Groundwater is an essential environmental resource, but one which is vulnerable to nitrate contamination from fertilizers, wastewater, and atmospheric pollution. Nitrogen in excess of ecosystem requirements is leached from soils into groundwater and stored on a range of time scales to create a ‘legacy’ of contamination for future generations^[1]. This is particularly the case in karst terrain where thin soils and dual permeability of the carbonate bedrock make them acutely vulnerable to water pollution^[2]. Over a quarter of the World’s population rely on groundwater contained within karst aquifers for drinking water supplies^[3]. Contamination with nitrogen is therefore causing major human and environmental health issues of global proportions^[2,4]. However, measurements of groundwater contamination in karst and quantification of storage times within aquifers is limited primarily to 21^st century observations. This means there is no knowledge of natural variability in groundwater nitrogen concentrations, and no information on pollution source or storage dynamics throughout the era of industrialization. The nitrogen content and its isotopic ratios in speleothem calcite provides a means through which these historical data can be obtained. Here, we present a two-year study investigating nitrogen biogeochemical cycling within a cave karst system located beneath pasture-land in northern Spain. We use concentration dynamics and stable isotopes in drip waters to inform nitrate source characteristics, biogeochemical processing and water routing through the karst system. We have also developed a methodology to measure stable isotope compositions in nitrate extracted from speleothem calcite. These data demonstrate the feasibility of using cave environments to study vadose zone nitrogen biogeochemical cycling, and the suitability of using speleothem carbonate for tracing nitrogen sources. Cave dripwater isotope values collected from Cueva Llanio, northern Spain, demonstrate an enrichment in δ¹⁵N-NO₃ from that observed in the local rainfall, due to the influence of nitrogen sourced from surface manure application. Based on a distinct change in δ¹⁸O-NO₃ values between rainfall and drip water, processes of nitrification are seen to dominate the biogeochemical cycle. Subtle changes in isotope values between cave chambers reflect closely the spatial changes in surface land use. Modern speleothem carbonate samples collected on glass plates over a period of 6 months beneath active drips contain nitrate at concentrations ranging between 2.5-10 ppm (NO₃-N). The nitrate isotopic composition of this speleothem carbonate directly reflects the composition of the drip waters, with no apparent fractionation during carbonate precipitation. Whilst a timeseries of nitrate dynamics has yet to be extracted from the speleothem record, these data present significant potential to obtain comprehensive records of groundwater nitrate loading and sourcing from stalagmite records. This is a major issue for policy makers and environmental legislators who rely on this information for implementation of policies to control groundwater pollution.

[1]Hansen, B. et al., Sci Rep, 7, 8566 (2017). [2] Ascott, M.J. et al. Nat Commun, 8 (2017). [3] Ford, D.C and Williams, P. Karst Hydrogeology and geomorphology. Wiley (2013). [4] Exner, M. E. et al. Water Resour Res, 50, 4474-4489 (2014).

Common speleothem-based paleoclimate proxies such as d18O and trace element ratios are capable of high temporal resolution but may be influenced by multiple climatic forcings that may result in ambiguity in their interpretation. The concentration of ferrimagnetic, Fe-oxide particles in speleothem laminae has been recognized as a potential tracer for local precipitation that can complement d18O data. However, the mechanism by which magnetic particles are incorporated appears to vary based on locality, motivating research to identify these enrichment processes before speleothem magnetism can be used as a reliable proxy. Further, the strength of correlation between magnetism and rainfall has not been quantified using historical rainfall measurements, in part due to the coarse temporal resolution of traditional rock magnetic sensors.

We use the recently developed quantum diamond microscope (QDM) to magnetically image both the central column and flank laminae of a speleothem with an imparted saturation isothermal remnant magnetization to map the distribution and properties of ferrimagnetic particles. The speleothem, collected in the Onça cave from Peruaçu National Park, Brazil and dated using U-Th, has an average deposition rate of 1.05 mm per year, allowing the QDM to image magnetism at sub-annual resolution. Following QDM imaging, we integrated the magnetic field intensity along laminae and binned at annual resolution to construct a relatively low-noise time series that can be compared to the local instrumental rainfall record available from 1913.

Our results demonstrate a statistically robust negative correlation (R2>0.33, P<1e-9) between local rainfall amount and the magnetic particle concentration in both the speleothem central column and flank. The negative correlation suggests that enhanced soil erosion due to vegetation retreat during dry intervals is the likely mechanism of magnetic particle enrichment, similar to interpretations in magnetic studies of other tropical speleothems. We further examined oxygen and carbon isotope compositions from a speleothem collected in a well-ventilated room from the same cave where variations in temperature and humidity drive significant kinetic isotope effects. Both central column and flank magnetism show a stronger correlation with rainfall compared to d18O or d13C, consistent with kinetic fractionation affecting these isotopic compositions. This comparison points to speleothem magnetism as a potential alternative for reconstructing precipitation in environments lacking isotopic equilibrium conditions.

Finally, our demagnetization sequence of a laboratory anhysteretic remanent magnetization permitted fingerprinting of the sources of ferrimagnetic grains based on grain populations characterized in previous environmental magnetism studies. We identified two populations of ferrimagnetic grains in the Onça speleothem, one associated with fine pedogenic magnetite and another with coarser detrital particles. These populations may imply multiple mechanisms of grain deposition in the speleothems, including drip water and aeolian transport. The existence of a correlation between magnetization and precipitation as reported above suggests that a single mechanism is dominant or that multiple depositional pathways covary to precipitation.

Based off these results, high-resolution magnetic imaging along with precise U-Th dating presents an opportunity to reconstruct local precipitation rates at annual intervals, complementing established proxies sensitive to a combination of precipitation and other effects.

In the Cradle of Humankind (Cradle) UNESCO World Heritage Site, South Africa, the chronology of calcite-aragonite speleothems is used for determining ages of hominin fossils preserved in Plio-Pleistocene clastic sediments deposited in dolomitic caves. The speleothems in the Cradle are mostly dated using either the U/Th disequilibrium or U-Pb dating methods [1,2]. While advancements in multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) has extended the U/Th dating range to 700 ka [3], the U-Pb dating method has been an essential tool, as many fossil occurrences are older than 1.5 Ma. However, speleothems suitable for U-Pb dating should have at least 1 ppm U and very low non-radiogenic Pb content (i.e., should not contain detrital matter) [2]. Here we show that the uranium-thorium-helium ((U,Th)-He) dating method combined with U/Th disequilibrium dating offers a promising alternative to U-Pb dating for speleothems with ages beyond the U/Th dating limit [4]. We establish three different methods of calculating (U,Th)-He ages as well as an additional method for calculating U/Th ages. Using calcite-aragonite speleothems, we reproduced within analytical uncertainty, the published U/Th or U-Pb ages for three drill core samples from the Sterkfontein cave in the range 2000-3000 ka, a hand sample from a speleothem deposit now exposed on surface at Swartkrans cave with an age of 1800 ka, and five underground hand samples from the Rising Star cave with ages in the range 100-800 ka. Subsamples that demonstrated loss of He or poor He retention did not yield reliable results, whereas subsamples with U concentrations as low as 0.1 ppm, severe diagenetic fabrics, and or contained detrital material were still able to produce reliable results. Most subsamples from the Rising Star cave yielded unexpected high initial (²³⁰Th/²³⁸U) activity ratios up to 19.7. These subsamples were not dateable by U/Th [1], without the additional input from He analysis because the finding of high initial (²³⁰Th/²³⁸U) activity ratios is in sharp contrast to the common assumption that calcite speleothems should have low initial (²³⁰Th/²³⁸U) activity ratios if their Th/U ratios are low. The high initial probably due to incorporation of Fe–Mn oxides-hydroxides dust, on which ²³⁰Th was previously adsorbed. This hypothesis is supported by high initial (²³⁰Th/²³⁸U) activity ratios found in soil Fe-oxide concretions from the eastern part of the Great Escarpment, South Africa, which were also dated by this method [5]. If not detected and corrected for, such high initial (²³⁰Th/²³⁸U) values can lead to inaccurate U/Th and U-Pb ages of speleothems. Our results show that the incorporation of He analysis in U/Th dating makes the (U,Th)-He dating method a viable potential alternative for dating speleothems where the application of U/Th or U-Pb is not possible.

References:

[1] Robbins et al., 2021. Chemical Geology, 567, 120108.

[2] Pickering & Edwards, 2021. Chemical Geology, 579, 120364.

[3] Cheng et al., 2013. Earth and Planetary Science Letters, 371, 82-91.

[4] Makhubela & Kramers, 2022. Quaternary Geochronology, 67, 101234.

[5] Makhubela et al., 2021, Chemical Geology, 580, 120368.

The world-wide unique Laser Ablation (LA) setup[1] coupled to the MICADAS Accelerator Mass Spectrometer (AMS) at ETH has opened unprecedented possibilities regarding the radiocarbon analysis of carbonate archives such as speleothems or shells. Such archives can be measured with relatively high resolutions, e.g. 125 (100) μm with a measurement speed of 33 (83) min/cm. This novel method allows producing nearly continuous radiocarbon records. Successful application has been proven for different types of carbonates, e.g. otoliths[2] and speleothems[3].

A continuous carbon flow of approximately 2.5 μg/min needs to be produced by the laser to ensure stable measurement conditions in the AMS. Accordingly, CaCO₃ is ablated at comparably high repetition rates of 200 Hz, which in turn requires a continuous movement of the sample relative to the laser beam through a positioning system (typical velocities: 5 – 20 µm/s). Single data points are produced at 10 sec intervals resulting in a dense dataset with well localised relative coordinates.

To make full use of the data sets, proper data processing is paramount. This includes (1) matching of the ¹⁴C data to individual growth layers, (2) signal extraction and noise reduction, and (3) correction for mixing and washout effects. Here, we show the current status of data processing improvements, especially the use of a Savitzky-Golay filter[4] in dealing with noisy data.

References

[1] C. Welte et al., ‘Novel laser ablation sampling device for the rapid radiocarbon analysis of carbonate samples by accelerator mass spectrometry’, Radiocarbon, vol. 58, no. 2, pp. 419–435, Feb. 2016, doi: 10.1017/rdc.2016.6.

[2] A. H. Andrews, C. Yeman, C. Welte, B. Hattendorf, L. Wacker, and M. Christl, ‘Laser ablation–accelerator mass spectrometry reveals complete bomb 14C signal in an otolith with confirmation of 60-year longevity for red snapper (Lutjanus campechanus)’, Marine and Freshwater Research, 2019, doi: 10.1071/mf18265.

[3] C. Welte et al., ‘Climatic variations during the Holocene inferred from radiocarbon and stable carbon isotopes in speleothems from a high-alpine cave’, Clim. Past, vol. 17, no. 5, pp. 2165–2177, Oct. 2021, doi: 10.5194/cp-17-2165-2021.

[4] A. Savitzky and M. J. E. Golay, ‘Smoothing and Differentiation of Data by Simplified Least Squares Procedures.’, Analytical Chemistry, vol. 36, no. 8, pp. 1627–1639, Jul. 1964, doi: 10.1021/ac60214a047.

In recent decades speleothems have become valuable climate archives and are of crucial importance for the investigation of past climate conditions. Albeit much progress has been made to understand speleothem proxies, it remains difficult to differentiate between a direct climate signal and variations, which occurred due to in-cave processes like prior carbonate precipitation (PCP), CO₂ degassing or exchange between dissolved inorganic carbon and cave air CO₂.
Here, we analyse the stable C and O isotopic composition as well as growth rate of contemporaneously growing speleothems from the same cave site. We argue that differences in those paleoclimate proxies arise mainly by differences of drip site specific conditions as climate conditions for pairs of contemporaneously growing speleothems can be considered to be similar. To better understand the observed differences in the isotopic composition and growth rate of contemporaneously growing speleothems, we investigate the in-cave processes by applying a speleothem isotope and growth model. The model is based on a Rayleigh process, which includes CO₂ degassing and CaCO₃ precipitation, oxygen and carbon buffering of the bicarbonate through water and cave air CO₂ and is able to calculate growth rates. The model accounts for CaCO₃ deposition in form of prior carbonate precipitation as well as CaCO₃ deposition at the speleothem top.
We find that CO₂ exchange processes and PCP are necessary to explain the observed isotopic and growth rate differences in contemporaneously growing speleothems. Thus, both processes have to be accounted for in speleothem-based paleoclimate reconstructions.

The speciation and mobility of trace metals in aquatic environments are tightly controlled by the interactions of metals with ligands present in natural organic matter. Under the slightly alkaline conditions typical of karst cave waters, organic ligands typically bind >90% of the total pool of Cu, Co, and Ni, leaving a marginal labile fraction consisting of hydrated ‘free’ cations and simple inorganic metal complexes. In consequence, organic metal complexes (OMCs) effectively control transition metal transport from the surface to caves (Hartland and Zitoun, 2018), where infiltrating waters deposit speleothems.

The distribution of trace elements between solution and crystal phases is conventionally described in terms of partition coefficients (k_d), estimated as the ratio of M/Ca_solid and M/Ca_solution. For alkaline earth metals (e.g., Mg, Sr, Ba), partitioning occurs at comparably predictable rates that favour the utilisation of these elements as palaeoenvironmental proxies. However, the kinetic inhibition of transition metals renders their apparent partitioning more ambiguous, as the supply of these metals for inclusion in the crystal phase is moderated by ligand action. The decay, or dissociation, of OMCs is thus expected to ultimately determine respective metal concentrations in the speleothem. Given that the residence time of the thin water film on top of a stalagmite is essentially governed by the impact of incoming drips, a mechanistic understanding of OMC dissociation kinetics offers an opportunity to quantitatively relate stalagmite metal concentrations to the drip rate.

In this study, we aim to improve our understanding of the partitioning pathways of selected transition metals (Ni, Co, Cu) with view to utilising the decay characteristics of OMCs for the reconstruction of past drip rates. We present results from cave-analogue and competitive ligand exchange experiments aimed at assessing elemental partitioning and OMC kinetics in synthetic solutions, as well as water samples from various New Zealand caves. Our study demonstrates how organic ligands regulate transition metal partitioning from dripwaters to speleothems, and provides first quantitative estimates of OMC decay rates that determine the supply of transition metals for inclusion in carbonate deposits.

References:

Hartland, A., Zitoun, R. (2018) Transition metal availability to speleothems controlled by organic binding ligands. Geochem. Persp. Let. 8, 22–25.