Reduction of oceanic temperature gradients in the early Eocene Southwest Pacific Ocean

Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology journal homepage: www.elsevier.com/locate/palaeo

Reduction of oceanic temperature gradients in the early Eocene Southwest Pacific Ocean Benjamin R. Hines a,⁎, Christopher J. Hollis b, Cliff B. Atkins a, Joel A. Baker a,c, Hugh E.G. Morgans b, Percy C. Strong b a b c

School of Geography, Environment and Earth Sciences, Victoria University of Wellington, New Zealand GNS Science, Lower Hutt, New Zealand School of Environment, The University of Auckland, New Zealand

a r t i c l e

i n f o

Article history: Received 29 August 2016 Received in revised form 17 February 2017 Accepted 26 February 2017 Available online 2 March 2017 Keywords: Paleoclimate Mg/Ca paleothermometry Early Eocene Climatic Optimum New Zealand East Coast Basin LA-ICP-MS

a b s t r a c t We present a Southwest Pacific Ocean paleothermometry transect using new and existing foraminiferal Mg/Ca data from six Eocene locations from eastern New Zealand to southwestern Campbell Plateau, spanning 10° of paleolatitude (43° to 53°S). Sea surface and seafloor temperatures (SST and SFT) have been calculated taking into account the partition coefficients of calcite and Paleogene Mg/Casw ratios, and used to determine values for the power component (H) of this relationship in foraminiferal calcite from paired proxy (δ18O and Mg/Ca) records from Canterbury Basin. This study presents the first Eocene paleotemperature record for the East Coast Basin and, in so doing, shows the absence of a meridional temperature gradient over 10° of latitude during the Early Eocene Climatic Optimum (EECO) and provides evidence for a very weak to absent gradient from the equatorial Pacific to 53°S at this time of extreme global warmth. The application of a new subsurface TEX86 calibration correlates remarkably well with Mg/Ca seafloor temperature trends from the continental slope setting (b 1000 m paleodepth) of the sites sampled in this study, suggesting that TEX86 is recording an upper ocean subsurface temperature record, rather than simply SST values. SST-SFT gradients demonstrate a notable reduction, decreasing to 3–5 °C across the EECO. The comparable ocean temperatures of the East Coast Basin, Canterbury Basin and DSDP Site 277 during this time support the intensification of a proto-East Australian Current (EAC), explaining the distribution of tropical sea temperatures extending into the high-latitude Southwest Pacific Ocean during the EECO. Seafloor temperatures (~1000 m paleodepth) likely correspond to a northern, intermediate water source during the early Eocene, before switching to a cooler southern-derived source in the middle to late Eocene. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Early Paleogene climate was characterised by temperatures that were significantly warmer than the present day, reaching a peak during the Early Eocene Climatic Optimum (EECO; 50–53 Ma). Proxy evidence from several localities indicates that the Southwest Pacific Ocean experienced a pronounced warming in the EECO that was followed by a stepwise cooling during the middle and late Eocene (Bijl et al., 2013; Hollis et al., 2012; Creech et al., 2010), culminating in the development of the first extensive Antarctic ice sheets in the earliest Oligocene (Shackleton and Kennett, 1975; Cramer et al., 2011). Paleogene proxy-based paleoclimate reconstructions from the southwest and tropical Pacific imply little to no latitudinal temperature gradient during the early Eocene, which is difficult to reconcile with the known climate dynamics and model studies (Bijl et al., 2013; Hollis et al., 2012).

⁎ Corresponding author. E-mail address: [email protected] (B.R. Hines).

http://dx.doi.org/10.1016/j.palaeo.2017.02.037 0031-0182/© 2017 Elsevier B.V. All rights reserved.

The thermodynamically controlled incorporation of magnesium into the calcite tests of foraminifera enables the derivation of past sea temperatures (e.g., Rosenthal et al., 1997; Lear et al., 2002; Eggins et al., 2003). However, post-depositional sedimentary contamination and diagenetic alteration of foraminiferal calcite can artificially bias Mg/Ca paleotemperature determinations (e.g., Barker et al., 2003; Kozdon et al., 2011). Application of the laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) method of determining elemental abundances for paleotemperature estimates provides the means to robustly assess issues associated with variable preservation, diagenetic coatings and sediment infilling of foraminifera tests by allowing the identification and removal of post-depositional effects and contaminated trace element depth profiles (Eggins et al., 2003; Creech et al., 2010). Mg/Ca sea temperature reconstructions for the Canterbury Basin have demonstrated good agreement with other paleotemperature proxies (Burgess et al., 2008; Hollis et al., 2012; Creech et al., 2010). In this study we extend the early Eocene record beyond the Canterbury Basin in order to develop a regional synthesis of Southwest Pacific sea temperature distributions and relate these to global circulation models.

42

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

An interval that spans the Paleocene–Eocene boundary at Deep Sea Drilling Project (DSDP) Site 277 on the Campbell Plateau (paleolatitude ~53°S) is combined with three East Coast Basin sections (paleolatitude ~43–44°S) exposed in the North Island of New Zealand, and previously published records from two localities in the Canterbury Basin (paleolatitude 55 and 56°S) in order to establish a new early Eocene sea temperature record spanning a ~ 10° latitudinal gradient. Palinspastic reconstructions place the New Zealand subcontinent between 42° and 54°S during the early Eocene, based on the paleomagnetic reference frame of Torsvik et al. (2012) as recommended by van Hinsbergen et al. (2015). This study adds new analyses to existing Mg/ Ca records from Hampden (Burgess et al., 2008; Hollis et al., 2012), and DSDP Site 277 (Hollis et al., 2015), with the addition of three new sites from the East Coast Basin.

2. Depositional setting and stratigraphy 2.1. Depositional setting The east coasts of the North and South Islands of New Zealand contains several hemipelagic Paleogene sedimentary successions from which moderately to well-preserved foraminiferal assemblages have been recovered. In this study, Mg/Ca proxy based paleo-sea temperatures have been obtained from three newly studied sections (Aropito, Tawanui and Tora) in the East Coast Basin, along with new data from Hampden Beach and DSDP Site 277, which are then integrated with two published records (mid-Waipara River, Hampden Beach and DSDP Site 277; Hollis et al., 2012, 2015) from the Canterbury Basin (Fig. 1). In particular, two sample suites were collected from the earliest to middle Eocene (Waipawan to Bortonian local stages) Wanstead Formation exposed in the Tawanui and Aropito sections in southern Hawke's Bay (Moore and Morgans, 1987), along with a lower to middle Eocene (Waipawan to Runangan) section at Pukemuri Stream, Tora, southeast Wairarapa (Hines et al., 2013). These sections were used to produce a new paleotemperature record for the central East Coast Basin (Fig. 2). Benthic foraminiferal paleodepth indicators for the East Coast Basin during the early to middle Eocene imply an upper bathyal (500– 1000 m) water depth for the Wanstead Formation at Aropito and Tawanui (Moore and Morgans, 1987; Kaiho et al., 1993), which is equivalent to the estimated paleodepth of 800 m for the Pukemuri Siltstone from the Tora section (Hines et al., 2013). These depths correspond to intermediate water. In the Canterbury Basin, the lower to Middle Eocene Ashley Mudstone exposed in the mid-Waipara River section has an estimated paleodepth of ~ 500 m (Hollis et al., 2012), which is also consistent with intermediate water. The Kurinui Formation exposed at Hampden Beach, Otago, was deposited at a shallower paleodepth of

200 m (Hollis et al., 2012). Benthic foraminiferal indicators from DSDP Site 277 imply an upper bathyal depositional depth of 800–1000 m. 2.2. Stratigraphy and age control The upper Paleocene (Teurian) to upper middle Eocene (Bortonian) sedimentary succession in the Aropito and Tawanui sections in southern Hawke's Bay is characterised by poorly bedded, calcareous, smectitic, green–grey mudstone of the Wanstead Formation (Moore and Morgans, 1987; Kaiho et al., 1993). A lower to upper Eocene (Mangaorapan to Kaiatan) succession is exposed in the Pukemuri Stream section at Tora, southeast Wairarapa. The lower part of the section (Mangaorapan to Heretaungan) comprises grey, poorly bedded mudstone and sandy mudstone of the Pukemuri Siltstone, which is unconformably overlain by Bortonian to Kaiatan, poorly bedded, highly calcareous, green–grey mudstone of the Wanstead Formation. Age control is based on the 2012 International Geological Timescale (GTS 2012; Gradstein et al., 2012). The ages assigned to New Zealand stage boundaries are based on Raine et al. (2015 – NZGTS2015). Foraminiferal, calcareous nannofossil and radiolarian datums were used to construct age–depth plots for all six sections in this study (Fig. 2; Supplementary Files 1 and 2). The linear sediment accumulation rates derived from these plots have been used to assign ages to our samples, which allows us to compare temporal trends between all six localities. The age–depth plots for DSDP Site 277 and the mid-Waipara and Hampden sections (Hollis et al., 2012, 2015) have been recalibrated to GTS 2012 and NZGTS2015 (Raine et al., 2015). 3. Analytical techniques 3.1. Sample preparation This study is based on 3251 new analyses of foraminiferal tests in 78 sediment samples collected from five sections, from the southern Hawke's Bay to Campbell Plateau. Sea surface temperature (SST) records of this study from the East Coast and Canterbury Basins were based on the Mg/Ca ratios of Morozovella crater, M. lensiformis, Acarinina primitiva and A. collactea, which inhabited a surface mixed-layer habitat (upper 200 m; Pearson et al., 2006). Samples from DSDP Site 277 are predominantly late Paleocene to Earliest Eocene in age and, as such, the species Morozovella aquea, M. subbotinae, Acarinina soldadoensis and A. coalingensis were analysed. The planktic genus Subbotina was used as an indicator of thermocline temperatures (~400 m; Pearson et al., 2006), but was not subdivided beyond genus level. The benthic species Cibicides eocaenus and C. truncatus were used to provide seafloor floor temperatures (SFT), which is thought to represent intermediate water in the lower bathyal setting of the East Coast Basin. Foraminifera were picked from the 150–300 μm fraction and individually washed in

Fig. 1. Location of stratigraphic sections used in this study. A) Map showing the present-day position of ODP Sites 865, 1124, 1172, 1209 and DSDP Site 277. B) Map of New Zealand showing the localities sampled in this study. TW = Tawanui and Aropito; TR = Tora; HD = Hampden Beach; 277 = DSDP Site 277. The location of mid-Waipara River (MW) is also shown.

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

43

Fig. 2. Biostratigraphic correlation of sections used in this study (LO – lowest occurrence; HO – highest occurrence). Tick marks indicate intervals sampled for this study. New Zealand stages have been calibrated to the 2012 International Geological Timescale (Raine et al., 2015). Age models and biostratigraphic datums for the Tawanui, Aropito, Tora, Hampden and DSDP Site 277 sections are presented in Supplementary Files 1 and 2.

ultra-pure (N18.2 m Ω) water and analytical grade methanol three times before being mounted on double-sided tape adhered to a disc of the NIST-SRM610 silicate glass standard. 3.2. LA-ICP-MS analysis Traditional bulk analytical methods require several foraminifera tests to be crushed and extensively cleaned using oxidizing (and/or reducing) reagents before being digested and analysed in solution (e.g., Marr et al., 2013). In contrast, the LA-ICP-MS method allows precise, in situ analysis of major and trace elements in foraminiferal calcite at the micron scale. The methods applied in this study follow those of Hollis et al. (2015), and a detailed outline of those is provided in Supplementary File 3. Stringent screening criteria were developed and applied to trace element depth profiles in this study in order to remove the effects of post-depositional diagenesis and silicate contamination. Specific zones within individual laser ablation depth profiles were screened for the effects of diagenetic alteration or detrital contamination by the identification of anomalous Mg/Ca, Al/Ca, Si/Ca, Ti/Ca, Mn/Ca and Sr/Ca ratios (after Barker et al., 2003; Eggins et al., 2003; Greaves et al., 2005). Selected elemental ratios are used for paleotemperature determinations (Mg/Ca) and indications of watermass changes (Zn/Ca, Ba/Ca). Calculated sea temperature estimates are based on screened Mg/Ca values averaged from multiple ablations of several individual specimens per species from each sample horizon. To provide further confidence in paleotemperature reconstructions, scanning electron microscope (SEM) imaging was used to qualitatively assess the state of individual foraminifera test preservation and identify zones of silicate

contamination, or indicators of diagenesis (e.g., dissolution and recrystallization; Supplementary File 3). These observations were supplemented by the application of several trace element proxies for silicate contamination (e.g., Al/Ca, Ti/Ca and Si/Ca) and diagenesis (e.g., Sr/Ca and Mn/Ca) that were applied to foraminifera in order to quantitatively assess the preservation of primary foraminiferal calcite and the associated paleoclimate signal. 3.3. Mg/Ca paleo-sea temperature calculations Mg/Ca-derived sea temperatures of Recent to Quaternary foraminifera are calculated using the exponential relationship between Mg/Ca and temperature (e.g., Lear et al., 2002; Eq. (1)). As the planktic foraminifera used in this study are extinct, SSTs were calculated using a general calibration based on the mean calcification temperatures of nine modern planktic species (A = 0.09, B = 0.38; Anand et al., 2003). SFTs were calculated using the calibration of Lear et al. (2002) based on three benthic Cibicidoides species (A = 0.109, B = 0.867).

Mg=Catest ¼

t¼t Mg=Casw

Mg=Cat¼0 sw

!  BexpAT

ð1Þ

The parameters necessary to calculate temperature (T) include the: (a) measured Mg/Ca ratio of foraminiferal calcite (Mg/Catest); (b) seawater Mg/Ca ratio at the time of calcification (Mg/Catsw= t) relative to the modern Mg/Ca seawater ratio (Mg/Catsw= 0); (c) species-specific calibration constants (A, B). Sea temperature reconstructions based on

44

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

early Eocene foraminifera have shown that a high (N3 mol/mol) Mg/ Casw value is necessary to reconcile Mg/Ca-derived paleotemperatures with those derived from δ18O when using Eq. (1) (Lear et al., 2002; Sexton et al., 2006a). Such high Mg/Casw values are at odds with several proxy studies (e.g., Dickson, 2004; Coggon et al., 2010) and models (e.g., Stanley and Hardie, 1998), but are consistent with modelled values from Wilkinson and Algeo (1989), and multi-proxy paleotemperature reconstructions (Hollis et al., 2012). Lower values of Paleogene Mg/ Casw (e.g., Stanley and Hardie, 1998; Dickson, 2004; Coggon et al., 2010) result in unrealistically high temperatures using Eq. (1). However, recent studies by Hasiuk and Lohmann (2010), Cramer et al. (2011) and Evans and Müller (2012) have shown that a power law relationship, rather than an exponential relationship, may better describe the relationship between Mg-partitioning and temperature in foraminiferal calcite (Eq. (2)). 0

1 B

H

A  Mg=Cat¼t expAT Mg=Catest ¼ @ sw t¼0 H Mg=Casw

ð2Þ

In applying this method, it is possible to derive realistic paleotemperatures using a lower Eocene Mg/Casw value that is consistent with proxy evidence. Eq. (2) can be rearranged to indirectly determine the power component (H), which relates the sensitivity of the calibration to the Mg content of the Eocene ocean for a particular species (Eq. (3)). ln

H¼

Mg=Catest

!

BexpAT     t¼t − ln Mg=Cat¼0 ln Mg=Casw sw

ð3Þ

Exponential and pre-exponential calibration constants from modern multi-species calibrations and paleotemperature values derived from oxygen isotopes can be utilised to estimate the power-law function (H) for extinct foraminifera. Using published data from Eocene foraminifera from Hampden Beach (Burgess et al., 2008; Hollis et al., 2012), mid-Waipara River (Creech et al., 2010; Hollis et al., 2012), ODP Site 865 (Kozdon et al., 2011) and Tanzania (Sexton et al., 2006a), for which Mg/Ca and δ18O data are available, we derived calibration correction constants (H) for the extinct species used in this study (Table 1). Paleotemperatures derived from δ18O measurements of foraminiferal calcite for both benthic and planktic taxa were calculated using a δ18Osw of − 1.3 for high-latitude sites (Hampden and mid-Waipara) and − 0.3 for low latitude sites (Tanzania and ODP 865) after Tindall et al. (2010) (see Supplementary file 3). In calculating the value of H, we have used an early Eocene Mg/Casw value of 1.6 mol/mol (Stanley and Hardie, 1998; Evans and Müller, 2012) and a modern Mg/Casw value of 5.17 mol/mol. The values of H determined for Paleogene

foraminifera in this study are an approximation that does not take into account the likely fine-scale variability in Mg/Casw values through the early Paleogene. The Mg/Ca–temperature calibrations of Anand et al. (2003) and Lear et al. (2002) have been used, although it is likely that the pre-exponential constant of Paleogene planktic foraminifera differed from that of modern taxa. The values of H calculated for Paleogene planktic foraminifera are lower than that of the modern taxon Globigerinoides sacculifer (H = 0.42; Hasiuk and Lohmann, 2010), which is attributable to differences in the Mg/Ca–temperature calibration, Mg partition coefficient of calcite (DMg) and Paleogene Mg/Casw ratio. Cramer et al. (2011) suggested that the value of H would be similar between the benthic foraminifera Cibicides sp. and Oridorsalis umbonatus, and this is supported by the comparable values calculated for O. umbonatus (Evans and Müller, 2012) and Cibicides sp. at midWaipara (Table 1). The value of H calculated for Morozovella crater and Cibicides sp. at Hampden is considerably lower (Table 1), which may be attributed to an uncertain oxygen isotope composition of seawater, due to the proximity of the locality to the paleo-coastline (Burgess et al., 2008; Tindall et al., 2010). Derivation of values for the power component (H) for application to Eocene Mg/Ca paleothermometry allows a Mg/Casw value consistent with model and proxy data to be applied to paleotemperature determinations, and results in temperature values that are consistent with independent proxy records. Once H has been determined, Mg/Ca-derived temperature values can be calculated using Eq. (4). Temperature values derived from Mg/Ca ratios of surface mixed-layer dwelling taxa used in this study (M. crater, M. lensiformis, A. primitiva, A. collactea, Acarinina sp.) are normalised to Morozovella crater in order to determine SSTs (after Creech et al., 2010). 0 h iH 1 t¼0 B½Mg=Catest   Mg=Casw C 1 T ¼ ln @ A h iH A t¼t B  Mg=Casw

ð4Þ

3.4. Error calculations Three types of error are relevant to absolute temperature values derived from Mg/Ca ratios: the analytical error (σ1); sample error (σ2); and the standard calibration error (σ3). The analytical error is accounted for in the data processing step, and typically produces very small uncertainties (±1–3%; 2 se) associated with counting statistics during ablation and data acquisition. The sample error pertains to the 95% confidence interval calculated for the mean temperature value obtained from multiple analyses within a single sample, and is determined by: σ X  t  pffiffiffi n

ð5Þ

Table 1 Calculated H values for selected Paleogene foraminifera. Where a range of paired Mg/Ca–δ18O measurements were available, minimum and maximum ranges for the H value are presented. *Mg/Ca data from Lear et al. (2002) and Rathmann et al. (2004). Calculated assuming no Eocene ice volume effect and using a δ18Ow value of −0.3 for low latitude sites (Tanzania [TDP] and ODP 865), and −1.3 for mid- to high-latitude sites (mid-Waipara River and Hampden Beach). Species

Age (Ma)

Location

Morozovella crater Morozovella crater Morozovella sp. Morozovella crassata Acarinina mcgrowrani Acarinina matthewsae Subbotina eoceanus Cibicides sp. Cibicides sp. Oridorsalis umbonatus

50.6–47.9 48.83 49–50 45 45 45 45 50.6–46.6 41.2–41.37 49

Mid-Waipara Hampden ODP 865 TDP 2 TDP 2 TDP 2 TDP 2 mid-Waipara Hampden –

Mg/Ca–δ18O data source

H Min

Max

Average

0.11 – – – – – – 0.29 0.10 0.44

0.53 – – – – – – 0.63 0.23 0.54

0.29 0.07 0.15 0.10 0.12 0.14 0.09 0.47 0.16 0.49

Creech et al. (2010); Hollis et al. (2012) Hollis et al. (2012) Kozdon et al. (2011) Sexton et al. (2006a) Sexton et al. (2006a) Sexton et al. (2006a) Sexton et al. (2006a) Creech et al. (2010); Hollis et al. (2012) Burgess et al. (2008) Evans and Müller (2012)*

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

where X is the sample mean, t is the inverse of the Student's t-distribution, σ represents the standard deviation and n is the number of analyses. The calibration error is the residual error of ± 1.6 °C on the regression of the multi-species calibrations established by Lear et al. (2002) and Anand et al. (2003). Each of these errors operates independently, thereby allowing the cumulative error to be determined by a simple quadratic expression: σ2 = σ21 + σ22 + σ23. Therefore, if all the three independent errors had a temperature uncertainty of ±1 °C, the overall uncertainty is ±1.7 °C.

4. Results 4.1. Foraminiferal trace element chemistry In order to generate a Southwest Pacific transect, paleotemperatures presented in this study are derived from 3251 new in situ analyses of 997 foraminifera, taken from previously unstudied localities (Tawanui, Aropito, Tora) and to supplement published records at mid-Waipara River, Hampden Beach and DSDP Site 277. Screened Mg/Ca ratios from 375 analyses of 125 foraminifera from ten samples obtained from the Wanstead Formation in the Aropito and Tawanui sections were used to construct an early to middle Eocene composite record for the southern Hawke's Bay. Mg/Ca ratios from 360 analyses of 120 foraminifera collected from the Pukemuri Siltstone at Tora, providing an early to middle Eocene record from the southern Wairarapa. Eleven samples from the early Eocene Mangaorapan and Heretaungan Stages were analysed from the Kurinui Formation exposed at Hampden Beach, with an additional 201 analyses of 67 foraminifera to supplement the record presented in Hollis et al. (2012). An additional 1316 analyses were conducted on 630 foraminifera collected from 40 samples from DSDP Site 277 to complement the 804 Mg/Ca analyses across 22 samples presented in Hollis et al. (2015). Planktic species typically produce lower Mg/Ca values than benthic species. The highest and lowest values are generally attributable to species of Cibicides and Subbotina, respectively. Morozovella and Acarinina commonly produce comparable Mg/Ca ratios, while Subbotina Mg/Ca ratios are typically lower than those of the other planktic species analysed in this study (Fig. 3). The benthic species Cibicides exhibits the greatest range of Mg/Ca values. Additional trace element/Ca ratios in foraminiferal calcite can potentially provide supplementary information on changes in seawater chemistry that may be related to environmental conditions other than temperature (Katz et al., 2010). The planktic species Morozovella, Acarinina and Subbotina from the Aropito, Tawanui and Hampden sections display similar Zn/Ca values and trends. Excluding two values recovered from Subbotina at Tawanui, Cibicides Zn/Ca ratios are consistently higher than those recorded for the planktic species, and display a marked increase up-section at Tawanui and Aroptio (Fig. 3). Likewise, the Ba/Ca record from Pukemuri Stream exhibits a clear offset between the benthic species (Cibicides) and the planktic species analysed in this study (Morozovella, Acarinina and Subbotina), with Cibicides having distinctly higher Ba/Ca ratios than the planktic species (Fig. 3). For the most part, Subbotina has Ba/Ca values higher than Morozovella and Acarinina, although these are distinctly lower than Cibicides. This is consistent with Zn/Ca and Ba/Ca values recorded in modern oceans, in which precipitation of barite in the upper levels of the ocean results in depleted barium concentrations in surface waters, and subsequent sinking and dissolution of barite resulting in increasingly enriched Ba concentrations at depth (Katz et al., 2010). Similar patterns of Zn/Ca ratios in the modern oceans exist due to the nutrient behaviour of this element. There are no clear Ba/Ca and Zn/Ca trends evident in foraminifera from Hampden, which is likely attributable to the shallow (200 m) depositional depth of these sediments, meaning that both planktic and benthic foraminifera populations were influenced by the same water mass.

45

5. Discussion 5.1. Applying the non-linear correction to Eocene Mg/Ca paleo-sea thermometry: Canterbury Basin The availability of independent, multi-proxy records from the Hampden and mid-Waipara sections in the Canterbury Basin enables an evaluation of Mg/Ca–temperature calibrations for Paleogene foraminifera. Mg/Ca ratios from the Canterbury Basin (Hollis et al., 2012) are here recalculated using the non-linear correction method outlined by Evans and Müller (2012), and compared to corresponding multi-proxy paleotemperature records derived from δ18O, TEX86 (TetraEther indeX of tetraethers consisting of 86 carbon atoms; Schouten et al., 2013) and the traditional linear method of Mg/Ca paleotemperature derivation (Eq. (1); which assumes high Mg/Catsw [e.g., 4.0 mol/mol; Hollis et al., 2012] and a invariant value for H; i.e., H = 1). For GDGT-based (glycerol dialkyl glycerol tetraethers) paleotemperature estimates, the BAYSPAR (Bayesian spatially-varying regression; Tierney and Tingley, 2014, 2015) SST TEX86 calibration has been applied. The BAYSPAR TEX86 calibration produces unreasonably high absolute SST estimates for the mid-Waipara record, although the temperature trends and relative offsets approximate those derived from inorganic proxies. The application of the BAYSPAR calibration to the Hampden TEX86 values plotted here produces SSTs that are in line with inorganic proxies in the middle Eocene. The lower bracketing value of H = 0.07 calculated from well-preserved Morozovella crater specimens from a single sample at Hampden (49 Ma; Table 1) provides SST estimates that are substantially lower than the Lear et al. (2002) calibration, and typically lower than corresponding δ18O temperatures. (Fig. 4a). The higher value of H = 0.2 is within 0.2 °C of calculated temperatures using the standard Mg/Ca– temperature calibration of Lear et al. (2002) assuming early Eocene Mg/Casw of 4.0 mol/mol at 49 Ma, as previously published in Hollis et al. (2012), and broadly agrees with SST estimates from oxygen isotopes (Fig. 4b). This is because the assumption of an invariant magnesium partition coefficient (i.e., H = 1) can be accommodated by assuming a high Mg/Casw ratio (e.g., Evans and Müller, 2012; Dunkley Jones et al., 2013). The value of H = 0.42 determined from the extant planktic foraminifera G. sacculifer provides the upper bracketing value in this example, although it produces peak temperatures during the EECO substantially higher than 30 °C (Fig. 4c), which are difficult to reconcile with independent organic and inorganic proxies and modelled temperature values (e.g., Hollis et al., 2012). Therefore, the adopted value of H for the planktic foraminifera M. crater and A. primitiva is 0.2. Similarly, the value of H for Cibicides sp. likely lies within the range from 0.16–0.48. An H value of 0.16 produces temperatures comparable (albeit marginally cooler by ~0.4–0.7 °C; Fig. 4a) to the standard calibration (Eq. (1); assuming 4.0 mol/mol Mg/Casw). An H value of 0.31 gives SFTs in agreement with temperatures derived from oxygen isotopes following the EECO (Fig. 4b). The highest bracketing value used (H = 0.48), gives considerably higher temperatures than those derived from oxygen isotope measurements and the standard Mg/Ca equation (Fig. 4c). The temperatures derived from the range of H values used for planktic species (0.07–0.42), results in a difference of 4.5 °C. Changes in the range of the H value for benthic foraminifera (0.16–0.48) result in a corresponding change in temperature of 3.3 °C. 5.2. Early Eocene paleotemperature record for the Southwest Pacific Ocean Applying the multiproxy comparisons from the Canterbury Basin to the wider dataset of this study, we adopt intermediate values of Hplanktic = 0.20 and Hbenthic = 0.31 to calculate paleo-sea temperatures from the planktic (Morozovella, Acarinina and Subbotina) and benthic (Cibicides) Mg/Ca values obtained in this study. The mean temperatures calculated from Acarinina are ~1–2 °C cooler than those of Morozovella (Fig. 5), which is consistent with the 0.7–2.6 °C offset between M. crater

46

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

Fig. 3. Screened trace element/Ca plots for Mg/Ca, Zn/Ca and Ba/Ca for planktic and benthic foraminifera. Mid-Waipara River data were taken from Creech et al. (2010). Shaded areas indicate the New Zealand Mangaorapan Stage (52–48.9 Ma).

and A. primitiva reported by Creech et al. (2010). There are two plausible explanations for this offset. There may be generic or taxon-specific differences in the degree of incorporation of Mg into foraminiferal calcite of these two taxa or, alternatively, there may be a seasonal or depthcontrolled difference in test precipitation between the two genera. For example, Morozovella shell growth is inferred to occur in surface waters during summer (Sexton et al., 2006b). Despite this offset, temperature estimates derived from Morozovella and Acarinina exhibit parallel trends throughout the record. The majority of the middle Eocene sea surface temperature (SST) record is based on Acarinina, as the species of Morozovella utilised in this study (M. crater and M. lensiformis) do not range above the early middle Eocene (45.7 Ma). Sea surface

temperatures peak at 30 °C in two Morozovella samples from DSDP 277 and two Acarinina samples from Tora, but typically do not exceed 27 °C between 51 and 49 Ma, followed by a gradual decline to ~ 22 °C at 42 Ma (Fig. 5). The seafloor temperature (SFT) record displays a similar trend to the SST record, reaching a peak of 16–21 °C at 52–50 Ma and declining to 11 °C by 45 Ma (Fig. 5). There is a peak of 14 °C between 43 and 42 Ma at Tawanui, at the same time as a ~1.5–2.0 °C warming in SST. An invariant Mg/Casw ratio (1.6 mol/mol) has been used for this study, although it is recognised that the ratio varied over the ~18.5 Myr spanned by this study, given the residence time of Mg and Ca in the ocean. Model and proxy studies indicate that Mg/Casw may

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

47

relationship outlined by Lear et al. (2002) (Eq. (1)), a 40% increase in the Mg/Casw value would result in a 2 °C decrease above that reported in this study. When applied, such an increase in Mg/Casw across the early Eocene effectively removes the offset between SFTs at 41–42 Ma derived from benthic δ18O and Mg/Ca recalculated from Burgess et al. (2008).

5.3. Comparison of Mg/Ca and GDGT-temperature proxies

Fig. 4. Canterbury Basin multiproxy records used in the estimation of H for early Paleogene foraminifera. Data sourced from: Burgess et al. (2008), Creech et al. (2010), and Hollis et al. (2012), with Mg/Ca values recalculated using bracketed values for H (A, low values [Hplanktic = 0.07, Hbenthic = 0.16]; B, intermediate values [Hplanktic = 0.20, Hbenthic = 0.31]; C, high values [Hplanktic = 0.42, Hbenthic = 0.48]), based on derived values for H in Table 1 and assuming an invariant Mg/Casw value of 1.6 mol/mol. Shaded areas show the errors associated with temperature values calculated using the Eq. (1).

have increased by 17–40% through this period (e.g., Wilkinson and Algeo, 1989; Stanley and Hardie, 1998; Lowenstein et al., 2001). This implies that the reported SSTs and SFTs for the middle Eocene may be overestimated, and that the cooling trend is likely greater than reported here. The non-linear correction method of Evans and Müller (2012) has the effect of decreasing the sensitivity of temperature determinations obtained from foraminiferal calcite to the influence of Mg/Casw, resulting in a 0.4 °C temperature decrease across this period, which is well within the error margin presented here. Applying the exponential

Throughout the early to middle Eocene portion of the record, the Bayesian spatially-varying regression (BAYSPAR; Tierney and Tingley, 2014, 2015) SST calibration for GDGT data produces high SST estimates across the records from ODP Site 1172, mid-Waipara and Hampden (Sluijs et al., 2011; Bijl et al., 2013; Hollis et al., 2012, 2014), with peak temperatures of up to 40 °C across the EECO, which is up to 10–15 °C warmer than corresponding Mg/Ca paleotemperatures (Fig. 6). Similarly, the BAYSPAR subsurface temperature calibration (subT), produces temperature estimates that are typically ~3 °C cooler than corresponding temperatures derived from the SST calibration, but remain substantially higher than SST temperatures derived by Mg/Ca paleothermometry between 48 and 52 Ma (Fig. 6). Temperatures derived from the BAYSPAR subT calibration are 1–5 °C cooler than temperatures from the TEXH 86 calibration of Kim et al. (2010). The BAYSPAR SST and subT calibrations produce a strong deviation in temperature across the Paleocene – Eocene boundary. Prior to the Paleocene–Eocene (P–E) boundary, the BAYSPAR calibrations record temperature values similar to sea floor temperatures rather than SSTs in the continental margin settings presented here (Fig. 6). The discrepancy between GDGT and carbonate-based temperature proxies across the EECO is well documented in the in the Southwest Pacific (Hollis et al., 2012), and may relate to differing archaea distributions during ultra-warm conditions of the early Eocene (Inglis et al., 2015). Varying oceanic regimes and seasonality have been demonstrated to shift GDGT distributions, complicating the TEX86 proxy (Taylor et al., 2013). Increased upwelling would be consistent with the shift of the TEX86 thermometer in line with benthic Mg/Ca temperatures, although evidence of this is not observed in the planktic Mg/Ca record. The GDGT-2/GDGT-3 ratio has been demonstrated to be more closely correlated to water depth than SST, reflecting a contribution from Thaumarchaeota (from which GDGT lipids originate) living deeper within the water column (Taylor et al., 2013). Following this line of reasoning, it is possible that there was a shift in either paleodepth, or Thaumarchaeota populations across the P-E boundary, resulting in a smaller contribution of GDGTs from deep sea Thaumarchaeota in the Eocene. The late Paleocene to middle Eocene temperature trend across the Southwest Pacific produced by GDGT temperature proxies effectively mirrors the trend in the sea floor temperatures derived from Mg/Ca paleothermometry. This suggests that the GDGT temperature calibration at these sites is affected by a contribution from archaea deeper in the water column, and/or that the temperature distribution is not entirely representative of sea surface temperatures. The recent reformulation of the TEXH 86 calibration by Ho and Laepple (2016) considers that derived temperatures are representative of subsurface, mixed ocean temperatures. The application of the Ho and Laepple (2016) calibration produces temperatures that are directly in line with benthic paleotemperatures derived by Mg/Ca analysis in this study. As the sites sampled for Mg/Ca analysis in this study are shallower than 1000 m paleodepth, this would place them within the range of paleodepths considered in the Ho and Laepple (2016) TEXH 86 reformulation. Therefore, this would support the hypothesis that TEX86 is representative of subsurface, rather than sea surface temperatures. Furthermore, the strong temperature deviation across the Paleocene – Eocene boundary shown by TEX86 SST calibrations (e.g. BAYSPAR, Kim et al., 2010) is not evident in subsurface temperatures derived from the Ho and Laepple (2016) TEX86 reformulation.

48

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

Fig. 5. Marine temperature trends from five localities plotted against age from the Late Paleocene to Middle Eocene (61 to 42 Ma). Temperatures for mid-Waipara River and Hampden Morozovella and Cibicides are recalculated from Creech et al. (2010) and Hollis et al. (2012) and plotted against deep sea temperatures derived from benthic δ18O (Cramer et al., 2009), which has been recalibrated to GTS 2012. Red and blue shaded area shows the errors associated with paleotemperature estimates of this study. Orange highlighted areas show the Paleocene–Eocene Thermal Maximum (PETM) and the Early Eocene Climatic Optimum (EECO).

The BAYSPAR SST and subT calibrations demonstrate extreme warming (Δ15 °C) across the PETM, similar to the ~10 °C warming shown by the Kim et al. (2010) TEXH 86 calibration across this same interval. However, the Ho and Laepple (2016) calibration produces temperature estimates that show a ~6 °C warming across the PETM, consistent with the increase in temperature demonstrated by the Mg/Ca record produced by this study. This corresponds directly with the increase across the PETM demonstrated by Dunkley Jones et al. (2013), and is more reasonable than the apparent 10–15 °C temperature increase displayed by other GDGTtemperature distributions across this interval (Fig. 6).

5.4. EECO meridonal temperature gradients Peak SSTs and SFTs across the EECO interval indicate no discernible temperature gradient in sea surface or intermediate water masses across 10° of latitude in the Southwest Pacific (Fig. 7). Peak SSTs at mid-Waipara, Hampden and ODP Site 1172 determined using the TEXH 86 calibration are consistently 6 °C higher than corresponding organic and inorganic temperature proxies (Fig. 7). The comparatively low SFT determined from δ18O at Hampden is conspicuous given the shallow depositional setting, and may be related to freshwater influence

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

49

H Fig. 6. Early Paleogene paleotemperature record for Southwest Pacific continental margin settings. Four calibrations (TEX86 (Kim et al., 2010), BAYSPAR SST, BAYSPAR SubT (Tierney and Tingley, 2014, 2015), and the Ho and Laepple (2016) subsurface (subT) reformulation) are applied to GDGT distributions for comparison to inorganic SST records determined in this study and Hollis et al. (2012). TEX86 data sourced from Bijl et al. (2013), Sluijs et al. (2011), Hollis et al. (2012, 2014), and Inglis et al. (2015).

on the δ18Osw at this neritic site. The apparent disparity between oxygen isotope-derived SST estimates from the equatorial Pacific ODP Site 865 (paleolatitude 2–6°N, paleodepth 1300–1500 m; Bralower et al., 1995) and ODP Site 1209 (paleolatitude 15–20°N, paleodepth 2500 m; Dutton et al., 2005), and southern, mid-latitude Mg/Ca-based temperature estimates in this study could be explained by diagenetic alteration in the low latitude localities. Kozdon et al. (2011) demonstrated that in situ δ18O measurements of Morozovella tests resulted in temperature estimates 4–8 °C higher than whole test estimates from ODP Site 865, due

to diagenetic cements. Factoring this offset into SST estimates for ODP Site 865, and including Mg/Ca SST estimates from Tripati et al. (2003), implies that there is no discernible meridional SST gradient during the EECO (Fig. 7). However, SST estimates from ODP Site 1209 are considerably cooler (~6 °C), which may also be attributed to diagenetic calcite cements (Kozdon et al., 2011). The lower SST and SFT estimates derived from ODP Site 1209 and Site 865 (Fig. 7) are highly unlikely to be the result of increased surface salinity affecting δ18Osw values and causing downwelling and low-

50

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

Fig. 7. Peak sea surface and seafloor temperatures across the EECO for the northern low-latitude to southern mid-latitude Pacific. BAYSPAR (SST) indicates the SST calibration of Tierney H and Tingley (2014, 2015). TEXH 86 (Kim) indicates the high-temperature calibration of Kim et al. (2010), and TEX86 (SubT) indicates the Ho and Laepple (2016) subsurface reformulation. ODP Site 1209 (Shatsky Rise, paleolatitude 15–20°N) oxygen isotope and Mg/Ca data sourced from Dutton et al. (2005). ODP Site 865 (Allison Guyot; paleolatitude 2–6°N) Mg/Ca and oxygen isotope data sourced from Bralower et al. (1995), Tripati et al. (2003) and Kozdon et al. (2011). Mid-Waipara River Mg/Ca, oxygen isotope and TEX86 data sourced from Hollis et al. (2012), and Creech et al. (2010). Hampden oxygen isotope and TEX86 data sourced from Hollis et al. (2012) and Inglis et al. (2015). ODP Site 1172 (Tasman Rise) TEX86 data sourced from Bijl et al. (2013) and Sluijs et al. (2011). *Indicates in situ δ18O measurements by Kozdon et al. (2011).

latitude deepwater formation in the Pacific (e.g. Tripati et al., 2003; Dutton et al., 2005), given the evidence for diagenesis, and ion probe data producing substantially lighter oxygen isotope values (Kozdon et al., 2011). Lower SFT values determined from Mg/Ca for ODP Site 1209 may be ascribed to the greater paledepth (Fig. 7). The close similarities in SST and SFT trends and magnitudes in the northern Canterbury (mid-Waipara) and the East Coast basins during the early to early middle Eocene (~53–46 Ma) indicate that the two regions were within the same oceanographic regime over this time. SST and SFT in southern Canterbury Basin (Hampden) from 51.5–52.3 Ma also agree with the northern records. However, termination of the EECO is marked by pronounced cooling at Hampden (Creech et al., 2010; Hollis et al., 2012). The abrupt cooling and termination of the EECO at Hampden (49.5 Ma; Hollis et al., 2012), corresponds with the sampled interval in both the mid-Waipara and East Coast Basin records (Aropito, Tawanui and Tora). The Tora SST record indicates that the termination of the EECO occurred around 48.2 Ma, comparable to the

timing of cooling at mid-Waipara (~ 48.5 Ma; Creech et al., 2010), although later than recorded at Hampden.

5.5. Decoupling of sea surface and sea floor temperature gradients The seafloor-to-sea surface temperature gradient derived from the late Paleocene to earliest Eocene and middle Eocene successions in the East Coast and Canterbury Basins consistently lie within the range of 9–12 °C, which is comparable to depths of 1000 m in the present midlatitude Southwest Pacific Ocean. The modern mean annual SST and SFT (at 1000 m) is 12–14 °C and 3–4 °C respectively, in the region to the east of New Zealand (Levitus et al., 2013), resulting in a 8–11 °C temperature gradient. However, the early Eocene, and particularly the EECO portion of the paleotemperature record presented here is marked by a notable reduction in seafloor-to-sea surface temperature gradients, from 10 to 12 °C, to an average of ~5 °C (Fig. 8). This appears to have

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

51

Fig. 8. Change in the sea surface temperatures relative to corresponding seafloor temperatures and the deep ocean δ18O compilation, showing changes in the surface – intermediate water, and surface to bottom water temperature gradients respectively. The modern mid-latitude Southwest Pacific sea surface (SST) to seafloor (SFT) temperature gradient (at 1000 m water depth) is shaded in grey. The positive offset across the PETM implies a greater warming of surface waters. The substantial decrease in temperature gradient across the EECO reflects increasing seafloor temperatures relative to SST, apparent in Figs. 5 and 6, followed by a gradual recovery of the sea surface to seafloor gradient in the middle Eocene.

an abrupt onset in the early Eocene prior to the EECO as evident from planktic and benthic paleotemperatures derived from Mg/Ca (Figs. 6 & 8). A substantial increase in the seafloor-to-sea surface temperature gradient occurs across the PETM in relation to the benthic deep-sea isotope record of Cramer et al. (2009) and intermediate water derived from Mg/Ca (Fig. 8). Given that all data across the PETM in this study is from DSDP Site 277, this likely reflects greater warming in the surface mixed layer than the SFT record at this site, resulting in the amplified seafloor-to-sea surface temperature gradient. However, in contrast, the EECO is marked by a substantial increase in SFT (initiating prior to the EECO), although the corresponding warming of surface water is less pronounced (Figs. 5 & 6). This long term SST cooling trend noted across the late Paleocene to middle Eocene in the composite Southwest Pacific record is consistent with the SST trend observed at ODP Site 1209 in the equatorial Pacific (Dutton et al., 2005). Within the EECO interval, the thermocline dwelling genera Subbotina provide paleotemperature estimates akin to those derived from the surface mixed layer dwellers, Morozovella and Acarinina, although data are sparse due to the lower preservation potential of Subbotina spp. In the interval following the PETM, Subbotina produce temperature estimates ~2 °C cooler than corresponding Morozovella and Acarinina, consistent with a deeper dwelling taxon. Following the EECO interval there is a gradual restoration of the sea surface to seafloor temperature gradient.

5.6. Southwest Pacific paleoceanography Paleotemperatures derived in this study are in broad agreement with the modelled depth–latitude heat distribution of SSTs and intermediate water temperatures during the early Eocene presented by Hollis et al. (2012). SST values for the EECO from this study are marginally higher than those modelled in Hollis et al. (2012), but are only plausible with a high atmospheric CO2 concentration (4480 ppm), summer bias to foraminiferal calcification and an intensification of the protoEast Australian Current (EAC) (e.g., Hollis et al., 2012), although it is noted that this model incorporates paleobathymetry 5° further south than the Torsvik et al. (2012) reference frame used here. The Hollis et al. (2012) model was run with a high atmospheric CO2 forcing (4480 ppm) which is substantially higher than pCO2 proxy data for this interval (~1000–2000 ppm; Anagnostou et al., 2016; Beerling and Royer, 2011). However, the NCAR CCSM3 model used by Hollis et al. (2012), has half the sensitivity of the HadCM and ECHAM models, therefore the 4480 ppm CO2 forcing is equivalent to 2240 ppm in these other models (Lunt et al., 2012), and more in line with proxy estimates. Previous studies of Eocene marine paleotemperature trends for this region have been limited to sites south of the Chatham Rise (Burgess et al., 2008; Bijl et al., 2013; Hollis et al., 2012; Creech et al., 2010; Sluijs et al., 2011). Throughout the Cenozoic and also presently, the Chatham

52

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

Rise has acted as a significant barrier to ocean circulation (e.g., Carter et al., 2004; Nelson and Cooke, 2001). Our study provides SSTs and SFTs for the region directly north of the Chatham Rise and confirms that the early Eocene meridional temperature gradient was extremely low across ~ 60° of latitude from the Equatorial Pacific (ODP Site 865; Tripati et al., 2003) to DSDP Site 277. This suggests that surface currents likely played a major role in the distribution of heat in the surface waters of the early Eocene Southwest Pacific Ocean. Intensification of surface currents during the EECO would enable efficient oceanic heat distribution, subsequently resulting in the apparent decreased thermal gradient from the equator to poles suggested by proxy records. Hollis et al. (2012) proposed an intensification of a proto-East Australian Current (EAC) and corresponding weakening of the Tasman Current (TC), in which an amplified proto-EAC results in the southward extension of warm, sub-tropical surface water masses into the high-latitude Southwest Pacific Ocean during the EECO (Fig. 9). The high temperatures of the East Coast Basin, which correspond with multi-proxy paleotemperature records from the Canterbury Basin (mid-Waipara River and Hampden Beach) and DSDP Site 277 during this time, support this hypothesis, and imply that this current also bathed northern New Zealand. An intensification of the EAC has been associated with warming of the South Pacific during Marine Isotope Stage 5e (MIS 5e; 125 ka), consequently producing a greater warming of surface waters to the east of New Zealand, including the East Coast of the North and South Islands (Cortese et al., 2013). This may provide an analogue for the early Eocene Southwest Pacific, and a plausible explanation for the distinct lack of surface water temperature variation across 10° of latitude along the eastern margin of New Zealand. The increase in benthic foraminiferal Zn/Ca and Ba/Ca ratios may indicate the influx of a new deepwater mass into East Coast Basin during the early late Eocene which corresponds with a coeval increase in seafloor Mg/Ca paleotemperatures of 4 °C. The timing of this event corresponds to a foraminifera origination event and increasing oxygenation

evident in benthic foraminiferal faunas at Tawanui at around 41 Ma (Kaiho et al., 1993), and the development of a proto-Subantarctic front (Nelson and Cooke, 2001), or alternatively, the deflection of a northern deep water mass down the eastern margin of the proto-New Zealand subcontinent (Carter et al., 2004). The latter scenario is favoured for the early Eocene, due to the correspondence between the SST and SFT records (which likely represents intermediate waters in this study), in conjunction with the absence of Subantarctic/Antarctic diatoms and distinct lack of dissolution of early Paleogene foraminifera in ODP Site 1124 (Carter et al., 1999, 2004). The Lutetian Stage (Porangan New Zealand Stage, 42.6–45.7 Ma) is absent in the Tora and mid-Waipara sections, and is considerably condensed at Hampden Beach, making it difficult to correlate changes in trace element chemistry beyond the southern Hawke's Bay. Evidence for changing paleoceanographic conditions throughout the southwest Pacific and the New Zealand sector of the Southern Ocean at this time (Nelson and Cooke, 2001), is coincident with decreasing εNd values in the deep Pacific Ocean, interpreted as an increasing influence of a Southern Ocean deepwater source (Thomas, 2004; Cramer et al., 2009). 6. Conclusions The LA-ICP-MS method of data collection and stringent screening criteria for foraminiferal trace element/Ca data provides a new, robust early Eocene temperature record for the New Zealand's East Coast Basin. As the first record from the southwest Pacific located north of the Chatham Rise, it is important for early Eocene climate reconstructions. Sea temperature estimates have been derived from Mg/Ca ratios after accounting for the partitioning of Mg into foraminiferal calcite and the lower Paleogene Mg/Casw values. Comparison of these new Mg/Ca paleo-sea temperatures with published multi-proxy records from the Canterbury Basin (δ18O, TEX86 and Mg/Ca) and DSDP Site 277 indicates that there was no latitudinal thermal gradient across

Fig. 9. Paleogeographic reconstruction of the New Zealand sub-continent during the Early Eocene (based on King et al., 1999; Torsvik et al., 2012; van Hinsbergen et al., 2015), showing peak temperatures for the EECO and the distribution of surface currents under an amplified proto-East Australian Current. TW = Tawanui and Aropito; TR = Tora; MW = mid-Waipara River; HD = Hampden Beach; 277 = DSDP Site 277. ODP Sites 1124 and 1172 are also plotted.

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

~10° of latitude from the East Coast Basin to DSDP Site 277 during the EECO. High sea surface temperatures derived from foraminiferal Mg/ Ca further imply an exceptionally low SST gradient between the Equatorial Pacific and southern mid-latitudes, which is difficult to reconcile with known climate dynamics and global circulation models, even assuming hyper-greenhouse conditions during the EECO. The application of the Ho and Laepple (2016) TEX86 SubT calibration correlates remarkably well with Mg/Ca seafloor temperatures from the continental slope setting (b1000 m paleodepth) of this study, supporting the hypothesis that TEX86 is recording an upper ocean subsurface temperature record, rather than solely SST values. This provides more reasonable Eocene temperature estimates, and largely removes the discrepancy between TEX86 SST records and carbonate based proxies (δ18O and Mg/Ca). Comparison of SST and SFT show reduced vertical and meridonal ocean temperature gradients during the EECO. This suggests that vigorous surface currents played a major role in the distribution of heat in the surface waters of the early Eocene Southwest Pacific Ocean. Intensification of a proto-East Australian Current (EAC) during the EECO provides an efficient means of oceanic heat distribution, subsequently resulting in the decreased thermal gradient from the equator to poles suggested by Southwest Pacific proxy records. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.palaeo.2017.02.037. Acknowledgments The authors would like to express their thanks to the VUW Geochemistry Laboratory, John Creech (VUW; STARPLAN), Katie Collins (VUW), Stephen Eggins (ANU), Euan Smith (VUW), Stephen Winch, and Randall McDonnell (GNS). This research was supported by the GNS Global Change through Time Programme and the Victoria University of Wellington Strategic Research Fund, grant number 202224. References Anagnostou, E., John, E.H., Edgar, K.M., Foster, G.L., Ridgwell, A., Inglis, G.N., Pancost, R.D., Lunt, D.J., Pearson, P.N., 2016. Changing atmospheric CO2 concentration was the primary driver of early Cenozoic climate. Nature 533 (7603). http://dx.doi.org/10.1038/ nature17423. Anand, P., Elderfield, H., Conte, M.H., 2003. Calibration of Mg/Ca thermometry in planktonic foraminifera from a sediment trap time series. Paleoceanography 18 (2):1050. http://dx.doi.org/10.1029/2002PA000846. Barker, S., Greaves, M., Elderfield, H., 2003. A study of cleaning procedures used for foraminiferal Mg/Ca paleothermometry. Geochem. Geophys. Geosyst. 4 (9):8407. http:// dx.doi.org/10.1016/j.quascirev.2004.07.016. Beerling, D.J., Royer, D.L., 2011. Convergent Ccenozoic CO2 history. Nat. Geosci. 4 (7). http://dx.doi.org/10.1038/ngeo1186. Bijl, P.K., Bendle, J.A., Bohaty, S.M., Pross, J., Schouten, S., Tauxe, L., Stickley, C.E., McKay, R.M., Röhl, U., Olney, M., Sluijs, A., Escutia, C., Brinkhuis, H., 2013. Eocene cooling linked to early flow across the Tasmanian Gateway. Proc. Natl. Acad. Sci. U. S. A. 110 (24):9645–9650. http://dx.doi.org/10.1073/pnas.1220872110. Bralower, T.J., Zachos, J.C., Thomas, E., Parrow, M., Paull, C.K., Kelly, D.C., Premoli Silva, I., Sliter, W.V., Lohmann, K.C., 1995. Late Paleocene to Eocene paleoceanography of the equatorial Pacific Ocean: stable isotopes recorded at ocean drilling program site 865, Allison Guyot. Paleoceanography 10 (4):841–865. http://dx.doi.org/10.1029/ 95PA01143. Burgess, C., Pearson, P.N., Lear, C.H., Morgans, H.E.G., Handley, L., Pancost, R.D., Schouten, S., 2008. Middle Eocene climate cyclicity in the southern Pacific: implications for global ice volume. Geology 36 (8):651–654. http://dx.doi.org/10.1130/G24762A.1. Carter, R.M., McCave, I.N., Richter, C., Carter, L., et al., 1999. Proceedings of the Ocean Drilling Program Initial Report 181. College Station, TX (CD-ROM and http:odp.tamu.edu/ publications/181IR). Carter, L., Carter, R.M., McCave, I.N., 2004. Evolution of the sedimentary system beneath the deep Pacific inflow off eastern New Zealand. Mar. Geol. 205:9–27. http://dx.doi. org/10.1016/S0025-3227(04)00016-7. Coggon, R.M., Teagle, D.A.H., Smith-Duque, C.E., Alt, J.C., Cooper, M.J., 2010. Reconstructing past seawater Mg/Ca and Sr/Ca from mid-ocean ridge flank calcium carbonate veins. Science 327:1114–1117. http://dx.doi.org/10.1126/science.1182252. Cortese, G., Dunbar, G.B., Carter, L., Scott, G., Bostock, H., Bowen, M., Crundwell, M., Hayward, B.W., Howard, W., Martinez, J.I., Moy, A., Neil, H., Sabaa, A., Sturm, A., 2013. Southwest Pacific Ocean response to a warmer world: insights from marine isotope stage 5e. Paleoceanography 28 (3):585–598. http://dx.doi.org/10.1002/palo.20052. Cramer, B.S., Toggweiler, J.R., Wright, J.D., Katz, M.E., Miller, K.G., 2009. Ocean overturning since the Late Cretaceous: inferences from a new benthic foraminiferal isotope compilation. Paleoceanography 24 (4). http://dx.doi.org/10.1029/2008PA001683.

53

Cramer, B.S., Miller, K.G., Barrett, P.J., Wright, J.D., 2011. Late Cretaceous–Neogene trends in deep ocean temperature and continental ice volume: reconciling records of benthic foraminiferal geochemistry (δ18O and Mg/Ca) with sea level history. J. Geophys. Res. 116. http://dx.doi.org/10.1029/2011JC007255. Creech, J.B., Baker, J.A., Hollis, C.J., Morgans, H.E.G., Smith, E.G.C., 2010. Eocene sea temperatures for the mid-latitude southwest Pacific from Mg/Ca ratios in planktonic and benthic foraminifera. Earth Planet. Sci. Lett. 299:483–495. http://dx.doi.org/10.1016/ j.epsl.2010.09.039. Dickson, J.A.D., 2004. Echinoderm skeletal preservation: calcite-aragonite seas and the Mg/Ca ratio of Phanerozoic oceans. J. Sediment. Res. 74 (3):355–365. http://dx.doi. org/10.1306/112203740355. Dunkley Jones, T., Lunt, D.J., Schmidt, D.N., Ridgwell, A., Sluijs, A., Valdes, P.J., Maslin, M., 2013. Climate model and proxy data constraints on ocean warming across the Paleocene–Eocene Thermal Maximum. Earth-Sci. Rev. 125:123–145. http://dx.doi.org/10. 1016/j.earscirev.2013.07.004. Dutton, A., Lohmann, K.C., Leckie, R.M., 2005. Insights from the Paleogene tropical Pacific: foraminiferal stable isotope and elemental results from Site 1209, Shatsky Rise. Paleoceanography 20 (3). http://dx.doi.org/10.1029/2004PA001098. Eggins, S., De Deckker, P., Marshall, J., 2003. Mg/Ca variation in foraminifera tests: implications for reconstructing palaeo-seawater temperature and habitat migration. Earth Planet. Sci. Lett. 212:291–306. http://dx.doi.org/10.1016/S0012-821X(03)00283-8. Evans, D., Müller, W., 2012. Deep time foraminifera Mg/Ca paleothermometry: nonlinear correction for secular change in seawater Mg/Ca. Paleoceanography 27 (4). http://dx. doi.org/10.1029/2012PA002315. Gradstein, F.M., Ogg, G., Schmitz, M. (Eds.), 2012. The Geologic Time Scale 2012 2-Volume Set. Elsevier, Oxford (1176 pp). Greaves, M., Barker, S., Daunt, C., Elderfield, D., 2005. Accuracy, standardization, and interlaboratory calibration standards for foraminiferal Mg/Ca thermometry. Geochem. Geophys. Geosyst. 6 (2). http://dx.doi.org/10.1029/2004GC000790. Hasiuk, F.J., Lohmann, K.C., 2010. Application of calcite Mg partitioning functions to the reconstruction of paleocean Mg/Ca. Geochim. Cosmochim. Acta 74 (23):6751–6763. http://dx.doi.org/10.1016/j.gca.2010.07.030. Hines, B.R., Kulhanek, D.D., Hollis, C.J., Atkins, C.B., Morgans, H.E.G., 2013. The Paleogene stratigraphy and paleoenvironment of Tora, southeast Wairarapa, New Zealand. N. Z. J. Geol. Geophys. 56:243–262. http://dx.doi.org/10.1080/00288306.2013.836112. van Hinsbergen, D.J., de Groot, L.V., van Schaik, S.J., Spakman, W., Bijl, P.K., Sluijs, A., Langereis, C.G., Brinkhuis, H., 2015. A paleolatitude calculator for paleoclimate studies. PLoS One 10 (6), e0126946. http://dx.doi.org/10.1371/journal.pone.0126946. Ho, S.L., Laepple, T., 2016. Flat meridional temperature gradient in the early Eocene in the subsurface rather than surface ocean. Nat. Geosci. 9 (8). http://dx.doi.org/10.1038/ ngeo2763. Hollis, C.J., Taylor, K.W.R., Handley, L., Pancost, R.D., Huber, M., Creech, J.B., Hines, B.R., Crouch, E.M., Morgans, H.E.G., Crampton, J.S., Gibbs, S., Pearson, P.N., Zachos, J.C., 2012. Early Paleogene temperature history of the Southwest Pacific Ocean: reconciling proxies and models. Earth Planet. Sci. Lett. 349–350:53–66. http://dx.doi.org/10. 1016/j.epsl.2012.06.024. Hollis, C.J., Tayler, M.J., Andrew, B., Taylor, K.W., Lurcock, P., Bijl, P.K., Kulhanek, D.K., Crouch, E.M., Nelson, C.S., Pancost, R.D., Huber, M., 2014. Organic-rich sedimentation in the South Pacific Ocean associated with Late Paleocene climatic cooling. Earth-Sci. Rev. 134:81–97. http://dx.doi.org/10.1016/j.earscirev.2014.03.006. Hollis, C.J., Hines, B.R., Littler, K., Villasante-Marcos, V., Kulhanek, D.K., Strong, C.P., Zachos, J.C., Eggins, S.M., Northcote, L., Phillips, A., 2015. Onset of the Paleocene-Eocene Thermal Maximum in the southern Pacific Ocean (DSDP Site 277, Campbell Plateau). Clim. Past Discuss. 11:243–278. http://dx.doi.org/10.5194/cpd-11-243-2015. Inglis, G.N., Farnsworth, A., Lunt, D., Foster, G.L., Hollis, C.J., Pagani, M., Jardine, P.E., Pearson, P.N., Markwick, P., Raynman, L., Galsworthy, A.M.J., Pancost, R.D., 2015. Descent towards the Icehouse: Eocene sea surface cooling inferred from GDGT distributions. Paleoceanography 30 (7):1000–1020. http://dx.doi.org/10.1002/2014PA002723. Kaiho, K., Morgans, H.E.G., Okada, H., 1993. Faunal turnover of intermediate water benthic foraminifera during the Paleogene in New Zealand. Mar. Micropaleontol. 23:51–86. http://dx.doi.org/10.1016/0377-8398(93)90053-Z. Katz, M.E., Cramer, B.S., Franzese, A., Hōnisch, B., Miller, K.G., Rosenthal, Y., Wright, J.D., 2010. Traditional and emerging proxies in foraminifera. J. Foraminiferal Res. 43: 406–414. http://dx.doi.org/10.2113/gsjfr.40.2.165. Kim, J.-H., van der Meer, J., Schouten, S., Helmke, P., Willmott, V., Sangiorgi, F., Koc, N., Hopmans, E.C., Damsté, J.S.S., 2010. New indices and calibrations derived from the distribution of crenarchaeal isoprenoid tetraether lipids: implications for past sea surface temperature reconstructions. Geochim. Cosmochim. Acta 74 (16): 4639–4654. http://dx.doi.org/10.1016/j.gca.2010.05.027. King, P.R., Naish, T.R., Browne, G.H., Field, B.D., Edbrooke, S.W., 1999. Cretaceous to recent sedimentary patterns in New Zealand. Institute of Geological and Nuclear Sciences Folio Series 1 (35 pp). Kozdon, R., Kelly, D.C., Kita, N.T., Fournelle, J.H., Valley, J.W., 2011. Planktonic foraminiferal oxygen isotope analysis by ion microprobe technique suggests warm tropical sea surface temperatures during the early Paleogene. Paleoceanography 26. http://dx.doi. org/10.1029/2010PA002056. Lear, C.H., Rosenthal, Y., Slowey, N., 2002. Benthic foraminiferal Mg/Ca paleothermometry: a revised core-top calibration. Geochim. Cosmochim. Acta 66: 3375–3387. http://dx.doi.org/10.1016/S0016-7037(02)00941-9. Levitus, S., Antonov, J.I., Baranova, O.K., Boyer, T.P., Coleman, C.L., Garcia, H.E., Grodsky, A.I., Johnson, D.R., Locarnini, R.A., Mishonov, A.V., Reagan, J.R., Sazama, C.L., Siedov, D., Smolyar, I., Yarosh, E.S., Zweng, M.M., 2013. The world ocean database. Data Sci. J. 12:WDS229–WDS234. http://dx.doi.org/10.2481/dsj.WDS-041. Lowenstein, T.K., Timofeeff, M.N., Brennan, S.T., Hardie, L.A., Demicco, R.V., 2001. Oscillations in Phanerozoic seawater chemistry: evidence from fluid inclusions. Science 294:1086–1088. http://dx.doi.org/10.1126/science.1064280.

54

B.R. Hines et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 475 (2017) 41–54

Lunt, D.J., Jones, T.D., Heinemann, M., Huber, M., LeGrande, A., Winguth, A., Loptson, C., Marotzke, J., Roberts, C.D., Tindall, J., Valdes, P., 2012. A model-data comparison for a multi-model ensemble of early Eocene atmosphere-ocean simulations: EoMIP. Clim. Past 8:1717–1736. http://dx.doi.org/10.5194/cp-8-1717-2012. Marr, J.P., Bostock, H.C., Carter, L., Bolton, A., Smith, E., 2013. Differential effects of cleaning procedures on the trace element chemistry of planktonic foraminifera. Chem. Geol. 351:310–323. http://dx.doi.org/10.1002/palo.20032. Moore, P.R., Morgans, H.E.G., 1987. Two new reference sections for the Wanstead Formation (Paleocene-Eocene) in southern Hawkes Bay. NZ Geol. Surv. Rec. 20, 81–87. Nelson, C.S., Cooke, P.J., 2001. History of oceanic front development in the New Zealand sector of the Southern Ocean during the Cenozoic: a synthesis. N. Z. J. Geol. Geophys. 44 (4):535–553. http://dx.doi.org/10.1080/00288306.2001.9514954. Pearson, P.N., Olsson, R.K., Huber, B.T., Hemleben, C., Berggren, W.A. (Eds.), 2006. Atlas of Eocene Planktonic Foraminifera, Cushman Found. For Foraminiferal Res., Spec. Publ. 41, Allen, Lawrence, Kans. Raine, J.I., Beu, A.G., Boyes, A.F., Campbell, H.J., Cooper, R.A., Crampton, J.S., Crundwell, M.P., Hollis, C.J., Morgans, H.E.G., 2015. Revised calibration of the New Zealand Geological Timescale: NZGT2015/1. GNS Sci. Rep. 2012 (39), 1–53. Rathmann, S., Hess, S., Kuhnert, H., Mulitza, S., 2004. Mg/Ca ratios of the benthic foraminifera Oridorsalis umbonatus obtained by laser ablation from core top sediments: relationship to bottom water temperature. Geochem. Geophys. Geosyst. 5 (12). http:// dx.doi.org/10.1029/2004GC000808. Rosenthal, Y., Boyle, E.A., Slowey, N., 1997. Temperature control on the incorporation of magnesium, strontium, fluorine, and cadmium into benthic foraminiferal shells from Little Bahama Bank: prospects for thermocline paleoceanography. Geochim. Cosmochim. Acta 61 (17):3633–3643. http://dx.doi.org/10.1016/S00167037(97)00181-6. Schouten, S., Hopmans, E.C., Damsté, J.S.S., 2013. The organic geochemistry of glycerol dialkyl glycerol tetraether lipids: a review. Org. Geochem. 54:19–61. http://dx.doi. org/10.1016/j.orggeochem.2012.09.006. Sexton, P.F., Wilson, P.A., Pearson, P.N., 2006a. Microstructural and geochemical perspectives on planktic foraminiferal preservation: ‘glassy’ versus ‘frosty’. Geochem. Geophys. Geosyst. 7 (12):218. http://dx.doi.org/10.1029/2006GC001291. Sexton, P.F., Wilson, P.A., Pearson, P.N., 2006b. Palaeoecology of late middle Eocene planktic foraminifera and evolutionary implications. Mar. Micropaleontol. 60. http://dx.doi.org/10.1016/j.marmicro.2006.02.006.

Shackleton, N.J., Kennett, J.P., 1975. Paleotemperature history of the Cenozoic and the initiation of Antarctic glaciation: oxygen and carbon isotope analyses in DSDP sites 277, 279, and 281. Initial Rep. Deep Sea Drill. Proj. 29, 743–755. Sluijs, A., Bijl, P.K., Schouten, S., Röhl, U., Reichart, G.-J., Brinkhuis, H., 2011. Southern ocean warming, sea level and hydrological change during the Paleocene-Eocene Thermal Maximum. Clim. Past 7:47–61. http://dx.doi.org/10.5194/cp-7-47-2011. Stanley, S.M., Hardie, L.A., 1998. Secular oscillations in the carbonate mineralogy of reefbuilding and sediment-producing organisms driven by tectonically forced shifts in seawater chemistry. Palaeogeogr. Palaeoclimatol. Palaeoecol. 144:3–19. http://dx. doi.org/10.1016/S0031-0182(98)00109-6. Taylor, K.W., Huber, M., Hollis, C.J., Hernandez-Sanchez, M.T., Pancost, R.D., 2013. Re-evaluating modern and Palaeogene GDGT distributions: implications for SST reconstructions. Glob. Planet. Chang. 108:158–174. http://dx.doi.org/10.1016/j.gloplacha.2013. 06.011. Thomas, D.J., 2004. Evidence for deep-water production in the North Pacific Ocean during the early Cenozoic warm interval. Nature 430 (6995):65–68. http://dx.doi.org/10. 1038/nature02639. Tierney, J.E., Tingley, M.P., 2014. A Bayesian, spatially-varying calibration model for the TEX 86 proxy. Geochim. Cosmochim. Acta 127. http://dx.doi.org/10.1016/j.gca.2013. 11.026. Tierney, J.E., Tingley, M.P., 2015. A TEX86 surface sediment database and extended Bayesian calibration. Sci. Data 2. http://dx.doi.org/10.1038/sdata.2015.29. Tindall, J., Flecker, R., Valdes, P., Schmidt, D.N., Marwick, P., Harris, J., 2010. Modelling the oxygen isotope distribution of ancient seawater using a coupled ocean–atmosphere GCM: implications for reconstructing early Eocene climate. Earth Planet. Sci. Lett. 292:265–273. http://dx.doi.org/10.1016/j.epsl.2009.12.049. Torsvik, T.H., Van der Voo, R., Preeden, U., Mac Niocaill, C., Steinberger, B., Doubrovine, P.V., van Hinsbergen, D.J., Domeier, M., Gaina, C., Tohver, E., Meert, J.G., 2012. Phanerozoic polar wander, palaeogeography and dynamics. Earth Sci. Rev. 114 (3):325–368. http://dx.doi.org/10.1016/j.earscirev.2012.06.007. Tripati, A.K., Delaney, M.L., Zachos, J.C., Anderson, L.D., Kelly, D.C., Harry, E., 2003. Tropical sea-surface temperature reconstruction for the early Paleogene using Mg/Ca ratios of planktonic foraminifera. Paleoceanography 18:1101. http://dx.doi.org/10.1029/ 2003PA000937. Wilkinson, B.H., Algeo, T.J., 1989. Sedimentary carbonate record of calcium-magnesium cycling. Am. J. Sci. 289:1158–1194. http://dx.doi.org/10.2475/ajs.289.10.1158.