Why, Where and How?: Investigating small mammals remains - Part 2

A word of introduction for Part 2

By Andrzej A. Romaniuk (PhD in archaeology, MSc in osteoarchaeology)

This blog post follows directly Why, Where and How?: Investigating small mammals remains PART 1. In the first part I explained a little bit about why and how we study micromammals, i.e. very small mammals like rodents or shrews. I also provided some explanations why the Orcadian environment is a surprisingly good environment to study those species, both currently and in the past. In this part, however, I will talk a little bit specifically about my own research on micromammal remains found at key Orkney archaeological sites, stressing the history of research, issues encountered, what the results are, and potential for later research.

As in Part 1, I would thank Dr Robin Bendrey (University of Edinburgh) and Dr Jeremy Herman (National Museums Scotland), whose ongoing support was crucial for the completion of this research. I would also like to thank Dr Gail Drinkall (Orkney Museum) for the interest in my work and support in obtaining necessary materials.

Figure 6 – Burnt Orkney vole mandible and partially burnt vole axis vertebra from Skara Brae. Burning can be differentiated from other sources of decolouration in a variety of ways, from as simple as visual analysis (burning includes firstly most exposed surfaces, including teeth; staining firstly affects the bone surface, often leaving teeth with their original colour) up to complex biochemical work (burning shows carbon oxidation on the surface with the loss of other elements, staining usually includes oxygen but with the presence of other elements, such as e.g. manganese). [Romaniuk et al. 2016, fig. 6]

How the research started?

My first encounter with Orcadian micromammals happened back in 2015. As a part of my Masters degree I decided to investigate micromammal remains retrieved from the famous Neolithic settlement of Skara Brae during Prof. David Clarke’s excavations back in the 70s. Results, published by the Royal Society in 2016 and later covered by the BBC, showed large assemblages in the site centre, commingled with human refuse and exhibiting signs of burning (Fig. 6). They also differed in structure to previously known natural accumulations by owls and other predators. Additionally, vole bones were found in the remains of several coprolites (fossilized faeces) of dog or human origin. I suggested, that observed patterns may be a result of an intentional accumulation of micromammals by humans, due to pest control and/or food processing and consumption. At that time I was very certain of the work I had done, but knew way more has to be done to be certain of these results.

Figure 7 – Location of all sites studied during the 2017-2021 PhD research, provided by the National Museums of Scotland (Edinburgh), Orkney Museum (Kirkwall) and Alder Archaeology (Perth). Settlement sites came from two key islands, Mainland, a centre of the archipelago, and Westray, a fringe island important for travellers sailing to Shetland and Faroe archipelagos. Each isle was represented by one site representing the Neolithic period (IV-II millennium BC, Skara Brae and Links of Northland) and one Norse and later Mediaeval period (VI – XV centuries AD, Birsay Bay and Tuquoy). Additionally, an iron-age broch of Bu (c. V century BC) and a Norse boat burial on Sanday (IX-XI century AD) were included to provide a contrast to the settlement sites. [Map by author]

How it developed into a PhD project?

Due to a favourable press reception and encouragement by staff at both the University of Edinburgh and National Museums Scotland, and later by Orkney Museum and Alder Archaeology, I continued my work on Orcadian micromammals as a part of my PhD. I firstly started by broadening the scope of my work to include new archaeological sites. I wanted to check whether the Skara Brae pattern repeats at other Neolithic sites, and how it differs from assemblages obtained from later sites (Fig. 7). I also started a side project on Skara Brae coprolites, aimed at providing more information about the provenance of those faeces and bone remains within them.

Figure 8 – Example of a difference between assemblages created by an owl, diurnal raptor and fox when micromammal skeletal elements are considered (data from Andrews 1990). With some caution, one can correlate studied micromammal remains with a specific predatory pattern. As it can be noticed, owls generally leave most complete specimens, especially skull bones, and a very small amount of loose teeth. In turn, foxes leave very small amount of skull remains but high amount of loose teeth, with most common finds being robust bones like femur or humerus, bones that tend to survive better the process of digestion. However, other factors can easily bias the results, putting them into question or completely invalidating them. It is especially common when other factors present in archaeology are considered (see Fig. 9) [Romaniuk 2022, Fig. 3.16].

Figure 9 – Visualisation of proportions between skeletal elements of four key anatomical areas (skull, vertebral column, front limbs and hind limbs), from expected outline for the whole specimen (upper left), two archaeological sites from different time periods (upper middle and right), and known stages of micromammal assemblage deterioration (lower row, data recalculated from Terry 2004). Large micromammal bone accumulations are often a result of predator activity, from owls to foxes and cats, with initial alterations (e.g. bone digestion, scattering of body parts, see Fig. 8) often unique to specific predatory species. However, once deposited, animal remains do not remain static, being a subject to a variety of factors. Those factors are varied (trampling, mechanical and chemical erosion, dispersal by the wind and water, gradual entry to a geological layer and eventual fossilization) and often affect the final outline of a studied assemblage. Does the difference between presented two sites shows exclusively the passage of time? Not really, as both sites may also represent different mode of deposition. In case of Skara Brae, between the closely built together roundish buildings, in case of Birsay Bay interior of the building. However, an additional impact of post-depositonal factors is highly likely. [Romaniuk 2022, Fig. 6.05 and 6.08].

How the PhD project changed over time?

However, with time I shifted towards a more methods-oriented approach. Some issues widespread in archaeological material are simply never encountered or considered by contemporary researchers. As there may be no references to a specific phenomenon, archaeologists themselves have to conduct experiments to establish their own references for identification, for example the impact of burning on bones under different conditions, like wildfire or intentional cooking. As I could not perform any on-field experimentation, I concentrated on working with statistics and statistical modelling to answer several specific questions, notably how representative micromammal remains are in comparison to modern references (Fig. 8), how factors common in archaeology affect the pattern of a deposition (Fig. 9) and how both affect identifiability of specific depositions, e.g. of predators. Sampling strategies employed during the excavations were also investigated (Fig. 10) as I realised some discrepancies between the sites may stem more from how they were retrieved rather than showcasing actual differences.

Figure 10 – Example of how the scale of sampling may affect the species composition of a single archaeological layer, given the percentage of the context covered. In a way, any archaeological dig is just a sample of the past, often very narrow in scope and biased towards specific finds (e.g. structural remains) at the expanse of the others. To further cut costs, what is excavated is sampled even further, which is often the case for harder to retrieve finds (including micromammals). That is why researching sampling and its impact on archaeology is crucial for understanding, whether what we see is actually a representation of the past population or rather a random selection of stuff that happened to be noticed and picked by the excavators. The issue may be further complicated due to sampling possibility to differently affect various types of data, e.g. species present and anatomical proportions, resulting in different thresholds for representativeness. In the example above we can clearly see, that any sampling not taking into account at least 40-50% of the original content may result at best in incorrect proportions, and at worst in a grave misinterpretation (below 24%). [Romaniuk 2022, Fig. 6.01 bottom].

What was the overall research conclusion?

The research generally confirmed the sequence of introductions known from previous studies but also showed a greater depth to this issue. While Orkney voles and field mice were introduced to Mainland roughly at the same time, around IV millennium BC, they definitely took different routes when colonizing other isles. Field mice, species more prone to scavenging near human habitation, arrived at Westray far later than voles, which prefer wild environments or fields of low vegetation. The Early Norse period saw both pygmy shrews and house mice populations already established through the archipelago. In contrast, no evidence for black rats, species widespread in European ports at that time, has been found. It may be due to those species being used to tropical vegetation and crowded harbours, not small fishing villages and unending fields and pastures.

Figure 11 – SEM image of an unerupted, still-developing first upper molar of a house mouse. In palaeoecology and archaeology any remains of juvenile micromammals are rare due to their general fragility, a point especially valid for still-forming teeth. This specific find survived deposition and later retrieval thanks to being completely encased by a bone in which it formed, which shattered during the inspection. Its presence is significant as it conclusively confirms house mouse nesting within human habitation in Birsay Bay, the site dated to Norse/Mediaeval times. Mice stay within the nest till the eruption of the first two out of three molars, and even at that point they start to roam next to the nest before attaining more adult features. This specific individual, to whom this molar belonged, died while still being cared for by its mother [Romaniuk 2022, Fig. 5.31].

Were there more specific results?

The rest of the results of my PhD are harder to convey, mostly because of their importance for the methodology of my discipline rather than Orcadian natural history. As expected, different types of deposits could be identified, from confirmed predators (owls, diurnal raptors and dogs) to evidence for accidental entrapment or natural mortality. Differences in age structure were found especially between Neolithic and Norse assemblages, with Birsay Bay showing predominance of young mice, perhaps all-year nesting (Fig. 11), and field mice from Skara Brae generally older, perhaps due to only seasonal infestation of human habitation. However, the location and age of an assemblage proved to be important during the assessment. Older accumulations, as well as open-space deposits, showed a higher degree of deterioration, not necessarily visible in general completeness of finds but more in the degree of fragmentation present as well as survivability of fragile bones (e.g. hand and feet bones or vertebra). Moreover, while some accumulations could come from a single predator or one specific event, there were cases with more than one predator responsible and/or more than one possible factor present. It was visible especially within the periods of active human habitation, with humans erecting structures driving accumulations, moving accumulations with other refuse or soil, and so on.

How a site-related results look like?

… But how does it actually relate to a specific archaeological site? The best example is how my PhD work reframed my understanding of Skara Brae. The evidence of multiple sources of deposition was indeed noted, with comparisons with other sites revealing a strong impact of deterioration and dispersal over time. Due to that, the comparisons with references I used had to be altered via statistical modelling. The adjusted methods revealed an off-site assemblage contemporary to the site coming most likely from an owl, with the majority of deposition within the site also likely coming from owls roosting or nesting on constructions. Considering plaster remains were found in the same depositions, the remains might have initially been deposited on a roof slope, falling down into comingled human refuse on its own or with parts of a deteriorating roof structure. However, dogs also deposited faeces with vole remains. It was to a minuscule degree, perhaps an accidental catch. The presence of burning was confirmed by checking element composition of affected bones surfaces, but considering the relative small amount of such finds it might have come to only occasional events, perhaps as accidental as with dog deposition.

Figure 12 – The right half of a rat mandible. Discerning closely related species using only visual clues can sometimes be impossible, especially if available material is only a singular bone or tooth. That is why archaeologists working with animal or plant remains sometimes need to use different methods, for example ones relying on biochemistry. Proteins, like collagen present in the bone, can differ in structure between specific species, and these differences can be noted e.g. by checking the weight of specific sections of the protein chain ( “peptides”) in the mass spectrometer (Buckley et al. 2009). In the case of black and brown rats there is a minor but easy-to-spot difference in one peptide, that is why shown mandible was successfully identified as coming from a brown rat by Dr David Orton and Dr Sam Presslee, from the University of York. [copyright National Museums Scotland]

Can those results be used in further research?

Optimally, conclusive research should provide ground for further studies or support other research in the field. Those studies can follow its initial success, further examining already gathered material, but even negative results can provide suitable grounds for worthwhile work. It proved to be the case for micromammal material found in a Viking boat burial at Scar, Sanday. Original excavations noted extensive signs of burrowing by a variety of species, notably otters, but I still had faint hopes I will find some micromammals contemporary to the boat burial itself. In the end, it turned out that the majority of finds relevant for my research were most likely deposited far later than boat burial itself. It was especially the case for the rat remains found, identified as brown rats (Fig. 12), introduced to Europe around the XVI century. However, burrowing itself was not necessarily recent. After the PhD was done one brown rat bone was radiocarbon dated to around the same time their introduction to the isles should have happened. If true, those finds could be a good DNA source for any research related to the dispersal of brown rats across Europe. Personally, I hope much more of such examples will become evident once my work is more known to the wider academic community.

Ending (for now)

It is all I can tell within a few short sentences. For those interested in further details, and not afraid of dull academic works, a recently defended PhD thesis “Rethinking Established Methodology In Micromammal Taphonomy: Archaeological Case Studies From Orkney, UK (4th millennium BC – 15th century AD)”, available online here: http://dx.doi.org/10.7488/era/1960. While I am not sure yet what the second part of year 2022 will hold for me, I do think about continuing my work on small Orcadian animals in one way or another. Perhaps I will try to utilize stable isotopes to get an insight into their diet, or DNA to check how it relates to modern specimens. I may also analyse their shape to see how they adapted to insular environments. Regardless, I think Orkney is the best place for such research for aspiring researchers, now and in the future.

Why, Where and How?: Investigating small mammals remains – Part 2