Sealinks Project in eastern Africa

On the coast of eastern Africa and islands in the adjacent Western Indian Ocean, extending from Kenya in the north to Madagascar in the south, Sealinks and other projects have taken a CPT approach to investigate a range of archaeological issues discussed below (Fig. 3). The first introduction of CPT in the region was in the Comoros Archipelago in 1980, where measured volumes of sediment from small excavations and adventitious sections in sites dating to 800–950 CE were sorted with water flotation and fine screens, studied under magnification, and presented quantitatively (Allibert et al., 1990; Wright et al., 1984, 1992). This early use of the CPT approach proved important for understanding settlement chronology, crop use, and site structure, later becoming the focus of a large-scale systematic study of the Sealinks Project beginning in 2008. While the main goal of the Sealinks Project was to examine early Indian Ocean trade and its relationship to biological exchange, a CPT approach offered insights into early dispersals, island colonisations, marine exploitation, and the spread of farming.. Since 2010, more than 50 small-scale trenches have been excavated at 22 individual sites across the study region, with the aim of rapidly building a broad temporal and spatial picture of an area that had seen little previous systematic research, with a particular paucity of archaeological science (Boivin et al., 2013). Sealinks research was conducted in collaboration with local communities and curators to define cultural resource management issues and to understand impacts on and threats to archaeological sites.

While the Sealinks Project focused on Holocene archaeology, the occupational histories at a few of the investigated sites extended into the Late Pleistocene, enabling insights into the early human occupation of the eastern African coast (Langley et al., 2016; Shipton et al., 2016, Shipton et al., 2018b), processes of continental island formation across the Late Pleistocene-Holocene boundary (Prendergast et al., 2016), and Late Quaternary landscape evolution (Kourampas et al., 2015). At Kuumbi Cave on Unguja island, CPT investigation placed adjacent to previous large-scale excavations provided stratigraphic control to enable the testing of claims of precociously early occupation and also presence of domestic fauna (Shipton et al., 2016). Notably, Sealinks CPT investigation led to the discovery and analysis of Panga ya Saidi, an extraordinary site with a 78,000-year-long archaeological record. Multidisciplinary investigations at Panga ya Saidi have shed light on early human occupation of tropical forests, early marine resource use, symbolic material culture and burial practice, stone tool technology, and long-term climate and ecosystem change (Blinkhorn and Grove, 2018; Clarkson et al., 2018; d’Errico et al., 2020; Faulkner et al., 2021; Goldstein et al., 2022a; Martinón-Torres et al., 2021; Roberts et al., 2020; Shipton et al., 2018b, 2021)..

The CPT approach enabled significant insights into the dispersal of food producers across coastal and island eastern Africa, linked to the large-scale migrations of early Bantu and Austronesian language speakers. The systematic application of flotation and the recovery of botanical remains and molecular analysis of faunal remains shaped our understanding of early foraging and farming economies and produced archaeological evidence for the spread of Austronesian speakers from Island Southeast Asia to the Comoros and Madagascar (Crowther et al., 2016a, 2016b, 2018; Culley et al., 2021a, 2021b; Helm et al., 2012; Ottoni et al., 2017; Prendergast et al., 2017a, 2017b; Shipton et al., 2013). These insights were enriched by the application of ancient DNA analysis to occasional human remains recovered through CPT, which contributed to an overall picture of delayed farmer dispersal to the coast and islands, followed by a long period of forager-farmer cohabitation with limited interaction and admixture, and ultimate swamping no of forager ancestry by people from incoming food-producing populations (Crowther et al., 2018; Skoglund et al., 2017; Wang et al., 2020).