The majority of rivers in southern England occupy valleys that were cut and partially infilled during the late Quaternary. These deposits underlie and partly confine the modern floodplains and strongly influence the modern channel structure and path. During the cold phases of the Quaternary, rivers adopted a braided form that was dominated by gravels. Finer deposits within these sediments are confined to elevated or abandoned channel areas. This is exhibited clearly in a previous MAS project at Harleyford (MAS, 2005) where older gravel deposits of a braided system had partly controlled the path of the Thames at Marlow, Buckinghamshire. This forced the river into an anastomosing channel style and distributed floodplain sediments adjacent to the main and palaeochannels around the elevated gravel deposit. The resulting gravel island, intermittently surrounded by water, was utilised for a Bronze Age cemetery. Secondary channels were very common in the past fluvial history of this part of the Thames. As these channels fill up with sediment there is a greater recurrence of overbank (flooding) events. As a result, the floodplain sediments become thicker and the floodplain increases in elevation. Cookham Paddock lies within the floodplain of the River Thames, close to the modern river channel and records some of these past events.

A consideration of the evidence for changes in climate during the Holocene for the UK, or NW Europe in general, is important for archaeological projects within the Thames floodplain or terraces. Climate changes relating to temperature and/or to the amount of precipitation will have a direct influence on run-off, base-levels and river activity. However, caution must be stressed, as during this period of time (the Holocene) there are concurrent changes in vegetation as a direct result of human intervention (discussed further in the next section).

Temperature and precipitation have both a direct and an indirect influence on fluvial activity. Also important are decreases or increases in seasonal contrast (colder summers/warmer winters) as well as the amount and type of vegetation. However, the latter is also controlled by human factors, not just by climate. Finding organic materials or human artefacts within these deposits can help to date periods of fluvial activity and inactivity. In return, understanding the timing and extent of fluvial activity helps in the prediction of the location and preservation potential of archaeological artefacts.

Two types of sequences can be recognised in the fluvial record: aggradational and erosional. Climate influences fluvial activity mostly through sediment supply and water discharge. The aggradational sequences (such as floodplain development) occur with an increase in sediment in a sufficiently low energy environment to allow the sediment to build up. In contrast, higher energy flow creates incision and erosion together with a high sediment load, which is deposited down river and as the current wanes. The high energy episode promotes an incision of variable magnitude such as a major channel cut or a minor cut, such as a scour. The incision is followed by a period of infill by gravels, sands, silts and clays and then by aggradation – a series of organic-rich silts and clays. The site at Cookham Paddock lies on the modern floodplain close to the River Thames. At this location the river is well confined with a stable, cohesive bank and by river management. However, at depth beneath the uppermost accretion (flood) levels remain the evidence of lateral movement of the main channel or palaeochannels. The dating of these events is of great interest in determining these major periods of coarse-grained and fine-grained aggradation. Robinson and Lambrick (1984) and Lewin et al. (2005) discuss the archaeological dating and sedimentological character on the Thames floodplain. They both suggest that progressive filling of the palaeochannels has occurred over the full period of the Holocene (rather than the view of many other workers that palaeochannel fill occurred predominantly in the early Holocene).

The availability of sediment together with the amount and energy of the water flow has controlled river characteristics over the Holocene. Table 1 and Figure 1 below is a composite of climatic information available for various divisions of Holocene time, which may help to interpret some dated deposits from the Thames river system.

(From MMU Global Climate Change Resources on www.ace.mmu.ac.uk)

This brief introduction to fluvial processes is only relevant to lowland Britain, and particularly to the Thames. River systems and sedimentation patterns will be very different in other parts of the country.

The pattern of early Holocene alluvial sedimentation is one of a transition from a braided to a meandering river system. A large number of anastomosing channels became more stable in their course as woodland became established in the water catchment area as well as on the floodplain. Relict braidplain surfaces gradually became overlain by fine lateral accretion of organic-rich clays and silts. Later in the Holocene, notably during the Neolithic periods onwards, human changes to water catchment and sediment dynamics considerably altered the river system. The most important human interference in the landscape was the felling of woodland. This allows more sediment erosion, which is delivered to the river system. The input of sediment to the channel encourages a more rapid surface flow and, with less channel space for water, and greatly increasing flooding. Where flow is increased sufficiently, the less cohesive river banks could be eroded and migrating channels result. This is currently a contentious issue at the moment and experts are divided as to precisely when alluviation occurred and whether it is anthropogenically or climatically-driven. For a discussion of this issue see: Howard and Macklin 1999; Macklin, 1999; Taylor et al. 2000; Lewin et al. 2005. Macklin believes that anthropogenic changes have masked climate changes – working to either intensify them or to blur them at various times. Other workers seem not to challenge this idea, but stress that anthropogenic changes are more pronounced in the later stages (FIII from c. 2000 years ago).

Recent work on Holocene river environments (such as Macklin, 1999) has shown that the earliest part of the Holocene (Flandrian substage I, FI) was a period of low sedimentation (as determined by the number of alluvial units per 100 years, Fig. 2). This early stability is suddenly halted by rapid sedimentation at 6000 years ago (coinciding with the Neolithic and large woodland clearance for farming land and for fuel and buildings. Increased overbank flooding (thicker floodplain clays and silts) appears at 5,500 years ago (Flandrian II), but this time is also coincident with a warmer period. Further woodland clearance occurs just after 4,000 years ago (FII) and continued to the present-day (FIII).

Three trenches were investigated on Cookham Church Paddock in September 2005 (see the main report for the trench locations and excavation details, and also Figs 3 and 4 overleaf). Augers were sunk to collect further sediment samples that provided data from beyond the trench boundaries or at a greater depth than the final trench base. Sediment samples were provided from Trench 2 from both Auger 2 and Auger 3 as follows:

A lithological description is given for each sample below, and one sample (T2, A2 at –2.45 m) was sieved. This sample was the only one suitable for dry sieving – the other samples had too high a clay component.

Below is a description of all auger samples. Colours were determined using the Munsell Color Chart on wet samples.

Colour: dark brown with a hint of orange-red (5YR 2/2)

Grain size: clayey silt

Grain composition: clays and quartz with <1% minute shell fragments.

Colour: moderate brown (5YR 3 /4)

Grain size: very fine clayey sand

Grain composition: clays (25%) and fine quartz (75%)

Colour: moderate yellow-brown (10YR 5/4)

Grain size: fine clayey sand

Grain composition: clays (10%) and fine quartz (90%)

Colour: Yellowish orange-brown (between 10YR 5/4 and 10YR 6/6)

Grain size: fine to medium sand (this sample was sieved – see the analysis below)

Grain composition: quartz (98%), shell and black rock fragments (1%), clay (1%)

Colour: dark brown (5YR 3/2)

Grain size: clayey silt

Grain composition: clays and quartz, with <1% shell fragments.

Colour: moderate orange-brown (10YR 6/4)

Grain size: very fine clayey sand

Grain composition: quartz (90%) and clay (10%)

Colour: moderate orange-brown (between 10YR 5/4 and 10YR 6/6)

(see below for other observations).

This sample was suitable for a grain size analysis by dry sieving and the results are shown in Table 3 below as well as the graph of Figure 5 overleaf.

Figure 5 Histogram of Auger sample A2, -2.45 m (data in Table 3, p.11)

Analysis of the data from the graph produced as Figure 5 produces these grain size parameters:

Mode: 250 microns

Median: 250 microns

Mean: 200 microns

Sorting: Poorly sorted (1.42)

An asymmetric graph, with a skewed distribution to the fine fraction.

This sand was eroded by a water flow velocity of at least 0.3 m s^-1 and deposited in the paddock area at 0.05 m s^-1

This is a distribution consistent with fill following a flood event that cut a channel.

Each of the sieve fractions were examined separately, as this sometimes shows up differences in size with mineral or clast type. In this instance, there was a uniformity of clast type throughout the sieve set:

1000 micron fraction: Mostly well-rounded, glassy quartz grains with well rounded chalk fragments. There was also rare (<1%) shiny black, very well rounded rock fragments probably lydite (a form of chert). There was one unidentifiable bone fragment measuring 9 mm x 1 mm.

500 micron fraction: Similar to the 1000 micron fraction, mostly well rounded glassy quartz (but with some more angular). Well rounded black rock fragments, probably lydite (1%) and one well rounded and abraded bivalve fragments (1%) and very rare, very rounded chalk fragments.

250 and 125 micron fractions: as above.

The sedimentary sequence examined from the Trench 2 auger traverse indicates three related processes: channel incision, channel fill, and floodplain aggradation. The coarser gravel at the base of the auger profile (where augering was forced to stop) represents the coarse fill immediately over an erosional incision, which is overlain by the sands deposited as the flow waned. From the main report, the excavation records show that there are minor cuts within this infill layer (contexts 005, 003, 004 and 006 successively) representing minor channels as they switch or migrate across the floodplain. Subsequent alluviation represents enhanced sedimentation and a higher water level.

The whole sequence from the basal auger gravels to the overlying sand, silts and clays shows a gradual fining upwards sequence (Figure 6). The clay content increases upwards as aggradation forces the change from coarse fill to floodplain sediment build-up. Given the substantial nature of the channel cut (Figure 4) and the coarseness of the gravel, a high-energy event is indicated. Overlying the gravels is the sand and clayey-sand layers, which indicate the waning flow followed by aggradational to the present-day.

The substantial channel cut is a notable event - standing apart from the usual river processes. Also, given the position of medieval and post-medieval finds from contexts 002 and 003 within the aggradational units, it is suggested that the gravel incision may be related to the intense cold period known as the ‘Little Ice Age’ (Fig 1). In this case, a combination of channel incision during the coldest phase, followed by higher water levels with a faster flow during the following warmer period would explain the sequence seen at Cookham Paddock. However, the sequence lithologically resembles older strata elsewhere in the area (Harleyford, MAS 2005), although this may be coincidental. Without firmer dating evidence it is uncertain which fluvial event this channel incision and fill can be assigned to.

Howard, A. J. and Macklin, M. G. 1999. A generic geomorphological approach to

archaeological interpretation and prospection in British river valleys: a guide

for archaeologists investigating Holocene landscapes. Antiquity 73: 527-41.

Lewin, J., Macklin, M. G., Johnstone, E. 2005. Interpreting alluvial archives:

sedimentological factors in the British Holocene fluvial record. Quaternary

Science Reviews, 24: 1873-1889.

Macklin, M. G. 1999. Holocene river environments in prehistoric Britain: human

interaction and impact. Journal of Quaternary Science, 14: 521-530.

MAS. 2005. Report on excavations at Harleyford 2002, Marlow, Buckinghamshire.

Marlow Archaeological Society internal report.

Munsell Colour Chart 1991. The Geological Society of America.

Robinson, M. A. and Lambrick, G. H. 1984. Holocene alluviation and hydrology in

the upper Thames basin. Nature 308. 809-814.

Taylor, M. P., Macklin, M. G., and Hudson-Edwards, K. 2000. River sedimentation

and fluvial response to Holocene environmental change in the Yorkshire Ouse

Basin, northern England. The Holocene, 10(2): 201-212.

Dr Jill Eyers, December 2005.