U-Pb geochronology of Quaternary pedogenic carbonates and teeth from archaeological sites of Zarqa Valley, Jordan

Dating Quaternary carbonates and phosphates beyond the range of U-series geochronology is challenging, but recent studies have shown potential using the method of LA-ICP-MS U-Pb geochronology. Terrestrial materials altered and produced by pedogenic processes are particularly challenging, as they are usually rich in common Pb and poor in U; however, they provide some of the rare opportunities to directly date important archaeological artefacts. Our aim is to advance the age constraints of the hominin presence in the Zarqa Valley, Jordan, an important geomorphological feature with the oldest hominin artefact outside Africa. Along with new Ar-Ar geochronology of sequence-bounding basalts and new palaeomagnetic constraints, our approach is to: (1) review published U-Pb data of 2019; (2) conduct LA-ICP-MS U-Pb analyses in various pedogenic carbonates (calcrete) that cap the artefact-bearing deposits along with mammoth fossil remains; and (3) review the regional ²³⁴U/²³⁸U activity ratio estimate required for calculation of accurate U-Pb ages.

A previous attempt to date hominin artifacts in the valley yielded an age of 1.98 ± 0.20 Ma for calcrete material, which is a weighted mean of several individual U-Pb isochrons derived from laminated calcretes, cracks, pore spaces and cement, and corrected by an initial ²³⁴U/²³⁸U activity ratio of 1.5 ± 0.5. We calculated an alternative, statistically favourable, age based on pooling the data into a single Tera-Wasserburg regression, resulting in 1.93 ± 0.44 Ma.

Our LA-ICP-MS analysis thus far has not yielded a new significant age, but we have focussed on understanding the U and Pb compositions in a range of materials, including calcretes, rhizoliths, fossil coatings and Mammalian fossils. Biogenic spherulite laminae in laminar calcrete have higher U (0.56 ± 0.064 ppm) and lower Pb contents (0.61 ± 0.15 ppm) compared to abiogenic cracks and cement (0.33 ± 0.084 and 0.9 ± 2 ppm, respectively). Although far from desirable for U-Pb geochronology, for now, the spherulite laminae seems to be the most suitable micromorphology to date calcrete. Coarse spar of rhizoconcretions have very low U contents (0.07 ± 0.35 ppm), leading to unfavourable U/Pb ratios. Two fragments of a mammoth molar from an artefact-bearing deposit have been analysed; one is part of the crown having dentine and enamel, and the other one is dentine of the root. The crown and root fragments have high U concentrations of 33 ± 11 and 13.7 ± 9.9 ppm, respectively, but with their high Pb contents they still yield a low spread in U/Pb ratios and very imprecise lower intercept ages. Using LA-ICP-MS mapping to better identify areas of alteration, we pooled U-Th-Pb spot data from regions with low Y (<5 ppm), Mn (<50 ppm), Al (<100 ppm), Th (< 0.05 ppm) and high P (> 7200 ppm), which correlated with the central portion of the dentine. These data yielded an age, using the combined U-Pb and Th-Pb calculation, of 1.9 ± 1.1 Ma; although very imprecise, it shows potential for further refinement.