Diagenetic uptake of rare earth elements by bioapatite, with an example from Lower Triassic conodonts of South China

Abstract

The distribution patterns of rare earth elements (REEs) are frequently used as proxies for ancient seawater chemistry or paleomarine environmental conditions. However, recent work has shown that diagenesis can lead to remobilization and inter-elemental fractionation of REEs, and that these effects often occur in conjunction with redox reactions in sediment porewaters. Here, we review existing literature on the diagenetic fluxes of REEs in marine sediments and porewaters in order to systematize existing knowledge on this subject. REEs undergo significant redistribution among sediment phases during both early and late diagenesis as a consequence of adsorption and desorption processes. Remobilization of REEs commonly leads to inter-elemental fractionation, variously leading to enrichment or depletion of the light, middle, or heavy REE fractions. Further, REE remobilization can be facilitated by redox changes, e.g., through reductive dissolution of host phases in suboxic and anoxic porewaters. Characteristic REE distribution patterns develop through these processes: (1) a ‘flat distribution’ signifying predominantly terrigenous siliciclastic influence, (2) a ‘middle-REE bulge’ probably due to adsorption of light and heavy REEs to Mn- and Fe-oxyhydroxides, respectively, and (3) ‘heavy-REE enrichment’ indicative of hydrogenous (seawater) influence (note: all patterns in this paper are normalized to the REE composition of average upper continental crust, or UCC).In the second part of this study, we undertake an analysis of the REE distributions in conodonts and whole-rock samples from West Pingdingshan, a Permian–Triassic boundary section in South China. Using ΣREE/Th and Y/Ho ratios, we show that almost all of the conodont samples have a strong diagenetic overprint, and that the hydrogenous REE fraction is small and not isolatable. Furthermore, the conodonts contain two diagenetic REE components, one characterized by low ΣREE (100–300ppm), high ΣREE/Th ratios (>1000), strong middle REE enrichment, and Eu/Eu* ratios of ~1.5–2.0, and the second by high ΣREE (300–2000ppm), low ΣREE/Th ratios (~20–30), little or no middle REE enrichment, and Eu/Eu* ratios of ~1.0. The first component exhibits a pronounced middle-REE bulge that represents an early diagenetic signature associated with suboxic conditions, possibly related to adsorption of REEs onto Fe–Mn oxyhydroxides in the shallow subsurface environment. The second component shows a flat REE distribution that is similar to both our whole-rock samples and average UCC, indicating derivation from REEs released from detrital siliciclastics (e.g., clay minerals), probably at a range of burial depths from shallow to deep. Failure of the conodont samples to yield an isolatable hydrogenous component demonstrates that bioapatite does not always preserve a primary marine REE signature. Given that bioapatite REEs have been widely used for this purpose, often on the assumption of minimal or no diagenetic influence, our findings are likely to necessitate a re-evaluation of the results of many earlier studies. In general, we counsel caution in inferring a hydrogenous origin for REEs in bioapatite owing to frequent diagenetic alteration of REE distributions.