Type 1 veins

Type 1 veins are dark green and consist of high-temperature amphibole (Fig. F173A–F173C). These record the early brittle deformation history beginning in amphibolite facies at temperatures >650°C.

Type 2 veins

Type 2 veins are coeval with brecciation, cataclasis, and mineralization at greenschist grade. They are most abundant in brecciated zones, consist of actinolite and clays (saponite), and exhibit a distinctive yellow-brown color (e.g., Fig. F173D). The color may be derived from the presence of clays forming much later in permeable zones, at temperatures <150°C.

Type 3 veins

The third and most pervasive type of veins are green to light green in hand sample, have fibrous textures, and consist of actinolite-tremolite and, possibly, minor talc and smectite (Fig. F173E, F173F). Tremolite and talc-tremolite (Fig. F173G) veins likely formed during the same alteration event as Type 3 veins and record a greenschist alteration event that affects nearly the entire section of recovered core.

The remaining vein types include carbonate (Fig. F173H), quartz, serpentine, and clays. These make up the latest and lowest temperature vein sets and crosscut all other types.

Expedition 305 vein description

Examples of the types of veins observed during Expedition 305 are shown in a series of core photographs and thin section photomicrographs (Figs. F174, F175, F176, F177, F178, F179, F180, F181). During Expedition 305, veins were classified into four types based on color: dark green (Fig. F174), pale green (Fig. F177), white (Fig. F178), and gray (Fig. F180). The first three types are present in mafic rocks, and the latter is present in ultramafic rock. This classification is not the same as that described in “Metamorphic petrology.”

Dark green veins (probably correlative with Expedition 304 Type 1 veins)

The dark green veins occur individually, in groups (Fig. F174A–F174D), and in parallel sets over a few tens of centimeters (Figs. F174E, F178C). They are commonly crosscut by the other types of veins. In thin section, the dark green veins commonly consist of amphibole and plagioclase (Fig. F179A, F179B), indicating that they formed in the amphibolite-facies regime. In the deeper sections of the core, these green veins are associated with more pervasive cataclasitic deformation of the surrounding rock (Fig. F175). Later brittle shear deformation tends to occur along the dark green veins, in which alteration minerals and plagioclase were fractured (Fig. F176). Some green veins have fibrous mineral infilling, indicating that they are fault veins.

Pale green veins (probably correlative with Expedition 304 Type 2 and Type 3 veins)

The pale green veins may be further divided into two types based on the presence of alteration halos. The orientation of the pale green veins without alteration halos varies from subhorizontal to steeply dipping, with both alteration zones having subparallel sets and randomly distributed sets of veins (Fig. F177A–F177C, F177E). The pale green veins with alteration halos sharply cut the gabbro, and alteration halos are typically a few centimeters thick (Fig. F177F). Apart from the alteration halos, the pale green veins are characterized by fiber textures of the mineral infilling, primarily talc and tremolite (Fig. F177B). Despite the variable dip of these veins, the fiber orientations are primarily subhorizontal (Fig. F177B) (see “Fault rocks,” below). In thin section, some pale green veins are observed to consist of epidote, chlorite, and actinolite (Fig. F179C, F179D). These mineral assemblages indicate that the veins formed at greenschist-grade conditions.

White veins

The white veins are present dominantly as sets of open cracks with carbonate and local sulfide infillings (Fig. F177A). The white veins crosscut the pale green (Fig. F178B) and dark green veins (Fig. F178C). Locally, large, steeply dipping white veins are present in Section 305-U1309D-136R-3 (Fig. F178D). The white veins are also associated with cataclasis (Fig. F178E, F178F). In thin section, white veins commonly crosscut pale green veins (Fig. F179C, F179D). Although most white veins are filled with carbonate, some are crack-seal veins with fibrous chlorite and, locally, prehnite (Fig. F179D). These mineral assemblages represent formation conditions that are at a lower temperature than the pale green veins described above, suggesting that they formed at the latest stage of the evolution history.

Gray veins and related fibrous veins in olivine-rich troctolites

Gray veins are present locally in olivine-rich troctolites. They do not show evidence for distinct deformation features (Fig. F180). In thin section, the veins consist of serpentine and chlorite, with chlorite present locally (Fig. F180F–F180J). These veins show little evidence for deformation, suggesting that they formed by replacement reactions along microcracks. The gray veins either crosscut or are crosscut by serpentine foliation.

Some fibrous gray veins have alteration halos, for example at ~1126 mbsf (Fig. F181). The fibrous minerals consist of tremolite, and the fibers are oriented obliquely from the wall toward the vein center (e.g., broken white line in Fig. F181), suggesting synkinematic shear during vein opening. For the example shown in Figure F181D, a normal sense of shear in the core reference frame was determined from the fibrous growth. In addition, some fibrous minerals exhibit weak plastic deformation.

Downhole distribution and orientation of veins

As illustrated in Figure F182, vein intensity generally decreases with depth in Hole U1309D, with superimposed “peaks” observed throughout. As described below, these peaks are present at depths where other data sets also indicate structural boundaries. In the upper 400 mbsf of Hole U1309D, intervals of relatively high vein intensity also coincide with the presence of ultramafic rocks and olivine-rich troctolites in Cores 304-U1309D-10R, 32R, and 56R. Based on the degree of alteration, these rock types apparently were more reactive than gabbroic rocks under the same conditions. A somewhat abrupt decrease in the overall vein intensity is observed below the sequence of faults in the interval between ~750 and ~785 mbsf (Fig. F182) and corresponds to the transition from structural Unit 2 to structural Unit 3.

Downhole plots comparing the intensity of veining in Hole U1309B and the upper 130 mbsf in Hole U1309D are shown in Figure F183. For the correlative depths, the vein intensity is greatest (1–3.5) in the upper gabbros of both holes (gabbro Zone 1 in Hole U1309D as defined in “Igneous petrology”) and then decreases (intensity ≤ 1) in gabbroic and troctolitic units (~64–94 mbsf in Hole U1309B and ~62–83 mbsf in Hole U1309D). In Hole U1309B, inferred faults separate regions with different vein intensity. The variation in vein intensity in the upper 130 mbsf of Hole U1309D is closely associated with the presence of Type 2 veins, which are present throughout the gabbro and troctolite and are most common in zones that experienced intense brittle deformation (Fig. F184). Vein intensity is also closely related to rock type, especially diabase. In both holes, diabase units show the least intense brittle deformation and vein intensity. Vein type in the diabase from both holes is restricted to Type 3 veins, blue-green actinolite + smectite veins (similar to Type 3), and a few Type 2 veins (Fig. F184). The presence of Type 2 veins in diabase is restricted to regions where brecciation was observed at the diabase margin in Cores 304-U1309B-5R and 6R and 304-U1309D-1R. These observations indicate that diabase units postdate Type 1 veins, are synchronous with Type 2 veins, and predate Type 3 veins.

For the core logged during Expedition 304, the compositional variability of veins with depth is shown in Figures F172 and F184. Type 1 veins are most abundant between 20 and 80 mbsf and absent between ~240 and 400 mbsf (Fig. F172A). By contrast, late-magmatic leucocratic dikes are most abundant between 195 and 400 mbsf (see also Fig. F141). The Type 2 and 3 veins are distributed throughout the upper 400 m of the core, although there are no Type 2 veins observed between ~80 and 130 mbsf (Fig. F172B, F172C). Talc-tremolite veins on fractures are generally restricted to depths of 140–290 mbsf; however, this vein type is similar to the “late tremolite (± talc) associated with olivine-rich rocks” shown in Figure F172C. Type 2 veins associated with brecciation are present throughout the upper 400 mbsf. The lowest temperature veins (serpentine, talc-tremolite, and late carbonate veins) are plotted on Figure F172D.

The vein sets record retrograde metamorphism, and crosscutting relationships indicate that some veins act as sites for fluid migration over a wide range of temperatures. The intrusion of thin leucocratic dikes (plagioclase + hornblende assemblages) was followed closely by Type 1 vein formation. These late-magmatic leucocratic dikes apparently acted as fluid pathways for much of the alteration history; they are commonly overprinted by Type 1 and Type 3 veins. Based on a limited number of observations, Type 2 veins predate Type 3 and talc-tremolite veins. Serpentine, carbonate, and quartz veins are younger and cut all other fabrics and vein types. Talc and carbonate veins commonly cut serpentine veins; these, together with prehnite and clay veins, are the youngest.

Histograms of vein dips for the upper 132 mbsf of Hole U1309D show a weak concentration in the range of 40°–75° (Fig. F185). To simplify the presentation of data for the upper 400 mbsf of Hole U1309D, average dip and the standard deviation are plotted on Figure F172E–F172G. Standard deviations about the mean dip orientation are 15°–23°. However, there is a suggestion that the dips of Type 2 and 3 and talc-tremolite veins covary across inferred fault zones at depths of ~255 and ~285 mbsf.

For the core logged during Expedition 305, the identified vein types (Fig. F186) show no clear trend in dip between 400 and 1415 mbsf. Figure F187 shows the dip distribution of fault veins (Fig. F187A, F187B), open cracks, cataclastic veins (Fig. F187C), and dark green (Fig. F187D), pale green (Fig. F187E), and white veins (Fig. F187F) logged during Expedition 305. The average dip of all fault veins is ~55°, and there is a progressive increase of the average dip from white to pale green to green and to faults with fibers, which have a steeper plunge (Fig. F187B). All veins, open cracks, and cataclastic veins have average dips of ~40°–55°, with large standard deviations and no clear trend (Fig. F187C–F187F).

Cataclastic deformation

Cataclastic deformation is divided into five categories:

1. Relatively thick zones of brecciation/cataclasis, including faults and semibrittle shear zones (Fig. F188A–F188D, F188E–F188G);
2. High-temperature diffuse cataclastic bands associated with late-magmatic leucocratic intrusions (Fig. F173D);
3. Low-temperature diffuse cataclastic bands associated with greenschist-grade (Type 2 and Type 3) veins (Fig. F188A);
4. Open fractures along low-temperature veins; and
5. Microfractures and veins associated with serpentinization (i.e., serpentine foliation) (Fig. F188H).

Fault rocks

Description of fault rocks from Expedition 304

In Hole U1309B, three fault zones are recognized. Two are at the boundaries between diabase and gabbro (lithologic Unit 23) and between peridotite and diabase (Units 32–33); the other is in the lower gabbro unit (Unit 48). Similar cataclasites were also observed in Core 304-U1309D-1R. However, there are also numerous examples of rocks that deformed by semibrittle mechanisms at greenschist-grade conditions (Fig. F188D, F188E, F188F). Most of the cataclastic deformation in gabbro units is associated with the intrusion of late-magmatic leucocratic veins with subsequent overprinting by greenschist-grade vein networks. Synmagmatic cataclastic deformation in diabase is associated with greenschist-grade hydrothermal brecciation (Fig. F188B, F188C). Microfracturing associated with alteration of both ultramafic rocks and olivine-rich gabbroic rocks resulted in the formation of a serpentine foliation (Fig. F188H).

High-temperature cataclasites are associated with late-magmatic leucocratic intrusions, which are almost always overprinted by greenschist-grade vein networks (Fig. F188G). Plagioclase is usually present as fine-grained, subhedral crystals (<0.2 mm in size). These breccias display a clast-matrix texture that formed during amphibolite- or greenschist-grade deformation. The clasts in these breccias range in size from 10 µm to 5 mm and comprise plastically deformed or brecciated gabbro and hornblende veins. In addition to these gabbro clasts, clasts of diabase and basalt are found in breccia zones in Hole U1309B and the upper 130 m of Hole U1309D (Fig. F188C). The matrix contains <10 µm diameter plagioclase and actinolite grains that locally form a weak cataclastic foliation. In contrast to intense cataclastic zones, diffuse zones of cataclasis do not show clast-matrix texture but instead display pervasive microcracking and incipient brecciation. Evidence for semibrittle deformation and alteration of plagioclase is commonly observed in high-temperature late-magmatic leucocratic zones of cataclasite (Fig. F188G), whereas in low-temperature cataclastic zones, isotropic growth of tremolite is common (Fig. F188F).

Description of fault rocks from Expedition 305

Four fault zones were identified during Expedition 305 by the presence of coherent cataclasite (Fig. F189). Apart from the brittle deformation associated with the detachment zone in the upper part of the core, the fault zone at ~750 mbsf represents the zone with the most significant slip recovered in Hole U1309D.

The most shallow fault zone recovered during Expedition 305 appears at 695 mbsf in Section 305-U1309D-141R-1. The recovered cataclasite is ~2 cm thick. It is well developed and indicates apparent reverse movement along the fault in the core reference frame. The lower boundary of the fault zone shows a sharp contact against cataclastic gabbro (Fig. F189A). In thin section, plagioclase fragments are present in a matrix of mostly chlorite and actinolite, suggesting that the related alteration occurred under greenschist-facies conditions.

A structurally deeper fault zone is located at 746 mbsf in Section 305-U1309D-152R-1 (Fig. F189B, F189C). Two pieces with well-developed ultracataclasite are present in Section 305-U1309D-152R-1. Foliation in the first cataclasite dips 55°, with overall reverse displacement in the core reference frame (Fig. F189B). The fault zone is at least 10 cm thick, but poor recovery precludes a precise measurement. The deeper fault zone is a minimum of several centimeters thick and shows well-developed planar fabrics. Reverse movement (core reference frame) can be determined from geometries of fragments similar to S-C fabrics in mylonite (Fig. F189C). In thin section, microstructural evidence also suggests reverse sense of movement in the core reference frame (Fig. F190A, F190B). Coarse plagioclase grains are highly fractured with moderate undulatory extinction, suggesting that the cataclasite developed by semibrittle mechanisms. Within the fault zone, plagioclase and, locally, amphibole fragments are present in a chlorite and actinolite matrix (Fig. F190C, F190D), indicating that alteration occurred under greenschist-facies conditions similar to those associated with the formation of pale green veins with chlorite and actinolite infilling. In this interval, only 0.8 m of core was recovered from 4.8 m of drilling. This poor recovery is probably related to the presence of lightly fractured fault rocks. It is therefore likely that a substantially wider zone of brittle deformation exists in this interval. Formation MicroScanner (FMS) images of the borehole walls acquired during logging show a south-dipping structure at ~765 mbsf slightly deeper in the hole, and an east-dipping structure is also recognized (see “Downhole measurements”). These structures may be related to the fault zone.

The third zone of cataclasis is located at 785 mbsf in Section 305-U1309D-161R-1 (Fig. F189D). It consists of 90% matrix with tiny fragments. The zone is subhorizontal, is a few centimeters in thickness, and has reverse-zone movement in the core reference frame. In thin section, pyroxene and plagioclase fragments are present in a carbonate matrix, indicating that the latest slip occurred under very low temperature conditions, probably <100°C.

The deepest zone of significant cataclasite was recovered at 1107 mbsf in Section 305-U1309D-230R-2 (Fig. F189E). The zone of very high strain ultracataclasite is thin (<10 mm), but the cataclasite zone is >10 cm thick and associated with veining around the fault zone. In thin section, protolith minerals are intensely fractured and bent/kinked with local calcite infilling. The mineral assemblages in the fault zone are similar to those in the fault zone at 785 mbsf.

Cataclasis is common along contacts between diabase intrusions and the adjacent gabbro (Fig. F191). In thin section, previously formed amphibole and plagioclase grains were fractured along the chilled margin of the diabase dikelet (Fig. F191B–F191E).

Serpentinization

Olivine-rich rock types are variably serpentinized; no systematic alteration of pyroxene to serpentine was observed (e.g., bastites). At scales ranging from core pieces (Fig. F192) to thin sections (Fig. F193), the degree of serpentinization is highly variable, ranging from <10% to >70%. Microstructural observations of moderately (<50%) serpentinized rocks indicate that alteration of olivine to serpentine initiated along irregular microfracture networks within coarse-grained olivine crystals (Fig. F193B). The serpentine-filled cracks in olivine commonly extend into adjacent pyroxene and plagioclase grains, where they may also be marked by serpentine infill. Serpentinization results in an irregular foliation, consisting of networks of branching and anastomosing serpentine veins with associated oxides (Fig. F193C). Some serpentinites are cut by talc-tremolite shear zones. One steeply southeast-dipping fault zone recognized in Hole U1309B (Figs. F115A, F188E) displays offset cleavages and oblique foliations that indicate a reverse sense of shear (in the core reference frame).

The process of serpentinization appears to be nominally static. Weakly altered or fresh olivine-bearing gabbros exhibit pervasively cracked olivine grains that are crosscut by an irregular network of fractures and little associated strain. Water percolation along these microfractures promoted serpentinization, with a progression from thin veins of serpentine along microcracks to progressive growth of these veins and, finally, a mesh texture. Fully serpentinized zones show pseudomorphs of olivine, and the magmatic textures, when recognizable, show no apparent shear displacement (Fig. F193A). In addition, microstructural observations suggest little or no dilation associated with serpentine vein formation in weakly altered rocks (Fig. F193B, F193D), as the magmatic texture of the surrounding, unaltered minerals (plagioclase and pyroxene) is not affected. Thus, serpentinization appears to have occurred without significant volume increase at the thin section scale, suggesting dissolved material was removed by the serpentinizing fluids during the hydration process.

The average intensity of cataclastic deformation is highest near the top of Hole U1309D, presumably related to detachment faulting, and becomes quite low (5 m average = <1) at depths below the fault zone at ~785 mbsf (Fig. F182). In the upper 100 m of Holes U1309B and U1309D, cataclastic deformation with an intensity of 3 and greater is concentrated in the upper gabbro units (Cores 304-U1309B-7R through 11R and 304-U1309D-4R through 8R) (Fig. F194), where many late-magmatic leucocratic dikes and low-temperature veins are present. Most diabase units have a cataclastic fabric intensity <1; a few diabase units in Cores 304-U1309B-2R through 4R and 304-U1309D-1R have an intensity >2. Although fault rocks with gouge or healed breccia cataclasite that are similar to those observed in Hole U1309B were not as abundant in Hole U1309D, the presence of a few locations with intense cataclastic deformation implies the existence of faults. These zones are shown as dotted lines in downhole figures. Deeper in Hole U1309D, several intervals with cataclastic fabric intensities ≥3 are present from 140 to 290, 600 to 650, 700 to 800, and at ~1100 mbsf.

Figure F195 shows the downhole dip magnitude of cataclastic fabrics above 130 mbsf in both holes. Diffuse late-magmatic leucocratic cataclastic zones are most abundant at depths above 60 mbsf in both holes. By contrast, the low-temperature diffuse cataclastic zones are distributed evenly. In general, the dips of low-temperature diffuse cataclastic zones are widely scattered, although local concentrations with more restricted dips are observed. For example, there is a local concentration of dips between ~20° and 60° at depths between 65 and 80 mbsf in Hole U1309D, which may correlate with a similar group observed at depths between 80 and 90 mbsf in Hole U1309B.

Figure F196 shows the downhole dip magnitude of cataclastic fabrics and rakes of mineral striae on low-temperature veins in Hole U1309D to ~400 mbsf. Dips of fault and semibrittle zones range from 10° to 60°, whereas striated fault veins dip as much as 85°. The fault veins are primarily distributed in the interval between 140 and 350 mbsf and show oblique slip to dip-slip displacement. Normal sense displacement is observed for almost all oblique slip faults. Dips of late-magmatic leucocratic cataclastic zones and low-temperature cataclastic zones are widely scattered; late-magmatic leucocratic intrusions are primarily distributed in three intervals: from 35 to 80 mbsf, 190 to 250 mbsf, and 290 to 400 mbsf.

Veins with fibers are good indicators of late-stage shear sense and slip. During Expedition 305, fault veins were identified (Fig. F197) corresponding to the dark and pale green veins (see “Expedition 305 vein description”). Although there is no apparent systematic dip or variation of dip with depth (Fig. F197A, F197B), >60% of the observed plunges of fibers are <30° and only ~15% of them plunge at angles >60°. Assuming no significant rotation since the formation of these fibers, these fault veins indicate primarily oblique-slip deformation, with no clear indicators of sense of shear.

Downhole distribution and orientation of open cracks and drilling-induced fractures

Open cracks identified along the core as not being related to drilling show a very low overall intensity (predominantly values of <1; see “Structural geology” in the “Methods” chapter), with no marked trends (Fig. F198). In several intervals of the core, open fracture intensity (particularly the 5 m running average) shows an anticorrelation with rock recovery (Fig. F198A) and correlation with the intensity of drilling-induced fractures (Fig. F198C). The intensity of open and drilling-induced fractures is relatively high in the interval between 720 and 750 mbsf, where the average recovery is low (<50%) and where several fault zones are identified. A zone of intense drilling-induced fractures is also found at ~1200–1300 mbsf, below the olivine-rich troctolite unit at ~1200 mbsf; this interval also coincides with a zone of low recovery (Fig. F198). There is no correlation between the intensity of open fractures and recovery in this interval. Open fractures appear to be one of the latest deformation events, and in the interval between 575 and 775 mbsf they are often associated with sulfides.

Downhole distribution and orientation of serpentine foliation

The vertical distribution of the dip and intensity of serpentine foliation are shown in Figure F199. Zones with low-to-moderate values (intensity ≤ 3) are concentrated in the upper part of the borehole (300–530 mbsf), along a broad zone ~30 m thick at ~560 mbsf and over the interval between ~1090 and 1235 mbsf corresponding to olivine-rich rock types (Fig. F199A). Between 675 and ~1000 mbsf, serpentine foliation is scarce and more localized and has very low intensity values. The dips of serpentine foliation are random, showing variations in orientation at scales <1 m (i.e., along individual core pieces). Similarly, the degree of serpentinization varies at length scales <1 m in intervals with similar olivine content. This heterogeneity in the structure and degree of serpentinization is also observed at the thin section scale (Fig. F193). There is little or no correlation between the degree of total alteration (see “Metamorphic petrology”) and serpentinization in the upper 900 m of the Hole U1309D, but the zone with intense serpentine foliation at ~1090–1235 mbsf corresponds to overall alterations of >50% with very strong local alteration gradients.

Correlation of cataclastic fabrics and rock type

Histograms of cataclastic fabric intensity (Cf) illustrate the relationships between rock types and cataclastic deformation (Figs. F200, F201). Whereas the majority of plutonic rocks of Hole U1309B are cataclastically deformed, the diabase and basalt units are largely undeformed. This suggests that diabases and basalts formed late with respect to detachment faulting. In contrast, below 400 m in Hole U1309D, diabase and oxide gabbro show higher intensities of cataclastic strain than other rock types (Fig. F201). The intensity of cataclastic deformation diminishes below the fault zone recovered at ~785 mbsf, suggesting that rocks above and below this boundary accommodated different magnitudes of brittle strain.

Locally, cataclasis also concentrates along leucocratic veins. Enhanced cataclasis in diabase, oxide gabbro, and leucocratic veins appears to reflect the semibrittle mode of emplacement for these rock types, which formed relatively late in the magmatic history. In Hole U1309D below 400 mbsf, olivine-rich rock types have undergone less overall cataclastic deformation than gabbros, which, in turn, bear less cataclastic deformation than oxide gabbros (Fig. F201). A significant proportion of the gabbro and olivine gabbro logged during Expedition 304 have Cf = >1, whereas oxide gabbros and ultramafic rocks are less cataclastically deformed (~20% have Cf = >1) and troctolites are even less. At this stage it is not clear whether this reflects the proximity of certain rock types to the detachment fault or another model needs to be invoked.

Correlation of cataclastic fabrics and logging data

We investigated the correlation between cataclastic deformation intensity and downhole logging results in Holes U1309B and U1309D (see “Downhole measurements”). Owing to the lack of paleomagnetic data in zones of poor recovery, we have few orientation data on fault zones, per se, to compare with logging data. However, several fault zones inferred from core observations in Hole U1309B (especially those at 60 and 80 mbsf) correspond to fault zones inferred from FMS images. In Figure F202A, we show the intensity of cataclastic fabrics in Hole U1309D averaged over 5 m intervals; the depths of possible fault zones are shown based on core observation. These locations correlate relatively well with the depths of faults inferred from resistivity and porosity data (Fig. F202B). In several cases, no rocks were actually recovered at the exact depth of the fault identified from the logging data. For example, the fault observed at 108 mbsf in the logging data corresponds to a depth where there was no recovery in Hole U1309D. However, in these cases, high cataclastic intensity, involving local brecciation, is observed in the adjacent rock.

Reorientation of structure data using paleomagnetic data

Paleomagnetic data were used to constrain the orientation of structural features in a geographic reference frame. The average inclinations of the remanent paleomagnetic vectors in samples recovered from the upper parts of both holes (all of Hole U1309B and the upper ~180 m of Hole U1309D) are amazingly consistent with that expected for the latitude of Atlantis Massif (Figs. F203, F204), suggesting that only modest tectonic rotations on a presumed ridge-parallel horizontal axis have occurred since the remanence was attained. Based on this hypothesis, we rotated cores for which both structural and paleomagnetic data were obtained about a vertical axis such that the declination of the remanent vector in each core was the same; only data from Hole U1309B and the upper 120 m of Hole U1309D were used in this analysis. The orientations of structures were placed in geographic coordinates by assuming that this vector plunges upward to the south (for reversed polarity) or downward to the north (for normal polarity). Caveats regarding this analysis must be kept in mind:

• Logging data indicate that Hole U1309B deviated 7° from vertical (it plunges 83° toward an azimuth of ~42°), but we did not account for this deviation in this preliminary analysis.
• The orientations of high-temperature structures may have rotated significantly between their formation (at magmatic temperatures) and the time when magnetic remanence was obtained (at ~520°–600°C). Modest counterclockwise rotations on axes subparallel to the ridge axis would not change the inclination of the remanent vector significantly. For example, a counterclockwise rotation of 20° about an axis oriented at 00°/010° (plunge/​trend) would produce no net change in the inclination of the remanent vector. Thus, significant rotation of structures can occur below the remanence acquisition temperature, although they are not detectable using paleomagnetic data. By contrast, a counterclockwise rotation of ~60° would produce a ~20° shallowing of the remanent inclination; such rotations would be resolved by paleomagnetic data.
• On a sample by sample basis, there may be as much as 20° uncertainty in the orientation of the remanent vector relative to the geographic poles owing to secular variation.

The orientations of crystal-plastic and magmatic structures reoriented using the paleomagnetic data are shown in Figure F205. No strong preferred orientation of magmatic fabrics is observed. This observation is consistent with our macroscopic examination of the core, through which we documented relatively large changes in the orientation of compositional banding across narrow intervals within individual pieces (e.g., in Sections 304-U1309D-12R-1 [Piece 6] and 12R-2 [Piece 9]). For crystal-plastic deformation, fabrics dip toward the west-southwest; lineations, where observed (the best examples of which were in Core 304-U1309D-8R), plunge downdip.

Stereonet diagrams displaying vein orientations from selected samples rotated into a common geographic reference frame using paleomagnetic data are shown in Figure F206. Type 1 veins in Hole U1309D preferentially dip to the north at ~25°, but no preferred orientation of Type 1 veins is evident. The orientations of Type 2 veins are apparently random in the upper 40–50 m of both holes. In contrast, at depths between 50 and ~80 mbsf, Type 2 veins dip ~35° to the east-northeast in both holes (Fig. F207). Type 3 veins dip predominantly toward the east in Hole U1309B (Fig. F207A); a similar concentration of Type 3 veins is observed for Hole U1309D, in addition to a small population of possibly conjugate veins that dip to the west (Fig. F207B). The orientations of Type 3 veins do not change significantly with depth. The similar orientations of Type 2 and Type 3 veins suggest that they formed in approximately the same stress field during upper to lower greenschist-grade conditions.

Cataclastic fabrics reoriented using paleomagnetic data are shown in Figure F208. Cataclastic zones generally dip shallowly to the north, whereas a majority of low-temperature cataclastic bands dip northeast. The orientations of late-magmatic leucocratic cataclastic bands are widely scattered. These trends are consistent between Holes U1309B and U1309D.

Summary of crosscutting relationships

Below we provide a summary of crosscutting relationships based on a combination of core description and thin section observations (from oldest to youngest; in this list, the symbol ≤ indicates “younger or the same age as”):

• a. High-temperature (~1200°C) deformation of peridotite
• b. Impregnation of residual peridotite by basaltic melt to form wehrlite
• c. Formation of cumulate dunites and magmatic flow of layered gabbros and troctolites
• d. ≤ c. Intrusion of coarse-grained gabbro cutting harzburgite in Hole U1309B
• e. Gabbro cutting troctolite (e.g., Core 304-U1309D-8R)
• f. Intrusion of oxide gabbros
• g. ≤ f. High-temperature, granulite-grade crystal-plastic flow concentrated in gabbros
• h. Late-magmatic leucocratic dikes and high-temperature hornblende veins cut crystal-plastic foliation
• i. Relatively widespread but low-strain veining and semibrittle deformation of gabbros (predominantly lower amphibolite through greenschist grade)
• j. Intrusion of diabase sills (predominantly lower amphibolite through greenschist grade)
(somewhere between i and k) Serpentinization of harzburgite
• k. ≤ j. Greenschist-facies brecciation and faulting
• m. ≤ l. late talc, prehnite, and carbonate veins
• n. Exhumation and denudation at low temperatures

Definition of structural units

Based on a combination of the spatial distribution of structural observations and other shipboard data, we defined structural units for the core logged during Expeditions 304 and 305. These structural units are intended to (1) be used as guides to evaluate which structures played a larger role in the tectonic evolution of the region and (2) define intervals of the cores that shared common deformation histories. Each of the structural unit boundaries coincides with a boundary identified through petrological or geochemical observations, relatively high cataclastic deformation intensity, and the location of faults identified in the logging data.

Hole U1309B

Hole U1309B contains three structural units (from bottom to top):

• Structural Unit 1a, a basalt-diabase brecciated unit;
• Structural Unit 1b, a generally coarse grained gabbro unit, which includes an intrusive contact with serpentinized peridotite; and
• Structural Unit 1c, a magmatically layered gabbro unit.

Hole U1309D

Upon drilling the rock below 130 mbsf in Hole U1309D, the scale of observations provides a different perspective for the definition of structural units. Based on the observations in Hole U1309D, three structural units were identified. In Figure F209, we illustrate the downhole intensity plots for magmatic, crystal-plastic, vein, and cataclastic fabrics averaged over 5 m intervals to 400 mbsf (Expedition 304). In addition, we show the faults that were identified by comparison of the cataclastic deformation intensity and the logging data during Expedition 304. Finally, we show the depths where there are significant concentrations of the different vein types by vertical bars. The boundary between the units (red lines in Figs. F209, F210, F211) are

• Structural Unit 1, 0–176.3 mbsf;
• Structural Unit 2, 176.3–785 mbsf (structural Unit 2 may be split into Subunits 2a and 2b with a boundary at 257.2 mbsf) (pink line in Fig. F209);
• Structural Unit 3, 785–1415.5 mbsf (structural Unit 3 may be split into Subunits 3a and 3b with a boundary at ~1100 mbsf).

In all cases, the boundaries should be considered as approximate.

As illustrated in Figure F209, the base of structural Unit 1 is defined by a relatively thick zone of crystal-plastic deformation and a zone of high vein and cataclastic fabric intensity. It also marks the initiation of a second zone of relatively high crystal-plastic and magmatic deformation intensity. The boundary between structural Units 1 and 2 includes the fault observed at 170 mbsf, which is well located by the logging data and marks the initiation of a new zone of high-temperature amphibole veins. The base of structural Unit 1 is coincident with the base of gabbro Zone 3 as defined in “Igneous petrology” (Expedition 304). Finally, the base of structural Unit 1 also defines the depth where a significant decrease in the inclination of the remanent paleomagnetic vector begins.

The base of structural Subunit 2a is defined by the depth at which the zone of relatively high crystal-plastic deformation stops and a zone of relatively high cataclastic deformation intensity that correlates well with the logging data. The base of structural Subunit 2a is also marked by an abrupt increase in the inclination of the remanent paleomagnetic vector (Fig. F210) and the base of gabbro Unit 5. The remainder of structural Unit 2 (Subunit 2b) is characterized by a generally low intensity of deformation fabrics.

The boundary between structural Units 2 and 3 at ~785 mbsf (Fig. F211) is defined by the base of a sequence of faults identified between ~700 and ~785 mbsf. This zone also marks the location of an abrupt decrease in both the intensity of veins and cataclastic deformation and a relatively abrupt decrease in Mg# (~600 mbsf). Structural Unit 3 is characterized by very low intensity of both high- and low-temperature deformation. The boundary between structural Subunits 3a and 3b (~1120 mbsf) is marked by a fault zone, a down-temperature deflection in the data recorded by the TAP tool (see “Downhole measurements”)—which suggests fluids are moving along this fault zone—and another rather abrupt decrease in the inclination of the remanent paleomagnetic vector (see “Paleomagnetism”).

Correlations between Holes U1309B and U1309D

Below are listed, in order of confidence, the pro and con arguments to correlate structures and rock types recovered in Hole U1309B and in the upper 130 m of Hole U1309D (Fig. F212).

1. Diabase units in Cores 304-U1309B-19R through 20R and 304-U1309D-15R through 16R.
• Pro: Correlation of MS and orientation of phenocryst alignment.
2. Lower layered gabbro sequence in Cores 304-U1309B-15R through 18R and gabbro Zone 2 in Cores 304-U1309D-11R through 14R (and perhaps 17R).
• Pro: both units have similar vein and brittle deformation fabrics, lithology, and magmatic deformation fabrics.
• Con: the sequence in Hole U1309D has a slightly higher crystal-plastic deformation intensity.
3. Upper gabbro sequence in Cores 304-U1309B-8R through 11R and gabbro Zone 1 in Cores 304-U1309D-4R through 10R.
• Pros: both units have similar intensities of cataclastic fabric, vein intensity, and crystal-plastic fabric. In addition, the orientations of veins and crystal-plastic foliations are similar in these two sequences.
• Cons: gabbros in Hole U1309D are more primitive and exhibit a higher intensity of magmatic fabric.
• Cross-examination: the observation that crystal-plastic deformation appears to postdate intrusion of gabbro into “troctolites” indicates that differences in lithology predate most of the deformation, and therefore juxtaposition, of these gabbro sequences.
  
4. Ultramafic rock in Cores 304-U1309B-11R and 304-U1309D-10R.
• Pros: similar structural position; similar lithology; the base of both units appears to mark a major change in deformation style and orientation of brittle fabrics.
  
• Cons: Although we have evidence for a fault beneath Core 304-U1309B-11R from both examination of the core and logging data, no evidence for a fault was observed at the base of either Core 304-U1309D-10R or the top of Core 304-U1309D-11R. Although these units represent the only ultramafic rocks recovered above 130 mbsf and they both show evidence for high-temperature mantle deformation, they are not strictly the same rock type (as described in “Geochemistry” and “Igneous petrology”).
  
5. Diabase units in Cores 304-U1309B-5R through 6R and 304-U1309D-2R through 3R.
• Pro: some correlation of phenocryst alignment, parallel to lower diabase unit.
• Con: correlation of phenocryst alignment is better between Cores 304-U1309B-5R through 6R and Core 304-U1309D-7R.
6. Diabase units in Cores 304-U1309B-12R through 13R and 304-U1309D-7R.
• Pro: general structure parallel to diabase units described above.
• Con: parallel phenocryst alignment is not obvious.

Based on these observations, we present a schematic cross section of this region in Figure F212. There is evidence for relatively large “brittle” faulting with offsets of tens of meters. The prominent zone of cataclasite at 81 mbsf in Hole U1309B correlates with a conductive horizon in the downhole FMS log. This fault dips 50°–64° southeast, with an apparent dip of up to 50° in the north-south direction. Cataclastic deformation and low-temperature alteration is observed near 36 mbsf in Hole U1309D (near the updip projection of the fault). In this case, using the uppermost and lowermost diabase units in Hole U1309B as markers, the diabase between 44 and 48 mbsf requires either a steeper dip than the other two diabases or a normal offset by the fault. Although the exact trace and timing of this fault is not resolvable from the core data, this fault is probably the best example of a steeply south-dipping fault and, possibly, part of an extensional system controlling syn- to post-greenschist-grade denudation of the central dome.

Discussion

Hole U1309D penetrated 1415.5 m into the footwall of a major oceanic detachment fault system. Some of the major structural questions which will eventually have to be resolved concerning the formation of Atlantis Massif are as follows:

• Did the initial unroofing of plutonic rocks occur along a shallow or steeply dipping detachment system (Karson, 1991; Tucholke et al., 1998)?
• Did the detachment system root deeply and undergo deformation at high temperatures (Tucholke et al., 1998) or at a gentle dip at low temperatures (MacLeod et al., 2002; Escartin et al., 2003)?
• Can denudation be explained by a single master fault, or is a system of faults required; does the plutonic history of the footwall block reflect the operation of a syntectonic normal fault?

Research results from Expeditions 304 and 305 should eventually help to resolve these problems.

Near-surface structures (upper 170 mbsf)

The upper ~170 mbsf of Hole U1309D shows several characteristics that differ from the rest of the hole:

1. The presence of fault schist with residual mantle origin (i.e., relict spinels) (see “Igneous petrology”).
2. An abundance of fine- and coarse-grained undeformed diabase and basalt.
3. More abundant plastic deformation at granulite facies.
4. Greater lithologic heterogeneity, including the presence of mantle rocks (upper 170 mbsf). None of these features are present farther downhole in Hole U1309D.

The history of faulting at lower-amphibolite through upper-greenschist conditions is not well represented in the upper 170 mbsf. Very low strain semibrittle deformation in the gabbros is primarily amphibolite to upper greenschist grade. Rare samples of talc-tremolite schist preserve semibrittle fault textures and provide evidence for syntectonic alteration of ultramafic units. These schists are preserved either within or near diabase sheets and basaltic breccias that postdate the alteration. In some cases, these deformation fabrics are cut by coarse-grained, isotropic tremolite, suggesting that fault zones became more localized at lower temperatures. Most greenschist-grade alteration is static, including post-tectonic actinolite distributed throughout the gabbros and younger diabase.

We cannot ascertain at this stage if the relative abundance of granulite grade deformation in the upper 170 mbsf is an early expression of the detachment faulting or if it possibly relates to strain gradients during intrusion of a pluton into an ultramafic environment.

Deformation below 170 mbsf

There is a lack of major discontinuities below 170 mbsf. Prominent zones of magmatic fabric and low deformation are located at ~220 mbsf (magmatic foliation; 40° dip), 560 mbsf (magmatic foliation; 60° dip), 660 mbsf (high crystal-plastic strain; 70° dip), 685–790 mbsf (three greenschist cataclastic fault/shear zones; 50°–60° dip), 1107 mbsf (greenschist grade, cataclastic), and 1300 mbsf (crystal-plastic strain, 70° dip). For a surprisingly large number of the plastic and cataclastic shear zones, reverse senses of shear in the core reference frame could be inferred.

Shear-zone indicators do not allow a simple relationship between the exposed detachment at the surface of Atlantis Massif (which presumably has a shallow dip and normal fault geometry). Internal deformation below 170 mbsf in Hole U1309D is localized along shear zones and developed as cataclasites, but these structures appear to be minor compared to the detachment fault and are not easily related to the deformation history of the detachment fault system. We conclude that relative to the presumably extremely high strain of the exposed detachment fault, the footwall block behaved as a generally coherent block, at least below ~170 mbsf. The detachment-related deformation appears confined to the upper 170 m; the rest of the structures observed correspond to either deformation prior to the detachment (i.e., emplacement in the lithosphere of the different rock types) or to local zones accommodating differential strain during denudation and/or flexure of the footwall.

High-temperature history

Only a minor fraction of all gabbroic rocks recovered from Hole U1309D shows a magmatic foliation (i.e., ~22%). This suggests a tectonically stable environment, unlike the dynamic environment of mantle flow turnover inferred by some for fast-spreading ridges and as mapped, for example, in the Oman ophiolite (Nicolas et al., 2000). Whether this implies a setting within the lithosphere or at the base of the lithosphere cannot be resolved at this stage. To this end, the resolution of the pressure of formation of the igneous rocks will be important.

Within the overall low strain history of the igneous environment, strain variations can nevertheless be discerned. Both macroscopic and microscopic observations lead to the conclusion that troctolites (excepting the olivine-rich troctolites) record the highest and oxide gabbros record the lowest magmatic foliation. Beyond this, the combined igneous lithology and microstructural evolution (Fig. F162) (see “Igneous petrology”) can be interpreted in such a way that between 400 and 1400 mbsf there are two major zones of troctolite at ~560 and ~1100 mbsf, both of which lie above highly evolved oxide gabbro. The lithology between ~650 and ~1100 mbsf may be interpreted as an intrusive sequence ranging, on the first order, from troctolite, olivine gabbro, and (noritic) gabbro to oxide gabbro. In this interval, magmatic foliation intensity decreases and crystal-plastic strain increases (combined log of coarser and finer grained neoblast sizes in Fig. F162) upward. Whether olivine-rich troctolite intervals below 1100 mbsf are part of this sequence is unresolved. The top of this interval is marked by steep shear zones and oxide gabbro. The base (including the olivine-rich troctolite) is marked by a recurrence and intrusion of oxide gabbro from below and is associated with higher crystal-plastic strain.

Observations of igneous contacts suggest that more fractionated rocks generally intrude more primitive rocks. This lends support to a model where magmatic accretion occurred by progressive underplating of igneous units as opposed to a model where new melt batches are sandwiched between existing plutons.

From microscopic to macroscopic scale, there is an abundance of features suggesting a late formation of clinopyroxene with respect to plagioclase (but not with respect to the oxide formation): the additional observation that clinopyroxene does overgrow a magmatic unit contact at interval 305-U1309-190R-1, 86 cm, plus more circumstantial evidence of abundant megacrystic clinopyroxene and the poor magmatic fabric development in coarse-grained gabbros compared to fine-grained gabbros. The relative timing between intrusion of oxide gabbro and clinopyroxene growth suggests that oxide gabbro emplacement was late.

Magmatic veins are present throughout the hole, but their abundance and lithology seems broadly related to the host rock, although local exceptions do occur. Plagioclase-rich veins are present near troctolitic gabbros, oxide-rich veins are present near oxide gabbros, pyroxenite veins are present near coarse pyroxene gabbros, and leucocratic veins are present near more trondhjemitic bodies.

Strain localization

To provide an initial estimate of the amount of strain accommodated by crystal-plastic deformation at depths above ~130 mbsf, we assigned the following strain magnitudes to rocks with different deformation intensities in a semiquantitatively manner:

• CPf-4 and CPf-3, shear strain (γ) = 10 (the aspect ratio of recrystallized plagioclase and augite bands suggests that this is a conservative estimate)
• CPf-2, γ = 2
• CPf-1 and CPf-0.5, γ = 0.25.

The largest uncertainty in these calculations is likely the strain magnitude in high-strain regions. With these assumptions, we estimate a total displacement accommodated by crystal-plastic deformation of 6 m (~10% of the recovered core) in Hole U1309B and 35 m (~60% of the recovered core) in U1309D. By comparison, using the same relationships of strain to CPf and the deformation intensity measured for gabbros in the top 50 m of Hole 735B on the SWIR, >500 m of displacement was accommodated by crystal-plastic deformation directly beneath the hypothesized detachment at Atlantis Bank (Dick, Natland, Miller, et al., 1999; MacLeod et al., 1999).

Brittle history

A major observation at Site U1309 is that footwall rocks have undergone independent low- and high-temperature deformation histories, with little overlap in regional amphibolite-grade deformation events within Hole U1309D. There are, however, instances of syntectonic brown amphibole grains within gabbroic shear zones in the upper part of Hole 1309D, as well as associated with oxide gabbro. The relative scarcity of strain in the amphibolite facies indicates that there is a gap in the deformation history through the entire section between the history related to igneous events (excluded here are the late diabases) and lower temperature, subamphibolite-facies deformation.

One of the most consistent indicators of internal deformation is veins with fibers that indicate direction of movement (vein faults). In most cases (>75%), these fault veins display oblique-slip deformation, with scant evidence for dip slip (normal or reverse) below 600 m. This episode of veining is likely to be late and associated with the uplift of the footwall and may indicate internal shear during uplift or relative lateral motion of structural blocks within the footwall.

Top of page   |    Previous   |    Next