Lost and future worlds: marine palaeolandscapes and the historic impact of long-term climate change

Alice Roberts explores the importance of coastal routes to our ancient ancestors, colonising the globe in the Palaeolithic, and looks at sites through prehistory which show how humans exploited and adapted to watery places, from Pinnacle Point in South Africa to the earliest humans in the Americas, and the Bronze and Iron Age landscapes of the Somerset levels and the Cambridgeshire Fens. Submerged landscapes hold the key to understanding so much about our ancestors, but can the ancients’ flexible approach to their environment also hold lessons for us, today?

Inundated and submerged prehistoric landscapes on the continental shelf have been studied internationally for decades, but, in the early years, only by individual scholars and local research groups.  The academic and political status of the topic has increased recently and the value of the research has been noted in the 72nd session of the Oceans and Law of the Sea report to the Secretary General of the United Nations for 2017. There are no formal national or multi-national commitments specifically to this research area, although numerous treaties, conventions and regional agreements are structured in such a way that submerged landscapes and prehistoric sites are included in their provisions. The definition of Underwater Cultural Heritage in the UNESCO Convention includes prehistoric sites, as does the United Nations Convention on the Law of the Sea. In Europe DG-MARE recognises the value of mapping submerged landscapes, and maps and data bases are being developed by EMODNet. The European Marine Board and the European Archaeology Council supported the publication of a joint report on the subject. Regional groups such as SINCOS and NSPRMF have supported work in the Baltic and the North Sea, while a longstanding sequence of INQUA-IGCP projects has been organised in the Black Sea. In the USA NOAA has a budget line for submerged landscapes, while the SPLASHCOS project supported by the COST Action Office integrated available data across the European seas with budget support for 4 years. As yet there are no permanent institutional or treaty-type commitments specifically aimed at supporting submerged landscape research, and this is a point of weakness. Everything still depends on individuals persuading their host institutions or agencies to provide support for the research. In spite of the high status of some of the support being provided, and the excellent research undertaken in many countries, the financial support is still minimal, and there are no long term commitments or funded institutions

Landscapes have evolved throughout geological time through a combination of geological and climatic processes. On the time scale of Homo sapiens the dominant physical changes in global landscapes are a consequence of the ice ages with their concomitant rise and fall in sea level and flooding and re-emergence of the continental shelfs and islands. Flood legends from many parts of the world attest that humans have attempted to explain their experiences of these changes within their framework of knowledge and the present discussion on the subject is one further step in that process. As ice sheets wax and wane, sea levels globally and on average, rise and fall in unison.  But the local or regional response can vary significantly from place to place due to the response of the planet to the redistribution of ice and water. The Earth deforms internally and surficially in response to these changes and the ocean surface re-adjusts to this new shape and associated changes in gravity.  This response is not instantaneous but lags behind the glacial cycle by amounts that are geography dependent. The result is a complex spatially and temporally variable global pattern of sea level change that cannot be adequately represented by simple concepts such as eustasy but one that contains important information on both the response of the Earth to forcing and of the glacial history. Such variability is indeed observed world-wide but the evidence is sufficiently incomplete and often imprecise to give a comprehensive description of this change. Thus physics-based models, founded in well-tested theories and observations from other geophysical and geological disciplines, have provided the means to interpolate between the disparate observations and to provide a predictive capability of sea level change and shoreline migration (always putting aside that the glacial cycle is not the only process that operates).  These models and their limitations will be discussed in this presentation, tested against observational evidence, and used to reconstruct coastal landscapes for some of the regions that may be relevant to some of the following presentations during this meeting.

Very high resolution (VHR) seismic imaging of buried palaeolandscapes in the southern North Sea poses a number of problems. The often sandy sea floor induces strong multiples that may obscure the data especially in shallow water (<10-15 m). Detailed imaging below sand banks is difficult because the short wavelength sound waves are quickly absorbed in the heterogeneous sandy sediments. Intertidal areas furthermore add additional challenges due to highly variable water depth (0-5m), wave action and strong currents. Within the framework of the SeArch project (www.sea-arch.be) an efficient survey methodology was developed for archaeological prospection of the subtidal and intertidal area. In the subtidal area this was done through the comparison of a range of seismic sources (covering a wide range of the frequency spectrum) and receiver combinations. The results show that sparker sources generally provide the best trade-off between image resolution and penetration depth. Surprisingly, multichannel recording significantly improved the image resolution already for shallow target depths. In the intertidal area a novel technique based on multi-transducer parametric echosounding was applied. This resulted in ultra-high resolution 3D images of the shallow sub-bottom (cm/dm range). Grid cell sizes as small as 20x20cm allowed to identify buried features in unprecedented detail. This opens important new perspectives for archaeological prospection in shallow water areas.

Occupation in Australia is now dated back to 50 ky BP, and for the bulk of this period sea level was lower than present.  Nearly one-third of Australia’s landmass, hence a significant part of the archaeological record, was drowned by the post-glacial transgression. Despite this, research aimed at finding submerged prehistoric archaeology on the drowned shelf is only just beginning. Over the next few years, a pioneering, multi-disciplinary study of submerged landscape archaeology will investigate the records of the now-submerged Pilbara coast in NW Australia (spanning 50 to 7 ky BP) and contribute a unique southern hemisphere insight into world prehistory.  For Australia’s NW Shelf, knowledge is increasing rapidly on its preserved drowned palaeoshorelines through use of high-resolution remote sensing data, palaeotidal and 3D modelling.  Computer-based visualisations, which overlay relative post-glacial sea levels on the modern bathymetry, help conceptual understanding of the changing coastal landscape.  However, the past development of coastal resources, their potential human use and the preservation of this in the archaeological record is primarily controlled not by sea level per se but by second and third-order effects associated with variations in coastal sedimentary environments.  Considering such physical processes in archaeological thinking will help and improve targeted prospection of submerged cultural sites, and will help stimulate new ways to include such information in conceptual models of human occupation and help us solve for the unknown.

Human DNA evidence of living populations suggests that between ~30,000 and ~15,000 yr BP, people paused in their migration into the New World from Siberia.  Where did they pause? Was the now-submerged Bering Land Bridge (BLB) a suitable place to live?  Vegetation reconstructions based on a variety of proxies (pollen, plant macrofossils, aDNA) indicate wide-spread herb-dominated tundra on the BLB, though there is intriguing evidence that woody taxa may have been more abundant than is generally assumed. Full glacial and late glacial pollen distribution maps from across Beringia, of which the BLB is only the central region, suggest in situ expansion of key taxa (cottonwood/aspen, spruce, pine, birch, and alder) and not migration from outside the region.  In addition, DNA results on modern spruce indicate long isolation of Alaskan populations from those growing south of the Laurentide ice sheet, suggesting that some Alaskan populations could have survived the last glacial age in Beringia.  Finally, some localities on the BLB may have been somewhat mesic, thus providing suitable habitat for the woody taxa to occupy.  However, issues with chronology, site taphonomy, and pollen-vegetation linkages do cloud the issue.  This presentation summarises current data on the BLB and adjacent regions and presents the pros and cons of the BLB as a suitable landscape for long-term human occupation.

The Pleistocene occupation of northwest Europe occurred during a period characterised by significant climatic changes. In Britain, hominin populations were present discontinuously through varied climates, from cool continental sites such as Happisburgh 1 and 3, to the balmy Mediterranean conditions of Pakefield. Whilst apparently able to survive through a range of conditions, the configuration of these landscapes, in particular the presence or absence of a land connection with the continent, would have had a fundamental impact on the density of hominin occupation in Britain at any given time. Similarly, the ecologies once present in the now-drowned North Sea Basin and Channel regions would presumably have played an important role in attracting and sustaining sporadic hominin populations. However, our understanding of the physical and environmental character of these landscapes, as well as the timing of transgressive and regressive periods, remains frustratingly murky, hindering the ability to imagine how these landscapes may have been used. A growing body of data from recent research and commercial offshore development is, however, beginning to address this issue: we can now start to piece together a more coherent picture of landscapes through time, but there is still an immense amount of work to do. By drawing some of this recent work together, this paper will discuss the potential configuration of these landscapes throughout the Pleistocene, the impact of fluctuating climatic changes upon them and the resulting implications for the hominins they sustained.

Events following the last glacial maximum (LGM) in high latitudes of North America illustrate the “push–pull” impacts of rising sea level and glacial recession on human and biotic communities.  These forcing mechanisms transformed northern North America resulting in: 1) separating the vast land connection between Asia and North America, 2) joining the Arctic and Pacific Oceans, 3) landward retreat (push) of human and near-shore biotic communities in response to rising sea level, and 4) the colonization (pull) of biota and humans in newly deglaciated regions.  Understanding and quantifying the timing and mechanisms of this dramatic reconfiguration of North America provides insights for the application of complex systems modeling to geographically predict and scale future “push-pull” events likely to result from climate change.