“Cryptic” diagenesis and its implications for speleothem geochronologies

Abstract

Speleothems are usually considered as one of the most amenable palaeoclimate archives for U-series dating. A number of studies in recent years, however, report cases of diagenetic alteration which compromises the use of U-series systematics in speleothems, resulting in inaccurate U-Th ages. Here we present the results of a high-resolution U-Th dating study of a stalagmite (CC26) from Corchia Cave in Italy where we document a number of departures from an otherwise well-defined age-depth model, and explore potential causes for these outliers. Unlike examples illustrated in previous studies, CC26 contains no visible evidence of neomorphism, and appears, at least superficially, ideally suited to dating. Good reproducibility obtained between multi-aliquot U-Th analyses removes any possibility of analytical issues contributing to these outliers. Furthermore, replicate analyses of samples from the same stratigraphic layer yielded ages in stratigraphic sequence, implying very localized open-system behavior. Uranium loss is suggested as a causative mechanism on account of the fact that all the outliers are older than their assumed true age. A limited number of micro-voids were observed under micro-CT analyses, and it is proposed that these were pathways for U loss. Uranium-loss modelling allows us to constrain the possible timing of diagenetic alteration and indicates that the precursor for the outlier with the largest age discrepancy (309%) must have been aragonite. This study indicates that visibly unaltered speleothems may still contain small domains that have experienced post-depositional alteration. Such “cryptic” diagenesis, as recorded in this stalagmite, has implications for the constancy of accuracy of the U-series dating technique, and suggests a need for careful examination of speleothems prior to dating, particularly in low-resolution U-Th studies

What do we really know about early diagenesis of non-marine carbonates?

“…Furthermore, even when no visible arguments are found for diagenetic alteration Ritter et al, 2015), for example, when replacement calcite inherits both textural and chemical properties of the precursor; very localized, open-system behavior may still cause resetting of an initial geochemical signal (Bajo et al, 2016). It is then only by a careful comparison of different geochemical proxies along the same transect that localized domains of post-depositional alteration can be identified (Bajo et al, 2016).…”

“…It is then only by a careful comparison of different geochemical proxies along the same transect that localized domains of post-depositional alteration can be identified (Bajo et al, 2016).…”

“…The high primary porosity allows the infiltration of waters that, at the surface, may still be involved in precipitation, but could change composition and initiate early diagenesis below (Frisia et al, 2002;Chafetz and Guidry, 2003). But also, very local scale, discontinuities (Bajo et al, 2016;Domínguez-Villar et al, 2017) and secondary porosity, like fractures, brecciation or desiccation cracks, can play an important role and provide efficient channels for infiltration, percolation, ascending fluids and lateral fluid circulation (Deitchman and Loheide, 2009;Frisia and Borsato, 2010;Gandin and Capezzuoli, 2014;Inverarity et al, 2016).…”

“…As a result, within the same structure, like for example a tufa build-up, microscopic textures may become dissolved and re-arranged in one place whereas in another the texture remains more or less unchanged (Golubić et al, 2008;Martín-García et al, 2009;Melim and Spilde, 2011;Bajo et al, 2016). In contrast, where clay minerals (e.g.…”

“…However, if fabric changes happen continuously or intermittently over a longer time, implying that the system has remained open, then a serious overprinting of paleoclimatic and time series information can be expected. Such open system behavior may not only affect -and possibly homogenize -stable isotopic carbonate signatures, but it may also be responsible for observed U/Th offsets and age reversals (Scholz et al, 2014;Claes et al, 2015;Bajo et al, 2016).…”