Soil erosion on buried archaeological sites in arable areas: a modelling approach.

Abstract

The Monuments at risk Survey (MARS) of England (Darvill and Fulton, 1998) concluded that the dominant agent of damage to archaeological sites is intensive agriculture. No such equivalent or similar study exists for Scotland. This study aimed to assess the threat of soil erosion posed to archaeological cropmark sites across an 80 x 60 km study area by quantitatively modelling soil erosion rates. Archaeological sites are widely distributed across lowland mid-Scotland and are clustered on arable land. Focus was placed on cropmark features since very little is known about them and damage rate is difficult to ascertain without excavation. 2849 registered (NMRS) archaeological sites are present in the study area, 1707 of which are cropmarks. To meet the aim, the total erosion budget was modelled in its component parts: water and tillage translocation. Firstly, water erosion and deposition were modelled using Desmet and Govers’s (1995) simple model accounting for field boundary structure and multiple flow directions. Secondly, tillage translocation was modelled using ARCTILL. The 137Cs tracer technique was applied at four field sites containing cropmark archaeological features. Transect based sampling was applied using 25 m x 25 m cells to coincide exactly with the GIS grid system. Derived erosion/deposition rates were then used to optimise the water and tillage models at each field site, from which a general optimised net model was defined and applied at the regional scale. The effect of field boundaries on patterns and magnitudes of potential overland flow and subsequent erosion/deposition was found to be significant and worthy of further research. The archaeological features at Loanleven (NO 058 252) and Littlelour (NO 479 444) were found to be under serious threat from erosion caused by ploughing practices up to -1.34 kg m-2 yr-1 (-1.14 mm yr-1) and -2.14 kg m-2 yr-1 (-1.34 mm yr-1) respectively. Tillage erosion on average has contributed 75% and 69% at the Loanleven and Blairhall (NO 116 280) sites respectively clearly demonstrating the significance of the process. The highest erosion rates were located on strongly convex slope sections, yet statistically were related only weakly. These loci were strongly correlated with topsoil depths. For the whole study area, the general optimised net model predicted 65% of all archaeological sites (2849 in total) as being on land experiencing net erosion. Of some 1707 cropmark sites, 63% were predicted as being on land experiencing net erosion. 547 cropmark sites (32% of cropmarks) and 1053 (37% of total) of all archaeological sites present within the study area exceeded the soil loss tolerance threshold (0.13 kg m-2 yr-1). This research underlines intensive agriculture as being the main damaging agent of buried archaeology across the study area.