Link to Projects page

Link to Concepts page

Link to Reviews page

Link to Publications Page























Manning, A. D., 1995, The Effect Of Roman Occupation On Vegetation In The North British Frontier Zone: A Palynological Study Of Ditch Deposits At Vindolanda Roman Fort, Northumberland, unpublished dissertation.

Manning, A., Birley, R., Tipping, R., 1997, Roman impact on the environment at Hadrian's Wall: precisely dated pollen analysis from Vindolanda, northern England, The Holocene 7, 175-186 (LARGE PDF FILE - MAY TAKE SOME TIME TO DOWNLOAD).

Manning, A. D., 1996, The effect of Roman occupation on vegetation in the north British frontier zone: a palynological study of ditch deposits at Vindolanda Roman Fort, Northumberland, The Geographical Journal 162, part 3, 352.

In the study of Roman frontiers in northern Britain one of the most contentious areas of debate has been the effect of the Roman invasion on the native agricultural economy of the region. Opinions differ as to who was responsible for, the motive behind and the timing of the large-scale clearance of woodland observed in many pollen diagrams. Some workers believe that the Romans invaded an essentially wooded landscape and cleared it primarily for military purposes. Others believe that by the time of the Roman arrival the landscape had already been cleared and was supporting mixed native agriculture. One of the most widely used techniques for the investigation of this debate has been radiocarbon dated mire-based pollen diagrams. The use of these “off site” diagrams is problematic because they lack temporal precision and true association with relevant archaeology.

In this study two “on site” pollen diagrams were prepared from ditch deposits at Vindolanda Roman fort, Northumberland. Both diagrams were dated by archaeological context and this chronological precision has provided insights into the timing of human impacts on the vegetation around Hadrian's Wall that are unobtainable from more conventional radiocarbon-dated stratigraphies. The first diagram was from pre-Hadrianic 1st century AD ditch deposits and pertains to the period just after the arrival of the Romans on the site. The second diagram was from a late 2nd century AD ditch after approximately 100 years of occupation.

The 1st century AD results suggest a pastoral landscape and indicate that deforestation was more likely to have been by native farmers than by Roman troops. The motive for this was probably to provide pasture for cattle as suggested by the pastoral pollen assemblage. This picture is in accordance with growing evidence from across the northern region that points to an agricultural intensification rooted in the 1st millennium BC. The results found at Vindolanda contradict the tendency in the past to overemphasise the role of the Romans in woodland clearance and landscape organisation. This overemphasis is probably in part due to the abundance of ‘obvious’ Roman, and the apparent paucity of, pre-Roman archaeology, in the frontier zone. A more balanced approach sees the Roman invasion and occupation within the wider context of agricultural expansion that began in the pre-Roman Iron Age.

On arrival the Romans appear to have continued a generally pastoral land use, with cereals probably being imported. In the early period of occupation, when the frontier was being established and consolidated, short term logistics seem to have been in play. However, as occupation continued and became more permanent, there appears to be a trend towards self reliance. The 2nd century AD results suggest a far more intensively farmed landscape, including evidence of cereal growth. It is postulated that this is related to forts on the frontier becoming more self-sufficient. This need for self reliance, driven by high transport costs, was probably facilitated by a combination of population increase on the site and the taxation of the natives’ cattle which, in turn, freed up more land around the fort for cereal production.

Secure archaeological association and dating of these “on site” pollen diagrams has enabled specific conclusions within narrow temporal and spatial limits. Claims that “off site” regional radiocarbon dated pollen diagrams can attain such specificity are of dubious validity. “On site” pollen analysis, although not without problems, provides the most promising method of associating palynological and archaeological data. The potential of “on site” pollen analysis in future investigations in the frontier zone is great. There is scope for further analysis not only at Vindolanda as excavation progresses, but also at other forts in the frontier zone.

This work was the basis of my Honours project at the University of Edinburgh.

Project supervisors

Dr Richard Tipping (University of Stirling)

Prof Andrew Dugmore (University of Edinburgh)


Return to Projects page