Wood exploitation for a major pre-colonial West African iron production centre (Bassar, Togo)

International audienceLocal iron production played a major role in the pre-colonial West African economy and certain metallurgical centres operated on an industrial magnitude. One of them is the Bassar area in central Togo where the onset of iron metallurgy is considered to date back to the 3rd-4th century BC, even though only from the 14th century AD onwards the area experienced distinct production intensification. High-level iron production lasted as late as the early 20th century. Bassar iron metallurgy was often considered to have caused massive deforestation and vegetation degradation, mainly due to wood exploitation for charcoal production, specifically to fuel iron smelting furnaces. The study presented here is the first to test this assumption using anthracological methods, supplemented by ethnohistorical interviews and archaeometallurgical investigations. The focus of our investigations is on two sites, Tchogma and Tatré. The charcoal assemblages at both sites and during all production periods are dominated by trees and shrubs with dense wood. According to our ethnohistorical interviews, hard wood species were considered particularly suitable for iron smelting, including several species with a strong capacity to re-sprout after coppicing. The taxon composition of the charcoal assemblages from metallurgical waste heaps generally corresponds to the oral traditions. In addition to the taxa mentioned by the interviewees, useful fruit trees such as the shea butter tree (Vitellaria paradoxa) were regularly used as fuel for ironworking. This demonstrates the exploitation of agroforestry parklands, which is providing evidence that arable land was already occupying a vast land surface of the Bassar area. The results of our study seem inconsistent with the previous idea of massive vegetation degradation due to the impact of wood exploitation for iron metallurgy in Bassar. Three main factors contradict previous general deforestation assumptions: 1. The anthracological results indicate the use of taxa with the ability to re-sprout, which probably allowed for fast woodland regeneration; 2. We consider the area to possess a higher biomass reproduction capacity in comparison to areas further north investigated in our previous studies; 3. The development of local craft specialisation scattered the impact of wood use on the Bassar landscape, thereby, preventing complete local deforestation. Our re-evaluation of Bassar slag amounts indicates fuel consumption was higher than previously thought. The over-exploitation of certain preferred taxa, particularly species with a restricted capacity to re-sprout, and local vegetation changes around the smelting sites is likely, although such modifications are not unequivocally reflected in the charcoal data