Link Between Oceanic Heat Transport, Thermohaline Circulation, and the Intertropical Convergence Zone in the Early Pliocene Atlantic

Home , Guinea Current

Link between oceanic heat transport, thermohaline circulation, and the Intertropical Convergence Zone in the early Pliocene Atlantic

K. Billups* Earth Sciences Department, University of California, Santa Cruz, California 95064, USA A. C. Ravelo Ocean Sciences Department, University of California, Santa Cruz, California 95064, USA J. C. Zachos Earth Sciences Department, University of California, Santa Cruz, California 95064, USA R. D. Norris Woods Hole Oceanographic Institution, Clark 117, MS-23, Woods Hole, Massachusetts 02543-1541, USA

ABSTRACT have been triggered directly or indirectly by the Planktonic foraminiferal oxygen isotope records from the western and eastern tropical restriction of surface-water flow through the Pacific and Atlantic Oceans suggest a southward shift in the Intertropical Convergence Zone Central American seaway. toward its modern location between 4.4 and 4.3 Ma. A concomitant shift in the carbon isotope compositions of Atlantic benthic foraminifera provides strong evidence for an increased thermo- METHODS haline overturn at this time. We suggest that the southward shift of the Intertropical Convergence Zone and associated change in trade-wind circulation altered equatorial surface hydrography, Analytical increased the advection of warmer and more saline surface waters into the subtropical and North Sample preparation and stable isotope analy- Atlantic, and contributed to thermohaline overturn. ses from ODP Sites 925, 851, and 959 follow standard procedures summarized in our previ- INTRODUCTION and Atlantic (Site 959, Norris, 1998) sites sample ous studies (Table 1). Analytical precision is Ocean heat transport from the equatorial re- the eastern equatorial surface hydrography, which better than 0.05‰ (δ13C) and 0.08‰ (δ18O). In gions to the high latitudes is a critical component is highly sensitive to changes in the intensity and accordance with our previous studies, Site 806 of the global climate system. The Atlantic Gulf direction of the trade winds. By characterizing Globigerinoides sacculifer (without a sac-like Stream, in particular, carries warm water north- changes in the surface hydrography at each site final chamber) and Globorotalia tumida were ward from the subtropics and supplies heat to (e.g., mixed-layer depth), we draw inferences picked from the 355–500 µm size fractions. high latitudes. In the western tropical Atlantic, about the character of the wind field and the posi- Age assignments for Sites 851 and 925 reflect the seasonal intensification of the southeast tion of the Intertropical Convergence Zone. tuning to Northern Hemisphere summer insola- trades, the direction of warm surface currents, We propose that in the early Pliocene a south- tion (Shackleton et al., 1995a; Tiedemann and and the associated migration of the Intertropical ward shift of the zone occurred between 4.4 and Franz, 1997, respectively). We use biostrati- Convergence Zone (Fig. 1) are linked to the 4.3 Ma. This event most likely reflects a change graphic data for Sites 806 (Mayer et al., 1993) amount of heat transport into the subtropics in the equatorial surface currents that allowed an and 959 (Norris, 1998) on the Berggren et al. (Philander and Pakanowski, 1986a, 1986b). increased proportion of warm water to flow north (1995) time scale. The time scales of Shackleton Observational and modeling studies suggest into the subtropics, thereby contributing to deep- et al. (1995a) and Berggren et al. (1995) are com- that northward heat advection and thermohaline water formation in the North Atlantic. The shift parable. Age uncertainties are <20 k.y. among overturn in the Atlantic were not as efficient until in the Intertropical Convergence Zone and asso- tuned age models and probably <100 k.y. among the closure of the Central American seaway be- ciated surface- and deep-ocean responses may records based on biostratigraphic data. tween 3 and 5 Ma (Keigwin, 1982; Mikolajewicz et al., 1993; Haug and Tiedemann, 1998). The on- set of large-scale Northern Hemisphere glaciation Longitude may have been closely linked to increased pole- 140˚E180˚W140˚W 100˚W 180˚W60140˚W 100˚W ˚W ward moisture transport as a result of a strength- summer winter ened Gulf Stream system (e.g., Stanley, 1995). 20˚N ITCZ We reconstruct equatorial sea-surface hydrogra- 10˚N NECC NECC 806 EUC ITCZ phy during the early Pliocene by using the oxygen 0˚N • EUC • 851/ SEC 846 • SEC • isotope composition of planktonic foraminifera 10˚S from four near-equatorial locations (Fig. 1, Table A B 1). The western Atlantic (Ocean Drilling Program

[ODP] Site 925, Billups et al., 1998a) and Pacific Latitude 20˚N 1 ITCZ NBCC (Site 806, this study) sites are located in regions NBCC 10˚N that supply warm and saline waters to the western NECC 959 ITCZITC GC GC • EUC • • EUC • boundary currents of these oceans. The eastern 0˚N 925/ Z Pacific (Site 851, Cannariato and Ravelo, 1997) 926/ SEC 10˚S 929 SEC summer C winter D *Present address: Department of Earth and Plan- 20˚S etary Sciences, Harvard University, 20 Oxford Street, 60˚W2040˚W2020˚W 0˚W ˚E 40˚W 20˚W 0˚W ˚E Cambridge, Massachusetts, 02138, USA. E-mail: [email protected]. Figure 1. Locations of Ocean Drilling Program Sites 806, 851, 846, 925, 926, 929, and 959. Italics 1GSA Data Repository item 9929, Table with ages, mark sites for deep-water reconstructions. Equatorial Pacific (A and B) and Atlantic (C and D) G. sacculifer δ18O, G. tumida δ18O, G. tumida– surface currents are as follows: SEC—South Equatorial Current; NECC—North Equatorial G. sacculifer ∆δ18O, is available on request from Docu- Counter Current; EUC—Equatorial Undercurrent (open arrows); NBCC—North Brazil Coastal ments Secretary, GSA, P.O. Box 9140, Boulder, CO Current; GC—Guinea Current; ITCZ—Intertropical Convergence Zone. Dashed arrows signify 80301-9140. E-mail: [email protected]. relatively weak and solid arrows indicate relatively strong transport of surface water.

Data Repository item 9929 contains additional material related to this article.

Geology; April 1999; v. 27; no. 4; p. 319–322; 4 figures; 1 table. 319 Intertropical Convergence Zone (Cannariato and Ravelo, 1997). In the western equatorial Atlantic, a steplike 0.2‰ decrease occurred in the planktonic δ18O values at 4.36 Ma (Fig. 3C). With the exception of events at 4.5 and 4.55 Ma, the G. sacculifer– N. dutertrei gradient remained relatively stable between 4.7 and 3.2 Ma. This steplike decrease in planktonic δ18O values most likely reflects a freshening of surface seawater in response to a southward shift of the Intertropical Convergence Zone (Billups et al., 1998a). At the same time (ca. 4.35 Ma) in the eastern Equatorial Sea-Surface Hydrography cause G. tumida does not occur in the Atlantic, Atlantic at Site 959, the G. sacculifer–N. dutertrei Reconstructions we rely on the δ18O values of Neogloboquadrina δ18O difference decreased significantly primarily Previous studies have demonstrated how the dutertrei, a thermocline dweller, to estimate past due to a decrease in the amplitude of N. dutertrei δ18O values of planktonic foraminifera can be changes in relative mixed-layer depth. δ18O fluctuations (Fig. 3D). These observations used to reconstruct past changes in the relative can be explained by a decrease in the thermal depths of the tropical mixed layer and thermo- RESULTS gradient within the upper thermocline (Norris, cline (Ravelo and Fairbanks, 1992; Ravelo and In the western equatorial Pacific (Site 806), the 1998). A weaker subsurface temperature gradient Shackleton, 1995; Cannariato and Ravelo, 1997; G. sacculifer–G. tumida δ18O gradient remained after 4.35 Ma more closely reflects the hydrogra- Norris, 1998). In environments with deep mixed relatively small and essentially stable between 3 phy within the modern Guinea Current (Norris, layers throughout the year, such as the western and 5 Ma (Fig. 3A). These results suggest that the 1998). In contrast, a stronger thermal gradient equatorial Atlantic and Pacific, the photic-zone thermal gradients in the photic zone remained prior to 4.35 Ma is more characteristic of up- temperature gradient (and hence the vertical small and the mixed layer remained deep. This welling regimes produced by the westward ad- foraminiferal δ18O gradient) is small, because observation is consistent with a stable, deep vection of surface water within the South Equa- the thermocline is below the base of the photic mixed layer observed today (e.g., Fig. 2A). torial Current (Voituriez and Herbland, 1982). zone (Fig. 2, A and C). In regions where the In the eastern equatorial Pacific G. tumida The shift in the thermal structure of the thermo- depth of the thermocline is more variable, such δ18O values increased between 4.4 and 4.3 Ma, cline at 4.35 Ma can be related to a decrease in the as the eastern equatorial Pacific and Atlantic, the reflecting a temperature decrease at the bottom influence of the South Equatorial Current at Site large photic-zone temperature range results in of the photic zone due to the shoaling of the 959, which in turn reflects a weakening of the relatively large foraminiferal δ18O gradients thermocline (Fig. 3B). A decrease in mixed-layer southeasterly trades and a southward shift of the (Fig. 2, B and D). At the Pacific sites we recon- depth is consistent with an oceanic response to a Intertropical Convergence Zone. struct photic zone temperature gradients by weakening or southward shift of the North Equa- using G. sacculifer, a mixed-layer dweller, torial Counter Current. Both changes could be DISCUSSION together with G. tumida, a deep dweller that is due to a decrease in the intensity of southeasterly Planktonic foraminiferal δ18O records from constrained to the bottom of the photic zone. Be- trade winds, allowing a southward retreat of the the tropical ocean provide a compelling case for a surface-ocean response to a southward shift in Temperature (˚C) the Intertropical Convergence Zone between 4.4 1020 30 1020 30 1020 30 1020 30 and 4.3 Ma. Other paleoclimate proxies have 0 documented evidence for a southward shift of the A 806 B 851 C 925 D 959 50 zone during the early Pliocene. Eolian grain-size March distributions across the eastern equatorial Pacific 100 -June (Hovan, 1995) and foraminiferal faunal assem- July 150 -Feb. blages in the western equatorial Atlantic (Chais- son and Ravelo, 1997) and eastern and western March 200 March -July Nov. Pacific (Chaisson, 1995) show patterns that sug- -June -June 250 July Aug. July gest that the Intertropical Convergence Zone was -Feb. -Feb. -Oct. located farther north during the late Miocene and 300 earliest Pliocene, prior to ca. 4 Ma. Our study W. eq. Pac. E. eq. Pac. W. eq. Atl. E. eq. Atl. 350 provides additional evidence for shifts in both (1.0‰) (1.9‰) (0.9‰) (1.4‰) ocean basins that suggest either a global change 400 in meridional temperature gradients or regional 1 0 -1 -2 1-20 -1 -2 1-20 -1 1 0 -1 changes in tropical sea-surface temperature dis- 18 δ O (‰) tributions that influence the position of the Inter- tropical Convergence Zone. Figure 2. Monthly water-column temperature profiles at locations of Ocean Drilling Program A southward shift of the Intertropical Conver- Sites 806 (A), 851 (B), 925 (C), and 959 (D) (Levitus et al., 1994), showing average annual mixed- layer depth as defined by 18 °C isotherm (solid arrows). Open arrows show planktonic gence Zone signifies a profound change in the foraminiferal core-top δ18O gradients (∆δ18O) between surface- and deeper-dwelling species direction of warm surface-water flow in the (Pacific [Pac.] Sites 851 and 806: Globigerinoides sacculifer–Globorotalia tumida; Atlantic [Atl.] Atlantic. In the modern tropical Atlantic, during Sites 925 and 959: G. sacculifer–Neogloboquadrina dutertrei). As result of predominantly south- the boreal summer when the zone lies at its north- easterly wind stress, mixed layer deepens from eastern to western basins of Pacific and Atlantic, producing east to west decrease in foraminiferal ∆δ18O (Pacific: 1.9‰ to 1.0‰; Atlantic: 1.4‰ to ernmost position, warm surface-water flow is 0.9‰;W. is western, E. is eastern, eq. is equatorial). For Site 959 core-top reconstruction we use directed to the west, across the equator, and then V29-144, which is located close to Site 959 (see Table 1). back to the east within the North Equatorial

320 GEOLOGY,April 1999 Counter Current and Equatorial Undercurrent -2 A -2 (EUC) (Fig. 1C). The NECC is comparable to the G. sacculifer Gulf Stream in mean velocity, amplitude, and -1 -1 width, and significantly modulates northward heat flux (Richardson and Reverdin, 1987; G. tumida 0 Philander and Pakanowski, 1986b). Large north- 0 ward tropical to subtropical heat fluxes occur 1 1 during the boreal winter when the North Brazil 0 Coastal Current flows continuously into the Gulf -2 B G. sacculifer of Mexico (Philander and Pakanowski, 1986a, 1 1986b) (Fig. 1D). -1 The modern seasonal cycle serves as an ana- 2 logue for interpreting early Pliocene climate thermocline shoaling 0 3 change. Prior to 4.4 Ma, a northern position of G. tumida the Intertropical Convergence Zone limited warm 1 O(‰) surface-water advection into the subtropical O(‰) 18 region of the Atlantic. Instead, warm surface- 18 -2 C sea surface freshening -2 G. sacculifer water flow was directed eastward in the North -1 Equatorial Counter Current. By 4.3 Ma, the Inter- -1 tropical Convergence Zone had shifted to the N. dutertrei 0 south, at least seasonally, diverting warm water 0 into the subtropics via the North Brazil Coastal 1 Current. We propose that, as a consequence, 1 northward oceanic transport of warm and salty -2 D stronger thermocline 0 surface water increased, ultimately contributing to North Atlantic deep-water formation. 1 -1 G. sacculifer Implications for Thermohaline Circulation 2 N. dutertrei and High-Latitude Warmth 0 weaker thermocline 3 Previous reconstructions of early Pliocene thermohaline circulation have demonstrated that 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 between 4.3 and 3.7 Ma, the relative flux of deep water forming in the North Atlantic was en- Age (Ma) hanced and warmer and more saline than today Figure 3. Ocean Drilling Program oxygen isotope records. A: Site 806. B: Site 851. (Billups et al., 1998b). These observations are de- C: Site 925. D: Site 959. Heavy lines represent δ18O gradients (∆δ18O). rived from benthic foraminiferal isotope records from Ceara Rise Sites 925 and 929 (Table 1, Fig. 4A) that span the modern mixing zone between North Atlantic deep water and Antarctic bottom A water in the western equatorial Atlantic. Rela- 2 925 (3.2-4.7 Ma)/ δ18 tively high O values at the shallower site in 926 (4.7-6 Ma) comparison to the deeper site, indicating warmer 929 2.5

and more saline North Atlantic deep-water con- O(‰) 959 ditions (Billups et al., 1998b), could be explained 18 846 by increased oceanic heat and salt transport. 3 However, it is the carbon isotope evidence of an warmer, more saline increase in the relative flux of North Atlantic North Atlantic deep water deep water occurring synchronously with a shift in the Intertropical Convergence Zone that hints B NADW at a direct link between North Atlantic deep-water formation and tropical surface-ocean circulation. 1 An increase in the relative North Atlantic deep- water flux between 4.4 and 4.3 Ma is evidenced by 13 0 an increase in the δ C values and a reduction in C(‰) the δ13C amplitude of benthic foraminifera from 13 the eastern equatorial Atlantic (Site 959) (Fig. 4B). -1 Site 959 is close to the upper mixing zone between North Atlantic deep water and intermediate water 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0 and is therefore sensitive to changes in relative Age (Ma) water-mass fluxes and thermohaline circulation. The benthic δ13C increase and the reduced δ13C Figure 4. A: Early Pliocene benthic (Cibicidoides) δ18O records from Ceara Rise Ocean Drilling δ18 amplitude signify an increased proportion of Program Sites 925 and 929 illustrating relatively high O values at Site 925 in comparison to Site 929 between 4.3 and 3.7 Ma. B: Late Miocene–early Pliocene δ13C records from equatorial North Atlantic deep water, because no equivalent Atlantic and Pacific (see Table 1) illustrating increase in Site 959 δ13C values at 4.3–4.4 Ma, which δ13C increase is recorded in mean ocean δ13C suggests increase in relative North Atlantic deep-water (NADW) flux.

GEOLOGY,April 1999 321 values (Pacific Site 846, Shackleton et al., 1995b) ACKNOWLEDGMENTS flow through the Central American Isthmus: (Fig. 4B). No change in the foraminiferal δ13C We thank E. Jansen (Site 806) and the Ocean Drilling Paleoceanography, v. 12, p. 429–442. Program for providing samples, J. Farrell for a con- Mikolajewicz, U., Maier-Raimer, E., Crowley, T. J., and values occurred at Site 925, most likely because structive review, and M. J. Harvey for technical support. Kim, K.-Y., 1993, Effect of Drake and Pana- the core of North Atlantic deep water remained This work was funded by a Joint Oceanographic Institu- manian gateways on the circulation of an ocean close to 3000 m between 6.0 and 3.2 Ma. tions/United States Science Advisory Committee model: Paleoceanography, v. 8, p. 409–426. (JOI/USSAC) fellowship and a Geological Society of Norris, R. D., 1998, Miocene-Pliocene surface water America grant to Billups, a JOI/USSAC grant to Norris, hydrography of the eastern equatorial Atlantic, in What Factors Triggered a Southward Shift of and National Science Foundation grants OCE-9458367- Mascle, J., et al., Proceedings of the Ocean the Intertropical Convergence Zone? 001 (to Zachos) and OCE-9510440 (to Ravelo). Drilling Program, Scientific results, Volume 159: The uplift of the Central American seaway College Station, Texas, Ocean Drilling Program, and/or a weakening of Southern Hemisphere tem- REFERENCES CITED p. 445–480. perature gradients may explain the observed Berggren, W. A., Hilgren, F. J., Langereis, C. G., Kent, Philander, S. G., and Pakanowski, R. C., 1986a, The D. V., Obradovitch, J. D., Raffi, I., Raymo, M. E., mass and heat budget in a model of the tropical surface- and deep-water changes. Mikolajewicz and Shackleton, N. J., 1995, Late Neogene Atlantic Ocean: Journal of Geophysical Re- and Crowley (1997) demonstrated that a partly (Pliocene-Pleistocene) chronology: New perspec- search, v. 91, p. 14,192–14,206. open seaway allows for an increase in the east- tives in high-resolution stratigraphy: Geological Philander, S. G., and Pakanowski, R. C., 1986b, A ward component of the surface-water flow in the Society of America Bulletin, v. 107, p. 1272–1287. model of the seasonal cycle in the tropical Billups, K., Ravelo, A. C., and Zachos, J. C., 1998a, Atlantic Ocean: Journal of Geophysical Re- eastern Pacific and equatorial Atlantic, which Early Pliocene climate: A perspective from the search, v. 91, p. 14,212–14,220. would result in a decrease of the sea-surface tem- western equatorial Atlantic warm pool: Paleo- Ravelo, A. C., and Fairbanks, R., 1992, Oxygen iso- perature gradients across basins and between ceanography, v. 13, p. 459–470. topic composition of multiple species of plank- basins. The exchange of water between the west- Billups, K., Ravelo, A. C., and Zachos, J. C., 1998b, tonic foraminifera: Recorders of modern photic ern Atlantic and eastern Pacific may have reduced Early Pliocene deep-water circulation in the zone temperature gradient: Paleoceanography, western equatorial Atlantic: Implications for v. 7, p. 815–831. zonal temperature and pressure gradients relative high-latitude climate change: Paleoceanography, Ravelo, A. C., and Shackleton, N. J., 1995, Evidence for to the modern ocean, allowing the meridional v. 13, p. 84–95. surface water circulation changes at ODP Site 851 component of the trade winds to dominate and Cannariato, K. G., and Ravelo, A. C., 1997, Plio- in the eastern tropical Pacific, in Pisias, N. G., forcing the Intertropical Convergence Zone to a Pleistocene evolution of eastern tropical Pacific et al., Proceedings of the Ocean Drilling Program, surface-water circulation and thermocline depth: Scientific results, Volume 138: College Station, northward position. As Pacific to Atlantic tropical Paleoceanography, v. 12, p. 805–820. Texas, Ocean Drilling Program, p. 503–514. surface-water flow became more restricted, zonal Chaisson, W. P., 1995, Planktonic foraminiferal assem- Richardson, P. L., and Reverdin, G., 1987, Seasonal temperature and pressure gradients were estab- blages and paleoceanographic change in the cycle of velocity in the Atlantic North Equatorial lished, and the Intertropical Convergence Zone trans-tropical Pacific Ocean: A comparison of Current as measured from surface drifters, shifted southward toward its modern position. west (Leg 130) and east (Leg 138), latest current meters, and ship drifts: Journal of Geo- Miocene to Pleistocene, in Pisias, N. G., et al., physical Research, v. 92, p. 3691–3708. Alternatively, weakening of the Southern Proceedings of the Ocean Drilling Program, Sci- Shackleton, N. J., and Hall, M. A., 1997, The late Hemisphere pole-to-equator temperature gradient entific results, Volume 138: College Station, Miocene stable isotope records, Site 926, in and associated relaxation of southeasterly trade Texas, Ocean Drilling Program, p. 555–597. Shackleton, N. J., et al., Proceedings of the Ocean winds can be indirectly related to restricted flow Chaisson, W. P., and Ravelo, A. C., 1997, Changes in Drilling Program, Scientific results, Volume 138: upper water-column structure at Site 925, late College Station, Texas, Ocean Drilling Program, through the Central American seaway. Haug and Miocene–Pleistocene: Planktonic foraminifer p. 367–373. Tiedemann (1998) proposed an increase in the assemblage and isotopic evidence, in Shackleton, Shackleton, N. J., Crowhurst, S., Hagelberg, T., Pisias, relative flux of upper North Atlantic deep water N. J., et al., Proceedings of the Ocean Drilling Pro- N. G., and Schneider, D. A., 1995a, A new late to Caribbean Site 999 related to a closing seaway gram, Scientific results, Volume 154: College Sta- Neogene time scale: Application to Leg 138 sites, at 4.6 Ma. The Southern Ocean may have gradu- tion, Texas, Ocean Drilling Program, p. 255–268. in Pisias, N. G., et al., Proceedings of the Ocean Gordon, A. L., 1981, South Atlantic thermocline venti- Drilling Program, Scientific results, Volume 138: ally warmed in response to upwelling of rela- lation: Journal of Geophysical Research, v. 86, College Station, Texas, Ocean Drilling Program, tively warm North Atlantic deep water (e.g., p. 1239–1264. p. 73–101. Gordon, 1981), reducing pole to equator tem- Haug, G. H., and Tiedemann, R., 1998, Influence of Shackleton, N. J., Hall, M. A., and Pate, D., 1995b, perature gradients, weakening the southeasterly Panamanian Isthmus formation on Atlantic Pliocene stable isotope stratigraphy of Site 846, Ocean thermohaline circulation: Nature, v. 393, in Pisias, N. G., et al., Proceedings of the Ocean trade winds, and allowing a southward retreat of p. 673–675. Drilling Program, Scientific results, Volume 138: the Intertropical Convergence Zone. Hovan, S., 1995, Late Cenozoic atmospheric circula- College Station, Texas, Ocean Drilling Program, tion intensity and climate history recorded by p. 337–353. CONCLUSIONS eolian deposition in the eastern equatorial Pacific Stanley, S. M., 1995, New horizons for paleontology, We suggest that changes in the surface- and Ocean, Leg 138, in Pisias, N. G., et al., Proceed- with two examples: The rise and fall of the Creta- ings of the Ocean Drilling Program, Scientific ceous supertethys and the cause of the modern ice deep-water hydrography in the equatorial results, Volume 138: College Station, Texas, age: Journal of Paleontology, v. 69, p. 999–1007. Atlantic were strongly coupled between 4.4 and Ocean Drilling Program, p. 615–625. Tiedemann, R., and Franz, S.-O., 1997, Deep-water cir- 4.3 Ma. The link appears to be increased north- Keigwin, L. D., 1982, Isotopic paleoceanography of the culation, chemistry and terrigenous sediment sup- ward heat advection and subsequent thermo- Caribbean and East Pacific: Role of Panama uplift ply in the equatorial Atlantic during the Pliocene, in late Neogene time: Science, v. 217, p. 350–352. 3.3–2.6 Ma and 5.4.5 Ma, 1997, in Shackleton, haline overturn. Changes in the sea-surface hy- Levitus, S., Burgett, R., and Boyer, T. P., 1994, World N. J., et al., Proceedings of the Ocean Drilling Pro- drography from the four sites suggest that a ocean atlas: Washington, D.C., National Oceanic gram, Scientific results, Volume 154: College Sta- southward shift in the Intertropical Convergence and Atmospheric Administration, U.S. Depart- tion, Texas, Ocean Drilling Program, p. 299–318. Zone occurred at this time, possibly as a response ment of Commerce, 99 p. Voituriez, B., and Herbland, A., 1982, Primary produc- to a restriction in surface-water flow through the Mayer, L. A., Jansen, E., Backman, J., and Takayama, tion in the tropical Atlantic ocean mapped from T., 1993, Climate cyclicity at Site 806: The the oxygen values of Equalant 1 and 2 (1963): Central American seaway. We propose that the GRAPE record, in Berger, W. H., et al., Proceed- Bulletin of Marine Science, v. 31, p. 853–863. position of the Intertropical Convergence Zone, ings of the Ocean Drilling Program, Scientific and associated tropical Atlantic surface currents, results, Volume 130: College Station, Texas, Manuscript received August 17, 1998 played an important role in moderating north- Ocean Drilling Program, p. 623–639. Revised manuscript received December 1, 1998 Mikolajewicz, U., and Crowley, T., 1997, Response of a Manuscript accepted December 8, 1998 ward heat advection, high-latitude warmth, and coupled ocean/energy balance model to restricted thermohaline overturn during the early Pliocene.

322 Printed in U.S.A. GEOLOGY,April 1999