232, THE WILTSHIRE ARCHAEOLOGICAL AND NATURAL HISTORY MAGAZINE 
It is suggested that the sediments of this zone 
started to accumulate after major forest clearance in 
the area (see below). The presence of arable pollen 
and weeds indicates a date post c. 4000 cal BC, i.e. 
Neolithic. Pollen data spanning this period in 
southern England are sparse and, where present, 
usually relate to areas away from the chalk where 
pollen preservation in more acid environments is 
better suited to more detailed palynological 
investigation. Nevertheless on the Chalk in Sussex 
(Thorley 1981), Dorset (Haskins 1978; Waton 1980; 
1982), Hampshire (Waton 1980; 1982) and the Isle 
of Wight (Scaife 1980; 1987) partial woodland 
clearance took place during the Neolithic. More 
extensive clearance took place subsequently in the 
Bronze Age with large areas of woodland cleared on 
the lighter soils of the Chalk and Greensand. This 
presents two possible interpretations for the 
Durrington pollen spectra in Durrington: 2. First, 
the relative absence of trees locally here, suggests a 
Bronze Age (or later) date. Second, the absence of 
pollen data from on or near the chalklands here 
presents the possibility that a large area was cleared 
of woodland at sometime during the Neolithic. 
Given the archaeological, pollen and molluscan 
evidence from Durrington Walls (Evans 1971) and 
mollusc evidence from Woodhenge (Evans and 
Jones 1979), it seems likely that zone Durrington: 2 
represents continuous and _ possibly intense 
Neolithic land use in this region. Furthermore, this 
poses the interesting possibility that such clearance 
and land use was responsible for the re-initiation of 
sedimentation in the Avon valley after a hiatus of 
possibly thousands of years. The removal of trees 
on the interfluves makes soil available for erosion 
and will have raised groundwater tables and 
increased surface sediment run-off (colluviation) 
into the valley bottom. This cause and effect has 
now been widely demonstrated from a number of 
British alluvial sites (Burrin and Scaife 1984; Scaife 
and Burrin 1983; 1985; 1992) and dry chalk valleys 
(Bell 1981; 1982; 1983; Allen 1988; 1992). 
Durrington: 3 (1.17-0.90m) 
A sherd of undiagnostic Roman pottery was 
recovered at 1.15m at the base of Durrington:3, 
clearly dates this zone to the Roman or post-Roman 
period. Since the pottery was a small, broken 
fragment, it could even have been incorporated at a 
later date. If the interpretation of zone Durrington: 
2 as Neolithic or Bronze Age is correct, there 
appears to be a substantial hiatus between these 
zones (perhaps spanning the later Bronze Age and 
Iron Age). In zone Durrington: 3, tree and shrub 
pollen become dominant in these fen carr peats. 
High values of Alnus and Corylus in more organic 
sediment and peat represent the growth of local 
alder dominated carr woodland on the floodplain 
here and is attested by the of ‘clusters’ of pollen 
found. This appears to represent a phase of stability 
in the catchment with lower water tables allowing 
the growth of a drier (fen carr) woodland and an 
absence of constant flooding. This phase is 
mirrored by a reduction in wetland herbs (largely 
Cyperaceae). 
There is some evidence of other woodland 
growth with Ulmus and a single record of Fraxinus 
(ash). Tilia continues to be represented with a 
mixture of degraded and non-degraded pollen 
grains. This indicates that some lime woodland 
remained on the drier areas of the river catchment. 
Since Tilia produces relatively small numbers of 
pollen and is insect pollinated, it is likely that it is 
under-represented in the pollen spectra (Anderson 
1973; Tauber 1965). 
The growth of more closed carr woodland on 
the Avon floodplain probably had a significant 
effect in reducing pollen input from the 
surrounding region on to the mire. Although 
reduced in numbers, many taxa remain and 
Plantago lanceolata and other typical anthropogenic 
pollen types are present. Cereal pollen and 
associated taxa present in the previous zone are 
largely absent. This may be interpreted as a real 
decrease in arable cultivation or, more likely, that 
the generally poorly dispersed pollen taxa have 
been ‘filtered out’ by the now substantial woodland 
growing on the river floodplain, but a presence of 
pasture is substantiated 
The substantial hiatus between zone 
Durrington: 2 and Durrington: 3 is not uncommon 
in alluvial sediments (e.g. Burrin and Scaife 1984; 
Scaife and Burrin 1992). The top of Durrington: 2 
may have been a land surface with perhaps 
seasonally waterlogged pasture on which the 
Roman pottery (at 1.15m) and charcoal (at 1.12m) 
was deposited. There is no visible effect of a fire on 
the vegetation/pollen spectra, nor was any evidence 
of pedogenesis noted in the sediments. Pedogenesis 
would be arrested if the floodplain remained wet or 
waterlogged. 
Durrington: 4 (0.90-0.68m) 
There was a return to an open floodplain 
environment with the demise of alder carr. Small 
numbers of Sphagnum spores may indicate localised 
