Subcrustal Lithosphere Research Articles

A special issue in memory of Richard St John Lambert: Preface

Richard St John Lambert, Professor at the University of Alberta and Chair of the Department of Geology there from 1970 to 1980, passed away December 6, 1994, at the age of 64, after a long and determined struggle against cancer. To us, he was a very good friend, a creative and thoughtprovoking colleague, and an outstanding raconteur with deep human insights. We are pleased and honoured to have been able to organize this set of papers in memory of Richard. Richard described himself as a chemical petrologist, with an acute interest in tectonics. The subjects addressed in the contributions in this special issue, and the selected bibliography presented below, display the breadth of interests that Richard enthusiastically pursued throughout his career. Richard was born in Trowbridge, England. He took a first-class honours degree and Ph.D. at Cambridge, completed in 1955. He then went to Leeds as an assistant lecturer. He accepted a post as university lecturer at Oxford in 1956, where he remained until 1970. In that year he accepted a chairmanship at the University of Alberta in Edmonton. His foresight, insistence upon the highest standards of scholarship and teaching, and emphasis on securing first-rate analytical facilities significantly reshaped the Department of Geology at Alberta. Richard published more than 100 refereed papers in scientific journals on a wide diversity of topics, including the geochronology of the global stratigrahic succession, the origin and age of the metamorphic complexes of the Scottish Highlands, the tectonic history and origin of igneous rocks in British Columbia, and the evolution of the subcrustal lithosphere. During the past decade, he became deeply involved in studies of Columbia River basalts and allied igneous rocks in the Pacific Northwest of the United States. We recall with great pleasure and a keen sense of loss, the excitement he brought to that research project, and the valuable insights he generated himself, or inspired in others. In 1991, Richard visited the Skaergaard Intrusion, Kangerdlugssauq

Subcrustal density inhomogeneities of Northern Eurasia as derived from the gravity data and isostatic models of the lithosphere

For the territory of Northern Eurasia (6°E–165°W; 30–75°N) the distribution of anomalous masses in the lithosphere has been estimated in accordance with the lithosphere isostatic model. The method of model construction is based on the admittance technique. The experimental admittance presents a relation between the part of the outer load uncompensated by the Moho undulations and the residual gravity field and is used to select the best model. The 1 × 1° averaged values of topography elevations, basement and Moho depths, sedimentary cover density and gravity anomalies have been used as initial data. According to the correlation equation relating the outer load and Moho depths, the mean density contrast between the lower crust and the subcrustal lithosphere is 0.43 g/cm 3, but the Moho undulation can not provide complete isostatic equilibrium. In some areas, the part of the outer load uncompensated by Moho undulations may be as large as 10 7 kg/m 2 and the residual gravity field is as intensive as + 260 mGal. Assuming that for loads of wavelength > 200 km, local isostatic compensation is valid, in accordance with the admittance analysis, the anomalous masses compensating for the part of the outer load, which is not compensated by Moho undulations, have to be located partly in the lower crust and in the subcrustal layer. The regional trend of anomalous compensating masses is negative under Western Europe, the Mediterranean, Eastern Asia and adjacent marginal seas, and positive under the East European Platform and Western and Central Asia. The local compensating masses correspond to particular tectonic units. The isostatic gravity anomalies of Northern Eurasia have been determined and the long-wave component of the field reflecting anomalous masses under the isostatic compensation level has been evaluated.

Orogen‐scale decollements

The continental lithosphere responds to stress by deforming as a generally layered medium. Deep seismic reflection data, coupled with a variety of ancillary geological and geophysical data, are interpreted to provide images of fault zones that tend to form moderately dipping ramp structures in mechanically rigid layers, and flat detachments in mechanically weak layers. This geometry is similar to ramp and flat structures observed at smaller scales in thrust and fold belts and leads to the interpretation that most orogens are underlain by orogen‐scale décollements in a manner that is analogous to so‐called “thin skin” deformation in sedimentary rocks. Décollements may occur within the crust (for example, near the base of a sedimentary section or in the middle crust), near the Moho, in the subcrustal lithosphere, or in the asthenosphere. Even intracratonic basement‐cored (“thick skin”) uplifts that occasionally occur in foreland regions such as the Wyoming province are probably large (crustal) scale versions of ramp/flat features observed in supracrustal rocks and are thus likely caused by the same fundamental tectonic processes.

Deep lithospheric structure across the Central African Shear Zone in Cameroon

SUMMARY Teleseismic P- and PKP-arrival times recorded by a network of 40 seismic stations deployed along a 300 km profile across the Adamawa Plateau at the northern end of the Volcanic Line in central Cameroon provide constraints on lithospheric thickness and anisotropy within the subcrustal lithosphere. These data indicate a thinned lithosphere beneath the Central African Shear Zone, where seismologically defined asthenosphere upwells from a depth of about 190 km to about 120 km in a relatively narrow belt. Thus it has only a low-amplitude effect on the observed gravity anomalies; the Bouguer gravity high over the Garoua Rift is consistent with crustal thinning beneath it. An abrupt change of the lithospheric thickness beneath the Northern Boundary Fault correlates with both the topographic relief and a distinct change of the orientation of relatively high- and low-velocity directions, which we infer to be due to anisotropy within the subcrustal lithosphere. This fault may represent an important accretionary suture zone dividing lithospheric blocks that originated in different tectonic settings and acquired different, frozen-in anisotropy.

Lithospheric thickness and velocity anisotropy—seismological and geothermal aspects

We compare inferences of regional variations in the lithospheric thickness in central Europe based on three types of observations. Seismologically determined thicknesses derived from representative P-wave residuals agree in most regions both with the magnetotelluric inferences of depths of the top of a layer of increased electrical conductivity and with the geothermally based calculations that allow inferences of the lithospheric thickness from surface heat-flow data. The seismological definition of the lithosphere as a high-velocity layer overlying the low-velocity zone in the upper mantle seems to be close to the geothermal concept in which the lithospheric base is defined by temperatures near partial melting, decreasing the seismic velocities and increasing the electrical conductivity. Interpretations of seismic data used for constraining geothermal models of the deep lithosphere assume velocity variations to reflect mainly changes in the mineral composition. However, recent observations of large-scale seismic anisotropy show clearly that, in the subcrustal lithosphere, velocity variations due to anisotropy caused by preferred orientations of olivine crystals can be much larger than the velocity variations caused by temperature or compositional changes.

Effects of rifting and subsidence on thermal evolution of sediments in Canada's east coast basins

Regional maps of lithospheric deformation and thermal history have been derived for the eastern continental margin of Canada. Subsidence associated with the rifting and cooling stages of rifted margin formation was calculated from gridded maps of sediment thickness and bathymetry along the Labrador, Grand Banks, and Nova Scotian margins. A two-layer lithospheric extension model was used to compute the deformation and thermal evolution of each region. Deformation results show that the crust and lower lithosphere have generally stretched by different amounts, and that either crustal or subcrustal lithospheric stretching dominates beneath the various basins. Thermal modelling results for the older Nova Scotian and Grand Banks margins show a strong correlation between thermal gradient, crustal stretching, and sediment thickness, and the predicted thermal gradient pattern for the younger Labrador margin correlates extremely well with predicted stretching of the still-cooling subcrustal lithosphere. Predictions of sediment maturity (vitrinite reflectance) of basin deposits were obtained from the derived time – temperature histories. Model results have been constrained with observations from individual boreholes and extrapolated away from these well-constrained areas into regions beyond the frontiers of present exploration. Results are presented as maps showing depths to present-day peak thermal maturity zones and the ages at which earliest post-rift sediments reached peak maturity levels. This reconnaissance approach has led to predictions of thermal maturity zones suitable for oil or gas generation in western Orphan Basin and beneath the continental slopes.

Teleseismic shear-wave splitting and deformations in Central Europe

We analyse the results of measurements of shear-wave splitting due to azimuthal anisotropy in the upper mantle of Central Europe, based on the SKS technique. The observations at seismograph stations in the eastern part of Germany are combined with the published estimates of the parameters of azimuthal anisotropy at neighbouring locations. These data are interpreted in terms of lattice preferred orientation of olivine which is caused by deformations in the upper mantle. Directions of polarization of the fast wave counted clockwise from north change from 50–70° in the western part of the region to 100–120° in the eastern part. These directions are close to the trend of the Hercynian fold belt and might suggest that anisotropy is frozen in the subcrustal lithosphere. However, the observed traveltime delays between the split waves are generally too large to be easily reconciled with the thermal and seismic estimates of thickness of the subcrustal lithosphere underneath Central Europe.

Multiple Enrichment Processes and their Relationships in the Subcrustal Lithosphere Beneath the Eifel (Germany)

Multiple Enrichment Processes and their Relationships in the Subcrustal Lithosphere Beneath the Eifel (Germany)

Plate tectonics, plate moving mechanisms and rifting

Abstract Analysis of a sequence of palaeo-reconstructions of the distribution of continents in the Western Hemisphere suggests that frictional forces exerted on the base of the lithosphere by the slowly convecting sub-lithospheric upper mantle play an important role as a driving mechanism of plate movements. With such a scenario, slab-pull and roll-back, ridge-push and deviatoric tensional stresses, related to upwellings of the asthenosphere as well as to lithospheric over-thickening in orogenic belts, can be considered as secondary, albeit important, plate moving forces. The present stress state of the globe and circumstantial geological evidence suggest that major compressional stresses can be transmitted over great distances through continental and oceanic lithosphere. Assembly of major continental masses (like Pangea) probably has an insulating effect on upper-mantle convection systems, causing their decay and reorganization. Development of new asthenospheric upwelling systems under mega-continents causes their break-up by development of deviatoric tensional stresses in the lithosphere and by exerting drag forces on its base. Extension of the lithosphere, culminating in its failure, is followed by passive advection of asthenospheric material into the space opening between the diverging plates to which it is accreted as oceanic lithosphere. At the same time developing ridge-push forces contribute to plate divergence. However, activity along seafloor spreading axes can terminate abruptly if far-field compressional stresses impede further divergence of the respective plates. This may explain the nearly contemporaneous decay of seafloor spreading axes in often distant areas during periods of plate boundary reorganization. Assuming a finite globe, the generation of new oceanic lithosphere at seafloor spreading axes has to be compensated for elsewhere by the subduction of commensurate amounts of oceanic lithosphere and/or shortening of continental crust and subduction of lower crustal material and the subcrustal lithosphere. Plate interaction, driven by mantle convection systems and their changes, ridge-push and slab-pull and roll-back, plays probably an all-important role in the development of intra-continental rift systems, the opening of new oceanic basins and the inception and development of subduction zones. Mantle plumes rising from the deep mantle do not appear to play a significant role in the development of intra-continental rifts though they probably contribute to a weakening of the lithosphere. The relative importance of the various processes contributing to the movement of lithospheric plates differs probably during the assembly and break up of Pangea-type mega-continents.

Global patterns of azimuthal anisotropy and deformations in the continental mantle

SUMMARY We present a summary of measurements of azimuthal anisotropy in the continental mantle based on the SKS technique and performed mostly with the active participation of the authors. The directions of polarization of the fast quasi-shear wave and the time delays between the quasi-shear waves are obtained at nearly 70 locations in all continents, except Antarctica. These data are interpreted in terms of lattice-preferred orientation of olivine which is caused by deformations in the mantle. The depth interval responsible for anisotropy is unknown but the data suggest that it may reach at least 300 km. The fast directions in SKS do not show clear correlation with the fast directions of the teleseismic P at the same seismograph stations. In the regions of present-day convergence the fast direction of anisotropy usually aligns with the plate boundary. This correlation implies that the direction of shortening is the same in the crust and the upper mantle. In the regions of rifting, the inferred direction of mantle flow usually aligns with the direction of extension in the crust. Outside the regions of recent tectonic activity we, most likely, observe a combined effect of frozen anisotropy in the subcrustal lithosphere and of recently formed anisotropy in the asthenosphere. On a global scale, in these regions there is a positive correlation between the absolute plate velocity directions and the fast directions of anisotropy. The correlation is especially strong in central and eastern parts of North America. A clear absence of any evidence of large-scale azimuthal anisotropy in the data of long-range refraction profiling for the upper 100 km of the mantle of that region implies that the effect in SKS is generated mainly at greater depths, in the asthenosphere. Orientation of olivine at these depths reflects recent and present-day flow in the mantle rather than processes of a distant geologic past.

Geodynamics of rifting and implications for hydrocarbon habitat

Abstract Rifting activity is probably mainly governed by frictional forces exerted on the base of the lithosphere by the convecting sub-lithospheric upper mantle, combined with deviatoric tensional stresses developing over upwelling branches of the asthenospheric convection system, and by far-filed stresses controlling the interaction of lithospheric plates moving relative to each other. Mantle plumes rising from the deep mantle do not appear to be a major driving force of rifting. Therefore, it is questioned whether a distinction between “active” and “passive” rifts is still justified. Lithospheric extension triggers by decompression partial melting of the lower lithosphere and upper asthenosphere. Magmatic processes contribute materially to thinning of the subcrustal lithosphere and probably also of the lower crust. This could explain the frequently observed discrepancy between upper crustal extension by faulting and lower crustal attenuation as defined by refraction and deep reflection seismic data. The duration of the rifting stage of palaeorifts and passive margins is highly variable. Depending on plate interaction, crustal separation can be achieved after as little as 7 Ma and as much as 280 Ma of rifting activity. The availability of crustal discontinuities does not appear to play a significant role in the time required for crustal separation to be achieved. The structural style of rifts is controlled by the physical state of the lithosphere, the mode and amount of its extension and the lithological composition of pre- and syn-rift sediments. Subsidence of rift basins is controlled by the balance between isostatic adjustment of the crust to lithospheric stretching and uplift induced by mechanical and thermal attenuation of the subcrustal lithosphere. Post-rift basin subsidence is mainly governed by the decay of rift-induced thermal anomalies; densification of the lower crust and intra-plate stresses provide overprinting effects. In view of this it is unlikely that quantitative analyses of post-rift basin subsidence give a direct measure of lithospheric extension. Tectonically active rifts, palaeorifts and passive margin basins contain major hydrocarbon provinces. Their hydrocarbon charge can rely exclusively on pre-rift, syn-rift or post-rift sedimentary sequences or a combination thereof. Maturation of source-rocks can be achieved during the syn- and/or post-rift stage of basin evolution. During rifting, conductive and convective heat transfer accounts for elevated geothermal gradients; these play an important role in the maturation of pre and syn-rift source-rocks; thermal gradients decrease asymptotically during the post-rift stage.

Crustal thinning during rifting: a possible signature in radiogenic heat production

The experimentally supported relationship between seismic velocity and radiogenic heat production was used to estimate the vertical distribution of heat sources in the lithosphere. Using these data we propose and compare geothermal models constrained by heat production (converted from seismic velocity) for ten selected extension zones in Africa, Asia, Europe and North America. Heat production in the crust sharply decreases with depth, but it can be shown that in the extension zones the vertical differentiation rate of heat sources increases in comparison with “normal” terranes. Strongly increased differentiation is also perceptible in the subcrustal lithosphere, where the rock material must have been enriched with radio-elements at the expense of deeper layers. Rifting thus represents a complex large-scale alternation of heat production in the lithosphere. The resulting distribution of radiogenic heat sources seems to be preserved during the entire life span of the rift and can be detected even in palaeorifts.

Dnieper-Donets palaeorift

The Devono-Carboniferous intra-cratonic Dnieper-Donets palaeorift transects the southwestern part of the East European platform in a NW-SE direction. It is characterized by linear bounding faults penetrating much of the crust, substantial crustal thinning and syn-rift volcanic activity. The Dnieper-Donets rift developed on cold Precambrian shield-type continental crust. Its rifting stage spans early Frasnian to Late Visean times. During its post-rift evolution an elongate, saucer-shaped basin subsided, the axis of which coincides with the trace of the underlying rift. The Palaeozoic Donets Rift is superimposed on a Late Proterozoic rift. A thermal perturbance of the asthenosphere is thought to have provided the driving mechanism for the development of the Palaeozoic Dnieper-Donets Rift. Rapid thinning of the subcrustal lithosphere in response to advecting asthenospheric melt was coupled with tensional fracturing of the crust and melting of its lower part. This caused an upward displacement of the Moho discontinuity. Post-rift subsidence of the Dnieper-Donets graben was governed by lithospheric cooling. The southeastern parts of this rift system were inverted during the Uralian orogeny.

The lithosphere in central Europe—seismological and petrological aspects

The lithosphere thickness in the Variscan belt of central Europe varies between about 60 and 150 km with typical values of 100–120 km. Our estimates, derived from directionally independent representative P-wave residuals, are in good agreement with magnetotelluric determinations of a layer with increased conductivity in the upper mantle. The large-scale anisotropies of the subcrustal lithosphere beneath four seismological stations determined from spatial variations of relative P residuals vary between 6.5 and 15.2% for P velocities; the S-wave anisotropies determined from SKS polarizations vary between 2.2 and 6.7%. These values are in reasonable agreement with the anisotropies of peridotites determined in laboratory. Systematic spatial variations of the directional terms of relative residuals in dependence on azimuths and incidence angles suggest the existence of large dipping anisotropic structures in the subcrustal lithosphere. The residual patterns at most stations in the Saxothuringicum, Rhenohercynicum and in the Massif Central imply northwesterly orientated dips of the anisotropic structures while stations in the Moldanubicum, the Alpine Foredeep and most of the Alps north of the Insubric line, suggest southeasterly orientated dips. In our interpretation the dipping anisotropic structures may represent paleosubductions which retain olivine preferred orientations originating from an ancient oceanic lithosphere. The Variscides of central Europe may thus represent a collision zone characterized by two systems of paleosubductions divergent relative to the suture between the Moldanubicum and the Saxothuringicum.

Seismic velocities beneath central and southern Britain from broad-band surface waves

Abstract Fundamental mode Rayleigh waves recorded across the UK broad-band network have been analysed to produce phase velocity curves in the period range of 12 to 85 s. An inter-station method has been used for two paths across southern England. The resulting mean phase velocity curve has been compared with a mean curve derived using three stations in central Britain. Lateral refraction characteristics of tripartite phase velocities have been analysed and shown to be less than 5° at periods of greater than 15 s. Forward modelling to S-wave velocity/depth profiles using a Hedgehog search procedure suggests that central Britain is underlain by a thicker crust than southern England, with a steady velocity increase down to the Moho. For southern England, low velocities are apparent either at the base of the crust or in the uppermost mantle. For both regions, sub-crustal lithosphere velocity is contrained to be between 4.6 and 4.7 km s −1 with a layer thickness of less than 140 km for southern England and between 80 and 160 km for central Britain. Both regions appear to exhibit low velocity zone (LVZ) development. However, for central Britain, more data are required to place better constraints on the magnitude of the LVZ channel. Velocities in the lower crust have been used to place constraints on Poisson's ratio which appears to decrease northwards from greater than 0.27, beneath southern England, to between 0.25 and 0.27 beneath central Britain. The distributions of P-wave velocities, S-wave velocities and Poisson's ratio have been interpreted of compositional variation and/or fluid content within the lower crust.

A heat production model of the crust and upper mantle

Experimental evidence on the relationship between seismic velocity and radiogenic heat production has been used to estimate the vertical distribution of heat sources in the continental lithosphere. The derived differentiation rate of the radioelements with depth has been investigated to depths ≤ 100 km in various world-wide tectonic provinces. The results show that more pronounced upward differentiation (steeper heat production distribution) corresponds to regions of higher heat flow and usually higher crustal temperatures. Thus, crustal distributions of radiogenic heat may be related to tectonic age and crustal evolution. The vertical distribution of heat production can be traced even in the subcrustal lithosphere and confirms the upward concentration of a substantial part of the radioelements into the crust at the expense of the upper mantle, reflecting the early stages of crustal separation. While in the uppermost crust the distribution of the heat sources seems to be controlled by uranium and thorium, in the upper mantle it is potassium which controls it. Data obtained may be important in the geothermal modelling, deep temperature calculation, and in determinations of Moho heat flow.

Low velocity layers in the subcrustal lithosphere beneath the Deccan Traps region of western India

Travel times and relative amplitude modeling of a seismic record section from the Koyna DSS profile in the Deccan Traps region has yielded a velocity model for the Indian continental subcrustal lithosphere. The observed record section, assembled with a reduction velocity of 6 km s −1 and using digitized data from analog deep seismic sounding (DSS) recorders, reveals a prominent wave group with an apparent velocity similar to but about 3 s after the reflected P phase from the Moho (PMP), at a reduced time of 7.5 to 6.5 s and recording distance range from 70 to 95 km. We interpret this strong amplitude phase as the reflection from an interface in the subcrustal lithosphere. Comparison of synthetic seismograms for a range of velocity models with the observed record section shows that the observed phase under investigation cannot be explained either by a reflected phase from a subcrustal velocity discontinuity with a single velocity increase or by the P-to-S converted phase (PMS) from the Moho. Subcrustal velocity models with a single low velocity layer (LVL) do not seem to give a satisfactory fit to the amplitude observations on the field record section. We prefer a subcrustal velocity model with alternating LVL (at least two with a velocity of 7.0–7.4 km s −1) separated by a thin high-velocity layer. In this model, the synthetic amplitudes of the reflected phase from the top of the deeper LVL at 56 km depth well match those of the observed reflection phase. It is inferred that the continental subcrustal lithosphere in this region of the Indian shield has a lamellar structure with significant vertical variation of structural and mechanical properties. The alternating LVL, occurring at relatively shallow depths below the Moho (velocity decreasing from 8.3 to 7.4 km s −1), may be associated with zones of weakness and lower viscosity. The LVL in the subcrustal lithosphere, as well as the LVL inferred in the upper and the lower crust (Krishna et al., 1989) together with the observed depth distribution of seismicity in this region, suggest a well-defined rheological stratification of the continental lithosphere with varying material properties.

Geochemical characterization of Pan-African dyke swarms in southern Sinai: from continental margin to intraplate magmatism

Abstract The chemical and mineralogical features of three Pan-African dyke swarms, occurring in the Wadi Feiran area, southern Sinai have been examined. Two swarms, striking approx. 35°, exhibit calcalkaline chemical properties. The oldest (S1) can be related to calcalkaline rocks from active continental margins by means of incompatible element plots displaying negative Nb-Ta anomalies. They dykes were emplaced synchronously with post-orogenic, alkaline granites to which they cannot be assigned chemically. A second dyke generation (L1) with the same strike shows apparent similarities with continental flood basalts (e.g. Columbia River basalts), indicating a growing contribution of continental components to melt chemistry. Subduction related features are displayed by negative Nb-Ta anomalies. The youngest dyke swarm (S2), striking approx. 135° and crosscutting S1, displays transitional to mildly alkaline properties without negative Nb-Ta anomalies. Its chemistry is fully consistent with intraplate basalts of continental areas. The source regions of the three swarms cannot be related via assimilation of crustal material because decreasing LILE percentages coupled with increasing amounts of Nb and Ta cannot be explained by this means. The overlap of M-values of rocks to be found with the three swarms indicates the unlikeliness of a development of the three magmas via fractional crystallization. The emplacement of two totally different types of dykes in the Wadi el Sheikh area requires a source region which is either chemically inhomogeneous in a more or less vertical scale or changing its composition and melting conditions with time. Although the dykes were emplaced in an extensional tectonic setting, the subcrustal lithosphere still produced calcalkaline melts, documenting its subduction-related magmatic history.

A kinematic model for the Late Cenozoic development of southern California crust and upper mantle

A model is presented for the young and ongoing kinematic deformation of southern California crust and upper mantle that accounts for both the known late Cenozoic geologic record and upper mantle structure. In this model, the high‐velocity upper mantle Transverse Ranges anomaly is created through the convergence and sinking of the entire thickness of subcrustal lithosphere. Using typical scaling relations, we estimate this anomaly to be ∼500°C colder and 1% more dense than average southern California mantle of the same depth. We infer that subcrustal lithosphere from both sides of the convergence zone participates nearly equally in the construction of the anomaly. The low‐velocity upper mantle anomaly beneath the Salton Trough region is attributed to high temperatures and 1–4% partial melt related to adiabatic decompression during mantle upwelling. Our model has crustal motions in the central Transverse Ranges region differing from mantle motions: While the subcrustal lithosphere south of the Transverse Ranges moves toward and sinks beneath the Transverse Ranges, the corresponding crust largely avoids convergence by adopting a more westerly direction of motion. The differing velocity fields necessitates of a decoupling zone immediately south of the Transverse Ranges near the Moho. We interpret the upper mantle anomalies as results of small‐scale convection beneath southern California, to which we attribute much of the recent tectonic activity in the region.

P-wave anisotropy in the lower lithosphere

Long-range controlled source seismic experiments yield detailed information about the velocity depth structure of the lower lithosphere. Between Ireland and Northern Britain three such profiles sample almost the same portion of the subcrustal lithosphere. Three zones with steep velocity gradients have been detected between 30 km and 90 km depth. Both the pattern and velocity of the arrivals are incompatible with isotropy. Preferential alignment of olivine crystals, with an azimuth ca. N25°E for the fast axis, could explain the observations, with the more highly aligned zones occurring in bands, or layers, separated vertically by zones in which the degree of alignment is slight or absent. We suggest that a shear heating mechanism may have played a part in producing these patterns. This deformation is most likely to be of Mesozoic or early Cenozoic age. It is argued that the upper mantle is not necessarily a strain marker for the last major orogenic episode, as recent findings have suggested [1], since it may undergo deformation which decouples from the brittle upper crust and hence is not “transmitted” to the Earth's surface.