Reconstructing the provenance of the hominin fossils from Trinil (Java, Indonesia) through an integrated analysis of the historical and recent excavations

In the early 1890s at Trinil, Eugène Dubois found a hominin skullcap (Trinil 2) and femur (Trinil 3, Femur I), situated at the same level ca. 10–15 m apart. He interpreted them as representing one species, Pithecanthropus erectus (now Homo erectus) which he inferred to be a transitional form between apes and humans. Ever since, this interpretation has been questioned—as the skullcap looked archaic and the femur surprisingly modern. From the 1950s onward, chemical and morphological analyses rekindled the debate. Concurrently, (bio)stratigraphic arguments gained importance, raising the stakes by extrapolating the consequences of potential mixing of hominin remains to the homogeneity of the complete Trinil fossil assemblage. However, conclusive evidence on the provenance and age of the hominin fossils remains absent. New Trinil fieldwork yielded unmanned aerial vehicle imagery, digital elevation models, and stratigraphic observations that have been integrated here with an analysis of the historical excavation documentation. Using a geographic information system and sightline analysis, the position of the historical excavation pits and the hominin fossils therein were reconstructed, and the historical stratigraphy was connected to that of new sections and test pits. This study documents five strata situated at low water level at the excavation site. Cutting into a lahar breccia are two similarly oriented, but asynchronous pre-terrace fluvial channels whose highly fossiliferous infills are identified as the primary targets of the historical excavations (Bone-Bearing Channel 1, 830–773 ka; Bone-Bearing Channel 2, 560–380 ka), providing evidence for a mixed faunal assemblage and yielding most of the hominin fossils. These channels were incised by younger terrace-related fluvial channels (terminal Middle or Late Pleistocene) that directly intersect the historical excavations and the reconstructed discovery location of Femur I, thereby providing an explanation for the relatively modern morphology of this ‘bone of contention’. The paleoanthropological implications are discussed in light of the current framework of human evolution in Southeast Asia.