Assessment of the controls on (²³⁴U/²³⁸U) activity ratios recorded in detrital lacustrine sediments

Abstract

Understanding landscape evolution on geological time scales has become increasingly important in the light of recent climate warming and intensified land use. Uranium isotopes ((²³⁴U/²³⁸U) activity ratios) can be used to reconstruct catchment-wide erosion from fine-grained detrital matter, as the lighter nuclide ²³⁴U is lost from grains <63 μm into surrounding pore space in a weathering profile, during transportation, and after final deposition. Thus, (²³⁴U/²³⁸U) activity ratios are a measure for the time elapsed since comminution of bedrock into detrital sediment. When applied to detrital grains in sedimentary deposits, the comminution age allows derivation of past variations in sediment residence time, i.e. the amount of time for which sediments are stored on hillslopes and transported.

Loss of ²³⁴U is mainly attributed to recoil of ²³⁴Th during α-decay of ²³⁸U. However, mobilisation of ²³⁴U is also controlled by preferential oxidation and leaching of ²³⁴U from detrital grains. Further considerations are required concerning the sediment mineralogy and uranium bound to authigenic or endogenic matter.

Here, we extend the discussion about the controls on previously published (²³⁴U/²³⁸U) activity ratios and uranium concentrations of detrital matter along a 5.43 m-long, Late Glacial to Holocene sedimentary record from Lake Ohrid (North Macedonia, Albania). Lake Ohrid is chosen to study how erosion responds to anthropogenic disturbances and climate variability, which are common through the Late Pleistocene and Holocene in the Mediterranean Region. We compared uranium concentration and isotope data to redox-sensitive, mineralogical and biogeochemical proxies. Mineralogical and biogeochemical data show no control on uranium concentration and (²³⁴U/²³⁸U) activity ratios. Rock magnetic redox proxies indicate uranium mobilisation from detrital matter in more oxic environments. No correlation between (²³⁴U/²³⁸U) activity ratios and rock magnetic data is observed, which implies that the redox environment has no control on the isotope signal. Monte Carlo simulations reveal that post-depositional preferential mobilisation of ²³⁴U has only negligible impact on the calculated sediment residence times. Our model implies that pre-depositional leaching can results in shorter sediment residence times but low (²³⁴U/²³⁸U) activity ratios during cold and dry intervals imply that recoil is likely the main processes explaining loss of ²³⁴U from the detrital grain. This is explained by the time scales studied herein in (>10,000 years), during which preferential leaching is seen to be not significant.

Introduction

Global greenhouse gas warming controlled erosion processes are predicted to affect 61% of the ice-free landmasses during the 21th century, a number that increases to 80% when land use is considered (Ostberg et al., 2018). A better understanding of landscape evolution is thus crucial to sustain one of our most important resources, which are soils. Uranium isotope compositions (²³⁴U and ²³⁸U) of fine-grained detrital matter are increasingly used to reconstruct erosion processes and landscape evolution on geological time scales (e.g. Dosseto et al., 2010; DePaolo et al., 2012; Dosseto and Schaller, 2016; Rothacker et al., 2018; Francke et al., 2019). The underlying geochemical conceptual model is based on on-going depletion of the daughter nuclide ²³⁴U from fine-grained detrital matter in the weathering profile, during transportation, temporary storage, and after final deposition (DePaolo et al., 2006; Maher et al., 2006). Uranium-234 becomes depleted from detrital matter by (a) direct recoil of ²³⁴Th, an intermediate product between ²³⁸U and ²³⁴U, (b) preferential leaching of ²³⁴U embedded in the recoil tracks, and (c) biotic or abiotic preferential oxidation of ²³⁴U, which can subsequently be leached from the grains (DePaolo et al., 2006; Dosseto et al., 2008; Suresh et al., 2013; Suresh et al., 2014a; Suresh et al., 2014b; Dosseto, 2015; Dosseto and Schaller, 2016). The depletion compared to the expected secular equilibrium (²³⁴U/²³⁸U = 1) in closed systems such as bedrock becomes significant in small detrital grains (usually <63 μm) with a large surface area to volume ratio. Mechanisms (b) and (c) are likely to be too insignificant to act as major controls on (²³⁴U/²³⁸U) activity ratios, but more evidence from paleoclimate and geological archives is still needed. In concert with quantitative estimates of the sample surface area (BET analyses), uranium isotope analyses can be used to calculated the comminution (DePaolo et al., 2006) or weathering ages (Dosseto and Schaller, 2016), defined as the time since the formation of detrital grains <63 μm by chemical or physical weathering. The palaeo-sediment residence time in the catchment equals the time between comminution and final deposition in a sedimentary basin (Francke et al., 2019, Fig. 1).

The application of the method requires some considerations concerning the geochemical behaviour of uranium and its nuclides in the catchment and after final deposition. Firstly, mobilisation of ²³⁴U from detrital matter causes surrounding pore fluids and freshwaters to become increasingly enriched in ²³⁴U (Plater et al., 1992; Robinson et al., 2004). Consequently, authigenic, endogenic, and organic matter is usually characterised by (²³⁴U/²³⁸U) activity ratios ≫ 1, which can mask the ²³⁴U depletion related to comminution if present in significant amounts. Thus, it has to be removed carefully from the bulk sediment, which is generally performed using chemical treatment methods (Lee, 2009; Martin et al., 2015; Francke et al., 2018). For instance, Francke et al. (2018) developed a protocol optimised to remove organic and authigenic phases without affecting the surface of detrital minerals, which retain the ²³⁴U depletion used to derive a comminution age. Secondly, the mobility of uranium isotopes can be redox-sensitive and the oxidised daughter nuclide ²³⁴U could be leached preferentially from the detrital grains leading to lower (²³⁴U/²³⁸U) activity ratios (Dosseto and Schaller, 2016; Kolodny et al., 2017). While uranium is too heavy for mass-dependent fractionation to be significant (e.g. Steiger and Jäger, 1977), there is increasing evidence that biotic or abiotic redox processes can cause mass-independent fractionation of ²³⁸U, ²³⁵U, and ²³⁴U (e.g. Fujii et al., 1989; Bigeleisen, 1996). In this case, ²³⁸U is preferentially reduced over lighter isotopes ²³⁵U and ²³⁴U, which could increase the (²³⁴U/²³⁸U) activity ratios of sediments. While the magnitude of mass-dependent related fractionation (<0.01%) is unlikely to be significant on ²³⁴U/²³⁸U isotope compositions, which fractionate at the percent level, the effect of redox-controlled mass-independent fractionation on ²³⁴U/²³⁸U remains to be tested (Cumberland et al., 2016; Kolodny et al., 2017). Furthermore, ²³⁴U can be preferentially oxidised and is thus more mobile than ²³⁸U as a result of ‘recoil-induced oxidation’ (Adloff and Roessler, 1991; Dosseto, 2015; Kolodny et al., 2017). This process could consequently decrease the (²³⁴U/²³⁸U) ratio. The impact of ‘recoil-induced oxidation’ and preferential leaching of ²³⁴U is difficult to assess, for example, by comparing to other U isotope systems such as ²³⁵U/²³⁸U, since other nuclei do not show α-recoil. Previous studies conducted on modern stream samples from Northern Australia have shown that preferential leaching of ²³⁴U can indeed bias (²³⁴U/²³⁸U) ratios, but the impact depends strongly on the chosen leaching parameters (Martin et al., 2019). The impact of post-depositional preferential leaching of ²³⁴U on (²³⁴U/²³⁸U) ratios in geological archives has not been studied until today. We consequently need to assess the role of redox processes and preferential mobilisation of ²³⁴U on the ²³⁴U/²³⁸U of detrital sediments to increase the reliability of comminution ages in sedimentary archives.

Herein, we present (²³⁴U/²³⁸U) activity ratios and uranium concentrations in Late Glacial to Holocene (integrating 5.43 m of sediment core) detrital matter from Lake Ohrid (North Macedonia, Albania, Fig. 2) previously published by Francke et al. (2019) to investigate potential impacts of (a) organic matter-dependent U fixation, (b) redox-controls, (c) mineralology, and (d) preferential leaching of ²³⁴U on the (²³⁴U/²³⁸U) activity ratio variability. This is achieved by comparing (²³⁴U/²³⁸U) activity ratios of chemically treated sediments (using the protocol of Francke et al., 2018) to redox-sensitive proxies (rock-magnetic), and biogeochemical data obtained from the same core. We then assess the impact of preferential leaching on estimated sediment residence times by introducing a new formulation that considers preferential ²³⁴U mobilisation prior to and after final deposition separately, and we compare these findings to previously published data by Francke et al. (2019).