… all these beloved and enduring things, which combine to make the real Gower which we have inherited, and which our children’s children will enjoy until the time when the glaciers flow south again, perhaps to envelop and disfigure our peninsula …
Horatio Tucker, Gower Gleanings
DESPITE THE apparently unchanging nature of the earth, the continents are in constant motion, travelling enormous distances over great periods of geological time. The continental plate of which Gower is part has been situated on, and even to the south of, the equator and has drifted slowly northwards over the past 425 million years. The rocks that make up the peninsula today have therefore been deposited under widely varying conditions, such as subtropical seas rich in corals, arid deserts and coastal swamps. The distinctive landscape, with its rocky coastal cliffs, sandy bays and rolling hills, is a direct result of the exposed solid rocks and the processes that have affected them during this long and complex history. Some of the rocks contain large numbers of fossil plants and these provide not only evidence of past plant communities, but also important information about plant evolution. In geological terms the present-day flora is only a passing phase in the area’s long history.
Gower is formed from very ancient rocks (Fig. 15, Table 1), and rocks younger than 290 million years ago are hardly represented. All were deposited as sediments, as sands, gravels and other fine-grained deposits in horizontal layers known as beds or strata, and were themselves the result of existing rocks being eroded, or of accumulations of organic remains. Small breaks in the deposition, or changes in the sediment type, can cause the beds to be visibly separated from each other. This interface between one bed and another is known as the ‘bedding plane’ and shows clearly when rocks have been folded due to earth movements. Folds where the bedding planes have been bent into an arch are known as ‘anticlines’ and downward folds are known as ‘synclines’. When erosion cuts across these folds anticlines show the oldest rocks at their axis, whereas younger rocks are present at the centre of synclines.
FIG 15. Geological map of Gower. (Gower Society)
Royal Commission on the Ancient and Historical Monuments of Wales for Figures
The oldest rocks exposed in Gower are characteristically red in colour and are known as Old Red Sandstone. They were laid down in the Devonian period, from 390 to 360 million years ago (although there appears to have been a lengthy gap in deposition) and occur only in the cores of major anticlines. The nature of any underlying ‘basement’ rocks is not clear. At this time the area lay in a region of sediment-laden rivers that crossed a wide plain between mountains to the north and the sea to the south, which lay over what is now Devon, the county giving its name to this period of the earth’s history. The climate at this time was tropical, with monsoon weather, and vast rivers carried sediment resulting from the intense erosion that was taking place in the mountains. These rivers deposited large amounts of material as they flowed to the sea and this eventually formed the sandstone rocks. A period of uplift and erosion in mid-Devonian times temporarily stopped this sedimentation. Although the Devonian is sometimes referred to as ‘the Age of Fishes’, other animals were present too, including insects, sharks, lungfishes and amphibians, the latter being the first vertebrates to walk on the land. They looked like large newts or salamanders. Increasing numbers of plants were also appearing on the land at this time and they were growing bigger and more complex than the first terrestrial plants.
TABLE 1. Rock types exposed in Gower, related to simplified geological time divisions. Age in millions of years before present. (Adapted from Aldhouse-Green, 2000)
| PERIOD | AGE (BP) | ROCKS PRESENT IN GOWER |
Quaternary |
1.8–0 |
Various glacial and postglacial drift deposits |
Neogene |
23–1.8 |
|
Paleogene |
65–23 |
|
Cretaceous |
142–65 |
|
Jurassic |
206–142 |
(Found offshore on seabed to the south) |
Triassic |
248–206 |
Remnant terrestrial deposits near Port-Eynon and rare fissure-fills within older rocks |
Permian |
290–248 |
|
Carboniferous |
354–290 |
Coal Measures in northeast Gower; Millstone Grit; Carboniferous Limestone |
Devonian |
417–354 |
Old Red Sandstone |
Silurian |
443–417 |
|
Ordovician |
495–443 |
|
Cambrian |
545–495 |
The tough, coarse sandstones and even coarser conglomerates form the high ground of Cefn Bryn, Llanmadoc Hill and Rhossili Down. Since it is only in the cores of major anticlines that these rocks are brought to the surface all the hills on the peninsula coincide with upfolds in the rocks underneath. Although it is not a general geological rule, in Gower anticlines always form hills and synclines always form valleys. This results from the fact that the older rocks are hard and the younger rocks are soft.
Around 360 million years ago at the beginning of the Carboniferous period there was a major rise in sea level, which covered almost all of what is now Wales, and the wide coastal plains of the Devonian landscape were drowned. After an initial period when mainly muddy sediments were deposited, the water became clearer, the amount of detritus reaching the area from the land reduced and limestones began to form. What is now Gower was then passing across the equator. Even today it is principally in such equatorial regions, where there is little sediment being deposited from the land, that limestone is formed. In the warm water lime (calcium carbonate) can be precipitated, and accumulates together with the remains of marine animals that secrete shells and skeletons. Without the sediment from the land to dilute them, these shells can form rocks on their own, although many limestones are composed of microscopic particles of lime, known as lime mud.
The Carboniferous Limestone in Gower is approximately 800 metres thick, with fine exposures in the southern cliffs, but northwards the sequence becomes gradually thinner and there may also be some parts missing. This suggests that the sea further north, as it was nearer the land, was shallower. The limestone consists of many different types, each with a different texture, thickness and group of fossils. Recent studies have shown that the deposition was controlled by a worldwide rise or fall in the sea level and not by a local or regional subsidence or elevation of the land. There were several advances and retreats of the sea from the south and layers of carbonates gradually built up under a very shallow sea that was rich in corals and brachiopods, sedentary shelled animals with feathery appendages to waft food particles to the mouth. The uppermost part of the Carboniferous sequence in Gower consists of marine shales and the muddy limestones of the Oystermouth Beds. These contain the widespread limestone fauna of brachiopods and corals, together with the rare trilobite Griffithsides spp. Brachiopods are the most abundant fossils, especially Martinia spp. and Spirifer spp. One of the Spirifer species is named Spirifer oystermouthensis after Oystermouth village, where it is found in Clements Quarry, the so-called ‘Black Lias Quarry’, named due to the alternations of dark, muddy limestones (which weather to white) with thin dark, calcareous shales. There are many theories about how this regular pattern was produced, but there is a strong possibility that it had some seasonal or climatic cause. Also present in the quarry, which is easy to access as it is currently used as a car park, is the small cornet-shaped shell Zaphentis oystermouthensis.
From time to time the carbonates were above sea level and there is evidence of erosion and of plant roots. Unusual breccias (rocks composed of angular fragments) suggest the presence of salts formed by the evaporation of sea water in a hot climate. Other deposits suggest lagoons and oolite, a limestone composed of small grains, comparable with those forming today in shoals around the Bahamas. In some levels there are algal tufts and mats that formed as calcareous layers built up by lime-secreting algae that favoured shallow, warm seas. These varying conditions produced different types of limestone, and six main rock units can be identified, all named after areas on the peninsula (Table 2).
During the Carboniferous period forces that were eventually to fold and fault the Gower rocks were beginning to bend the earth’s crust upwards. This increased the runoff from the land, which in turn resulted in river deltas swamping the limestone sea with sediment. In comparison to other areas of Glamorgan, where the rocks show that there were alternating periods of non-marine and marine conditions, the sea over Gower at this time, some 320 million years ago, was comparatively deep, and fine shales were deposited with marine animals preserved within them. This fine material, known as Namurian Shales, is relatively soft and was subsequently eroded to form the bays of Oystermouth, Oxwich and Port-Eynon. The best section through the Namurian rocks occurs along Barland Common Stream, near Bishopston, where the rock sequence contains a number of bands with marine fossils. The section is of historical interest as it was originally recorded by de la Beche and described in the first Geological Survey Memoir published in 1846.
TABLE 2. Simplified stratigraphy of Gower limestone. (Adapted from Lowe, 1989)
The environmental conditions of the Namurian continued into the latter part of the Carboniferous period, with the formation of the Coal Measures. The coal seams originated as peat formed in low-lying swamps on the coastal plains, and coal formation may be linked to changes in sea level. The Coal Measures consist of sandstones, shales and coals arranged in a repeated sequence, indicating that forests grew, were buried by shales as the land subsided and then by sands as the rivers deposited sediments. On top of the river sand soil developed and eventually the forest became established again. In the northeast of the peninsula coal occurs in seams running diagonally from Blackpill to Llanrhidian.
The freshwater and brackish swamps, marshes and lagoons of this period of Gower’s history supported rich forests and there were large numbers of fern-like plants, the seed ferns, together with the true ferns that occur today. One of the most conspicuous features of the flora at this time was the presence of large tree-sized ferns in groups that are now only represented by small herbaceous types, the most spectacular being the horsetails and lycopods. Such plants inevitably broke up and fossil fragments consisting of discrete pieces of roots, stem and leaves are common.
The great coalfield basin dominates the geology of South Wales and Gower forms part of the southern rim of the massive South Wales Coalfield syncline. The peninsula should therefore consist of strata that dip northwards and become progressively older towards the south. This is not the case, because the structure of the southern part of the coalfield is much more complex than it appears to be from the outcrop pattern, and on the southern limb of the syncline there is a series of tight folds. The axes of these folds lie roughly east–west, like the axis of the coalfield itself, and this shows that the pressure that created these folds was coming from the south. One of the best places to see the structures formed during these earth movements is at Bracelet Bay, where a plunging anticlinal fold has been eroded by the sea to create a gently curving pattern of limestone beds in the wave-cut platform. This fold, known as the Langland anticline, and associated structures such as faults and veins, can also be seen in the cliffs on the eastern side of Caswell Bay.
Rocks of the latter part of the Carboniferous period are not present in Gower. Around this time a prolonged period of uplift and erosion, lasting for some 80 million years, affected much of what is now Wales and adjacent regions. During this time, through the Permian and well into the Triassic period, the Coal Measures and older strata were folded into the present coalfield basin.
Life flourished during the Carboniferous and Permian. Crinoids, ammonites, corals and fish diversified in the seas, while amphibians and reptiles continued their invasion of the land. After more than 100 million years of relative stability, however, the end of the Permian (248 million years ago) saw the largest extinction event in the earth’s history, an event far more devastating than the later and much more famous Cretaceous extinction, when the dinosaurs died out. Around 75 per cent of known species of land animals and 96 per cent of marine animals disappeared forever, together with early corals and trilobites. Many causes have been suggested for the Permian extinctions including fluctuations in sea level, a change in the salinity of the ocean, and volcanic activity. The most important factor, however, seems to have been climate change.
By the end of the Triassic period Gower lay in northern tropical latitudes and the climate was hot and semi-arid. The Triassic was a fresh beginning for life on earth and new species evolved to fill the gaps left by the Permian extinction, with the first dinosaurs evolving towards the end of this period. Throughout the Triassic, river systems running south across what is now Glamorgan drained into a lake or lakes whose northern shore lay close to the present Bristol Channel. From about 220 to 210 million years ago a complex of mudstones and evaporates was deposited as a result of the rapid evaporation of these lakes. In some areas these sediments have penetrated and filled fissures in the Carboniferous Limestone. The lakes appear to have been surrounded by an arid treeless hinterland, since there are few fossils of plants and insects. Only one small area and a few infilled fissures near Port-Eynon remain as evidence of a similar Triassic cover in the peninsula. In the 1690s this ‘red ochre’ deposit from the age of the early dinosaurs was discovered by John Lucas of Port-Eynon, who considered it a ‘seam or deposit of paint material’. Lucas ‘employed men to dig therefore to the great well being and benefit to himself and to the men and he possessed much wealth in moneys, and did buy skiff at Swainsey and Bristol to beare ye paint material away to number of five …’ This source of material was exploited for centuries and the skiffs carried it to Cardiff for sale. According to the accounts of the Penrice Estates the use of this paint was the origin of the Great Western Railway livery of brown and ochre. The 1938 edition of the company magazine recording that they used 10,000 pounds (4,536 kilograms) of ‘levigated raw ochre’ a year for their paints.
Marine conditions gradually returned some 210 million years ago, reaching their full extent in the early Jurassic period. The seas were full of life, including ammonites and marine reptiles such as ichthyosaurs and plesiosaurs. During this period the early Jurassic sea surrounded islands of Carboniferous Limestone, such as Gower, and the Blue Lias Limestone was deposited as muds and lime muds in a relatively quiet sea, the beds now covering much of the Vale of Glamorgan. Purer and lighter limestones originating in shallow water then succeeded the Blue Lias. It has long been claimed that thick layers of younger Jurassic rocks, plus succeeding deposits from the Cretaceous period, including the chalk, originally covered the whole of Wales. No evidence of these younger rocks remains, but it does seem likely that the chalk, formed mainly from the remains of microscopic plankton called coccoliths, may have covered much of the country, with the exception perhaps of small areas of Mid and North Wales. It appears that this late Cretaceous deposit was removed by erosion shortly after it was formed. No Jurassic or Cretaceous rocks remain in Gower today, although Jurassic strata have been found offshore to the south of the peninsula and they also occur to the east in the Vale of Glamorgan.
The general topography of Gower has developed since the beginning of the Tertiary period when Wales was subjected to a long period of periodic uplift. As a result, around Swansea and in the Vale of Glamorgan there is a series of stepped, but highly dissected, coastal platforms at 60, 90, 120 and 180 metres, which are considered to be remnants of marine erosion surfaces, or ‘wave-cut platforms’. Like similar planation surfaces in the uplands they truncate pre-existing geological structures. Remnants of the 180m surface are represented in Gower by the Old Red Sandstone ridges of Cefn Bryn, Llanmadoc Hill and Rhossili Down. Much of the peninsula is at a lower level and like the Vale of Glamorgan extends to a little over 120 metres. This 120m platform is not very clear, but can be seen near Pen-clawdd, Three Crosses and Clyne Common. The southern cliffs represent the lowest platform, later erosion along fault lines producing the dry valleys known locally as ‘slades’. The general appearance of the peninsula is therefore a series of plateaus with occasional hills rising above the surface.
The general shape of the south Gower coast is determined by a number of pitching folds (folds with tilted axes) in the Carboniferous Limestone that result in anticlinal headlands with cliffs and synclinal bays with sandy beaches. The anticlines occur at Mumbles Head, Pwlldu Head and Oxwich Point, while synclines form the major bays of Oxwich and Port-Eynon (Fig. 16). The many smaller bays such as Limeslade and Bracelet have, in contrast, generally been formed by erosion along faults running north–south (Table 3).
In some locations on the peninsula displacement of the rocks has occurred along a reverse fault, a break in the rock where one mass of rock has pushed up over another. Such faults tend to run parallel with the axis of the folds and are called ‘thrusts’ in Gower, although this is strictly not the correct use of the term. One such thrust occurs on Cefn Bryn where it is marked by a hollow on the north side and crest of the anticline that is followed by the road from Cilibion to Reynoldston. Because both the anticlines and the thrusts in Gower were formed by pressure from the south they tend to lie parallel to one another. The peninsula owes much of its shape to geological faults. Professor Neville George, Senior Lecturer in Geology at the University of Swansea, noted in 1933 that ‘Almost every bay, inlet, gully, mere, cave or sound is eroded along a joint or a fault or a fold or a soft bed.’ Faulting has, for instance, resulted in the small coves and bays at Langland, Caswell, Pwlldu and Limeslade (Fig. 17). Other faults run between the tidal islands of the Mumbles. Each island is also bisected by a smaller fault. The faults may contain interrupted iron veins, which indicate incremental openings of the fault or fissure.
FIG 16. The anticlinal headland of Oxwich Point and the synclinal Oxwich Bay. (David Painter)
David Painter for Figures
TABLE 3. The four key factors in Gower geology. (Adapted from Bridges, 1997)
| FACTOR | EXAMPLES |
1. East–west anticlines and synclines. |
Mumbles Head Port-Eynon Bay |
2. Anticlines producing hills and synclines producing wide valleys. |
Llanmadoc Hill Oxwich Bay |
3. East–west thrust faults where one mass of rock has pushed up over another. |
Cefn Bryn |
4. North–south faults, which may give rise to typically narrow bays where they meet the coast. |
Limeslade Bay Three Cliff Bay Western scarp of Rhossili Down |
Iron is an important characteristic of many faults in Gower, and mineral veins associated with them usually contain calcite and haematite. The blood-red haematite ore was probably derived from the underlying Old Red Sandstone by the action of the hydrothermal solutions from which the vein minerals were crystallised. One of the larger faults, which runs across Mumbles Hill and whose southern end forms Limeslade Bay, was mined for iron from the Roman period until the nineteenth century. Between 1845 and 1899 the remaining iron ore was blasted out, broken into fragments and barrowed to small ships that took it across Swansea Bay to Swansea and Briton Ferry. The Swansea Guide of 1851 notes that ‘a valuable vein of iron ore … has been worked, much to the disfigurement of this romantic spot.’ A narrow trench, known as ‘the Cut’, was left across the hill. It was almost completely backfilled with rock hollowed out of Mumbles Hill when an underground sewage works was constructed in the 1930s. The remaining part of the trench can be clearly seen today from the viewpoint at the top of the hill, near the coastguard radio mast.
Rhossili Down is a rather exceptional feature; unlike the other high points on the peninsula the axis of the anticline is aligned north–south rather than east–west. This is because the rocks that make up the ridge were rotated along the line of the Llangennith fault. The west-facing scarp of Rhossili Down is formed along the Broughton fault and the Port-Eynon thrust lies to the south. It is therefore surrounded on three sides by faults, the steep slopes on the west, south and east being fault faces.
This then is the sequence of the Gower rocks and the forces that have shaped them into the peninsula. Because it is the oldest rocks, the Old Red Sandstone, that are now exposed on high ground at Cefn Bryn, Llanmadoc Hill and Rhossili Down it is clear that in these places the combined thickness of the Carboniferous Limestone, Millstone Grit and Coal Measures has been removed over the last 280 million years. The action of wind and water, assisted by plant roots, chemical breakdown and glaciers, has taken away an incredible 4,600 metres of rock.
FIG 17. Rocky shore near Pwll Du showing a cross-section of a fold. (Harold Grenfell)
David Painter for Figures
The whole of Gower was affected by ice on numerous occasions although, as in many places, the current shape of the land can only be related to the last two glaciations. Despite this the peninsula has one of the most complete glacial records in Britain and is therefore critical to the understanding of the Quaternary period. Advances and retreats of the polar ice caps were interrupted by interglacial periods lasting many thousands of years during which the climate was much milder, in many cases even warmer than it is today. Most of the ice in the penultimate glaciation appears to have been derived from source areas in the mountains of central and southern Wales. However, the ice in western Gower, along with that in Pembrokeshire and the Vale of Glamorgan, formed part of a very extensive ice sheet that encroached inland from the southern Irish Sea. This is clearly indicated by the glacial debris, which north of Cefn Bryn contains a greater amount of material originating from the South Wales coalfield while south of the ridge the rock types suggests an Irish Sea origin.
As the ice melted during what is known as the Ipswichian interglacial period, between 130,00 and 120,000 years ago, sea levels rose about 6 to 9 metres above present levels. Subsequent falls in sea level left behind beach deposits cemented with calcium carbonate, which are known as ‘raised’ beaches but should not be confused with the true raised beaches found in Scotland. Although the beaches and their platforms have been eroded since their formation their origin is shown by the rounded shingle and the shells they contain. At Foxhole, near Southgate, for example, at the foot of the low cliff, there is an elevated platform on the Carboniferous Limestone. On this wave-cut platform is preserved one of the best examples of the ‘Patella beach’, which is composed of sand, rounded limestone fragments and fossils of the common limpet Patella vulgata, held together by calcareous cement (Fig. 18). In Gower this deposit of shelly shingle (in which periwinkles and dog-whelks Nucella lapillus are also very common) is widespread and is normally cemented into a hard conglomerate, which rests on a narrow, wave-cut platform.
Many of the Gower caves open onto the platform of the raised beach, and it is probable that they were enlarged by wave action at the same time as the platform was created. Associated with the beach deposits are the well-known cave deposits, which have yielded bones of animals such as straight-tusked elephant Palaeoloxodon antiquus, hippopotamus Hippopotamus amphibius and soft-nosed rhinoceros Dicerorhinus hemitoechus, showing that the climate was much warmer than that of today. The importance of these deposits lies in their relationship to the raised beaches and the information this provides on the climatic changes in the late Devensian. Minchin Hole at Pennard contains a particularly important sequence of deposits and is regarded as the ‘type site’ for the Patella beach, which has been commonly used as a ‘marker horizon’ throughout southwest Britain. The cave is also unique in South Wales in containing two raised beaches of different ages superimposed in a single section. Investigations in the 1980s suggested that these beaches represent two separate interglacial periods, and because of this Minchin Hole is a nationally important site for studies of the Pleistocene (Fig. 19).
FIG 18. The ‘Patella beach’ at Foxhole, near Southgate. (Harold Grenfell)
The last glaciation, known as the Devensian, which reached a maximum about 18,000 years ago, destroyed many of the features created by the previous glaciation and is responsible for most of the glacial landforms preserved in Gower today (Table 4). This time all of the ice was derived from Wales and large glaciers occupied Carmarthen and Swansea bays. Mammals typical of the Devensian period include mammoth Mammuthus primigenius, woolly rhino Coelodonta antiquitatis, horse Equus caballus, reindeer Rangifer tarandus, arctic fox Alopex lagopus, spotted hyena Crocuta crocuta and abundant small mammals such as arctic lemming Dicrostonyx torquatus.
FIG 19. Minchin Hole, a nationally important site for studies of the Pleistocene. (Harold Grenfell)
The limit of this last glaciation in South Wales has been the subject of much debate. By about 22,000 years ago, much of Wales had been overrun by ice, and large glaciers occupied Carmarthen and Swansea bays. Much of south Gower, however, lay beyond the limits of the ice sheet and the deposits found above the raised beach sediments at Worms Head record these cold conditions. These ‘head’ deposits consist of angular limestone fragments prised from the slopes and cliffs above by the action of frost. During this cold period much of the ground would have been permanently frozen (permafrost) – only the upper layers would have thawed out, and during such thaws, soil and other loose materials would have slipped down the slope (solifluction) to lie with the broken limestone fragments above the raised beach deposits. These different types of sediment make Worms Head an outstanding site for scientists studying the ice ages and the way climatic changes occur through time. The evidence from Worms Head complements that obtained from Rhossili Bay, where there is more direct evidence for glacial activity during this period.
TABLE 4. The development of the Gower landform. (Adapted from Bridges, 1997)
The narrow strip of land below the scarp of Rhossili Down is the prime example of a ‘solifluction bench’ formed by large quantities of surface material sliding down the scarp at the end of the last glaciation (Fig. 20). This formed an apron which the sea has now eroded, forming a low cliff of loose material. Similar examples occur at the foot of the limestone cliffs between Slade and Oxwich Point. The slopes of the other Gower hills also have the appearance of having been affected by solifluction, with a smooth flowing shape and few rock outcrops. On the south-facing outcrops of the Carboniferous Limestone, frost shattering and scree formation occurred and some glacial debris was deposited.
While it is clear that the ice covered the northeast of the peninsula, doubts remain about the precise limit of ice in west Gower. Most of the available evidence comes from the northern end of Rhossili Down where a large exposure of shelly gravels occurs. This material is thought to be the ‘outwash’ of an ice sheet in Carmarthen Bay. Whiteford Point is a shingle ridge that may have originated as the remnants of the terminal moraine, a continuous line of debris left by the glacier that occupied the Loughor valley. This glacier greatly modified the landscape by providing an abundant supply of silts, clays, sands and boulders that have been reworked during the subsequent rise in sea level to form the basis of the marshes today. It is also likely that ice reached the coastal cliffs on the north shore, clearing frost-shattered debris and solifluction material. Ice sheets and glaciers also brought many large rocks into the peninsula, the most obvious of which is Arthur’s Stone. Utilised in the Neolithic period to form a chambered tomb, the 24-tonne conglomerate capstone contains brown clay-ironstone, coal fragments and traces of fossil plants, which do not occur in the Devonian conglomerates elsewhere in the peninsula. This confirms that the capstone was derived from the northern outcrop of the Millstone Grit and carried to Gower by the ice.
FIG 20. Solifluction bench below Rhossili Down and Rhossili Rectory, now a holiday cottage owned by the National Trust. (Harold Grenfell)
David Painter for Figures
After the glaciations the sea level rose again and at its maximum extent low cliffs were cut in Oxwich and Swansea bays. Since then sediments accumulating in front of these low cliffs has isolated them from the sea and sand dunes have formed. Present marine activity is restricted to a limited zone at the foot of the soft cliffs. In contrast the limestone cliffs were largely shaped in the Pleistocene when sea levels were high. The north Gower cliffs, which are now some distance from the sea, are in the main ‘fossil’ cliffs, while those on the south coast are best described as ‘relict’ because although they were largely shaped at the same period as those of north Gower in places they are still subject to erosion by the sea. Periglacial conditions were also present immediately following the last glaciation, but the retreat of the ice was rapid and much of the peninsula is thought to have been ice-free by 14,000 to 15,000 years ago. There is no reason to believe, however, that the ice will not return. Despite global warming, the present conditions represent only a brief respite in a predominantly glacial age.
At the end of the Pleistocene period, 10,000 years ago, sea levels were well below the present, probably at least 22.5 metres lower. Early in the postglacial period the area of the present-day Bristol Channel would have been occupied by a large river in the middle of a wide plain of birch tundra. From the end of the last glaciation to the early Neolithic, about 5,700 years ago, sea levels in Britain rose steadily as the water which had been locked up in the ice sheets was slowly released and covered these coastal woodlands. Offshore from Rhossili are drowned river valleys, the channels of which, although partially infilled with sediment, are generally 5 metres deeper than the adjacent seabed. The general trend therefore has been for low-lying coastal margins to become inundated, as shown by the submerged peat beds in areas along the coast. In some cases tree remains are also found, the so-called ‘submerged forests’. Describing Oystermouth in Swansea Bay at the end of the seventeenth century, Isaac Hamon wrote ‘The sea hath encroached upon a great part of the low grounds of this parish, as appears by the roots of trees, and whole trees that lyes in the sands and other tokens.’ Other remains of the forest have also been uncovered in the past in Port-Eynon and Broughton bays.
Comparatively little research has been carried out on the remains of this past habitat, much of the interest being centred on the geology and animal fossils present. As Neville George (1930) remarked, ‘it cannot be too strongly emphasised that much of the interest of the Forest rests in its being a forest, and that a considerable portion of the material consists of an abundant and varied flora; an investigation by some competent botanist of this rich and practically untouched assemblage could not fail to prove exceedingly illuminating.’ Species recorded include silver birch Betula pendula, hazel Corylus avellana, alder Alnus glutinosa, elder Sambucus nigra, deergrass Trichophorum cespitosum, rushes Juncus spp., irises Iris spp. and spurges Euphorbia spp. Leaves of pedunculate oak Quercus robur have also been found. This group of plants is characteristic of low-lying wet habitats with dry land nearby and represents a similar habitat to that existing today in parts of Oxwich marsh. Insect remains include the wing-cases of beetles such as the dung beetle Geotrupes vernalis, which appears to have been fairly common and was probably associated with the presence of larger animals, including roe deer Capreolus caprea, red deer Cervus elaphus and ox Bos taurus.
The grinding action of ice flow and glaciers, when the majority of northwest Europe lay under thick ice sheets, formed sands and gravels. After the last glaciation, some 6,000 to 8,000 years ago, a period of global warming produced large volumes of meltwater, which transported these materials into the river systems and near coastal waters. When the sea level rose it reworked these sands and gravels and pushed them shorewards, until they lay at or near the heads of bays. These sediments are not being significantly renewed by natural processes and are therefore a finite resource. Although the Bristol Channel once contained vast reserves of this ‘glacial outwash’ research has shown that there is a long-term and generally westward transport of sand out of the area, leaving behind a largely rocky seabed with a sparse covering of sediment.
The high tidal range of the Bristol Channel results in wide expanses of sand being exposed at low water, and when conditions were suitable some of this was blown inland, leading to the development of coastal dune systems. Near the northwestern end of Whiteford Point peat, containing leaves, roots and other plant material, is exposed just below the high-tide level and represents a former land surface. It has not been dated, but Bronze Age remains have been recorded on top of the clay and beneath the dunes that fringe Broughton Bay, with Roman remains on the overlying sand. This indicates that sand began to move inland in significant quantities in the Iron Age, around the first millennium BC. This process continued for centuries, although there was a considerable increase in storminess in northwest Europe from AD 1200, which continued after a slight improvement in the fifteenth and early sixteenth centuries into the seventeenth and eighteenth centuries. The seventeenth and eighteenth centuries have been described as the ‘Little Ice Age’ because of the increase in the size of the European mountain glaciers that took place during this time. The environmental evidence is backed up by documents that record devastating sand movements between the thirteenth and fifteenth centuries, corresponding with the period of climatic deterioration. In addition, studies of tidal levels show that the general level of the tide was steadily increasing during the twelfth and thirteenth centuries, with a peak around 1433, and that it remained very high during the two following centuries before diminishing. This combination of increasingly stormy weather and higher tides played a significant part in determining the fate of many coastal settlements in South Wales.
Incursions of sand are thought to be responsible for settlements at Rhossili, Penmaen and Pennard being abandoned during the thirteenth and fourteenth centuries, although it appears that their abandonment was more the result of a loss of interest in maintaining them, rather than a result of the settlements being rapidly submerged beneath deep sand. Sand movement is usually sufficiently gradual for land to remain in use if necessary. At Rhossili, for example, there seems to be evidence for the abandonment of the lower village on economic grounds alone, land on the plateau being about twice as productive as that on the solifluction bench where the deserted settlement was located. The presence of a strip field in the medieval field complex adjacent to the medieval village called ‘Sandylands’ suggests that windblown sand affected both parts of the manor and not just the area below Rhossili Down. The besanding of settlements such as these, on cliff tops situated well above sea level, is not a common phenomenon, but studies have shown that winds of sufficient speed to carry sand up to the top of cliffs can be expected to occur about eighty times a year, even in today’s climate. Once sand is airborne it remains so until it is caught by a suitable surface, or the local wind speed falls below a critical velocity, which is dependent on particle size. In this situation it is the local topography that determines where the deposition of sand takes place (Table 5).
TABLE 5. Besanded sites in Gower. (Adapted and simplified from Toft, 1988)
There are two documents relating to sand at Pennard (Fig. 21), one from the fourteenth century and another 200 years later. The first, from 1317, has been generally accepted as marking the beginning of the advance of the sand. In the document William de Breos III granted hunting rights on the ‘sandy waste at Pennard’. Excavations at a medieval house site close to Pennard Castle, however, revealed that sand was present long before the fourteenth century. The second document is preserved among the archives of All Souls College, Oxford and is a petition written in 1535 by Harry Hopkin, the vicar of Pennard, to the king’s commissioners at Swansea. The commissioners were engaged in compiling a comprehensive valuation of ecclesiastical benefices to provide the Crown with an up-to-date assessment of their income as a basis both for imposing on them a perpetual tax (amounting to one-tenth of their annual net incomes) and also for levying the whole income of every benefice for the first year following the appointment of a new vicar.
It was this valuation in the deanery of Gower that prompted Hopkin to submit the petition, because the encroachment of drift sands that affected Pennard and some other coastal areas of South Wales at this time had substantially reduced the value of the benefice. Indeed the advance of drift sand along the coast of Glamorgan was sufficiently serious to be the subject of an Act of Parliament of 1554 (An Acte touchynge the Sea Sandes in Glamorganshyere) and its contribution to the decline of the coastal borough of Kenfig is well documented. The sand had seriously affected the parish for many years before the petition was written. Because of the sand tenants had had their rent reduced in 1478, and in 1517 the church was excused its royal levy. Eventually it was abandoned during the 1520s in favour of a new site 2 kilometres to the east, along with the ‘mansion house’ appertaining to it, by which Hopkin presumably meant the priest’s house or vicarage. The glebe lands belonging to the church had also been covered over together with many other tenements. All this is described in the first part of the petition:
To the kinge’s comysshionerz nowe being at Swansey.
In his humble wyse shewyth unto your masterships your oratour Sir Harry Hopkin, vicar of Penarth in the denery of Gower within the dyosys of Saint David, that where the said churche and the vicaraige with all the glebe landes to the same belonging or appertaynyng is utterly and clerly destroyd and overgon with the dryfe sandes of the see. And not only the said church with the mansion house and the glebe landes whyche to the same appertained, but also diverz and many other tenements whyche were within the said paryshe be in like wyse destroyed and decayed by reason of the said dryfe sandes, in such wyse the proffites of the said churche is gratly decayed …
FIG 21. Pennard Castle, a site ‘besanded’ during the thirteenth and fourteenth centuries. (Harold Grenfell)
Records, including further Acts of Parliament concerned with coastal protection, show that the stormy conditions and subsequent drifting of sand continued until the eighteenth century.
Soil formation in Britain began as the ice cover diminished, and it is continuing today. An earlier origin has been proposed for the red clay-rich material that lies on top of the Gower limestone, but in most cases this appears to have been produced by glacial ‘smearing’, or by solifluction. The soils (Table 6) broadly reflect the distribution of the underlying geology of Carboniferous Limestone, Millstone Grit and the Coal Measures, but are more locally influenced by a covering of glacial, or periglacial, deposits. South and west Gower is covered with gravely loams, while finer-textured soil is widespread in the northeast. Brown earths are the main soils in the complex till and glaciofluvial deposits that occur in south and west Gower. These soils, which cover most of the peninsula, are agriculturally important and are deep and loamy with rounded stones throughout the soil profile. Although they are derived from relatively acidic Carboniferous rocks, with a pH of around 5.0, their acidity has been lowered by the application of lime ‘manure’ and basic slag.
It was a common practice for Gower farmers to mix lime with proportions of earth to produce what was termed ‘marl’. Lime making was one of the earliest industrial activities in the peninsula, the material being produced by heating limestone in kilns and then treating it with water to form calcium hydroxide. The Oxwich jurors wrote in 1632 that it was their right ‘to digge lime-stones for to repaire … and to burn lime’, while at Landimore in 1639 it was recorded that ‘there be Quarries of Limestones where Tenants time out of mind have used to burn their Lime for the composting of their lands’. Most of the surviving limekilns were constructed largely in the period between 1750 and 1850 to burn limestone in order to make lime for local use: to spread on the fields as a soil improver, to make mortar for construction, lime-wash for painting buildings and for many other uses. Lime burning had all but died out by the turn of the twentieth century and the kilns (Fig. 22) were abandoned. Walter Davies described the application of lime in the first volume of his book, a General View of the Agriculture and Domestic Economy of South Wales (1814–15):
In Gower, west of Swansea, sound soil on limestone; some rich land near the Mumbles, letting for 3 1., 3 1. 10s., and 4 1. per acre. The Duke of Beaufort grants three lives’ leases, esteemed here the best preservatives of land, ‘as the farmer finds a warmer interest in the soil’ … Course of leaseholders for three lives: 1. Wheat on fallow, limed on the first ploughing with from 50 to 100 stacks or horse-loads (150 to 300 bushels) per acre, and as much dung on the third tilth; 2. Barley; 3. Oats; 4. Barley, with dung and coal-ashes, and lay down for four years without seeds … Course of tenants at will, in the same tract: 1. Wheat on fallow, with half the above quantity of lime and dung, or less than that; 2. Barley; 3. Oats; 4. Barley; and lay down for four years; or as some do, 5. Oats, and oats as long as any can be had.
Despite the details of the application rate it is difficult to estimate how much lime was actually applied. The imperial bushel, legally established in Great Britain in 1826, was a dry measure of 8 gallons (36 litres) used principally for grain and fruit, but it also had a great variety of other values in local use, varying not only from place to place but in the same place according to the kind or quality of the commodity in question. Frequently it was no longer a measure of capacity, but a weight of flour, wheat, oats or potatoes.
On the Coal Measures in the northeast of the peninsula the soils are heavy and impermeable and this, together with gentle slopes and the humid climate, causes these soils to be wet for much of the year. Unimproved soils are extremely acidic with pH values around 4.0. On improved land the values are higher, commonly around 5.5 to 6.5.
FIG 22. Limekiln at Mewslade. Lime making was one of the earliest industrial activities in the area. (Harold Grenfell)
TABLE 6. The soils of Gower. (Adapted and simplified from Wade et al., 1994)
| UNDERLYING ROCK OR OTHER PARENT MATERIAL | DOMINANT SOIL GROUP | SOIL CHARACTERISTICS |
Dune sand |
Raw sands |
Calcareous windblown sand; thin humic topsoils only present on stabilised dunes and frequently buried. Gleyed soils and peat in hollows. |
Limestone and Triassic ‘gash’ breccia |
Brown earths |
Well-drained loamy, or loamy over clayey soils, shallow in places, especially on steep slopes and crests with ‘rendzina’ in rocky areas. |
Fluvioglacial sands and gravels |
Brown earths |
Deep well-drained loamy soils. |
Millstone Grit and Coal Measures |
Brown podzolic soils |
Well-drained loamy soils over sandstone, usually on steep slopes. Heavy clays on shales – slowly permeable, seasonally waterlogged soils with a peaty surface, usually on valley floors. |
Devonian conglomerate and sandstone |
Podzols |
Well-drained, very acid sandy soils with a bleached subsurface horizon over conglomerate. Associated with less acid well-drained reddish loamy soils over sandstone and siltstone. |
Till |
Stagnohumic gley soils |
Heavy stony clays – very acid, slowly permeable seasonally waterlogged loamy soils with a peaty surface horizon. |
Estuarine alluvium |
Alluvial gley soils |
Loamy alluvial soils with high groundwater. Saline and partially anoxic in marshes. Sand or soft mud on intertidal flats. |
The soils on the ridges of the Old Red Sandstone outcrops are mainly rocky, coarse loamy podzols. They are found under heathland vegetation and below a surface accumulation of acid plant debris there is a thick bleached sandy horizon which overlies a dark zone of humus, iron and aluminium accumulation. This passes to a yellowish red horizon enriched with hydroxides of iron and aluminium. The associated fine loamy brown earths and brown podzolic soils lie mainly under grassland with associated bracken Pteridium aquilinum and their distribution is related to the underlying rocks. The podzols are very acidic with pH values less than 4.0, while the associated brown earths and brown podzolic soils are only slightly less acid, with pH values around 4.5.
Well-drained fine loamy and silty brown earths occur along the south Gower coast and are commonly shallower than 30 centimetres on the cliffs themselves. The calcareous material and the relatively low rainfall, together with additions of lime and basic slag on the agricultural land, result in these soils having a near-neutral pH of 6.5 to 7.0. Brown podzolic soils are well-drained coarse loamy soils that can be distinguished from the duller-coloured brown earths by layers of bright orange-brown subsoil, which contain larger proportions of iron and aluminium hydroxide.
All the soils on the salt marshes are affected by the high water table and are classed as alluvial gley soils. They have different profiles depending on slight differences in elevation and there are four types recognised, each with a distinctive vegetation community. At the lowest point on the shore there is an extensive area of common saltmarsh-grass Puccinellia maritima and the upper 20cm of the soils show some structural development. On the mid-shore under red fescue the subsoil structures are well developed and the upper 20cm is almost completely decalcified. The highest zone of the marsh, on the landward edge with sea rush Juncus maritimus, the soils are completely decalcified with a pH value of 5.5 in the upper 50cm.
The raw sands of the dunes are composed largely of quartz grains and shell fragments and except on the inland areas where they have been fixed by vegetation they are unstable and there is no visible structure below the top few centimetres. As the dune surfaces are very unstable fresh sand frequently buries old topsoils causing layering. On fixed dunes there can be more than 5cm of topsoil. In dune slacks humic-sandy gley soils form, with surface layers rich in organic matter, overlying gleyed sand. Stabilised dunes on cliff tops such as those at Pennard, where there is little accretion of fresh sand, are sometimes decalcified below 30 cm and the soils are non-calcareous, resulting in an interesting flora.
The plant life of every area, including Gower, is characteristic of the rock, soil and climate. Plants can change their growth patterns relatively quickly, but because changes in the soil occur over long periods there must be an equivalent slow change in plant growth. Plants adapted, for example, to growing on the coarse loamy podzols of the commons may not be able to survive on the brown earths covering the limestone. Some plants, however, can grow equally well on both acid and neutral soils if there is reduced competition. Many existing habitats in the peninsula have been largely formed by human activity, but they are so old and so traditional that they have come to be regarded as natural. So commons and heathlands are described in this book as plant communities, even though they need constant management to exist in their accepted form. The intervention of people is the subject of the next chapter.
