Combined U-series dating of cave pearls and mammal fossils: Constraint on the age of a late middle pleistocene Ailuropoda–Stegodon fauna from the Diaozhongyan Cave, Guangxi, South China

Abstract

To extend the range of materials that can be used to provide chronological constraint in archaeological and palaeontological settings, we explore the use of cave pearls that form in shallow pools. Here, we present U-series dating results of cave pearls found in cave sediments with mammal fossils collected from systematic excavations at the Diaozhongyan Cave (“DZY”), located in Guangxi Zhuang Autonomous Region, southern China. Similar to soda straw stalactites, cave pearls can be transported after formation and subsequently incorporated into sediments. Thus, the U-series ages of the outermost layers of cave pearls are expected to be older than their host sediments and associated archaeological and paleontological materials. In conjunction with U-series dating of mammal teeth, which can provide a reliable minimum age, U-series ages of cave pearls can provide a maximum age limit of the deposits. Thus, dating of the mammal teeth and cave pearls from DZY reasonably constrains the first occurrence of the DZY Ailuropoda-Stegodon fauna to a period between 205.6 ± 1.4 ka and 231.0 ± 12.8 ka. These age constraints agree well with previous estimates for the time range of the late Middle Pleistocene Ailuropoda-Stegodon fauna in southern China. We propose that greater attention should be paid to cave pearls, which are commonly found in cave deposits that were laid down at least in part due to fluvial activity, and can be used for U-series dating to better constrain the age of the associated cave deposits.

Introduction

Building reliable chronologies for vertebrate paleontological assemblages across time and space are critical to understand regional and global aspects of Quaternary mammal evolution. One such method, uranium-series (or U-series) dating, has been widely applied to establish a general chronological framework for Quaternary mammalian evolution (e.g., Shen et al., 2002; Grün et al., 2008, 2010; Price et al., 2011, Price et al., 2013; Michel et al., 2016; Westaway et al., 2017; Hershkovitz et al., 2018). The main advantage of this method is that it can provide highly precise and accurate absolute age determinations from 0 to 500 ka by utilizing the radioactive decay chain of ²³⁸U–²³⁴U–²³⁰Th isotopes (Edwards et al., 1987). Speleothems, such as layered flowstones, calcite veins and fillings, in situ stalagmites, coralloid speleothems (“cave popcorn”), and soda straw stalactites are the most commonly used materials for U-series dating (e.g. Zhao et al., 2001; St Pierre and Zhao, 2009, St Pierre and Zhao, 2012, St Pierre et al., 2013; Liu et al., 2010; Ibrahim et al., 2013; Aubert et al., 2014, 2017; 2018; Bae et al., 2014; Liu et al., 2015; Sutikna et al., 2016). When used in combination, it is possible to derive minimum and/or maximum age estimates for the fossil assemblages in stratigraphic contexts.

Limitations for using the U-series dating method in archaeological sites do exist, however For instance, materials appropriate for dating must be present and their stratigraphic association to the fossil assemblages should be understood. Furthermore, direct U-series dating of bones and teeth commonly yields what are thought to be minimum age estimates that are younger than the true ages of the fossils (e.g. Pike et al., 2002; Sambridge et al., 2012; Grün et al., 2014; Louys et al., 2016; Van den Bergh et al., 2016; Bartsiokas et al., 2017). Fresh bones and teeth contain little to no U; they scavenge U effectively, but exclude Th, from the burial environment during the early diagenesis period. After that, the radioactive decay of ²³⁸U starts to produce ²³⁰Th. As such, the U-series ages of bones and teeth represent their buried ages, and are generally younger than their true ages (Pike et al., 2002; Sambridge et al., 2012; Grün et al., 2014; Louys et al., 2016). It is important to find alternative materials to provide reliable dates from cave sites using U-series. One such material that has not received much attention is “cave pearls”.

Cave pearls are spheroidal or subprolate grains which are formed by concentric growth of calcium carbonate layers around a nucleus (e.g. Gradziñski and Radomski, 1967; Donahue, 1969; Nader, 2007; Jones, 2009; Melim and Spilde, 2018). In modern karst cave systems, they usually form in two types of settings: low energy shallow pools bordered by rimstone dams and high energy shallow splash pools (Jones and Kahle, 1986; Jones and MacDonald, 1989; Hill et al., 1997; Melim and Spilde, 2011, 2018). The rate of growth of their cortical laminae with apparent different colors are controlled by the episodic influxes of variable water flows that result from fluctuations in water levels, water chemistry, and terrigenous detrital materials that are present in the pools (Jones, 2009; Melim; Spilde, 2018). In certain instances, cave pearls may move from the pools by extremely powerful water activity and incorporated into cave sediments. In cases like these, cave pearls can act as a chronological marker for stratified cave deposits. Cave pearls appear to have received little attention in earlier geochronological studies. Here, we present a pioneering study of the U-series ages of cave pearls from stratified deposits in the Diaozhongyan (“DZY”) Cave, Guangxi Zhuang Autonomous Region, southern China (Fig. 1). The aim of this study is not only to determine the age of the DZY Ailuropoda–Stegodon fauna, but also to demonstrate the utility of a combination of U-series dating of mammal fossils and cave pearls to constrain the timing of a mammal fauna from cave deposits.