreline features that are now above present-day sea level across Scotland and northern England. These old shoreline features have clearly been uplifted, and those further inland have generally risen more than those near to the present-day coast, so that they now form a dome-like structure. This dome is broadly centred on Rannoch Moor, which was one of Scotland’s main ice centres during the last glaciation. We can therefore usefully identify a Rannoch Rebound Dome as an active feature of local Earth movement, resulting from unloading of the crust of western Scotland in response as the Devensian ice melted.






FIG 43. General trends of crustal movement, relative to sea level, averaged over the last 5000 years.






FIG 44. The elevations and gradients of various old shoreline features along a horseshoe-shaped traverse across northern Britain, suggesting the Rannoch Rebound Dome.


CHAPTER 6

Area 1: Galloway

AREA 1 EXTENDS FROM AYR in the northwest down to Dumfries in the southeast (Fig. 45 (#ulink_198b0e89-2d40-58d2-8ddb-d43a9f9d99df)). It lies mainly within the western half of the Southern Uplands, a terrain of rolling hills, bounded to the west by the Firth of Clyde and its numerous sandy bays. The Southern Uplands Fault crosses the northern half of the Area, separating the Southern Uplands from the generally lower-lying ground of the Midland Valley, with its volcanic hills and important coal reserves (Fig. 46 (#ulink_09dd45a9-6c18-5f65-a6b5-d38fb3378ef5)).

People have inhabited this Area for thousands of years, and it was an important gateway between England and Ireland. There are many remains of human occupation dating from prehistoric times to the present day, ranging from Mesolithic fish traps to medieval burghs and castles. Another interesting feature of Area 1 is the unusual place names, particularly in the Southern Uplands themselves, where Old Norse, Gaelic and Celtic influences can be seen. Examples are the Rig of the Jarkness and the Dungeon of Buchan.






FIG 45. Location map for Area 1.






FIG 46. Natural and man-made features of Area 1.

STORIES FROM THE BEDROCK

Geologically speaking, Area 1 lies mainly within the Southern Uplands terrane, sandwiched between the Iapetus Suture in the south and the Southern Uplands Fault in the north (see Fig. 20 (#ulink_92e7c4dd-e904-5c91-a5eb-a77514394804), Chapter 4 (#ub72c883b-08de-5367-91ab-f3d6ef483997)). The bedrock of the Southern Uplands (Figs 47 (#ulink_5f5e697c-fb37-5248-acae-75ecf871e10e), 48 (#ulink_5cf891da-8d88-5bdd-8ef5-ca222d5bb5f5)) is mostly altered Ordovician and Silurian sedimentary rocks, deposited between 490 and 420 million years ago on the floor of the Iapetus Ocean (see Chapters 1–5 (#u661596f2-c933-517a-abc9-8c82715c3114)). The sediments which make up this bedrock were swept off a nearby continental shelf and down the continental slope in turbid (muddy, cloudy) currents – underwater avalanches, more dense than the surrounding sea water – that were probably earthquake-triggered. On reaching the flat ocean floor, the entrained sediments in each avalanche gradually settled – coarse sands first, followed by fine sand, and then the much slower deposition of clay and mud. In this way, each turbid current resulted in a graded bed, from coarse-grained at the bottom to fine-grained at the top, and as the process repeated itself many times, a thick sequence of such beds built up. Small amounts of limestone were also deposited, along with volcanic material such as pale ash layers. Graptolites – small, now-extinct marine animals – are common in the fine-grained sediments of the Southern Uplands. Their rapid evolutionary changes of form mean they have become very useful time markers for determining the relative ages of different sedimentary beds, especially when combined with studies of the folds and faults, in reconstructing the origins of the Southern Uplands.






FIG 47. Simplified geology and hill-shaded topography for Area 1.






FIG 48. Timeline of bedrock and surface-layer events in Area 1.

Prior to the Caledonian mountain building, the crustal foundations of Scotland and England were separated by the Iapetus Ocean. Around 490 million years ago, this ocean began to be destroyed by subduction: oceanic crust moved down into the mantle beneath the Grampian Highlands, and then beneath the Midland Valley (see Chapter 4). A small fragment of this oceanic crust escaped subduction, being instead thrust up onto the margin of Scotland, to be �welded’ onto the Midland Valley by around 470 million years ago. Today, this small but intensively studied area of complexly interfolded rock units outcrops around Ballantrae, the so-called the Ballantrae Complex. Rocks characteristic of the deep sea and oceanic crust are found – sediments such as black shale and chert, basalt lavas with pillow structures, ash, sheets of dykes and upper mantle rocks. The latter originated at depths of up to 40 km in the Earth’s crust, and are today coarse-grained (mafic) gabbros and serpentinite.

The main deformation of the Southern Uplands terrane occurred during the later stages of the Caledonian mountain building, between the mid-Ordovician and the early Devonian. As oceanic crust continued to be subducted, the sediments which today make up the Southern Uplands were scraped off the ocean floor along a series of thrust faults and stacked up in a pile against the edge of the Midland Valley (Fig. 28 (#ulink_97ce30e1-e360-557b-93ae-e0a8135abc77)). During this deformation, the sediments became tightly folded and weakly metamorphosed: fine-grained mudstone and siltstone became slate, while cement within sandstones recrystallised to produce a tough, hard rock (greywacke). Today, bedding in the Southern Uplands is aligned in a general northeast/southwest direction and dips very steeply to the southeast, and northeast/southwest faults divide the region into numerous fault blocks.

By the start of the Devonian (around 415 million years ago), the major deformation of the Southern Uplands had ceased and Scotland and England were welded along the Iapetus Suture. It was around this time that the major granite bodies of the Southern Uplands (Fig. 47 (#ulink_5f5e697c-fb37-5248-acae-75ecf871e10e)) were emplaced: partial melting at the base of the thickened crust produced liquid magma, which then rose up into the upper crust where it slowly solidified to form coarse-grained igneous bodies (plutons). As the overlying rocks were subsequently removed by erosion, three major plutons were revealed in the Southern Uplands. The most northerly of these is the hourglass-shaped Loch Doon intrusion, said to be one of the finest examples in Scotland of a concentrically zoned pluton: the interior of the body is silica-rich (felsic) granite, separated from the outer silica-poor grey granodiorite which makes up most of the body by a transition zone. Similar well-developed concentric zonation is seen in the eastern half of the Criffel–Dalbeattie body on the south coast, although overall this body is much less compositionally evolved (i.e. it has a lower silica content) than the other Southern Uplands granites. It is also the most deformed: originally oval, its western part has been distorted southwards by complex faulting. Porphyrite dykes and sills commonly surround the main intrusion (e.g. at Black Stockarton Moor), made up of large crystals embedded in a fine, glassy groundmass. Between the two, the roughly oval Fleet pluton was intruded around 390 million years ago (Devonian) into a broad ductile shear zone, making it the youngest reliably dated Caledonian pluton in mainland Scotland (Fig. 49 (#ulink_87059e73-9e67-58ac-9415-3f87fed46f49)). It is also the most evolved of the Southern Uplands intrusions, consisting entirely of granite, and is the only intrusion whose magmas were sourced wholly from the melting of metamorphosed sediments (rather than igneous rocks). Because of both its young age and its evolved composition, this pluton has more in common with the Lake District and Northern Ireland granites than with those of Scotland, and it has been suggested that these areas shared a magma source.






FIG 49. Cairnsmore of Fleet, 711 m (10 km east of Newton Stewart), viewed from the southeast. This mountain landscape has been created by erosion of the Fleet granite intrusion. (В© Lorne Gill, Scottish Natural Heritage)

As these hot granite bodies were emplaced, their heat baked the surrounding rocks, creating an encircling metamorphosed zone (an aureole) 1 km or, in the case of the Criffel–Dalbeattie intrusion, even 2 km wide. These aureoles are often rich in mineral veins, deposited by hot circulating fluids released by the crystallising granite. Gold, silver, copper, lead and zinc are common, particularly around the Fleet intrusion, and over 60 copper and iron-rich carbonate veins have been located northwest of the Criffel–Dalbeattie pluton.

Volcanic vents active during the early Devonian are also present in the area, although they are generally poorly preserved. An exception is the large vent at Shoulder o’ Craig, 17 km southwest of Castle Douglas, on the Dee estuary. The headland here is principally made up of a vent-filling intrusion breccia, which consists of Silurian sandstone and siltstone clasts within a basalt (mafic) matrix. Both vent rock and country rock are cut by very potassium-rich dykes, indicating a magma source deep within the mantle. These dykes often have irregular shapes, and one dyke in the area is known as the �Loch Ness Monster’ due to its particularly bizarre outcrop pattern. On a regional scale, this area presents a bit of a conundrum, as volcanic vents, mantle-derived dykes and granite plutons, i.e. igneous rocks from all depths within the crust, were intruded around the same time (between around 415 and 400 million years ago, earliest Devonian), and are now seen at the same level of erosion.

Further north, the late Silurian and early Devonian was the time when a series of basins first began to develop in what would become the Midland Valley, as crustal tension caused movement on the Highland Boundary and Southern Uplands faults. At this time (around 420 to 400 million years ago), Scotland lay in the interior of a large continent some 20 degrees south of the equator, and in this environment the new Caledonian mountains were eroded rapidly because soil-binding plant cover had not yet evolved. Rivers and streams washed the sediment into the developing Midland Valley basins, forming coarse conglomerates, red sandstones and mudstones, collectively called the Lower Old Red Sandstone. Volcanic rocks (associated with crustal extension) are common in the upper 600 m of the Lower Old Red Sandstone, where lava sheets (predominantly andesite) are intercalated with river and lake sediments, mostly sandstones. Today, principal outcrops include a 400 m-thick lava pile underlying the Carrick Hills and a 600 m-thick lava pile in the Dalmellington area (20 and 30 km east of Girvan, respectively).

The Carrick Hills lava pile is particularly well exposed along the coast around Dunure (10 km southwest of Ayr). This coastal section has been studied for over a century in an attempt to unravel the complex relationships between the lava and intervening sediment; the upper and lower surfaces of andesite (mafic) sheets are often very irregular, with bulbous, finger-like protrusions that extend upwards and downwards into the sediment, or have become detached completely, forming zones of lava pillows. In places, lava engulfs patches of sediment; elsewhere, the lava is surrounded by sediment. The andesite sheets are generally well jointed, and these joints are often filled with hardened sandstone. Despite these contorted relationships, lamination in the sandstones is generally intact, save for a small zone near the contact. Such irregular contacts are thought to result from the sills being intruded into wet, unconsolidated sediment; as hot magma was emplaced, it vaporised water at the magma–sediment contact, fluidising the sediment in a narrow zone next to the contact. This vapour and its entrained sediment then flowed away along the hot contact surface, offering very little resistance to the magma and allowing bulbous protrusions to form. Likewise, the liquid magma could not push directly against the wet host sediment, and so this sediment remains largely undeformed, except at the contact zone. After intrusion, large amounts of water vapour were trapped in sediment enclaves and at contact zones. As the andesite then cooled, it contracted and cracked, often resulting in a sudden decrease of pressure in the sediment. This led to explosive boiling of the water, fluidising the sediment and blasting it along the fractures and cooling joints. Vesicles (cavities formed by gas bubbles) are also very common in the lavas, generally now infilled by minerals such as quartz, agate or chalcedony precipitated by circulating groundwaters.

By the middle Devonian (400 to 385 million years ago), further earth movements resulted in uplift and erosion of much of the sediment laid down in early Devonian times, and some of the underlying Ordovician and Silurian. The main granite bodies probably became exposed at the surface during this time, as evidenced by the clasts of Criffel–Dalbeattie granite found in Upper Old Red Sandstone deposits in Area 2 to the east. These late Devonian deposits are rare in Area 1, only outcropping near Dalmellington in a thin strip north of the Southern Uplands Fault.

The Caledonian Mountains had been largely eroded by the start of the Carboniferous, around 360 million years ago, although the Southern Uplands still formed a considerable upland area. Throughout the following 60 million years of the Carboniferous, deposition occurred mostly in the lowlands of the Midland Valley and the Solway Firth basins in marine or coastal-plain environments. Sea levels varied, resulting in the deposition of limestones, sandstones, mudstones and coal, often arranged in �cycles’ of varying layers, as shallow seas and river estuaries gave way repeatedly to swampy forests. Towards the end of the deposition of the Lower Carboniferous, the Southern Uplands had been sufficiently lowered by erosion to be breached by the sea along what is today Nithsdale, and the Midland Valley and Solway Firth basins were linked. Coal deposits were laid down under swampy conditions in the Carboniferous, and are today found around Sanquhar and Thornhill and in the larger Ayr Basin. These sedimentary basins were defined by numerous northwest-trending normal faults. The Carboniferous was also a time of renewed igneous activity, after the quiet of the mid- and late Devonian. This activity was associated with faulting and basin formation, and continued intermittently for some 100 million years until mid-Permian times. Today, lavas, volcanic plugs and sills from this time underlie much of the high ground in the Midland Valley. Hot fluids associated with this igneous activity resulted in mineral veins forming, and in many cases these have been economically important for the region. Gold, silver and lead have been mined for centuries from the well-known mining district around the Lowther Hills and Leadhills (20 km north of Thornhill, Fig. 46 (#ulink_09dd45a9-6c18-5f65-a6b5-d38fb3378ef5)). Leadhills has been designated a Site of Special Scientific Interest (SSSI) because of the variety of rare lead minerals present. Lead smelting in the Leadhills area has left its mark on the countryside, in the form of old tips, abandoned machinery and poisoned vegetation.

By the end of the Carboniferous, Scotland had drifted northwards from the equator and the climate changed from tropical to arid. Throughout the Permian (between 300 and 250 million years ago), Scotland had a desert climate in which the red sandstones and conglomerates of the New Red Sandstone were deposited, often on top of Carboniferous rocks as sedimentary basins continued to subside. Today, significant outcrops of Permian sediments are found between Loch Ryan and Luce Bay (near Stranraer), in the southern and central parts of Nithsdale and east of Ayr.

During the Mesozoic, sea levels were at times up to 300 m higher than today, and shallow-water sediments are likely to have been deposited at least in the Midland Valley. However, no Mesozoic rocks are preserved today, showing that, overall, the last 250 million years have been a time of net erosion in Area 1, as in much of Scotland.






FIG 50. South end of Ailsa Craig. The highest point of the island is 338 m above sea level. The term �Paddy’s Milestone’ has been applied, because the island is a marker by sea between the Clyde ports and those of Ireland. The paddle steamer Waverley is close to the shore, which is fringed by a raised beach marking the recent uplift of the island relative to sea level. (© David Law)

The youngest bedrock in this Area underlies the small but remarkable island of Ailsa Craig, some 15 km northwest of Girvan (Fig. 50 (#ulink_aa7379e2-a84e-5957-8cf4-007e3146be66)). The island is the deeply eroded remains of a volcanic plug, emplaced at the start of the Tertiary (around 60 million years ago) into gently dipping Permo-Triassic rocks. The intrusion is a fine-grained granite, whose unusual minerals give the rock a characteristic bluish colour. Columnar jointing is very prominent around the island, as are quarries from which the rock has been extracted to manufacture the famous polished curling stones (or �ailsas’).

MAKING THE LANDSCAPE

In early Tertiary times, sea-floor spreading in the North Atlantic was accompanied not only by the eruption of lavas in the Tertiary Volcanic Province (including the intrusion of the Ailsa Craig microgranite), but by widespread uplift across much of the Scottish mainland. The Southern Uplands and Highlands were once again uplifted, while the Midland Valley, lying on the periphery of these two blocks, became relatively lowered. The uplift, and the more modest episodic uplift events of the later Tertiary, were accompanied by vigorous denudation, often concentrated along lines of geological weakness such as faults and softer sedimentary units. In the generally warm, wet climate of the Tertiary, the intervening phases of tectonic stability were times of deep bedrock weathering that enhanced the pre-existing relief, widening valley floors and basins and resulting in the development or extension of erosion surfaces. In this way, the main landscape features seen today were initiated during the Tertiary: an erosion surface between 400 and 600 m in elevation developed across the Southern Uplands, dissected by numerous river valleys. The final form of the Southern Uplands owes much to glacial erosion, but the Tertiary erosion surface is still apparent as the smooth, rounded hills tend to be at uniform heights at approximately this elevation. The projecting hills of the Southern Uplands tend to be underlain by more resistant material, which would have formed topographic features during the Tertiary before being moulded by glaciers. Examples are the higher hills of the Lowther Hills or Leadhills (in places over 750 m high), which tend to be made of tougher and more resistant quartzites and thick beds of grit, whereas the thinner greywackes and shales have been weathered into gentler rolling hills. Further west, the highest hills of the Southern Uplands are found around the Loch Doon granite, although they are not underlain by the granite itself. This will be examined later, when looking at the effects of glacial erosion on the landscape.






FIG 51. Elevation map of Area 1, showing the main river valleys and upland areas.

The rolling hills of the Southern Uplands are interrupted by the broad valleys of the rivers Cree, Dee and Nith, which flow roughly southeast off the high ground into the Solway Firth (Fig. 51 (#ulink_0038c7b5-86de-59b0-86a7-9ea4660af799)). Another prominent area of low ground oriented roughly northwest to southeast has been flooded by Loch Ryan and Luce Bay, and therefore separates the Rhins peninsula from the mainland. It is obvious in Figures 47 (#ulink_5f5e697c-fb37-5248-acae-75ecf871e10e) and 51 (#ulink_0038c7b5-86de-59b0-86a7-9ea4660af799) that the river valleys of the Cree and Dee are aligned roughly parallel to large northwest/southeast-trending faults, and it seems likely therefore that the more easily weathered rocks in the fault zone provided a relatively easy pathway for river erosion, probably as early as the Tertiary and certainly more recently. It is also clear from Figure 47 (#ulink_5f5e697c-fb37-5248-acae-75ecf871e10e) that Luce Bay–Loch Ryan and Nithsdale are in part underlain by Devonian to Permian sedimentary rocks. These rocks are softer than the surrounding Ordovician and Silurian rocks, and have been more extensively weathered to form the low ground seen today. In effect, the Nith is once again flowing down what would have been a valley at least as far back as Carboniferous times, when a sedimentary basin became established running at right angles to the northeast/southwest-trending major faults, such as the Southern Uplands faults and the general folding of bedrock.






FIG 52. Hill-shade map of southwestern part of Area 1. The Southern Uplands Fault shows up well, as do other erosional and depositional features due to glaciation. Note the drumlins north and west of Patna Hill.

The fault which is most obvious in the landscape is the large Southern Uplands Fault. River and stream valleys have been preferentially eroded along the fault over much of its length. The fault is particularly prominent at its southwestern end (Fig. 52 (#ulink_7b1f5143-141b-5238-bf11-4293cff1d9a7)), where it splits into two (the southern Glen App Fault and northern Stinchar Valley Fault: Fig. 47 (#ulink_5f5e697c-fb37-5248-acae-75ecf871e10e)). Preferential weathering along the Glen App Fault has resulted in the remarkably steep-sided, linear valley of Glen App, whilst the more curved line of the Stinchar Valley Fault has been excavated and now underlies Stinchar Valley. In broader terms, the Southern Uplands Fault separates the generally higher, hillier ground of the Southern Uplands from the lower-lying, flatter ground of the Midland Valley. This change in topography is not, however, generally clear-cut across the fault, as Carboniferous and Permian sedimentary rocks infiltrate into the Southern Uplands along the Nith valley, as described above, whilst igneous rocks are relatively common just north of the Southern Uplands Fault within the Midland Valley and, as described later, have often resisted erosion to form hills comparable to those found just south of the fault.

Glacial landscape development

Whilst the broad outlines of the present Scottish landscape had probably been established by the end of the Tertiary, its detailed configuration owes much to events of the Quaternary period. During the last million years, ice sheets have repeatedly expanded to cover much of Scotland, including Area 1. These ice sheets flowed radially outwards from centres in the Highlands and Southern Uplands and were powerful agents of erosion and deposition, moulding the uplands, scraping sediments from the lowlands and locally depositing great thicknesses of boulder clay (till).

The most recent glacial episode, the Devensian, reached its coldest about 25,000 to 20,000 years ago, when an ice sheet centred on the Western Highlands and Southern Uplands had expanded to cover most of Scotland and all of Area 1. The broad pattern of ice flow during this time is shown in Figure 53 (#ulink_4e497bf7-251f-5d19-940c-bea67c775b74): the thickest ice was centred on the Southern Uplands, and it flowed radially outwards from an ice divide that extended from Merrick in the west to the Lowther Hills in the east. Ice flowing northwards into the Midland Valley came up against southwards-flowing Highland ice, forcing ice to flow east and west across the low ground of central Scotland.

Landscape modification by glacial erosion

Glacial erosion has played an important role in creating the final shape of the landscape seen today. Most of Area 1 was extensively ice-scoured throughout the course of the Pleistocene glaciations, and the land surface present at the end of the Tertiary became heavily modified. Glacial erosion in this Area is most obvious in the uplands, which have been extensively ice-scoured. The mountains around the Loch Doon and Carsphairn igneous intrusions have an, albeit very rounded, Alpine form, with corries, rounded arГЄtes and intervening glacial troughs. These large-scale landforms were produced over the course of multiple glaciations, in particular by local valley and cirque glaciers during early and late phases of glaciation. The intensity of glacial erosion, at least during the Devensian glaciation, decreased eastwards towards the Lowther Hills, where more localised erosion took place: powerful ice streams continued to deepen the main valleys, but the intervening ridges and plateau were relatively unmodified. The corries that are relatively common in this southwestern part of the Southern Uplands are largely absent from the northeastern Southern Uplands. This could reflect a difference in climate from west to east, with the glaciers of the warmer, wetter southwest flowing much more vigorously, and hence eroding more than those of the colder, drier northeast, where the ice was frozen directly onto the rock and was therefore unable to scour deeply.






FIG 53. Generalised map of ice flow during the Devensian.

It is often the case in Scotland that, where granite intrusions are present in the bedrock, they are visible in the landscape because they form topographic highs. The relationship does not, however, seem to be so clear-cut in Galloway, where the different granite bodies have responded differently to both Tertiary and glacial erosion. The Cairnsmore of Fleet and Dalbeattie bodies do seem to be loosely associated with topographic highs: round, smoothed hills in the case of the Fleet body, and some of the only elevated ground on the south coast in the case of the Dalbeattie body (Figs 47 (#ulink_5f5e697c-fb37-5248-acae-75ecf871e10e), 51 (#ulink_0038c7b5-86de-59b0-86a7-9ea4660af799)). However, in both cases only part of the intrusion seems to have resisted erosion to form elevated ground: the western half of the Fleet body and the southeastern margin of the Dalbeattie. Elsewhere, the elevation of the land is not discernibly different to that underlain by the surrounding Palaeozoic metamorphics. However, a contrast is seen in the �texture’ of the land, with those areas underlain by granite or granodiorite having a much more �smoothed’ appearance, probably reflecting the more uniform nature of the rock, and its response to erosion, than the surrounding folded Silurian rocks. Likewise, the Carsphairn intrusion underlies the large hill of Cairnsmore of Carsphairn, with its knock-and-lochan topography and craggy faces. Again, however, there is no change in topography at the contact between the pluton and the surrounding rock, and the area north and east is almost equally as mountainous.

Something very different is associated with the Loch Doon granite. This area formed the centre of accumulation for the local icecap during the Devensian glaciation, and likely earlier, with glaciers moving outwards onto lower ground with an approximately radial pattern of flow. As such, it has been subject to intense glacial erosion, both under an extensive icecap during the glacial maximum and by local valley glaciers during early and late phases of glaciation. It is obvious from Figure 52 (#ulink_7b1f5143-141b-5238-bf11-4293cff1d9a7) that the Loch Doon pluton itself has resisted this erosion much less than the baked Ordovician sediments which surround it. These tough hornfels (baked sediments) today underlie the distinctive elevated ridge of peaks which almost completely surrounds the Loch Doon pluton. These hills include the Rhins of Kells, with Corserine (814 m) on the eastern flank and Merrick on the western flank (Fig. 51 (#ulink_0038c7b5-86de-59b0-86a7-9ea4660af799)), which at 843 m elevation is the highest point in Scotland south of the Highlands. The view from the summit of Merrick is exceptional, from Ben Cruachan northeast of Oban, across to the Paps of Jura and then south across the Isle of Man and the Lake District to Snowdonia. The granodiorite which makes up much of the Loch Doon pluton has a tendency to form the low boggy ground between these hills, averaging around 300 m elevation. Glacial rock basins have been gouged out of this granodiorite, and today they are flooded to form the numerous small lochs which are seen within the boundaries of the intrusion, such as Loch Enoch. Meanwhile, the granite which forms the centre of the intrusion is obviously more resistant than the granodiorite, and underlies a ridge of high ground including the hills of Craiglee, Hoodens and Mulwarchar (692 m).

Throughout these uplands, bare, scoured rock is relatively common, particularly around the Loch Doon and Carsphairn area, although the jagged peaks and cliffs common in the Highlands are mostly absent. Where present, this bare rock provides an interesting contrast to the otherwise rolling moorland. One particular craggy outcrop southwest of Loch Enoch has been named the �Grey Man of Merrick’, because of its resemblance to a man’s face when seen from the side. Another interesting feature is the so-called �Devil’s Bowling Green’ on Craignaw, a remarkably flat, smooth glaciated rock surface strewn with rounded boulders.

Elsewhere in Area 1, the ground is much lower-lying, and the effects of glacial erosion are not so obvious. This ground, already low at the end of the Tertiary, has been further lowered and smoothed by the passage of ice. Where hills are present, they tend to have been rounded by ice scouring, and often have a streamlined shape. An excellent example is the island of Ailsa Craig (Fig. 50 (#ulink_aa7379e2-a84e-5957-8cf4-007e3146be66)). Although the island has been considerably modified by subsequent marine action (which produced its precipitous cliffs, discussed below), its overall shape is round, but elongated from north to south, reflecting southerly ice flow. The granite has clearly resisted erosion much more effectively than the soft Permo-Triassic sandstones into which it was intruded. Ailsa Craig was positioned in the path of many different ice streams during the Devensian glaciation, and these ice streams carried blocks of Ailsa Craig microgranite in the direction they were flowing. As the microgranite has a very distinctive composition, these blocks are easy to identify, and they have proved very useful in tracing flow directions of the last ice sheet across Britain. Indeed, the Ailsa Craig boulder train is one of the most famous and largest in the British Isles, extending south across the Irish Sea and parts of England and Wales as far as Pembroke, and westwards to Ireland.

North of the Southern Uplands Fault, the relatively soft sedimentary bedrock of Devonian to Permian age has generally been heavily weathered and eroded by the passage of ice, forming a low-lying landscape. However, the sedimentary units are punctuated by horizons of lava and numerous plugs of fine-grained igneous rock, only the largest of which are shown in Figure 47 (#ulink_5f5e697c-fb37-5248-acae-75ecf871e10e). As a result, the generally low-lying landscape is frequently interrupted by rounded hills, underlain by the more resistant units. This is well illustrated in the Straiton area (20 km northeast of Girvan), where the craggy hill tops of Bennan Hill and Craig Hill are underlain by more resistant Devonian lavas, with Devonian sediments underlying the gentler slopes of the surrounding area. Likewise, Mochrum Hill (Fig. 52 (#ulink_7b1f5143-141b-5238-bf11-4293cff1d9a7)), near Maybole, is underlain by the eroded remains of a large Devonian volcanic vent, around 1 km in diameter. The vent is filled with agglomerate (coarse angular blocks of volcanic material), which has resisted erosion to form the prominent, rounded hill, whilst the surrounding Lower Old Red Sandstone is much softer and lower-lying. The sandstone in this area is feldspar-rich, and has weathered to produce particularly fine arable soils. Younger volcanic rocks are also common, such as the Permian, agglomerate-filled volcanic neck underlying Patna Hill, just northeast of Patna. There are more than 20 such vents in the Patna–Dalmellington area, many of which are responsible for small topographic features. As these intrusions are often basaltic (mafic), they have weathered to produce nutrient-rich soils, the so-called �Green Hills’ of Ayrshire.

In places, prominent hard bands within the sedimentary rocks are also associated with rounded, glacially scoured hills. An example is the �Big Hill of the Baing’ southeast of Straiton (20 km northeast of Girvan), an elongated, faulted ridge of Ordovician boulder conglomerate. More extensive outcrops of this conglomerate occur in the Girvan–Ballantrae area, where, along with the Ballantrae Complex, they underlie higher, hillier ground than the softer rocks further south.

Landscape modification by glacial deposition

Much of Area 1 is relatively low-lying, and here the effects of glacial erosion are more subtle than in the high ground of the Southern Uplands: the ground level was lowered, pre-existing Tertiary valleys were deepened and the low hills were moulded and streamlined. Equally important in the formation of today’s landscape in these lowland areas was glacial deposition: on deglaciation, great thicknesses of till were deposited and today glacial till, sand and gravel mantles much of the lowlands. These deposits have a range of surface forms, including eskers, kames, outwash terraces and, in particular, drumlins.

Drumlin swarms are important landscape features throughout the lowlands of this Area, tending to broadly correspond with the arrows on Figure 53 (#ulink_4e497bf7-251f-5d19-940c-bea67c775b74). They mantle much of the Rhins of Galloway, the Machars, the Glenluce, Ballantrae and Girvan districts and Nithsdale. They also make up much of the land surface of the Midland Valley, being responsible for the rather intriguing, �hummocky’ texture that is so characteristic when viewed from the air, or on a simple hill-shade map (Fig. 52 (#ulink_7b1f5143-141b-5238-bf11-4293cff1d9a7)). The drumlin swarms in these areas produce a distinctive landscape of low hills, typically around 30 m high and 300 m long, all oriented in the same direction and with similar shapes – blunt at one end and tapered at the other, rather like an egg. This streamlined shape is produced by deposition at the base of a flowing glacier: drumlins often have a core of rock or glacial till, and as sediment-laden ice flows over these obstructions, material is deposited downstream of the core, where it is relatively sheltered from ice erosion. As this process repeats itself, a streamlined mound is gradually produced, with a tapered end pointing downstream and a blunt end pointing upstream. One is aware of the whaleback shape of the drumlins that make up these swarms from the ground, but an aerial view allows the best appreciation of their three-dimensional streamlined form. Excellent examples are seen, for example, around Newton Stewart and in the New Galloway district. Smaller swarms are also present in the uplands.

The broad Carsphairn Valley cuts across some of the highest ground of the Southern Uplands, with the Loch Doon hills to the southwest and the Cairnsmore hills to the northeast. Reconstructions of former ice-flow directions in the valley indicate that, during the Late Glacial Maximum, a northeast/southwest ice divide was located across its central part, passing from Cairnsgarroch summit through Craig of Knockgray to the Cairnsmore Hills. The thickest and most extensive till deposits present in the Carsphairn Valley are found around the area of this ice divide, which seems somewhat contradictory. Horizontal ice flow is minimal at ice divides, and the till cannot therefore have been deposited when the divide existed, so the source and age of this till is an interesting question. The answer seems to be that the till was deposited during or before the glacial maximum, during the growth of the Late Devensian ice sheet. At the start of the Late Devensian glaciation, ice would have initially accumulated in the corries and trough heads northeast and southwest of the Carsphairn Valley. As the glaciation advanced, these glaciers expanded and finally converged in the valley bottom, and as their flow was impeded till would have been deposited. During the subsequent glacial maximum, the preservation of this till beneath great thicknesses of ice is likely due to its location under the ice divide, as although the ice sheet expanded and thickened, the slow rates of ice movement meant the ice had little erosive power here.

This till, deposited during the growth of the Late Devensian ice sheet and preserved under the ice during the glacial maximum, was then remoulded during a late stage of glaciation into a set of interesting landforms – rogen, or ribbed, moraine, which consists of sinuous, 20 m-high, elongated ridges that run perpendicular to the valley axis. The mechanism by which these till ridges formed, perpendicular to the down-valley direction of Devensian ice flow, is another interesting point. A likely scenario is that the rogen moraines represent ridges of sediment produced by thrusting (by compression) or fracturing (by extension) at the base of the ice sheet. For this to happen, the flow speed of the lower part of the ice must have varied downstream: a sudden speeding-up would produce fracturing by extension; a sudden slowing-down, such as upstream of an obstacle, would produce thrusting by compression. This would have been most likely to happen during a late stage of ice-sheet deglaciation, when faster, more concentrated flow occurred within the main valleys. The most recent episode recorded by this till involved the drumlinisation of the rogen moraine, as the original landforms became elongated down-valley to varying degrees.

Important amounts of sediment were also deposited by sediment-charged meltwaters flowing out from retreating glaciers, referred to as glaciofluvial deposits, and present in a variety of forms. Sediment may accumulate in channels, ponds and lakes trapped between lobes of glacier ice or between a glacier and the valley side. Where such sediment has a ridge or mound form, it is termed a kame; where it is a flat-topped mound, it is termed a kame terrace, and is likely to have been deposited in a lake. When sub-glacial meltwater drains through tunnels, the sub-glacial stream may deposit sediment as a surface stream would, but confined to the tunnel. The result is a long sinuous ridge of gravel, termed an esker. Outwash plains often build up downstream of the melting glacier, large plains of poorly sorted, stratified sediment deposited by braided streams. Sequences of terraces are often seen in these plains, formed by river incision. Kettle holes are also common features, formed by blocks of ice that become buried in outwash sediment, and then melt to leave behind a depression. Many of these kettles have been infilled with sediments, particularly peat, during the post-glacial times, but some are still visible today as small isolated lakes or deep water-filled depressions in boggy areas that were once the low-lying outwash plains.

A famous example of �kame-and-kettle topography’ is found in Nithsdale, north of Dumfries. Particularly on the eastern side of the valley, there are many short, linear glaciofluvial ridges separated by depressions and hollows. The relative relief between ridge crest and depression is usually between 8 and 25 m, and the ridges are relatively short, with very few being over 500 m long. The extensive gravel pit at Kilblane, for example, is developed in three such kame ridges. The coarse sediment that makes up these ridges, and other linear kame ridges in this part of Nithsdale, was probably deposited in meltwater channels that flowed between ice-cored ridges parallel to the ice margin. As the ice ridges melted, the sediment-filled channels became inverted to produce the kames seen today. Kame terraces are also seen on both sides of the Nith, with the best developed just east of Duncow (8 km north of Dumfries) at an altitude of around 55 m. Further north, in the mid-part of the Nith valley (south of Thornhill), a similar kame-and-kettle topography is seen. The glaciofluvial deposits of Nithsdale account for the rather large number of sand and gravel pits seen just north of Dumfries, now often flooded. Glaciofluvial deposits are relatively common elsewhere in Areas 1 and 2, such as in many of the valleys on the south side of the Southern Uplands and in the area around Stranraer.

The mapping of glaciofluvial deposits in the Nith Valley has allowed the reconstruction of the pattern of glaciofluvial drainage which developed during a late stage of deglaciation. The result shows a narrow marginal zone of ice-cored ridges and troughs in the north, feeding meltwater and sediments to the ice front north of Dumfries. This ice front is marked by a terminal moraine across the Nith valley, which is crossed by the River Nith in a gorge in Dumfries. Further southeast, the drainage fed outwash systems in the Lochar Water and Nith valleys. Today, this outwash plain underlies much of the uniformly flat surface southwest of Dumfries.

A closer look at Late Devensian ice-flow directions

The broad ice-flow pattern shown in Figure 53 (#ulink_4e497bf7-251f-5d19-940c-bea67c775b74) is useful, but presents a highly generalised picture; in reality, ice-flow directions over the course of the Late Devensian were somewhat more complicated. The large number of streamlined glacial deposits found throughout Area 1, particularly in the lowlands, has allowed a much more detailed reconstruction of Devensian ice flow. A recent study looked in detail at the glacial features present in the western part of Area 1, in particular at drumlins, erratic trains and glacial striae. It was found that several generations of these features can often be seen superimposed on one another, recording multiple passages of ice from different ice centres. These changing flow directions are summarised in Figure 54 (#ulink_02913ac2-f5e9-515a-a5aa-d79b9c21ae51), and record the changing relative strengths of the Southern Uplands and Highlands ice centres.

Some of the earliest features in the western Southern Uplands indicate that ice from the Highlands was initially dominant during the Late Devensian, when it streamed southwards from the Firth of Clyde and crossed the Glenluce lowlands, producing north/south lineations. This Highland ice was then replaced over much of Area 1 by Southern Uplands ice, as shown by a southwest-oriented flow set running across Glenluce and the southern Rhins. A similar story is recorded by till deposits in the southern Midland Valley, around the margins of the Southern Uplands. For example, a vertical section cut by the River Nith at Nith Bridge, just south of Cumnock, reveals three tills deposited during the Late Devensian and separated by glaciofluvial sands and gravels. These tills have been carefully studied, and the bottom two were both found to have been deposited by Highland ice, which probably flowed across central Ayrshire from the Firth of Clyde area. The topmost till, by contrast, was deposited by ice originating in the Southern Uplands. There were, therefore, at least two distinctive phases of ice movement across central Ayrshire, with an initial advance of Highland ice being succeeded by Southern Uplands ice. The evidence at Nith Bridge matches similar evidence found across the southern Midland Valley, and the story indicated by streamlined landforms further southwest. It seems, therefore, that Highland ice initially expanded to encroach on the Southern Uplands, and that it was only as glaciation progressed that Southern Uplands ice became more dominant in Area 1. Further south, another major ice centre was established in the Lake District, and converging drift lineations at the tip of the Machars peninsula mark the confluence of this and Southern Uplands ice.

At the coldest stage of the last glaciation, around 20,000 years ago, Highland ice one again played a role – an ice stream flowed out from the Highlands and along the western seaboard of Area 1. As the ice moved down the Firth of Clyde, it scraped marine deposits off the sea bed and re-deposited them further south. These shelly deposits are found, for example, on top of a 10 m-high shore platform around the Mull of Galloway. The Highlands ice sheet also brought glacial erratics of the distinctive Ailsa Craig microgranite and Arran granite southwards, found today throughout the Rhins peninsula.






FIG 54. More detailed examination of local ice-flow directions during the Devensian (LGM is the Late Glacial Maximum that occurred late in the Devensian). After Salt and Evans, 2004

Following the glacial maximum, climate began to warm, and both the Southern Uplands and Highland ice centres contracted. Some time after Highland ice had retreated from the western seaboard, a phase of local ice expansion interrupted the general waning, and Southern Uplands ice once again flowed across the western part of Area 1. For much of the western, lowland parts of this Area, this would be the last time they were ice-covered, and the ice sheet left behind the extensive drumlin swarms described above. This Southern Uplands ice flowed southwestwards across the Rhins of Galloway, bringing with it erratics of Loch Doon granite. It also flowed roughly westwards across the Ballantrae and Girvan districts. In the Ballantrae area, flow was somewhat valley-contained, showing that the ice sheet was much thinner than it had been during the glacial maximum. Further north around Girvan, the lineations are particularly notable for their cross-cutting relationships, probably produced by the slight shifting in the main ice-dispersal centre with time.

The most recent flow set was also produced during the waning of the Southern Uplands ice sheet, again during a minor re-advance. This time, valley glaciers radiated out from a Southern Uplands dispersal centre located around Merrick, down valleys such as Nithsdale, Glenluce and northwards into the Midland Valley. Again, erratic trains from distinct granite outcrops around the Galloway area provide useful trackers, along with moraine ridges and drumlins. A local surge of Highland ice down Loch Ryan also occurred at this time, and a prominent moraine at the head of Loch Ryan (the Stranraer moraine) marks the outer limit of this re-advance.

Pollen and beetle records indicate that temperatures in Area 1 may have risen to as warm as present by around 13,000 years ago, by which time southwest Scotland must have been completely deglaciated. Temperatures then fell sharply around 12,000 years ago, culminating in the Loch Lomond Stadial. This climatic deterioration was accompanied by the return of glaciers to parts of the Southern Uplands, although these glaciers were of very limited extent, generally being confined to the highest corries. Where glaciers developed, they bulldozed earlier till deposits into moraine ridges. A fine example of one such moraine is seen at Loch Dungeon, just southeast of Corserine. Steep cliffs of Silurian sediments rise from the southeast shore of the loch, and a subsidiary corrie of Corserine opens out on the northwest shore. A glacier emerged from this corrie, and its terminus is marked by a large terminal moraine to the west of the loch and by a shallow area within the loch itself.

Even in unglaciated upland areas, ice growth often caused extensive frost shattering. The scree and loose rock this produced is still visible today, particularly on summits and upper slopes, and has often been modified by subsequent flow to form a series of lobes and sheets. Elsewhere in the lowlands, the cold climate of the Loch Lomond Stadial made itself felt through the development of permafrost, as evidenced today by features such as ice wedge casts, seen most commonly in gravel pits. Evidence for periglacial disturbance and movement of the soil (solifluction) is also widespread on lowland slopes, usually affecting 1–2 m depth of soil, though in the valley floors of the Southern Uplands, great thicknesses of solifluction deposits have accumulated.

Post-glacial landscape development

At the end of the Loch Lomond Stadial, a temperature rise of around 7 В°C occurred within just 700 years, marking the start of the current Flandrian (or Holocene) period. Although the effects of glaciation still dominate much of the landscape, in the 10,000 years since the disappearance of the last glaciers the land surface has been slowly adjusting to non-glacial conditions. These changes are particularly evident in areas of high relief, where glacial retreat exposed a bare rock landscape with over-steepened slopes. Soon after deglaciation, this landscape began adjusting to the new conditions, with rock falls, debris flows and reworking of glacial sediments. As the landscape re-equilibrated and soils and stabilising vegetation became established, it seems that these processes almost stopped, as shown by the vegetated, relict nature of most of the talus slopes, debris cones and alluvial fans in this Area.

Today, Area 1 is notable for its variety of river types and sizes, reflecting contrasts in relief and catchment size throughout the Area. Most of the main rivers originate in the high ground of the Southern Uplands and drain southwards, including the Nith, Cree and Dee (Fig. 51 (#ulink_0038c7b5-86de-59b0-86a7-9ea4660af799)). The upland tributaries of these rivers are akin to mountain torrents, becoming wandering gravel-bed rivers in their middle reaches as the relief becomes more subdued. Most of the major rivers have highly sinuous, meandering courses in their lower reaches, and drain into silty estuaries in the Solway Firth (Fig. 55 (#ulink_4cf1a00e-33ae-5a42-93e7-7b3926fd97d8)). The River Ayr, in the Midland Valley, has one of the most meandering courses in this Area, as it weaves across a flat lowland strewn with glacial deposits. Many of the lower reaches of the watercourses in this Area have been embanked to prevent flooding of adjacent land, and some of the smaller rivers show signs of having been straightened in the past.

The first vegetation to become established early in the Holocene was a juniper-dominated community, followed by birch and hazel around 9000 years ago, and then by oak and elm during the middle Holocene. Pine forest was present in the Galloway Hills, but was never the dominant species in this Area. From around 5000 years ago, human activity first began to have a significant impact on the landscape, primarily through forest clearance to make way for agriculture. There is evidence that woodland began to be progressively replaced by peat around 5000 years ago, and that by around 4200 years ago forest cover had essentially disappeared from the Area, replaced by blanket mire. This deforestation is thought to have led to enhanced soil erosion, with an increase in slope failure, debris flow activity and river incision. Peat has been the most widespread soil type in Area 1 since around 4000 years ago, in the form of blanket bog (including the internationally important Silver Flowe Bog in the low ground of the Loch Doon intrusion), or drier heather-covered slopes.

More recently, damming is another way in which humans have significantly altered parts of Area 1, flooding valleys to create reservoirs. The Galloway hydroelectric scheme was built between 1930 and 1936, and was the first of its kind in Scotland. Although small compared to some of the later Highland schemes, it is a model of unobtrusive and ecologically sensitive hydroelectric engineering, and is studied by engineers from around the world. Making use of water principally stored in Loch Doon, Clatteringshaws Loch and Loch Ken, the scheme includes eight dams, 12 km of tunnels, aqueducts and pipelines together with six power stations along 130 km of river. Whilst Loch Doon is on the site of a natural loch, damming has increased the water level by some 9 m, submerging various small islands. Before the Loch Doon dam was built, Loch Doon Castle, a thirteenth-century castle originally located on an island in the centre of the loch, was moved, stone by stone, to the adjacent bank where it now stands. Elsewhere, dam building flooded valleys, thereby significantly altering the landscape. Such reservoirs include Clatteringshaws Loch and Loch Ken on the Water of Ken. Loch Ken is now a major nature reserve and a breeding ground for many varieties of wild birds.






FIG 55. Tidal marshes showing typical, highly sinuous channels, where the wavelength of the channels cut in the muddy sediments reflects the tidal discharges involved. (В© Patricia & Angus Macdonald/Aerographica/Scottish Natural Heritage)

Man has also altered the landscape by mining and quarrying activities, particularly prevalent in the Midland Valley where large opencast coal workings are still operational today in places, such as east of Patna. Recent clean-up efforts have greatly reduced the impact of colliery tips (bings) on the landscape, such as northeast of Girvan, where they have been landscaped and forested. Local stone has also been quarried for building stone, roadstone and crushed rock aggregate. An important source of building stone is the area around Mauchline, from which the attractive orange-red Permian sandstone has been extracted. Stone from this area has been widely used throughout the UK and Ireland, and even shipped to the USA. The granites have also been economically important for the region – the Glasgow and Liverpool docks, for example, were constructed using Criffel–Dalbeattie granite. Glacial sands, silts and gravels have also frequently been quarried, often leaving their mark on the landscape with flooded gravel pits.

Today, the Area is generally very wet and mild, as the North Atlantic Drift maintains higher temperatures than those found on the east coast. Indeed, plants normally associated with more southerly latitudes are found on the Rhins, along with dolphins and basking sharks off the coast. Much of the Southern Uplands in this Area lie within the Galloway Forest Park, managed by the Forestry Commission, and as well as rolling moorland, conifer plantations are common on the shallow, poor soils. The main river valleys (such as the Urr, Dee, Cree and Nith) provide a contrast to this rolling moorland, providing much of the good arable land of the Southern Uplands.

The formation of the coastline

As described in Chapter 5 (#u661596f2-c933-517a-abc9-8c82715c3114), sea level in much of Scotland has not been constant, but has risen and fallen according to the interplay between global sea level and the elevation of the land. Global sea level decreases during an ice age, as water is locked up within the ice, and rises again as this ice melts. Meanwhile, during glacials, the crust becomes depressed locally under ice sheets, sinking into the mantle, and slowly rebounds once this ice has melted. The sea level at the coast at any one time therefore depends on the interplay between these two effects, and in the past sea level in Area 1 has been both higher and lower than at present. When sea level has remained constant for long enough, shorelines formed – marked today by erosional features such as rock-cut platforms backed by cliffs, or by beach deposits. Where subsequent uplift has outstripped global sea-level rise, these shorelines now take the form of raised beaches, raised deltas, raised estuarine deposits (known in Scotland as carse), and raised rock platforms and cliffs. Good examples of these features are found in this Area. Shore platforms are relatively common on stretches of rocky coastline, and are particularly well developed, for example, between the Heads of Ayr and Turnberry and around Ballantrae. In the latter area, numerous small caves and gullies often delineate the foot of the cliff, and occasional raised sea stacks rest on the raised platform. Further south, a shore platform can be traced around most of the Mull of Galloway at approximately 10 m above present sea level, and deposits of glacial till on top of it show that it predates the Devensian glaciation. In places, several shorelines are present, such as some 8 km southwest of Girvan, where two old shorelines give the skyline a stepped profile. These shorelines occur as a series of benches, commonly cut into boulder clay. Mafic, dolerite dykes have been etched out by the sea during the formation of each of these shorelines, and today stand as raised sea stacks.

In the lower-lying coastal areas, past increases in sea level involved the flooding of sometimes large patches of land. For example, along the Ayrshire coast the late glacial sea was nearly 30 m higher than at present, and so the sea would have come several kilometres inland from its present position. Along the Solway coast in the south of the Area, large sections of the present shore are backed by raised estuarine deposits of silt and clay, which provide valuable records of this sea-level change over the last 15,000 years. Good examples occur flanking the heads of the Cree and Fleet estuaries, which in this area lie between 7 and 10 m above present sea level.

The sediments within cores taken from the carselands flanking the Cree estuary have been analysed and dated, and have proved very useful in reconstructing the sea-level history of the Solway Firth, as summarised in Figure 56 (#ulink_8bebc753-f365-59ed-bb54-d07aa50aaf50). These deposits show that sea level rose to cover this area by around 9600 years ago, followed by a rise to the so-called Main Post-glacial Shoreline by 6500 years ago, when sea levels were between 7 and 10 m above present and some 9 m of estuarine deposits were laid down in the Cree area. Sea level then fell from the uppermost carse surface to its present level as the land continued to rebound, whilst worldwide sea volume changed very little.

The large-scale shape of the coastline is controlled by a number of factors, many of which are in turn related to one another. Rising sea levels cause valleys to become flooded, forming bays and islands, whilst the location of these valleys often reflects rock type (i.e. hardness) and structure (both the presence of jointing within a rock unit and lines of weakness, such as faults). Climate and tidal energy also play an important role, controlling wave energy environments and terrestrial processes, such as sediment supply. The amount of sediment supply to a coast of course depends on the availability of that sediment, and in this regard glaciation has been very important. At the end of the Devensian, great volumes of glacial debris were deposited on the continental shelf. As sea level recovered to present levels, this sediment was reworked and moved towards the shore, to form the basis of our present beach and sand-dune systems. Today, where sediment supply is abundant, the coastline is currently advancing seawards, whilst in areas where sediment supply is in decline, the coastline is usually retreating through erosion. Humans, also, can have an impact – for example through building, quarrying, constructing sea defences and trampling of stabilising vegetation.






FIG 56. Sea-level curve for the Solway Firth. (Data from Smith et al. 2003, Transactions of the Royal Society of Edinburgh: Earth Sciences, 93, 301–31)

The number of large bays is one of the more obvious features of the coastal strip of Area 1. The main ones are Loch Ryan and Luce Bay, which together define the Rhins of Galloway, and Wigtown Bay further east. Numerous smaller bays are present along the south coast, including the Water of Fleet and Kirkcudbright Bay on the southeast side of Wigtown Bay. In general, these bays are located in the lower-lying ground of this Area, and therefore do not reflect a large difference in rock strength between headland and bay. Instead, they seem to roughly coincide with the large northwest/southeast faults shown in Figure 47 (#ulink_5f5e697c-fb37-5248-acae-75ecf871e10e), and with the outlets of major rivers. The exception is the Loch Ryan–Luce Bay pair, which may have been the site of a Tertiary river, now partially flooded by the sea.

Igneous rocks underlie some of the more prominent headlands on the west coast of Area 1, such as the Carboniferous vent rocks that make up the Heads of Ayr, a prominent headland some 6 km southwest of Ayr. The vent has been intruded into a colourful mix of Early Carboniferous sediments, including limestone, grey-green shales and red and green sandstones. Ash erupted from the Heads of Ayr vent is thought to underlie the knoll on which Greenan Castle has been built, around a kilometre up the shore towards Ayr. Where Silurian and Ordovician sedimentary rocks have been cut by cliff sections, the result is often rather impressive because of the exposed folding within the bedrock. Particularly spectacular �textbook’ examples of rock folds and other structures are found on the Machars peninsula, for example at Back Bay, just south of Monreith on the west coast, and around the Isle of Whithorn in the southeast. The fold structures at Back Bay not only illustrate two fold generations with a second set of folds superimposed on a first, but represent one of the most dramatic large-scale exposures of major re-folded folds in the UK.

Perhaps some of the most dramatic coastal scenery in Area 1 is found on the island of Ailsa Craig (Fig. 50 (#ulink_aa7379e2-a84e-5957-8cf4-007e3146be66)). Despite being only 1.2 km wide, the island is nearly surrounded by 340 m-high cliffs. The height of these cliffs reflects the exposed nature of the island to marine erosion, the strength and resistance of the bedrock to this erosion, and also the topographic high left behind here after the retreat of the last glaciers. For the most part, the foot of these cliffs is now between 5 and 10 m above sea level, and so marks a raised shoreline. A distinctive triangular raised beach is located on the eastern side of the island, fringed by storm ridges and an associated spit. These landform features highlight the importance of prevailing wind and storm direction, as they were caused by the westerly winds and waves since post-glacial times: sediment is deposited on the sheltered side of the island. The importance of exposure is also seen on the Rhins of Galloway, where the exposed western coast is generally rugged with steep cliffs and occasional inlets, in contrast to the calmer eastern coast with its sandy beaches.

South of the Southern Uplands Fault, the coast is generally rocky, with low cliffs and only small local beaches or small stretches of shingle. Cliffs are particularly well developed along the exposed western coast of the Rhins, reaching a maximum of 120 m in height near Dunman. This cliff line is largely inactive now, having been raised clear of wave action by crustal uplift. On the north coast of the Solway Firth in particular, the rocky coastline is punctuated by a number of large bays, generally river estuaries. The more sheltered conditions within these bays and the ample sediment supply have allowed wide expanses of sand-flat, mud-flat and salt-marsh to accumulate, and together the flats and marshes of the Solway Firth provide one of the largest continuous areas of intertidal habitat in Britain. Remnants of formerly more extensive lowland peat bogs are developed on raised estuarine sediments. These include the nationally important Lochar Moss and Moss of Cree.

Beaches are generally more common and extensive in this Area than in the Highlands of Scotland. The main reason for this is the relative abundance of sediment, primarily glacial sediment laid down on the near-shore shelf during the Devensian glaciation, and then driven onshore during the Flandrian sea-level rise, where it became stranded as sea levels fell again across the region. The relatively mild winds and waves experienced by the lowlands then meant that this abundant sediment source has remained fairly stable, and many of the beaches in this Area are still accreting today, rather than eroding. Old dune deposits are relatively common on the coastal strip south of Troon on the north edge of the Area, and sands were formerly worked from pits in these dune sands northwest of Monkton. Now, most of the dune deposits on the coast around Ayr are covered by golf courses.

Large expanses of tidal sand-flats are found along the southern coast of Area 1, and at low tide many kilometres of sand are exposed, such as at Mersehead Sands (20 km south of Dumfries). Further inland, the reworking of raised beach deposits and other sandy sediments by the wind has, in some coastal locations, produced extensive spreads of sand dunes, now for the most part anchored by coarse grass or forestry. The largest beach-dune system in southwest Scotland is found at the head of Luce Bay, home to a complex array of dune-related landforms, and still actively accreting today. The entire peninsula of the Rhins acts as a huge breakwater from the currents of the North Channel, creating the relatively calm waters of Loch Ryan and Luce Bay.

Salt-marshes are typically developed on low, raised beaches of sand or shingle and display a complex topography of pans, creeks and terraces. The salt-marshes of the Cree estuary in Wigtown Bay are particularly well developed, sandwiched between extensive sand-flats seawards and reed-swamps and emerged estuarine deposits (carse) landwards. The extensive carse deposits of the Cree estuary show that sedimentation has prevailed here over most of the last 10,000 years, despite changing sea levels. The estuary is well sheltered by Burrow Head to the south, and this has produced a largely unidirectional wave climate in which sediment is brought into the bay, with little subsequent removal. It appears that this system still operates, since many of the sand-flat and salt-marsh systems are accreting today. The presence of Sellafield-derived radionuclides attached to the sediment in the Cree and Water of Fleet sandbanks confirms the Outer Solway as a major sediment source, whilst important amounts of mud within the Cree mouth itself suggest a more fluvial source, further enhanced by active reworking of sediment from the carse deposits.


CHAPTER 7

Area 2: Southern Borders

AREA 2 COVERS THE CENTRAL PART of the Southern Uplands, sandwiched between Areas 1 (Galloway) and 5 (Edinburgh) (Fig. 57 (#ulink_47424c5b-7d09-59f4-b9f5-ee0eac35342b)). It straddles the southern part of the English–Scottish border, extending from Annandale and Moffatdale in the west to the Cheviot Hills in the east (Fig. 58 (#ulink_5dc1ce89-5134-5a8f-a2e3-be76ef0fd668)). The discussion provided here is restricted to features on the Scottish side of the border.

In the northern half of the Area, the River Teviot drains the relatively low-lying ground of Teviotdale, which in turn feeds into the larger Tweed Basin of Area 5 (Chapter 2 (#uc0e147ab-0db8-5cd3-a46c-c03cd0831740), Fig. 7 (#ulink_7473398e-e912-5841-b9c0-27bb75749122)). Teviotdale is bounded to the east by the Cheviot Hills and to the west by the Moffat Hills of the Southern Uplands. These rounded hills reach heights of over 800 m, and form a broad region of elevated ground which dominates the landscape in the northwestern part of this Area. The Solway Firth with its adjacent low-lying plains is the main feature of the southwestern corner of Area 2, along with the broad valleys of the Annan and Esk, which penetrate northwards into the hills of the Southern Uplands.






FIG 57. Location map for Area 2.






FIG 58. Natural and man-made features of Area 2.

STORIES FROM THE BEDROCK

Area 2 is principally made up of hard Ordovician and Silurian sediments, which are also the oldest rocks in the Area (Figs 59 (#ulink_52db03e9-d893-5ca4-a784-de798678b239), 60 (#ulink_c79c2c08-ea50-54c2-abd6-817f62198f72)). These sandstones (greywackes), siltstones and shales were deposited between around 490 and 420 million years ago on the floor of the extensive Iapetus Ocean that, during this time, formed an Atlantic-scale ocean with Scotland and England on opposite margins. Around 490 million years ago, this ocean began to become smaller, as oceanic crust became subducted beneath the Grampian Highlands, and the Ordovician and Silurian sediments were scraped off the ocean floor and stacked up in a pile against the Midland Valley (Fig. 28 (#ulink_97ce30e1-e360-557b-93ae-e0a8135abc77)). This deformation created a fold system across the uplands of this Area, with the fold axes trending northeast to southwest. Superimposed on this larger structure are countless minor parallel folds, which give a �grain’ to the oldest bedrock of the Area. During these movements, the sediments were altered to brittle sandstones and slates.






FIG 59. Simplified geology and hill-shaded topography for Area 2.

By the Devonian (around 410 million years ago), the major uplift and deformation of the Southern Uplands had ceased and the crustal materials of what are now Scotland and England became �welded’ along the Iapetus Suture. The mountainous ground of the Southern Uplands then began to be eroded, and large rivers flowed across the Area, depositing sandstones, siltstones and occasional conglomerates on lower-lying plains fringing the mountains. This Area lay in the interior of a large continent, and sediments accumulated on river flood plains where oxidising conditions led to the deposits being stained with red iron oxide. These Devonian sediments were deposited on the eroded surface of older, deformed Silurian and Ordovician rocks. This surface was far from uniform – in places, it had a relief of elevation of over 100 m. As a result, the thickness of the Devonian sediments varies from place to place, in part reflecting the fact that these river sediments were filling hollows in the topography. These deposits, collectively termed the Old Red Sandstone, today underlie a sizeable area just west of the Cheviot Hills.






FIG 60.Timeline of bedrock and surface-layer events in Area 2.

The early Devonian was also a time of igneous activity, and Devonian igneous rocks are often associated with interesting landscape features in Area 2. The largest area underlain by Devonian igneous bedrock is the Cheviot Hills in the northeast. Here, great thicknesses of sub-horizontal lava flows (predominantly mafic andesite) rest unconformably on Silurian marine sediments. This lava was extruded from an igneous centre that today straddles the border between Scotland and England. The centre experienced several stages of igneous activity: an initially explosive volcano was followed by a large out-pouring of lava, which today, after erosion, covers around 600 km2 of the Earth’s surface. This large area of lava was then intruded at depth by felsic granite, which after millions of years of erosion has been exhumed and today underlies the highest part of the Cheviot Hills, over the border in England. Elsewhere in the Area, much smaller Devonian volcanic vents and intrusions are relatively common.

By the start of the Carboniferous, around 360 million years ago, the Caledonian Mountains had been largely eroded, although the Southern Uplands still formed a considerable upland area. Throughout the following 60 million years of the Carboniferous, deposition in this Area occurred mostly in the lowlands of the Solway Firth Basin and along the border with England. Marine and coastal plain environments dominated, although varying relative sea levels resulted in the deposition of a variety of sediments, including limestones, sandstones, mudstones and coal. The Carboniferous sediments near Glencartholm, just south of Langholm (Fig. 58 (#ulink_5dc1ce89-5134-5a8f-a2e3-be76ef0fd668)), contain one of the richest Carboniferous fish faunas in Great Britain, and indeed in the world. The site is remarkable for the number of species of fish discovered, around 35, several of which are unique to the site. Some Carboniferous to Permian-aged igneous rocks are also found in this Area, generally in the form of small intrusions of coarser-grained rock.

By the end of the Carboniferous, Scotland had drifted northwards from the equator and the climate became more arid. Throughout the Permian (between 290 and 250 million years ago), red desert siltstones, sandstones and conglomerates were again deposited by winds and rivers. Fossils are not particularly common, but at Locharbriggs, just northeast of Dumfries, numerous reptile footprints have been discovered, generally heading in a southwards direction (very possibly towards the nearest source of water, in the Solway Firth Basin). The hot and arid climate of the Permian continued into the Triassic, and boulders, pebbles and sands continued to be washed and blown from the higher ground of the Southern Uplands down into sedimentary basins in the south and west of Area 2. These Triassic units tend to be mostly red, yellow and brown mottled sandstones, and comprise the Sherwood Sandstone Group. Together, both Permian and Triassic units make up the New Red Sandstone Supergroup.

In this Area, the New Red Sandstone was deposited in a series of basins, preserved in what are today the valleys of Nithsdale and the lower part of Annandale. These basins generally have a north-northwest/south-southeast trend, broadly perpendicular to the regional Caledonian trend. The Dumfries and Lochmaben basins are fault-bounded, and gravity studies suggest that Permian strata in these basins reach thicknesses of over 1000 m. They are flanked to the south by the much larger Solway Firth Basin, whose axis overlies the Iapetus Suture. The basement of this large basin is made up of tightly folded Ordovician and Silurian strata, overlain by up to 6 km of Devonian to Lower Jurassic fill.

No Mesozoic rocks are preserved in Area 2, although they would certainly have been present once: by the end of the Cretaceous, the sea had risen to cover all but the highest topography then present in this Area, and marine sediments would have been deposited during this time. However, during subsequent uplift and erosion all these younger rocks were washed away, such that during the last ~230 million years this Area has been subject to net erosion. The youngest bedrock present in this Area consists of Tertiary dykes, which despite being very small can be responsible for distinct walls in some landscapes. These dykes are long, thin intrusions of igneous rock, injected from igneous centres in western Scotland, such as Mull, around 60 million years ago.

MAKING THE LANDSCAPE

Tertiary erosion

In early Tertiary times, widespread uplift occurred across much of Scotland, particularly in the west. The Southern Uplands were again uplifted, and, in the warm, wet climate of the time, the result was vigorous weathering and erosion. As in the rest of Scotland, this weathering initiated some of the largest-scale landscape features seen today: the main upland and lowland areas became either defined or enhanced, and most of the main river valleys were initiated. Although there is not an exact correlation, bedrock has obviously played a role in determining the characteristic of the landscape – the high ground of the Southern Uplands is underlain by relatively hard Silurian strata, whilst the softer Devonian to Triassic rocks underlie generally lower ground. The bedrock of the Cheviot igneous centre has also resisted erosion, forming the Cheviot Hills (Fig. 61 (#ulink_419321f5-4a9b-5530-bbcc-da0be3d7ab54)).

The Southern Uplands would have been a relatively high table land at the start of the Tertiary. The subsequent erosion cut into this plateau, carving out valleys and watersheds which today define the hills. Although glacial erosion has been very important in creating the ultimate shape of the landscape, the remnants of this plateau surface are still visible in the flat or rounded hill tops, which all lie at similar elevations.

In places, weathering has clearly been concentrated along lines of geological weakness, such as faults and softer sedimentary units. The remarkably linear valley of Moffat Water, for example, coincides with a large Caledonian fault and a thin band of softer Ordovician shale. Likewise, the major valley of Annandale appears to have been eroded in the softer Permian sedimentary rocks, whilst the thin band of Silurian rocks separating the Permian of Dumfries and Lochmaben is obviously more elevated. The main rivers in Area 2 are the Tweed, Annan, Esk, Ettrick, Yarrow and Teviot, and the sources of all but the last lie in the higher hills in the northwest of the Area. Many of these rivers flow along northeast/southwest-trending valleys, parallel to the underlying folded strata, the water having taken advantage of either faults or weaker bands within the succession. Another important valley is found in the southern part of the Area: Liddesdale runs approximately northeast to southwest, with Liddel Water defining the Scottish–English boundary for several kilometres before joining the Esk. The development of this valley, too, appears to have been influenced by the underlying geology, as its location coincides with an area of heavily faulted and soft Carboniferous sedimentary rocks.






FIG 61. Digital elevation model of Area 2 with the main river valleys shown, as well as main upland areas. Sub-areas discussed in later maps are indicated by rectangles with red borders.

Evidence that the Solway Firth area acted as a basin for sedimentation in Permian and Triassic times has been outlined above. It seems very likely that it was acting as a river valley during Tertiary times.

Landscape development during the Quaternary glacials

The final form of the landscapes in Area 2, as in the rest of Scotland, owes much to the action of ice and ice meltwater. During the last 2 million years, ice sheets have repeatedly expanded to cover much of Scotland. These ice sheets flowed radially outwards from centres in the Highlands, the Southern Uplands and the Lake District, and were powerful agents of erosion and deposition.

The most recent glacial stage reached its coldest in the Late Glacial Maximum about 20,000 years ago, at which time Area 2 was overrun by Southern Uplands ice flowing roughly eastwards across the Area from an ice divide located over the outer Solway Firth (Fig. 62 (#ulink_5c10b331-96d7-56cc-92cb-3a0664ff07e3)). This flow is recorded by west/east-oriented drumlins along the low ground adjacent to the Solway Firth, and by a beautifully preserved array of ice-streamlined ridges along the Teviot valley. This flow direction must have been influenced by the presence of a strong ice dispersal centre in the Lake District that deflected the Southern Uplands ice to flow eastwards along the inner Solway Firth. Ice also accumulated in the Moffat Hills in the northwest of this Area, and a small but independent ice centre, powerful enough to withstand the pressure of the main ice mass, was present in the Cheviot Hills.

After the glacial maximum, the Lake District ice centre decreased in size and strength and Southern Uplands ice became dominant throughout the region. In the northeastern half of the Area, ice continued to flow roughly northeastwards down Teviotdale, but in the southwestern half, flow directions were reversed: drumlins and glacial striae indicate that ice flowed from the uplands around Moffat southwestwards down into the Solway Firth. As deglaciation continued, ice flow became valley-contained, and a final flow phase is recorded by north-northeast/south-southwest-oriented drumlins in the lowlands around the Solway Firth. Mapping of glacial deposits west of Annan and north of Gretna shows that ice then receded up the Solway Firth, as the Solway glacier retreated.

As mentioned above, the Cheviot Hills acted as a small but independent centre of ice accumulation and glacier dispersion during the Devensian glaciation. During this time, Cheviot ice flowed radially outwards from the centre, shielding the higher parts of the massif, particularly those above about 300 m, from the ice from the Solway Firth and Tweed basins. The lower peripheral hills, meanwhile, were overwhelmed and smoothed by this ice, and blocks of bedrock from the Tweed and Solway areas were deposited by the ice up to altitudes of around 300 m. Above this level in the central parts of the massif, soft, deeply rotten bedrock is relatively common and there are tors, erosional relics, over the border in England. These areas have clearly been glaciated, as they are often overlain by till, and yet they have survived erosion. Perhaps at the centre of Cheviot ice dispersal, the preservation favoured net deposition rather than erosion.






FIG 62. Ice flow in Area 2 at different times during the Devensian (LGM is Late Glacial Maximum).

The brief return to glacial conditions which occurred during the Loch Lomond Stadial (around 13,000 to 11,500 years ago; see Fig. 39 (#ulink_fb953f5d-69f7-5726-b236-2e0f344fd41f)) had only a limited effect in this Area. Glaciers returned only to the highest ground, such as around Broad Law, but even here the largest was only a few kilometres long. The position of these glaciers is clearly represented by terminal moraines – ridges of glacial till bulldozed by the Loch Lomond Stadial glaciers. Good examples are found, for example, at Loch Skene (Fig. 63 (#litres_trial_promo)), where a series of terminal moraines records fluctuations in glacier growth during retreat. The terminus of one ice lobe from this time has been located at the head of the Grey Mare’s Tail waterfall, at an altitude of around 450 m. Another terminal moraine forms a prominent ridge some 250 m long that runs parallel to the northeastern side of the loch (�the Causey’). Indeed, Loch Skene owes its presence to these moraines, which effectively dam the loch outlet.

Glacial modification of the uplands

Glacial erosion has played an important role in creating the final shape of the landscape. Most of Area 2 was extensively ice-scoured throughout the course of the Quaternary glaciations, and the land surface present at the end of the Tertiary became heavily modified. In the uplands, moving ice lowered the valley bottoms and smoothed the mountains, creating the high broad hills seen today. This erosion was not as intense as that to which the western Southern Uplands were subjected, and whilst the main valleys were deepened by powerful ice streams, the intervening ridges and plateaus escaped deep scouring. In general, the uplands of this Area therefore lack the �Alpine’ form seen in the Highlands of Scotland: corries, arêtes and troughs are not as common nor, when present, as pronounced (Fig. 63 (#litres_trial_promo)




Конец ознакомительного фрагмента.


Текст предоставлен ООО «ЛитРес».

Прочитайте эту книгу целиком, купив полную легальную версию (https://www.litres.ru/peter-friend/scotland/) на ЛитРес.

Безопасно оплатить книгу можно банковской картой Visa, MasterCard, Maestro, со счета мобильного телефона, с платежного терминала, в салоне МТС или Связной, через PayPal, WebMoney, Яндекс.Деньги, QIWI Кошелек, бонусными картами или другим удобным Вам способом.



Если текст книги отсутствует, перейдите по ссылке

Возможные причины отсутствия книги:
1. Книга снята с продаж по просьбе правообладателя
2. Книга ещё не поступила в продажу и пока недоступна для чтения