Deep water in the late Maastrichtian ocean



[1] Stable isotopic data from benthic foraminifera indicate the occurrence of at least three deepwater masses in the late Maastrichtian ocean. Given mean oceanic δ18Ow of −1.0‰, the temperature of the coolest intermediate-depth waters was 5°–7°C, that of the deepest waters was 10°C, and that of the warmest intermediate waters was 13°–15°C. The cool intermediate-depth water mass probably originated in the high-latitude Southern Ocean. The deepest waters originated at least partly in the northern Atlantic. The source region for the warmest intermediate-depth water mass is unknown. Although much of the late Maastrichtian deep water was probably preconditioned for winter sinking by low- or middle-latitude evaporation, no more than ∼11% of late Maastrichtian deep water could have been directly actuated by low-latitude sea surface evaporation. At least in the southern Atlantic and Indian Oceans, heat transport by upwelling of deep water was not the primary cause of mild sea surface and coastal temperatures.

1. Introduction

[2] Deep water in the modern ocean is formed by the dual processes of cooling and salinization. At present, cooling of high-latitude surface waters is the dominant process in deep water formation [Sverdrup et al., 1942]. The deepest water (Antarctic Bottom Water (AABW)) is formed at the surface of the high-latitude Southern Ocean, and much intermediate-depth water (North Atlantic Deep Water (NADW)) forms at high latitudes in the North Atlantic. Nonetheless, salinization remains an important secondary process. Evaporation preconditions eastern Mediterranean surface water for winter sinking and outflow as Mediterranean intermediate water, low-latitude evaporation preconditions Gulf Stream waters for sinking in the North Atlantic Norwegian Sea, and exclusion of brine from shelf ice contributes to the density of high-latitude surface waters [Chamberlin, 1906; Sverdrup et al., 1942].

[3] Chamberlin [1906] reviewed the processes that presently contribute to deep water formation, noted the occurrence of subtropical or warm-temperate, high-latitude climates during much of the geologic past, and proposed that during those periods, deep-sea circulation may have been reversed: “In these periods of warm polar temperatures there is reason to believe that the polar-temperature effects fell below the low-latitude concentration effects, and that therefore the deep oceanic circulation was actuated by the dense waters of the evaporating tracts.” He further suggested that low-latitude sinking of deep waters was balanced by upwelling at high-latitudes. In closing his discourse, Chamberlin proposed that “aided by the enshrouding mantle of vapors that must have arisen from such a body of [upwelled] water, it is conceived the mild temperatures requisite for the maintenance of the recorded life through the polar night may have been thus maintained.” The seductiveness of this briefly articulated model is difficult to overstate. In a few short paragraphs, Chamberlin posed an elegant paradigm for deep oceanic circulation of warm-climate intervals and effectively proposed that the hypothetical deep circulation would have sustained the high-latitude warmth that defines those climate intervals.

[4] The hypothesis of low-latitude evaporative deepwater formation was resurrected by Berger[1979], Arthur and Natland[1979], and Brass et al. [1982]. It has since proven a popular paleoceanographic paradigm [Hay, 1988]. Saltzman and Barron [1982]interpreted the oxygen isotopic (δ18Oc) signature of inoceramid shells to indicate the Late Cretaceous occurrence of low-latitude deepwater formation, at least in the South Atlantic Basin. Prentice and Matthews [1988]interpreted δ18Oc differences between the tests of tropical planktonic foraminifera and those of benthic foraminifera to indicate the dominance of low-latitude deep water formation throughout most of the Cretaceous and Paleogene. Stott and Kennett [1990] interpreted the lack of a K/T boundary shift in benthic δ13Cc values of Maud Rise (Antarctica) to indicate that the Southern Ocean was not a significant source of early Paleocene deep waters. Kennett and Stott [1990] examined the δ18Ocsignals of planktonic and benthic foraminifera from the same Maud Rise sites. On the basis of those signals, they inferred that warm saline deep waters from low latitudes underlay cooler surface waters at southern high latitudes throughout most of the Paleogene. Although popular, such interpretations are not universally accepted. For example, E. Barrera and coworkers interpreted the δ18Oc signatures of planktonic and benthic organisms at southern high-latitude sites to indicate that the Southern Ocean was a locus of deep water formation through much of the Late Cretaceous [Barrera et al., 1987; Barrera and Huber, 1990]. MacLeod and Huber[1996] built on the latter interpretation by proposing that low-latitude sources of deep water were supplanted by high-latitude sources in the middle Maastrichtian. In contrast, Frank and Arthur [1999] inferred that deep water formed at high latitudes throughout the entire Maastrichtian. Corfield and Norris [1996] interpreted the δ13Ccvalues of benthic foraminifera to indicate that deep waters formed in the subtropical North Atlantic throughout the latest Cretaceous and Paleocene, much as North Atlantic Deep Water does today.

[5] In order to better assess the relative importance of low-latitude and high-latitude deepwater formation in late Maastrichtian (latest Cretaceous) oceans we determined the δ18Ocof well-preserved Maastrichtian benthic foraminifera (Gavelinellaand Nuttallides) from 16 Deep Sea Drilling Project (DSDP) and Ocean Drilling Project (ODP) sites of the Atlantic, Indo-Pacific, and Southern Oceans (Figure 1). The sites ranged in paleolatitude from 36°N to 70°S and in paleodepth from 1 to 3.5 km. Most late Maastrichtian deep water would have occurred in this depth range (∼80% of the modern ocean occurs at <3.5 km water depth) [Pilson, 1998]. As discussed below, we have assessed the possible effects of mean oceanic paleosalinity, mean oceanic δ18O(δ18Ow), and water mass evaporative histories on deepwater paleotemperature and paleosalinity estimates.

Figure 1.

Deep Sea Drilling Project sites and Ocean Drilling Project sites sampled for analysis of Maastrichtian deep-water properties. Base map is a 66.0-Ma reconstruction from the Ocean Drilling Stratigraphic Network (ODSN) Plate Tectonic Reconstruction Service (

2. Materials and Methods

[6] In order to disaggregate and clean the foraminifera, bulk sediment samples were processed as described by D'Hondt and Arthur [1995]. Clean specimens of benthic Gavelinellaand Nuttallides were identified and handpicked using a reflecting light microscope. The foraminiferal samples were isotopically processed and analyzed using the instruments and procedures described by D'Hondt and Arthur [1996]. Five to 15 foraminifera were analyzed for each measurement. One to 10 samples were analyzed from each site (88 samples total). The δ18Ocand δ13Cc compositions of foramineral tests are reported in per mil (‰) notation with respect to the Vienna Peedee belemnite (VPDB) standard (Table 1) using NBS-19 as a primary reference (δ18OPDB= −2.20 and δ13CPDB = −1.96). On the basis of replicate samples, analytic precision is ±0.08‰ (1σ) for δ18Oc and ±0.04‰ (1σ) for δ13Cc.

Table 1. Oxygen and Carbon Isotopic Values of Benthic Gavelinellaand Nuttalides From the Upper Maastrichtian (Upper C30N) Interval of Various DSDP and ODP Sites
SiteCore-Section (cm Interval)TaxonEstimated Paleolatitude, degEstimated Paleodepth, mbsfδ13C, ‰δ18C, ‰
20C6-5 (66–67)N. truempyi3118871.72−0.03
20C6-5 (66–67)G. beccariformis3118871.680.26
20C6-5 (83–84)N. truempyi3118871.73−0.05
20C6-5 (83–84)N. truempyi3118871.510.10
20C6-5 (83–84)G. beccariformis3118871.350.21
20C6-5 (98–99)N. truempyi3118871.520.06
20C6-5 (98–99)G. beccariformis3118871.530.15
30516-1 (100–102)N. truempyi826091.320.45
30516-3 (100–102)N. truempyi826091.040.44
30516-4 (100–102)N. truempyi826091.250.46
35731-2 (50–51)N. truempyi3311280.890.12
35731-2 (50–51)G. beccariformis3311280.79−0.07
35731-2 (75–76)N. truempyi3311281.190.05
35731-2 (75–76)G. beccariformis3311281.18−0.11
35731-3 (25–26)N. truempyi3311280.970.22
35731-3 (25–26)G. beccariformis3311280.99−0.04
36318-5 (100–102)N. truempyi3017751.190.41
36318-5 (100–102)G. beccariformis3017751.300.32
36319-2 (100–101)N. truempyi3017751.140.49
36319-2 (100–102)G. beccariformis3017751.360.51
36319-4 (99–100)N. truempyi3017751.450.61
36319-4 (99–100)G. beccariformis3017751.410.52
38413-5 (48–49)G. beccariformis3634021.680.76
38413-5 (104–105)G. beccariformis3634022.040.64
38413-6 (109–110)G. beccariformis3634022.080.32
39013-1 (99–101)N. truempyi2625731.740.67
465A3-5 (28–30)N. truempyi1014161.080.06
465A3-5 (28–30)G. beccariformis1014162.25−0.28
4653-5 (28–30) G. beccariformis1014161.240.25
465A3-5 (58–60)N. truempyi1014161.230.08
465A3-5 (58–60)G. beccariformis1014161.300.11
465A3-5 (88–90)N. truempyi1014161.23−0.07
465A3-5 (88–90)G. beccariformis1014161.24−0.03
525A41-1 (97–98)N. truempyi3910631.300.20
525A41-1 (97–98)G. beccariformis3910631.340.35
525A41-3 (78–79)G. beccariformis3910631.540.37
525A41-4 (143–144)N. truempyi3910631.340.25
525A41-4 (143–144)G. beccariformis3910631.560.33
52733-3 (43–44)N. truempyi3829011.240.37
52733-3 (43–44)G. beccariformis3829011.460.53
52733-3 (103–104)N. truempyi3829011.300.60
52733-3 (103–104)G. beccariformis3829011.460.69
52733-4 (33–34)N. truempyi3829010.860.33
52733-4 (33–34)G. beccariformis3829010.960.47
52832-4 (100–102)N. truempyi3823121.630.09
52833-1 (98–101)N. truempyi3823121.280.22
52833-1 (98–100)G. beccariformis3823121.410.40
52833-2 (100–102)N. truempyi3823121.030.27
52833-3 (100–102)N. truempyi3823121.560.38
577A13-2 (109–111)cf. Nuttalides823511.011.27
577A13-3 (44–46)cf. Nuttalides823510.961.08
577A13-3 (52–54)cf. Nuttalides823510.761.32
577A13-3 (136–138)cf. Nuttalides823510.891.13
577A13-4 (45–47)cf. Nuttalides823510.901.12
661A13H-4 (146–148)N. truempyi332430.610.42
689B25X-6 (54–55)G. beccariformis7011151.260.94
689B25X-6 (97–98)N. truempyi7011151.220.80
689B25X-6 (97–98)G. beccariformis7011151.351.02
689B26X-1 (102–103)N. truempyi7011151.331.22
689B26X-1 (102–103)G. beccariformis7011151.431.04
690C16X-1 (58–59)N. truempyi7021531.510.96
690C16X-1 (58–59)G. beccariformis7021531.240.92
690C16X-1 (99–100)N. truempyi7021531.290.61
690C16X-1 (99–100)G. beccariformis7021531.250.82
690C16X-2 (100–102)N. truempyi7021531.610.71
761B22X-3 (99–101)N. truempyi4224831.060.93
761B22X-3 (99–101)G. beccariformis4224831.150.91
761B22X-4 (53–55)N. truempyi4224831.090.85
761B22X-4 (53–55)G. beccariformis4224831.210.61
761B22X-4 (99–101)N. truempyi4224831.120.85
761B22X-4 (99–101)G. beccariformis4224831.200.77
761B22X-5 (53–55)N. truempyi4224830.870.78
761B22X-5 (53–55)G. beccariformis4224831.330.71
761B22X-5 (102–104)N. truempyi4224831.120.93
761B22X-5 (102–104)G. beccariformis4224831.520.78
766A10R-2 (111–113)N. truempyi4533711.380.87
766A10R-2 (111–113)G. beccariformis4533711.420.38
766A10R-2 (131–133)N. truempyi4533711.530.65
766A10R-2 (131–133)G. beccariformis4533711.560.42
766A10R-3 (11–13)N. truempyi4533711.420.59
766A10R-3 (11–13)G. beccariformis4533711.450.35

[7] Isotopic paleotemperatures were calculated using a modified form of the δ18O equation of Erez and Luz [1983]:

equation image

where δ18Oc is the oxygen isotopic composition of foraminiferal calcite, δ18Omwis the oxygen isotopic composition of mean deep water, Δδ18Olw= [(SmSl)(Δδ18OwSw)] is an adjustment for the surface evaporation and precipitation history of the local water mass (the water mass in contact with the foraminiferal calcite at the time of calcification), Sm is the salinity of mean deep water (‰ SMOW, unit of measure omitted[United Nations Educational, Scientific, and Cultural Organization (UNESCO), 1981a]), Sl is the salinity (psu) of the local water mass, and Δδ18Ow/ΔSwis the estimated ratio of changes in δ18Owand S that resulted from the surface evaporation and precipitation history of the local water mass. Oxygen isotopic estimates of the ocean and atmosphere are reported in per mil notation with respect to the SMOW standard. Because the stable isotopic signals of Gavelinellaand Nuttallides appear close to equilibrium values [Shackleton et al., 1984], we did not correct our isotopic results for taxon-specific vital affects.

[8] Paleodensity estimates were calculated with the One-Atmosphere International Equation of State of Seawater (Millero and Poisson [1981], as corrected by UNESCO [1981b]

equation image

where ρ0 is the density of pure water (kg m−3), S is salinity (psu), and tis temperature (°C). The density of pure water is calculated with the following equation (Millero and Poisson [1981], as corrected by UNESCO [1981b]):

equation image

For water masses of different depths, these equations provide potential density estimates, as the effect of increasing pressure with depth is not considered.

[9] For sites assumed to overlie oceanic crust, paleodepth estimates are based on the subsidence curves and isostatic adjustment procedures of Sclater et al. [1985], seafloor age estimates of R. D. Müller et al. (Digital Isochrons of the World's Ocean Floor, 1992, available on the World Wide Web at, and sedimentary data from the DSDP and ODP Initial Reports [Maxwell et al., 1970; Larson et al., 1975; Supko et al., 1977; Benson et al., 1978; Bolli et al., 1978; Tucholke et al., 1979; Thiede et al., 1981; Moore et al., 1984; Heath et al., 1985; Ruddiman et al., 1988, 1989; Barker et al., 1988, 1990; Haq et al., 1990; Gradstein et al., 1990, 1992; von Rad et al., 1992]. The paleodepth estimate for the one site assumed to overly continental crust (ODP Site 761B) is based on its present-day water depth and an isostatic adjustment for sedimentary overburden. All paleodepth estimates assume that late Maastrichtian sea level was 200 m above that of the present-day level [Haq et al., 1987]. Paleolocation estimates are based on visual backtracking of seafloor isochrons from R. D. Müller et al. (Digital Isochrons of the World's Ocean Floor, 1992, available on the World Wide Web at

2.1. Stratigraphic Control

[10] The sampled sequences lie within the uppermost C30N paleomagnetic interval. This interval occurs within the upper Abathomphalus mayaroensisplanktonic foraminiferal zone and the lower portion of the Micula murus nannofossil zone (data from DSDP and ODP Initial Reports). Because magnetostratigraphy and nannofossil stratigraphy are not available at all 16 sites, relative ages on submillion-year timescales were estimated by assuming that (1) the M. murus first occurrence (FO) preceded the 30N/29R paleomagnetic reversal by 400–600 kyr [Herbert and D'Hondt, 1990] and (2) the Cretaceous/Tertiary (K/T) boundary followed the 30N/29R magnetic reversal by 310–350 kyr [D'Hondt et al., 1996]. The duration of the interval from the M. murusFO to the 30N/29R reversal was derived from the Milankovitch chronostratigraphy of Herbert and D'Hondt [1990], and the duration of the interval from the 30N/29R reversal to the K/T boundary was derived from the Milankovitch chronostratigraphy of D'Hondt et al. [1996]. Given these estimated durations, sample ages can be estimated for the 16 sites by (1) assuming constant accumulation rates between the M. murus FO and the K/T boundary and (2) extrapolating via any two of the three stratigraphic data. By these criteria, all of the sampled sequences are age equivalent on the scale of less than two hundred thousand years.

2.2. Foraminiferal Preservation

[11] Stable isotopic ratios of foraminiferal calcite are typically changed by carbonate diagenesis (i.e., recrystallization or authigenic growth of carbonate). In order to qualitatively assess the degree of diagenetic alteration of our benthic foraminiferal samples, representative specimens were examined by scanning electron microscope for evidence of recrystallization and carbonate overgrowth. All samples used in this study were composed of free tests with little or no subscrilling of chambers. At 3000 times magnification, tests from different sites exhibit slightly different degrees of visible alteration. The degree of alteration appears to largely depend on depth of burial. The foraminifera from shallowly buried sequences generally exhibit primary surface texture; fine pores are open, and bilamellar wall structures remain visible (Figure 2). However, foraminifera from the most deeply buried sequences show clear evidence of some postdepositional crystallization. For example, the interior surface of a broken test from DSDP Site 357 contains small but abundant blocky crystalline overgrowths (Figure 2).

Figure 2.

Scanning electron photomicrographs of representative benthic foraminifera. Photomicrographs in left column are from DSDP Site 384 (core-section 13–5, 147–148 cm): (top) Gavelinella specimen; (middle) close view of Nuttalides specimen; (bottom) the same Nuttalides specimen. Photomicrographs in center column are from ODP Site 689B (core-section 26X–1, 47–48 cm): (top) Gavelinella specimen; (middle) close view of the Gavelinella specimen; (bottom) Nuttalidesspecimen. Photomicrographs in right column are from DSDP Site 357 (core-section 31-2, 124–125 cm): (top) Gavelinella specimen; (middle) close view of the Gavelinella specimen; (bottom) Nuttalides specimen. Scale bars for the top and bottom rows are 100 microns. Scale bar for the middle row is 10 microns.

[12] The effect of diagenesis on oxygen isotopic values should depend on the proportion of calcite that resulted from postdepositional (secondary) crystallization, the temperature of secondary crystallization, and the δ18O value of the pore water where secondary crystallization occurred [e.g., Killingley, 1983; Schrag et al., 1995]. Because temperature of secondary crystallization increases with increasing sediment depth [Lawrence, 1989; Louden, 1989] and pore water δ18O values decrease with sediment thickness [Lawrence and Geiskes, 1981; Lawrence, 1989], the δ18O of benthic foraminiferal tests that have been significantly altered by postdepositional crystallization should be negatively correlated to burial depth.

[13] Such diagenetic effects do not appear to have greatly altered the δ18O values included in this study. For example, some of the benthic foraminifera from intermediate paleo–water depths and recent burial depths <200 m below seafloor exhibit mean δ18O values as negative as those of more deeply buried foraminifera from similar paleo–water depths (Figure 3). Furthermore, the foraminifera of the deepest paleo–water depths exhibit similar δ18O values, regardless of burial depth (Figure 3). These comparisons suggest that the δ18O values described by this study generally have not been greatly altered by secondary crystallization.

Figure 3.

Mean δ18Ocvalues of late Maastrichtian (late C30N) epibenthic foraminifera from 16 DSDP and ODP sites (Table 1). Solid circles mark values from burial depths of <200 m. Open circles mark values from burial depths of >200 m.

3. Benthic Isotopic Signals (the Minimum Number and Sources of Late Maastrichtian Deep Water Masses)

[14] As Figure 3 illustrates, the mean δ18Ocof benthic foraminifera varies from site to site; the minimum mean value occurs at DSDP Site 357 (0.03‰), and the maximum value occurs at DSDP Site 577A (1.18‰). At shallower paleodepths (1–2.5 km), oxygen isotopic values varied by almost 1.2‰ between sites (n= 10 sites) (Figures 3 and 4a). At deeper paleodepths (2.5–3.5 km), oxygen isotopic values were much more stable (ranging from 0.42 to 0.67‰) (n = 6 sites) (Figure 3 and 4b).

Figure 4.

(a) Mean δ18Ocvalues of epibenthic foraminifera from sites with paleodepths of 1.0–2.5 km below seafloor. (b) Mean δ18Ocvalues of epibenthic foraminifera from sites with paleodepths of 2.5–3.5 km. Base map and DSDP/ODP site locations are as in Figure 1.

[15] Southern Ocean sites and near-equatorial Pacific Site 577A exhibit the most positive δ18Ocvalues (0.8–1.2‰) (Figure 4a). Isotopic data suggest that winter cooling of surface waters in the Southern Ocean was the likeliest source of the intermediate-depth paleowaters that bathed these sites. Although the mean δ18Ocof epibenthic foraminifera from Site 577A is slightly more positive than that of foraminifera from the Southern Ocean sites, the most positive values from Southern Ocean ODP Sites 689B and 690C are essentially the same as those from Site 577A (Table 1). Furthermore, the mean δ13Cc of epibenthic foraminifera from the southernmost sites was significantly more positive than that of near-equatorial Site 577A (Figure 5a). Finally, the δ18Ocsignals of these benthic foraminifera overlap those of coeval 18O-enriched planktonic foraminifera from southern high-latitude sites (i.e., two Heterohelix globulosa samples from Site 690C exhibit δ18Ocvalues of 0.7 and 0.9‰ [D'Hondt and Arthur, 1995, 1996]. The overlap between these planktonic δ18Ocvalues and cooccurring benthic values supports the interpretation that this intermediate-depth water mass formed in the high-latitude Southern Ocean. The slight discrepancy between these planktonic values and the most positive benthic values is not critical to this interpretation. That discrepancy could easily have resulted from a small number of factors, including (1) local formation of intermediate-depth water when cool-season surface waters were slightly cooler than those inhabited by the H. globulosapopulation, (2) inclusion of warmer-season specimens in the planktonic isotopic sample, and (3) formation of intermediate-depth water slightly poleward of Site 690C.

Figure 5.

(a) Mean δ13Cc values of epibenthic foraminifera from sites with paleodepths of 1.0–2.5 km below seafloor. (b) Mean δ13Cc values of epibenthic foraminifera from sites with paleodepths of 2.5–3.5 km. Base map and DSDP/ODP site locations are as in Figure 1.

[16] The most negative δ18Ocvalues (0.0–0.1‰) characterized benthic foraminifera from intermediate paleodepths of the eastern South Atlantic and near-equatorial Pacific DSDP Site 465A (Figure 4a). At present levels of geographic resolution, the paucity of sites with mean values in this range probably precludes definitive identification of the source of the paleowaters that were in contact with these foraminifera.

[17] Oxygen isotopic data allow two interpretations of the deepest late Maastrichtian water mass (Figure 4b): (1) it was a relatively even mixture of the intermediate-depth paleowater masses that were in contact with benthic foraminifera characterized by more positive (∼1.0‰) and negative (∼0.0‰) δ18Ocvalues, or (2) it originated independently of those two paleowater masses and simply has not been sampled at intermediate paleodepths in its source region. At least in the northwestern Atlantic, benthic foraminifera in contact with this deepest paleowater mass were characterized by unusually positive δ13Ccvalues (1.7–1.9‰) (Figure 5b). Such positive δ13Ccvalues were typically not achieved by benthic foraminifera in contact with either of the paleowater masses interpreted from the intermediate-paleodepth δ18Ocdata (Figure 5a). Since the δ13C of total dissolved carbon (TCO2= CO2 + HCO3 + CO3-) decreases with water mass age as a function of organic carbon oxidation at depth, it appears unlikely that the northwestern Atlantic deep water was a mixture of the two intermediate-depth water masses.

[18] In short, these isotopic data collectively suggest that there were at least three deep water masses in the late Maastrichtian ocean: two intermediate-depth water masses characterized by near-equilibrium δ18Ocvalues of ∼1 and 0‰, and a deeper water mass characterized by a mean near-equilibrium δ18Oc value of 0.5‰. The waters in contact with benthic foraminifera characterized by δ18Oc values that approximate 1.0‰ probably originated in the Southern Ocean. The deepest waters, in contact with benthic foraminifera characterized by a δ18Ocvalue of 0.5‰, originated at least partly in the northwestern Atlantic. The source of waters in contact with benthic foraminifera characterized by δ18Ocvalues that approximate 0.0‰ remains unknown. This could have been winter water formed at low middle latitudes in Tethys or other basins.

4. Temperature and Salinity of Late Maastrichtian Deep Waters

4.1. A Mean Paleotemperature Estimate for the Late Maastrichtian Deep Ocean

[19] The mean temperature of the deep ocean can be estimated from the mean δ18Owof seawater and the δ18Oc of calcite in equilibrium with the average deep ocean. The stable isotopic signals of Gavelinella and Nuttallidesappear close to equilibrium values [Shackleton et al., 1984]. Hence, although we do not know the exact δ18Ocof calcite in equilibrium with the mean late Maastrichtian ocean, we can approximate it with the mean δ18Oc of benthic foraminifera from the 16 sites analyzed for this study (0.51‰).

[20] Estimates of the mean δ18Owof Late Cretaceous seawater typically assume that there was no globally significant Late Cretaceous ice volume and that the mean δ18O value of the terrestrial surface hydrosphere has not changed over the last 100 million years [Shackleton and Kennett, 1975]. This assumption results in a mean δ18Owvalue of about −1.0‰ for late Maastrichtian seawater. Given mean δ18Owof −1.0‰ and benthic foraminiferal tests in equilibrium with their paleoenvironments, a mean benthic foraminiferal δ18Ocvalue of 0.51‰ indicates a mean deep-ocean paleotemperature of 10.2°C.

4.2. Paleotemperature and Paleosalinity Estimates for Late Maastrichtian Deep Water at Individual Sites

[21] Extending paleotemperature estimates to individual sites requires another level of assumption. Since different deepwater masses have different evaporative histories, the effects of evaporation and precipitation on the δ18Owvalues and salinities of deep waters are not constant from site to site. For example, the mean δ18Ow value of deepwater masses in the modern open ocean ranges from 0.12‰ (North Atlantic Deep Water) to −0.45‰ (Antarctic Bottom Water) [Craig and Gordon, 1965]. Similarly, mean deepwater salinity in the modern open ocean ranges from 34.6 psu in the North Pacific to 35.0 psu in the North Atlantic [Levitus et al., 1994]. Deep waters entering the open ocean from the Red Sea and the Mediterranean Sea are much more 18O-enriched and much more saline. The Red Sea is characterized by a deepwater δ18Owvalue of 1.95‰ and deepwater salinity of 40.6 psu [Craig, 1966]. The Mediterranean basin contains deep water with δ18Ow values of 1.55–1.9‰ [Thunell et al., 1987] and salinity of ∼38.4–38.7 psu [Sverdrup et al., 1942] (deep Mediterranean δ18Owand salinity decrease to ∼1.45‰ and 37.75 psu, respectively, in the Strait of Gibraltar where Mediterranean deep water mixes with North Atlantic waters during formation of Mediterranean Overflow Water). Hence estimation of paleotemperatures at different sites requires consideration of evaporation-precipitation balances.

[22] To a closer approximation the deep ocean has about the same potential density everywhere. In the modern open ocean, potential density (at 1 atmosphere of pressure) ranges from 1027.7 kg m−3(North Pacific) to 1027.9 kg m−3 (North Atlantic) (based on temperature data from Levitus and Boyer [1994] and salinity data from Levitus et al. [1994]). In semienclosed basins, potential density deviates further from open ocean values. For example, it averages 1028.7 kg m−3 in the Red Sea and reaches an extreme of 1029.1 kg m−3 in the eastern Mediterranean. These interbasinal differences in potential density are dwarfed by the interbasinal differences in deepwater salinity. Hence, given an estimate of the relationship between δ18Oc and paleosalinity (Δδ18OwS), the average paleotemperature and paleosalinity of any individual site can be most closely approximated by calculating the paleosalinity and paleotemperature values at which the mean δ18Ocof Gavelinella and Nuttallidesfrom that site corresponds to the potential paleodensity of mean deep water at 1 atmosphere of pressure.

[23] An estimate of mean paleosalinity is necessary in order to calculate the potential paleodensity of mean Maastrichtian deep water. The mean paleosalinity of the late Maastrichtian ocean can be provisionally estimated by assuming that there was no globally significant late Maastrichtian ice volume and that the mean salinity of the terrestrial surface hydrosphere has not changed over the last 70 million years (similar to the standard δ18Ow assumption). Given a lack of significant ice volume and a constant salt balance in the surface hydrosphere, late Maastrichtian oceans were characterized by a mean paleosalinity of ∼34 psu. If the late Maastrichtian deep ocean was characterized by a mean paleotemperature of 10.2°C and a mean paleosalinity of 34 psu, its mean potential density was 1026.13 kg m−3 (equivalent to a σt of 26.13).

[24] The relationship between δ18Ocand paleosalinity (Δδ18OwS) must be estimated before we can use paleodensity to infer the average paleotemperature and paleosalinity of any individual site. For an isolated deepwater mass this relationship is determined by effects of evaporation and precipitation that occurred during the surface water history of the deepwater mass. Mixing considerations aside, the salinity of a surface marine water mass is simply determined by the mean salinity of seawater and the local balance between evaporation and precipitation. In contrast, the δ18Owvalue of a surface water mass is controlled by the mean δ18Owof seawater, the degree of evaporation from the water mass, the degree of precipitation to the water mass, the δ18Owvalue of the evaporated water, and the δ18Owvalue of the precipitated water. The lighter isotope (16O) is preferentially evaporated and the heavier isotope (18O) preferentially rains out (resulting in atmospheric concentration of 16O by Rayleigh distillation); consequently, the δ18O value of early stage precipitation is relatively close to that of seawater, whereas subsequent precipitation is progressively depleted in 18O [Dansgaard, 1964; Craig and Gordon, 1965]. The relative rain out of 16O and 18O is strongly dependent on temperatures of evaporation and condensation for two reasons. First, the equilibrium ratio of 18O/16O in liquid to 18O/16O in water vapor increases with decreasing temperature [Dansgaard, 1964]. Second, a cold atmosphere holds less water vapor than a warm atmosphere and consequently is generally characterized by greater relative atmospheric “distillation.” These effects lead to strong positive correlations between atmospheric temperature and the δ18O values of water vapor and precipitation. For example, at mean air temperatures approaching 30°C, the δ18O value of precipitation approaches that of mean seawater (0‰), whereas at a mean air temperature of −50°C the δ18O value of precipitation approaches −50‰ [Dansgaard, 1964].

[25] Low-latitude waters of the modern open ocean and marginal seas define a Δδ18Ow/ΔSwslope of ∼0.35 [Railsback et al., 1989]. The slope of Δδ18OwSw is as low as 0.11 in much of the low-latitude open Atlantic [Craig and Gordon, 1965]. At these latitudes the Δδ18OwSw relationship is defined by the low level of isotopic fractionation undergone by the surface ocean during early stage evaporation and precipitation at relatively high temperatures [Craig and Gordon, 1965]. In contrast, midlatitude and high-latitude waters of the modern open ocean define a Δδ18OwSw slope of ∼0.5 [Railsback et al., 1989]. The steepest surface water Δδ18OwSwslope in the modern ocean (0.6) is defined by midlatitude and high-latitude Atlantic surface waters [Craig and Gordon, 1965]. In these regions the slope of Δδ18Ow/ΔS is primarily controlled by the high level of isotopic fractionation undergone by late stage precipitation at low condensation temperatures [Craig and Gordon, 1965]. A world with warmer high-latitude waters (i.e., the late Maastrichtian) would probably have been characterized by high-latitude precipitation with δ18O values more positive than that of the present (due to decreased isotopic fractionation at higher temperatures). Consequently, the steepest Δδ18OwSw slopes of the late Maastrichtian ocean were probably <0.5.

[26] On the assumption that tropical evaporation was a dominant cause of deepwater formation in ancient oceans (Cretaceous and Ordovician), a Δδ18OwSw slope of 0.35 has been used to develop estimates of deepwater paleotemperatures [Railsback et al., 1989; Woo et al., 1992]. In previous studies we assumed the same Δδ18OwSw slope to estimate late Maastrichtian surface water paleotemperatures [D'Hondt and Arthur, 1995, 1996]. Given a Δδ18OwSw slope of 0.35, a late Maastrichtian deepwater potential density of 1026.13 kg m−3 allows deepwater paleotemperature and paleosalinity as low as 6°C and 33.2 psu (DSDP Site 577: mean δ18Oc value of 1.2‰) and as high as 13.7°C and 34.8 psu (DSDP Site 357: mean δ18Ocvalue of 0.0‰) (Figure 6 and Table 2).

Figure 6.

Estimates of mean late Maastrichian paleotemperature and paleosalinity values for the 16 DSDP and ODP sites. These estimates assume a mean deep-ocean density of 1026.13 kg m−3 (Table 2).

Table 2. Estimates of Mean, Minimum (Site 577), and Maximum (Site 357) Deep-Sea Paleotemperatures and Paleosalinities Given Different Assumptions of Mean Oceanic Paleosalinity, Mean Oceanic δ18OWand Δδ18OWS
Mean Oceanic δ18O, ‰Mean Oceanic Paleosalinity, psuΔδ18O/ΔSMean Deep Ocean (16 Sites)aSite 577aSite 357a
Paleotemperature, °CPaleodensity, kg m−3Paleotemperature, °CPaleosalinity, psuPaleotemperature, °CPaleosalinity, psu
  • a

    Mean deep ocean has a benthic δ18O (‰) of 0.51, Site 577 has a benthic δ18O of 1.18, and Site 357 has a benthic δ18O of 0.03.


4.3. Sensitivity of Paleotemperature and Paleosalinity Estimates to Estimates of Mean Paleosalinity, Mean δ18Ow, and Δδ18Ow/ΔS

[27] Section 4.2suggests that estimation of ancient seawater temperatures from δ18Ocvalues is fraught with assumptions. The sensitivity of paleotemperature estimates to most of those assumptions can be readily assessed by calculating paleotemperature estimates under different conditions of mean paleosalinity, mean δ18Ow, and Δδ18OwSw. Such calculations demonstrate that over the range of δ18Ocvalues found in this study, deepwater paleotemperature estimates are insensitive to the assumed mean paleosalinity (given standard estimates of Maastrichtian δ18Ow) (Table 2). In contrast, on the scale of 3°–4°C, deepwater paleotemperature estimates are sensitive to 1.0‰ differences in the estimate of mean oceanic δ18Ow. On the scale of 1°–2°C, intersite differences in paleotemperature estimates are also sensitive to 1.0‰ differences in the estimate of mean oceanic δ18Ow (Table 2).

[28] Changes in the estimates of mean oceanic paleosalinity and mean oceanic δ18Owaffect paleotemperature estimates similarly throughout the range of δ18Ocvalues. For example, with an assumed mean oceanic δ18Owof 0.0‰, all δ18Oc values correspond to higher paleotemperature estimates than with an assumed mean oceanic δ18Owof 1.0‰ (Table 2). In contrast, a change in the assumed slope of Δδ18Ow/ΔSwleaves mean estimates of paleosalinity and paleotemperature unchanged but changes end-member estimates in opposite directions. For example, an assumed Δδ18OwSwslope of 0.5 results in negative δ18Ocvalues corresponding to warmer paleotemperatures and higher paleosalinity values than the standard Δδ18Ow/Δ Sw slope of 0.35. However, the same difference in assumed Δδ18Ow/Δ Swslopes results in positive δ18Oc values corresponding to cooler paleotemperatures and lower paleosalinity values than the standard Δδ18Ow/Δ Sw slope of 0.35. In any case, the dependence of deepwater paleotemperature and paleosalinity estimates on the slope of Δδ18Ow/Δ Sw is relatively low for Δδ18Ow/Δ Sw values in the range of 0.2–0.5 (Table 2).

[29] Barring a relatively large (= 1.0‰) error in the standard assumption of mean oceanic δ18Ow[Shackleton and Kennett, 1975], the preceding calculations suggest that the most positive mean δ18Oc value in our study corresponds to a paleotemperature estimate of 5°–7°C (Site 577A), the most negative mean δ18Oc value corresponds to a paleotemperature estimate of 13°–15°C (Site 357), and the average value of all 16 sites corresponds to a paleotemperature estimate of ∼10°C.

5. Discussion

[30] As previously discussed, the coolest (∼6°C) intermediate-depth water mass originated at least partly in the high-latitude Southern Ocean. The source of the warmer (∼14°C) intermediate-depth water mass remains unknown. Epibenthic δ13Cc signals suggest that the deepest (10°C) water originated at least partly in the northern Atlantic drainage.

[31] Our data do not allow definitive identification of all the regions where late Maastrichtian deep water was formed. In particular, our lack of extratropical Pacific data prevents us from determining the presence or absence of late Maastrichtian deepwater formation at high latitudes in the Pacific Ocean. Similarly, our lack of data from epicontinental seaways prevents us from assessing the contributions of specific epicontinental seas to late Maastrichtian deepwater formation.

[32] Despite these limits on our ability to identify all possible deepwater sources, we can use these data to quantitatively assess the general role of low-latitude evaporation in late Maastrichtian deepwater formation. In addition, the geographic coverage of our data allows us to partially assess the role of deep waters in maintaining surface warmth at southern high latitudes.

5.1. Possible High-Temperature (Low-Latitude?) Sources of Deep Water

[33] It is possible that the 14°C intermediate-depth water mass was a mix of warm low-latitude water and cooler high-latitude waters. It is also possible that the deepest (10°C) water mass resulted in part from mixing of the two intermediate-depth water masses. Given these possibilities, we must consider the possibility that tropical surface waters provided an evaporative source of some late Maastrichtian deep waters. Modeling the deepest (10°C) water mass as a mix of the coolest high-latitude waters and tropical surface waters (the canonical low-latitude deepwater source) provides an upper bound estimate for the fraction of deep water that originated at low latitudes by evaporative processes.

[34] Oxygen isotopic signals of near-surface planktonic foraminifera suggest that the warmest late Maastrichtian surface waters were characterized by a δ18Ocvalue of about −1.4‰ (−1.42‰ at North Atlantic western gyre margin DSDP Site 390A) [D'Hondt and Arthur, 1996]. If this tropical surface water was a source of late Maastrichtian deep water, it must have approached the mean paleodensity of 1026.13 kg m−3. With mean oceanic paleosalinity of 34 psu and Δδ18Ow/Δ Sw slope of 0.35, a δ18Ocvalue of −1.4‰ and a paleodensity estimate of 1026.13 kg m−3correspond to a paleosalinity estimate of 40.6 psu and a paleotemperature estimate of 29.5°C. Given deepwater paleodensity of 1026.13 kg m−3, a mean deepwater paleosalinity of 34 psu, and a minimum deep water paleosalinity of 33.2 psu (Table 2), the hypothetical 40.6 psu deepwater source could have provided only 11% of mean late Maastrichtian deep water.

[35] Several lines of evidence suggest that the above calculation overestimates the possibility of a purely evaporative deepwater source. First, in the modern world, the paleosalinity required for this hypothetical deepwater source is never approached in the open ocean and is only attained in the midlatitude Red Sea and smaller semienclosed evaporative environments. Hence the hypothesis of low-latitude evaporative deepwater formation is difficult to reconcile with GENESIS atmospheric general circulation model climate simulations that suggest net annual precipitation in the near-equatorial intertropic convergence zone was even higher during the Maastrichtian than at present (P. Fawcett, personal communication, 1996). Second, our sampling of widely distributed sites failed to reveal any deep waters inhabited by benthic foraminifera with δ18Ocvalues close to those of near-surface planktonic foraminifera from tropical western gyre margins [cf. D'Hondt and Arthur, 1996]. Third, as previously noted, carbon isotopic evidence suggests that the 10°C deep waters originated at least partly in the midlatitude or high-latitude North Atlantic.

[36] Although a strictly evaporative tropical source could not have been a major source of deep Maastrichtian waters, it remains possible that the warmest intermediate water mass was created by midlatitude winter cooling of either shallow epicontinental seas or (sub)tropical surface waters translated poleward by western gyres. In either case, the water would have been preconditioned for sinking by warm-season evaporation (much as the Mediterranean, Red Sea, and North Atlantic deep waters are today). Under such circumstances, with a cool deepwater end-member of 6°C and mean deep water of 10°C, the 14°C end-member could have constituted as much as 50% of late Maastrichtian deep water (at least in regions outside of the northwestern Atlantic basin).

[37] This possibility returns us to Chamberlin's [1906] hypothesis. It appears likely that low-latitude or midlatitude evaporation played an important role in preconditioning some surface waters for sinking during the Maastrichtian (as it does today in the North Atlantic, Mediterranean, and Red Sea). Hence up to 50% of late Maastrichtian deep oceanic circulation could be loosely described as having been indirectly “actuated by the dense waters of the evaporating tracts.” In a stricter sense, however, such deepwater formation was probably directly actuated by seasonal cooling of surface waters that originated in evaporative regions.

5.2. Possible Effect of Deep Waters on High-Latitude Warmth

[38] Creation of deep waters by low-latitude evaporative brine formation has been proposed as a path of increased heat transport from low to high latitudes during warm climate intervals [e.g., Chamberlin, 1906; MacLeod and Huber, 1996]. Previous studies have estimated that sustenance of Eocene high-latitude warmth solely by deep oceanic heat transport would have required as much as 60–80 sverdrups of warm deepwater formation [Crowley, 1991; Sloan et al., 1995]. Such a rate of deepwater formation would have significantly exceeded the total modern deepwater formation of ∼40 sverdrups [Sloan et al., 1995]. Estimates of late Maastrichtian (C30N) warm-season surface water temperatures at southern high-latitude ODP Sites 689B and 690C range from 10° to 12.2°C [D'Hondt and Arthur, 1996]. This paleotemperature range approximates those of the middle Eocene and late Paleocene (10°–12°C) and approaches that of the early Eocene (14°–15°C) [Zachos et al., 1994]. Consequently, sustenance of late Maastrichtian high-latitude warmth by poleward flow of deep waters would probably have required similarly high rates of low-latitude deepwater formation.

[39] The complete absence of late Maastrichtian paleotemperature estimates for high-latitude Pacific waters prevents us from directly addressing the possibility that some heat was carried poleward by deep water in the Pacific Ocean. However, we can largely rule out the possibility that heat was carried poleward by deep water in the Atlantic and Indian portions of the Southern Ocean. As previously discussed, late Maastrichtian intermediate-depth water at southern high-latitude Atlantic Sites 689B and 690C (70°S) and Indian Ocean Site 761 (42°S) was characterized by a temperature in the range of 5°–7°C (given a mean oceanic δ18O value of −1.0‰). The relative coolness of this water indicates that it did not originate equatorward of these sites. Consequently, at least in these high southern-latitude regions, mild high-latitude sea surface and coastal temperatures cannot be attributed to surface outcropping of warm deep waters from lower latitudes.


[40] All isotopic analyses were undertaken at the Isotope Biogeochemistry Laboratory of Pennsylvania State University. We thank D. Walizer for laboratory assistance and M. Pilson for comments on δ18Owand the evaporative histories of modern deep water masses. We thank P. Johnson and N. Gorbea for assistance with scanning electron microscopy and plate layout. R. Pockalny shared his expertise in plate back tracking and graphical manipulation. Thoughtful comments by L. C. Sloan, T. C. Moore Jr., and an anonymous reviewer helped to improve the manuscript. The plate tectonic reconstruction used in our figures is from the Ocean Drilling Stratigraphic Network (established by GEOMAR, Research Center for Marine Geosciences, Kiel, and the Geological Institute of the University of Bremen). This study was funded by the Marine Geology and Geophysics Program of the U.S. National Science Foundation.