Optically stimulated luminescence (OSL) dating of quartzite cobbles from the Tapada do Montinho archaeological site (east-central Portugal)

The burial age of an alluvially deposited cobble pavement at the Tapada do Montinho archaeological site (east-central Portugal) is investigated using optically stimulated luminescence (OSL) dating. Measurements on the cobbles (quartzite clasts) were carried out on intact slices and large aliquots (similar to 8?mm) of quartz grains (63300?mu m), both recovered from the outer 1.5-mm surface of the cobbles. The recycling ratio, recuperation and dose-recovery tests show that the single-aliquot regenerative-dose (SAR) protocol is applicable to both rock slices and quartz grains; both have similar luminescence characteristics. The variation in the natural OSL signal with depth below the cobble surface using intact slices from two different cobbles shows that both were bleached to a depth of at least similar to 2?mm before deposition. A model of the variation of dose with depth fitted to data from one of the cobbles gives a burial age of similar to 19?ka and also predicts the dose-depth variation at the time of deposition. Ages based on rock slices suggest that one cobble surface, and the inner parts of two other cobbles experienced a resetting event at similar to 45?ka, consistent with the age control. However, the surfaces of the other cobbles all record light-exposure events in the range 26 to 14?ka, suggesting that some of the cobbles were exposed to daylight perhaps more than once in this period. Given the shallow burial depth and unexpectedly young ages of the surrounding and overlying finer-grained sediment, it is suggested that phases of light exposure following surficial erosion are probably responsible for this underestimate. Nevertheless, it is remarkable that we can identify and quantify four events (two light exposures of different durations and two sequential burial periods) in the dose record contained within a single clast, and this suggests that the luminescence dating of rock surfaces may prove, in the future, to be at least as important as sand/silt sediment dating.