Prehistoric landscapes of the North Sea

Dutch fishermen often discover remains of extinct large mammals such as woolly mammoths and aurochs, indicating that the North Sea was once inhabited by hunter-gatherer communities. Archaeological research both onshore and offshore seeks to map these ancient landscapes. Offshore efforts use seismic and geotechnical surveys to locate prehistoric campsites, focusing on specific landscape features. The research that is described in this article aims to understand the evolution of ancient environments and enhance the preservation of the North Sea’s maritime cultural heritage.

As Dutch fishermen have known for decades, the North Sea region was once a barren landscape populated by large mammals long extinct. The fishermen frequently find the remains of these mammals – woolly mammoths, rhinos, aurochs, Irish elk (giant deer) and reindeer – in their nets. Artefacts made of bone or antlers and even human remains are also sometimes part of the by-catch. These finds prove that migrating hunter-gatherer communities populated the vast North Sea area at the end of the last Ice Age (some 13,000 years ago) and the Early Holocene (11,650 to 8,700 years ago). Remains are also found due to other activities such as sand extraction for beach replenishment along the Dutch coastline. Collectors scouring the beaches for fossils and hidden artefacts recently found hyena coprolites (fossilized faeces) and a 50,000-year-old Neanderthal tar-hafted flint tool.

Under Dutch law, a developer of an area must determine whether remains of potential archaeological value are affected by the planned activities. Archaeological research is conducted in successive steps, starting with a desk study and followed by field surveys to explore, map and valuate archaeological sites. This stepwise approach applies to both onshore and offshore research, yet there are some marked differences.

Onshore, a wealth of historic, archaeological and geological information is available to build a detailed predictive archaeological model. Field surveys generally focus on the tracing and mapping of pre- and proto-historic settlements. Simple tools such as an Edelman hand auger and a gouge auger suffice to meet this objective, after which the archaeological site can be valuated by digging trial pits.

Offshore, our knowledge of the shallow geology is limited. Although last century’s geologists did a tremendous job by producing geological maps utilizing seismic data, the amount of borehole data to ground truth this is too limited to reach the level of accuracy and detail needed for geoarchaeological research. Research programmes are currently being carried out and cooperation sought by the Geological Survey of the Netherlands at the European level to fill the knowledge gaps.

Submerged prehistoric landscapes

In the early stages, offshore archaeological research primarily focused on the tracing of historic wreck sites and WWII aircraft. That has changed. The Dutch Cultural Heritage Agency (RCE) has placed the research of submerged prehistoric landscapes high on its agenda, underscoring the importance of gaining a better understanding of “an important and sometimes overlooked element of the maritime cultural heritage of the southern North Sea.”

Prehistoric hunter-gatherer campsites are generally small, featuring a sparse array of flint artifacts and hazelnuts, for instance. Such sites are extremely difficult and costly to trace in the subsurface of the North Sea area. Offshore, the primary aim of archaeological research is therefore not to trace settlements but to search for the specific parts of the landscape in which hunter-gatherer campsites are found. Sand dunes and ridges, outcrops of boulder clay and higher grounds along fresh-water brooks are the preferred locations. It is therefore imperative to obtain a picture of the prehistoric landscapes that are hidden beneath the seafloor.

Abundant detailed information of the subsurface is collected during seismic and geotechnical surveys carried out for the development of offshore wind farm zones. Commissioned by RVO, Periplus Archeomare used this data to assess the submerged prehistoric landscapes of IJmuiden Ver Wind Farm Sites alpha and beta (IJVWFS alpha & beta).

Layers and sediments

Figure 1 shows a cross-section of IJVWFS alpha and beta, based on seismic data acquired by GEOxyz. In Figure 1, we see that Unit A consists of Holocene mobile sands of the Bligh Bank Member. Sand ridges and dunes are part of this unit, which has a planar erosional base. In the southern Brown Bank area, Unit A is absent. Here, Unit C is exposed at depths below 34m LAT. Unit C consists of Early Glacial lagoonal and shallow marine deposits of the Brown Bank Member, with laminae and layers of very fine sand, silt, clay and detritus. According to the Flemish Bight geological map, the top of the Pleistocene sequence consists of the Brown Bank Member, with locally a less than one-metre-thick cover of the Boxtel Formation. Unit B has a thickness of some 5.5 metres and a highly irregular base that sharply truncates underlying Unit C, the Brown Bank Member. Unit B appears to have been formed by the deposition of homogenous sediments in a high-energy environment with strong erosion of the then existing landscape. According to the geological maps of the area, Unit B is initially interpreted as Early Holocene tidal deposits of the Naaldwijk Formation.

A channel feature can be seen in the north-eastern part of the cross-section, shown in plan view in Figure 2. The channel incises both Unit B and Unit C and is therefore younger than these units. Peat is found locally in the upper parts of Unit B and the channel infill. The presence of a channel feature and peat bed raises several questions. Has the channel feature formed through incision by a fresh-water brook in a terrestrial landscape or a tidal channel in a near-coastal marine environment? What is the age of the peat? What is the timing of incision and infill? What did the surrounding landscape look like? What is the age, depositional environment and lithostratigraphy of the units that are incised by the channel?

To answer these questions, the sampling strategy focused on the channel feature and its surrounding landscape. Three vibrocore locations were selected at three locations (nine in total) to sample the channel infill and sediments of the bordering landscape. Sediments of Units B, C and D were sampled at three other locations to obtain optimal vertical coverage of the sedimentary sequences in the area.

Evolution of landscapes

The description of vibrocores is not limited to the lithology. Special interest is paid to sedimentary structures to assess the depositional environment, the character of layer boundaries (erosive versus non-erosive), and phenomena that point to secondary processes such as soil formation, rooting, ripening and decalcification of clay, bioturbation and cryoturbation.

The aim of the research is to picture the evolution of aquatic and terrestrial landscapes. To meet this objective, microfossils (foraminifers and ostracods), diatoms and palynomorphs (pollen, spores, algae, fungal parts and insect remains) were extracted from selected sediment samples for detailed research.

Ostracods, foraminifers and diatoms are microorganisms that live in aquatic environments. The fossilized remains of these organisms are found in sediments. Environmental variables including substrate, water depth, temperature, salinity, organic matter and dissolved oxygen determine whether the habitat is fit for different species of ostracods, foraminifers and diatoms. From the diversity and abundances of the identified species, the aquatic environment can therefore be deduced.

Airborne pollen and spores are deposited on land or at the bottom of lakes, lagoons, brooks, salt marshes or the seabed. The different types of pollen found in a sample give an impression of the vegetation of the surrounding landscape and climatic conditions during deposition. The pollen distributions are correlated with known pollen zones, to judge whether the pollen distributions fit the initial lithostratigraphic interpretation.

Plant remains, peat and molluscs are collected from the sediments for radiocarbon dating. As radiocarbon dating is restricted to sediments that are younger than 50,000 years, optically stimulated luminescence dating (OSL) is used to date sandy sediments that include sands older than 50,000 years. This works as follows: quartz grains that are covered by other sediments after deposition store the energy that is emitted by surrounding natural radioactive minerals in their crystal lattice. The amount of energy stored in a single quartz grain is therefore equivalent to the time that has passed since its burial, provided the grain is not exposed to light. OSL uses this characteristic of quartz to determine the time of deposition.

Vibrocore IJV506 sampled the central part of the channel feature (see Figure 3). From 2.07m to 3.22m, the infill consists of non-calcareous silty fine sand (see Figure 4). The rooted sands are interpreted as Late Glacial fresh-water brook deposits of the Singraven Member | Boxtel Formation. Radiocarbon (13.8 ± 0.2 cal ka BP) and OSL dating (14.1 ± 0.8 cal ka BP) support the inferred Late Glacial age, pointing to deposition during ‘warm’ Bølling-Allerød interstadials.

The Early Holocene climate warming led to a rise in sea level and inundation of the North Sea area. Groundwater was pushed up in bordering coastal zones, and fens, marshes and swamps developed in which peat was deposited. A bed of such peat (Basal Peat Bed | NIBA) is found at 1.90m to 2.07m, covering the Pleistocene sands of the Boxtel Formation (BX). Brackish water clay (Velsen Bed | NAVE) that conformably covers the peat mirrors the first marine ingression in the area. Peat and organic clay were deposited between 9.4 and 9.0 cal ka BP, when an open landscape with pine, birch, grasses and sedges transformed into a more densely forested landscape with mixed pine-birch-oak-hazel woods.

Conclusions

The outcome of the IJV506 vibrocore analysis has major implications for the geogenesis of the area. Evidently, the Late Glacial brook is younger than the deposits of Unit B which it incised. The inferred Holocene age is therefore ruled out for Unit B. On geological maps, the Brown Bank Member forms the top of the Pleistocene sequence. However, the semi-transparent seismic facies of Unit B do not fit the Brown Bank Member, whereas the plan-parallel sub-horizontal strata of Unit C do. The deposits of Unit B post-date the Early Glacial Brown Bank Member and predate the Late Glacial Singraven Member.

The top view of the base of Unit B shows a morphology that resembles that of a braided river system. Correlation with palaeogeographical maps of the North Sea area results in a light bulb moment. During the cold Early Pleniglacial, the westward extent of the Rhine catchment area appears uncertain (see question marks in Figure 5). The scarcely vegetated landscape and peak discharges during the summer months turned the Rhine into a braided river, which transported large quantities of meltwater and sediment to the North Sea area. It now seems likely that a northern tributary of the Rhine crosscut the IJVWFZ during the Early Pleniglacial. The high-energy braided river deposited poorly sorted sands (Unit B) and incised the then exposed Early Glacial Brown Bank Member (Unit C). A sand sample from Unit B (IJV501) reveals an OSL dating of 69 ka, which supports the Early Pleniglacial age. The Geological Survey of the Netherlands plans to analyse the heavy mineral assemblages in the sand, to test whether the sands indeed were deposited by the Rhine. If so, Unit B will be mapped as the Kreftenheye Formation instead of the Brown Bank Member.

The geoarchaeological analysis of 12 vibrocores has changed our view on the evolution of prehistoric landscapes in this part of the North Sea area. The research findings cast archaeological finds in the Brown Bank area in a new light. In 2005, a fisherman found a decorated bison bone in his nets, south of the Brown Bank. The artefact was dated 13.5 cal ka BP, which proves that modern humans lived in the area at the end of the last Ice Age. The fresh-water brook in the northern part of the research area dates from that same period, making it tempting to imagine that the brook valley was visited by the man or woman who produced this artefact.

The presence of the Rhine in this part of the North Sea during the Early Pleniglacial impacts the possible landscape use by humans, with a Neanderthal skull fragment and tar-hafted flint artefact both found in the context of Pleistocene Rhine deposits. Future geoarchaeological research for offshore developments will, bit by bit, aid our understanding of the evolution of prehistoric aquatic and terrestrial landscapes in the North Sea area.