Introduction

Citrus fruits are nowadays a major component of the Mediterranean landscape, and the most important fruit production in the world, but the history of their diversification and spread from their primo-domestication areas (south-western China and north-eastern India) towards the Mediterranean remains poorly documented. The AGRUMED research project (directed by G. Fiorentino and V. Zech-Matterne) aims to address questions regarding ancient species of citrus fruits, their cultivation and uses, as well as the agents, routes and tempo of their dispersal. The project comprises an international network of laboratories and associated researchers working in the fields of archaeobotany, ethnobotany, archaeology, genetics, molecular biochemistry and agronomy.

Hypothesis and first results

DNA genetic markers have successfully established the fundamentals of citrus fruit phylogeny (Nicolosi et al. 2000; Barkley et al. 2006; Garcia et al. 2013; Luro et al. 2012). Diversity within this botanical genus originates from four species or taxonomic groups: citrons (C. medica L.), pomelos (C. maxima (Burm.) Merr.), mandarins (C. reticulata Blanco) and papedas (wild progenitors). After a phase of allopatric evolution in separate geographical districts, these four species generated, in cross-border areas, the domestic forms currently cultivated today: sweet orange (C. sinensis L.), sour orange (C. aurantium L.), lemon (C. limon (L.) Burm.) and lime (C. aurantifolia (Christm.) Swing.). These hybrids—considered as new species—were multiplied through an apomixis reproductive process (adventitious nucellar embryony) and became even more diversified by mutation. Their introduction to the Mediterranean created a secondary centre of diversification and assisted the appearance of new types of citrus fruit.

Figure 1.

Study area and sampled sites.

Recent discoveries in south-western Italy support the hypothesis of an early introduction of citrus trees to Italy, during the pre-Archaic period. Pollen of Citrus type has been found, dating from 896–657 cal BC onward, in the reference profile of the harbour city of Cumae, one of the earliest Greek foundations in the Campania region (Bui Thi Mai & Girard 2010). Archaeological knowledge about the precocity and wide extent of Phoenician and Punic trade along the southern coasts of Italy and the related islands raises the idea that these traders could have played a role in the propagation of Citrus species. Recent finds in Pompeii and Rome establish the existence of at least Citrus medica and C. limon in Italy, from the end of the first century BC (Pagnoux et al. 2013) (Figure 1).

Taking this evidence into account, the aims of the AGRUMED project are, on the one hand, to set new identification criteria (based on geometrical morphometrics) applicable to archaeobotanical remains, which are morphologically quite distant from recent cultivars. Moreover, we would like to test the genetic markers best adapted to aDNA short sequences. On the other hand, the project focuses on archaeological investigations at strategic sites located on potential dispersal routes. Within the framework of several excavation programmes, we aim to improve the recovery of archaeobotanical remains, especially through water-sieving under a 0.5mm mesh onwards.

Tools for the identification and characterisation of citrus fruit remains

The shape, size and organisation of the cell patterns render the distinction between citron and lemon, and other types of citrus fruit, possible and reliable. Such criteria have been successfully applied to the determination of archaeobotanical remains from Italy (Pagnoux et al. 2013). A geometrical morphometric approach is planned on eight species: citron, mandarin, pomelo, sweet and sour orange, lemon, lime and chinotto (x Citrus aurantium var. myrtifolia Raf.). The dataset includes 30–50 specimens per species. Using digitised outlines of the seeds, parametrical descriptors will be defined and entered into a barycentre matrix, on which statistical analyses will be performed.

The botanical remains preserved by carbonisation or waterlogging sometimes contain very small quantities of aDNA (e.g. see Schlumbaum et al. 2008) and require specific extraction techniques.

The National Institute for Agronomic Research (INRA) and Centre de coopération internationale en recherche agronomique pour le développement (CIRAD) research team have developed hundreds of genetic markers from gene or genome sequences (Garcia et al. 2013). Among the SNP and InDel markers, we will select the most informative ones and adapt them to the very fragmented matrices typical of archaeological samples. Specimens from herbal and wood collections of the INRA conservatorium will be used for this purpose.

Figure 2.

Top: mineralised seeds of citrus fruit from the temple of Venus, Pompeii (© V. Zech); 1) cell patterns of the archaeological seeds at the magnification of X1000. They are similar to those of C. medica; 2) comparison with the cell patterns of Citrus medica var. Etrog (X488) 3) comparison with the cell patterns of Citrus limon var. Frost Eureka (X488) (© S. Pont/V. Zech).

Geographical setting

The project's study covers the potential places of diversification and dispersal of the citrus fruit towards and within the Mediterranean: Cyprus, Greece, Malta, Italy, Egypt, Morocco, Spain and southern France.

Diverse domestic, military and other cultural sites, spanning the Archaic to the Islamic periods, have been specifically sampled to retrieve fruit/seed remains and charcoals of Citrus (Figure 2). These sites present long chronological sequences, and their potential interest is enhanced by the fact that other citrus remains have been found in the same regions or in the same kind of contexts. Written sources reinforce the interest of medieval sites located in these areas as relays for the transmission of these species.

The archaeobotanists and geneticists associated with the AGRUMED project would be interested in studying any available material from other sites in these Mediterranean areas. You can contact the project via: girolamo.fiorentino@unisalento.it and zech@mnhn.fr.