Pollen analysis has provided a basic framework for the environment of the Roman Iron Age. Despite attempts by some palynologists to suggest otherwise (e.g. Dumayne and Barber 1994), major forest clearance linked to the extension of agriculture is a late Iron Age phenomenon unrelated to the arrival of Roman forces (Hanson 1996; Tipping 1997). What remains in debate is the extent of (and motives behind) that clearance, and the size and, indeed, location of the Caledonian forest which features in Roman literary accounts as early as the writings of Pliny in the mid-first century AD (Breeze 1992).
Although a lot of basic environmental work has been done in relation to individual recently-excavated fort sites (e.g. Dickson 1989; Clapham 2007), providing evidence of diet and food supply, there has been no systematic, synthetic study. The major debate concerns the extent to which primary bulk foodstuffs (cereals and, to a lesser extent, meat) were obtained locally, rather than imported. Unfortunately, animal bone evidence from Scotland is notorious for its paucity, linked to the general acidity of the soils, making it hard to assess any evidence for stock improvement or the intriguing suggestion by Stallibrass (2009) that droving may have been an element of supply patterns. Wherever possible local supply of cereals seems to have been preferred (Manning 1975), and there are very occasional hints in the archaeobotanical evidence from fort sites which lend support to such a suggestion, including the appearance of unprocessed or only partly processed cereals. On the supply side, recent work on environmental evidence from indigenous settlement sites in the Lothians indicates that emmer and spelt wheat were potentially available locally, with barley in abundance (Huntley 2000; Huntley and O'Brien 2009; Lelong and MacGregor 2007).
Though there are limited examples of field systems known outside forts (e.g. Carriden, Croy Hill, Rough Castle, Auchendavy, Inveresk, Castledykes), the character of their relationship with the fort remains speculative. Some have been sampled by excavation, and some dated to the Roman period as a result (e.g. Inveresk; Cook 2004), but there has been no systematic study. The fields at Auchendavy, if correctly identified as such from the limited sampling, lie north of the Antonine Wall (Dunwell et al. 2002, 274-279).
Aerial view centred on the cropmarks of the field system and Roman temporary camp at Castledykes, South Lanarkshire © RCAHMS
Excavation suggests that most fort sites were provided with internal wells, though clay water-pipes have been recorded at Newstead, and traces of aqueducts noted at Fendoch and Carpow. However, there has been no systematic attempt to identify external water sources. For sustenance in material terms, seen in the evidence for craft production, see section 5.6.
Environmental data from both Roman military and late pre-Roman and Roman Iron Age indigenous settlement sites needs to be synthesised.
Trial excavation is needed to confirm the relationship of the field system and temporary camp at Castledykes and further geophysical survey is required at Auchendavy to establish the extent of the postulated field system.
Systematic study of the wider landscapes of forts is required.
Appropriate levels of standard, systematic environmental sampling should be undertaken for any excavation projects, with a particular attention paid to those with waterlogging or good bone preservation.