The tropical rainforest environment is nutritionally poor and tricky to navigate as compared to open habitats. This poses challenges for human subsistence. There has been little evidence to suggest that human populations relied on rainforest resources before the start of the Holocene, 10,000 years ago. Roberts et al. analyzed earlier fossil human and animal tooth enamel from Sri Lanka. The diet of these humans suggests rainforest rather than open-habitat foraging. Thus, humans were effectively exploiting rainforests in Sri Lanka since at least 20,000 years ago throughout periods of considerable climatic and environmental flux.Science, this issue p. 1246 Human occupation of tropical rainforest habitats is thought to be a mainly Holocene phenomenon. Although archaeological and paleoenvironmental data have hinted at pre-Holocene rainforest foraging, earlier human reliance on rainforest resources has not been shown directly. We applied stable carbon and oxygen isotope analysis to human and faunal tooth enamel from four late Pleistocene–to–Holocene archaeological sites in Sri Lanka. The results show that human foragers relied primarily on rainforest resources from at least  20,000 years ago, with a distinct preference for semi-open rainforest and rain forest edges. Homo sapiens’ relationship with the tropical rainforests of South Asia is therefore long-standing, a conclusion that indicates the time-depth of anthropogenic reliance and influence on these habitats.

Abstract The interpretation of Acheulean landscape use through the analysis of artefact distributions over a range of environmental settings is vital for understanding early hominin behaviour. Such approaches have been successfully applied in areas such as East Africa and South Africa, where long-term and sustained archaeological research has led to the development of robust environmental frameworks within which to analyse hominin land-use patterns. Much less is known about Acheulean hominin behaviour in the Arabian Peninsula, which is increasingly being recognised as an important area for studying early hominin dispersals and adaptations to new environments. To address this lacuna, we have undertaken the first multi-scale systematic survey of Acheulean occupation evidence at Dawadmi, in the centre of the Arabian Peninsula. Specifically, we carried out systematic transect surveys over a large andesite dyke at Saffaqah, on which the majority of 26 known Acheulean sites are associated, as well as across narrow drainage channels, desert pavements and hills located within 5 km of the dyke. Survey transects also crossed neighbouring dykes and adjacent landscape units with a 25 × 20 km area. Our surveys at Saffaqah have led to the discovery of 14 new Acheulean sites. Initial lithic analyses reveal differences between sites in terms of typology, but further work on the assemblages is required to determine if these differences are behavioural or a product of post depositional processes. A broad regional survey was undertaken to identify the full extent of Acheulean activity around Dawadmi. This led to the discovery of a further 22 sites. There is a strong correspondence between Acheulean sites and fine-grained andesite dykes, which were major lithic raw material sources. No Acheulean sites in the study area were found away from dykes or their adjacent landscape units. Based on dyke distributions, the geographic range of Acheulean activity is estimated to be 100 × 55 km, making Dawadmi one of the largest Acheulean landscapes in the world.