DISCUSSION AND CONCLUSIONS

The five sites cored during Leg 207 recovered sedimentary sequences that record the sedimentary and paleoceanographic history of the tropical Atlantic. The sedimentary pile can be divided into three broad styles of deposition—synrift clastics, restricted marine "black shales," and open marine chalk and calcareous claystones (Fig. F8). Synrift deposits show a variety of lithologies, and important questions remain regarding depositional water depth, terrigenous sources, transport paths, and possible differences in age within and among the sites. Pelagic deposition included long intervals of relatively constant sedimentation recovered at multiple sites suitable for detailed chronostratigraphic and paleoceanographic studies. In addition, the changing proportions of terrigenous and biogenic components, the distribution of regional hiatuses, and benthic isotopic studies should help constrain the evolution of circulation in the tropical Atlantic. The most dramatic lithologic features recovered during Leg 207, however, are the sedimentological expression of the P/E and K/T boundary events and black shale deposition encompassing several OAEs.

Paleocene/Eocene Boundary

The P/E boundary is placed at the base of a ~30-cm-thick greenish clay bed that is a unique lithology at each site where it is present (Fig. F9A). This layer is provisionally interpreted as the record of the shoaling of the CCD following the P/E boundary event and is expected to coincide with both the benthic foraminiferal extinction event and the carbon isotopic excursion. The 1–5 cm of the Paleocene below the clay shows a progressive increase in clay content toward the boundary. At Site 1260, the clay is relatively thick and distinctly laminated, but this fabric is not seen at the other sites. Either the record at Site 1260 is exceptionally complete or there were paleogeographic or paleobathymetric differences in expression of the event on Demerara Rise. A 12- to 20-cm-thick red interval begins 2–20 cm above the P/E boundary at Sites 1258, 1259, and 1261 but not at Sites 1257 or 1260, documenting additional site-to-site variability.

Magnetic susceptibility measurements suggest that the clay-rich part of the boundary sequence ranges from ~1 m in thickness at Site 1261 to ~2 m in thickness at Site 1258. The differential thickness indicated by the magnetic susceptibility high may reflect depth-related changes in carbonate preservation or variations in sedimentation rate. The P/E interval at all sites displays pronounced cyclicity in physical property measurements and sediment color. Cyclicity is particularly pronounced at Site 1259, where a short-period cycle with a ~25- to 30-cm wavelength is modulated by a ~1- to 1.5-m cycle that resembles the modulation of the orbital precession cycle with a period of ~21 k.y. by the 100-k.y. eccentricity cycle. The pervasive cyclicity in physical property records from the lower Eocene offers the possibility not only of refining the chronology of the P/E boundary event but also cross-checking the results among sites.

Cretaceous/Tertiary Boundary

Compared to the P/E boundary, the lithologic character of the K/T boundary is remarkably consistent at the three sites in which it was recovered (Fig. F9B). At each site, sediments ~0.5 m above and below the boundary have a yellowish brown tinge, but the causal relationship of these colors to the boundary is unclear. The boundary itself occurs within an interval of pelagic sediment with no visible change in depositional style until 0.3 cm below an ejecta layer marking the boundary. The ejecta layer is 1.7–1.9 cm thick and is composed of normally graded green spherules. Spherules decrease in diameter from ~2 mm at the base to ~0.25 mm at the top; the uniformity of this layer among sites suggests settling of ejecta material through the water column without subsequent reworking (i.e., primary air fall deposit). The spherule bed sits on a ~3-mm-thick gray fine-grained homogeneous layer, probably composed of Maastrichtian nannofossils. Possible interpretations of this layer include (1) fine material suspended or elutriated by impact shaking, (2) a thin layer of reworking between the time of the impact and the arrival of the first spherules, and (3) a diagenetic artifact related to rapid burial and alteration of the spherules. The spherule bed is overlain by a clay-rich interval containing characteristic earliest Tertiary foraminifers and nannofossils as well as bloom taxa. Carbonate content gradually increases over ~0.5 m, and the basal Danian seems to be expanded, suggesting that details of the paleoceanographic conditions in the earliest Tertiary may be preserved on Demerara Rise.

Black Shales

Tens of meters of organic-rich laminated sediments (black shale) were recovered at all sites during Leg 207 (Fig. F8). These rocks represent the local equivalent of widespread organic-rich sedimentation in the southern part of the mid-Cretaceous North Atlantic (Kuhnt et al., 1990). They include the globally recorded OAE 2 around the Cenomanian/Turonian boundary.

On Demerara Rise, black shale sedimentation began abruptly in the Cenomanian and extended into the Campanian. The black shales unconformably overlie Albian–Cenomanian shallow-water siliciclastic sediments, including a unit tentatively interpreted as a tidal flat deposit. Some of the Albian claystones have a high organic carbon content, but organic matter composition and a lack of lamination distinguishes them clearly from the younger black shales. During the Coniacian, and especially in the Santonian, oceanographic conditions in the region were apparently less stable, as indicated by intercalated glauconitic and bioturbated intervals. Bioturbation affected large intervals of the organic-rich claystones and destroyed the lamination. Furthermore, probable tectonic instability on Demerara Rise led to widespread mass wasting, resulting in numerous debris flows affecting the Coniacian–Campanian part of the black shale sequence. The lithologic transition from the black shales into the overlying Campanian pelagic deposits at some sites occurs over several meters of glauconite-rich, bioturbated claystone. At other sites, bioturbated and laminated sediments are in a sharp contact. Regardless, the basal Campanian chalks typically contain some glauconite, quartz, and abundant zeolite-replaced radiolarians.

The black shales are characterized by the following three facies:

1. Laminated organic-rich claystones with variable carbonate content (5–50 wt%) and up to 30 wt% TOC of marine origin; this facies is particularly well developed in the upper Cenomanian–Turonian part of the black shale interval;
2. Lighter laminated to finely bedded foraminifer wackestone to packstone occurring as either light–dark cycles in organic-rich claystones or as discrete beds with a sharp base and a gradual transition into overlying sediments; these beds may represent tempestites (storm deposits?) or turbidites; and
3. Glauconitic bioturbated intervals that might represent periods of oxygenation. Burrows in rocks of this facies extend deeply into the underlying sediment (up to 1 m). These glauconite-rich intervals can be correlated between holes and, perhaps, among the shallower sites. No equivalent has been found at the deeper sites.

Well-preserved fish debris and phosphatic nodules (~2 cm) are ubiquitous but are particularly concentrated in the organic-rich claystones. These occurrences may form discrete layers/intervals or are scattered in the background sediment. Diagenetic calcite formation is common at the shallower sites. Calcite replaces organic-rich sediment in the form of cone-in-cone structures up to several centimeters thick.

Black shale sediments seem to have been deposited in a shallow to moderately deepwater environment. Existing seismic lines do not support a silled basin model but rather an intensified oxygen-minimum layer impinging on Demerara Rise. This interpretation is further supported by rare glauconite-rich bioturbated intervals at the shallower sites marking the weakening or even retreat of the oxygen minimum zone.

Of the five sites drilled, multiple sequences of Cenomanian and Turonian black shales, Campanian–Maastrichtian chalk, as well as Paleocene, lower Eocene, and middle Eocene chalk were recovered at each. In aggregate, the recovered sections form a continuous record of tropical sedimentation from the late early Cenomanian (~98 Ma) to the Santonian (~83.5 Ma) and from the late Campanian (~76 Ma) to the late middle Eocene (~38 Ma). The oldest sedimentary rocks recovered during Leg 207 are early and middle Albian claystones (Site 1258), and the youngest are Pliocene–Pleistocene clay-rich nannofossil oozes (Site 1261). At all but one site (1261), the Paleogene was exposed at the seafloor or by <90 m of burial beneath Neogene sedimentary cover.

Neogene or Oligocene calcareous ooze is present at the seafloor at all the Leg 207 sites and consists mostly of winnowed foraminifer sands. The thickest Neogene sequence is present at Site 1261, where 370 m of Pleistocene, Pliocene, and upper Miocene strata is present above chalk of late middle Eocene age. The Neogene section at Site 1261 was spot cored through most of its thickness but appears to comprise an expanded sequence of lower Pliocene and upper Miocene deposits that form a drape over the shallower parts of Demerara Rise shoreward of the Leg 207 drill sites. Elsewhere, drilling encountered lower Miocene and lower Oligocene calcareous ooze below a thin veneer of Pleistocene foraminifer nannofossil ooze. At Site 1259, nearly 125 m of Oligocene and lower Miocene strata are present, but at least the upper 28 m represents a remobilized sequence that was slumped or eroded and redeposited in inverse stratigraphic order. Redeposited lower Oligocene calcareous ooze also is present at Site 1260. The E/O boundary is unconformable or highly condensed at all sites.

There are widespread Paleogene unconformities in the upper Eocene, the lower middle Eocene, and upper Danian (Figs. F10, F11). However, with the exception of the upper Eocene, the other hiatuses are mostly represented by sediment in at least one of the sites. Most notably, the widespread lower Eocene–middle Eocene hiatus, which is found virtually everywhere in the North Atlantic, is represented by apparently continuous sedimentation at Site 1258. Indeed, recovery at Site 1258 appears to include the most complete lower Eocene section cored by DSDP and ODP in the tropical oceans (Figs. F10, F11). In the Cretaceous, there is a condensed surface or unconformity across the Santonian–lower Campanian and another between the upper Albian and lower Cenomanian (Figs. F12, F13). Further biostratigraphic work may well recover zonal markers for parts of the lower Campanian in the sequence of glauconitic-rich horizons at the contact between the Turonian–Santonian black shales and the upper Campanian chalk. The age of the Albian–Cenomanian unconformity may be diachronous with the oldest Cenomanian section at Sites 1258 and 1260, whereas Cenomanian sedimentation began slightly later in the late Cenomanian at the other sites.

Preservation of microfossils is highly variable at all Leg 207 sites. Calcareous microfossils are best preserved in parts of the Albian at Site 1258, the black shale sequence, and the Paleocene—all sequences rich in clay. In some cases, particularly in the black shales, foraminifers are preserved with translucent skeletons and primary microstructure. Much of the lower and middle Eocene suffers from extensive recrystallization of calcite. Siliceous microfossils (mainly radiolarians) are present in parts of the middle and lower Eocene and Campanian and in spots in the black shales and underlying Albian claystone. Middle Eocene siliceous nannofossil chalks at Site 1260 are particularly notable for their expanded sequence of middle Eocene radiolarian zones with excellent preservation.

Paleocene/Eocene Boundary

During Leg 207, we recovered the P/E boundary at all five sites, with a remarkable 10 cores spanning the boundary interval. At all sites, the P/E boundary is represented by dark green clay-rich beds that form a sharp contact with underlying chalk. The chalk below the boundary is frequently well lithified and shows a transition to slightly darker sediment, often laminated, ~1–2 cm below the first dark green clay. The green beds themselves are either massive (e.g., Sites 1257–1259) or display fine laminations (e.g., Sites 1260 and 1261) in the lower parts. Bioturbation gradually returns between 5 m (Site 1261) and ~30 cm (Site 1259) above the base of the green bed.

All five sites showed the same typical succession of calcareous nannofossils across the P/E boundary interval (Zone NP9). Abundant Zygrhablithus bijugatus, Discoaster multiradiatus, and rhombs (interpreted as Rhomboaster) characterize the early Eocene (upper part of Zone NP9 = NP96). Neochiastozygus junctus is also abundant and continues without rhombs but with occasional fasciculiths to the P/E boundary. Below the P/E boundary fasciculiths replace N. junctus and Z. bijugatus in dominance and preservation improves considerably.

Samples taken from the massive green beds at Sites 1258 and 1259 contain few calcareous foraminifers but common, poorly preserved radiolarians. Calcareous microfossils begin to appear at about the same level that the massive or laminated sequence begins to become bioturbated. The foraminiferal assemblages are characteristic of dissolved faunas, consisting mostly of small benthic foraminifers. However, as preservation improves, a distinct fauna of planktonic foraminifers appears. These species include M. allisonensis, A. sibaiyensis, and A. africana, all of which have been named the excursion fauna (e.g., Kelly et al., 1996) because they have been previously recognized only in the interval of the ¹³C anomaly. Leg 207 sites appear to contain a few additional, as yet unnamed, species of the excursion fauna. In addition, the foraminiferal assemblage from Site 1259 includes a clavate species, P. paleocenica, which has thus far been described only from coastal Senegal and the equatorial Pacific (Site 1220). About 2 m above the base of the green beds at Site 1259, the excursion fauna appears to be rare or absent, but several other foraminifer species are present that may be undescribed. These include a nearly biconvex variant of Morozovella aequa.

Cretaceous/Tertiary Boundary

Although Demerara Rise is located some 3500 km southeast of and therefore upwind from the Chicxulub impact site, ejecta layers ~1.7–1.9 cm thick were cored at the K/T boundary in six holes at three sites (Fig. F14). This is the first such discrete K/T ejecta layer reported from the South American continent proper.

The ejecta usually rests on a 1- to 2-mm-thick white calcareous layer (Fig. F14) composed of Cretaceous microfossils that were possibly suspended into the water column by the seismic shock of the impact. The ejecta itself is typically laminated, sometimes consisting of sublayers of darker and lighter color with different concentrations of spherules, black to light green in color. These are often graded upward, and individual spherules may range in size from 1 to 2.5 mm in diameter.

Above the ejecta layer, the dark green boundary clay is relatively soft, low in carbonate, and sometimes drilling disturbed. In the most complete sections (at Sites 1258 and 1259) (Fig. F14), the greenish chalk beneath contains the latest Maastrichtian-age calcareous nannofossils Micula prinsii and Micula murus along with the corresponding planktonic foraminifers Abathomphalus mayaroensis and the rare Plummerita hantkenoides, which suggests that the Leg 207 sites preserve an unusually complete record of latest Cretaceous paleoclimate and biotic evolution.

Immediately above the boundary clay, the "disaster forms" Braarudosphaera and Thoracosphaera are present, but other nannofossils are difficult to extract from the more lithified Danian chalk farther uphole. Instead, planktonic foraminifers provide better age control for the biotic "recovery," beginning with rather small specimens of P. eugubina and associated forms that distinguish the basal Tertiary Zone P and hence progressing upsection through Subzone P1b or Zone P2 before a disconformity intervenes.

Black Shales

At each of the five sites drilled during the course of Leg 207 a sequence of claystones rich in organic matter, referred to here as black shales (Fig. F13), was recovered. The black shale sequence has been attributed to a Cenomanian–earliest Campanian age based on biostratigraphic observations (Fig. F12). Throughout the depth transect covered by the five sites, the thickness of the black shales is relatively consistent (Site 1261: shallow, 89 m; Site 1259: shallow, 56 m; Site 1260: intermediate, 93 m; Site 1257: deep, 57 m; and Site 1258: deep, 56 m).

The black shales rest on a variety of Albian–earliest Cenomanian shallow marine to nonmarine sediments that include:

Site 1261: upper Albian–lower Cenomanian silt–sandstones,

Site 1259: lower Cenomanian silt–sandstones,

Site 1260: lower–middle Albian claystones,

Site 1257: upper Albian siltstones, and

Site 1258: upper–middle Albian black shales and claystones.

The onset of the black shale sedimentation occurred, except at Site 1258, in the Cenomanian over a period of ~3 m.y. (Fig. F13). Because of the low resolution, poor preservation, and absence of both calcareous nannofossils and planktonic foraminifers in the lower part of the black shale succession at most sites, an accurate age assignment of the basal part has not been made. The current biostratigraphic data cannot solve the question of whether the onset of the black shales at the various sites is synchronous or diachronous over a period of ~3 m.y. in the Cenomanian. A very dark black shale interval, intercalated in the lower part of the black shale succession of the shallow sites (1259 and 1261) is of late Cenomanian–early Turonian age and may equate to OAE 2.

The top of the black shale succession is diachronous, with ages ranging from latest Turonian (Site 1258) to mid-Coniacian (Site 1260) to late Santonian (Site 1261) to early Campanian (Sites 1257 and 1259). The different ages of the cessation of black shale deposition are partly caused by postdepositional erosion and/or condensation.

The magnetic polarity zonation in Eocene–Campanian sediments could be resolved at nearly all Leg 207 sites. Paleomagnetic investigations began with a shipboard phase, followed by an intensive shore-based analysis of ~800 minicores at the paleomagnetism laboratory at the University of Munich, Germany. In addition to revealing a more reliable magnetostratigraphy for each site, the detailed shore-based analyses of these discrete samples allowed insight into the magnetic properties and magnetic mineralogy of the sediments and a determination of paleolatitudes.

The preliminary results suggest that this region of Demerara Rise, now at ~9°–10°N latitude, was at approximately the same northern latitude during the Albian, then drifted southward during the latest Cretaceous to a location at or just south of the paleoequator during the Paleocene–middle Eocene. The near-equatorial setting implied that polarity assignments required analysis of the changing magnetic vectors as each sample underwent progressive thermal demagnetization to remove present-day secondary overprints. The resulting polarity zonations, when constrained by shipboard biostratigraphy, could generally be unambiguously correlated to the magnetic polarity timescale. Additional shore-based paleomagnetic sampling, especially within selected stratigraphic intervals that displayed cyclic alternation of sediment, increased the resolution of the polarity chron boundaries. The combined magnetostratigraphy from shipboard and shore-based studies is included for each site in this volume:

1. At Site 1257, we resolved Chrons C17n–C20r (middle Eocene), Chrons C24r–C26n (late Paleocene), and Chrons C31r–upper C33n (Maastrichtian/Campanian boundary).
2. At Site 1258, we obtained a high-resolution record of Chrons C20r–C26r (middle Eocene–late Paleocene) and Chrons C29r to potentially C33r (Maastrichtian–late Campanian).
3. At Site 1259, Chrons C18r–C24r (middle Eocene–latest Paleocene) are well defined, and a tentative assignment was made for Chrons C29r–C31r (Maastrichtian).
4. At Site 1260, we obtained a high-resolution record of Chrons C18r–C21r (middle Eocene), and lower-resolution identification of Chrons C3n–C26n (early Eocene–late Paleocene) and C29r–C31r (Maastrichtian–late Campanian).
5. At Site 1261, Chrons C19r–C21r (middle Eocene), Chrons C24r–C27n (early Eocene–early Paleocene), and Chrons C29r–C31r (Maastrichtian) were delimited.

The main goals for postcruise paleomagnetic studies are to utilize sediment magnetic characteristics to examine bottom environments during the latest Paleocene and earliest Eocene, to obtain astronomical calibration of the durations of Eocene–Campanian polarity chrons, and to compile a detailed paleolatitude motion of Demerara Rise.

Linear sedimentation rates (LSRs) were derived from age-depth models at each of the Leg 207 sites with current stratigraphic resolution. At all sites, the sediment sequences contain intervals of apparently continuous sedimentation separated by distinctive hiatuses, condensed intervals, and/or mass flows. Each of the hiatuses recognized comprises, based on biostratigraphic dating, at least 1 m.y.

Linear Sedimentation Rates of Cretaceous Black Shales

The absence of biostratigraphic zonal markers, as well as the extensive Normal Polarity Superchron C34n, restricted the age assignment in most of the Lower–Upper Cretaceous sequences recovered during Leg 207 (Albian–Santonian). Therefore, LSRs calculated for these intervals should be considered as imprecise estimates. These intervals include the black shale sequences (Cenomanian–Santonian) that were characterized by LSRs of 3–5 m/m.y. at Sites 1257, 1258, and 1259 and slightly higher values of ~8.5 m/m.y. at Sites 1260 and 1261 (corresponding to 0.3–0.9 cm/k.y. in Fig. F15). In addition, mass accumulation rate (MAR) calculations were limited by the lack of reliable dry bulk density data at Sites 1258 and 1259.

Significant LSR and MAR Increases across the K/T Boundary

Average LSRs increased markedly across the K/T boundary interval at all sites (Fig. F15). They were at least 1.5 and up to 2.8 times higher during the Paleogene (Paleocene–Eocene) than rates recorded during the latest Cretaceous (Campanian–Maastrichtian). There is no discernible change in LSRs across the P/E boundary interval at all Leg 207 sites (Fig. F15). Late Paleocene and early Eocene LSRs ranged from 7 to 15 m/m.y. (Fig. F15), rates typical for pelagic nannofossil chalk, which is the dominant lithology in this interval. The middle Eocene interval is characterized by a decrease in LSRs from 11 m/m.y. (during the early Eocene) to 3 m/m.y. at Site 1257, whereas a relative increase in LSRs is recorded for the same time interval at Sites 1260 (from 12 to 20.5 m/m.y.) and 1261 (from 7 to 9 m/m.y.) (Fig. F15).

MAR calculations, using the LSRs and average dry bulk density data, allow for a better assessment of sedimentation processes because the influence of compaction has been taken into account. This effect is clearly illustrated at Site 1261, where overburden by Neogene sediments on the Paleogene chalks is greatest and, hence, is reflected in lowest LSRs calculated for the Paleogene sequences on Demerara Rise. However, MARs are comparable to those found at the less deeply buried sediments at Site 1257 (Fig. F15). In contrast, at Site 1258, where the change in LSR across the K/T boundary was least pronounced, MARs only increased by 10% during the Paleocene and Eocene, compared to 50% using LSR only (Fig. F15). Nevertheless, the conclusion that sedimentation rates, hiatuses notwithstanding, were higher overall during the Paleogene than during the Late Cretaceous still holds when MARs are considered.

Glimpses of Neogene Sedimentation Rates: Site 1261

Neogene oozes recovered during Leg 207 are too thin relative to sampling density at four of the five sites for a meaningful sedimentation rate to be calculated. In addition, slumps and hiatuses occur in these sequences. At Site 1261, however, the upper 368 m contains a thin Pleistocene interval separated by a gap of ~3 m.y. from an interval of upper Miocene–Pliocene nannofossil clay and nannofossil ooze. Sedimentation rates for this Miocene–Pliocene section are the highest calculated for Demerara Rise (65 m/m.y.) (Fig. F15). This interval of high sedimentation rates sits above a thick (60 m) series of mass flow deposits that occupy a stratigraphic position corresponding to a >30-m.y. gap in pelagic sedimentation.

Intervals of complete recovery were obtained by multiple coring, particularly through the middle–upper Eocene and Cenomanian–Campanian (Fig. F16). Rotary coring and moderate recovery prevented sampling the complete stratigraphy everywhere, however. Where overlap existed in the Paleocene–lower Eocene (Fig. F17) and Campanian and Maastrichtian (Fig. F18) sequences, magnetic susceptibility was generally the best correlation tool, reflecting varying percentages of carbonate and clay. In the Cenomanian–Santonian black shales, GRA bulk density allowed good correlation among holes where organic-rich laminated claystones alternate with more calcareous well-indurated sediments (Fig. F19).

In the lower and middle Eocene, splices were created for Sites 1258, 1259, and 1260 (Fig. F16). At these sites, the splice sections span the P/E boundary. Splices also were constructed that span the K/T boundary at Sites 1259 and 1260, where the boundary is conformable and contains the ejecta layer (Figs. F17, F18). A splice also was generated for Site 1257, where the base of the Paleocene is an unconformity. It was possible to create a splice section that covered the entire Campanian–Maastrichtian succession at Sites 1259 and 1260. For Sites 1259, 1260, and 1261, sampling splices were created that cover all or nearly all of the Cretaceous black shales.

Cenomanian–Santonian sequences, extraordinarily rich in organic carbon, were recovered at all five sites cored during Leg 207. The finely laminated, dark-colored organic-rich claystones typically contain 2–15 wt% organic carbon, and they range in thickness from 56 m at Sites 1258 and 1259 to 93 m at Site 1260. These thick and extensive black shales are part of the global burial of huge amounts of organic carbon on the seafloor during OAEs 2 and 3.

Organic geochemical properties of the black shales measured during Leg 207 reveal aspects of the exceptional conditions of organic matter production and preservation involved in their formation. The results of Rock-Eval pyrolysis show that the bulk of the organic matter originates from algal and microbial primary production (Fig. F20). Extractable biomarker hydrocarbon compositions suggest that relative proportions of algal and microbial contributions of organic matter vary in different parts of the black shale sequences (see Forster et al., this volume). Hydrocarbon distributions from some parts are dominated by the acyclic tetraterpenoid lycopane, which is indicative of microbial productivity (Brassell et al., 1981). In other parts, the algal C₁₅–C₁₉ n-alkanes dominate. The presence of both algal and microbial biomarkers suggests that organic matter production was enhanced by a consortium of primary producers. Brassell et al. (1981) postulate expansion of an intensified oxygen minimum zone into the photic zone led to black shale deposition at DSDP Site 144, which was recored as Site 1257 during Leg 207. This paleoenvironment probably permitted the coexistence of algae and the photosynthetic microbes that function best under dysaerobic and anaerobic conditions.

Organic matter is thermally immature, as shown by both the high Rock-Eval hydrogen index and low T_max values (Fig. F20) and also by the dominance of nonrearranged steranes (see Forster et al., this volume). In addition, improved organic matter preservation is implied by C_organic/N_total ratios that increase to ~40 as organic carbon concentrations increase (Fig. F21). Values above ~20 are usually considered typical of land-plant organic matter, but such elevated ratios are also common to Cretaceous black shales (Meyers, 1997). Consequently, most of the elevated C/N values that mimic those of land-derived organic matter are likely to be the result of retarded selective alteration of algal and microbial organic matter. A likely scenario is that nitrogen-rich components were more readily degraded than other organic matter components during sinking of organic matter through a strongly developed oxygen minimum zone, thereby elevating the C/N ratio of the surviving organic matter (Twichell et al., 2002).

Contributions of land-derived organic matter are also evident in the black shales. Some of the n-alkane distributions of the 10 samples that compose the biomarker survey are dominated by the C₂₉ and C₃₁ components diagnostic of land-plant waxes (see Forster et al., this volume). These hydrocarbon compositions do not indicate a dominance of land-plant organic matter in the black shales, inasmuch as land plants contain larger proportions of hydrocarbons than do algae and microbes (e.g., Meyers, 1997), but they do confirm the presence of continental plant debris. However, the proportion of land-plant material appears to become important in some of the lower Cenomanian black shales from Site 1260 in which C/N ratios are between 40 and 60 (Fig. F21).

Concentrations of interstitial methane in the sediment sequences at all five Leg 207 sites increase abruptly in the black shales, where they reach their peak values. The possible relation between sediment organic matter contents and gas concentrations was tested by measuring the organic carbon concentrations of the headspace sediment samples at three of the sites. A rough correspondence exists between higher TOC and greater gas concentrations (see Meyers et al., this volume). Marked excursions from a simple linear relation suggest that the type of organic matter, and not simply the amount, affects gas generation from the black shales. Because of the low thermal maturity of organic matter at all five sites (Fig. F20) and the predominance of methane in the interstitial gases, the origin of the gases is probably from in situ microbial activity. Dramatic decreases in methane concentrations above the black shale units suggest that methane oxidation, which consumes interstitial sulfate, proceeds in overlying units (see "Interstitial Water Geochemistry," below, and Meyers et al., this volume). Moreover, active generation of gas must exist in the black shales to replace losses from that migration and to maintain the elevated gas concentrations in these lithostratigraphic units.

Interstitial waters from 152 samples were collected at all five sites cored on Demerara Rise, covering a depth range from the sediment/seawater interface to 648 mbsf. This comparably dense interstitial water sampling program was conducted to understand the impact of relatively deep-seated organic matter–rich black shale sequences on diagenetic processes along the paleoceanographic transect. At all sites (except Site 1261), sulfate and ammonium gradients are essentially linear from the sediment/seawater interface to the top of the black shale sequence (Fig. F22A, F22B). These results indicate the following:

1. Nearly 100 m.y. after deposition, the black shale sequence continues to act as a bioreactor that dominates interstitial water chemistry.
2. Sulfate reduction is of minor importance at shallower depth intervals.
3. The resulting downhole profiles are controlled by the existence of one major stratigraphic sulfate sink and ammonium source (Cretaceous black shales) and simple compensatory diffusion from and/or to the sediment/seawater interface. We interpret the linearity of these profiles to reflect minimal accumulation of sediments younger than middle Eocene age at Sites 1257–1260. Methane diffusing upward from the black shales may furnish metabolic activity above the black shale unit, possibly anaerobic methane oxidation.

Despite the reducing character of the sedimentary column at all Leg 207 sites, only very low pore water concentrations of manganese and iron are attained in the black shales (Fig. F22D, F22E). Our favored working hypothesis for this observation is that these redox-sensitive metals were completely remobilized during or shortly after the host organic matter–rich units were deposited, implying conditions of synsedimen-tary oxygen depletion. In contrast, high dissolved barium levels (>300 µM) are attained in the black shales (Fig. F22C), indicating ongoing mobilization of barium where sulfate concentrations are lowest. Uphole diffusion of barium and downhole diffusion of sulfate from the sediment/water interface give rise to authigenic barite formation in the overlying Campanian–Paleogene sediments (see "Lithostratigraphy").

The other prominent features seen in the Leg 207 interstitial water data set are chloride anomalies (Fig. F22F). At Sites 1257, 1259, and 1261, we see increases in chloride concentration downhole >60% above standard seawater. Based on the depth profiles obtained at Site 1257 (where data are available from significantly below the black shales), we infer that the shale sequence acts as an aquifer for brines at these sites. At Sites 1258 and 1260, we observe relatively low salinity and chloride concentration anomalies between ~300 and 500 mbsf (up to 17% freshening relative to seawater). Low-chlorinity anomalies such as those seen at Site 1258 are not easy to interpret with confidence on the basis of shipboard data alone. Significant concentrations of methane in headspace gas analyses (>50,000 ppmv) (see "Organic Geochemistry") are consistent with the anomalies having been caused by dissociation of gas hydrates. Alternative explanations for the chloride anomalies are clay dehydration reactions and dilution by meteoric water. The former possibility seems unlikely given the lithologies encountered, but the latter possibility cannot be excluded even though the nearest landmass is located ~350–400 km away.

Four of the five sites were logged during Leg 207 (Holes 1257A, 1258C, 1260B, and 1261B), all cores were passed through the MST, and discrete samples were measured for index properties and velocity. Downhole logging runs included the triple combo tool string with the Multi-Sensor Spectral Gamma Ray Tool (MGT) and the FMS-sonic and WST tool strings, acquiring borehole diameter and wall imagery, formation acoustic velocity, density, porosity, electrical resistivity, natural gamma radiation, and direct acoustic traveltimes. Logging data from Leg 207 show a high degree of correlation with the physical property data, which includes index properties, P-wave velocities, GRA densities, and NGR emissions.

Index property trends, corroborated by the porosity, density, and velocity logs, generally indicate normal consolidation from the surface to the depth of the K/T boundary. Significant perturbations in these profiles are coincident with the P/E and K/T boundaries as well as various hiatuses and mass flow deposits. At Sites 1257–1260, acoustic velocity increases linearly with depth until the P/E boundary (Fig. F23). In the interval between the P/E and K/T boundaries, velocity varies about a fixed baseline, with cyclical fluctuations reflecting changes in sediment composition and degree of lithification. These variations are well represented in the high-resolution FMS images, MGT gamma ray and porosity logs, and MST GRA densities, all providing an excellent opportunity for cyclostratigraphic analysis. At Site 1261, a 310-m-thick Neogene sequence of normally consolidated clay-rich material overlies a 60-m-thick debris flow and a highly indurated middle Eocene sequence that is unique in character to this particular site (Fig. F23).

Across the K/T boundary and through most of the Cretaceous, all sites show a high degree of similarity in the physical property and downhole logs. Highly lithified Maastrichtian chalks gradationally change to low-velocity, high-porosity calcareous claystones with glauconite-rich horizons at their base. This change is readily tracked in both the downhole MGT natural gamma ray logs and the NGR records from the MST.

Through the black shale intervals, the wireline data provide a continuous record of the stratigraphy that is readily interpreted in terms of organic- and clay-rich intervals interbedded with carbonate-cemented layers. High-resolution MST data through the shales shows excellent agreement with the downhole logs. Coupled with high recovery through these sequences, the generation of a core-log composite stratigraphy at all logged sites should be possible. Figure F24 shows the black shale sequence in Hole 1260B, with the carbonate-cemented layers readily picked out with low gamma ray and porosity and high resistivity (white bands in the FMS images), density, velocity, and photoelectric absorption cross-section index (Pe) values. Strong physical property contrasts characterize the transition from the black shales to the underlying Albian synrift sediment. These latter rocks are characterized by very high velocity and density values.

The seismic stratigraphy for the Demerara Rise study area is characterized by five key reflectors (Reflectors O, A, B, B´, and C) that define four main seismostratigraphic units (Units 1–4) and one minor unit (Unit Q). A seismic track map with seafloor bathymetry is presented in Figure F2. In each site chapter, physical property data (in particular, velocity and density) derived from laboratory measurements on core and from downhole logging were used to generate synthetic seismograms. These seismograms were matched to seismic reflection profiles across each site to provide quantitative ties from core and logging results, which are in the depth domain to seismic data that are in the time domain. The results of these ties are summarized in Figure F25. Although these data are presented in each site chapter, there was no attempt to correlate seismic reflection profile data to lithology. Using the information derived from the synthetic seismograms, seismic profiles across each drill site have been depth migrated and matched against the summary lithologic column (Figs. F26, F27, F28, F29, F30). These results permit the seismic profiles to be interpreted with certainty and allow lithostratigraphic units to be correlated site to site and regionally across Demerara Rise.

Reflector C is clearly defined as the base of the black shales, and the underlying sediments of Unit 4 are Albian-age claystone, clayey siltstone, and sandstone. Site 1260 sampled the deepest below Horizon C (40 m). It is uncertain whether any of these sediments were truly synrift or whether they represent undeformed shallow-water clastic sediments following rifting. Horizon C is presently at ~2650 mbsl at Site 1261 (the shallowest site) and dips 0.7° to the north (Fig. F31). At Site 1258 (the deepest site), Horizon C is at 3700 mbsl.

The complete sediment sequence from the seafloor to Horizon C is >1000 m thick south of Site 1261. It gradually thins to the north and rapidly thins near the flanks of the rise. It is ~200 m thick on the upper slope of the flank and then rapidly drops to zero, as the sediment is not retained on the very steep slopes (see Fig. F4 in Shipboard Scientific Party ["Site Survey and Underway Geophysics"], this volume).

Seismic Unit 3 lies between Reflectors C and B. Horizon B´ is within seismic Unit 3, and it correlates with the top of the black shale sequence (lithostratigraphic Unit IV) at each site, except Site 1259. At this location, it correlates slightly above Unit IV, probably due to an expanded gradational sequence of organic-rich glauconitic claystone above the black shales. The top of seismic Unit 3 is Reflector B, a regional high-amplitude reflector that correlates with the K/T boundary at every site. This surface dips rather uniformly at ~1° to the northwest. The thickness of Unit 3 is locally variable but, in general, increases from ~120 m in the south (Site 1261) to >300 m thick in the northwest between Sites 1258 and 1260.

Seismic Unit 2 lies between Reflectors B and A. Horizon A correlates everywhere to the top of lithostratigraphic Unit II, a variable nannofossil chalk sequence that is mostly Eocene but possibly as young as early Miocene. Horizon A is believed to result from an episode of extensive erosion during the early Miocene. The surface of Reflector A dips toward the northwest and is at ~2100 mbsl at Site 1261 and 3200 mbsl at Site 1258. This surface appears to be channelized, but seismic line density is insufficient to discern much detail. Unit 2 is highly variable in thickness as a result of differential amounts of erosion but, in general, thins toward the northwest. At Site 1260, for example, the seismic unit is 180 m thick, whereas at Site 1259 it is 400 m thick.

Reflector A forms the base of seismic Unit 1. The top of Unit I is typically the seafloor because Reflector O and seismic Unit Q are rarely present. Unit 1 correlates with unlithified or semilithified sediment, such as the 300-m-thick succession of Miocene–Pliocene nannofossil clay at Site 1261. The unit is absent at other sites. Seismic lines show that its thickness is highly variable because its lower bounding surface is Reflector A. Its maximum thickness is >350 m, and it thins toward the west and northwest.

Seismic Unit Q rarely is resolved on seismic reflection profiles, but at Site 1261 it is nearly 30 m thick (Fig. F30). At this location, Reflector O correlates with the base of lithostratigraphic Subunit IA, a Pleistocene nannofossil ooze. Reflector O represents an unconformity, separating Pleistocene and younger sediment from the middle Miocene. Reflections in Unit 1 sometimes truncate against Reflector O.

In summary, strong physical property contrasts in lithostratigraphic units, as described from Leg 207 drilling results on Demerara Rise, translate to excellent correlation with seismic units and even with some single-reflection events. Depth-migrated seismic reflection profiles consistently match summary lithologic columns as a result. Seismic reflection data can be used to confidently correlate from site to site and across the Demerara Rise study area.

The mid-Cretaceous (Albian) to Neogene history of Demerara Rise reflects multiple depositional episodes during the late rifting and drifting stage of a continental margin that saw different sedimentological regimes separated by several hiatuses of varying duration. Transform movements related to the opening of the Equatorial Atlantic Gateway are believed to have influenced the sedimentary history during the Cenomanian–Campanian stages.

The oldest sediments recovered during Leg 207 are early Albian in age. At the deepest site of the depth transect (Site 1258), they are represented by organic carbon–rich claystones with phosphatic pebbles. Ammonites and microfossils recovered from this interval indicate open marine conditions in a probable epicontinental basin. Upslope (at Site 1260), an increase of siliciclastic detritus indicates a sedimentary environment closer to the sediment source. Probable tidal flat deposits of unknown age (Albian?), represent the oldest sequence recovered at the shallow end of the transect (Site 1259). Shallow-water environments are also indicated by marine quartz sandstone with occasional ammonite casts and shell debris recovered at the shallow sites (Sites 1261 and 1259) and at Site 1257, where they unconformably overlie middle–upper Albian open marine clays. These sandstones are late Albian–late Cenomanian in age. Early Cretaceous shallow-water carbonates and Aptian–Albian open marine marls have been recovered in industry wells farther south. These data suggest that Demerara Rise was a submarine high, separated from South America by a shallow-marine epicontinental basin.

Black shale deposition at Demerara Rise started in the middle Albian at the deep end of the transect. At this location, sediments are overlain by a lower–middle Cenomanian sequence of storm-induced layers intercalated with black shales, which might be laterally equivalent to the quartz sandstone upslope. No upper Albian sediments have been identified. Elsewhere, middle–upper Cenomanian black shales overlie a prominent unconformity that separates them from the underlying Albian–lower Cenomanian clays and sandstones, which represent the synrift deposits on Demerara Rise.

Cyclic successions of middle–upper Cenomanian organic matter–rich black shale with abundant phosphatic pebbles and fish remains and laminated foraminiferal packstone are believed to represent the first shallow hemipelagic sediments on Demerara Rise. They reflect high surface water productivity and dysoxic bottom water conditions. The thickness of the Cenomanian sequence on Demerara Rise changes greatly (8–80 m). This may be the result of either varying subsidence histories of different small slope basins or the middle–late Cenomanian sea level rise and the consequent transgressive onlap. However, the onset of black shale deposition at the shallow sites dates as late Cenomanian. There, dark TOC-rich shales yield very low diversity reduced planktonic faunas and calcareous nannoplankton of the genus Eprolithus, which indicate a very shallow-marine to lagoonal setting.

High-productivity conditions, probably related to local upwelling, continued until the early Campanian, with most extreme conditions during OAE 2. Clayey bentonite layers indicate the proximity of volcanoes. Occasional glauconite-rich horizons are interpreted as either an oxygenation event or reduced sedimentation rates. Debris flow deposits, slumped intervals, and condensed horizons occur in the upper Turonian–lower Campanian section. Black shale deposition on Demerara Rise ends in the Santonian–early Campanian.

No middle–late early Campanian sediments were recovered, and the upper Campanian yields glauconite-rich horizons indicating reduced sedimentation rates. It is believed that the cessation of black shale sedimentation, low sedimentation rates, and hiatuses are related to oceanographic modifications and transform motions following the opening of the Equatorial Atlantic Gateway. Oxic conditions were established by the late Campanian, when sedimentation on Demerara Rise changed from hemipelagic to pelagic. The abundance of radiolarians in the Campanian indicates high surface water productivity. The cyclic pattern of trace fossil abundance suggests repeated recurrence of variations in bottom water oxygenation.

Maastrichtian–Oligocene sediments at Demerara Rise consist of pelagic deep marine nannofossil and foraminiferal chalks and oozes deposited in an upper bathyal setting. The succession records the lithologic consequences of the K/T boundary impact and the abrupt global warming associated with the P/E boundary. The high abundance of radiolarians in middle Eocene sediments is related to upwelling conditions. Because of a major erosional event in the early Miocene, Oligocene sediments on Demerara Rise are often reworked or reduced in thickness.

Neogene sediments on the slopes of Demerara Rise are rather patchy. As a probable result of Amazon plumes, upper Miocene–Pliocene sediments are clay rich and sedimentation rates can be very high (Site 1261). However, no upper Pliocene and only very thin veneers of Pleistocene and Holocene clay-rich sediments are present.