Rifting of a continent and birth of a new oceanic spreading center are fundamental yet poorly understood parts of the plate tectonic cycle. Rifted margins are commonly classified into two types, volcanic and nonvolcanic (Mutter et al., 1988; White and McKenzie, 1989; White et al., 1987), although a relatively continuous range of margin types that vary in character according to tectonic stress, lithospheric strength, and mantle conditions is likely (Mutter, 1993). Two principal models of lithospheric extension have been proposed for nonvolcanic margins. In the pure shear model (McKenzie, 1978), crustal thinning is relatively uniform across a rift; brittle deformation causes thinning and faulting of the upper crust and ductile deformation thins the lower crust. This model predicts that conjugate margins will have generally similar crustal thickness, structure, composition, and subsidence history. Progress in modeling of continental rift structure and extensional tectonics, however, together with observations of significant asymmetries in conjugate margins, suggests that many rifts may develop by a simple shear mechanism (e.g., Lister et al., 1986, 1991; Rosendahl, 1987; Wernicke, 1985). Simple shear predicts an upper plate margin consisting of weakly structured upper continental crust with a rift stage history of uplift and a lower plate margin dominated by highly structured lower continental crust and a history of subsidence. Melt generation and attendant volcanism, compared to a pure shear environment, is probably minimal (Buck, 1991; Latin and White, 1990).

As continental plates separate, crustal thinning, volcanism, faulting, uplift, subsidence, and sedimentation profoundly modify the structure of the rifted margins. To understand these processes we need detailed information on the resulting geological record, particularly the basement architecture and the overlying sedimentary framework. Furthermore, in order to evaluate the role of pure shear, simple shear, or other mechanisms of rift extension, it is essential to examine the geological record of conjugate rifted margins. This is best done by acquiring and analyzing wide-angle reflection/refraction and vertical-incidence reflection data along carefully chosen conjugate margin transects and then by sampling critical sections by drilling.

In the early 1990s, the Joint Oceanographic Institutions for Deep Earth Sampling (JOIDES) North Atlantic Rifted Margins Detailed Planning Group recommended the Newfoundland and Iberia conjugate margins as high-priority drilling targets to understand the evolution of nonvolcanic rifts (Fig. F1). These margins presented several advantages for such a study:

1. They are considered to be representative of nonvolcanic rifting.
2. Rifting is complete, so the entire rift history can be studied.
3. The along-rift spatial relations of crustal conjugates are well constrained in plate reconstructions.
4. Sediments are comparatively thin, so important basement targets can be imaged by seismic reflection/refraction and they are accessible to the drill.
5. The locations are logistically convenient, thus facilitating access.

By design of the JOIDES advisory and planning structure, extensive drilling (Ocean Drilling Program [ODP] Legs 149 and 173) was conducted on the Iberia half of the rift (Fig. F1). This complemented earlier drilling (ODP Leg 103) on the western margin of Galicia Bank, and it was supported by extensive geophysical work (e.g., Whitmarsh et al., 1990, 1996; Pinheiro et al., 1992; Reston et al., 1995; Whitmarsh and Miles, 1995; Reston, 1996; Pickup et al., 1996; Discovery 215 Working Group, 1998). An early drill site from Deep Sea Drilling Project (DSDP) Leg 47B (Site 398) also provided valuable information near the Leg 149/173 transect. These studies, summarized below, provided surprising results about the composition and origin of crust in the transitional zone between known continental and known oceanic crust on the Iberia margin. They also raised major questions about how the Newfoundland–Iberia rift developed and whether rifting was symmetrical or asymmetrical. To answer these questions, it is necessary to investigate the structure and evolution of the conjugate Newfoundland margin, which was the major objective of Leg 210. The following section outlines the geological setting of the Newfoundland–Iberia rift and summarizes the results of a site survey that was conducted in preparation for drilling on the Newfoundland margin.

Rift Development and Basement Structure

The Newfoundland and Iberia margins first experienced significant extension in Late Triassic time when rift basins initially formed within the Grand Banks and on the Iberia margin (Lusitania Basin) (Figs. F2, F3). The Grand Banks basins accumulated siliciclastic "redbed" sediments, and these were succeeded by deposition of evaporite deposits, which extended into the earliest Jurassic in both the Grand Banks and Lusitania basins (Jansa and Wade, 1975; Wilson, 1988; Rasmussen et al., 1998). A second prolonged rift phase in the Late Jurassic through Early Cretaceous extended the crust in several subbasins, but it ultimately focused extension between the Grand Banks and Iberia (Tankard and Welsink, 1987; Enachescu, 1987; Wilson, 1989). This culminated in continental breakup and formation of the first oceanic crust no later than Barremian to Aptian time. Excepting the Southeast Newfoundland Ridge at the southernmost edge of the rift, no significant thickness of volcanic rocks or magmatic underplating is known to be present in the rift. Thus the system is considered to be nonvolcanic.

Plate reconstruction of the Newfoundland–Iberia conjugate margins at the time of Anomaly M0 (Barremian/Aptian boundary; ~121 Ma) (Fig. F3) provides a regional overview of the rift and the conjugate margins. At this time, thick continental crust of Flemish Cap was close to extended continental crust of Galicia Bank at the northern end of the rift. To the south, geophysical studies and magnetic anomaly identifications suggest that ocean crust was present, extending landward from a seafloor-spreading axis to at least Anomaly M3 (see shaded area in Fig. F3). Along-axis, the Newfoundland–Iberia rift can be roughly divided into three segments: (1) a northern segment containing Flemish Cap and Galicia Bank, (2) a central segment bounded on the south by the Newfoundland Seamounts and Tore Seamount, and (3) a southern segment extending south to the Southeast Newfoundland Ridge and the present-day Gorringe Bank off southwestern Iberia. Each of these is reviewed below.

In the northern segment, Flemish Cap has full continental crust thickness of ~30 km (Funck et al., in press) and it is separated from the shallow Grand Banks by thinned continental crust under the Flemish Pass Basin and Flemish Cap Graben (Enachescu, 1987). Galicia Bank is extended continental crust that has a maximum thickness of ~20 km in its central part and thins to zero thickness at its western margin; it is separated from Iberia by the Galicia Interior Basin, which contains rifted continental crust that is thinned to ~10 km (González et al., 1999; Perez-Gussinye et al., 2003). Anomaly M0 appears to occur just seaward of the edges of continental crust in these conjugate segments (Srivastava et al., 2000), but there are no older M-series magnetic anomalies present. At the seaward margin of Galicia Bank (ODP Site 637), the westward transition from continental to ocean crust is marked by a prominent ridge composed of serpentinized peridotite (Boillot et al., 1987, 1995).

On the Iberia margin the southern limit of this rift segment lies roughly at the southern edge of Galicia Bank. In this location, the shallow crust of the bank is expressed in a series of rift-parallel ridges that plunge to the south and lose a large portion of their amplitude beneath the southern Iberia Abyssal Plain. It is along this transition that the Leg 149/173 transect was drilled. The seaward edge of known continental crust passes southeastward through this transect near ODP Site 1069, then probably to the south toward Estremadura Spur (Fig. F3). On the Newfoundland margin, continental crust reaches seaward at least to the Flemish Hinge near Flemish Cap and to a hinge line at the eastern edge of Salar-Bonnition Basin to the south. Seaward of the hinge, no along-strike change has been identified in basement structure that would correlate with the structural change at the southern margin of the conjugate Galicia Bank.

The central segment on both margins has an abrupt transition from shallow continental shelf to deep basin, although there are rift basins beneath the continental shelves that are completely filled with sediments (Fig. F3). On the Newfoundland margin these are the Jeanne d'Arc, Carson, and Salar-Bonnition basins, and the Lusitania Basin is present on the Iberia margin. These basins are in extended continental crust that reaches seaward an uncertain distance beneath the continental slope and rise, and they contain evaporites of Triassic age (Jansa et al., 1980; Austin et al., 1989). Farther seaward, the basement is deep and is considered to be "transitional" crust out to a point where magnetic anomalies M3 to M0 are identified (Fig. F4).

The origin of the transitional crust has been a matter of intense debate. Structural trends in the basement are oriented north-northeast to northeast, subparallel to the M0 rift axis. Srivastava et al. (2000) suggested that magnetic anomalies as old as M17 (early Berriasian; ~140 Ma) are present in this segment, but it is unclear whether the low-amplitude anomalies represent polarity reversals or are related to the basement relief. In the transition zone at the northern margin of the segment, Leg 149 and 173 drilling recovered serpentinized peridotites from basement (Fig. F4). In seismic data to the south (line IAM9; Fig. F4), a thin (1.0–2.5 km) unreflective basement layer is observed overlying a more reflective layer (Pickup et al., 1996). This upper layer also has been interpreted to be serpentinized upper mantle peridotite, grading downward into less altered and unaltered peridotite. Seismic refraction experiments there seem to agree, defining a low-velocity "crust" that is only 2–4 km thick and that overlies a layer with velocities of ~7.1–7.7 km/s that is thought to be partially serpentinized peridotite (Whitmarsh et al., 1990; Discovery 215 Working Group, 1998; Dean et al., 2000). In the conjugate central Newfoundland Basin, similar "crustal" thicknesses and velocity structure have been detected in the transition zone (Srivastava et al., 2000). The transition zone width in the central segment is ~150 km on both margins.

The southern segment is like the central segment in that it has deep and thin crust in the transition zone and the zone is ~130–150 km wide. On the Newfoundland side, refraction data of Reid (1994) indicate the presence of very thin crust (~2 km) over apparently serpentinized mantle (7.2–7.5 km/s) that extends at least ~50 km east of the seaward hinge of the Salar-Bonnition Basin. Very similar results have been reported for the conjugate Iberia transitional crust beneath Tagus Abyssal Plain (Pinheiro et al., 1992). Srivastava et al. (2000) suggested that the transitional crust in the southern segment is oceanic and that it is Tithonian (Anomaly M20; ~145 Ma) in its oldest part. The landward portions of the southern segment differ from the central and northern segments in that there are no major proximal rift basins in the continental crust, excepting the southern Salar-Bonnition Basin on the Newfoundland side. On the Newfoundland margin, the southeasternmost Grand Banks is intact continental crust that has been in a subaerial or shallow-shelf environment throughout the Mesozoic and Cenozoic (Jansa and Wade, 1975).

Just as the three rift segments differ from one another, the conjugate sides of each segment also show dissimilarities. In the northern segment, the major distinction is in the amount of crustal extension (i.e., the thick, intact crust of Flemish Cap vs. the extended and structured crust of Galicia Bank). In the central and southern segments, the major differences are in crustal depth and crustal roughness. The Newfoundland transitional basement averages a kilometer or more shallower than the Iberia basement (Fig. F4), even when corrected for sediment loading. In addition, Newfoundland basement is relatively smooth compared to that off Iberia, where >1 km of basement relief is common.

Structural trends in the transitional basement (Fig. F3) tend to show some convergence toward the north. This, together with the northward-narrowing zone of ocean crust in the M0 reconstruction, suggests that the rift may have opened from south to north, which is consistent with a stage pole of opening a short distance north of the rift (Whitmarsh et al., 1990; Srivastava et al., 2000). Considering the segment-to-segment differences in extent of rifting in the shallower continental crust, the southern part of the rift may have switched from continental rifting to seafloor spreading relatively early and abruptly, while the northern part experienced prolonged continental extension and a delayed change to normal seafloor spreading.

Insights from Newfoundland Basin Geophysical Survey

Additional constraints on basement structure and sedimentary stratigraphy of the Newfoundland transition zone were obtained in 2000 during the Study of Continental Rifting and Extension on the Eastern Canadian Shelf (SCREECH) program (Ewing Cruise 00-07). In this program, multichannel seismic (MCS) and Ocean Bottom Hydrophone/Seismometer surveys were made in three major transects across the Newfoundland margin (Figs. F3B, F5, F6). Each transect extended from full-thickness continental crust on the landward end seaward to known oceanic crust beyond magnetic Anomaly M0. Transect 2 was located so that it is conjugate to the Leg 149/173 drilling on the Iberia margin (Fig. F3B), and it is along this transect that Leg 210 drilling was conducted (Fig. F7). To provide regional perspective, the principal results for all three transects are summarized below.

Transect 1

Transect 1 across Flemish Cap is in a position conjugate to Leg 103 drilling conducted on the seaward margin of Galicia Bank (Fig. F3B). It shows that continental crust thins rapidly from ~30 km beneath Flemish Cap to ~2 km beneath the lower continental slope at Flemish Hinge (Funck et al., in press; Hopper et al., in press). Farther seaward, probable ocean crust appears first with oceanic Layer 2/3 velocity structure, and it is 3–4 km thick. It then changes eastward to Layer 2 velocity structure and is only 1 km thick. This crust reaches to slightly beyond Anomaly M0, where more normal thickness ocean crust is present. Both the thin continental and thin ocean crust overlie a layer of probably serpentinized mantle (V_P = 7.6–7.9 km/s) that is 3–5 km thick.

Transect 2

Transect 2, as noted above, is conjugate to Leg 149 and 173 drilling on the Iberia margin and is the focus of drilling during Leg 210 (Fig. F3B). On this transect continental crust thins quickly seaward beneath the continental slope from 30 km to ~7–8 km over a distance of 60 km, and then over the next 50 km it thins more slowly to ~5 km at the Flemish Hinge (Fig. F3). Beyond this, transition zone crust out to Anomaly ~M3 is only 3–5 km thick. Like part of transect 1, this crust has velocities characteristic of oceanic Layers 2/3, but there appears to be no significant underlying zone of possibly serpentinized mantle (Nunes, 2002). This contrasts with transect 1 and with the velocity structure on the conjugate Iberia margin, where a thick zone of serpentinized mantle appears to be present at and south of the Leg 149/173 drilling transect. Also unlike the Iberia conjugate, transect 2 seismic reflection data show no unreflective upper basement layer that might be highly serpentinized peridotite.

Transect 3

Transect 3 exhibits still another set of basement structures. Although continental crust thins rapidly from 35 to <10 km under the continental slope near the seaward edge of Salar-Bonnition Basin (Fig. F3B), thin (<5 km) continental crust above serpentinized mantle appears to reach seaward 50 km into the transition zone. The remainder of the transition zone to the east has extremely thin (~2 km thick) "crust" that may be either a serpentinized layer of exhumed mantle or thin ocean crust (Lau et al., 2003).

A common feature of all three transects is that there is thin crust and apparently very limited magmatism in the transition zone. The transects differ, however, in how magmatism was expressed, in distribution and character of tectonic extension, and in development of serpentinized "lower crust." Thus, it appears that the balance between tectonic extension and limited magmatism was heterogeneous both along and across the rift.

Sedimentary Seismic Sequences

Basal Sequence (Seismic Sequence A)

In the transition zone off Newfoundland there is a very flat, high-amplitude reflection (U) that closely overlies or intersects basement (Figs. F4, F8, F9). This reflection reaches some 600 km south to north in the basin and up to ~150 km across the transition zone. Where traced landward it merges with the Lower to mid-Cretaceous Avalon unconformity on the Grand Banks (Tucholke et al., 1989). At its seaward edge it normally pinches out on crust of about Anomaly M3 age (Hauterivian–Barremian) (Fig. F9). These features suggest that the horizon has an Early Cretaceous age.

The sedimentary sequence between basement and U is defined here as seismic Sequence A. It is seismically laminated, exhibits strong and laterally coherent reflections, and is up to ~0.5 s (two-way traveltime) thick along the proximal part of the margin (Figs. F8, F9, F10). At many locations, basement below U is not identified as a distinct reflection but as a downward disappearance of reflections (Fig. F8A, F8B). This indicates that the impedance of Sequence A is high and is probably close to that of the underlying basement. The sequence is often faulted near its seaward margin, but it is seldom faulted at other locations in the transition zone (Tucholke et al., 1989). The apparent age of U and underlying sediments is similar to that of the Blake-Bahama Formation (Hauterivian–Barremian limestones, capped by Horizon ) in the western North Atlantic Basin (Tucholke and Mountain, 1979; Jansa et al., 1979). The reflection character is also similar, although the U horizon is a much stronger reflection than Horizon . These features suggest that Sequence A may be equivalent to the Blake-Bahama Formation in the western North Atlantic Basin.

Locally, U appears to truncate the underlying basement (Tucholke et al., 1989). The possible truncations, areal extent, and flatness of the horizon led Tucholke et al. (1989) to suggest that it could be an unconformity eroded at or above sea level on extended continental crust. This interpretation has been evaluated with thermal-mechanical modeling (B. Tucholke and N. Driscoll, unpub. data). The results indicate that if erosion occurred at sea level, it would have to be on continental crust at least ~18 km thick, given reasonable upper mantle temperatures of 1300°–1400°C. Thus, if the reflection originated in this way, it must be a synrift unconformity and not a "breakup unconformity."

U and Sequence A can also be compared to a similar deep reflection sequence across probable continental crust of southern Galicia Bank on the Iberia margin (see Figs. F28, F29]). The sequence there is capped by the "orange" reflection and it has been drilled at Site 398 (Shipboard Scientific Party, 1979). It is represented by lithologic Unit 4C, of Aptian age, which is composed of bioturbated mudstones, turbiditic mudstones, thin laminated and cross-bedded fine-grained sandstones and siltstones, and debris flows or mud flows. The unit was interpreted as a fining-upward submarine fan sequence that began to be deposited in Barremian time, with fan deposition gradually waning during the late Aptian. Seismic profiles and the mapped distribution of this sequence (seismic Unit 4 of Réhault and Mauffret [1979]) show that it fills in basement depressions (Fig. F30). A deeper lithologic, just unit above basement (lithologic Unit 5), consists of nannofossil limestones interbedded with laminated mudstones. Velocity in lithologic Units 4 and 5 at Site 398 probably increases with depth, but a mean assumed velocity of 3.59 km/s results in good fit between seismic reflections and a synthetic seismogram based on borehole physical property data (Bouquigny and Willm, 1979). Thus, velocities and impedance in the deep part of the section are high and may be close to those of the underlying basement, much as interpreted on the Newfoundland margin. However, the regional reflectivity of the orange reflection and the upper part of the sequence appear to be lower than the reflectivity of Sequence A on the Newfoundland margin.

Shallower Sedimentary Sequences

Reflection profiles in the Newfoundland Basin show five principal seismic sequences in the section above U. These are differentiated from one another by broad changes in reflection character, which in turn suggest general changes in depositional conditions and/or deformation. The sequences are summarized below, from base to top.

1. Seismic Sequence B. This sequence is conformable to the underlying U reflection and extends upward to the base of a strongly laminated zone near middepth in the sedimentary column (Fig. F8). In the landward part of the transition zone it has a thickness of ~0.5 s two-way traveltime. Reflections in Sequence B have low amplitude compared to reflections in the underlying and overlying sequences, but they are still relatively well defined and are mostly laterally continuous over distances of tens to hundreds of kilometers. Reflections tend to be more coherent and readily traced in the lower part of the sequence than in the upper part, where they are sometimes disrupted or even chaotic. At some locatfions in the upper part of Sequence B, reflections show seaward downlap and landward onlap of reflections that suggest local control of depositional patterns (e.g., in a fan-channel system). The top of the sequence is marked by an apparent unconformity that truncates progressively deeper beds in a landward direction. The predicted age of this sequence is mid-Cretaceous, which would include black shales equivalent to the Hatteras Formation farther south in the main North Atlantic Basin.
2. Seismic Sequence C. Sequence C exhibits a series of very strong, flat, coherent reflections that are easily traced laterally for distances of 100 km or more. The interval represents ~0.3 s of reflection time and occurs midway in the sedimentary section. Reflection character of this sequence indicates that it consists of interbedded high- and low-velocity layers (e.g., turbidites). In its landward portions, the upper part of Sequence C remains strongly layered in its lower section but its upper section expands and contains chaotic reflections; these reflections appear to be caused by debris flows and other downslope mass movements.
   The reflection that marks the top of Sequence C, as discussed below, appears to be equivalent to Horizon A^u in the western North Atlantic, suggesting that Sequence C is Eocene at its top; it probably extends into the Paleocene or possibly the Upper Cretaceous at its base. The following formations would be included in this sequence, from base to top: black shales of the Hatteras Formation, reddish and multicolored pelagic shales of the Plantagenet Formation (limestones of the Crescent Peaks Member), and siliceous shales and cherts of the Bermuda Rise Formation (Jansa et al., 1979).
3. Seismic Sequence D. This sequence is characterized by reflections with distinctive pinch-and-swell morphology that indicates current-controlled deposition and formation of sediment waves, much like Oligocene–Miocene seismic sequences along the eastern margin of North America to the south (Mountain and Tucholke, 1985). The sequence is thickest close to the margin (~0.4 s reflection time) and sediment waves are best developed there. Seaward, it thins to ~0.1 s. A number of strong reflections are well developed and continuous through the sequence, but weaker intervening reflections are often broken up, particularly in the lower part of the sequence and in its thinner section away from the margin. The semichaotic character of these reflections suggests that debris flows and mass wasting deposits may form part of the sequence.
   The base of this sequence is interpreted to be equivalent to Horizon A^u in the western North Atlantic south of Newfoundland Basin. There, the horizon is a widespread unconformity that was eroded when strong abyssal circulation (Deep Western Boundary Current [DWBC]) developed in the basin (Tucholke and Mountain, 1979). On the Newfoundland margin the reflection shows truncation of the underlying irregular beds of upper Sequence C beneath the inner continental rise, but farther seaward it is mostly conformable to deeper bedding. The source of bottom water for the DWBC is thought to be in the sub-Arctic/Arctic seas, so the Newfoundland Basin is the "gateway" region through which this current flowed southward and it may contain an important record of how abyssal circulation developed in the North Atlantic Ocean. The DWBC is interpreted to have developed in the latest Eocene to early Oligocene (Miller and Tucholke, 1983; Davies et al., 2001). The predicted age of Sequence D is lower Oligocene at the base, extending up into the Miocene at the top.
4. Seismic Sequence E. This sequence is well developed all along the Newfoundland margin. It is consistently thicker close to the margin (~0.8 s two-way traveltime) and thins seaward to as little as ~0.2 s beneath the outermost continental rise and abyssal plain. Sequence E has a very distinctive seismic signature. It is marked by undulating and contorted reflections that usually can be traced for only limited distances. Some reflections have the form of poorly developed sediment waves. In its upper part the sequence contains channels that are filled with chaotic debris, and other portions also show chaotic signature that probably represents rapid deposition of mass-wasting deposits. The sequence is also permeated by small-throw normal faults, almost none of which extend into the underlying or overlying sequences. These features are commonly developed in abyssal fans; they suggest that the sediments were deposited rapidly while trapping pore fluids and that they later failed as the sediments dewatered. Similar seismic sequences appear along the U.S. East Coast margin (Mountain and Tucholke, 1985) and on other margins across the globe in the middle Miocene to Pliocene–Pleistocene. They appear to document a global period of margin progradation (Bartek et al., 1991).
5. Seismic Sequence F. This is the topmost seismic sequence in the Newfoundland Basin and it consists of reflective, flat-lying turbidites that form an abyssal plain seaward of the lower continental rise. The turbidites interfinger with and lap landward onto underlying fan Sequence E. Thus the base of the sequence is time-transgressive, becoming younger landward. Within the turbidites, it is common to observe chaotic beds up to ~0.1 s thick and extending for many tens of kilometers. These appear to be debris flows that originated on the continental rise fan. The predicted age of the turbidities is late Pliocene to Quaternary.

Drilling objectives for Leg 210 were twofold. The primary objective was to sample the deep sedimentary structure and basement in order to investigate early rift development. A related objective was to study the shallower stratigraphy and to elucidate the postrift sedimentation processes and paleoceanographic history of this gateway between the North Atlantic and the sub-Arctic sea. The background for each of these objectives is summarized below.

Origin of Transitional Crust

Drilling results from the Iberia margin and geophysical data from both sides of the Newfoundland–Iberia rift show clearly that the rift was characterized by very limited volcanism, that there are marked asymmetries between margin conjugates, and that there is significant structural variability along strike between rift segments. These features are particularly manifested in the transition zones between known continental and known oceanic crust on the opposing margins. We have posed three hypotheses to explain the crustal structure and basal stratigraphy in the Newfoundland and Iberia transition zones and the cross-rift asymmetries between these zones (e.g., Tucholke et al., 1999) (Fig. F11). Leg 210 provided the first direct test of these hypotheses by drilling in the transition zone along transect 2 on the Newfoundland margin (Figs. F3, F4, F7, F8B).

Hypothesis 1: Newfoundland Transition Zone Is Highly Thinned Continental Crust

Newfoundland transitional crust is shallower and has less roughness than Iberia crust, and it could be the upper plate in an asymmetric detachment system (Fig. F11B). According to this hypothesis, the lower Iberia plate east of Anomaly ~M3 would be exhumed lower continental crust and mantle (e.g., Whitmarsh et al., 2001). Strong thinning of the Newfoundland crust without significant brittle extension might be possible if the lower crust thinned by ductile flow (e.g., Driscoll and Karner, 1998) (Fig. F11). This should be reflected in rapid synrift subsidence of the Newfoundland basement. If U corresponded to a subaerially eroded unconformity, the rapid subsidence would be recorded in the sedimentary section above the reflection.

There are two other possible explanations of U. One is that it corresponds to the top of basalt flows emplaced either subaerially (i.e., it is synrift on continental crust) or on the seafloor (Enachescu, 1988). If melt was extracted from the rising lower plate and emplaced in the Newfoundland upper plate (see, e.g., Fig. F11C), the exhumed Iberia mantle could be virtually melt-free, as has been suggested by existing Iberia drilling (Whitmarsh and Sawyer, 1996). Although it seems unlikely that smooth basalt flows could be as widespread as is indicated by the distribution of U, there are documented instances where such flows are known to be extensive (e.g., Larson and Schlanger, 1981; Driscoll and Diebold, 1999).

Another explanation is that U corresponds to the top of high-velocity sedimentary deposits that were shed from the Grand Banks, probably in Early Cretaceous time. As already noted, this sequence could be similar to the Aptian fan deposits recorded on the conjugate Iberia margin, although the stronger seismic signature on the Newfoundland margin suggests much higher-velocity beds that possibly are very coarse grained or carbonate rich.

Hypothesis 2: Transition Zones Reflect Extreme Extension in an Amagmatic Rift

According to this hypothesis, continental extension proceeded under nearly amagmatic conditions to a state where only mantle was exposed, and at some point an asymmetric shear developed within the exposed mantle (Fig. F11C). This hypothesis differs from the one above in that Newfoundland transitional crust would be exhumed, probably serpentinized mantle. U could not correspond to a subaerial unconformity because it would be impossible to uplift extending mantle to sea level. The U–basement interval could correlate with basalt flows, with melt generated from the rising lower plate in an asymmetric extensional system, or it could be high-velocity sedimentary beds, as noted above.

Hypothesis 3: Newfoundland Transitional Crust Was Formed by Ultra-Slow Seafloor Spreading

Slow seafloor spreading (Fig. F11D) is known to expose lower crust and mantle (e.g., in the Labrador Sea [Chian and Louden, 1995; Osler and Louden, 1995]), and it could explain the transitional crust in the Newfoundland Basin. However, symmetrical ultra-slow seafloor spreading in the rift seems unlikely because it does not explain the extensive mantle exposures off Iberia, nor does it explain the asymmetries in crustal structure and depth between the conjugate transition zones. It is possible that extension first exposed mantle in the rift, that ultra-slow seafloor spreading then occurred on the Newfoundland side of the rift, and that this ocean crust was then isolated on the Newfoundland margin by an eastward jump of the spreading axis (Fig. F11D). This hypothesis precludes U from corresponding to a subaerial unconformity because it would overlie thin ocean crust. As in the above hypotheses, U might correlate with either basalt flows or the top of high-velocity sedimentary beds.

Sedimentary History and Paleoceanography

Rifting between Labrador and Greenland, and between Greenland and Eurasia (Rockall Trough), began in the Early Cretaceous, leading to Late Cretaceous seafloor spreading in the Labrador Sea and Paleocene spreading east of Greenland (e.g., Srivastava and Roest, 1999; Eldholm et al., 1990). The Newfoundland–Iberia rift was a gateway between the main North Atlantic and these developing ocean basins, so it is in a key position to investigate sedimentary history and paleoceanographic links through the northward-expanding ocean basins.

Two features of the predicted sedimentary record above U were of particular interest during Leg 210. The main basin of the adjacent North Atlantic was accumulating black shales of the Hatteras Formation in Barremian–Cenomanian time, followed by deposition of the Plantagenet Formation under oxygenated seafloor conditions in the Late Cretaceous (Jansa et al., 1979). The Newfoundland Basin Cretaceous sedimentary record provided an opportunity to examine whether this record of reduced and then increased ventilation of the deep basin extended northward into the developing ocean basins, as well as information on the timing of that record. It also provided important information on paleobiogeography in a zone where Tethyan and boreal flora and fauna were expected to have mixed.

The second feature of interest was the upper Eocene–lower Oligocene sedimentary record, which could contain important information on the first development of strong abyssal circulation in the North Atlantic. As already noted, the source of the bottom water for this developing circulation has been interpreted to be the sub-Arctic seas and the timing has been estimated as latest Eocene to early Oligocene (Miller and Tucholke, 1983; Davies et al., 2001). However, these predictions are based largely on the occurrences of hiatuses in boreholes farther south in the North Atlantic, and the lack of sedimentary records in the critical intervals there makes it difficult to verify the predictions. New data from the gateway region could help to constrain the source and timing of the circulation event.

The most direct and productive method to test our hypotheses about basement structure and the deep reflection sequence in the Newfoundland transition zone was to drill a deep hole (up to ~2200 m) into that zone. Such a hole would also recover an expanded Cretaceous and Tertiary sedimentary record with which to investigate the paleoceanographic history in this gateway between the sub-Arctic seas and the main North Atlantic Basin. The prime site that was selected to accomplish these objectives is proposed Site NNB01A (Site 1276), located in the westernmost edge of the abyssal plain at the foot of the Newfoundland continental rise. In the event that our basement and deep sedimentary objectives could not be achieved at this site and sufficient drilling time remained during Leg 210, we also developed a series of alternate sites that extend seaward to known oceanic crust that could be drilled in order to partially satisfy our objectives.

Depths of major seismic horizons, including basement, were predicted from semblance velocity analysis of multichannel seismic reflection data obtained along the drilling transect during the SCREECH site survey program in 2000. This analysis indicated that the major drilling objectives at proposed Site NNB01A (i.e., U and basement) were at depths of ~1860 and 2080 meters below seafloor (mbsf), respectively (Table T1). Sampling to these great depths in one drilling leg was considered to be an ambitious goal, but the objectives were deemed to be achievable by following a plan of drilling, casing, and logging the hole as shown in Figure F12. To improve our chances of meeting the objectives within the time available for drilling, it was agreed that the upper 800 m of the hole would be drilled without coring. Thus, we expected that the first sediments recovered from the hole would be of Oligocene age, above Horizon A^u.

Actual hole conditions and time constraints necessitated modification of this plan in real time during Leg 210. Nonetheless, we were able to follow part of the drilling and casing plan at proposed Site NNB01A (Site 1276) (Fig. F32), and we achieved a significant part of our objectives. In addition, we drilled a short hole ~95 m into basement at proposed alternate Site NNB04A (Site 1277). Results of both sites are summarized below.