Separating the signal from the noise: multi-site reconstructions of Holocene vegetation change (the Tilia and Ulmus declines) in lowland Britain

An extensive review has been undertaken for a major Holocene pollen stratigraphic event, the Tilia decline, from lowland Britain. The data available has been critically re-examined with a view to establishing the associated palynological, spatial and temporal trends. The decline is generally agreed to be a non-synchronous event caused by anthropogenic activity. In this re-examination only sites where declines (individual sites may exhibit several declines) have been directly dated or where an age can be interpolated have been included. Of the 164 declines examined 44% can be attributed to processes relating to the depositional environment (e.g. breaks in sedimentation, paludification, marine inundation). Analysis of the remaining declines suggests a number of important spatial / temporal trends in anthropogenic activity (e.g. sites where the soils are calcareous or loamy, as against sandy, are cleared earlier, with clearance activity declining during a period of climatic deterioration), which remain after problematic sites are excluded. The well-known Ulmus decline is also discussed, demonstrating that current assumptions over the processes (e.g. the role of disease, climate change, human activity or interaction of several of these) cannot be effectively tested. It is shown that differences in the nature of the sites, methodologies adopted, etc., obscure the signal. We argue that these methodological and interpretive problems pose the most important constraints when using multi-site comparisons and syntheses for studies of Holocene vegetation change in lowland England.