Fine and coarse-quartz SAR-OSL dating of Last Glacial loess in Southern Romania

Previous single aliquot regenerative optically stimulated luminescence (SAR - OSL) dating studies of representative loess sections in southeastern Romania revealed severe discrepancies among the ages obtained on fine (4-11 mu m) and coarse (63-90 mu m) grains of quartz. The current study aims at expanding these investigations by extending the area of study to the more westerly loess deposits in the Lower Danube Basin. The luminescence properties of the grain sizes of quartz extracted from 18 samples collected from the Last Glacial loess layer at Lunca section (Wallachian Plain) are examined and compared. Investigations confirm the reliability of the SAR - OSL protocol previously applied to Romanian loess (preheat at 220 degrees C for 10 s, cutheat at 180 degrees C and elevated temperature OSL). Despite this, the obtained equivalent doses on coarse quartz are higher than those obtained on fine material for most samples, as in the case of our previous studies. These results into fine quartz OSL ages that are significantly lower compared to the coarse quartz age results. However, the ages obtained for the uppermost two samples are in very good agreement. Ages spanning from 19 +/- 2 ka to 43 +/- 4 ka on fine quartz and from 20 +/- 2 ka to 54 +/- 6 ka on coarse quartz have been obtained for the loess deposited during Last Glacial period. Based on the fine as well as the coarse quartz chronology significantly low sedimentation rates were computed for the Last Glacial Maximum period and MIS 3. OSL ages and magnetic susceptibility measurements indicate that the loess layer corresponding to the Early Glacial period (MIS 4) is very thin at this site and affected by pedogenesis. Regarding the investigation into the cause of the observed age discrepancy we present evidence through time resolved investigations that the observed behaviour is typical of quartz and cannot be accounted by feldspar contamination. We confirm once more that the growth of the OSL signal in nature does not correspond to the laboratory generated SAR dose response curve that is best described by a sum of two saturating exponential functions. The differences observed between the natural and the laboratory dose response for the two quartz fractions are believed to be a cause for the different chronologies reported. (C) 2014 Elsevier Ltd and INQUA. All rights reserved.