Seawater calcium isotope ratios across the Eocene-Oligocene transition Elizabeth M. Griffi th1,2, Adina Paytan2, Anton Eisenhauer3, Thomas D. Bullen4, and Ellen Thomas5 1Department of Geology, 221 McGilvrey Hall, Kent State University, Kent, Ohio 44242, USA 2Institute of Marine Sciences, University of California−Santa Cruz, Santa Cruz, California 95064, USA 3Leibniz-Institut für Meereswissenschaften (IFM-GEOMAR), Wischhofsraße 1-3, D-24148 Kiel, Germany 4Branch of Regional Research, Water Resources Division, U.S. Geological Survey, MS 420, 345 Middlefi eld Road, Menlo Park, California 94025, USA 5Department of Geology & Geophysics, Yale University, P.O. Box 208109, New Haven, Connecticut 06520-8109, USA, and Department of Earth and Environmental Sciences, Wesleyan University, Middletown, Connecticut 06459, USA ABSTRACT a period of time longer than the oceanic Ca resi- During the Eocene-Oligocene transition (EOT, ca. 34 Ma), Earth’s climate cooled signif- dence time, seawater δ44/40Ca would decrease icantly from a greenhouse to an icehouse climate, while the calcite (CaCO3) compensation to a new steady-state value (Sime et al., 2007; depth (CCD) in the Pacifi c Ocean increased rapidly. Fluctuations in the CCD could result Fantle, 2010). from various processes that create an imbalance between calcium (Ca) sources to, and sinks Alternatively, a scenario of increased weath- from, the ocean (e.g., weathering and CaCO3 deposition), with different effects on the isotopic ering and alkalinity input to the ocean could composition of dissolved Ca in the oceans due to differences in the Ca isotopic composition of cause a deepening of the CCD without a change various inputs and outputs. We used Ca isotope ratios (δ44/40Ca) of coeval pelagic marine bar- in seawater δ44/40Ca, if compensation occurred ite and bulk carbonate to evaluate changes in the marine Ca cycle across the EOT. We show over a period shorter than the oceanic Ca resi- that the permanent deepening of the CCD was not accompanied by a pronounced change in dence time (Opdyke and Wilkinson, 1988; seawater δ44/40Ca, whereas time intervals in the Neogene with smaller carbonate depositional Kump and Arthur, 1997; Coxall et al., 2005; changes are characterized by seawater δ44/40Ca shifts. This suggests that the response of seawa- Tripati et al., 2005; Merico et al., 2008). ter δ44/40Ca to changes in weathering fl uxes and to imbalances in the oceanic alkalinity budget The δ44/40Ca values of bulk marine carbon- depends on the chemical composition of seawater. A minor and transient fl uctuation in the ates decreased by 0.6‰ close to the EOT, in Ca isotope ratio of bulk carbonate may refl ect a change in isotopic fractionation associated agreement with scenario 1 (De La Rocha and with CaCO3 precipitation from seawater due to a combination of factors, including changes in DePaolo, 2000). If the observed change of 0.6‰ temperature and/or in the assemblages of calcifying organisms. represents seawater δ44/40Ca, it implies a 50% increase in Ca weathering relative to sedimenta- INTRODUCTION Processes that result in deepening of the tion fl ux (De La Rocha and DePaolo, 2000), too Calcite (CaCO3) sedimentation in the ocean CCD (and carbonate accumulation in the deep large to reconcile with modeling results or mass represents the largest sink for calcium and car- sea) combined with changes in seawater δ44/40Ca balance calculations that account for increased bon in the combined atmosphere-biosphere- include the following. (1) An increase in CaCO3 carbonate burial associated with the observed ocean system, connecting the global carbon saturation (and preservation) is caused by an changes in the CCD (Lyle et al., 2008; Merico and Ca cycles (Milliman, 1993; Ridgwell and imbalance between Ca input to the ocean and et al., 2008). The 0.6‰ shift may at least in part Zeebe, 2005). Determining fl uctuations in the output as CaCO3 sediments (Rea and Lyle, be due to changes in species-dependent isotopic rate and locations of CaCO3 sedimentation 2005), and is accompanied by a decoupling fractionation caused by changes in calcifying and their association with carbon cycle per- of the Ca and carbon cycles. This requires an plankton species, and may not represent seawa- turbations over climate transitions provides increase in the fl ux of Ca to the ocean relative ter (see following discussion). important information on the behavior of the to alkalinity (e.g., through increased weather- The δ44/40Ca of marine (pelagic) barite is a coupled Ca-C biogeochemical system (e.g., ing of carbonate rocks relative to silicate rocks; reliable recorder of seawater δ44/40Ca (Griffi th Lyle et al., 2008). However, reconstructing Heuser et al., 2005), leading to an increase in et al., 2008a, 2008b). Marine barite precipitates global changes in CaCO3 sedimentation pat- Ca concentration in seawater, with a decrease inorganically in seawater, with a constant offset terns during geologically rapid events is com- in seawater δ44/40Ca (De La Rocha and DePaolo, (Ca isotopic fractionation) from dissolved Ca in − σ plicated, because preservation is not spatially 2000). (2) An increase in silicate weathering seawater of 2.01‰ ± 0.15‰ (average 2 mean), homogeneous and the calcite compensation rates (e.g., due to glaciation), reduces atmo- unaffected by temperature (1–14 °C; Griffi th et depth (CCD) is affected by local (as well as spheric CO2, increases alkalinity in the oceans, al., 2008b). Thus barite may record seawater global) processes (van Andel, 1975; Iglesias- and deepens the CCD (Zachos et al., 1999; δ44/40Ca more predictably than biogenic carbon- Rodriguez et al., 2002). Ravizza and Peucker-Ehrenbrink, 2003; Zachos ates. We reconstructed the isotope ratio of Ca in In the equatorial Pacifi c, the CCD (the depth and Kump, 2005). If changes occurred over a seawater (from marine barite) and of its major where rates of CaCO3 dissolution and deposition time period longer than the residence time of sink (bulk CaCO3) in order to evaluate these sce- 1 are equal, below which CaCO3 is not preserved Ca in the oceans, and there was no net change narios (see the GSA Data Repository ), and to in the sediment) increased by 1200–1500 m in total marine carbonate sedimentation, sea- constrain the amount of CaCO3 deposited in the in <300 k.y., a unique event (e.g., van Andel, water Ca concentrations would increase, with oceans relative to the input of Ca to the ocean 1975; Coxall et al., 2005; Lyle et al., 2005) that increasing seawater δ44/40Ca (Sime et al., 2007). occurred during the Eocene-Oligocene transi- (3) If the CCD deepening was associated with 1GSA Data Repository item 2011210, supplemen- tary text, fi gures and tables, is available online at tion (EOT), the transition from a warm green- a smaller globally averaged fractionation dur- www.geosociety.org/pubs/ft2011.htm, or on request house to a cold icehouse world (Katz et al., ing carbonate sedimentation resulting from a from [email protected] or Documents Secre- 2008; Lear et al., 2008; Liu et al., 2009). shift from aragonite to calcite precipitation for tary, GSA, P.O. Box 9140, Boulder, CO 80301, USA. © 2011 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected]. GEOLOGY,Geology, July July 2011; 2011 v. 39; no. 7; p. 683–686; doi:10.1130/G31872.1; 2 fi gures; Data Repository item 2011210. 683 via weathering, making assumptions regard- Time (m.y. ago) DISCUSSION ing the isotopic composition of Ca sources and In order to constrain the possible processes sinks (Skulan et al., 1997; Zhu and Macdougall, controlling the Ca isotopic composition of sea- 32.0 32.5 33.0 33.5 34.0 34.5 35.0 35.5 1998; De La Rocha and DePaolo, 2000). 36.0 water and carbonate sediments, we used the Oligocene Eocene barite and carbonate data in a simplifi ed model METHODS 0.4 of the marine Ca cycle, allowing us to defi ne A relatively high resolution record of the 0.2 both the seawater δ444/40Ca and the fractionation δ44/40Ca of barite (representing seawater) and 0.0 factor associated with carbonate sedimenta- coeval bulk CaCO (Ca sink) across the EOT tion, Δ44/40Ca . Interpretations using this model 3 ) –0.2 sed was constructed using sediments collected by O require that the system is in isotopic steady state, –0.4 the Deep Sea Drilling Project and Ocean Drill- a reasonable assumption given that we did not –0.6 δ44/40 ing Program at Sites 574 and 1218 in the equa- (% Ca observe signifi cant changes in seawater Ca torial Pacifi c Ocean. Barite was extracted using –0.8 (as recorded in barite) over this time. a sequential leaching process (Paytan et al., 44/40 Bulk carbonate δ44/40Ca is controlled by sea- δ –1.0 δ44/40 Δ44/40 1993) and screened for purity using scanning –1.2 water Ca and Cased. Since seawater Ca electron microscopy with energy dispersive –1.4 isotopes in our record did not change by more X-ray spectroscopy. Barite samples were pre- than 0.14‰ (average 2σ of barite data) for –1.6 mean pared for Ca isotope analysis following meth- periods of time longer than approximately ods in Griffi th et al. (2008b), and bulk carbon- 0 half the oceanic Ca residence time, ~0.5 m.y., ate samples were prepared following methods (wt%) the transient fl uctuation in the bulk carbon- 40 3 in Fantle and DePaolo (2005). Ca isotopic com- ate record must indicate a transient change in 80 Δ44/40 positions were determined by thermal ioniza- Cased that did not result in a measurable tion mass spectrometry using the double-spike 0.2 CaCO change in the global isotopic composition of δ44/40 ) technique, and reported in Ca (‰) relative O 0.6 seawater (i.e., greater than our analytical reso- to modern seawater (see the Data Repository; lution of ~0.14‰).