NEOLITHIC AND LATER PREHISTORIC LANDSCAPE OF THE AVON VALLEY 233 
growth although taxa represented are likely to be 
those from the less acid end of the range (eg. S. 
plumulosum) and flushed habitats. It can be noted 
that such a community of meadowsweet and sedges 
is present today in localised areas of the River Avon 
floodplain. There is also a corresponding increase 
in dry land herbs and a marked increase in taxa 
indicative of arable cultivation (cereal type, 
Polygonaceae etc.) and ruderals (especially Plantago 
spp.). Pteridium aquilinum (bracken) reaches its 
highest values in this pollen profile and as such, the 
whole zone is indicative of an open landscape 
showing the effects of intense and widespread land 
use (arable, pasture and possibly wet meadow 
pasture on the floodplain) in the local area. It seems 
likely that this zone is medieval, reflecting a period 
of intense land use. 
RELATIONSHIP OF THE 
ARCHAEOLOGY TO THE 
SEDIMENTS 
The floodplain stratigraphy and biostratigraphical 
record will reflect important human activity in the 
area and especially that of the nearby monuments of 
Durrington Walls (Wainwright 1971) and 
Woodhenge (Evans and Wainwright 1979). Human 
activity can have a profound impact on the fluvial 
hydrology of river catchments (Evans 1992) and in 
many cases is responsible for varying degrees of 
alluviation. The sediment architecture of river 
valleys show that increased erosion and 
sedimentation occur in response to prehistoric 
human activities such as increased agricultural 
pressure (Burrin and Scaife 1988). In contrast, 
during periods of interfluve soil stability caused by 
woodland growth, sediment input to river systems 
is absent or markedly diminished. Thus major 
periods with littke human intervention in the 
landscape, such as the Mesolithic, may not be 
represented in the sequence. Consequently, 
lithostratigraphical units, or parcels of sediments 
(sensu Needham 1991) may be separated by hiatuses 
spanning considerable time. On this part of the 
“Avon floodplain the lower energy levels of 
overbank deposition is confirmed by the semi- 
organic character of the sediments, indicating im 
situ deposition of organic matter on a herb rich 
floodplain or wet meadow. It is clear that a number 
of natural and anthropogenic factors are 
responsible for the variations noted. 
It is unlikely that there is a direct causal 
relationship between the Durrington: 1 deposits 
and the impact of the essentially hunting and 
gathering communities of the Upper Palaeolithic 
and very early Mesolithic periods. It is, however, 
clear from other studies throughout Britain and 
Europe that the initiation of organic and inorganic 
sedimentation can occur at this time. Evidence for 
similar late Devensian/early Flandrian conditions 
has been provided by Evans (1971) at Durrington 
Walls, and been noted at a number of southern 
English sites (cf. Scaife 1980; 1982; 1987; Scaife and 
Burrin 1992). Thus, we can see the pollen 
fluctuations of Durrington: | as reflecting natural 
environmental changes. The cessation of 
sedimentation at the top of Durrington: 1 can be 
viewed as a response to either increasing soil 
stability caused by the dominance of deciduous 
woodland on the interfluves, or through the drier, 
continental-type climate of the Boreal period (c. 
8500-6000 BC). It is, however, clear that during the 
hiatus between Durrington: 1 and Durrington: 2, 
essentially representing the Mesolithic period, that 
there was major environmental change. 
The herb dominated Durrington: 2 suggests a 
Neolithic or post-Neolithic date since it is generally 
accepted that cereal cultivation arrived with the 
Neolithic at c. 4000 cal BC. Unfortunately we have 
no absolute dates for this zone, and as such we must 
compare and contrast this data set with studies of 
pollen and mollusca at Durrington Walls (Evans 
1971) and Woodhenge (Evans and Jones 1979) that 
provide data on the local Neolithic. 
The molluscan data from the pre-henge 
environment at Durrington Walls shows a phase of 
prehistoric woodland clearance and cultivation of 
possible middie Neolithic date (Evans 1971, 335). 
Pollen data (Dimbleby in Evans 1971, 334) is 
superficially discordant with the molluscan 
evidence in showing open vegetation dominated 
not by grasses but by ferns and bracken attributed 
to non-contemporaneous pollen and _ spore 
assemblages. In view of the extremely low pollen 
sum analysed, and the poor pollen-preserving 
conditions of chalk soil, it is perhaps more relevant 
to consider the molluscan evidence rather than the 
impoverished pollen data from Durrington Walls. 
The pollen record reveals some evidence of hazel 
woodland prior to clearance and anthropogenic 
activity. The open landscape postulated by Evans 
compares favourably with the evidence of 
Durrington: 2 noted above. This complete open- 
ness of the landscape during the Neolithic can now 
be regarded as unusual for this period (Allen 1997) 
