45. PLIOCENE PALEOCLIMATIC AND PALEOCEANOGRAPHIC HISTORY OF THESOUTH ATLANTIC OCEAN: STABLE ISOTOPIC RECORDS FROM LEG 72

DEEP SEA DRILLING PROJECT HOLES 516A AND 5171

Kathleen A. Leonard,2 Department of Geology, University of South Carolina, Columbia, South CarolinaDouglas F. Williams, Department of Geology and Belle W. Baruch Institute, University of South Carolina,

Columbia, South Carolinaand

Robert C. Thunell, Department of Geology, University of South Carolina, Columbia, South Carolina

ABSTRACT

Stable isotopic analyses (18O/16O, 13C/12C) of both benthic and planktonic foraminifers from DSDP Holes 516Aand 517 appear to reflect changing paleoclimatic and paleoceanographic conditions in the South Atlantic Ocean duringthe Pliocene. The δ 1 8 θ records of Cibicides wuellerstorfi and Globigerinoides sacculifer do not vary in phase with oneanother, indicating that the δ 1 8θ of C. wuellerstorfi is reflecting changing glacial ice volumes, but the δ 1 8 θ of G. sac-culifer is being influenced more by other factors, such as variable sea-surface temperatures and salinities.

A net depletion in 1 8O and a series of very low-amplitude fluctuations characterize the late Pliocene planktonic δ 1 8 θrecord. The stability of this section may be related to increasing surface-water temperatures in middle southern latitudesduring northern hemisphere glacial ice buildup caused by differences in the phasing of the orbital precessional cycle be-tween the northern and southern hemispheres.

Hole 516A is presently located in an area of strong δ13C gradients in the water column, and a similar situation prob-ably prevailed during the Pliocene. A record of the difference between the δ13C of C. wuellerstorfi and G. sacculifer(Δ13C) reflects times of increased or decreased similarity in δ13C between Circumpolar Water and subtropical surfacewaters throughout the Pliocene. Periods of increased or decreased Δ13C may reflect: 1) changes in water mass bound-aries because of expansion or contraction of a particular water mass, or 2) changes in the partitioning of 13C within thevarious components of the global carbon reservoir. These changes may be associated with glacial/interglacial climaticchanges, such as the initiation of northern hemisphere glaciation.

INTRODUCTION

Paleoclimatic and paleoceanographic trends and var-iations through time can be effectively studied throughchanges in the oxygen and carbon stable isotopic com-position of foraminifers in deep-sea sediments (Emili-ani, 1955, 1966; Savin, 1977; Shackleton and Opdyke,1977). Oxygen and carbon stable isotopic analysis ofsediments recovered at Holes 516A and 517 duringDSDP Leg 72 in the western South Atlantic Ocean pro-vide a relatively continuous history of the PlioceneEpoch from 4.6 to 1.8 Ma. Several factors affect the18Q/16O ratio of calcite precipitated from seawater, in-cluding: 1) temperature, 2) the isotopic composition ofseawater, which is affected by changing glacial ice vol-ume, 3) salinity, 4) depth habitat, 5) vital effects, and6) partial dissolution or diagenesis. The influences ofdepth habitat and vital effects are minimized by analyz-ing a limited size fraction of a particular species. The ex-cellent preservation of the foraminifers at Holes 516Aand 517 indicate that diagenesis or dissolution is negligi-ble. Therefore, temperature, ice volume, and salinityshould be the three major factors affecting variations inthe δ 1 8θ of foraminifers.

Barker, P. F., Carlson, R. L., Johnson, D. A., et al., Init. Repts. DSDP, 72: Wash-ington (U.S. Govt. Printing Office).

2 Present address: Marathon Oil Company, Houston, Texas.

The oxygen isotopic variation in the composition ofpost-Miocene benthic foraminifers is considered to re-flect changes in glacial ice volume (Shackleton and Ken-nett, 1975; Savin et al., 1975; Woodruff et al., 1981).The effects of temperature and salinity on oxygen iso-topic fractionation in benthic foraminifers remain rela-tively constant because the bottom waters are formedfrom high-latitude surface waters of relatively constanttemperature and salinity. Oxygen isotopic variations inpost-Miocene planktonic foraminifers also predominant-ly reflect changing glacial ice volumes, as well as themore variable surface-water temperatures and salinities(van Donk, 1976; Savin, 1977).

The carbon isotopic composition of calcite precipi-tated from seawater appears to vary with the 13C/12C ra-tio of total dissolved carbon dioxide ( Σ C O 2 ) . Total dis-solved carbon dioxide is approximately 89% dissolvedbicarbonate (Kroopnick, 1974); therefore, variations inthe δ13C of foraminiferal calcite should dominantly re-flect variations in the δ13C of dissolved HCOf. The pa-leoceanographic changes affecting δ13C variations ofdissolved bicarbonate are not well understood, but sev-eral theories have been proposed. Shackleton (1977)postulated that carbon isotopic events in the late Pleis-tocene are caused by changes in the volume of the conti-nental biomass, which are related to the glacial/intergla-cial aridity cycles. Broecker (1982) related variations inthe δ13C of dissolved HCO3~ to the effects of glacial/in-

895

K. A. LEONARD, D. F. WILLIAMS, R. C. THUNELL

terglacial sea-level fluctuations on the volume of organ-ic matter buried in shelf sediments and the associated re-moval of phosphate (POf) from the oceans. Scholle andArthur (1980) have related variations in the δ13C of car-bonate precipitated from seawater to organic carbon bur-ial rates in deep-sea sediments. Times of increased or-ganic carbon burial would be times of increased δ13C indissolved bicarbonate. Duplessy and others (1980) havepresented evidence that <513C variations in the Atlanticare a result of changing circulation rates, caused by apossible decrease in the production of North AtlanticDeep Water (NADW) during northern hemisphere glaci-al events. Variations in the degree of stratification of thewater column and its effects on upwelling rates couldalso influence the δ13C values of surface waters.

Studies of carbon isotopic records have proposed dif-ferent combinations of the above theories to explainpaleoceanographic events reflected in excursions ofthe δ13C record of calcite precipitated from seawater.Bender and Keigwin (1979) observed that a significantnegative shift in δ13C occurred during the late Miocene.They related this 13C decrease to a decrease in upwellingrates or to changes in abyssal circulation. Vincent andothers (1980) also noted that this negative late Mioceneδ13C event had occurred, and they proposed that the 13Cdepletion reflected an organic carbon transfer to the

oceans from shelf areas exposed during regressions.This transfer may have been accompanied by a changein deep circulation patterns and ocean fertility.

Changes in oceanic circulation patterns or rates shouldalso be reflected in the distribution of various watermass properties. This detailed comparison of Pliocenebenthic and planktonic stable isotopic records fromDSDP Hole 516A and planktonic isotope data fromDSDP Site 517 attempts to determine if the δ 1 8 θ andδ13C fluctuations in and between Circumpolar Water(CPW) and surface waters in the South Atlantic are re-lated to significant climatic or oceanographic events.

SITE LOCATION AND HYDROGRAPHY

DSDP Holes 516A and 517 were drilled with the hy-draulic piston corer (HPC) during Leg 72 in the westernSouth Atlantic Ocean. Hole 516A is located at 30° 17' S,35°17'W. Site 517 is located to the west of 516A at30°56'S, 38°02'W. Both sites are located on the RioGrande Rise (Fig. 1). The Rio Grande Rise is a majoraseismic ridge whose origin is most likely related to frac-ture zones associated with the Mid-Atlantic Ridgespreading center. These fracture zones were probablyactive during the initial rifting of the South AtlanticOcean during the Early Cretaceous (Francheteau andLePichon, 1972).

30°

• DSDP sitesArgentine Basin

35° S45° W 40° 35° 30° W

Figure 1. The location of DSDP Holes 516A and 517, which were hydraulically piston cored during Leg 72on the Rio Grande Rise in the South Atlantic Ocean.

896

STABLE ISOTOPIC RECORDS AND PLIOCENE HISTORY

At a water depth of 1313 m, Hole 516Alies within thetransition between Antarctic Intermediate Water (AAIW)and the upper branch of CPW. The deeper site, 517(2963 m), lies within NADW and is mainly a product ofseveral sources in the northernmost Atlantic. Wheresouthward-flowing NADW meets CPW flowing north,NADW separates the CPW into distinct upper and low-er branches (Reid et al., 1977; Broecker and Takahashi,1980). All of these water masses can be traced and dis-tinguished based on unique characteristics derived fromtheir original source (Broecker and Takahashi, 1980).Particular water masses also exhibit characteristic rangesin δ 1 8 θ and δ13C (Fig. 2, after Kroopnick, 1980a, b).Data from GEOSECS Stations 59 and 60 and Cato Ex-pedition Stations 10 and 14 in the western South Atlanticindicate that Hole 516A is presently situated within anarea of strong hydrographic gradients at the base of thethermocline within both the oxygen and salinity minimaand within a nutrient (POf5 and NOf) maximum (Reidet al., 1977). The depth of Hole 516A (1313 m) is alsoassociated with a strong δ13C gradient centered at thebase of the thermocline (Kroopnick, 1980a).

Apparently, present-day Atlantic surface circulationpatterns became established during the Pliocene afterthe emergence of the Panama isthmus (Berggren andHollister, 1977; Keigwin, 1978, 1979; Keigwin, 1982).Backtracking indicates that the Rio Grande Rise hadsubsided to approximately its present water depth by thePliocene (Thiede, 1977; Barker, this volume). Therefore,Pliocene hydrography and circulation in the South At-lantic are assumed to have been very similar to the mod-ern patterns at Hole 516A. Because of its location with-in an area of strong hydrographic gradients, Hole 516Ashould be a sensitive indicator of expansions or contrac-tions of the NADW, CPW, and AAIW during the Plio-cene.

METHODS

Isotopic Analyses

The planktonic foraminiferal species, Globigerinoides sacculifer,was selected for oxygen and carbon isotopic analyses from the 300-to-355-micron size fraction at intervals of 50 cm or less from the Pliocenesections of Holes 516A and 517. The isotopic samples were limited toa particular morphotype of G. sacculifer, a compact form that doesnot possess a large irregular sac as a final chamber. The benthicforaminiferal species, Cibicides (Planuliná) wuellerstorfi from the>150-micron size fraction was also analyzed from the Hole 516Asamples. In a few samples, this species was not abundant enough forisotopic analysis. When this was the case, C. kullenbergi was ana-lyzed. Analyses of both benthic species from the same samples indi-cate that these two species are isotopically similar and may be usedinterchangeably within experimental error (Table 1) (Duplessy et al.,1980; Woodruff et al., 1981; Keigwin, 1982). The benthic foraminiferswere also sampled every 50 cm or less at Hole 516A.

In general, foraminifers do not deposit their calcite tests in oxygenand carbon isotopic equilibrium with seawater. A number of species,however, exhibit a consistent offset relative to the δ 1 8 θ of ambientseawater and the δ 1 8 θ of dissolved bicarbonate. Both G. sacculifer(Williams et al., 1977) and C. wuellerstorfi (Graham et al., 1981;Belanger et al., 1981) appear to deposit their tests close to, but consis-tently offset from, equilibrium. This characteristic allows researchersto base the qualitative paleoclimatic and paleoceanographic interpre-tations on the isotopic variation of these species through time.

The samples for isotopic analysis were processed and analyzed ac-cording to the techniques of Williams and others (1977). Each samplewas ultrasonically cleaned in deionized water. Then the benthic fora-minifers were lightly crushed, and both planktonic and benthic fora-minifers were roasted in vacuo at 380°C for 60 minutes. The oxygenand carbon isotopic composition of the CO2 gas (derived by reactingthe samples with 100% phosphoric acid at 50°C) was analyzed by useof a VG Micromass 602 D isotope ratio-mass spectrometer.

The benthic and planktonic isotopic results (Tables 1, 2, and 3) arereported in terms of the PDB standard by the repeated analysis of thetwo isotopic standards, NBS-20 and TKL-1. The δ 1 8 θ and δ13C valuesfor NBS-20 with respect to PDB are 4.18% and -1.06%, respectively(Craig, 1957). The δ 1 8 θ and δ13C values for TKL-1 are -4.26 and-1.69%, respectively (Blattner and Hulston, 1978). Analytical pre-cision based on the routine analysis of a standard carbonate powder is±0.09%0 for δ 1 8 θ and ±0.07% for δ13C The average standard devia-tion from the mean (lα) for replicate analyses of benthic foraminiferal

GEOSECS stations

s r» in c > σ > f ^ c o i o c M o σ > r • io co «- o TZç>in in in «s•θ*i• ^r *r * m m m m mm OCM

c ΣCO2

GEOSECS Western-Atlantic section

Contours in ‰I

40° 50° 60° 70° 80° N

Figure 2. The present-day distribution of the δ13C of ECO2 (%) in the western Atlantic Ocean (from Kroopnick, 1980a).

897

K. A. LEONARD, D. F. WILLIAMS, R. C. THUNELL

Table 1. Oxygen and carbon isotopicresults of Cibicides wuellerstorfifrom Hole 516A.

Table 2. Oxygen and carbon isotopicresults of Globigerinoides saccu-lifer from Hole 516A.

Core-section(interval in cm)

Age(Ma) a

δ 1 8 θ δ 1 3 C(PDB‰) (PDB‰)

3-2, 50-523-2, 70-723-2, 120-1223-3, 20-223-3, 44_464-1, 4-64-1, 4-64-1, 4-64-1, 20-224-1, 20-224-1, 70-724-1, 120-1224-2, 20-224-2, 70-724-2, 100-1024-2, 100-1024-2, 120-1224-3, 20-224-3, 70-724-3, 94-965-1, 10-125-1, 20-225-1, 70-725-1, 120-1225-2, 20-225-2, 70-725-2, 120-1225-3, 20-225-3, 34-365-3, 34-365-3, 64-665-3, 70-725-3, 80-825-3, 80-825-3, 98-1006-1, 7-96-1,20-226-1, 50-526-1,50-526-1, 70-726-1,70-726-1, 70-726-1, 120-1226-2, 20-226-2, 70-726-2, 120-1226-3, 20-226-3, 70-726-3, 120-1227-1, 20-227-1, 70-727-1, 120-1227-2, 20-227-2, 70-727-2, 120-1227-3, 20-227-3, 70-727-3, 70-72b

8-1, 20-228-1, 20-22b

8-1, 70-728-1, 120-1228-1, 120-122b

8-2, 20-22b

8-2, 70-728-2, 70-72b

8-2, 120-1228-3, 20-228-3, 20-228-3, 70-728-3, 120-1229-1, 20-229-1, 70-72b

9-1, 120-122b

9-1, 138-1409-1, 138-1409-2, 10-129-2, 20-229-2, 70-729-2, 120-1229-3, 20-229-3, 70-729-3, 120-122

2.712.732.772.802.822.902.902.902.912.912.952.993.023.063.083.083.103.143.173.193.233.233.273.303.343.373.413.483.503.503.543.553.573.573.593.663.683.723.723.753.753.753.823.843.873.893.913.933.963.984.004.024.054.074.094.114.144.144.184.184.204.224.224.254.274.274.294.314.314.314.364.384.404.434.434.434.444.444.474.514.544.584.61

2.882.172.762.912.622.962.302.772.862.892.862.712.432.432.702.522.532.712.432.492.442.202.362.472.232.322.152.232.372.202.462.352.402.182.372.502.242.112.312.341.902.252.252.442.112.272.342.372.272.401.972.182.162.181.901.982.042.272.262.232.332.082.182.382.242.272.152.292.332.372.422.772.342.313.013.113.062.302.262.112.202.292.12

1.130.420.870.850.890.870.830.740.900.751.070.860.620.571.151.070.760.710.510.861.010.890.680.800.790.690.600.500.910.750.970.871.090.901.020.890.580.860.830.780.780.750.660.770.600.720.670.700.640.570.780.710.840.660.730.770.850.490.580.600.790.990.850.830.870.721.060.790.750.951.020.760.360.640.960.820.540.900.900.821.010.870.86

a Age based on interpolation between detailedbiostratigraphic zonation of planktonic fora-minifers.

b Cibicides kullenbergi was analyzed for thesesamples.

Core-section(interval in cm)

3-2, 40-423-2, 50-523-2, 70-723-2, 70-723-2, 70-723-2, 120-1223-3, 20-223-3, 20-223-3, 20-223-3, 44-464-1, 4-64-1, 20-224-1, 20-224-1, 70-724-1, 120-1224-1, 120-1224-2, 20-224-2, 20-224-2, 70-724-2, 70-724-2, 100-1024-2, 120-1224-2, 120-1224-2, 120-1224-2, 120-1224-3, 20-224-3, 20-224-3, 70-724-3, 94-965-1, 10-125-1, 10-125-1, 20-225-1, 70-725-1, 120-1225-2, 20-225-2, 70-725-2, 120-1225-3, 20-225-3, 34-365-3, 64-665-3, 70-725-3, 80-825-3, 98-1006-1, 7-96-1, 20-226-1, 20-226-1, 50-526-1, 70-726-1, 120-1226-2, 20-226-2, 70-726-2, 120-1226-3, 20-226-3, 70-726-3, 120-1227-1, 20-227-1, 70-727-1, 120-1227-2, 20-227-2, 20-227-2, 20-227-2, 120-1227-2, 120-1227-3, 20-227-3, 70-728-1, 20-228-1, 70-728-1, 120-1228-2, 20-228-2, 70-728-2, 120-1228-3, 20-228-3, 70-728-3, 120-1229-1, 20-229-1, 70-729-1, 120-1229-1, 138-1409-2, 10-129-2, 20-229-2, 70-729-2, 98-1009-2, 120-1229-3, 10-129-3, 20-229-3, 34-369-3, 54-569-3, 70-729-3, 120-122

Age(Ma)a

2.702.712.732.732.732.772.802.802.802.822.902.912.912.952.992.993.023.023.063.063.083.103.103.103.103.143.143.173.193.233.233.233.273.303.343.373.413.483.503.543.553.573.593.663.683.683.723.753.823.843.873.893.913.933.963.984.004.024.054.054.074.094.094.114.144.184.204.224.254.274.294.314.344.364.384.404.434.434.444.444.474.494.514.544.544.554.574.584.61

δ 1 8 θ(PDB‰)

0.330.390.310.610.300.220.330.410.530.390.230.050.250.140.630.410.430.210.100.270.320.280.280.790.450.350.580.130.13

-0.090.310.210.290.360.440.570.580.540.080.310.640.560.440.260.570.620.280.310.410.260.160.170.230.210.150.260.310.110.140.150.300.190.410.300.400.410.430.370.400.210.650.400.510.190.140.290.250.220.410.090.13

-0.060.100.190.410.160.050.300.20

δ 1 3 C

(PDB‰)

2.032.091.641.931.681.921.911.821.882.021.931.742.031.711.801.981.981.731.911.691.942.001.831.831.781.841.982.072.112.221.911.752.312.132.082.222.082.011.941.882.102.071.861.832.121.871.992.121.881.621.841.921.932.302.152.122.101.982.242.241.962.042.092.172.112.161.982.411.961.962.132.252.422.432.382.182.541.941.972.472.582.312.402.192.622.232.152.452.41

1 Age based on interpolation between detailedbiostratigraphic zonation of planktonic fora-minifers.

898

STABLE ISOTOPIC RECORDS AND PLIOCENE HISTORY

Table 3. Oxygen and carbon isotopicresults of Globigerinoides sacculi-fer from Site 517.

Core-section[interval in cm)

7-1, 20-227-1, 70-727-1, 120-1227-2, 20-227-2, 70-727-2, 120-1227-3, 20-227-3, 70-728-1, 20-228-1, 20-228-1,70-728-1, 70-728-1, 120-1228-1, 120-1228-2, 20-228-2, 70-728-2, 120-1228-3, 20-228-3, 70-728-3, 120-1229-1, 20-229-1, 70-729-1, 70-729-1, 120-1229-1, 120-1229-2, 20-229-2, 20-229-2, 70-729-2, 70-729-3, 20-229-3, 20-229-3, 20-229-3, 20-229-3, 20-229-3, 70-729-3, 120-12210-2, 20-2210-2, 70-7210-2, 120-12210-3, 20-2210-3, 70-7210-3, 70-7210-3, 70-7211-2,20-2211-2, 70-7211-2, 120-12211-3,20-2211-3,70-7211-3, 120-12212-1, 20-2212-1, 70-7212-1, 120-12212-1, 120-12212-2, 20-2212-2, 70-7212-2, 120-12212-2, 120-12212-3, 20-2212-3, 20-2212-3, 20-2212-3, 70-7212-3, 70-7212-3, 120-12212-3, 120-122

Age(Ma)a

1.761.801.841.881.931.972.012.052.132.132.172.172.222.222.262.302.352.402.442.492.532.582.582.622.622.672.672.722.722.812.812.812.812.812.862.912.962.972.993.003.013.013.013.073.093.103.113.133.143.153.163.183.183.193.203.213.213.233.233.233.243.243.253.25

δ 1 8 θ(PDB‰)

-0.04-0.10-0.01

0.04-0.08

0.020.04

-0.10-0.09

0.12-0.09-0.05-0.12

0.09-0.09

0.110.270.390.200.140.050.120.290.280.360.430.53

-0.20-0.14

0.040.160.400.110.11

-0.110.030.080.020.090.170.070.550.020.050.19

-0.11-0.08

0.08-0.13

0.200.290.010.020.340.620.220.020.570.490.440.490.310.350.10

á13c(PDB%»)

1.862.041.652.002.121.942.181.792.101.781.911.572.201.811.872.191.791.941.771.932.032.282.011.941.942.001.982.412.112.172.092.092.032.132.132.262.112.042.031.922.022.101.871.932.212.312.152.142.202.131.901.691.732.081.942.372.332.172.052.102.072.181.711.64

a Age before present based on interpolation be-tween detailed biostratigraphic zonation ofplanktonic foraminifers.

samples is ±0.11 for δ 1 8 θ and ±0.05 for δ13C. The lα values forplanktonic analyses are ±0.10% for δ 1 8 θ and ±0.09% for δ13C.

Stratigraphy of Holes 516A and 517

The planktonic foraminiferal assemblages at Holes 516A and 517are well preserved and exhibit little evidence of mixing, reworking, orsevere dissolution. The planktonic foraminiferal zonation establishedfor the Rio Grande Rise region of the South Atlantic Ocean by Berg-gren (1977) was used to zone both sites biostratigraphically. The timeframework is based on extrapolation between planktonic foramini-feral biostratigraphic datum levels (Berggren, 1977) that have previ-ously been correlated with the paleomagnetic time scale (Ryan et al.,1974; Saito et al., 1975). The absolute ages assigned to the biostrati-

graphic events are those suggested by Thunell (1981), which are basedon the revised polarity time scales of McDougall (1977) and Mankinenand Dalrymple (1979). The Pliocene/Pleistocene boundary is assumedto fall within the Olduvai normal event (Haq et al., 1977) and is as-signed a date of approximately 1.80 Ma according to the time scale ofMcDougall (1977).

At Hole 516A, the Pliocene section extends from foraminiferalZone PL1 to PL5 (Fig. 3). A hiatus occurs within Section 516A-3-2,resulting in the loss of Zone PL6. This gap is indicated by the occur-rence of a PL5 fauna including Globorotalia fniocenica and G. exilisin Sample 516A-3-2, 50-52 cm, immediately followed by the first ap-pearance datum (FAD) of the Pleistocene species G. truncatulinoidesin Sample 516A-3-2, 20-22 cm (Berggren, Aubry, and Hamilton, thisvolume).

The top of paleomagnetic Chron 5, equivalent to the Miocene/Pliocene boundary, was used to calculate sedimentation rates withinZone PLla. Placement of the top of Chron 5 at 50.4 m subbottom isbased on the absence of one Gilbert normal event as indicated by boththe polarity reversals and on the generalized biostratigraphic place-ment of the Miocene/Pliocene boundary (Hamilton and Suzyumov, inpress). The time scale at Site 517 is based solely on the foraminiferalbiostratigraphy and covers the period from planktonic foraminiferalZone PL3 to the Pliocene/Pleistocene boundary (Fig. 4). The forami-niferal assemblages indicate no disruption.

RESULTS

Pliocene Planktonic Oxygen Isotopic Record ofHoles 516A and 517

The planktonic isotopic records from Holes 516Aand 517 are plotted together against time in order toprovide a history of the Pliocene from approximately4.6 to 1.8 Ma (Fig. 5). Some uncertainty in correlationexists between the planktonic isotopic records fromHoles 516A and 517 (Fig. 5), most probably because ofthe limited resolution of the biostratigraphically derivedtime scale. Possibly, shifting the curves ± 100,000 yearswould improve the observed correlations. Undetectedhiatuses and some subjectivity in determination of bio-stratigraphic zonal boundaries may also contribute tothe apparent inconsistencies between Holes 516A and517. The data can be described in terms of generaltrends and specific isotopic events. A series of low-am-plitude fluctuations characterize the base of the δ 1 8θrecord. This trend is interrupted by an 18O enrichmentof 0.55% from approximately 4.5 to 4.3 Ma. A gradualenrichment of 0.5% begins at 3.9 Ma and is followed bya depletion of the same magnitude. A change in thecharacter of the record occurs at approximately 3.2 Ma,when more rapid variations in isotopic composition oc-cur in contrast to the gradual, long-term trends before3.2 Ma. Two positive isotopic events of 0.65%0 and0.4‰ occur at 2.7 and 2.4 Ma, respectively. The remain-ing section of the δ 1 8θ curve exhibits variations of-0.15%, resulting in a very stable late Pliocene record.

Pliocene Planktonic Carbon IsotopicRecord of Holes 516A and 517

The early Pliocene exhibits a 1.0% depletion in 13Cfrom approximately 4.6 to 3.9 Ma. This trend ends in anabrupt negative shift of 0.7% (Fig. 5). After this event,the δ13C values are relatively constant, but remain lessthan those before the shift. A second abrupt depletionof 0.6% occurs at approximately 3.3 Ma. The subse-quent values for Hole 516A also remain offset relativeto those before the event. The remainder of the δ13C

899

K. A. LEONARD, D. F. WILLIAMS, R. C. THUNELL

δl180(PDB%.) δt3C(PDB%o)

1.0 0.5 2.5 2.0 1.5

15-

35-

40-

45-

50-

55-

10

11

12

13

Iδ

1

•2

o Foraminifer

T

o

δ

1zones

N22

1 TT" T i l

TT!c

ma

rga

•21o

1oü

c

1•c

T

ii-§ sδ ^

oCD

<S 5

PL5

PL4

PL3

PL2

PL1c&b

PL1a

2.91

3.10

3.41

3-82 -

S

4.44

5.44

Figure 3. The δ 1 8 θ and δ13C values for the planktonic foraminiferal species Globigerinoides sacculifer. The biostratigraphic zonation is basedon planktonic foraminifers, and the paleomagnetic zones were interpreted from the magnetic polarity reversals (Hamilton and Suzyumov,this volume) plotted to sub-bottom depth for DSDP Hole 516A. Areas of diagonal lines in polarity logs indicate intervals for which polaritywas unclear.

curve exhibits fluctuations on the order of 0.5%0 or less,with no net change.

Pliocene Benthic Isotopic Record of Hole 516A

Several trends and events are evident in the oxygenisotopic record of Cibicides wuellerstorfi (Fig. 6). Thisrecord is initially characterized by a sharp enrichment of0.75% at approximately 4.4 Ma., followed by a long-term depletion of 1.15‰ ending at approximately 4.1Ma (Fig. 6). An oscillatory increase in δ 1 8 θ of 0.55%0

occurs from 4.1 to approximately 3.8 Ma. Fluctuationsof 0.35%0 and less characterize the δ 1 8θ record from 3.8to 3.4 Ma. A trend towards enrichment in 1 8O begins atapproximately 3.4 to 3.2 Ma. This trend offsets the re-maining δ 1 8 θ values from those earlier in the Pliocene.

The top of the Hole 516A record (from approximately2.8 Ma) contains one light value at the onset of higher-amplitude fluctuations of 0.65 to 0.75%0.

The carbon isotopic record of C. wuellerstorfi ap-pears to be relatively stable, fluctuating about a mean ofO.79%o, but several events can be observed (Fig. 6). Theearliest significant δ13C event is an abrupt negative peakcentered at approximately 4.4 Ma. After this peak, theδ13C values decrease 0.45‰ over a period of approxi-mately 380,000 years. A relatively stable section of therecord ends in a well-defined enrichment peak at ap-proximately 3.7 Ma. This peak is followed by threeevents of similar character. The first is an enrichment of0.5%0 over approximately 250,000 years, ending in anabrupt depletion of the same magnitude. Immediately

900

STABLE ISOTOPIC RECORDS AND PLIOCENE HISTORY

δ

3 d20

254

30-|

40 H

45 H

50 H

6

7

8

9

10

11

12

1

2

31

2

31

2

31

2

31

2

31

2

3

6 1 8 O (PDB %« δ 1 3 C (PDB‰

0.5 -0.5 2.5 2.0 1.5]§•.§ Foraminifer& 2 zones

I T

& I

T

T

ÖS

T

N22

PL6

PL5

PL4

PL3

1.8

2.3

2.9

3.1

Figure 4. The δ 1 8 θ and δ13C values for the planktonic foraminiferal species Globigerinoides sacculifer and the biostratigraphiczonation for Site 517 based on planktonic foraminifers plotted against sub-bottom depth.

after this abrupt shift, there is a second enrichment,which terminates in an abrupt depletion of approxi-mately 0.5‰. This relationship, an abrupt depletionpreceded by a more gradual enrichment, is repeated athird time at approximately 3.1 Ma. The remaining sec-tion of the δ13C record is relatively stable until it ends ina sharp negative peak centered at 2.7 Ma.

The absolute differences between the δ 1 8 θ values(Δ18O) and the δ13C values (Δ13C) of C. wuellerstorfi andGlobigerinoides sacculifer were determined. Increasingand decreasing values of Δ18O and Δ13C should reflectdifferences and similarities, respectively, between inter-mediate and surface-water δ 1 8 θ and δ13C values throughtime. The Δ18O record (Fig. 7) is predominantly influ-enced by variations in benthic δ 1 8 θ record. Benthic andplanktonic carbon records, however, show a somewhatgreater covariance than the oxygen isotopic records (Fig.6). Both δ13C records exhibit a gradual depletion in 13Cthat occurred during the early Pliocene. This covarianceis punctuated by two abrupt planktonic depletions ofapproximately 0.73%0 and 0.75‰. It seems that, betweenapproximately 3.5 and 3.2 Ma, the benthic and plankton-ic δ 1 8 θ values varied out of phase. After this, changes inΔ13C are dominated by the greater magnitude of thebenthic isotopic fluctuations. These changing relation-ships between and planktonic δ13C values (Δ13C) may

be related to variations in water mass properties, bound-aries throughout the Pliocene, or both.

DISCUSSION

Pliocene Paleoclimatic History Interpreted fromOxygen Isotopes

Pliocene paleoclimatic trends and events are inferredfrom the δ 1 8θ records of benthic and planktonic fora-minifers. The oxygen isotopic results from Cibicideswuellerstorfi and Globigerinoides sacculifer of Hole516A do not correlate well. In some instances, they arein phase, in others out of phase, and often the relation-ship appears to be random. This pattern suggests thatmany times during the middle to late Pliocene, the oxy-gen isotopic composition of planktonic foraminiferswas influenced by factors other than ice volume, such aschanging sea-surface temperatures or salinity variations(other than those related to changing glacial ice volume).

Planktonic δ 1 8 θ values are generally considered to bemore variable than benthic δ 1 8 θ because of the morefrequent fluctuations in regional and global surface-water temperatures. Modern benthic δ 1 8 θ values aresubject to less of a temperature effect because of themore constant temperatures of intermediate and deep-water masses. Therefore, the δ 1 8 θ record of C. wueller-

901

K. A. LEONARD, D. F. WILLIAMS, R. C. THUNELL

δ18O(PDB%o)

1-0 0.5 0

δ 1 3 C(PDB‰)

2.5 2.0 1.5

1.8

2.0

2.2

2.4

2.6

2.8

~ 3.0

3.2

3.4

3.6

3.8

4.0

4.2

4.4

4.6 -

< " > Hole 517

Hole516A

Figure 5. The planktonic oxygen and carbon isotopic records of Globi-gerinoides sacculifer at DSDP Holes 516A (solid line) and 517(dotted line) spliced together based on extrapolation of constantsedimentation rates between planktonic biostratigraphic zones.

storfi at Hole 516A should be a reliable indicator of theglacial/interglacial climatic events from 4.6 to 2.7 Ma,although bottom temperature changes may play a smallrole in older δ 1 8θ records.

One of the most obvious series of events in the benthicδ 1 8 θ record is the sharp, positive peak (0.75%0) centeredat approximately 4.4 Ma (Fig. 6), followed by a decreasein δ 1 8 θ of approximately 1.15%o ending at 4.1 Ma. Oscil-lations of 0.25 to 0.55%0 are superimposed over this de-crease. This series of events was possibly related to gla-

cial fluctuations in the Antarctic ice cap. The east Ant-arctic ice sheet began forming at approximately 14 Ma,and both the east and west Antarctic ice sheets were es-tablished by 7 Ma (Mercer, 1978; Woodruff et al., 1981).The positive peaks centered at 3.7 and 3.5 Ma may berelated to one of the first Patagonian glaciations (3.5Ma) (Mercer, 1976), which may have been associatedwith an Antarctic glacial advance. This enrichment in18O is also reflected in the planktonic data by the in-crease in δ 1 8 θ between approximately 3.7 and 3.5 Ma.

A significant planktonic event occurs from approxi-mately 3.5 to 3.2 Ma, where a depletion of 0.5%0 occurs.This event does not correlate with the benthic "ice vol-ume" record, which suggests that the planktonic eventis possibly related to regional changes effected by restric-tion of surface equatorial circulation caused by the emer-gence of the Panama isthmus at approximately 3.5 Ma(Berggren and Hollister, 1977).

The benthic trend towards increasing δ 1 8θ values,which begins at approximately 3.2 Ma, probably reflectsthe initiation of permanent northern hemisphere glacia-tion (Shackleton and Opdyke, 1977; Shackleton andCita, 1979). Beginning at approximately 2.8 Ma, ac-cording to the benthic δ 1 8 θ record, there occurred twohigher-frequency, higher-amplitude (0.7%0) glacial/in-terglacial fluctuations, which are analogous to the in-crease in scale of the glacial events characteristic of thelate Pliocene (Shackleton and Opdyke, 1977).

The planktonic oxygen isotopic record of Hole 516Adoes not have the same features as the benthic isotoperecord since 3.2 Ma. (Fig. 6). The nature of the δ 1 8 θevents changes from the more gradual trends immedi-ately preceding 3.2 Ma to more rapid changes in δ 1 8 θafter, but no significant offset or enrichment in 18O isrecorded. In Hole 517, two events of higher amplitudeoccur at approximately 2.7 and 2.4 Ma in an otherwisestable late Pliocene section (Fig. 5). In the remainder ofthe planktonic record (from 2.3 to 1.8 Ma) δ 1 8 θ varia-tions are 0.15%0 or less, resulting in a very stable latePliocene section.

At Site 517, the expected 18O enrichment that oc-curred during the late Pliocene and was caused by theinitiation of northern hemisphere glaciation is not ob-served, possibly because of a slight surface-temperatureincrease in the region of the Rio Grande Rise resulting inan 18O depletion. The two major factors that influencethe isotopic composition of foraminiferal calcite, theisotopic composition of seawater (which is related tosalinity) and the temperature of seawater, are apparent-ly acting in opposite directions at this time, resulting in astable δ 1 8θ record.

Dampening of the late Pliocene isotopic signal at Site517 may perhaps be explained by the interaction of theorbital parameters and their effects on ocean warmingand glacial ice volumes. The distribution of incomingsolar radiation is predominantly controlled by the threeorbital variables: precession of the equinoxes, obliquity(tilt of the earth's axis), and eccentricity (the ellipticityof the earth's orbit). The growth and decay of continen-tal ice sheets are dependent on the amount and the dis-

902

STABLE ISOTOPIC RECORDS AND PLIOCENE HISTORY

2.6 -

2.8

3.0

3.2

3.4

3.6

3.8

4.0

4.2

4.4

4.6

3.0— r —

2.5— I —

δ1 8O(PDB%o)

2.0 A . 1.0

δ13C(PDB%o)

0.5 0.0 -0.5 2.5— i —

2.0—T~

1.5

—r~1.0

—r~0.5

— I —0.0

—I

G. sacculifer

C. wuellerstorfi 4 C. wuellerstorfi

Figure 6. The δ 1 8 θ and δ13C values for Cibicides wuellerstorfi and Globigerinoides sacculifer plotted against time for Hole 516A. Thetime scale for Hole 516A was derived through extrapolation of constant sedimentation rates between biostratigraphic zones.

tribution of summer insolation at high latitudes, wherethe ice accumulates, and are therefore a function of or-bital variations (Hays et al., 1976). Periods of decreasedseasonal contrast (cool summers and warm winters) fa-vor the buildup of glacial ice. Because the precessionalcycle is out of phase between the northern and southernhemispheres, times of reduced seasonal contrast in thenorth are times of increased seasonal contrast in thesouth (warm summers and cold winters) (Broecker andvan Donk, 1970). This combination may actually resultin a warming of mid-latitude southern hemisphere sur-face waters during periods of northern hemisphere gla-cial ice buildup and climatic cooling.

The CLIMAP (1976) reconstructions (comparingmodern sea-surface temperatures to those of the glacialmaximum 18,000 years ago) support the hypothesis thatmid-latitude southern hemisphere surface waters be-came warmer during northern hemisphere glacial events.In these reconstructions, several areas in the southernhemisphere, including the area of the Rio Grande Rise,exhibit temperature anomalies. Surface-water tempera-tures in the vicinity of the Rio Grande Rise either re-mained at their warm interglacial values or increasedduring the last increase in northern hemisphere ice vol-ume at the glacial maximum.

Pliocene Paleoceanographic Events Related toCarbon Isotopes

Variations in the degree of similarity or difference be-tween the δ 1 8 θ of benthic and planktonic foraminifersat Hole 516A may reflect changes in the δ 1 8 θ of inter-mediate and/or surface water masses as well as changesin the positions of transition zones between particularwater masses. The location of DSDP Site 516 within anarea of strong hydrographic gradients (especially δ13Cgradient) at the transition between AAIW and CPWshould make it a sensitive indicator of variations asso-ciated with these two water masses. Because of the exis-tence of a strong δ13C gradient at Site 516, only changesin δ13C of 0.4%0 and greater are considered to reflectsignificant changes in water masses.

The core of NADW at approximately 3000 m depthpresently has a δ13C maximum of > 1,0%0 relative toPDB (Fig. 2). The core of CPW presently records a δ13Cminimum of <0.2% between 1300 and 1700 m depth atapproximately 50°S latitude. The δ13C at Hole 516A ap-proximated from GEOSECS data (Kroopnick, 1980a) is0.65%0 at a depth of approximately 1300 m and 1.8%0 at50 m, the calcification depth of Globigerinoides saccu-lifer. These values reflect an average modern Δ13C of

903

K. A. LEONARD, D. F. WILLIAMS, R. C. THUNELL

approximately 1.15%0. The average Pliocene Δ13C isequal to 1.29 ± 0.26%0. Because this is within the rangeof modern values, it can be assumed that the distributionof δ13C in the Pliocene South Atlantic was similar to themodern distribution. This assumption is also supportedby sedimentologic, faunal, and other isotopic evidencethat modern intermediate and deep circulation patternswere established by the early Pliocene and that modernsurface circulation became fully established during themid-Pliocene (Berggren and Hollister, 1977; Keigwin,1982).

Evidence from core-top samples suggests that the dis-tribution of the δ13C of dissolved bicarbonate in sea-water is reflected in the δ13C distribution in foramini-fers. Belanger and others (1981) and Graham and others(1981) present evidence that δ13C values of Cibicideswuellerstorfi exhibit a consistent offset from the δ13C ofdissolved bicarbonate; therefore, this species may beuseful in determining relative changes in the δ13C of sea-water HCO^~. Data from Curry and Lohman (1982) andWilliams and Healy-Williams (1982) indicate that theδ13C of benthic foraminifers has trends similar to theδ13C of HCO3- within water masses and across watermass boundaries within the Vema Channel. These stud-ies suggest that changes in Δ13C based on the carbon iso-

topic composition of foraminifers should reflect changesin the distribution of δ13C and therefore changes in wa-ter mass properties or patterns through time.

A number of the significant (>0.4%0) Δ13C events atHole 516A (Fig. 7) can be explained by a combinationof several events, such as variations in circulation rates,differing rates of production of particular water masses,stratification of the water column, or changes in theΔ13C of oceanic HCO^~. Several of the Δ13C events arebetter understood by examination of the relative isotop-ic changes in individual benthic and planktonic δ13Crecords (Fig. 6). Rapid negative changes in the plank-tonic δ13C record in the early Pliocene (at approximately4.3-4.2 and 3.9-3.8 Ma), which do not appear in thebenthic δ13C record, most probably reflect times whenthe intermediate and surface waters became more simi-lar in isotopic composition.

The above situation, in which the planktonic δ13Crecord approaches the benthic record with only minorvariations recorded by the benthic data, could be relatedto decreased stratification in the surface waters. Thismight occur as a result of increased upwelling of inter-mediate waters having lower δ13C values (Kroopnick,1980a, b). Upwelling of waters more depleted in 13Cshould result in a decrease in the δ13C values of the

3.0

Δ 1 8 O(%o)

2.5 2.0 1.5 2.0

Δ1 3C(%o)

1.5 1.0 0.5 0.02.6 -

12.8

;3.o

3.2

3.4

3.6

3.8

4.0

4.2

4.4

4.6 -

Figure 7. The difference between planktonic and benthic δ 1 8 θ values (Δ18O) and the dif-ference between planktonic and benthic <513C values (Δ1 3C) at Hole 516A plottedagainst time.

904

STABLE ISOTOPIC RECORDS AND PLIOCENE HISTORY

planktonic foraminifers. The paleoceanographic eventclosest in time to the first rapid 13C depletion at approxi-mately 4.43 Ma is the pronounced northward shift ofthe Antarctic convergence that occurred in the earlyPliocene (Berggren and Hollister, 1977). There is a pos-sibility that this event had an effect on oceanic mixingrates, which in turn led to increased upwelling.

The initiation of northern hemisphere glaciation at3.2 Ma (as indicated by the oxygen isotopes) is ac-companied by an increased difference between benthicand planktonic δ13C values mostly because of a 0.5%decrease in the benthic record. There is evidence thatrates of NADW production may have decreased duringnorthern hemisphere glacial events of the late Pleisto-cene, resulting in sluggish Atlantic circulation (Streeterand Shackleton, 1979; Boyle, 1981). In this case, therewould be two ways to decrease the δ13C at the transitionbetween CPW and AAIW. First, if circulation rates de-creased because of decreased production of NADW,then each water mass would be an "older" water massat any given site. Because of the effects of respirationand organic matter oxidation, an older water mass willhave a lower O2 content and a greater CO2 concentra-tion with more negative δ13C values. Secondly, with de-creased NADW production, the CPW might expand togreater depths and lower latitudes. CPW is depleted in13C; therefore, its expansion should carry water withlower δ13C values further north into the area of Hole516A at 1313 m.

After this first response to the initiation of northernhemisphere glaciation at 3.2 Ma, a negative trend inplanktonic δ13C is reflected in a trend towards moresimilar Δ13C values throughout the late middle Pliocene.This increased similarity could be a response to theexpansion of the Antarctic water masses (CPW andAAIW) during the contraction of NADW, which in turnmay have resulted in decreased water column stratifica-tion and increased upwelling of 13C-depleted waters.Surface waters more depleted in 13C would thus be re-flected in the carbon isotopic composition of Globigeri-noides sacculifer.

SUMMARY AND CONCLUSIONS

Comparison of oxygen and carbon isotopic recordsof benthic and planktonic foraminifers from DSDPHole 516A reveals general paleoclimatic and paleo-ceanographic trends that occurred on the Rio GrandeRise during the Pliocene. The first trend is a series ofvariations in δ 1 8 θ inferred to be related to variations inthe Antarctic ice cap throughout the early to middlePliocene. The second is the initiation of northern hemi-sphere glaciation at approximately 3.2 Ma. This eventexhibits the characteristic net 18O enrichment after theinitiation, as well as the previously observed increase inscale of the later Pliocene glacial/interglacial events atapproximately 2.8 Ma (Shackleton and Opdyke, 1977).

After the initiation of northern hemisphere glaciationat 3.2 Ma, the planktonic δ 1 8 θ record contrasts with thebenthic record. The net depletion and dampened isotopicsignal reflected in the isotopic composition of Globigeri-noides sacculifer may be related to increased surface-

water temperatures in middle southern latitudes duringnorthern hemisphere glacial ice buildup. The reciprocalinfluences of the orbital parameters on oceanic tempera-tures and glacial ice growth and decay between the twohemispheres may be recorded by the net depletion andlow-amplitude fluctuations of the late Pliocene δ 1 8 θ rec-ord.

The δ13C record at Hole 516A reflects times of in-creased or decreased similarity between CPW and sur-face water δ13C values, which may be associated withpaleoclimatic events. For example, at approximately3.2 Ma, a decrease in the difference in δ13C between in-termediate and surface waters is reflected in a 13C de-pletion of benthic foraminifers. This could be related todecreased production of NADW causing sluggish Atlan-tic circulation and a contraction of the core of NADW.Diminished NADW production would allow the expan-sion of CPW, which would introduce waters lower inδ13C into the area of Hole 516A. The oxygen and carbonisotopic records of Holes 516A and 517 allow for thediscrimination between global paleoclimatic events andregional paleoceanographic events that influenced thedistribution of intermediate water masses within thesouthwestern Atlantic Ocean during the Pliocene.

ACKNOWLEDGMENTS

The authors would like to thank Michael Arthur, Richard Fillon,Lloyd Keigwin, and Warren Prell for constructive criticism of themanuscript; David Johnson for assistance in obtaining the samples forthis study; and Joyce Goodwin and Donna Black for typing the manu-script in its various forms. This research was supported by NationalScience Foundation Grants OCE 80-08239 and DPP 80-23696. This iscontribution number 493 of the Belle W. Baruch Institute for MarineBiology and Coastal Research.

REFERENCES

Belanger, P. E., Curry, W. B., and Matthews, R. K., 1981. Core-topevaluation of benthic foraminiferal isotopic ratios for paleo-oceanographic interpretations. Palaeogeogr. Palaeoclimatol. Pa-laeoecol., 33:205-220.

Bender, M. L., and Keigwin, L. D., Jr., 1979. Speculations about theupper Miocene change in abyssal Pacific dissolved bicarbonateδ13C. Earth Planet. Sci. Lett., 45:383-393.

Berggren, W. A., 1977. Late Neogene planktonic foraminiferal bio-stratigraphy of the Rio Grande Rise (South Atlantic). Mar. Micro-paleontol, 2:265-313.

Berggren, W. A., and Hollister, C. D., 1977. Plate tectonics and pa-leocirculation—commotion in the ocean. Tectonophysics, 38(1-2):11-48.

Blattner, P., and Hulston, J. R., 1978. Proportional variations of geo-chemical δ 1 8 θ scales—an interlaboratory comparison. Geochim.Cosmochim. Acta, 42-59.

Broecker, W. S., 1982. Glacial to interglacial ocean chemistry duringglacial time. Geochim. Cosmochim. Acta, 46:1689-1705.

Broecker, W. S., and Takahashi, T., 1980. Hydrography of the cen-tral Atlantic. III. The North Atlantic deep-water complex. Deep-SeaRes., 27A:591-613.

Broecker, W. S., and van Donk, J., 1970. Insolation changes, ice vol-umes, and the 18O record in deep sea cores. Rev. Geophys. SpacePhys., 8:169.

CLIMAP Project Members, 1976. The surface of ice-age earth. Sci-ence, 191:1131-1137.

Craig, H., 1957. Isotopic standards for carbon and oxygen and cor-rection factors for mass-spectrometric analysis of carbon dioxide.Geochim. Cosmochim. Acta, 12:122.

Curry, W. B., and Lohmann, G. P., 1982. Carbon isotopic changes inbenthic foraminifera from the western South Atlantic: reconstruc-tion of glacial abyssal circulation patterns. Quat. Res., 18:218-235.

905

K. A. LEONARD, D. F. WILLIAMS, R. C. THUNELL

Duplessy, J. C , Moyes, J., and Pujol, C , 1980. Deep water forma-tion in the North Atlantic Ocean during the last ice age. Nature,286:479-482.

Emiliani, C , 1955. Pleistocene paleotemperatures. J. Geol., 63:538., 1966. Isotopic paleotemperatures. Science, 154:851.

Francheteau, J., and LePichon, X., 1972. Marginal fracture zonesas structural framework of continental margins in South AtlanticOcean. Am. Assoc. Pet. Geol. Bull., 56:991-1007.

Goodney, D. E., Margolis, S. V., Dudley, W. C , Kroopnick, P., andWilliams, D. F., 1980. Oxygen and carbon isotopes of recent cal-careous nannofossils as paleoceanographic indicators. Mar. Mi-cropaleontol., 5:31-42.

Haq, B. U., Berggren, W. A., and Van Couvering, J. A., 1977. Cor-rected age of the Pliocene/Pleistocene boundary. Nature, 269:483-488.

Hays, J. D., Imbrie, J., and Shackleton, N. J., 1976. Variations in theearth's orbit: pacemaker of the ice ages. Science, 194:1121-1132.

Keigwin, L. D., Jr., 1978. Pliocene closing of the isthmus of Panama,based on biostratigraphic evidence from nearby Pacific Ocean andCaribbean Sea cores. Geology, 6:630-634.

, 1979. Late Cenozoic stable isotope stratigraphy and paleo-ceanography of DSDP sites from the east equatorial and centralNorth Pacific Ocean. Earth Planet. Sci. Lett., 45:361-382.

_, 1982. Isotopic paleoceanography of the Caribbean and eastPacific; role of Panama uplift in late Neogene time. Science., 217:350-353.

Kroopnick, P., 1974. Correlations between 13C and ECO2 in surfacewaters and atmospheric CO2. Earth Planet. Sci. Lett, 22:397-403.

, 1980a. The distribution of 13C in the Atlantic Ocean. EarthPlanet. Sci. Lett., 49:469-484.

_, 1980b. Isotopic fractionations during oxygen consumptionand carbonate dissolution within the North Atlantic Deep Water.Earth Planet. Sci. Lett, 49:485-490.

McDougall, I., 1977. The Present Status of the Geomagnetic PolarityTime Scale: Canberra, A. C. T. (Research School for the Earth Sci-ences, Australian National University), Publ. No. 1288.

Mankinen, E. A., and Dalrymple, G. B., 1979. Revised geomag-netic polarity time scale for the interval 0-5 m.y.B.P. J. Geophys.Res., 84:615-626.

Mercer, J. H., 1976. Glacial history of southernmost South America.Quat Res., 6:125-166.

, 1978. Glacial development and temperature trends in theAntarctic and in South America. In Van Zinderen Bakker, E. M.(Ed.), Antarctic Glacial History and World Paleoenvironments:Rotterdam (A. A. Balkema), pp. 73-93.

Reid, J. L., Nowlin, W. D., Jr., and Patzert, W. C , 1977. On thecharacteristics and circulation of the southwestern Atlantic Ocean./. Phys. Oceanogr., 7:62-91.

Ruddiman, W. F., and Mclntyre, A., 1981. Oceanic mechanisms foramplification of the 23,000-year ice-volume cycle. Science, 212:617-627.

Ryan, W. B. F., Cita, M. B., Rawson, M. D., Burckle, L. H., andSaito, T., 1974. A paleomagnetic assignment of Neogene stageboundaries and the development of isochronous datum planes be-tween the Mediterranean, the Pacific and Indian oceans in order toinvestigate the response of the world ocean to the Mediterraneansalinity crisis. Riv. Ital. Paleontoi, 80:631-687.

Saito, T., Burckle, L. H., and Hays, J. D., 1975. Late Miocene toPleistocene biostratigraphy of equatorial Pacific sediments. InSaito, T., and Burckle, L. H. (Eds.), Late Neogene Epoch Bound-aries: New York (Micropaleontology Press), pp. 226-234.

Savin, S. M., 1977. The history of the earth's surface temperatureduring the past 100 million years. Ann. Rev. Earth Planet Sci., 5:319-355.

Savin, S. M., Douglas, R. G., and Stehli, F. G., 1975. Tertiary marinepaleotemperatures. Bull. Geol. Soc. Am., 86:1499-1510.

Scholle, P. A., and Arthur, M. A., 1980. Carbon isotopic fluctuationsin Cretaceous pelagic limestones: potential stratigraphic and petro-leum exploration tool. Am. Assoc. Pet. Geol. Bull., 64:67-87.

Shackleton, N. J., 1977. Carbon-13 in Uvigerina: Tropical rainforesthistory and the equatorial Pacific carbonate dissolution cycles. InAndersen, N. R., and Malahoff, A. (Eds.), The Fate of Fossil FuelCO2 in the Oceans: New York (Plenum Press), pp. 401-427.

Shackleton, N. J., and Cita, M. B., 1979. Oxygen and carbon isotopestratigraphy of benthic foraminifers at Site 397: detailed history ofclimatic change during the late Neogene. In von Rad, U., Ryan,W. B. F., et al., Init. Repts. DSDP, 47, Pt. 1: Washington (U.S.Govt. Printing Office), 433-455.

Shackleton, N. J., and Kennett, J. P., 1975. Paleotemperature historyof the Cenozoic and the initiation of Antarctic glaciation: oxygenand carbon isotope analyses in DSDP Sites 277, 279 and 281. //iKennett, J. P., Houtz, R. E., et al., Init. Repts. DSDP, 29: Wash-ington (U.S. Govt. Printing Office), 743-756.

Shackleton, N. J., and Opdyke, N. D., 1977. Oxygen isotope and pa-leoclimatic evidence for early northern hemisphere glaciation. Na-ture, 270:216-219.

Streeter, S. S., and Shackleton, N. J., 1979. Paleocirculation of thedeep North Atlantic: 150,000-year record of benthic foraminiferaand oxygen-18. Science, 203:168-171.

Thiede, J., 1977. Subsidence of aseismic ridges: evidence from sedi-ments on Rio Grande Rise (southwest Atlantic Ocean). Am.Assoc. Pet Geol. Bull., 61:929-940.

Thunell, R. C , 1981. Late Miocene-early Pliocene planktonic fora-miniferal biostratigraphy and paleoceanography of low-latitudemarine sequences. Mar. Micropaleontol., 6:71-90.

van Donk, J., 1976. 18O record of the Atlantic Ocean for the entirePleistocene Epoch. In Cline, R. M., and Hays, J. D. (Eds.),Investigation of Late Quaternary Palaeoceanography and Palaeo-climatology: Boulder, CO (Geol. Soc. Am. Mem.), 145:147-163.

Vincent, E., Killingley, J. S., and Berger, W. H., 1980. The magneticEpoch-6 carbon shift: a change in ocean's 13C/12C ratio 6.2million years ago. Mar. Micropaleontol., 5:185-203.

Williams, D. F., and Healy-Williams, N., 1982. Vertical isotope gra-dients in Recent benthic foraminifera and the water column of theVema Channel. EOS, Trans. Am. Geophys. Union, 63:359.

Williams, D. F., Sommer, M. A., and Bender, M. L., 1977. Carbonisotopic compositions of recent planktonic foraminifera of the In-dian Ocean. Earth Planet. Sci. Lett., 36:391-403.

Woodruff, F., Savin, S. M., and Douglas, R. G., 1981. Miocene sta-ble isotope record: a detailed deep Pacific Ocean study and itspaleoclimatic implications. Science, 212:665-668.

Date of Initial Receipt: July 2, 1982

906