The scope of Strategic Environmental Assessment of Irish Sea Area SEA6 in regard to prehistoric archaeological remains
Executive Summary
The SEA6 area comprises the UK sector of the Irish Sea, that is the bulk of the Irish Sea itself, and part
of the coastal waters of Northern Ireland. In order to understand and prioritise the nature of prehistoric
archaeological sites which might occur on the sea floor this report considers the context of all the
adjacent land masses, including the Irish Republic, and to the south the Celtic Sea and western
Channel. Prehistoric submarine archaeological remains back to a date of about 225,000 years ago,
Palaeolithic, Mesolithic and Neolithic, could occur with low probability in many parts of the SEA6 area.
Palaeolithic archaeological sites as old as 225,000 years Before Present (BP) occur at locations on the
Welsh coast, with a great density of sites from the later Mesolithic and Neolithic. Some sites therefore
pre-date the last interglacial high sea level, and although they were covered by the Devensian ice
sheet, material inside caves survived.
The penultimate glaciation, the Anglian covered the whole Irish Sea, and southern Wales and Ireland,
but the last one the Devensian, left the southern Irish Sea exposed, and also left southern Wales and
the southern rim of Ireland exposed. The deep channel known as the Celtic Trough is the cumulative
effect of over-deepened scour by ice sheets and meltwater tunnelling, partially infilled by glacial and
marine sediments. It is deepest at the northern end. Occupied caves co-existing with the last ice sheet
have been found in southern Wales, but not in southern Ireland. Particular attention is given in this
report to Oxygen Isotope Stage 3, 64,000-22,000 years before present, for which sophisticated
reconstructions of the climate and vegetation of western Europe have recently been made.
There are also Palaeolithic sites in southern England, the Channel Islands, and Brittany. At the time of
the maximum glaciation of the Devensian, people were probably living on the floor of the English
Channel and the exposed plains of the Celtic Sea, as well as in Cornwall, and on the ice edge in
southern Wales itself. Submerged prehistoric sites have been found in the English Channel. The reoccupation of Wales and Ireland could have been as much from the south, as from the east, when the
ice retreated. This could explain some of the highly developed Mesolithic sites in Ireland which seem to
pre-date the equivalent structures on the British mainland. The rising sea level separated Ireland from
the rest of Britain at about 12,000 years BP, although brief land contact may have been re-established
around 10.5-9.5 ka BP due to isostatic uplift.
Evidence from the northern North Sea and the Russian Arctic suggests that some prehistoric peoples
occupied the exposed shelf area during late glacial periods utilising Inuit-style survival methods, and
butchering marine mammals. If this proves to be the case, there may be unexpected occurrence of
early prehistoric sites, Late Palaeolithic, on the Atlantic shelf, and in the southern Irish Sea, where
people could have been hunting marine mammals. Pipe entrenching is the process in the oil and gas
industry which is most likely to disturb prehistoric archaeological deposits. Commercial site
investigation using acoustics and coring could provide beneficial new archaeological data.
Reinterpretation
of existing acoustic data should be carried out whenever possible in order to reveal the
Pleistocene low sea level landscape and drainage. The paper concludes with tentative suggestions for
discussion of protocols and a reporting regime.
2. Overview of known and likely areas with prehistoric archaeological
remains, with mapped indications of relative likelihood of the presence of
remains (sensitivity mapping) and with hotspots identified.
2.1 During the last million years the British landmass has been connected by dry land to the
mainland of Europe for far more time than it has been separated by sea, and the Irish Sea was dry land
or blocked by ice sheets for much of this period (Fig. 1). The earliest occupation of the British mainland
by hominids, Homo heidelbergensis, occurred about 500,000 years Before Present (BP) (Pitts and
Roberts, 1997), and recent evidence suggests that it could be as early as 700,000 years BP. In Wales
three stages of human occupation can be identified, with early Neanderthals at Pontnewydd Cave near
Llandudno dating to about 225 ka (thousand years) BP, Coygan Cave with classic Neanderthal at about
50 ka BP near Tenby, and Paviland Cave with modern humans at about 26 ka BP, on the Gower
peninsula. Of these sites Coygan and Paviland are very close to the present shore, while Pontnewydd
is 8km inland. In addition, Kendrick's Cave, on the Great Orme peninsula near Llandudno, has
revealed Late Upper Palaeolithic materials dated to 10 ka BP. Lynch et al. (2000, p.9, Fig. 1.4) map 28
major Palaeolithic sites in Wales, of which approximately half are either on or very close to the coast.
The proximity of occupied Palaeolithic caves so close to the shore suggests that sites of similar age
might be found on the continental shelf. In principle human and proto-human artefacts may therefore
have been deposited in sediments or caves on the continental shelf at any time in the last half million
years whenever the glacial control of world sea level caused the floor of the Irish Sea to be exposed.
However, the facts that the ice sheets of the last two glaciations both scoured the floor of the Irish
Sea, and that great thickness of infill and sediments has accumulated in some areas since the last
glaciation, make the survival and discovery of sites problematic.
2.2 The Ancient Human Occupation of Britain (AHOB) project led by Chris Stringer at the Natural
History Museum has analysed the various phases during which hominids could cross into the British
isles, and when they were either isolated, or absent (Fig. 2.). From this figure it can be seen that
Britain was cut off from the mainland of Europe only briefly for about 10 ka at the last interglacial 125
ka BP, and again for longer at about 200-230 ka BP, and briefly at 330 ka and 400 ka. At all other
times in the last half million years you could walk across the English Channel, or live there. At the
same time, large parts of Scandinavia and northern Britain were covered in ice sheets, which
sometimes extended as far south as the Thames, and covered almost the whole of Wales and Ireland
(Figs.3, 4). Thus the favourability for hominids living on the British Isles was determined as much by
the proximity of the ice sheet as the contact with France or the Netherlands. Human sites do not exist
at all on the British Isles between the dates 21-13 ka BP. The occupation of the English Channel,
Ireland and the floor of the Celtic Sea at this time is possible, and people would then migrate back to
Ireland and the floor of the Irish Sea when the ice melted, since people were certainly driven out of
these areas completely when the ice expanded to its maximum. On the other hand, if people were
living on the floor of the Celtic Sea, or on the Atlantic margins, the migration routes may have been
more complex. In Stringer's analysis, Pontnewydd appears as the type site for the last interglacial,
stressing the importance of this region for the whole of the British Isles.
2.3. Van Andel and Davies (2003) have published a multi-disciplinary analysis of the climatic
fluctuations during Oxygen Isotope Stage 3, approximately 60-24 ka BP, and the consequent effects
on the distribution of Neanderthal and anatomically modern humans (AMH). The study consists of a
concatenated sequence of models describing the temporal and regional variation of temperature,
precipitation, seasonal variability and extremes, snow cover, wind speed, vegetation, fauna, wind-chill
factor, and habitability for hominids. During OIS-3 the Greenland ice core data GISP2, (Meese et
al,.1997; Johnsen et al, 2001) show rapid fluctuations of temperature of the order of 5-100C every few
thousand years, the so-called Dansgaard/Oeschger oscillations. The models are run on a 60 x 60km
grid resolution (Van Andel and Davies, 2003, p.58) and this necessarily limits the accuracy, as well as
there being some uncertainties as to how one model output relates causally to the next model.
Nevertheless, the sequence of calculations and plotted maps, correlated with summaries of known
major archaeological sites, provides a thought-provoking analysis. The maps and calculations should
be used in future as a starting point in studies which attempt to understand where people would have
been living on the sea floor at times of low sea level. Unfortunately Van Andel and Davies (2003) do
not take into account any archaeological data from the seabed, or any of the known occurrences of
fossil fauna or human occupation sites on the sea floor of North West Europe.
2.4 The ice sheet of the penultimate glaciation, the Anglian, covered the whole of Ireland and
Wales, and the floor of the Irish Sea (Fig. 3). The Anglian glacial maximum was at about 300-350 ka
BP. The last glaciation, the Devensian, with a maximum at 20 ka BP, was slightly smaller in extent,
and the ice sheet stopped just short of the southern coasts of Wales and Ireland, leaving small strips
of the present coastal lands free of ice (Fig. 4). The great majority of Palaeolithic sites in Wales are
on Carboniferous Limestones, suggesting that they survive in caves. Most open sites would have
been destroyed by later ice movements. Also, rather curiously, some of the southern Welsh
Palaeolithic sites were very close to the ice margin, suggesting that AMH, who arrived in central and
northern Europe about 30 ka BP, could adapt efficiently to living in very cold conditions.
2.5 The possibility or probability that humans were living in the northern borders of England and
around the fringes of Scotland before 10,000 BP is being investigated at the University of Newcastle
by Dr Penny Spikins through a project entitled "Submerged Archaeological Landscape Team" (SALT).
A post-graduate dissertation by Miriam Cantley entitled "Is there a convincing argument for late-glacial
occupation of Northern Britain?" (University of Newcastle, web-site, 2004) is directly relevant to the
present assessment. This work is not yet complete. The question of pre-glacial maximum occupation
of Scotland was discussed in the reports on SEA4 and 5 by Flemming (2003, 2004).
2.6 The Mesolithic period, starting about 10,000 years BP, introduces the technologies of hut
construction, more sophisticated tools, microliths, sophisticated hunting techniques, recoverable
canoes, and more evidence of fishing and the use of coastal resources. Forests were spreading over
the British Isles, with birch, juniper, and pine progressively giving way to larger deciduous trees, and
finally the landscape was dominated by oak, elm, lime, ash and alder by 7,000 years BP. Mesolithic
peoples used fire to clear forest and improve the scope for hunting. Because the sea level was still
rising, sites which are now situated on the coast would actually have been several miles inland, and
hunters were probably roaming the coast plains which are now submerged (Lynch et al., 2000, p. 28).
The presence of drowned forests and submerged freshwater peats at several locations on the Welsh
coast confirms the preservation, at least in part, of this drowned terrestrial landscape (Steers, 1948,
p.125, 140, 144-5).
2.7. Anatomically modern humans (AMH) were present in Ireland by about 10 ka BP in the early
Mesolithic, and the transition to Neolithic farming only took place in 6000 BP, the last agricultural
revolution in Europe apart from northern Scandinavia ((Malone, p.11). The population density during
the Mesolithic was extremely low, about 1 person per km2, with variation from 5/km2 in the densest
areas, and 1/20km2 in the lowest densities (Malone, p.18). Before the peak of the last glaciation the
fauna of Ireland included woolly mammoth, brown bear, arctic fox, the Irish giant deer, and reindeer.
Archaeologists have made extensive searches to discover signs of human occupation before the
glacial maximum, but nothing definite has been found. Harbison (1988, p.17) describes some
interesting flints which were supposed at various dates to be Palaeolithic, but concludes that none of
them were both genuinely Palaeolithic and in situ. Curiously, a Palaeolithic worked flint found at Mell,
near Drogheda, was probably ??dislodged from a layer perhaps a quarter of a million years old.
Mitchell?s interpretation was that it was a piece of Palaeolithic hunter?s waste which had probably
been deposited somewhere in the base of the Irish Sea before being brought {inland} by icemovement??
(Harbison, 1988, p.17). The analysis so far establishes the broad picture that there was
pre-glacial maximum occupation of Wales, with archaeological material surviving close to the coast,
while artefacts identified beyond doubt as in situ only establish AMH occupation of Scotland and
Ireland after the Last Glacial Maximum (LGM).
2.8 Figures 5 and 6 (Lambeck, 1995, Shennan et al., 2000b) show the sequence of ice sheet limits,
coastline, and the impact of rising sea level on the British Isles, including Ireland, for the melting of the
Devensian ice cap. Lambeck (1995) and Shennan et al. (2000a, 2000b) have produced models
which combine the compensation for the addition and removal of the weight of ice (Glacial isostatic
correction) and for the removal and addition of weight of water (Hydro-isostatic correction) during the
rising sea level (see Figs 5 and 6). In Figures 5(a)-5(d) we see the north British ice cap melting
rapidly from 22,000 to 14,000 years ago. As the weight of ice is removed the land rises faster than
the global sea level, so that the area of dry land increases throughout this period, both northwards
and south-westwards. By 12,000 years BP the sea is beginning to overflow the land (Fig. 5(e)) and,
although a small ice cap forms briefly around 10,000 years BP, the sea continues to rise faster than
the land, forming deep bays and gulfs penetrating into the North Sea, isolating Dogger Bank, and
separating the Straits of Dover about 7,000 years BP (Fig. 5(h)). In the Irish Sea by 18,000 BP the
Celtic Trough appears as an over-deepened basin filling with meltwater from the north, as a
periglacial lake (Fig. 5(b)). This situation continues to 14,000 BP with an extensive land-mass to the
south linking Ireland, Wales, Cornwall, the Celtic Sea basin, the English Channel and Brittany.
2.9 By 12 ka BP a channel of sea water breaks through at north and south, and Ireland is
separated from Britain and the European mainland. The Isle of Man remains connected to the
mainland until 10,000 BP. A variation of this scenario is provided by Tappin et al. (1994, p.87, Fig.
65), with the initial breakthrough earlier at about 16,000 BP, followed by the isostatic uplift of the land
overtaking the sea level rise from 10,500 to 9,500 years BP producing land connections from
Cornwall to Waterford in Ireland, and then from Bardsey Island in Wales to Wicklow Head in Ireland.
This sequence is derived from the analysis of sediment deposits on the floor of the Irish Sea, rather
than the generalised eustatic+isostatic model for the whole north-west shelf.
2.10 Fitzhugh (2002) sets out the strong evidence for early human exploitation of the food resources
of the circum-polar zone, using life-strategies similar to modern Inuit or Eskimos. Zhokhov Island, north
of Siberia, in the Laptev Sea, is the northernmost Arctic site occupied at 8400 years BP (Pitulko, 2001).
Excavations at the Mamontovaya Kurya site on the Usa River, inside the Arctic circle, revealed stone
tools and carved mammoth tusks nearly 40,000 years BP (Pavlov et al., 2001). The exploitation of
marine mammals, especially seals, walruses, and cetaceans must be considered for peoples living in
circum-polar conditions. Anyone who has seen a walrus haul-out will know how clumsy the animals are
on the beach. They would have been the most attractive prey for any peoples who chose to live on the
northern or north-west margins of Europe during peak glaciation, or as early as, say, 12-14,000 years
BP. The reported recovery by Dutch fishermen of walrus bones showing signs of cut-marks and
butchery from 56o North in the central North Sea (Klaas Post, personal communication) strongly
suggests this possibility. This type of culture may correlate with the otherwise curious retrieval of a
lithic artefact off the Viking Bank from a depth of 145m (Long et al., 1986). Wickham-Jones has pointed
out (2003, personal communication) that the availability of large quantities of fat from marine mammals
is an important component of a glacial climate diet, since the hunter-gatherer diet inland tends to be too
lean to support human survival in extreme cold.
2.11 Pitulko et al. (2004) show that modern humans were living in the Russian Arctic at 72o North on
the delta of the Yana River about 30,000 years BP. Tools of stone and bone were found. The area
was never covered by thick ice, and remained suitable for large herbivores throughout the last
glaciation. This site and others of similar age (Pavlov et al., 2001) show that people were living in the
high Arctic before the last glacial maximum. In the region of Scandinavia, in the areas where the ice
cap and glaciers were presumably similar to present conditions at the last interglacial, any population
would have been forced to migrate outwards as the ice thickened and grew in extent. While humans
may have exploited floating sea ice and the peri-glacial tundra for terrestrial and marine mammals, they
cannot have existed very close to the ice cap itself, or in the land areas actually covered by ice. This
argument applies equally to any population of Scotland after the last interglacial, if any. Thus
determination of the limit of the ice sheet at different dates is critical to understanding where people
might have lived, exploiting an Inuit-style of life on the outer margins of the continental shelf during the
glacial maximum. The failure to find pre-Devensian sites in most of the northern UK, other than a few
cave sites, is in marked contrast to the Russian experience. Professor Ole Gron (personal
communication, 2005) has expressed the view that this failure may be due the earlier archaeological
materials being buried beneath a thick sterile layer below the post-Devensian occupation site. In
Russian excavations it is not uncommon to excavate through a post-glacial site, then several metres of
deposits with no anthropogenic signs, and then to find a deeper pre-glacial archaeological site below
that. Such persistence is not common in western Europe.
2.12 In northern peri-glacial conditions the availability of protein for prehistoric peoples close to the
shore was higher than in the hinterland (Momber, 2000, 2001; Fischer, 1995). At glacial maximum
when the sea did not penetrate into the Irish Sea or North Sea area any inhabitants would have
depended on large mammals such as mammoth and reindeer. Typical maps and discussions of the
food base tend to emphasise the availability of terrestrial mammals on the continental shelf (e.g.
Barton, 1997, p.134). Fischer (1995, 2002) has added the importance of coastal fish and shellfish.
Later, Mesolithic peoples would have benefited from the resources of wetlands and estuaries.
Flemming (1996) summarises the reasons for prehistoric peoples being attracted to the coast, and
estimates that, as sea level fell, vegetation and fauna would colonise the exposed land close to the
shoreline within a few decades. Bailey (2003) has recently summarised the strong case for coastal
dwelling during prehistoric times.
2.13 Human remains in south Wales have been found a few km from the ice front (Woodcock, 2000,
p.404; Lynch et al., 2000, p.8), so cold itself was not a deterrent. Palaeolithic hunters required fresh
water, food supplies, a supply of flints, bone and wood to make weapons and tools, some timber for
fires, shelter, skins for clothing, and a secure position which might have to be defended, with good
routes of access, and the option to move or migrate with the seasons, or with changing supplies of fish,
shellfish, or mammals. Mesolithic settlements with constructed dwellings were often positioned so as to
be convenient to fish traps and fish weirs on the coast. Knowledge of these requirements has been
used with great success by archaeologists in the UK and Denmark to predict and interpret submerged
Mesolithic sites (Andersen, 1980; Pedersen et al., 1997; Momber, 2001; Coles 1998, 1999, 2000;
Fischer, 2004; Skaarup and Gron, 2004). Hunting kill sites, flint quarries, flint-knapping sites,
settlements, camps, shell middens, charcoal from fires, and shelters, tend to cluster round shorelines,
estuaries, lagoons, headlands and promontories.
2.14 This places a premium on identifying accurately the ice limits, shorelines and rivers at each
date, and especially those shorelines where the sea level was locally constant for hundreds or
thousands of years, relative to the local land. Under these conditions rivers would tend to create
stable estuaries, and perhaps barrier bars or lagoons and wetlands, waves would erode substantial
rock terraces, cliffs, and caves, and shallow water sediments or peat could accumulate. Because of
the doming of central Scotland the previous shorelines with terraces and caves have been uplifted in
many areas, and several occupied caves are known on raised terraces around Oban (Wickham-
Jones, 1994 p.71-73). The shorelines of Northern Ireland and Scotland on both sides of the North
Channel are still uplifting as part of the outer fringe of the Scottish post-glacial isostatic rebound.
2.15 Taphonomy is the study of the changes which occur to deposits after primary burial.
Archaeological materials may be covered by metres of sediments which protect them indefinitely, or
eroded by ice, eroded by rivers, eroded and scattered by surf action on a beach, eroded by bottom
action of storm waves in shallow water, eroded by tidal currents, chemically altered, or disturbed by
trawling, dredging, entrenching, or drilling. There is insufficient space in this report to discuss all the
processes, conditions, and topography which are most favourable in every combination of
circumstances for the survival of an archaeological artefact in situ which is submerged for at least part
of its existence. The typical conditions for the survival of known submerged archaeological prehistoric
sites are presented in a table by Flemming (1983, p.161-163) classified as Ria, Lagoon, Estuary,
Sheltered alluvial coast, Exposed accumulating beach, Submerged sea caves, Karstic caves, and
Islands and archipelagos. Each site is classified in terms of depth, age, tidal range, current, wind
fetch, and estimated wave action. Peat and submerged forests are important indicators, and Figs 19
and 20, in Louwe Kooijmans (1970/1), illustrate the widespread occurrence of peat on the floor of the
North Sea.
2.16 The topography, seabed geology, and unconsolidated sediments of the Irish Sea are described
by O'Cofaigh and Evans (2001), Tappin et al. (1994) and Jackson et al. (1995). See also Holmes and
Tappin (2005) in the SEA6 technical report series, especially Figures 7 and 13. The detailed analysis
of the regional sediments, with the probability of finding submerged prehistoric sites, will be discussed
in Section 2 27. In broad terms the Quaternary sediments are less than 50m thick over most of the
SEA6 area, with the exception of the central axis of the Celtic Trough, where they can be as much as
300m thick (Fig. 7). The Celtic Trough is flanked on the west by the Irish Platform, and on the east by
the Eastern and Welsh Platforms, where the water depth is usually less than 60m, and the
Quaternary deposits are partially absent, especially around Anglesey. The Celtic Trough itself, and
scattered elongated depressions on the shelves, are interpreted as of glacigenic origin, with three
successive generations of incisions (Jackson et al., 1995, p.85). The Quaternary deposits consist of
glacial tills, clay, pebbles, mud, and some interglacial and artic-like glaciomarine beds. Detailed
cross-sections are shown by Jackson et al. (1995, p.90-92) and Tappin et al. (1994, p.80-81).
Because of the relatively shallow water, and the strong tidal current regime, much of the seafloor is
covered in active, mobile, modern marine bedforms (Fig. 8), gravel furrows, gravel waves, sand
ribbons, and sand waves. North of 53o N the seabed shows relict bedforms from terrestrial
periglacial conditions (Fig. 9) including polygonal periglacial patterned ground and roches moutonn?e.
2.17 Tappin et al. (1994, p. 86-91) describe the relict seabed features and modern marine
sediments south of 53o N. Within the SEA6 sector there are Pleistocene glacial deposits forming
seabed outcrops, fault scarps, infilled channels, and the so-called "sarns". In Cardigan Bay there are
three low ridges of cobbles and boulders stretching out to sea hundreds of metres perpendicular to
the coast. In legend these have been supposed to be the remains of ancient buildings. North (1957)
showed comprehensively that these features are natural glacial features. Tappin et al.( 1994, p.86)
describes them as clast-supported clayey diamictons, covered by gravel, cobbles, and boulders.
Garrard and Dobson (1974) conclude that they are late glacial median moraines formed by the
glaciers descending from the Cambrian Mountains. Tappin et al. (1995, p.86) comment that the sarn
ridges overlie deep incisions, and are similar to the so-called St Patrick's Causeway on the south
coast of County Wexford leading to the Saltees Islands. The origin of these ridges, while undoubtedly
glacial, may be a complex result of ice-rafting, flooding, and moraine material.
2.18 In view of the work of Pitulko et al. (2004) it is important to consider the effect of sea water
rising over archaeological deposits in permafrost, which would indicate the possibility of good
preservation of artefacts. Although other factors also apply, for example ice scour, glacial erosion,
frost shattering, and normal subaerial erosion processes, the critical period for survival of an
archaeological deposit is the time when the surf zone starts to impact on the site, and the ensuing few
hundred years as the sea level rises over the site, and coastal shallow water waves are breaking over
the site, or washing into a cave mouth. Favourable factors for survival in the deposit area include:
- Very low beach gradient and offshore gradient so that wave action is attenuated and is
constructional in the surf zone.
- Minimum fetch so that wave amplitude is minimum, wavelength is short, and wave action on the
seabed is minimum.
- Original deposit to be embedded in peat or packed lagoonal deposits to give resistance and
cohesion during marine transgression. Drowned forests and peat are good indicator
environments.
- Where deposits are in a cave or rock shelter, roof falls, accumulated debris, concretions, breccia,
conglomerate formation, indurated wind-blown sand, all help to secure the archaeological strata.
- Local topography contains indentations, re-entrants, bays, estuaries, beach-bars, lagoons, nearshore
islands, or other localised shelter from dominant wind fetch and currents at the time of
transgression of the surf zone.
- Frozen ground or permafrost enclosing archaeological deposit at time of inundation.
2.19 This brief analysis demonstrates that survival or destruction of an archaeological deposit,
whether originally inland or on the coast, depends acutely upon the local topography within a few
hundred metres or a few km of the site. Generalised coarse resolution maps tend to omit the details
which show the necessary local topographic clues. The BGS 1/250,000 maps, although they are
primarily designed to present sediment data, provide a much more accurate representation of
topography, with isobaths at 10m intervals, than the Admiralty Charts. Additional high resolution
swath bathymetry would be enormously valuable in detecting probable sites. It is no coincidence that
the most prolific area of proven submerged Mesolithic sites is between the islands of the Danish
archipelago, where many hundreds of sites have been mapped and sampled by the National Museum
Maritime Archaeological Institute, and the National Forest and Nature Agency, assisted by amateur
divers (e.g. Skaarup and Gron, 2004). Further submerged Baltic sites have been discovered in
sheltered waters off the coast of northern Germany (Lubke, 2001, 2002). The Bouldnor Cliff site in
the lee of the Isle of Wight on the Solent is protected in the same way. Off Gibraltar a hook-shaped
submerged promontory contains caves facing inwards towards the land which would be protected
from waves while the sea level rose (Flemming, 1963; 1972), and similarly protected sea caves have
been found in the Bay of Villefranche (Flemming, 1972). The ability to reconstruct the conditions
under which Irish Sea archaeological sites were formed and buried has recently been improved by the
sophisticated analysis techniques of Praeg (2003) and Gaffney (2004). Praeg (op. cit.) has used
seismic imaging to detect buried glacial tunnels under modern sediments. Gaffney (op.cit.) has reinterpreted extensive sub-bottom seismic records to detect the changes in sediment characteristics
indicating buried river valleys. This technique has exposed a wide meandering river draining
northwards from the north-east flank of the Dogger Bank in the North Sea, and is being tested on
other parts of the UK shelf.
2.20 The factors in the previous paragraphs are those which promote survival of the original deposit
in situ. However, if an archaeological deposit is buried under 5-10m of mud or sand it will not be
discovered, except in very unusual circumstances. Thus the final requirements for survival and
discovery are:-
- Low net modern sediment accumulation rate so that the artefacts are not buried too deeply.
- No fields of sand waves or megaripples over the site.
- Ideally, a slight change in oceanographic conditions so that the site is being gently eroded to
expose deposits when visited by archaeologists. (This factor is sufficiently common in known sites
to be a serious factor, and should not be regarded as an unlikely fluke).
2.21 Potential discovery "hot-spots" in the SEA6 cannot be listed exhaustively at this stage. The
steps needed to create high resolution local sensitivity maps can be identified, and are discussed later
in this section. In principle the key factors are:-
25
- "Fossil" estuaries and river valleys.
- The flanks of banks and ridges which have been proven to have peat layers, or which are likely to
have peat layers.
- Valleys, depressions, or basins with wetland or marsh deposits.
- Nearshore creeks, mudflats, and peat deposits.
- "Fossil" archipelago topographies where sites would have been sheltered by low-lying islands as
the sea level rose.
- Niche environments in present coastal zones, wetlands, intertidal mudflats, lochs, and estuaries.
- Caves and rock shelters in re-entrant bays, fossil erosional shorelines, submerged rocky shores
protected by other islands, or in archipelagos.
- Deposits of sediments formed within, or washed into rocky gullies and depressions.
- "Fossil" coastal sites comparable by analogy to modern Inuit migratory sites, adjacent to sea ice,
giving access to marine mammals as a food resource.
- Areas of permafrost containing archaeological deposits which were then inundated, and protected
by other factors listed above.
2.22 The changes in and survival of an archaeological site, and the chances of discovery, depend
on the present conditions of winds, waves, and currents in the area, and the water movements on the
seabed.