As discussed earlier, rifting of strong, possibly subcontinental lithospheric mantle may have persisted until latest Aptian to earliest Albian time (ca. 112–110 Ma). We here consider this to be the end of the second period of transitional extension (TE2). In the following postrift period, extension localized at a well defined spreading axis and magmatic seafloor spreading ensued.
The most significant postrift magmatism, other than seafloor spreading, encountered in the Newfoundland Basin is in the form of two alkaline diabase sills drilled at Site 1276, which is located in TE1 probably near the seaward edge of thinned continental crust (Van Avendonk et al., 2006) (Figs. F1, F4). The main body of the lower sill, in which the hole bottomed, is >18 m thick (1719.2 to >1736.9 mbsf). It is accompanied by at least four other thin (3–31 cm) sills that were cored 7–14 m above the main sill and may be apophyses of the sill (Tucholke, Sibuet, Klaus, et al., 2004). The upper sill is ~10 m thick (1612.7–1623 mbsf). Both sills were intruded into uppermost Aptian to lowermost Albian sediments at an estimated ~90–160 m and 200–270 m above basement, respectively, based on interpretation of the seismic reflection record (Fig. F3) and seismic velocities at the drill site (see Shillington et al., this volume). Whole-rock 40Ar/39Ar radiometric dates indicate that the upper sill was intruded first, at 104.7 Albian on the Gradstein et al., 2004, timescale), and the lower sill was subsequently intruded at 95.9 nian) (Table T1) (Hart and Blusztajn, 2006). Considering these ages in the context of the plot of sediment age vs. subbottom depth for Site 1276 (Tucholke, Sibuet, Klaus, et al., 2004), the upper sill was intruded beneath a minimum of ~260 m of overburden, and the lower sill was intruded beneath at least ~590 m of overburden (neither value accounts for postintrusion compaction). Karner and Shillington (2005) estimated intrusion depth of 0–556 mbsf based on reconstruction of porosity-depth relations, which is similar to these values.
Geochemical studies of the sills (Hart and Blusztajn, 2006) indicate that they are alkali basalt-hawaiite (differentiated basanite) and that they have experienced significant alteration, as evidenced for example by extremely high Ba contents (up to 4900 ppm) in abundant acicular secondary apatite, particularly in the upper sill. Isotopic and trace element data show that the two sills are not strictly co-genetic but originated from separate magmatic events, consistent with their ~7.5-m.y. age difference. The source of the magma is uncertain, although the sills clearly were not derived from MORB-type asthenospheric upper mantle. Hart and Blusztajn (2006) concluded from the isotope geochemistry that the magma probably was derived from an enriched plume source in the mantle, possibly contaminated by a component of continental material.
The age of the lower sill is not significantly different from that of a trachyte from Scruncheon Seamount (97.7 foundland Seamounts 200 km south of the drill site (Table T1) (Sullivan and Keen, 1977). Trace element geochemistry of the sills is also similar to that of basalts from the Newfoundland Seamounts (Hart and Blusztajn, 2006). These commonalities suggest a shared source for the igneous rocks. Duncan (1984) proposed that at ~100–90 Ma the Newfoundland Seamounts and the southern Newfoundland Basin passed over the Madeira and Canary hotspots, respectively, with the southern Newfoundland Basin also crossing the Azores hotspot some 20 m.y. later. Thus, both available geochemical data and plate-kinematic data are consistent with the idea that plume-related magmatism was responsible for intrusion of the sills.
The extent to which other parts of the Newfoundland Basin were affected by sill injection is unclear, but it appears that this magmatism may have been a basin-wide phenomenon. At Site 1276, the depth of the upper sill is coincident with the U reflection or Aptian event (Fig. F3), and the lower sill also coincides with a strong, deeper reflection (Shillington et al., this volume). Few features in the seismic reflection record are clearly diagnostic of sills. Minor disruptions of the U reflection 2–3 km to either side of the drill site might represent locations where melt was injected into the basal sedimentary column (Fig. F3). Such disruptions are not uncommon elsewhere in the basin. In other cases, short segments of the U reflection show increased amplitude (Fig. F3, left side) and could mark sill locations. At a few other places such short, high-amplitude reflection segments occur near the U reflection but are separated from it (Fig. F9), and these very likely are sills (Tucholke et al., 1989).
On a broader scale, the U reflection itself has unusually high amplitude throughout most of the basin, to the degree that it commonly masks most deeper structure (Tucholke et al., 1989); it also is generally much stronger than the correlative "orange reflection" (Aptian event) on the conjugate Iberia margin. The strength of the U reflection is something of a puzzle because at Site 1276 there is little evidence for unusually high impedance in the sediments at that level (excluding sediments that have been hydrothermally altered by the sill injection) (Tucholke, Sibuet, Klaus, et al., 2004). We would expect to see such sediments represented at Site 1276 if they give rise to the reflection throughout the basin. On the other hand, if sills were consistently injected at this level, they could explain the strength of the reflection. Laterally extensive sills are well known in the geological record. For example, the Carboniferous Great Whin Sill in northern Britain extends over 5000 km2 (Johnson and Dunham, 2001). This sill mainly follows bedding planes, with occasional splitting; it has been suggested that once magma has reached a level of hydrostatic equilibrium at shallow subbottom levels, it can flow laterally with little frictional resistance, much like subaerial flood basalts (Francis, 1982). Considering this mechanism, multiple sills could have been emplaced at a relatively uniform depth in the level-bedded sediments of the Newfoundland Basin. If this was the situation for the older event (the shallower sill at Site 1276), then younger sills might be mostly restricted to the underlying section, as observed at Site 1276. In the future, detailed analysis of lateral changes in amplitude of the U reflection and the underyling section, combined with structural interpretations, may help to resolve the occurrence and distribution of sills.
If there was widespread intrusion of sills, and their emplacement was related to passage of the Newfoundland Basin over one or more mantle plumes as discussed above, it may help to explain why the U reflection is much stronger than the equivalent reflection event on the Iberia margin. According to Duncan (1984), the axis of the Mid-Atlantic Ridge did not pass over the hotspots until ~70 Ma, so the Iberia plate was isolated from their possible magmatic effects until ~40 m.y. after the Aptian event. Even then, the hotspots were located under the southernmost edge of the Iberia plate and the northwestern African plate, so they would have had little effect on the Iberia side of the rift.
The older, upper sill was injected into calcareous, turbiditic grainstones, siltstones, and claystones (see Tucholke, Sibuet, Klaus, et al., 2004, for details). Its upper contact is preserved as a chilled margin juxtaposed against a thin (<1 cm) layer of baked sediment. Prominent megascopic effects of hydrothermal alteration are observed for ~66 cm above the contact, in the form of porphyroblasts composed of pure stoichiometric calcite (T. Pletsch, pers. comm., 2007; note that the porphyroblasts were incorrectly reported in the Leg 210 Initial Reports as consisting of albite, quartz, alkali feldspar, and magnesian chlorite, which actually comprise the surrounding sediment). The porphyroblasts decrease in size and frequency with distance from the contact. A subvertical, crenulated vein filled with calcite and pyrite in this interval is interpreted to have precipitated from fluids circulating through fractures as the sill was injected. The crenulation, together with compaction of sedimentary laminae around the porphyroblasts, provides evidence that the sill was injected into relatively unconsolidated sediments. Later compaction shortened the vein by ~21% (Karner and Shillington, 2005). The lower contact of the sill was not recovered; although the sill shows a chilled margin, the subjacent sediments show no thermal overprint in smear slides and it is likely that a section of the sedimentary record was not recovered in the cores.
Pross et al. (2007) used sporomorph colors to assess thermal alteration above the sill. They determined that thermal effects extend upward ~20 m, but the strongest alteration occurs in an aureole within ~4.2 m of the sill. They also observed reduced thermal alteration in an interval ~5–6 m from the sill, which suggests small-scale convection in the circulating hydrothermal fluids.
Near the bottom of Site 1276, the upper contact of the younger, lower sill was not recovered, and most contacts of the overlying thin sills are missing or poorly preserved because of drilling disturbance. The principal evidence for hydrothermal metamorphism is observed in Sections 210-1276A-98R-2 to 98R-CC between a pair of thin sills at ~1711 mbsf and the top of the main sill at ~1719 mbsf. Here, a ~2-m section of claystone, mudstone, and minor calcareous sandstone was affected by contact metamorphism. The alteration produced a variety of effects in different beds, including (1) hornfels textures with porphyroblasts that are visually similar to those overlying the upper sill but that have not been chemically analyzed; (2) nearly total replacement by albite, sudoite, nonstoichiometric calcite, pyrite, and zeolite; (3) color alteration; and (4) thermal alteration of organic matter (Tucholke, Sibuet, Klaus, et al., 2004). Veins in the sediment contain stoichiometric calcite with pyrite, and extremely light carbon isotope composition of the carbonates suggests a carbon source in the adjacent organic-rich sediments (T. Pletsch, pers. comm., 2007).
Thus overall, strong hydrothermal metamorphic effects associated with the sills appear to have been very restricted. They occurred only within a few meters of the two major sills and within a few centimeters to decimeters of the minor, thin sills. However, lesser effects may have extended 10–20 m from the sills.
Other postrift magmatism, unrelated to seafloor spreading, is present within the Newfoundland-Iberia rift, but it is widely scattered and very limited in volume (Table T1). On the Newfoundland margin, a 96-Ma porphyritic monzodiorite dike was penetrated in the Emerillon C-56 well (Jansa and Pe-Piper, 1986), and alkaline dikes and volcanic rocks dating to ~100–70 Ma are documented on the southern Iberia margin (Pinheiro et al., 1996). In the deep basin, minor igneous rocks associated with basement peridotites have a similar wide range of apparent ages at ODP Sites 1070 and 1277 and along the northwest Galicia margin (Table T1). However, ages on all these samples were determined on plagioclases, and it is possible that the ages do not represent original magmatism but were reset by late-stage hydrothermal alteration.
Jansa and Pe-Piper (1988) attributed the 96-Ma dike in the Emerillon well to intraplate fracture or fault reactivation related to opening of the Labrador Sea, and they suggested that most such igneous activity occurred at times of significant plate-motion changes. However, if the widely scattered (>30 m.y.) igneous ages in the deep Newfoundland-Iberia rift (Table T1) represent true ages of magmatic events, then minor magmatism would appear to be relatively common and not necessarily linked to major plate-tectonic events. Furthermore, these ages could indicate that intraplate stress locally reactivates off-axis faults more often than heretofore recognized, particularly in areas of exhumed peridotite where weak, serpentinized faults can fail easily. Where the faulting was combined with a thermally thin plate or underlying plume activity, widely distributed minor magmatism might result. It is highly unlikely that this would ever be recognized on a normal mid-ocean ridge where igneous rocks are abundant, whereas these rocks stand out uniquely in peridotite basement and thus tend to be extensively studied and documented.
A separate group of postrift igneous rocks has been documented on Gorringe Bank (Table T1). This feature lies along the Iberia/Africa plate boundary, which is the former Newfoundland-Gibraltar Fracture Zone. According to plate-kinematic arguments, this boundary was locked from Late Cretaceous to middle Eocene time as Iberia moved with Africa, but subsequently it has been active and notably compressive in its eastern (Gorringe) part (Srivastava et al., 1990b). On the other hand, tilted sedimentary strata around Gorringe Bank and related highs suggest compression across the boundary since Late Cretaceous time (Féraud et al., 1982). The Late Cretaceous magmatic episodes indicated by isotopic ages of igneous rocks from this feature may reflect intermittent effects of plate-boundary tectonics. Féraud et al. (1986) argued that a primary alkaline magmatic event at ~65–67 Ma may have reset older rocks to ~111 Ma and ~75–84 Ma ages. However, the 77-Ma U-Pb zircon age on ferrogabbro reported by Schärer et al. (2000) (Table T1) appears to represent a primary igneous event. The ~65- to 67-Ma alkaline event suggested by Féraud et al. (1986) could reflect the influence of the Madeira and Canary hotspots, over which the region probably passed beginning at ~70 Ma (Duncan, 1984).