Radiocarbon ages were calibrated using the IntCal09 calibration curve (Reimer

et al., 2009) and probabilities were summed using OxCal version 4.1 (Bronk Ramsey, 2009). To remove the effects of the variation in the gradient of the calibration curve and in alluvial unit preservation, the probability distribution for anthropogenic alluvium dates was divided by the probability distribution for all 844 dates within the radiocarbon database to give a relative probability distribution, following Hoffman et al. (2008) and Macklin et al. (2010). The resulting probability curves were then normalized by dividing each date by the highest probability in the data set. Relative probability trans-isomer distributions have been plotted with the frequencies of dates in 100-year intervals, calculated using the mid-point of the 2σ calibrated age range. Fig. 1 shows the location of sites in the UK where Holocene fluvial units have been 14C dated. AA has been identified at 93 out of 256 (36%) of these sites. This is not to say that alluviation at 163 locations

has not also been affected by anthropogenic activity, but using our strict criteria this is not registered using the information reported in publications. 130 out of 844 dated UK fluvial units (15%) can be classified as AA. Anthropogenic alluvium is recorded only at one site in the Scottish Highlands and is probably under-represented in eastern England and the English Channel catchments, as well as in tidally influenced river reaches because of the lack of 14C-dated Holocene fluvial units. Only two 14C-dated selleck AA units are classified as colluvial and debris flow deposits. The oldest AA unit is dated to c. 4400 cal. BP (Early Bronze Age) and there is an apparent 1500 year lag between the adoption of agriculture in the UK, as recorded by direct 14C dating of cereal grains (Stevens and Fuller, 2012), and its impact on floodplain sedimentation (Fig. 2). There

is, however, no correspondence between accelerated lake sedimentation – attributed to anthropogenic activity (Edwards and Whittington, 2001) – and AA, except at c.1000 cal. BP. Furthermore, selleck products episodes (c. 6000, 5000 and 3000 cal. BP) where lake deposition rates increase between the beginning of the Neolithic and the end of the Bronze Age, do not correspond with periods of notable cereal cultivation as identified by Stevens and Fuller (2012). Indeed, they coincide with troughs in the independently summed probability distribution of cultivated plant food and suggest that the primary cause of accelerated sedimentation was not related to arable farming. Alternatively, climate change and/or over-grazing in these mostly small catchments in northern and western Britain and Ireland could have been contributing factors.