Floodplain environmental change since the last glacial maximum in the Lower Kennet Valley, South-Central England
Collins, Philip E. F.
PublisherUniversity of Reading
MetadataShow full item record
The lower Kennet Valley terraces reflect fluctuations in river regime through the Quaternary. Sediments and valley floor geomorphology provide a detailed record of Devensian and Flandrian fluvial behaviour and environmental change. Two major surfaces occur, the floodplain and the Beenham Grange Terrace (2 - 3 m above the floodplain). Levelling showed that the terrace south of the river, underlain by deposits of two Early Devensian interstadials, was part of the Beenham Grange level, and not the Thatcham Terrace as previously proposed. The fluvial succession consisted of the subflood plain Midgham Peat Formation overlying the Woolhalllpton Gravel Formation The latter also rises to form the surface of the Beenham Grange Terrace adjacent to the floodplain. A subunit of the Woolhampton Formation, the Heales Lock Gravel Member, thickened over a large depression in the local Tertiary bedrock and was associated with syndepositional subsidence which aided excellent subfossil preservation in the sediments. A major channel fill within the member - the Wasing Sand Bed - contained Betula with a distinctive herb flora which, with 14C data, indicates a correlation with the Late Devensian Windermere Interstadial. Lithostratigraphy, biostratigraphy and 14C data from gravels resting unconformably on the Wasing Sand indicated a Loch Lomond Stadial age. The sedimentology of the Midgham Peat Formation reflected early Flandrian reductions in flow competence associated with silt, peat and tufa deposition. Regional woodland expansion was followed by alder can' closing of the floodplain during the Atlantic. A subsequent hiatus ended ,with Sub-boreal (?Bronze Age) flooding and deposition, perhaps anthropogenically induced. An upper silt reflects lower energy flooding in an agricultural landscape. To avoid taphonomic difficulties macrofossil interpretations utilised only well-preserved material. Pollen taphonomy was more problematic. A pilot study in subarctic Canada suggested that, while fluvial pollen assemblages were not exactly equivalent to regional pollen rain through local inputs and sorting, vegetation changes were identifiable.