High-precision dating of Abrigo de La Boja (Murcia, Spain): Implications for the Middle-to-Upper Palaeolithic transition and the culture-stratigraphy of Upper Pleistocene Iberia

La Boja is a rock-shelter located in Murcia (Spain) whose c. 7 m infilling spans the last 50,000 years of the Pleistocene. Though punctuated by a few episodes of stasis or erosion, sediment accumulation proceeded at a regular rate, while human occupation was intermittent and light. This pattern allowed for the preservation of 31 well-separated Occupation Horizons whose stratigraphic integrity is demonstrated by stone tool refitting, the pristine preservation of hearths and earth ovens, and the age-depth consistency of dating results. Verification of these criteria warranted Bayesian modelling of the sequence, which was carried out under OxCal and delivered a high-resolution chronostratigraphy based on 56 age measurements obtained by OSL and radiocarbon. Our main conclusions are as follows: substantial human use of south-east Spain’s arid interior during MIS 2 and MIS 3 was restricted to interstadials; during stadials, visits were fleeting and left behind an exceedingly scant record (except perhaps in the Solutrean, when harsh conditions prevailed globally but short-lived pulses of increased humidity occurred locally); the Mousterian’s basal metre, featuring Levallois reduction and Soyons points, formed through the 50–55 ka (thousands of years ago) interval; the region’s earliest Upper Palaeolithic is the Evolved Aurignacian, which replaced the Middle Palaeolithic during the 37.1–37.4 ka interval (thereby supporting that, in southern and western Iberia, Neandertals persisted for longer than elsewhere in Eurasia); the transition to the Gravettian occurred within the 34–35 ka interval and there is no evidence that it was demic rather than simply technological; the existence of a distinct Lower Solutrean phase spanning the 25.0–25.5 ka interval is confirmed; in the region, the 20–22 ka slot is occupied by the Solutreo-gravettian, which implies that claims for the French Badegoulian to have extended as far south as the Spanish Levant cannot be supported.