River terraces and corresponding archaeological finds

(Words, marked in red are linked at  the glossary of this website) 

River terrace sequences are a helpful medium in the chronostratigraphy of fluvial landforms. Archaeological correlate from the river terrace horizons could be dated either by relative or by absolute dating methodology. Both dating methods have their applications and limited possibilities depending on the occurrence of conjoinable material (Mishra et al. 2007)

Surface finds

In case of surface finds, absolute dating methods (OSL, ESR )are of coarse impossible, so only the relative age of a river terrace is an appropriate framework for the chronostratigraphy.(1)
In case of surface finds,we must keep in mind, that different geological processes after deposition (2) played a dominant role in the disturbance of a possible chronostratigraphy related to the original  river terrace.(3). Moreover, the geology of the river terrace has to be considered, to be sure we deal with an original  depositional terrace instead of an erosional terrace.
Surface finds, described at this website, do not really come from the "surface" ( image), as eroded horizons lie on top, and more important still are visible and identifiable. In the local and more the regional stratigraphy there's a difference in the gravel carrying horizons in the northern part of the concerning Belgian Kempen area  and the southern part of it. In the southern regions we distinguish a mix of Cromerian(?) -type sand ("Winterslag - zanden")  in the gravel layers. Natural disturbance, erosion ( e.g. by former human activities)  and disconformity of the layers causes disturbed contexts in the upper horizons where artifacts can be found.

Surface finds found at eroded surfaces do correspond with a certain horizon in the local stratigraphy , but top horizons could be  mixed

Methods in the MA project

The "MA project" has started 2008 and is still ongoing. It's gain was to reveal early human presence along the Maas/ Meuse river, roughly between the Belgian communities of Visé - Lanaye in the south and Opoeteren on the Belgian west-bank in the north and between the southern location of Eijsden and the northern location of Elsloo on the Dutch east-side of the river.
For several reasons the "MA project" only was mainly  focused on the upper terraces of the river Maas /Meuse.
In the first place, the high terrace (= plateau edge)  is the oldest incision made by the river, so a former riverbank would be positioned in the nearby locations of these levels.
Secondly, eroding surfaces could show remains of former human activity, such as tools, carbon or ( in theory) bone material. In the third place, eroding surfaces would show possible layered  stratigraphies, providing information about the geological processes in the past and after- deposit -processes of the plateau edge.
The age of the upper terraces along the Maas /Meuse river in the examined area is depending on the longitude: to the north and the west the upper terraces are ( much) younger because of uplift of the eastern land form units, so the southern and eastern river terraces were build during older Pleistocene stages. There is of coarse a corresponding sequence in the north - south division, where the older river terraces in Southern Limburg simply were located further from the rive, up to 20 km. In this case, the Gravenvoeren terrace, dating back to 590.000 BP, has about the same age as the Kempen Plateau, both formed during Late Cromerian glacials.
Because of intense weathering processes after the first incisions of the river, the high terraces are often eroded and horizontally "deformed" ( lower river terraces were covered by eolian and colluvial deposits and do  have a much better preservation), so the present upper terrace edge is not exactly the former Pleistocene river bank, as this is in many cases covered by eolian sand depositions from the Saalian Complex, so it is often  hard to tell at which point exactly the plateau terrace would begin to slope down(4)
This made it difficult to predict where archaeological remains could be found. The first and main determinants for finding possible archaeological locations were the eroded surfaces.
Many such places were visited, varying from road paths in forests, forest clearing areas, (semi-) arable lands and building sites. Notice, that all these locations are influenced by human activities today. Naturally eroded surfaces of the plateau edges are relative scarce.(5)
Only at those  locations where the ( local) stratigraphy could be detected, some reconstruction of a global view of the former Maas/ Meuse riverbank was possible ( see image below this article).
In most cases, the horizontal stratigraphy could be derived from a local vertical stratigraphy(6) in the nearby area ( < 2 km), by comparing samples from both stratigraphies and by determining specific horizons in the stratigraphy, i.e. the first vast (unmixed, pure) eolian sand deposition above the reworked sand depositions (sands and gravels).
Despite of this information there is not yet much to say about the former Pleistocene channel edges, except the fact, geological processes are more clear now ( see : results) and the extrapolation from archaeological find-spots, the Maas/ Meuse river had a level decline of about 15 -17 meters over 20 kilometers between the pre- Saalian period and the late Holsteinian ( roughly between 350.000 BP - 450.000 BP).
In the next step of the project, the palaeo- geographic situation is important, to answer questions about the choice of settlement and activity locations and the possibilities / limitations  of this particular  palaeo- landscape. For this it is important to have in mind the generally accepted idea of dispersal pathways and  settlements of early hominids in the river basins of Europe during the Middle Pleistocene (Fleagle, 2010). The Kempen area is no exception in environmental potential for early hominids or early human settlers. The ultimate step in the MA project would be the most important: the search for any remains of the Pleistocene fauna and /or  early hominids in the region.

Some important results concerning the relation river terrace - artifacts in the Kempen area

Locally observed stratigraphies of the MA6- site showed a slowly downward sloping Pleistocene channel edge during the formation of the As palaeosol. In a slope gradient of MA6 (7), with orientation west- east ( towards the river), the Middle Pleistocene ( currently eroded ) channel edge, consisting of small sized sub -angular gravels, was visible over more than 100 meters. The finds of in situ artifacts from some of these horizons was important as the artifact types match with artifacts found at other locations from the upper terrace.
Two major aspects were important in the determination of the artifacts original  periods of  use/ deposition. At first, the artifacts were 'integrated' in the gravel deposits with a similar appearance like the naturally gravels. The only important difference for the tools were the strikes made by humans, to transform them into tools. Finds of both processed quartzite and quartz as well as blank material demonstrates  the abundance of material for the tool- makers at the sites. Secondly, the gravel depositions in all find spots contain a gravel type that is different in composition and presence. It is reducible to two different horizons, one containing small gravels / sands containing a black flint type, the other consisting of coarse sands with more large pebbles, overlaying a (younger) ( dark yellow - yellow) clay horizon.
The relation with the upper terrace and its possible relative age has become clear because
* the similar appearance of both artifacts and gravels/ pebbles from its corresponding gravel horizon  
* the limits of the size of the artifacts were corresponding with size limits of the gravels where they were found
* though long term use might be possible, no palimpsests or mixed periods could be detected, though the assemblages are not diagnostic for some period
* The artifacts had no signs of post- deposition transport ( no rounded edges or different patinas)but did show alteration, possibly mainly from the pre- deposit period
* The absence of any diagnostic artifacts, pointing into any other period than the Lower/ middle Palaeolithic, such as processed flint, prepared platforms, artifacts in a Levallois technique,  has been decisive in assigning a Lower /Middle Palaeolithic - Middle Pleistocene period for the use/ deposition of the artifacts.

The key indicators, for the attribution of the discovered  surface artifacts to the Lower Palaeolithic, were based on both the assemblage - type and the find location at the upper terrace, with the fact the artifacts do not show much abrasion and weathering indicating they could not have been moved over long distance and were processed at the same bed of the river.  Though Upper Palaeolithic and Mesolithic finds are rather 'common' in the area, these could be excluded by the above listed arguments.(8)
For a different situation I will describe the situation more to the south, south of Maastricht (NL) in the Belgian area of Caestert ( Visé) (9).
The upper terraces of the Maas /Meuse river at Caestert (B) have been used for a very long period. Evidence from stone tool objects show a site use from (Early-) Middle -Palaeolithic to the late Neolithic.(10)
Why is this not the case in the sandy Kempen area?

When the river build new terraces the distance between upper river terrace and the river became bigger. Not only the distance was an important fact, the slope gradient was also important. Between the Middle terrace (Eisden -Lanklaar, formed during the Saalian ) and the upper terrace we find a steep gradient over 25 meters. For humans in the Kempen region after the Saalian periods, ( i.c. >300.000 and younger, during pleniglacial- periods, the Upper Palaeolithic, Mesolithic, Neolithic) the upper terrace only was an interesting place when there was ( running) water available ( sources, brooks, ponds).
Indeed, at the Schootsheide (Rothem) we find both circumstances on the Saalian terrace: the Bergerven a large pond could be the water-supply for Mesolithic hunters, settling at a high place nearby such ponds.(11)
Another very important fact for the long term habitation of Caestert is, this site is located between the Maas/ Meuse river and the Jeker / Geer affluent, which  made it a very special place, together with the abundance of silex in the region. Since the middle terrace in this area has been dug out for the local Albert- canal there is nothing to say about a continuation of habitation on  the local middle terrace. The example demonstrates the differences in environmental positions along the Maas/ Meuse river, influencing prehistoric humans settlements strategies.

Simplified schematic river terraces showing the relation between the fluvial landforms and corresponding periods.

Early humans  could  easily choose pebbles in the large variety of gravel types and dimensions  on the former Maas / Meuse beaches, just like in present times similar beach types are visible ( Helle brook, Eupen, Belgium)

Image below: Generalized reconstruction of the declination of the Maas/ Meuse river during Hoslteinian and pre- Saalian period, based on archaeological evidence in the region. The view is general, but it's impossible to make a better image  with the data available. Present south - north decline of the Maas/ Meuse river is about 1 m./ km; during the Tertiary the decline is estimated 1,5 m/km , based on current tertiary levels over 20 km (south-north 90m - 60m);  in case of the 17 m decline over 20 km during pre- Saalian/ post Elsterian conditions, based on corresponding archaeological finds, this ratio gives an estimated 0.85 m /km, implicating a slow meandering river during the interglacial conditions of the periods of habitation.


Notes from the text
(1) S. Mishra et al.; Fluvial deposits as an archive of early human activity(2007) i.c. the Kempen area: Gullentops 1974; Zagwijn 1996; Busschers et al. (84-1: 25)  2005; Sedimentary architecture and optical dating of Middle and Late Pleistocene Rhine-Meuse deposits – fluvial response to climate change, sea-level fluctuation and glaciation PDF (NJG) 
(2) Processes like eolian sand covers, erosion, solifluction, cryoturbation. etc.
(3) See  the Sado River Drainage Survey  in Portugal and the British project IGCP 449 as examples for the problems encountered in the diagnostic properties and the dating of surface finds.
(4) These processes are  also depending on which side of the river was made by erosional or depositional processes: during several glacial periods,  parts of the depositions again were removed by braided rivers. So often the current plateau edge is definitely not the former riverbank.
(5) Even the "naturally" eroded surfaces were influenced by present activities , such as trampling of cattle ( in meadows). The only naturally eroded surfaces in the area were tree falls, so limited surfaces.
(6) Corresponding stratigraphy from local sand pits.
(7) MA6, the best studied location in the area, is located in the Maasmechelen area, corresponding to the stratigraphy of As. See the article on this website MA6 the possible Holsteinian context
(8) See for an overview of corresponding sites from these episodes in the Kempen- region: Daniel Cahen et Paul Haesaerts; D Watteyne;  Peuples chasseurs de la Belgique préhistorique dans leur cadre naturel. Bruxelles : Patrimoine de l'Institut royal des sciences naturelles de Belgique, 1984
(9) This region is totally different in geological context: cretaceous bedrock and loess cover.
(10) In the ASAC (Vaals) a very large collection of surface finds (>8000 )from Caestert a.s. there are diagnostic artifacts from the Lower Palaeolithic: chopper- chopping tools ( quartz), Middle -Palaeolithic Levallois- technique based artifacts and smaller hand core  -tools,  Mesolithic microliths and bladelets (imported flint), Neolithic cores and blades and a polished handaxe made of local flint from the late Neolithic. A main geological difference is the fact, at Caestert the middle terraces, which originate to the same period as the upper terrace at the Kempen area, are cleared by former river dynamics.During the Holsteinian, it is possible these terraces were intact and inhabited.
(11) See e.g. : Marc De Bie / Marijn Van Gils Uitgestrekte mesolithische site-complexen in de Kempen. Ravels Witgoor en Opglabbeek Ruiterskuilen-Turfven (boorcampagne 2002) -PDF

No comments:

Post a Comment