Coupling of marine and continental oxygen isotope records during the Eocene-Oligocene transition

While marine records of the Eocene-Oligocene transition indicate a generally coherent response to global cooling and the growth of continental ice on Antarctica, continental rec ords indicate substantial spatial variability. Marine Eocene-Oligocene transition records are marked by an ~+1.1‰ foraminiferal δ18O shift, but continental rec ords rarely record the same geochemical signature, making both correlation and linking of causal mechanisms between marine and continental records challenging. Here, a new high-resolution continental δ18O record, derived from the freshwater gill-breathing gastropod Viviparus lentus, is presented from the Hampshire Basin, UK. The Solent Group records marine incursions and has an established magnetostratigraphy, making it possible to correlate the succession directly with marine records. The V. lentus δ18O record indicates a penecontemporaneous, higher-magnitude shift (>+1.4‰) than marine records, which reflects both cooling and a source moisture compositional shift consistent with the growth of Antarctic ice. When combined with “clumped” isotope measurements from the same succession, about half of the isotopic shift can be attrib uted to cooling and about half to source moisture change, proportions similar to marine forami niferal rec ords. Thus, the new record indicates strong hydro logical cycle connections between marine and marginal continental environments during the Eocene-Oligocene transition not observed in continental interior records. INTRODUCTION The Eocene-Oligocene transition, onset at ca. 34 Ma, represents a climatic regime change from “greenhouse” conditions to “icehouse” conditions. Physical evidence (various; e.g., Davies et al., 2012) indicates the initiation of large-scale Antarctic glaciation during the Eocene-Oligocene transition, with permanent ice sheets formed on Antarctica for the first time between 34.0 and 33.65 Ma, peaking at the onset of an oxygen isotopic event known as Oi-1 (Miller et al., 1991; Zachos et al., 1996, 2001; Coxall et al., 2005; Pälike et al., 2006, supplement), which is an ~+1.1‰ shift that reflects a combination of temperature and salinity/ ice-volume change (Zachos et al., 2008). Both benthic (Coxall et al., 2005; Coxall and Wilson, 2011) and planktonic (Pearson et al., 2008) forami nifera record a coherent spatial and temporal isotopic response, which implies globalscale changes to Earth’s oceans and global-scale climatic cooling. In contrast, continental records are spatially variable (Sheldon, 2009; Zanazzi et al., 2015), with regional differences in the magnitude of temperature change (e.g., Zanazzi et al., 2007 vs. Retallack, 2007), the magnitude of precipitation change (Sheldon and Retallack, 2004; Abels et al., 2011), and even whether precipitation increased or decreased (Sheldon et al., 2009). The widely recognized positive isotope excursion seen in marine foraminiferal dO records of the Eocene-Oligocene transition glaciation is rarely found in terrestrial records (cf. Zanazzi et al., 2007; Zanazzi and Kohn, 2008), which instead often show no discernible shift in dO (e.g., Sheldon et al., 2012) or a delayed shift relative to marine records (Zanazzi et al., 2007). In addition, it can be difficult to link marine and terrestrial records of the Eocene-Oligocene transition for two reasons: (1) Continental successions are most often preserved in endorheic basins, far from marine incursions that would make direct age comparison possible, and (2) oxygen isotope records from continental interiors can be complicated by a variety of non-temperaturerelated factors (e.g., changing circulation patterns, orographic effects) or may respond to local, rather than global hydrologic cycle drivers (e.g., Sheldon et al., 2012). Thus, sections in continental strata that span the EoceneOligo cene transition and that are well calibrated to the geomagnetic polarity time scale and to marine geochronology are rare. The English Hampshire Basin Solent Group (Fig. 1) was deposited in a coastal floodplain environment. It has documented magnetostratigraphy, sequence stratigraphy, mammalian and charophyte biostratigraphy, and brief calcareous nannoplankton events (Hooker, 1987, 2010; Sille et al., 2004; Gale et al., 2006; Hooker et al., 2009) that allow good calibration to marine records through the entire late Eocene (Priabonian) and early Oligocene (Rupelian). The depositional rate of the Solent Group strata is high, ranging from ~3 to 10 cm k.y. (Hooker et al., 2009), which facilitates high-resolution sampling. The only significant hiatus was during the glacial maximum following the Oi-1 event, caused by major sea-level fall. Hren et al. (2013) recently published a “clumped isotope” paleotemperature reconstruction of the Eocene-Oligocene transition from the Solent Group based on the prosobranch gastropod Viviparus lentus (Solander), where they found: (1) the magnitude of mean annual air temperature change (~4–6 °C) was comparable to North Atlantic sea-surface GSA Bulletin; Month/Month 2015; v. 1xx; no. X/X; p. 1–9; doi: 10.1130/B31315.1; 4 figures; Data Repository item 2015314.; published online XX Month 2015. [email protected] For permission to copy, contact [email protected] © 2015 Geological Society of America as doi:10.1130/B31315.1 Geological Society of America Bulletin, published online on 14 September 2015