Though isotope, proteomic, and aDNA analyses all, to various degrees, are destructive, some sampling approaches can however minimise object damage. Depending on the analytical approach, certain sampling methods can leave objects seemingly intact, retaining their shape, mass and molecular integrity, with minor surface modifications only visible via high resolution imagery. Following paragraphs will outline some of the ‘quasi non-destructive’ sampling approaches along with their respective advantages and limitations – still, it is recommended to discuss these sampling methods with professionals before conducting any sampling. This section is merely meant to point out minimally invasive sampling methods that could potentially pave the way for the analysis of objects not suitable for destructive sampling.

The eraser extraction method

The eraser extraction method (EEM) utilises an PVC eraser, a common component of any office stationery, which is rubbed against the surface of a chosen archaeological object. The electrostatic charge created by the rubbing extracts the protein from the object into the eraser crumble, which is collected afterwards for further analysis. This method is somewhat commonly used within proteomics, especially for ZooMS. The EEM ensures bone taphonomy and other features are intact at a macroscopic level, although on a microscopic level EEM can leave traces similar to use-wear. However, the EEM approach has proven to generate low-quality MALDI-TOF MS spectra from palaeolithic bone specimens as well as archaeological Iroquoian bone artefacts. The eraser method has also been conducted on 7–80-year-old herbarium specimens for DNA extraction, which demonstrated that a non-destructive procedure is possible within the field of DNA. Concerning ancient bone material, it is unclear which purposes and what material the EEM would be suitable for in DNA studies.

The polishing film method

Polishing film with fine grit sizes (14,000-600) is a novel method, which in a sense is like the EEM. Polishing film is rubbed in circles against the hard tissue surface, extracting bone material which is fastened to the polishing film. The film is then retrieved for analysis. Surface alterations are visible for the eye, though minimal. The method has for now been used within proteomics and could be a good sampling alternative for the sampling of bone artefacts or other high status hard tissue objects.

Membrane box

The membrane box relies on its two flexible polyurethane membranes, which by closing the box embraces the bone specimen and should statically make loose collagen strands attach themselves to the membranes either by opening and closing the box or leaving the box closed for a considerable time, leaving the artefact intact. The surface is then swabbed with ammonium bicarbonate and ready for gelatine extraction. However, the method has proven to be less successful than most sampling approaches. The yield of collagen seems to be reliant on the bone specimen’s time of contact with the membranes, though the general preservation of the bone as usual plays an important role. Therefore, timewise, the placement of the specimen in the membrane box can highly vary, from weeks to years. It is also recommended that the bone specimen should, prior to membrane box sampling, not have been stored with any other bone artefacts, which often is not the case or difficult to prove otherwise, due to cross contamination risks. Additionally, the membrane box itself is costly compared with other sampling equipment.

Bag approach

Similar to the membrane box approach, the bag approach relies on loose collagen strands attaching themselves from the surface of the bone to the inside of a plastic bag. The strands of bone residue are collected with ammonium bicarbonate for further analysis. Though the approach is less costly than the membrane box approach, it has similar disadvantages, such as no or poor taxonomic identifications.

Ammonium bicarbonate

For a range of bone artefacts it is possible to submerge the artefact partly or fully into ammonium bicarbonate (AmBic). The artefacts would, however, need to be of a minor size, for an easier submergence, and have no consolidation, varnish, and/or marker traces which could interfere with the spectrum. The artefacts would be submerged for approx. 24 hours in AmBic, whereafter the artefact can be left to dry and restored again, with no macroscopic modifications, though it is unclear if any changes occur on a molecular level. Note that bones of a porous state or less dense mineral structure will function as a sponge, and the amount of AmBic should therefore be increased.

This section will provide a few examples of possible applications of the methods presented in the flowchart and outlined in the Limitation chapter. Hopefully this will inspire future studies, and further usage and development of the presented methods.

Fig. 2, Overlapping areas of δ18O and 87Sr/86Sr isotopes signals matching three individuals (Laffoon et al. 2017).

δ18O and 87Sr/86Sr

Though δ18O and 87Sr/86Sr analyses come with a range of challenges, e.g. sufficient baselines reflecting the local and extra-local geology and biosphere, it has been widely applied on prehistoric organic material. The two isotopic analysis approaches can be used to get a better understanding of human migration patterns on a general and individual level. Again, it is important to stress that the analyses exclude areas of origin, and should be perceived as bottom-up methods, even though the geographical area of interest can be narrowed down, by combining δ18O and 87Sr/86Sr analyses. Fig. 2 is a good example of how to approach provenance analysing using δ18O and 87Sr/86Sr analyses. However, even when overlapping the matching areas of δ18O and 87Sr/86Sr values, a good understanding of the archaeological context is still required to correctly interpret the area where an individual grew up. Still, by successive sampling of e.g. a tooth it is possible to trace mobility patterns of an individual's first years of living to their teens. Though it might not be possible to point out the exact spot geographical placement, the results can be used to reflect on groups and individuals’ mobility.

Zooarchaeology by Mass Spectrometry (ZooMS)

ZooMS has been especially helpful in studying taxonomically identifying bone fragment assemblages and bone artefacts where diagnostic features are absent. Bone assemblages from archaeological sites can thus be compared with the morphologically identified bone assemblage, providing new insight into fauna composition of a given site. New species might prevail within the fragmentary assemblage, not found within the morphologically identified collection, adding yet another layer of complexity. Additionally, ZooMS has been successfully utilised as a screening tool for hominins in caves with a high degree of  bone fragmentation. As for bone artefacts, species identifications can provide stimulus to further discussion about past strategies involving the selection/opportunism of taxa for the manufacture of bone artefacts. Lastly, the method has proved important for the distinction of sheep and goat bones in husbandry studies for a better understanding of such practices in prehistory and historical times.

Stable isotope analysis (δ13C and δ15N)

δ15N and δ13C isotope analyses have been commonly used as an indicator of dietary tendencies. δ15N values explain the relative trophic level, while δ13C values provide information on the intake of C3 (highlands and lowlands) and C4 (lowlands and aquatic) plants. Combined these methods have been used to explore the diet of prehistoric people, e.g. the clear change in dietary practices within Scandinavian Mesolithic and Neolithic groups, changing from a marine and occasionally freshwater based diet to that of an almost exclusively terrestrial based diet. Interestingly, the methods have also been applied on Sus scrofa found near Ertebølle sites, late Scandinavian Mesolithic, showing enriched carbon and nitrogen values suggesting a diet of marine intake. This has been implied to indicate that the management of Sus scrofa developed independently in Scandinavia of contact with Neolithic societies.

Shotgun proteomics

Palaeoproteomics or ancient proteomics has a wide range of applications, encompassing phylogenetic reconstruction, taxonomic identification, sex determination, study of diseases, and the study of bacterial proteins, among others. Recently, endogenous ancient proteins have been extracted from the denisovan Xiahe mandible, more specific from the dentine, dated to be at least 160,000 years old. Phylogenetic reconstruction of the proteome determined the mandible to be closely related to the great apes, including that of humans, neanderthals, and denisovans, especially that of the latter. This would otherwise not be possible with aDNA analysis, as no DNA could be recovered. This is just one example out of many, setting high prospects for proteomic studies in the future.

As the flowchart above shows, biomolecular techniques are able to provide valuable information regarding a wide range of different research subjects. And, even though the increased integration of biomolecular techniques within archaeological research is sure to help answer many existing and new research questions, they have their limitations. In general, biomolecular research is expensive and requires destructive sampling and each technique has its own restrictions as well. In order to clarify which and how biomolecular techniques can be best used to investigate your research question, we will briefly outline the limitations of each method.

δ18O and 87Sr/86Sr

Both δ18O and 87Sr/86Sr are often used to investigate the geographic origin of an organism and both suffer from the same main limitation: specificity. Although it highly depends on local circumstances, an area with the same δ18O or 87Sr/86Sr values can be sizable and diverse. This is why δ18O and 87Sr/86Sr values are usually used to exclude particular areas as potential geographic origins, rather than identifying one or several locations as the geographic origin. 

ZooMS

The main limitation of ZooMS is the specificity of its taxonomic identifications. ZooMS identifications are usually limited to genus level, although in some cases ZooMS can be as specific as species or as general as family level. Fortunately, based on the archaeological context the ZooMS identification can often be further specified, but this is a matter of interpretation. A second limitation of ZooMS is that there are not yet collagen biomarkers available for every species. Although the dataset of publicly available collagen biomarkers is continuously expanding, there is still a bias towards mammals in general and European fauna specifically. Therefore it is good practice to check if there are collagen biomarkers available for the species you would expect to find. A regularly updated list of published collagen biomarkers from the University of York can be found here.

Stable isotope analysis (δ13C and δ15N)

A common interpretation of differences in δ13C values is whether an organism ate plants from an open (e.g. steppe) or closed (e.g. dense forest) landscape. However, just as with δ15N, it is vital to have knowledge of the δ13C values of the archaeologically relevant plant species. In order to correctly interpret δ13C data, it is necessary to understand which of potentially exploited plant species are C3 or C4 plants.

It is important to understand that the relationship between δ15N values and the organism’s place in the food chain is relative. An organism will have a higher δ15N level than the food it consumes, so in order to correctly identify its place in the food chain, the δ15N values of the ‘food’ must also be known. The δ15N of the primary producers in ecosystems can vary through time and location, so absolute δ15N values might not be comparable between samples from different periods and regions.

Shotgun proteomics

The main limitation of shotgun proteomics is the limit of our ability to extract protein from samples. The state of preservation, such as exceptionally old age, may prevent the extraction of proteins. Some substrates, like ceramic matrix, have also proven more challenging for recovery of proteins. In these and similar cases, where it is uncertain whether protein extraction will be successful, it may be sensible to first screen for the presence of proteins using a different technique. A limitation more specific to species identification is that proteomics are more restricted, compared to DNA, in the level of specificity they can provide. Shotgun proteomics can be more specific than ZooMS, but DNA analysis remains the most specific of all taxonomic identification methods.

aDNA analysis