Assessing the potential of stable calcium isotope ratios for paleotemperature reconstruction from studies of laboratory-precipitated inorganic carbonates and modern corals, fish otoliths and foraminifera

Abstract

Determining paleo-seawater composition and temperature is critical for reconstructing Earth’s surface conditions through time and has major implications for understanding the origin and evolution of life on Earth. Paleo-seawater composition is inferred from the chemical and iso-topic compositions of biogenic and inorganic chemical precipitates that form out of seawater. Some of the widely used geochemical and isotopic tracers that have been measured in marine carbonates for determining the physical and chemical attributes (e.g., temperature, pH, salinity) of the past oceans include Sr/Ca, Mg/Ca, Li/Ca, Ba/Ca, 18O, 47, 11B, etc. However, each geochemical and isotopic proxy used for determining paleo-seawater temperature has its limi-tations and requires addressing four main issues: (i) inter-species variability in biogenic car-bonates, (ii) variation in past seawater composition, (iii) post-depositional alteration of sam-ples, and (iv) temperature-calibration of the proxy. Hence, an important component of paleo-climate research involves the development of the robust geochemical and isotopic proxies which are minimally affected by factors (e.g., pH, growth rate, vital effect, etc.) other than temperature. Recent improvements in mass spectrometry have resulted in high-precision Ca isotopic measurements (44/40Ca) which allows us to investigate the applicability of 44/40Ca as a paleotemperature proxy. Due to the high abundance of Ca in inorganic and biogenic car-bonates, the 44/40Ca value is expected to be unaffected by the diagenetic alteration of geolog-ical archives which makes the 44/40Ca composition of carbonates a powerful tool for paleocli-mate research. In this thesis, Ca isotopic measurements were performed on various inorganic and biogenic carbonates along with measurements of other geochemical (Sr/Ca, Mg/Ca, Li/Ca, Ba/Ca, B/Ca) and traditional isotopic proxies (18O, 13C) to understand the efficacy of these different viii proxies in reconstructing water temperatures. A novel sample-loading technique, involving a Re-Ta double filament assembly along with the use of tantalum oxide as an activator, was developed for Ca isotopic measurements using thermal ionization mass spectrometry (TIMS). Using this loading technique, the analytical reproducibility for 44/40Ca was better than 0.08‰ (2SD). Elemental ratios in selected biogenic carbonates were measured using a quadrupole inductively coupled plasma mass spectrometer (ICPMS). Using a cold-plasma technique and bracketing using synthetic and international carbonate standards, high precision data were obtained. For example, the external precision was better than ± 0.03 (SD) for Sr/Ca (mmol/mol) and ± 0.025 (SD) for Mg/Ca (mmol/mol). To understand the effect of temperature and the roles of pH and precipitation rate on Ca isotopic composition of inorganic carbonates, calcite precipitation experiments were conducted in a controlled-laboratory environment at different temperatures (5, 10, 20, 30, 40 and 50 °C). At each temperature, the effects of precipitation rate, pH and calcite dissolution rate on Ca isotopic composition of the precipitated carbonate were investigated. It was observed that the calcite dissolution rate is a function of temperature; at low temperatures (5-10 °C), the Ca isotopic fractionation was maximum (44/40Ca = -0.84 to -1.07‰; 44/40Ca = δ44/40Cacrystal - δ44/40Caparent solution) and the associated dissolution rate was low (0.5×10-7 mol/m2/h). In contrast, at relatively high temperatures (40-50 °C), the calculated dissolution rate was higher (25×10-5 mol/m2/h) while the Ca isotopic fractionation was the low (44/40Ca = -0.1 to -0.07). The results from the precipitation experiments suggest that the temperature-controlled variation in 44/40Ca is ~0.02 ‰/°C in inorganic calcites. The 44/40Ca-T relationship proposed in this study (44/40Ca = 0.024 ± 0.003 × T(°C) -1.08 ± 0.11) for inorganically precipitated carbonates can be used to investi-gate significant temperature changes in the geological past. The proposed 44/40Ca-T relation-ship was applied to published Ca isotopic data for cap carbonates to infer that the seawater ix temperature variability was between -17.9 C to 24.5 °C during Neoproterozoic snowball Earth events and these temperature estimates are consistent with published temperature estimates based on climate/ice sheet models. The second part of the thesis focuses on biogenic carbonates. To understand the efficacy of different geochemical and isotopic proxies, multi-elemental (Sr/Ca, Mg/Ca, Li/Ca, Ba/Ca, B/Ca) and isotopic (44/40Ca, 18O, 13C) measurements were performed on the same set of samples drilled from fourteen consecutive low-density and high-density bands in a Porites sp. coral which was collected from the Kavaratti island of the Lakshadweep archipelago in the Arabian Sea. For this coral, the temperature sensitivity of 44/40Ca was determined as ~0.094 ‰/°C. With the present analytical precision and proposed calibration equation, it is proposed that Ca isotopic composition of corals can be used to reconstruct paleotemperatures with a temperature resolution of ~1°C. The temperatures estimated from the geochemical and isotopic analyses were compared with satellite sea-surface temperatures over seven years. The results indicate that a combination of Sr/Ca, Li/Ca and 44/40Ca measurements in corals provide the most accurate sea-surface temperature. A new set of temperature-proxy relationships have been proposed for the Porites sp. in the Arabian Sea. The proposed proxy-SST relationships are: (i) Sr/Ca (mmol/mol) = 10.46 (±0.04) - 0.054 (±0.002) × SST (°C), (ii) Li/Ca (mol/mol) = -0.134 × SST + 9.99 and (iii) 44/40Ca SRM915a (‰) = 0.094 (±0.01) × SST -1.95 (±0.39). To further evaluate the applicability of Ca isotope ratios as a paleotemperature proxy, 44/40Ca was meas-ured in fish otolith samples. Otolith samples are commonly found in the fossil record and are important archives for paleoclimate research. Seasonal growth bands in otolith can yield high-resolution temperature and paleo-water chemistry where other archives like coral or foraminif-era are not available. Elemental ratios (Sr/Ca, Mg/Ca and Ba/Ca) along with 44/40Ca were measured in otolith samples from six different species of fish collected from different x geological locations with a wide range of temperatures from 2 to 25 °C. These samples were previously characterized for their 18O, 13C, and clumped isotope (47) compositions. The temperature sensitivity of 44/40Ca in fish otoliths (~0.02 ‰/°C) was determined for the first time in this thesis. The results suggest that Ca stable isotopes in fish otoliths can be used as a proxy for paleotemperature reconstruction. There is limited information on the temperature dependence of Ca isotopic fractionation in foraminifera. While some studies have reported a strong temperature dependency of 44/40Ca values (0.24 ‰/°C) in G. sacculifer culture samples, other studies have reported negligible temperature sensitivity of 44/40Ca values in foraminifera. The discrepancy between these stud-ies suggests a complex biomineralization pathway in foraminifera as well as the effects of additional factors on the 44/40Ca composition of foraminifera. To further evaluate the temper-ature sensitivity of 44/40Ca in foraminifera, 44/40Ca values were measured in samples which were also characterised for their 18O, 13C, Mg/Ca and Sr/Ca. Paired measurements of isotopic and elemental ratios were performed in benthic (C. wuellerstorfi), thermocline planktic (G. truncatulinoides and G. inflata) and other planktic species (G. ruber, G. bulloides, O universa, G. siphonifera) of foraminifera collected from the North Atlantic Ocean using a sediment trap. Some of these samples were earlier analysed for Mg/Ca and 18O measurements and used for the Mg/Ca thermometry calibration by Anand et al. (2003). In this study, we have analysed selected samples from Anand et al. (2003), showing a wide range in Mg/Ca values (0.99-4.62 mmol/mol), for their 44/40Ca compositions. When the species G. ruber and those with esti-mated habitat temperatures less than 3 °C are excluded, 44/40Ca values show a significant (r2 = 0.76) relationship with temperature [44/40Ca (‰) = 0.35 exp (0.055 × T (°C)], with temper-ature dependency of ~0.055 ‰/°C for 44/40Ca in foraminifera, thereby suggesting the possi-bility of using 44/40Ca in foraminifera as a paleotemperature proxy with a precision of ± 1.5 xi °C. There is a well-defined positive correlation between 44/40Ca and Sr/Ca for a few selected species, suggesting similar temperature dependent Sr partitioning and 44/40Ca fractionation in foraminifera. However, higher Sr/Ca ratio and low 44/40Ca in G. ruber suggests that the com-position of this surface-dwelling species may be affected by precipitation rate. Our study sug-gests that in addition to Mg/Ca, 44/40Ca in foraminifera can be used for accurate paleotemper-ature reconstruction. Overall the results reported in this thesis document the applicability, use-fulness, and limitations of Ca stable isotopes as a paleotemperature proxy.