We present the results of a comprehensive structural analysis of the core recovered from Hole 735B during Leg 176. Five separate categories of observations were recorded: magmatic deformation, crystal-plastic deformation, brittle deformation, microstructural characteristics, and crosscutting relationships. Details of each structural classification scheme are given in "Structural Geology" in the "Explanatory Notes" chapter. We discuss the first four of these categories individually, presenting the relevant data with minimal interpretation. We then discuss the relationships, both temporal and spatial, between the categories before presenting initial interpretations in the context of the tectonic setting of Hole 735B. We conclude with a general summary of the structural geology of Hole 735B.
There is a considerable variation in the structures observed throughout Hole 735B, both in style and in position. Many of the characteristic trends observed in the upper 500 m of the hole during Leg 118 were also observed lower in the hole during Leg 176; however, there are a number of different features, notably the occurrence of significant reverse-sense shearing and a general decrease in deformation downhole. Wherever possible, we compare the section logged during Leg 176 with the data collected during Leg 118 to provide a complete structural overview of Hole 735B.
A majority of the rocks recovered from Hole 735B have coarse- to medium-grained hypidiomorphic-granular, intergranular, and subophitic textures with poikilitic pyroxene crystals and no preferred mineral alignment. Approximately 20% of the recovered material contains a variably developed magmatic foliation defined by the preferred orientation of elongate plagioclase laths and locally of pyroxene crystals. The relative proportions of different magmatic fabric intensities as defined in "Overview of Macroscopic Core Description" (in "Structural Geology") in the "Explanatory Notes" chapter and the percentage of the total recovery that each intensity value represents are shown in Figure F77. Rocks in which no igneous texture is preserved were recorded as magmatic fabric intensity = 0. Weak magmatic foliations predominate, whereas moderate to strong fabrics are only sporadically developed (e.g., Sections 176-735B-120R-1 through 120R-3, and 120R-7). Figure F78A displays the variation of fabric intensity as a running average over the entire core. In general, a weak magmatic foliation is variably developed throughout the core, but localized zones of well-developed foliations occur. Two intervals, ~1200 to ~1300 mbsf and ~1400 to ~1500 mbsf, are almost entirely isotropic.
Magmatic foliations vary in both strike and dip with depth. As illustrated in Figure F78B, there is no systematic variation between dips and depth, though the majority have dips between 20º and 50º toward 90º in the core reference frame (Fig. F79B). Marking of the core for splitting into working and archive halves by the Structural Geology Group followed a convention in which the predominant fabric in the rock (magmatic or crystal-plastic foliation, igneous layering, etc.) was dipping toward 90º in the core reference frame. Shipboard paleomagnetic declinations cluster around 250º in the core reference frame throughout the entire interval cored during Leg 176 (see "Paleomagnetism"). Thus, the clustering of magmatic foliation poles indicates that magmatic fabrics have a preferred orientation. Magmatic lineations, defined by elongate pyroxene crystals in the foliation plane, were noted in Section 176-735B-169R-4, 0-128 cm. A single lineation was measured from this section and is oblique to the dip direction with a trend and plunge in the core reference frame of 143º/33º.
An obliquity between multiple magmatic foliations was only observed in Section 176-735B-191R-3 at 38 cm, where an olivine gabbro with a weak magmatic fabric was intruded by a microgabbro. Eight centimeters away from the contact, the foliation within the olivine gabbro dips toward the contact. In the olivine gabbro immediately adjacent to the microgabbro, plagioclase and pyroxene crystals are subparallel to the margin of the microgabbro, but they do not appear to be deflected into this orientation. One interpretation of this geometry is that intrusion of the microgabbro occurred while the olivine gabbro was partially molten, allowing crystals to be realigned subparallel to the contact. Generally, the microgabbros have a weak magmatic foliation subparallel to the contacts.
The orientation of igneous layering was measured wherever continuous and planar contacts separate different layers. Fifty-six contacts separating modal or textural igneous layers were suitable for measurement and dip on average 24º (Fig. F80A). The modal and/or textural differences were often contained within centimeter-scale patches with irregular contacts indicating a lack of continuous layering. The scarcity of igneous layers is illustrated in Figure F80B, which displays the distribution of the 56 measurable layers (i.e., those with planar contacts) per 50-m intervals down the hole.
Layers with subparallel magmatic foliations occur in Sections 118-735B-83R-7, 176-735B-169R-5, 170R-4, 170R-7, 171R-1, 190R-2, and 195R-8. Generally, however, layers in rock cored during Leg 176 contain no magmatic foliation. Obliquities between magmatic foliations and layers were observed twice. In Section 176-735B-169R-5, centimeter-scale anorthositic and pyroxenitic layers dip 10°-20º less than a magmatic foliation of moderate intensity. Magmatic foliations crosscutting layers were not observed.
Igneous contacts that correspond to boundaries separating different lithologic intervals are interpreted as intrusive (41% of all contacts), gradational (37%), not recovered/preserved (16%), or tectonic (6%; see "Igneous Petrology"). The orientations of planar, intrusive, and tectonic contacts were measured; irregular intrusive and gradational contacts were not measured but noted in the Comments section of the MAGMATIC.XLS spreadsheet (see "Appendix" in the "Leg 176 Summary" chapter). With the exception of tectonic contacts, dips range from 0º to 90º, average 36º, and display no systematic downhole trends (Fig. F81). Tectonic contacts defined by crystal-plastic shear zones separating different igneous intervals occur from 450 to 590 mbsf (Cores 118-735B-82R through 88R and 176-735B-89R to 102R) and from 950 to 1150 mbsf (Cores 176-735B-148R to 169R). In the second of these intervals, the tectonic contacts shallow with depth from 55º at 950 mbsf to 7º at 1150 mbsf and display reverse-sense displacement (Fig. F81). Near the base of the hole, thin and sinuous microgabbros intrude the coarse-grained olivine gabbro along steeply dipping to vertical intrusive contacts. In contrast, microgabbros higher in the core have contacts with shallow to moderate dips.
Structures preserved in the rocks recovered from Hole 735B demonstrate that crystal-plastic deformation occurred over a wide range of conditions during emplacement and cooling of the gabbroic crust at the Southwest Indian Ridge. Fabrics were logged using a semiquantitative deformation intensity scale (Table T2, in the "Explanatory Notes" chapter) that ranges from undeformed (fabric intensity = 0) to ultramylonitic (fabric intensity = 5). The orientation of these fabrics was measured in the core reference frame, and where possible we noted the sense of shear of the deformation and whether deformation occurred in association with hydrothermal alteration.
The variation in the intensity of crystal-plastic deformation with depth is illustrated in Figure F82A. Macroscopic observations of deformation intensity from Leg 118 logged using the same intensity scale (Dick et al., 1991a) are also shown in Figure F82A. For comparison, the deformation intensity logged during Leg 176 is plotted beside that measured by Dick et al. (1991a) for the interval between 450 and 500 mbsf (Fig. F83); this series of cores from Hole 735B was relogged during Leg 176 in transit to the site. Data from Leg 176 are 11-cell running averages of the deformation intensity log (see PLASTIC.XLS spreadsheet in "Appendix" in the "Leg 176 Summary" chapter), whereas data from Leg 118 are "visual averages" of ~1- to 2-m-long sections (H.J.B. Dick, pers. comm., 1997). Despite the differences in logging techniques, the correlation between data from Legs 118 and 176 is good.
Crystal-plastic
deformation in Hole 735B is highly localized with the most intense deformation observed in
the intervals 0-50 and 450-600 mbsf (Fig. F82A).
Intervals with very little or no crystal-plastic deformation occur at 280-400, 700-800,
1180-1325, and 1400-1500 mbsf. At depths between ~450 and 700 mbsf, the regions with the
highest intensity of crystal-plastic deformation were later overprinted by brittle faults.
The strongly localized nature of crystal-plastic deformation is further illustrated by
calculating the percentage of the core from Leg 176 that displays a particular deformation
intensity. As illustrated in Figure F84, 77% of
the recovered rocks exhibit no crystal-plastic fabric (fabric intensity = 0) and 16% show
only a weakly defined foliation (fabric intensity = 1). Similarly, 71% of rocks recovered
during Leg 118 have no crystal-plastic fabric and 15% are weakly foliated (Dick et al.,
1991a). However, a significant number of intervals have high intensity deformation. For
example, a strong mylonitic fabric (fabric intensity = 
4) occurs in 127 intervals logged during Leg 176, of which 16 are
ultramylonites (fabric intensity = 5). One of the most plastically deformed zones is at
the base of a ~20-m-thick shear zone located between 945 and 964 mbsf (Fig. F85); the transition from mylonitic to undeformed
rock is just 10 cm below the base of the mylonite.
In general, there are
no systematic variations in the dip of the crystal-plastic foliation with depth in the
core from 0 to 1500 mbsf (Fig. F82B), with the
exception that mylonitic foliations (Pf = 4) tend to have a shallow dip at depths below ~1050 mbsf (Fig. F86A; see also Dick et al., 1991a). Although there is
no systematic trend in the dip of the foliation with depth, there is a strong
concentration of foliations that dip ~30º in the direction of 090º (i.e., poles that
plunge to 270º) in the core reference frame (Fig. F80);
090º in the core reference frame is defined in "Structural
Measurements" in the "Explanatory Notes" chapter. This relationship is primarily
a manifestation of the structural team's protocol in orienting the core for splitting.
However, the strong preferred orientation of both the deformation fabric and the average
declination of the remnant magnetic vector (Fig. F111)
indicates that the foliation has a consistent orientation in the hole. The orientations of
the foliation can be placed into a geographical reference frame by assuming that the
declination of the remnant magnetic vector dips toward the south. In this case, the
crystal-plastic foliation would predominantly dip toward the north and, therefore, toward
the ridge axis. This hypothesis is supported by initial analysis of FMS data reported in "Downhole Logging."
The distribution and orientation of both semi-brittle and retrograde (primarily amphibolite to transitional greenschist grade) shear zones with depth in the hole are illustrated in Figure F86B. The amphibolite to transitional greenschist grade shear zones were identified by the presence of deformed amphibole in hand specimen and confirmed by microscopic observations (Figs. F87, F88F, respectively). We use the term "semi-brittle" to describe rocks in which both brittle and plastic deformation fabrics were observed. A fairly high concentration of amphibole-bearing shear zones occurs at depths between 450 and 600 mbsf. Amphibole veins are also abundant in this interval (e.g., Figs. F47, F87); the correlation between amphibole veins and crystal-plastic deformation is described in detail in "Discussion". The highest density of amphibole-bearing shear zones occurs in the strongly deformed region directly above the fault observed at a depth of 490 mbsf (Fig. F86B). Semi-brittle shear zones occur throughout the hole, but these features are most abundant in regions adjacent to faults or major shear zones, especially the fault located at 690 mbsf.
One of the most surprising features of the crystal-plastic deformation fabrics from Hole 735B is the large number of shear zones with reverse sense of shear in the core reference frame. As illustrated in Figure F86C, the reverse shear zones are concentrated within and below the 20-m-wide shear zone located between 945 and 964 mbsf, and also directly above the fault located at 690 mbsf. Between 450 and ~680 mbsf, most of the shear zones have a normal sense of shear. An example of a reverse shear zone from Section 176-735B-155R-1 is illustrated in Figure F89. We were conservative in our assignment of sense of shear; in many cases, analyses based on asymmetric recrystallization tails were ambiguous. In addition, in many of the mylonitic rocks it was difficult to identify a lineation. The best kinematic indicators in the rocks from Hole 735B were deflected foliations (e.g., Fig. F89) and the orientation of oblique deformation fabrics (i.e., S-C fabrics).
A number of scenarios may be invoked to explain apparent reverse-sense kinematic relations in the higher temperature shear zones, including rotation of normal-fault blocks associated with displacement on low-angle detachments after the Hole 735B massif cools below the Curie temperature. Analysis of paleomagnetic data from Leg 118 indicates a 20º rotation of the tectonic block in which Hole 735B was drilled (Pariso et al., 1991). Assuming that this rotation occurs on a shallowly plunging east-west axis, and therefore that the crystal-plastic foliation predominantly dips toward the north, shear zones with a reverse sense of shear that dip less than ~20º (e.g., Fig. F89) could represent passively rotated normal-sense shear zones. However, a significant proportion of the reverse shear zones have dips in the range of 30°-60º in the core reference frame. These zones generally display macroscopic textures very similar to those exhibited by the gently dipping shear zones and would require block rotations significantly larger than indicated by the paleomagnetic analyses to be normal-sense shear zones. Thus, there is no explanation for all of the reverse shear zones, based on the current structural and magnetic data.
In numerous locations in Hole 735B, a magmatic foliation is overprinted by a crystal-plastic deformation fabric. The variations with depth of the average crystal-plastic fabric and magmatic fabric intensities are directly compared in Figure F79A. In general, there is a good positive correlation between the intensities of the two fabrics; however, there are numerous exceptions to this relationship. Obviously, no magmatic fabric is preserved in regions with a very strong crystal-plastic fabric (e.g., in the intervals 550-700 and 930-960 mbsf). In addition, several portions of the core exhibit a strong magmatic foliation and little or no plastic deformation fabric (e.g., the intervals 710-730, 960-1000, and 1075-1125 mbsf). The strongest correlation between the two fabrics is observed in the intervals 830-920 and 1300-1400 mbsf. Based on the texture in the core, we interpret these zones to show a moderate crystal-plastic overprint of a magmatic fabric; in regions with no crystal-plastic overprint, subhedral and nonrecrystallized plagioclase is observed in addition to sharp "corners" in interstitial pyroxene. In contrast, where a crystal-plastic overprint is observed, the pyroxene is more elongate with "rounded" edges, and the plagioclase is recrystallized (identified by a sugary texture in hand specimen). Microscopic observations discussed later in this section are generally in agreement with the macroscopic observations made on the core.
Brittle deformation recorded in the Leg 176 cores is divided into two major types, there being magmatic and lower temperature cataclastic deformation. Brittle magmatic deformation is represented by felsic magmatic or intrusion breccias and veins concentrated in the upper 70% of the hole (to ~1100 mbsf), at a variety of scales from micro- to macroscopic. Brittle features attributed to lower temperature cataclastic deformation at Site 735 include localized microfracturing, discrete faults, zones of cataclasite, veins, and joints. Thick zones of hydrothermal breccia with greenschist-facies mineralization similar to those reported in the lower portions of the Leg 118 section of Hole 735B (Stakes et al., 1991) are absent in the lower 1000 m cored during Leg 176. Direct correlation of lower temperature cataclastic deformation with data gathered during Leg 118 is not possible because occurrences of faults and cataclasites were not recorded during that leg.
Veins and associated breccias of magmatic, or probable magmatic, origin occur throughout the upper 1100 m of Hole 735B, diminishing in significance toward the bottom of the cored interval (Fig. F90A). These felsic veins and magmatic breccias are distinguished from veins of hydrothermal origin by the presence of the following mineral assemblage: plagioclase ± oxide ± clinopyroxene ± amphibole ± biotite ± alkali feldspar ± quartz ± titanite ± apatite ± zircon (see "Igneous Petrology" and "Metamorphic Petrology").
The felsic material ranges in macroscopic character from narrow 1- to 2-mm-wide veins (Fig. F91A) to centimeter-scale veins and irregular patches. Felsic patches greater than 1-2 cm in width commonly have diffuse, irregular, nonparallel boundaries throughout their length, and they typically host abundant rotated xenoliths of the adjacent wall rock (Fig. F92A, F92B). These breccia contacts have a broad scatter in dip down the core (Fig. F90B), with no significant preferred orientation.
At least two generations of felsic material are recognized in the core. The earliest is characterized by plagioclase (±oxide) that infiltrated and locally brecciated the host gabbro (Sample 176-735B-110R-2, 120-149 cm). On a local scale, there is a correlation between regions rich in plagioclase and areas with moderate to strong crystal-plastic deformation (Samples 176-735B-114R-3, 48-94 cm; 116R-7, 0-12 cm; 117R-2, 8-15 cm; 119R-1, 78-90 cm; 137R-5, 10-14 cm; and 145R-6, 101-113 cm: Fig. F92C). Younger, more evolved felsic material typically intruded after the high-temperature deformation (Samples 176-735B-110R-3, 0-11 and 111-142 cm; 118R-2, 38-41 cm; 118R-6, 120-129 cm; 147R-7, 98-102 cm; 156R-3, 18-20 cm: Fig. F92A, F92B).
Cataclastic features resulting from lower temperature brittle deformation in Hole 735B range in scale from micro- to macroscopic. The predominant cataclastic fabrics are discrete faults and zones of associated cataclasite of variable thickness, fracture intensity, and mineralogy. Based on thickness and fracture intensity of recovered cataclastic rocks, brittle deformation is primarily localized along two faults of unknown displacement (560 and 690-700 mbsf) marked by relatively low core recovery (<60%), and several minor zones of cataclasis (including 1076 and 1100-1120 mbsf). Both of the lower temperature faults overprint granulite and amphibolite grade shear zones (Fig. F86B) and were active at greenschist grade conditions. The lower 50 m of core obtained during Leg 118 and relogged at the beginning of Leg 176 includes another zone of significant cataclasis and faulting at ~490 mbsf, with several minor zones between ~460 and 495 mbsf. In addition, the fault zone at 560 mbsf is marked by a zone up to 5 m thick of significantly increased porosity and decreased density and resistivity (see "Downhole Logging").
Discrete planar faults with associated gouge, breccia, cataclasite, and ultracataclasite, cut all rock types and are associated with brittle and semi-brittle deformation in cores from Leg 176. More than 600 discrete faults and/or zones of cataclasis were logged using the intensity scale outlined in "Deformation Intensities" in the "Explanatory Notes" chapter (Fig. F93A, F93B). Faults and associated cataclastic deformation in rocks from Hole 735B are concentrated in the upper 50% of the hole (to ~720 mbsf) and are virtually absent below ~1400 m.
Intense cataclasis is extremely localized downhole into zones of variable thickness that range from less than 1 mm up to 5.5 cm (Samples 176-735B-97R-4, 78-84 cm, and 118R-1, 53-54 cm). Fault rocks of all intensities (see "Deformation Intensities" in the "Explanatory Notes" chapter) were recovered, including gouge (Sample 176-735B-97R-4, 127-130 cm), breccia and oxide-rich breccia (Samples 118-735B-87R-1, 98 cm; 176-735B-101R-1, 66-67 cm; 164R-1, 112-113 cm), cataclasite (176-735B-97R-4, 122-127 cm), and ultracataclasite (Sample 118-735B-86R-2, 0-3 cm). The most common brittle tectonic features are, however, small-offset "microfaults" filled with calcite (predominantly above 560 mbsf; Fig. F91C), amphibole (Fig. F88F, F91B), and/or smectite (below ~1050 mbsf). The pale-green smectite veins from the central and lower parts of Hole 735B (Cores 176-735B-127R through 131R, and Cores 176-735B-172R through 180R) often bear slickensided striae of variable orientation. Although these faults are common, they did not accommodate significant displacement, as the delicate smectite veins maintain subparallel margins. Indicators of the existence of faults not recovered include brecciated fragments (i.e., Samples 176-735B-98R-1, 9-30 cm; 100R-2, 0-20 cm; and 103R-4, 0-11 cm) and rubble (i.e., Samples 176-735B-99R-1, 0-5 cm; 131R-1, 76-84 cm; 162R-3, 122-134 cm; and 170R-3, 0-10 cm) with slickensided striae.
Mineral assemblages associated with the brittle tectonic structures range from calcite, amorphous silica, and prehnite to chlorite, epidote, actinolite, and secondary plagioclase, implying a range of conditions associated with cataclastic deformation (see "Metamorphic Petrology"). The alteration mineral assemblage characteristic of the zone of intense faulting and cataclasis in Section 176-735B-97R-4 (121-126 cm; 560 mbsf; see Fig. F92D) includes fractured clasts of gabbro with talc and amphibole after olivine, hornblende and actinolite after clinopyroxene, and secondary plagioclase. These altered clasts are enclosed in a fine-grained matrix dominated by chlorite and amphibole, indicative of greenschist facies metamorphic conditions during deformation.
Downhole variation in fault dip shows no consistent pattern (Figs. F93C, F94). However, the concentration of poles with a 90º plunge indicates a preferred subhorizontal fault orientation within the lower 1000 m of Hole 735B. Fault displacements observed on the core face were treated as components of dip-slip displacement, either normal or reverse (see "Structural Measurements" in the "Explanatory Notes" chapter). Displacement parallel to the trace of the fault was generally measured using displaced planar features, such as grain boundaries, vein walls, and/or igneous contacts. Fracture surfaces with slickenside orientations were used when available to differentiate between strike slip (pitch 0°-10º), oblique slip (pitch 11°-79º), and dip slip (pitch 80°-90º) movement. Of the logged faults, only a small number (7% total) provide information on relative displacement (Fig. F95A). A greater number of the faults were noted on fracture surfaces and vein walls, which allowed the determination of pitch, and therefore relative slip of these minor faults. Of those determined (Fig. F95B), oblique-slip was by far the dominant slip sense (8.8%), followed by nearly pure dip-slip (2.8%), and strike-slip (2.3%). However, the sense and magnitude of displacement associated with the discontinuities at 490, 560, 690, and 1100 mbsf are unknown. The discontinuity at 560 m clearly corresponds with the seismic reflector of Swift et al. (1991; Fig. F96).
In cores taken from Hole 735B during Leg 176, localized brittle and crystal-plastic deformation are closely associated spatially at numerous locations. Macroscopic and microstructural textural relations together indicate that in most instances brittle deformation overprints crystal-plastic deformation (Samples 176-735B-117R-2, 62 cm; 137R-3, 134-135 cm; and 146R-5, 139 cm). During visual core description, the term "semi-brittle" was assigned to features where both brittle and crystal-plastic structures coincide and the chronology of deformation could not be determined.
We measured the intensity and orientation of veins, joints (fractures), and subhorizontal microfractures (SHMs) using the intensity scale outlined in "Structural Geology" in the "Explanatory Notes" chapter. The intensity of these planar features is a measure of their average frequency in a 10-cm interval. The majority of joints were found to have minor mineralization along them and, for both structural and metamorphic analyses, these were logged as veins. For this reason the intensity and orientation of jointing is incorporated into the vein data.
The intensity of veining varies considerably throughout the cored interval, as shown in Figure F97A. This figure incorporates data from Leg 118 (H.J.B. Dick, pers. comm., 1997); the interval from 450 to 500 mbsf was logged during both legs, and comparison of the two data sets across this interval allows conversion of the uppermost 450 m data to the Leg 176 intensity scale. In the upper part of the hole, from 0 to 800 mbsf, there is a zone of fluctuating high vein intensity (up to an average of >10 veins per meter) that gradually decreases downhole. This is followed by an interval between 800 and 1100 mbsf of low vein intensity (<2 veins per meter). Below 1100 mbsf, the vein intensity is characterized by short intervals, typically 5-20 m, of high vein intensity separated by similar-sized intervals of little or no veining. This pattern continues down to the bottom of the core to 1508 mbsf. The variation in vein dip with depth is shown in Figure F94B; the abundance of data makes trends difficult to observe.
There are 17 different types of veins, which can be separated into four broad groups: (1) magmatic veins--compound felsic (plagioclase ± clinopyroxene ± amphibole); (2) amphibole veins; (3) carbonate veins--mostly calcite; and (4) smectite/zeolite veins (see "Metamorphic Petrology"). Magmatic veins occur throughout the section, but are most abundant in the region between 900 and 1100 mbsf. Both amphibole and carbonate veins are abundant in the interval between 450 and 750 mbsf but rarely occur below 750 mbsf. Smectite veins and, lower in the section, zeolite veins are concentrated into two major intervals (580-850 and 1050-1508 mbsf). These are the most abundant vein types in the entire core; however, the considerable width of magmatic veins makes them more significant in terms of the percentage of the cored interval. A more detailed analysis of vein minerals can be found in "Metamorphic Petrology".
In a few places there are crosscutting relationships between different vein types; in general amphibole veins postdate magmatic veins, whereas carbonate and smectite/zeolite veins postdate both the magmatic and amphibole veins.
The dip distribution and dip variation with depth for each of the major vein types is shown in Figure F98. Both the magmatic and the carbonate veins show an approximately normal distribution of dips, with a maximum in the 40°-60º range. As found in the upper 500 m of the hole (Robinson, Von Herzen, et al., 1989), the amphibole veins generally have steeper dips with a maximum between 60º and 80º. A sequence typical of the upper part of the cored interval (450-750 mbsf), illustrated in Figure F87A, shows a number of steeply dipping, parallel amphibole veins orthogonal to the foliation. A more detailed consideration of the relationship between the amphibole veins and the crystal-plastic foliation is given in "Discussion". Smectite and zeolite veins generally have a shallower dip, as shown in Figure F98G, with a concentration between 20º and 40º. Typical moderately dipping, parallel smectite veins are shown in Figure F87B. Note that below 1400 mbsf there are a significant number of vertical/subvertical veins, generally of smectite, zeolite, and chlorite; similar orientations only occur in the upper parts of the hole in the form of irregular, subvertical clay-lined fractures (Dick et al., 1991a).
The variation of vein dip with depth, averaged over 50-m intervals is shown in Figure F99. The average vein dip decreases downsection from ~70º at 450 mbsf to <20º at 1508 mbsf. This intriguing trend is accompanied by the shift from amphibole-dominated veining at the top to smectite/zeolite-dominated veining farther down the hole. However, there are several exceptions, notably a cluster of very shallow dips at about 1100 mbsf (due to a number of subhorizontal magmatic veins) and the vertical-dipping chloritic veins below 1400 mbsf mentioned earlier.
The temporal evolution
of vein orientations can be traced using the vein mineralogy. A down-temperature sequence
of magmatic-amphibole-chlorite-carbonate-smectite/zeolite (from ~900º to <100ºC)
indicates a general shift from steep vein orientations to shallower vein orientations. It
also indicates a temporal shift in fracture location downhole. Although initial inspection
of the data suggests a considerable change in the stress conditions with time, closer
analysis shows that many of the shallower, later formed smectite/zeolite veins occur in
conjugate sets. The orientation of the maximum principal stress direction (
1) calculated from these
conjugate sets generally plunges at >65º. The vertical chlorite/smectite/zeolite veins
found below 1400 mbsf could have formed in the same stress regime as the shallower dipping
smectite/chlorite veins. Therefore, there may have been no substantial shift in the
principal stress orientations during the formation of the veins in this section, although
there are no conjugate vein sets in the upper parts of the core.
The variation in the intensity of the subhorizontal microfractures with depth is shown in Figure F100. No effort was made to interpret the distribution of these features. Similar small, white, subhorizontal joints and veins were also described during Leg 118 (Shipboard Scientific Party, 1989); all are probably drilling-induced or unloading features.
The gabbros of Hole 735B display a wide variety of microstructures, ranging from nearly undeformed igneous textures to late brittle features, illustrating the progression of deformation with time and cooling consistent with that interpreted macroscopically. For each shipboard thin section, we recorded the type and intensity of deformation on an intensity scale (Table T4 in the "Explanatory Notes" chapter) from 1 to 6 (plastically undeformed to mylonitic or cataclastic); short descriptions of the main microstructural features are also included in the thin-section descriptions (see the "Core Descriptions" contents list).
Igneous textures that are completely free of any crystal-plastic overprint are extremely rare; where preserved, they are mainly found in fine-grained rocks. In the more common coarser grained rocks, the most pristine igneous textures consist of plagioclase with a few mechanical, tapered twins, undulose extinction, and/or subgrains (Fig. F101A, F101B, F101D, F101E, F101F). Olivine is often slightly deformed, as indicated by the common development of subgrains (Fig. F101F). Clinopyroxene is always undeformed. Fe-Ti oxides appear as aggregates or as interstitial phases, surrounding other grains or filling triple junctions. Igneous textures range from poikilitic (Fig. F101B), with mostly euhedral grains, to equilibrated (Figs. F101C, F101D, F101F), with mostly anhedral grains, 120º triple junctions, and curviplanar grain boundaries. A magmatic foliation (Fig. F101A), defined by shape-preferred orientation of plagioclase and pyroxene, is present in about 22% of the core (see "Magmatic Structures"), although it is sometimes difficult to see, especially in thin sections, because of coarse grain size. The transition from igneous textures to crystal-plastic deformation textures is marked by the increasing abundance of mechanical twins, subgrains, and minor recrystallization in plagioclase.
Crystal-plastic deformation often results in extensive dynamic recrystallization of olivine (Fig. F102A, F102C, F102D) and plagioclase (Fig. F102B, F102C, F102D). Clinopyroxene is typically undeformed, or slightly kinked in the less-deformed rocks, although we observed some subgrains in the lower 50 m of the core. In more deformed facies, clinopyroxene shows greater recrystallization and appears as porphyroclasts with recrystallized mantles and tails (Fig. F102C, F102D). Where Fe-Ti oxides are abundant in the highly deformed intervals, they appear as flattened ribbons (Fig. F102C), with subgrains and recrystallized grains that can be seen in reflected light. The smallest recrystallized grains are observed in the highly localized deformation bands (Fig. F102C).
As temperature decreased, crystal-plastic deformation was localized into narrow shear zones, which often grade into semi-brittle faults. All minerals appear as a fine-grained recrystallized matrix surrounding porphyroclasts (Fig. F88A, F88B, F88C, F88D, F88E). Plagioclase porphyroclasts are strongly deformed, displaying subgrains, undulose extinction, subgrain boundaries, and microcracks (Fig. F101A, F101B, F101C, F101D, F101E). Amphibole frequently replaces clinopyroxene in recrystallized tails around porphyroclasts (Fig. F88A, F88F) or occurs as highly deformed, very fine-grained bands (Fig. F88B, F88F). Semi-brittle shear zones are often associated with Fe-Ti oxides (see "Crystal-Plastic Structures"), the latter including numerous small recrystallized ribbons and individual grains of plagioclase (Fig. F88C, F88D, F88E). In some thin sections, Fe-Ti oxides locally crosscut crystal-plastic fabrics, including aligned recrystallized plagioclase grains, and then fill microcracks in plagioclase porphyroclasts (Fig. F88D).
Numerous brittle features are observed in Hole 735B (see "Brittle Deformation"), ranging from early magmatic veins and breccias to late hydrothermal veins and faults. Figure F91 illustrates a few brittle microstructures.
Microstructural observations of the deformation fabrics in gabbros from Hole 735B are consistent with structures observed macroscopically. This conclusion is supported by a comparison of macroscopic fabric intensities to thin-section analyses (Fig. F103). For example, in rocks for which no magmatic foliation is observed macroscopically, 150 of 168 display no magmatic foliation at the thin-section scale. Similarly, for rocks with a moderate to strong magmatic fabric, 23 of 29 display a magmatic foliation at the thin-section scale (Fig. F103A). A stacked histogram of observations of the intensity of crystal-plastic deformation illustrates that the dominant thin-section intensity corresponds to the macroscopic deformation intensity (Fig. F103B). For example, 100 of 147 rocks with no macroscopic crystal-plastic fabric (macroscopic fabric intensity = 0) show no evidence of plastic deformation at the thin-section scale, with the exception of the development of mechanical twins in plagioclase or subgrains in olivine or plagioclase. Similarly, all rocks with a macroscopic fabric intensity greater than 2 show either porphyroclastic (30%-70% recrystallization) or mylonitic (>70% recrystallization) fabrics at the thin-section scale (Fig. F103B). Not surprisingly, the greatest discrepancies between the macroscopic and microscopic observations arise for rocks with weak deformation fabrics. A number of factors, including coarse-grain size and heterogeneous deformation, are likely causes for these discrepancies.
The structures described in the gabbros of Hole 735B result from the evolution of deformation conditions with time and range from magmatic fabrics to late brittle faults and veins. Crosscutting relationships between the structures outlined above are described in hand samples as part of the VCDs (see "Overview of Macroscopic Core Description" (in "Structural Geology") in the "Explanatory Notes" chapter, and file X-CUT.XLS in "Appendix" in the "Leg 176 Summary" chapter) they are briefly summarized below. We then discuss correlations between some of the structures described in the previous subsections.
Igneous textures, with or without a magmatic fabric, are overprinted by crystal-plastic fabrics. The latter is always localized into relatively narrow zones (see "Crystal-Plastic Structures"). The stronger the foliation, the thinner the zone of plastic deformation. High-temperature crystal-plastic foliations are locally overprinted by lower temperature/higher stress crystal-plastic shear zones, often semi-brittle, and/or associated with hydrothermal alteration, as recognized both in hand specimen and thin section. Many of these have a reverse sense of shear in the core reference frame; some of them grade into faults. However, between 450 and 700 mbsf, most shear zones are normal. Brittle deformation always postdates the crystal-plastic deformation, except for early felsic veins and associated magmatic breccias (see "Brittle Deformation"). Later, more evolved felsic veins and associated breccias generally crosscut the crystal-plastic fabrics and are not plastically deformed. Other brittle features include hydrothermal breccias, veins with different mineral assemblages (see "Metamorphic Petrology" and "Brittle Deformation"), and faults. Faults have different crosscutting relationships with veins, depending on whether they developed after crystal-plastic or semi-brittle shear zones, or after veins (e.g., faults overprinting smectite veins are very common in the lower half of the hole; see "Brittle Deformation"). In the last 50 m of the Leg 118 cores, which were relogged during Leg 176, a few faults appear to be overprinted by crystal-plastic deformation (e.g., in Section 118-735B-85R-6); only one fault in Leg 176 cores was overprinted by crystal-plastic deformation (in Section 176-735B-147R-6).
Fine-grained intervals (microgabbros) are present throughout the entire core recovered at Hole 735B. The microgabbros display various types of contacts with the coarse-grained host rock, from very diffuse and irregular to sharp and straight, with various dips (Fig. F104). In the lowest 100 m, they are subvertical, generally a few centimeters thick (e.g., Fig. F104C, F104D). In many instances, they are clearly intrusive, exhibiting a weak to moderate magmatic foliation parallel to the contacts with the undeformed coarse-grained host rock. In Section 176-735-191R-3, a layer of microgabbro, with a magmatic foliation parallel to the contact, intrudes and crosscuts at a high angle a pre-existing magmatic foliation in the coarse-grained host rock (see "Magmatic Structures"). Microgabbros do not intrude coarse-grained gabbro previously deformed plastically (i.e., no crystal-plastic foliation is crosscut by a contact with a fine-grained interval). On the other hand, crystal-plastic foliations locally overprint both the fine- and coarse-grained materials (e.g., Section 176-735B-191R-1). These relationships suggest that intrusion of microgabbros predated significant crystal-plastic deformation. The variable style of the contacts indicates that microgabbro intrusion took place in hot, sometimes partially molten host rock.
An intriguing feature observed throughout portions of the weakly deformed intervals of the core is a positive correlation between both the intensity and orientation of crystal-plastic fabrics and pre-existing magmatic foliations (Fig. F79). This is particularly evident in the intervals 830-920 and 1300-1400 mbsf, where magmatic foliations are locally overprinted by relatively weak crystal-plastic fabrics. As discussed in the previous section, thin-section observations are generally consistent with interpretations made on hand specimen. We analyzed all 40 thin sections of rocks that displayed a crystal-plastic overprint of a magmatic foliation in hand specimens. Of these, the majority illustrated a magmatic foliation defined by plagioclase that contains deformation twins, subgrains, and minor recrystallization, and olivine that exhibits weak subgrain development and minor recrystallization. Pyroxene in these samples may be locally recrystallized, contain subgrains, or exhibit slight undulose extinction, but it is generally interstitial and undeformed. No magmatic foliation could be identified in a subset of the 40 samples; the majority of these samples were >30% recrystallized, thus a preserved magmatic foliation could not be identified.
There are a number of explanations for the correlation in intensity and orientation between magmatic and weak crystal-plastic fabrics. Crystal-plastic deformation may occur easily because of the existence of a mechanical anisotropy (i.e., a magmatic foliation). For example, the easiest slip system to activate in plagioclase is the b-plane in the [a] direction (i.e., [010],[100]; Ji and Mainprice, 1988), which is parallel to the long dimension of crystals aligned during formation of a magmatic foliation.
Foliation development may occur across the transition from magmatic to crystal-plastic conditions as temperatures decrease and the melt phase is either expelled by compaction or crystallizes in situ. Such overprinting relationships have been described in granitoids (e.g., Blumenfeld and Bouchez, 1988; Paterson et al., 1989; Miller and Paterson, 1994) and are interpreted to form first by magmatic strain, and then by crystal-plastic processes as the melt fraction decreases and the framework minerals accommodate further strain.
As was also observed during Legs 118 and 153 (Dick et al., 1991a; Agar and Lloyd, 1997), there is generally a strong association between regions rich in oxide and regions with moderate to strong crystal-plastic deformation. The variation in oxide content with depth in Hole 735B is compared to the variation in deformation intensity in Figure F105. Although the association is strong, there are numerous exceptions to the relationship, especially toward the bottom of Hole 735B. For example, in the interval between 1350 and 1320 mbsf, there are several regions with a moderate deformation fabric intensity and low proportions of oxides. Similarly, in the interval between 1200 and 1300 mbsf, there are several regions with a considerable amount of oxide and little or no crystal-plastic deformation; a good example of this situation occurs in Section 176-735B-153R-6 (1002 mbsf), in which a 30-cm-thick oxide-rich interval is free of crystal-plastic foliation, except for a weak foliation, developed over a few centimeters at the bottom, adjacent to the boundary with an undeformed olivine gabbro below. These observations suggest that the strong association between the presence of oxide and crystal-plastic fabric intensity is related to the localization of deformation in regions with a high concentration of highly fractionated melts. Once deformation is localized in a melt-rich region, the existence of low effective pressure may induce "magma-fracturing," potentially leading to a local increase in porosity and, therefore, permeability, promoting enhanced melt migration along the shear zone. In many occurrences, crystal-plastic and/or cataclastic deformation is concentrated in oxide-rich intervals (Fig. F88C, F88D, F88E); the absence of annealing in these textures suggests that this deformation occurs in the solid state, well below the gabbro solidus (see also Agar and Lloyd, 1997).
Both amphibole-bearing shear zones and amphibole veins are abundant at depths between 450 and 700 mbsf, suggesting a correlation between these features at a 100-m scale (see also Cannat et al., 1991; Dick et al., 1991a). The highest density of amphibole veins occurs in a region with the highest intensity of amphibolite to transitional greenschist grade semi-brittle deformation that is later overprinted by cataclastic deformation at lower temperature greenschist grade (Figs. F86B, F106 ). These observations indicate that deformation remains localized in the same region from granulite to lower temperature greenschist grade conditions. The strong correlation between deformed intervals and amphibole veins is much weaker or absent at length scales less than ~10 m (Fig. F106). The angle between the dip of amphibole veins and crystal-plastic foliations is shown in Figure F107. A majority of amphibole veins are roughly perpendicular to the crystal-plastic foliation, although the angle is variable.
Gabbroic rocks recovered in Hole 735B display a much wider spectrum of structures and microstructures than gabbros of fast-spreading environments. The upper gabbros cored during Leg 147 in Hess Deep (Gillis, Mével, Allan, et al., 1993) and the gabbros of ophiolites inferred to derive from fast-spreading ridges, such as the Oman ophiolite (e.g., Nicolas, 1989), do not have extensive crystal-plastic fabrics. The lack of planar and continuous igneous layering also contrasts with the layered gabbros observed in the lower two-thirds of the Oman ophiolite gabbro crust (Nicolas et al., 1996). Some structural characteristics such as the lack of well-developed layering, size of igneous intrusions, and large-scale tectonics of Hole 735B rocks are similar to observations in ophiolites inferred to have formed in a slow-spreading environment, such as the Trinity ophiolite (e.g., Boudier et al., 1989) or the Josephine ophiolite (e.g., Alexander and Harper, 1992). However, the general decrease in crystal-plastic deformation with depth in the core has not been described for these ophiolites.
Gabbroic rocks cored during Leg 176 at Hole 735B display magmatic, crystal-plastic, and brittle deformation features, as well as associated crosscutting relationships consistent with synkinematic cooling and extension in an ocean ridge environment. The following observations provide a basis for interpreting the conditions of deformation during evolution of this block of lower oceanic crust:
Additional observations relevant to the structural evolution of the Hole 735B tectonic block include the following:
These observations indicate that the processes that control crustal accretion at slow-spreading ridges are strongly influenced by localized deformation at conditions ranging from magmatic to low-temperature cataclastic. The correlation between structural domains and igneous intervals are consistent with this segment of the Southwest Indian Ridge not having a steady-state magma source. Rather, intrusion and deformation are likely episodic phenomena that may occur separately or synchronously, but at different rates. In many cases, zones of localized deformation remained active over a wide range of conditions (e.g., cataclastic overprint of oxide-rich crystal-plastic shear zones that were initially active under partially molten conditions) and possibly for a much longer duration than igneous intrusion. The abundance of crystal-plastic deformation is in contrast to observations at faster spreading rates (e.g., Hess Deep and the Oman ophiolite) and is generally consistent with the paradigm of the spreading rate dependence of crustal accretion at ocean ridges.
