Evidence of forest management and arboriculture from wood charcoal data: an anthracological case study from two New Caledonia Kanak pre-colonial sites
- First Online:
- Cite this article as:
- Dotte-Sarout, E. Veget Hist Archaeobot (2017) 26: 195. doi:10.1007/s00334-016-0580-0
- 187 Downloads
Archaeological wood charcoal analysis or anthracology has been applied for the first time in New Caledonia as part of an interdisciplinary research program examining Kanak pre-colonial landscape management in a valley on the northeast coast of the subtropical Pacific island. In contrast to previous hypotheses, this study demonstrated that when the Kanak traditional cultural complex emerged around ad 1000, following the initial 2,000 years of human presence on the island, the vegetation cover showed few signs of deforestation and the tropical rainforests were still prominent. The vegetation surrounding Kanak settlement sites evolved during the first half of the 2nd millennium ad towards a more open but more complex composition that included useful taxa. This was interpreted as showing a form of forest management and possible arboriculture (the cultivation and management of trees). However, in sharp contrast to its Melanesian neighbours, very little is known about arboricultural practices in New Caledonia. Through the interpretation of data from two sites in particular, a discussion of the two-step analytical process is used to argue for the existence of arboricultural practices associated with these sites: (1) using anthracological data to reconstruct the vegetational landscape, together with ethnobotanical, ethnohistorical and archaeological data; (2) allowing for the recognition of specific practices of forest domestication, based on the manipulation of plants and of spatial patterns of forests. At a time when archaeology is engaged in a process of post-colonial re-evaluation of its schemes of interpretation, it seems timely for archaeobotany to try integrating more indigenous systems of representation into its analyses. The approach presented here is an effort in this direction, trying to make sense of new data by reading them through a local lens.