Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Earthquake engineering wikipedia , lookup
Interferometric synthetic-aperture radar wikipedia , lookup
Post-glacial rebound wikipedia , lookup
Geomorphology wikipedia , lookup
Shear wave splitting wikipedia , lookup
Oceanic trench wikipedia , lookup
Plate tectonics wikipedia , lookup
Rheology and Tectonic Significance of Serpentinite Sheared serpentinite, Oman Greg Hirth1 and Stéphane Guillot2 1811-5209/13/0009-107$2.50 DOI: 10.2113/gselements.9.2.107 S erpentinites occur in many active geologic settings and control the rheology of the lithosphere where aqueous fluids interact with ultramafic rocks. The crystal structure of serpentine-group minerals results in diagnostic physical properties that are important for interpreting a wide range of geophysical data and impart unique rheological behaviors. Serpentinites play an important role during continental rifting and oceanic spreading, in strain localization along lithospheric strike-slip faults, and in subduction zone processes. The rheology of serpentine is key for understanding the nucleation and propagation of earthquakes, and the relative weakness of serpentinite can significantly affect geodynamic processes at tectonic plate boundaries. KEYWORDS : seismicity, ocean–continent transition, seafloor, subduction zone, serpentinization INTRODUCTION Plate tectonics is controlled by a combination of ductile flow and frictional resistance in fault zones. At plate boundaries, where deformation is localized along centimeter- to kilometer-scale shear zones, the influence of serpentinite on tectonic processes is linked to its unique rheological properties. While the rheological weakness of serpentinite relative to other rocks has been appreciated through field observations in various geological settings, such as along slow-spreading oceanic ridges, along strike-slip faults, and in subduction zones, the quantification of its influence on tectonic processes requires an understanding of the factors that control ductile flow and frictional resistance. Experimental observations constrain these properties and highlight the rheological behaviors that are important for understanding the role of serpentinization in strain localization in the lithosphere, its role in both fault creep and dynamic fault rupture, and the influence of the dehydration of serpentinite on the seismicity of subducting slabs and the exhumation of high- to ultrahigh-pressure rocks. RHEOLOGICAL PROPERTIES OF SERPENTINITE Since the pioneering work of Raleigh and Paterson (1965), who documented how dehydration reactions in serpentinite result in weakening and strain localization, numerous experimental studies have been conducted to investigate the physical properties of the serpentine minerals. These 1 Department of Geological Sciences, Brown University Providence, RI 02912, USA E-mail: [email protected] 2 CNRS, Institut des Sciences de la Terre, Université de Grenoble 1381 rue de la Piscine, 38041 Grenoble cedex, France E-mail: [email protected] E LEMENTS , V OL . 9, PP. 107–113 studies provide fundamental data for understanding fracture and plastic flow, friction, and the role of dehydration reactions on fault rheology. Fracture and Plastic Flow Both brittle and crystal plastic de for mat ion pro cesses a re strongly controlled by the anisotropic properties of serpentine. Dislocation glide (a deformation process where strain is accommodated by stress-driven motion of strike-slip fault, crystal defects) is easy along the basal plane of serpentine minerals, leading to ductility of serpentinite at low temperatures. Nonetheless, experimental studies have produced apparently contradictory results regarding high-temperature flow behavior. Several studies indicate that serpentinites deform by macroscopically ductile processes at high pressure (Hilairet et al. 2007 and references therein). In contrast, Chernak and Hirth (2010) observed localized faulting under hightemperature/high-pressure conditions, although these authors document macroscopically ductile flow at strains less than ~0.1 to 0.2. These results can be reconciled by considering the von Mises strain compatibility criterion, which states that five independent slip systems are required to accommodate homogeneous flow of a polycrystalline material. Dislocation glide on the basal plane and kinking can accommodate a limited amount of deformation, but to achieve higher strains additional processes are required. Experimental studies on serpentinite (Chernak and Hirth 2010) indicate that microcracking initiates at moderate strains, leading to a semibrittle rheology for which strength remains pressure sensitive. Thus, while anisotropy in plasticity facilitates dislocation glide at low temperatures, this process is insufficient to promote fully crystal plastic deformation at high strain. Flow laws for antigorite (a high-pressure type of serpentine) are reported by Hilairet et al. (2007). Based on analysis of the stress exponent (n) and the ratio of strength to confi ning pressure [(σ1 – σ3 )/σ3 ], they concluded that samples at 4 GPa deformed by dislocation creep (where the motion of crystal defects is thermally controlled by diffusive processes) while samples at 1 GPa deformed by dislocation glide. These data represent results of pioneering high-pressure deformation experiments, though the strains in the experiments of Hilairet et al. are below those for which semibrittle processes were observed in the study by Chernak and Hirth (2010). The observation of a power-law relationship between stress and strain rate (i.e. ė ∝ σn with n ≈ 3) suggests that deformation is limited by dislocation 107 A PR IL 2013 climb (the diffusive process that generally controls dislocation creep), which requires rapid diffusion of all components of the crystal structure (i.e. Si, Mg, O, and H for serpentine). However, this seems unlikely at the temperature (down to 200 oC) of these experiments. Nonetheless, the high strain-rate sensitivity of serpentine deformation has important implications for geologic processes. A The tendency for strain localization during semibrittle flow (where deformation is accommodated by a combination of brittle and ductile processes) of serpentine has important implications for understanding the stability of fault slip when an earthquake rupture propagates into a serpentinized region. Constraining the conditions where strain localization occurs is also important for extrapolating lab data to natural conditions, owing to the trade-offs between effective viscosity and shear zone width (e.g. Wada et al. 2008). Escartín et al. (1997) documented a transition from localized to distributed deformation with an increase in confi ning pressure from 200 to ~400 MPa at room temperature. They inferred that the transition occurred when the bulk sample strength became less than the frictional strength. However, with increasing temperature at constant pressure, a transition back to localized semibrittle deformation is observed (Chernak and Hirth 2010). The transition back to localized semibrittle behavior at high temperature suggests that the temperature dependence of the friction coefficient is greater than the temperature dependence of the bulk sample strength. While the details of semibrittle flow remain unresolved, the observation of a deformation-induced alignment of minerals [i.e. lattice-preferred orientation (LPO)] in both experimental and natural samples of serpentinite (e.g. Katayama et al. 2009; Bezacier et al. 2010) has important implications for interpretating seismic anisotropy (FIG. 1). Serpentine has a high compressibility normal to the basal plane (parallel to the c-axis), which leads to a low isotropic P-wave velocity (Vp) (e.g. Bezacier et al. 2010). Similarly, the shear modulus (ratio of shear stress to shear strain) parallel to the basal plane is very low. Thus, where a strong LPO arises owing to deformation, the elastic anisotropy leads to strong seismic anisotropy (Katayama et al. 2009). Nonetheless, the limited knowledge of elastic properties makes interpretation of seismic data nonunique. For example, minor amounts of lizardite/chrysotile and/or LPO can strongly impact the Vp /Vs ratio and the magnitude of shear wave splitting (e.g. Bezacier et al. 2010). In addition, the magnitude of shear wave splitting depends on both the strength of LPO and the width of the anisotropic region that waves pass through. Thus, if strain is highly localized under geologic conditions, shear wave splitting will be negligible. Friction Examining the frictional behavior of fault materials over a range of sliding velocities is essential for understanding the dynamics of stress evolution and slip during earthquakes. Phyllosilicate minerals are often inferred to control the behavior of creeping sections of faults (e.g. Moore and Rymer 2007). Friction experiments on serpentinite document velocity-strengthening behavior (which promotes aseismic fault slip) at tectonic displacement rates (e.g. Reinen et al. 1994), consistent with this hypothesis. Experiments also indicate a decrease in the friction coefficient (µ) of antigorite with increasing temperature. Reinen et al. (1994) measured µ = 0.5–0.85 for antigorite at room temperature; Moore et al. (1996) determined µ in the range of 0.4–0.6 for antigorite gouge at temperatures of 25–194 °C. Chernak and Hirth (2010) determined values of µ from 0.1 to 0.35 at temperatures of 400–550 °C. In E LEMENTS B C (A) Photomicrograph (crossed polarizers) of an antigorite sample deformed in the lab at 300 oC and 1 GPa. (B) Contoured, lower-hemisphere stereographic projections showing the lattice-preferred orientation (LPO) of antigorite with a [001] maximum oriented perpendicular to the shear plane (the shear plane pole is horizontal in the pole figure) and a maximum of poles to (100) oriented parallel to the slip direction. (C) Anisotropic velocity imparted by this LPO: (LEFT) contours of P-wave velocity; ( MIDDLE) contours of % anisotropy of S-wave velocity; ( RIGHT) polarization direction of shear waves. After Katayama et al. (2009) FIGURE 1 general, theories for rock friction suggest that µ should be relatively insensitive to temperature (e.g. Scholz 1990). In adhesion theory, µ is related to the ratio of the shear strength of asperity contacts to the indentation hardness of the asperities (FIG. 2). A lack of temperature dependence for µ is inferred based on the notion that the processes that control the shear strength of asperities are the same as those that control indentation hardness (e.g. Scholz 1990). However, when considering a strongly anisotropic material, another possibility arises. Assume that the basal planes of phyllosilicate grains at asperity contacts are dominantly aligned parallel to the fault surface. In this scenario, the shear strength (e.g. controlled by dislocation glide) may be significantly lower than the indentation hardness (e.g. controlled by semibrittle flow). The temperature dependence of basal slip would be expected to be greater than that for indentation hardness, resulting in temperature dependence and a lower value for µ. High-velocity friction experiments on serpentinite document dramatic dynamic weakening and dehydration at slip velocities above ~0.1 m/s (see Kohli et al. 2011 and 108 A PR IL 2013 A B C Schematics of a fault surface (shown perpendicular to the fault plane) in serpentinite at different scales. (A) The fault surface is rough, making contact only at asperities. (B) The basal planes of serpentine grains (dashed lines) are aligned parallel with the fault surface at asperities. (C) The shear deformation of asperities is accommodated by dislocation glide (as represented by symbols parallel to dashed lines). Grains are much stronger for deformation normal to the fault zone, resulting in a low, temperature-dependent coefficient of friction. FIGURE 2 references therein). These results indicate that a material exhibiting velocity-strengthening behavior at tectonic displacement velocities (~10 -9 m/s) may become dynamically unstable at near-seismic slip velocities after only a few millimeters of displacement (FIG. 3). Such behavior may be particularly important for understanding processes that control the dynamics of great subduction earthquakes, which likely propagate across patches that creep during interseismic periods. When the asperity-contact lifetime during frictional slip is short compared to the time it takes to diffuse heat away from asperities, the resulting “flash heating” can produce dramatic weakening. Kohli et al.’s experiments indicate that weakening occurs at a critical weakening velocity (Vw), at which dehydration of serpentine to talc occurs at asperities (500–700 o C). Based on extrapolation of Kohli et al.’s data, Vw is about 25 mm/s at an ambient T = 300 °C. This analysis suggests a possible explanation for the spectrum of fault-slip behaviors observed along fracture zones and subduction zones, where the alteration of mantle peridotite produces serpentinites. The effectiveness of dynamic weakening depends strongly on the width of the actively deforming zone. Thus, an understanding of how deformation localizes in serpentinite during interseismic deformation remains critical for constraining the significance of dynamic weakening. (Jung et al. 2004) suggest that antigorite dehydration can at least generate microseismicity. However, Chernak and Hirth (2010) observed macroscopically ductile flow and velocity-strengthening behavior in dehydrated antigorite at 1.5 GPa and 700 °C, challenging the hypothesis that intermediate-depth earthquakes result directly from dehydration embrittlement. Motivated by these results, Chernak and Hirth (2011) conducted tests in which temperature was ramped across the dehydration boundary, reasoning that earthquake-like instabilities would be most likely if samples dehydrated at high load under conditions where strain was already localized. No frictional instabilities were observed; rather, stress relaxed stably over several minutes, providing experimental documentation of stable fault slip during dehydration reactions. Similar interpretations were made based on the lack of detectable acoustic emissions during dehydration experiments on antigorite in a multianvil device (Gasc et al. 2011). These observations suggest that a mechanism other than dehydration may be responsible for intermediate-depth earthquakes (see discussion and references in Chernak and Hirth 2011). One possibility is that the migration of fluids produced via dehydration promotes seismogenic failure in the surrounding unaltered rocks. Alternatively, earthquakes may nucleate via thermally induced viscous flow instabilities. In this case, the fi ne-grained and polycrystalline products of dehydration reactions may promote the initial strain localization required for the thermally induced viscous instability. GEOLOGICAL OBSERVATIONS Since their discovery on the seafloor of the central Atlantic in the 1970s, serpentinites have been recognized worldwide in different active tectonic settings. Due to their unique physical properties, described above, serpentinites play a major role in strain localization. The Ocean–Continent Transition and Slow-Spreading Ridges Continental rifting and subsequent ocean-ridge spreading involve extensional faulting, exhumation of subcontinental mantle, and magmatism, processes that reflect the tendency for localized deformation in the lithosphere. Dehydration Embrittlement Dehydration embrittlement (where fluids resulting from dehydration reactions induce brittle deformation) has become the accepted mechanism invoked to explain intermediate-depth seismicity (e.g. Hacker et al. 2003). Several issues are important for the application of these results to higher pressures. First, the canonical experiments were conducted at confi ning pressures where the Clapeyron slope of the dehydration reaction is positive (where dehydration results in a volume increase, promoting microcracking) and the thermal stability of serpentine is relatively modest (inhibiting plastic flow processes). Second, the experiments did not show evidence for stick– slip behavior (or velocity weakening), which is required to nucleate earthquakes. Recent studies have extended these results to higher pressure. Acoustic emissions during deformation experiments at P = 1–6 GPa and T = 550–820 °C E LEMENTS Influence of sliding velocity on the frictional behavior of antigorite (after Kohli et al. 2011). At tectonic slip rates, the behavior is velocity strengthening, resulting in creep. However at seismic slip rates, samples exhibit dramatic dynamic weakening, which could facilitate earthquake rupture through aseismically creeping patches. 109 FIGURE 3 A PR IL 2013 The denudation of serpentinized mantle along the ocean– continent transition (OCT) and by seafloor spreading elicits the following questions: when and how is lithosphere serpentinized and what role does serpentinization play during rifting processes and ongoing slow-spreading-ridge activity? Along the western Iberia margin, the OCT forms a 100 km wide basin dominated by completely (95–100%) to moderately (50%) serpentinized peridotite (e.g. Afilhado et al. 2008). Gravimetric and seismic profiles indicate that the serpentinized mantle is about 5 km thick. Locally, tectonic breccia and semibrittle gouge composed of gabbro, amphibolite, and serpentinite clasts in a matrix of calcite- and chlorite-rich cataclasite are recovered during drilling. The geochemical signatures of primary minerals indicate that the serpentinized mantle is of depleted, subcontinental type. The oxygen isotope signature and the absence of antigorite in these rocks suggest temperatures less than 300 °C during fluid infi ltration (Skelton and Valley 2000). Later infi ltration of cooler seawater (<100 °C) along brittle normal faults accompanied the seafloor exhumation of the mantle. Mohn et al. (2012) proposed the following scenario of mantle exhumation: Crustal thinning is fi rst accommodated by pure shear along conjugate, crustal-scale shear zones and by ductile flow of the middle crust (FIG. 4A). The upper mantle is progressively thinned and replaced by deeper, hotter mantle impregnated by mafic magmas (FIG. 4B). Faults cut from the surface into the mantle, and seawater infi ltrates down to the upper mantle, initiating moderate-temperature serpentinization in the lizardite stability field (FIG. 4 C ). A Serpentinization weakens the upper mantle layer, promoting strain localization and allowing the normal faults in the distal margin to root at low angle. The ongoing extension leads to buoyant flow of partially serpentinized mantle along low-angle detachment faults, resulting in the horizontal stretching of the subcontinental mantle at the seafloor (FIG. 4D). B At slow- to ultraslow-spreading ridges complexes (OCCs) (FIG. 5). The domes exhibit corrugated surfaces that are interpreted to be caused by exposed detachment faults. OCCs are usually younger than 2 Ma and, when active, show hydrothermal activity and moderate seismicity down to 3–4 km depth along the main detachment fault. OCCs cover areas of ~200 to 500 km 2, extending ~25 km along the ridge axis and ~15 km perpendicular to the ridge axis. The corrugated surfaces include serpentinized peridotite and altered gabbro; some basalt is observed close to the axial ridge (FIG. 5). Seismic studies show relatively high velocities (Vp > 7 km/s) within a 2 to 3 km thick zone parallel to the surface, suggesting that this zone is dominantly composed of a mixture of gabbro and partially serpentinized peridotite (Karson et al. 2006). Deformation is localized in a ~100 to 200 m thick shear zone at the footwall of the detachment zone, and the rest of the massif is mostly undeformed (Karson et al. 2006). Far from the shear zone, serpentinites are massive and sparsely crosscut by joints. Partially serpentinized peridotites retain porphyroclastic textures acquired during high-temperature mantle deformation. Mesh-textured serpentinites show variable densities of crosscutting serpentine veins. Within the detachment faults, the mineralogy is more complex; the presence of talc, amphibole, and chlorite indicates the addition of silica, calcium, and aluminum from hydrothermal fluids derived through seawater– gabbro interactions. Serpentinization along the detachment faults involves at least several stages of alteration and veining (e.g. Andreani et al. 2007). Early serpentine (<40 mm/y), 20 to 25% of the seafloor is composed of serpentinite (Cannat et al. 2009). These areas of outcropping serpentinite are associated with domes, up to 500 meters high, known as oceanic core C FIGURE 4 Dynamic evolution of a rifted continental margin up to the formation of the ocean– continent transition (modified after Mohn et al. 2012). UCC = upper continental crust; MCC = middle continental crust; LCC = lower continental crust; OCT = ocean–continent transition; LT = low-temperature; HT = hightemperature; dark green = lower oceanic crust; blue to green = partially serpentinized upper mantle; yellow = marine sediments D E LEMENTS 110 A PR IL 2013 mation in a closed chemical system. The last generation of veins show a granular texture formed by micron-scale lamellae of Al-rich lizardite. These compositions and microtextures require a change in ambient conditions, with significant water circulation at shallow depths (<2 km) and temperatures of <200 °C. Strike-Slip Faults (B OTTOM) Schematic 3-D block diagram of an oceanic core complex (OCC) (after Cannat et al. 2009). (UPPER LEFT) Corrugated surface of an OCC (area shown is 10 × 5 km). (U PPER RIGHT) Photomicrograph (crossed polarizers) of a weakly serpentinized sample of troctolite from IODP Hole U1309D (Bar scale: 200 µm). PHOTO IODP, EXPEDITION 305 FIGURE 5 veins are irregular and thin (<0.5 mm) and exploit preexisting cracks in olivine. The local presence of antigorite and Fe-rich brucite suggests that early serpentinization occurred at about 300–400 °C at a low water/rock ratio. Second-generation veins, composed dominantly of lizardite or chrysotile and magnetite, are usually interconnected and a few centimeters in length, and they form the primary network of serpentine veins. These veins are often oriented parallel to the detachment fault, suggesting that they opened during tectonic unroofi ng. The third generation of veins contain oblique synkinematic fibers that record a shear component; the fibers are typically chrysotile and polygonal serpentine and document incremental defor- A Large-scale strike-slip faults can either slip by aseismic creep or move abruptly during earthquakes. Weak fault behavior can be caused by high fluid pressures, locally high geothermal gradients, or the presence of weak materials such as serpentinite. The San Andreas Fault (SAF), one of the most active strike-slip faults on Earth, extends at least 15 km into the crust and is marked at the surface by a complex network of crushed and fractured rocks a few hundred meters to several kilometers wide (FIG. 6). Just north of the Parkfield segment, the SAF creeps at a rate of 28 mm/y. Serpentinized ultramafic rocks derived from the Franciscan Complex mélange have been associated with fault creep, as well as low fault strength, in this region (e.g. Moore and Rymer 2007). The SAFOD project (San Andreas Fault Observatory at Depth) sampled and instrumented the SAF 9 km northwest of Parkfield. At about 3 km depth, the drill hole intersects the Great Valley Formation, a sedimentary unit rich in serpentinites associated with the Coast Range ophiolite (e.g. Moore and Rymer 2007). Microstructural observations from core samples show evidence of deformation across the damaged zone. The deforming zones contain serpentinite clasts and highly sheared siltstones. Moore and Rymer (2007) also reported talc in the cuttings. This discovery was noteworthy, as the frictional strength of talc at elevated temperatures is low enough to meet the constraints on the shear strength of the fault, and its stable sliding behavior is consistent with fault creep. (A) Distribution of serpentinite outcrops in central California and localization of major faults with creeping (aseismic) segments. SF, San Francisco; SB, Santa Barbara; SJ, San Jose; M, Monterey; P, Parkfield; C, Cholame; O, SAFOD; ID, Idria; SAF, San Andreas Fault; HF, Hayward Fault; CF, Calaveras Fault; SYF, Santa Ynez Fault; LPF, Little Pine Fault. (B) Geology of the SAFOD Borehole with location of cuttings (yellow circle) containing serpentinite-derived clay minerals. Red lines indicate active fault strands. (C) Scanning electron microscope image of gouge showing microshear zones in serpentinite-derived clay minerals (SAFOD Photography Atlas). Width of field of view = 2 mm B FIGURE 6 C E LEMENTS 111 A PR IL 2013 Recent investigations on the main 200 m thick damage zone at ca. 3 km depth at SAFOD show that the creeping area consists of clasts of serpentinite and sedimentary rock dispersed in a matrix of saponite (Mg-rich smectite clay), which is one of the weakest phyllosilicates. Saponite is a low-temperature (<150 °C) product of metasomatic reactions between the quartzofeldspathic wall rocks and serpentinite blocks in the fault (e.g. Lockner et al. 2011). Thus, serpentinite-derived clay minerals alone appear sufficient to explain the low fault strength and creeping behavior of the San Andreas Fault. However, considering the temperature stability field of saponite and the geothermal gradient within the drill hole, it is unlikely that this very weak mineral is present deeper than 4 km along the fault zone. Thus, stable creep and low strength in the SAF down to a minimum depth of 10 km must reflect the presence of other low-strength minerals (such as talc), elevated fluid pressure, or other deformation mechanisms (e.g. pressure solution) in serpentinites or quartzofeldspathic rocks. Subduction Zones Decoupling between the mantle wedge and the subducting slab is inferred from modeling of heat flow data down to 75 ± 15 km (Wada et al. 2008). The decoupled zone corresponds to the stability of phyllosilicates such as serpentine, whose weakness and/or ductility can explain a rheological weakening of the interface (Hilairet et al. 2007; Wada et al. 2008). Hydrous silicates (e.g. serpentine, amphibole, talc, chlorite, lawsonite) destabilize at increasing pressure and temperature in subduction zones and release water at different depths (Hacker et al. 2003). With ~13 wt% water in its structure, serpentine is an efficient carrier and source of water at depth during subduction (see Evans et al. 2013 this issue). Dehydration of serpentine has also been hypothesized to play an important role in the origin of double seismic zones, observed in some subduction zones, and in the origin of intermediate and deep seismicity along the slab (e.g. Yamasaki and Seno 2003). However, as noted above, the exact mechanism by which dehydration promotes seismicity remains a matter of debate. In addition, high-resolution seismic tomography indicates that seismic velocities in the lower seismicity plane are better explained by seismic anisotropy of anhydrous peridotite than by the presence of serpentinite (Reynard et al. 2010). Indirect petrological evidence for serpentinization at subduction zones is provided by the intimate association of serpentinites with eclogites (high- and ultrahigh-pressure rocks). Guillot et al. (2009) estimated that almost 30% of Phanerozoic eclogitic massifs worldwide are associated with serpentinites. Evidence for prograde metamorphism of serpentinite includes the transformation of lizardite to antigorite (FIG. 7) and the presence of metamorphic olivine and enstatite associated with Ti clinohumite, which reflects partial dehydration above 450 °C. The rheological properties of serpentine may help explain the mechanical evolution of subduction zones. Single-sided subduction appears to require a weak hydrated slab interface and high slab strength (Gerya et al. 2008). Tectonic settings that favor serpentinization of strong mantle rocks near the surface (such as the OCT and fracture zones) are thus good candidates for the location of subduction initiation. Serpentinites can lubricate the nascent interplate fault zone, facilitating asymmetric plate movement. Similarly, during exhumation of high-pressure or ultrahigh-pressure rocks, the opposite trajectories of exhumation and subduction require a decoupling zone within the subducting slab. A serpentinized layer such as the OCT prior to subduction may become a decoupling zone between the oceanic crust and underlying lithosphere. Moreover, the buoyancy of serpentinite likely contributes to eclogite exhumation because the average density of eclogite and serpentinite mélange is lower than that of anhydrous mantle peridotite. CONCLUSION Our understanding of the rheological properties of serpentinite has advanced significantly during the last ten years, and improved geophysical observations allow detection of serpentinite bodies in active geological A low-velocity zone above the high-velocity anomalies of a cold subducting slab has been imaged by seismic tomography below the Cascadia forearc and is interpreted as partially serpentinized mantle wedge (Bostock et al. 2002). The hydrated subduction interface, called the serpentinite channel, is more difficult to detect as it is probably thinner than a few kilometers. Receiver functions show a slow layer forming on top of the subducting plate below Japan at depths between 80 and 140 km, a structure interpreted as evidence for a serpentinite layer (Kawakatsu and Watada 2007). Bezacier et al. (2010) argued that serpentinite layers with kilometer-size folds could result in an orientationindependent anisotropy, with bulk anisotropy close to the mean value of 0.05–0.06 s km−1. In that case, the observed S-wave delay times (up to 1 s or more) would require a 20 km thick serpentinite layer. Serpentinite mud volcanoes containing clasts of serpentinized mantle peridotite in the Mariana forearc provide direct evidence of serpentinites in subduction zones (e.g. Fryer et al. 1999). The faults in the upper plate link mud volcanism to the main decollement fault and provide a route for buoyant serpentine-rich mud. Both the unconsolidated serpentine and the serpentinized peridotite clasts are dominated by chrysotile ± lizardite and antigorite that formed at 350–450 °C and 600–700 MPa. E LEMENTS Schematic diagram illustrating the geological context during subduction-related serpentinization (COURTESY OF F. D ESCHAMPS) and associated photomicrographs (COURTESY OF S. SCHWARTZ). With increasing depth, low-grade chryzotile-lizardite (Lz) transforms into high-grade antigorite (Atg). Below 50 km depth, antigorite breakdown releases fluids and metamorphic olivine (Ol) crystallizes. 112 FIGURE 7 A PR IL 2013 settings. In spite of emerging consistency among laboratory, geophysical, and geological observations, uncertainties remain concerning the rheological laws used to fit the data and the extrapolation of the laws to natural strain rates. Future work will focus on comparing experimental deformation mechanisms with natural examples and on investigating the processes that control the brittle–ductile transition and strain localization. These properties are key for understanding the nucleation and propagation of earthquakes and the significance of localized deformation REFERENCES Afi lhado A, Matias L, Shiobara H, Hirn A, Mendes-Victor L, Shimamura H (2008) From unthinned continent to ocean: The deep structure of the West Iberia passive continental margin at 38°N. Tectonophysics 458: 9-50 Andreani M, Mével C, Boullier A-M, Escartín J (2007) Dynamic control on serpentine crystallization in veins: Constraints on hydration processes in oceanic peridotites. Geochemistry Geophysics Geosystems 8: doi: 10.1029/2006GC001373, 24 pp Bezacier L, Reynard B, Bass JD, SanchezValle C, Van de Moortèle B (2010) Elasticity of antigorite, seismic detection of serpentinites, and anisotropy in subduction zones. Earth and Planetary Science Letters 289: 198-208 Bostock MG, Hyndman RD, Rondenay S, Peacock SM (2002) An inverted continental Moho and serpentinization of the forearc mantle. Nature 417: 536-538 Cannat M, Sauter D, Escartín J, Lavier L, Picazo L (2009) Oceanic corrugated surfaces and the strength of the axial lithosphere at slow spreading ridges. Earth and Planetary Science Letters 288: 174-183 Chernak LJ, Hirth G (2010) Deformation of antigorite serpentinite at high temperature and pressure. Earth and Planetary Science Letters 296: 23-33 Chernak LJ, Hirth G (2011) Syndeformational antigorite dehydration produces stable fault slip. Geology 39: 847-850 Escartín J, Hirth G, Evans B (1997) Nondilatant brittle deformation of serpentinites: Implications for Mohr-Coulomb theory and the strength of faults. Journal of Geophysical Research B 102: 2897-2913 Evans BW, Hattori K, Baronnet A (2013) Serpentinites: What, why, where? Elements 9: 99-106 Fryer P, Wheat CG, Mottl MJ (1999) Mariana blueschist mud volcanism: Implications for conditions within the subduction zone. Geology 27: 103-106 Gasc J, Schnubnel A, Brunet F, Guillon S, Mueller H-J, Lathe C (2011) Simultaneous acoustic emissions monitoring and synchrotron X-ray diffraction at high pressure and temperature: Calibration and application to serpentinite dehydration. Physics of the Earth and Planetary Interiors 189: 121-133 along crustal-scale strike-slip faults, in subduction zones, and in oceanic environments. ACKNOWLEDGMENTS We acknowledge funding from the National Science Foundation (USA) and Labex OSUG @2020Carbon. Comments and suggestions by the editors and M. Cannat, D. Cluzel, G. Mohn, and J. Nakajima were helpful in improving this article. Funiciello F (eds) Subduction Zone Dynamics. Springer-Verlag, Berlin, pp 175-204 Hacker BR, Abers GA, Peacock SM (2003) Subduction factory: 1. Theoretical mineralogy, densities, seismic waves speed, and H 2O contents. Journal of Geophysical Research B 108: doi:10.1029/2001JB001127 Hilairet N, Reynard B, Wang Y, Daniel I, Merkel S, Nishiyama N, Petitgirard S (2007) High-pressure creep of serpentine, interseismic deformation, and initiation of subduction. Science 318: 1910-1913 Jung H, Green HW II, Dobrzhinetskaya LF (2004) Intermediate-depth earthquake faulting by dehydration embrittlement with negative volume change. Nature 428: 545-549 Karson JA, Früh-Green GL, Kelley DS, Williams EA, Yoerger DR, Jakuba M (2006) Detachment shear zone of the Atlantis Massif core complex, Mid-Atlantic Ridge, 30°N. Geochemistry Geophysics Geosystems 7: doi: 10.1029/2005GC001109 Katayama I, Hirauchi K-I, Michibayashi K, Ando JI (2009) Trench-parallel anisotropy produced by serpentine deformation in the hydrated mantle wedge. Nature 461: 1114-1117 Kawakatsu H, Watada S (2007) Seismic evidence for deep-water transportation in the mantle. Science 316: 1468-1471 Kohli AH, Goldsby DL, Hirth G, Tullis T (2011) Flash weakening of serpentinite at near-seismic slip rates. Journal of Geophysical Research B 116: doi: 10.1029/2010JB007833 Lockner DA, Morrow C, Moore D, Hickman S (2011) Low strength of deep San Andreas fault gouge from SAFOD core. Nature 472: 82-85 Mohn G, Manatschal G, Beltrando M, Masini E, Kusznir N (2012) Necking of continental crust in magma-poor rifted margins: Evidence from the fossil Alpine Tethys margins. Tectonics 31: doi: 10.1029/ 2011TC002961 Moore DE, Rymer MJ (2007) Talc-bearing serpentinite and the creeping section of the San Andreas fault. Nature 448: 795-797 Moore DE, Lockner DA, Summers R, Shengli MA, Byerlee JD (1996) Strength of chrysotile-serpentinite gouge under hydrothermal conditions: Can it explain a weak San Andreas fault? Geology 24: 1041-1044 Raleigh CB, Paterson MS (1965) Experimental deformation of serpentinite and its tectonic implications. Journal of Geophysical Research 70: 3965-3985 Reinen LA, Weeks JD, Tullis TE (1994) The frictional behavior of lizardite and antigorite serpentinites: Experiments, constitutive models, and implications for natural faults. Pure and Applied Geophysics 143: 317-358 Reynard B, Nakajima J, Kawakatsu H (2010) Earthquakes and plastic deformation of anhydrous slab mantle in double Wadati-Benioff zones. Geophysical Research Letters 37: doi: 10.1029/2010GL045494 Scholz CH (1990) The Mechanics of Earthquakes and Faulting. Cambridge University Press, Cambridge, 471 pp Skelton ADL, Valley JW (2000) The relative timing of serpentinisation and mantle exhumation at the ocean–continent transition, Iberia: constraints from oxygen isotopes. Earth and Planetary Science Letters 178: 327-338 Wada I, Wang K, He J, Hyndman RD (2008) Weakening of the subduction interface and its effects on surface heat flow, slab dehydration, and mantle wedge serpentinization. Journal of Geophysical Research B 113: doi: 10.1029/2007JB005190 Yamasaki T, Seno T (2003) Double seismic zone and dehydration embrittlement of the subducting slab. Journal of Geophysical Research B 108, doi: 10.1029/2002JB001918 PFA LABWARE Essential for ultra trace analysis in geochemistry Gerya TV, Connolly JAD, Yuen DA (2008) Why is terrestrial subduction one-sided? Geology 36: 43-46 Guillot S, Hattori KH, Agard P, Schwartz S, Vidal O (2009) Exhumation processes in oceanic and continental subduction contexts: a review. In: Lallemand S, OUR EXPERIENCE … YOUR PROFIT! E LEMENTS 113 A PR IL 2013 www.ahf.de :: [email protected]