This paper presents dating results of geological (speleothems, tufa, soil, and sediment), biological (mollusks and botanical), as well as hydrogeological samples from Croatia, Slovenia, Bosnia and Herzegovina, Turkey, and China. Included are results of samples measured by gas proportional counting (GPC) in the Zagreb lab until abandonment of this technique in 2007, as well as results of several series measured by both GPC and liquid scintillation counting (LSC) methods.

New attempts are presented to determine the age of large Pleistocene mammals excavated at Starunia, ~130 km southeast of Lviv, Ukraine. This remarkable discovery made at the beginning of the 20th century included a complete carcass of woolly rhinoceros (No. 2), fragments of 3 woolly rhinoceroses (Nos. 1, 3, and 4) and remnants of numerous specimens of other fossil fauna and flora. Although attempts to date paleontological findings from Starunia site go back to the early 1970s, the results obtained before 2006 are somewhat misleading, mostly due to unresolved contamination problems. Comprehensive cleaning of the samples adopted in the framework of this study was aimed at removal of 2 potential sources of contamination: (i) radiocarbon-free hydrocarbons abundant at the burial site; and (ii) allochthonous organic materials containing contemporary carbon that were used in the past during preservation of the dated specimens. Two types of samples have been analyzed for their 14C content in the framework of the present study: (i) fragments of bones and teeth collected from specimens stored or exposed in the Natural History museums in Lviv and Kraków; and (ii) samples of terrestrial macrofossils retrieved from sediment cores obtained during the 2007–2008 field campaigns in the Starunia area. 14C analyses of collagen were supplemented by measurements of its elemental C/N ratio and 13C/12C and 15N/14N isotope ratios. Three 14C dates obtained for rhinoceros No. 2 span the age range from 35.3 to 40.0 ka BP, in agreement with the minimum age estimated from macrofossils. The mean value of 37.7 ± 1.7 ka BP falls in the range of ages reported for big Pleistocene mammals from other locations in Europe. The bones of rhinoceros No. 3, which were found in close vicinity to those of rhinoceros No. 2, reveal a 14C age of 36.7 ± 0.6 ka BP. The δ15N and δ13C values obtained for collagen extracted from bones and teeth belonging to rhinoceroses Nos. 1, 2, and 3 are in a broad agreement with analogous literature data for large Pleistocene mammals found in other sites in Europe, North America, and Siberia.

Two soil profiles from northeast India, one from Bakrihawar, an agricultural land, and the other from Chandipur, a virgin hilly area from Assam, are investigated to understand the organic carbon dynamics of the area. Due to frequent flooding, the Bakrihawar soil has accumulated a higher clay content than that of Chandipur. The carbon content is less than 1% by weight in both the sites. The higher clay content is responsible for relatively more soil organic carbon at Bakrihawar. The mean δ13C values at both sites reflect the values of the overlying vegetation. At Bakrihawar, both rice cultivation (C3) and natural C4 grasses contribute to higher mean enriched values of 13C relative to Chandipur, where the surface vegetation is mostly of C3 type. The turnover time of organic carbon, estimated using the residual radiocarbon content, depends strongly on the soil particle size distribution, especially the clay content (i.e. it increases with clay content). To the best of our knowledge, this is the first soil carbon dynamics study of its kind from northeast India.

Paleoenvironmental and paleogeomagnetic tracing studies from Chinese loess 10Be have progressed in recent years (Zhou et al. 2007a,b, 2010). In this approach, 10Be flux determined from sediment concentration and accumulation rate may be used to recover information about paleomonsoon rainfall rates as well as past variations in the geomagnetic field strength. However, these methods require that a correction be made for residual undecayed 10Be in remobilized dust. To better understand the feature of the 10Be signals related to the remobilized dust, we report the first observational study on 10Be concentration of modern falling dust using the 3MV multi-element accelerator mass spectrometer (AMS) in the Xi’an AMS Center. Ten samples collected at Ansai observation station (109°19'E, 36°51'N) in northern China from May 2008 to June 2009 are measured along with 3 chemical blanks. The results clearly show that the 10Be content of modern falling dust is relatively uniform, with a mean value of 1.21 × 10sup8/sup atoms/g, a measurement similar to that of Chinese loess (Zhou et al. 2007a) and to the value found in the study by Shen et al. (2009) on dust 10Be falling near Dingbian, China (1.25 ± 0.06 × 10sup8/sup atoms/g). Despite the fact that modern dust flux is much higher in spring relative to summer in northern China, 10Be concentration in falling dust remains fairly constant. In addition, we find that dust 10Be concentration is roughly independent of the local precipitation changes. This feature might be considered as an analogue to improve our understanding on the fundamental information of the source component contained in loess 10Be records and its spatial/temporal distribution features.

This paper describes a method for effective separation of the pure binder fraction of lime mortars for reliable radiocarbon dating. The methodology allows removal of the detrital carbonate fraction and the unburnt limestone particles, obtaining particles of under 1 µm. The extracted fraction ensured that all carbonate has been generated by slaked lime carbonation. Consequently, the measured carbon corresponds to atmospheric carbon. The proposed method allows to obtain pure datable binder, simplifying considerably the performance of radiometric measurements because dating other grain-size fraction is unnecessary. In order to prove the effectiveness of binder refining, the extraction method has been applied to 5 lime mortars of different archaeological periods from the perimeter walls of Santa María la Real parish church (Zarautz, northern Spain).

Single- and multiphase models have been used extensively in construction of chronologies. We model more gradual transition between phases with a trapezoid model since it better reflects the nature of the information that goes into the model prior. We find that a simple trapezoid model has a bias that does not reflect prior knowledge, and thus propose an addition of a noninformative element to the prior. We also present an alternative parameterization, which transforms the current abrupt transition model into a model that allows for gradual changes. The addition of a noninformative prior ensures model flexibility. We evaluate these Bayesian models using 2 case studies.