In the southern part of Kyushu Island in southern Japan and the small islands further south the earliest pottery is found beneath the Satsuma tephra, which has been well dated to ca. 12,800 cal BP. Here we focus on Incipient Jomon pottery, 14,000/13,500–12,800 cal BP, from the Sankakuyama I site on Tanegashima Island. Previous visual analysis of the fabrics suggested that about half of the vessels were not locally made. In this study, we conducted both ceramic and raw material petrography and an electron microprobe study on samples of pottery. Our results indicate that pottery was mainly produced in situ, away from the coast, but that there is some clear non-local material which came from either Yakushima Island or Kyushu Island proper. Yakushima has no reported Incipient Jomon sites. There should be undiscovered sites on Yakushima if pottery circulated from there rather than Sankakuyama I residents embedding production in their logistical moves on Yakushima. Minor signatures of non-local geology in locally produced pottery are probably the result of volcanic eruptions and sea currents. Pottery production began when Tanegashima was disconnected from Kyushu and probably about to be separated from Yakushima. During the Incipient Jomon period, Tanegashima had become isolated. We conclude that pottery producers were hunter-gatherers who were mainly sedentary, living in a mild environmental with ecotone properties. They occasionally engaged in costly communication and exchange, which may have involved transporting pottery by watercraft and on foot, to buffer risks. Our study is among the first to investigate the pottery economy of the late Pleistocene and the decisions made by its producers and users in response to environmental variability and change. The research contributes to the debate on the origins of pottery and the Upper Paleolithic to Neolithic transition.
