The Isles of Scilly

THE EARLY LANDSCAPE

During glaciation, land south of the ice sheet would have been bare tundra, cold, with sparse vegetation and probably few animals (Yalden, 1999), and certainly few that are still found in Britain today. It is difficult to imagine what Scilly was like at the time of the earliest human visitors, who were probably Mesolithic hunter-gatherers who left only a few flints as evidence of their passing. We have already seen that the islands would have been a considerably larger landmass than the present-day scatter of islands. Much of the land was covered in birch woodland, sparse grassland and marshy land with sedges. These conditions of the Mesolithic period persisted across southern Britain, then part of Continental Europe, and most of the steppe species that were present then have either died out or retreated to more northerly areas. During the Neolithic period people may have started to settle in Scilly and begun clearing the land, but pollen evidence shows some forest clearance was followed by woodland regeneration and agricultural decline. There are a few artefacts from this time, but it is likely these were only temporary occupations (Ratcliffe & Johns, 2003).

EARLY MAMMALS AND OTHER FAUNA

Very few remains of the early fauna of the islands have been found, but one small rodent, the root or Pallas’s vole Microtus oeconomus (very similar to our field vole M. agrestis) was present, as was the red deer Cervus elaphus, and both were still present in Scilly in the Bronze Age.

Modern Scilly is poor in mammal species, and the written records are sparse. Bones found in the Iron Age sections of the excavations on Nornour included Scilly shrew, wood mouse or long-tailed field mouse Apodemus sylvaticus and root vole. The first two are still extant in Scilly, but the root vole is believed to have become extinct at some later period, no remains having been found after Romano-British times (Turk, 1984; Ratcliffe & Straker, 1996). Root voles are no longer found in Britain although there are isolated (relict) populations still in the Netherlands, Scandinavia and eastern Europe (Mitchell-Jones et al., 1999). In late summer 1978 my daughter and I found a vole mandible and two molars in storm debris on the boulder beach below the Porth Killier Bronze Age midden, along with scraps of bone and shards of coarse pottery. These were sent to the Natural History Museum, where the vole remains were identified as root vole (Gordon Corbet, in litt.). Later, in 1982, I had the opportunity to go to Hungary and was able to visit root vole habitat near Lake Kolon, in Kiskunsági National Park. This is an area of rough grassland and Phragmites swamp, which would seem to be typical habitat of the vole. An interesting note by Mitchell-Jones et al. (1999) is that root voles migrate from wetland to dunes or drier habitats in winter, and even into houses.

Other animal remains that have been found from archaeological sites include seals, various cetaceans, red deer, toad Bufo bufo (an amphibian no longer native in Scilly), as well as numerous fish and bird species and domestic horse, ox and sheep, all from Bronze Age sites. Roe deer Capreolus capreolus, seals, cetaceans and domestic animals have been recorded from Iron Age/Romano-British sites. At coastal sites seals, small cetaceans and fish were clearly an important part of the diet of the human inhabitants. Although not listed among the remains of fish and molluscs that have been recorded, the boulder beaches and rock pools can support several species of easily caught edible fish, for example rocklings, the larger gobies and grey mullet Chelon sp. Today the large freshwater pools on the islands also contain very large eels Anguilla anguilla, and these may also have been present in the past.

Further discussion on the early mammal fauna is included in Chapter 15, and prehistoric and historical records of birds are discussed in Chapter 16.

THE EARLY HABITATS

Of particular interest is Thomas’s (1985) description of the reconstructed palaeoenvironment of the early Scillonian landscape and the mapping of four main types of habitat. Some of the evidence for this he based on the pollen records, which unfortunately were limited to the few peat deposits and archaeological digs, and also on the distribution of some significant plants in Lousley’s Flora (1971). The four habitat types he described were stream-drained marsh, woodland, sand dune and open ground (including heath).

There are still marshlands in Scilly today, although they are nowhere near their former extent. Some of the land now under the sand flats between the islands could have been low-lying and boggy, but all that remains now are small wetlands at Higher Moors, Holy Vale and Lower Moors on St Mary’s, now much contracted in area. Even as recently as the 1960s there were wet fields from near Porthloo Pool and Rose Hill through to Lower Moors with yellow iris Iris pseudacorus, lesser water plantain Baldellia ranunculoides and hemlock water-dropwort Oenanthe crocata. Although these areas are still there they are now much drier and less species-rich. Another similar wetland area is now flooded and forms the Great Pool on Tresco. All the other streams and marshy areas are now lost under the sea, but some can still be traced from the geological record. Between Teän and St Martin’s is the deep channel of Teän Sound, which probably marks the route of a prehistoric stream.

One of the most interesting theories propounded by Thomas is his mapping of the ancient woodland cover on the islands by looking at the distribution of woodland species in Lousley’s Flora. From the pollen samples analysed by Dimbleby (1977) from Innisidgen and by Scaife (1984) from Higher and Lower Moors it would seem that Scilly was once covered in woodland. This woodland consisted of oak Quercus robur and birch Betula spp. with an understorey of hazel Corylus avellana and alder Alnus glutinosa (probably where there were wetter areas). Pollen evidence also included some ash Fraxinus excelsior and traces of yew Taxus baccata, and later hornbeam Carpinus betulus and elm Ulmus sp. Virtually nothing of this woodland is evident today, but support for the pollen evidence and what it tells us about former woodland distribution can be extrapolated from the present-day distribution of plants (known as ancient woodland indicators) that have strong ancient woodland associations, for example wood spurge Euphorbia amygdaloides and wood dock Rumex sanguineus (Kirby, 2004). In his Flora of the Isles of Scilly Lousley (1971) comments on a number of these woodland plants that were growing in non-woodland habitats. These fall in very neatly with the pattern of woodland 2000 years ago, as demonstrated by Thomas (Fig. 12).

Since 1971 additional plant records have reinforced the pattern. So it is possible to visualise the kind of woodland that may have grown on the islands at the time, possibly similar to the present-day Wistman’s Wood on Dartmoor, with stunted, twisted trees, wind-pruned into shape and only able to reach any reasonable height where they are sheltered in the hollows between the hills – as happens with the elms in Holy Vale today. The ground cover may have been open, with many of the species that still exist in Scilly. The trees and exposed rocks would also have supported luxuriant ferns and bryophytes. Other evidence of the ancient woodlands that existed on Scilly are the numbers of buried tree trunks that have been found on Tresco in the past, and the few oak Quercus sp. trees and woodland plants in the area still known as Tresco Wood. There are also records of submerged tree trunks on St Mary’s and, more reliably, St Martin’s.

FIG 12. The present distribution of AWI (ancient woodland indicator) plants may indicate where woodland existed before the submergence. (Updated since Thomas, 1985)

The work carried out between 1989 and 1993 by the Cornwall Archaeological Unit (Ratcliffe & Straker, 1996) on a number of cliff-face and intertidal deposits also provided exciting additional evidence for the deciduous Mesolithic/Neolithic forests. In addition, the CAU found further evidence that these forests were being replaced by heathland, grassland and cultivated plants by the Late Iron Age, as people began to have an impact on the land.

The other main habitats, sand dunes and heathlands, are still present. Many of the dunes have been flattened and have become vegetated with grassland and scrub, or are now cultivated fields. Heathlands and grasslands have resulted from the management of the open habitats over centuries. The land would have been utilised in many ways, from grazing for animals to the stripping of turf from the heath to use as fuel – once the inhabitants had cleared most of the woodland.

Pollen analysis

Pollen analysis of samples from Higher and Lower Moors (Scaife, 1984) shows the distribution of pollen and spores in four levels of the peat below the two mires on St Mary’s. Later work investigated more areas of peat (strictly not really peat but humic silts) at Porth Mellon, St Mary’s, and Par Beach, St Martin’s (Ratcliffe & Straker, 1996). The earliest levels are mostly tree and shrub pollens, oak, birch, hazel, some ash, elm and willow Salix sp., also some grass species, sedges, bracken Pteridium aquilinum, other ferns and some aquatic plants. These all point to a landscape with woodland, mire and open-water habitats. The record for Lower Moors has less tree pollen and may fit the theory that the ancient woodland was distributed mainly on the north and east of the island (Thomas, 1985). Pollen samples also show there was some further clearance of the secondary woodland that regenerated after the earlier clearances. This coincided with the more open landscape and evidence of arable, heathland, mire and coastal habitats associated with the Iron Age and Romano-British communities then inhabiting the islands. John Evans (1984), excavating an Iron Age field system at Bar Point, found the kind of plant remains that would be expected to follow after most of the woodland had been cleared. These charred fragments were mostly plants of heathland or acid grassland: grasses, ribwort plantain Plantago lanceolata, vetch Vicia sp., Galium, Medicago, broom Cytisus scoparius, elder Sambucus nigra, gorse Ulex sp., false oatgrass Arrhenatherum elatius, as well as oak.

EARLY AGRICULTURE

It was not until about 2000 BC that the first settlements may have started on Scilly (Ashbee, 1974; Thomas, 1985). The evidence from pollen deposits, and from sources such as middens and other archaeological deposits, shows changes in the palaeoenvironment after the clearing of the forest, resulting in more open landscapes with grass and heathland species (Ratcliffe & Straker, 1996). These clearances coincided with the early settlement of Scilly, when it is presumed the woodland was cleared for timber and firewood, as well as to open up areas for cultivation. The growing of cereals (and possibly other plants that produce edible seeds) and the use of the plough must have happened quite early after the arrival of the first settlers. The excavation of the Iron Age site at Bar Point graphically illustrates the agricultural practice at that time (Evans, 1984), with evidence of stone banks to produce individual fields – not apparently for keeping stock in, but to demarcate ownerships. Cultivation marks left by the rip-ard (a primitive plough) and mattock as well as burnt charcoal from gorse were found, and also hoof prints of domestic animals – horses, cattle, and sheep or goats. Pollen analysis indicated open land without trees or shrubs (although a small amount of oak, alder and hazel pollen was found in one location), and the presence of cereal pollen points to cereal cultivation. Ribwort plantain was ‘in consistent but low abundance’, which suggests cultivation and grazed land. The evidence for deforestation is based on the association of ribwort plantain with pasture and cultivation and its intolerance of competition with woody plants (Godwin, 1975).

Once the islands had been successfully settled one presumes a period of stability and expansion of the human population followed. This is when the great changes in the landscape, the vegetation and animal life in Scilly would have really begun. As the woodland was cleared, perhaps leaving only small areas, more and more land would have been broken up and cultivated to meet the demands of the increasing human population. Farm stock would have been grazed on the open land, cliffs, dune grasslands, coastlands or around the marshy areas. Bell heather Erica cinerea, ling Calluna vulgaris, reed Phragmites australis and bracken would have been cut for bedding and thatching. Beaches would have been scoured to search for edible seaweed, molluscs and fish. Widespread cultivation and grazing became prominent during the second millennium BC, with many evidences of farming in the pollen and charcoal plant remains, querns for grinding grain, bones of farm animals and so on

FIG 13. The remains of an Iron Age and Romano-British settlement are still visible on Halangy Down, St Mary’s. March 2006. (Rosemary Parslow)

(S. Butcher, in litt.). Also the field walls, settlements and lynchets across some fields indicate the impact on the landscape. Cultivation of the islands must have continued over many centuries, as is demonstrated near the Romano-British settlement at Halangy Down (Fig. 13), where Charles Thomas and Peter Fowler found traces of earlier fields with ones on a different alignment underneath.

There are numerous field walls that are now only seen at low water spring tides in the sea off Samson, Teän and some other islands (Fig. 14). These all point to there having been much more land available for grazing or cultivation before it was lost to the sea with sea-level rise. Their presence also suggests that not all the low-lying land was marsh; but only reasonably accessible land would have been divided into fields.

FIG 14. At low tide the lines of ancient field walls are visible in the sand flats off Samson. June 2002. (Rosemary Parslow)

Arable cultivation

Early cereal crops on Scilly included six-rowed (and possibly two-rowed) barley Hordeum vulgare in both the hulled and naked varieties, which have been identified from several archaeological sites. Other crops that have been recorded are emmer wheat Triticum dicoccum, Celtic bean Vicia faba var. minor and oats Avena sp. The origins of domestication have been identified in the area of the Near East some 9,000-10,000 years ago, and barley and other crop species must have gradually made their way to Europe with the early inhabitants and through trade (Smith, 1995).

There is evidence of arable cultivation in the Isles of Scilly at least as far back as the Bronze Age. On Samson archaeologists found a cache that consisted of a cleaned crop of naked barley, probably six-rowed barley, confirming this was grown in the Early Bronze Age. Six-rowed barley (and possibly also two-rowed barley) was clearly an important crop in Bronze and Iron Age Scilly both in the hulled and naked forms – although most of what has been found is the naked variety. Hulled barley has been found at Middle and Late Bronze Age sites. Barley has also been identified from post-medieval deposits at Steval Point, St Mary’s. Emmer wheat and Celtic bean were found with both kinds of barley at sites at Porth Cressa, St Mary’s, and Porth Killier, St Agnes. At Halangy Down, besides the older lynchets beneath those of the Romano-British settlement, a pot was found nearby with impressions of grain in the clay. One intriguing find was of a large number of seeds in the post-occupation layers of a hut at West Porth: these were dated to Late Bronze Age/Early Iron Age and are of great interest as they were of common arable weeds (Ratcliffe & Straker, 1996). Unfortunately there are very few records or remains of plants to give more than a hint of the wild flora of arable fields.

Arable weeds identified from Bronze Age sites in Scilly include vetches, knotgrass Polygonum sp., chickweed Stellaria sp., black bindweed Fallopia convulvulus, small nettle Urtica urens, corn spurrey Spergula arvensis and red goosefoot Chenopodium rubrum. A curious occurrence was ploughman’s spikenard Inula conyza, a plant more usually growing on calcareous soils. It is possible that some plant seeds (for example corn spurrey and knotgrass) were eaten in a kind of porridge. Certainly weed seeds continued to be found from archaeological sites after the Bronze Age, as did emmer and barley. Unfortunately there is not a great deal of evidence of Iron Age cultivation. Only two sites are recorded, Shipman Head, Bryher, and Halangy Porth, St Mary’s (Ratcliffe & Straker, 1996).

There is little information on arable plants from the Late Roman and Early Medieval periods, although there are unspecific barley and oat records from the seventh/eighth-century layer and wheat and barley from the tenth to thirteenth centuries. Weed seeds, most of which could be arable weeds, were also recorded. Some of the plants that are still found in Scilly are believed to have been early introductions as contaminants of seeds or goods brought in from the Continent, even from as far away as the Mediterranean.

LIMPET MIDDENS

There are shell middens in many places on the Isles of Scilly, refuse heaps that in some cases date back to the Bronze Age. They contain bones and other rubbish thrown out by the inhabitants, especially common limpet Patella vulgata shells. There seems to be a correlation between the size of the limpet shells and the living conditions of the people at that time: the smaller the shell apparently the tougher life had become for the people, as the shells were being harvested before they had reached full size. It is reported that they are very chewy and uninteresting, but presumably they were at least an easily accessible food and protein source. However, the University of Bristol Expedition in 1978 declared limpets were delicious in a risotto!

Limpet shell middens have been found at many archaeological sites on Scilly, on any islands that have been inhabited at any time. Middens were often in use up to the nineteenth century, and frequently can be dated by pottery shards and other remains layered in with the shells and other rubbish. When the inhabitants of Samson were evacuated from the island in 1839 they left behind a huge pile of limpet shells. Some of the beach pebbles used to knock limpets off the rocks were also found. There is evidence that some limpets were collected for fishing bait, but it is also very probable they were eaten even in quite recent times (Ratcliffe & Straker, 1996). Although limpets were the main species taken, remains of other edible molluscs have also been identified from the middens. These are mostly the species that would be expected: cockles, scallops and topshells, all of which are still common around the islands.