Late holocene palaeoecology and human impact on the environment of northern Britain

Pollen diagrams from nine mire sites in northern Britain have been constructed to assess the record of human impact on vegetation over the last 3,000 years. Of particular interest is the effect of the Roman invasion and occupation of northern Britain on vegetation, especially that related to the construction of the Hadrianic: and Antonine Walls, forts and roads. Pollen analysis was undertaken to investigate whether the impact was widespread across the frontier zone or was related to the proximity to Roman Walls and forts. During the Roman occupation of northern Britain between A. D. 71 and A. D. 411, large amounts of timber and turf were needed for the construction of forts and other structures and it is estimated that a fort 4 acres in size would have needed 22,000 cubic feet of timber for its construction (Keppie, 1986). Woodland was also cleared to create a military clear zone, for the production of lime, for cooking and to create open land for agriculture. The effect of these activities on the environment of the frontier zone has not been intensively studied and few pollen diagrams concerned with the anthropogenic impact on the vegetation have previously been produced from the area (Barber, 1981; Turner, 1979; Davies and Turner, 1979). Therefore, a regional reconstruction of vegetation history has been undertaken and was long overdue. The results of high-resolution pollen analysis of the nine sites, supported by 25 radiocarbon dates, have demonstrated that the first major and permanent clearance of vegetation at certain sites occurred during the Iron Age, and this is followed by a second clearance relating to the Roman occupation. At Fozy Moss, Northumbria, no Iron Age clearance occurs and the first major clearance occurs at the time of the Roman occupation. The dramatic response of the grass pollen curves and the relatively low level of arable indicators, is in accord with the archaeological evidence for the Roman impact being one of woodland clearance for military purposes rather than for settled agriculture. Results indicate that the degree of clearance is related to distance from Roman structures. Sites close to Hadrian's and the Antonine Wall reqord the highest levels of grass pollen - 80% at Fozy Moss; 55% at Letham Moss: 40% at Walton Moss and Fannyside Muir. Dogden Moss, situated in the inter-wall area, also displays high grass levels - 50% - attributable to the moss being situated in an area amenable for agriculture. For sites at greater distances from the major archaeological structures, Gramineae percentages are lower. At Carsegowan Moss, Wigtonshire, Glasson Moss, Cumbria and Cranley Moss, Lanarkshire, the Gramineae curves barely reach 30%. This, indicates that the Roman occupation was not as intense as at the other sites. Ellergower Moss, Dumfriesshire, is in an area outside the influence of the Roman occupation and here Gramineac levels do not begin to rise until much later. Thus the hypothesis - that the proximity of an area to the Roman structures will be reflected in a greater degree of forest clearance - is upheld by the above results, but this appears not to have been the first major human impact on the landscape of the region. Ibc record of vegetation change is described up to the present day and evidence for prehistoric agriculture discussed. The concepts of pollen source areas, inter-site variability, agricultural indicators and agricultural indicies are explored in relation to the palynological results.