GIAP at the 19th Conference of the International Workgroup for Palaeoethnobotany

The GIAP’s archaeobotany team is presenting in the 19th Conference of the International Workgroup for Palaeoethnobotany (IWGP2022), held from 13-17th June in the Czech Republic, with multiple contributions.

If you’re registered, you can access an interactive programme and check the posters online:

SESSION 1.3 Plants structuring society: trade and globalisation
Poster presentation.

Food and plant trade in the Roman period: an examination of the archaeobotanical remains from Guissona, Spain

Theoni Baniou1, Esther Rodrigo2 and Nuria Romaní2 

1Landscape Archaeology Research Group (GIAP), Catalan Institute of Classical Archaeology (ICAC), Tarragona, Spain
2Serra Húnter Fellow, Autonomous University of Barcelona, Spain

This presentation will discuss new archaeobotanical finds recovered from the Roman city of Guissona, Spain, that bring new information on diet and food trade in the western part of the Roman Empire. The Roman town of Iesso was founded in the beginning of the 1st century BC following the basic Roman urban design of this period and it is considered one of the most important sites in the area. Some of the key features of the city that renders it of high significance to bioarchaeological studies are five wells that have preserved material in anaerobic conditions. Such conditions are extremely rare in the Mediterranean while in the Iberian Peninsula, archaeobotanical studies of such contexts in the Roman period are very few with Iesso offering the first waterlogged results in Catalonia. In this presentation we focus on two wells excavated in 2017 and 2020. The first of these wells was located just outside the city walls and was associated with a building interpreted as likely a hostel, offering accommodation and food to travellers. The second one is found within the city walls, in close proximity to the public bath complex of the city. The on-going archaeobotanical study of the two wells has revealed a rich assemblage of foods that demonstrates several differences between the two different spaces. The analysis of the finds of the two wells offers ultimately new insights into the circulation of goods during the Roman period, inform on their differential access and investigates the potential reasons behind this.

Macro-Theme 1. Session: 1.4. Round table: Method development within archaeobotany
Poster presentation 

Feeding the Gap: A modern experimental cultivation project towards the creation of new tools that explore past agriculture

Alexandra E.T. Kriti1, Alexandra Livarda1, Ioannis Mylonas2, Elissavet Ninou3 and Hector A. Orengo1

1 Landscape Archaeology Research Group (GIAP), Catalan Institute of Classical Archaeology (ICAC), Tarragona, Spain
Hellenic Agricultural Organization-Dimitra, Institute of Plant Breeding and Genetic Resources, Thermi, Thessaloniki, 57001, Greece
Department of Agriculture, International Hellenic University (IHU), Sindos, Thessaloniki, 57400, Greece

In this paper we aim to present a new experimental cultivation project in the north of Greece 

and how this can be used towards the development of new tools that will allow investigation of agriculture in the past. The project focuses on barley (Hordeum vulgare L.) and involves experimental cultivation under field conditions of 22 accessions under different cultivation regimes at the Institute of Plant Breeding and Genetic Resources, based at Thermi Thessaloniki, north Greece. The barley genotypes are originating from various regions of contemporary Greece and constitute the traditional landraces of each region up to the point of the Green Revolution. We have also included a modern cultivar from Thessaloniki, which is used as a control reference. Different cultivation regimes are examined by differentiating input levels of water and manure simulating traditional agricultural practices. By conducting daily observations and measurements on all the plants during their maturity, we are accessing the landraces’ response to the climatic and environmental conditions, as well as their potential impact on the phenotype of the grain and its dietary value. After presenting the experiment we discuss the on-going application of a combination of various methods, such as stable isotope analysis, geometric modern morphometrics and artificial intelligence, on the harvested grains and how these can be then potentially applied to archaeological assemblages to inform on past agriculture. 

Session: 1.4. Round table: Method development within archaeobotany
Oral presentation. Wednesday, 15th June 2022.

A pilot study of a new 3D GMM-based machine learning algorithm for the identification of crop landraces and agricultural practices

Alexandra Livarda1, Hector A. Orengo1, Michael Wallace 1,2, Peter Martin3, Alexandra Kriti1

1 Landscape Archaeology Research Group (GIAP), Catalan Institute of Classical Archaeology (ICAC), Tarragona, Spain
Department of Archaeology, University of Sheffield, UK
3 Agronomy Institute, Orkney College, University of the Highlands and Islands, UK

This paper presents a new methodological approach for the identification of crop landraces and agricultural regimes. Geometric modern morphometric (GMM) analyses of plant remains based on two-dimensional data have been proved promising as a means to differentiate between crop landraces. This approach can be considered a ‘direct’ analysis of crop remains, applied to individual specimens in the archaeological record. Another major development in direct analyses of crop remains has been stable isotope analyses, used effectively to infer past agricultural practices. Cost and the destructive nature of analysis are barriers to the widespread adoption of stable isotope analysis.

In this paper we are presenting a new approach of direct analysis of grains, employing three-dimensional GMM in combination with artificial intelligence that offers an innovative, non-destructive and low-cost new tool that uses the full complexity of grain morphology. We first briefly explore the potential of this new tool to distinguish different agricultural regimes and then present our first application on distinguishing landraces. We use as an example bere, a six-row barley landrace associated with the highlands and islands of Scotland, and other British and Scandinavian barley landraces. Using modern specimens of known origin and growing regimes, we present the results of their 3D shape recording and analysis using GMM. We then outline a series of measures, ratios and indices, which were used in machine learning-based analysis to produce a new algorithm for the automated classification of grain shape according to landrace. These results are then compared with those obtained by 2D-GMM, applied previously by Wallace and colleagues (2018), and their efficacy is discussed. 

Tags: , ,