Sediments on Shatsky Rise show that deep ocean circulation fluctuated in a highly regular fashion in the Paleogene. The record from Sites 1209, 1210, 1211, and 1212 is strongly cyclic as a result of regular changes in the properties of the deep waters bathing Shatsky Rise. Decimeter-scale alternations in color that correspond to minor changes in clay and carbonate content are relatively faint in cores but striking in color reflectance records. Magnetic susceptibility records also show these fluctuations clearly (Fig. F46). The cycles occur in bundles that can be easily correlated between sites. The cycles are likely produced by minor changes in the dissolution of carbonate in the deep ocean. Variation in the flux of eolian material may have also played a minor role.

Preliminary biostratigraphy suggests that the frequency of cycles corresponds to orbital periodicities. The predominant frequencies appear to be short- (100 k.y.) and long- (400 k.y.) period eccentricity. However, the nature of cycles varies among sites and through time in subtle ways. The amplitude of the cycles does not appear to correlate well with paleodepth. In fact, the two shallowest sites, Sites 1209 and 1210, show the highest amplitude cycles for significant periods of time. Magnetic susceptibility and color reflectance data indicate that the two shallowest sites with the most expanded Paleogene records also have the most cycles. This suggests that the two deeper sites, Sites 1211 and 1212, have a number of condensed intervals and/or diastems.

The lack of a clear correlation of cycle amplitude with depth suggests that the cycles are not the result of a simple dissolutional scenario in which deep waters were more corrosive. If this relationship held, the deepest site should have the largest amplitude fluctuations between the dissolved and less dissolved end-members. Instead, the cycles may represent repeated shoaling and deepening of the boundary between two water masses, with the boundary lying close to the depths of the shallower two sites (Sites 1209 and 1210) at some times and in deeper waters close to Site 1211 at others. This situation would produce higher-amplitude chemical and physical fluctuations (i.e., corrosiveness, and oxygenation) at depths close to the water mass boundaries than in shallower and deeper water where conditions were less variable. Alternatively, the cycles might represent variation in surface-water productivity that produced subtle changes in chemical composition. For instance, the shallower sites might have had more variable productivity for a number of reasons. Additional compositional and isotopic data are required to shed further light on the origin of Paleogene cycles.

At this preliminary stage, the cyclic Paleogene record has helped considerably in the construction of composite sections. These composites show that multiple coring at the Southern High sites has completely sampled the Paleogene sedimentary section (see "Physical Properties, Downhole Measurements, and Core Logging" in "Specialty Syntheses"). Furthermore, with the exception of a gap close to the Oligocene/Miocene boundary, the combination of Sites 1209, 1210, 1211, and 1212 of Leg 198 has recovered a composite section from the Pleistocene down to below the K/T boundary.

Sediments cored on Shatsky Rise show evidence of a strong deep-ocean response to warming in the LPTM. The LPTM interval was cored at four sites on the Southern High (Sites 1209, 1210, 1211, and 1212). Double and triple coring at these sites recovered a total of 10 separate LPTM records: three each from Sites 1209 and 1211, and two each from Sites 1210 and 1212 (Fig. F54). The Paleocene–Eocene boundary interval was also recovered at Site 1208 on the Central High. The range of present depths, from 2387 m at Site 1209 to 2907 m at Site 1211 provide a 520-m depth transect to observe the sedimentary response to this abrupt warming event as a function of depth. Although the Site 1208 sequence is highly condensed and it is not currently possible to determine whether the LPTM is present, some inferences can be made from this section that extend the transect some 440 m to 960 m.

At the Southern High sites, the LPTM corresponds to an 8- to 23-cm-thick layer of clayey nannofossil ooze with a sharp base and a gradational upper contact. The clay-rich layer is generally yellowish brown in color and is often bioturbated into the underlying sediment. At several sites an extremely thin (1 mm) dark brown clay seam lies at the base of the LPTM. Carbonate contents have been measured in detail across the LPTM at Site 1210. These data show a decrease from 96 to 89 wt% CaCO₃ at the base of the event, a decrease that would involve a substantial increase in dissolution. Color reflectance and magnetic susceptibility data (Figs. F39, F46) allow detailed correlation between holes and sites in this time interval.

Preliminary biostratigraphic investigations show that the event lies toward the top of nannofossil Zone CP8 and planktonic foraminiferal Zone P5. At several sites, rare specimens of Gavelinella beccariiformis, a benthic foraminiferal species that goes extinct at the onset of the LPTM (i.e., Thomas, 1990) were found below the event. The abrupt decrease in the nannofossil Fasciculithus that occurs just above the LPTM in other sections (Bralower et al., 1995; Aubry et al., 1996; Monechi et al., 2000) lies near the top of the clay-rich layer. The stratigraphic level of the top of the LPTM is currently undefined, but the FO of the nannofossil Discoaster diastypus and the LO of the plantkonic foraminifer Morozovella velascoensis indicate that the record in all sections is apparently complete, with the exception of Hole 1211B. This biostratigraphy shows that the LPTM interval at Sites 1209, 1210, and 1212 are condensed compared to continental margin records from the Atlantic and Tethys (e.g., Kennett and Stott, 1991; Norris and Röhl, 1999) but somewhat expanded compared to other deep-sea sites such as Site 865 on Allison Guyot (Bralower et al., 1995) and Site 527 on Walvis Ridge (Thomas et al., 1999). At Site 1211, the LPTM interval was recovered across the break between two sections in Holes 1211A and 1211C. The clay-rich bed is more prominent and condensed than in the other records (Fig. F54). In Hole 1211B, the basal clay seam appears to be present, but the occurrence of D. diastypus 1–2 cm above this level and the concurrent decline in the abundance of Fasciculithus strongly suggests an unconformity. The LPTM in both holes at Site 1212 appears complete; however, the presence of Globanomalina pseudomenardii suggests a slight unconformity immediately below the event.

The LPTM interval at all of the sites contains a clear record of nannofossil and planktonic foraminiferal assemblage transformation at this time of major environmental upheaval (Fig. F22). One of the dominant nannolith genera, Fasciculithus, is replaced by Zygrhablithus bijugatus, a nannolith that is often a highly abundant or dominant component of Eocene assemblages. The genus Discoaster is highly abundant, likely as a result of warming or increased oligotrophy (Bralower, unpubl. data). Calcispheres, possible resting cysts produced by calcareous dinoflagellates at times of environmental instability, are also found. Planktonic foraminiferal assemblages within the clay-rich interval contain an ephemeral group of ecophenotypes or short-lived species of the genera Acarinina and Morozovella (Kelly et al., 1996). These "excursion" taxa include both end-member as well as transitional morphologies.

The depth transect strategy of Leg 198 was specifically designed to address the response of the ocean to the greenhouse forcing mechanism proposed for the LPTM. This warming is generally thought to have resulted from input of a massive burst of methane into the ocean-atmosphere system (e.g., Dickens et al., 1997). Methane is the only agent that can explain both the warming and the rate of carbon isotopic change at the onset of the excursion. The oceanic response to this methane input is predictable but currently untested (e.g., Dickens, 2000). Regardless of how the transfer to the ocean took place, oxidation of methane would generate CO₂, which would lower the dissolved CO₃^2– content of seawater and cause a dramatic shoaling in the depth to the lysocline and CCD. This response should be recorded in changes in carbonate content and preservation in all marine sections. Over a depth range, shallow sections should show less change in dissolution and carbonate content than deep sections.

Nannofossil preservation below the event in all of the Southern High sites is moderate indicating that the sites were located in the broad range of the lysocline. All sites show a general deterioration in preservation at the onset of the event and abundant 10- to 20-crystals that are thought to have been derived from precipitation of dissolved carbonate are found in smear slides. The detailed response of fossil preservation is complex and different from site to site, as is the absolute change in carbonate content as indicated by reflectance data.

The general changes in lithology suggest a transition from paleodepths in the shallower sites that were less sensitive to changes in carbonate solubility in the deep ocean (Sites 1209, 1210, 1212) to those that were at depth ranges highly sensitive to changes across the LPTM (Sites 1208 and 1211) (Fig. F54). The decrease in carbonate content and deterioration in nannofossil preservation are evidence for an abrupt rise in the level of the CCD and lysocline during the LPTM. Determination of the magnitude of this change awaits postcruise analysis; however, the preliminary lithostratigraphic and biostratigraphic results from the LPTM interval in the Shatsky Rise depth transect are highly consistent and thus support the expected ocean response to massive methane input.

Sediments recording the response of the tropical Pacific ocean to cooling in the Eocene–Oligocene transition were recovered across a large depth range on Shatsky Rise. The Eocene–Oligocene transition was cored at Sites 1208, 1209, 1210, and 1211 in a total of nine holes. The boundary between the two epochs is identified by the LO of the planktonic foraminiferal genus Hantkenina at 33.7 Ma. Preliminary nannofossil and planktonic foraminiferal biostratigraphy suggests that the boundary interval is complete. At the Southern High sites (Sites 1209, 1210, and 1211), the transition records a gradual, subtle but distinctive change, over a 4- to 7.5-m interval in the uppermost Eocene and lowermost Oligocene, from light brown to tan nannofossil ooze with clay to a light gray to white nannofossil ooze (Fig. F32). This transition is associated with marked changes in color reflectance data (Fig. F28) but fairly minor changes in percent carbonate from about 90 to 96 wt% (in Hole 1210A). Superimposed on the gradual change in color and increase in carbonate contents are marked cycles that appear to represent orbital rhythms (see "Physical Properties, Downhole Measurements, and Core Logging" in "Specialty Syntheses").

The lithologic record of the Eocene–Oligocene transition at Site 1208 on the Central Rise is markedly different from the Southern High sites. A lithologic transition from a dark brown zeolitic claystone with extremely low carbonate content to a gray-orange nannofossil ooze was observed in an identical stratigraphic position to the gradual changes observed on the Southern High. However, the Site 1208 transition is far more condensed than the other records, occurring over an interval of 1–2 cm. The Eocene/Oligocene boundary is associated with a marked increase in sedimentation rates at all of the sites.

In general, the lithologic change in all of the sites is accompanied by a general improvement in microfossil preservation; late Eocene nannofossils in most of the sections show a high degree of etching, whereas early Oligocene assemblages show less dissolution and slightly more overgrowth. Planktonic foraminifers are rare and badly fragmented in the upper Eocene, but preservation improves and abundance increases in the Oligocene. At Site 1208, the deepest site in the transect that contains the transition interval, planktonic foraminifers are largely absent from the sequence. The benthic foraminifer Nuttallides truempyi, the LO of which shortly precedes the Eocene/Oligocene boundary, is found below the color transition. Sparse nannofossils below the boundary are extremely etched.

The age model for the Eocene–Oligocene transition at all of the Leg 198 sites is currently preliminary, and thus the exact correlation of the change in carbonate content with the series boundary and the sharp Oi-1 cooling event (33.15–33.5 Ma) (e.g., Miller et al., 1991; Zachos et al., 1996) has yet to be determined. However, the preliminary data show that the prominent change in lithologic signature occurs just before or within the time of cooling. Moreover, because the exact location of the Oi-1 event is currently unknown, we have yet to determine whether orbital cycles can be detected within this event.

The distinctive color change in all of the Leg 198 records reflects a pronounced deepening in the CCD in the Eocene–Oligocene transition. In the latest Eocene, the CCD on Shatsky Rise was between the paleodepths of Sites 1208 and 1211, probably closer to the former site based on the sporadic occurrence of nannofossils. After the event, the depth was substantially greater than Site 1208. This significant change is observed in other ocean basins (e.g., Zachos et al., 1996) and possibly reflects an increase in mechanical and chemical weathering rates on continents associated with cooling. Alternatively, deepening of the CCD may be associated with an intensification of deep-ocean circulation and a consequent decrease in the age of deep waters. The magnitude of the change in the transition interval as shown by color reflectance data varies as a function of depth (see Fig. F28). This is consistent with the deepening of both the lysocline and CCD.

Our current understanding of the changes in climate and circulation in the Eocene–Oligocene transition is based almost entirely on records from the mid- and high-latitude Atlantic and Indian Oceans (Miller et al., 1991; Zachos et al., 1996). Changes observed across this transition in the tropical-subtropical Pacific from the Shatsky Rise depth transect have the potential to add another dimension to this understanding.

A surprising side note to the major Cretaceous and Paleogene objectives on Leg 198 was the recovery of two expanded late Neogene sections on the Northern High of Shatsky Rise at Site 1207 and the Central High at Site 1208. These sections could be important from both stratigraphic and paleoceanographic perspectives. The two sections have apparently complete upper middle Miocene (~12 Ma) to Holocene sections that are composed of nannofossil ooze and nannofossil clay. Both sections have a mixture of nannofossils and significant amounts of diatoms (10%–40% at Site 1207; 5%–20% at Site 1208), and minor amounts of foraminifers, radiolarians, and silicoflagellates. Numerous discrete ash horizons are found at both sites, predominantly in the Pliocene–Pleistocene interval.

Sedimentation rates at Site 1207 average 18.4 m/m.y. from the Holocene to latest Miocene. Rates at Site 1208 range between 22 and 42 m/m.y. from the Holocene to the early late Miocene (Fig. F42). These rates are far higher than typical pelagic sedimentation. The detrital clay and silt component of the sediment may have been delivered by eolian transport. However, a large component of the sediment could have been delivered by sediment drift.

Marked cyclic variations are observed in MST and color reflectance data throughout the upper Miocene to Holocene section at both sites. These variations are expressed as strong lithologic cycles that have frequencies at the decimeter to meter scale. These cycles are marked by relatively subtle to sharp color changes that are associated with variations in the amount of clay, pyrite, and different biogenic particles. The darker interbeds tend to have more abundant diatoms and clay, more dissolved nannofossil assemblages, and more abundant pyrite. The lighter interbeds contain fewer diatoms, less clay, and a better-preserved nannofossil assemblage. The cycles might represent variations in carbonate dissolution as a result of changes in the depth of the lysocline, fluctuations in biosiliceous production, or a combination of the two. Preliminary biostratigraphy indicates that these variations represent eccentricity and obliquity cycles. The section recovered at Site 1208 contains an extraordinary expanded magnetic stratigraphy extending back to Chron C5An in the late middle Miocene (Fig. F55). This section also shows great promise for establishing a paleointensity record for the North Pacific.

The Site 1207 and particularly the Site 1208 Neogene section has significant potential for establishing a high-resolution biochronology and astrochronology back to at least the late Miocene. This potential derives from the combination of siliceous and calcareous microfossil biostratigraphy, high-resolution magnetostratigraphy, a marked orbital cyclicity, numerous ash layers with potential for radiometric dating, and a potential magnetic paleointensity record. With high sedimentation rates and foraminifers found throughout the section, the sites also have significant potential for reconstructing climate and paleoceanography of the northwestern Pacific over the last 12 m.y.

During Leg 198, we cored the first igneous rocks from Shatsky Rise, probing the top of the igneous pile. Coring in Hole 1213B terminated in mafic igneous rocks on the flanks of southern Shatsky Rise. In all, 46.4 m of igneous section was drilled, with 33.4 m of recovered core (72%). Six cores, 198-1213B-28R through 33R recovered mostly massive diabase and minor basalt from three subunits (IVA to IVC), with each subunit thought to be a separate sill. The igneous rocks are predominantly hypocrystalline, fine-grained diabase (97.6%) with a small amount of sparsely phyric, aphanitic basalt (2.0%) at contacts. The diabase groundmass consists mainly of euhedral to subhedral plagioclase and intervening subhedral pyroxene and olivine, with minor glass. Alteration in the igneous section ranges from minor to moderate. Plagioclase and pyroxene crystals are locally altered to clay, and in thin section, glassy groundmass has been ubiquitously devitrified and/or altered to clay minerals. Basalt occurs at subunit "chilled" contacts, symmetrically disposed around fragments of metasediment that mark the subunit boundaries. From the chilled contacts, the basalt grades toward more coarse-grained diabase in the unit centers, where the groundmass approaches medium grained.

The sills must be early Berriasian age or younger, since this is the age of the host sediment. Paleomagnetic data show two subunits that have fairly steep positive magnetic inclinations, whereas the third, basal subunit has a lower, negative inclination, implying both normal and reversed magnetic polarities are recorded in the igneous section. This mixture indicates that the sills must have formed either before or after the Cretaceous Long Normal Superchron (i.e., the Cretaceous Quiet Period, 121–83 Ma). On the seismic profile along which Hole 1213B was drilled, the seismic "basement" has an odd character that may be related to the presence of intrusive, rather than extrusive, igneous rock at the sediment-igneous rock contact. The "basement" reflector, that being the deepest continuous seismic horizon, is weaker than elsewhere on Shatsky Rise, and other, stronger reflectors occur beneath it. These deeper reflectors were not considered "basement" because they are not continuous all along the line, as is the weaker, shallower horizon. The cored section suggests that the weak "basement" horizon denotes the top of the sills, whereas the deeper reflectors may be the top of the extrusive lava pile.

The stratigraphy of sites drilled on the Southern, Central, and Northern Highs during Leg 198 can be integrated to interpret the geologic history of Shatsky Rise from its formation as a LIP in the Late Jurassic and Early Cretaceous, through multiple depositional episodes and different sedimentological regimes separated by short and sometimes quite lengthy hiatuses (Fig. F42). These gaps were produced by erosion and dissolution, but distinguishing between these two processes can be difficult. Comparison of the stratigraphy of the cored sections of sites on the Northern and Central Highs shows similarities and differences with the sequence cored on the Southern High. This suggests that a combination of local and regional-scale processes controlled sedimentation. The following history is compiled from the stratigraphic data gathered during Leg 198. Although we have cored a complete record from the Southern High, drilling was terminated in the mid-Cretaceous on the other highs.

Although Leg 198 did not core "true" extrusive basement, igneous rocks intruded soon after the formation of extrusive basement were recovered at Site 1213. These sills are widespread above basement in the Pacific (e.g., Larson, Schlanger, et al., 1981; Plank, Ludden, Escutia et al., 2000) and argue for a multistage origin for the igneous foundations of Shatsky Rise. The latest Jurassic and Cretaceous history of the rise was dominated by fairly continuous deposition on the Southern and Northern Highs (Fig. F42). Unconformities found in more than one location suggest an interval of erosion on the Southern High during much of the latest Hauterivian and Barremian. The late Cenomanian to Santonian interval was also characterized by sporadic sedimentation, as noted by Sliter (1992) (Fig. F56), and widespread erosion on the Central and Southern Highs, at least in most locations (Sites 1208, 1212, and 1214). The deepest site in the depth transect (Site 1213) may have also rested below the CCD (e.g., Thierstein, 1979). The sequence on the Northern High appears to be complete in this interval. Continuous deposition resumed in the shallower sites in the Campanian and Maastrichtian whereas dissolution likely continued in deeper locations; an unconformity above the upper Campanian at Sites 1207 and 1208 was likely caused by a regional erosional episode at that time or in the early Cenozoic.

Deposition in the early Paleogene was nearly continuous at the shallow sites of the Southern High (Sites 1209–1212), but even at these sites, a lengthy hiatus in the later part of the Paleogene continued into the early Miocene in most locations. The onset of this hiatus lies in the middle Eocene at Site 1212 and in the Oligocene at the other shallower sites. At the shallower sites, this hiatus was likely a result of removal of sediment by erosion during the late Oligocene and early Miocene. Very slow sedimentation characterized the late early and early middle Miocene (except at Sites 1213 and 1214 where no sediment from this interval was recovered), and the rate of sedimentation increased through the rest of the Neogene at most locations. Sites on the Northern and Central Highs lay in a different oceanographic regime in the late Neogene with a significant biosiliceous sediment component as well as clastic material concentrated into sediment drifts. Continuous sedimentation on the deep rise (Sites 1213 and 1214) began again in the early Pliocene.

Like other open-ocean plateaus and rises, Shatsky Rise shows many aspects of stratigraphy and sedimentation that are different from other open-ocean, abyssal settings. Open-ocean plateaus and rises are prone to erosion, and sedimentation rates are not as high as areas along the continental margins. Thus, their sedimentary sections are not as thick, and older deposits often rest at shallow burial depths. This has important implications for fossil preservation and the fidelity of paleoceanographic proxies. Because open-ocean deposits tend to be mostly biogenic and are less complex sedimentologically, their geologic record is often easier to interpret than in areas along continental margins. Vertical changes in physical and chemical oceanography can be determined using depth transects down the flanks of the rises. The cycles recovered on Shatsky Rise are a good example. Leg 198 has proven that oceanic plateaus hold some of the best records for investigating climate change through extended intervals of geologic time, particularly based on a depth transect approach.