Deep-sea paleotemperature record of extreme warmth during the Cretaceous Brian T. Huber* Department of Paleobiology, NHB-121, Smithsonian Institution, Wasliington, D.C. 20560, USA Richard D. Norris* Department of Geology and Geophysics, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543, USA Kenneth G. IVJacLeod* Department of Geological Sciences, University of Missouh, Columbia, Missouh 65211, USA ABSTRACT Oxygen isotope analyses of well-preserved foraminifera from Blake Nose (30?N paleo- latitude, North Atlantic) and globally distributed deep-sea sites provide a long-term pa- leotemperature record for the late Albian-Maastrichtian interval that is difficult to rec- oncile with the existence of significant Cretaceous ice sheets. Given reasonable assumptions about the isotopic composition of Cretaceous seawater, our results suggest that middle bathyal water temperatures at Blake Nose increased from ~12 ?C in the late Albian through middle Cenomanian to a maximum of 20 ?C during the latest Cenomanian and earliest Turonian. Bottom waters were again ~12 ?C during the middle Campanian and cooled to a minimum of 9 "C during the Maastrichtian. Correlative middle bathyal fo- raminifera from other ocean basins yield paleotemperature estimates that are very similar to those from Blake Nose. Comparison of global bottom-water temperatures and latitu- dinal thermal gradients suggests that global climate changed from a warm greenhouse state during the late Albian through late Cenomanian to a hot greenhouse phase during the latest Cenomanian through early Campanian, then to cool greenhouse conditions dur- ing the mid-Campanian through Maastrichtian. Keywords: Cretaceous climate, greenhouse effect, oxygen isotopes, ocean circulation, polar regions. INTRODUCTION Greenhouse conditions prevailed during the Cretaceous. Isotopic data, paleobiogeography of terrestrial and marine organisms, leaf phys- iognomy, and distribution of climatically sen- sitive sediments indicate temperatures gener- ally warmer than present, especially at high latitudes (e.g., Barron, 1983; Herman and Spi- cer, 1997; Frakes, 1999). Despite evidence for warmth, conditions at high latitudes have been debated. Continental ice sheets in Antarctica have been invoked to explain short-term glob- al eustatic sea-level fluctuations during the Cretaceous (Matthews and Poore, 1980; Abreu et al., 1998) and the occurrence of glen- donites and diamictites in Lower Cretaceous sediments deposited at high paleolatitudes (Kemper, 1986; Frakes et al., 1992). More- over, oxygen isotope data from belemnites and other macroinvertebrates could indicate at least seasonally cold temperatures and limited polar ice caps (Sellwood et al., 1994; Ditch- field et al., 1994). Evidence for warmth and lack of any unequivocal Cretaceous glacial de- posits convinced most workers that the Cre- taceous was effectively ice free, but recent studies have revived the glacial hypothesis as an explanation for high-frequency sea-level changes and have cited strontium (StoU and Schrag, 1996) and oxygen isotopic data (Mill- er et al., 1999; StoU and Schrag, 1996, 2000) as evidence. Resolution of the Cretaceous ice- sheet debate requires new oxygen isotope data that can be reliably used for Cretaceous pa- leotemperature reconstructions. To that end we have generated 8'^0 data from monospecific samples of benthic fora- minifera and two to three species of planktic foraminifera for the late Albian-Maastrichtian (ca. 101-65 Ma) using samples from Ocean Drilling Program (GDP) Sites 1049 and 1050 (Fig. 1). We compare these data with com- posite foraminiferal 8'*0 records from south- ern high-latitude deep-sea sites to test whether they reveal similar long-term climate trends. We also combine new 8'*0 data from addi- tional deep-sea sites with previously published 8'*0 data to reconstruct latitudinal surface- and bottom-water temperature gradients for 1- 2 m.y. time slices within the late Albian, Cen- omanian-Turonian boundary interval, mid- Campanian, and mid-Maastrichtian. Our re- sults indicate that greenhouse conditions prevailed throughout the late Albian-Maas- trichtian, and bottom-water and high-latitude temperatures were too warm to be consistent with significant ice buildup at polar latitudes. MATERIAL AND METHODS Sediments cored at Blake Nose Sites 1050 and 1049 were deposited at -1200 and 1500 m water depths, respectively, and are com- posed of clay- and organic carbon-rich sedi- ments alternating with chalky marls in the up- per Albian, and of foraminiferal-nannofossil chalk in the Cenomanian-Maastrichtian inter- val (Norris et al., 1998). The sequence at Site 1050 is complete from the upper Albian through uppermost Cenomanian, but an ~0.5 m.y. hiatus occurs across the Cenomanian- Turonian boundary (Huber et al., 1999), and multiple hiatuses occur within the Turonian- Maastrichtian interval (Fig. 2). The sequence at Site 1049 includes mid-Campanian and Maastrichtian pelagic chalk separated by an ~5 m.y. unconformity. Stable isotopic analyses were obtained from single taxon separates of one or two epifaunal benthic foraminiferal species and two to three *E-mails: Huber?huber.brian@nmnh.si.edu; Norris?morris@whoi.edu; MacLeod?macleodk@ missouri.edu. Figure 1. Paleogeographic reconstruction (after IHay et al., 1999) for 94 IVIa showing locations of sites discussed in this study. ? 2002 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or editingd Geology; February 2002; v. 30; no. 2; p. 123-126; 3 figures; Data Repository item 2002008. 'geosociety.org. 123 0) S o = OT N O : CO CO u:^ en c 0) D) -1 CD o c o i2 03 C3 o o O CO Q. r h ai cn O ei i; en U) es ,2 F JFS- vy CD 0) u o i; S c c Q) o o fll u o O Q o 03 J- C ^?R o 3 1- > Si CO -C o ns E o rQiCh O 5 o m c O (1) f CO F < fll s ail CD 75 .. 1_ 85 ?- 95 ?? E 100 ?? . - ^ ?-I- S^ ^iP^} Site1049 I I to I I ? Benthic foraminifera > Deeper planktic foraminifera o Shallower planktic foraminifera 90 ?? 1- CTBI H?1?I 1 1- t%o- o Solid = benthic > Hollow = planktic k > ? Site 511' > ? Site 511^ 0.07%o for 8'^0 and methods of data correction were 'GSA Data Repository item 2002008, Tables 1-7, is available from Documents Secretary, GSA, EO. Box 9140, Boulder, CO 80301-9140, editing @geosociety. org, or at www.geosociety.org/pubs/ft2002.htm. 124 GEOLOGY, February 2002 described by Ostermann and Curry (2000). Sample values are reported relative to the Vi- enna Peedee belemnite (VPDB) standard. Pa- leotemperatures were calculated using the equation of Erez and Luz (1983) assuming (1) a value of ~1-0%CSMOW for the mean isotopic composition of seawater in a nonglacial world (Shackleton and Kennett, 1975) (SMOW is standard mean ocean water), (2) that the fo- raminiferal calcite formed in isotopic equilib- rium with Cretaceous seawater, and (3) that the Zachos et al. (1994) correction for modem- day latitudinal variation in S'^Osg^^atg^ is ap- propriate to use in estimates of Cretaceous sea-surface temperatures (SSTs). Foraminiferal shells from Blake Nose are generally well preserved, but several levels in the upper Albian and upper Cenomanian- lower Turonian of Site 1050 show significant shell recrystallization and a varying amount of calcite infilling and overgrowth (e.g., Huber et al., 1999). Such specimens were excluded from our stable isotopic study or cleaned of infilling chalk prior to stable isotopic analysis. Specimens from the other deep-sea sites dis- cussed in this study show excellent preserva- tion or minor to moderate shell recrystalliza- tion, and no calcite infilling or overgrowth. RESULTS AND DISCUSSION Late Alblan-Maastrichtian Temperatures at Blake Nose Oxygen isotope results indicate that middle bathyal waters at Blake Nose were very warm throughout the late Albian-middle Turonian, with temperatures averaging 16 ?C (Fig. 2). The coolest temperatures recorded in this interval occur in the late Albian and middle Cenoman- ian, when bottom waters reached ~12 ?C. The warmest deep-water temperatures are recorded in the latest Cenomanian and early Turonian, during which temperatures averaged 19 ?C. Bottom waters continued to be very warm dur- ing the Coniacian, averaging 17 ?C, then cooled to an average of 12 ?C during the middle Cam- panian and further cooled to an average of 12 ?C during the Maastrichtian. Minimum temper- atures of 9 ?C were recorded within the Maas- trichtian at 71.2 Ma and 67.8 Ma. Latitudinally adjusted 8'*0 values recorded by multiple planktic species from Blake Nose provide a range of estimated temperatures of the surface waters (Fig. 2). Species yielding the most depleted 8'*0 are assumed to best ap- proximate SSTs, and those yielding the most enriched values are assumed to have lived close to the thermocline. Our results indicate that subtropical SSTs were similar to those of the modem day during the late Albian-Turonian and middle and late Maastrichtian, but cooler (or more enriched in '^O) during the Conia- cian-early Maastrichtian. Small differences in 8'*0 values between planktic and benthic taxa during the late Cenomanian-early Turonian, Coniacian, and mid-Campanian suggest a poor- ly developed thermocline and strongly mixed surface waters during those times. Paleotemperatures estimated for bottom and surface waters at Sites 511 and 690 in the sub- Antarctic are consistent with the paleotempera- ture history inferred for bottom waters at Blake Nose (Fig. 2). Warm conditions prevailed during the late Albian-Cenomanian and Coniacian-ear- liest Campanian, extreme warmth occiured dic- ing the Turonian, and the coolest temperatures were reached during the Maastrichtian. Latitudinal Thermal Gradient Reconstructions Time-slice comparison of oxygen isotopic data from this and previous studies permits characterization of surface- and bottom-water latitudinal temperature gradients for four well- correlated Cretaceous time intervals (Fig. 3). Whereas the number of sites that can be com- pared is relatively few, their broad latitudinal distribution provides better coverage than has been obtained from other climate proxies. Despite differences in location and depth of the sites analyzed, bottom-water temperatures estimated for the time slices reveal surprising consistency across latitude and significant var- iation through time. For the 100-99 Ma time slice, middle bathyal paleotemperatures at all sites averaged 16 ?C and differed by <1 ?C. At Blake Nose, bottom waters were 3 ?C cool- er at middle bathyal Site 1050 than at upper bathyal Site 1052 (-1000 m shallower than Site 1050). In the 94-92 Ma time sUce, mean values for bottom waters at middle bathyal sites as well as deeper Site 551 were warmer than in the late Albian, ranging from 18 to 19 ?C. For the 75.4-76.5 Ma and 66.5-68.5 Ma intervals, bottom-water temperatures in all the ocean basins and across latitudes were much cooler than in the early Turonian, with esti- mates for most sites ranging from 9 to 11 ?C. The four Cretaceous time-slice reconstmc- tions show SST gradients that are considerably lower than the mean modern-day SST gradient of ~0.40 ?C per degree of latitude (Fig. 3). Tropical SSTs for the 99-100 Ma time slice are close to those of the modem day, but SSTs at high latitudes are much warmer than at present, resulting in an average SST gradient of 0.25 ?C per degree of latitude. The 92-94 Ma time shce indicates extremely warm northern and southem high-latitude SSTs and an asymmetrical and nearly flat thermal gra- dient that averaged 0.10 ?C per degree of lati- tude. Although the benthic temperature estimates and other geological evidence indicate extreme global warmth at this time, such a uniform dis- tribution of ocean siuface temperatures is not supported by existing theories of ocean and at- mospheric heat transport. Therefore, isotopic ev- idence for extreme polar warmth needs to be substantiated by additional high-latitude data. In the absence of evidence for significant diagenetic alteration of foraminiferal calcite at the highest latitude sites (Huber et al., 1995), regional fac- tors affecting the oxygen isotope ratio of the am- bient seawater (8^^,) could account for overesti- mation of high-latitude SSTs. Higher evaporation rates in the tropics would increase the 8^^, of tropical surface waters while increased precipitation of isotopically depleted water would lead to '*0 depletion in high-latitude siu'- face waters. This explanation is supported by cli- mate modeling studies (e.g., Poulsen et al., 1999; Hay and DeConto, 1999) and evidence for increased storm intensity during the Cretaceous (Ito et al., 2001). In addition, surface-water 8'*0 values at Sites 511 and 551 may have been in- fluenced by coastal runoff of isotopically de- pleted precipitation from bordering emergent ter- rains. Such landmasses are variably depicted in different paleogeographic reconstructions (e.g., Scotese, 1997; Hay et al., 1999) because of un- certainties in the geologic record. Tropical SSTs estimated for the 75.4-76.5 Ma and 68.5-66.5 Ma reconstructions range from 10 to 8 ?C cooler than modern day SSTs at these locations. For the 75.5-74.5 recon- struction latitudinal SST gradients are quite low, averaging ~0.13 ?C per degree of lati- tude, because of relatively warm high-latitude paleotemperatures. However, cooler high lati- tudes result in a higher SST gradient of ~0.22 ?C per degree of latitude for the 68.5-66.5 Ma time slice. Because some shell recrystalliza- tion is evident in all foraminiferal samples from low-latitude deep-sea sites during the Campanian-Maastrichtian time intervals, dia- genetic shift of the foraminiferal S'^O toward more positive ratios could account for the ap- parent cooler than modern temperatures and diminished vertical 8'^0 gradients (Wilson and Opdyke, 1996). PALEOCLIMATIC AND PALEOCEANOGRAPHIC IMPLICATIONS Our data suggest that during the late Albian to mid-Cenomanian there was a warm green- house climate with warm bottom waters and moderately low latitudinal thermal gradients. A transition to a hot greenhouse climate oc- curred in the latest Cenomanian as bottom- water temperatures reached 20 ?C and tem- peratures at high latitudes approached those of low latitudes. This hot greenhouse state con- tinued into the early Campanian, but was fol- lowed by a cool greenhouse climate when bot- tom-water temperatures dropped below 12 ?C, and low-latitude surface-water temperatures dropped below those of the modern day. Although the Blake Nose record suggests bottom-water cooling excursions of 3-5 ?C at several middle through Late Cretaceous inter- GEOLOGY, February 2002 125 vals, SST estimates from these same intervals do not exhibit changes that parallel the deep- water temperature trends, which would be ex- pected if changes in the 8'*0 composition of global seawater had occurred during those times. Short-term variability observed among planktic 8'^0 values on Blake Nose (MacLeod et al., 2001; Wilson and Norris, 2001) suggests that high-resolution sampling might increase the range of SSTs estimated for Site 1050. However, even high-frequency S'^O variation should be correlated with benthic 8'*0 excur- sions if there were a shift in 8^^05,5,5,^2,5,^ T 1?^ global cooling. In addition, late Albian-Maas- trichtian bottom waters and high-latitude sur- face waters seem too warm to be compatible with the ice-sheet scenarios proposed by Miller et al. (1999) and StoU and Schrag (1996, 2000). The time-slice reconstructions for the late Al- bian, mid-Campanian, and late Maastrichtian show that surface- and bottom-water tempera- tures intersect at southern high latitudes, sug- gesting that the south polar region was a primary source of bottom-water formation during those times. However, the reconstruction for the late Cenomanian-early Turonian shows no conver- gence of bottom- and surface-water temperatures and, therefore, no evidence for deep-water con- vection at high latitudes. The large vertical 8'*0 gradients and extreme warmth at high latitudes, warm bottom-water temperatures, and maximum flooding of continental interiors during the late Cenomanian-early Turonian point to the most favorable conditions for deep-water formation at low latitudes of any time during the past 250 m.y. Additional information, including carefully scrutinized stable isotopic and other paleocli- matic proxy data, from this supergreenhouse pe- riod is needed to better understand the d3Tiamics of an ocean-climate system that was so different from that of today. ACKNOWLEDGMENTS We thank the Ocean Drilling Program for sup- plying samples used in this study, and P. Wilson, K. Bice, E. Barron, and W Hay for constructive com- ments. Funding for this project was obtained from the Smithsonian Institution's Scholarly Studies Pro- gram (to Huber) and National Science Foundation grant OCE-9819050 (to Norris). REFERENCES CITED Abreu, VS., Hardenbol, I., Haddad, I., Haddad, G.A., Baum, G.R., Droxler, A.W, and Vail, PR., 1998, Oxygen isotope synthesis: A Cretaceous ice-house?, in de Graciansky, PC, et al., eds., Mesozoic and Cenozoic sequence stratigraphy of European basins: Society for Sedimentary Geology Special Publication 60, p. 75-80. Barrera, E., and Huber, B.T., 1990, Evolution of Antarctic waters during the Maestrichtian: Foraminifer oxygen and carbon isotope ratios, ODP Leg 113, in Barker, PF, Kennett, I.P, et al.. Proceedings of the Ocean Drilling Pro- gram, Scientific Results 113: College Station, TX, Ocean Drilling Program, p. 813-823. Barrera, E., and Savin, S.M., 1999, Evolution of late Campanian-Maastrichtian marine climates and oceans, in Barrera, E., and lohnson, C, eds., Evolution of the Cretaceous ocean- climate system: Geological Society of Amer- ica Special Paper 332, p. 245-282. Barron, E.I., 1983, A warm, equable Cretaceous: The nature of the problem: Earth-Science Re- views, v. 19, p. 305-338. Cande, S.C, and Kent, D.V, 1995, Revised calibra- tion of the geomagnetic polarity timescale for the Late Cretaceous and Cenozoic: loumal of Geophysical Research, v. 100, p. 6093-6095. D'Hondt, S., and Arthur, M.A., 1996, Late Creta- ceous oceans and the cool tropic paradox: Sci- ence, v. 271, p. 1838-1841. Ditchfield, PW, Marshall, I.D., and Pirrie, D., 1994, High latitude palaeotemperature variation: New data from the Tithonian to Eocene of James Ross Island, Antarctica: Palaeogeogra- phy, Palaeoclimatology, Palaeoecology, v. 107, p. 79-101. Erez, J., and Luz, B., 1983, Experimental paleotem- perature equation for planktonic foraminifera: Geochimica et Cosmochimica Acta, v. 47, p. 1025-1031. Fassell, M.L., and Bralower, T.J., 1999, Warm, equable mid-Cretaceous: Stable isotope evi- dence, in Barrera, E., and Johnson, C, eds.. Evolution of the Cretaceous ocean-climate system: Geological Society of America Spe- cial Paper 332, p. 121-142. Frakes, L.A., 1999, Estimating the global thermal state from Cretaceous sea surface and conti- nental temperature data, in Barrera, E., and Johnson, C, eds., Evolution of the Cretaceous ocean-climate system: Geological Society of America Special Paper 332, p. 49-57. Frakes, L.A., Francis, I.E., and Syktus, J.I., 1992, Climate modes of the Phanerozoic: Cam- bridge, UK, Cambridge University Press, 274 p. Hay, W.W., and DeConto, R.M., 1999, Compari- son of modern and Late Cretaceous merid- ional energy transport and oceanology, in Barrera, E., and Johnson, C, eds., Evolution of the Cretaceous ocean-climate system: Geological Society of America Special Pa- per 332, p. 283-300. Hay, W.W., DeConto, R., Wold, C.N., Wilson, K.M., Voigt, S., Schulz, M., Wold-Rossby, A., Dullo, W.-C, Ronov, A.B., Balukhovsky, A.N., and Soding, E., 1999, Alternative global Creta- ceous paleogeography, in Barrera, E., and Johnson, C, eds., Evolution of the Cretaceous ocean-climate system: Geological Society of America Special Paper 332, p. 1-47. Herman, A., and Spicer, R.A., 1997, New quanti- tative palaeoclimate data for the Late Creta- ceous Arctic: Evidence for a warm polar ocean: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 128, p. 227-251. Huber, B.T, Hodell, D.A., and Hamilton, C.P, 1995, Mid- to Late Cretaceous climate of the south- ern high latitudes: Stable isotopic evidence for minimal equator-to-pole thermal gradients: Geological Society of America Bulletin, v. 107, p. 1164-1191. Huber, B.T., Leckie, R.M., Norris, R.D., Bralower, T.J., and CoBabe, E., 1999, Foraminiferal as- semblage and stable isotopic change across the Cenomanian-Turonian boundary in the sub- tropical North Atlantic: Journal of Foraminif- eral Research, v. 29, p. 392-417. Ito, M., Ishigaki, A., Nishikawa, T, and Saito, T, 2001, Temporal variation in the wavelength of hummocky cross-stratification: Implications for storm intensity through Mesozoic and Ce- nozoic: Geology, v. 29, p. 87-89. Kemper, E., 1986, Allgemeine und regionale geo- logic BR Deutschland und nachbargebiete, tektonic, stratigraphie, palaontologie: Geolo- gisches Jahrbuch, v. 96, p. 5-185. MacLeod, K.G., Huber, B.T, Pletsch, T, Rohl, U., and Kucera, M., 2001, Maastrichtian forami- niferal and paleoceanographic changes on Milankovitch time scales: Paleoceanography, v. 16, p. 133-154. Matthews, R.K., and Poore, R.Z., 1980, Tertiary 5'*0 record and glacio-eustatic sea-level fluc- tuations: Geology, v. 8, p. 501-504. Miller, K.G., Barrera, E., Olsson, R.K., Sugarman, P.J., and Savin, S.M., 1999, Did ice drive early Maastrichtian eustasy?: Geology, v. 27, p. 783-786. Norris, R.D., Kroon, D., and Klaus, A., 1998, Pro- ceedings of the Ocean Drilling Program, Ini- tial reports. Volume 171B: College Station, Texas, Ocean Drilling Program, p. 1-749. Ostermann, D.R., and Curry, W.B., 2000, Calibra- tion of stable isotopic data: An enriched gas S'^O standard used for source gas mixing de- tection and correction: Paleoceanography, v. 15, p. 679-697. Poulsen, C.J., Barron, E.J., Peterson, W.H., and Wil- son, PA., 1999, A reinterpretation of mid- Cretaceous shallow marine temperatures through model-data comparison: Paleocean- ography, V. 14, p. 679-697. Scotese, C.R., 1997, Continental drift (seventh edi- tion): Arlington, Texas, PALEOMAP Project, 79 p. Sellwood, B.W., Price, G.D., and Valdes, PL, 1994, Cooler estimates of Cretaceous temperatures: Nature, v. 370, p. 453^55. Shackleton, N.J., and Kennett, J.P, 1975, Paleotem- perature history of the Cenozoic and the ini- tiation of Antarctic glaciation: Oxygen and carbon isotope analysis in DSDP Sites 277, 279, and 280, in Kennett, J.P, et al., Initial reports of the Deep Sea Drilling Project, Vol- ume 29: Washington, D.C., U.S. Government Printing Office, p. 743-755. Stoll, H.M., and Schrag, D.P, 1996, Evidence for glacial control of rapid sea level changes in the Early Cretaceous: Science, v. 272, p. 1771-1774. Stoll, H.M., and Schrag, D.P, 2000, High-resolution stable isotope records from the Upper Creta- ceous rocks of Italy and Spain: Glacial epi- sodes in a greenhouse planet?: Geological So- ciety of America Bulletin, v. 112, p. 308-319. Stott, L.D., and Kennett, J.P, 1990, The paleocean- ographic and paleoclimatic signature of the Cretaceous/Paleogene boundary in the Antarc- tic: Stable isotopic results from ODP Leg 113, in Barker, PF, Kennett, J.P, et al., Proceedings of the Ocean Drilling Program, Scientific Re- sults 113: College Station, TX, Ocean Drilling Program, p. 829-848. Wilson, PA., and Norris, R.D., 2001, Warm tropical ocean surface and global anoxia during the mid-Cretaceous period: Nature, v. 412, p. 425-429. Wilson, PA., and Opdyke, B.N., 1996, Equatorial sea-surface temperatures for the Maastrich- tian revealed through remarkable preserva- tion of metastable carbonate: Geology, v. 24, p. 555-558. Zachos, I.e., Stott, L.D., and Lohmann, K.C, 1994, Evolution of early Cenozoic marine tempera- tures: Paleoceanography, v. 9, p. 353-387. Manuscript received June 19, 2001 Revised manuscript received October 1, 2001 Manuscript accepted October 17, 2001 Printed in USA 126 GEOLOGY, February 2002