What was shipped in Greek amphoras? A reevaluation through DNA analysis

One of the big stories covered by archaeology blogs last month was the announcement that a team of researchers had determined the ancient content of Greek amphoras through the analysis of residual DNA.  News of this discovery appeared in this article in Science and this one in Nature, and both summarized a technical article now in press in Journal of Archaeological Science: B.P. Foley, M.C. Hansson, D.P. Kourkoumelis, T.A. Theodoulou, “Aspects of Ancient Greek Trade re-evaluated with amphora DNA evidence,” Journal of Archaeological Science (2011).

The abstract of the article by Foley et al. runs as follows:

Ancient DNA trapped in the matrices of ceramic transport jars from Mediterranean shipwrecks can reveal the goods traded in the earliest markets. Scholars generally assume that the amphora cargoes of 5th–3rd century B.C. Greek shipwrecks contained wine, or to a much lesser extent olive oil. Remnant DNA inside empty amphoras allows us to test that assumption. We show that short ∼100 nucleotides of ancient DNA can be isolated and analyzed from inside the empty jars from either small amounts of physical scrapings or material captured with non-destructive swabs. Our study material is previously inaccessible Classical/Hellenistic Greek shipwreck amphoras archived at the Ministry of Culture and Tourism Ephorate of Underwater Antiquities in Athens, Greece. Collected DNA samples reveal various combinations of olive, grape,Lamiaceae herbs (mint, rosemary, thyme, oregano, sage), juniper, and terebinth/mastic (genus Pistacia). General DNA targeting analyses also reveal the presence of pine (Pinus), and DNA from Fabaceae (Legume family); Zingiberaceae (Ginger family); and Juglandaceae (Walnut family). Our results demonstrate that amphoras were much more than wine containers. DNA shows that these transport jars contained a wide range of goods, bringing into question long-standing assumptions about amphora use in ancient Greece. Ancient DNA investigations open new research avenues, and will allow accurate reconstruction of ancient diet, medicinal compounds, value-added products, goods brought to market, and food preservation methods.

Although the Science and Nature pieces were useful, I wanted to have a look at the article myself.  The staff at Messiah College’s Murray Librarygenerously got me a copy (thanks!).

In the article, Foley et al. begin with the question of how do we know what was carried in those large ancient transport jars called ‘amphoras.’  Our ancient literary sources are sporadic but suggest that olive oil and wine were very common with other commodities carried with a little less frequency.  Stamps and graffiti on the amphoras sometimes reveal content but these are rare.  Occasionally archaeologists have discovered enough physical residues left on the amphora to apply some kind of chemical analysis.  Generally, though, most scholars assume that olive oil and wine were carried most often as these are, after all, the agricultural products of early modern Mediterranean societies.

The authors propose in this article a new way of determining ancient amphora contents: collect DNA samples of genetic material trapped in the surface of amphoras and try to match their sequences with genetic families (species and genera) of plants common to the Mediterranean today: olive, grape, etc…  The authors used two different techniques for extracting the DNA (cotton swabs and scraping) that are inspired in part from modern forensic science (think, police investigations and crime labs).

The sample included 9 amphoras provided by the Ephorate of Underwater Antiquities at the Hellenic Ministry of Culture & Tourism.  This “random sample” of amphoras represented vessels of Classical and Hellenistic date that were literally fished from the sea over the last two decades and given to the Greek Ministry of Culture.  Although a couple of these probably originated from Northern Aegean contexts, most (n=7 of 9) are “Corinthian B” amphoras produced (according to the article’s authors) in Corfu, but associated with production centers in regions of the Ionian sea, S. Italy, and even now N. Africa.  Corinth is often connected to the production of Corinthian B amphoras although this is a matter of debate (see Carolyn Koehler, “A Brief Typology and Chronology of Corinthian Transport Amphoras” and recent discussion here). 

Figures from “A Brief Typology and Chronology of Corinthian Transport Amphoras

Here, as an archaeologist, I wanted some sense of provenience: where exactly were these amphoras fished from the sea?  How were these amphoras selected?  Why all Classical and early Hellenistic?  Why so many Corinthian B vessels?  Some explanation of the selection process would have grounded this study in some spatial context.  As it is, the sample is floating in space.

The point of the article, of course, is not to make historical conclusions but to advertise a new technique: one can analyze trapped DNA in submerged amphoras to determine ancient contents.  This kind of enterprise has dramatic repercussions for our understanding of ancient trade and the value of submerged cargoes.

Foley et al. draw several interesting conclusions about their amphoras.  Olive sequences (oil? olives?) are common in the amphoras (6 of 9), as are grape sequences (5 of 9); juniper sequences appear in most (8 of 9).  Then there are a range of other additives that appear frequently: ginger, walnuts, legumes, beech, citrus, mint, thyme, etc…  One can only conclude (as the authors do) that there was no “pure” wine or olive oil product but that various other spices and plants were used in the production to add flavor and preservatives for the transport.  Olive oil and wine were not undiluted but “value-added products”.

The results also suggest that amphora reuse for different products (olive and wine) was rather common.  There is nothing new, of course, about the idea that amphoras carried many different products (see, for example, this collection of sources) or had many forms of reuse (see, for example, Frontinus’ Stratagems 4.7.9-10 for the hurling of amphoras containing pitch or vipers in naval battles!).  But the ability to map out the local content of a local product has ramifications for our knowledge of trade.

But one wonders: how much does this cost?