In several convergent tectonic settings burial or exhumation appear to occur rapidly, at least on the order of 1 cm/ yr, ie. at plate tectonic velocities. Burial at rates of 0.1, 1 or 10 cm/ yr would bury 35 km MOHO to 70 km in 35, 3.5 or 0.35 Myr, respectively. Rapid thickening of crust is possible only if the deformed portion of lithosphere has significantly lower integrated strength than the colliding intending continent. This can result from an original hotter geotherm and/or from crustal lithology of low strength with respect to the surrounding lithosphere and is favoured if heat is continuously supplied at the base of the thickened root by syn-convergent magmatism. During the thickening process the integrated strength of the root continuously increases, forcing sub-root mantle downwards displacing soft upper mantle. This occurs when the velocity of downwards movement is significantly higher than heat conduction. At the end of the thickening process (ca. 8 Myr) the velocity of downward movement slows down and vanishes at about 70 km. when the root reaches strength equilibrium with the converging blocks, even if thermally unstable. Thermal relaxation of the root promotes decrease of the integrated strength/viscosity leading to the collapse of the orogenic root which becomes rapidly extruded over the shoulders of converging continents. The velocity of shoulder extrusion is controlled by the width of the weak zone and integrated rheology and may reach exceptionally high elevation rates of about 0.7 to 1 cm /yr. We argue that continental deep roots can develop only if the external heat supply exists. Thus the back-arc regions in front of subduction zones with enhanced calc-alkaline magmatism are the best candidates for formation of deep and hot continental roots. Such conditions are extremely favourable for rapid burial and rapid exhumation of crustal materials.
Orogeny in the Betic-Rif mountain belt is in two main stages. The first comprises the collisional stage of the Betic-Ligurian subduction system dipping W-ward under E-ward drifting Iberia. It may have taken place within the period c. 50-30 Ma; age constraints are weak. The second stage was initiated after slab break-off at c. 24 Ma and is characterized by fast uplift/cooling and concomitant extensional tectonics. In the central part of the Betic Cordilleras cooling rates of not less that 500°C/m.y. are demonstrated by comparing closure ages for a series of thermochronometers. For the whole of the Betic Cordilleras a period of 22-17 Ma is suggested for this extensional tectonic stage, but for discrete areas a much faster development within 1-2 m.y. seems indicated. Magmatism connected to the mountain belt is mainly post-collisional with extrusion/intrusion ages ranging from c. 20 to 5 Ma. The extensive calc-alkaline volcanic complex in Sierra de Gata shows eruption ages ranging from 12 to 10 Ma, but ion microprobe U-Th-Pb zircon dating shows that the pre-eruptional magmatic development started as far back as 17 Ma. Sr-isotopic relations and zircon age systematics for the well defined Caldear volcanic group indicate mingling and mixing of magma batches. Magmatism is thought to result from crustal anatexis under the influence of the HT-régime created by in-flow of HT-asthenosphere into the widening gap above the sinking lithospheric slab. This regional HT-régime may also provide a mechanism for the extensional tectonics: it might have caused thermal weakening in the crustal section facilitating tectonic extrusion under the influence of compression between rigid lithospheric plates according to a model outlined by Thompson et al. (1997).
Thompson AB, Schulman K & Jezek J, Geology, 25, 491-494, (1997).
At approximately 46° 30'S, along the Peru-Chile Trench, the Chile Rise, a mid-oceanic spreading centre, is being subducted beneath Andean-type continental lithosphere at the Chile Triple Junction (CTJ). The upper plate to subduction presently shows several topographic and kinematic effects of this 'ridge-trench collision'. These include the 1000 km trench parallel dextral strike-slip Liquiñe-Ofqui Fault (LOF), that extends northward of the CTJ, a large difference in the maximum elevation of the Andean divide to the north and south of the present day latitude of the CTJ, a region of low seismicity or 'seismic gap', and large areas of back-arc plateau lavas. This study, as part of a multi-disciplinary program in the region, applies fission track (FT) analysis to assess the thermal and denudational history of the overriding plate in relation to the late Tertiary evolution and northward migration of the CTJ, climate induced differential E-W erosion and denudation across the southern Andes, and the influence of the LOF.
At the time of writing, 21 zircon and 19 apatite FT ages have been obtained. The majority of the zircon ages from the Patagonian Batholith are Cretaceous to Jurassic in age, and likely reflect cooling to below ca. 230°C at shallow crustal levels shortly after intrusion. The largely unstudied metamorphic basement of this region, to the east of the batholith, yield zircon FT ages of 250-260 Ma, indicating metamorphism must have occurred before the late Permian. Several zircon ages close to the Peru-Chile trench on the western edge of the batholith give younger ages between 23 and 42 Ma. This may reflect a greater depth of erosion of the western side of the batholith. Along the present day main southern Andean divide the apatite fission track ages show ages ranging from 9 Ma in the south of the studied area, to ca. 5 Ma as one approaches the LOF in the north. The apatite FT ages increase to both the east and west of this divide to about 25-30 Ma. Initial interpretation of this data suggests that late Cenozoic denudation has been controlled by the N-S topographic expression of the Andes themselves. However, both the LOF and perhaps to a lesser extent ridge subduction (initiated at about ca. 6-8 Ma between 48° and 47°S) may also have been influential. By the time of presentation, more FT ages, including more from close to the LOF, will have been obtained. This will then allow the rates and timing of the denudational response of the southern Andes to the effects of ridge subduction and the arc-parallel LOF to be better assessed.
Partial melting during orogeny potentially causes mechanical weakening of the crust and leads to chemical differentiation of the continental crust. In order to investigate the physical impact of these processes on dynamic evolution of orogenic belts, and in particular the temporal relationship between melting and crustal deformation during plate convergence, we designed a thermomechanical numerical model using the finite element method simulating the behavior of partially molten material. In the model, two-dimensional vertical plane-strain deformation of the upper part of the system (the crust) is driven by the imposed velocity at its base (the subducting slab), and for each time step the mechanical properties for each element are derived from the calculated temperature distribution. The thermal evolution of the zone of thickened crust is mainly controlled by the competition between cooling due to advection of cold subducted material beneath the crust and internal production related to radioactive decay concentrated in the zone of thickened crust. Generation of partially molten material is represented in the model by modifying the viscosity and density of the original solid over a range of temperature corresponding to melting. Several models of viscosity and density evolution as a function of temperature (taken as a proxy for melt fraction) are tested. Estimation of the bulk viscosity and in particular the bulk density of the partially molten material requires a liquid-solid separation model. The assessment of the geodynamical pertinence of the model comprises an investigation of the effect of differential erosion, convergence velocity, melt properties, melt segregation, presence of weak horizons and variations of lateral boundary conditions. The generation of a low-viscosity and low-density pocket in the zone of thickened crust results in intra-crustal mechanical decoupling and in the development of gravitational instabilities.
Eclogitic rocks near Faro, Yukon-Tanana terrane, Canada, occur as three separate lenses embedded within glaucophane-bearing mica schists. The freshest eclogitic lens (lens III) contain garnet, omphacite, quartz and rutile with only local evidence for partial breakdown under blueschist-facies conditions. Garnet zoning and inclusion patterns (omphacite) support a prograde evolution to high temperatures and pressures (~690oC and > 1.5 GPa). In the altered domains, the presence of blueschist hydrous minerals and the absence of plagioclase + clinopyroxene symplectites developing after omphacite + quartz indicate that high-pressures prevailed during cooling down to at least 540oC. A remarkable feature of the garnet cores is the presence of minute garnet "inclusions" of a distinctly different composition and hence origin. Preservation of a sharp compositional gradient along the garnet host - garnet inclusion interface and the results of diffusion modeling in a closed system indicate that the prograde evolution and near-isobaric cooling event took place over a very short time-scale of less than 0.2 Ma. These results provide crucial constraints on the rates of thermal equilibration of accretionary wedges at the onset subduction.
Much of what we understand about the thermal and mechanical properties of the continental crust comes from integrated studies of orogenic zones. The Himalaya, being a spectacular example of collisional orogenesis, constitutes a natural laboratory for such studies. However, despite significant research over the last two decades, our understanding of the thermal behaviour of continental crust during burial is limited. With the advent of reliable and precise chronometers such as the garnet Sm-Nd system (that record prograde ages as well as thermal information) this situation is slowly changing (e.g. Vance and Harris, 1998; Vance and Mahar, 1998). In this contribution we compare thermal information recovered from the North-western Himalaya of India with that of Pakistan and draw important conclusions concerning the variation in the time scale of prograde metamorphism and anatexis along the orogenic arc.
New garnet-whole rock Sm-Nd isotopic data from northern Pakistan and the Nanga Parbat Haramosh Massif (NPHM), indicate that garnet-grade metamorphic conditions were reached in this part of the Himalaya around 44 Ma, at least 7 Ma earlier than further east in Zanskar or Garhwal Himalaya (Vance and Harris, 1998; Prince et al. 1997). Garnet growth intervals of < 3 Ma combined with textural information, recovered from the High Himalayan metapelites of the NPHM, also indicates that garnet- grade conditions were reached considerable faster here than in the central Himalaya. A new 38 Ma garnet-whole rock Sm-Nd age for a deformed garnet leucogranite recovered from Garhwal Himalaya indicates that here crustal melting occurred approximately synchronously with peak metamorphism. 50 Ma leucogranites from Northern Pakistan (Smith et al. 1994; Zeitler and Chamberlain, 1991) indicate that this is also the case further west, and hence this "prograde" crustal melting also shows a 7 - 10 Ma diachronicity. However, so far no such along arc age variation has been recognised in the age distribution of the main High Himalayan leucogranite suite.
These data therefore provide additional support for a diachronous collision initiating first in the north-west some time in the Late Palaeocene/Early Eocene. Following burial and thermal re-equilibration, peak metamorphism and wet melting followed a similar diachronicity, with peak conditions being reached earlier, and more rapidly, in the west. The absence of a similar along-arc variation in Miocene melting places further constraints on the mechanisms and timing of crustal anatexis in the Himalaya.
Prince CI, Vance D, Harris NBW, Terra Nova, 9, 486, (1997).
Smith HA, Chamberlain CP, Zeitler PK, Jnl. Geol, 102, 493, (1994).
Vance D, Harris NBW, accepted Geology, (1998).
Vance D, Mahar E, Mineral. Petrol, 132, 225, (1998).
Zeitler PK, Chamberlain CP, Tectonics, 10, 729, (1991).
Ultimately deformation and metamorphism are related to plate motion but plate velocities are often quite constant for long periods whereas deformation, metamorphism and magmatism at the plate margins are episodic. The whole concept of "phases" of deformation and metamorphism implies that the process is not continuous. The critical question is then which values in this range of possible rates is actually of interest. Clearly the long-term average is not. Neither are the rates of very short time-scale seismic events. It is the rates related to individual deformation phases or structures (e.g. a particular shear zone or fold) or to specific periods of mineral growth that we hope to determine.
Rates of deformation and metamorphism are not independent. With the possible exception of contact metamorphism due to magmatically advected heat, metamorphism involves movement of material through isotherms, which is only possible if there are velocity gradients and therefore equivalent strain rates. For near surface deformation, rates can sometimes be estimated from stratigraphic relationships, but at depth estimations are only possible using isotopic dating methods. The isotopic reeqilibration necessary to apply these methods reflects intracrystalline diffusion, cystallization or recrystallization, which are variously determined by temperature, strain and chemical environment (in particular, fluid access). It is not always clear which of these interrelated rates we are actually attempting to measure. Until now, estimates of rates have generally involved interpolation between discrete ages rather than ages determined along continuous growth profiles, the exceptions being core-rim garnet ages (e.g. Vance & O'Nions, 1992) and current work on fibre growth in pressure shadows (Müller, this symposium).
For velocity fields dominated by horizontal components (e.g. strike slip zones) the maximum average velocity (ignoring short term effects due to stored elastic strain) is on the order of the plate velocities, i.e. centimetres per year. Strain rates are then entirely determined by the length scale over which the velocity gradient is distributed. The vertical velocity component due to buoyancy forces can be dramatically different from plate velocities. In subduction/collision zones, buoyancy forces will act against subduction of material riding on the downgoing plate and return flow (i.e. exhumation) can produce relative velocities far exceeding plate velocities, especially relative to the lower plate. This is reflected in short time spans between high-pressure metamorphism and exhumation to the near surface. Measuring such "high" exhumation rates is limited by the resolution of the geochronological methods, the effects of strong heat advection, rapid near surface cooling, and topography, so that estimations at this upper end are only qualitative. The rheological response of rocks (e.g., evidence for high differential and mean stress, fold geometries reflecting an important elastic component etc.) can also be indicative for high strain rates, but the evidence again remains qualitative.
Vance D & O'Nions K, Earth Planet. Sci. Lett., 114, 113-129, (1992).
Granulite-facies metamorphism and partial melting in shallow crustal levels are described from several orogens. However, most thermal models for lithosphere thickened by collision predict high-temperature metamorphism and melting only in the lower crust. In order to understand how very high temperatures can be attained in thickened mid to upper continental crust, the P-T paths of high-temperature low-pressure (HT-LP) metamorphic rocks must be time-calibrated as accurately as possible.The Bayerische Wald within the Bohemian Massif is a shallow crustal level of an orogenic belt displaying HT-LP metamorphism and partial melting. Migmatites are characterised by a clockwise P-T path, prograde biotite dehydration melting reactions in the absence of an aqueous fluid, peak conditions of 850-900°C/0.5-0.7 GPa and subsequent decompression and cooling. Due to the high temperatures, only the cooling part of the P-T path can be dated. U-Pb dating on monazite, zircon and titanite and Ar-Ar laser dating on hornblende, biotite and K-feldspar were performed. Single grains were dated and all chronometers were used on one sample or outcrop, respectively, when possible. In order to detect regional variations in the cooling paths, various locations were chosen. The results suggest regional variations in the HT history and indicate an overall similar LT evolution. Single zircon tips grown in the presence of a melt in leucosomes of migmatites at 800-850°C display concordant U-Pb ages of 325.9±1.6 Ma (all errors 2s) in a western area and of 323.4±1.3 Ma in an eastern area. Single monazites from leucosomes and mesosomes of the same samples show the same U-Pb ages as the zircons, respectively (326.1±1.6 - 325.3±1.9 Ma and 321.8±0.7 - 323.3±1.0 Ma) and have excess 206Pb due to 230Th disequilibrium. These two observations indicate the monazite U-Pb ages to record crystallisation rather than cooling. In several cases mesosomes contain monazite grains or cores with apparent 206Pb/ 238U ages of 338-426 Ma. U-Pb ages of titanites are uniform at 321±2 Ma and coincide with the highest Ar-Ar ages of hornblende (315.6±3.5 Ma - 321.9±3.3 Ma). Ar-Ar ages of biotite are also uniform at 309.7±4.3 Ma - 313.3±3.2 Ma. K-feldspars that display plateaux for incremental heating experiments yield ages between 293.6±4.6 Ma and 305.1±3.6 Ma, 319.7±3.1 Ma in one case. The results show the HT-LP rocks of the Bayerische Wald to have cooled at comparatively slow rates. The high peak temperatures attained in these shallow crustal levels require an external heat source. The slow cooling rates suggest that the HT-LP rocks were not rapidly removed from their external heat source or vice versa after peak metamorphism.
The Rhodope zone, situated between the Dinarides-Hellenides to the SSW and the Balkan belt to the north, occupies most of northeastern Greece and southern Bulgaria. In central Rhodope of northern Greece, kyanite eclogites, concordantly intercalated in orthogneisses, record a multi-stage development of mineral assemblages formed during different stages of metamorphism within a single tectono-metamorphic cycle. The eclogite-facies assemblage was: garnet+omphacite+kyanite+zoisite+phengite+rutile. Minimum PT conditions of HP metamorphism were 19 kbar, 700°C (Liati and Seidel, 1996). Sapphirine, Fe-Mg-spinel, corundum, and högbomite formed during early exhumation (ca 16.5 kbar, 800°C). During this matamorphic stage, highest temperatures were reached and partial melting of the orthogneisses occurred. Amphibolite-facies overprint followed at 8-11 kbar, 580-690°C.
Magmatic and metamorphic zircon domains from different rock types were analysed by SHRIMP. Ion microprobe work was assisted by cathodoluminescence (CL) images of the zircon crystals: (a) hydrothermally grown zircon domains from a deformed prograde quartz vein in paragneisses (first formed probably at 250-300°C, 3 kbar) yielded an age of 45.3±0.9 Ma, which corresponds to the time of a low-T prograde stage of metamorphism. (b) zircons from the kyanite eclogites yielded an age of 42.2±1.0 Ma. Zircons from these rocks are homogeneous in CL and therefore completely reset during metamorphism, probably due to the presence of fluids derived by H2O liberating reactions, close to the P-peak. The age of 42.2 Ma corresponds therefore to a time close to the P-peak. (c) magmatic zircon cores from the orthogneiss country rock of the eclogites yielded Hercynian ages of 294±8 Ma (formation of the protolith) whereas the CL-bright metamorphic rims gave 42.0±1.1 Ma, almost identical to the age of the enclosed eclogites. (d) oscillatory zoned zircons from a leucosome in a migmatized orthogneiss, probably precipitated from the partial melt, yielded 40.0±1.0 Ma, which corresponds to the time that the rocks experienced highest temperatures. (e) zircons from a pegmatite (formed probably at ca 300°C, 3 kbar) crosscutting the schistosity of amphibolites yielded a crystallization age of 36.1±1.2 Ma. Thus, the whole tectono-metamorphic cycle above ca 300°C, 3 kbar lasted from 45.3 Ma to 36.1 Ma, i.e. ca 9 Ma. A correlation between the SHRIMP results and the P-T-t path constructed on petrological grounds reveals that the average heating and burial rates during subduction, above ca 300°C and 3 kbar, were >94°C/Ma and >15 mm/a, respectively. Regarding cooling and exhumation rates, also above 300°C and 3 kbar, these were >128°C/Ma and >7.7 mm/a, respectively. These data have important implications for the rheological properties of the lithosphere, as heat and mass transfer must have been considerably faster than generally thought so far.
Liati A & Seidel E, Contrib. Mineral. Petrol, 123, 293-307, (1996).
Rates of deformation within natural fault zones are only poorly constrained. Geodetic surveys across faults or displacement rates from sea floor spreading yielded strain rate estimates between 10-13 to 10-15 s-1 (Pfiffner & Ramsay, 1982). Mylonite zones, however, may be characterized by significantly higher strain rates (Schmid, 1975).
Strain fringes (or pressure shadows) around pyrite in carbonate mylonites have been used as key strain markers in order to reconstruct the incremental strain history of shear zones (e.g., Aerden, 1996). Being unique examples of synkinematically grown minerals, strain fringes can also be used for reliable dating of deformation provided they contain appropriate mineral assemblages. Dating of individual increments of strain fringes with the aim of calculating strain rates was attempted using recently developed microsampling techniques (Kelley et al., 1994; Müller, 1998).
Two examples of large strain fringes (5 - 10 mm), from the northern Pyrenees (Aerden, 1996) and the Eastern Alps (Fritz, 1988), have been chosen for dating of individual increments. They consist of thin, parallel fibres composed of carbonate, quartz, chlorite and/or white mica. The presence of the latter two minerals makes the application of Rb-Sr microsampling and laser-ablation 40Ar/39Ar techniques possible. Preliminary ages of individual increments have been obtained for one example from the Eastern Alps (Graz Paleozoic, Upper Austroalpine unit). These strain fringes record W-directed thrusting. Samples for Rb-Sr dating were directly prepared from thick section (~35 µm) under microscope control (microsampling). Two individual strain increments (1.4 and 1.8 mm long) were sampled from a 6 mm long strain fringe grown on one side of a 2 mm pyrite cube. Each increment consisted of parallel fibres composed both of carbonate (calcite) and quartz-white mica; the different minerals were sampled and analyzed independently. The carbonate provides an estimate of the 87Sr/86Sr-initial ratio, whereas the quartz-white mica fibres should represent the more radiogenic phase. Both the small samples (weights between 18 and 41 µg), the low Rb and Sr concentrations (4-15 ppm) and the moderate 87Rb/86Sr-ratios of the quartz-white mica fibres resulted in relatively imprecise ages. Ages of 32.6±2.6 Ma and 39.9± 3.8 Ma (95% c.l.) for the increment next to and away from the pyrite, respectively, are consistent with the antitaxial growth direction of the fibres towards the pyrite. These first and preliminary data are encouraging for the final calculation of strain rates of natural fault zones that will be presented at the meeting.
Aerden D, J Struct Geol, 18, 75-91, (1996).
Fritz H, Geodin Acta, 2, 53-62, (1988).
Kelley SP, Arnaud NO & Turner SP, Geochim Cosmochim Acta, 58, 3519-3525, (1994).
Müller W, Diss ETH No. 12580, 135 p, (1998).
Pfiffner OA & Ramsay JG, J Geophys Res, 87 B1, 311-321, (1982).
Schmid SM, Eclog Geol Helv, 68, 247-280, (1975).
The tectonostratigraphically lowest unit of Moldanubia is the Monotonous Series (MS). It is mainly composed of migmatic Crd-gneiss showing an HT/LP-overprint (720°C, >4.5 kbar) that distinctly differs from that of the other Moldanubian units. This pattern is spatially related with the Southern Bohemian Pluton. The Southern Bohemian Main Thrust separates the MS from the hanging Bunte Series (BS). The base of the latter comprises Proterozoic orthogneisses (Dobra, Stráz and Svètlík) that are overlain by paragneisses, orthogneisses, calcsilicate- and graphite-bearing schists, marbles and amphibolites. This lithology points to a shelf- and slope-derived vulcano-sedimentary series. The hanging Gföhl unit (GU) contains mainly amphibolites, the widespread Gföhl gneiss and, at highest position, felsic granulites. Gföhl gneiss and granulites are locally associated with mantle-derived rocks. The predominant granitic lithology of the GU points to an origin at an active continental margin. An unambiguous tectonic separation of BS and GU is hardly supported by field evidence. Moreover, the rocks in both units attained their common, last, HT-overprint (aH2O<<1.0, 700-800°C, 7-11 kbar, anatexis) during HT-, nearly isothermal decompression (HT-ITD). Therefore, the tectonic juxtaposition of BS and GU took place before this overprint. An earlier HP-overprint is suggested by textural, paragenetic and literature (>16 kbar) data. At completion of the HT-ITD, the already stacked, hot Moldanubian pile experienced isobaric cooling (IBC at c. 6 kbar, 700-500°C) that is related with the development of a transpression zone between Moldanubia and Moravosilesia and the thrusting of the former over the latter.
Recently, zircon data around 340 Ma were interpreted as the age of the HP-overprint within the Hercynian. Nevertheless, the available zircon data from Moldanubian granulites and gneisses vary considerably (370-333). Sm/Nd-data from clinopyroxenites spread at the upper and monazite data from felsic rocks at the lower limit of this range. The distribution of the data shows a pronounced frequency maximum around 340 (345-335). However, the lower half of this peak coincides notably with Ar/Ar-cooling ages from hornblende and muscovite (341-334). The latter are common for Moldanubian and Moravian rocks constraining, thus, the thrusting of Moldanubia over Moravosilesia. Similar coincidences of, mostly, zircon and Ar/Ar-data are observed more or less throughout the Central European Hercynian. Contrary to the above interpretation, the data distribution in Austria is clear evidence that the "magic number" 340 is the age of effective closure of most geochronology systems, rather than the age of HP- or even MP-metamorphism. This age correlates well with the onset of IBC that followed the HT-ITD of several kbar (Petrakakis, 1997). On the ground of the observed data distribution, ages around 370 mark the older, HP-overprint and ages of 360-350 the last HT-overprint.
Petrakakis K, J. Met. Geol, 15, 203-222, (1997).
Thirty fold hinge azimuths determined by examining subsurface contours of two stratigraphic levels on fifteen en echelon anticlinal structures in the Los Angeles and Ventura areas. At successively deeper levels, hinge azimuths of an en echelon fold systematically deviate to a new orientation and acquire smaller angles to the shear direction. The change in fold azimuths is interpreted as the consequence of the difference in duration of deformation and rock properties between the stratigraphic levels studied.
Strains were computed from the angle of rotation measured in fold azimuths. They were used for the construction of the strain rate-age diagrams. Within the Newport-Inglewood fault zone of the study area the strain rate is in the order of 0.9 to 2x10-14 per second for rocks from seven to twelve million years old. However, away from the fault zone the strain rate is approximately constant at 2.5-3.0 10-15 per second for rocks from six to fourteen million years old. These rates imply a 15 mm/yr displacement rate by distributed strain along the San Andreas for the late Neogene. This rate is about one-third of the current 50 mm/yr rate of displacement along the Pacific and the American plates. Displacement by distributed strain is an alternative explanation for the 'missing' plate motion that is manifested in insufficient slip rates obtained from strike-slip offsets along the fault.
Rb-Sr isotope studies have raised the possibility that silicic magmas are formed rapidly (<1% of the longevity of a magmatic centre) and that these magmas may reside in a chamber for long periods >100 kyr (ea. Halliday et al., 1989, Davies et al., 1994). The preservation of isochronous relationships in rocks erupted over protracted periods and mineral-glass age relationships further imply that the magmas remained within a chamber isolated from crustal interaction and magma mixing. Considerable debate has been generated concerning the possibility of maintaining thermal balance of magmatic systems over such timescales (Sparks et al., 1991; Mahood, 1991). We will critically review the evidence for rapid magma formation and extended storage times and particularly concentrate on new work that provides constraints from two chemically unrelated isotopic systems (U-Th disequilibria and Rb-Sr). We will argue that the coherence of these combined studies generally supports rapid magma formation and long storage times. Magmas stored for > 10 kyr are also shown to undergo crystal fractionation of minor phases that may disrupt the U-Th systematics of the magmas but does not effect the Rb-Sr system. Finally, available data will be reviewed to assess the importance of extended magma storage time in upper crustal chambers and the physical parameters that control the storage times (i.e., chamber size, tectonic setting, magma production rate, magma composition etc.).
Davies GR, Halliday AN, Mahood GA & Hall CM, E. P. S. L, 125, 17-37, (1994).
Halliday AN, Mahood GA, Holden P, Metz J, Dempster TJ & Davidson JP, E. P. S. L, 94, 274-290, (1989).
Mahood GA, E. P. S. L, 99, 395-399, (1991).
Sparks RJS, Huppert HE & Wilson CJN, E. P. S. L, 99, 387-389, (1991).
Magmatic reservoirs that evolve by open-system processes over prolonged time-scales are subject to the dynamic interplay of melt transport, fractional crystallization and magma storage. The chemically distinct Rb-Sr and U-Th disequilibrium isotope systems provide independent age constraints to address the timescales over which the physical processes operate.
The temporal evolution of post-caldera rhyolites from the Long Valley system is characterised by a constant change to more mantle-like Sr-Nd-Pb isotope signatures. Open magma chamber modelling places limitations on country rock assimilation and requires high rates of chemical fractionation compared to magma recharge. Sr isotopes record a distinction between high- and low-silica rhyolites which could represent magma batches isolated at different times from a constantly replenished open magma chamber that records a progressive decrease in 87Sr/86Sr. Young high-silica rhyolites (110 ka) define glass and internal Rb-Sr isochrons indicating fractionation of a feldspar-rich assemblage with initial 87Sr/86Sr of 0.70630 about 150 ± 40 kyr prior to eruption. This age coincides with a major chemical diversification of the system with contemporaneous low- and high-silica rhyolites and thermal rejuvenation, i.e. mafic magmatism. Major and minor phases record well define linear arrays in U-Th equiline diagrams with no true age significance (e.g. 235 ± 1 ka). These arrays represent mixing between minerals containing variable portions of minor phases formed at different times. 230Th/238U disequilibria of low- and high-silica rhyolites are controlled by fractionation of the minor phases zircon and allanite and indicate crystallisation of minor phases between 140 ± 1 and 254 ± 3 ka. Mineral crystallization occurred rapid and at discrete events (e.g. feldspars ± 40 kyr; zircons ± 1 kyr) possibly revealing repeated episodes of undercooling.
Despite the long-term open system evolution indicated by the Sr-Nd-Pb data, the combined Rb-Sr and U-disequilibrium age constraints confirm variable, but extensive residence times for rhyolites which in turn implies the existence of a stable, stratified upper portion of the magma chamber isolated from major convection in the reservoir for timescales of 150 kyr or more.
Experimental fluid dynamics and theoretical modelling widely contribute to constrain the processes operating in magma chambers. Nevertheless, combining these studies with geochemical data is still a challenge. In this respect, the U-decay series make it possible to link data and theoretical models, by providing direct information on magma processes and their timescales. This is a consequence of the short half-life of 230Th, 226Ra, and 231Pa isotopes (75200yr, 1622yr and 32400yr respectively) and of the ability of the U-decay chain to record recent fractionation events during mantle melting and/or magma crystallization. A closed- and an open-system models, both taking into account226Ra radioactive decay and crystal fractionation, are developed here for explaining the (226Ra/230Th) disequilibria for the November 1978 Ardoukoba eruption lava series, and permit to calculate precise magma crystallization times beneath the volcano.
Th, Sr and U isotope (measured by TIMS) indicate that the lavas are derived from a homogeneous mantle source. Constant (230Th/238U) and (235U/231Pa) disequilibria have been found in all the lavas and were interpreted as similar conditions of melting for all the magmas. Trace element trends show that the basalts belong to a single crystallization series. Sr, Sc, Ni, Ba, U, and Th versus La trends have thus been modelled with 30% crystallization (49%plag, 39%ol, 12%cpx). The bulk rocks exhibit a negative correlation between (226Ra/230Th) and Th contents which cannot be entirely explained by the fractional crystallization model (with DRa=DBa). As a consequence, the 226Ra-230Th disequilibria variations must result at least partially from 226Ra decay. Two types of magma chamber dynamical models were proposed: a closed-system model, represented by several reservoirs crystallizing independently, and an open-system model represented by a zoned magma chamber refilled by continuous input of fresh parental magma. Both models are based on mass balance calculation, and take into account continuous crystallization, 226Ra decay, and continuous magma input (for the open-system). The two models perfectly fit the data in the (226Ra/230Th) versus Th plot, but are characterized by two different time parameterizations. When reported in a (226Ra/230Th) versus time plot, the solutions for the two models allow to estimate relative crystallization ages directly from the 226Ra-230Th disequilibria measured in the Ardoukoba volcanic rocks. The open-system model systematically yields older ages for a given Ra-Th disequilibrium.
An age of 850 ± 40 yr is inferred from the closed-system model, and the open-system model yields an age of 1840 ± 90 yr for the most differentiated lava. In both cases, the calculated ages correspond to the age of the initiation of the magma crystallization. In the open-system, this age also represents the residence time of the first magma injected in the chamber.
Studies of U-series nuclides in arc lavas often reveal linear correlations on the (230Th/232Th)-(238U/232Th) isochron diagram. Such a correlation from historical lavas in the SVZ of south Chile was interpreted as an isochron reflecting the time elapsed between a U addition by fluids to the mantle source and magma eruption, being ~20 ky. These lavas have a uniform 87Sr/86Sr implying a relatively homogeneous magma source. Elsewhere, samples forming such correlation have variable Sr, Nd and Pb isotope ratios that may indicate that the linear variations are rather mixing lines than isochrons. Moreover, young arc lavas such as those from the SVZ have 226Ra-excess over 230Th that may result from the most recent fluid addition recorded by the lava compositions.
The amount of fluids from dehydration of a subducting oceanic lithosphere is likely to be related to the age of the slab, which may determine both its thermal structure and degree of hydration. This is supported by a fair correlation between the maximum observed (238U/230Th) for a given arc and the age of the subducting oceanic crust. In contrast, the minimum ( 238U/230Th) for each setting is relatively constant regardless of the ocean crust maturity.
Melts of subducted sediments with rutile and garnet in the residue should have (230Th/238U), (226Ra/230Th) and (231Pa/ 235U) higher than one, which upon mixing with a fluid enriched mantle melt (with U and Ra enriched relative to Th and Pa) would generate a mixing line on the isochron diagram. Excess 226Ra are indeed observed in lavas with (230Th/238U) lower, equal or higher than unity. The maximum transfer time for primary arc magmas would then be less than 8 ky.
Granitic intrusions present large aspect ratios, with an horizontal major axis about 4-8 times the vertical one. Geophysical studies reveal two main types of plutons. Flat-floored plutons (the most common) are rather thin (3-4 km) and extend in every horizontal direction with a gently dipping floor toward several root zones. These contrast with the thick (>10 km) wedge-shaped plutons, more elongated in one direction with steep walls, and a floor with no apparent deepening toward a feeder zone. We demonstrate that the two fold discrimination results of a switch in the stress pattern caused by magma emplacement and cooling. Modelling of cooling and crystallization shows that the rate to which rigidity (50% crystals) is achieved (100-200 mm/y) is faster than tectonic rates (10-50 mm/y) so magma can sustain and modify the ambiant stress pattern. Magma is preferentially emplaced into the plane (<sigma>1 <sigma>2) perpendicular to the least principal stress component (<sigma>3). This plane is initially vertical, except for compressional conditions. Magma intrusion induces dilation which causes a local re-organization of the stress field, by increasing the minor and intermediate principal stress components. When they overcome the major principal stress components (the lithostatic load in case of extension), a drastic change in the orientation of the opening plane results, switching from vertical to horizontal position. This constitutes a change from vertical dike-shaped intrusions to sub-horizontal laccoliths. Our model examines all deformation patterns ranging from radial tension to radial compression. In strongly extensional regions, the continual exchange of the intermediate stress components explains the rounded shape of intrusions. Specific transtensional or transpressional regimes are also considered. The analysis suggests that the concept of magmas rising to a level of neutral buoyancy is not applicable to many settings. Rather, we suggest that the feedback between magma intrusion and the local stress pattern controls the geometry of magma emplacement.
Transpressional structural zones are tabular ideally -viscous regions limited sideways by strong walls and beneath by a rigid plate. Transpressional tectonics occurs in zones of oblique convergence. We try to link both concepts and show consequences for elevation rates of rocks, strain accumulation and structural patterns in general. Our calculations show that for a plate motion velocity of 1 cm/yr and a transpressional belt wider than 100 km, the vertical strain rate does not exceed a value of 0.3 x 10-14 /s for any angle of convergence. Only very narrow belts show higher strain rate values but never exceed a value of 0.3 x 10-13 /s. However, numerous field studies have shown that in transpressional belts major discontinuities exist where higher displacement velocities are responsible for high strain accumulations and exceptionally rapid exhumation of rocks. Three types of strain partitioning can be defined: 1) directed ductile strain partitioning: 2) dispersed ductile strain partitioning: 3) viscosity partitioning. For 1), lateral displacement is accomodated by simple shear (SS) zones and belt perpendicular shortening accomodated by pure shear (PS) zones. Model for 1) is fully controlled by external boundary conditions and preexisting structures, and allow us to obtain the degree of horizontal stretching for any angle of convergence. Typical are regions of vertical fabrics of low strains limited by narrow simple shear zones with subhorizontal lineation and high finite strains. Originally deeper rocks will be observed only in PS domains. For 2), lateral SS zones with lesser PS and perpendicular PS zones with lesser SS, result in SS domains preserving horizontal stretching even when vertical motion has occurred. SS zones can form at almost frontally convergent zones, where they can accomodate the whole or part of lateral displacement and large amounts of finite strain, and also enjoy significant vertical elevation. Case 2) is most applicable to deformation of soft lower crustal rocks representing weak zones surrounding rigid cratonic basement blocks. For 3) the type of deformation is equal in both competent and incompetent zones. They differ only in strain rate and consequently in accumulated finite strain. Low viscosity domains will exhibit also higher elevation rates than rheologically stronger domains and deeper rocks will reach the erosion surface earlier in these zones. Model 3) is applicable to any convergent region involving lithospheric domains of different initial composition and temperature.
Earthquakes and fractures which occur in the upper crust attest of its brittle behaviour. However, observations of natural structures, such as folds and cleavage, which also develop throughout the entire upper crust, are evidence for a ductile and viscous behaviour. These two mechanisms appear at two time scales: fast for brittle deformation, slow for viscous processes.
Natural observations of relationships between pressure solution and fractures in sandstones, mica flakes indentations, or indented pebbles as well as recent experimental results of pressure solution with an indentor technique are presented. All these observations and experiments indicate that these two mechanisms can interact and a transition between them can occur (Gratier, 1993; Renard et al., 1997).
A simple model of brittle-ductile deformation is developed and applied to limestone indented pebbles. It is shown that cycles of slow deformation can alternate with short-time brittle episodes. This new model takes into account the complex mechanism of deformation by pressure solution where a crucial parameter is the grain contact structure that can trap a water film allowing for matter diffusion from the grain contact to the pore (Renard and Ortoleva, 1997).
It is suggested that geologic modelling of the upper crust (creep or compaction laws) must integrate these complex mechano-chemical interactions (Gratier et al., 1999). In particular, the viscosity of the upper crust can be modified if brittle deformation and pressure solution occur simultaneously.
Gratier JP, Geophysical Research Letters, 20, 1647-1650, (1993).
Renard F & Ortoleva P, Geochimica et Cosmochimica Acta, 61, 1963-1970, (1997).
Renard F, Ortoleva P & Gratier JP, Tectonophysics, 280, 257-266, (1997).
Gratier JP, Renard F & Labaume P, Journal of Structural Geology, submitted
As in many other migmatites, melt in migmatites in the Fletcher Creek area, Kimberleys, West Australia, initially segregated into thin mm-cm scale veins, now leucosomes. Small layer-parallel leucosomes are abundant and are found in the most fertile metasedimentary layers. Dm-scale and larger leucosomes are distinctly less abundant, also occur in less fertile local lithologies and are not restricted to layer-parallelism.
Melt bearing veins are mobile when a critical normal stress gradient along their length is exceeded. Mobile veins move by opening at one end and closure at the rear end. Tectonic stress gradients, which can reach values in the order of one MPa/m, can cause dm-scale melt veins to move over short distances, while buoyancy of the melt can cause rapid upward mobility of veins that are tens of metres in length. Merging of mobile melt veins leads to accumulation of melt. This step-wise mechanism can explain the accumulation of melt from initial leucosomes to batholith-scale melt volumes.
The consequences of this accumulation model were investigated with Monte Carlo computer simulations.
At a low deformation rate, accumulation is slow. Only at relatively high melt fractions do the low mobility melt veins merge and accumulate to form large enough melt batches that can ascend through the crust by their own buoyancy. The melt chemistry of these extracted melt volumes is close to the batch melting curve. At a high deformation rate, accumulation is fast. Large enough melt volumes to ascend the crust are already formed at low melt fractions. Early melt volumes have a batch-melting signature, but ongoing melting produces a transition towards a fractional melting signature in extracted melts.
Thus, at a high deformation rate, more granites can be produced and these exhibit a wider range in chemical composition than granites produced at a low deformation rate.
We use non-linear materials compressed parallel to layering to investigate the growth rate and stress versus strain behaviour of folding in single and multiple competent layers, embedded in a viscous matrix. In a first series of tests we study growth rates in single layer folds exhibiting strain hardening or softening, embedded in a weaker viscous matrix. The growth rate of strain softening folds is twice that for strain-hardening folds. Maximum amplification rate for semi-brittle folds is reached at 20% and for ductile folds at 30% bulk shortening. Strain-hardening single layer folds exhibit remarkable hinge-thickening and limb-thinning, at the early stages of progressive shortening. This stage is characterised by a gently sloping growth rate curve, which steepens as deformation proceeds. At final stages, growth rate diminishes leading to hinge fracturing at high limb dips. Folding in a strain-softening single layer initiates with a sinusoidal wave form coinciding with a gently increasing growth rate curve. With time fold shapes become chevron-like when faulting nucleates at the hinges. This process results in abrupt steepening of the growth rate curve.
In second series of test, we study the role of viscosity contrast and ratio of incompetent to competent layer thickness in controlling fold growth rate. For a high viscosity ratio of competent to incompetent layer and low to moderate ratio of incompetent to competent layer thickness, the structure develops regular chevron fold forms with a high growth rate at the initial stages of folding. With time, thickening of hinges and thinning of fold limbs takes place as the incompetent material migrates from the limbs to the hinges, resulting in reduced growth rate. Decrease in ratio of incompetent to competent layer thickness (for the same viscosity contrast) results in enhanced growth rate of the fold perturbations.
By studying growth rate of folding in non-linear material analogues it may be possible to study the growth rate of natural folds during their growth to finite amplitudes.
The Kola Peninsula, wich occupies the northeastern part of the Baltic Shield, is largely composed of Archaean and Proterozoic gneisses (Mitrofanov, 1995). Rifting during Hercynian time was associated with the intrusion of many alkaline rock complexes (Khibina, Kovdor, Lovozero, Kurga, Seblyavr, Vuoriyarvi, etc), lamprophyres (the coast and islands of the Kandalaksha Gulf) and kimberlite explosion bodies (Terskii Bereg). These alkaline complexes constitute part of the Kola Alkaline Province (Dudkin and Mitrofanov, 1993).
The isotope studies performed by Kramm (Kramm et al., 1993, 1994) provided the short time span of Paleozoic alkaline magmatism in the NE Fennoscandia. During a period of 360-380 Ma the largest agpaitic complexes of Khibina and Lovoozero were formed simultaneously with numerous carbonatite intrusions of Kovdor, Turiy Mys, Sokli, Afrikanda, etc.
The Seblyavr alkaline ultrabasic massif with carbonatites and the Kurga intrusion have been studied by the Rb-Sr method (Gogol et al., 1998). The Rb-Sr data (WR and minerals) for Seblyavr gave ages of 410±7 Ma (pyroxenite) and 408±7 Ma (carbonatite). A Rb-Sr isochron for Kurga pyroxenite, alkali syenite and mineral separates yielded an age of 404±8 Ma (Gogol et al., 1998).
The diamond-bearing kimberlite from the Terskii Bereg area, south Kola Peninsula, yielded a preliminary Rb-Sr age (WR and minerals) about 404 Ma, similar to the ages for the Seblyavr and Kurga massives.
We have also analysed rocks and extracted mineral fractions from the host ijolite of the Kukisvumchorr apatite-nepheline deposit (Khibina) and Vuoriyarvi massif. A Rb-Sr age of Kukisvumchorr ijolite (WR and minerals) is 371±6 Ma. A Rb-Sr age of Vuoriyarvi carbonatite is 375±7 Ma. These ages fall into the interval of 360-380 Ma (Kramm at al., 1993).
Thus, the ages of the Kurga and Seblyavr massives and the Terskii Bereg kimberlites expand the time span of alkaline magmatism in NE Fennoscandia. These massives and the kimberlite bodies of the Terskii Bereg area are related to the initial phase of Paleozoic tectonic and magmatic activation in the North Atlantic Realm during Caledonian orogeny.
Kramm U, Kogarko LN, Kononova VA, Vartainen H, Lithos, 30, 33-44, (1993).
Dudkin OV & Mitrofanov FP, Geochem. Int, 31, 1-11, (1993).
Kramm U & Kogarko LN, Lithos, 32, 225-242, (1994).
Mitrofanov FP (Editor), Geology of the Kola Peninsula (Baltic Shield), 144 pp, (1995).
Gogol OV, Bayanova TB, Balaganskaya EG & Delenitsin AA, Abstracts of ICOG-9, Beijing, 45 p, (1998).
The Crystal Size Distributions (CSD) in rocks reflect their crystallization kinetics. The CSD are in particular related to the thermal evolutions of the samples because (i) the nucleation and growth rates depend on the temperature and (ii) the cristallization influences the temperature by the release of latent heat. The eruption of the Makahopuhi lava lake in 1965 provided a unique opportunity of studying the thermal and petrological evolution of a magmatic system during the crystallization. Samples taken from drill holes at different depths and different dates are available. The CSD have already been optically studied by Cashman and Marsh (1988) and Marsh (1998) for one drill hole, and nucleation and growth rates have been deduced from their interpretation. The early studies considered the growth rate as constant with grain size (Kirkpatrick,1977; Cashman and Marsh, 1988), and did not discuss the CSD with respect to the dynamics of a lava lake. Marsh (1998) includes a variable growth rate in his model and makes the hypothesis that small crystals disappear by their annexation by larger ones. We used SEM images for the reconstruction of CSD of plagioclase in the Makaopuhi lava lake samples. The use of this technique permits to take into account the smallest grains, which were not considered previously. These CSD are interpreted in the light of the lava lake thermal evolution model of Davaille et Jaupart (1993). The drill holes sample different parts of the lake (i) at depth, a convective layer, with more than 50% liquid, (ii) a rigid layer where crystallization takes place in a closed system (iii) near the surface, an entirely solid layer. The lower layer is probably that where the characteristics of the CSD are acquired. In the intermediate layer, the CSD for samples taken at different depth are not parallel, which indicates that the growth rate depends on the grain size. In the upper layer, the differences between the CSD are believed to reflect the thermal evolution of the lake through time, the crystallization being much more rapid at the surface than at depth.
Cashman KV, Marsh BD, Contrib Mineral Petrol, 99, 292-305, (1988).
Davaille A, Jaupart C, Geophys Res Lett, 20, 1827-1830, (1993).
Kirkpatrick RJ, Klein L, Uhlmann DR, Hays JF, J Geophys Res, 84, 3671-3676, (1979Marsh BD).
Marsh BD, J Petrol, 39, 553-599
The Rungwe volcanic field is located between the Rukwa, Karonga and Usangu basins within the Western branch of the East-African rift system. Stratigraphical, geochronological data (Harkin, 1960; Ebinger et al, 1993) and our own observations, supported with laser fusion and stepwise-heating Ar-Ar dating, allow to constrain the Mio-Pliocene and Quaternary evolution of its volcanic activity.
Primitive (olivine and pyroxene melanephelinites, basanites, basalts), differentiated alkaline (phonolitic basanites, phonolites) and differentiated moderately alkaline (hawaiites, mugearites, benmoreites, trachytes) magmatic rocks were distinguished based on their petrography and major element compositions.
The wide variations of 208Pb/204Pb, 206Pb/204Pb (Furman, Graham, 1994), 87Sr/86Sr, Zr/Sm, Hf/Sm, Ce/Pb (Ivanov et al., 1998) in primitive rocks suggest a mixing of magmas, produced by a melting of three mantle-derived components. One of the components is thought to be an asthenospheric mantle (AM) component with radiogenic lead composition 206Pb/204Pb>18.9 and values of 87Sr/86Sr~0.7045, Zr/Sm~15-20, Hf/Sm~0.43, Ce/Pb~9-11. Another component, having higher 87Sr/86Sr~0.7056, lower 206Pb/204Pb and elevated 208Pb/204Pb ratios, represents an ancient enriched lithospheric mantle (ELM). The third component is recognised through its low Zr/Sm~1.3, Hf/Sm~0.07, high Ce/Pb~20-35, heterogeneous 87Sr/86Sr~0.7046-0.7052 and low 206Pb/204Pb, 208Pb/204Pb ratios. Such signatures had been attributed to a metasomatised lithospheric mantle (MLM) peridotites, significantly reworked by aqueous and/or phosphor-rich fluids with amphibole and/or apatite growth (Ivanov et al., 1998).
Primitive magmas of the AM component are characterised by a higher degree of partial melting (F~5%) and lower pressure origin (P~20-25 kbar). Primitive magmas of the ELM component originated under conditions of lower F~2-4% and higher P>30 kbar. Lowest values of F~1-2% and intermediate P~30 kbar have been calculated for partial melts of the MLM component. Alkaline differentiated melts fractionated from magmas with AM component signatures. Moderately alkaline differentiated melts evolved due to a combined process of fractional crystallisation and lower crust assimilation from the parental melts of any mantle-derived components.
Geological and our new geochronological data reveal a temporal shift of the AM component from the Rukwa towards Karonga and Usangu basins. The lower crust assimilation is noticed only at places were magmas of the AM component had erupted. The deepest ELM component was only melting in the Quaternary beneath the central part of the Rungwe province (volcanoes Rungwe, Ibuli hill etc.). While one can find geochemical signatures of the MLM component in any erupted primitive magmas. This suggests why extension and subsidence of the Rukwa basin is decreasing with time, while rifting of the Karonga and Usangu basins becomes more intensive.
Ebinger C J, Deino A L, Tesha A L, Becker T & Ring U, J. Geophys. Res., 98, 17821-17836, (1993).
Furman T & Graham D, Mineral. Mag., 58A, 297-298, (1994).
Harkin D, Mem. Geol. Surv. Tanganyika, 172, (1960).
Ivanov A V, Rasskazov S V, Boven A, Andre L, Maslovskaya M N & Temu E B, Petrology, 6, 208-229, (1998).
Tectonic history of the Indian oceanic basins is generally well established, however some areas remain unknown. It is particularly the case of the north Mascarene Basin, located south of the Seychelles Archipelago, east of Madagacar and west of the Mascarene Ridge.
A compilation of all bathymetric, seismic, magnetic and gravimetric data, including recently acquired data, allows to propose a new morphostructural map of whole the Mascarene Basin, which is used to study its evolution during the Cretaceous.
The Mascarene Basin is segmented by numerous fracture zones, which right-lateral offsets of 75 to 465 km, and about 75 magnetic anomalies are identified from 83 to 61 My. The fossil spreading center is well observed in bathymetric data in the southern and middle parts of the basin.
The opening of the Mascarene Basin took place in two periods. The first period (A34-A31) corresponds to a slow to intermediate spreading rate (55 km/My) in a ENE-WSW direction. The second one (A30-A27) is characterized by an ultra-fast spreading rate (150 km/My). Seafloor spreading ceased from north to south from 65 My to 59 My, which implies deformation along fracture zones in the east flanck of the basin with a total among of 650 km.
The relative timing of the Variscan and post-Variscan processes in the Ruhla Crystalline Complex (RCC) - situated in the NW part of borderzone between the Bohemian Massif and South German Block - are well known. However, the absolute timing of the late Variscan cooling and the timing and magnitude of Mesozoic inversion remain speculative. To help clarify these outstanding problems zircon and apatite fission track (FT) data have been obtained from various basement rock units of the RCC.
Previous geochronological data and the overlying sedimentary record provide clear evidence that the RCC consisted of three blocks that underwent different exhumation histories during Late Carboniferous / Early Permian times. The eastern and western blocks comprise previously dated metamorphic and granitic rocks overlain by molassic sedimentary and volcanic rocks of Late Carboniferous age (<300 Ma). The central block of the RCC (Ruhla Horst), in contrast, was affected by Late Carboniferous / Early Permian plutonism between 295 and 288 Ma, and is unconformably overlain by younger Late Permian (Zechstein) sedimentary rocks. This interpretation is confirmed by the results of zircon FT analysis. The ages from the central block overlap at around 272 Ma, indicating a moderate cooling rate after granite intrusion, before being exposed at the surface before ca. 256 Ma. The eastern and western blocks give zircon FT ages mainly around 300 Ma, and indicate a two stage cooling of this block after granite intrusion at ca. 335 Ma to exposure at the surface at around 300 Ma.
The apatite data show that the majority of the Mesozoic inversion of the RCC occurred during the Late Cretaceous 'sub-Hercynian' inversion phase of NW Europe, between 85-75 Ma. As all the apatite ages represent samples that lay below 110±10°C before ca. 90 Ma, then around 3-4 km of overburden must have been removed from above the present day exposure level of the RCC rocks since this time. The sedimentary overburden likely to have overlain the eastern and western blocks of the RCC are enough to account for the FT ages. However, in the central block of the RCC it is estimated that the sedimentary pile before 90 Ma was at most ca. 1900 m thick. This implies that a considerable section of Variscan rocks were also eroded during inversion, especially north of the Klinge Fault. Basement samples from close to the Late Permian unconformity are presently being analysed to test whether this is the case, or whether the sedimentary pile that overlay the central block before ca. 90 Ma was much thicker than is presently assumed.
A complex series of brecciated metamorphic, and magmatic rocks at Mount Jackson, eastern Palmer Land, Antarctica record a history of at least five phases of tectonic and volcanic fragmentation associated with movement on a large-scale strike-slip fault. Brecciated megacrystic granite of Lower Jurassic age and sheared aplogranite of unknown age host magmatic rocks emplaced and brecciated during multiple episodes of magmatism and deformation, forming a complex at least 1 km in width. The sequence of deformation and brecciation is as follows: 1) Megacrystic granite and aplogranite mylonitized by reverse shear show transition to granite breccia with cm- to dm-scale clasts. The field evidence seems to point clearly to a post-magmatic, tectonic origin for this. 2) Gabbro cross-cutting this breccia is dextrally mylonitized on a west-striking fabric that may have originated synmagmatically; gabbro was subsequently brecciated. Rafts of the granite breccia in this gabbro are folded, indicating a substantial time-gap between brecciation events. 3) One of two rhyolitic tuffisite phases cut brecciated gabbro. This tuffisite has a diverse xenolith load and appears partially melted. It shows no ductile fabric. 4) Basic dykes with extremely irregular contacts cut the gabbro and the early rhyolitic tuffisite. The basic dykes are brecciated internally and clasts show a internal ductile foliation that is misoriented within breccia. Brecciated basic dykes are cut by probable mid-Cretaceous alkaline basic dykes. 5) A final episode of brecciation is represented by the second of two phases of acid tuffisite dykes. These cut or incorporate clasts of all the lithologies described above. Multiple episodes of tectonic and/or magmatic brecciation have been recorded elsewhere. For example, on the Devonian Cobequid shear zone in Nova Scotia, Koukouvelas et al. (1996) have identified 26 phases of fault movement and magmatism or vein development. The eastern Palmer Land breccia system sits on the boundary between crustal domains of very different affinity. To the west the rocks are dominated by largely mantle-derived arc plutons (Wareham et al. 1997); to the east the rocks are of more continental affinity (Wever and Storey 1992). The breccias lies along strike and on a probable restraining bend of a major ductile reverse shear zone along the boundary between these domains. Multiple brecciation events, particularly with good evidence of tectonism as seen northeast of Mount Jackson, is strong evidence for a predominantly tectonic origin. Repeated emplacement and brecciation of magma may have resulted from channelling of magma through the fault system and provided the glue to preserve evidence of brittle fault activity at mid to shallow crustal levels.
Koukouvelas I, Pe-Piper G & Piper DJW, Geol. Mag, 133, 285-298, (1996).
Wareham CD, Millar IL & Vaughan APM, Cont. Min. Pet, 128, 81-96, (1997).
Wever HE & Storey BC, Tectonophys, 205, 239-259, (1992).
The south Bohemian granulites are divided into four granulite massifs (Blansky les, Kristanov, Prachatice and Lisov) which are incorporated in the Moldanubian nappe. The granulite bodies are mostly composed of leucocratic felsic granulite gneisses which regularly alternate with banded types composed of biotite and garnet-containing mafic bands and felsic bands. The granulite bodies contain blocks and xenoliths of ultramafic rocks.
The structural profile through largest granulite body (Blansky les massif-BLM) reveals both a flat-laying foliation and steep foliations marked by a biotite preferred orientation. In the cross-section, this foliations has a flower structure geometry and overprints an older steep N-trending foliation. These two planar structures can be interpreted in terms of homogenous thickening of the crust, continually compressed between rigid lithosperic plates, exhumed by vertical extrusion folowed by a horizontal shoulder displacement. Zones of granulites with horizontal structures and strong retrogressive changes of garnet to kyanite, garnet to sillimanite or garnet to biotite are restricted to the easternmost edge of BLM.
These strongly retrogressed granulites (ga-bt) with biotite foliation parallel to the fabric of the adjacent Moldanubian metasediments of the Varied Group are cut by slightly deformed granite dykes. Deformation of these granite dykes was coeval with the main deformational phase recorded in metasediments of the Varied Group and with the retrogression in the granulites. In order to define the youngest geological history of the granulite bodies, the two-mica granite from the Kremze Valley was dated, employing the conventional multigrain U/Pb method. Zircon points for equant to oval zircon grains define near concordant ages at ca. 320 ±1 Ma. The prismatic and euhedral needles of air-abraded zircon fraction have upper intercepts at 979±30 Ma, 1219±30 Ma and 1533±44 Ma. The obtained near concordant age of ca. 320 Ma is interpreted as the time of magmatic zircon crystallization from a granitic magma. The ages of upper intercepts probably reflect the presence of inherited cores assimilated during the granite formation and ascent. The Early Carboniferous age was verified by single grain evaporation 207Pb/206Pb method applied to the same slightly deformed granite of the Køemze Valley. The results fall in the range of ages between ca. 340 and 539 Ma. Both the BSE images and evaporation of zircons suggest the presence of inherited cores in zircons.
The obtained ages of ca. 320 - 340 Ma correspond to the minimum age of retrograde metamorphism, deformation and uplift of the granulites in the southern part of the Bohemian Massif. Similar ages of ca. 330 Ma were recorded from the granitic intrusions in the Saxonian Granulite Complex (Bauman et al., 1996).
Baumann N, Pilot J, Werner CD & Todt W, J. Con. Abs, 1(1), 50, (1996).
The Saxothuringian Erzgebirge is part of the metamorphic basement of the internal Mid-European Variscides. In lithological regard, postkinematic granites and lamprophyres are closely associated with metamorphic rocks, of which the interpretation is complicated by local intercalations of granulite- and eclogite- relics. Based on recent zircon-data, the age of the high-pressure metamorphism is about 341 Ma (Kröner & Willner 1998). Subject of our work is the chronological record of orogenic processes (quantification). Cooling ages were determined by means of the K-Ar- and Rb-Sr-methods. Working hypothesis was the idea that individual cooling-ages of different metamorphic and magmatic units might be used to model tectonometamorphic processes (simulation). Independent of method, technique and mineral, only Variscan ages, interpreted as cooling-ages, could be established. Phengites from the high-pressure gneisses as well as phengites and amphiboles from basic intercalations yield cooling ages of c. 340 Ma, identical with zircon evaporation data of 341 Ma obtained by other workers. This finding proves fast cooling causing phengite and amphibole ages to indicate the waning stages of the high-pressure event and the zircons to date its peak. Together with their gneissic host rocks, the granulites and eclogites were reheated 330 Ma ago, due to stacking of hot micaschists, garnet-phyllites and red-gneisses. Subsequently they were completely exhumed. Thermal modelling and diffusional considerations show that an uplift rate of 5 mm/a would have been sufficient to obtain gneiss ages of 340 Ma in the upper structural levels.
Kröner A, Willner AP, Contrib Mineral Petrol, 132, 1-20, (1998).
Rapid uplift histories of metamorphic belts are well known, especially from the younger orogens. Rapid burial is often assumed in relatively high-pressure rocks, but can not always be documented. The Corner Brook Lake belt is a thin-skinned foreland fold-thrust belt that experienced earliest ductile deformation, burial up to intermediate to high pressure amphibolite facies and cooling through 400°C within 10 my.
Late Precambrian to Early Ordovician clastic and calcareous sediments of the Fleur de Lys Supergroup and thin slices of their Grenvillian basement underlie the Corner Brook Lake area of western Newfoundland. They represent rocks of the Laurentian continental margin in the internal Humber zone, that were incorporated in a west-vergent fold-thrust belt and metamorphosed during the Silurian. A ca. 600 m thick, amphibolite grade imbricate stack of supracrustal rocks and thin basement slices overlies low metamorphic equivalents and is itself overlain by the many kilometres thick, metaclastic Yellow Marsh thrust slice. Highest metamorphic grades (650°C, 11 kbar) occur at the floor thrust of the Yellow Marsh sheet and decrease rapidly away from this structure. Rocks were affected by an early phase of brittle thrusting, followed by ductile simple shear lasting until peak metamorphism, during which the regionally dominant foliation (S1) was formed. Local shearing with creation of a second foliation (S2) and two phases of folding post-date peak metamorphism, the first contractional (F3), the youngest most likely associated with extensional structures (F4).
Timing of the metamorphic and deformational events was constrained by: 1) a syn-tectonic pegmatite which intruded during S1 at 434+2/-3 Ma (U-Pb zircon), constraining the age of earliest ductile deformation; 2) monazite in amphibolite grade pelite yielded 430 ±2 Ma, indicating the age of the temperature peak of metamorphism; 3) hornblende and muscovite Ar-Ar data yield ages around 425 Ma and 423 Ma representing cooling through respectively ca. 500°C and 400°C. A muscovite cooling age (Ar-Ar) of 413 Ma in the Yellow Marsh thrust slice could relate to significant, post peak movements. Interpretation of microstructures around peak metamorphic minerals suggests that the rocks were undergoing compression during F3 folding after peak metamorphism, indicating that contractional deformation of these rocks continued during exhumation. Ar-Ar spot measurement of syn-kinematically recrystallised muscovite in F3 folds further constrain the timing of this phase of deformation and its relationship to other regional deformation events.
Post-peak cooling in the Corner Brook Lake belt occurred at a rate of ca 30°C/Ma. This is a cooling rate that is not uncommon in young orogenic belts, but sharply contrasts the uplift rates of 1-2°C/Ma found in cores of Proterozoic, deeply exhumed orogens. These contrasting uplift rates are likely to reflect differences in exhumation processes during and after convergence.
The Sibai core complex in the Eastern Desert of Egypt is one of several magmatic and/or metamorphic core complexes aligned parallel to the Najd fault system. It is bounded by sets of sinistral strike-slip and interlinking normal faults arranged in a pull-apart geometry. The complex almost entirely consists of synkinematically intruded granitoids overlayn by Panafrican ophiolitic and supracrustal rocks. Exhumation and extensive magmatic activity were controlled by late Panafrican wrench tectonism following island arc accretion.
Relative ages of magmatic and kinematic events are deduced from cross cutting relations of both, distinct granitoid bodies and gradually evolving faults. Magmatic ages of the granitoids (U/Pb dating of single zircons; evaporation technique) confirm the field observations and allow absolute dating of extension and magma intrusion. Magmatic ages range from 659 - 645 Ma and can be related to the rate of shear extension using the map scale fault geometry.
40Ar/39Ar radiometric ages of synkinematically grown white mica from the strike-slip faults document cooling of the entire complex below the 350° isograde. Data cluster around c. 582 Ma. Final exhumation can be indirectly dated by the formation of a molasse basin adjacent to the dome, since pebbles of internal granitoids are found in the base conglomerate.
Combining all observations and data sets allows to reconstruct the entire history of the late Proterzoic Sibai core complex. Rates of shear extension and exhumation can be estimated from radiomertic and supplementary field data.
The Southern East Greenland volcanic margin is characterised by a 1000 m elevated coast line rising inland to more than 2000 m elevation at the inland ice. The terrain consists of mainly Archaean to Proterozoic basement. Kangerlussuaq (at 68ºN) was in theTertiary the location of the present Icelandic plume, which gave rise to voluminous magmatic activity.
Fission track analysis of apatites is used to constrain the extent and amount of magmatism through the thermal overprint, address the effect of plume passage on thermal history and reveal the late thermotectonic evolution (below c. 125ºC) of rift shoulders compared to stable continent.
The obtained FT ages show magmatism to be the main cause for the thermal anomaly found in the central Kangerlussuaq. At inner Kangerlussuaq and south of Kangerlussuaq the present surface was never exposed to temperatures higher than c. 125º since Caledonian (c. 60º in Neogene). Thus, thermal history restricts the possible thickness of overlying basalts S of Kangerlussuaq and doming in the Kangerlussuaq area which could be related to plume passage from a northwesterly direction.
FT analysis and thermal modelling from three studied profiles south of Kangerlussuaq shows ongoing exhumation at elevated rates since Tertiary times and an overall increase in FT ages with increasing elevation and distance to coast line. It is, however, difficult to resolve whether this represents constant exhumation rates of the landscape or a tilting away from rift shoulders in the Neogene. Moreover, the position and slopes of the paleo partial annealing zones show that thermal differences between profiles were established as a result of different crustal levels and thermal structure for the individual profiles before c. 300 Ma. Thus, extension of the Caledonian orogeny South of Kangerlussuaq may be responsible for the lack of Caledonian ages close to Kangerlussuaq.
The SongChay massif is the largest igneous complex in the southwestern corner of the South China block (SCB). Located at the border between North Vietnam and China, it has been thought to represent an exhumed piece of the Precambrian basement of this block. The SongChay complex is chiefly made of migmatites, porphyroid granites and augen-gneisses resulting from the deformation of these granites. These rocks form a 40 km wide dome with flat foliations in the core passing to steeper ones at the borders of the massif. In the core, a well defined stretching lineation strikes 5°E on the average, and shear criteria, imply a top to the north sense of shear. Multi-system geochronology was conducted on a mildly deformed orthogneiss. Seven zircon fractions are analysed for U-Pb. Linear regression of all data yields a normal discordia with an upper intercept age of 428±5 Ma, corresponding to the emplacement age of the granitic protolith. Such a middle Silurian age is consistent with the regional lower Devonian unconformity. One Rb-Sr isochron, based on one whole rock, one Kfeldspar and 5 muscovite fractions defines an age of 206±10 Ma, while another, based on the same whole rock and Kfeldspar, and 3 biotite fractions yields an age of 176±5.3 Ma. 39Ar/40Ar measurements on the same micas give plateau ages of 209±9 Ma and 190±8 Ma respectively. Together with P-T estimates on a nearby garnet micaschist, which bracket the main ductile deformation to have occurred at temperatures between 550 and 650°C, these results suggest that shearing took place during the Upper Triassic around 210 Ma. Similar ages are found at the western and northwestern margins of the SCB (SongPan and QinLing). The Kfeldspar 39Ar/40Ar age spectrum suggests a rapid Upper Jurassic (=140 Ma) cooling event, that could be associated with active subduction and magmatism along the SE margin of the SCB at that time. If continuous cooling is assumed, the model cooling history calculated from the Kfeldspar spectrum shows a second rapid cooling at =30 Ma, and temperatures dropping below 120°C after 50 Ma ago. However, the model cooling history calculated from apatite fission tracks data is incompatible with the latter. Temperature dropping below 120°C before 80 Ma ago. The only way to obtain compatible results with both methods is to envisage a re-heating episode younger than 50 Ma. In that case, fast cooling at 40 Ma is rapidly followed by re-heating culminating at 100°C 22 Ma ago. This Tertiary re-heating event is possibly associated with shear heating along the Ailao Shan - Red River strike-slip fault that bounds the SCB to the south, 40 km away from the SongChay complex.
The Northampton Block (NB) is a high-grade gneiss terrain west of the Darling Fault in Western Australia. It is dominated by granulite facies metasediments, intruded by porphyritic granites before peak metamorphism and later by pegmatites. The recorded P-T path consists of a near-isobaric heating-cooling cycle starting and ending at ca. 650 oC. The prograde path culminated at ca. 850 oC and 6-7 kbar.
SHRIMP U-Pb ages of monazites from a high-grade metasediment and zircon rims gave indistinguishable Grenvillean ages of ca. 1070 Ma. The monazite age of a strongly retrogressed sample is indistinguishable at ca. 1076 Ma, suggesting fast initial cooling. The youngest inherited zircons in metasediments and the porphyritic granite are ca. 1150-1170 Ma old, which gives a maximum age for their deposition into the Northampton palaeo-basin. The porphyritic granites intruded between 1094 Ma (zircon age) and 1082 Ma (monazite age), i.e. 10-20 Ma before peak metamorphism. Late pegmatites with staurolite-bearing contact aureoles intruded at ca. 1055 Ma, when the host rocks cooled near-isobarically towards a stable geotherm.
It follows from published and new age constraints that the average rate of heating was ca. 10oC/Ma during prograde metamorphism. Average cooling rates were similar during the first 20 Ma, decreasing to 2oC/Ma between 1055 and 980 Ma and 1oC/Ma until dolerite intrusion at 800-650 Ma. The near-isobaric prograde and retrograde P-T vectors are interpreted as the result of magmatic underplating in a strike-slip setting. This is consistent with structural data, which indicate transtension followed by dextral transpression, with the (proto-)Darling Fault acting as the major lineament.
Two metasediments and one porphyritic granite display consistent age spectra of detrital zircons (n=62), with clusters at 1150-1170, 1300-1350, 1400, 1490-1520, 1620 and 1700-1770 Ma. The absence of Archaean zircons is remarkable in view of the vicinity of the Yilgarn Craton. Each observed age group can be matched to specific metamorphic or igneous events in the Albany-Fraser Belt in Western Australia. We infer that the sediments were derived entirely from the Albany-Fraser Belt or its equivalents in India and East Antarctica. During the time of sedimentation and granulite metamorphism in the NB (a timespan of <80 Ma), the high grade rocks of the Albany-Fraser orogen underwent cooling. The NB is interpreted as a collision-induced foreland rift that developed simultaneously with uplift of the Grenvillean Albany-Fraser Belt and adjacent areas, in a setting that shows similarities with the Alpine Rhine Graben.
The Tertiary magmatic rocks of the Sierra Madre del Sur (SMS) are broadly distributed south of the Trans-Mexican Volcanic Belt (TMVB) of Miocene-Quaternary age and extend to the southern continental margin of Mexico. The magmatic activity in the SMS was originated at a time characterized by significant changes in the plate interactions as a result of the formation of the Caribbean plate and the southeastward displacement of the Chortis block along the continental margin of southwestern Mexico. The magmatic rocks of the SMS define two belts of NW orientation. The first is represented by the nearly continuous coastal plutonic belt (CPB), which consists of batholiths and stocks of predominantly silicic composition. The second belt is inland of the first and consists of discontinuous volcanic fields of basaltic andesite to rhyolitic composition. This second belt has been named the Tertiary volcanic province of southern Mexico (TVPSM) and extends from Michoacan to the Isthmus of Tehuantepec, Oaxaca.
The TVPSM has been studied in the central and southeastern most part of Oaxaca. The volcanic sequences in the study region are constituted predominately by silicic tuffs with minor rhyolitic to andesitic lava flows interbedded with lacustrine deposits. The entire sequence is intruded by dacitic to andesitic hypabyssal rocks. Geochemical data are very similar to other volcanic regions of the TVPSM, such as western Oaxaca and northeastern Guerrero. These rocks define a clear calc-alkalic tendency and have an intermediate- to high-K content. Chondrite normalized REE patterns display typical magmatic arc behavior, with LREE enrichment relative to HREE. Low initial strontium ratios along with trace element geochemistry suggest a mantle wedge source for the magmas with different degrees of magmatic differentiation and/or crustal contamination. Previously reported K-Ar age determinations for rocks of this region range from 20.6 near the city of Oaxaca to 15.0 Ma in Nejapa (easternmost part of Oaxaca). Other K-Ar dates from northwestern Guerrero and western Oaxaca show an eastward decrease in the age of magmatic activity in the TVPSM (38.2±1.0 to 26.5±0.5). The magmatism in the study region represents the latest activity in the SMS before the initiation of the E-W trending volcanism of the TMVB. Therefore, the location of these volcanic rocks in Oaxaca puts then in a key position for understanding the tectonic changes in the kinematic relations between the Cocos and North American plates.
Cooling ages of different mineral isotope systems infer a cooling path which is the base for calculation of the exhumation rate of metamorphic rocks. However, such calculations should incorporate not only the closure temperature concept, but should also consider that fluid interaction, deformation mechanisms, recrystallization, and mineral reactions during cooling are capable to (partially) reset isotope systems. In metamorphic complexes with a complicated prolonged tectonometamorphic history mineral assemblages may preserve the record of several stages. Within minerals of such complexes the distribution of major elements and isotope ratios may be heterogeneous hampering estimations of PT-conditions and age determinations. Consequently isotopic mineral data of such complexes may not reflect the cooling path.
This is demonstrated in the granulite terrain of Southern Calabria (Serre) using already published and new geochronological data (performed at Zentrallabor. Geochron. Münster). In this terrain different mineral ages ranging from 292-296 Ma (U-Pb monazite data) to 108-114 Ma (Rb-Sr biotite data) have been previously interpreted as to reflect the very slow cooling of lower crustal rocks (Schenk, 1990).
In a felsic metablastic grt-plg-qtz±bio granulite Sm-Nd investigations of garnet reveal significant isotope heterogeneities. Inclusion free idiomorphic garnet grains (0.25-0.375 mm) show Sm-Nd garnet - whole rock data of 238 ± 7 Ma. In contrast, idiomorphic garnet grains (0.25-0.375 mm) which were abraded to 25% of their original volume, yielded a Sm-Nd garnet - whole rock data of 279 ± 7 Ma. These data delineate that (i) at > about 280 Ma the garnet cores ceased to exchange Nd isotopes indicating a minimum age of garnet crystallization in the investigated rock, and (ii) a later resetting of the Sm-Nd system of the garnet rims. This may have been caused by interaction with intergranular fluids and/or material-exchange with neighbouring phases (plagioclase) during later stages. In the investigated sample evidence for fluid infiltration is given by discrete fluid inclusion-decorated microcracks.
Such features of deformation processes after granulite facies metamorphism are more pronounced in neighbouring granulite samples, where microstructures indicate (i) annealing and grain growth of garnet, plagioclase, (ii) deformation processes at elevated temperatures (synkinematic crystallization of sillimanite, biotite, garnet, cordierite in shear zones), (iii) followed by discrete shearing under LT-conditions (cf. Kruhl & Huntemann, 1991; Kruhl, in prep.). Resetting of the Sm-Nd system in the garnet rims may be attributed to these later deformation processes, presumably during emplacement tectonics of (meta)granulites against neighbouring complexes metamorphosed at lower grade. These results emphazise that the Sm-Nd date of the bulk garnet fraction does not reflect a cooling age, and does not indicate a point on the T-t-path as was earlier suggested. Microstructures and mineral data corroborate a polyphase history of the Calabrian granulite terrain, in contrast to the earlier suggested long lasting cooling history.
Kruhl JH & Huntemann T, Geologische Rundschau, 80/2, 289-302, (1991).
Schenk V, Salisbury & Fountain (eds). Exposed cross-sections of the continental crust. Kluwer Academic Publishers, 21-42, (1990).
The four tectonic units belong to the Eastern Pennine nappes. The heterogeneously deformed Tambo- and Suretta nappes are polycyclic and polymetamorphic nappes from the Briançonnais zone. Early Permian metagranites are intruded in their basement. They are overlain by a autochthonous Permo-Mesozoic cover and underlain by the Chiavenna and Gruf units. The Chiavenna unit consists of ultramafic/mafic rocks with a subcontinental-oceanic mantle origin. The Gruf unit consists of granulites and migmatites of presumably pre-Alpine age. The Tertiary intrusives Bergell and Novate intrusions cross-cut the nappe pile.
Based on the analyze of deformation, kinematics, metamorphic assemblages, absolute time markers and mineral ages, a four-phase tectonometamorphic evolution during Tertiary collision is proposed. In particular the convergence rate during the subduction event (D1 phase) and the exhumation rates (D2-D4 phase) are evaluated and discussed. Tectonic vs. erosional influence is distinctly recognizable at the velocity of the processes and the inclination of the slope in the PTt-diagram.
The first phase (D1) shows top to the NW-moving shear zones and mineral recrystallized under HP-conditions at 13 kbar. It is interpreted as nappe stacking during the late Eocene subduction. Fossils in Pennine Flysches indicate a maximum age of 37 Ma for the onset of the subduction of the Briançonnais zone, while dating of white micas in the D1-mylonites show 38 Ma. The Helvetic Flysch deposits at 34 Ma delimits the end of the D1-phase of the Briançonnais zone. Depending the subduction geometry, results rate between 2 and 8 cm/y.The second phase (D2) creates the main NNE-dipping schistosity with a top to the E-shearing. Decompression of 7 kbar is related to early Oligocene syn-orogenic E-W extension due to buoyancy disequilibrium caused by the anomalous thickness of the crust. D2-structures are cross-cut by the Bergell granodiorite (30 ma). This intrusion is D3-deformed. The exhumation rates around 8 mm/y, taking into account erosion and strain. The unroofing of the Tambo and Suretta nappes shows more than 30 km horizontal displacement. The third phase's (D3) steep schistosity and lineation are associated to E-W striking folds, which are responsible for the Oligocene differential uplift with major elevation in the Bergell area. It provokes erosional exhumation at rates between 2.5 and 0.6 mm/y and induces rapid cooling. The D4-deformed Novate intrusion (25 ma) cross-cuts the D3-structures and indicates the end of the D3-phase.
As precise geochronology of deformation fabrics is a critical factor in tectonic reconstructions it is desirable to date minerals directly within the microstructure in order to relate mineral ages to distinct deformational events. The high-spatial resolution of the 40Ar/39Ar UV-laser technique allows in-situ analysis of closely spaced muscovites aligned parallel to the main mylonitic foliation (S foliation) and mylonitic shear bands (C´surfaces). This avoids pitfalls inherent in separating populations of mineral grains based solely on their physical properties (e.g., grain size) rather than on their microstructural setting and their relation to metamorphic mineral paragenesis. The sample studied comes from a ca. 3 km wide crustal scale shear zone (Pogallo Shear Zone, Southern Alps, Italy) that accommodated oblique sinistral shear and exhumation of the Ivrea-Verbano Zone during Early Jurassic time. A comparative study of muscovite grains situated both within the S and C' surfaces indicates that deformation exerts strong control on the diffusion behavior of 40Ar* within muscovite grains. The irregular pattern of apparent ages within individual, deformed grains suggests that the classical static diffusion model is not appropriate for predicting intragranular 40Ar* distribution. The C' oriented muscovites are somewhat younger and record a narrower time interval than the muscovites within the S foliation. This is consistent with microstructural investigations indicating that C' microshears nucleate during the late, retrograde stages of mylonitic deformation. Once present, they accommodate slip contemporaneously with continued shearing on the S foliation (Platt & Vissers, 1980). Because these C' oriented grains are also smaller, they are more prone to Ar-loss during post-tectonic cooling, explaining their more pronounced age peak in the age histogram. In contrast, the broader age distribution of the S oriented muscovites reflects the superposition of syn- and post-mylonitic thermal regimes; the older end of the age spectrum for these muscovites corresponds to the age of mylonitization in the Pogallo Shear Zone, whereas the younger part of the age spectrum records post-mylonitic cooling. Our results of in-situ 40Ar/39Ar laser ablation dating revealed that mylonitization and nucleation of microshears is associated with segmentation of the muscovite grains, reducing the effective grainsize and increasing the characteristic diffusional length of Ar within the grains. In the case of the muscovites analyzed here, this lowered the closure temperature some 80-100°C compared to that predicted by diffusion theory for undeformed grains.
Platt JP & Vissers RLM, Journal of Structural Geology, 2, 397-410, (1980).
In order to determine the effect of simple shear deformation on argon retention in muscovites a detailed 40Ar/39Ar study using an excimer ultraviolet (UV) laser (l= 248 nm) was undertaken on a mylonitic Variscan pegmatite from the Siviez-Mischabel nappe (Penninic Alps, Wallis, Switzerland). The pegmatite is associated with eclogites having preserved amphibolite- and greenschist- facies mineral assemblages and for which a pre-Alpine age has been assumed on the basis of field and petrological relationships. The mylonitic direction and sense of shear in the pegmatite is given by large pre-existing muscovites with an asymmetric "mica fish" shape. One of these porphyroclasts (8.6 mm long and 3.6 mm wide) was oriented perpendicular to the rock foliation and its (001) cleavage. 182 in-situ high spatial resolution UV laser ablation 40Ar/39Ar analyses, approximately 50 x 70 mm, yielded two distinct age groups, one at 360-330 Ma, the age of Variscan metamorphism, and another group centered about 260 Ma. Compared to the staircase-shape conventional 40Ar/39Ar step-heating spectrum of grain separates from the same sample the former group is a little higher than the flat part of the spectrum (320-330 Ma for 90% 39Ar released) and the latter is identical to the first step. Microprobe investigations were done but no chemical anisotropy was found. An age contour map of this porphyroclast exhibits subparallel zones of roughly equal age over 2/3 of its area that are parallel to the direction of the mylonitic simple shear deformation. This spatial distribution of argon is incompatible with a simple volume diffusion pattern. They are related to micro-shearbands that act as preferred microstructurally-controlled pathways for argon loss (short-circuit diffusion) that remain open at significantly lower temperatures than for undeformed mica of the same physical grain size. During the ongoing deformation, mica acts as a solid body and the "mica fish" undergo very limited rotation. Beyond this point micro-shearbands form in a parallel relationship with the bulk shear sense and crosscut the "mica fish" at an angle of 30° to the (001) cleavage. Some incipient shear bands can be observed in thin section and are related to a development in the late stage of deformation. However many of them are observed only with the 40Ar/39Ar data. These results underscore the importance of simple shear deformation on the distribution of Ar* in a mica lattice.
Very iron-rich gabbros were dredged from the massif exposed beneath a major extensional detachment fault in the inside corner at the intersection between the Mid-Atlantic Ridge and the Atlantis Transform Fault close to 30°N (Blackman et al., 1998). The gabbros are associated with peridotites, and probably represent a small magmatic body intruded into upper mantle peridotite (Cannat, 1993). These ferrogabbros contain plagioclase (An45-An50), clinopyroxene (Mg#55-56), orthopyroxene (Mg#43-47), ilmenite and magnetite, and locally some interstitial iron-rich olivine (Fo31). This represents one of the most evolved chemical compositions known for the Atlantic Ocean. The pyroxene contains well-developed exsolution lamellae. Backscattered electron images show that the clinopyroxene consists of coarse augite with inverted pigeonite lamellae on (001). After exsolution as pigeonite from the augite host, the coarse (001) pigeonite lamellae, which are up to 5 mm in width have inverted to orthopyroxene. The low Ca-pyroxene is inverted pigeonite with coarse augite exsolution up to 15 mm in width on (001) and secondary fine exsolution lamellae (up to 1 mm) on (100). The exsolution lamellae nucleate and grow during the cooling of the rock, and from the bulk composition of the original pyroxenes (which indicate crystallisation close to 1100°C), the composition of the individual phases now (which indicate a final equilibration close to 800°C), the widths of the lamellae and the profiles in Ca content on traverses across the lamellae, it is possible to estimate a cooling rate for that temperature interval (Miyamoto and Takeda, 1994). The detachment fault which caps the inside corner massif has a total exposed length in the extension direction of 8 km, and dips down towards the nearby spreading axis. Seafloor morphology indicates that the slip rate on the fault was (or perhaps is) about half the total extension rate at this latitude, perhaps as much as 10 mm/yr, suggesting a slip duration for the fault of less than 1 million years. Estimation of cooling rates for these gabbros gives a powerful constraint on tectonic models for the evolution of detachment faults and of inside corner areas in this tectonic environment.
Blackman DK, Cann JR, Janssen B & Smith DK, J. Geophys. Res, 103, 21315-21333, (1998).
Cannat M, J. Geophys. Res, 98, 4163-4172, (1993).
Miyamoto M & Takeda H, Earth Planet. Sci. Lett, 122, 343-349, (1994).
Cation diffusivities in pyroxenes are important parameters for retrieving cooling histories of metamorphic rocks or kinetics of magmatic processes from measured zonations. Diffusion data are scarce, however, since the diffusivities are low even at magmatic temperatures and hence difficult to derive experimentally (Freer et al., 1982). Here I present data of Fe-Mg interdiffusion rates in orthopyroxene that are based on concentration gradients at the rims of a harzburgite xenolith from La Palma (Canary Islands). This xenolith had resided in a magma reservoir for tens of years prior to eruption, a time during which prolonged contact with alkalic magma caused formation of polycrystalline reaction rims and adjacent diffusion zones (Klügel, 1998). The diffusion zones are characterised by strongly decreasing Mg-number in olivine and orthopyroxene from typical mantle values (0.90-0.93) in the xenolith's interior to 0.80-0.85 at the rims. The width of the diffusion zones is 0.10-0.15 mm in orthopyroxenes and 1.3-1.6 mm in adjacent olivines. A simple diffusion model is used to calculate the average rate of Fe-Mg interdiffusion in orthopyroxene relative to olivine (Chakraborty, 1997) at a temperature range of 1050-1200°C as inferred for the host magma. If the reaction rims are neglected in the calculations, the model yields a lower bound for the Fe-Mg interdiffusion coefficient D(Fe-Mg) which is 2 to 2.3 log unit below that of olivine, e.g. 10-19 to 10-20 m2s-1 at 1100°C and IW-buffered oxygen fugacity. If the width of the reaction rims is taken into account, the calculated diffusivities increase by up to 0.7 log unit and overlap with the extrapolated values of Ganguly and Tazzoli (1994) that are based on theoretical formulations of the Fe-Mg interdiffusion process. The results are also similar to the maximum diffusion coefficients found experimentally for Fe, Mg, Ca and Al diffusion in diopside (Freer et al., 1982). The anisotropy of D(Fe-Mg) in orthopyroxene predicted by Ganguly and Tazzoli (1994) cannot be quantified from the zonations measured so far because diffusion interfaces are geometrically complex and include reaction rims which affected the length of the diffusion profiles to an extent not precisely known. A comparison of D(Fe-Mg) in orthopyroxene to that in Fe-rich aluminosilicate garnet (calculated after Ganguly et al., 1998) suggests that both are similar within one order of magnitude at 1050 to 1200°C.
Freer R, Carpenter MA, Long JVP & Reed SJB, Earth. Planet Sci. Lett, 58, 285-292, (1982).
Klügel A, Contrib. Mineral Petrol, 131, 237-257, (1998).
Chakraborty S, J. Geophys. Res, 102, 12317-12331, (1997).
Ganguly J & Tazzoli V, Am. Mineral, 79, 930-937, (1994).
Ganguly J, Cheng W & Chakraborty S, Contrib. Mineral Petrol, 131, 171-180, (1998).
Plagioclase-rich cumulates and restites are likely to be present near the continental crust/mantle boundary, significantly contributing to the overall positive Eu-anomaly of the lower continental crust. During collision tectonics, such plagioclase-rich rocks may be converted into grossular (grt), diopside (cpx) and kyanite (ky)-bearing assemblages (ky-eclogites and grospydites). In order to determine eclogite producing reactions and to investigate the partitioning behaviour of trace elements we performed high pressure, piston cylinder experiments on natural, plagioclase-rich rocks from lower Austria as well as on equivalent, synthetic Fe-free compositions. In order to study the influence of H2O on the reaction kinetics the experiments were performed under nominally "dry" (0.15 wt.% H2O) and "wet" (2.2 wt.% H2O) conditions, the latter reflecting the maximum water content observed in natural grospydites. Anhydrous starting materials for all experiments were glasses, which in a second set of experiments were doped with trace elements (REE, Rb, Sr; Th, Ta, Nb, Zr, Hf, U) at the 500 ppm level. Experiments were performed at 1100-1200°C, 3.0-3.5 GPa for 4 and 7 days, at the Bayerisches Geoinstitut (Contract No. ERBFMGECT980111 to D.C. Rubie). The products of the syntheses were analysed using the electron microprobe (Hannover) and will be analysed using secondary ion mass spectrometry to determine the distribution of trace elements among the reaction products. We have determined the effects of P, T, volatile content and experiment duration on the crystalline run products. Complete crystallisation of grt, cpx and ky only occurred under "wet" conditions on the timescale of our experiments. Under dry conditions, however, no significant crystallisation had occurred after 4 days. In the presence of water, grospydite assemblages form over the entire P, T range of our experiments. The new data allow an improvement of the existing grt/cpx geothermometers for high-Ca-Al protoliths.
Index of EUG 10 Volume
Further EUG 10 Information
Index of the Journal of Conference Abstracts
Cambridge Publications Home Page
Last Updated on Wednesday, March 17, 1999.
© 1997 Cambridge Publications