18. isotopic and faunal record of paleogene deep …€¦ · 18. isotopic and faunal record of...

Rea, D.K., Basov, I.A., Scholl, D.W., and Allan, J.F. (Eds.), 1995Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 145

18. ISOTOPIC AND FAUNAL RECORD OF PALEOGENEDEEP-WATER TRANSITIONS IN THE NORTH PACIFIC1

Dorothy K. Pak2 and Kenneth G. Miller3

ABSTRACT

The climatic change from early Eocene warmth to the onset of Oligocene continental glaciation may have been associated withchanges in deep-water circulation. The North Pacific Transect (Ocean Drilling Program Leg 145) recovered Paleogene carbonatesat Site 883 (2000 m paleodepth; upper Paleocene to lower Oligocene) and Site 884 (3300 m paleodepth; lower Eocene to lowerOligocene), providing a middle and upper Eocene isotope record from the North Pacific. These new data, combined with existingdata from Site 577, allow the construction of a late Paleocene and Eocene composite Pacific isotopic record.

Interbasinal isotopic comparisons indicate that during the early middle to middle middle Eocene (ca. 51-46 Ma) the SouthernOcean was the sole source of deep water to the North Pacific. However, during the early Eocene (ca. 57-52 Ma) and the middlemiddle Eocene to Oligocene (ca. 46-36 Ma), there were two sources of deep water: a cool, nutrient-depleted source originating inthe Southern Ocean and a warmer, more nutrient-enriched source probably originating in mid-latitudes.

Deep-water circulation changes have the potential to profoundly affect benthic foraminifer communities. During the earlyEocene, the deep Site 884 was characterized by a distinct abyssal (>3000 m) agglutinated foraminiferal faunal assemblage. Theabyssal assemblage disappeared near the early/middle Eocene boundary, coinciding with both the first Cenozoic deep-watercooling step and an inferred shift in circulation from a two-source to a one-source mode. We suggest that benthic foraminiferassemblages responded to both rapid cooling and decreased corrosivity as the circulation regime shifted from one mode to another.

INTRODUCTION

Early Eocene (-52 Ma) climate was the warmest of the Cenozoic,with deep-water temperatures of 10°-15°C (Shackleton and Boersma,1983; Miller et al, 1987a; Zachos et al., 1994). By the earliest Oligo-cene, deep-water temperatures had cooled to ~2°-4°C and major icesheets had begun to build on Antarctica (Wise et al., 1985; Shackleton,1986; Barron et al., 1991). Although this trend was one of the mostimportant climate transitions of the Cenozoic, significant gaps exist inour understanding of its nature and timing. In this study, we examinebenthic foraminifer isotopic and faunal records from three PacificEocene deep- and intermediate-water sections. We provide the firstNorth Pacific middle and upper Eocene isotope record. As the largestglobal reservoir of deep water (and of carbon), the Pacific recordrepresents a critical aspect of circulation that has been unavailable untilnow. Using these records, we address three questions: (1) Was climaticcooling accompanied by shifts in deep- or intermediate-water circula-tion? (2) What was the vertical temperature structure of the Eocenedeep North Pacific? and (3) Were changes in benthic foraminiferassemblages related to deep-water cooling and/or circulation changes?

Eocene Deep-Water Formation

Changes in deep-water circulation can alter climate through theinterhemispheric transport of heat and salt, as well as by regulatingthe atmospheric CO2 content (e.g., Broecker and Denton, 1989). Inthe modern ocean, deep waters form in polar regions during sea-iceformation (as in Antarctic Bottom Water [AABW]) or as a result ofintense cooling and evaporation of warm surface waters advected intopolar oceans (as in North Atlantic Deep Water [NADW]). The evapo-rative heat released during NADW formation plays a vital role in themodulation of local climate and changes in the production rate of

1 Rea, D.K., Basov, LA., Scholl, D.W., and Allan, J.F. (Eds.), 1995. Proc. ODP, Sci.Results, 145: College Station, TX (Ocean Drilling Program).

2 Lamont-Doherty Earth Observatory and Department of Geological Sciences,Columbia University, Palisades, NY 10994, U.S.A.

3 Department of Geological Sciences, Rutgers University, New Brunswick, NJ 08855,U.S.A.

NADW were strongly linked to late Pleistocene glacial-interglacialcycles (e.g., Broecker and Denton, 1989).

Isotopic data indicate that during most of the Eocene polar surfacewaters were warm (>10°C), whereas equatorial temperatures mayhave remained relatively stable (Shackleton and Boersma, 1983;Zachos et al., 1994). The low thermal gradient and absence of polar seaice has led to debate about the modes of formation of Eocene deepwaters. Specifically, were the high latitudes cold enough to be a majorsource of bottom waters? Alternatively, the deep ocean may have beenventilated by warm saline deep water (WSDW) formed by high evapo-ration rates at mid-latitudes. Although several authors have argued thatEocene paleogeographic and climatic conditions favored the forma-tion of a mid-latitude WSDW (Chamberlin, 1906; Brass et al., 1982;Barron, 1987; Barron and Peterson, 1991), little isotopic or faunalevidence exists to support this idea. Reconstructions of the verticaltemperature gradient in the Eocene ocean have been hindered by thelack of appropriate depth transects. One exception has been a study ofthe vertical isotopic gradient across intermediate-water depths in theSouthern Ocean (Kennett and Stott, 1990) showing that late Eoceneδ 1 8 θ values at the deeper site were lower than at the shallower site. Thisobservation led the authors to suggest that the deeper site was influ-enced by WSDW during most of the Eocene. However, Kennett andStotfs (1990) deepest site had an Eocene paleodepth of approximately2000 m and does not reflect conditions in the deep basin. Mead et al.(1993) subsequently studied the late Eocene to Oligocene verticalisotopic gradient in the subantarctic in order to determine whether ahalothermal water mass, if present, was primarily restricted to interme-diate or deep depths. They concluded that WSDW influenced the deepsubantarctic from approximately 2000 to 3400 m and was overlain bya colder water mass at intermediate (1500-2000 m) depths. Additionalvertical transects from deep (>3000 m) through intermediate waterdepths are necessary to test the presence and importance of WSDWduring the Paleocene and Eocene.

Interbasinal comparisons of late Paleocene to early Eocene δ 1 8θand δ13C values indicate that, although there were two early Eocenesources of deep water with different δ 1 8 θ signatures, the SouthernOcean remained an important source throughout the late Paleocene toearly Eocene (Pak and Miller, 1992). Similarly, Southern Oceanplanktonic and benthic foraminifer δ 1 8θ values were similar through-

265

D.K. PAK, K.G. MILLER

out the Eocene (Barrera and Huber, 1991; Zachos et al., 1994), indi-cating that the Southern Ocean was a source of deep water at this time.

Neither the sources of Eocene deep water nor the vertical tempera-ture structure of the Eocene ocean is well constrained. However, sub-stantial isotopic evidence exists for Eocene deep-water cooling. Iso-topic and faunal data indicate that early Eocene climates were thewarmest of the Cenozoic. Deep waters at this time were as warm as15°C. Fossil remains of tropical faunas have been found as far north asAlaska (Wolfe, 1978), indicating high-latitude continental warmth.Subsequent to the early Eocene peak warmth, climate cooled through-out the remainder of the Eocene (Savin et al., 1975; Shackleton andKennett, 1975; Savin, 1977; Miller et al., 1987a). The benthic foramin-ifer δ 1 8 θ data available indicate that the Eocene deep-water coolingtook place as a series of at least three steps (Miller, 1992): near theearly/middle Eocene boundary, near the middle/late Eocene boundary,and in the earliest Oligocene. Although the earliest Oligocene shift iswell established (Savin et al., 1975; Shackleton and Kennett, 1975;Miller et al., 1987a), the two middle Eocene steps are not well con-strained in terms of either timing or magnitude (Miller, 1992). Theearly/middle Eocene δ'8O increase is characterized as a rapid increaseof ~l.O‰ beginning in Chron C22n (Shackleton et al, 1984; Miller,1992) and continuing through C21 (Katz and Miller, 1991; Miller,1992), thus spanning the early/middle Eocene boundary. The middle/late Eocene δ 1 8 θ increase was also approximately l.O‰ in magnitude(Shackleton and Kennett, 1975; Oberhànsli et al, 1984; Keigwin andCorliss, 1986). The timing of this step is poorly constrained; however,based on biostratigraphic relationships at subantarctic Site 702, theincrease occurred in the latest middle Eocene (Katz and Miller, 1991).The timing of the early Oligocene shift is constrained to within C13n,subsequent to the last occurrence (LO) of Hantkenina spp., clearlyplacing the shift within the earliest Oligocene (Keigwin, 1980; Miller,1992). The step was characterized by a benthic foraminifer δ 1 8 θ in-crease of ~1.0‰-1.5‰. A coeval planktonic foraminiferal δ 1 8 θ in-crease of ~l‰ suggests that at least 0.3‰-1.0‰ of the increase isattributable to ice growth, leaving approximately 0.5‰-1.2%o (2°-6°C) attributable to cooling (Miller et al, 1987a).

Eocene Benthic Foraminifer Faunal Events

Deep-water circulation and temperature changes may have had aprofound effect on benthic foraminifer communities. The most impor-tant benthic foraminifer turnover of the Cenozoic occurred in the latestPaleocene, when approximately 50% of the benthic fauna becameextinct within a few thousand years (Kennett and Stott, 1991). Thisevent was identical in magnitude and timing in the Southern, Pacific,and Atlantic oceans, and was linked to probable deep-water warming andcirculation change (e.g., Tjalsma and Lohmann, 1983; Thomas, 1990;Katz and Miller, 1991; Kennett and Stott, 1991; Pak and Miller, 1992).

A second benthic foraminifer turnover took place during the latemiddle Eocene to earliest Oligocene (P14-P18) (Tjalsma and Lohmann,1983; Corliss and Keigwin, 1986; Miller et al., 1992; Thomas, 1992).Unlike the Paleocene/Eocene extinction event, the middle Eoceneturnover was characterized by a more gradual increase in first and lastoccurrences. In addition, there were important late middle Eocenefaunal abundance changes. The relative abundance of the cosmopoli-tan deep-water species Nuttallides truempyi decreased dramaticallyuntil its last occurrence near the end of the Eocene. In the late middleEocene, Atlantic (Miller, 1983; Tjalsma and Lohmann, 1983; Müller-Merz and Oberhànsli, 1991), Caribbean (Tjalsma and Lohmann, 1983),and subantarctic (Miller et al., 1992) assemblages dominated by Nut-tallides truempyi were replaced by those dominated by Cibicidoidespraemundulus, Oridorsalis spp., Gyroidinoides, and Globocassidu-lina subglobosa. In the Southern Ocean, there was a middle Eocenedecline in the abundance of heavily calcified buliminids (Thomas, 1992).The decrease in the relative abundance of N. truempyi was related toprogressive deep-water cooling throughout the middle and late Eocene(Corliss and Keigwin, 1986; Miller et al., 1992; Thomas, 1992).

Paleogene benthic foraminifer faunal events were also character-ized by bathymetric migrations. In a study of Paleogene benthic fora-minifer faunas of the Atlantic and Caribbean, Tjalsma and Lohmann(1983) noted that faunal assemblages had wide bathymetric tolerancesbefore the late Paleocene extinction. Subsequent to this extinction, twobathymetrically distinct biofacies developed. In the Atlantic andCaribbean, the Nuttallides-àommaiεá assemblage gradually becamerestricted to abyssal depths (>3000 m), whereas the C. praemundulus-dominated assemblage colonized the bathyal region (<3000 m). Thesebiofacies maintained their approximate depth distributions until theend of the middle Eocene. However, in the Pacific and Southern oceansNuttallides truempyi remained a common element of bathyal andlower bathyal assemblages (Thomas, 1990; Pak and Miller, 1992).

Although Paleogene faunal events are relatively well studied inthe Atlantic and Southern oceans, few Pacific Paleogene sites yieldhigh-quality benthic foraminifer records. This study provides infor-mation regarding the timing and extent of the middle Eocene turnoverin the North Pacific and Eocene bathyal and abyssal depth migrations.

METHODS

The North Pacific Transect Leg 145 recovered Paleogene carbon-ates at two sites on the flank of Detroit Seamount (Fig. 1): Site 883(5ril .908'N, 167°46.128'E, 2384 m present depth) and Site 884(51°27.026'N, 168°20.228'E, 3825 m present depth). Samples wereobtained from Sites 883 and 884 at approximately 150-cm intervals(1 sample per section), corresponding to a sampling interval of about100-300 k.y. Each sample was disaggregated in a sodium metaphos-phate and tapwater solution, washed through a 63-µm sieve, and ovendried. Benthic foraminifers were picked from the >150-µm fractionand mounted on reference slides. One sample per section from Site883 and all samples from Site 884 were picked for benthic foraminiferfaunal analysis. In general, between 100 and 300 specimens werepicked per sample. Samples with fewer than 100 individuals wereomitted from the quantitative analysis.

We conducted Q-mode principal component and varimax factoranalyses on the relative abundance (percentage) data, in order toexamine faunal trends in depth and time. Principal components analy-sis uses a cosine theta matrix to determine eigenvalues and eigenvec-tors, normalizing the data to unit length. As a result, Q-mode analysisemphasizes the most abundant taxa.

Our identifications follow the taxonomy of Tjalsma and Lohmann(1983) and van Morkhoven et al. (1986). We follow Berggren andMiller (1989) in recognizing the following benthic foraminifer depthzones: upper bathyal (200-600 m), middle bathyal (600-1000 m),lower bathyal (1000-2000 m), upper abyssal (2000-3000 m), andlower abyssal (>3000 m). Paleodepths for the Detroit Seamount siteswere estimated assuming simple thermal subsidence and empiricalage-depth curves ("backtracking"; e.g., Berger and Winterer, 1974;Detrich et al., 1977; Sclater et al., 1977). The paleodepths of Sites 883and 884 were calculated using a basement age of 70 Ma and asubsidence parameter (k) of 300. Eocene sediments at Site 883 weredeposited in water depths of approximately 1700-2000 m, whereassimilarly aged sediments at Site 884 were deposited in water depthsof 3000-3300m.

Isotopic analyses were performed on the taxa Nuttallides truem-pyi, Cibicidoides praemundulus, and Cibicidoides eocaenus. Sam-ples were ultrasonically cleaned for 2-4 s and roasted at 370°C undervacuum for 1 hr. Isotopic measurements were conducted at Lamont-Doherty Earth Observatory using a Carousel-48 automatic carbonatepreparation device attached to a Finnigan MAT 251 mass spectrome-ter. Analytical error (1 σ), based on 30 analyses on the carbonatestandard NBS20, was 0.07%o and 0.04%o for δ 1 8 θ and δ13C, respec-tively. The data are reported referenced to PDB (Table 1).

We apply no species corrections to our Cibicidoides data. How-ever, our compilation of 125 paired analyses of Paleocene and EoceneNuttalides truempyi and Cibicidoides spp. shows that Nuttallides is

266

ISOTOPIC AND FAUNAL RECORD

60°N

150°E 180° 150°

Figure 1. Map showing locations of sites used in this study.

120°

lower than Cibicidoides by O.l‰ in δ 1 8 θ and ~0.3‰ in δ13C (Fig. 2).These results are consistent with the results of similar compilationsby Katz and Miller (1991) and Pak and Miller (1992). Therefore, sothat we could compare Nuttallides data with Cibicidoides data, wehave adjusted the Nuttallides data by the above correction factors.

One sample per section (-150 cm) was analyzed for percentage ofcarbonate. Carbonate analyses were conducted on 0.5-g subsamplesusing the vacuum-gasometric method of Jones and Kaiteris (1983).This method yields a precision of <1%.

STRATIGRAPHY

Stratigraphic control was provided by the revised shipboard nan-nofossil biostratigraphy (Rea, Basov, Janecek, Palmer-Julson et al.,1993). Additional datum levels were provided by planktonic (thisstudy) and benthic (this study, using the datum levels of Berggren andMiller, 1989) foraminiferal biostratigraphies, as well as oxygen iso-topic stratigraphy (this study). We employ the time scale of Berggrenet al. (1985) to provide ages of the zonal boundaries, with the Paleo-cene/Eocene boundary of Aubry et al. (1988). The ages of our sam-ples were estimated by linear interpolation between datum levels(Table 2). Cande and Kent (1992) have made major revisions (>2m.y., in some cases) to much of the Paleogene geomagnetic polaritytime scale. However, changes in the numeric ages of biostratigraphicdatum levels do not affect the relative timing of events discussed here.

Site 883 recovered upper Paleocene (nannofossil Zone NP7/8 un-differentiated) to upper Oligocene (nannofossil Zone NP25) carbon-ates (Fig. 3). Within this sequence, at least three hiatuses or condensedzones are identified. The well-known latest Paleocene benthic fora-minifer extinction event (e.g., Tjalsma and Lohmann, 1983; Thomas,1990; Katz and Miller, 1991; Pak and Miller, 1992) occurred at approx-imately 813.22 m (between Samples 145-883B-85X-3, 105-110 cm,and 145-883-85-4, 10-15 cm), providing an upper Paleocene datumlevel. However, the uppermost Paleocene and lowermost Eocene sec-tion is highly condensed or missing, where approximately 3 m.y. ofsection is compressed into approximately 1 m (L. Beaufort, pers.comm., 1993). A second short (<l m.y.) hiatus separates the lowerEocene (Samples 145-883B-85X-3, 129-134 cm, to 145-883B-82X-4,95-100 cm; 813.19-784.65 mbsf) from a thick middle Eocene (Sam-ples 145-883B-82X-4, 49-54 cm, to 145-883B-75X-1, 30-35 cm;

784.20-712.90 mbsf). A long hiatus (~5 m.y.) separates the middleEocene section from the lower Oligocene.

The stratigraphy at Site 884 is uncertain for much of the sectionwhere spotty carbonate precluded shipboard biostratigraphies basedon calcareous microfossils. The situation is also complicated by zonesof slumping and core disturbance. However, using a combination ofnannofossil, planktonic, and benthic foraminifer biostratigraphies,we identify a thick lower Eocene section (Fig. 4, Cores 145-884B-87X through -83X). The middle Eocene appears to be condensed atSite 884, comprising only Cores 145-884B-81X and -82X. The mid-dle/upper Eocene boundary is tentatively placed in the lower part ofCore 145-884B-80X (747.3 mbsf; between Samples 145-884B-81X-1, 138-145 cm, and 145-884B-80X-5, 75-80 cm) based on the lastoccurrence (LO) of the benthic foraminifer taxon Aragonia arago-nensis. It appears that a thick upper Eocene section is present at Site884, extending from the base of Core 145-884B-80X to the middle ofCore 145-884B-75X. We suspect that the Eocene/Oligocene bound-ary is unconformable, based on the sedimentological evidence (colorchange, slumping and reworking of benthic foraminifers). However,the biostratigraphic evidence indicates that the hiatus, if any, is short(all nannofossil zones are represented).

RESULTS

Faunal Changes at Site 883

Site 883 had an Eocene paleodepth of approximately 1800 m(lower bathyal). Benthic foraminifer assemblages are similar to depth-calibrated lower bathyal benthic assemblages in the Atlantic (e.g.,Tjalsma and Lohmann, 1983). Thus, faunal results are consistent withthe interpretation that deposition took place within the lower bathyalzone throughout the Paleogene. Examination of the benthic foramin-ifer ranges and the Q-mode factor results indicates that the upperPaleocene to lower Oligocene section at Site 883 is characterized bythree faunal assemblages.

The upper Paleocene section, represented by a single sample (Sam-ple 145-883B-85X-4, 10-15 cm; 813.50 mbsf), was too thin to pro-duce meaningful results in the quantitative analysis. However, thestratigraphic ranges reflect the late Paleocene turnover event (Fig. 5).Of the 33 taxa present in our upper Paleocene sample, 12 occur onlywithin this sample, including Coryphostoma midwayensis, Neoeponides

267


Table 1. δ 1 8 θ and δ13C data from Sites 883 and 884.

Core, section,interval (cm)

Site 884 Cibicidoides spp.:145-884B-

73X-2, 103-10873X-3, 129-13473X-4, 75-8073X-5, 33-3874X-1,0-574X-7, 10-1575X-3, 128-13375X-4, 75-8080X-1,53-5880X-4, 119-124

Site 884 Nuttallides truempyi:14S 884R

75X-4, 75-8075X-5, 140-14575X-6, 84-8975X-7, 12-1777X-3, 73-7877X-5, 2-779X-3, 114-11979X-5, 113-11880X-4, 119-12480X-5, 75-8081X-1, 138-14581X-2, 138-14581X-3, 110-11581X-4, 50-5582X-1,52-5782X-3, 107-11282X-3, 107-11282X-4, 79-8482X-6, 134-13983X-1,72-7783X-2, 26-3183X-3, 10-1583X-4, 51-5683X-6, 51-5683X-7, 65-7084X-1,95-10084X-2, 52-5784X-5, 100-105

Site 883 Cibicidoides spp.:145-883B-

69X-4, 95-10069X-5, 95-10069X-6, 95-10070X-3, 92-9770X-5, 95-10071X-4, 100-10571X-5, 95-10071X-6, 100-10572X-1,95-10072X-2, 95-10072X-3, 95-10072X-4, 95-10072X-5, 95-10073X-1, 95-10073X-2, 95-10073X-3, 50-5573X-3, 95-10073X-4, 95-10073X-4, 140-14573X-5, 95-10073X-6, 95-10074X-1,95-100

Depth(mbsf)

673.53675.29676.25677.33680.60689.70694.48695.45738.93744.09

695.45697.60698.54698.82713.23715.52732.94735.93744.09745.15749.48750.98752.20753.10758.22761.77761.77762.99766.54768.12769.16770.50772.41775.41776.40777.95779.02784.00

660.25661.75663.23668.35671.35679.60681.10682.60684.65686.15687.65689.15690.65694.25695.75696.80697.25698.75699.20700.25701.75703.95

Age(Ma)

34.5734.6534.6934.7434.8936.6537.6037.7239.6339.86

37.7237.8137.8637.8738.5038.6039.3739.5039.8639.9040.7641.3441.8242.1744.1845.5745.5746.6350.9151.0651.1551.2751.4551.7251.8151.9552.0552.50

24.4325.8627.9032.3733.9034.5134.6234.7334.8935.0035.1135.2235.3335.6035.7135.7935.8235.9335.9636.0436.1536.31

δ 8 θ

1.481.721.531.661.561.720.910.310.660.29

0.260.400.100.270.680.430.600.190.130.390.240.020.270.320.000.130.190.07

-0.38-0.12-0.05

0.040.310.380.360.470.14

-0.33

1.681.382.151.852.041.731.251.92.64.82.62.34.39.51.58.76.68.56.69.55

1.62.98

δ 1 3 c

0.500.600.400.400.900.700.800.500.400.50

0.320.31

-0.030.260.380.370.260.560.190.700.18

-0.160.400.49

-0.21-0.17-0.10

0.070.420.08

-0.08-0.10

0.080.030.01

-0.100.050.13

0.280.040.630.06

-0.030.13

-0.060.080.540.920.770.680.800.820.790.880.770.630.890.820.910.95

Core, section,interval (cm)

74X-1, 102-10774X-2, 95-100

Site 883 Nuttallides truempyi:145-883B-

75X-1,30-3575X-1,30-3575X-1,95-10075X-2, 13-1875X-4, 95-10075X-5, 94-9976X-1, 54-5976X4,95-10076X-2, 19-2476X-2, 95-10076X-3, 95-10076X-4, 95-10076X-5, 95-10077X-1, 33-3877X-1, 33-3877X-1, 95-10077X-2, 33-3877X-2, 95-10077X-3, 95-10077X-3, 135-14077X-4, 95-10077X-5, 3 8 ^ 377X-5, 55-6078X-1, 52-5778X-1,52-5778X-1, 122-12778X-2, 100-10578X-3, 50-5578X-3, 95-10078X-3, 140-14578X-4, 51-5678X-4, 51-5678X-4, 94-9978X-4, 127-13278X-5, 91-9678X-6, 25-3078X-6, 95-10079X-1,94-9979X-1, 122-12779X-2, 50-5579X-2, 92-9779X-2, 109-11479X-3, 95-10079X-3,140-14580X-CC81X-CC82X-1,30-3582X-1, 95-10082X-1, 122-12782X-2, 50-5582X-2, 95-10082X-3, 95-10082X-4, 95-10082X-5, 95-10082X-6, 95-10083X-3, 92-9784X-1,45-5084X-1,65-6985X-2, 105-11085X-3, 105-11085X-4, 10-15

Depth(mbsf)

704.20705.45

712.80712.80713.55713.96716.34717.83720.64721.05721.79722.55724.05725.55727.05730.23730.23730.85731.73732.35733.85734.23735.35736.28736.45740.22740.22740.92742.20743.20743.65744.10744.71744.71745.14745.47746.61747.45748.15750.44750.72751.50751.92752.09753.45753.90769.20779.20779.50780.15780.42781.20781.65783.15784.65786.15787.65793.12799.65799.85811.45812.95813.50

Age(Ma)

36.3336.43

42.3042.3042.4342.5142.9343.2043.7043.7743.9044.0344.3044.5744.8445.4045.4045.5145.6745.7846.0546.1246.3146.4846.5147.1847.1847.3047.5347.7147.7947.8747.9847.9848.0548.1148.3248.4748.5949.0049.0549.1949.2649.2949.5349.6150.1551.3551.4151.5351.5851.7351.8252.1054.0254.0754.1254.3254.5654.5654.9857.0958.17

δ 1 8 o

1.841.84

0.230.180.15

-0.130.130.120.140.140.12

-0.050.08

-0.110.270.140.160.020.150.110.120.13

-0.14-0.02

0.18-0.03

0.090.17

-0.200.16

-0.080.150.000.03

-0.010.03

-0.100.130.04

-0.09-0.31

0.05-0.49-0.13

0.030.05

-0.57-0.21-0.60-0.96-0.77-0.60-0.69-0.74-0.99-0.92-0.90-0.95-0.99-1.09-0.91-0.78

0.10

δ 1 3 c

1.351.29

0.250.180.19

-0.100.05

-0.140.060.040.22

-0.030.01

-0.300.01

-0.05-0.05-0.17

0.13-0.07-0.13

0.07-0.16

0.020.06

-0.08-0.02

0.00-0.16-0.15-0.23-0.12-0.18-0.17-0.21-0.17-0.32-0.25-0.12-0.07

0.12-0.14

0.16-0.26-0.04

0.080.050.350.680.380.500.580.510.52

-0.160.260.400.470.09

-0.18-0.29-0.23

1.43

lunata, Neoeponides hillebrandti, Pullenia coryelli, Spiroplectamminasubhaeringensis, Stensioina beccariiformis, Tappanina selmensis, Tri-taxiahavanensis, Tritaxiatrilatera, Bolivinoides delicatulus, and Bulim-ina midwayensis. Most of these species are characteristic of latest Pale-ocene benthic foraminifer assemblages described from other sequences(e.g.,TjalsmaandLohmann, 1983; Miller et al., 1987b; Thomas, 1990;Pak and Miller, 1992). The thin, possibly incomplete, nature of thissection precludes determination of the timing of the turnover.

A lower to middle middle Eocene biofacies is recognized in Q-mode Factor 1 (explaining 36% of the total variance) between Sam-ples 145-883B-85X-3, 129-134 cm (813.19 mbsf), and -78X-1, 52-57 cm (740.22 mbsf; Fig. 6). Factor 1 closely reflects the percentageof Nuttallides truempyi (factor score 95.0), which was generally highin the early and early middle Eocene. Bulimina semicostata/jarvisi

contributes high negative loadings on Factor 1 (-14.6), reflectingresponse to changes in N. truempyi abundance in the early and earlymiddle Eocene. Both of these species diminished in importance in themiddle middle Eocene.

Factor 2 (explaining 15% of the total variance) reflects a middlemiddle Eocene to Oligocene biofacies (Fig. 7). Taxa that contributehigh loadings on Factor 2 include Oridorsalis (60.8), Gyroidinoides(48.2), and Cibicidoides praemundulus (39.8). Percentages of thesetaxa increased from the middle middle through late middle Eocene aspart of a well-known lower bathyal and abyssal faunal abundancechange (e.g., Miller, 1983; Tjalsma and Lohmann, 1983; Müller-Merz and Oberhánsli, 1991; Miller et al., 1992). The early EoceneNuttallides truempyi assemblage was replaced in the late middleEocene by an OridorsalislGyroidinoides, C. praemunduluslGlobo-


cassidulina subglobosa biofacies (e.g., Tjalsma and Lohmann, 1983;Miller et al., 1992). Our data indicate that the timing of this changewas similar in the Atlantic (e.g., Tjalsma and Lohmann, 1983, fig. 50)and Pacific oceans. At Site 883, G. subglobosa does not load heavilyon Factor 2 (-6.8) but, instead, experiences an abundance peak in themiddle middle Eocene (-745-735 mbsf) where it temporarily domi-nates the fauna (Fig. 7).

In the Southern Ocean the middle to late Eocene transition ismarked by the last appearance of the heavily calcified buliminids in thelower bathyal region (Thomas, 1990). At Site 883 several heavily cal-cified species of Bulimina (including Bulimina macilenta, B. calla-hani, and B. velascoensis) last occur in the late middle Eocene (be-tween Samples 145-883B-75X-4, 95-100 cm, and -74X-2, 95-100cm; 716.34-705.45 mbsf). Asingle specimen of Bulimina semicostata/jarvisi within the Oligocene section (Sample 145-883B-73X-4, 95-100 cm; 698.75 mbsf) may be the result of mixing.

The stratigraphic ranges of Site 883 species show an apparentincrease in LOs within Core 145-883B-75X (Fig. 5). However, be-cause approximately 5 m.y. is missing between the middle Eoceneand the lower Oligocene in this core, Site 883 does not show evidenceof an Eocene/Oligocene mass extinction event but may reflect anextended interval of FOs and LOs as reported elsewhere (e.g., Milleret al., 1992).

Faunal Changes at Site 884

Site 884 lies approximately 1300 m deeper than Site 883, havingan abyssal Eocene paleodepth of -3100 m. Comparisons of earlyEocene assemblages from this deep Pacific site with those at Atlanticsites (Tjalsma and Lohmann, 1983; Miller, 1982/3) are consistentwith its abyssal paleodepth, including high abundances of Nuttallidestruempyi, Abyssamina spp. and Buliminella cf. grata. However, thelower abyssal Pacific assemblages differ from those in the Atlantic bythe lower abundances of Alabamina dissonata and Clinapertina spp.and by the high relative abundances of agglutinated taxa. Examina-tion of the Q-mode factor results reveals two distinct biofacies withinthe lower Eocene to Oligocene section, represented by Factors 4(explaining 5% of the total variance) and 2 (explaining 21% of thetotal variance; Figs. 8,9). Factors 1 (48%) and 3 (15%) are dominatedby Nuttallides truempyi and Globocassidulina subglobosa, respec-tively, and are not shown here.

Factor 4 of the Site 884 data identifies a distinctive lower Eoceneagglutinated biofacies (Fig. 8; -810-770 mbsf). Taxa with high load-ings include Bathysiphon sp. (-73.5), Cribrostomoides subglobosus(-31.4), Cydammina sp. (-27.1), and Spiroplectammina spectabilis(-24.1). Although agglutinated species never comprise more than60% of the assemblage, they are an important component of the lowerEocene fauna, as is apparent from the close relationship betweenFactor 4 and the relative abundance of agglutinated taxa (Fig. 8).Agglutinated taxa become less important near the lower/middleEocene boundary.

The early Eocene agglutinated assemblage coexisted with a cal-careous abyssal assemblage. As at Site 883, Nuttallides truempyidominates the calcareous biofacies, contributing up to 40% of thetotal assemblage. The dominance of Nuttallides truempyi is evidentin Q-mode Factor 2 (Fig. 9), to which Nuttallides truempyi contrib-utes high positive loadings (30.6).

Oridorsalis spp. contributes high negative loadings to Factor 2(-88.0; Fig. 9), as this genus becomes increasingly important in themiddle and upper Eocene (-760-715 mbsf). Other important calcare-ous taxa included in Factor 2 are Globocassidulina subglobosa, Ci-bicidoides praemundulus and Gyroidinoides spp. However, unlikeOridorsalis, no clear shift in importance occurs in the latter taxa.

Although the stratigraphic ranges indicate that a benthic foramin-ifer turnover occurred near the Eocene/Oligocene boundary (Fig. 10;13 species last occur between Samples 145-884B-75X-5, 140-145cm, and -75X-3, 128-133 cm), we suspect that the upper Eocene and

Table 2. Age-model parameters, Sites 883 and 884.

Datum level

Site 883:LO Reticulofenestra bisectaLO Reticulofenstra umbilicalate Eocene δ 1 8 θ shiftLO Chiasmolithus solitusLO Discoaster sublodoensisFO Nannotetrina spp.FO Discoaster sublodoensisFO Discoaster lodoensisLO Fasciculithus spp.Benthic extinction eventFO Discoaster mohleri

Site 884:LO Reticulofenstra umbilicaLO Ericsonia formosalate Eocene δ 1 8 θ shiftLO Discoaster saipanensisLO Aragonia aragonensisLO Morozovella aragonensisLO Morozovella caucasicaF0 Cibicidoides grimsdalei

Age(Ma)

23.833.836.8*42.349.750.352.355.056.958.1**60.2

33.834.936.8*37.740.045.951.954.0

Depth(mbsf)

659.6670.0710.9712.8754.4774.2784.2812.0812.9813.2821.4

656.5680.8694.9695.0747.3762.4767.3800.4

Notes: LO = last occurrence and FO = first occurrence. Single asterisk (*) = age cali-brated to Site 690 (Kennett and Stott, 1990); double asterisks (**) = age calibratedto Site 577 (Pak and Miller, 1992).

lower Oligocene sections are separated by an unconformity. We notethe presence of sediment slumping and mixing as well as faunalreworking and suggest a hiatus between Samples 145-884B-75X-4,75-80 cm, and -75X-3, 128-133 cm (-694.96 mbsf; see "Stratig-raphy" section, this chapter).

Early Eocene Faunal Depth Migrations

Comparison of the stratigraphic ranges and factor analysis resultsof Sites 883 (lower bathyal) and 884 (abyssal) reveal that Nuttallidestruempyi dominated the lower bathyal and abyssal realms in thenorthern Pacific during the early Eocene to middle middle Eocene asnoted by Thomas (1992). During the middle middle Eocene, therelative abundance of Nuttallides truempyi decreased in the lowerbathyal region, and was replaced by higher relative abundances ofOridorsalis, Globocassidulina subglobosa, Cibicidoides praemundu-lus, and Gyroidinoides. In the upper abyssal region, the relative abun-dance of Nuttallides truempyi became highly variable after the mid-dle Eocene. Variations in the late Eocene percentage of N. truempyiappear to be in response to changes in Oridorsalis, G. subglobosa,and C. praemundulus or vice versa (Fig. 11). In the lower bathyalregion, most species of heavily calcified buliminids last occur in thelate middle Eocene. In the upper abyssal region, there is no clearpattern of buliminid last occurrences; Bulimina semicostata/jarvisiand Bulimina tuxpamensis last occur in a sample tentatively assignedto earliest late Eocene (Sample 145-884B-80X-5, 75-80 cm; 745.15mbsf), whereas Bulimina impendens persists until the early Oligo-cene. In general, heavily calcified buliminids are less abundant in theupper abyssal than the lower bathyal region.

Among the less abundant taxa, we find that, during the early Eo-cene, there was a distinctive upper abyssal agglutinated assemblage,which persisted until the early/middle Eocene boundary. In the modernocean, the distribution of agglutinated faunas is primarily controlled bythe availability of calcium carbonate (Greiner, 1969). Because manyforms utilize a solution-resistant organic cement, agglutinated fora-minifers are favored in regions with more corrosive deep waters (seeGradstein and Berggren, 1981, for review). Our data indicate thatPacific deep waters were more corrosive during the early Eocene thanthe middle Eocene. This is consistent with reconstructions of theEocene Pacific calcite compensation depth (van Andel et al., 1975).

Abyssaminids, including Abyssamina poagi, Abyssamina quad-rata, and Quadrimorphina profunda, were present in both the lowerbathyal and abyssal realms in the Pacific until the early/middle Eo-

269


3

Cibicidoidβs δ 1 8 θ

2

1

U

-1

y = -0.29 + 0.94X r*=0.90

• M£

4

D

•

•D

•

O

A

A

«

a

Site 690

Site 738

Site 401

Site 702

Site 525

Site 527

Site 528

Site 384

Site 550

Site 549

-

- 1 0 1 2 3Cibicidoides δ13C

Figure 2. Paired Nuttallides truempyi and Cibicidoides spp. oxygen and carbonisotope analyses. Data is from Kennett and Stott (1990) (Site 690), Barrera andHuber (1991) (Site 738), Pak and Miller (1992) (Site 401), Katz and Miller(1991) (Sites 702 and 384), Shackleton et al. (1984) (Sites 525,527, and 528),and Pak and Miller (unpubl. data) (Sites 550 and 549). Regression equationsare drawn in the solid line.

cene boundary, at which time they became restricted to abyssal depths(Fig. 12). A similar migration event took place in the Atlantic, whereabyssaminid species displayed wide bathymetric ranges in the earlyEocene and gradually became restricted to deep-water sites in themiddle Eocene (Tjalsma and Lohmann, 1983).

Pacific Paleogene Isotope Record

Together, Sites 883 and 884 provide relatively complete coverageof the Eocene isotopic record, providing a previously unattainableopportunity to monitor Pacific deep-water isotopic changes. The deepPacific Ocean is the largest part of the global carbon budget, compris-ing >50% of the modern reservoir (Baumgartner and Reichel, 1975).This was particularly true during the Eocene, when the Pacific wasmuch larger than it is today. As a result, inferences regarding deepocean circulation and ocean nutrient inventories rely on the Pacificrecord to approximate mean ocean values.

We compare Site 883 oxygen and carbon isotopic values to thoseat nearby Shatsky Rise Site 577 (32°26.52'N, 157°43.40/E, 2675 mpresent depth, Eocene paleodepth -2000 m; Miller et al., 1987b; Pakand Miller, 1992). Site 577 provides a better late Paleocene to earlyEocene record, but lacks middle-upper Eocene sediments (Miller etal., 1987b; Pak and Miller, 1992). In intervals of overlap, Site 883δ 1 8θ and δ13C values are similar to Site 577, indicating that the recordsmay be combined to form a composite isotope record for the Pacific(Figs. 13, 14). Although we are concerned with the possibility of dia-genesis at the deeply buried Detroit Seamount sites, the close agree-ment between Sites 883 and 577 (burial depth -150 m) in both carbonand oxygen isotope values indicates that diagenesis has not beensevere. If significant recrystallization had taken place, we would ex-pect anomalously low δ 1 8 θ values in the affected intervals (Vernaud-Grazzini et al., 1979). We see no such anomalous values throughoutthe studied section.

The composite Pacific record provides a relatively complete iso-topic record for the late Paleocene to the early Oligocene. However,hiatuses were detected, such as a probable short gap across the Eo-cene/Oligocene boundary (see "Stratigraphy" section, this chapter).Additional hiatuses may be undetected because chronostratigraphiccontrol on these sections is tentative (i.e., many of the calcareousmarker taxa are dissolved and the sections lack magnetostratigraphiccontrol). In addition, core recovery was poor in some intervals (e.g.,Cores 145-883B-83X to -85X). Despite these limitations, the com-posite δ 1 8θ and δ13C records (Figs. 13 and 14) provide a first approxi-mation of Pacific deep-water isotopic changes.

Comparison between the Pacific δ 1 8θ record and the Atlantic ben-thic δ 1 8 θ record (Miller et al., 1987a, fig. 1) indicates that both theamplitude and magnitude of the well-known isotopic shifts are pre-served in the Pacific record. Benthic foraminifer δ 1 8 θ values decreasedfrom ~0.5‰ in the late Paleocene to peak negative values of 1.0‰ inthe early Eocene. A sharp δ 1 8 θ increase of almost l.O‰ occurred nearthe early/middle Eocene boundary. This increase marked the begin-ning of a gradual increase in δ 1 8 θ that continued throughout the re-mainder of the Eocene. The δ 1 8 θ values increased sharply by ~1.5%eacross the Eocene/Oligocene boundary, reaching values of ~1.8‰. Nodistinct increase is noted in the late middle Eocene, as reported in theAtlantic (Miller, 1992) and Southern oceans (Kennett and Stott, 1990;Katz and Miller, 1991; Miller, 1992), although this could be the resultof poor coverage between ca. 41 and 39 Ma. In general, δ 1 8 θ valuesappear to increase over this interval (Fig. 13).

Isotopic comparisons between Sites 883 and 884 provide an oppor-tunity to examine depth-related differences in the Eocene deep Pacific.Specifically, we compared the late early Eocene and middle Eoceneintervals of overlap between the lower bathyal Site 883 and the upperabyssal Site 884 (Figs. 13 and 14). During the middle Eocene therewere no significant differences in either δ'8O or δ13C of benthic fora-minifer calcite, suggesting that the deep North Pacific (between -2000m and 3400 m) was not strongly stratified at this time. Comparisonsbetween late early to early middle Eocene (-52.5-51 Ma) isotopes atSites 883 and 884 show no significant differences in δ 1 8θ. However,δ13C values at Site 884 were approximately 0.5‰ lower than at Site883, suggesting late early to early middle Eocene (-52-51.5 Ma)deep-water-mass stratification in the North Pacific.

We compare our composite Pacific isotope record to the SouthernOcean record of Site 690 (Maud Rise, 65°09.63'S, l°12.30'E, 2925 mpresent depth, Eocene paleodepth -2000 m; Kennett and Stott, 1990)to address possible Eocene deep-water circulation changes (Figs. 15,16). We note that the chronology of the Eocene section at Site 690 iscomplicated by hiatuses, as well as the absence of planktonic forami-niferal index species. As a result, several different interpretations of theEocene magnetics have been published (Kennett and Stott, 1990;Speiss, 1990; Thomas, 1990). These three interpretations differ indetail, but agree that all magnetic reversals between Chrons C22n andC25n are represented at Site 690. Based on calcareous nannofossilevidence, M.-P. Aubry et al. (unpubl. data, 1994) have made an inter-

270

C/5oü

ery

ooΦ

oc

o

'cΦ

XI

nife

r

εo

sno

e

COücσo

foss

i

occCO

c

δ 1 8 θ2 1 0


δ1 3C-1

700-

Q.ΦQ

750- «

800-

> '

Figure 3. Stratigraphic summary of Hole 883B showing series, core recovery, benthic foraminifer zonation, calcareous nannofossil zonation, and uncorrectedisotope data. Open squares represent data from Nuttallides truempyi; solid squares represent Cibicidoides spp.

pretation that is substantially different, in that they think that there areat least three major hiatuses in the lower Eocene section. An age modelbased on the M.-P. Aubry et al. (unpubl. data, 1994) stratigraphy (Table3), indicates that very little lower Eocene section is preserved at Site690 (Figs. 17,18). If the Aubry et al. stratigraphy is correct, few com-parisons can be made between the early Eocene Pacific and SouthernOcean isotopic records. An additional Southern Ocean record wouldbe needed to validate some of the conclusions drawn here for the earlyEocene. However, in this study we employ the published magneto-stratigraphy-based age model of Kennett and Stott (1990) to comparethe Pacific isotope record to the Southern Ocean record of Site 690.

During the late Paleocene (58-55 Ma), the Pacific and Southernoceans were remarkably similar in δ 1 8θ, indicating similar deep-water temperatures and salinities. At the same time, the Pacific hadlower δ13C values than did the Southern Ocean, supporting the con-clusions of Pak and Miller (1992) that the Southern Ocean was thedominant source of deep water during the late Paleocene.

During the early Eocene, Pacific δ 1 8 θ values were 0.5%c-0.75%olower than Southern Ocean values, whereas δ13C values were 0.5‰-1 Xf/oo lower. These data indicate that, although the deep Pacific andSouthern oceans were bathed by different water masses, the SouthernOcean remained a source of colder and/or more saline deep waters.This is consistent with previous results (Miller et al., 1987b; Katz andMiller, 1991; Pak and Miller, 1992).

Table 3. Age-model parameters, Site 690, after stratigraphy of M.-P.Aubry (pers. comm., 1994).

Datum levelAge(Ma)

Depth(mbsf)

50.3-53.7

-55.5

56.758.6

123.63132.16

135.91

149.2185.47

Base of C2InSediment missing: level between NP 12 (reversed) and NP 14a

(normal)Sediment missing: level younger than LO Tribrachiatus

contortus, older than LO Discoaster lodoensis, within C24nlrLO Tribrachiatus bramletteiTopofC25n

Note: LO = last occurrence.

By providing constraints on the Pacific δ13C composition, theDetroit Seamount isotopic record allows the reconstruction of middleEocene deep-water circulation. From the early middle to middlemiddle Eocene (49-44 Ma), δ 1 8 θ values in the Pacific and Southernoceans were identical. Pacific δ13C values for the same interval were0.5%o-0.75‰ lower than in the Southern Ocean. This relationshipindicates that, as in the late Paleocene, the Southern Ocean repre-sented the single dominant source of deep water during the earlymiddle Eocene.

271


ocΦO

Φ

OOΦ

0

[hie

min

i!

C CO

-3 f/1

are

ono

fo;

S ^δ i β 0 δ 1 3 C

O DC

700

XI

ε

£750-

800-

AB8 NP22

NP21

NP19/NP20

NP18'

NP7?

T3CDCoN

P10-P13

P8-P9

Figure 4. Stratigraphic summary of Hole 884B showing series, core recovery, benthic foraminifer zonation, calcareous nannofossil zonation, planktonic foraminiferzonation, and uncorrected isotope data. Open squares represent data from NuttalUdes truempyi; solid squares represent Cibicidoides spp.

From the middle middle Eocene to the Eocene/Oligocene bound-ary (46-36.5 Ma), the relationship between Pacific and SouthernOcean δ 1 8 θ and δ13C returned to a configuration much like that of theearly Eocene. Southern Ocean δ 1 8 θ was ~0.5‰ heavier than in thePacific in both δ 1 8θ and δ13C. These data indicate that the SouthernOcean was bathed by a cooler, more nutrient-depleted (i.e., closer tosource) deep water than that which influenced the deep Pacific. It alsoimplies that, although relatively warmer waters filled the SouthernOcean in the early middle Eocene, by the late middle Eocene South-ern Ocean waters were again relatively cooler.

DISCUSSION

Isotopes

Carbon isotopic reconstructions indicate that the Southern Oceanwas a source of deep water throughout the late Paleocene and most ofthe Eocene, with two possible exceptions: near the Paleocene/Eoceneboundary, and near the early/middle Eocene boundary. At these twotimes, Pacific and Southern Ocean δ13C values converged. At all othertimes, Southern Ocean δ13C values were significantly (0.5‰-0.75%c)higher than Pacific δ13C values. Given that the δ13C of foraminifercalcite decreases with increasing nutrients in deep water (Belanger etal, 1981; Graham et al., 1981), relatively high δ13C values indicate

that the Southern Ocean was proximal to a more nutrient-depletedsource of deep water than the Pacific.

A Southern Ocean/Pacific δ'3C gradient does not necessarily implythat the two basins were influenced by different deep-water masses.The gradient may simply reflect aging along a circulation path from anutrient-depleted source area to the more nutrient-enriched "end of thecirculation route." However, oxygen isotopic results help constraindeep-water source areas. Unlike δ13C, δ 1 8 θ values are not affected byaging; a deep-water mass retains its δ'8O signature unless its tempera-ture and/or salinity is changed by mixing or diffusion. Deep-water(2000-3400 m) δ 1 8 θ values between the Southern and North Pacificoceans were identical in the late Paleocene, early middle Eocene, andearly Oligocene. This similarity supports the conclusion that the twobasins were bathed by a water mass originating in a single source areaproximal to the Southern Ocean.

Our comparisons between the Southern and Pacific ocean isotopicdata support the results of previous studies. Isotopic reconstructionsof Miller et al. (1987b), Katz and Miller (1991), and Pak and Miller(1992) all indicated that the late Paleocene Southern Ocean was asource of relatively nutrient-depleted, cool deep water. Previous stud-ies (Kennett and Stott, 1990; Katz and Miller, 1991; Pak and Miller,1992) suggested that the flux of Southern Ocean deep water wasreduced or eliminated for a brief period in the latest Paleocene at thetime of the much-heralded carbon isotopic excursion (see Kennett

272


5-Si sT3 . _ J -

3& S I

!lü

680

& 1*1 SJ**

s i « .Sj,

§ liüfiii.T - O <u <u a S 5> S r £

in p ii j p flipi1 βl § §N S i !* i ! l i ! i§ Si s l l is i l

s nil atl ii Hi if mir-

S a.à£ S S.<J «.S 3*6, ̂ J g£,,; w

•a-a a•s.

I §s SSJ s §^ g^ | ^s |^ g s J•52 «§

&s.S¥|Ss -SE» to t<, t<] t>i tn-S Sn 3 g

1t θ

4

4

1

11

1

4 1

4

1

I

4

4

1

1

11

() 1

4I 1

1

<

j4

4

1

1 <

(

) 1

4I

I

I

1

1

1 4

i

1 41 1

i 1

1 4

1

1

I

1 4I

I

11

4

1

1 1

(

4

4

iii

I 4

4ii

4

1 1

4

4

i1 4

1<

>

1

| t1

1

t I

1 i

| •

1 1

t i

I {

I (

1i>1

!

t

*

i

4

(i

1

1

I

1

1

1

1

4

4

1

1I

(

|

|

1

|

|

1

( 11 |

1 O 1

1

1

I1

1

i 4

1 l

4

4

1

t 4

44

4

}

1

i

t 1

I

l <

l

I

i

l

Ii

t 1» <

1 14

<

(11

1 {<

) 4

1 4

4

4

4

J

14

4

4

4

11 1

> 1i 4

1

1

1t 1

1 1

11

I 4i ii i> <

1 1

> 4

1 I

» |

1

4

4

4

1

i 1

II

1

1

r• i

1| I

I)

11 1

1

1

) {

•

•

|

|

1\\

|

|

\

1

1

1

i

1 iiii

il

ii

\

I i

<

<

i

Ii

T (4

i

1

|j

{

1 4

| 4

i1 4

•

•

i

•

•

)

(

I

1 4

I

I\

\

i

I

i

4

4

4

i

14

fr 4

1 4

4

i4I

i

ft

ft

(

(

1|

|11 1

1

1

•

1

4

<

|

4

1<

1

ft i

i

\

I

ft 4

ft <

ft 4ft 1

1

ii

4

I

1

\

1

11

1

4

I

t1

I I

II {

1

1 1

1 1

» 4

I

\

I

1

1

tI

Ift 4

i 4

l

4

4

4

4

4

4

4

4

4

1> i

4

1t 1

I <

1

1

4

l

I

4

I1 |

r 1

• (

t11

1I

<

I 4

1 1

| 4

ft 4

; 'i

1 1

1 1

<414

1

1

4

I

4

4

4

I

1

4

4<

1

I

t I

i

<

1

II

<

4

I

820

Figure 5. Stratigraphic range chart of benthic foraminifers from Hole 883B. Circles indicate the position of samples. Open circles indicate single occurrences of species.

and Stott, 1991, for discussion of the excursion). Although our datalack the resolution to address the isotopic excursion in particular, weprovide evidence to support a major deep-water circulation change atthis time.

Early Eocene (57-52 Ma) δ'3C values indicate the continued sup-ply of a nutrient-depleted Southern Ocean source of deep water. How-ever, oxygen isotopic values diverged through this interval, withSouthern Ocean values ~0.5%c higher than Pacific values. This rela-tionship indicates that two sources of deep water must have operatedsimultaneously, providing different 618O signatures to the Pacific andSouthern oceans. Although the isotopic reconstructions suggest thatthe cool, nutrient-depleted source originated in the Southern Ocean,we cannot establish the source region of the relatively warm, nutrient-enriched water mass that influenced the Pacific sites. We interpret thatthe North Pacific was not the source, because the relatively low δ13Csignature suggests an "old" water mass. On the other hand, the rela-tively low 618O signature suggests a warmer (or less saline) water than

the Southern Ocean. Much attention has been given to the possibilityof a source of warm saline deep water originating in the Tethyanregion (Brass et al, 1982; Woodruff and Savin, 1989; Kennett andStott, 1990; Pak and Miller, 1992). Our data support, but do notconfirm, the presence of an early Eocene Tethyan deep water. Sites ofEocene age from within the Tethyan region are needed to identifyTethyan outflow.

The early/middle Eocene boundary is marked by a sharp δ 1 8 θincrease. This increase is widely recognized (see Miller, 1992, forreview) and represents the first step of the long Cenozoic coolingtrend. Comparisons between the Pacific and Southern Ocean isotopicrecords suggest that the cooling step was accompanied by a deep-water circulation change. Concurrent with the cooling step, Southernand Pacific ocean δ 1 8θ and δ13C values converged. The δ13C conver-gence was short term (1 m.y.), after which the Southern Ocean-Pacific gradient returned. The δ 1 8 θ convergence lasted ~6 m.y. (52-46 Ma). In the middle middle Eocene (-46 Ma), Southern Ocean

273


Factor 1

0 0.2 0.4 0.6 0.8

Nuttallidestruempyi (%)

10 20 30 40

Bulimina Globocassidulinasemicostata/jarvisi (%) subglobosa (%)

10 20 30 40 10 20 30 40680

700

720

780

800

8201

-14.6

r~

///

-6.8

CD

CDOOCD

α>cCD

8LU_CD•o;•σ

E

Φ

cΦüO

LU

"8

Figure 6. Q-mode Factor 1 of the Site 883 data set compared with the relative abundance (percentage) of species contributing to this component. Factor scores ofeach species are shown in the respective columns.

δ'8O returned to values ~0.5‰ higher than Pacific δ 1 8 θ. This relation-ship persisted until at least the Eocene/Oligocene boundary(-36.5 Ma).

On the basis of our Pacific-Southern Ocean isotopic comparisons,we suggest that the Eocene ocean alternated between two dominantcirculation modes. The first, represented by the late Paleocene andearly middle Eocene, is characterized by similar Pacific and SouthernOcean δ'8O values, indicating a single source of deep water. Rela-tively high Southern Ocean δ'3C values suggest that the source origi-nated in or near the Southern Ocean. The second circulation mode isrepresented by the early Eocene and the middle middle to late Eocene,and is characterized by relatively high δ 1 8 θ values in the SouthernOcean. This relationship suggests two sources of deep water: a coolsource, originating in the Southern Ocean, and a warm source, prob-ably originating in the mid-latitudes.

Benthic Foraminifers

Although previous studies (primarily from Atlantic sites; e.g.,Tjalsma and Lohmann, 1983) have shown that benthic foraminiferfaunal assemblages became increasingly depth stratified throughoutthe early and middle Eocene, the Detroit Seamount sites do not showsuch a stratification. Although the two sites are separated by approxi-mately 1300 m in water depth, assemblages at the two sites wereremarkably similar. Both sites show a middle Eocene change fromassemblages dominated by Nuttallides truempyi to assemblages domi-

nated by Oridorsalis, Globocassidulina subglobosa, Cibicidoidespraemundulus, and Gyroidinoides spp.

Isotopic comparisons between Sites 883 and 884 support the con-cept of a relatively undifferentiated Eocene deep North Pacific. Wherethe sequences overlap (early/middle Eocene boundary, middle Eocene,earliest Oligocene), both oxygen and carbon isotopic values are virtu-ally identical (Figs. 13 and 14). This indicates that the same water massbathed regions of 2000 and 3300 m (middle Eocene paleodepths ofSites 883 and 884, respectively) in the North Pacific. Both the faunaland isotopic data suggest that there was little deep-water stratificationin the middle to late Eocene and early Oligocene.

The one exception is the early Eocene, when the deep Site 884 wascharacterized by a distinctive agglutinated fauna not observed at Site883. Similar deep-sea agglutinated faunas have been described fromPaleogene abyssal sequences in the Labrador and North seas(Gradstein and Berggren, 1981; Miller et al., 1982; Kaminski et al.,1989). The Labrador and North sea faunas comprise diverse aggluti-nated assemblages in Maastrichtian through Eocene sediments. Theirdisappearance from the Labrador and North seas is attributed to theincrease in bottom-water oxygenation and decrease in bottom-watercorrosivity associated with the onset of vigorous abyssal circulationin the late Eocene/early Oligocene (Gradstein and Berggren, 1981;Miller et al., 1982; Kaminski et al., 1989).

Similarly, we suggest that Pacific deep waters were more corro-sive during the early Eocene and less corrosive during the middleEocene. This is supported by the carbonate percentage data, which


Factor 2

0.2 0.6

Cibicidoides GlobocassidulinaOridorsalis spp. (%) Gyroidinoides spp. (%) praβmundulus (%) subglobosa (%)

10 20 30 40

820

Figure 7. Q-mode Factor 2 of the Site 883 data set compared with the relative abundance (percentage) of species contributing to this component,

each species are shown in the respective columns.

Factor scores of

Agglutinated foraminifers (%)

0 10 20 30 40 50

Factor 4-0.8 -0.6 -0.4 -0.2 0

-=-750-

5% explained

Ba<hysiphon -0735Cribrostαnoides subglobosvs -0J14Cydamrrina sp. -0-271Spr<spleαaimmi speclaMb 0.241

Factor 2

-1.0 -0.8 -0.6 -0.4 -0.2

Nuttallidβs truempyi (%) Oridorsalis sp. (%)

10 20 30 10 20 30 40 50

800

Figure 8. Relative abundance of agglutinated taxa at Site 884 compared withQ-mode Factor 4 of the Site 884 data set.

Figure 9. Q-mode Factor 2 of the Site 884 data set compared with the relativeabundance (percentage) of species contributing to this component. Factorscores of each species are shown in the respective columns.

275


70011

4

4

4

1I

1 1

<

<

(

(

1

1

(

| |

1

|

1

1 i

{

I

1 <

I

1

i

1

1

|(i

' 1i

<

I

1

\ *

1

1

1 1

1 1

H1

1 1

[

| 1

1 1

1

i -

> I

i

4

|

i

- *

1 1

l

ii

i

i

i

i

11

: ,

!

1

{

{

(

~ i

I

1

1

{

i '

(

1

11

1> 1

1» 1

I i

I

1

l

P

I

4

* 4

i

i

i

I

1

- ^

•

1

: «

i

•

ni

i

)

i

i

, i

ii11

1 1

i

i

•

i

i

i

* {t 4

i

! i

' !

i i

1 4

1 1

i

i

1 l

I '

• 1

1i

1 i

1 11 1

i

(

1

l

i

M

1 1

1 1

1t

I 1

I 1t 1

\

)1 1

1

1I 1

11

i (

l

l

t

1

1

(

1

(

1

1> i

1

<

1

1

(

4

1

(

' 1

1 1

; i

i

, i11

> i

<

i

>i

11i

i <

i i

i i

ii

i i

I

<

> ii

> i

i i

i

i

i

i

)

i

i

•

. ^

|

i

i

I

<

i

i

i

i

ii

; :

i

i

i

yo

LU<DQ.Q.

"DC

•isLU

Figure 10. Stratigraphic range chart of benthic foraminifers from Site 884. Pluses indicate the presence of suspected reworked taxa.

show generally less carbonate in the early Eocene than in the youngersection (Fig. 17). However, although very low carbonate values cor-relate with high percentages of agglutinated foraminifers, postdepo-sitional dissolution is clearly not the sole factor controlling the earlyEocene agglutinated assemblage, because (1) dissolution-resistantcalcareous species (e.g., Gyroidinoid.es; Corliss and Honjo, 1981) arenot preferentially preserved in this section; and (2) the percentage ofearly Eocene agglutinated taxa is high even when the percentage ofcarbonate is relatively high (Fig. 19). We think that the presence ofthe early Eocene agglutinated assemblage reflects the ability of thesespecies to survive under more corrosive conditions, rather than post-depositional dissolution of calcareous species below the lysocline.The agglutinated assemblage decreased dramatically near the early/middle Eocene boundary (Fig. 8). There was a simultaneous impor-tant benthic foraminiferal depth migration as abyssaminid speciesmigrated from lower bathyal to abyssal depths (Fig. 12).

The faunal changes near the early/middle Eocene boundary arelinked to deep-water changes. The first Cenozoic deep-water coolingstep, as indicated by the benthic foraminifer δ 1 8θ record, was coinci-dent with the faunal changes. In addition, the first Eocene convergenceof Pacific and Southern ocean δ13C values, which we interpret as thechange from a two-source to a one-source circulation system, alsooccurred at the early/middle Eocene boundary. We speculate that ben-

thic foraminifer assemblages responded to both deep-water coolingand circulation change. After the early/middle Eocene cooling wasmore gradual, perhaps allowing the benthic fauna to become moregradually conditioned to the colder environment. Unfortunately, we donot yet have isotopic data from Site 884 that will enable us to determinethe relative temperature of the deep site during the early Eocene.Nuttallides truempyi also apparently responded to the deep-watercooling trend. As noted by Miller et al. (1992) in subantarctic Site 702,the relative abundance of N. truempyi decreased during the early andmiddle Eocene in concert with the increase in 518O (Fig. 20).

The absence of the agglutinated assemblage from the shallowersite suggests that it preferred the presumably colder conditions of theabyssal region. Therefore, we cannot attribute its early/middleEocene decline to deep-water cooling. Instead, we think that thefaunal change was caused by changes in other deep-water parameters,probably decreased corrosivity, as the circulation system shifted fromone mode to another.

Deep-water circulation changes have the potential to affect ben-thic foraminifer communities profoundly. The dramatic latest Paleo-cene benthic foraminifer extinction event was strongly linked todeep-water warming and a probable shift in circulation (e.g., Kennettand Stott, 1990; Thomas, 1990; Katz and Miller, 1991; Pak andMiller, 1992). The latest Paleocene faunal changes coincided with a

276


20 40 60650

700 •

750 •

800 •

850

1 8ELLJ

Nuttallidβs truempyi (%)

C. praβmundutus/G. subglobosa/Oridorsalis (%)

Figure 11. Comparison between the relative abundance of Nuttαllides truempyi(solid circles) and the combined relative abundance of Cibicidoides prαemun-dulus, Globocαssidulinα subglobosα, and Oridorsαlis spp. (open circles) atSite 884.

divergence of Pacific and Southern ocean δ 1 8θ values (Fig. 15), whichcan be interpreted as a shift from a one-source circulation mode to atwo-source mode. We also note that the early Eocene was bracketedby anomalous short-term carbon isotopic convergences, which mayherald circulation shifts, as well as by both benthic foraminifer faunalevents and oxygen isotopic events.

CONCLUSIONS

1. Leg 145 Sites 883 and 884 provide the first detailed middle andupper Eocene isotope record from the North Pacific, although poorrecovery and hiatuses still plague our records. Despite these limita-tions, these new data allow the construction of a composite Pacificδ 1 8 θ record that is similar in both amplitude and magnitude to theAtlantic δ'8O record. The Cenozoic cooling trend is punctuated by atleast two strong δ 1 8θ increases: a step of ~l .0%c near the early middleEocene boundary, and a step of ~1.5%o at the Eocene/Oligoceneboundary. In addition, a third δ'8O increase of lesser magnitude maybe present near the middle/late Eocene boundary.

2. Eocene climatic cooling was accompanied by several shifts indeep-water circulation. Interbasinal isotopic comparisons indicatethat the Eocene ocean alternated between two dominant circulationmodes. During the early Eocene (-57-52 Ma) and the middle middleEocene to Oligocene (-46-36.5 Ma), two sources of deep water werein operation: a cool, nutrient-depleted source originating in the South-ern Ocean and a warmer, more nutrient-enriched source probablyoriginating in the subtropics. However, during the early to middlemiddle Eocene (52-46 Ma), the deep North Pacific was influenced bya single source of deep water originating in the Southern Ocean. Ourdata indicate that, although there were times of probable mid-latitudedeep-water production, the Southern Ocean remained a source regionthroughout most of the Paleogene.

3. North Pacific faunal changes indicate that early Eocene deepwaters (2000-3300 m) were distinctly depth-stratified. However, iso-topic and faunal changes indicate that there was little deep-water-mass stratification in the middle Eocene. This is in contrast to Atlanticforaminifer faunal studies (i.e., Tjalsma and Lohmann, 1983) which

Abyssaminid group (%)

10 20 30

Figure 12. Comparison of the relative abundance of abyssaminid group fora-minifers (Abyssαminα poαgi, Quαdrimorphinα profundα, and Abyssαminαquαdrαtα) at Sites 883 (solid circles) and 884 (open circles), showing themigration of these taxa from the upper abyssal to the lower bathyal realm nearthe early/middle Eocene boundary.

suggest that benthic foraminifers became progressively depth-strati-fied after the latest Paleocene extinction event.

4. Changes in Eocene benthic foraminifer assemblages were re-lated to both deep-water cooling and circulation events. The early/mid-dle Eocene migration of abyssaminid species from lower bathyal toabyssal depths coincided with the first Cenozoic cooling step. Also, therelative abundance of Nuttαllides truempyi decreased in response tolong-term Eocene cooling. Finally, the disappearance of the earlyEocene agglutinated assemblage from the North Pacific coincided withan inferred shift in circulation from a two-source to a one-source mode.

ACKNOWLEDGMENTS

Reviews by E. Thomas and J. Kennett greatly improved the manu-script. We thank M. Katz, L. Beaufort and M.-P. Aubry for helpfuldiscussions. This work was partially supported by a JOI/USSACshipboard Ocean Drilling Fellowship and by JOI/USSAC grant 145-20713. This is Lamont-Doherty Earth Observatory contribution num-ber 5350.

REFERENCES*

Aubry, M.-P., Berggren, W.A., Kent, D.V., Flynn, J.J., Klitgord, K.D.,Obradovich, J.D., andProthero, D.R., 1988. Paleogene geochronology: anintegrated approach. Pαleoceαnogrαphy, 3:707-742.

Abbreviations for names of organizations and publications in ODP reference lists followthe style given in Chemical Abstracts Service Source Index (published by AmericanChemical Society).

277


Barrera, E., and Huber, B.T., 1991. Paleogene and early Neogene oceanog-raphy of the southern Indian Ocean: Leg 119 foraminifer stable isotoperesults. In Barron, J., Larsen, B., et al., Proc. ODP, Sci. Results, 119:College Station, TX (Ocean Drilling Program), 693-717.

Barron, E.J., 1987. Eocene equator-to-pole surface ocean temperatures: asignificant climate problem. Paleoceanography, 2:729-739.

Barron, E.J., and Peterson, W.H., 1991. The Cenozoic ocean circulation basedon ocean General Circulation Model results. Palaeogeogn, Palaeoclima-tol., PalaeoecoL, 83:1-28.

Barron, J., Larsen, B., and Baldauf, J.G., 1991. Evidence for late Eocene toearly Oligocene Antarctic glaciation and observations on late Neogeneglacial history of Antarctica: results from Leg 119. In Barron, J., Larsen,B., et al., Proc. ODP, Sci. Results, 119: College Station, TX (Ocean DrillingProgram), 869-891.

Baumgartner, A., and Reichel, P., 1975. The World Water Balance: New York(Elsevier).

Belanger, P.E., Curry, W.B., and Matthews, R.K., 1981. Core-top evaluationof benthic foraminiferal isotopic ratios for paleoceanographic interpreta-tions. Palaeogeogr., Palaeoclimatol, PalaeoecoL, 33:205-220.

Berger, W.H., and Winterer, E.L., 1974. Plate stratigraphy and the fluctuatingcarbonate line. In Hsü, K.J., and Jenkyns, H.C. (Eds.), Pelagic Sedimentson Land and Under the Sea. Spec. Publ. Int. Assoc. Sedimentol., 1:11—4-8.

Berggren, W.A., Kent, D.V., and Flynn, J.J., 1985. Jurassic to Paleogene: Part2. Paleogene geochronology and chronostratigraphy. In Snelling, NJ.(Ed.), The Chronology of the Geological Record. Geol. Soc. LondonMem., 10:141-195.

Berggren, W.A., and Miller, K.G., 1989. Cenozoic bathyal and abyssal cal-careous benthic foraminiferal zonation. Micropaleontology, 35:308-320.

Brass, G.W., Southam, J.R., and Peterson, W.H., 1982. Warm saline bottomwater in the ancient ocean. Nature, 296:620-623.

Broecker, W.S., and Denton, G.H., 1989. The role of ocean-atmosphere reor-ganizations in glacial cycles. Geochim. Cosmochim. Acta, 53:2465-2501.

Cande, S.C., and Kent, D.V., 1992. A new geomagnetic polarity time-scale forthe late Cretaceous and Cenozoic. /. Geophys. Res., 97:13917-13951.

Chamberlin, T C , 1906. On a possible reversal of deep-sea circulation and itsinfluence on geologic climates. J. Geol., 14:363-373.

Corliss, B.H., and Honjo, S., 1981. Dissolution of deep-sea benthonicforaminifera. Micropaleontology, 27:356-378.

Corliss, B.H., and Keigwin, L.D., 1986. Eocene-Oligocene paleoceanography.In Hsü, K.J. (Ed.), Mesozoic and Cenozoic Oceans. Am. Geophys. Union,Geodyn. Ser., 15:101-118.

Derrick, R.S., Sclater, J.G., and Thiede, J., 1977. The subsidence of aseismicridges. Earth Planet. Sci. Lett., 34:185-198.

Gradstein, F.M., and Berggren, WA., 1981. Flysch-type agglutinated forami-nifera and the Maastrichtian to Paleogene history of the Labrador andNorth seas. Mar. Micropaleontol, 6:211-268.

Graham, D.W., Corliss, B.H., Bender, M.L., and Keigwin, L.D., Jr., 1981.Carbon and oxygen isotopic disequilibria of Recent deep-sea benthicforaminifera. Mar. Micropaleontol, 6:483-497.

Greiner, G.O., 1969. Recent benthonic foraminifera: environmental factorscontrolling their distribution. Nature, 223:168-170.

Jones, G.A., and Kaiteris, P., 1983. A vacuum-gasometric technique for rapidand precise analysis of calcium carbonate in sediments and soils. J.Sediment. Petrol, 53:655-660.

Kaminski, M.A., Gradstein, F.M., and Berggren, WA., 1989. Paleogene benthicforaminifer biostratigraphy and paleoecology at Site 647, Southern Labra-dor Sea. In Srivastava, S.P., Arthur, M.A., Clement, B., et al., Proc. ODP,Sci. Results, 105: College Station, TX (Ocean Drilling Program), 705-730.

Katz, M.E., and Miller, K.G., 1991. Early Paleogene benthic foraminiferalassemblages and stable isotopes in the Southern Ocean. In Ciesielski, RF.,Kristoffersen, Y., et al, Proc. ODP, Sci. Results, 114: College Station, TX(Ocean Drilling Program), 481-512.

Keigwin, L.D., and Corliss, B.H., 1986. Stable isotopes in late middle Eoceneto Oligocene foraminifera. Geol. Soc. Am. Bull, 97:335-345.

Keigwin, L.D., Jr., 1980. Palaeoceanographic change in the Pacific at theEocene-Oligocene boundary. Nature, 287r:722-725.

Kennett, J.P., and Stott, L.D., 1990. Proteus and Proto-oceanus: ancestralPaleogene oceans as revealed from Antarctic stable isotopic results: ODPLeg 113. In Barker, RF, Kennett, J.P., et al., Proc. ODP, Sci. Results, 113:College Station, TX (Ocean Drilling Program), 865-880.

, 1991. Abrupt deep-sea warming, paleoceanographic changes andbenthic extinctions at the end of the Palaeocene. Nature, 353:225-229.

Mead, G.A., Hodell, D.A., and Cielsielski, RF, 1993. Late Eocene to Oligo-cene vertical isotopic gradients in the South Atlantic: implications forwarm saline deep water. In Kennett, J.P., and Warnke, D. (Eds.), TheAntarctic Paleoenvironment: A Perspective on Global Change. Am. Geo-phys. Union, Antarc. Res. Ser., 60:27^8.

Miller, K.G., 1983. Eocene-Oligocene paleoceanography of the deep Bay ofBiscay: benthic foraminiferal evidence. Mar. Micropaleontol, 7:403-440.

, 1992. Middle Eocene to Oligocene stable isotopes, climate, anddeep-water history: the Terminal Eocene Event? In Prothero, D., andBerggren, W.A. (Eds.), Eocene-Oligocene Climatic andBiotic Evolution:Princeton, NJ (Princeton Univ. Press), 160-177.

Miller, K.G., Fairbanks, R.G., and Mountain, G.S., 1987a. Tertiary oxygenisotope synthesis, sea-level history, and continental margin erosion.Paleoceanography, 2:1-19.

Miller, K.G., Gradstein, FM., and Berggren, W.A., 1982. Late Cretaceous toearly Tertiary agglutinated benthic foraminifera in the Labrador Sea.Micropaleontology, 28:1-30.

Miller, K.G., Janecek, T.R., Katz, M.E., and Keil, D.J., 1987b. Abyssalcirculation and benthic foraminiferal changes near the Paleocene/Eoceneboundary. Paleoceanography, 2:741-761.

Miller, K.G., Katz, M.E., and Berggren, W.A., 1992. Cenozoic deep-seabenthic foraminifera: a tale of three turnovers. In BENTHOS '90: Sendai(Tokai Univ. Press), 67-75.

Müller-Merz, E., and Oberhánsli, H., 1991. Eocene bathyal and abyssalbenthic foraminifera from a South Atlantic transect at 20-30° S. Palaeo-geogr., Palaeoclimatol, Palaeoecol, 83:117-171.

Oberhánsli, H., McKenzie, J., Toumarkine, M., and Weissert, H., 1984. Apaleoclimatic and paleoceanographic record of the Paleogene in the centralSouth Atlantic (Leg 73, Sites 522, 523, and 524). In Hsü, K.J., LaBrecque,J.L., et al., Init. Repts. DSDP, 73: Washington (U.S. Govt. Printing Office),737-747.

Pak, D.K., and Miller, K.G., 1992. Paleocene to Eocene benthic foraminiferalisotopes and assemblages: implications for deepwater circulation. Paleo-ceanography, 7:405^122.

Rea, D.K., Basov, I.A., Janecek, T.R., Palmer-Julson, A., et al., 1993. Proc.ODP, Init. Repts., 145: College Station, TX (Ocean Drilling Program).

Savin, S.M., 1977. The history of the EarüYs surface temperature during thepast 100 million years. Annu. Rev. Earth. Planet Sci., 5:319-355.

Savin, S.M., Douglas, R.G., and Stehli, EG., 1975. Tertiary marine paleotem-peratures. Geol. Soc. Am. Bull, 86:1499-1510.

Sclater, J.G., Abbot, D., and Thiede, J., 1977. Paleobathymetry and sedimentsof the Indian Ocean. In Heirtzler, J., Bolli, H.M., Davies, T.A., Saunders,J.B., and Sclater, J.G. (Eds.), Indian Ocean Geology and Biostratigraphy.Am. Geophys. Union, 25-59.

Shackleton, N., and Boersma, A., 1981. The climate of the Eocene ocean. J.Geol. Soc. London, 138:153-157.

Shackleton, NJ., 1986. Paleogene stable isotope events. Palaeogeogr.,Palaeoclimatol, PalaeoecoL, 57:91-102.

Shackleton, NJ., Hall, M.A., and Boersma, A., 1984. Oxygen and carbonisotope data from Leg 74 foraminifers. In Moore, T C , Jr., Rabinowitz,P.D., et al., Init. Repts. DSDP, 74: Washington (U.S. Govt. Printing Office),599-644.

Shackleton, NJ., and Kennett, J.P., 1975. Paleotemperature history of theCenozoic and the initiation of Antarctic glaciation: oxygen and carbonisotope analyses in DSDP Sites 277, 279, and 281. In Kennett, J.P., Houtz,R.E., et al., Init. Repts. DSDP, 29: Washington (U.S. Govt. Printing Office),743-755.

Spiess, V., 1990. Cenozoic magnetostratigraphy of Leg 113 drill sites, MaudRise, Weddell Sea, Antarctica. In Barker, RF., Kennett, J.P., et al., Proc.ODP, Sci. Results, 113: College Station, TX (Ocean Drilling Program),261-315.

Thomas, E., 1990. Late Cretaceous through Neogene deep-sea benthic fora-minifers (Maud Rise, Weddell Sea, Antarctica). In Barker, RF., Kennett,J.P., et al., Proc. ODP, Sci. Results, 113: College Station, TX (OceanDrilling Program), 571-594.

, 1992. Cenozoic deep-sea circulation: evidence from deep-seabenthic foraminifera. In Kennett, J.P., and Warnke, D. (Eds.), The AntarcticPaleoenvironment: A Perspective on Global Change. Am. Geophys.Union, Antarct. Res. Ser., 56:141-165.

Tjalsma, R.C., and Lohmann, G.P., 1983. Paleocene-Eocene bathyal andabyssal benthic foraminifera from the Atlantic Ocean. MicropaleontolSpec. Publ, 4.

278


van Andel, T.H., Heath, G.R., and Moore, T.C., Jr., 1975. Cenozoic history andpaleoceanography of the central equatorial Pacific Ocean: a regionalsynthesis of Deep Sea Drilling Project data. Mem.—Geol. Soc. Am., 143.

van Morkhoven, F.P.C.M., Berggren, W.A., and Edwards, A.S., 1986. Ceno-zoic cosmopolitan deep-water benthic foraminifera. Bull. Cent. Rech.Explor.-Prod. Elf-Aquitaine, Mem. 11.

Vergnaud Grazzini, C, Muller, C , Pierre, C, Létolle, R., and Peypouquet, J.P,1979. Stable isotopes and Tertiary paleontological paleoceanography in thenortheast Atlantic. In Montadert, L., Roberts, D.G., et al., Init. Repts.DSDP, 48: Washington (U.S. Govt. Printing Office), 475^91.

Wise, S.W., Gombos, A.M., and Muza, J.P., 1985. Cenozoic evolution of polarwater masses, southwest Atlantic Ocean. In Hsü, K.J., and Weissert, H.J.

(Eds.), South Atlantic Paleoceanography: Cambridge (Cambridge Univ.Press), 283-324.

Wolfe, J.A., 1978. A paleobotanical interpretation of Tertiary climates in theNorthern Hemisphere. Am. J. Sci, 66:694-703.

Woodruff, F., and Savin, S.M., 1989. Miocene deepwater oceanography.Paleoceanography, 4:87-140.

Zachos, J.C., Stott, L.D., and Lohmann, K.C., 1994. Evolution of earlyCenozoic marine temperatures. Paleoceanography, 9:353-387.

Date of initial receipt: 4 April 1994Date of acceptance: 18 August 1994Ms 145SR-121

CD

O Λ

ou

40-

50-

3

Φ

sOσ>ö

ΦC

8oLU

2ΦüOLUΦ

T3T3

ε

oce

ne

LUΦ

öΦ(0

Q.

•

2

m

/

3!

1

3—?*1 ^ _

— α

j-(

01

-n

>

« » •

I

-1

-

•

~ ç-

cf"

30

40

0)

<

50

-1

δ 1 3 c1

Olig

oce

neI.

Eoce

nem

idd

le E

oce

ne

e.

Eo

cene

Pa

leo

.

Figure 13. Oxygen isotopic records of Pacific Sites 883 (solid circles),(squares), and 577 (open circles); data from Pak and Miller (1992).

60

Figure 14. Carbon isotopic records of Pacific Sites 883 (solid circles), 884(squares), and 577 (open circles); data from Pak and Miller (1992). Nuttallidestruempyi data are corrected by 0.3%o for comparison with Cibicidoides data.

279


δ 1 3 C

CDCS)

<

30-

40-

50-

•

-1

ΦC

8og>δ

ΦCΦUO

LU

—

ΦCΦUO

LU_Φ

"O3

ΦCΦ

8LU

Φ

dΦ

Pacific

0

•9=

fj

a?

1 2

_ v erΛ— y•

— o - ^ _ _

>xT^ Southern Ocean

^ J B <O<,

^ o

3

-

•

Figure 15. Comparison of the composite Pacific oxygen isotopic record (solid

circles; includes data from Sites 883, 884, and 577) with the Southern Ocean

record from Site 690 (open circles); data from Kennett and Stott (1990).

Figure 16. Comparison of the composite Pacific carbon isotopic record (solid

circles; includes data from Sites 883, 884, and 577) with the Southern Ocean

record from Site 690 (open circles); data from Kennett and Stott (1990).

Figure 17. Comparison of the composite Pacific oxygen isotopic record (solid

circles) with the Southern Ocean record (open circles) using the Aubry et al.

(unpubl. data, 1994) stratigraphy.

280


60-1-Figure 18. Comparison of the composite Pacific carbon isotopic record (solidcircles) with the Southern Ocean record (open circles) using the Aubry et al.(unpubl. data, 1994) stratigraphy.

650

700 -

750 -

800 -

85010 20 30 40

N. truempyi (%)60

Figure 20. Comparison of the relative abundance of Nuttallides truempyi (opencircles) with the oxygen isotopic record (solid circles) at Site 883.

650

Agglutinated foraminifers (%)0 20 40 60 80

CaCO20 40 60 80 100

700

S.750CD

Q

850

Figure 19. Comparison of the relative abundance of agglutinated taxa withpercentage of carbonate at Site 884.

281

18. isotopic and faunal record of paleogene deep …€¦ · 18. isotopic and faunal record of...

Documents