PALEOCEANOGRAPHY, VOL. 8, NO. 4, PAGES 527-547, AUGUST 1993

ON STABLE ISOTOPIC VARIATION AND EARLIEST

PALEOCENE PLANKTONIC FORAMINIFERA

Steven D'Hondt

Graduate School of Oceanography, University of Rhode Island, Narragansett

James C. Zachos

Earth and Marine Sciences, University of California, Santa Cruz

Abstract. Extant planktonic foraminifera display positive covariance between •513C signals and test size. As documented by other studies, primary causes of increased •513C values with increased test size may include increased reliance on ambient CO2 for calcification at larger test sizes, decreased kinetic fractionation during calcification at larger test sizes, and increased photosymbiotic activity in larger symbiont-bearing planktonic foraminifera. Planktonic foraminiferal •5180 values also often covary with test size, although the direction of this covariance is taxon dependent. Possible explanations for relationships between •5180 signals and test size include changing habitat depth over ontogeny, correlations between adult test size and environmental conditions, and changing isotopic disequilibrium with size, ontogenetic stage, or photosymbiont density. In order to assess the magnitude and implications of similar size dependence in earliest Paleocene planktonic foraminifera, we measured the stable isotopic signals of multiple size fractions of 10 earliest Paleocene species. All of these taxa exhibit a strong positive correlation between •513C and test size. The slope and magnitude of this trend varies between species, with Woodringina claytonensis displaying the largest shift (1.1%0 over a 130 gm range in mean sieve size) and Guembelitria cretacea displaying the smallest (0.2 %0 over a 38 gm range). By analogy with modern planktonic foraminifera, this general relationship between •513C and size probably resulted from increased reliance on ambient CO2 for calcification at larger test sizes. The high magnitude of this shift in some taxa may reflect either photosymbiotic enhancement of the general trend or relatively greater changes in the proportions of metabolic and ambient CO2 used for calcification at different test sizes.

Copyright 1993 by the American Geophysical Union.

Paper number 93PA00952. 0883-8305/93/93 PA-00952510.00

Failure to account for relationships between test size and •513C signals can lead to underestimation of early Paleocene surface ocean •513C values by 1%0 or more. These size-related •513C effects provide an alternative explanation for decreases in whole-rock •513C values and some decreases in planktonic-to- benthic foraminiferal •513C gradients documented at marine K/T boundary sequences. At all size fractions, the 10 Paleocene taxa display a very limited interspecies range of •5180 derived paleotemperatures. Despite this limited range, paleobiogeographic patterns and •5180 signals appear to provide realistic estimates of relative paleodepth and seasonal affinities of earliest Paleocene planktonic foraminiferal species. Earliest Paleocene •5180 and biogeographic data are consistent with a general trend of surface-to-deep diversification of microperforate planktonic foraminifera following the K/T boundary. Such a trend may simply result from exploitation of a near-surface open-ocean habitat by the epicontinental K/T survivor G. cretacea.

INTRODUCTION

Stable isotopic ratios provide a powerful and widely used tool for reconstructing fossil paleoceanographic and paleoclimatic conditions from fossil marine carbonates. In particular, oxygen isotopic ratios of coccolithophorid and planktonic foraminiferal calcite are commonly used to estimate sea surface paleotemperature and paleosalinity, while carbon isotopic ratios have been used to estimate carbon fluxes between different major carbon reservoirs. For example, oxygen isotopic ratios of marine carbonates have been used to estimate changing sea surface and deepwater paleotemperature across the Cretaceous/Tertiary (K/T) boundary [Douglas and Savin, 1973; Boersma and Shackleton, 1981; Hsii et al., 1982; Perch-Nielsen et al., 1982; Shackleton et al., 1984a; Williams et al., 1985; Zachos et al., 1985, 1989a, b; Zachos and Arthur, 1986; Oberhansli, 1986; D'Hondt and Lindinger, 1988; Keller

527

528 D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera

and Lindinger, 1989; Herman, 1990; Stott and Kennett, 1990a, b; Corfield et al., 1991]. Additionally, carbon isotopic ratios have been used to reconstruct •513C values of total dissolved carbon in seawater and to estimate changes in the carbon balance between surface and deep oceans. For example, the apparent equilibration of •513C in surface and deepwater carbonates at the K/T boundary has been commonly interpreted as indicative of a radical decrease in marine primary productivity at that time [Hsii et al., 1982; Hsii and MacKenzie, 1985; Zachos et al., 1985, 1989a, 1992; Zachos and Arthur, 1986; Arthur et al., 1987; D'Hondt and Lindinger, 1988; Keller and Lindinger, 1989; Barrera and Keller, 1990; Stott and Kennett, 1990]. These oxygen and carbon isotopic data and interpretations underlie paleoclimatic and paleoceanographic models which estimate early Paleocene atmospheric pCO2 [Hsii and McKenzie, 1985; Caldeira et al., 1990], ocean alkalinity [Caldeira et al., 1990], and the balance between terrestrial and marine carbon fluxes [Kump, 1991].

Either explicitly or implicitly, such paleoceanographic studies assume that foraminiferal and/or nannofossil carbonate

faithfully records relative changes in equilibrium isotopic conditions. This assumption can be problematic: stable isotopic signals may vary within taxa and from taxon to taxon, depending on paleoenvironmental associations and relative "vital" effects. Several studies have documented

intertaxonomic variation in isotopic vital effects of extant calcareous nannoplankton [Dudley and Goodney, 1979; Dudley et al., 1986; Goodney et al., 1980; Paull and Thierstein, 1987]. Others have demonstrated strong interspecies and intraspecies variation in isotopic signals of extant planktonic foraminifera [i.e., Emiliani, 1971; Berger et al., 1978]. For example, some planktonic foraminifera shift their carbon isotopic signal with size by the same magnitude that separates ambient isotopic values of surface and deep waters [Berger et al., 1978; Oppo and Fairbanks, 1989; Spero et al., 1991].

Detailed analysis of such variation helps to constrain paleoceanographic interpretation. It documents which planktonic taxa are most suitable for oxygen isotopic reconstruction of sea surface paleotemperature and paleosalinity. It is also necessary for accurately estimating carbon isotopic variation between different surface and deep marine carbon reservoirs. Additionally, measurement of stable isotopic variation within and between taxa provides an important key for determining relationships between fossil organisms and their environment. For example, intraspecies carbon isotopic variation may record photosymbiont activity in planktonic foraminifera [Spero et al., 1991], while interspecies oxygen isotopic variation can be used to document the interplay between fossil planktonic foraminifera and paleoenvironment on both ecological and evolutionary timescales [Douglas and Savin, 1978; Boersma and Premoli Silva, 1989; Spero and Williams, 1988; Corfield and Cartlidge, 1991; Schweitzer and Lohmann, 1991].

In this study, we document and interpret stable isotopic variation within and between species of earliest Paleocene planktonic foraminifera. To this end, we briefly review stable isotopic variation in modern planktonic foraminifera, present evidence for such variation in earliest Paleocene taxa, and discuss some paleoceanographic and paleobiologic implications of stable isotopic variation in the earliest Paleocene species.

STABLE ISOTOPIC VARIATION IN EXTANT PLANKTONIC FORAMINIFERA

Carbon Isotopes

Modem planktonic foraminiferal species generally exhibit strong within-taxon trends of increasing test •513C signals with increasing test size [Savin and Douglas, 1973; Berger et al., 1978; Kahn, 1979; Currey and Matthews, 1981; Duplessy et al., 1981; Erez and Honjo, 1981; Williams et al., 1981; Bouvier-Soumagnac and Duplessy, 1985; Spero and DeNiro, 1987; Spero and Williams, 1988; Oppo and Fairbanks, 1989; Ravelo, 1991; Spero et al., 1991]. Over an increase in test diameter of a few hundred micrometers (gm), whole-test •513C signals can increase by as much as 1.5 to 2.0 %o [Berger et al., 1978; Oppo and Fairbanks, 1989; Spero et al., 1991]. The slope and magnitude of this trend varies between taxa, with extant Globigerinoides species exhibiting much greater •513C increases than extant Globorotalia over the same approximate range in test diameter [Berger et al., 1978]. Within a species, the rate of •513C increase generally decreases with increasing mean test size [Berger et al., 1978; Bouvier-Soumagnac and Duplessy, 1985; Oppo and Fairbanks, 1989]. Similar trends have been observed in late Paleocene Morozovella, Acarinina, and Subbotina species [Shackleton et al., 1985].

Even for larger specimens, the test •513C values of many planktonic foraminiferal species may not equilibrate with ambient seawater values [Kahn, 1979; Schweitzer and Lohmann, 1991 ]. Nonetheless, to the extent that they depart from equilibrium values by a constant amount, they accurately record geographic or temporal variation in seawater •513C values. Prior studies suggest that the intraspecies relationship between •513C and size does not significantly vary between different open-ocean regions [Bouvier-Soumagnac and Duplessy, 1985] or between glacial and interglacial samples [Oppo and Fairbanks, 1989]. Based on Neogloboquadrina dutertrei and Globorotalia menardii captured in plankton tows, •513C values of tests from narrowly constrained size fractions appear to faithfully record relative changes in •513C values of the ambient seawater TCO2 [Bouvier-Soumagnac and Duplessy, 1985].

At least three models can be invoked to explain the positive correlation between test size and •513C signals of planktonic foraminiferal calcite. These are (1) decreased metabolic rates at larger test sizes resulting in lower kinetic •513C fractionation during calcification [Berger et al., 1978; Kahn, 1979; Turner, 1982; McConnaughy 1989a, b], (2) increased photosymbiotic activity in larger specimens [Spero and Williams, 1988; Oppo and Fairbanks, 1989], and (3) a shift in the relative dominance of internal (respiratory) and external (seawater) carbon pools as the mass of carbonate required for calcification increases with increasing test size [Berger et al., 1978; Erez, 1989]. Other factors that can affect calcite •513C signals include the •513C value of food consumed by the foraminifer [Kahn, 1979; Spero et al., 1991], ontogenetic migration over a range of water depths [Shackleton et al., 1985], and the amount of calcite crust added in late ontogeny [Schweitzer and Lohmann, 1991].

The positive correlation between test size and •513C is displayed by both species with relatively shallow-water life cycles (i.e., Globigerinoides ruber, O. universa) and species that migrate toward deep C12-enriched water over ontogeny

D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera 529

(i.e., Globorotalia hirsuta, Globorotalia truncatulinoides) [Berger et al., 1978; Bouvier-Soumagnac and Duplessy, 1985; Hemleben et al., 1985, 1989; Ravelo, 1991]. The presence of this positive trend in taxa that migrate to deeper waters over ontogeny suggests that it is only weakly modified by ontogenetic variation in watermass •513C composition. Furthermore, despite the influence of food •513C values on the respiratory carbon pool, mass-balance calculations suggest that it is unlikely that changing respiratory •j13C values can drive large increases in test •513C with size [Kahn, 1979; Spero et al., 1991].

The trend toward increasing test •513C with test size is displayed by photosymbiotic species (i.e., Orbulina universa, Globigerinoides sacculifer) and by symbiont-free species (i.e., G. truncatulinoides) [Berger et al., 1978; Hemleben et al., 1985; Oppo and Fairbanks, 1989; Ravelo, 1991; Spero et al., 1991 ]. The ubiquity of this trend in both groups of planktonic foraminifera indicates that it does not solely derive from photosymbiotic activity. Nonetheless, in some taxa, photosymbiont activity does strongly affect the relationship between test •513C values and test size. Through a series of culturing experiments, Spero et al. [Spero and DeNiro, 1987; Spero and Williams, 1988; Spero et al., 1991 ] demonstrated a close link between photosymbiosis and test •513C composition; for O. universa, mean diameter, shell weight and •j13C content of the terminal chamber all vary strongly with symbiont activity. For this taxon, specimens grown with no photosymbiont activity display terminal chamber •513C values that are depleted by 1.6 to 2.3 %0 relative to specimens grown under conditions of maximum photosymbiont activity [Spero and DeNiro, 1987; Spero et al., 1991]. On the basis of their work with O. universa, Spero et al. [ 1991] developed a quantitative model for interpreting planktonic foraminiferal •513C signals. In applying their model to another photosymbiotic species, Globigerinoides sacculifer, they successfully predicted a clear relationship between test size and test •513C values. However, in comparison to fossil material, their predicted values overestimate test •513C signals for both large and small size fractions of G. sacculifer [Berger et al., 1978; Oppo and Fairbanks, 1989]. Possible causes of this apparent offset include overestimation of photosynthetic rates in larger specimens, overestimation of the environmental •513C value during calcification, and underestimation of the respiratory CO2 contribution during calcification in smaller specimens [Spero et al., 1991].

At present, the most parsimonious general explanation of the positive correlation between planktonic foraminiferal test size and •513C values appears to be increased reliance on the external carbon pool or decreased kinetic fractionation at larger test sizes, whether driven by decreasing metabolic rates over ontogeny or simply by the increased amount of carbonate required for calcification by larger individuals. This explanation is supported by the presence of this positive correlation in most or all extant planktonic foraminifera [Berger et al., 1978; Kahn, 1979; Hemleben et al., 1985; Ravelo, 1991 ] and by the tendency of some taxa to approximate ambient seawater •513C values at larger test sizes (i.e., symbiont free O. universa [Spero and Williams, 1988]). In any case, the magnitude of this positive trend is strongly modified by symbiont activity in photosymbiotic species,

with high levels of such activity leading to greatly enriched test •513C values [Spero et al., 1991].

Oxygen Isotopes

Most analyses of the relationship between seawater •5180 and the •5180 of planktonic foraminiferal tests are based on comparison of local seawater properties to •jl 8 0 signals of specimens captured in plankton tows, or on comparison of •5180 values in sediment trap or core-top assemblages to equilibrium values calculated for the overlying water column. In general, such studies suggest that •jl 8 0 signals in adult planktonic foraminifera appear to correlate closely with ambient temperature and •jl 8 0 values [Berger, 1978; Currey and Matthews, 1981; Duplessy et al., 1981; Erez and Honjo, 1981; Bouvier-Soumagnac and Duplessy, 1985]. Despite this general correlation, whole-test •5180 values of some taxa may depart from isotopic equilibrium.

Several modern taxa exhibit strong increases in •5180 with increasing test size. This shift is up to 2 %0, varying with taxon, size range, seasonal production and the degree of late ontogenetic crust development [Emiliani, 1971; Berger et al., 1978; Douglas and Savin, 1978; Kahn, 1979; Curry and Matthews, 1981; Erez and Honjo, 1981; Schweitzer and Lohmann, 1991]. Berger et al. [ 1978] noted several possible explanations for this relationship, including (1) systematic change in depth of habitat over ontogeny [Emiliani 1954, 1971], (2) correlation between adult test size and environmental conditions, i.e., ambient temperature [B6 and Duplessy, 1976], and (3) changing isotopic disequilibrium over ontogeny [Vergnaud-Grazzini, 1976; Kahn, 1979]. Culturing experiments and studies of specimens caught in plankton tows and sediment traps suggest that all three factors apply, although perhaps to a greater or lesser extent for different taxa and under different oceanographic conditions [Bouvier-Soumagnac and Duplessy, 1985].

Within single sediment samples, some extant taxa exhibit small decreases in •5180 with increasing terminal test size [Emiliani, 1954; Berger et al., 1978; Wefer et al., 1983; Bouvier-Soumagnac and Duplessy, 1985]. For example, the •5180 value of N. dutertrei has been observed to decrease by 0.46 + 0.14 %0 with test diameter increasing from 300-450 to 600-750 !.tm fractions, whereas O. universa •5180 values can decrease by 0.22 + 0.06 %0 with an increase in terminal test diameter from 450-600 gm to 600-750 gm [Bouvier- Soumagnac and Duplessy, 1985]. Wefer et al. [1983] and Bouvier-Soumagnac and Duplessy [ 1985] note that N. dutertrei only displays this trend in upwelling regions. Wefer et al. [ 1983] hypothesize that smaller adult specimens are associated with low temperatures, high seawater nutrient contents and active upwelling while large specimens occur during nonupwelling intervals that are characterized by higher temperatures and lower nutrient contents. Following B6 et al. [ 1973] and Deuser et al. [ 1981 ], Bouvier-Soumagnac and Duplessy [ 1985] further note that the negative trend between •5180 and test size in adult O. universa also reflects an association of smaller specimens with colder ambient conditions. These interpretations are at least partly consistent with recent culturing experiments that indicate O. universa attains larger terminal test sizes at 25øC than at 12ø-18øC,

530 D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera

Globigerinella siphonera reaches larger final test sizes at 23.5øC and 29øC than at 13ø-16øC, and G. sacculifer attains its largest mean terminal test size at optimum conditions (26.5øC, rather than 14ø-16øC or 31øC) [Bijma et al., 1990].

Several factors lead us to expect the whole-test •5180 signals of most species to be somewhat 180_depleted relative to equilibrium values in the environment of final chamber calcification. These include (1) ontogenetic migration of planktonic foraminifera downward through watermasses of increasingly positive/5180 equilibria (combined with retention of early ontogenetic calcite in adult foraminiferal tests), (2) possible incorporation of metabolic CO2 into the foraminiferal test [Kahn, 1979], (3) kinetic discrimination against 180 and 13C during rapid carbonate precipitation [McConnaughey, 1989a, b], and (4) discrepancies between assumed and actual equilibrium conditions. On the basis of studies of specimens captured in plankton tows and sediment traps, many species do exhibit adult whole-test/518 0 values that are slightly lower than values assumed for equilibrium conditions [Kahn, 1979; Fairbanks et al., 1980; Erez and Honjo, 1981; Bouvier- Soumagnac and Duplessy, 1985]. In counterpoint to this general trend of 18 0 depletion in planktonic foraminiferal tests, both Fairbanks et al. [1980] and Erez and Honjo [1981] noted that deep dwelling species captured in plankton tows can display/518 0 signals that are greater than apparent equilibrium values (by up to 1.3 %0). Both of the latter studies suggested that positive departures from apparent equilibrium may reflect discrepancies between actual and assumed equilibrium conditions, possibly resulting from foraminiferal migration to shallower waters.

In most of these studies, the •5180 signals of most taxa are offset from calculated equilibrium values by 0 to 0.35%0. Such limited departures from equilibrium are consistent with the results of laboratory culturing experiments [Erez and Luz, 1983; Bouvier-Soumagnac and Duplessy, 1985; Spero and Williams, 1988]. For example, while cultured O. universa precipitates its test slightly below (0.15 to 0.4 %0) [Bouvier- Soumagnac and Duplessy, 1985; Spero, 1992] or at/5180 equilibrium [Spero and Williams, 1988], cultured G. sacculifer appears to calcify its test in oxygen isotopic equilibrium [Erez and Luz, 1983].

Isotopic Depth Ranking

Oxygen isotope ratios are commonly used to estimate the relative depth of extant and fossil planktonic foraminiferal habitats in the water column [Emiliani, 1954, 1971; Lidz et al., 1968; Douglas and Savin, 1978; Shackleton and Vincent, 1978; Fairbanks et al., 1980, 1982; Boersma and Shackleton, 1981; Durazzi, 1981; Erez and Honjo, 1981; Boersma and Premoli Silva, 1983; Poore and Matthews, 1984; Keller, 1985; Keigwin and Corliss, 1986]. These studies are generally based on whole-test/5180 signals and rely on the increase in equilibrium •5180 values that results from decreasing seawater temperatures with increasing water depth. Such studies are complicated by (1) size-dependent variation in •5180 signals, (2) the common, but generally small and generally negative, departure of whole-test/518 0 signals from equilibrium values, (3) salinity effects, and (4) seasonal or interannual variation in thermocline structure and planktonic foraminiferal production.

Depth rankings based on oxygen isotopes do not directly predict the absolute water depth of foraminiferal paleohabitats. With the exception of a few well-controlled studies of extant taxa [i.e., Schweitzer and Lohmann, 1991 ], they are also limited to approximating the average relative depth of calcification for adult populations; they generally do not account for variation in vertical migration over ontogeny or for individual migration away from the mean depth of calcification. Despite these caveats, oxygen isotopic rankings do roughly approximate the relative depth rankings of adult planktonic foraminifera. For adult specimens, isotopic depth rankings are generally consistent with the results of stratified plankton tows [Berger, 1969; Kahn, 1979; Fairbanks et al., 1980, 1982; Hemleben et al., 1989] and current models of planktonic foraminiferal life cycles [Emiliani, 1971; Hemleben et al., 1989].

Paleoceanographic Use of t•13C and 8180 Signals in Recent Planktonic Foraminifera

The preceding discussion illustrates that planktonic foraminifera do not directly record absolute values of seawater /513C. Because planktonic/513C signals increase with increasing test size and/5180 values can also vary significantly with changing test size, isotopic analysis of size-variable populations can mask or lead to the spurious appearance of significant variation in paleohabitat stable isotopic values. Additionally, the analysis of a multispecimen sample provides average/513C and/5180 values for the population analyzed but does not account for individual variation in stable isotopic signals. Despite these problems, narrowly constrained size fractions of monospecific samples do record relative variation in mean/513C values of ambient seawater TCO2 [Bouvier- Soumagnac and Duplessy, 1985; Oppo and Fairbanks, 1989; Spero et al., 1991; Spero, 1992]. Furthermore,/5180 signals of modem adult populations generally correlate closely with mean habitat temperature and/5180, with most/5180 signals falling within 0 to 0.4 %0 of calculated equilibrium values [Berger et al., 1978; Currey and Matthews, 1981; Duplessy et al., 1981; Erez and Honjo, 1981; Erez and Luz, 1983; Bouvier- Soumagnac and Duplessy, 1985; Spero and Williams, 1988; Spero, 1992]. Hence well-controlled monospecific samples can be used to trace relative geographic and temporal variation in average equilibrium/5180 and •513C values of specific planktonic foraminiferal paleohabitats (i.e., summer sea surface).

STABLE ISOTOPIC VARIATION IN EARLIEST PALEOCENE PLANKTONIC FORAMINIFERA

Foraminiferal Methods and Taxonomy

Bulk sediment samples were soaked in a solution of 40 g sodium hexametaphosphate and 201 distilled deionized water, lightly buffered with 58 % ammonium hydroxide to equilibrate the pH to 7. Disaggregated samples were washed with tap water in 38 and 63 gm sieves, and oven dried overnight at temperatures less than 50 ø C. This process was repeated 2 to 3 times in order to fully disaggregate and clean the foraminiferal samples.

D'Hondt and Zachos' Isotopic Variation and Planktonic Foraminifera 531

...

.

.3!;: .•:

,

Figure. 1. Scanning electron photomicrographs of the earliest Paleocene taxa included in this study. Scale bar is 100 gm. Top row (left to right): Parvularugoglobigerina eugubina (from site 577, core 12, section 5, 115-117 cm), Guembelitria cretacea (from 577, 12-5, 115-117 cm), Woodringina claytonensis (from 577, 12-5, 94-96 cm), and Woodringina hornerstownensis (from 577, 12-4, 34-36 cm). Middle row

(left to right): umbilical and dorsal views of Globoconusa daubjergensis, Eoglobigerina eobulloides, and Parasubbotina aff. pseudobulloides (all from site 528, core section 31-6, 141- 143 cm). Bottom row (left to right): Chiloguembelina midwayensis (from 577, 12-4, 34-36 cm), Parasubbotina pseudobulloides (from 577, 12-4, 54-56 cm), and Praemurica taurica (from 577, 12-4, 34-36 cm).

In order to determine the relationship of stable isotopic signals to test size, we picked multiple specimens of each taxon from several different size fractions. These size fractions

were taken at 1/4 phi intervals (12 to 55 gm intervals, increasing with increased mean diameter). For each taxon, these size fractions ranged from a minimum sieve size of 63 grn to the largest 1/4 phi sieve size which contained representatives of that taxon (i.e., 106 gm for Globoconusa daubjergensis, 355 gm for Praemurica taurica). For interspecies comparisons, we sampled this size range for all abundant taxa present in three earliest Paleocene samples: a lower Parvularugoglobigerina eugubina (Ptx) zone sample of DSDP site 577, and lower Parasubbotina pseudobulloides (Pla) zone samples of DSDP sites 528 and 577 [D'Hondt and Keller, 1991 ].

This study follows the taxonomic concepts of Olsson et al. [ 1992] for species that exhibit cancellate surface texture (Eoglobigerina eobulloides, Parasubbotina pseudobulloides, Parasubbotina aff. pseudobulloides, and Praemurica taurica) and

of S. D'Hondt (unpublished manuscript, 1991) for the microperforate taxa (Guembelitria cretacea, Woodringina claytonensis, Woodringina hornerstownensis, Chiloguembelina midwayensis, Parvularugoglobigerina eugubina, and Globoconusa daubjergensis) (Figure 1). Where these authors have relatively broad species concepts, we restricted the analyzed populations to more narrowly defined morphotypes in order to minimize the effect of phenotypic variation on our isotopic results. For example, P. eugubina was restricted to very low-spired microperforate forms with an umbilical to extraumbilical aperture and five chambers in the outer whorl, while E. eobulloides was restricted to forms with

a cancellate wall structure, an umbilical or slightly extra umbilical aperture, and only four chambers in the outer whorl. As defined in this study, P. aff. pseudobulloides is limited to morphotypes with four to four and a half chambers in the outer whorl and an extra umbilical aperture. P. aff. pseudobulloides was referred to as Morozovella moskvini (Shutskaya) by D'Hondt and Keller [ 1991] and is morphologically intermediate

532 D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera

between E. eobulloides and P. pseudobulloides. The taxon discussed here as P. taurica was referred to as Morozovella

inconstans by D'Hondt and Keller [ 1991].

Stable Isotopic Methods

In preparation for isotopic analyses, individual planktonic foraminifera were hand picked, sonicated in distilled water, rinsed with ethyl alcohol, and roasted in vacuo at 380øC. Analyses were performed with an automated Kiel carbonate extraction device coupled to a Finnegan MAT-251 mass spectrometer. The carbonate reaction and distillation procedures were carried out in the extraction device, which reacts 0.01 to 0.03 mg of carbonate sample in individual reaction vessels with 3 drops of phosphoric acid at 75øC and then isolates the CO2 in a single step distillation. The CO2 is then introduced directly into the MAT-251 mass spectrometer for measurement. NBS 18, 19, and 20 as well as an in house standard LV-2 were measured on a daily basis to monitor instrument calibration and analytical accuracy. All isotope values are expressed in the (15) notation where

•513C = [13C/12Csample- 13C/12Cstd] / [13C/12Cstd] x 103

relative to the PDB standard (Appendix A). Corrections were made for 170 interference [Craig, 1957]. Precision of replicate analyses was better than _-+0.1 for both oxygen and carbon compositions.

Calculation of the carbonate mass used in each analysis allows us to directly compare size-1513C relationships of taxa that diverge in their relationship of mean diameter to mean mass (foraminifera of similar test diameter may differ greatly in test mass, depending on such factors as tightness of coil and thickness of chamber wall [Schweitzer and Lohmann, 1990]). It also allows estimation of the incremental effect of chamber

addition on mean [13C values (regardless of size-related changes in test morphology). In order to estimate the 1513C of calcite added with increasing mass, we assumed that the maximum gas pressure evolved in a carbonate analysis is directly proportional to the mass of analyzed calcite. This simply assumes that (1) the mass spectrometer sample chamber was maintained at the same volume for each

maximum pressure measurement and (2) the analyzed calcite was totally acidified. This relationship has been quantified for the mass spectrometer used in this analysis. For the A intake line of the mass spectrometer,

M = 9.79 x 10( 0'001 x Pmax), [r 2 = 0.984],

whereas for the B line,

M = 10.589 x 10( 0.001 x Pmax), [r2= 0.986].

For both equations, M is the total carbonate mass (gg) acidified in that analysis and Pmax is the maximum gas pressure (gbar) evolved in that analysis. The equations are defined by samples that evolved maximum pressures of 250 to 870 gbar (A line) and 250 to 780 gbar (B line). The high correlation coefficients indicate that Pmax can be used to

directly estimate M over that range of maximum pressures.

Given these relationships, we can use the 1513C-mass equation of Oppo and Fairbanks [1989] to quantify within species relationships between [13C and carbonate mass:

1513Ci = A + B x ln(M/S)i

where M is the total mass (gg) of carbonate acidified for a single analysis, S is the number of individual specimens included in that analysis, i = 5, 10, 15...n increments of M/S, •513Ci is the calculated •513C of a whole foraminifer with mass proportional to (M/S)i, and A and B are regression coefficients taken from the line fit of the natural log of M/S to the measured [13C. Using the calculated values of [13C i, we can then estimate the [13C of the calcite added with an increase in mass [Oppo and Fairbanks, 1989]:

1513Cca [i-1-->i] x ((M/S)i - (M/S)i-1) = (1513Ci x(M/S)i) - (1513Ci_1 x (M/S)i-1)

where 1513Cca [i-1-->i] is the mass of calcite added as M/S increases from increment i-1 to increment i.

Isotopic paleotemperatures were calculated using the [18 0 equation of Erez and Luz [ 1983], assuming a mean 15180 value of 1.2 (SMOW) for earliest Paleocene seawater.

Estimating Postdepositional Effects

Sedimentary disturbance. Marine sediments can be mixed or reworked by a number of processes, including bioturbation, drilling disturbance, downslope transport, and current winnowing. Such disturbance clearly can affect stable isotopic results by mixing faunas of different ages and/or paleoenvironments. Since there is no visible evidence of such reworking in these samples, any sedimentary mixing probably does not exceed background levels. The background level of sedimentary disturbance in these lowermost Paleocene samples can be estimated by calculating the degree of faunal mixing immediately below and above the K/T boundary (which underlies our site 528 sample by 0.9 m and both of our site 577 samples by less than 2.5 m).

On the basis of relative abundance counts of 250 to 430

planktonic foraminifera, the degree of sedimentary mixing at site 528 is minimal. Essentially 0 % of the fauna 1 cm below the K/T boundary and 100 % of the fauna 1 cm above the boundary are typical Paleocene taxa [D'Hondt and Keller, 1991; S. D'Hondt and T. D. Herbert, unpublished manuscript, 1993]. At site 577, the situation is more complicated. In the 60 cm immediately underlying the K/T boundary, Paleocene taxa dominate planktonic foraminiferal relative abundance counts (85 to 97 %), but foraminifera constitute less than 1% of the total sediment [Zachos and Arthur, 1986; D'Hondt and Keller, 1991]. On the basis of counts of 425 to 1100 specimens, Paleocene taxa consitute 100 % of the foraminifera in Lower

Paleocene samples at site 577, where 20 to 30 % of the total sediment is comprised of foraminifera [Gerstel et al., 1986, 1987; Zachos and Arthur, 1986; D'Hondt and Keller, 1991]. A worst case estimate derived from these data suggests that about 5 % of the sediment at the site 577 K/T boundary is disturbed downward by drilling or bioturbation. At both sites, it appears likely that sedimentary disturbance plays only a minor role in modifying earliest Paleocene stable isotope records.

D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera 533

Diagenetic Alteration. Diagenetic alteration of foraminiferal calcite can potentially alter initial stable isotopic values 2.0. [Matter et al., 1975' Killingley, 1983; Arthur et al., 1989]. Several criteria can be applied to assess the extent of diagenetic

• 1.6, alteration in fossil foraminifera, including elemental and stable isotopic ratios [Baker et al., 1982; Richter and DePaolo, 1988]. Because diagenesis generally lowers the Sr/Ca ratio of foraminiferal calcite, Sr/Ca comparison with well preserved 1.2. coeval specimens provides one such test of diagenetic alteration [Delaney, 1983; Hess et al., 1986]. Sr/Ca ratios in 0.8

planktonic foraminifera from site 577 core sections 12-4 and 0 12-5 are consistent with those of coeval well preserved specimens [Hess et al., 1986] (compare to Graham et al. 2.2 [1982]). No equivalent data are available for site 528. On kiloyear time scales, 87Sr/86Sr ratios equilibrate throughout

• 1.8

the world ocean. For this reason, postdepostional alteration of marine carbonates can also be assessed by comparison of strontium isotope ratios to well-preserved coeval material. The 87Sr/86Sr ratios in planktonic foraminifera of these sites 1.4 528 and 577 core sections are consistent with those of

contemporaneous well preserved specimens from other sites [Hess et al., 1986' Zachos and Arthur, 1986; Martin and l'ø0 MacDougall, 1991].

Foraminifera from all three samples are free of carbonate 2.2 infilling, but marked by some surficial overgrowth or recrystallization. This overgrowth is most noticeable on

• 1.8, specimens of microperforate taxa from site 577 (Figure 1). However, due to oxygen, carbon, and strontium isotopic consistency with coeval planktonic foraminiferal samples from other sites [Boersma et al., 1979; Boersma, 1984; Hess et al., 1.4- 1986; Zachos and Arthur, 1986; D'Hondt and Lindinger, 1988; Martin and MacDougall, 1991], we believe that carbonate

1.0

overgrowth has not seriously altered the isotopic values of the 0 site 528 and 577 foraminiferal samples. The depth of burial of these samples ranged from 106.7 and 109 m at site 577 to 406.2 m at site 528.

Oxygen and carbon isotope ratios of planktonic foraminiferal calcite also vary with the degree of test dissolution [Wu and Berger, 1989]. Usually, such dissolution preferentially affects smaller (thin walled) and surface dwelling specimens [Wu and Berger, 1989]. In general, dissolution decreases the influence of early ontogenetic calcite on whole-test stable isotopic values and biases death assemblages toward larger specimens and cold water faunas. Individual earliest Paleocene specimens exhibit low to moderate test dissolution at site 528 and exhibit

no noticeable dissolution at site 577 [D'Hondt and Keller, 1991]. At both sites, the presence of large populations of G. cretacea and other relatively small, fragile, and surface dwelling forms suggests that dissolution has not strongly affected earliest Paleocene planktonic foraminiferal populations.

Patterns of Stable Isotopic Variation

Carbon Isotopes. We examined 10 earliest Paleocene planktonic foraminiferal taxa for size-related variation in •513C signals. All 10 taxa exhibit clear trends of increasing whole- test •513C values with increasing size (Figure 2). The magnitude of size-related variation in •513C varies from species to species, ranging from 0.2 to 1.1%0. In general, taxa characterized by small individuals display relatively negative

(a)

2;0 3;0 400

(b)

ß []

[] []

[] []

[]

[] ß ß

400

(c)

O• GI•,A A A A

B

Or

1;0 2;0 3;0 400

mean test diameter (gm)

Figure 2. Plots of •513C versus test size for earliest Paleocene taxa. Test size is measured in 1/4 phi sieve increments and plotted on the x axis at the lower bound of each 1/4 phi increment. (a) data from a Pla zone sample of DSDP site 577 (core section 12-4 (34-36 cm)). (b) data from an older Pla zone sample of DSDP site 528 (core section 31-6 (141-143 cm)). (c) Plx zone sample of DSDP Site 577 (core section 12- 5 (115-117 cm)). Note the positive relationships between test size and •513C exhibited by the planktonic taxa (solid circles, P. taurica, solid squares, P. pseudobulloides; open triangles, W. claytonensis; open circles, W. hornerstownensis; horizontally striped squares, C. midwayensis; dotted squares, P. aft. pseudobulloides; solid triangles, E. eobulloides; vertically striped squares, G. daubjergensis; horizontally striped circles, P. eugubina and open squares, G. cretacea). The remaining species are benthic (N, N. truempyi; G, G. cf nacatochensis; A, A. velascoensis; B, B. cf midwayensis and; and O, O. velascoensis).

•513C signals and a narrow range of size-related •513C variation (0.2 to 0.4 %0: G. cretacea, C. midwayensis, G. daubjergensis). While E. eobulloides and the biserial Woodringina species exhibit the greatest •513C increase, only P. pseudobulloides and P. taurica appear to approach stable

534 D'Hondt and Zachos' Isotopic Variation and Planktonic Foraminifera

•513C values. Although different taxa are characterized by size- related •513C trends of different magnitude and slope, those intraspecies trends do not appear to change significantly in slope and magnitude between different samples at the same site (Figure 3).

Comparison of •513C signals to mean test mass underscores strong differences between taxa in their magnitude of •513C increase with equivalent increases in mass. For example, the Woodringina species, E. eobulloides and P. eugubina clearly exhibit much steeper increases in •513C than co-occurring taxa do over the same range of mass (Figure 4, Table 1). In comparison, P. pseudobulloides and P. taurica exhibit very small increases in •513C values over a wide range of larger test sizes (Figure 4).

When the estimated effect of early ontogenetic calcite is subtracted from whole-test planktonic signals, whole-test •513C signals appear to lag those of the incremental calcite added with increasing test size (Figure 5). For example, in the zone Pla sample from site 577, the •513Ci value of each taxoffs largest size fraction is between 0.1 and 0.5 %o lower than the •513Cca value of the calcite added between the largest and second largest size fractions (Figure 5). In effect, this suggests that size-related trends in the •513C signals of whole foraminiferal tests can greatly underestimate actual differences between the •513C values of the smallest specimens measured and the values of the last calcite added. One implication of this result is the conclusion that whole-test •513C signals underrepresent early Paleocene seawater values by at least 0.1 to 0.5 %o--if the •513Cca values closely approached watermass

1.4

1.0

0.6

ß i i

1.0 1.4 1.8

Site 577, core 12-4, 34-36 cm

Figure 3. Comparison of •13C values from Woodringina claytonensis of two site 577 samples. The linear fit is defined by the equation y = 0.093175 + 1.1479x. r 2 = 0.977. The approximately unitary slope suggests a nearly constant relationship between environmental •13C values and the •13C values recorded by W. claytonensis tests of different size fractions. Note that the value of the y intercept is slightly greater than that of the x intercept. This suggests a slight increase in the mean •13C value of site 577 W. claytonensis habitats between the temporal intervals represented by these two samples.

2.0 ÷ (a)

A O

0.8

10

2.2'

1.8

1.4

1.0

(b)

2.2'

1.8

1.4

(c) A

A

1.0

0 :• • • • 10

mean test mass (gg)

Figure 4. •513C values plotted against mean test mass. (a) species from DSDP site 577 (core section 12-4 (34-36cm)). (b) species from DSDP site 528 (core section 31-6 (141-143 cm)). (c) species from site 577 (core section 12-5 (115-117 cm)). Symbols are the same as in Figure 2. Note that P. pseudobulloides and P. taurica attain the greatest mass, although W. claytonensis and W. hornerstownensis reach higher •513C values. For further explanation, see the text and Table 1.

•513C values in larger foraminiferal size fractions. If these •513Cca values did not reach equilibrium with paleowater values, then early Paleocene whole-test •513C signals underrepresent seawater values by even greater magnitudes.

Stable isotopic signals from multiple size fractions of several benthic taxa were also measured (Figure 2). Of the three analyzed sediment samples, the site 577 Zone P{x sample provided the greatest number of monospecific benthic samples spanning a wide range of size fractions (> 150 to >355 gm). None of these exhibited clear trends between •513C and test size, and most displayed relatively stable values over the analyzed range of test sizes (Figure 2). It remains uncertain whether smaller specimens of these taxa would display a clear relationship between •i13C and test size.

D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera 535

TABLE 1. Regression Coefficients tbr the Line Fit of the Natural Log of Mean Test Mass (gg) to/518C Value ( %0 PDB)

Taxon A B r2

DSDP site 577 12-4 (34-36 cm) Woodringina claytonensis 1.5454 0.95605 Woodringina hornerstownensis 1.4171 1.1329 Chiloguembelina midwayensis 1.1869 0.42655 Parasubbotina pseudobulloides 1.4905 0.27841 Praemu rica tau rica 1.3811 0.38868

DSDP site 577 12-5 (115-117 cm) W. claytonensis 1.8933 1.1572 Parvularugoglobigerina eugubina 1.8332 0.81106

DSDP site 528 31-6 (141-143 cm) Globoconusa daubjergensis 1.3909 0.35322 Guembelitria cretacea 1.6838 0.50339

C. midwayensis 1.2573 0.35179 Parasubbotina aff. pseudobulloides 1.7859 0.52995 Eoglobigerina eobulloides 1.6544 0.81328

0.966 0.964

0.963 0.795 0.829

0.962

0.925

0.648

0.886

0.888 0.986 0.968

Interspecies •513C gradients can be calculated from the maximum •513C values exhibited by each planktonic species within a sample. These gradients range from 0.6 %0 in the site 577 zone Pla sample to 0.8 %0 in the site 577 Zone sample to 1.1%o in the site 528 zone Pla sample. In two of the three samples, these interolanktonic gradients are greater

2.5

2.0

1.5

1.0

0.5

ø ooø o

o

••_•..•======= ....

O_ffi •r

10

mean mass (micrograms)

Figure 5. Examples of the difference between mean •513C values of whole tests and estimated •513C values of calcite added with increased mean mass. Symbols for whole-test values are the same as in Figure 2. These represent mean whole-test values at 0.25 microgram increments of mass, while the overlying open symbols represent the estimated •513C values of the calcite added at each increment. Trends are figured for three species from site 577, 12-4 (34-36). Note that P. pseudobulloides displays the smallest difference between whole-test •513C values and •513C values of calcite added at all size fractions. For further explanation, see the text, Figure 4, and Table 1.

than planktonic-to-benthic •5f3C gradients calculated from the most enriched planktonic and benthic signals (0.2 %0 in the site 577 zone Plx sample, 0.35 %0 in the site 528 zone Pla sample, and 0.8 %0 in the site 577 zone Pla sample). This discrepancy suggests one or more of the following: (1) the ambient vertical or seasonal •513C gradient was greater in near- surface waters than the total •513C difference between surface and deep waters, (2) vital effects in these planktonic and/or benthic taxa lead the planktonic-to-benthic •513C gradient to underrepresent the actual surface-to-deep •513C gradient, and/or (3) planktonic vital effects cause the interplanktonic •513C difference to overestimate the true difference in ambient •513C values of different planktonic foraminiferal habitats.

Oxygen Isotopes. The range of •5180 values exhibited by each planktonic species is between 0.2 and 0.5 %0 (Figure 6). The total range of •5180 values displayed by all planktonic species within each lowermost Paleocene sample ranges from 0.7 %0 in the site 577 Zone Po• sample to 0.9 %0 in the site 577 zone P1 a sample. If we limit our comparison to the •5180 signals exhibited by the largest size fraction of each taxon, these interspecies ranges are even lower; 0.3 %0 in the 577 zone Po• sample, 0.6 %0 in the site 528 sample, and 0.5 %0 in the site 577 zone Pla sample. Despite these low interspecies •5180 differences and relatively high intraspecies variability, the different taxa are characterized by distinct •5180 signals: (1) in the zone Po• sample, G. cretacea and W. claytonensis exhibit lower •5180 values than P. eugubina, and (2) in the zone Pla samples, G. cretacea and the Woodringina species display lower values than G. daubjergensis, C. midwayensis and the cancellate species (Figure 6).

All three samples are characterized by relatively low planktonic-to-benthic •5180 gradients (Figure 6). In the sample from mid-latitude site 528, the •5180 values of E. eobulloides overlap those of co-occurring benthic taxa. The

536 D'Hondt and Zachos' Isotopic Variation and Planktonic Foraminifera

1.0-

0.5-

m 0.0.

• -0.5.

-1.5

0

1.0

(a)

1(;o 2(;0 3(•o 400

0.5

0.0'

-0.5

-1.0

-1.5

(b)

A • $ G,N

0 1;0 200 300

1.0

0.5

• 0.0

• -0.5

-1.0

-1.5

B A Or (C) OAr A A Or •r n

G G Or G G,G G

¸ ¸ ¸

o ,;o

Or

mean test diameter ([tm)

Figure 6. Plots of •5180 vs test size for earliest Paleocene taxa. (a) Data from site 577, 12-4 (34-36 cm). (b) data from site 528, 31-6 (141-143 cm). (c) Data from site 577, 12-5 (115-117 cm). Symbols are the same as in Figure 2. For further explanation, see the text.

400

400

largest planktonic-to-benthic •5180 difference in this sample is 0.9 %0. Planktonic-to-benthic gradients are slightly higher in the near-equatorial site 577 samples. These samples are characterized by maximum planktonic-to-benthic gradients of 2.2 %0 and differences of 0.8 to 0.9 %0 between their most

enriched planktonic •5180 signals and most depleted benthic signals (Figure 6). Several previous studies also document relatively small differences between benthic •5180 values and the most enriched planktonic values in Maestrichtian and lower Paleocene samples [Douglas and Savin, 1978; Boersma et al., 1979; Boersma and Shackleton, 1981; Boersma, 1984; Stott and Kennett, 1990a, b]. Together with the low interplanktonic •5180 differences, this indicates relatively low surface-to-deep •5180 differences and/or fairly large •5180 disequilibria in the benthic and/or planktonic taxa analyzed.

DISCUSSION

Interpreting the 613C Trends

As illustrated in our results, earliest Paleocene planktonic foraminifera display increased •513C values at larger test sizes. In this respect, they closely resemble extant taxa. Comparison with modem taxa suggests that this general size-related trend was driven primarily by increased reliance on ambient CO2 for calcification at larger test sizes [Berger et al., 1978; Erez, 1989]. The absence of a positive correlation between test size and •5180 values suggests that this general trend was probably not driven by decreased kinetic fractionation during calcification at larger test sizes [Berger et al., 1978; McConnaughy, 1989a, b]. P. pseudobulloides and P. taurica do not exhibit the large magnitude of size-related •513C change that is characteristic of modem photosymbiotic taxa [Berger et al., 1978; Spero and Williams, 1988; Oppo and Fairbanks, 1989]. For this reason, it appears unlikely that these largest earliest Paleocene forms relied heavily on photosymbiotic carbon sources. Unlike those taxa, E. eobulloides,

Woodringina spp., and P. eugubina do exhibit large size-related •513C increases at all test sizes (Figures 2, 4, and 5; Table 1). It appears likely that either photosymbiotic activity was important to the carbon budget of these taxa or, relative to co- occurring taxa, they used a much higher proportion of respiratory to ambient carbon for calcification at small sizes and/or a much lower proportion of respiratory to ambient carbon at larger sizes.

Such size-related trends probably cause whole-test •513C signals of some or all earliest Paleocene taxa to underrepresent paleoenvironmental •513C values. They probably also result in interplanktonic •513C differences that overestimate real •513C gradients between different earliest Paleocene planktonic foraminiferal habitats. As discussed previously, in modem taxa, the smallest size fractions may be most reliant on metabolic CO2 for calcification. This is consistent with our earliest Paleocene results, where the smallest size fractions of almost all of the examined taxa display more negative •513C values than did benthic foraminifera in the same or

stratigraphically contiguous samples. Even at their largest test sizes, both G. cretacea and G. daubjergensis consistently display maximum •513C values that are more negative than the maximum values of both benthic foraminifera and the other

planktonic taxa within the same samples (Figure 2). In the site 528 sample, the •513C signals of all C. midwayensis size fractions are also more negative than the signals of co- occurring benthic samples (Figure 2).

The negative •513C difference between these last three planktonic taxa and co-occurring benthic samples can be reconciled by accepting one or more of the following: (1) a strong negative vital effect in the planktonic taxa, (2) a positive vital effect in the benthic taxa, or (3) negative •513C gradients betwhen planktonic and benthic habitats. We interpret this difference to reflect a size-related negative vital effect in the planktonic taxa, combined with a relatively low background surface-to-deep •513C gradient. Three factors underscore the importance of size-related vital effects to these negative planktonic-to-benthic differences. First, of these earliest Paleocene planktonic species, G. cretacea, G.

D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera 537

daubjergensis and C. midwayensis are characterized by the smallest maximum size fractions. Second, G. cretacea and G.

daubjergensis are relatively most abundant in shallow marine environments and the former exhibits the most negative adult bl 8 0 values in its planktonic assemblages (Figure 6) [Smith and Pessagno, 1973; Boersma et al., 1979; Boersma, 1984; Leckie, 1987; D'Hondt and Keller, 1991; Liu and Olsson, 1992]. These biogeographic and oxygen isotopic characteristics suggest a near-surface habitat. Such a near- surface habitat precludes these taxa from exhibiting equilibrium •i13C values that are more negative than deepwater values. Even in a low-productivity ocean, the surface-to-deep carbon cycle and preferential uptake of 12 C by photosynthetic organisms prohibits near-surface equilibrium •i13C values from being globally more negative than deepwater •13C values. Third, like extant planktonic taxa, modem deep-sea benthic taxa generally display negative •il 3 C vital effects and are unlikely to display •i 13C values that are more positive than planktonic foraminifera in isotopic equilibrium with the overlying water column [Belanger et al., 1981; Graham et al., 1981; Vincent et al., 1981; Grossman, 1987]. The effect of low background •i13C gradients on these negative planktonic- to-benthic differences is illustrated by the C. midwayensis to benthic difference, which is negative in the zone Pla sample from site 528 but positive in the slightly younger zone Pla sample from site 577 (Figure 2).

These problems underscore the need to consistently use tightly constrained planktonic foraminiferal samples in measuring stable isotopic trends. Comparison to modern taxa suggests that relative temporal and geographic trends may be tracked by using any consistent and narrowly defined size fraction of a single taxon [Bouvier-Soumagnac and Duplessy, 1985; Oppo and Fairbanks, 1989]. This point is supported by the fact that Woodringina claytonensis exhibits parallel •i13C trends in different earliest Paleocene samples at site 577 (Figure 3). However, comparison between stable isotopic signals of different taxa or size fractions requires detailed understanding of stable isotopic size-dependence and intertaxon variation. For example, tracking relative change in near- surface equilibrium values requires stable isotopic analysis of a single near-surface taxon or of taxa whose relative vital effects are well understood. For fossil species, the influence of vital effects on •13C trends can be minimized by analyzing a narrow size range of a taxon that exhibits relatively stable •i13C values over that size range. Larger specimens of P. pseudobulloides and P. taurica appear to meet this criterion (Figure 4).

Interpreting the (5180 Results

The interplanktonic •18 0 differences in our lowermost Paleocene site 528 and 577 samples agree closely with those in other low and middle latitude samples of Maestrichtian and early Paleocene age (0.65 to 0.9 for Maestrichtian faunas and 0.7 to 1.1 for earliest Paleocene faunas) [Douglas and Savin, 1978; Boersma and Shackleton, 1981; Boersma, 1984]. If we assume earliest Paleocene planktonic taxa to accurately reflect equilibrium bl 8 0 values and no within-sample intertaxon paleosalinity variation, these interplanktonic differences are equivalent to very small ranges in average paleohabitat

temperatures. In our samples, the maximum paleotemperature difference between the largest specimens of each taxon varies from 1.3 øC in the site 577 zone Po• sample, to 2.2 øC in the site 577 zone Pla sample to 2.6 øC in the 528 zone Pla sample. Because these are within-sample differences, they are not sensitive to changing assumptions of global ice volume. However, they are sensitive to intraspecies vital effects, intraspecies variation in seasonality and paleohabitat depth, and local paleosalinity variation. If we accept a possible vital effect of 0 to 0.35 %0 for any or all of of these taxa, then these estimated paleotemperature differences could differ from actual mean interspecies gradients by +1.5 øC. Despite the possibility of such vital effects, these are low interplanktonic differences relative to modem tropical interplanktonic differences (3.6 %0 in the tropical Indian Ocean) [Williams et al., 1977]. They more closely resemble modern subpolar interplanktonic differences in magnitude (1.0 to 1.6 %0 in the southern Indian Ocean) [Williams et al., 1977]. These small interplanktonic differences and low planktonic-to-benthic •180 differences of the lowermost Paleocene samples suggest low surface-to-deep temperature and salinity gradients in early Paleocene temperate and tropical oceans.

As noted in our results, different earliest Paleocene taxa are

generally characterized by distinct •i 180 signals. However, differences between these specific •18 0 signals are usually small (Figure 6). In the site 528 and 577 samples, several taxa differ by less than the 0.35 %0 offset attributable to many modern taxa. For example, the •i 180 signals of the largest size fractions of G. cretacea, W. claytonensis and P. eugubina are within a 0.3 %0 range at Site 577. Similarly, the bl 8 0 values of the largest fractions of C. midwayensis and the various cancellate species consistently fall within a 0.2 %0 range. Other earliest Paleocene species exhibit greater differences; at site 528, G. cretacea differs from G. daubjergensis, C. midwayensis, E. eobulloides, and P. aft pseudobulloides by 0.4 to 0.6 %0 and at site 577, the Woodringina species differ from C. midwayensis by almost 0.5 %0 (Figure 6). Comparison to previous data indicates that the relative order and magnitude of these specific •18 0 signals remains relatively constant from site to site [Boersma et al., 1979; Boersma, 1984]. In total, these data indicate that, while some of the interspecies variation in •i 180 may be due to vital effects, differences between the most 180_enriched and 18 O_ depleted species largely reflect small but real interspecies habitat variation.

The interplanktonic •i 180 differences are consistent with both depth-partitioning of planktonic foraminiferal habitats and seasonal habitat differentiation. Given paleobiogeographic data, the former explanation is preferable for most of these taxa. The 180_depleted G. cretacea is relatively most abundant in nearshore and shelf sequences, where the 180_enriched p. eugubina and C. midwayensis are less common [Olsson, 1970; Keller, 1988, 1989; D'Hondt and Keller, 1991; Liu and

Olsson, 1992]. These isotopic and biogeographic data suggest a shallower water habitat for G. cretacea than for the other two

taxa. Biogeographic data and interplanktonic •i180 differences indicate a slightly colder and deeper water habitat for cancellate taxa. Cancellate species are much less abundant in lowermost Paleocene shelf sequences than in most coeval open-ocean sequences [Keller, 1988, 1989; D'Hondt and Keller, 1991; Liu

538 D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera

and Olsson, 1992]. They are also generally less abundant at low latitudes than at high and midddle latitudes [Boersma and Premoli Silva, 1983; D'Hondt and Keller, 1991].

The relationships between fi180 signals and paleobiogeographic affinities are less sharply defined for Woodringina species and G. daubjergensis. While Woodringina tests are relatively 180_depleted ' their relative abundance is higher in open-ocean sequences than shelf sequences [D'Hondt and Keller, 1991; Liu and Olsson, 1992]. Their highest known relative abundance is at mid-Pacific site 577 [D'Hondt and Keller, 1991 ]. It appears likely that Woodringina species preferred a warm-water open-ocean habitat. In contrast to Woodringina species, G. daubjergensis exhibits an unusually positive fi180 signal, given its paleobiogeographic abundance patterns. It comprises up to 69% of lower Paleocene assemblages in Gulf Coast shelf sequences but is much less abundant in open ocean sequences [Keller, 1988, 1989; D'Hondt and Keller, 1991; Liu and Olsson, 1992]. Its highest open-ocean abundance is at high latitude southern ocean ODP site 750A, where it constitutes 11 to 34 % of the planktonic foraminiferal fauna [Liu and Olsson, 1992]. Its lowest relative abundance is at tropical Pacific DSDP site 577, where it comprises less than 2 % of the fauna [D'Hondt and Keller, 1991]. These isotopic and paleobiogeographic data can be reconciled by assuming a slightly cooler surface-water habitat for G. daubjergensis than for G. cretacea and Woodringina. If it did inhabit a cooler sea- surface niche, the difference between its •5180 signal and the more negative signals of co-occurring G. cretacea and Woodringina species may reflect seasonal variation in sea- surface temperature and/or ambient •5180 values.

Despite these specific differences in fi180 signal and paleogeographic affinity, very few of these taxa display clear trends of increasing •5180 with increasing test size (Figure 6). This suggests that they did not migrate upward though strong temperature or salinity gradients over ontogeny. It is unclear whether this primarily reflects limited ontogenetic migration or weak vertical and seasonal temperature and salinity gradients in the earliest Paleocene. In the site 577 zone Pla sample, the biserial taxa display small decreases in fil 8 0 with increases in test size. Comparison to modern taxa suggests that possible explanations of this tendency include (1) an association of smaller adult specimens with colder paleoenvironments [B6 et al., 1973; Deuser et al., 1981, 1987; Wefer et al., 1983; Bouvier-Soumagnac and Duplessy, 1985; Bijma et al., 1990], and (2) increased deviation from isotopic equilibrium at larger test sizes, resulting from higher photosymbiotic activity in larger specimens [Spero, 1992].

Implications for the Strangelove Ocean

Post-K/T boundary "Strangelove" ocean scenarios are supported by three lines of evidence: (1) decreased •513C differences between fine-fraction carbonate and benthic

foraminifera, (2) decreased •13C differences between planktonic and benthic foraminifera, and (3) decreased carbonate accumulation and increased carbonate preservation immediately following the K/T boundary [Arthur, 1979; Boersma and Shackleton, 1981; Hsti et al., 1982; Shackleton et al., 1984; Hsti and MacKenzie, 1985; Zachos et al., 1985, 1989a, 1992;

Gerstel et al., 1986; Zachos and Arthur, 1986; Arthur et al., 1987; Smit and van Kempen, 1987; D'Hondt and Lindinger, 1988; Keller and Lindinger, 1989; Barrera and Keller, 1990; Stott and Kennett, 1990a, b; D'Hondt and Keller, 1991]. Our results have clear implications for the first two lines of evidence.

Fine-fraction or bulk-rock carbonate fil 3 C signals are generally assumed to be dominated by nannofossil carbonate and to represent surface water •513C values. These assumptions are already known to be problematic; calcareous dinoflagellate and coccolithophorid •513C signals may differ from equilibrium values by as much as 6.0 %o [Goodney et al., 1980; Paull and Thierstein, 1987]. This disequilibrium effect appears to be especially strong in modem samples dominated byThoracosphaera [Paull and Thierstein, 1987] and can be expected to affect lowermost Paleocene samples of low and middle latitudes, which generally differ from upper Maestrichtian samples by containing abundant Thoracosphaera [Perch-Nielsen, 1979; Perch-Nielsen et al., 1982; Thierstein and Okada, 1979; Pospichal, 1991; Ehrendorfer and Aubry, 1992]. This effect can be exacerbated by the size-related planktonic foraminiferal vital effect documented here. Even if earliest Paleocene nannofossils secreted calcite in equilibrium with near-surface seawater, the presence of smaller or deeper dwelling planktonic foraminifera will lead bulk-rock or fine- fraction samples to underrepresent sea surface •513C values. Where either Thoracosphaera or smaller planktonic foraminifera constitute a large proportion of such samples, planktonic vital effects can lead to an apparently negative planktonic-to-benthic fi13C gradient. Such negative differences occur in several earliest Paleocene stable isotopic records and invariably result from comparison of benthic isotopic records to those of bulk carbonate, fine carbonate, or small planktonic foraminifera [Boersma and Shackleton, 1981; Hsti et al., 1982; Shackleton et al., 1984; Zachos et al., 1985, 1989a; Zachos and Arthur, 1986; D'Hondt and Lindinger, 1988; Keller and Lindinger, 1989; Barrera and Keller, 1990; Stott and Kennett, 1990].

Some K/T stable isotopic studies have attempted to counter these problems by measuring the stable isotopic signals of monospecific planktonic foraminifera [i.e., Zachos et al., 1989a, 1992; Barrera and Keller, 1990; Stott and Kennett,

1990]. Our results suggest that several factors must be considered in undertaking this approach. First, as discussed previously, all size fractions of some P0 and P. eugubina Zone taxa clearly underrepresent paleoequilibrium fil 3 C values (i.e., G. cretacea, G. daubjergensis, and C. midwayensis). This also applies to smaller size fractions of the other earliest Paleocene species and may apply to all size fractions of all earliest Paleocene planktonic foraminiferal taxa (unless the effect of respiratory carbon incorporation on test •13C was fortuitously countered by photosymbiotic effects). Second, because cancellate species exhibit fil 8 0 signals and biogeographic abundance patterns that suggest relatively deepwater habitats, they probably do not directly record sea surface stable isotope values. Third, analysis of taxa characterized by relatively low increases in • 13Cc a will minimize the influence of size-related vital effects on apparent trends in relative •13C variation of the earliest Paleocene. In basal zone P1 a, the cancellate

species P. pseudobulloides and P. taurica are characterized by

D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera 539

Site 577 $13 C (PDB) Magnetobiostrat. •'• 1 2

planktic

12-5,117-1

this study

+fine fraction (<63 !•m) . P. eugubina [] W. hornerstownensis 0 G. stuarti + P. pseudobulloides [] R. rotundata t• M. uncinata x p. taurica © S. triloculinoides

[] C. midwayensis :• P. petaloidea

Figure 7. The stippled area covers the "Strangelove ocean" interval of low planktonic-to-benthic gradients measured by Zachos et al. [ 1989]. The horizontal bars in C29R and C29N represent the range of •513C values obtained for planktonic foraminifera in this study [samples 577, 12-4 (34-36) and 577,

such low increases in •513Cca at larger size fractions (Figure 5).

Different studies suggest that planktonic-to-benthic •513C gradients decrease at the K/T boundary by different amounts at different sites. It decreases by at least 0.7 %0 at near-equatorial Pacific DSDP site 577 [Zachos et al., 1989a] (Figure 7), about 1.2 %0 at an epicontinental site near present day Brazos River, Texas [Barrera and Keller, 1990], and at least 0.5 %0 at high latitude southern ocean ODP sites 689B, 690C, and 750A

[Stott and Kennett, 1990; Zachos et al., 1992]. All of these studies relied on size fractions broad enough to be susceptible to size-related vital effects (for earliest Paleocene records, the site 577 study relied on planktonic foraminifera greater than 125 gm in diameter, the site 750A study used specimens greater than 150 gm in diameter, and the other two studies relied on specimens greater than 63 gm in diameter). The site 577 analyses relied on P. pseudobulloides for an earliest Paleocene record, whereas Eoglobigerinafringa (= E. eobulloides) and P. pseudobulloides were used at sites 689B

•13 C (PDB) 1 2

benthic fine fraction

-- Aragonia • Gavelinella & Gyroidinoides

A Nuttallides

12-5 (115-117)]. Aragonia, Gavelinella, Gyroidinoides, and Nuttallides are benthic foraminiferal genera. The other 10 taxa are planktonic species. Figure modified from Zachos et al. [1989].

and 690C, E. eobulloides was used at site 750A, and Heterohelix globulosa was used at the Brazos core. Additionally, different taxa were used to define the Cretaceous records at the different sites (Rugoglobigerina rotundata and Globotruncanita stuarti at site 577, Heterohelix globulosa and Globigerinelloides species at sites 689B, 690C and 750A, Heterohelix globulosa at the B[azos core).

Given the species and size f•fictions analyzed in these studies, only at sites 577 and 750A can we presently rule out the possibility that Paleocene size-related •513C variation was a primary cause of the apparent K/T decrease and earliest Paleocene increase in planktonic-to-benthic •513C gradients (Figure 7). Furthermore, it remains possible that the •513C signals of some or all of the previously analyzed Cretaceous taxa were (1) positively offset by photosymbiotic activity, (2) negatively offset by size related trends, and/or (3) offset from the Paleocene taxa by differences in either mean paleohabitat depth or seasonal watermass affinities. The magnitude, rates of change, and regional variation of K/T boundary and earliest

540 D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera

Paleocene surface-to-deep 513C gradients can be better constrained by (1) developing stable isotopic records from narrowly defined size fractions of planktonic foraminifera, (2) comparing 513C signals of latest Maestrichtian and earliest Paleocene planktonic foraminifera from similar paleohabitats (as defined by 5180 signals and paleobiogeographic patterns), and (3) determining the relationship between 513C and test size for relevant Maestrichtian taxa (in order to rule out strong photosymbiotic effects and to determine the range of test sizes characterized by relatively stable intraspecies 513C values).

Implications for Depth-Related Planktonic Foraminiferal Diversification Models

Earliest Paleocene taxa that are closely related and morphologically similar also display similar isotopic signatures. In the site 528 sample, E. eobulloides and P. aff. pseudobulloides resemble each other in their 5180 and 513C signals. Similarly, the cancellate species in the site 577 Pla sample resemble each other in this regard, as do the Woodringina species in the same sample (Figures 2 and 6). To the extent that these isotopic relationships reflect paleoenvironmental associations, they suggest that closely related morphologically similar species generally inhabited similar habitats.

Despite these intra-generic similarities, less closely related taxa do not consistently resemble each other in their stable isotopic signatures. For example, among the microperforate taxa, G. cretacea and the Woodringina species display more negative 5180 signals than P. eugubina, G. daubjergensis, and C. midwayensis. Yet, on morphologic and stratigraphic grounds, G. cretacea appears to have been immediately ancestral to P. eugubina, G. daubjergensis, and Woodringina claytonensis and ancestral to W. hornerstownensis and C. midwayensis via W. claytonensis [Olsson, 1970, 1982; Premoli Silva, 1977; Smit, 1982; D'Hondt, 1991; Li and Radford, 1991; Liu and Olsson, 1992]. Within this clade, speciation events appear to be marked by little or no change in 5180 signals or by a shift to more positive values in the descendant species. An example of the former case is the radiation of biserial Woodringina species from triserial G. cretacea, all exhibiting similarly negative 5180 values. Examples of the latter case include the evolution of trochospiral P. eugubina and G. daubjergensis from G. cretacea and of C. midwayensis from W. claytonensis, with all three descendant taxa displaying more positive 5180 values than their immediately ancestral taxa (Figure 6) [Boersma et al., 1979; Boersma, 1984].

In part, these relationships are consistent with previous models that explicitly couple planktonic foraminiferal diversification and depth of habitat [Hart, 1980; Caron and Homewood, 1983]. These models were originally applied to the evolutionary history of Mesozoic [Caron and Homewood, 1983] and Mesozoic-Cenozoic foraminifera [Hart, 1980]. Both hypotheses follow the speculation of Douglas and Savin [ 1978] in predicting that, during planktonic foraminiferal evolutionary radiations, trochospiral species evolve from shallow-water ancestors to fill progressively deeper water niches. While the isotopic signals of G. cretacea and its earliest Paleocene descendants are consistent with this general

pattern, the applicability of such evolutionary models to the earliest Paleocene foraminiferal radiation is tempered by several caveats. First, these isotopic data simply test possible relationships between planktonic foraminifera and a limited array of local paleoenvironmental characteristics. Most models of planktonic foraminiferal diversification also predict or assume a variety of factors not addressed here. These include extrinsic physical parameters, such as upper water-column stability [Hart, 1980; Caron and Homewood, 1983; Leckie, 1989] and intrinsic biological characteristics, such as reproductive effort and efficiency of nutrient assimilation [Caron and Homewood, 1983]. Second, these models were originally formulated to explain hypothetical patterns of diversification among trochospiral taxa (i.e., Eoglobigerina, Parasubbotina), not serial taxa (i.e., Guembelitria, Woodringina, and Chiloguembelina). In fact, Hart [ 1980] assumed all serial taxa to dwell in shallow water, including Chiloguembelina species. Third, a general surface-to-deep diversification trend among earliest Paleocene microperforate taxa could simply result from exploitation of a near-surface open-ocean habitat by the epicontinental K/T survivor, G. cretacea. Until habitat depths of cancellate taxa and their ancestors are better documented, the applicability of depth- related diversification models to the entire earliest Paleocene foraminiferal radiation will remain unclear.

CONCLUSIONS

Earliest Paleocene planktonic foraminifera exhibit a strong trend toward increased 513C with increased test size. The slope and magnitude of this trend varies between taxa, with Woodringina species displaying the largest shift (1.1%0 over a 130 gm range in mean sieve size) and Guembelitria cretacea displaying the smallest (0.2 %0 over a 38 gm range). Analogy with recent planktonic foraminifera suggests that this general 513C size-dependence probably resulted from increased reliance on respiratory CO2 for calcification at smaller test sizes. The cancellate species P. pseudobulloides and P. taurica are characterized by small increases in 513C at larger size fractions. In contrast, Eoglobigerina eobulloides, Woodringina spp., and Parvularugoglobigerina eugubina exhibit relatively large size-related increases in 513C at both small and large test sizes. Photosymbiotic activity may have been important to the carbon budget of the latter taxa.

These size-related effects provide an alternative explanation for decreases in whole rock 513C values and some decreases in

planktonic-to-benthic foraminiferal 513C gradients documented at open-marine K/T boundary sequences. They also provide a likely explanation of the negative fine-fraction to benthic and planktonic-to-benthic 513C gradients of the earliest Paleocene. For these reasons, accurate paleoceanographic interpretation of planktonic foraminiferal, bulk-rock, or fine-fraction isotopic records requires thorough analysis of the relative abundance, size distribution, and isotopic vital effects of planktonic foraminifera and nannoflora in the samples analyzed.

Despite limited interspecies 5180 differences and size- dependent 513C trends, paleobiogeographic patterns and 5180 signals appear to provide realistic estimates of the relative water mass affinities of earliest Paleocene planktonic foraminiferal species. Isotopic and paleobiogeographic data

D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera 541

suggest shallower habitat depths for G. cretacea and Woodringina species than for P. eugubina, C. midwayensis, P. taurica, E. eobulloides, and Parasubbotina species. These data also suggest a relatively shallow water but cooler season habitat for G. daubjergensis. These results are consistent with a general trend of surface-m-deep diversification of microperforate taxa following the K/T boundary.

See Table AI which shows stable isotopic and size date for monospecific planktic and benthic foraminiferal samples from the three earliest Paleocent sediment samples.

Acknowledgments. We thank J. Bijma, G. P. Lohmann, B. Luz, and H. Spero for perceptive discussions of modem

planktonic foraminiferal ecology and stable isotopic vital effects, and for generously sharing the results of their ongoing research on extant planktonic foraminifera and/or stable isotopes. We also thank M. A. Arthur and K. C. Lohmann for discussion of stable isotopes and W. A. Berggren and R. K. Olsson for helpful perspective on Paleocene taxonomy and biogeography. Thoughtful reviews by A. Boersma, K. G. Miller, R. K. Olsson, and H. Spero improved the final draft of this manuscript. P. Johnson and N.M. Gorbea provided welcome assistance with scanning electron photomicroscopy and plate layout (respectively). All isotopic analyses were performed at the University of Michigan Stable Isotope Laboratory. This research was supported in part by the Petroleum Research Fund of the American Chemical Society (grant PRF-23488-G). Samples were provided by the Ocean Drilling Program. This is GRIPS publication 4.

TABLE A1. Stable Isotopic and Size Data

Taxon Size Fraction 3180, 313C, um PDB PDB

Number of

Specimens Maximum Analyzed (S) Pressure, Pmax

Spectrometer Inlet Line

Site 528 31-6 (141-143 cm)

Chiloguembelina midwayensis

Guembelitria cretacea

Globoconusa daubjergensis

Parasubbotina aft. pseudobulloides

Eoglobigerina eobulloides

Gavelinella cf. nacatochensis

63 to 75 -0.55 1.05

75 to 90 -0.37 1.09

90 to 106 -0.44 1.10

106 to 125 -0.32 1.20

125 to 150 -0.29 1.33

63 to 75 -0.68 1.37

75 to 90 -0.70 1.33

90 to 106 -0.53 1.54

106 to 125 -0.69 1.61

125 to 150 -0.79 1.68

63 to 75 -0.55 1.09

75 to 90 -0.43 1.22

90 to 106 -0.37 1.37

106 to 125 -0.36 1.39

125 to 150 -0.22 1.82

150 to 180 -0.19 1.84

180 to 212 -0.19 2.05

212 to 250 -0.17 2.16

63 to 75 -0.27 1.10

75 to 90 -0.34 1.22

90 to 106 -0.35 1.35

106 to 125 -0.20 1.63

125 to 150 -0.13 1.76

150 to 180 0.06 1.90

180 to 212 -0.18 1.89

212 to 250 -0.28 2.03

180 to 212 0.03 1.69

180 to 212 0.08 1.57

250 to 300 0.14 1.46

65 180 B

60 213 A

35 242 B

25 264 A

13 210 B

67 140 A

55 210 A

40 240 A

25 227 A

14 173 A

65 205 A

55 236 A

40 235 A

11 158 A

17 268 B

12 218 A

8 325 B

4 344 A

65 198 A

55 189 A

40 250 B

25 282 A

16 290 B

12 283 A

8 286 B

6 262 A

3 262

4 381

2 509

542 D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera

TABLE A1. (continued)

Taxon Size Range um

•180

PDB

Number of

• 13C Specimens Maximum Spectrometer

PDB Analyzed (S) Pressure, Pmax Inlet Line

Nuttallides truempyi

Site 577, 12-4 (34-36 cm)

Woodringina claytonensis

Woodrin g ina ho rne rstownensis

Chiloguembelina midwayensis

Parasubbotina pseudobulloides

Praemurica taurica

Site 577, 12-4 (34-36 cm)

180 to 212

180 to 212

250 to 300

63 to 75

75 to 90

90 to 106

106 to 125

125 to 150

150 to 180

180 to 212

63 to 75

75 to 90

90 to 106

106 to 125

125 to 150

150 to 180

180 to 212

63 to 75

75 to 90

90 to 106

106 to 125

125 to 150

150 to 180

63 to 75

75 to 90

90 to 106

106 to 125

125 to 150

150 to 180

180 to 212

212 to 250

250 to 300

63 to 75

75 to 90

90 to 106

106 to 125

125 to 150

150 to 180

180 to 212

212 to 250

250 to 300

300 to 355

0.09 1.76 4 460

-0.05 1.71 3 262

0.17 1.81 2 526

-1.02 0.98 65 234 B

-1.13 1.05 50 223 A

-1.12 1.24 50 346 B

-1.17 1.41 49 426 B

-1.24 1.54 27 398 A

-1.35 1.68 10 250 A

-1.34 1.89 8 203 B

-0.91 0.90 65 318 A

-0.98 1.05 35 331 A

-1.03 1.21 30 294 B

-1.11 1.44 28 393 B

-1.06 1.60 19 394 A

-1.34 1.71 13 370 B

-1.33 1.81 8.5 291 A

-0.45 0.92 65 240 B

-0.51 1.00 60 303 B

-0.68 1.11 35 263 A

-0.57 1.17 25 399 B

-0.75 1.29 12 287 A

-0.88 1.33 9 270 B

-0.83 1.49 65 230 B

-0.75 1.30 55 211 B

-0.97 1.29 36 273 A

-0.86 1.43 25 287 B

-0.82 1.59 12 274 A

-0.83 1.58 9 275 B

-0.78 1.62 6 272 A

-0.95 1.67 4 277 B

-0.95 1.77 2 232 A

-0.90 1.28 65 207 A

-0.80 1.19 60 275 A

-0.89 1.14 40 279 A

-1.06 1.24 19 283 B

-0.89 1.47 11 275 A

-1.18 1.51 9 264 B

-1.01 1.63 6 289 A

-0.91 1.72 7 429 B

-1.08 1.75 4 445 A

-1.04 1.63 3 329 B

Nuttallides truemp¾i 180 to 212 0.39 1.11 4

D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera 543

TABLE A1. (continued)

Taxon Size Range •180 •13C um PDB PDB

Number of

Specimens Maximum Spectrometer

Analyzed (S) Pressure, Pmax Inlet Line Site 577, 12-5 (115-117 cm)

Woodringina claytonensis

Guembelitria cretacea

Parvularugoglobigerina eugubin.

Gavelinella cf nacatochensis

Aragonia velascoensis

Bulimina cf midwayensis

Osangularia velascoensis

63 to 75

75 to 90

90 to 106

106 to 125

125 to 150

150 to 180

180 to 212

63 to 75

75 to 90

90 to 106

106 to 125

63 to 75

75 to 90

90 to 106

106 to 125

125 to 150

150 to 180

180 to 212

150 to 180

150 to 180

180 to 212

180 to 212

212 to 250

250 to 300

150 to 180

180 to 212

212 to 250

250 to 300

300 to 355

150 to 180

180 to 212

150 to 180

150 to 180

180 to 212

250 to 300

300 to 355

> 355

-1.06 1.17 65 230

-1.24 1.38 55 257

-1.25 1.43 40 230

-1.09 1.77 50 493

-1.16 1.86 32 433

-1.06 2.04 16 389

-1.12 2.24 10 276

-1.07 1.37 413

-1.21 1.31 438

-1.01 1.46 454

-1.18 1.48 371

-0.92 1.33 95 482

-0.59 1.52 55 286

-0.6 1.75 60 569

-0.79 1.79 45 578

-0.76 1.88 35 601

-0.83 1.91 13 298

-0.91 1.96 14 290

0.45 1.75 4 199

0.26 2.02 5 260

0.23 1.86 3 246

0.25 1.98 4 285

0.49 2.00 2 266

0.27 1.93 1 288

0.67 1.75 4 225

0.86 1.76 4 291

0.67 1.68 2 353

0.67 1.72 2 396

0.64 1.78 1 300

0.89 1.79 4 236

0.56 1.42 3 227

0.76 1.37 4 252

0.53 1.57 5 260

0.63 1.52 3 267

0.48 1.01 2 332

0.85 1.75 1 372

0.53 1.86 1 432

REFERENCES

Arthur, M. A., Sedimentologic and geochemical studies of Cretaceous and Paleogene pelagic sedimentary rocks: The Gubbio sequence, Ph.D. dissertation, Princeton Univ., Princeton, N.J., 1979.

Arthur, M. A., J. C. Zachos, and D. S. Jones, Primary

productivity and the Cretaceous/Tertiary boundary event in the oceans, Cretaceous Res., 8, 43-45, 1987.

Arthur, M. A., W. E. Dean, J. C. Zachos, M. Kaminski, S. Hagerty-Reig, and C. Elmstrom, Geochemical expression of early diagenesis in Middle Eocene-Lower Oligocene pelagic sediments in the southern Labrador Sea, site 647, ODP leg 105, Proc. Ocean Drill Proj., Sci. Results, 105, 111-136, 1989.

544 D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera

Baker, P. A., J. M. Gieskes, and H. Elderfield, Diagenesis of carbonates in deep sea sediments - evidence from Sr/Ca ratios and interstitial dissolved Sr data, J. Sediment. Petrol., 52, 71-82, 1982.

Barrera, E. and G. Keller, Foraminiferal stable isotope evidence for gradual decrease of marine productivity and Cretaceous species survivorship in the earliest Danian, Paleoceanography, 5, 867-890, 1990.

Barrera, E., B. T. Huber, S. M. Savin, and P. N. Webb, Antarctic marine temperatures: Late Campanian through Early Paleocene, Paleoceanography, 2, 21-47, 1987.

B6, A. W. H., and J. C. Duplessy, Subtropical convergence fluctuations and Quaternary climates in the middle latitudes of the Indian Ocean, Science, 194, 419-422, 1976.

B6, A. W. H., S. M. Harrison, and L. Lott, Orbulina universa d'Orbigny in the Indian Ocean, Micropaleontology, 19, 150-192, 1973.

Belanger, P. E., W. B. Currey, and R. K. Matthews, Core-top evaluation of benthic foraminiferal isotopic ratios for palaeoceanographic interpretation, Palaeogeogr. Palaeoclimatol. Palaeoecol., 33, 205-220, 1981.

Berger, W. H., Ecological patterns of living planktonic foraminifera, Deep Sea Res., 16, 1-24, 1969.

Berger, W. H., J. S. Killingly, and E. Vincent, Stable isotopes in deep sea carbonates: Box Core ERDC-92, west equatorial Pacific, Oceanol. ActOr, 1,203-216, 1978.

B ijma, J., W. W. Faber, Jr, and C. Hemleben, Temperature and salinity limits for growth and survival of some planktonic foraminifers in laboratory cultures, J. Foraminiferal Res., 20 (2), 95-116, 1990.

Boersma, A., Campanian through Paleocene paleotemperature and carbon isotope sequence and the Cretaceous-Tertiary boundary in the Atlantic Ocean, in Catastrophes and Earth History, edited by W. Berggren and J.A. van Couvering, Princeton University Press, Princeton, New Jersey, 247-278, 1984.

Boersma, A., and I. Premoli Silva, Paleocene planktonic foraminiferal biogeography and paleoceanography of the Atlantic Ocean, Micropaleontology, 29, 355-381, 1983.

Boersma, A., and I. Premoli Silva, Atlantic Paleogene biserial heterohelicid foraminifera and oxygen minima, Paleoceanography, 4 (3), 271-286, 1989.

Boersma, A., and N.J. Shackleton, Oxygen- and carbon- isotope variations and planktonic-foraminifer depth habitats, Late Cretaceous to Paleocene, Central Pacific, Deep Sea Drilling Project sites 463 and 465, Init. Rep. Deep Sea Drill. Proj., 62, 513-526, 1981.

Boersma, A., N.J. Shackleton, M. Hall, and Q. Given, Carbon and oxygen isotope variations at DSDP site 384 (North Atlantic) and some paleotemperatures and carbon isotope variations in the Atlantic Ocean, Init. Rep. Deep Sea Drill. Proj., 43, 695-717, 1979.

Bouvier-Soumagnac, Y., and J. C. Duplessy, Carbon and oxygen isotopic composition of planktonic foraminifera from laboratory culture, plankton tows and recent sediment: Implications for the reconstruction of paleoclimatic conditions and of the global carbon cycle, J. Foraminiferal Res., 15, 302-320, 1985.

Caldeira, K., M. R. Rampino, T. Volk, and J. C. Zachos, Biogeochemical modelling at mass extinction boundaries: Atmospheric carbon dioxide and ocean alkalinity at the K/T boundary, in Extinction Events in Earth History, edited by E.G. Kauffman and O. H. Walliser, Springer-Verlag, Berlin, 333-346, 1990.

Caron, M., and P. Homewood, Evolution of early planktic foraminifers, Mar. Micropaleontol., 7, 453-462, 1983.

Corfield, R. M., and J. E. Cartlidge, Isotopic evidence for the depth stratification of fossil and recent Globigerinina: a review, Hist. Biol., 5, 37-63, 1991.

Corfield, R. M., J. E. Cartlidge, I. Premoli Silva, and R. A. Housley, Oxygen and carbon isotope stratigraphy of the Palaeogene and Cretaceous limestones in the Bottaccione Gorge and the Contessa Highway sections, Umbria, Italy, Terra Nova, 3, 414-422, 1991.

Craig, H., Isotopic standards for carbon and oxygen and correction factors for mass spectrometric analysis of carbon dioxide, Geochim. et Cosmochim. Acta, 21,133-149, 1957.

Currey, W. B., and R. K. Matthews, Equilibrium 18 0 fractionation in small size planktic foraminifera: Evidence from Recent Indian Ocean sediments, Mar. Micropaleontol., 6, 327-337, 1981.

D'Hondt, S. L., Phylogenetic and stratigraphic analysis of earliest Paleocene triserial and biserial planktonic foraminifera, J. Foraminiferal Res., 21 (2), 168-181, 1991.

D'Hondt, S., and M. Lindinger, An extended Cretaceous-Tertiary boundary stable isotope record: Implications for paleoclimate and the nature of the K/T boundary event, Global Catastrophes in Earth History, 40- 41, Lunar Planet. Inst., 1988.

D'Hondt, S., and T. D. Herbert, Marking time at the South Atlantic Cretaceous-Tertiary boundary,. Geol. Soc. Am. Abstr. Programs. 23, A179, 1991.

D'Hondt, S., and G. Keller, Some patterns of planktic foraminiferal assemblage turnover at the Cretaceous-Tertiary boundary. Mar. Micropaleontol, 17, 77-118, 1991.

Delaney, M. L., Foraminiferal trace elements: Uptake, diagenesis, and 100 m.y. paleochemical history, Ph.D. dissertation, MIT-WHOI joint program, Woods Hole, Massachusetts, 253 pp., 1983.

Deuser, W. G., Seasonal variations in isotopic composition and deep-water fluxes of the tests of perenially abundant planktonic foraminifera of the Sargasso Sea: Results from sediment-trap collections and their paleoceanographic significance, J. Foraminiferal Res., 17, 14-27, 1987.

Deuser, W. G., E. H. Ross, C. Hemleben, and M. Spindler, Seasonal changes in species composition, mass, size, and isotopic composition of planktonic foraminifera settling into the deep Sargasso Sea, Palaeogeogr., Palaeoclimatol., Palaeoecol., 33, 103-127, 1981.

Douglas, R. G., and S. M. Savin, Oxygen and carbon isotope analyses of Cretaceous and Tertiary foraminifera from the central North Pacific, Init. Rep. Deep Sea Drill. Proj., 17, 591-607, 1973.

Douglas, R. G., and S. M. Savin, Oxygen isotopic evidence for the depth stratification of Tertiary and Cretaceous planktic foraminifera, Mar. Micropaleontol., 3, 175-196, 1978.

Dudley, W. C., and D. E. Goodney, Oxygen isotopic content of coccoliths grown in culture, Deep Sea Research, 26, 495-503, 1979.

Dudley, W. C., P. Blackwelder, L. Brand, and J.-C. Duplessey, Stable isotopic composition of coccoliths, Mar. Micropaleontol., 10, 1-8, 1986.

Duplessy, J. C., A. W. H. B6, and P. L. Blanc, Oxygen and carbon isotopic composition and biogeographic distribution of planktonic foraminifera in the Indian Ocean, Palaeogeogr., Palaeoclimatol., Palaeoecol., 33, 9-46, 1981.

D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera 545

Durazzi, J. T., Stable-isotope studies of planktonic foraminifera in North Atlantic core-tops, Palaeogeogr., Palaeoclimatol., Palaeoecol., 33, 157-172, 1981.

Ehrendorfer, T., and M.-P. Aubry, Calcareous nannoplankton changes across the Cretaceous/Paleocene boundary in the Southern Indian Ocean (site 750), Sci. Results, Proc. Ocean Drill. Proj., 120, part 2, 451-470, 1992.

Emiliani, C., Depth habitats of some species of pelagic foraminifera as indicated by oxygen isotope ratios, Am. J. Sci., 252, 149-158, 1954.

Emiliani, C., Depth habitats of growth stages of pelagic foraminifera, Science, 173, 1122-1124, 1971.

Erez, J., Calcification physiology in foraminifera: palaeo- oceanographic implications,. Terra abstracts, 1, 86, 1989.

Erez, J., and S. Honjo, Comparison of isotopic composition of planktonic foraminifera in plankton tows, sediment traps and sediments, Palaeogeogr. Palaeo. climatol. Palaeoecol. 33, 129-156, 1981.

Erez, J., and B. Luz, Experimental paleotemperature equation for planktonic foraminifera, Geochim. Cosmochim. Acta, 47, 1025-1031, 1983.

Fairbanks, R. G, P. H. Wiebe, and A. W. H. B6, Vertical distribution and isotopic composition of living planktonic foraminifera in the western North Atlantic, Science, 207, 61-63, 1980.

Fairbanks, R. G., M. Sverdlove, R. Free, P. H. Wiebe, and A. W. H. B6, Vertical distribution and isotopic fractionation of living planktonic foraminifera from the Panama Basin, Nature, 298, 841-844, 1982.

Gerstel, J., R. C. Thunell, J. C. Zachos, and M. A. Arthur, The Cretaceous-Tertiary boundary event in the North Pacific: Planktonic foraminiferal results from Deep Sea Drilling Project site 577, Shatsky Rise, Paleoceanography, 1 (2), 97- 117, 1986.

Gerstel, J., R. Thunell, and R. Ehrlich, Danian faunal succession: planktonic foraminiferal response to a changing marine environment, Geology, 15, 665-668, 1987.

Goodney, D. E., S. V. Margolis, W. C. Dudley, P. Kroopnick, and D. F. Williams, Oxygen and carbon isotopes of recent calcareous nannofossils as paleoceanographic indicators, Paleoceanography, 5, 31-42, 1980.

Graham, D. W., B. H. Corliss, M. L. Bender, and L. D. Keigwin, Jr, Carbon and oxygen isotopic disequilibrium of recent deep-sea benthic foraminifera, Mar. Micropaleontol., 6, 483-497, 1981.

Graham, D. W., M. L. Bender, D. F. Williams and L. D. Keigwin, Strontium-calcium ratios in Cenozoic planktonic foraminifera, Geochim. Cosmochim. Acta, 46, 1281-1292, 1982.

Grossman, E. L., Stable isotopes in modern benthic foraminifera: A study of vital effects, J. Foraminiferal Res., 17, 48-61, 1987.

Hart, M. B., A water-depth model for the evolution of the planktonic Foraminiferida, Nature, 286, 252-254, 1980.

Hemleben, C., M. Spindler, I. Breitinger, and W. G. Deuser, Field and laboratory studies on the ontogeny and ecology of some globorotaliid species from the Sargasso Sea off Bermuda, J. Foraminiferal Res., 15 (4), 254-272, 1985.

Hemleben, C., M. Spindler and O. R. Anderson, Modern planktonic foraminifera, 363 pp., Springer-Verlag, New York, 1989.

Herbert, T. D., and S. L. D'Hondt, Precessional climate cyclicity in Late Cretaceous-Early Tertiary marine sediments: A high resolution chronometer of Cretaceous-

Tertiary boundary events, Earth Planet. Sci. Lett., 99 (30), 263-275, 1990.

Herman, Y., Selective extinction of marine plankton in the Paratethys at the end of the Mesozoic Era: a multiple interaction hypothesis, in Global Catastrophes in Earth History, edited by V.L. Sharpton and P. D. Ward, Geol. Soc. Am. Spec. Pap., 247, 531-540, 1990.

Hess, J., M. L. Bender, and J.-G. Schilling, Evolution of the ratio of strontium-87 to strontium-

86 in seawater from Cretaceous to present, Science, 231, 979-984, 1986.

Hsti, K. J., J. A. McKenzie, and Q. X. He, Terminal Cretaceous environmental and evolutionary changes, in Geological Implications of Impacts of Large Asteroids and Comets on the Earth, Geol. Soc. Am. Spec. Pap., 190, 317-328, 1982.

Hsti, K. J., and J. McKenzie, A "Strangelove" ocean in the earliest Tertiary, in The carbon cycle and atmospheric C02: natural variations Archean to Present, edited by W. S. Broecker and E. T. Sundquist, Geophys. Monogr. Ser., vol. 32, AGU, Washington, D.C., 487-492, 1985.

Kahn, M. I., Non-equilibrium oxygen and carbon isotopic fractionation in tests of living planktonic foraminifera, Oceanol. Acta, 2, 195-208, 1979.

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

Keller, G., Depth stratification of planktonic foraminifers in the Miocene ocean, in The Miocene

ocean: paleoceanography and biogeography, edited by J.P. Kennett, Geol. Soc. Am. Mem., 163, 177-195, 1985.

Keller, G., extinction, survivorship, and evolution across the Cretaceous/Tertiary boundary at E1 Kef, Mar. Micropaleontol., 13, 239-263, 1988.

Keller, G., Extended Cretaceous/Tertiary boundary extinctions and delayed population change in planktonic foraminifera from Brazos River, Texas, Paleoceanography, 4 (3), 287- 332, 1989.

Keller, G. and M. Lindinger, Stable isotope, TOC and CaCO3 record across the Cretaceous-Tertiary boundary at E1 Kef, Tunisia: Palaeogeogr. Palaeoclimatol. Palaeoecol., 73, 243- 265, 1989.

Killingley, J. S., Effects of diagenetic recrystallization on 180/16 0 values of deep sea sediments, Nature, 301, 594- 596, 1983.

Kump, L. R., Interpreting carbon-isotope excursions: Strangelove oceans, Geology, 19, 299-302, 1991.

Leckie, R. M., Paleoecology of mid-Cretaceous planktonic foraminifera: A comparison of open ocean and epicontinental sea assemblages, Micropaleontology, 33 (2), 164-176, 1987.

Leckie, R. M., A paleoceanographic model for the early evolutionary history of planktonic foraminifera, Palaeogeogr. Palaeoclimatol. Palaeoecol., 73, 107-138, 1989.

Lidz, B., A. Kehm, and H. Miller, Depth habitats of pelagic foraminifera during the Pleistocene, Nature, 217, 245-247, 1968.

Li, Q., and S.S. Rad•ford, Evolution and biogeography of Paleogene microperforate planktonic foraminifera, Palaeogeogr. Palaeoclimatol. Palaeoecol., 83, 87-115,

1991.

Liu, C., and R. K. Olsson, Evolutionary radiation of microperforate planktonic foraminifera following the K/T

546 D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera

mass extinction event, J. Foraminiferal Res., 22 (4), 328-346, 1992.

Martin, E. E., and J. D. MacDougall, Seawater Sr isotopes at the Cretaceous/Tertiary boundary, Earth Planet. $ci. Lett., 104, 166-180, 1991.

Matter, A., R. G. Douglas, and K. Perch Nielsen, Fossil preservation, geochemistry and diagenesis of pelagic carbonates from Shatsky Rise, N.W. Pacific, Init. Rep. Deep Sea Drill. Program, 32, 891-907, 1975.

McConnaughy, T., 13 C and 18 0 isotopic disequilibrium in biological carbonates, I, Patterns, Geochim. Cosmochim. Acta, 53, 151-162, 1989a.

McConnaughy, T., 13C and 18 0 isotopic disequilibrium in biological carbonates: II, In vitro simulation of kinetic isotopic effects, Geochim. Cosmochim. Acta, 53, 163-171, 1989b.

Oberhansli, H., Latest Cretaceous--Early Neogene oxygen and carbon isotopic record at DSDP sites in the Indian Ocean, Mar. Micropaleontol., 10, 91-115, 1986.

Olsson, R. K., Planktonic foraminifera from base of Tertiary, Millers Ferry, Alabama, J. Paleontol., 44, 598-604, 1970.

Olsson, R. K., Cenozoic planktonic foraminifera: A paleobiogeographic summary, in Foraminifera Notes for a Short Course, vol. 6, edited by T. W. Broadhead, (organized by M. A. Buzas and B. K. Sen Gupta), University of Tennessee Department of Geological Sciences Studies, Knoxville, 1982.

Olsson, R. K., C. Hemleben, W. A. Berggren, and C. Liu, Wall texture classification of planktonic foraminifera genera in the Lower Danian, J. Foraminiferal Res., 22 (3), 195- 213, 1992.

Oppo, D. W., and R. G. Fairbanks, Carbon isotope composition of tropical surface water during the past 22,000 years, Paleoceanography, 4, 333-351, 1989.

Paull, C. K., and H. R. Thierstein, Stable isotopic fractionation among particles in Quaternary coccolith-sized deep sea sediments, Paleoceanography, 2 (4), 423-430, 1987.

Perch-Nielsen, K., Calcareous nannofossils from the Cretaceous between the North Sea and the

Mediterranean, Int. Union Geol. Sci. Series A, 6, 223-272, 1979.

Perch-Nielsen, K., J. A. McKenzie, and Q. X. He, Biostratigraphy and isotope stratigraphy and the 'catastropic' extinction of calcareous nannoplankton at the Cretaceous/Tertiary boundary, in Geological Implications of Impacts of Large Asteroids and Comets on the Earth, Geol. Soc. Am. Spec. Pap., 190, 353-372, 1982.

Premoli Silva, I., The earliest Tertiary Globigerina eugubina Zone: Paleontological significance and geographical distribution, Mem. 2nd Congr. Latinoam. Geol., 3, spec. publ. 7, 1541-1555, 1977.

Poore, R. Z., and R. K. Matthews, Oxygen isotope ranking of Late Eocene and Oligocene planktonic foraminifers: implications for Oligocene sea-surface temperatures and global ice-volume, Mar. Micropaleontol., 9, 111-134, 1984.

Pospichal, J. J., Calcareous nannofossils across Cretaceous/Tertiary boundary at Site 752, eastern Indian Ocean, Proc. Ocean Drill. Program, Sci. Results, 121, 395-413, 1991.

Pospichal, J. J. and S. W. Wise, Jr., Maestrichtian calcareous nanofossil biostratigraphy of Maud Rise ODP Leg 113 Sites 689 and 690, Weddell Sea, Proc. Ocean Drill. Proj. Sci. Results, 113, 465-487, 1990.

Ravelo, C., Ph.D. thesis, Columbia Univer., Palisades, N.Y., 1991.

Richter, F. M., and D. J. DePaolo, Diagenesis and Sr isotopic evolution of seawater using data from DSDP 590B and 575, Earth Planet. Sci. Lett., 90, 382-394, 1988.

Savin, S.S., and R. G. Douglas, Stable isotope and magnesium geochemistry of Recent planktonic foraminifera from the South Pacific, Geol. Soc. Am. Bull., 84, 2327-2342, 1973.

Schweitzer, P. N., and G. P. Lohmann, Ontogeny and habitat of modem menardiiform planktonic foraminifera, J. Foraminiferal Res., 21 (4), 332-346, 1991.

Shackleton, N.J., and E. Vincent, Oxygen and carbon isotope studies in Recent foraminifera from the southwest Indian Ocean, Mar. Micropaleontol., 3, 1-13, 1978.

Shackleton, N.J., M. A. Hall, and A. Boersma, Oxygen and carbon isotope data from Leg 74 foraminifers, Init. Rep. Deep Sea Drill. Proj., 74, 599-612, 1984a.

Shackleton, N.J., and members of the shipboard scientific party, Accumulation rates in Leg 74 sediments, Init. Rep. Deep Sea Drill. Proj., 74, 621-644, 1984b.

Shackleton, N.J., R. M. Corfield and M. A. Hall, Stable isotope data and the ontogeny of Paleocene planktonic foraminifera, J. Foraminiferal Res., 15(4), 321-336, 1985.

Smit, J., Extinction and evolution of planktonic foraminifera at the Cretaceous/Tertiary boundary after a major impact, in Geological Implications of Impacts of Large Asteroids and Comets on the Earth, Geol. Soc. Am. Spec. Pap., 190, 329-352, 1982.

Smit, J. and Th. M. G. van Kempen, Planktonic foraminifers from the Cretaceous/Tertiary boundary at DSDP Sea Drill. Proj. Site 605, North Atlantic, Init. Rep. Deep Sea Drill. Proj., 93, 549-553, 1987.

Smith, C. C., and E. A. Pessagno, Planktonic foraminifera and stratigraphy of the Corsicana Formation (Maestrichtian) North-central Texas, Cushman Found. Spec. Publ., 12, 68 pp., 1973.

Spero, H. J., Do planktic foraminifera accurately record shifts in the carbon isotopic composition of seawater •;CO27, Mar. Micropaleontol., 19 (4), 275-285, 1992.

Spero, H. J., and M. J. DeNiro, The influence of symbiont photosynthesis on the/5180 and/513C values of planktonic foraminiferal shell calcite, Symbiosis, 4, 213- 228, 1987.

Spero, H. J., and D. F. Williams, Extracting environmental information from planktonic foraminiferal •513C data, Nature, 335, 717-719, 1988.

Spero, H. J., I. Lerche, and D. F. Williams, Opening the carbon isotope "vital effect" black box, 2, quantitative model for interpreting foraminiferal carbon isotope data, Paleoceanography, 6, 639-655, 1991.

Stott, L. D., and J.P. Kennett, The paleoceanographic and paleoclimatic signature of the Cretaceous/Tertiary boundary in the Antarctic: Stable isotopic results from ODP Leg 113, Proc. Ocean Drill. Proj., Sci. Results, 113, 829-848, 1990a.

Stott, L. D., and J.P. Kennett, New constraints on early Tertiary palaeoproductivity from carbon isotopes in foraminifera, Nature, 342 (6249), 526-529, 1990b.

Thierstein, H. R., and H. Okada, The Cretaceous/Tertiary boundary event in the North Atlantic, Init. Rep. Deep Sea Drill. Proj., 43, 601-616, 1979.

Turner, J. V., Kinetic fractionation of carbon-13 during c alcium carbonate precipitation, Geochim. Cosmochim. Acta, 46, 1183-1191, 1982.

D'Hondt and Zachos: Isotopic Variation and Planktonic Foraminifera 547

Vergnaud-Grazzini, C., Non-equilibrium composition of shells of planktonic foraminifera in the Mediterranean Sea, Palaeogeogr. Palaeoclimatol. Palaeoecol., 20, 263-276, 1976.

Vincent, E., J. S. Killingly, and W. H. Berger, Stable isotope composition of benthic foraminifera from the equatorial Pacific, Nature, 289, 639-642, 1981.

Wefer, G., R. B. Dumbar, and E. Suess, Stable isotopes of foraminifers of Peru recording high fertility and changes in upwelling history, in Coastal Upwelling, Its Sediment Record edited by J. Thiede and J. Suess, NATO Conf. Ser. Mar. Sci., IV, 295-311, 1983.

Williams, D. F., M. A. Sommer II, and M. L. Bender, Carbon isotopic compositions of Recent planktonic foraminifera of the Indian Ocean, Earth Planet. Sci. Lett., 36, 391-403, 1977.

Williams, D. F, A. W. H. B6, and R. G. Fairbanks, seasonal stable isotopic variations in living planktonic foraminfera from Bermuda plankton tows, Palaeogeogr. Palaeoclimatol. Palaeoecol., 33, 71-102, 1981.

Williams, D. F., R. C. Thunell, D. A. Hodell and C. Vergnaud-Grazzini, Synthesis of Late Cretaceous, Tertiary, and Quaternary stable isotope records of the South Atlantic based on Leg 72 DSDP core material, in South Atlantic Paleoceanography, edited by K. J. Hsfi and H. J. Weissert, Cambridge University Press, Cambridge, England, 205-242, 1985.

Wu, G., and W. H. Berger, Planktonic foraminifera: Differential dissolution and the Quaternary stable isotope record in the west equatorial Pacific, Paleoceanography, 4, 181-198, 1989.

Zachos, J. C., and M. A. Arthur, Paleoceanography of the Cretaceous-Tertiary boundary event: Inferences from stable

isotopic and other data, Paleoceanography, 1 (1), 5-26, 1986.

Zachos, J. C., M. A. Arthur, R. C. Thunell, D. F. Williams, and E. J. Tappa, Stable isotope and trace- element geochemistry of carbonate sediments across the Cretaceous/Tertiary boundary at Deep Sea Drilling Projects hole 577, leg 86, Init. Rep. Deep Sea Drill. Proj., 86, 513- 532, 1985.

Zachos, J. C., M. A. Arthur, and W. E. Dean, Geochemical evidence for suppression of pelagic marine productivity at the Cretaceous/Tertiary boundary, Nature, 337 (5), 61-64, 1989a.

Zachos, J. C., M. A. Arthur, and W. E. Dean, Geochemical and paleoenvironmental variations across the Cretaceous/Tertiary boundary at Braggs, Alabama, Palaeogeogr. Palaeoclimatol. Palaeoecol., 69, 245-266, 1989b.

Zachos, J. C, M.-P. Aubry, W. A. Berggren, T. Ehrendorfer, and F. Heider, Magnetobiochemostratigraphy across the Cretaceous/Paleogene boundary at ODP Site 750A, Southern Kerguelen Plateau, Proc. Ocean Drill. Proj. Sci. Results, 120, part 2, 961-977, 1992.

S. D'Hondt, Graduate School of Oceanography, University of Rhode Island, Narragansett, Rhode Island, 02882.

J. C. Zachos, Department of Geological Sciences, University of Michigan, Ann Arbor, MI 48109.

(Received April 14, 1992; revised March 31, 1993;

accepted April 16, 1993.)