- General models of Iron Age agriculture suggest a move towards more intensive farming, which has the potential to create a surplus. This surplus could then be traded or exchanged, allowing for the acquisition of exotic goods or materials and the ability of certain members of society to specialise in the production of craft industries.
- The production of grain in ample quantities would, of course, have been a benefit to the Roman invaders when they began their conquest of much of mainland Britain in AD 43. There has been much discussion over whether the souterrains – several of which have been confirmed and even excavated in South West Scotland, particularly Auchrannie – functioned as storage for grain surplus which could then be either traded with the Roman military or exacted by them as tribute. The chronological relationship between Romans and souterrains is key, and the link is not entirely proven. Instead, souterrains may represent an element of yet another distinct form of monumental roundhouse which developed in the middle Iron Age, just prior to the Roman invasion.
- Palaeoenvironmental evidence is often recovered during the excavation of Iron Age settlements. Charred wheat and barley are common finds: both were recovered from the settlement site at the Curragh South, Girvan, while small quantities of barley were found at Woodend Farm, Johnstonebridge. Evidence from the waterlogged deposits encountered on crannog sites is, however, particularly rich, indicating that a wide range of resources were being utilised. As well as the arable crops, wild plants such as hazelnuts might still be eaten on a seasonal basis. Animal bone also tends to survive well on crannogs, which is particularly useful to archaeologists, as soil conditions in South West Scotland usually result in its disappearance. A wide range of species may be found on these sites, though early discoveries should be treated with caution as crannog sites were often subjected to multiple phases of occupation spanning considerable centuries. Nonetheless, evidence for sheep/goat, cattle, pig and horse, as well as wild fauna such as deer have been recovered from such sites.
- Occasionally, modern excavations of sites which reveal waterlogged features have revealed a comparable range of species. Excavations at the upland farmstead enclosure site of Long Knowe, Eskdale, revealed the bones of sheep, cattle and horse, along with cut twigs of predominantly willow which could have been used as fodder (Mercer 1981). Despite the excellent preservation of organic material, no cereal grains were recovered, suggesting an emphasis on pastoral production. Evidence for the use of wild species is rare, but a fragment of deer antler was recovered from the settlement site at the Curragh South.
- Some caution should be engaged, however, with regards to the early crannog finds, as crannogs are often very long-lived sites and refuse may have accumulated over a period which extends to the earlier medieval period and right back to perhaps even the Late Bronze Age. The structural elements of crannogs also suggest some degree of woodland management, and certainly a good understanding of the properties inherent in various tree species. Oak is used for the main beams of many crannog structures in South West Scotland, with alder stakes used to stabilise the brushwood heaps that created a sound, dry base for the ‘packwerk’ type crannogs so commonly found in the region.
- Coupled with the anecdotal evidence provided by Roman and – to a much lesser extent – Celtic sources and the actual physical evidence provided by palaeoenvironmental remains, is the nonetheless solid evidence inferred by wider environmental data. This shows evidence for a significant decline in tree cover across South West Scotland. Trees had largely been cleared from the landscape, even early in the Iron Age, giving way to a mixed agricultural regime of arable and pasture farming. The site of Uppercleuch, Annandale, for example, revealed a cobbled yard and enclosure consistent with a largely pastoral economy focused upon stock rearing. Pollen from weed species commonly found in pasture and grassland supported this theory, but charred grains of barley, oats and in particular, emmer wheat were also recovered in small quantities (Terry 1993).
- A peak in activity appears to have occurred in the later Bronze Age, with arable farming extending even into higher altitudes, on hill slopes that today we would dismiss as marginal land. Ongoing climatic deterioration coupled with peat expansion meant that arable crops were no longer viable at these upland sites, but it is difficult to reach any more than a very blunt understanding of when and how this deterioration would have impacted local communities and how their responses would have changed over time.
- In upland areas towards the eastern limits of the region, relict prehistoric landscapes can still be identified by the presence of roundhouse footings, small cairns of field-cleared stone and even cultivation marks which reveal the use of a small hand-held plough or ard. These are difficult to date, and some may even date back to the Bronze Age, but their use appears to have extended onwards into the Iron Age. One example, which was recently surveyed and partially excavated, was at Coats Hill, Moffat. Here, an area of relict landscape revealed 40 small cairns, two annular structures, and a larger, more substantial cairn to the north. No dating evidence was recovered, but the presence of two enclosed settlements – of similar form and morphology to that characteristic of eastern Dumfries – allowed a 1st to 2nd century AD to be inferred for at least some of the activity on the site. It is entirely possible that – even when climatic factors meant that arable cultivation was no longer viable – this land would still have been used as pasture, for the grazing of cattle and sheep, perhaps on a seasonal basis.
- There has been some suggestion that increased pressure on land during the Iron Age led to an increase in the formal division of land and, perhaps, an increase in territorial claims by family groups or communities. This is reflected in the appearance of land divisions, which take the form of fields bounded by low stone walls or ditches. Some of these still survive in association with well-preserved sections of relict prehistoric landscapes.
- Several examples of these Iron Age landscapes been studied. Excavations at Hayknowes Enclosures targeted linear boundaries and a rectilinear enclosure. These were in the vicinity of a double-ditched circular enclosure with a monumental gateway and internal roundhouse. These features were potentially contemporary with (and associated with) the settlement, which dated to the mid- to late pre-Roman Iron Age (Gregory 2001a).
- Another area which was subject to detailed study was centred upon the hillfort site of Castle O’er and a nearby enclosure at Over Rig. Excavations on both sites and detailed field survey of their hinterland suggested that the hillfort at Castle O’er experienced a period of settlement hiatus between the last century of the 1st millennium BC and the 3rd or 4th century AD. During this time, it appears instead to have been incorporated into a wider system of linear boundaries and enclosures that the excavators suggested may have been associated with the corralling and movement of livestock. It was also argued that the nearby large enclosure site at Over Rig may have formed an important part of this larger system (Mercer 2018, 234). Excavations at the site revealed a brief period of occupation coeval with the Roman military occupation of the area, with the enclosure perhaps providing a neutral space where cattle and/or ponies might have been gathered, either for exchange with the Romans, or offered as tribute.
- Finally, it has been suggested that the appearance of rectilinear enclosures like Rispain Camp may have been associated with the creation of such land division; the square or rectilinear enclosures were required to be this shape, as opposed to circular, in order to be accommodated within a linear field system, defined either by low walls or ditches.
- While this is certainly possible, it should be remembered that square and rectilinear enclosures appear to be a localised settlement form, while linear field boundaries are more widespread and, in some cases, have a demonstrably greater antiquity which stretches far beyond the Iron Age. Excavations near the Iron Age settlement at Warden’s Dykes, Gretna, explored a series of linear cropmarks, thought to be part of a field system associated with the settlement. The ditches which formed the land divisions were re-cut during the Roman Iron Age, that is, in the Middle to Late Iron Age, but the origins of these features were more ancient. It is possible that the fields were laid out during the early Iron Age, but a Bronze Age origin could not be ruled out, and even a Neolithic date was possible.
- For comparison, work undertaken on the field systems of Machrie North, Arran by John Barbour demonstrated that similar features had potential origins in the Early Bronze Age, and were associated with later Bronze Age roundhouse structures (see Chapter 6 Chalcolithic and Bronze Age).