Skip navigation

You are in:  Home » Research » Research projects » Controls on interglacial sedimentation

Interglacial fluvial sediments  

Interglacial fluvial sediments have been recorded from many sites in lowland Britain.  This region occupies the south and east of the country and contrasts physiographically in geology, Pleistocene history, tectonic setting, relief and river flow regimes from upland northern and western Britain.  These factors have resulted in variations in river channel pattern, sedimentation and river evolution in upland valleys.  Interglacial sediment sequences in the lowlands have been studied for two reasons, primarily to provide biostratigraphy in otherwise unfossiliferous sediments and secondarily to support a palaeoenvironmental interpretation of contemporaneous conditions.  An examination of the literature reveals that the bulk of the investigations of interglacial fluvial sediments has been to identify and interpret their fossil assemblages.  In the past this was only occasionally accompanied by some limited interpretation of the sediments that contain them.  This loss of important information on fluvial activity, much of which could have been used to inform the discussions of fossil preservation and representation at a site (i.e. taphonomy), is unfortunate.  However, more recent works have included fuller discussions and interpretations of the sediments themselves.

Interglacial fluvial sequences in the region generally occur within more complex aggradations, frequently underlying terrace or floodplain surfaces.  The most common situation repeatedly found is one in which predominantly fine, often laterally discontinuous, interglacial sediments are intercalated between coarse clastic (i.e. gravel and sand) sediments of cold stage, braided stream origin; the so-called ‘gravel-interglacial-gravel’ sequence.  This sequence, originally noted by Wymer and modified by Green & McGregor was developed as a model relating “terrace formation to climatic fluctuation” by Bridgland & Allen and Bridgland.  In this model the sequence is interpreted as representing respectively: (i) stream incision; (ii) deposition of cold-climate gravel and sand; (iii) deposition of minor channel-fills or of remnants of overbank sediment, preserved within the cold-phase sediments; (iv) downcutting under cold-climate leaving floodplain deposits isolated as terrace remnants; (v) renewed aggradation of cold-climate gravel and sand; (vi) and finally formation of the terrace surface and return to (i).

Whilst this model seems applicable in several cases, it does not apply universally to all depositional sequences underlying terrace surfaces, the majority of which demonstrably lack any evidence for interglacial sedimentation.  Indeed many fluvial accumulations were deposited exclusively under periods of cold-climate, e.g. in the Thames Valley and the Solent system. There is no basis for assuming that, if interglacial sediments are absent from a particular gravel and sand unit, this apparent three-fold sequence still applies, implying a subsequent removal of previously present interglacial sediments.  Nevertheless, where interglacial sediments do occur the model seems to hold, broadly describing but not explaining their relative chronological occurrence and deposition.

The figure shows a selection of sequences of typical interglacial fluvial sediments from 6 – 7 events throughout the region.  The sequences have been selected not only to demonstrate the variety of sediment facies, but also the thickness of units and the length of the time they represent.  Setting aside initially the time span, based upon the  substage scheme, it is apparent that sediment facies vary considerably in grain-size, thickness and lateral continuity.  The most abundant sediment is organic detritus-dominated mud (detritus mud) with varying quantities of inorganic material.  This sediment is frequently rich to very rich in fossil plant and animal remains, the latter being scattered throughout the sediment body, or locally occurring as discrete beds.  The muds are often interstratified with sand laminae or beds, occasionally with coarser sediment that may include the remains of trees and large vertebrates. Gravel-dominated units occur but are very rare (e.g. at Ardleigh, Essex).  By contrast, fines (clay and silt) – dominated units are very common and may be very calcareous in places where they are commonly termed marls.

Sediment thicknesses vary conspicuously, although it is clear from the figure that the predominant sediments are generally less than 2 m thick and often show lateral variability both in thickness and composition.  Locally greater thicknesses are found but they are exceptional.  Of particular interest are the infills of enclosed floodplain funnel-shaped hollows, several metres deep, such as the Last Interglacial –age sequence at Elsing, Norfolk and the Holocene sequence in Lambeth - Southwark, central London.  Such hollows have been attributed to bedrock chalk solution, possibly associated with local spring activity.  

In terms of sequence, no one pattern predominates.  Fining-upwards profiles are known, e.g. in the lower part of the Swanscombe sequence which appears to represent channel infill and subsequent development of a floodplain surface by vertical build-up of fines.  Such surfaces may subsequently be subjected to pedogenesis.  At Sugworth and Little Oakley, lateral facies changes from cross-stratified sand or silty sand to inorganic fines interbedded locally with organic sediment indicate that the sequences could represent point-bar accumulations.  At West Runton, the sequence overwhelmingly comprises organic-rich silts, rich in plant macrofossils that are interbedded with sands containing abundant mollusc shells.  Both articulated and disarticulated vertebrate remains are common and the complex sediment relationships suggest that large mammals were responsible for disturbing the sediments (bioturbation).  The disturbance of sequences by mammals is also known from other localities such as Barrington, Cambridgeshire.  Transitions from inorganic to organic fine sediments are common, indicating accumulation in a depression or channel that becomes cut off from clastic inwash.  However, reverse changes also occur implying a renewed input of channel flow possibly arising from channel avulsion, or an exceptional flood transfering mineral sediment onto a floodplain.

It is particularly noticeable that the longer sequences at Histon Road, Mannings Farm, Highbury, Barrington, Ilford and Swanscombe all include significant thicknesses of inorganic sediment, mostly fines but also sands and locally even gravel, occasionally cross-stratified.  These sequences are markedly less fossiliferous, often suggest sheet-like sedimentation, and are frequently partially oxidised.  In terms of timing, these inorganic sequences consistently represent the second half of interglacials and continue into the subsequent cold stage.

By contrast, the richly organic detrital sediments and marls represent the first half to two-thirds of interglacial time, on the basis of the pollen zonation.  These sediments represent the majority of preserved interglacial sequences.  They fill depressions or channels often resting directly on and cloaking the surface of cold stage coarse clastic deposits or bedrock surfaces beneath. Channel-like infills are particularly common, as are locally included highly mixed accumulations of organic and fossiliferous material.  Such an accumulation at Woolpack Farm, Cambridgeshire, described as ‘diamicton-like’ by Gao, was attributed to a substantial flood event from major storms or collapse of a beaver dam. The organic deposits frequently show lateral variation in age within the interglacial event.  This is seen at sites such as Ardleigh and Wretton where accumulations represent the different substages of the interglacial, yet all occur at similar level.  This variability is thought to typify accumulation in a floodplain complex (see below).

Next page