Thesis: The Implementation of SedaDNA Analysis in Roman Archaeology in the Netherlands

Written by: Mirjam Rijpma

In order to finish my master’s degree, I wrote a master thesis in cooperation with the Radboud university and Constructing the Limes (C-LIMES). During my thesis I researched the implementation of sedimentary DNA analysis in Roman archaeology in the Netherlands. During my thesis I was supervised by Astrid van Oyen (Radboud University) and Arjen de Groot (University of Wageningen).

At the end of last year I got interested in sedimentary DNA (sedaDNA) after hearing about it at one of the C-LIMES lectures at the Roman symposium. Later that year, when it became time to pick my thesis topic, the first thing that came to mind was sedaDNA. Because of my background in technical medicine I love the biological aspects of archaeology, and I’m always looking for fun technical challenges. This makes sedaDNA analysis a topic in which all my interest come together perfectly.

The thesis “The Implementation of SedaDNA Analysis in Roman Archaeology in the Netherlands” is a study focussed on exploring the potential use of sedaDNA analysis in Roman archaeology, and the steps needed to be taken in order to implement the technique in archaeological research in the Netherlands. The research started with an analysis of archaeobotany and archaeozoology in the Netherlands, in order to define the techniques, applications and limitations within the research sector. To explore these aspects, I held expert interviews with archaeobotanist Laura Kooistra and archaeozoologist Maaike de Groot. In the next chapter I researched which of the limitations found in bioarchaeology could potentially be resolved by implementing sedaDNA analysis. However, to use sedaDNA analysis as an addition to current research, the technique needs to be further developed. Therefore the next part of the thesis focussed solely on sedaDNA, for this part I also conducted an expert interview with geneticist Fabricio Furni. First I gave an overview of the whole process, from sampling to interpreting results. For every step written in the thesis, I explained in-depth which limitations presented itself. In the last part of the research I took all the current limitations that sedaDNA analysis faces, and presented potential solutions or further research opportunities, in order to set out the steps that need to be taken in the future to implement the technique in Roman archaeology in the Netherlands.

An important limitation of conventional bioarchaeological research techniques is the dependence on available floral or faunal material. Especially floral material is heavily influenced by its surrounding conditions such as the soil type or moisture levels. On the other hand, small faunal remains, like fish or bird bones, can be difficult to find on site. Because of this it is possible that species that were present on site in the Roman period are missed in the bioarchaeological analysis because there is no material left to collect.

To determine species that are hard to collect on site we can use sedaDNA. This form of DNA is present in soil, after plant remains compost, animal remains decay, or when animals secrete. So, since all species that live somewhere leave their traces in the soil, in theory it would be possible to make a more accurate reconstruction of biodiversity using DNA. Using this we can answer research questions about agriculture, husbandry, diet, mobility, and health in the Roman Netherlands.

Before sedaDNA is applicable in practical research there is still work to be done. Because the DNA is not only preserved in the soil, but is also about 2000 years old, the material is often degraded and has sometimes migrated through the soil layers. Additionally, there is a high risk of contamination during the sampling, processing, and analysis of the DNA. The combination of these factors can cause an underrepresentation of certain species in the analysis results even though they are present in the sample, or can result in species that are not actually present in the sample to show up in the analysis results. During the last step in the process, bioinformatics, some problems can occur with the taxonomic resolution of the sample. How well certain species can be detected based on their DNA is dependent on the database with reference DNA. Some species, like cows, that are useful for practices like husbandry and agriculture generally get more attention when constructing a reference database than animals like insects. Because of this, certain DNA fragments are more easily recognized when comparing it to the database, because there is more reference material available. Additionally, quantifying species in the sample is also difficult based on soil samples.

In order to make sedaDNA analysis easier to implement in Roman archaeology we have to learn to deal with these challenges. First of all, we need to further research the effect of different soil types on the preservation and migration of sedaDNA. This will help to understand the way in which DNA degradation develops, but will also provide context for the sedaDNA analysis. Contamination also has to be minimised by working under sterile protocols during sampling and processing of the samples. In order to improve the taxonomic resolution we could use a capturing technique. This technique targets specific genomes, which makes it easier to distinguish genomes out of a larger sample. This can be useful when the research question is centered on a specific species. Apart from this, we should also aim to further add to the reference database. Finally, in order to improve the general resolution of sedaDNA analysis we should account for false positive and false negative values in the dataset. This can be done by implementing a form of statistical modelling known as “site occupancy detection models”.

In terms of logistics there are a few challenges as well. At the moment sedaDNA analysis is still expensive and the process consumes a lot of energy. In the Netherlands there is only a handful of sedaDNA specialists in archaeology, so lab work often has to be done abroad. This costs time and energy. Apart from that, data processing on such large scale requires a lot of energy in general. This issue does not only apply to sedaDNA analysis specifically, but applies to all big datacentres. Nonetheless it is good to stay aware of this.

We still have a long road ahead of us when we want to apply sedaDNA analysis as a research technique in Roman archaeology in the Netherlands. When the limitations and solutions that are presented in this thesis are researched further sedaDNA analysis could form an addition to current research techniques in archaeobotany and archaeozoology. Together with archaeobotany and archaeozoology sedaDNA analysis could contribute to the reconstruction of biodiversity and landscape in the Roman Netherlands. This way we can answer research questions about agriculture, husbandry, diet, mobility, and health even more accurately.

Download and read her thesis here: MA-Thesis-Mirjam-Rijpma-15-0611