15N is inversely correlated with total organic carbon in OAEs, suggesting that the layers contain marine carbon deposited when nitrogen fixation was abnormally elevated in the surface waters.
Prerift sediments have been cored in Hole 639D on the western margin of Galicia Bank, where Tithonian shallow-water limestone and marlstone with lesser sandstones and claystones are capped by Tithonian to ?Berriasian dolomites (Boillot, Winterer, Meyer, et al., 1987). These were interpreted by Boillot et al. to have been deposited under shelf to lagoonal conditions prior to rifting of the underlying crust.
In contrast to Hole 639D, Tithonian sediments on the southern edge of Galicia Bank at Sites 1065 and 1069 under the southern Iberia Abyssal Plain are dominated by gray turbiditic to hemipelagic clays and claystones, with thin conglomerates and clasts of shallow-water limestones and ?Paleozoic metasediments (Whitmarsh, Beslier, Wallace, et al., 1998). At Site 901 the Tithonian sediments are olive-black clays that are locally silty, dolomitic, and rich in plant debris (Sawyer, Whitmarsh, Klaus, et al., 1994). These sediments differ markedly from the coeval red to gray-green limestones of the Cat Gap Formation, which was deposited farther south in the deep western North Atlantic Basin (Fig. F4). The dearth of carbonate and presence of turbidites in the Newfoundland-Iberia rift suggest that deposition was at least at moderate depths, rather than in shallow water. Studies of benthic foraminifers at Site 901 and calcareous nannofossils at Sites 1065 and 1069, together with evidence from the lithofacies, indicate that deposition was under neritic, dysoxic conditions in a marginal shelf basin at Site 901 (Collins et al., 1996) and in a restricted interior basin with little communication with the open ocean at Sites 1065 and 1069 (Concheryo and Wise, 2001). Thus, the crust along the southern Galicia margin probably was rifting by Tithonian time, creating a series of extensional graben that were relatively isolated from the main North Atlantic Basin. Thin laminae of nannofossil ooze appeared at Sites 901 and 1065 in the latest Tithonian, suggesting increasing connections with the open ocean, although dysaerobic conditions still persisted at the seafloor (Concheryo and Wise, 2001).
The Early Cretaceous (Berriasian to early Barremian) in the main North Atlantic Basin was a period characterized by a deep calcite compensation depth (CCD) and by deepwater deposition of calcareous sediments comprising the Blake-Bahama Formation (Fig. F4) (Jansa et al., 1979). The same conditions appear to have been present within the Newfoundland-Iberia rift. Upper Berriasian to lower Valanginian nannofossil chalks were deposited at Site 1069 and Valanginian–Barremian breccias containing basement clasts in a chalk matrix were deposited at Site 1068 to the east (Whitmarsh, Beslier, Wallace, et al., 1998), while upper Hauterivian–lower Barremian nannofossil limestones with interbedded mudstones were deposited at Site 398 (Sibuet, Ryan, et al., 1979). On the western Galicia margin to the north, ?Berriasian, presumably shallow-water dolomites were rifted and succeeded by deposition of Valanginian–Barremian marlstones in deepening rift basins (Boillot, Winterer, Meyer, et al., 1987). These were augmented by input of silty and sandy turbidites (in places arkosic) during the Valanginian–Hauterivian as the rift topography was eroded.
A marked change in basin conditions occurred during the Barremian as the CCD shoaled and the deep basin became less oxygenated. Sediment facies changed to predominantly gray-green to black hemipelagic claystones (locally carbonate or silica rich) that commonly were interrupted by coarser-grained turbidites from the continental margins. These sediments are ubiquitous on both the Newfoundland and Iberia margins, even up through the Cenomanian, and they are equivalent to the Hatteras Formation in the main North Atlantic Basin to the south (Fig. F4) (Tucholke and Vogt, 1979; Jansa et al., 1979).
A notable facies consisting of breccias occurs atop the peridotite basement highs drilled at Sites 897, 899, 1070, and 1277 (Sawyer, Whitmarsh, Klaus, et al., 1994; Whitmarsh, Beslier, Wallace, et al., 1998; Tucholke, Sibuet, Klaus, et al., 2004). These breccias contain varying proportions of sedimentary, serpentinite, gabbro, and basalt clasts and appear to have been deposited by mass flows. At least some of these deposits probably were sourced from local topography (Gibson et al., 1996), although it has been argued that at Sites 897 and 899 they were derived from more distal sources before the basement blocks were faulted and uplifted (Comas et al., 1996). Ages of the deposits are late Barremian to late Aptian (Site 897), early Aptian (Site 899), and late Aptian (Site 1070). No age-diagnostic fossils were found in the mass flows at Site 1277, so it is uncertain when they were deposited (Tucholke, Sibuet, Klaus, et al., 2004).
The widespread reflection marking the Aptian event and the interpreted cessation of rifting near the Aptian/Albian boundary has been drilled in only three holes where it is intact (Sites 398, 641, and 1276). In other holes it correlates with a significant hiatus or it is not present, primarily because the holes were drilled on basement highs. At Site 398, Aptian sediments below the reflection are black to green-gray mudstones interbedded with debris flows, mud-supported conglomerates, and turbidites, contrasting with overlying Albian claystone and mudstone that is dominantly laminated or burrowed (Sibuet, Ryan, et al., 1979). Tucholke et al. (2007) suggested that this sharp reduction in mass flows was caused by subsidence of source areas when in-plane stresses relaxed at the end of rifting. At Site 641, Aptian sediments are dominated by marlstones and limestones, and they are succeeded by Albian black to gray-green claystones (Boillot, Winterer, Meyer, et al., 1987). Because Aptian sediments elsewhere in the rift are dominated by dark claystones, this change probably reflects subsidence of the seafloor, possibly coupled with a further rise in the CCD. At Site 1276 there appears to be no significant change in lithology that can be associated with the Aptian event (U reflection) (Tucholke, Sibuet, Klaus, et al., 2004). As discussed above (see "Sills at Site 1276"), the strong reflection at this site is due to the presence of the upper diabase sill.
It is noteworthy that the reflectivity of the Aptian event and the underlying sediments around many of the peridotite ridges on the Galicia margin is exceptionally strong adjacent to the ridges but fades with distance from the ridges (Fig. F6) (Tucholke et al., 2007). Tucholke et al. (2007) proposed that the highly reflective sediments were deposited by extensive mass wasting from the weak, serpentinized peridotites; the mass wasting was strongly curtailed when in-plane stresses were relaxed at the end of rifting and the peridotite ridges subsided. Overall, the reduction in mass wasting at the end of rifting (leading to deposition of finer-grained sediments), combined with subsidence (and thus a change from deposition of calcareous to noncalcareous sediments), may largely account for the pronounced reflectivity and rift wide distribution of the reflection that marks the Aptian event.
Gray-green to black mudstones and claystones with lighter, commonly marly intervals continued to accumulate until the end of Cenomanian to early Turonian time, as they did in the North Atlantic Basin farther to the south (Fig. F4). In the Newfoundland-Iberia rift, these sediments are represented primarily at Sites 398, 641, and 1276.
At Site 1276 the lower (1502–1719 mbsf) and upper (1067–1130 mbsf) parts of this succession are dominated by muddy gravity-flow deposits with a wide variety of sedimentary structures, whereas the central part is largely burrowed hemipelagic mudstone (Tucholke, Sibuet, Klaus, et al., 2004; Shirai et al, this volume) (see Sawyer and Fackler, this volume, for an automated method used to associate core lithology to measured properties). Sediments in the middle and upper parts of the succession accumulated at rates of ~18–23 m/m.y. whereas the lower part was deposited at much faster rates of ~69 m/m.y. to as much as 105 m/m.y. (Tucholke, Sibuet, Klaus, et al., 2004; Ladner, this volume). Overall dark color of the sediments and lack of burrowing in the upper and lower intervals indicates low-oxygen conditions (dysoxic to anoxic) at and/or below the seafloor, and this is supported by elevated concentrations of redox-sensitive trace metals in the sediments (Robertson, this volume; Arnaboldi and Meyers, this volume). Reducing conditions in the central, more burrowed part of the section were either less severe or restricted mostly to the subsurface, or both.
Total organic carbon (TOC) in the Albian–Cenomanian sequence at Site 1276 varies from 0 to 11.7 wt%, with generally low values (~1 wt% or less) in the upper part of the sequence increasing to normally >1 wt% in the lower, Albian part (Tucholke, Sibuet, Klaus, et al., 2004;
Arnaboldi and Meyers, this volume). Most TOC values >2 wt% occur in black shales represented by oceanic anoxic events (OAEs). Five OAEs have been recognized: (1) the Cenomanian–Turonian OAE 2 ("Bonarelli" event); (2) the "Mid-Cenomanian Event;" and (3) to (5) the Albian OAE 1b, OAE 1c, and OAE 1d events (Fig. F4). In addition, another possible OAE with total organic carbon as high as 3.2 wt% was detected in the middle Albian section.
Arnaboldi and Meyers (this volume) report that the organic carbon throughout the section was derived from both terrestrial and marine sources, although there appears to be a dominance of marine carbon in OAE 2 and OAE 1b. In addition, 15N is inversely correlated with total organic carbon in OAEs, suggesting that the layers contain marine carbon deposited when nitrogen fixation was abnormally elevated in the surface waters.
In gray-green to black claystones of the same age on the Iberia margin, TOC values are predominantly low at Site 398 (<1 wt%) (Sibuet, Ryan, et al., 1979), but higher average values (2–4 wt%) are present at Site 641 (e.g., Boillot, Winterer, Meyer, et al., 1987; Dunham et al., 1988). As at Site 1276, the TOC is a mixture of terrestrial and marine carbon, with higher values generally occurring in black layers. OAE 2 is present at both Sites 398 and 641, and it is dominated by marine carbon, with TOC values as high as ~10–13 wt% (Dunham et al., 1988; Thurow et al., 1988).
Terrestrial sources provided most of the sediment deposited in the Newfoundland Basin during this period (Robertson, this volume), and the primary source was probably the adjacent Grand Banks. Studies of fluid inclusions in detrital quartz grains from this and the overlying sedimentary sections (Shryane and Feely, this volume) show that the fluid signatures are typical of those in detrital quartz derived from granitic sources. Hiscott (this volume) studied paleoflow directions determined from ripples and grain fabric in turbidites and deduced that the currents probably flowed primarily to the north-northeast. This suggests a source area on the southeastern Grand Banks in the area of the Avalon uplift. This area was elevated from at least Middle Jurassic to Cenomanian time (Jansa and Wade, 1975; Grant et al., 1988), as evidenced by the long-ranging hiatus represented by the Avalon unconformity in this area. Surprisingly, there is little evidence for flows from Flemish Cap to the north, perhaps because of the much smaller area of this feature or because it was submerged at the time. Sandstone compositions (Marsaglia et al., this volume) are consistent with one major terrigenous source on the Grand Banks, and relatively low potassium feldspar contents distinguish the Site 1276 sandstones from those deposited off Iberia at the same time. The composition resembles a "recycled orogen" provenance (Dickinson, 1985), suggesting that the sandstones were recycled from foreland basin sediments originally deposited during the Paleozoic (Marsaglia et al., this volume). White micas from turbidites in the Site 1276 Albian sequence have 40Ar/39Ar ages between 250 and 340 Ma (Wilson and Hiscott, this volume). Wilson and Hiscott interpret the direct source of these micas to be the Meguma Terrane at the southeast end of the Grand Banks, possibly originally emplaced during the Acadian orogeny (~375–415 Ma) but possibly later reactivated and metamorphosed during the Alleghenian orogeny (~260–350 Ma). However, they also note that the original provenance could have been Iberia if Iberia sediments were shed westward into foreland basins (now under the easternmost Grand Banks) during the Alleghenian orogeny.
Deep-basin conditions in the rift changed dramatically near the Cenomanian/Turonian boundary from dysoxic/anoxic to well oxygenated, and this oxygenated state mostly persisted through the late Paleocene. The deep basin remained well below the CCD. The correlative facies in the main North Atlantic Basin is the dominantly reddish to multicolored shales of the Plantagenet Formation (Jansa et al., 1979). The cause of the change to well ventilated deep basins and presumably increased deep circulation is uncertain, but it has been suggested that it relates to a deep-basin connection that became established between the North and South Atlantic at about this time (Tucholke and Vogt, 1979).
At Site 1276, change to a well oxygenated seafloor occurs in the middle to upper Turonian. It is marked by a change to deposition of reddish, highly bioturbated, siliciclastic muddy sandstones and sandstones, both of which have very low TOC values (0–0.4 wt%). In addition to the low TOC values, strongly oxidizing conditions are indicated by locally high concentrations of MnO (Robertson, this volume). These sediments were deposited at very low rates (<2 m/m.y.) until earliest Campanian time (Tucholke, Sibuet, Klaus, et al., 2004); there is poor biostratigraphic control in this interval and hiatuses may be present. There was no significant change in sandstone detrital modes in the Upper Cretaceous section compared to the deeper Albian–Cenomanian sediments (Marsaglia et al., this volume), and white mica in the Turonian sediments shows the same age distribution and presumed source area as the underlying Albian section (Wilson and Hiscott, this volume). Thus, the sandstones continued to be deposited from gravity flows from the Grand Banks, but the extremely low sedimentation rates indicate that this source was highly attenuated, probably because of elevated eustatic sea level in the Late Cretaceous (Haq et al., 1988). The general scarcity of fines also indicates that the sediments were strongly winnowed by abyssal currents.
Reddish to multicolored sediments were also deposited on the conjugate Iberia margin at this time, although they are mostly pelagic claystones to hemipelagic silty claystones (Sibuet, Ryan, et al., 1979; Boillot, Winterer, Meyer, et al., 1987; Whitmarsh, Beslier, Wallace, et al., 1998). In contrast to Site 1276, these sediments show little indication of modification by deep currents. Thus, the currents appear to have been westward intensified, primarily affecting only the Newfoundland margin.
The lowermost Campanian to upper Paleocene sediments at Site 1276 consist of low-TOC, largely unburrowed reddish to multicolored claystones and mudstones that accumulated at rates of ~4–6 m/m.y.; they contain a low proportion (~20%) of coarser, more calcareous beds that were introduced mostly by mass flows and are concentrated in the lower part of the section (Tucholke, Sibuet, Klaus, et al., 2004). The carbonates herald increasing input of calcareous sediments that predominated in the late Paleocene and Eocene. The shift away from deposition of coarse siliciclastic sediments suggests declining input of terrigenous sediment from the adjacent Grand Banks, even though sedimentation rates increased. This apparent paradox may be explained if we consider the abundant fines in the sediments. These were not winnowed away as they were in the underlying sandstones, suggesting a marked reduction or cessation of vigorous abyssal circulation beginning in the early Campanian. The lower and particularly the middle Paleocene part of the section contain numerous dark gray to black beds (Fig. F4), most of which are associated with gravity flows. Although these sediments are low in TOC (mostly <1 wt%), the dark colors indicate reducing conditions, at least below the sediment surface, and they also support the idea that deep circulation was sluggish and the basin was not well oxygenated during this time. A similar observation was made at DSDP Sites 386 and 387 on the Bermuda Rise (Tucholke and Vogt, 1979), which indicates that this phenomenon may have extended throughout the deep basins of the North Atlantic.
Campanian to Paleocene sediments on the Iberia margin at Sites 1068 and 1069 are broadly similar to those at Site 1276 (Whitmarsh, Beslier, Wallace, et al., 1998), but at the ~1-km shallower seafloor of Site 398 the sediments are dominated by red-brown marly chalk (Sibuet, Ryan, et al., 1979). Although dark gray-green beds associated with gravity flows are present at these sites, they are not as dark as the Paleocene beds at Site 1276 and thus appear not to be as reduced. All these sites were drilled on basement highs, and thus the sediments were deposited on shallower, possibly more oxygenated seafloor than that at Site 1276.
At many drill sites in the main North Atlantic Basin the upper Maastrichtian is represented by the carbonate-rich Crescent Peaks Member of the Plantagenet Formation (Fig. F4), which is interpreted to have been deposited during a deep excursion of the CCD (Tucholke and Vogt, 1979). This facies is represented on the Iberia margin by the marly nannofossil chalks at Site 398 and by nannofossil chalks at Site 1068, but it is not present at Sites 1069 and 1070 (Sibuet, Ryan, et al., 1979; Whitmarsh, Beslier, Wallace, et al., 1998). The last three sites all penetrated this stratigraphic interval at ~5900 m below present sea level, so the same facies might be expected at each site. However, because the sites were drilled on basement highs it's possible that highly calcareous sections originally deposited at Sites 1069 and 1070 were later removed by mass wasting. Maastrichtian sediments at Site 1276, penetrated at ~5550 m below present sea level, are dominantly mudstones and claystones similar to underlying and overlying beds. Sedimentation rates did not change during this period (Tucholke, Sibuet, Klaus, et al., 2004), so a pulse of pelagic carbonate deposition would not have been diluted and masked by increased sediment input from shallow water. Thus, it appears that there was significant asymmetry in CCD depth between the two margins of the rift, with Site 1276 below the CCD while deeper seafloor on the Iberia margin was above it.
One unusual sedimentary component, consisting of mafic and felsic volcaniclastic debris within turbidites in Core 210-1276A-15R, was recognized in the middle Paleocene (~56 Ma) (Urquhart et al., this volume) section at Site 1276. Mica 40Ar/39Ar ages from the turbidites range from 55 to 74 Ma (plus one 186-Ma sample), with 8 of the 13 ages dating to 55–61 Ma (Wilson and Hiscott, this volume). Marsaglia et al. (this volume) also detected minor volcanic debris in samples ranging from Santonian to Eocene age at Site 1276, and they found hitherto unrecognized mafic and felsic(?) volcanic detritus in Iberia cores at Site 897 (Upper Cretaceous?), Site 1068 (lower Eocene), and Site 1069 (upper Paleocene). There was scattered magmatic activity around the rift during the Late Cretaceous to Paleocene (Table T1), but there is little evidence for subaerial volcanism that could have dispersed significant amounts of ash over the basins. Thus, it seems likely that most of these occurrences were from local sources. A significant exception may have been related to the opening of the Norwegian-Greenland Sea, which produced voluminous amounts of volcanic rocks (Sinton and Duncan, 1998; Hopper et al., 2003) and might have been a source of subaerial volcaniclastic ash at ~55–56 Ma (Storey et al., 2007). Given the indications for weak bottom circulation in the middle Paleocene noted above, however, it seems unlikely that volcaniclastic debris would have been transported into the Newfoundland-Iberia rift by deep currents.
In contrast to the findings of Marsaglia et al. (this volume), Robertson (this volume) conducted geochemical studies of the fine-grained "background" sediments enclosing the turbidites at Site 1276 and found that volcaniclastic debris appears to be isolated in the turbidites, indicating derivation from a local source rather than from more widely dispersed volcanism. Concentration of micas in the turbidites and their relatively limited age grouping within the Paleocene (Wilson and Hiscott, this volume) also are consistent with a local provenance. One possible source area is the Newfoundland Seamounts to the south, which would be consistent with transport directions of turbidity currents (albeit in the Albian section) that were determined by Hiscott (this volume). Although the only known volcanism at the seamounts occurred in the Cenomanian (Table T1), it is possible that volcanism continued there or recurred at a later time. It is also possible that the volcaniclastic debris was derived from undocumented sources on the Grand Banks. Previously unsuspected igneous rocks have been penetrated in exploration wells there (e.g., Cenomanian dikes in the Emerillon C-56 well) (Table T1), and there could be other, younger intrusive or volcanic rocks present that are yet to be discovered.
In the main North Atlantic Basin, upper Paleocene to middle Eocene sediments are characteristically biosiliceous, cherty claystones that characterize the Bermuda Rise Formation (Fig. F4). The time equivalent sequence at Site 1276, however, contains only a minor biosiliceous component. It consists of carbonate grainstones and marlstones with subordinate mudrock and represents a record dominated by deposition from mass flows, largely below the CCD (Tucholke, Sibuet, Klaus, et al., 2004). The lower part of the sequence is dominated by siliciclastic turbidites, and the proportions and types of siliciclastic grains indicate sources similar to those of the underlying section (Marsaglia et al., this volume). In the middle Eocene section, white-mica ages continue to group at 250–340 Ma, but a new group (402–428 Ma) of white micas appears, creating a bimodal mica-age distribution (Wilson and Hiscott, this volume). This distribution suggests addition of another source of siliclastic detritus, either because of shifting sediment supply routes or because of erosional exposure of older rocks. The upper part of the upper Paleocene to middle Eocene sequence shows a strong shift to largely calcareous (grainstone) turbidites (Marsaglia et al., this volume), indicating progressive development of carbonate-bank deposits in source areas on the adjacent shallow continental margin. Mineralogy of the smectites in the fine-grained sediments suggests that the source areas were relatively warm and humid (Tucholke, Sibuet, Klaus, et al., 2004).
A very diverse and abundant assemblage of mostly reworked, large benthic foraminifers was identified in the uppermost Paleocene to lowermost middle Eocene part of the section (Georgescu et al., submitted [N2]). This assemblage indicates the presence of a carbonate shelf in the adjacent source area of the Grand Banks, consistent with the above interpretations of source area based on sediment composition. Most of the benthic taxa are representative of the Caribbean bioprovince, and Georgescu et al. (submitted [N2]) propose that the warm climate of the Paleocene/Eocene Thermal Maximum, together with biodispersal by the proto-Gulf Stream, accounts for their presence that this high latitude.
The conjugate Iberia margin (Sites 398, 897, 1068, and 1069) similarly shifted to accumulation of carbonate-rich sediments during this period (Sibuet, Ryan, et al., 1979; Sawyer, Whitmarsh, Klaus, et al., 1994; Whitmarsh, Beslier, Wallace, et al., 1998), although the bulk of the carbonate consists of nannofossils. Only Site 398 shows a significant biosiliceous component, with siliceous marly chalk and radiolarian mudstone deposited during the late middle to early late Eocene. Thus, overall, the sedimentary record within the widening rift demonstrates productive surface waters, but the plankton were dominantly calcareous, in contrast to the rich siliceous fauna and flora farther south in the North Atlantic.
Sediments deposited during the middle Eocene to early Oligocene were generally less calcareous than those in the underlying section. Site 1276 accumulated claystones and mudstones below the CCD, interrupted by occasional coarser turbidites (Tucholke, Sibuet, Klaus, et al., 2004). Benthic foraminifers in the nonturbidite sediments of this section are primarily agglutinated forms comprising faunal associations similar to those in Eocene sediments previously cored in the Labrador Sea (ODP Leg 105) and off Iberia (ODP Legs 149 and 173) (Takata, this volume). The dearth of calcareous forms in the "background" sediments is consistent with deposition on a seafloor below the CCD. Claystones, silty claystones, and nannofossil claystones with interbedded turbidites also were deposited on the Iberia margin, with only Site 398 accumulating largely marly nannofossil chalk with lesser mudstone (Sibuet, Ryan, et al., 1979).
The middle Eocene to lower Oligocene section at Site 1276 should contain the event that correlates with the Horizon AU reflection farther south in the western North Atlantic Basin. Along that continental margin Horizon AU commonly shows unconformable relations with underlying reflections and it correlates with a major hiatus near the Eocene/Oligocene boundary (Tucholke and Mountain, 1979; Miller and Tucholke, 1983). It has been interpreted to record the initiation of strong abyssal circulation in the North Atlantic Ocean, probably correlating with the first significant flow of cold bottom waters over the Greenland-Scotland Ridge from the Norwegian-Greenland Sea (Miller and Tucholke, 1983). One candidate for the Horizon AU equivalent at Site 1276 is at ~7.02 s two-way traveltime (AU1, Fig. F3), one of several coherent reflections below the top of a sequence of strong, relatively level reflections at Site 1276. Traced landward, it becomes the top of this reflection sequence, and the overlying reflections become less coherent and splay into a landward-thickening wedge (Fig. F3). A second likely candidate (AU2) is at the top of this wedge. This reflection separates the wedge from overlying beds that show thickening–thinning patterns suggestive of sediment waves; the wedge itself nearly pinches out at Site 1276, where the AU2 reflection is at ~6.96 s reflection time.
Shillington et al. (this volume) constructed synthetic seismograms using core physical property data from Site 1276, and their correlations to the seismic reflection record suggest that AU1 falls at 864.7 mbsf, matching the lithologic Unit 1/2 boundary at the site (Tucholke, Sibuet, Klaus, et al., 2004). Wood et al. (submitted [N3]) examined the nannofossil biostratigraphy of this interval in detail. They concluded that the Unit 1/2 boundary is represented by a condensed interval or hiatus that spans between 1.2 and 6.9 m.y. and is contained within the middle Eocene (47.3–40.4 Ma). They also noted that cooler-water taxa are more abundant above this level than below, which could indicate a change in circulation patterns, although the depth interval they analyzed is very restricted (<3 m) and may not be representative. Wood et al. (submitted [N3]) compared their stratigraphic results with Site 398 on the Iberia margin and noted a comparable break in the record there at the same age. Because of the nearly identical stratigraphic gaps at these widely separated locations, they suggested that the condensed intervals/hiatuses on the two margins were caused by high eustatic sea level that restricted sediment supply to the deep basins. This is consistent with Site 1276 shipboard sediment descriptions that define an increased hemipelagic component in lithologic Unit 1 compared to Unit 2 and that note an absence of any obvious structures that could be related to abyssal current effects (Tucholke, Sibuet, Klaus, et al., 2004). Considered together, these results suggest that AU1 does not mark the initiation of strong abyssal circulation and therefore is not equivalent to Horizon AU. In this case the sedimentary wedge between AU1 and AU2 probably was formed by downslope sediment movement rather than by current-controlled deposition. Its restriction to a near-slope position may reflect limited sediment input associated with elevated sea level.
The AU2 reflection, which underlies wavelike bedforms in the seismic reflection record (Fig. F3), appears to be a better match to Horizon AU, and this is our preferred interpretation. According to the seismic/borehole correlation of Shillington et al. (this volume), this reflection falls at the very top of the cored sediment section at 800 mbsf. Sediments in samples from Sections 210-1276A-1W-4 and 1W-CC are dated by calcareous nannofossils and dinocysts as uppermost Eocene to lower Oligocene (Tucholke, Sibuet, Klaus, et al., 2004). Although these samples were from a wash core (753–800 mbsf), it seems likely that the core catcher did recover sediments from the 800-m level; the next core (210-1276A-2R) is also upper Eocene to lowermost Oligocene (palynomorphs) or upper Eocene (nannofossils). Thus, the best available evidence is that the interpreted Horizon AU and the initiation of strong abyssal circulation in the North Atlantic dates very close to the Eocene/Oligocene boundary, as first suggested by Miller and Tucholke (1983).
