Chronology of the shell bar section and a discussion on the ages of the Late Pleistocene lacustrine deposits in the paleolake Qarhan, Qaidam basin

Abstract

Shell bar, composed of abundant fossil shells of Corbicula fluminea muller and Corbicula largillierti philippi and located at the southeastern end (36°30′N, 96°12′E) of the paleolake Quarhan, is one of the most prominent features in the Qaidam basin. It is the highest site where such species of fossil shells have been found in the Late Pleistocene age. A 2.6-m-thick fresh profile was manually excavated to determine the formation ages and the scope of the high paleolake levels. Accelerator mass spectrometry (AMS), conventional radiocarbon dating, and sector inductively coupled plasma-mass spectroscopy 230Th methods were used to investigate the reliability and accuracy of dating results in different laboratories. Ages of various components (e.g. acid residual and acid soluble fraction of the organic matter) from the same sample were determined. Age differences of a variety of materials (e.g. organic matter, fossil shells, and salt crystals), and age differences at the same sampling position were evaluated. Dating shows that the AMS ages given by alkali residue and acid soluble fractions vary greatly, e.g., from 124.5 cm upwards. The dates given by the acid soluble fraction were normal. Down the column, the dates show a reverse pattern and those of alkali residues, especially the lower part of the section, show an unstable pattern. These imply that organic matter had been influenced by two separate processes. One possible explanation is that the alkali residual fraction most likely had been contaminated by dead carbon-bearing reworked material because the study section is located near the edge of the paleolake and could have been easily influenced by old eroded deposits. Another is that the acid soluble fraction could have been contaminated by upward-flowing groundwater containing soluble organic matter. The amount of this soluble organic material should be very small because the strongest age reversal is in the lower part, where the TOC content remains low, meaning that the change of soluble organic fraction alters neither TOC content nor the δ13C dramatically. It is concluded that a uniform mega-paleolake developed in the Qaidam basin in the northeastern Tibetan plateau between 39.7 and 17.5 14C kaB.P. During the period when the high paleolake level of Qarhan was formed, the huge paleolake covered a vast area with dramatic lake level fluctuations. It is found that the ages determined may be influenced by either radioactive 14C variations of repeated deposition or content variations in the atmosphere. The ages given by fossil shells are 15–18 ka older than those given by organic matter. The large differences between the ages of fossil shells and organic matter might have resulted from the large water areas and huge water volume and the special location of the high-elevation Qaidam basin. This study also shows the reliability and accuracy of the 230Th dating method on the salt crystals but further study is needed to determine whether this method could be applied to the study area.