The project focused on a topo-chronological reassessment of iron smelting on Elba in Antiquity and the environmental impacts of ancient metallurgy on the island.


Elba Island was—together with the major Etruscan seaport Populonia—the centre of Etruscan and Roman iron smelting in the Western Mediterranean region in classical Antiquity. The island’s rich iron ore deposits played an important role in ancient economy, especially in the armed conflicts between Phocians, Carthaginians, Etruscans, and Romans.

The earliest archaeological traces of iron smelting on Elba date apprx. to the 5th century BCE, although other research tendencies suggest an earlier beginning. According to the current state of the art, all smelting sites were abandoned on the island in the first half of the 1st century BCE and should have shifted to the continent. The reasons for both the dislocation and the end of smelting are under discussion. Besides a senatus consultum prohibiting all mining in Italia, a popular hypothesis is that smelting on the island was no longer possible due to a lack of fuel wood to run the furnaces. Nevertheless, a reliable scientific basis for this assumption is still missing.

In close cooperation between Classical Archaeology and Physical Geography the project focused on (1) a topo-chronological reassessment of iron smelting on Elba in Antiquity and (2) the environmental impacts of ancient metallurgy on the island. Besides deforestation, also the emission of soot—the ancient Greek name of Elba is Aithále, ‘sooty’—and the disposal of thousands of tons of slag all over the island is an issue. Thus, the reconstruction of human–landscape interactions was the overarching objective of the project.

The methodological approach includes archaeological and geographical techniques of data collection, namely: document and literature research, survey of smelting sites and iron slag accumulations, sediment percussion drilling, and on-site soil sampling. Slags are radiocarbon-dated as absolut-chronological tool for an independent age control of existing relative-chronological models; sediments were macroscopically described, and analysed in the lab for physical parameters, bulk chemistry, and elemental composition. Interdisciplinary interpretation of the results included archaeological data and sedimentological evidence.


Chronological results more or less confirmed the existing chronology, but also offered new insights into the beginning and end of iron smelting on the island: Although all ore deposits are located in the eastern part of Elba, there was also an early smelting site (3rd century BCE) more than 30 km away on the western coast. Thus, the common model of a complete deforestation of the eastern slopes, followed by a shift of smelting to the west to exploit new fuel wood resources, has to be questioned. Neglected finds from research in the 1950s that gained little attention until now increase the doubts. A re-examination of old archaeological data show that iron smelting on Elba might not have ended in the 1st century BCE, but continued in the 1st century CE.

Core data offer insights in the landscape history of a coastal plain, where Roman (and Medieval) smelting sites were located in the hinterland. The plain is strongly influenced by the Holocene/Versilian Transgression—as indicated by paludal sediments. Alluvial aggregation begun c. 5000 BP. A colluvial layer and alternating high magnitude flood layers indicate accelerated soil erosion – potentially after deforestation – during Roman times. Geochemical data show that trace metal(loid)s are enriched in the sediments, but are mainly of geogenic origin and associated with autochthonous, pedogenic oxides. Nevertheless, a sedimentary smelting signature—characterized by an enrichment of arsenic and allochthonous iron—shows a potential impact of iron smelting on the island.