STRUCTURAL GEOLOGY

We present the results of structural observations of the core recovered from Hole 1274A, followed by a discussion of some of the problems in interpreting the crystal-plastic deformation. Three categories of observations were recorded in spreadsheet format (see the "Supplementary Materials" contents list): crystal-plastic deformation, alteration veins, and brittle deformation. These were supplemented by microstructural observations in 34 thin sections. The peridotites from this site were the freshest drilled during this leg (as much as 40% fresh olivine in harzburgite). However, pervasive alteration has led to some ambiguity in the interpretation of observed features. Details of the structural classification scheme for each feature are given in "Structural Geology" in the "Explanatory Notes" chapter.

High-Temperature, Low Strain–Rate Deformation

The harzburgites and dunites from Hole 1274A have undergone very little high-temperature crystal-plastic deformation. They are coarse grained and have mostly protogranular textures with only a few examples of weakly porphyroclastic harzburgite (~3% of the core). The protogranular textures are characterized by smooth, curved grain boundaries and range from simple equigranular grains (Fig. F36A) to grains with protruding lobes and included olivine (Fig. F36B, F36C). The grain interiors have minor broad, undulose extinction of some grains and, occasionally, subgrain boundaries in olivine. The average olivine and orthopyroxene grain size is difficult to determine, but most of the volume appears to be occupied by grains >5 mm, ranging to 20 mm in diameter. Small vermicular spinel grains associated with orthopyroxene or clinopyroxene are ubiqitous; much larger lobate spinels are also present in some thin sections. These textures may be interpreted as the result of grain boundary–controlled diffusion creep during mantle upwelling and melting. However, the lobate, intergrown shapes of many grains (Fig. F36D) together with the great abundance of delicate interstitial and intragranular late magmatic spinel grains (see "Igneous and Mantle Petrology") suggests that some of the earlier fabrics may have been overprinted during late-stage melt-rock reaction at the base of the lithosphere.

Some indications of high-temperature deformation are present within the harzburgites. These include kink bands with neoblasts in orthopyroxene, as well as neoblasts along some grain boundaries (Fig. F37A, F37G). Relatively closely spaced subgrain boundaries in olivine are also present (Fig. F37C). In interval 209-1274A-5R-2, 17–20 cm, arrays of kink band boundaries with clinopyroxene exsolution have formed in some orthopyroxene grains (Fig. F37B). Patches of orthopyroxene grains that are significantly smaller than the average grain size are also present in some sections (Fig. F37D, F37E, F37F). Intergranular clinopyroxene and intergrowths of olivine, orthopyroxene and clinopyroxene, and vermicular spinel suggests that the formation of these fine-grained patches is concurrent with the formation of the protogranular texture. Only rarely do high-temperature deformation features begin to dominate the Hole 1274A harzburgite fabrics, producing narrow intervals of weakly deformed harzburgite (crystal-plastic deformation intensity = 1) (Figs. F38, F39; Table T4).

High Strain–Rate Deformation

A single weathered pebble of peridotite mylonite recovered at the top of Core 209-1274A-2R appears out of place, as none of the adjacent pieces of core show the effects of weathering or a similar style of deformation. Therefore, it is possible that the mylonite is a piece of rubble that fell into the hole.

This mylonite consists of orthopyroxene porphyroclasts in a matrix of finely recrystallized olivine (Fig. F40A). Individual orthopyroxene porphyroclasts have length:width ratios of 5:1 (Fig. F40C) and often consist of complex arrays of kink bands (Fig. F40B) formed subparallel to the foliation plane. Relatively coarse, less recrystallized patches of olivine constitute enclaves or even augen within the finer-grained olivine matrix. Once formed, the finer-grained matrix can deform by diffusion creep rather than dislocation creep and is therefore weaker than the remaining enclaves of unrecrystallized olivine (Jaroslow et al., 1996). Locally, undeformed enclaves of olivine and spinel are also present in the mylonite, with the spinel grains terminating abruptly at the margins of the enclave, indicating that the mylonitic deformation took place after spinel crystallization (Fig. F40D).

Abyssal peridotite mylonites similar to this one constitute ~10% of dredged abyssal peridotites, mostly from oceanic transforms (Dick, 1989) and represent high-temperature (granulite to lower amphibolite) fault zones (Jaroslow et al., 1996). Given the style of deformation and its location at the top of the section, this fragment may indicate that the uppermost few meters of the outcrop are part of a mylonitic shear zone associated with unroofing of the peridotite block sampled in Hole 1274A.

Crystal-Plastic Foliations

Fifteen crystal-plastic or protogranular shape fabrics were measured in the Hole 1274A harzburgites. These fabrics were rare and hard to find, and all were determined by examination of the core with the binocular microscope. In many cases these foliations were very weak. These fabrics are defined by smaller intergranular pyroxenes and rare arrays of lineated spinels at a high angle to the serpentine foliation (cf. Ceuleneer and Cannat, 1997). It is not clear whether they represent crystal-plastic deformation or late crystallization of intergranular pyroxene and spinel from a melt migrating through the peridotite—or some combination of both. Because of their weak development, measuring the orientation of these fabrics on anything other than the cut face of the cores was impossible. Therefore, the measurements are of apparent dips in the cut face of the core, which was generally oriented parallel to the dip of the pervasive serpentine foliation in the Hole 1274A peridotites (see below). The apparent dips are shown as a pole figure in the core reference frame (Fig. F41A).

A plot of poles to the foliations, rotated into a common reference frame using the measured paleomagnetic declination, shows that they appear to be coherently oriented (Fig. F41B). This is a surprising result, as a simple plot of apparent dips, measured in the cut face of the core, would likely give a random scatter within a cone of error in the stereo plot. This result may imply some systematic relationship between the pyroxene and spinel shape fabrics and the serpentine foliation as the core was systematically split parallel to the dip of the serpentine foliation. This could reflect formation of the pyroxene and serpentine fabrics in similarly oriented stress fields, preferred formation of the serpentine fabric along preexisting planes of weakness related to the pyroxene fabric, or a coincidental clustering of data points from a small number of difficult and uncertain measurements.

With two exceptions in Section 209-1274A-18R-1 at 94.6 and 95.4 mbsf (expanded depth), the poles to foliation form a girdle in Figure F41B. The two outlying points represent foliations measured in peridotites just above the major fault gouge found at 94 mbsf and thus may reflect rotations due to late faulting. Another foliation, measured 40 cm higher in the same core, dips at 30°, similar to the 25° dip measured in the overlying Section 209-1274A-17R-2. There is no systematic variation in the dip of the measured foliation with depth in Hole 1274A. Thus, the girdle of poles in Figure F41B might represent outcrop-scale folding on the scale of tens of meters.

Postintrusion Crystal-Plastic Deformation

Late oxide gabbro or gabbronorite dikes were found in several intervals in the Hole 1274A core, concentrated near the bottom. Secondary minerals largely replace the dikes, and many are completely rodingitized. They preserve sharp contacts with peridotite, with pseudomorphed pyroxene locally perpendicular to these contacts. This indicates intrusion of the gabbroic dikes into the mantle peridotite at relatively high temperatures (probably >600°C). Patches of relict clinopyroxene and plagioclase are often undeformed but in other cases record a weak to moderate crystal-plastic deformation with the formation of abundant plagioclase deformation twins (Fig. F42A) and patches of polygonal neoblasts replacing strained plagioclase grains (Fig. F42B). In at least one case, partially replaced Fe-Ti oxides are intergranular to clinopyroxene neoblasts. The average deformation grade of the gabbros = 0.5, the highest of any lithology recovered, and is consistent with observations at Sites 1268 through 1272 that strain is partitioned into the gabbros relative to the peridotites during high-temperature deformation. A downhole plot of crystal-plastic deformation intensities is shown in Figure F38.

Cross-Fiber Serpentine Foliation

Cross-fiber serpentine foliation at Site 1274 is variable and appears to vary with the degree and style of serpentinization. Harzburgite containing more than ~30% fresh olivine is cut by numerous fine (<0.1 mm wide) anastomosing serpentine and magnetite veins that form pervasively through the body of the rock during the main phase of serpentinization. They can have preferred orientation or occur randomly. Where veins have a preferred orientation, they define a fairly strong foliation visible in thin section. Harzburgite containing less than ~30% fresh olivine is generally dominated by strong ribbon texture serpentine (O'Hanley, 1996), which defines a consistent planar to slightly anastomosing foliation (Fig. F43). Alternating zones of highly serpentinzed and fresher material appear to form layered bands in the core ranging 1–3 cm in width. These bands define a faint foliation in many locations that is generally parallel to fine serpentine veins and ribbons. These serpentine bands are deflected around clusters of large pyroxene porphyroclasts or are deflected into areas with lower than average pyroxene proportion. In some locations the serpentine bands are aligned perpendicular to the predominant long axis of pyroxene grains. Cross-fiber serpentine foliation is present in most dunite samples as serpentine ribbons and magnetite veins. These define a planar to slightly anastomosing foliation in most samples.

The harzburgites and dunites of Hole 1274A contain four generations of alteration veins (Fig. F44). The first consists of planar to curved, thin (<1 mm wide) black magnetite-serpentine veins (Fig. F44D, F44F) that are more readily seen in the dunites (e.g., Section 209-1274A-8R-1) and at the top of the core (Section 1R-1) where they are not obscured by later generations of veins. The second generation consists of rare planar (<2 mm wide) pale green-white picrolite veins (Fig. F44D, F44E) that are best developed in the dunites (e.g., Section 209-1274A-8R-1) and at the base of the core next to the fault zones (Cores 20R to 25R).

The third and most abundant generation of veins consists of small (1 cm long) planar to sigmoidal chrysotile veins (Fig. F44A). These veins occur throughout the harzburgites within the core but are less numerous in the dunite horizons (40–59 and 100–120 mbsf) (Figs. F44D, F45). They are composed of white chrysotile and occasionally white chysotile with pale green picrolite centers. The veins are dilational, often defining a distinct foliation (Fig. F44A, F44G) and sometimes form conjugate sets (Fig. F44B). The foliation dips 20°–90°, most commonly 30°–60°, with the dominant dip decreasing downcore (Fig. F45). Rarely, the veins show no foliation and instead form an irregular mesh or anastomosing network of veins (Fig. F44C).

The final generation of veins consists of late aragonite-carbonate veins (Fig. F44G) and associated oxides that fill planar to irregular brittle fractures (up to 8 cm long) and crosscut all three earlier generations of veins. The carbonate veins are confined to the upper 90 m of the core (Fig. F45) above the major fault zones in the bottom of the hole. Crosscutting relationships for all four generations of veins are shown in Figure F44D, F44E-F44G.

The intensity and orientation of veins were measured using the intensity scale outlined in "Structural Geology" in the "Explanatory Notes" chapter. The intensity of these veins is a measure of their average frequency in a 10-cm piece of core. The variation in total vein intensity with depth for Hole 1274A is shown in Figure F38. Alteration vein intensity is generally low (intensity 2.5) and is significantly lower than the vein intensity in the peridotites at Sites 1268, 1270, and 1271. However, the vein intensity is greater than that in the altered peridotites of Hole 1272A. This may simply be an artifact of the high degree of alteration shown throughout Hole 1272A, which may have obliterated some chrysotile veins. The average alteration vein intensity in Hole 1274A is slightly lower in the top 50 m of the hole than in the rest of the hole. This variation corresponds to the degree of alteration shown in the core (see "Metamorphic Petrology"). Some local peaks in the intensity curve appear to coincide with peaks in the brittle deformation intensity curve, suggesting that some of the veining can be correlated with brittle deformation. However, pervasive serpentinization appears to have preceded the brittle deformation shown by these rocks.

A total of 41 chrysotile foliation orientations that could be reoriented with the paleomagnetic data were measured in the core. Figure F46 shows a lower hemisphere plot showing the poles to the chrysotile foliation, restored with the declination of the stable remnant magnetization pointing north as discussed in "Structures in Peridotite and Gabbroic Intrusions" in "Mantle Upwelling, Melt Transport, and Igneous Crustal Accretion" in the "Leg 209 Summary" chapter. The poles to the chrysotile foliation are remarkably clustered, and the data suggest that the mean dip of the veins is to the south-southeast at ~40° with a few conjugate vein sets dipping steeply north-northeast (the reference frame that we used to reorient the data). These are parallel to the orientations shown by the fractures and cataclastic zones within the core. The orientation of the chrysotile veins may reflect the regional (extensional) stress field during serpentinization, consistent with the later formation of cataclastic gouge zones.

The lower one-third of Hole 1274B contains several horizons of serpentine mud and breccia interpreted to be fault gouge over intervals 209-1274A-15R-1, 34–36 cm; 17R-1, 57–59 cm; 18R-2, 6–13 cm; 18R-2, 26–36 cm; 20R-1, 113–121 cm; 22R-1, 58–76 cm; 23R-2, 0–133 cm; 24R-1, 0–74 cm; 24R-1, 107–113 cm; 24R-1, 118–120 cm; 25R-1, 12–25 cm; 25R-1, 40–49 cm; 26R-1, 5–17 cm; and 26R-1, 33–36 cm. Gouge/breccia intervals are partially cohesive and initially had the consistency of stiff clay when recovered from the core barrel. As it dried, the gouge became cohesive and friable. The gouges are dominantly matrix supported, with occasional narrow (<10 cm wide) clast-supported horizons. These horizons contain <10% to >70% clasts. Clasts are mainly subangular to subrounded serpentinized peridotite together with rare metagabbroic rocks and range in size from 0.04 to 5 cm. Clast size is generally finer in matrix-rich horizons and coarser in clast-rich horizons (Fig. F47). The matrix in most intervals is composed of serpentine mud and/or clays; however, some intervals have a carbonate-rich matrix (e.g., intervals 209-1274A-23R-2, 106–120 cm, and 24R-1, 0–8 cm).

Clast-rich and clast-poor gouges appear to have subhorizontal contacts with the surrounding peridotite. Some intervals of gouge have a weak to moderately strong subhorizontal foliation, but a few intervals have a steeply dipping foliation (up to 60°) with undulatory variations (Fig. F48). It is unknown to what degree this foliation was caused or disturbed by drilling.

Brittle deformation is generally low in the upper 74 mbsf of Hole 1274A above the major gouge units (Fig. F38) and restricted to several narrow zones of minor fracturing near serpentine veins. Cores 209-1274A-15R and below have high degrees of brittle deformation. Several intervals of gouge recovered from below 74 mbsf are very narrow but may represent wider fault zones with poor recovery. Wide gouge zones (>2 m thick) in Sections 209-1274A-23R-2 and 24R-1 probably represent a major fault zone. It is possible that all of the gouge samples collected from Sections 209-1274A-15R-1 through 26R-1 represent a single major fault system with numerous splays. Brittle shear zones and fractures measured in Hole 1274A generally dip between 20° and 60° to the southeast in the reference frame that we used to reorient the data (Fig. F46), subparallel to the chrysotile veins.

The absence of an unequivocally identifiable mantle deformation fabric is somewhat surprising, since models of mantle flow at mid-ocean ridges suggest pervasive deformation at the transition from vertical flow beneath the ridge to horizontal flow away from the ridge. At the hand sample scale in rocks from this hole, identification of high-temperature foliation in peridotite is difficult to separate from the pervasive chrysotile foliation. At the thin section scale, the coarse grain size often prevents identification of high-temperature foliation.

The subtle high-temperature crystal-plastic foliation is a crystal shape fabric resulting from elongation of orthopyroxene grains by slip along (001) and the elongated shape of otherwise undeformed protogranular pyroxene grains. Elongate protogranular pyroxene grains were occasionally parallel to elongate spinels. Elongation was generally observed in smaller interstitial pyroxenes that are texturally distinct from larger (2–7 mm), more abundant equant pyroxene grains.

The serpentine foliation surrounds enclaves of relatively fresh peridotite (Fig. F44A). These enclaves, generally 2–5 mm wide and elongate parallel to the serpentine foliation, are easily mistaken for stretched pyroxene porphyroclasts (Ceuleneer and Cannat, 1997). Examination under a binocular microscope and in thin section showed that when a crystal shape fabric was present it was subperpendicular to the serpentine foliation in the cut face of the core and occasionally parallel to elongate spinel trails. In many cases, individual pyroxene grains extend across the fresh peridotite enclaves into highly altered zones. These pyroxenes are pseudomorphed in the altered zones and are fresh in the enclaves.

The serpentine foliation is defined by chrysotile veins and may have formed in the following way. Alteration of olivine to serpentine is accompanied by a significant volume expansion (up to 40%) (O'Hanley, 1996). Olivine is preferentially altered compared to pyroxene, leading to stress concentrations around pyroxene grains. The differential stress is accommodated by dilation cracks that are filled with chrysotile. If pyroxene is relatively sparse and randomly distributed in the rock, a random pattern of veins is produced (Fig. F44C). If the pyroxene is strongly clustered or forms bands, the chrysotile veins become concentrated into the more olivine-rich bands aligned parallel to the pyroxene bands (Fig. F44A).

To summarize, the peridotites show a very weak high-temperature, low-strain mantle deformation fabric. They were intruded by gabbroic rocks near the base of the thermal boundary layer at high temperatures (probably >600°C). Deformation of the gabbroic rocks may have been coincident with the formation of the mylonite inferred to come from the top of the hole. These deformation events indicate the onset of strain localization at decreasing temperature, which ultimately produced the major fault gouges in the lower part of Hole 1274A. At lower temperatures, concurrent with the faulting, the rocks were partially altered by a succession of serpentinization events. The last recorded event is fracturing and seawater circulation indicated by carbonate veins.