Constraints on Earth System Functioning at the Paleocene‐Eocene Thermal Maximum From the Marine Silicon Cycle

Abstract

The Paleocene-Eocene Thermal Maximum (PETM, ca. 56 Ma) is marked by a negative carbon isotope excursion (CIE) and increased global temperatures. The CIE is thought to result from the release of 13 C-depleted carbon, although the source(s) of carbon and triggers for its release, its rate of release, and the mechanisms by which the Earth system recovered are all debated. Many of the proposed mechanisms for the onset and recovery phases of the PETM make testable predictions about the marine silica cycle, making silicon isotope records a promising tool to address open questions about the PETM. We analyzed silicon isotope ratios (δ 30 Si) in radiolarian tests and sponge spicules from the Western North Atlantic (ODP Site 1051) across the PETM. Radiolarian δ 30 Si decreases by 0.6‰ from a background of 1‰ coeval with the CIE, while sponge δ 30 Si remains consistent at 0.2‰. Using a box model to test the Si cycle response to various scenarios, we find the data are best explained by a weak silicate weathering feedback, implying the recovery was mostly driven by nondiatom organic carbon burial, the other major long-term carbon sink. We find no resolvable evidence for a volcanic trigger for carbon release, or for a change in regional oceanography. Better understanding of radiolarian Si isotope fractionation and more Si isotope records spanning the PETM are needed to confirm the global validity of these conclusions, but they highlight how the coupling between the silica and carbon cycles can be exploited to yield insight into the functioning of the Earth system.

Silicon Isotope Signatures of Radiolaria Reveal Taxon-Specific Differences in Isotope Fractionation

“…The new δ 30 Si rad data of different radiolarian taxa presented here reveal values ranging from +0.6 to +1.8 and are thus in the range of previously published data from surface sediments (+0.7 to +1.3 ; Abelmann et al, 2015) and sediment cores (+0.3 to +2.0 ; Hendry et al, 2014;Abelmann et al, 2015;Fontorbe et al, 2016Fontorbe et al, , 2017Fontorbe et al, , 2020. However, our results show systematic isotopic differences between the mean δ 30 Si signatures of the different radiolaria taxa of up to 0.9 (mean δ 30 Si Acro = +0.7 ; mean δ 30 Si Dictyo = +1.6 ).…”

“…From the 125-250 µm fraction, between 100 and 200 radiolarian tests were handpicked under a light microscope. Following previous studies (Hendry et al, 2014;Abelmann et al, 2015;Fontorbe et al, 2016Fontorbe et al, , 2017Fontorbe et al, , 2020, first, radiolarian samples were picked to represent the typical bulk radiolarian community, i.e., mixed radiolarian samples. The mixed radiolarian samples mainly contained species dwelling between 0-50 m (e.g., Spongurus sp., Tetrapyle octacantha) and 50-150 m (e.g., Lamprocyclas sp., Dicytocoryne sp.)…”

“…This comparison was used to evaluate differences in the Si cycling of deep and intermediate (subsurface) water masses and potential changes in their DSi concentration. δ 30 Si rad has been mainly applied as paleoceanographic proxy for DSi concentrations based on specimen retrieved from marine sediments since the last glacial maximum (last ∼25 kyrs BP; Hendry et al, 2014) and since the Palaeogene (30-60 Myrs ago; Fontorbe et al, 2016Fontorbe et al, , 2017Fontorbe et al, , 2020. However, to apply δ 30 Si as a direct proxy for DSi concentrations of specific water masses or water depth, detailed knowledge of the fractionation factor and a proper calibration for DSi concentrations has been lacking.…”

“…There is potential for radiolarian δ 30 Si (δ 30 Si rad ) to be used as a tool to constrain surface to mid-depth DSi concentrations, which will allow a more detailed reconstruction of the past silicon cycle. However, only a few studies have been carried out on the stable isotope composition of radiolarian opal (Egan et al, 2012;Hendry et al, 2014;Abelmann et al, 2015;Fontorbe et al, 2016Fontorbe et al, , 2017Fontorbe et al, , 2020. Radiolarian Si isotope fractionation factors are challenging to constrain given that it has so far not been possible to grow radiolarians through a reproduction cycle in culture (Kouduka et al, 2006;Krabberød et al, 2011), and short-term culture experiments have only been successful for certain species, such as spongodiscids (Sugiyama and Anderson, 1997;Ogane et al, 2010).…”

“…−0.5 to −0.9 ; Abelmann et al, 2015), which have been estimated to be similar to the fractionation range for modern diatoms. Studies that focus on the past δ 30 Si rad record thus have to revert to assumptions about the Si fractionation in radiolaria (Hendry et al, 2014;Fontorbe et al, 2016Fontorbe et al, , 2017Fontorbe et al, , 2020.…”