A wide range of biotic observations suggest that substantially higher mid-latitude and polar temperatures relative to today prevailed during certain intervals of Earth history (e.g., mid-Cretaceous and early Paleogene), with tropical temperatures throughout the past ~150 m.y. probably at least as warm as today (Adams et al., 1990; Crowley and North, 1991). ¹⁸O paleothermometry in deep-sea foraminiferal calcite supports the existence of these past warm climates (Fig. F9). These data show that deep and surface waters in the Cretaceous Antarctic during these intervals were significantly warmer than today (e.g., ~15°C for sea-surface temperatures [SSTs]) (Huber et al., 1995). In contrast, broadly contemporaneous SSTs estimated in this way for the tropics are generally no warmer and sometimes much cooler (mininum of ~12° to 18°C) than today (Shackleton, 1984; Barrera, 1994; D'Hondt and Arthur, 1996). Such cool tropical SSTs contradict not only biotic observations but also basic theories of tropical ocean-atmosphere dynamics (Crowley, 1991). Attempts to simulate Cretaceous climates using numerical general circulation models (GCMs) have consistently demonstrated that (1) high levels of atmospheric CO₂ (four times present) are needed to explain the warm polar SSTs derived from ¹⁸O paleothermometry, and (2) this level of greenhouse forcing also yields increases in tropical SSTs beyond those indicated by ¹⁸O data sets (e.g., Manabe and Bryan, 1985; Barron, 1995; Bush and Philander, 1997; Poulsen et al., 1999). Explanations for the apparent paradox of the cool-tropical greenhouse fall into two basic categories: (1) models of past warm climates fail to account adequately for polar ocean and/or atmospheric heat transport, and (2) tropical ¹⁸O SST estimates are misleading (Zachos et al., 1994; Crowley and Zachos, 2000).

Many artifacts plague existing records of tropical SST, including their extremely low resolution, misidentification of true surface-dwelling species of foraminifers, and the susceptibility of epipelagically secreted calcite to early diagenetic alteration in favor of artificially low SSTs (Douglas and Savin, 1975; Killingley, 1983; Schrag et al., 1995). Recent studies demonstrate that ancient carbonates (even highly metastable minerals) can be remarkably well preserved and yield ¹⁸O SSTs for the tropics that are significantly warmer than those provided by diagenetically suspect material (Pearson and Shackleton, 1995; Wilson and Opdyke, 1996; Norris and Wilson, 1998). These studies show that foraminifers recovered from sections with shallow burial depths and/or clay-rich lithologies display excellent textural preservation and include epipelagic fauna yielding tropical ¹⁸O SSTs that match or, in some cases, exceed those measured today, thereby suggesting a thermal response to greenhouse forcing in the tropics.

The concept of a greenhouse mid- to Late Cretaceous period is well supported by models of Earth's tectonic history. These models indicate that the mid- to Late Cretaceous was a time of exceptional rates of seafloor spreading and intraplate volcanism. This pulse in global oceanic crustal production is hypothesized to have caused increases in the levels of atmospheric carbon dioxide and global sea levels via increases in global oceanic ridge volumes, magmatic outgassing, and metamorphic decarbonation reactions (Schlanger et al., 1981; Larson, 1991; Berner, 1994). Fundamental problems, however, remain in terms of our understanding of these Cretaceous environments and their Paleogene equivalents. For example, the timing of maximum rates of crustal cycling and inferred carbon dioxide levels significantly pre-dates Cretaceous climatic optima as perceived from existing paleothermometric records (e.g., Larson, 1991; Clarke and Jenkyns, 1999). This discrepancy suggests that additional factors to the geochemical carbon cycle may play an important role in determining Cretaceous climate (e.g., the hydrologic cycle). Furthermore, new high-resolution bulk carbonate ¹⁸O records from classic land sections in Italy reveal large positive excursions that have been interpreted in terms of short glaciations superimposed on the middle of the Cretaceous greenhouse (Stoll and Schrag, 2000). Sedimentological and biotic records show no support for this hypothesis, but these records are of insufficient temporal resolution to provide a categorical test. Similarly, our highest-resolution long-term ¹⁸O record from deep-sea sites for the Cretaceous comes from diagenetically altered bulk carbonate and has a temporal resolution of ~1 sample/200 k.y. (Clarke and Jenkyns, 1999). The best existing corresponding record from separates of planktonic foraminiferal calcite is also diagenetically suspect (it comes from chertified deeply buried chalks in the Pacific) and is of very low resolution (<1 sample/m.y.; Aptian through Santonian) (Barrera, 1994).

Results from DSDP Site 144 indicate that Demerara Rise sediments contain well-preserved microfossils of mid-Cretaceous to Oligocene age. The following scientific questions will be addressed using well-preserved microfossils from the Demerara Rise transect:

1. What is the history of changes in atmospheric CO₂ levels from mid-Cretaceous to Paleogene time? Well-preserved microfossils from Demerara Rise would provide an ideal means to evaluate this question using exciting new proxies (e.g., Pearson and Palmer 1999).
2. What is the history of tropical SSTs in the tropical Atlantic? The presence of Demerara Rise within the core of the tropics throughout the entire Cretaceous and Paleogene provides a way to evaluate the relative strength of greenhouse forcing over long time periods. It is important to establish whether the persistent problem of tropical overheating in simulations of past warm climates is an artifact of poor SST records or the result of the existence of some tropical thermostatic regulator.
3. What evidence is there for rapid ocean warming associated with extreme perturbations in the geochemical carbon cycle (e.g., Cretaceous OAEs and LPTM)? High-resolution records from Demerara Rise across these events would also provide a way to test the hypothesis that C_org burial during OAEs acted as a negative feedback for global warming.
4. Are hypothesized mid-Cretaceous glaciations real? Answers to this question have important consequences to (a) the long-standing problem of the mechanism responsible for perceived changes in global sea level prior to the icehouse and (b) our understanding of the stability of greenhouse climates.

The Paleogene record is rife with critical boundaries that offer significant opportunities for understanding the dynamics of greenhouse gas release, warm climate stability, biotic turnover associated with climate transitions, and extraterrestrial impacts. For example, the early Eocene warm period (~50–53 Ma) is the most extreme interval of global warming in the past ~80 m.y., but little is known about the number of hyperthermals within it, the range of temperatures, or their effects on biotic evolution (Thomas and Zachos, 1999). The Eocene warm period is succeeded by a long shift toward the lower temperatures and increased ice buildup of the late Eocene and Oligocene (Fig. F9), whose history and consequences for ocean circulation, carbon cycling, and biotic evolution are only vaguely understood. Finally, extraterrestrial impacts in the early middle Eocene and the late Eocene offer the opportunity to study the climatological and biotic effects of impacts that were too small to precipitate global mass extinctions but were large enough to have engendered global changes in climate. Below we discuss two events that have a particularly good likelihood of being preserved on Demerara Rise.

Late Paleocene Thermal Maximum

The transient global warming near the end of the Paleocene is one of the best candidates for greenhouse warming in the geologic record. A growing body of evidence implicates a massive release of greenhouse gases into the atmosphere and ocean as a cause for ~5°–7°C warming in the Southern Ocean and subtropics, a 35%–50% extinction of deep-sea benthic foraminifers, and widespread carbonate dissolution in the deep oceans record (e.g., Zachos et al. 1993; Koch et al. 1995; Dickens et al. 1997). Recent studies utilizing high-resolution stable isotope analyses (Bains et al. 1999) and orbitally tuned chronologies (Norris and Röhl 1999) suggest that carbon release occurred in a series of short steps (lasting a few thousand years) punctuated by catastrophic shifts in ¹³C and ocean temperature. Although these new data support the idea that the carbon may have been sourced from methane hydrate reservoirs, considerable uncertainty remains about how the carbon was released, what triggered the different phases of release, and what the biotic and climatological response to the input of large amounts of greenhouse gas.

It is unknown whether there are complete records of the late Paleocene on Demerara Rise, but there are good indications that a well-preserved LPTM could be recovered. DSDP 144 spot cored through 50 m of late Paleocene calcareous ooze. A nearly 40-m coring gap between the upper Paleocene and middle Eocene oozes allows for up to ~90 m of Paleocene strata at DSDP Site 144. DSDP 144 was drilled where the Paleogene section is greatly condensed near the northern escarpment of Demerara Rise, but a relatively expanded (~300–400 m thick) sequence of Paleocene and Eocene chalk and ooze is present inboard of the escarpment. Furthermore, seismic data suggest that Paleogene sediment drifts exist along the northern escarpment where we propose to drill.

Eocene–Oligocene Transition

The Eocene/Oligocene boundary represents an important point in the transition from the greenhouse world of the Cretaceous and early Paleogene into the late Paleogene–Neogene icehouse. The Eocene–Oligocene transition appears to record a dramatic growth in Antarctic ice sheets, but the rarity of complete sections across the boundary have limited our understanding of the dynamics of this important step in to the modern icehouse world (Miller et al., 1991). This transition is marked by a large rapid increase in the in benthic foraminiferal calcite ¹⁸O record in earliest Oligocene time (Oi-1) (Fig. F9). This excursion was first ascribed to a 5°C temperature drop associated with the onset of thermohaline circulation, but more recently, Oi-1 has been associated with the onset of continental ice accumulation on Antarctica. This conclusion is in accord with a recent Mg/Ca paleothermometry record developed for benthic foraminifers, which shows no significant change corresponding to Oi-1 (Lear et al., 2000). On this basis it has been hypothesized that a lack of moisture available for snow precipitation on Antarctica rather than excess warmth prevented ice accumulation immediately prior to this time. However, our most complete records of the Eocene/Oligocene boundary come from mid-latitude sites (DSDP 522 and ODP 744). Low-latitude records of the Eocene–Oligocene transition and the Oligocene–Miocene transition are needed to fully evaluate the competing roles played by global cooling and ice growth in the transition from the Cretaceous greenhouse into the Neogene icehouse.

The Demerara Rise is known to have expanded records of the early Oligocene to early Miocene at shallow burial depth. A widespread network of submarine channels on the outer Demerara Rise is of early Miocene age. Oligocene strata were recovered in both DSDP 144 and industry well Demerara A2-1. Dating the channeled surface, which does not cut deeply into underlying strata, is important to understand the transient change in climate and sediment supply that produced the submarine channels. The presence of lower Oligocene ooze in DSDP 144 suggests that a much more expanded record of the Oligocene, and possibly the upper Eocene, may be present in the Paleogene sediment drifts present along the northern escarpment.

The opening of the equatorial Atlantic gateway was driven by the separation of Africa and South America and is widely hypothesized to have had a significant effect on both oceanic circulation patterns and heat transport over wide areas of the Cretaceous Atlantic. Yet, the timing of the opening of this gateway remains poorly constrained. Based on the biogeographic distribution of foraminifers and cephalopods, a shallow-water passage is thought to have been initiated between the North and South Atlantic oceans at some time during the Albian (Moullade and Guerin, 1982; Förster, 1978; Wiedmann and Neugebauer, 1978; Moullade et al., 1993). Results from ODP Leg 159 on the eastern side of the equatorial Atlantic gateway suggest that a strong relationship existed between stepwise deepening and widening of the gateway and black shale deposition on the west African margin from the Albian to the Turonian (Wagner and Pletsch, 1999). Cessation of black shale deposition in the Upper Cretaceous is interpreted to result from increasingly vigorous circulation between the North and South Atlantic, hence marking the transition from a Mesozoic longitudinal circulation system through the Tethyian and the central Atlantic oceans to a more Cenozoic-like oxidizing latitudinal circulation pattern through the Atlantic gateway.

Analysis of the subsidence history of Demerara Rise will contribute to interpretations of the history of the opening of the equatorial Atlantic gateway. High-resolution sampling of distinct time slices across a range of paleowater depths will help to constrain the following questions:

1. What were the timings of the establishment of oceanographically significant through-flow in the equatorial Atlantic gateway (i.e., the onset of through-flow of upper intermediate and deeper water masses)? Results will help to determine the paleoceanographic consequences of connecting the previously restricted South Atlantic to the North Atlantic–Tethyian realm.
2. What was the specific role played by the equatorial Atlantic gateway in controlling the development of Cretaceous black shale deposition? By comparing the high-resolution multiple OAE records from Demerara Rise with earlier DSDP mapping it will be possible to evaluate whether this gateway merely controlled OAE sedimentation in the more restricted South Atlantic or whether its influence extended to the tropical North Atlantic–Tethys.
3. What is the long-term history of cross-equatorial heat transport into the North Atlantic? The Demerara Rise is positioned in an ideal location to sample meriodonal circulation and delivery of heat northward from Cretaceous to Oligocene time (Fig. F1).