Menu

Evaluating thermal and non-thermal controls on benthic foraminiferal element/Ca ratios: a global core-top revised calibration

Benthic foraminiferal element/Ca ratios are widely used proxies to reconstruct past ocean conditions. For example, Mg/Ca and Mg/Li ratios are applied to reconstruct deep-sea temperature, while Sr/Ca and B/Ca are used to estimate past seawater carbonate chemistry. However, their interpretation remains challenging as other environmental factors can influence their variability, which is still not fully understood. Improving these reconstructions is therefore essential for better evaluating the evolution of ocean environments under different climate regimes. To address it, we compiled published and unpublished core-top element/Ca data for multiple benthic foraminiferal species (see Figure 1 below from Nauter-Alver et al., 2025). We then applied a multivariate linear regression model to quantify species-specific sensitivity to both in situ deep-sea temperature and calcite saturation state (Ωcalcite). In the model. we also explored the influences of dissolved inorganic carbon (DIC) and salinity on the element/Ca variability, but we found that these parameters did not improve the explained variance in the dataset.

Figure 1: Site Location of the core-top samples included in this study. From Nauter-Alves et al. (2025)

Our results show that (see also Figure 2 below from Nauter-Alver et al., 2025): (1) the benthic foraminiferal species included in this study are less sensitive to temperature when Ωcalcite is included in the models; (2) when Mg/Ca exhibits a carbonate effect, this needs to be corrected in order to obtain more precise temperature reconstructions; (3) variability in Sr/Ca and B/Ca is primarily driven by changes in Ωcalcite, although Sr/Ca shows a secondary temperature dependence in some cases; and (4) temperature reconstructions are associated with lower uncertainty when using Mg/Li compared to Mg/Ca.

Figure 2: Specie-specific temperature and Ω -2calcite/Ωcalcite regression coefficients for (a) Mg/Ca,  (b) Sr/Ca, and (c) B/Ca. (d–f) Show equivalent plots based on alternative regression models in which nonsignificant parameters have been removed. From Nauter-Alves et al. (2025).

Based on these findings, we developed a processing MATLAB tool called “ElCarBenthic”, which uses species-specific calibration coefficients to estimate deep-sea temperature and carbonate chemistry parameters from fossil benthic foraminiferal geochemistry data. The model combines benthic foraminiferal Mg/Ca or Mg/Li with Sr/Ca or B/Ca measurements and allows correction of data for secondary effects, including the long-term changes in seawater chemistry. In addition, this tool also propagates uncertainties providing robust estimates of deep-sea temperature and carbonate chemistry (see Figure 3 from Nauter-Alves et al. (2025) for an example).

Figure 3: Example application of ElCaRBenthic to reconstruct temperature and seawater carbonate chemistry at ODP Site 806 over the last 15 Myr (western Equatorial Pacific, 2.5 km, data from Lear et al., 2015). (a) Mg/Ca‐derived temperature assuming constant Ωcalcite.(b) Mg/Ca-derived temperature corrected for changes in Ωcalcite in the case of C. mundulus and L. wuellerstorfi. (c) Sr/Ca-derived Ωcalcite.