Site 899, located over a basement high 18.5 km east-southeast of Site 897, presented a significantly more disturbed sedimentary section than the equivalent units at Site 897. The acoustic basement at Site 899, capped by unusual serpentinite breccia units (see "Lithostratigraphy and Petrology" section, this chapter), also displays an abundance of structural features that provide some insight into the formation, emplacement, and post-emplacement history of this unit.
The relative abundance of deformation features within the upper lithostratigraphic Units I, II, and III at Site 899 contrasts with the relative scarcity of structures in the equivalent units at Site 897. Post-depositional features observed at Site 899 include dipping beds, folded and contorted sediments, and microfaults. The distribution of deformation structures with depth is apparently not systematic, but observations were limited by core recovery and by the absence of marker horizons that display offset and disturbance.
Cores recovered between the seafloor and 229 mbsf in Hole 899A did not display any clear deformation structures, with the rare exception of bedding dips up to 10° (e.g., Core 149A-899A-14R). The absence of structures may result from their destruction by drilling disturbance, but it also suggests a tectonically quiet post-depositional history during the last 18 m.y.
Structural features were evident within sediments from Hole 899B. In the upper part of the cored interval of Hole 899B (Cores 149-899B-1R to -7R; 230.5-292.7 mbsf), the most common features were zones of highly disturbed sediment, which occurred at a frequency of about one per core. These were characterized by a lack of stratification (in contrast to the adjacent sediments), contortion of color variations, and fissility in the clay-rich sediments (Interval 149-899B-5R-3, 55-85 cm; Fig. 33). Because the consolidation state of these sediments was similar to that of the adjacent undisturbed sediments, we discount drilling disturbance as an explanation. These features could be distinguished from soft-sediment deformation by the high fissility of the core, suggesting that the disturbance post-dated burial. Bedding dips up to 10° also were observed in the upper part of the cored interval of Hole 899B, particularly in the structurally disturbed cores (e.g., Interval 149-899B-7R-7, 6-8 cm).
Microfaults occurred within several cores and were defined most clearly by offsets in color boundaries associated with burrows and depositional cycles (e.g., Interval 149-899B-10R-4, 43-49 cm; Fig. 34). Although several microfaults were observed in the upper cores of Hole 899B, they were best developed below 283 mbsf (Core 149-899B-7R and deeper). The sense of displacement was not always apparent, but both normal and reverse senses of motion were observed, with a maximum measurable displacement of 2 to 2.5 cm.
The top of Core 149-899B-16R marks the top of a new structural domain composed dominantly of serpentinized peridotite breccias in lithostratigraphic Unit IV. This enigmatic unit displays characteristics of both sedimentary and igneous rocks, and descriptions of lithology and "stratigraphy" are given elsewhere (see "Lithostratigraphy and Petrology" section, this chapter). Several structural aspects of this unit merit discussion, as they may bear on the origin of the breccias, as well as on emplacement and post-emplacement history. These include textural observations of syn-emplacement brittle and ductile deformation of the clasts and matrix, interactions between the breccia units and the underlying sediments and cobbles, and post-emplacement brittle fracture expressed by veining. In addition, evidence exists for pre-emplacement deformation of clasts within and below the breccia units.
The breccia units consist of angular fragments dominated by serpentinized peridotite, with a continuous size distribution in hand specimen and thin section (see "Lithostratigraphy and Petrology" section, this chapter). Aside from the occurrence of large (>l m) clasts, the breccia units are homogeneous and nearly structureless. However, subtle fabrics are present within the matrix that help to constrain the mode of origin and emplacement of this unit.
The shape and distribution of the clasts within the breccias record a history dominated by brittle fracture. Clasts are angular, particularly the finer grains, and show pervasive fracturing. Boundaries between clasts and matrix are sharp and often characterized by a reduction in the matrix grain size. Disaggregated grains having fragment trails were observed at both hand-specimen and thin-section scales (Interval 149-899B-26R-1, 50-64 cm; Fig. 18). Some clasts show two symmetric trails that wrap around the clast (Interval 149-899B-24R-3, 88-93 cm). This type of geometry demonstrates a rotational deformation that occurred in a shear regime. The process by which fragmentation took place apparently is frozen in Sample 149-899B-21R-1, 12 cm (Fig. 35), where two grains impinge, fracturing the smaller one. The preservation of fragment trails and the close association of lithologically similar grains suggest that the process of fragmentation was active during the last stages of emplacement.
Matrix textures generally were homogeneous and poorly sorted; however, rare occurrences of grain-size zoning were evident in some intervals (e.g., Interval 149-899B-23R-2, 92-106 cm; Fig. 36).
The texture of the matrix is dominated by brittle features. However, local "flow" fabrics are expressed as narrow shear zones composed of fine, fragmented, and rotated grains (e.g., Sample 149-899B-21R-3, 67 cm; Fig. 37), and occasionally are localized at grain boundaries (e.g., Sample 149-899B-21R-4, 43 cm; Fig. 38). In Interval 149-899B-26R-1, 22-45 cm, a vertical planar fabric has been bent into broad open folds (Fig. 39). These structural features suggest that the matrix flowed, an apparent inconsistency within the well mixed, dominantly brittle fabric that dominates the fabric in the breccia units. Examination of thin sections suggests that the grain fragments may lie within a "ductile" clay-rich matrix, possibly derived from the breakdown of serpentine. The flow fabrics are not uniformly distributed, but appear to occur preferentially near the bases of the interpreted breccia units, apparently near breccia/sediment contacts (see "Lithostratigraphy and Petrology" section, this chapter).
A clear penetrative deformation of the breccia was observed infrequently. A subhorizontal planar fabric near the top of the breccia unit (Sample 149-899B-16R-1, 128 cm) is associated with rotated clasts, alignment of elongate clasts, and a 2-cm-thick shear zone of high intensity deformation (Fig. 40). In Interval 149-899B-28R-1, 110-130 cm two parallel subhorizontal shear zones were observed; a 3-cm-thick shear zone occurs at the boundary of a serpentinized peridotite clast, affecting both matrix and clast, and another 2-cm-thick zone occurs in the clast, deflecting the previous serpentinite foliation (Fig. 41).
The breccias and other basement rocks recovered from Hole 899B are intercalated with sediments (see "Lithostratigraphy and Petrology" section, this chapter). In rare intervals, the sediment/breccia contact can be directly observed. An excellent example lies in Interval 149-899B-29R-1, 30-70 cm, where blocks of the breccia unit appear to have been fractured and injected by clay-rich mud (Fig. 24). The breccia fragments are intersected by calcite veins, which are apparently continuous with the mud. Our preliminary observations were not adequate for determining (1) if this contact is a normal depositional or weathering boundary or (2) if it represents the disaggregation of the breccia resulting from injection of the mud.
Below the breccia units, intervals consisting of mixed blocks of serpentinized peridotite, microgabbro, basalt, and sediment are thought to represent accumulations of allochthonous blocks. Some evidence exists for deformation within the discontinuous sediment intervals. A silty claystone at Interval 149-899B-30R-1, 5-15 cm, shows millimeter-scale mineralized microfaults, along which burrow marks have been folded and displaced in a normal sense. Meso-scale faults, having offsets of less than 1 cm, were observed within a thin laminated gray sandstone unlike any other lithology at this site (Interval 149-899B-35R-1, 103-122 cm; Fig. 26). This unit resembles siltstones that were recovered at Site 897, where similar microfaults were observed (see "Site 897" chapter, this volume). In Interval 149-899B-27R-1, 41-57 cm, the juxtaposition and textures of altered serpentinized peridotite and claystone suggest shear deformation in both lithologies (Fig. 23). The lithologies and the structural relationships are similar to those described for the Barremian sediments recovered at Site 897 (see "Site 897" chapter, this volume), suggesting that the two features may have similar histories. These contact zones also involve sand-sized soft serpentine that contains mafic or ultramafic clasts (see "Lithostratigraphy and Petrology" section, this chapter), which have been sheared and folded.
Except for some relatively fresh pyroxenite and pyroxene-rich peridotite clasts having a primary coarse-grained texture, most of the serpentinized peridotite clasts in Unit IV display evidence for serpentinite recrystallization that pre-dates the formation of the breccias. Except in some of the clasts, this recrystallization has overprinted previous deformation. Figure 42 displays a small clast of serpentinite that is highly sheared (Sample 149-899B-20R-3, 51 cm). The serpentine recrystallization is contemporaneous with intense fracturing that led to the development of veins of dark serpentine and an opaque mineral (magnetite?). This episode of veining was clearly pre-emplacement, as the veins are truncated at the clast boundaries. Rare calcite veins confined to the clasts also were observed, but the relative timing of the calcite and serpentine phases was not immediately evident.
Below the breccia units, near the bottom of Hole 899B (Interval 149-899B-34R-1, 6-37 cm), several pieces of mylonitized microgabbro were recovered. The largest piece displays a vertical foliation associated with a C-S fabric (Fig. 43). A vein parallel to the foliation shows clear evidence of intense shearing. The low recovery of this lithology, and the association with undeformed microgabbro in Interval 149-899B-34R-1, 38-73 cm, suggests that this fabric occurs within an allochthonous block. Cores 149-899B-36W, and -37R also contain some highly sheared chlorite schists of uncertain origin that are associated with microgabbro and basaltic pieces.
The breccia units and the underlying crystalline rocks are highly veined, which reflects a history of post-emplacement deformation. Two major vein mineralogies were identified: (1) serpentine generally confined within clasts and (2) calcite throughout both matrix and clasts. A third type of vein filled by material having a clastic texture was recognized only near the top of the breccia units and at depth in Core 149-899B-35R.
The serpentine veins were dominantly observed within clasts, particularly of serpentinized peridotite (e.g., Interval 149-899B-22R-2, 32-40 cm; Fig. 44). Most of these appeared to be primary features associated with late-stage, low-temperature fracturing of the serpentinized peridotite. However, several medium-to-dark green serpentine veins were observed to traverse the matrix and to transect clast boundaries (Interval 149-899B-20R-2, 40-50 cm). These enigmatic veins are rare and generally have been overprinted by calcite veins, but suggest that an early episode of brittle fracturing of the breccia unit may have occurred.
Calcite veins of various thicknesses are distributed unevenly throughout the breccia units (see discussion in "Lithostratigraphy and Petrology" section, this chapter) and correlate closely with the degree of calcitization of the unit. Large, sheeted, calcite vein complexes tend to be localized along the boundaries of large clasts within the breccias and exhibit multiple stages of crystallization. In some cases (e.g., Interval 149-899B-18R-5, 55-70 cm), calcite appears to have crystallized along preexisting veins composed of white serpentine, occasionally overgrowing (or replacing) the primary mineral (Fig. 45). Elsewhere, calcite is associated with green serpentine veins, some of which propagate through the matrix (Interval 149-899B-18R-2, 0-20 cm). Secondary veins spawned by the larger calcite veins can be observed to crosscut the pre-existing vein filling. This generation of calcite veins was observed to crosscut the fine intraclast serpentine veins and to transect clast boundaries and clearly was formed after lithification of the breccia units. These veins may have accommodated minor displacement, as suggested by offset fractured clasts (e.g., Interval 149-899B-21R-4, 39-2 cm; Fig. 46).
Calcite veins also were observed below the breccia units within crystalline clasts (e.g., Core 149-899B-31R; Interval 149-899B-35R-1, 25-32 cm) and, rarely, in association with sediments. The contact between sheared claystone and sheared, soft serpentine (Interval 149-899B-27R-1, 41-57 cm; Fig. 23) is marked by a thin calcite vein. The boundary between fluidized sediment and breccia at Interval 149-899B-29R-1, 30-70 cm, displays a vein that apparently is continuous with the apparently fluidized sediment, and cuts across the fractured breccia (Fig. 24). The absence of calcite veins elsewhere in the Unit IV sediments is significant, as this may constrain the relative timing of emplacement of the different lithologies observed within Unit IV. The last type of vein is filled with a fine material having granular texture. This type occurs rarely near the top of Unit IV (Sections 149-899B-16R-1 and -2) and at the very bottom of the hole in an apparently allochthonous block of diabase (Interval 149-899B-35R-1, 26-39 cm; Fig. 47). This vein type was observed to crosscut calcite veins, resulting in slight displacement (Fig. 46) or fragmentation (Fig. 48) of the preexisting vein. The significance of this vein is not yet clear.
The deformation history of Unit IV is partly preserved within the primary texture of the brecciated units, as well as in the post-emplacement fracturing and veining of the rock. The angularity of the fragments and clasts within the breccia units, the in-situ fragmentation and grain trails within the matrix, and the scarcity of ductile deformation within the matrix all point to a dominant ongoing brittle process during emplacement. The presence of solitary and symmetric tails on fractured clasts points to the translation and rotation of grains during breccia emplacement. The overall texture of the matrix might be described as "cataclastic" in the broadest sense, as it is characterized by brittle fracture and grain-size reduction. In tectonic settings, this fabric indicates shear deformation under low normal stress conditions, typically resulting from high fluid pressures.
The rare occurrences of grain flow within an apparently ductile medium (e.g., Interval 149-899B-26R-1, 22-45 cm) and ductile shear zones (Interval 149-899B-28R-1, 110-128 cm) point to local heterogeneities in rheology, possibly correlated with local variations in fluid pressure, composition, or degree of alteration.
The degree of deformation and alteration in the sediments intercalated with the breccias, and to a lesser degree in the crystalline rocks, may also be used to constrain the history and mode of emplacement of the breccias. A variety of sedimentary lithologies occurs over relatively short intervals, yet none shows clear evidence for significant shear deformation comparable to that evident in the breccia formation. Moreover, the mineralogy of the sediments suggests that they have not been thermally altered. This demonstrates that the relatively high-temperature mylonitization evident within several isolated crystalline blocks occurred prior to the breccia formation. Although some contacts between the breccias and sediment do suggest a small degree of mixing between them, the sedimentary units do not appear to have been brecciated. These observations point to relatively low-temperature conditions during emplacement, with only minor shear at the breccia/sediment boundaries. However, the sand-sized soft serpentine is clearly deformed, and critical intervals may have been missed because of limited core recovery, which restricted these interpretations.
Finally, the sequence of events in which the breccia units were deformed can be deduced from the superposition of textures and from the sediment ages. The primary textures of clasts within the breccias indicate serpentine recrystallization and a low-grade metamorphic event that mostly overprinted any prior deformation. However, before these two events, some evidence exists for the basement having undergone a structural evolution comparable to that interpreted for the basement rocks at Site 897 (see "Site 897" chapter, this volume). Subsequent brittle deformation and grain fragmentation accompanying the formation of the breccias have resulted in grain-size reduction and nearly complete lithologic mixing. The ensemble was fractured and veined at least once following lithification. In addition, there is a notable lack of veins within the sediments which may indicate that the fracturing and veining in the breccia may have predated the juxtaposition of the breccia and sediments. Alternatively, the lack of veins in the sediments may indicate more about the source of the vein fluids or the ease with which the two units fractured, than about the timing of emplacement. The ages of the sediments suggest an age no older than Barremian for emplacement of the serpentinized peridotite in Core 149-899B-34R and no older than early Aptian for emplacement of the Upper Breccia Unit (see "Lithostratigraphy and Petrology" section, this chapter). This is consistent with emplacement during the last stages of rifting along the western Iberia margin.