Throughout the project, we conducted a series of activities to explore long-term cultivation practices of the three C4 species under study: we conducted experimental cultivation under controlled conditions, we performed ethnoarchaeological interviews with local traditional farmers, we modelled ethnoarchaeological data and applied these models to archaeological evidence, and we performed archaeobotanical analysis of millets from selected sites.
Three seasons of experimental cultivation were carried out, two at ICRISAT (Patancheru, Hyderabad, Telangana, India), and one at Universitat Pompeu Fabra (Barcelona, Spain) using an approach that enabled precise control over watering conditions for plants grown under different regimes, facilitating the study of their responses to water treatments and their transpiration rates. The data collected provided insights into the relationship between plant watering and the archaeological proxies under investigation. The analysis included carbon and nitrogen stable isotopes composition of grains (fresh and charred), total biosilica extracted from plant tissues, phytolith morphotypes abundance and ratio, and stable oxygen and stable silicon isotopic composition of phytoliths. Results were modeled and used to examine watering practices in samples from the Indus Civilisation (Sindh, Pakistan) and Pre-Aksumite and Aksumite contexts (Tigrai, NE Ethiopia) and neolithic contexts (Central Sudan).
Ethnographical interviews with local farmers were conducted in Pakistan, Ethiopia, and Sudan. In Pakistan, we explored an area near the Khirtar mountains in Sindh, where farmers cultivate rain-fed sorghum and pearl millet. In Ethiopia we collaborated with sorghum and finger millet farmers in Tigrai. Additionally, in Sudan, we interviewed farmers cultivating rain-fed pearl millet and décrue sorghum in the Khartoum State. The ethnographic work revealed similarities in decision-making processes among farmers dealing with scarce and unpredictable rainfall, as well as specific adaptations unique to each region. This information allowed us to construct a global model of the cultivation of three specific crops (pearl millet, finger millet, and sorghum) by combining a cross-cultural approach with the data collected during interviews.
Finally, we analyzed samples of phytoliths and seeds from several archaeological sites in drylands, primarily in Pakistan, Sudan, and Ethiopia, but also in Egypt and South Africa. The integration of data and models developed in the experimental and ethnographic work packages enabled us to reconstruct water management practices and land-use strategies associated with millet cultivations at these archaeological sites.
