Rethinking Holocene landscape evolution at Olympia (Greece): challenging the palaeolake hypothesis in favour of an open fluvial system

The lower Alpheios valley hosts one of the most emblematic archaeological landscapes of the Mediterranean world, where long-term environmental dynamics shaped the development and functioning of the sanctuary of Olympia. Understanding the Holocene evolution of the Olympia plain is therefore central to geoarchaeological interpretations of river dynamics, landscape accessibility, and site preservation. Recently, the traditional fluvial interpretation of the Olympia plain has been challenged by a new palaeoenvironmental model invoking a persistent palaeolake. Here, we evaluate the Holocene evolution of the lower Alpheios valley by integrating published chronostratigraphic, sedimentological and palaeoecological datasets within a newly developed basin-scale geomorphological, sediment-budget and regional source-to-sink framework. The stratigraphic record lacks the laterally continuous, consistently fine-grained sedimentary package expected for a long-lived lacustrine basin. It is instead dominated by mixed grain-size assemblages and repeated moderate-to high-energy depositional facies. Sediment-budget analysis reveals a first-order volumetric mismatch between the maximum storage capacity of the proposed palaeolake (∼0.15 × 10 9 m 3 ) and the ∼2.1–2.5 × 10 9 m 3 of Holocene sediment stored in the Alpheios delta and lower floodplain, exceeding potential lake storage by approximately 14–17-fold. A regional source-to-sink comparison of seven Peloponnesian fluvial systems independently confirms this interpretation, showing that the Alpheios sediment budget is fully consistent with an open fluvial system, whereas the palaeolake scenario would imply anomalously low sediment transfer efficiency outside the regional envelope. Comparative microfaunal and grain-size evidence further demonstrates that assemblages previously interpreted as lacustrine are fully compatible with dynamic floodplain environments. Palaeohydrological reconstruction of the presumed harbour site further indicates that the water levels required by the lacustrine scenario would have flooded large parts of the sanctuary during Antiquity, rendering a harbour interpretation implausible. Geomorphological and lithological constraints preclude long-term natural damming in the Flokas sector, while chronostratigraphic marker horizons point to localised late-Holocene block uplift, offering a parsimonious alternative explanation for elevation offsets and stratigraphic discontinuities previously attributed to lake-level fluctuations. Collectively, these results indicate that the Olympia plain functioned throughout the Holocene as part of an open (unimpounded) fluvial system, with the Alpheios River during Antiquity likely flowing in proximity to the southern margin of the sanctuary rather than along the shore of a persistent lake.