In a recent paper published by Scientific Reports, we detailed the results of our reinvestigation of Pathirajawela, an open-air site in Sri Lanka’s Southern province. Previously dated to ca. 74,000 years ago, we showed that the site was most likely occupied by people from around 25,000 to 1000 years ago. Our findings provide insights on how human groups that settled the site responded to climatic and environmental changes from the Last Glacial Maximum onward (i.e. the end of the last Ice Age), specifically how they adapted to changing ecosystems linked to sea level transgressions and coastal evolution in the last 25,000 years. We also argue for the anthropogenic nature of shell midden deposit in the site, a feature observed in other archaeological sites in the southern Sri Lankan coast. This midden deposit, dated to ca. 4,800 years ago showed people focused on collecting a single shell taxon, a saltwater/estuarine clam called Meretrix casta, as humans took advantage of new resources offered by the formation of new ecosystems resulting from higher sea levels during the mid-Holocene.

The Pathirajawela site was first discovered by Dr. Siran Deraniyagala during his 1970s survey of archaeological sites in southern Sri Lanka. He noted two general types of stone tools in the site: large flakes and scrapers that he recorded more in the lower levels and smaller blades in the upper sediment layers of the site. These blades (‘microliths’) bear resemblance to the stone tools that were being found in cave sites in the interior forest regions of Sri Lanka. Sediment samples from the lower levels were dated via thermoluminescence to ca. 70,000 years ago and this led Dr. Deraniyagala to hypothesize that the lower levels of Pathirajawela could yield perhaps the earliest evidence of human occupation of the island.

Following this initial study, we decided to revisit the site and conduct a two-month long excavation in late 2019. The team consists of researchers from the Max Planck Institute (Germany), the University of Sri Jayewardenepura and the Department of Archaeology of the government of Sri Lanka, with me and Dr. Oshan Wedage of the University of Sri Jayewardenepura leading the excavation. It was an intense two month, where we had to work almost everyday in the field from seven in the morning to three in the afternoon, continuing in our basecamp with artefact processing and cataloguing and sediment sieving and floatation until around ten at night. The site is located within the Bundala National Park, a protected wildlife area where we encountered many animals including the occasional spotted deer, elephants and crocodiles and the many toque macaques that watch us while we excavate. By the end of the two-month excavation, we have removed almost 17 cubic meters of sediments, reaching a depth of around 4.5 meters and recovered around 5000 stone artefacts, more than half of which we 3D recorded using a total station. We also sampled almost 1200 kg of bivalve shells.

The next years were devoted to the analyses of the materials we recovered. COVID-19 delayed our plans significantly, as most of our materials were left in institutions in Sri Lanka. But we were working with amazing, dedicated researchers that went out of their way to make sure that our research progressed. Dr. Pat Faulkner from the University of Sydney led the painstaking analyses of the more than 1000 kg shells we collected. He and his team looked at, weighed and classified a total of 125,000 pieces of shells. Dr. Chris Clarkson studied the almost 5000 pieces of stone artefacts we recovered. Colleagues from the Spanish National Centre for Human Evolution conducted Optically stimulated luminescence (OSL) dating and Magnetic Susceptibility reading of sediment samples we managed to export out of Sri Lanka. The export process itself was tricky, as OSL dating requires the samples not to be exposed to light or any source of radiation- quite complicated when you need to take 2-3 international connecting flights (i.e. not checked-in and hand inspected, following EU protocols on import of sediments, etc). 

The OSL dates suggest that Pathirajawela records human occupation starting from 25,000 years ago and not from 74,000 as previously estimated. We provide an explanation for this in the paper- noting the challenges in the older thermoluminescence dating method, particularly considering the type of sediment found in the site. We describe in detail the changes in patterns of stone artefact manufacture in Pathirajawela, from the use of bigger flaked tools to the utilization mostly of microliths. We estimate the shift to have occurred around 14,000-15,000 years ago. We modelled the distribution of stone artefacts in the stratigraphy, using the 3D coordinates we have of almost 2000 stone artefacts. We recorded three major phases of artefact accumulation dating to roughly 5000, 14,000 and 25,000 years ago. Interestingly, we noticed that these phases correspond to the same sediment units where we recorded high magnetic susceptibility measurements, which we hypothesize to correspond to periods with increased moisture or vegetation.