The more than 100 ha wide reclamation area was meticulously searched on hands and knees [by Berghuis], collecting vertebrate remains visible to the naked eye, Harold Berghuis and co-workers explained in detailing a procedure that would eventually return the first underwater hominin fossils ever to be found in the drowned plains of Sundaland a landscape that, until today, had existed only in models and legend.
The Madura Strait, a narrow waterway separating Java from Madura Island, today conceals a Pleistocene river valley that once teemed with life. Over 6,000 fossils bones of Stegodon, Komodo dragons, deer, hippos, and two crucial fragments of a Homo erectus skull were drawn from the seabed during marine sand dredging for a construction project near Surabaya. The discovery, as reported in four papers in Quaternary Environments and Humans, represents a milestone for paleoanthropology in Southeast Asia: the first direct proof of hominins in a drowned environment that was the Sundaland heartland.
The valley fill was dated by optically stimulated luminescence (OSL) to between 163,000 and 119,000 years ago, or during the penultimate glacial period. Sundaland was then a big lowland, its rivers slicing through savannas and forests, linking Java to the Asian continent. The ecological wealth of the region is highlighted by the fossil record: Komodo dragons now threatened and restricted to a few islands were once widespread as apex predators, and Stegodon, an extinct elephantine giant, and deer species point toward a mixture of habitats.
What makes the Madura Strait discovery notable is not just its underwater setting but also the behavior it reveals. Among the animal remains, scientists found cut marks and breaks proof that here Homo erectus practiced selective hunting of healthy bovids and turtles, and processed bones for marrow. “Among our new finds are cut marks on the bones of water turtles and large numbers of broken bovid bones, which point to hunting and consumption of bone marrow,” Berghuis reported to Archaeology Magazine. This is the first known evidence of turtle hunting in Southeast Asia, and targeting prime-age bovids is a practice previously associated with more recent human species on the Asian mainland.
This provokes intriguing questions regarding contact and even gene flow among archaic groups. Although previous models were that Javanese Homo erectus existed in isolation, the new data indicate potential interaction with other hominins like Denisovans or Neanderthals, whose presence on the Asian mainland is well documented. The complex paleogeography of the region controlled by a pattern of glaciation and sea-level change would have periodically brought populations together and kept them apart, encouraging exchange and isolation.
The technical feat behind such findings is no less incredible than the fossils themselves. Underwater archaeology in this context is based upon industrial-scale dredging, followed by labor-intensive manual sorting and sediment analysis. Contractors dug out approximately 177 million cubic feet of sediment through a trailing suction hopper dredger, which disintegrates and suctions seabed material for land reclamation. The recovered fossils, from the reclamation area and core samples, were dated and placed in context by OSL and stratigraphic correlation with established river systems.
The environmental narrative is no less captivating. The river valleys of Sundaland in the late Middle Pleistocene were teeming with life, featuring perennial water sources and all-year-round edible vegetation, shellfish, and fish. But that landscape was not permanent. As the last glacial maximum dissipated, mean sea level worldwide increased by more than 120 meters, fueled by two huge meltwater pulses MWP1A and MWP1B between 14,500 and 11,000 years ago (GMSL increased from ~−122 m to −1 m). The sharp increases drowned the lowlands, isolating habitats and forcing migrations. In Southeast Asia, Sundaland’s terrestrial area declined by approximately 50% from the Last Glacial Maximum to mid-Holocene, radically restructuring human and animal populations.
Genomic analyses now show that these environmental catastrophes left a genetic trace in contemporary Southeast and South Asian populations. Admixture between Malaysian Negritos and South Asian Austroasiatic populations coincident with the flooding of Sundaland implies that sea-level rise drove migrations and established novel population compositions. The results, enabled by high-coverage whole-genome sequencing and sophisticated paleogeographic modeling, illustrate how environmental change can promote both extinction and innovation.
The Madura Strait site also presents a unique possibility of reconstructing a Pleistocene ecosystem in its complexity. The existence of river sharks, hippos, and water buffalo, in combination with terrestrial megafauna and reptiles, gives us an image of a dynamic, interrelated world. Such reconstructions rely on merging sedimentology, paleontology, and geochemistry disciplinary areas that, in this instance, came together because of the coincidence of engineering and archaeology.
For scholars and historians of ancient human civilization, the Madura Strait find is a reminder that the best pages of our history could still be hidden beneath the oceans. Berghuis’ thought about it was, “We present the results of our studies in four extensive, richly illustrated articles, creating a unique window into the drowned Sundaland of 140,000 years ago.”
The fossils, presently preserved at the Bandung Geological Museum, provide a tangible connection to an altered world ravaged by climate, sea, and the restless innovativeness of early humans.
