Scientific objectives, introduction, and background

Scientific objectives

The Pacific Equatorial Age Transect (PEAT) program (Fig. F1) was designed to achieve an age transect along the paleoequatorial Pacific spanning early Eocene through middle Miocene crustal basement ages, with Paleocene/Eocene and late Miocene to recent intervals being covered by previous Ocean Drilling Program (ODP) Legs 138 and 199 (Pisias, Mayer, Janecek, Palmer-Julson, and van Andel, 1995; Lyle, Wilson, Janecek, et al., 2002). Drill sites target specific time slices of interest (Fig. F2) at locations that provide optimum preservation of calcareous sediments (Figs. F1, F3, F4, F5, F6, F7). The overall aim was to obtain a continuous well-preserved sediment section that addresses the following primary scientific objectives:

  1. To detail the nature and changes of the CCD over the Cenozoic in the paleoequatorial Pacific;

  2. To determine the evolution of paleoproductivity of the equatorial Pacific over the Cenozoic;

  3. To validate and extend the astronomical calibration of the geological timescale for the Cenozoic, using orbitally forced variations in sediment composition known to occur in the equatorial Pacific, and to provide a fully integrated and astronomically calibrated bio-, chemo-, and magnetostratigraphy at the Equator;

  4. To determine temperature (sea surface and bottom water), nutrient profiles, and upper water column gradients;

  5. To better constrain Pacific plate tectonic motion and better locate the Cenozoic equatorial region in plate reconstructions; and

  6. To make use of the high level of correlation between tropical sedimentary sections and existing seismic stratigraphy to develop a more complete model of equatorial circulation and sedimentation.

Additional objectives include:

  1. To provide information about rapid biological evolution and turnover rates during times of climatic stress;

  2. To improve our knowledge of the reorganization of water masses as a function of depth and time, as our strategy also implies a paleodepth transect (Fig. F3);

  3. To develop a limited north–south transect across the paleoequator, caused by the northward offset of the proposed sites by Pacific plate motion, providing additional information about north–south hydrographic and biogeochemical gradients; and

  4. To obtain a transect of mid-ocean-ridge basalt (MORB) samples from a fixed location in the absolute mantle reference frame and to use a transect of basalt samples along the flow line that have been erupted in similar formation-water environments to study low-temperature alteration processes by seawater circulation.


As the world's largest ocean, the Pacific Ocean is intricately linked to major changes in the global climate system that took place during the Cenozoic. Throughout the Cenozoic, Pacific plate motion has had a northward component. The Pacific Ocean is unique in that the thick sediment bulge of biogenic-rich deposits from the currently narrow equatorial upwelling zone is slowly moving away from the Equator. Hence, older sections are not deeply buried and can be recovered by drilling. Previous Legs 138 and 199 were remarkably successful in giving us new insights into the workings of the climate and carbon system, productivity changes across the zone of divergence, time-dependent calcium carbonate dissolution, an integrated astronomically age-calibrated bio- and magnetostratigraphy, the location of the intertropical convergence zone (ITCZ), and evolutionary patterns for times of climatic change and upheaval. Together with older Deep Sea Drilling Project (DSDP) drilling in the eastern equatorial Pacific, both legs also helped delineate the position of the paleoequator and variations in sediment thickness from ~150°W to 110°W longitude.

Legs 138 and 199 were designed as latitudinal transects across the paleoequator in order to study changing patterns of sediment deposition across equatorial regions at critical time intervals. As we have gained more information about the past movement of plates and when in Earth's history "critical" climate events took place, it became possible to drill an age transect ("flow line") along the position of the paleoequator in the Pacific, targeting important time slices where calcareous sediments have been preserved best and the sedimentary archive in general allows us to reconstruct past climatic and tectonic conditions. Consequentially, the PEAT program will sharpen our understanding of extreme changes of the CCD across major geological boundaries during the last 53 m.y.

During most of the Paleogene the CCD was between 500 and 1300 m shallower than today. Thus, a very shallow CCD makes it difficult to obtain well-preserved sediments during these stratigraphic intervals because initial thermal subsidence of the ridge crest is rapid. Nevertheless, the careful coring and site location strategy of current PEAT Expedition 320/321 allowed us to drill the most promising sites and to obtain a unique sedimentary biogenic sediment archive for time periods just after the Paleocene/Eocene boundary event, the Eocene cooling, the Eocene–Oligocene transition, the "one cold pole" Oligocene, the Oligocene–Miocene transition, and the Miocene. These new cores and data will significantly contribute to the objectives of the IODP Extreme Climates Initiative and provide material that we were unable to recover during previous legs.


The circulation of the equatorial surface ocean is inescapably linked to the trade wind system. The equatorial Pacific is the classic "world ocean" example of this linkage. It is dominated by wind-driven circulation and is largely unfettered by ocean boundaries. Here, the Equator itself is characterized by a narrow zone of divergence that results from the change in the sign of the Coriolis effect and gives rise to a band of high biologic productivity (Fig. F6). The strength of the equatorial circulation and of this divergence is linked to the strength of the trade winds, which are in turn strongly tied to the global climate system. Variations in global climate, interhemispheric differences in temperature gradients, and marked changes in the ocean boundaries are all imprinted on the biogenic-rich sediments that are accumulating in the equatorial zone. The PEAT program was designed to provide an understanding of equatorial Pacific circulation, carbonate production, deposition, and dissolution for the last ~53 m.y. at a scale where orbital forcing can be resolved. Combined with seismic reflection data (Lyle et al., 2006, 2002) following in the vein of Mitchell et al. (2003) and synthesized with earlier drilling (e.g., Moore et al., 2002, 2004; Lyle et al., 2005; Moore, 2008a), we can reconstruct equatorial Pacific history with high confidence and substantially improve upon work from the early stages of DSDP and recent ODP legs.

Deciphering the sedimentary history of the equatorial Pacific has been greatly simplified by favorable motion of the Pacific plate. Throughout the Cenozoic, the movement of the Pacific plate has had a northward latitudinal component of ~0.25°/m.y. This northward movement transports equatorial sediments gradually out from under the zone of highest sediment delivery, resulting in a broad mound of biogenic sediments (Fig. F8). This transport prevents the older equatorial sections from being buried deeply beneath the younger sections as the crust moves northward. The diminished overburden resulting from this transport allows relatively good preservation of biogenic sediments and minimizes burial diagenesis. In addition, it allows us to core nearly all sections using the advanced piston corer (APC). The northward tectonic displacement, however, is not so large that a traverse of the equatorial zone (within 2° latitude of the Equator) was too rapid to record a reasonable period of equatorial ocean history. Drill sites typically remain within the equatorial zone for 10–20 m.y. before passing beyond the northern edge of high biogenic sedimentation. Older equatorial sections are thus buried beneath a thin veneer of younger sediments as the crust moves northwestward.

In his summary of DSDP results in the equatorial Pacific, van Andel (1975) gave a general view of the development of the equatorial mound of sediments in the Pacific Ocean, based mostly upon three early DSDP legs (5, 8, and 16). He showed how both temporal and spatial variation in sediment accumulation rates resulted from plate movement, varying biologic productivity at the equatorial divergence and carbonate preservation. The buildup of the Pacific equatorial mound of sediment has been more recently documented and discussed by Mitchell (1998) and Mitchell et al. (2003) (Fig. F8).

Drilling across the Pacific equatorial mound was addressed again some two decades after the van Andel (1975) compilation when an equatorial latitudinal transect along 10 Ma crust was drilled during Leg 138 (Pisias, Mayer, Janecek, Palmer-Julson, and van Andel, 1995) and then again a decade later when a similar transect along 56 Ma crust was conducted during Leg 199 (Lyle, Wilson, Janecek, et al., 2002). The newer drilling, coupled with major advances in geochronology, has documented the remarkable correlation of paleoceanographic events over thousands of kilometers in the equatorial Pacific, caused by the large scale of equatorial Pacific circulation (Fig. F9). It was thus possible during the PEAT program, with the addition of a relatively small number of new sites, to build detailed reconstructions of equatorial Pacific circulation throughout the Cenozoic.

Earlier drilling missed most of this detail because of the lack of important drilling technologies such as APC coring, multisensor track (MST) correlation, and core-log integration that allowed collection of relatively undisturbed sediments, rebuilding of a continuous sediment column from individual cores, and correlation to seismic reflection data. Together with an improved knowledge of the plate tectonic regime, these advances allow us to locate areas of enhanced depositional rates associated with the paleoequator. Combining multiple sites along the Equator will result in a detailed record from the Pleistocene to the Paleocene. These records will be invaluable for the continued development of the Cenozoic timescale as well as for the paleoceanographic information they contain.

Excellent sections were obtained during Legs 138 and 199, on which detailed orbital tuning of the geologic timescale has been carried out. These sections give a much clearer picture of variations in sedimentation rates, isotopic evolution of the oceans, biologic evolution and zoological provenance, variations in carbonate preservation, and variations in geochemical fluxes that result from paleoceanographic and paleoclimate changes. Parts of the Cenozoic timescale still require further refinement and verification of the proposed orbital tuning. The timescale older than the late Eocene has not yet been calibrated sufficiently, even though evidence of orbital frequencies appears in parts of the records recovered from this older interval (e.g., Norris and Röhl, 1999; Röhl et al., 2001).

To develop a detailed history of the Pacific equatorial current system, the strategy pursued during the most recent ODP legs (138 and 199) was to drill along a line of equal oceanic crustal age, thus obtaining an approximate north–south transect across the major east–west currents during time intervals of particular interest.

During the Paleocene and Eocene, the shallow CCD prevented deposition of carbonate except on crust at relatively shallow depths. Drilling near the paleoposition of the ridge crest at the critical time interval allows the recovery of the shallowest sections available in the pelagic oceans and thereby assures the best possible preservation of the carbonate sediments recovered. As the crust cools and sinks, the seafloor on which the sediments are deposited approaches the lysocline and CCD. Thus, the best preserved part of the sections recovered in such "time line" transects is restricted by the depth at which carbonate dissolution significantly increases, as well as by the northward movement of sediment sections out of the region of high equatorial productivity. This limitation was exemplified by the results from ODP Leg 199, which recovered only limited amounts of carbonate prior to the Eocene/Oligocene boundary (e.g., at ODP Site 1218 on 42 Ma crust).

For the PEAT program, we planned to overcome this limitation of the time line strategy by pursuing an equatorial age transect, or flow line strategy (Figs. F1, F3), to collect well-preserved equatorial sections through the Cenozoic while also making use of the Pacific plate motion to add an oblique latitudinal transect across all time slices.

We drilled a series of sites on the paleoequator at key intervals in the evolution of the Cenozoic climate. These intervals span the extremely warm times of the early Eocene, through the cooling of the late Eocene through Oligocene and the early Miocene time of relatively warm climates (or low ice volume), and into sections deposited during the development of the major Southern and Northern Hemisphere ice sheets (Fig. F2). There are very few previous drill sites that match our site selection criteria. Each site was selected to be close to the geographic paleoequator and on crust aged slightly older than the age intervals of particular interest.

In this way we will be able to track paleoceanographic conditions at the paleoequator in the best preserved sediments obtainable. We can also make use of the high level of correlation between tropical sediment sections and seismic stratigraphy to develop a more complete model of equatorial productivity and sedimentation.

Understanding the interplay between the CCD, CaCO3 dissolution, and productivity

The Pacific Ocean, specifically the equatorial upwelling zone, is the largest oceanic source of CO2 to the atmosphere and controls atmospheric CO2 levels (Dore et al., 2003). The release and uptake of CO2 is the direct consequence of calcium carbonate deposition and the interplay between nutrient supply, carbonate dissolution, surface water productivity, and export of biogenic carbonate from surface waters to the sediment pile. Distinguishing between the effects of carbonate dissolution and productivity has been a field of intense study in the past. An important objective of the PEAT program is to address the detailed workings of depth-dependent carbonate dissolution, which is intricately linked to the climate system and paleoceanography. In the standard model for carbonate dissolution, accumulation rates locally decrease linearly from a lysocline down to a CCD, reflecting a linearly increasing rate of dissolution. The depth of both of these mappable surfaces varies spatially and in time as a result of climatic and physical processes. The equatorial Pacific is one of the classic areas where the lysocline-CCD model was first developed, but little subsequent effort had been made to test it—a necessary step, considering that the functional form of dissolution is now known to depend in a more complex way on organic carbon burial and water mass properties. The age transect will provide the necessary additional data with which to test the carbonate paradigm and recover previously unavailable carbonate material from important Paleogene time slices in the Pacific.

Specifically, the recovery of shallowly buried carbonate sediments from near the paleoequatorial upwelling zone will contribute significantly toward separating the various processes that affect carbonate deposition and preservation and will reduce some of the processes that affect climatic proxy records, such as diagenetic recrystallization (Pearson et al., 2001). Neogene productivity has been strongly oriented parallel to the Equator, so differences in carbonate thicknesses at a common latitude but differing depths will permit the effect of dissolution to be isolated (Lyle, 2003; Mitchell et al., 2003; Mitchell and Lyle, 2005). In addition, the strategy adopted in this program will provide new data throughout the Cenozoic with which it will be possible to map the spatial evolution of the equatorial CCD with time (see "Results and highlights") because the northward component of the Pacific plate movement results in the multiple recovery of the same time slice in different sites but with a slightly different paleolatitude (Fig. F3).

Recovering more detailed records from the best possible material will also allow a better understanding of physical processes that might affect or hinder our interpretation of carbonate proxy records, such as the "carbonate ion effect"—an observed and modeled influence of carbonate ion concentration on stable isotope fractionation in carbonate (Spero et al., 1997; Zeebe and Wolf-Gladrow, 2001; Lear et al., 2008).

Preliminary work with seismic data (Mitchell et al., 2003) revealed a surprising lack of correlation between dissolution and depth in the westerly region of this study area. Our aim is to develop a more extensive three-dimensional (3-D) model for the stratigraphy of equatorial Pacific deposits that links all existing core data using a grid of high-resolution seismic reflection profiles, including more recent data from the PEAT site survey AMAT-03 (Lyle et al., 2006). A numerical stratigraphic model will then be used to assess carbonate dissolution and, in particular, the spatial pattern of sharp changes in dissolution, such as the extremely abrupt change in the CCD at the Oligocene/Eocene boundary, which has been linked to a possible abrupt onset of continental weathering. The sediment archive recovered during the PEAT program will allow the application of the substantial array of carbonate-based proxies with which the wider regional seismic study can be constrained and calibrated.

Reconstructing paleoceanographic properties and sea-surface temperature

A large number of paleoceanographic interpretations rely on obtaining proxy data such as stable isotope measurements, elemental ratios such as Mg/Ca, sea-surface temperature estimates from faunal distributions and isotope data, alkenone proxies, TEX86, geochemical productivity, and so on. In turn, a very large number of these measurements rely on the presence of biogenic calcium carbonate. For the Pacific Ocean, the PEAT drilling strategy was designed to recover this important material with the best possible preservation and the least amount of diagenetic effects for long intervals throughout the Cenozoic.

Spatial range considerations

In order to recover the best preserved and most complete carbonate record from the paleoequatorial Pacific, the age transect siting strategy necessarily implies a restricted north–south transect, even though the northward movement of the Pacific plate does allow us to recover identical time slices multiple times at different paleolatitudes (separated by several degrees). However, we note that the regional seismic study to be developed as part of our site survey work gives us the opportunity to integrate data from older drill sites with new drilling. The site survey linked the new sites to key drill sites from DSDP Legs 9 and 85 and ODP Legs 138 and 199. Combining data from these expeditions and surveys will allow us to construct a site-to-site correlation and, finally, a Pacific "megasplice" of high-resolution data spliced together to cover most of the Cenozoic.

Paleomagnetic objectives

One important aspect of the PEAT program is the recovery of high-quality paleomagnetic data so that attempts to improve existing geological timescales (Gradstein et al., 2004) can be extended further back in time. Results from Leg 199 demonstrate that these records can be recovered from near-equatorial carbonate sediments (e.g., Lanci et al., 2004, 2005). Almost all magnetic reversals from the Paleogene through the present were recovered during Leg 199. However, neither biogenic carbonate sediments through most of the Eocene nor for ages younger than the lower Miocene were recovered during Leg 199. Thus, although the paleomagnetic record during these times was of high quality, global stratigraphic correlation is hindered by the lower mass accumulation rate, the absence of a detailed isotope stratigraphy, and sparser biostratigraphic control. In order to facilitate the development of an integrated magneto- and biostratigraphic framework with a stable isotope stratigraphy (necessary to enable global correlation), recovery of magnetic reversals within carbonate sediment is desirable. In addition, further detailed paleomagnetic, magnetostratigrapic, and magnetic rock fabric data, most importantly from the Eocene, will help resolve the suggestion that the geographic Equator, as determined from the biogenic sediment bulge, might not coincide with the paleoequator position backtracked with a fixed-hotspot reference frame (Moore et al., 2004; Tarduno, 2003; Parés and Moore, 2005).

Ancillary benefits (MORB, basement)

Our drilling aimed to recover basement samples at all sites. A transect of MORB samples from a fixed location in the absolute mantle reference frame is a unique sample suite for mantle geochemists. A transect of basalt samples along the flow line that have erupted in similar environments will be of interest for low-temperature alteration studies (e.g., Elderfield and Schultz, 1996).

Site selection strategy and site targets

The time slices to be drilled during the PEAT program were chosen to cover the overall climatic history of the Cenozoic and target particular times of marked changes in the climatic regime. The spacing of the sites was determined by what we know of the Cenozoic evolution of the lysocline from previous drilling. Where the CCD is particularly shallow, the spacing in time of age transect sites needs to be closer than where the CCD is deep (Fig. F3). As a guide, Site 1218 was drilled on 42 Ma crust during times when the CCD was near 3.3 km. Nannofossil oozes were deposited at this location to ~37 Ma before the crust at this site sank below the CCD. An age separation between drill sites of 2–5 m.y. is a maximum for the shallow CCD of the Eocene; for good preservation of foraminifers even closer spacing should be used. The results of our paleoequator reconstruction and drill site locations are shown in Figure F1.

Site location strategy

In pursuing the history of the equatorial Pacific Ocean through both time line and flow line transects, we have two major advantages over the efforts that took place in the earlier days of scientific ocean drilling. Although previous drill sites targeted the general area, they mostly did not fulfill all of our criteria in terms of (1) a sufficient number of holes to obtain a continuous record, (2) modern coring technology to obtain undisturbed sediments, (3) location inside the paleoequatorial zone, or (4) location on the right crustal age to ensure the presence of calcium carbonate at the targeted time slice. We positioned Sites U1331 through U1338 somewhat south of the estimated paleoequatorial position at their target ages (Fig. F6) to maximize the time that the drill sites remained within the equatorial zone (i.e., ±2° of the equator), to allow for some error in positions (evidence suggests a southward bias of the equatorial sediment mound relative to the hotspot frame of reference [Knappenberger, 2000]), and to place the interval of maximum interest above the basal hydrothermal sediments.

To determine the site and site survey locations, we used the digital age grid of seafloor ages from Müller et al. (1997), heavily modified and improved with additional magnetic anomaly picks from Petronotis (1991) and Petronotis et al. (1994) as well as DSDP/ODP basement ages. For this grid, each point was backrotated in time to zero age, using the fixed-hotspot stage-poles from Koppers et al. (2001) and Engebretson et al. (1985) and the paleopole data from Sager and Pringle (1988). From the backtracked latitudes for each grid point we then obtained the paleoequator at the crustal age by contouring all backrotated latitudinal positions.

Eocene (Sites U1331–U1334)

The Eocene was a time of extremely warm climates that reached a global temperature maximum near 52 Ma, a period around the Early Eocene Climatic Optimum (EECO) (Fig. F2) (Zachos et al., 2001a; Shipboard Scientific Party, 2004). From this maximum there was gradual climatic cooling through the Eocene to the Eocene/Oligocene boundary. There appears to have been a slight reversal to this trend in the middle Eocene near 43 Ma and in the late Eocene at 34–36 Ma, just prior to the pronounced drop in oxygen isotopes that marks the Eocene/Oligocene boundary and one of the most dramatic changes of the CCD (Fig. F3).

Throughout the Eocene, the CCD lay near a depth of 3.2–3.3 km, albeit with potentially significant short-term fluctuations (Lyle et al., 2005). Thus, recovering well-preserved carbonate sediments from the equatorial region is a substantial challenge but not impossible if the depth of the East Pacific Rise lay near the global average of 2.7 km. We presently lack calcareous sediments from the region of the equatorial circulation system during this time of maximum Cenozoic warmth (Zachos et al., 2001a), elevated atmospheric pCO2 concentrations (Lowenstein and Demicco, 2007), and a shallow early Eocene CCD estimated between 3200 and 3300 m water depth (Lyle, Wilson, Janecek, et al., 2002; Lyle et al., 2005; Rea and Lyle, 2005). The Eocene equatorial upwelling system appears to differ from the modern equatorial upwelling regime by having strong secondary upwelling lobes ~10° in latitude away from the primary equatorial region (Figs. F10, F11). These lobes produced a much broader region of (relatively) high productivity than is present today.

Early and middle Eocene (Sites U1331 and U1332; ~53 and 50 Ma crust)

During Leg 199 a north–south transect was drilled across the equatorial region on ~56 Ma oceanic crust. Sites on this transect had generally drifted below the CCD by 52–53 Ma. Thus, we presently lack calcareous sediments from the region of the equatorial circulation system during the time of maximum Cenozoic warmth. Site U1331 is located on crust with an estimated age of ~53 Ma in order to intercept the interval between 53 and 50 Ma in basal carbonate sediments above the shallow early Eocene CCD (4200–4300 m). This interval was poorly sampled during Leg 199.

Average (noncarbonate) accumulation rates in the early Eocene were moderate, only slightly increasing in some of the more northern sites on the Leg 199 transect (Sites 1215 and 1220). What is particularly interesting in the Leg 199 records is that the very shallow CCD of this early Eocene time appears to deepen to the north, perhaps suggesting a northern source for the bottom waters. Sites targeting this time interval would ideally give us sediments with sufficient carbonate material to better constrain the isotopic and biotic characteristics of the near-surface equatorial waters.

During the early Eocene, a very shallow CCD and typical rapid tectonic plate subsidence of young crust near the shallow ridge crest conspired to make the time window during which carbonate is preserved short (~2–5 m.y.). Thus, although good records of pelagic carbonates during and just after the Paleocene/Eocene Thermal Maximum (PETM) were recovered during Leg 199 (Lyle, Wilson, Janecek, et al., 2002; Nunes and Norris, 2006; Raffi et al., 2005), the time period of the EECO (Zachos et al., 2001a) is not well sampled.

Sites U1331 (53 Ma crust) and U1332 (50 Ma crust) aim to provide the sedimentary archive to address several important questions that relate to causes and responses of the true Cenozoic "Greenhouse" world: the Eocene was a time of extremely warm climates that reached a maximum in temperatures near 52 Ma (Zachos et al., 2001a). From this maximum there was gradual climatic cooling to the Eocene/Oligocene boundary. Good paleomagnetic stratigraphy from Leg 199 sites allowed a much-improved calibration of nannofossil and radiolarian biostratigraphic datums throughout the Eocene. A north–south transect across the equatorial region at ~56 Ma was drilled during Leg 199. Although good records of pelagic carbonates during and just after the PETM were recovered at Leg 199 sites (Lyle, Wilson, Janecek, et al., 2002; Raffi et al., 2005; Nunes and Norris, 2006), the time period of the EECO (Zachos et al., 2001a) and the shallowest CCD is not well sampled.

In combination with Site U1331, which is located on 53 Ma crust, Site U1332 is located on crust with an expected age of ~50 Ma to intercept the interval between 50 and 48 Ma in biogenic sediments above the CCD. Thus, Site U1332 forms the second oldest time slice component of Expedition 320/321.

Middle and late Eocene (Site U1333; 46 Ma crust)

Good paleomagnetic stratigraphy at Leg 199 sites allowed a much improved calibration of nannofossil and radiolarian biostratigraphic datums (Pälike et al., 2005, 2006b; Raffi et al., 2005, 2006; Nigrini et al., 2006). From the combined information, a more detailed picture emerged of temporal variations in sediment accumulation through the middle and upper Eocene of the tropical Pacific. These data showed an increase of a factor of up to 2–3 in accumulation rates of siliceous ooze in the middle Eocene (41–45 Ma).

There are also several notable periods of highly fluctuating CCD associated with intervals in which carbonate is preserved as deep as 4000 m water depth, or ~700 m deeper than the average Eocene CCD (Lyle, Wilson, Janecek, et al., 2002; Lyle et al., 2005; Rea and Lyle, 2005; Bohaty et al., 2009). These fluctuations occur immediately prior to the Middle Eocene Climatic Optimum (MECO), which is associated with CCD shoaling (Bohaty and Zachos, 2003; Bohaty et al., 2009). Such fluctuations in the CCD are similar in magnitude to those at the Eocene/Oligocene boundary (Coxall et al., 2005). High siliceous sedimentation rates occur near an apparent short reversal in the middle Eocene cooling interval. It is difficult to interpret the cause of such a substantial change in silica flux during a very warm climatic regime.

The primary objective of coring at Site U1333 is to recover a complete sequence of carbonate sediments spanning the middle Eocene to Oligocene so we can evaluate changes in temperature and structure of the near-surface ocean, bottom water temperatures, and the evolution of the CCD.

One of the additional objectives of the PEAT program was to provide a depth transect for several Cenozoic key horizons, such as the Eocene–Oligocene transition (Coxall et al., 2005) targeted at Sites U1331–U1334. Site U1333 forms the third deepest paleodepth constraint, with an estimated crustal paleodepth of <4 km and a paleolatitude ~3° north of the paleoequator during the Eocene–Oligocene transition.

Eocene/Oligocene Boundary (Site U1334; 38 Ma crust)

Site U1334 sediments are estimated to have been deposited on top of late middle Eocene crust with an age of ~38 Ma, and this site targets the events bracketing the Eocene–Oligocene transition with the specific aim of recovering carbonate-bearing sediments of latest Eocene age prior to a large deepening of the CCD that occurred during this Greenhouse-to-Icehouse transition (Kennett and Shackleton, 1976; Miller et al., 1991; Zachos et al., 1996; Coxall et al., 2005). The Eocene–Oligocene transition experienced the most dramatic deepening of the Pacific CCD during the Paleogene (van Andel, 1975), which has now been shown by Coxall et al. (2005) to coincide with a rapid stepwise increase in benthic oxygen stable isotope ratios interpreted to reflect a combination of growth of the Antarctic ice sheet and decrease in deepwater temperatures (DeConto et al., 2008; Liu et al., 2009).

So far the most complete Eocene/Oligocene boundary section recovered from the equatorial Pacific has been Site 1218 on 42 Ma crust; however, it is far from pristine. Carbonate percentages drop markedly below the boundary and reach zero near 34 Ma during a time of apparent global shoaling of the CCD just prior to the Eocene–Oligocene transition and CCD deepening (Bohaty et al., 2008). The absence of carbonate in this interval prevented the recovery of information about paleoceanographic conditions prior to the Eocene–Oligocene transition and also has implications for the interpretation of paleotemperature proxies such as Mg/Ca ratios in foraminifer shells that were bathed in waters with very low carbonate ion concentrations (Lear et al., 2008; Elderfield et al., 2006). The integrated stratigraphy from Site 1218 has been correlated to the planktonic foraminifer marker extinction of the genus Hantkenina in exceptionally well preserved shallow clay-rich sediments from Tanzania by Pearson et al. (2008), who demonstrated that the Eocene/Oligocene boundary falls within the middle plateau of the stable isotope double-step described by Coxall et al. (2005) just prior to the base of Chron C13n.

Data from Site 1218 allowed astronomical time calibration of the entire Oligocene (Coxall et al., 2005; Wade and Pälike, 2004; Pälike et al., 2006b), but the lack of carbonate in the uppermost Eocene at this site made the detailed time control now available for the Oligocene much less certain for the late Eocene. Site U1334 is located on crust with a basement age of ~38 Ma and crossed the paleoequator shortly thereafter. It was located to provide the missing information about the crucial chain of events prior to and during the Eocene–Oligocene transition.

Oligocene (Site U1336; ~32 Ma crust)

Site U1336 targets the Oligocene and is located on lower Oligocene crust. This interval of time is noted for its markedly heavy benthic oxygen isotopes (Fig. F2) and its relatively deep CCD (Fig. F3). There was probably ice on Antarctica during this interval, but not the large ice sheets found there later in the middle Miocene. There is no compelling evidence for ice sheets in the Northern Hemisphere during the Oligocene and early Miocene. Thus, there was apparently a time of low global ice volume, cold bottom waters, a cold South Pole, and a relatively warm North Pole. This scenario of a "one cold pole" world has given rise to speculation on the impact of interhemispheric temperature imbalance on pole-to-Equator temperature gradients and on the symmetry of the global wind systems. The extent to which such an imbalance may have affected the trade winds, the position of the ITCZ, and the seasonal shifts in this zone should be seen in the wind-driven currents of the equatorial region.

The older low-resolution DSDP data indicate relatively high but variable sediment accumulation rates during this interval and better carbonate preservation south of the Equator (van Andel, 1975). The highest accumulation rates encountered (>15 m/m.y.) from the Leg 199 transect occurred in the lower part of the Oligocene, but these were at sites north of the Oligocene Equator or on relatively old (and therefore deep) crust. Thus we expected a better preserved, thicker carbonate section at the Oligocene Equator. Studies of Oligocene sections from Leg 199 and other ODP sites (e.g., Paul et al., 2000; Zachos et al., 2001a; Billups et al., 2004; Pälike et al., 2006a) indicate the presence of strong eccentricity and obliquity cycles in carbonate preservation and suggest a strong (southern) high-latitude influence on the carbonate record. These cycles are leading to the development of an orbitally tuned timescale that reaches back to the base of the Oligocene (Pälike et al., 2006b). Such a timescale will make it possible to develop a detailed picture of equatorial geochemical fluxes and of the degree of variability in the equatorial system of the Oligocene.

Latest Oligocene–earliest Miocene (Site U1335; 26 Ma crust)

Site U1335 was designed to focus on paleoceanographic events in the late Oligocene and into the early and middle Miocene, including the climatically significant Oligocene–Miocene transition and its recovery. In conjunction with Sites U1336 and U1337, Site U1335 was also designed to provide a latitudinal transect for early Miocene age slices. A significant several million year long rise in the oxygen isotope record (Lear et al., 2004; Pälike et al., 2006b) at the end of the Oligocene is closely followed by a relatively short, sharp increase in oxygen isotope values. This increase has been interpreted as a major glacial episode (Mi-1) (Fig. F2) (Paul et al., 2000; Zachos et al. 1997, 2001a, 2001b; Pälike et al., 2006a) and correlated to a pronounced drop in sea level (Miller et al., 1991). The Mi-1 event is very close to the Oligocene/Miocene boundary and has now been astronomically age calibrated in several ocean basins (Shackleton et al., 2000; Billups et al., 2004). Although there are clear periodic isotopic signals indicating major changes in ice volume, ocean temperatures, and/or ocean structure, this biostratigraphic boundary has always been somewhat of an enigma. Unlike the major changes in the isotopic stratigraphy, the biostratigraphies of the planktonic microfossils show very little change at all across this boundary. In fact, it is one of the most difficult epoch boundaries to pick using solely microfossil biostratigraphy.

At Leg 199 Sites 1218 and 1219 this interval was well recovered; however, carbonate preservation still presented a problem for foraminifer stratigraphy. Both sites were deep and well within the lysocline, making the application of temperature proxies such as Mg/Ca ratios in foraminifer tests more difficult (Lear et al., 2008). At the time Miocene-Oligocene sediments were deposited, Site 1218 already resided on 18 Ma crust and was ~4100 m deep. Site 1219 was on ~34 Ma crust and was ~4500 m deep. There was a relative increase in the large diatoms near this boundary in the siliceous coarse fraction, suggesting increased productivity; however, detailed high-resolution flux rates across this interval have yet to be determined. A well-recovered section on the latest Oligocene Equator near the late Oligocene ridge crest was targeted at Site U1335 and should provide both the resolution and the preservation required to better describe the changes in the equatorial ocean taking place at this time.

Miocene (Site U1337; 24 Ma crust)

Site U1337 was proposed for drilling to focus on the paleoceanographic events in the early and middle Miocene. The latest Oligocene through the middle Miocene appears to have been a time of relative warmth comparable to the latest Eocene. However, the short-term variability in the isotopic record of the early to middle Miocene is larger than that of the Eocene and may indicate more variability in climate and in global ice volume. The climatic "optimum" at ~15 Ma comes just before the major development of ice sheets on Antarctica and the marked increase in ice-rafted debris in circum-Antarctic sediments. The early Miocene also marked a major evolutionary change from the relatively static Oligocene planktonic biota. In the equatorial Pacific, the early Miocene marked the beginning of abundant diatoms in the stratigraphic record (J. Barron, pers. comm., 2003) and thus may represent a major change in carbon cycling as well.

The only major ocean boundary change proposed for the time near the Oligocene/Miocene boundary was the opening of the Drake Passage to deep flow; however, there is some debate as to the exact timing of this event (Barker, 2001; Pagani et al., 1999; Lawver and Gahagan, 2003; Scher and Martin, 2006; Lyle et al., 2007) and its direct impact on the tropical ocean is uncertain. It may be that, as in the Eocene/Oligocene boundary section, the link lies in the shallow intermediate waters that provide nutrients to lower latitude upwelling regions. For the equatorial region, an even more pertinent question is "What changes were occurring in the Miocene tropical ocean that led to this burst of Miocene evolution?"

Middle Miocene (Site U1338; ~18 Ma crust)

In principle, the age transect strategy of this proposal would not be complete without data from the Pliocene–Pleistocene. However, in addition to limitations of cruise length, near-paleoequatorial records have already been targeted by ODP Legs 138 (Pisias, Mayer, Janecek, Palmer-Julson, and van Andel, 1995) and 202 (Mix, Tiedemann, Blum, et al., 2003); these records provide information about the development of Northern Hemisphere glaciation. Our last site (U1338) focuses instead on the interesting events following a middle Miocene maximum in deposition (van Andel, 1975).

Site U1338 was proposed for drilling to focus on the paleoceanographic events following a middle Miocene maximum in deposition (van Andel, 1975). In addition, large changes in the state and frequency of continental glaciation have recently been described within the middle Miocene (after 14 Ma) (Holbourn et al., 2005; Abels et al., 2005; Raffi et al., 2006). There is a wide latitude range of CaCO3 deposition during the earliest Neogene, with a relatively sharp transition to a narrower CaCO3 belt after 20 Ma (Lyle, 2003). CaCO3 mass accumulation rates in the central equatorial Pacific recovered from the 18–19 Ma "famine" and in the period between 14 and 16 Ma reached a second maximum in carbonate deposition, which is also evident in the seismic stratigraphy of the equatorial sediment bulge (Knappenberger, 2000; Mitchell et al., 2003). We designed Site U1338 to recover an equatorial record at the early middle Miocene sedimentation maximum.