The application of sedimentary ancient DNA analysis to archaeological sediments for the reconstruction of palaeoenvironments

Abstract

Sedimentary ancient DNA (hereafter sedaDNA) analysis is now a common method employed for palaeoenvironmental reconstruction in studies of paleoecology. It is most often applied as a palaeoecological technique when analysing lake sediments, marine sediments or permafrost in Arctic or Sub-Arctic locations, as these have proved favourable contexts for DNA preservation and stratigraphic stability. However, sedaDNA analysis is also beginning to be more widely applied to fluvial and terrestrial environments in temperate locations, which increase the opportunity for application of the technique to archaeological sites and sediments. If such applications can be shown to be successful, then sedaDNA analysis will be a powerful tool for understanding the archaeological record.
This thesis focuses on the use of sedaDNA metabarcoding, and in one instance shotgun sequencing, to better understand the palaeoenvironments at significant UK Prehistoric archaeological sites. The three sites are: Blick Mead, Wiltshire, Killerby, North Yorkshire, and Seven Springs, Martlesham, Suffolk. Applications of sedaDNA to terrestrial contexts are not without their specific challenges and limitations, which will be discussed in detail later in this text, the method is therefore applied in conjunction with more traditional proxies such as plant macrofossils, mollusca, coleoptera, and pollen. These methods complement and, in some cases, overlap with the genetic evidence to provide a more complete assessment of biodiversity at any given site. As such, at each location chosen for study, sedaDNA was part of a multiproxy approach.
In the three papers presented in this thesis, sedaDNA was shown to be preserved fairly well in temperate lacustrine, fluvial and terrestrial sediments and corroborated much of the existing archaeological faunal and lithic evidence found at each site. At all three sites, the molecular evidence increased the number of taxa than would have otherwise been found using only pollen and/or macrofossil data. SedaDNA improved reconstructions of floristic diversity, and sedaDNA was able to detect plant taxa that are frequently missing or underrepresented in other environmental and faunal records, such as aquatic plants and insect pollinated trees. This allowed a re-evaluation of archaeological hypotheses at each site, leading to insights surrounding site selection strategies and the exploitation of local floral and faunal resources.
The findings from this thesis also highlighted the importance of careful site selection when sampling for sedaDNA in terrestrial environments. At both Blick Mead and Killerby, the sampled horizons were largely protected from post-depositional mixing of sediment. At Martlesham, dating provided evidence of a mixed context within the peat, but the addition of shotgun sequencing and map-damage assessment confirmed that the sedaDNA found was ancient DNA. Furthermore, the preservation of DNA in peatland was found to be variable, with limited species representation at Killerby but a rich community seen from the peatland at Martlesham. To detect such problematic factors, the papers from this thesis showed that sedaDNA studies in terrestrial environments required a thorough chronological model and understanding of stratigraphic depositional environments.
At the two sites from Southern England, Martlesham and Blick Mead, sedaDNA analysis suggested the presence of rich floodplain carr-woodland and meadow environments. At Blick Mead, this fed into wider questions of Mesolithic-Neolithic transitional vegetation disturbance. Whilst the large plant assemblage found at Martlesham of over 60 taxa revealed the species richness of an Early Neolithic environment in SE England, and the data suggested use of the area for forest farming. The site from North Yorkshire, Killerby, showed evidence of a large post-glacial lake that underwent hydroseral succession in the beginning of the Holocene. The later Cladium-sedge fen that developed provided the environment for Mesolithic archaeological activity found at the site, either via timber constructions or the exploitation of sedges for fuel and consumption.
Overall, the sites mentioned in this project show the potential of this novel method of environmental reconstruction for use of environmental archaeology. The findings of the three papers improved understanding of Lateglacial vegetational change, Late Mesolithic forest coverage and Neolithic farming practices, and the project as a whole outlined where further study using sedaDNA could benefit outstanding research questions. However, the methodological problems identified, such as the overrepresentation of local species, poor representation of mammal taxa, variable preservation in different sedimental contexts and possibility of contamination, mean that sole applications of sedaDNA are not yet recommended at archaeological sites for best practice. Instead, sedaDNA analysis, in particular metabarcode data, is best used as an accessible and powerful complementary environmental proxy for application in an increasing variety of archaeological sedimentary contexts.

More information

Identifiers

ORCID for Samuel Michael Hudson: orcid.org/0000-0001-8914-0998

ORCID for Paul Hughes: orcid.org/0000-0002-8447-382X

ORCID for Antony Brown: orcid.org/0000-0002-1990-4654

Catalogue record

Export record