Tuesday, March 6, 2012

Bubbling over: archaeological lipid analysis and the Irish Neolithic: Review

[** If you like this post, please make a donation to the IR&DD project using the button at the end. If you think the review is useful, please re-share via Facebook, Google+, Twitter etc. **]

I recently attended one of the PCC Lunchtime Seminar Series talks (Booms and Busts in Europe’s Earliest Farming Societies given by Prof. Stephen Shennan) at the School of Geography, Archaeology and Palaeoecology at QUB. It was a fascinating talk that I enjoyed very much and I decided that I would make every effort to attend the next one. Yesterday (6th March 2012) we had Dr. Jessica Smyth, currently of the Organic Geochemistry Unit, in the School of Chemistry, at the University of Bristol, speaking on the topic of lipid analysis and their application to the study of the Irish Neolithic.

She began by introducing the topic of lipid analysis, explaining how it involves the investigation of surface and embedded fats to reconstruct past diets. This field of research was largely pioneered by Prof. Richard Evershed and Smyth explained how she is working closely with him on this post-doctoral research project. She explained that one of the thrusts of her PhD research (Neolithic settlement in Ireland: new theories and approaches, completed at UCD) was to provide a counterpoint to the traditional narrative of the Irish Neolithic. In the past much of the debate had centred on the megalithic tombs and the relatively small number of excavated houses. However, by the early to mid 2000s, much new and challenging evidence was being uncovered as part of the large infrastructural projects and the topic was ripe for re-evaluation (See my take on the benefits of the ‘Celtic Tiger’ years here). One of the bounties of the ‘Celtic Tiger’ period was the vast increase in recovered ceramics, many of which may have been suitable for lipid analysis. Following on from her PhD, Smyth applied to the Marie Curie Actions foundation and was awarded a Marie Curie Intra-European Fellowship for Career Development for lipid analysis for the ‘SCHERD’ project. ‘SCHERD’ stands for “Study of Cuisine and animal Husbandry among Early farmers via Residue analysis and radiocarbon Dating” and, as these things go, is a pretty snappy title for the project and gets all the main points across.

The first question that Smyth posed was ‘why is it important?’ She emphasised that the Neolithic is a significant stage in human development with the introduction of megalithic architecture, farming, and generally increased social complexity. Lying at the edge of Europe, Ireland is also chronologically and geographically at the end of this process, which makes an even more interesting research proposition. She explained how her research interests partially overlapped with those of the LeCHE group, a confederation of researchers examining the origin of dairying in Neolithic Europe through the use of lipid analysis, DNA etc. However, their research focuses on central and south-eastern Europe, leaving Ireland previously unstudied.

The aims of the SCHERD project are to firstly identify the contents of the pots via lipid analysis. This is particularly important in the case of Ireland where generally poor preservation of animal bone has constricted our ability to understand some of the developments and mechanisms of Neolithic agriculture. In this way the analysis of lipids may be developed for use as a proxy for the economy of the past. It is also hoped to distinguish patterns of regional variation that may shed light on differing diets, economy and traditions.  It is also hoped that, similar to the research carried out by Mukherjee et al. (2008) for England, that diachronic changes may be discerned, such as a change in emphasis on cattle rearing to pig production. Another aim of the project is to provide more secure date markers for the Irish Neolithic by directly dating surface and absorbed residues from selected vessels in key assemblies. In this way, it is hoped to add to or, assist in the revision of, Bayesian chronologies for the Neolithic, a process begun with the volume Gathering Time: Dating the Early Neolithic Enclosures of southern Britain and Ireland. Smyth also highlighted the problem that much of our chronology of the Irish Neolithic is relatively poor as it is based on typological assessments, chiefly of changing pottery styles. For all of these reasons, the chronology requires urgent revision and the SCHERD project hopes to play a prominent role in that process.

Smyth outlined her sampling regime of 15 key sites, spread across the entirety of the Irish Neolithic. In the Early Neolithic her sites include two enclosures (Donegore Hill, Co. Antrim, and Magheraboy, Co. Sligo) and five houses (including Ballygalley, Co. Antrim, and Monanny, Co. Monaghan). The sites from the Middle Neolithic included one enclosure (Tullahedy, Co. Tipperary), two pits (including Goodland, Co. Antrim), and one pit/spread. The late Neolithic sites examined by Smyth include one enclosure (Ballynahatty, Co. Down), one spread (Longstone, Co. Kildare), one house (Ballynacarriga, Co. Cork), and one pit (Lowpark, Co. Mayo). To ensure that a statistically viable sample was investigated, where possible, a minimum of 30 sherds were examined from each site. As Smyth points out, the lipid analysis process is destructive. However, she argues that considering the potential rewards in terms of new data and insights, the removal of 2-3g of pottery per analysed sherd is a pretty minimal price to pay.

The process of preparing the sherd is as follows: the surface is cleaned with a modelling drill to remove all forms of surface contamination. To release the lipids from the clay matrix of the pottery, the sample is ground into a fine powder. The resulting dust is placed in furnaced glass phials and solvents are added to release the lipids. The solvents used are most usually of the form of chloroform/methanol compounds. After some time the resulting complex mixture is purified and separated. The resulting Total Lipid Extract (TLE) amounts to approximately 2ml. This is kept in refrigerated storage and all analyses are conducted using small portions of this core sample. The TLE is further treated to make it suitable for use in various tests, including use in a mass spectrometer or a gas chromatograph. Only once all of these steps have been completed can analysis actually begin.

Smyth explained that one of the research aims pursued by Prof. Evershed was that extensive experimental work was necessary to provide a library of lipid signatures that would allow researchers to confidentially infer the former presence of fats. She showed a number of chromatograms showing Triacylglycerol signatures (TAGs). She also explained how, over time, these degrade into Diacylglycerols (DAGs) and then into Monoacylglycerols. Finally, these will degrade into simpler, free fatty acids. Lipids recovered from archaeological sherds are most usually of this latter type. It is estimated that of the original volume of deposited lipids, only 1% is likely to survive. Some of the sites investigated show excellent rates of survival (such as Tullahedy, Co. Tipperary), though survival rates are thought to be related to variables such as climate and temperature etc. As an example, Smyth cited research work carried out in dry countries, such as Greece. Here, less than 10% of sherds retained lipid traces. Experimentation has shown that there is vast and rapid degradation of the lipid remains on sherds stored in aerobic conditions over a 40 day period. It is for this reason that the sherds selected for this form of analysis should come from sealed, anaerobic contexts. Smyth also noted that the storage of sherds in plastic bags (extremely common these days) may also leave a lipid signature on the pottery. Her concern is that while it is not an insurmountable problem, it does need to be noted and it may also mask genuine archaeological signals.

In examining what makes a high quality sample, Smyth noted that upper body sherds and rim sherds have shown the best level of survival. At face value, this seems counter intuitive as this is the region in least contact with any boiling liquids in the pottery vessel. As noted above, sherds from sealed, anaerobic contexts are also preferred over finds from the plough soil etc. It is also important that finds are well archived and are retained with their meta data, indicating their contact with plastics and other forms of post-excavation processing. While, from a chemical point of view, the need for sherds from well stratified deposits is not essential, it is of the highest archaeological importance so that the fullest data may be relayed back into our site models. Finally, it is always good to have a complementary faunal record or other proxy data against which to compare results.

Addressing the question of the feasibility of this project, Smyth noted that earlier work on the Tullahedy, Co. Tipperary, assemblage reported a rate of five out of six analysed sherds retaining lipids. At another Middle Neolithic site on Rathlin Island, Co. Antrim, 50 sherds were analysed. This showed a recovery rate of 28 of the 50 analysed sherds containing lipid residues. Smyth was also keen to stress that, with 14 months left to run, she is still at a relatively early stage in her project and the results she has thus far obtained are preliminary in the extreme. With regard to the site at Donegore Hill, Co. Antrim, the recently published monograph and the Gathering Time volumes have provided a secure dating range. Given the enormous quantity of recovered ceramics (c. 45,000 sherds), only the rim sherds were formally analysed. These were estimated to represent approximately 1,500 individual vessels. Unfortunately these sherds were unavailable for lipid analysis. As none of the body sherds were identified to individual vessels, a selection of fabric types was selected, spread across the entirety of the site. At the Neolithic house site at Ballygalley, Co. Antrim, the upper body and rim sherds were available for analysis, though the results of these analyses are yet to be completed. The enclosed Middle Neolithic settlement at Tullahedy, Co. Tipperary, produced a very rich cereal assemblage – in fact it is the most abundant assemblage of its type throughout the entire Irish Neolithic. However, the site had remarkably poor bone preservation. Earlier investigations of some sherds from Tullahedy confirmed the presence of free fatty acids on a high proportion of the examined sherds. While early results are promising, she was quick to stress that further analysis, including isotope ratio analysis, would be necessary to confirm them.

Looking towards the future of the SCHERD project, Smyth hopes to document the changing nature of Irish agriculture over time. Evidence from such sites as the Céide Fields, Co. Mayo, and Kilshane, Co. Dublin - the latter with its exceptionally large Late Neolithic assemblage of animal bone – have been taken to suggest that there was a move away from a cereal-based economy to a stock rearing economy over the course of the Neolithic. Smyth believes that with enough good quality samples, and sufficient time on the various machines, the SCHERD project can make significant contributions to the state of our knowledge.

In the question and answer session at the end of the presentation, Smyth was asked if it was possible to date lipids directly, using radiocarbon dating. She replied that it did appear to be possible, though they had not attempted it so far. For example, some of the sherds, especially those from Donegore Hill, had produced over 5mg of lipid residue, more than enough for an AMS date, so the potential is definitely there. Smyth was also asked if it was possible to distinguish between sheep and cattle in the lipid residues. She explained that it was, though the analysis of subtle differences in the signatures of Triacylglycerol (TAGs) and Diacylglycerols (DAGs). One other contribution related to the description by Peacock, during an 18th century tour of the Hebrides. He had witnessed the native islanders creating coil-built pottery vessels and firing them in open hearths – essentially the same manner as in the prehistoric past. However, Peacock noted that before they were placed in the hearths, they were filled with milk, which was allowed to evaporate during the firing process. The question then was that could this form of activity be what we are witnessing, rather than the continued use of the vessel to hold a variety of different cooked meals. Smyth admitted that it could complicate matters, but that it was most likely that the firing process would not just evaporate the milk, but destroy any lipid residues in the process. One further intriguing problem was raised during this discussion. This was the discovery of certain keytones in some samples that indicate that the contents were fired to a temperature over 300°. While it is thought that the average open heath would struggle to achieve this temperature, the reality is that little food would survive for long even if it could be achieved.

As Smyth stressed throughout this presentation, she is currently only able to present preliminary results. Nonetheless, she has already produced some interesting, and potentially controversial, findings. I, for one, eagerly await her return trip to Belfast when she has been able to fully verify her findings and process more samples.

Notes:
I hope that I have done justice to Jessica Smyth’s lecture and managed to convey the gist of her ideas and results. Nonetheless, I do sincerely apologise if I have misrepresented or misquoted the speaker. If so, please feel free to contact me, and I will endeavour to set the record straight.

I have not been able to give the full references to any of the literature referenced in the lecture – sorry.

Some photographs of the 1984 season of excavations at Donegore Hill are available for public viewing, at The William Dunlop Archaeological Photographic Archive.


[** If you like this post, please consider making a small donation. Each donation helps keep the Irish Radiocarbon & Dendrochronological Dates project going! **]

If you’re going to do any shopping through Amazon, please go via the portal below. It costs you nothing, but it will generate a little bit of advertising revenue for this site!