Ireland: Submerged Prehistoric Sites and Landscapes 11 Kieran Westley and Peter Woodman Abstract archaeology. Realistically, the Irish shelf is Evidence of Ireland’s drowned landscapes and likely characterised by pockets of preserva- settlements presently comprises 50 sites tion, which makes detection and study of sub- spread across the entire island. These com- merged landscapes difficult but not impossible. prise mainly intertidal find spots or small col- A range of potential routes of investigation are lections of flint artefacts. A handful of fully identifiable, including site-scale archaeologi- subtidal sites are known, generally from near- cal survey, landscape-scale seabed mapping, shore regions and consisting, with one excep- archival research and community tion, of isolated single finds. Evidence of engagement. organic remains is also sparse, with the excep- tion of Mesolithic and Neolithic wooden fish Keywords traps buried in estuarine sediments under Submerged prehistoric sites · Intertidal Dublin. The relatively small number of sites is archaeology · Holocene sea-level rise · probably due to lack of research as much as Submerged forests · Fish traps · Maritime taphonomic issues, and thus the current evi- adaptation dence hints at the potential archaeological record which may be found underwater. Such evidence could contribute to knowledge of the coastal adaptations and seafaring abilities of 11.1 Introduction Ireland’s earliest inhabitants. Nonetheless, taphonomic considerations, specifically relat- The island of Ireland lies to the west of Great ing to Ireland’s history of glaciation, sea-level Britain, separated from it by the Irish Sea. At change and also modern oceanographic condi- present, it is politically divided between Northern tions likely limit the preservation of sub- Ireland (part of the United Kingdom) and the merged landscapes and their associated Republic of Ireland and, as such, its historic envi- ronment is recorded and managed under two K. Westley (*) separate jurisdictions. However, in this chapter, School of Geography and Environmental Sciences, the submerged archaeological record of the entire Ulster University, Coleraine, Northern Ireland, UK island will be considered as a whole, for the sim- e-mail: [email protected] ple reason that this political boundary did not P. Woodman (Deceased) exist in prehistory. The aim of this chapter is to Department of Archaeology, University College review the extant evidence for submerged archae- Cork, Cork City, Republic of Ireland © The Author(s) 2020 221 G. Bailey et al. (eds.), The Archaeology of Europe’s Drowned Landscapes, Coastal Research Library 35, https://doi.org/10.1007/978-3-030-37367-2_11 222 K. Westley and P. Woodman ological landscapes around Ireland, discuss any Carden 2016). The former presents the strongest patterns visible in the data, identify the potential available evidence for, at the very least, a human contribution of these landscapes to Irish archae- incursion into Ireland during the final Palaeolithic. ology and, finally, comment on potential avenues Even if the possibility of Palaeolithic occupa- of future investigation. The timeframe under con- tion, or at least pioneer incursion(s), cannot be sideration focuses on the Irish Mesolithic and the ruled out, the earliest incontrovertible evidence early Neolithic (c. 10,500–5500 cal BP) because, of full-scale human colonisation and settlement as demonstrated below, most submerged land- comes during the Mesolithic. This is the site of scape evidence from Ireland likely dates to this Mount Sandel, an occupation site situated close interval. to the north coast and dated to c. 9800 cal BP Before proceeding to the evidence, a brief (Bayliss and Woodman 2009; Woodman 2015). review of the archaeological, sea-level and pal- By comparison, Britain has Late Glacial hunter- aeogeographic background is needed. This will gatherers present throughout the GI-1 interstadial not only set the scene for readers unfamiliar with (c. 14,600–12,900 cal BP; Pettitt and White Irish prehistory but also identify issues in the 2012), and even after near-total depopulation in extant evidence base which, as will become the first half of the cold GS-1/Younger Dryas sta- apparent later, are directly relevant to the study of dial (c. 12,900–11,500 cal BP), humans had submerged prehistoric landscapes. More detailed returned from c. 12,000 cal BP onwards in the recent overviews of the Irish Mesolithic can be form of Epi-Ahrensburgian hunters (Pettitt and found in Woodman (2015) and Warren (2017), White 2012; Warren 2017). while a similarly recent overview covering sub- The reasons for the missing Irish Palaeolithic merged Quaternary palaeolandscapes and tapho- remain unclear. For the pre-Last Glacial nomic issues can be found in Westley and Maximum (LGM), this could be partly explained Edwards (2017). by the scouring effect of multiple glaciations and the fact that Ireland’s location may have been too peripheral for anything other than sporadic pio- 11.2 Archaeological Background neer expeditions. However, for the Late Glacial and Mesolithic, there is still a delayed occupation Ireland’s prehistoric record bears some similari- relative to Britain which cannot be explained by ties with neighbouring Britain but also some purely taphonomic factors and could relate to major differences. One key distinction is the lack social and ecological factors, not least the of Palaeolithic evidence compared to Britain, ‘impoverished’ nature of the Irish fauna (see which was occupied as far back as c. 0.78– Woodman 2015 and Warren 2017 for a fuller 0.99 million years ago and is evidenced by well- account of these issues). preserved in situ sites as well as extensive Ireland’s earliest occupants likely arrived secondary context assemblages (Pettitt and White from Britain, bringing with them similar toolkits 2012; Bailey et al., Chap. 10, this volume). The and strategies. In Britain, the early Mesolithic (c. Irish evidence so far consists of four isolated 11,500 to 10–10,500 cal BP) is typified by forms lithic finds (two handaxes and two flakes) of such as obliquely blunted points, broad blade Lower to Middle Palaeolithic appearance. microliths and slender uniserial bone/antler However, the provenance of the handaxes is barbed points. By c. 10,500–10,000 cal BP, these uncertain while the flakes are derived from assemblages gave way to different forms, charac- reworked contexts (Woodman 2015; Warren teristically narrow blade or geometric microliths 2017). Most recently, a cut-marked bear patella and flat, squat barbed bone/antler points (the later from Alice and Gwendoline Cave (Co. Clare) Mesolithic; Tolan-Smith 2009; Warren 2017). was radiocarbon dated to 12,810–12,590 cal BP The first thousand years of the Irish Mesolithic along with a second cut-marked bear vertebra (c. 9800–8800 cal BP) show some similarities dated to 11,080–10,400 cal BP (Dowd and with British later Mesolithic assemblages (as 11 Ireland: Submerged Prehistoric Sites and Landscapes 223 expected given the colonists’ likely origins), 11.3 Sea-Level most evidently in the form of geometric micro- and Palaeogeographic liths, but also differences, such as distinctive Change small core and flake axes not found in contempo- rary British sites. Once established in Ireland, the Like much of north-west Europe, Late Pleistocene trajectory of the Irish Mesolithic departs further. and Holocene relative sea-level (RSL) change From c. 8800 to 8600 cal BP, tool assemblages around Ireland was driven by the interplay become dominated by a ‘macrolithic’ industry of between global glacio-eustatic ocean volume large blades, flakes and distinctive butt-trimmed change and local- to regional-scale isostatic forms which have no British counterpart, except response to ice loading during the LGM and on the Isle of Man. Bone artefacts also differ as unloading thereafter. The resulting impact on the barbed points of the British Mesolithic are not RSL change was particularly complex as Ireland present in Ireland; instead, slender points without is located in a transitional zone between uplifting barbs were used (Woodman 2012, 2015). formerly glaciated and collapsing forebulge areas Effectively, this means that the British Late (Edwards and Craven 2017; Westley and Edwards Mesolithic can be regarded as equivalent to the 2017). Irish Earlier Mesolithic, while the Irish Later At its maximum extent (c. 27,000–23,000 cal Mesolithic has no parallels in mainland Britain BP), ice entirely covered Ireland and its conti- (Warren 2015a). For the rest of this chapter, nental shelf as far west and south as the Atlantic unless otherwise stated, the terms ‘Earlier’ and shelf break and Celtic Sea shelf edge, respec- ‘Later Mesolithic’ will be used by reference to tively. Thickness, duration and timing of maxi- the Irish typochronology. mum extent of ice cover all varied spatially As elsewhere in north-west Europe, the (Clark et al. 2012; Ballantyne and O Cofaigh Mesolithic was characterised by mobile popula- 2017). Where ice was thickest and longest last- tions hunting and gathering terrestrial, lacus- ing, isostatic depression and subsequent postgla- trine, riverine and, to some extent, marine cial rebound were greatest. This is exemplified by resources (note, however, questions regarding Ireland’s north-east corner and was further the level of coastal adaptation which will be dis- enhanced by proximity to the regional centre of cussed later). This way of life was then replaced ice loading in Fenno-Scandinavia. Moving south by the Neolithic with its new agricultural prac- and south-west across the island, isostatic tices, burial traditions, built structures and mate- rebound diminished with distance from the cen- rial culture. Dating of the key transition interval tre of ice loading (Westley and Edwards 2017). has been refined to 5700–5570 cal BP and com- This resulted in RSL histories which vary spa- prised a 40–100 year ‘boom’ in Neolithic settle- tially and are often non-monotonic, i.e., charac- ment evidenced by the rapid spread of distinct terised by variations in the rate of RSL change rectangular houses, appearance of domestic and, in some places, complete oscillations.