洞穴地点骨化石铀系年龄可信度的讨论*

骨化石是铀不平衡系测年法广泛应用而又颇有争议的研究对象。文章通过对典型铀加入模式的计算,指出以α能谱的精度,二种铀系年龄差异的显著性表明近期内有较大量铀的迁移,但其在误差范围内的一致不能保证样品构成封闭体系。以往积累的数据表明,大多数洞穴地点骨化石给出了二法一致的铀系年龄,但钙板与下伏骨化石铀系年龄大多差异显著且与地层顺序矛盾,即使被次生碳酸盐岩包裹,多半骨化石的铀系年龄仍显著偏低于其包裹体。此外,相当一部分样品的U-Th同位素比难以用简单的铀后期加入或淋失来解释。基于对次生碳酸盐岩铀系年龄可信度的认识,骨化石总体上不构成封闭体系,所载铀系年代信息只具有限的分辨率。

DISCUSSIONS ON THE RELIABILITY OF U-SERIES DATING OF FOSSIL BONES FROM KARSTIC CAVES

U-series dating of fossil bones has been widely applied, but its reliability has long been a subject of controversy. With the precision attainable with classical alpha spectrometry (±1σ, 3.0 ％ for ²²⁷ Th/²³⁰ Th, 2.5 ％ for ²³⁰ Th/²³⁴ U, 1.0 ％ for ²³⁴ U/²³⁸ U and hence 4.0 ％ for ²³¹ Pa/²³⁵ U) and according to two frequently quoted open system models, linear U addition and 10 ％ U recent addition, this paper presents two figures comparing the real ages and corresponding apparent ²³⁰ Th/²³⁴ U, ²³¹ Pa/²³⁵ U and ²²⁷ Th/²³⁰ Th ages of a sample. From these comparisons, conclusions can be made that while a statistically significant discrepancy between the dates of two U-series methods attests a relatively recent and important uranium migration, their consistency within error range cannot be cited as a sufficient condition to guarantee the samples closed-system behavior. Compared with open-air sites, limestone caves are generally considered to provide better conditions for preserving fossil bones. In the past two decades, in the course of dating Chinese archaeological sites mainly with intercalated speleothem formations, from 17 cave localities we have accumulated 46 alpha spectrometric ²³⁰ Th/²³⁴ U dates on fossil samples, among which 33 have been analyzed in parallel with ²²⁷ Th/²³⁰ Th. The results demonstrate that a greater part (25 out of 33) of these fossil samples give ages of two U-series methods consistent within ±1.5σ. However, at five sites where the dates on bones can be checked by comparing with overlying flowstone layers, three give substantially younger dates on bones, which is at variance with the depositional order. Fossil bones with calcite coating were considered as preferable materials for dating. However, four out of six samples we have studied give substantially younger dates in comparison with their encasing calcite. Based on the widely accepted reliability of relatively pure and dense speleothem calcites, bones yield mostly underestimated U-series dates in cave settings.Our results show also aberrant U-Th isotopic ratios for quite a few fossil specimens, the abnormality being difficult to account for with simple models of posterior U addition or loss. Besides, as compared with associated speleothem calcite a notable feature of fossil samples is their generally high ²³⁴ U/²³⁸ U ratios, of which no decreasing tendency is discernible with increasing antiquity of sites.To account for all these phenomena, we believe that fossil bones as a whole do not conform to a closed system. Rather uranium may cycle into or out of fossil bones in response to changing environmental conditions, such as pH and Eh of the surrounding pore water, and no known condition can securely protect them from being affected. If the U migration event is relatively recent and sufficiently important, two U-series methods will give discordant results. Otherwise with relatively stable geochemical conditions, a temporary equilibrium may be established, leading to apparently "good" samples with concordant dates of two U-series methods. With only two independent equations from two U-series decay schemes, it is literally impossible to solve precisely the chronological signal controlled by a variety of random parameters. Thus, in spite of its continued usefulness particularly for sites without any other dating material, the U-series dating of fossil bones should be of limited reliability.