South Devon Coastal Group: Introduction

1. Coastal Geology and Coastline Planform - References Map

This coastline is broadly defined by the large, shallow arcuate embayment of Lyme Bay to the east and a succession of smaller, headland-confined, bays south of the Teign estuary. It is dominated by cliffs, rocky platforms and mostly coarse clastic beaches, but "soft" shoreline features are associated with the region's two major estuaries. Here, banks, bars and shoals have accumulated both landward and seaward of estuary mouths. Dawlish Warren is a major spit structure substantially composed of sand, with superimposed dunes.

The major coastline discontinuities are provided by headlands and estuary re-entrants. The latter are the product of Holocene submergence of former seaward extensions of the major regional river valleys (Section 3). All of the headlands are the result of long-term differential erosion with combined marine and sub-aerial denudation selecting the most resistant rock lithologies. The most prominent salients, such as Hope's Nose, Berry Head and Start Point have evolved through a long succession of relative sea-level changes and major climatic fluctuations during the Quaternary period (see Quaternary unit).

The rock substrate between Start Point and Hope's Nose is composed of very approximately east to west striking units of Devonian gritstones, slates, sandstones and crystalline limestones. The latter account for both Berry Head and Hope's Nose, though local spatial changes in rock type and their resistance are due to faulting, folding and thrusting induced by tectonic movements. In the extreme south, mica and horblende schists account for the Start Point promontory. Many minor coves and small bays have been eroded into areas where rock strength is reduced by lines or zones of weakness. The extent of this exploitation is also partly determined by degree of exposure to wave energy.

North of Oddicombe Beach, Torquay, the open coastal planform between estuary inlets is less indented than it is further south. From here, eastwards to Lyme Regis, there is a nearly conformable sequence of Permo-Triassic and Jurassic strata, with an overall eastwards dip interrupted locally by faulting movements. The Permo-Triassic strata are predominantly sandstones, conglomerates, breccias and mudstones. These frequently support near-vertical cliffs, but landslipping occurs in the area immediately west of Budleigh Salterton where sandy, permeable strata overlie an impermeable, clayey lithology. East of Sidmouth, Jurassic and (overstepping unconformable) Cretaceous sandstones and limestones, overlying Triassic mudstones and marls, generate large-scale, complex rotational and translational slides and slips, as well as active mudflows. This response is particularly well-developed in the vicinity of Beer Head and in the Axemouth to Lyme Regis Undercliff. Local variations in dip, imposed by subsidiary folds, dictate the scale, detail and frequency of coastal slope failures.

The only sectors of coastline where geological structure and rock lithology fail to make a contribution to geomorphic character are where barrier beaches have been driven shorewards under conditions of rising sea-level but remain partially detached from the shoreline. The finest example is Slapton Sands, between Torcross and Strete, which has trapped the Higher Ley, Slapton Ley and Lower Ley lagoons in front of the ancestral coastline. Much smaller scale examples occur along the Torquay frontage (e.g. seawards of Torre Abbey), and arguably at the mouths of the Otter and Axe rivers. The role of geological controls on coastal morphology is also suppressed or modified where there are substantial defence and protection structures (Sims, 1998). This has yet to have any significant effect on shoreline change, but will eventually do so as 'hardened' frontages remain static in relation to adjacent erodible and dynamic sections of coast. This is particularly the case where there are several unconnected protected sections, e.g. within Tor Bay, where 45% of the coast has been defended by sea walls, revetments, armouring and gabions.

2. Hydrodynamic Regime - References Map

Tidal Regime

Mean spring tidal range on the open coast becomes progressively larger from east to west. A range of 3.6m at Lyme Regis gradually increases to 4.7m at Start Point. Intermediate locations have ranges of 3.8m (Exmouth); 4.1m (Teignmouth); 4.2m (Brixham) and 4.4m (Dart estuary entrance). Tidal currents are generally weak (less than 0.5m.s^-1) within the major bays, but at, and immediately seawards, of the mouths of the Exe and Teign estuaries maximum (spring) ebb current velocities can exceed 3.0m.s^-1. A tidal vortex (gyre) is set up by the protrusion of Start Point into the offshore rectilinear residual tidal streams of the English Channel that generates velocities as high as 2.0-2.2m.s^-1 during a part of the tidal cycle. Tidal eddies, with an anticlockwise circulation, occur at and seawards of the mouths of the Exe and Teign. In each of these locations, tidal currents play a significant role in complex partially closed systems of sediment circulation. Offshore, net eastwards flood current flow is confined to nearshore areas of relatively shallow water; mean ebb current flow is directed westwards. These characteristics are the outcome of the constrictions imposed on the regional tidal amphidromic system by the shape and depth of the English Channel (Posford Duvivier, 1998).

Wave Climate

The regional wave climate is dominated by Atlantic swell waves that approach from the south-west. These suffer (i) attenuation due to the wide extent of relatively shallow water, especially in Lyme Bay; (ii) refraction imposed by the sequence of major headlands between Start Point and Straight Point; (iii) diffraction due to outer estuary banks and shoals. Wave penetration into the estuarine inlets is excluded by the presence of offshore banks (ebb tidal deltas) and enclosing spits. Higher energy waves generated over fetch areas to the east and south east operate over relatively short periods, especially during the winter. These often cause significant short-term reductions of beach volume and elevation.

Arber (1940), Hydraulics Research (1989) and others, have noted that the asymmetric quasi-zetaform planshape of the major bays are adjustments to the regional wave climate. The longer, smoother segments face the direction of dominant wave approach, whilst the shorter, more indented lengths of shoreline occur to the immediate lee of major headlands (usually to the east). Variations in relative rock resistance are probably a more direct control of bay shape.

Changes in shoreline orientation resulting from both large-scale planform and variation at smaller spatial scales determine the precise nature of wave climate at specific locations. Thus, for example, the coast between Hope's Nose and the mouth of the Exe estuary is significantly more protected than that east of Straight Point.

Characteristic mean significant inshore wave heights have not been calculated in any detail for this coastline, as a whole; some site-specific data is available (e.g. Hydraulics Research, 1992) and is quoted in the appropriate regional sections. Values range from 5.3m to less than 1.0m (for annual return frequencies), according to location. Posford Duvivier (1998) report the deployment of a wave-measuring buoy for a one year period, starting in July 1997, off Start Point. Data from this programme was not, however, available for this report. As most previous work has employed numerical modelling, using hindcasting approaches based on regional wind speed data, this new data set will be a valuable contribution to quantifying wave climate. The relative frequencies of wind and swell waves, of a range of heights and from different directions, are presented as rose diagrams in Posford Duvivier (1998). This is based on Meteorological Office synthesised offshore data for the period 1988-1997 and illustrates the importance of mean significant wave heights, between 0.5 to 3.0m, characteristic of waves propagating across easterly, south-easterly and southerly fetches.

3. Shoreline Evolution - References Map

Sea-level movements, in response to the major climatic fluctuations of the Pleistocene period, have had a major impact on present day coastal geomorphology. This is also the case for the coastal hinterland, where a sequence of ten fluvially and marine eroded partial erosion surfaces have been mapped within the region's major drainage basins (Orme, 1960; Brunsden, et al 1964; Kidson, 1977). The highest (and presumed to be the oldest) surfaces of apparent marine origin are at approximately 220m O.D. It is difficult, even with several millions of years available, to reconstruct such a high sea-level. The only alternative is to assume post-formation tectonic elevation, but this would be likely to create differential rates of uplift within such a large region accommodating numerous pre-existing lines of structural weaknesses. Yet the field evidence does not support this approach, and there is no clear evidence for tectonic/structural displacement of much more recently created raised shorelines. This problem, which is not entirely academic, has yet to be resolved. It applies equally to other areas of western Britain and north-west France.

The modern coastline is likely to have evolved through various stages of low and high sea-levels during the Quaternary. The most recent, and continuing, (Holocene) stage of sea-level rise has therefore re-occupied an ancestral coastline, though it has added new features of both submergence and emergence.

Low Sea-Levels - References Map

During several extended periods of cold climate in the Quaternary period sea-level fell by up to at least 100m, exposing extensive areas of the contemporary sea bed. During these periods, intense periglacial (artic tundra) conditions prevailed throughout the South-West Peninsula (Cullingworth, 1982). This produced a large yield of sediment, through weathering and slope mass movement. Summer melt-season, linear erosion by much higher gradient rivers would have been strong, thus spreading large quantities of sediment across what are now Lyme, Tor and Start Bays. This has been mapped in some detail by Clarke (1969), Hails (1974, 1975) and Hails, et al (1975), all of whom identify large areas of pre-Holocene coarse clastic (gravel) and sand spreads. This sediment has been available for post last glacial (Devensian stage) wave-constructed - beach and spit building by rising sea-level. However, it is evidently a finite reserve that is not being replenished under contemporary conditions.

Evidence for Quaternary low sea-levels derives mainly from buried river channels; submerged rock platforms and cliffs, and offshore/nearshore sediment stratigraphies. Several authors have reported largely infilled buried channels incised into bedrock; most are below or are extensions of the estuarine lower reaches of modern rivers. Durrance (1974) used borehole investigation and seismic survey to locate a buried channel beneath the Teign estuary. This occurs at -10.2m O.D. at the estuary head, and -22.9m O.D beneath the Denn, and is infilled with fluvial gravels and silty clays. There are also three sets of corresponding terraces at -10; -14 and -23m O.D. Durrance regards this channel as late Devensian in age. Beneath the lower Exe valley and the Exe Estuary, there are two distinct buried channels (Durrance, 1969, 1974). The lowest and oldest is cut into bedrock to below -50m O.D., and the sediment infill of the younger channel to an approximate depth of -30m O.D. Both are associated with a river terrace sequence, which occur at -5.8m; -10.4m; -14.0m and -22m O.D. below the modern channel between the distal end of Dawlish Warren and Exmouth (Durrance, 1974). The more recent channel is probably contemporary with that below the Teign. Shallow coring and stratigraphical interpretation of offshore sediments between the mouths of the Exe and the Teign (Clarke, 1969) suggests the possibility that the proto-Teign was a tributary of a southward-flowing extension of the River Exe. A buried channel at -46m O.D. offshore Hope's Nose may link with this shore-parallel course of a proto-Exe. If this is the case, it might provide an approximate position from which the modern coastline has receded.

A buried channel entrenched into bedrock, and only partially infilled, at the mouth of the River Dart occurs at -52m O.D. (reported by Durrance, 1974). This has also been detected upstream, at -3.8m O.D., at Maypool. It is not known if there was offshore confluence of the ancestral Dart and Exe rivers during the Devensian, or any earlier stage. No information on buried channels along the East Devon coast has been reported.

Whilst there is no detailed picture of the offshore relief of Lyme, Babbacombe, Tor and Start Bays, some significant local features have been reported. Clarke (1969) identified a buried cliffline, with a base at -48m O.D. close to Berry Head and a distinct north-south, rock 'scarp' approximately coincident with the outer limit of Tor Bay. Donovan and Stride (1975) Kelland (1975) and Hails (1975) note three distinct cliff-like breaks of slope in Start Bay at -43m; -54m and -65m O.D. The shallowest of these is 10m in height and is closely parallel to, though several kilometers seawards of, the modern coastline. Rock-cut platforms at -9m and -15m O.D. also occur in Start Bay (Hails, 1975), but their extent is uncertain because of extensive concealment by seabed sediments. The age of these submerged platforms and cliffs is unknown. Donovan and Stride (1975) proposed that all are pre-Quaternary in origin, as sea-level still stands during the Pleistocene were not of sufficient duration to create such substantial features cut into resistant bedrock. Kelland (1975), however, is more prepared to entertain a middle or late Pleistocene age for the higher elevation submerged cliffs and - by inference - adjacent platforms. They may, however, be multi-stage features due to re-occupation as sea-levels fluctuated throughout the Quaternary period. The effect, if any, of subsequent tectonic and/or isostatic displacement has not been determined but crustal movement at the present time is relatively very slight.

Higher Sea-Levels - References Map

The evidence for shoreline features - principally platforms, beaches and cliffs - that were formed at elevations higher than modern sea-level is more accessible, and has attracted considerable research. Raised beaches and platforms occur between +0.5m and +11m O.D. and form a discernable sequence along parts of Tor Bay and southern Start Bay. Orme (1960, 1962) describes several well-developed surfaces of marine planation between 23m and 4.5m O.D. immediately inland of Start Bay, and at 8.0m and 4.0m O.D., in the Torquay area. All are cut across inclined rock structures, and are best preserved on resistant limestones, schists and intrusive dolerites. Degraded, sometimes multi-faceted or composite, backing slopes are regarded as former cliffs co-terminous with contemporary platforms. Orme (1960) also notes less well defined evidence of raised shorelines between (i) Western Combe Cove and Blackstone Point (south west of the Dart estuary mouth), cut across the Dartmouth Slates at approximately 3.6m O.D.; and (ii) Hallsands to Start Point, cut across micaschist at 3.6 to 8.0 m O.D., but subsequently fragmented by frost and corrosion weathering and marine erosion of major lines of weakness (as below the site of the former village at Hallsands). North of Slapton Sands, the +8m O.D. cliffline between Pilchard Cove and Strete Gate is now being incorporated into the modern shoreline. If there was a shore platform in front, it has been destroyed by later erosion.

Mottershead, et al (1987) provide a detailed morphological and litho-stratigraphical account of shore platforms and raised beach deposits between +8 and +12m O.D. around Tor Bay. They are best exposed at Hope's Nose, Shoalstone Point and on the offshore island of Thatcher's Rock, where they have been planed across Devonian Limestone. Amino-acid racemization has been applied to fossil mollusca in raised beach sediments (Bowen, 1986; Keen, 1995), indicating formation over a range of oxygen isotope stages. These approximately date to the penultimate and ultimate full Pleistocene interglacials. It is probable that all of the platforms were re-occupied at least once during the middle and late Pleistocene (Keen, 1998). Dune sands and periglacial 'head' deposits overlying beach sediments in contact with bedrock surfaces represent subsequent cold climate (low sea-level) conditions. Orme (1962) has interpreted many characteristic abandoned and composite coastal slope profiles along the South Devon coast in terms of these fluctuating climate and mass movement processes.

An independent approach to securing the geochronology of relative sea-level movements prior to the last (Devensian) glacial stage and postglacial (Holocene) period is uranium-series dating of stalagmite (speleothem) floors that seal marine sediments in an elevational sequence of solution caves below the present ground surface of Berry Head (Proctor and Smart, 1991). The advantage of this type of evidence is that it has been less susceptible to post-depositional weathering modification and erosional loss than has exposed raised beach material. There are three main altitudinal groups, shown by speleothem and marine deposits to record quasi-stable freshwater/marine zones of mixing controlled by higher sea-levels that were either slowly transgressive or nearly constant. Uranium-series absolute dates of speleothem deposits correlate closely with those derived from aminostratigraphy, thus confirming the established timescale of environmental changes.

Holocene (Postglacial) Sea-Level Rise - References Map

Much of the evidence for the most recent, and continuing, recovery of sea-level, and its effects on the evolution of coastal geomorphology in the south-west peninsula has been integrated and summarised by Everard, et al (1964); Kidson (1977) and Heyworth and Kidson (1982). Within this broader context, knowledge of the Holocene history of the coastlines of Lyme Bay and Start Bay is relatively limited and fragmentary. Clarke (1969) employed 14C dates of organic horizons revealed in shallow cores of offshore seabed sediments to propose that sea-level was at -43m O.D. at approximately 9,500 years B.P., and that the rate of relative rise was in the order of 1.5m per century between 9,000 and 7,000 years B.P. This would have resulted in a rate of westwards shoreline migration in Tor Bay of 7.6m.a^-1. Radiocarbon dates from buried intertidal freshwater and brackish peats are available from North Hallsands, Slapton Ley and Teignmouth; another comes from the very occasionally exposed submerged forest at Blackpool Sands. Unfortunately, these absolutely dated horizons are too few, and too geographically scattered, to enable a reliable reconstruction of the rate of relative sea-level rise from approximately 6,000 years B.P. If there has been any differential crustal movement responsible for relative displacement of Holocene deposits, the poverty of evidence does not register it. It is, however, presumed that in common with adjacent areas, the rate of sea-level rise reduced to nearly that of the present (1-1.5mm.a^-1) between 5,500 and 5,000 years B.P. Clarke (1969) and Kidson (1977), using evidence from the upper Exe estuary, independently conclude that mean high water was at about -4.0m at 4,000 years B.P., and -3.0m O.D. one thousand years later. A marine clay in the lower Axe valley has been tentatively dated as Romano-British, suggesting a period of rapid transgression during that period.

Using foraminiferal assemblages and peat horizons in gravity-cored sea-bed sediments offshore Berry Head, Clarke (1969) reconstructed a former tidal embayment, infilled with mudflats and saltmarsh, protected by a sand spit and backed by an extensive aeolian dune field dating to about 9,500 years B.P. He interpreted this as the mouth of a south-flowing proto-Exe, which was established during a brief sea-level stillstand. There is insufficient evidence to trace the progressive inundation of this river system as the Holocene progressed. However, it is likely that the sand composing the spit, dunes and offshore bars were reworked to provide part of the sediment cover that conceals bedrock and composes the complex of shoals and banks south and east of the mouth of the Exe at the present time.

Clearer evidence for the contribution of rising sea level to coastal evolution comes from a segmented set of coarse clastic barrier beaches between Hallsands and Strete, Start Bay (Morey, 1976, 1983; Job, 1993). From the provenance of material making up these substantial accretion forms, it is plausible that they were originally part of an uninterrupted 'super' barrier several kilometers seawards of the present coast when sea-level was some -40m to -45m O.D. at circa 10,000 years B.P. They have subsequently transgressed shorewards, trapping lagoons such as the Higher and Lower Ley at Slapton and at Beesands. Morey (1976, 1983) has proposed that a prototype of Slapton Sands formed seawards of a shallow estuary backed by a cliffline dating to the last interglacial (Ipswichian stage). Job (1993) envisages several independent barrier islands, which later became a contiguous structure as they moved both upwards and landwards. From the oldest radiocarbon date from the lagoonal sediments behind the Slapton barrier, it may be inferred that it was at approximately its present position by about 3,000 years B.P. Some confirmation of early to mid Holocene barrier migration is provided by the presence of relict gravel deposits in Start Bay (Hails, 1975). They are similar in composition to the modern barrier beach, with a high percentage of clasts composed of flint, chert and quartzite. None of these can have derived from updrift or local sources, thus excluding littoral transport as a significant source of sediment supply. This critical fact also excludes the possibility of in situ barrier development due to spit elongation.

The segmented barriers that have sealed the mouths of former valleys at Beesands and North Hallsands are approximately the same age as Slapton Sands; peat at the base of the Beesands barrier has been dated to 4,700 years B.P. The relationship between the offshore Skerries Bank to barrier beach development is problematic. If this feature pre-dates the Holocene, it would have had a significant impact on barrier migration due to wave refraction effects. Hails (1975) suggests that without its presence, the "rollover" process would have been faster, with Slapton Sands progressing to the base of the former cliffline. The composition of the Skerries Bank clearly indicates that it is not a relict barrier structure (Hails, 1975), but its precise mode(s) of formation have not been determined.

There is no evidence of any contemporary supply of sediment from offshore sources to sustain the barrier beaches of Start Bay. They must, therefore, be regarded as relict, closed sediment systems. Significant losses and crest cut back in the winters of 1995/6 and 2000/01 at Slapton Sands suggest a negative sediment budget.

The concept of proto-barriers migrating landwards, encountering raised topography (former clifflines) and blocking river mouths, is applicable to several other sites along the south and east Devon coastline. It applies convincingly to Blackpool Sands (possibly a segment of the mid-Holocence Start Bay system), and to several smaller examples in Tor Bay, e.g. at Torre Abbey. Dawlish Warren, a substantial former 'double' spit that has apparently grown eastwards across the entrance to the Exe estuary, may owe something of its mid to late Holocene development to barrier type emplacement. In this case, however, much of the offshore to onshore sediment supply has been sand. Coarse clastic beaches and spits eastwards of Straight Point might also be re-examined as possible barriers, especially in situations where longshore transport appears to be an insufficient source of supply.

Shoreline evolution in historical times is not well documented. A substantial length of this coastline remains natural, but human influences include (i) accelerated estuarine sedimentation due to catchment land management practices and resource exploitation, such as the removal of tree cover and mining; (ii) coastal quarrying; (iii) removal of beach sediments for road metal, facing of buildings, etc; (iv) dredging of navigation channels with consequent spoil removal and dumping; and (v) shoreline defence and protection. The latter has been undertaken since the mid/late nineteenth century to protect urban development and infrastructure, much of it in the form of "hard" engineering to protect recreational and tourist amenities and resources. Examples vary from the seawalls and promenades of Torquay (some 45% of the frontage of Tor Bay is now protected) to the combination of walls, revetments, rock armour, groynes and dune creation and management at Dawlish Warren and the Exmouth frontage.

4. Littoral and Estuarine Sediment Transport and Sediment Budgets - References Map

The open coastline is divided into a sequence of discrete littoral transport sub-cells by the presence of both headlands and estuary mouths (tidal passes). Several of these sub-cells are relatively very small, some little more than large confined beaches. Examples are numerous between Hope's Nose and Langstone Point; along the East Devon coast and north of the Dart estuary. Within the type of sub-cell, input is largely derived from cliff and shoreline erosion, although offshore to onshore transport has been inferred in some cases. By-passing of most headlands by bedload transport of coarse sediments seems to be excluded, except perhaps under very high energy wave conditions. Weak tidal currents make no evident contribution to the movement of even fine material. Net pathways of longshore transport have been determined for longer, less interrupted, beaches (eg. Slapton Sands; Teignmouth; east of Sidmouth). These are northwards or eastwards, according to coastline orientation, but are relatively weakly defined. This is often because frequent reversals of net direction occur in response to changes in incident wave approach. Gross rates of longshore transport are thus significantly greater than net rates, but in only a few locations have either been quantified. Some beach systems exhibit apparent quasi-equilibrium longshore transport, eg. Sidmouth and northern Slapton Sands.

As previously explained, substantial proportion of beach volume is the product of mid to late Holocene sea-level rise inducing off to onshore barrier migration. This is best exemplified by Slapton Sands, Beesands and Hallsands fronting Start Bay, but may apply to several other beaches, particularly east of Straight Point. Dawlish Warren and other, smaller, spits (eg. those enclosing the Otter and Axe estuaries) are probable partial barrier structures, though sustained either at present or in the past by longshore drift. Littoral sediment budgets are, therefore, substantially dependent on the availability of these 'fossil' (relict) stores.

Sediment transport at the entrances to the major spit protected estuaries of the Exe and the Teign is complex. In the case of the Teign, there are several major and minor sandbanks that have an apparently cyclical pattern of anticlockwise rotation, whose movement is the result of both tidal and wave-induced sediment entrainment. Linkages to littoral transport are not yet fully understood. Bank movement and channel migration in the outer Exe estuary exhibit comparable but less clear-cut temporal behaviour; in this case there is probable interlinkage with phases of net accretion, erosion and stability of Dawlish Warren. It is uncertain, in either case, if the local sediment budget is self-contained; on balance, this is more probable for the Teign than the Exe, but there are several unresolved research problems.

These sandbank complexes seawards of the mouths of the Exe and Teign estuaries are the product of long-term ebb delta construction and are most appropriately classified as sediment stores. Much smaller examples exist at the exits of several other tidal inlets, e.g. the Otter, Sid and Axe. The Dart estuary, by contrast, occupies a submerged deeply-incised , entrenched valley and has no bar-built or bank features. The Skerries Bank, in southern Start Bay, is a discrete accumulation that has no contemporary links with the littoral transport system and is essentially a relict geomorphic form.

Sediment transport and sedimentation processes within the region's estuaries have not been systematically investigated. In all cases, there is input of both marine and fluvially-transported material. Given that ebb-dominated tidal regimes characterise each estuary, sediment introduced by flood tide transport is unlikely to penetrate far upstream. From the fact that all estuaries exhibit net post mid-Holocene sedimentation, following their creation by earlier sea-level rise, it may be inferred that there has been substantial long-term sediment input ie they are sediment sinks.

Coast protection structures inhibit new inputs of material and control rates of longshore transport. In view of naturally slow rates of coastline recession and weak net longshore drift for extensive lengths of this coastline, their cumulative effects, over more than a century in some places, have been comparatively small. This, however, depends on the continuity of defences, which varies between zero and nearly 60% for the frontages of specific units. Defences have a more important role at critical locations of long-term erosion, particularly along the seaward face of Dawlish Warren, the Exmouth frontage and in front of the Dawlish to Teignmouth coastal railway line.

5. Nearshore and Offshore Bathymetry, Sediments and Sediment Transport - References Map

Most of the offshore sea bed of Lyme Bay, including the embayments of Babbacombe and Tor Bays, has a gentle seawards gradient with little relief. The offshore slope steepens considerably between Berry Head and Scabbacombe Head and around Start Point. Within Start Bay, there is a rapid fall to a mean depth of 10m beyond the maximum low water boundary but the northern and central areas then shelve at a low angle. In the southern part of the Start Bay, the south-west to north-east orientated 5km long Skerries Bank produces the only regionally significant submarine relief feature. Its seaward boundary is in parts nearly coincident with a marked break of slope cut into bedrock. This occurs at a water depth of approximately 20m (Donovan and Stride, 1975; Hails, 1975). There are other submerged cliffs and reef-like forms both northwards and further offshore in Start Bay.

Within Lyme Bay, the 10m isobath is approximately 900-1000m distant from, and parallel to, the coastline, whilst the 20m isobath varies between 2 and 3 km from the shore. South of the mouth of the Exe, the 20m submarine contour defines the outer limits of Tor and Start Bays but is much closer inshore between Sharkham Point and the mouth of the Dart Estuary, and off Start Point. At this last location, the 50m isobath is within 3km of the cliff line.

Sea bed sediments in Lyme Bay are predominantly gravelly muddy sands or muddy sands, although a more complex pattern of mixed sediments exists immediately south of the coastline between Beer Head and Lyme Regis. The sediment veneer rarely exceeds 0.5m in depth, and in places (e.g. directly offshore the Axmouth-Lyme Regis cliff complex) there are exposures of bedrock (Posford Duvivier, 1998). It is not known if these are the result of non-deposition or removal by sea bed currents subsequent to their accumulation. Over a restricted area of banks and channels west and south west of the exit of the Exe estuary, sediment thicknesses are between 1 and 10m, made up largely of slightly gravelly muddy sands (Posford Duvivier, 1998, 1999; Posford Duvivier and British Geological Survey, 1998). Buried channel infills of up to 38m in thickness have been reported, e.g. offshore the Exe estuary and in outer Tor Bay (Cullingford, 1982). In Start Bay, the dominant sediment type is sand, with some discrete areas of muddy sands and gravelly sands (Hails, 1975). Clarke (1969) has provided a detailed description of offshore sediments between the mouths of the Dart and Exe estuaries based on an extensive programme of surface sampling and shallow coring. Most cores exhibited a thin superficial sandy or muddy sand layer overlying either gravels or sandy gravels with a significant (>15%) percentage of biogenic debris. Both of the latter types of deposits are rarely exposed at the sea bed, and probably pre-date the Holocene sea-level transgression. The presence of fragmented shelly material is a probable indication of reworking in shallow water conditions, although contemporary addition of shell debris has been inferred by Merefield (1982). Glauconite and detrital quartz grains in sub-surface sediments were presumed by Clarke (1969) to derive from the erosion of Permo-Triassic bedrock.

It has been suggested (Joint Nature Conservation Committee, 1996) that large areas of sea bed sediments in Lyme Bay and seawards of the 20m isobath south of Hope's Nose are 'lag' (winnowed) deposits. This is easiest to accept where there are local accumulations of gravels and sandy gravels containing flint, sandstone, chalk and limestone clasts. However, recent and modern abrasion of the clay bedrock beneath shallow water in Lyme Bay could create fine material that occupies the interstices between coarse clastic particles (Pingree, et al, 1983). It is therefore likely that gravel and coarse sandy deposits are relict, (probably pre-Holocene in age) and immobile under the present day seabed hydrodynamic regime. This is apparent from the encrustation of large clasts, including those of boulder size, by bryozoans, barnacles and serpulids. Over large seabed areas beneath water depths exceeding 12-15m, there is no convincing evidence that contemporary deposition is occurring (Clarke, 1969). High turbidity values in the shallow water of inner Lyme Bay, reported by Pingree et al (1983), indicates the potential for some deposition of fine grained sediments where depths are less than 10m. In the latter case, abrasion by shoaling waves may be inferred. In deeper water, if sediment transport is taking place it is likely to involve the movement of fine sand by residual tidal currents. The presence of rippled sand, sand waves and sand ribbons has been reported (Joint Nature Conservation Committee, 1996), but there has been no systematic mapping of these features. Isolated shell beds (e.g. south of Lyme Regis and north-east of Hope's Nose), as well as the incorporation of shelly debris in superficial sediments, also infer some current transport at the sea bed. This is likely to take place approximately parallel to the modern shoreline in a net eastward direction nearshore and a net westward direction offshore. Maximum tidal current velocities are not thought to exceed 0.4 to 0.5m.s^-1 (Posford Duvivier, 1998), but notable exceptions occur - where tidal gyres (vortices) are created by headlands and at estuary inlets. Pingree and Maddock (1979) describe a numerically modelled vorticity effect at Start Point that might generate residual offshore moving currents with speeds of up to 2.2ms^-1. This would promote offshore transport of suspended sediment towards the Skerries Bank; however, it is unlikely to be a contributory cause of bank formation, but rather a partial product of its existence. Clarke (1969) has identified so-called "closed" basins of either non-deposition, or erosion, of sea-bed sediments offshore Hope's Nose and Berry Head. These may be the outcome of past, and possibly present day, headland tidal gyres, though there is no proof of current transport. The anticlockwise circulation patterns set up by tidal vortices in the outer Exe and Teigh estuaries are associated with tidal currents with maximum velocities of 3.0 ms^-1.

Pingree, et al (1983) describe a system of tidal gyres east and west of Portland Bill. The latter has a strong offshore (i.e. clockwise) component, thus it is possible that a compensating current flow may exist within central Lyme Bay. However, this is likely to be weak and therefore incapable of generating any significant net off to onshore sediment transfer. Some studies have suggested that the offshore sediment reservoir might act as a limited feed to the nearshore littoral sediment transport system. Merefield (1982; 1984) proposes large-scale clockwise transport of sand, and shell debris, between Otterton Ledge and Hope's Nose. This is regarded as a possible input to Dawlish Warren and the sandbank complexes at the Teign and Exe estuary mouths. However, a proportion of this material is moved into water depths greater than 50m and is lost to the sedimentary sink of the western English Channel. These conclusions are based on analysis of the carbonate content of offshore, beach and estuarine sediments. Given the spatial and temporal restrictions of sampling, they must be regarded a speculative. However, an offshore to onshore input of fine sand to the outer Teign estuary is considered a possibility by COAST-3D researchers (HR Wallingford, 2001). Posford Duvivier (1998) have proposed a much more spatially restricted, essentially nearshore, clockwise transport of sand between the Budleigh Salterton shoreline and the outer Exe estuary.

There is a small input of material to the offshore sediment store from the dumping of sediment removed from both capital and maintenance dredging of navigation channels in the lower Exe and Teign estuaries. The quantity varies from year to year, averaging some 300,000 tonnes a^-1 in the early to mid 1990s. It is probable that an increasing proportion will, in future, be used for beach and estuary recharge (Joint Nature Conservation Committee, 1993). There are several licensed dumping sites between Torquay and Lyme Regis, some of which are used mostly for sewage sludge. All but one are several kilometres from the coastline and in water depths exceeding 25m. The exception is the Ness, at Teignmouth, but it has not been used in recent years. Given the quantities involved, this input is considered a negligible one, with no impact on either offshore or nearshore sediment budgets.

6. References - Map

Bowen D Q et al (1986) Amino Acid Geochronology of Raised Beaches in South West Britain, Quaternary Science Reviews, 4, 279-318.

Brunsden D et al (1964) Denudation Chronology of Parts of South-Western England, Field Studies, 2(1), 115-132.

Clarke R H (1969, published, 1970) Quaternary Sediments off south-east Devon, Journal of the Geological Society (London), 125, 277-318.

Cullingford R A (1982) The Quaternary, in E M Durrance and D J C Laming (Eds) The Geology of Devon, University of Exeter Press, 273-280.

Donovan D T and Stride A H (1975) Three Drowned Coastlines of Probable Late Tertiary Age Around Devon and Cornwall, Marine Geology, 19(3), m³5-M40.

Durrance E M (1969) The Buried Channels of the Exe, Geological Magazine, 106(2), 174-189

Durrance E M (1974) Gradients of Buried Channels in Devon, Proc Ussher Soc, 3(1), 111-119.

Everard C E et al (1964) Raised Beaches and Marine Geomorphology, in K F G Hoskins and G J Shrimpton (Eds) Present Views of Some Aspects of the Geology of Cornwall and Devon, Penzance: Royal Geological Society of Cornwall, 283-310.

H R Wallingford (1992) Lyme Bay, Dorset, Wave Recording. Interim Report covering the Period 24 July 1992 to 24 October 1992. Report EX2684. 3 pp and Tables, Appendices.

H R Wallingford (2001) COAST 3-D; Final Volume of Summary Papers, Report TR121.

Hails J R (1975) Some Aspects of the Quaternary History of Start Bay, Devon, Field Studies, 4(2), 207-222.

Heyworth A and Kidson C (1982) Sea-Level Changes in South West England and Wales, Proc Geologists' Association, 93, 91-111.

Hydraulics Research Ltd (1989) A Macro Review of the Coastline of England and Wales, Volume 6: The South West Coast, Report SR 192.

Job D A (1993) The Start Bay Barrier Beach System, in: T P Burt (Ed) A Field Guide to the Geomorphology of the Slapton Region, Field Studies Council (Slapton Ley Field Studies Centre), FSC Occasional Publication No 27, 14-19.

Joint Nature Conservation Committee (1993) An Inventory of UK Estuaries, Vol 2: South-West Britain, Peterborough: JNCC.

Joint Nature Conservation Committee (1996) Coasts and Seas of the United Kingdom, Region 10, South-West England, Peterborough; JNCC.

Keen D H (1995) Raised Beaches and Sea-Levels in the English Channel in the Middle and Late Pleistocene: Problems of Interpretation and Implications for the Isolation of the British Isles, in: R C Preece (Ed) Island Britain: A Quaternary Perspective, London: Geological Society, Special Publication No 96, 63-74.

Keen D H (1998) The Quaternary History of the Dorset, South Devon and Cornish Coasts, in: S Campbell et al (Eds) Quaternary of South-West England, London: Geological Conservation Review, Publication No 14, 157-169.

Kelland N C (1975) Submarine Geology of Start Bay determined by Continuous Seismic Profiling and Core Sampling, Journal of the Geological Society (London), 131(1), 7-17.

Kidson C (1977) The Coast of South-West England, in C Kidson and M Tooley (Eds) The Quaternary History of the Irish Sea, Geological Journal, Special Issue No 7, 257-298.

Merefield J R (1982) Modern Carbonate Marine-Sands in Estuaries of Southwest England, Geological Magazine, 119(6), 567-580.

Merefield J R (1984) Modern Cool-Water Beach Sands of Southwest England, Journal of Sedimentary Petrology, 54(2), 413-424.

Morey C R (1976) The Natural History of Slapton Ley Nature Reserve, IX: The Morphology and History of the Lake Basins, Field Studies, 4(3), 353-368.

Morey C R (1983) The Evolution of a Barrier-Lagoon System - A Case Study from Start Bay, Proc Ussher Society, 5, 454-459.

Mottershead D N, Gilbertson D D and Keen D H (1987) The Raised Beaches and Shore Platforms of Tor Bay: A Re-Evaluation, Proc. Geologists' Association, 98(3), 241-257.

Orme A R (1960) The Raised Beaches and Strandlines of South Devon, Field Studies, 1(2), 109-130.

Orme A R (1962) Abandoned and Composite Seacliffs in Britain and Ireland, Irish Geography, 4(4), 279-291.

Perkins J W (1971) Geology Explained in South and East Devon, Newton Abbot; David and Charles, 192 pp.

Pingree R D and Maddock L (1979) The Tidal Physics of Headland Flows and Offshore Tidal Bank Formation, Marine Geology, 32, 269-289.

Pingree R D, Mardell G T and Maddock L (1983) A Marginal Front in Lyme Bay, Journal of the Marine Biological Association of the UK, 63, 9-15.

Posford Duvivier (1998) Lyme Bay and South Devon Shoreline Management Plan, 3 Volumes. Report to Lyme Bay and South Devon Coastal Group.

Posford Duvivier and British Geological Survey (1998) SCOPAC Sediment Inputs Research Project. Phase 3: Erosion of Coastal Platforms and Long-Term Sedimentary Deposits, Report to SCOPAC.

Posford Duvivier (1999) SCOPAC Research Project. Sediment Inputs to the Coastal System; Summary Document, Report to SCOPAC, 54 pp and 11 Appendices.

Proctor C J and Smart P L (1991) A Dated Cave Sediment Record of Pleistocene Transgressions on Berry Head, Southwest England, Journal of Quaternary Science, 6, 233-244.

Sims P (1998) Coastline Erosion, Protection and Management in Devon and Cornwall, in: M Blacksell et al (Eds) Environmental Management and Change in Plymouth and the South-West, University of Plymouth Press, 73-92.