Long-term archaeological perspectives on new genomic and environmental evidence from early medieval Ireland

https://doi.org/10.1016/j.jas.2019.04.001Get rights and content

Highlights

  • A database containing over 8000 radiocarbon dates of human activity in Ireland has been assembled.

  • Contextualizing the data for the period 400 to 1200 CE reveals a pronounced oscillation in activity throughout the landscape.

  • A long-term population model has been developed for Ireland using these radiocarbon data and historical records.

  • We suggest haplotype admixture in Ireland took place in the context of long term population decline.

  • These results mirror recent palaeoisotope studies of intensification.

Abstract

Using archaeological data, this paper investigates past population trends in Ireland as a response to recent genomic studies that have identified admixture signals in the genomes of Irish people caused by historically-recorded migration events. Among these was Norse settlement in the 9th-10th Centuries CE, which has a greater than expected signal in the contemporary population of the island. Here, we contextualise these discoveries using a large database of recently discovered archaeological sites with radiocarbon dates that we have analysed using Kernel Density Estimation techniques. We argue that the Viking migrations occurred following a 300-year period of population decrease in Ireland. This new, data-driven synthesis of the archaeological record contrasts with previous accounts of early medieval Ireland as a period of ever-growing expansion and progression. However, this new interpretation is also aligned to evidence for economic and environmental change, including recent discoveries concerning the soil nitrogen cycle and agricultural intensification. We compare historical evidence for Viking migrations to later episodes of migration between Britain and Ireland, where more details are known about the size of the incoming groups, ultimately wishing to confront the opinion that past population sizes cannot be fathomed for cultures without documentary records. Through comparison with historic analyses and census records, we make broad estimates of absolute population size in Ireland since prehistoric times, including during these demographic events, and argue that much value is added to genomic evidence for migration when these points in time are contextualised in terms of evolving population trends.

Introduction

Human population dynamics is an important window into the mechanisms underpinning societies, informing us about yet more fundamental processes including environmental pressure, evolution, epidemiology and the availability of resources. Human population size in particular has had a major influence on ecosystem function today and at least since the Mid-Holocene (e.g. Ruddiman, 2003). Until recently, the long-term history of human population was guesswork as historic records only cover recent centuries. However, independent work in ancient DNA, modern genomics, paleoecology, and data-driven archaeology are together beginning to address this gap in our understanding.

To this end, this paper forms an archaeologists’ response to the recently published studies of the genomic structure of modern Irish and British populations (Byrne et al., 2018; Gilbert et al., 2017), focusing on how archaeometric and genomic data can together advance our understanding of the dynamics of the early medieval population in Ireland (400–1200 CE). We also discuss recently-published evidence that the archaeological signals of population change are present in the Holocene nitrogen cycle in Ireland, and look before and after the early medieval period, contextualising these trends over the longer-term.

Our particular focus is on Ireland (Fig. 1), uniquely suited to this study because the growth of the ‘Celtic Tiger’ economy in the 1990s and 2000s led to an extraordinary number of archaeological rescue excavations of quasi-random samples of the landscape, unbiased by the pre-existing research interests of archaeologists (Armit et al., 2014; McLaughlin et al., 2016). The basic premise underlying the use of archaeology in this way is that larger human populations tend to generate more detectible archaeological signals than smaller ones, and frequency of archaeological materials can be used to model population trends if certain conditions are met and the data are treated appropriately (Edinborough et al., 2017; Timpson et al., 2014).

Genomic studies of population structure are made possible by haplotypes (blocks of genetic information) that are shared within and across populations in patterns that uniquely segregate with discrete localities, or, by proxy, certain points in time. New approaches differ from previous attempts to come to terms with Ireland's genetic landscape because they make use of new methods that analyse the entire genome of living people, and can predict geographical affinity using genetic data alone, implying that local genetic patterns have persisted throughout many generations. It seems that the main effect of the limited scale of migration and exogamy in human history is that people from nearby regions tend to be more closely related than people from further away. In Britain and Ireland, this is exemplified by the kinship shared across the shortest sea crossing between northeast Ireland and southwest Scotland, and the genetic similarities between English people and those of Ireland's eastern coast. Genomic studies (Byrne et al., 2018; Gilbert et al., 2017) have been able to estimate when episodes of migration occurred by modelling the decay of haplotypes shared between modern Irish individuals and modern proxies for migrating individuals. The independent insight from modern genomics presents stimulating archaeological implications, and an opportunity to tackle tricky problems of migration and absolute population in past societies through collaborative dialogue.

These genomic signals offer a tantalising glimpse at a rather murky period in Ireland's past. Through admixture modelling, the origins of genomic affinities between Irish and Scandinavian populations can be dated to during the Viking Age, 900–1200 CE (Byrne et al., 2018; Gilbert et al., 2017; Leslie et al., 2015). Former estimates of Viking ancestry in Ireland based on Y-chromosome haplotypes detected little presence of Scandinavian-type haplotypes (McEvoy et al., 2006). For a time, and bolstered by the inadequately-powered Y-chromosome haplotype data, narratives of acculturation between native and small numbers of incoming Scandinavian immigrants were popular (e.g. Knudson et al., 2012), but these models must now be revised. The size of the contemporary native population is a key parameter which must be investigated in order to understand the causes, scales and effects of the migrations. For this we turn to an analysis of the archaeological record.

Section snippets

Methods

We built a database of radiocarbon data derived from archaeological excavations in Ireland (n = 8805 dates, see supplementary data) drawing upon other published sources (Bevan et al., 2017; Chapple, 2015; McLaughlin et al., 2018; McLaughlin, 2018). Each entry in the database consisted of a radiocarbon age determination (14C yrs ± standard deviation), and geographic attributes. For early medieval samples (locations shown in Fig. 1), the database also included categorical variables describing the

Archaeological proxy evidence for population trends

The density models (Fig. 2) reveal widespread decline spanning 700 and 1100 CE, the centuries before and during the Viking migrations. There are some regional variations, but it seems the decline occurred more-or-less simultaneously throughout the island. It also manifested in every archaeological context we have studied. This is an important point because some contend that radiocarbon date densities are influenced by changes in human behaviour and research bias (Torfing, 2015). Research bias

Conclusions

Existing literature on early medieval Ireland makes clear that the island contains a bountiful archaeological landscape, with traces of a distant but familiar past occurring in virtually every neighbourhood. Tens of thousands of early medieval sites can still be traced in the Irish countryside (Fig. 1) in the form of stone buildings, earthworks and newly-discovered settlement and industrial sites found across the entire island (Stout, 2017). This points to a large rural early medieval

Data, code and materials

The supplementary datasets contain all the 14C data and meta-data needed to calculate the density models used in this analysis. A full bibliographic index to these data is beyond the scope of this paper but readers can cross-reference the 14C laboratory codes with the following sources:

Chapple, R. M. (2015) Irish Radiocarbon and Dendrochronological Dates, https://sites.google.com/site/chapplearchaeology/irish-radiocarbon-dendrochronological-dates.

Archaeological Site Index to Radiocarbon Dates

Funding

EH is supported by the Arts Humanities Research Council (United Kingdom) Northern Bridge Doctoral Training Partnership.

Acknowledgements

Maps were drawn using NaturalEarth (coastline) and SRTM (elevation) data; the locations of settlement sites were provided by the public Sites and Monuments Records of the Republic of Ireland and Northern Ireland. We thank the very many archaeologists – virtually all those working in Ireland – whose work we summarise here and the efforts of the Council for British Archaeology, Heritage Council (Ireland), Transport Infrastructure Ireland, Mr Robert Chapple, Ms Isabel Bennett, the National

References (41)

  • L.M. Cassidy et al.

    Neolithic and Bronze Age migration to Ireland and establishment of the insular Atlantic genome

    Proc. Natl. Acad. Sci. Unit. States Am.

    (2016)
  • R.M. Chapple

    Catalogue of Radiocarbon Determinations & Dendrochronology Dates (December 2015 Release)

    (2015)
  • L. Coyle-McClung

    A Palynological Investigation of Land-Use Patterns in First Millennium AD Ireland

    (2012)
  • L. Coyle-McClung

    The late Iron age lull – not so late Iron age after all!

    Emania

    (2013)
  • L.M. Cullen

    Population trends in seventeenth-century Ireland

    Econ. Soc. Rev.

    (1974-5)
  • A. Dooley

    The plague and its consequences in Ireland

  • K. Edinborough et al.

    Radiocarbon test of demographic events in written and oral history

    Proc. Natl. Acad. Sci. Unit. States Am.

    (2017)
  • E. FitzPatrick

    Native enclosed settlement and the problem of the Irish 'Ring-fort

    Mediev. Archaeol.

    (2013)
  • E. Gilbert et al.

    The Irish DNA atlas: revealing fine-scale population structure and history within Ireland

    Sci. Rep.

    (2017)
  • J. Goldstone

    Revolution and Rebellion in the Early Modern World

    (1991)
  • View full text