Journal of the Geological Society, London, Vol. 160, 2003, pp. 655–676. Printed in Great Britain.
Exhumation of the Ronda peridotite and its crustal envelope: constraints from thermal modelling of a P – T –time array J. P. P L AT T 1, T. W. A R G L E S 2 , A . C A RT E R 1 , S . P. K E L L E Y 2, M . J. W H I T E H O U S E 3 & L. LONERGAN4 1 Research School of Earth Sciences at UCL-Birkbeck, London WC1E 6BT, UK (e-mail:
[email protected]) 2 Department of Earth Sciences, Open University, Walton Hall, Milton Keynes MK7 6AA, UK 3 Laboratory for Isotope Geology, Swedish Museum of Natural History, Box 50007, SE-104 05 Stockholm, Sweden 4 Department of Earth Science and Engineering, Imperial College of Science, Technology and Medicine, Prince Consort Road, London SW7 2BP, UK Abstract: The Ronda peridotite in the Betic Cordillera of southern Spain is a relic of the sub-orogenic lithospheric mantle that was exhumed during earliest Miocene time from about 66 km depth. Overlying crustal rocks show an apparently coherent metamorphic zonation from high-pressure granulite-facies rocks at the contact to unmetamorphosed rocks 5 km higher in the structural sequence, indicating drastic tectonic thinning of the whole orogenic crust during exhumation. P–T paths from the peridotite and its crustal envelope indicate decompression with rising temperature to shallow depths. U–Pb ion microprobe dating of zircon, Ar/Ar dating of hornblende, Ar/Ar laserprobe dating of muscovite and biotite, and fission-track analysis of zircon and apatite reveal that cooling was extremely rapid in the interval 21.2–20.4 Ma. One-dimensional thermal modelling of the array of P–T–time paths indicates that an asthenospheric heat source at an initial depth of about 67.5 km is required to explain heating during exhumation, and that the main period of exhumation lasted 5 Ma, starting at around 25 Ma. Exhumation must therefore have directly followed removal of most, but not all, of the lithospheric mantle beneath the Betic orogen, and was coeval with a period of late orogenic extension that profoundly modified the crustal structure and created the present-day Alboran Sea in the western Mediterranean. Keywords: Betic Cordillera, absolute age, exhumation, thermobarometry, extensional tectonics.
the rocks approach the surface. Analysis of rates of exhumation therefore requires detailed evidence for the shape of the P–T path, and on the timing of zircon growth with respect to that path. Proposed mechanisms for the exhumation of high-P and ultrahigh-P rocks fall essentially into two groups: those that involve circulation of material within an active zone of tectonic convergence (e.g. Cloos 1982; Platt 1986; Chemenda et al. 1995; Burov et al. 2001), and those that require large-scale extension of the whole crust and lithosphere after thickening has ceased (e.g. Platt & Vissers 1989; Andersen et al. 1994). These should produce distinguishable P–T paths: during synorogenic corner-flow circulation, rocks will be refrigerated by the cool downgoing slab, and exhumational P–T paths will involve cooling (Cloos 1982; Ernst 1988). Post-orogenic extension, on the other hand, should produce roughly isothermal P–T paths over a large part of the exhumational trajectory, and temperature may rise during decompression as a result of radioactive heat generation (England & Thompson 1984; Huerta et al. 1999) or of changes in the thermal structure of the underlying mantle (Platt & England 1994; Platt et al. 1998). Thermal modelling of the P–T path can therefore provide constraints on possible exhumation mechanisms. Such models, however, have to allow both for the large uncertainties on P–T determinations (usually about 508 in temperature and 200 MPa in pressure), and the effects of local complexities in the geological setting and history. To help overcome this problem, we present here the results of thermal modelling of a well-dated P–T array that extends through the entire orogenic crust and upper-mantle section. This allows us to place much
There are two outstanding and related geodynamic problems concerning the exhumation of high-pressure and ultra-highpressure rocks. One is the rate of exhumation; the other is the cause and mechanism of exhumation. Rates of exhumation as high as 30 km Ma1 , comparable with rates of plate motion, have been suggested for high-P and ultra-high-P rocks in the Alps (e.g. Gebauer et al. 1997; Amato et al. 1999) and the Himalayas (O’Brien et al. 2001), as well as for the Ronda peridotite in southern Spain (Sa´nchez-Rodrı´guez & Gebauer 2000), mainly on the basis of zircon U–Pb ages. Some of these estimates have been based on assumptions that are open to question. These are (1) that the peak temperature and peak pressure of metamorphism necessarily coincide; (2) that zircon U–Pb ages correspond to the peak pressure of metamorphism; (3) that cooling tracks linearly with exhumation. Both theory (e.g. England & Thompson 1984; Platt & England 1994) and observation (e.g. Buick & Holland 1989; Platt et al. 1998; Argles et al. 1999) show that in many tectonic settings peak temperature is reached on the exhumation path at pressures substantially below peak pressure. This largely invalidates the first assumption. Zircon growth may be a result of a number of different processes, including crystallization from a melt phase (Roberts & Finger 1997), Ostwald ripening, and decompressional breakdown of silicates such as pyroxene and garnet (Fraser et al. 1997). None of these processes justify the second assumption. The third assumption is inherently improbable, because the time constant for re-equilibration of a disturbed thermal gradient scales with the square of the distance to the disturbance. If heat loss is to the surface, most of the cooling will occur towards the end of the exhumation history, as 655
656
J. P. P L AT T E T A L .
tighter constraints on the tectonic and thermal evolution of the lithosphere during late orogenic extension than has previously been possible.
The Ronda peridotite and its crustal envelope The mechanism of emplacement of large bodies of subcontinental mantle lherzolite within orogenic edifices, such as the Ronda and Beni Bousera peridotites of southern Spain and northern Morocco, remains controversial. The association of these bodies with a strongly zoned sequence of crustal metamorphic rocks, grading from granulite-facies pelitic gneiss at the contact upwards to unmetamorphosed rock over a structural thickness of a few kilometres, has suggested a relationship between their emplacement and the late history of extension, exhumation and decompressional metamorphism in the Betic–Rif orogen (e.g. Argles et al. 1999). The Internal Zones of this orogen, together with the crust underlying the adjacent Alboran Sea, comprise the Alboran Domain (Fig. 1). This was a Paleogene collisional orogen, which underwent rapid and substantial extension in late Oligocene to early Miocene time, resulting in the subsidence of much of the region below sea level (Platt & Vissers 1989). Extension and thinning of the Alboran Domain occurred during continuing convergence between Africa and Iberia, and was accommodated by shortening in a peripheral thrust-belt that defines the Betic–Rif arc (Platt et al. 2003a). The cause of this pattern of simultaneous extension and shortening is widely
debated: current ideas include back-arc extension driven by the westward rollback of a subducting slab of oceanic lithosphere (e.g. Lonergan & White 1997); convective removal of the lithospheric root beneath the collisional orogen, causing an increase in elevation and gravitational potential energy that drove the extension (Platt & Vissers 1989); delamination of the subcontinental lithosphere, peeling off either to the west (Comas et al. 1992) or to the east (Docherty & Banda 1995); or the break-off of a subducting slab (Blanco & Spakman 1993; Carminati et al. 1998). Seismological evidence increasingly points to the removal of lithospheric mantle in the region (Seber et al. 1996; Calvert et al. 2000), but it is difficult to use present-day mantle structure to place firm constraints on processes that acted more than 20 Ma ago. Analysis of the P–T–t (where t is time) history of exhumed rocks can be useful in this respect, however, as the various mechanisms have different consequences for the lithospheric and hence thermal structure of the region.
A P – T – t array through the orogenic crust and upper mantle We have assembled thermobarometric and radiometric age data for the Ronda peridotite, the zoned metamorphic sequence that lies above it in the Carratraca area, and nearby upper-crustal rocks around Malaga (Figs 2–4). The data from each level in the sequence, and the associated uncertainties, are discussed separately, and summarized in Tables 1–5 and Figs 5 and 6. Except where stated otherwise, all thermobarometric data have been previously published, and the radiometric data are all new. Analytical methods and detailed discussion of the radiometric results are presented in a Supplementary Publication, which can be obtained from the Society Library or the British Library Document Supply Centre, Boston Spa, Wetherby, West Yorkshire LS23 7BQ, UK as Supplementary Publication No. SUP 18189 (7 pages).
Peridotite The peridotite bodies in the internal zones of the Betic–Rif arc (Fig. 1) are mainly spinel lherzolites, which form sheet-like bodies up to several kilometres thick interlayered within highgrade metamorphic rocks of crustal origin (e.g. Barranco et al. 1990; Torne´ et al. 1992). The largest body (Sierra Bermeja in
Fig. 1. Tectonic map of the Alboran Sea and the adjacent mountain belts. Prebetic and Subbetic Zones are Miocene thin-skinned thrust belts developed in Mesozoic–Tertiary platform and basin sequences on the southern rifted margin of Iberia. External Rif is the equivalent thrust belt on the northern rifted margin of Africa. The Flysch Domain comprises Cretaceous–Tertiary siliciclastic sequences deposited in a deep basin between Africa and Iberia. The Internal Betic and Rif Zones, together with the crust beneath the Alboran Sea, constitute the Alboran Domain, a collisional terrane developed in Late Mesozoic–Paleogene time from continental crustal fragments of uncertain provenance.
Fig. 2. Tectonic map of the westernmost Betic Cordillera. Patterns are identified by the labels on the figure.
E X H U M AT I O N O F T H E RO N DA P E R I D OT I T E
657
Fig. 3. (a) Map of the Carratraca massif (after Argles et al. 1999). Locations of the cross-section and samples are indicated. The non-metamorphic units are traditionally assigned to the Malaguide Complex, all other rock units to the Alpujarride Complex. (b) Section across the northern side of the Carratraca massif. Patterns for rock types are the same as in (a). The steep northerly tilt is related to the Internal– External Boundary thrust just north of the area. Discrete normal faults that cut the metamorphic sequence are indicated; most other boundaries acted as extensional ductile shear zones at some point in the exhumational history.
southern Spain) is broadly zoned, with a narrow zone of mylonitic garnet-bearing peridotite on the NW margin, passing structurally downwards into foliated spinel lherzolite, then granular spinel lherzolite, and plagioclase peridotite in the south (Obata 1980; Van der Wal & Vissers 1993, 1996). Graphite pseudomorphs after diamond indicate that the peridotite has at some stage of its history been at depths greater than 150 km (Pearson et al. 1989; Davies et al. 1993), but osmium isotope systematics suggest that the body has been isolated from the convecting mantle since 1.36 Ga (Reisberg & Lorand 1995). Van der Wal & Vissers (1993) proposed that part of the exhumational history took place in the early Mesozoic as a result of the
separation of Eurasia from Gondwanaland, and that the presentday association of different peridotite facies in the massif results from Tertiary exhumation during the later stages of Alpine orogeny. This brought the peridotites from about 66 km depth to mid-crustal levels accompanied by significant heating, producing the textures and plagioclase-bearing assemblages seen in the southern part of the body, together with textural evidence for partial melting and melt percolation (Van der Wal & Bodinier 1996; Lenoir et al. 2001). The timing of this evolution is poorly constrained, but Sm–Nd dating of garnet and clinopyroxene (Zindler et al. 1983) and U–Pb dating of zircon (Sa´nchezRodrı´guez & Gebauer 2000) from garnet pyroxenite layers in the
658
J. P. P L AT T E T A L .
Fig. 4. Map of the Montes de Malaga with fission-track sample locations. Geology simplified and modified from Instituto Geolo´gico y Minero de Espan˜a, Ardales (1038), Alora (1052) and Malaga (1053) 1:50 000 map sheets. Lambert Grid. L, M, U in the legend refer to the lower, middle and upper tectonic units distinguished in the Malaguide Complex in this paper (see text).
Table 1. New thermobarometric data for the Carratraca garnet gneisses T (8C) 730 50 830 50 790 60
P (MPa) 1170 110 830 80 430 150
Phases
Calibrations
C234: Grt core, Cpx core Holland (1980, 1983); Krogh Ravna (2000) C234: Grt plateau, Hbl core, Plag matrix (AN) Kohn & Spear (1990) ; Holland & Blundy (1994)y C234: Grt rim, Opx, Plag (in corona, AI) Harley (1984a); Bhattacharya et al. (1991)
Letters in parentheses refer to specific analyses in Table 3. With a(An) from Holland & Powell (1992), garnet activities from Ganguly et al. (1996). y T estimate from Gnt–Hbl thermometer (Graham & Powell 1984) gives 820 8C.
Table 2. Summary of thermobarometric data for the Carratraca crustal sequence Rock unit Ronda peridotite Spinel tectonite (A) Recrystallization front (B) Granular peridotite (C) Carratraca crustal sequence Garnet gneiss high P (D) Garnet gneiss medium P (E) Garnet gneiss low P (F) Granoblastic gneiss Fibrolite schist early (G) Fibrolite schist main foliation (H) Fibrolite schist low P (I) Fibrolite schist cooling (J) Andalusite schist high P (K) Andalusite schist low P (L) Andalusite schist cooling (M)
Assemblage Ol Opx Cpx Sp Grt Ol Opx Cpx Sp Ol Opx Cpx Sp Pl Grt Cpx Grt Hbl Pl Grt Opx Pl Crd Grt An Spl Grt St Ky Pl Ru Grt St Sil Bi Musc Grt Sil Bi Musc And Bi Musc Grt St Ky Bi Musc Grt St And Bi Musc St And Bi Musc
T ( 8C)
P (MPa)
Ref.
851 84 1180 1225 1040 1200
1700 1900 1480 1780 700 1050
1 2 1
730 50 830 50 790 60 810 20 500 600 610 30 630 10 515 640 540 20 550 15 515 620
1170 110 830 80 430 150
3 3 3 4 4 4 4 4 4 4 4
900 1000 800 100 500 80 150 360 730 80 440 60 150 460
Letters in parentheses refer to the P–T fields shown in Figure 6. References: 1, van der Wal & Vissers (1993); 2, Lenoir et al. (2001); 3, this paper; 4, Argles et al. 1999.
Ronda peridotite give 21.5 1.8 Ma and 19.9 1.7, respectively, and a mean age of 25 1 Ma has been obtained by Lu– Hf dating of garnet and clinopyroxene from a similar layer in the Beni Bousera massif (Blichert-Toft et al. 1999). The interpretation of these ages is problematical, but they probably suggest cooling from high temperature in the earliest Miocene. A composite P–T path was published by Van der Wal &
Vissers (1993) for the evolution of the Ronda peridotite, based on data from the Sierra Bermeja (Fig. 5). This was constructed using data from different parts of the massif, and does not necessarily represent the actual path followed by any single body of rock. Temperatures were determined mainly from various calibrations of the two-pyroxene thermometer, the Ca in orthopyroxene and the Al in orthopyroxene thermometers, and
E X H U M AT I O N O F T H E RO N DA P E R I D OT I T E
659
Table 3. Mineral chemistry on Carratraca sample C234iii (mafic granulite) Mineral/analysis: Description: SiO2 TiO2 Al2 O3 FeO MnO MgO CaO Na2 O K2O Sum No. of oxygens Si Ti Al Fe2þ Mn Mg Ca Na K Sum
Grt CC15 core
Grt CC3 plateau
Grt PR rim
Cpx L core
Opx AJ corona
Amph BD matrix
Plag AO matrix
Plag AN matrix
Plag AI corona
39.28 0.09 21.68 24.48 0.71 7.25 7.51 – – 101.00 12 3.003 0.005 1.954 1.565 0.046 0.826 0.615 0 0 8.015
40.15 0.11 22.51 19.58 0.34 9.01 9.51 – – 101.21 12 3.001 0.006 1.983 1.224 0.022 1.004 0.762 0 0 8.001
38.51 0 22.03 28.98 1.39 7.4 2.11 – – 100.42 12 2.988 0 2.015 1.88 0.091 0.856 0.175 0 0 8.005
49.03 0.64 6.94 8.04 0.12 12.15 21.18 0.43 0.06 98.59 6 1.846 0.018 0.308 0.253 0.004 0.682 0.854 0.031 0.003 3.999
51.77 0.11 1.15 27.1 0.61 18.29 0.71 0.02 0.04 99.80 6 1.979 0.003 0.052 0.866 0.02 1.042 0.029 0.001 0.002 3.994
44.49 1.86 10.75 10.96 0.11 13.39 11.91 1.55 1.04 96.06 23 6.605 0.208 1.881 1.361 0.014 2.962 1.895 0.446 0.197 15.568
54.61 – 28.24 0.2 – – 10.57 5.26 0.14 99.02 8 2.485 0 1.515 0.008 0 0 0.515 0.464 0.008 4.994
47.77 – 32.72 0.13 – – 15.95 2.15 0.05 98.77 8 2.213 0 1.787 0.005 0 0 0.792 0.193 0.003 4.992
46.02 – 33.99 0.19 – – 17.32 1.5 0.04 99.06 8 2.136 0 1.86 0.007 0 0 0.861 0.135 0.002 5.002
Analyses were selected from a population of at least 10 for each texturally distinct category as being closest to the mean for that population. Garnet analyses were selected on the basis of stoichiometry and location (i.e. extreme rim, inner rim, plateau or core) from several line transects across the largest grains. An analytical error of 1% is assumed on all the analyses presented here in estimating uncertainties on P–T estimates.
pressures from the inferred stable mineral assemblages in peridotites and pyroxenites (Van der Wal 1993). Point 1 on the curve (Fig. 5) represents the crystallization of the assemblage in the foliated spinel peridotites, after exhumation from the garnet peridotite facies (inferred to have taken place during early Mesozoic rifting), and before the start of the Alpine orogenic history. Point 2 represents the formation of garnet-bearing peridotitic mylonite on the western (upper) boundary of the Sierra Bermeja, inferred to represent the effects of subductionrelated cooling or burial during Betic orogeny. Point 3 represents the start of thermal annealing of the spinel tectonites during the exhumation history, and point 4 the formation of granular spinel and plagioclase peridotites and the beginning of partial melting. The rocks on the upper margin of the massif (points 2 and 3) in this interpretation were exhumed apparently without much heating, a very different thermal history from that experienced by the granular peridotites in the south. The formation and preservation of the peridotite mylonite (point 2) in particular presents problems of interpretation. Van der Wal & Vissers (1993) described spinel crystals rimmed by a kelyphitic intergrowth of spinel and orthopyroxene, which they interpreted to represent the breakdown products of garnet that had formed around pre-existing spinel during a subduction-related cooling or burial event. The mechanism by which these rocks escaped subsequent annealing during the thermal event that affected much of the rest of the massif is not clear, nor why these mylonites are located along the upper surface of the peridotite body at the contact with the crustal sequence. Lenoir et al. (2001) suggested that there was in fact a smooth thermal gradient across the Ronda massif, and that preservation of the spinel tectonites and mylonites in the upper part of the body is a result of the transient nature of the thermal event. They suggested that the pyroxene compositions, on which the thermobarometry was based, were not fully re-equilibrated during this event. Based on additional pyroxene thermobarometry, they calculated that condi-
tions on the recrystallization front within the massif, where the spinel tectonites become modified by grain growth and melt percolation, were around 1180–1225 8C, at pressures corresponding to the transition from the ariegite to the seiland subfacies of the spinel peridotite facies (c. 1600 MPa at 1200 8C). Bearing in mind the compositional and textural evidence for disequilibrium in the spinel tectonites, we propose the following conservative interpretation of the P–T history of the massif. The temperature determined by two-pyroxene thermometry of 851 84 8C for mylonitic peridotites on the northern margin of the massif (Van der Wal 1993) is likely to represent conditions prior to the major thermal event that accompanied exhumation. The neoformation of garnet in the mylonites suggests that the massif at this time lay near the boundary between the garnet and spinel peridotite facies. This corresponds to point 2 in Figure 5; we have interpreted the uncertainties fairly broadly as area A in Figure 6. During exhumation and heating the body passed through the conditions defined at the recrystallization front by Lenoir et al. (2001) (area B in Fig. 6): the pressure constraint is approximate as it depends on the composition of the garnet clinopyroxenite (ariegite) layers in the peridotite. The massif subsequently passed through the spinel–plagioclase peridotite facies boundary at temperatures between the wet and dry solidus (1040–1200 8C) (area C in Fig. 6). The origin of the garnet peridotite mylonite presents a particularly interesting problem. Van der Wal & Vissers (1993) interpreted it as representing a fossil subduction zone, where cold lithosphere was carried beneath the massif. In view of its position along the upper contact of the body, in contact with lower-crustal rocks metamorphosed at a significantly lower pressure (,1500 MPa), we propose here that it is more likely to represent an extensional detachment formed during exhumation, which placed mantle rocks from around 66 km depth in contact with the base of the crust. In this interpretation, the preservation of the garnet mylonites is a result of their having been protected
660
J. P. P L AT T E T A L .
Table 4. Summary of radiometric age data for the Montes de Malaga area and the Carratraca crustal sequence Sample no.
UTM Co-ords
Montes de Malaga Upper unit MM10 542.2, 236.4 MM20 525.5, 250.7 MM21 521.7, 252.8 MM11 546.4, 245.8 Middle unit MM22 524.5, 248.5 MM9 548.3, 236.5 MM18 528.3, 242.5 MM19 527.0, 248.2 MM12 545.4, 245.4 MM15 542.6, 239.7 Lower unit MM3 551.2, 239.3 MM14 543.3, 243.9 MM16 532.8, 248.2 PB324 533.1, 246.2 Carratraca area CAR11 498.7,256.3 CAR9 499.1, 299.8 Shear zone PB436 499.1, 299.6 Shear zone CAR5 499.1, 299.6 Shear zone PB437 501.6, 256.1 CAR14 501.5, 255.3 PB374 501.5, 255.3 PB375/CAR16 503.1, 255.8 Shear zone PB376/CAR17 503.3, 255.7 PB377 505.4, 254.9 Shear zone PB378/CAR19 505.1, 254.6 PB439 505.6, 253.7 PB379 505.6, 253.7 TAE11 505.6, 253.7 PB380 505.6, 253.7 Shear zone PB381 505.4, 253.4 PB382B 505.4, 253.4 PB383 505.4, 253.4 PB383B 505.4, 253.4
Rock type
P–Tr P–Tr P–Tr P–Tr
U–Pb Zr
Hbl
White mica
Bi
FTzr
FTap
269 22 185 24 189 14 20.9 3.8
sandstone sandstone sandstone sandstone
Late Pal greywacke Late Pal greywacke Late Pal greywacke E. Pal clastic rocks E. Pal clastic rocks E. Pal clastic rocks
355 50
Phyllite and quartzite Phyllite and quartzite Phyllite and quartzite And phyllite
22.7 1.8 22.1 1.6 21.9 2.8
315 39 278 42 170 20
150 36 15.1 1.8 17.5 3.4 22.5 3.4 24.7 3 15.7 1.6 13.4 2.2
18.6 1.6 22.0 4.0 25.0 4.0
Pal greywacke Pz greywacke
280 19
Psammite
18.9 3.0
Psammitic schist
19.0 4.0
18.0 6.0
Quartzite Quartzite Qtz–Ky–Mu vein And quartzite
21.1 4.4
20.8 5.2 16.4 2.8
19.9 .8 21.7 .4
21.3 .3
Fibrolite schist Fibrolite schist
20.0 0.8 21.2 0.6
21.0 2.0
17.8 2.4
21.5 1.5
20.9 .4
22.0 4.0
14.9 2.2 17.2 2.0 22.4 3.0
Sill gneiss Sill gneiss Migmatitic gneiss Migmatitic gneiss Leucosome Garnet gneiss Mafic gneiss Garnet gneiss
19.2 4.6
21.1 .4 20.5 0.7
22.9 0.5
20.4 2.4 21.0 0.6
21.7 3.8
19.5 2.0 21.2 0.3 21.4 1.8
19.2 8.4 15.9 3.8
Samples are arranged in approximate structural order within the two areas. Single-grain ion microprobe U/Pb age. U/Pb ages published by Platt et al. (2003b) and Platt & Whitehouse (1999).
from heating by juxtaposition against cooler rocks of the lower crust, after which they followed the same P–T path as the latter.
Crustal envelope The rocks above the peridotite bodies consistently show a distinctive and remarkable metamorphic zonation, from garnetrich pelitic granulite at the contact upwards through sillimanite gneiss and migmatite, fibrolite schist, andalusite schist, phyllite, to unmetamorphosed rock at the top of the sequence (Loomis 1972a; Torres-Rolda´n 1981; Balanya´ et al. 1993, 1997; Tubı´a et al. 1993; Bouybaoue`ne et al. 1998; Argles et al. 1999). Granulite, migmatite and high-grade schist, gneiss and marble also form a zoned metamorphic complex beneath the peridotite in the Sierra Alpujata (Tubı´a et al. 1997). The zoned sequence lying above the Ronda peridotite was originally interpreted as a thermal contact aureole around a high-
temperature diapiric ultramafic intrusion (Loomis 1972a, 1972b), but this interpretation is now untenable for several reasons. (1) The higher-grade rocks of the inner aureole show evidence for decompression from high-P and high-T conditions during strong deformation, and the gradient in metamorphic pressure across the sequence is substantially greater than its present-day thickness can account for (Balanya´ et al. 1993; Tubı´a et al. 1997; Argles et al. 1999). This suggests that the present-day structure is primarily a result of substantial extension and thinning of an originally much thicker sequence with a depth-related P–T gradient. (2) Thermal modelling by Argles et al. (1999) demonstrates that a peridotite sheet with the geometry and dimensions of the present-day Ronda massif could not have been responsible for the thermal evolution of the zoned sequence. (3) The petrological evidence from the peridotite itself indicates that temperature rose during its emplacement into crustal levels (Van der Wal & Vissers 1993; Lenoir et al. 2001). The late thermal
E X H U M AT I O N O F T H E RO N DA P E R I D OT I T E Table 5. Ar/Ar data for the Montes de Malaga area and Carratraca crustal sequence; amounts of 40
PB324 muscovite J ¼ 0.009651
Ar/39 Ar
38
Ar/39 Ar
37
Ar/39 Ar
36
Ar/39 Ar
39
661
Ar 3 1012 cm3 STP 39
Ar
40
Ar /39 Ar
0.013 0.014 0.020 0.013 0.012 0.013 0.013 0.012 0.013 0.013 0.012 0.014 0.013
0.000 0.003 0.000 0.000 0.000 0.003 0.029 0.016 0.017 0.004 0.021 0.012 0.047
0.0007 0.0038 0.0163 0.0051 0.0049 0.0039 0.0025 0.0020 0.0008 0.0036 0.0020 0.0027 0.0041
24.3 55.4 12.6 35.4 31.1 21.4 97.7 77.1 103.2 122.3 64.0 53.3 116.4
1.559 1.147 0.687 0.448 0.759 1.028 1.022 1.154 1.389 1.143 0.926 1.091 1.082
26.9 19.9 11.9 7.8 13.2 17.8 17.7 20.0 24.0 19.8 16.1 18.9 18.7 18.6
1.3 0.7 2.6 1.0 1.1 1.5 0.4 0.4 0.5 1.3 0.4 0.5 0.3 1.8
2.47 3.38 3.45 3.35 6.18 2.11 2.83 2.28 2.00
0.017 0.030 0.030 0.018 0.023 0.017 0.016 0.015 0.016
0.000 0.000 0.000 0.003 0.000 0.002 0.000 0.000 0.000
0.0011 0.0047 0.0057 0.0038 0.0142 0.0007 0.0028 0.0003 0.0000
86.0 24.0 31.3 56.1 96.4 108.4 94.7 107.8 66.1
2.14 2.00 1.76 2.22 2.00 1.91 2.01 2.19 1.99
20.8 19.4 17.1 21.6 19.4 18.6 19.5 21.3 19.3 19.9
0.5 1.9 1.4 0.8 0.5 0.4 0.4 0.4 0.6 0.8
2.53 2.69 2.64 2.91 2.58 3.52 2.87
0.037 0.038 0.041 0.048 0.041 0.043 0.040
0.002 0.000 0.000 0.059 0.020 0.012 0.022
0.0010 0.0015 0.0013 0.0024 0.0020 0.0045 0.0020
58.5 35.2 38.4 25.6 28.0 184.9 162.4
2.22 2.24 2.27 2.20 1.98 2.20 2.28
21.5 21.7 22.0 21.3 19.2 21.4 22.1 21.7
0.6 0.9 0.9 1.3 1.2 0.2 0.2 0.4
2.566 2.284 2.596 2.507 2.522 2.648 2.590 2.604 2.524
0.0274 0.0296 0.0282 0.0273 0.0285 0.0274 0.0254 0.0267 0.0255
0.0009 0.0114 0.0005 0.0017 0.0000 0.0012 0.0000 0.0000 0.0069
0.0009 0.0002 0.0016 0.0010 0.0013 0.0015 0.0011 0.0015 0.0011
101.8 17.3 81.5 50.1 127.4 69.0 101.5 111.5 29.4
2.310 2.212 2.124 2.210 2.151 2.198 2.272 2.148 2.186
22.37 21.42 20.57 21.40 20.84 21.29 22.00 20.81 21.17 21.300
0.45 1.71 0.37 0.59 0.25 0.32 0.20 0.19 0.60 0.30
2.254 2.156 2.262 2.238 2.271 2.282 2.337
0.0166 0.0185 0.0162 0.0152 0.0161 0.0156 0.0162
0.0103 0.0043 0.0077 0.0099 0.0032 0.0056 0.0056
0.0005 0.0002 0.0007 0.0001 0.0009 0.0010 0.0008
18.5 47.0 38.4 32.2 52.9 39.4 35.8
2.120 2.100 2.068 2.199 2.013 1.999 2.091
20.54 20.35 20.03 21.30 19.50 19.37 20.26 20.000
0.92 0.58 0.70 0.54 0.65 0.69 0.48 0.80
2.392 2.453 2.702 2.282
0.0142 0.0156 0.0224 0.0164
0.0127 0.0362 0.0136 0.0110
0.0003 0.0007 0.0015 0.0009
45.7 32.1 25.8 24.2
2.295 2.259 2.269 2.026
22.22 21.88 21.97 19.63 21.50
0.60 0.55 1.05 0.71 1.50
2.421 2.327 2.322 2.323 2.383 2.258
0.0225 0.0181 0.0231 0.0234 0.0211 0.0175
0.0282 0.0037 0.0080 0.0077 0.0063 0.0100
0.0009 0.0004 0.0005 0.0001 0.0005 0.0006
46.8 89.9 95.3 67.1 58.6 48.5
2.156 2.217 2.179 2.286 2.235 2.079
Weighted mean age PB375 muscovite J ¼ 0:0054052
Weighted mean age PB375 biotite J ¼ 0:0054003
Weighted mean age PB376 biotite J ¼ 0:0054005
Weighted mean age PB377 muscovite J ¼ 0:0054003 Weighted mean age PB375 biotite J ¼ 0:0054003
Weighted mean age
1.766 2.268 5.497 1.957 2.196 2.188 1.757 1.733 1.616 2.198 1.506 1.881 2.283
Weighted mean age PB374 muscovite J ¼ 0:0054052
Age (Ma)
20.88 0.58 21.47 0.23 21.10 0.22 22.13 0.41 21.64 0.47 20.14 0.37 21.20 0.60 (continued on next page)
662
J. P. P L AT T E T A L .
Table 5. (continued) 40
PB378 biotite J ¼ 0:005405
Ar/39 Ar
38
Ar/39 Ar
37
Ar/39 Ar
36
Ar/39 Ar
39
Ar
40
Ar /39 Ar
0.028 0.029 0.030 0.031 0.029 0.030 0.029
0.025 0.033 0.026 0.020 0.000 0.000 0.000
0.0003 0.0002 0.0002 0.0002 0.0000 0.0001 0.0009
88.7 139.9 170.0 169.0 91.8 32.6 9.2
2.10 2.25 2.11 2.18 2.07 2.04 2.07
20.3 21.8 20.4 21.2 20.0 19.8 20.1 20.9
0.4 0.3 0.2 0.2 0.5 1.5 5.2 0.4
2.48 2.61 2.48 2.44 2.59 2.44 3.05 2.90 2.63 2.53
0.048 0.046 0.036 0.040 0.042 0.042 0.047 0.045 0.045 0.044
0.000 0.000 0.000 0.003 0.000 0.022 0.000 0.003 0.015 0.019
0.0016 0.0011 0.0007 0.0009 0.0010 0.0007 0.0030 0.0019 0.0015 0.0015
6.8 9.2 67.8 200.7 24.1 58.5 23.5 110.5 186.1 200.9
2.01 2.29 2.26 2.16 2.29 2.23 2.16 2.34 2.20 2.09
19.5 22.2 21.9 20.9 22.2 21.7 21.0 22.7 21.3 20.3 21.1
7.7 5.7 0.8 0.3 2.2 0.9 2.2 0.5 0.3 0.4 0.4
1.462 4.267 1.675 1.585 1.274 1.777 1.452 1.279
0.014 0.015 0.014 0.014 0.014 0.012 0.014 0.014
0.0099 0.0239 0.0109 0.0150 0.0002 0.0060 0.0650 0.0469
0.0011 0.0106 0.0015 0.0012 0.0006 0.0017 0.0008 0.0004
124.4 23.8 255.9 132.9 312.9 55.6 198.6 180.2
1.134 1.134 1.232 1.241 1.107 1.263 1.213 1.166
19.6 19.6 21.3 21.4 19.1 21.8 21.0 20.1 20.5
0.3 1.3 0.2 0.3 0.2 0.4 0.2 0.2 0.7
2.45 3.14 2.26 2.39 2.37 2.27 2.24 2.44 2.38
0.042 0.052 0.026 0.040 0.033 0.039 0.039 0.041 0.041
0.009 0.031 0.000 0.000 0.000 0.000 0.000 0.000 0.016
0.0016 0.0032 0.0002 0.0006 0.0005 0.0002 0.0002 0.0007 0.0011
256.6 219.6 55.0 38.9 152.6 314.4 242.3 261.0 134.6
1.99 2.20 2.19 2.21 2.21 2.20 2.19 2.24 2.06
19.3 21.3 21.2 21.4 21.4 21.4 21.2 21.7 20.0 21.0
0.2 0.3 0.8 1.1 0.3 0.2 0.2 0.2 0.3 0.6
2.64 2.28 1.92 2.50 3.65 3.44 3.15 2.35 2.17
0.011 0.011 0.010 0.011 0.012 0.012 0.010 0.011 0.012
65.971 90.242 65.953 15.470 8.120 14.962 10.707 9.116 13.558
0.0004 0.0001 0.0001 0.0003 0.0011 0.0010 0.0010 0.0002 0.0001
99.8 56.9 68.6 184.1 89.5 51.7 52.2 61.6 49.5
2.52 2.24 1.90 2.41 3.31 3.15 2.85 2.28 2.13
24.5 21.7 18.4 23.3 32.0 30.4 27.6 22.1 20.7 19.5
0.5 0.6 0.5 0.3 0.4 0.7 0.8 0.6 1.2 2.0 1430.0
Weighted mean age TAE11 muscovite J ¼ 0:009634
Weighted mean age PB381 biotite J ¼ 0:005405
Weighted mean age PB382 hornblende J ¼ 0:005405
2.19 2.32 2.16 2.25 2.08 2.08 2.34
Weighted mean age PB379 biotite J ¼ 0:005405
Age (Ma)
Isochron age 40 Ar/36 Ar intercept J values shown with samples have an error of 0.5% and are quoted against GA1550 biotite standard with an age of 98.79 Ma.
evolution of the peridotite and its crustal envelope may therefore have a common cause in an elevated heat flow from the underlying mantle, rather than the one being the cause of the other. Radiometric data from the high-grade schist and gneiss surrounding the peridotite bodies consistently indicate rapid cooling in the early Miocene (Zeck et al. 1992; Platt et al. 2003b; Monie´ et al. 1994; Andriessen & Zeck 1996; Sosson et al. 1998; Platt & Whitehouse 1999), and this is generally interpreted to indicate the timing of the main phase of extensional exhumation. A well-developed example of the zoned metamorphic sequence occurs on the north side of the Carratraca peridotite massif (Fig. 3a), the easternmost of the peridotite outcrops that make up the
Ronda body. The envelope consists of garnet gneiss, migmatite, sillimanite gneiss, fibrolite schist, andalusite schist and quartzite, phyllite and unmetamorphosed Palaeozoic to Cretaceous sedimentary rocks. Argles et al. (1999) published thermobarometric data for this sequence, now some 5 km in structural thickness, and concluded that it represents the dramatically thinned remains of the orogenic crust, originally some 50 km thick, that lay above the peridotites. The sequence is cut by numerous faults and shear zones of extensional origin, the presence of which can be inferred both on petrological and structural grounds (Fig. 3b). Extensional deformation appears have been approximately coaxial during the early ductile stages, but a clear top-to-the-NE shear sense is evident during the transition to brittle deformation (Argles et al. 1999).
E X H U M AT I O N O F T H E RO N DA P E R I D OT I T E MPa 2000
663
garnet peridotite ariegite
2 1600
1 spinel peridotite
1200
seiland
3
4 800
plagioclase peridotite 400
wet solidus 600
800
1000
dry solidus 1200
T˚C
Fig. 5. Composite P–T path for the Ronda peridotite, after Van der Wal & Vissers (1993). 1, Foliated spinel peridotite; 2, peridotite mylonite with kelyphite after garnet; 3, ariegite dyke close to annealing front in spinel peridotite; 4, granular spinel and plagioclase peridotite. The ariegite and seiland subfacies of the spinel peridotite facies are distinguished on the basis of the mineral assemblage in mafic layers, either garnet bearing (ariegite) or spinel bearing.
Garnet gneiss. Strongly foliated porphyroblastic pelitic gneiss forms a selvedge up to 100 m thick in sharp (but in some cases ductile) contact with peridotite on the northern margin of the Carratraca massif. The domainal texture of the gneiss preserves assemblages from several distinct stages in their metamorphic history, as illustrated in Figure 7a. Inclusions of kyanite, rutile and white mica in the cores of large (up to 4 cm) garnet porphyroblasts attest to an early high-pressure stage (A in Fig. 7a). Garnet rims contain abundant sillimanite needles (B in Fig. 7a), and show higher grossular contents than the cores (Whitehouse & Platt 2003), probably reflecting partial melting in the matrix during this stage of growth (Spear & Kohn 1996). Leucosome development associated with the main foliation wrapping the garnets supports this evidence. This foliation (C in Fig. 7a) is defined by sillimanite pseudomorphs after kyanite, implying decompression during its formation. Low-pressure Crd–An–Herc assemblages occur as reaction coronas on garnet (D in Fig. 7a) and in a millimetre-scale shear zone (E in Fig. 7a) with orthopyroxene formed from garnet breakdown. Thermobarometric results on the pelitic gneiss were published by Argles et al. (1999), showing roughly isothermal decompression from around 1400 to 400 MPa at temperatures of 750– 800 8C. We have re-examined these results for the following reasons. First, major element profiles across the garnets are remarkably flat; this suggests diffusional homogenization, which is likely to compromise the results of inclusion thermobarometry. Second, much of the plagioclase and K-feldspar in the pelitic gneiss is likely to have crystallized from melt at a late stage and under low pressure; their compositions cannot therefore be used to constrain the early history of these rocks. Third, it is likely that much of the biotite in the matrix of the pelitic gneiss has undergone compositional modification during decompression, and the low-temperature assemblages show strong textural evidence for disequilibrium. We conclude that thermobarometric results for the pelitic gneiss are in general unlikely to be reliable, particularly for the high-P part of the history. The pelitic gneiss contains small isolated boudins of mafic gneiss. The textural evidence and mineral chemistry indicate that diffusional modification was more restricted in these boudins, which may therefore provide a more reliable route for defining
Fig. 6. Inferred P–T conditions and paths for the Ronda peridotites and their crustal envelope in the Carratraca massif. A, Foliated spinel peridotite and garnet-bearing peridotite mylonite. B, Granular spinel with ariegite layers at recrystallization front. C, Plagioclase peridotite. D, Grt– Cpx cores in mafic boudin from garnet gneiss. E, Grt–Hbl–Pl in mafic boudin. F, Grt–Opx–Pl in mafic boudin. The temperature constraint from a Crd–Grt–An–Spl granoblastic gneiss is shown by the smaller ellipse, at a nominal pressure of 450 50 MPa. G, Grt–St–Ky–Pl–Rt assemblage in fibrolite schist. H, Grt–St–Ky–Bt–Ms assemblage in fibrolite schist. I, Grt–Sil–Bt–Ms assemblage in fibrolite schist. J, And– Bt–Ms assemblage in fibrolite schist. K, Grt–St–Ky–Bt–Ms in andalusite schist. L, Grt–St–And–Bt–Ms assemblage in andalusite schist. M, St–And–Bt–Ms assemblage in andalusite schist. Phase boundaries as follows (given in the up-T direction). Fields A–C are interpretations made in this paper based on data from Van der Wal & Vissers (1993) and Lenoir et al. (2001), and D–F from data presented in this paper. All other fields after Argles et al. (1999). Univariant equilibria: 1, garnet–spinel peridotite; 2, ariegite–seiland subfacies boundary; 3, spinel–plagioclase peridotite; 4, dry peridotite solidus; 5, wet peridotite solidus; 6, Ms þ Qtz ! Kfs þ Als þ H2 O; 7, St þ Ms þ Qtz ! Grt þ Bi þ Al2 SiO5 þ H2 O; 8, Bi þ Al2 SiO5 þ Qtz ! Crd þ Grt þ Kfs þ L; 9, Ms þ Grt ! Bt þ Crd þ And; 10, Chl þ Ms þ Grt ! Bt þ St þ Qtz; 11, Ky–Sill; 12, And–Sil; 13, Ky–And. Abbreviations from Kretz (1983).
the P–T path. We have therefore carried out further thermobarometric studies on a sample (C234) from the unfoliated core of one of these boudins (see the Supplementary Publication for analytical methods). An early high-P assemblage of Grt–Cpx– Ttn Qtz initially reacted to form hornblende (sensu lato) from pyroxene breakdown, and subsequently by a reaction of the form Grt þ Cpx þ Qtz þ Na Pl þ H2 O ¼ Hbl þ Ca-Pl:
(1)
Later coronal Opx-Pl (An80 ) symplectites on garnet formed by the reaction Grt þ Qtz ¼ Opx þ Ca-Pl:
(2)
Along with ilmenite rims on titanite, these undeformed coronas correspond to the mainly static Crd–An–Herc stage seen in the pelitic gneiss. X-ray mapping of mineral chemistry was
664
J. P. P L AT T E T A L .
Fig. 7. Metamorphic textures in the garnet gneisses. (a) Textures in the pelitic gneiss, B203. Photomicrograph in plane-polarized light showing examples of the separate domains selected for petrographic analysis and thermobarometry. A, Inclusions of high-pressure minerals in garnet cores. Thermobarometry compromised by diffusional re-equilibration. B, Garnet rim with sillimanite needles and zircon. Included zircon from these rims has been dated at 21.2 0.3 Ma. C, Main sillimanite-grade foliation reflecting strain during decompression. D, Largely post-tectonic compound coronas of cordierite and anorthite þ hercynite spinel between garnet and sillimanite. E, Low-pressure discrete mylonite seam with garnet breaking down to orthopyroxene in a fine-grained, low-pressure assemblage. (b) Textures in the mafic boudin, C234. A, Decompression-related Opx–Pl coronas between garnet and quartz towards the edge of the mafic boudin. B, Ilmenite rimming titanite grain, with extensive garnet breakdown as in A. C, Nearer the core of the boudin, garnet and clinopyroxene are separated by amphibole, and later Opx–Pl coronas. Interstitial plagioclase is zoned, becoming more calcic from core to rim. Quartz is less abundant.
used to confirm the viability of selective thermobarometry on the separate assemblages. Large clinopyroxene grains are unzoned except at extreme rims, and small idiomorphic garnet inclusions in clinopyroxene show no obvious zonation except very thin Mnrich rims, probably as a result of minor garnet resorption. Garnets outside pyroxene grains have cores preserving relics of zoning in major elements (similar in composition to garnets included in clinopyroxene), as well as flat plateaux and thin, diffusionally modified rims. Hornblende is unzoned, and contains a few garnet inclusions with no discernible reaction between the two phases. Another indication of the restricted diffusion in this sample is the proximity of two texturally and chemically distinct
generations of plagioclase; symplectite plagioclase is of uniform calcic composition (An80 ), whereas interstitial grains have more sodic interiors (An55 ) and thin, perfectly concentric calcic rims merging into the edges of the coronal symplectite. We infer that the sodic cores predated the coronas, which were probably coeval with the calcic rims. The sodic plagioclase probably formed part of the higher-pressure assemblage, but unequivocal textural evidence for this is lacking. P–T estimates and calibrations are summarized in Tables 1 and 2, and the mineral chemistry data are collated in Table 3. The most reliable high-P estimate (1170 110 MPa) is given by the jadeite content of clinopyroxene (Holland 1980, 1983), the
E X H U M AT I O N O F T H E RO N DA P E R I D OT I T E
665
Fig. 7. (continued )
only system to have resisted resetting during decompression and cooling (D in Fig. 6), but this is a minimum estimate, as there is no evidence for plagioclase in the assemblage. We have estimated the temperature at this stage from Fe–Mg exchange between garnet and clinopyroxene (Krogh Ravna 2000) at 730 50 8C. The lack of extensive retrograde diffusional zoning in garnet suggests that this estimate is meaningful, but some reequilibration is likely to have occurred, so the estimate is at best a maximum value, and may be too high. Attempts to correct for this re-equilibration (Fitzsimons & Harley 1994) were unsuccessful. Peak-T during decompression is better constrained (at around 830 50 8C) from coexisting hornblende, garnet plateaux and interstitial plagioclase (Graham & Powell 1984; Holland & Blundy 1994). An intermediate pressure (830 80 MPa) from garnet plateaux and matrix plagioclase (Kohn & Spear 1990; Holland & Powell 1992; Ganguly et al. 1996) records growth of this assemblage during decompression (E in Fig. 6). P–T conditions for the low-pressure stage were determined on Opx–Pl coronas and extreme garnet rims at 790 60 8C, 430 150 MPa (F in Fig. 6) (Harley 1984; Bhattacharya et al.
1991). This result agrees with estimates (using a KFMASH grid in equilibrium with melt; Waters 1991) from a granoblastic Crd– Grt–An–Spl gneiss (D204i) at the peridotite contact, 330 m SE from C234, of around 810 8C (Argles et al. 1999), as well as the widespread development of hercynite spinel in the low-P assemblages. The P–T path defined by thermobarometry appears to involve some heating during decompression, from around 730 8C during the high-P stage to around 800 8C during the formation of the symplectitic coronas, but the uncertainties on our estimates are too large to determine precisely how much heating there was or when it occurred. Some additional constraints on the form of the P–T path, however, can be gained by consideration of the Caenriched rims of the garnets in the pelitic gneiss, which formed in the sillimanite field, and were probably associated with partial melting. This stage of garnet growth must have occurred at a pressure of less than 900 MPa, being the pressure at which our estimated P–T path (Fig. 6) intersects the kyanite–sillimanite univariant curve. A likely cause of both garnet growth and melting during decompression is the reaction
666
J. P. P L AT T E T A L .
Bt þ Sill þ Qtz ¼ Crd þ Kfs þ L
(3)
which has a positive dP/dT (Johnson et al. 2001) and would be crossed by the decompression path for the garnet gneiss at about 650 MPa. The study by Johnson et al. (2001) provides Fe–Mg isopleths for garnet composition in the P–T region of interest, which are essentially isothermal in the Grt–Bt–Sill field, but change abruptly to a low negative dP/dT slope in the Grt–Crd– Sill field. Garnet Fe–Mg zoning profiles in the garnet gneiss B203 are flat except for a narrow outermost rim where Fe/(Fe þ Mg) increases markedly. X(Fe) values in the sillimanite-bearing garnet rims imply temperatures of c. 820 8C, in good agreement with our estimates for peak temperature from the mafic boudin. The flat Fe/Mg profiles could result either from peak-T homogenization, or from garnet growth parallel to Fe–Mg isopleths in the Grt–Crd–Sill divariant field until the extreme outer rim where the P–T path cuts across Fe–Mg isopleths. In either case the profiles appear to imply rising T during decompression at least until the sillimanite-bearing rims had formed, at P ,650 MPa. We have an extensive radiometric dataset from the garnet gneiss, summarized in Table 4. Zircons have small (c. 50 ìm) inherited cores with a wide range of Precambrian to Palaeozoic ages, and broad metamorphic overgrowths that give a very precise ion-microprobe U–Pb mean age of 21.2 0.3 Ma (Platt & Whitehouse 1999; Platt et al. 2003b). The garnet cores contain only inherited zircons, whereas zircons from the sillimanitebearing garnet rims have overgrowths, as do those in the matrix. Rare earth element distributions in the zircons and garnets strongly suggest that the zircon overgrowths formed in equilibrium with garnet (Whitehouse & Platt 2003). We conclude that the zircon rim age corresponds to the latest stages of garnet growth, after decompression into the sillimanite field, and at the time of peak temperature and partial melting (Fig. 6). Our new Ar/Ar and zircon fission-track ages for the garnet gneiss all lie close to 21 Ma (Tables 5 and 6), indicating that cooling to less than 300 8C was extremely rapid, occurring over less than 1 Ma. The minimum rate of cooling that is consistent with the age determinations is over 600 8C Ma1 , which is comparable with values estimated by Zeck et al. (1992), Monie´ et al. (1994) and Andriessen & Zeck (1996). More importantly, perhaps, a minimum exhumation rate during the final stages of the process is constrained by the zircon U–Pb age, which corresponds to a pressure of at least 400 MPa (Fig. 6), and hence constrains the exhumation rate to be at least 15 km Ma1 . Sillimanite gneiss and migmatite. The main assemblage in these rocks (Qtz–Pl–Kfs–Sil–Bt–Crd–Ilm) formed under low-P and high-T conditions (640 40 8C, 400 MPa) towards the end of the decompressional stage. Relict kyanite, garnet and rutile provide evidence for an earlier high-P stage, but equilibrium assemblages suitable for thermobarometry are not preserved. Radiometric ages for these rocks are comparable with those from the garnet gneiss (Table 4), and indicate similar rates of exhumation and cooling. Fibrolite schist. Medium-grade schist with an early Grt–St–Ky– Pl–Rt assemblage, a Sill–Bt-bearing main foliation and late andalusite porphyroblasts provides a clearly defined P–T path (G–J in Fig. 6) showing decompression from around 1000 MPa to ,350 MPa, at constant or slightly rising temperature (Argles et al. 1999). Ar/Ar and zircon fission-track ages close to 21 Ma (Table 4) indicate cooling at essentially the same time as the higher-grade rocks.
Andalusite schist and quartzite. Graphitic mica schist with an early assemblage of Grt–St–Ky, overprinted by late andalusite and biotite porphyroblasts, provides a decompressional P–T path (K–M in Fig. 6) from around 900 MPa to ,400 MPa, at constant or slightly rising temperature (Argles et al. 1999). These rocks are overlain by a sequence of quartzites with kyanite-bearing and locally andalusite-bearing veins. Ar/Ar and fission-track ages indicate cooling a little before 21 Ma (Table 4), perhaps very slightly earlier than the highest-grade rocks. Phyllite. Blue–grey phyllites of probable Permo-Triassic age from this section have not provided useful P–T data, although some are kyanite bearing. Similar rocks in the Jubrique section west of the Sierra Bermeja preserve structures interpreted by Balanya´ et al. (1997) as pseudomorphs after carpholite; and carpholite, chloritoid and locally kyanite have been reported from low-grade phyllites elsewhere in the Internal Betic Cordillera and Internal Rif (Goffe´ et al. 1989; Azan˜o´n & Goffe´ 1997; Michard et al. 1997). This suggests peak pressure of around 700–800 MPa for temperatures in the range 200–350 8C. Ar/Ar crystallization ages from white micas in low-grade phyllites in the eastern and central Betics lie in the range 30–50 Ma (Platt et al. in press), whereas cooling ages from rocks that have been above the closure temperature for Ar in white mica almost invariably lie in the range 19–22 Ma (Monie´ et al. 1994). Low-grade to unmetamorphosed rocks. Largely unmetamorphosed rocks of Late Palaeozoic to Cretaceous age lie at the top of the structural sequence. These have traditionally been assigned to a separate nappe complex (Malaguide) from the remainder of the sequence, and there is no doubt that they must originally have formed an assemblage of thrust slices above the metamorphic rocks below, as there is a duplication of stratigraphy (e.g. Devonian limestones above kyanite-bearing phyllites of presumed Permo-Triassic age). Thrust stacking, however, predated the extensional thinning that produced the present structure, and we therefore treat these unmetamorphosed rocks as the top of the pre-extensional orogenic sequence. The present-day lower contact appears to be extensional in origin (Tubı´a et al. 1993), as the rocks rest on phyllites that were metamorphosed at .700 MPa. Locally, thin slices of rocks showing incipient metamorphism are present along the contact, representing relics of the missing part of the original metamorphic section. To gain some insight into the thermal history of the missing part of the section, we have examined the much thicker sequence assigned to the Malaguide Complex in the vicinity of Malaga (Fig. 4). This area lies immediately eastwards of Carratraca, and hence approximately in the displacement direction during the motion on the detachment. Zircon and apatite fission-track data from both these bodies of Malaguide rocks are presented in Table 6 and are discussed in more detail in the Supplementary Publication. Zircon fission-track analyses from Palaeozoic greywacke sandstone at the top of the Carratraca sequence indicate a Late Carboniferous age, and probably reflect post-Variscan cooling (Table 6). Apatite fission-track analyses indicate latest Oligocene to earliest Miocene ages, and are likely to reflect the same exhumational event that controls the cooling ages of the metamorphic rocks below. Interestingly, the lowest thrust slice in this sequence gives a zircon fission-track age in the same range, indicating that this rock was buried sufficiently deeply to be heated above the zircon fission-track annealing window (about 310 8C), and then exhumed and cooled in the early Miocene (see
Rock type
Phyllite and quartzite Late Pal. greywacke
E. Pal clastics
Permo-Triassic sandstone Permo-Triassic sandstone Late Pal. greywacke
MM16 MM18
MM19
MM20 MM21 MM22
9 19 25 14 13 15 7 11 18 14 16 25 21 22 20 26 23 21 5 24 22 20 13 30 20 20 32 21 22 48 35 20 30 40 18 19 24 25 25 26 12
A Z A Z Z A A Z A A Z A Z A A A A A A A Z A Z Z A A A A A Z Z A Z Z A Z A A A Z A
nc
4.255 1.085 1.097 1.188 1.085 1.085 4.269 4.279 1.097 4.284 4.294 1.085 4.303 1.085 1.085 1.085 4.323 1.085
1.241 0.423 1.142 0.423 0.423 0.124 1.510 0.423 1.064 1.064 0.424 1.064 0.425 1.064 1.510 1.161 1.161 1.161 1.161 1.510 0.423 1.161 0.423
rd
3039 4509 4560 4938 6015 6015 3039 3039 4560 3039 3039 6015 3039 6015 6015 6015 3039 6015
6882 3039 6331 3520 3039 6882 8347 3520 5091 5091 3039 5091 1340 5091 8347 6429 6429 6429 6429 8347 3520 6429 3520
nd
Dosimeter
2.611 2.172 1.710 2.365 0.311 0.404 2.772 8.730 2.283 2.213 11.92 0.325 14.78 0.379 1.801 2.941 16.95 1.239
0.655 13.72 0.287 2.221 2.967 1.235 0.197 3.946 0.158 0.431 4.911 0.244 3.300 0.147 0.206 0.232 0.271 0.110 0.066 0.053 2.900 0.319 4.669
rs
2139 376 284 1967 249 331 3500 4392 425 954 7750 177 3277 371 1585 2478 4629 309
158 2871 262 474 514 63 72 700 188 175 1307 335 1340 198 328 258 361 193 23 98 1495 215 906
ns
Spontaneous
3.112 3.308 2.235 1.734 2.800 3.002 3.360 1.365 2.875 2.747 0.998 3.341 1.420 3.313 1.833 2.888 1.286 1.571
6.344 1.290 2.234 3.125 4.346 1.429 2.408 4.864 1.707 4.410 6.227 2.495 4.061 1.783 3.061 2.039 2.609 1.034 0.681 0.823 3.534 2.667 5.664
ri ni
2549 5726 3711 1442 2243 2459 4242 687 5351 1184 649 1820 315 3246 1613 2434 351 392
1530 270 2037 667 753 729 881 863 2061 1791 1657 3423 1649 2395 4871 2272 3475 1814 236 1603 1822 1800 1099
Induced
15 ,1 60 50 3 0 ,1 ,1 60 ,1 ,1 15 6 ,1 ,1 ,1 9 ,1
80 50 40 ,1 ,1 40 95 ,1 30 ,1 10 40 ,1 75 77 50 60 30 7 6 10 60 70
P÷2
7.8 25.0 4.9 7.2 24.0 75.0 14.6 20.9 0.1 21.7 22.0 21.1 17.4 22.2 24.7 10.8 17.9 29.0
0 7.8 11.5 16.8 26.4 18.5 0 29.3 11.4 28.8 10.2 13.6 25.1 0.03 0.2 5.6 0.7 16.5 0 25.5 8.5 0.9 0
RE (%)
Age dispersion
22.7 0.9 13.4 1.1 15.1 0.9 269 11 .20.9 1.9 24.7 1.5 22.1 0.8 170 10 15.7 0.8 21.9 1.4 315 18 17.5 1.7 278 21 22.5 1.7 185 12 189 7 355 25 150 18
22 2 280 19 25 2 18.9 1.5 19 2 18 3 20.8 2.6 21.1 2.2 16.4 1.4 19.2 2.3 21 1 17.8 1.2 22 2 14.9 1.1 17.2 1.0 22.4 1.5 20.4 1.2 21.7 1.9 19.2 4.2 15.9 1.9 21.4 0.9 23.5 1.7 21.6 1.0
Central age (Ma) 1ó
1.21 1.40 1.54 1.48 1.13
1.9
100 73 75 36 90
1.16 2.61 1.73 0.94
11.56 0.28
1.25 13.51 0.21 9.70 0.26 10.89 0.17
14.51 0.19
12
33 98 100
46
Triass þ Cretaceous components 14.48 0.19 1.01 30
14.33 0.12 14.46 0.17 12.34 0.18 14.06 0.25 14.28 0.12
14.18 0.37
108
61
1.44
13.90 0.14
10 17
1.27
1.46 1.82
13.94 0.49 12.52 0.46
7
91 41 24 26 20
0.80
13.80 0.33 No lengths
49
10
nt
1.36 1.42 1.14 1.12 1.22
1.48
14.14 0.21
13.60 0.14 14.52 0.23 14.77 0.24 14.62 0.22 15.05 0.28 No confined tracks 14.75 0.16
1.32
SD
13.33 0.44
Mean track length (ìm)
A, apatite; Z, zircon; nc , number of crystals; nt , number of tracks. Track densities are 3 106 tracks cm2 . Analyses are by external detector method using 0.5 for the 4ð/2ð geometry correction factor. P÷2 is probability for obtaining ÷2 value for v degrees of freedom, where v ¼ number of crystals 1. RE%: random effects model standard deviation (Galbraith & Laslett 1993). Further analytical details are given in the Supplementary Publication.
Phyllite and quartzite E. Pal clastics
Late Pal. greywacke Permo-Triassic sandstone Permo-Triassic sandstone
MM9 MM10 MM11/2 MM12 MM14 MM15
Carratraca crustal sequence CAR11 Greywacke CAR9 Greywacke CAR9 PB436 Psammite CAR5 Grt–mica schist CAR5 PB437 Quartzite PB437 CAR14 Quartzite CAR16 And quartzite CAR17 Fibrolite schist CAR17 CAR19 Sill gneiss CAR19 PB439 Sill gneiss PB379 Migmatitic gneiss PB380 Leucosome PB381 Garnet gneiss PB383 Garnet gneiss PB383b Garnet gneiss PB383b PB384 Granitic gneiss PB384b Granitic gneiss Montes de Malage area MM3/1 Phyllite and quartzite
Sample number
Table 6. Fission-track analyses for the Carratraca crustal sequence and the Montes de Malaga area
E X H U M AT I O N O F T H E RO N DA P E R I D OT I T E 667
668
J. P. P L AT T E T A L .
the Supplementary Publication for more detailed discussion of the fission-track data). The Malaguide Complex in the Malaga area can be divided into three units on the basis of lithostratigraphy and thermal evolution (Fig. 4). The lowest, which lies in contact with medium-grade metamorphic rocks comparable with the andalusite schists of the Carratraca area, consists of dark grey graphitic phyllite and quartzite of probable early Palaeozoic age, with a well-developed cleavage defined by fine-grained metamorphic white mica. The unit is cut by numerous basaltic dykes of late Oligocene age (Platzman et al. 2000). Post-cleavage andalusite and biotite are locally developed, indicating metamorphic temperatures in excess of 400 8C. Andalusite-bearing phyllite from this unit has yielded an Ar/Ar age of 18.6 1.8 Ma (Table 5). Zircon fission-track analyses indicate cooling in the early Miocene (Tables 3 and 6). The middle unit of the Malaguide Complex consists of Late Palaeozoic clastic and calcareous sediments, with occasional Late Oligocene dykes (Platzman et al. 2000). These have a good slaty cleavage in pelitic rocks and a well-developed crenulation cleavage in places. Zircon fission-track analyses indicate ages in the range 170–355 Ma, which is likely to indicate variable degrees of partial resetting of Palaeozoic detrital or cooling ages. Apatite fission-track analyses reflect cooling in the early Miocene. Alpine metamorphic temperatures therefore lay between the top of the apatite and the top of the zircon fission-track annealing windows (120–310 8C). The upper unit consists of red continental clastic rocks of presumed Permo-Triassic age, overlain by Triassic carbonate rocks. With one exception, apatite fission-track analyses give Permian to Jurassic ages, suggesting partial resetting of late Palaeozoic detrital ages. These rocks have therefore not been exposed to temperatures in excess of the top of the apatite fission-track annealing window (c. 120 8C). In view of the distinct thermal histories of these three units, we propose that they are now separated from each other by extensional detachments of comparable age to those identified in the Carratraca region. These rocks may therefore represent the upper part of the vertically thinned orogenic pile, now partly excised along the detachment in the Carratraca region.
Thermal modelling of the P – T – t array The P–T histories of the Ronda peridotite and the overlying crustal rocks pose some important questions that bear on the tectonic evolution of the entire Betic–Rif orogen, and on the process of late orogenic extension more generally. The rocks at all crustal levels show evidence for decompression along paths that were either isothermal or involved a small degree of heating, and the onset of cooling occurs only at very shallow levels (10– 15 km depth) towards the end of their exhumation history. Both phenomena require a significant heat source during exhumation: without additional heat input, cooling will start at depths of 20– 25 km even at very high rates of exhumation (see below). Given the lack of evidence for significant magmatism in the area, the likely heat sources are either radiogenic, or elevated heat flux from asthenospheric mantle emplaced at a relatively shallow level beneath the orogen. The first question is therefore to distinguish between these two possible heat sources. If the heat source is asthenospheric, the next question is to determine at what level it was emplaced. Finally, the rate, timing and mechanism of exhumation need to be determined. We have carried out thermal modelling of the P–T–t array, with the idea that any explanation for the thermal and exhuma-
tional history of the peridotites and their crustal envelope should satisfy the evidence from all the exposed structural levels. In principle, this should provide a more rigorous test of rates of exhumation and proposed heat sources than modelling a single P–T path. The models are simple 1D thermal calculations. There is insufficient evidence about the pre-exhumational geometry of the system to justify more elaborate 2D or 3D models, and given the short vertical length scale (a few tens of kilometres) particularly during the critical late stages of the exhumational history, 1D models are likely to be a good approximation to reality. Ruppel et al. (1988) have also shown that the difference in P–T paths calculated using 1D and 2D solutions for the case of asymmetric extension of the lithosphere along low-angle faults amounts to only a few per cent. The modelling approach is similar to that presented and investigated in some detail to address comparable questions at Ocean Drilling Program (ODP) Site 976 in the Alboran Sea by Platt et al. (1998). We make the following assumptions as boundary conditions for the calculations. (1) Thickened orogenic crust and lithosphere was produced by contractional deformation of the Alboran Domain in the Early Tertiary. A post-contractional crustal thickness of 55 km, corresponding to a pressure of 1457 MPa in crust with an average density of 2700 kg m3 , is assumed; this is based on the petrological evidence for metamorphism of now-exposed crustal rocks at pressures up to around 1500 MPa at various locations in the Betic Cordillera (Go´mez-Pugnaire & Ferna´ndez-Soler 1987;Tubı´a & Gil-Ibarguchi 1991; Puga et al. 1999). The timing of this event is poorly constrained, but available radiometric and stratigraphic data suggest a late Eocene age (Monie´ et al. 1991; Vissers et al. 1995). Recent arguments that Miocene U–Pb ages on zircon relate to subduction events (Sa´nchez-Rodrı´guez & Gebauer 2000;Sa´nchez-Vizcaı´no et al. 2001) are difficult to sustain: the geological evidence clearly indicates that they relate to stages in the exhumational history of the Alboran Domain. Zircon growth at this time was likely to have been a result of partial melting or of decompression reactions during exhumation. (2) The thermal gradient after crustal thickening is likely to have been around 10 8C km1 (2.65 MPa 8C 1 ), based on thermobarometry on early high P/T ratio rocks preserved in the orogen (Go´mez Pugnaire & Camara 1990; Azan˜o´n & Goffe´ 1997; Puga et al. 1999). The thermal gradient subsequently increased, probably as a result of radiometric heating: as discussed below, the pre-exhumational P–T conditions probably lay on a gradient of around 14 8C km1 (1.9 MPa 8C 1 ). (3) We have assumed that the principal process of exhumation was continuous stretching of the whole orogenic crust. This is constrained by the geological evidence that most levels in the original orogenic crust are still present, albeit in attenuated form, and that the subcontinental mantle was exhumed. It is also supported by the fact that the crust of the Alboran Domain is now around 15 km thick on average, and that it subsided below sea level shortly after extension began. Although significant detachments are present (e.g. the basal contact of the Malaguide complex), most of the exhumation was achieved by distributed extension. Deformation was non-coaxial, with top-to-the-east or ENE sense of shear (Tubı´a et al. 1993; Argles et al. 1999), but this does not affect the calculations. We have modelled this by applying a vertical stretch of one-third, which reduces the crustal thickness to 18 km (the reason for choosing this value is discussed below). Continuous deformation does not bring rocks to the Earth’s surface, so the final stages of exhumation were modelled by unroofing, which we envisage to have occurred by some combination of erosion and normal faulting in the brittle
E X H U M AT I O N O F T H E RO N DA P E R I D OT I T E
upper crust. By far the largest discontinuity within the section is that between the unmetamorphosed Malaguide section at the top and underlying kyanite-bearing phyllite and quartzite. This could correspond to about 25 km of missing section. Slip on this detachment must, however, have largely or completely post-dated the distributed stretching phase. Had it taken place earlier, the underlying rocks would have cooled rapidly during the early stages of decompression, which is not observed. The decompression path (isothermal or with slight heating to pressures as low as 300 MPa, in the andalusite stability field) requires that the detachment became active only after the footwall rocks had risen to a depth of about 12 km. As the footwall rocks originated at a depth of about 35 km, this justifies our choice of a vertical stretch of one-third for the ductile stretching phase. We have therefore modelled the unroofing stage by slip on a detachment initiated at 10 km depth, followed by erosion once the detachment footwall has been brought to the surface. The rate of exhumation during the unroofing stage is tightly constrained: as discussed above, the garnet gneiss travelled from at least 15 km depth to near-surface conditions in about 1 Ma. The first question we address is the heat source during exhumation. Calculations of the possible level and timing of radiogenic heating in the post-contractional crust are shown in Figure 8. A period of 70 Ma is allowed between crustal thickening and final exhumation: this implies a Late Cretaceous (90 Ma) age for the contractional event, which is as early as the geological evidence allows. Radiogenic heat production in the crust is set at the comparatively high level of 4.0 3 1010 W kg1 (1.08 3 106 W m3 ), appropriate for an orogenic crust with a high content of pelitic sediment. For each calculation we show the P–T and T–t paths (Figs 7 and 8) for rocks at initial depths (pressures) of 37.5 km (992 MPa), 40.25 km (1067 MPa), 50.9 km (1348 MPa) and 66.25 km (1821 MPa), representing the andalusite schist, fibrolite schist, garnet gneiss and peridotite, respectively. These are maximum reasonable depths for each point on the array, and hence maximize the effect of radiogenic heating. Figure 8a shows the effect of continuous very slow exhumation (by stretching followed by unroofing as described above), over the whole 70 Ma period. The rocks show heating during exhumation, as observed, but reach their maximum temperatures fairly early in the exhumation history, and the maximum temperatures reached are in general significantly below observed values. This is most marked for the mantle peridotites, which reach a maximum temperature of 770 8C, far below the observed value of about 1200 8C. The temperature– time plot (Fig. 8d) shows how this model also fails to satisfy the radiometric data. Figure 8b shows the effect of very rapid exhumation after a long post-contractional pause during which radiogenic heating allows the crustal rocks to warm up. By the end of the pause, model temperatures in the crust are similar to or even exceed observed values, but there is no heating during exhumation because of the short time scale involved. Mantle temperatures are well below observed values. Even at the high rate of exhumation applied (average 9 km Ma1 at the level of the garnet gneiss), the rocks start to cool once the pressure drops below about 600 MPa (22 km depth). Figure 8c involves a 30 Ma pause followed by slow exhumation during the remaining 40 Ma of the post-contractional period. The pause allows crustal rocks to rise to sufficiently high temperature that heating during exhumation brings them close to observed values. The relatively slow rate of exhumation, which is necessary to allow radiogenic heating during decompression, also results in an early onset of cooling. As a result, peak
669
temperatures occur at much higher pressures than is actually observed, and the model fails to satisfy the radiometric data (Fig. 8d). These three calculations clearly demonstrate that radiogenic heating alone cannot explain critical features of the observed P– T paths, in particular the tendency to reach maximum temperature at very shallow depths, and the very late onset of cooling. This conclusion does not conflict with those of other workers (e.g. England & Thompson 1984; Ruppel & Hodges 1994; Huerta et al. 1999), who have modelled P–T paths with heating during decompression caused by radiogenic heat production. In the case of the Carratraca area (and in the Alboran Sea core modelled by Platt et al. 1998), heating continues until exhumation has brought the rocks to very shallow depths (10–15 km). In no case has radiogenic heating been shown to produce this effect. It should be noted that Platt et al. (1998) showed that this conclusion holds for a wide range of possible values of radiogenic heat production. In Figure 9 we show the results of investigating the effect of removing the lithospheric mantle at various depths and replacing it by asthenosphere at a temperature of 1300 8C, followed immediately by rapid extensional exhumation. A 31.4 Ma postcontractional pause is allowed for radiogenic heating, at the end of which the lithospheric mantle is removed below a depth of 67.5, 83 or 101 km. Lithospheric removal at 67.5 km depth reproduces both the maximum temperatures and the shapes of the P–T paths for the various elements in the array fairly well, although model temperatures in the mantle are still rather low (this is discussed further below). Lithospheric removal at depths greater than 67.5 km fails to produce significant syn-exhumational heating at crustal levels, and temperatures reached at the modelled depth for the Ronda peridotite are far too low. We conclude from these experiments that conductive heating from asthenosphere emplaced at 67.5 km depth provides a viable explanation for the overall pattern of heating during exhumation. Lithospheric removal at any shallower depth than this is ruled out by the preservation within the system of lithospheric mantle from 66 km depth. The final question to be addressed concerns the rate of exhumation. Changing the rate of exhumation after lithospheric removal changes both the amount and the timing of the synexhumational heating, and hence the shape of the decompressional P–T path, but these are also affected by the pre-exhumational depth of the rock. Peak pressures, and hence preexhumational depths, are rather poorly constrained, however, as the assemblages are poorly preserved. For the purposes of the modelling we have assumed a linear thermal gradient of 14 8C km1 at the start of exhumation through the crust and upper mantle (Fig. 10). This gradient is reasonably consistent with the effects of radiogenic heating for 31.4 Ma in a thickened crust with a post-contractional linear gradient of 10 8C km1 and thermal productivity of 4.0 3 1010 W kg1 (Fig. 9). It assigns somewhat higher peak pressures for the three crustal points in the array than indicated by the thermobarometry, and a preexhumational temperature to the peridotite that is within the observational uncertainties, but may be on the high side for lithospheric mantle at 66 km depth. The reasons for choosing this gradient are, first, that it is consistent with P–T conditions at the start of exhumation in Alpujarride schists elsewhere in the Betic Cordillera (e.g. Azan˜o´n & Alonso-Chaves 1996; Garcı´a-Casco & Torres-Rolda´n 1996; Tubı´a et al. 1997), and, second, it provides the most conservative approach to the modelling. Syn-exhumational heating by conduction from below is favoured by higher initial depths and lower initial temperatures, as these both
670 MPa 1800
J. P. P L AT T E T A L . 600
800
1000
T°C
MPa 1800
A
600
800
1000
B
T°C
MPa 1800
1600
1600
1600
1400
1400
1400
1200
1200
1200
1000
1000
1000
800
800
800
600
600
600
400
400
400
200
200
200
1000 T°C 900
D
U-Pb zircon
800
B
700 600
C A
500
hornblende Ar-Ar
400
biotite Ar-Ar zircon FT
300 200
apatite
100 time Ma 0
increase the thermal gradient between the rock and the heat source. Choosing an initial thermal gradient of 14 8C km1 assigns the highest reasonable initial pressures (for the determined starting temperatures) to the P–T array, and hence favours the process we are trying to test. The amount of syn-exhumational heating is more sensitive to the pre-exhumational depth of the rock than it is to the rate of exhumation, whereas the timing of the heating is more sensitive to the rate of exhumation. We have taken advantage of this to investigate the effect of our choice of initial P–T conditions for the various elements of the model array, so as to ensure that these choices do not influence our study of the rate of exhumation. Figure 10 presents calculated P–T paths for a range of preexhumational depths along the 14 8C km1 gradient, all for the same average rate of exhumation during the ductile stretching phase (9 km Ma1 at the level of the garnet gneiss). Based on these experiments, we have chosen for each crustal level the preexhumational depth that best matches the observed peak temperatures and amount of syn-exhumational heating. These depths and associated P–T points were then used for the investigation of the rate of exhumation, as follows: 36.8 km, 974 MPa, 520 8C (andalusite schist); 41.0 km, 1086 MPa, 580 8C (fibrolite schist); 50.9 km, 1348 MPa, 720 8C (garnet gneiss). The starting point for the peridotite is taken as 67.5 km, 1790 MPa, 920 8C. The investigation of the rate of exhumation is presented as P– T and T–t plots in Figures 11 and 12. Six calculations were carried out, four of which are presented, showing the results of exhuming the garnet gneiss over periods ranging from 9.4 to
600
800
1000
T°C
C
Fig. 8. Model P–T and T–t paths for post-contractional radiogenic heating during exhumation. Thickened crust (55 km), with a post-contractional thermal gradient of 10 8C km1 , and radiogenic heat production of 4.0 3 1010 W kg1 , is assumed. P–T paths are shown for post-contractional depths of 37.5 km, 40.25 km, 50.9 km and 66.25 km. Observed P–T conditions from Figure 6 for the four crustal and mantle depths are shown in grey for comparison. (a) Continuous slow exhumation (by stretching followed by unroofing as described in the text), over 70 Ma (Run 221). (b) Post-contractional pause of 60 Ma, followed by rapid exhumation (Run 222). (c) Pause of 30 Ma followed by slow exhumation over 40 Ma (Run 223). (d) T–t paths for 50.9 km initial depth for scenarios shown in (a)– (c). Radiometric data for the garnet gneiss are shown in grey. These scenarios produce P–T paths with the temperature peak and the onset of cooling much earlier than observed. Scenarios shown in (a) and (c) cannot be fitted to the observed radiometric data. Peclet numbers in these experiments were 10 (a), 200 (b) and 20 (c) (see text).
3.5 Ma. The model parameters are shown in Table 7. The exhumation rate is controlled in the model by the Peclet number, which is the non-dimensional vertical velocity at the base of the 125 km thick stretching layer in units of thickness per characteristic thermal time for the layer. Exhumation rates vary with depth, and also with time during homogeneous stretching at a constant strain rate: the exhumation rates for the garnet gneiss indicated in Table 7 are average values. The calculations are identified by the Peclet number in Figures 11 and 12. As noted above, none of the calculations we have carried out can reproduce the temperature of 1200 8C determined by Lenoir et al. (2001) from the recrystallization front in the Ronda peridotite (B in Fig. 6). The reason for this is simple: the temperature produced by conductive equilibration at the interface between two large bodies is close to the average of their two temperatures. If asthenospheric mantle at 1300 8C is brought into contact with lithosphere at a temperature of 900 8C, the temperature at the boundary will not exceed 1100 8C. If we accept the thermobarometric data of Van der Wal & Vissers (1993) and Lenoir et al. (2001), some additional heat transfer, perhaps by percolating melt (Van der Wal & Bodinier 1996), may be required to explain the observed conditions. The calculations do, however, provide at least a qualitative explanation for heating during exhumation in the mantle rocks. It should be noted that the P–T path at this level is not strongly sensitive to the exhumation rate. Tighter constraints on the rate of exhumation come from the crustal rocks. At a Peclet number of Pe ¼ 75 (average exhuma-
E X H U M AT I O N O F T H E RO N DA P E R I D OT I T E
67.5 km
700
800
900
1000 1100
T˚C
83 km 101 km MPa 1400 1200 1000 800 600 400 200
Fig. 9. Model P–T paths produced for a 30 Ma post-contractional pause, followed by convective removal of lithosphere at depths of 67.5, 83 or 101 km, followed by rapid extensional exhumation (Pe ¼ 200). Only lithospheric removal at a the shallowest depths is capable of producing P–T paths comparable with those observed in the Carratraca area. Runs 226–228.
tion rate 4.1 km Ma1 at the level of the garnet gneiss) peak temperature and the onset of cooling occur too early on the exhumation path at all crustal levels (Fig. 11). Examination of the P–T and T–t curves for the level of the garnet gneiss shows that cooling starts at a depth of 700 MPa, 3 Ma before the time of zircon growth (Figs 11 and 12). This is inconsistent with the evidence that the zircon grew during the formation of the sillimanite-bearing garnet rims, at a time of rising temperature (see above). Pe ¼ 125 (average exhumation rate 6.4 km Ma1 ) produces a better fit to the shape of the P–T curves, but cooling at the level of the garnet gneiss still starts too early, at 22 Ma. At Pe ¼ 200 (average exhumation rate 9.2 km Ma1 ) the P–T curves fit the data well, and the peak temperature in the garnet gneiss occurs within the uncertainty of the zircon age. Analysis of the T–t curves at higher Peclet numbers is difficult because of the very short time scales involved. Conclusions about the timing of the temperature peak are very sensitive to the way the curves are fitted to the radiometric data in Figure 12. The very high rate of denudation, and the mode of denudation by slip on a detachment fault, dictate that the cooling paths for the three crustal levels are very steep and close together. These curves cannot be fitted perfectly to the radiometric data. The most obvious outlier is the apatite fission-track dataset, for which the mean of the whole dataset is shown in Figure 12. The individual sample ages are variable, and it is likely that exhumation had slowed significantly by the time the rocks reached a few kilometres below the level of the present surface, and that the different sample ages reflect the topography developed at that
500 500
600
700
800
900
1000 1100
T˚C
MPa 1800
MPa 1800
1600
1600
1400
1400
garnet gneiss
1200
fibrolite schist andalusite schist
671
600
700
800
900
1000 1100
T˚C
Pe 75 125 200 400
1200 1000
1000
800
800
600
600
400
400
200
200
Fig. 10. Model P–T paths calculated for a range of pre-exhumational depths along the 14 8C km1 gradient for the three crustal points in the array, following convective removal of lithosphere at 67.5 km depth and rapid exhumation (Pe ¼ 200). This provided the basis for the choice of starting depths and P–T conditions in the experiments shown in Figure 11. Runs 210–212.
Fig. 11. Model P–T paths during exhumation for different initial depths in the orogenic lithosphere, compared with thermobarometric data from the peridotites and their crustal envelope (from Fig. 6). The curves are identified by the Peclet number during stretching (see Table 6). Inflections in the curves mark the end of the stretching phase, and the onset of the unroofing stage, during which exhumation was achieved by slip on a detachment initiated at 10 km depth, followed by erosion, at a rate of 15 km Ma1 . (See text for details and discussion.) Runs 300–306.
672
J. P. P L AT T E T A L .
the thermal peak and final exhumation for Pe ¼ 200 (1.2 Ma). Hence the timing of the thermal peak cannot be used to judge the viability of Pe values .200. The amount of syn-exhumational heating decreases at high Pe values, and this provides some indication of an upper limit to Pe. At Pe ¼ 400, the amount of heating is less than the observed values at all crustal levels. Based on all these considerations, our best estimate for Pe is 200 (þ100/50), corresponding to an exhumation rate of about 9 km Ma1 for the garnet gneiss, a total exhumation time of 5 1 Ma, starting at c. 25 Ma, and a strain rate around 13 3 1015 s1 . This strain rate compares well with estimates by Hanne et al. (2003) based on the inversion of subsidence data from the Miocene basins in the Alboran region.
Discussion and conclusions
Fig. 12. Modelled T–t paths during exhumation for crustal rocks, compared with radiometric data. The curves are identified by the Peclet number during the stretching phase (see Table 6). Exhumation rate during final cooling is constrained to be 15 km Ma1 (see text). The biotite age shown for the garnet gneiss is the mean of all analyses from the three garnet gneiss and sillimanite gneiss samples. The fission-track ages are means of all the data for the three metamorphic zones (20.9 1.2 Ma for zircon, 18.3 1.4 Ma for apatite), and may underestimate the true variance.
Table 7. Durations, strain rates and exhumation rates for the exhumation models Peclet number 75 100 125 150 200 300 400
Time (Ma) 9.4 7.4 6.5 5.8 4.85 3.96 3.52
Strain rate (s1 ) 4.8 3 1015 6.4 3 1015 8.0 3 1015 9.6 3 1015 12.8 3 1015 19.2 3 1015 25.5 3 1015
Exhumation rate (km Ma1 ) 4.1 5.4 6.4 7.3 9.2 12.1 14.3
The Peclet number is the non-dimensional variable used in the model calculations to determine the stretching rate: it dictates the vertical velocity at the base in units of thickness per characteristic thermal time for the 125 km thick stretching layer. The time column gives the total time for exhumation of the garnet gneisses from 51 km to the surface: this includes 1.14 Ma for the denudation stage. The strain rate is the average extensional strain rate for the garnet gneisses during the ductile stretching stage: it should be noted that this value changes with both time and initial depth. The exhumation rate is the average value for the garnet gneiss during the ductile stretching stage: during the denudation stage the rate is fixed at 15 km Ma1 .
time. The other data point that cannot easily be fitted is the Ar/ Ar age from the andalusite schist zone. The fit shown in Figure 12, which brings the garnet gneiss close to the surface at 20.4 Ma, is most consistent with the data; but earlier dates for final exhumation up to 21.0 Ma would also be compatible with the Ar/Ar and fission-track data. This uncertainty is comparable with the duration of the denudation stage (1.1 Ma for the garnet gneiss), and is also comparable with the time difference between
Syn-exhumational heating and the late onset of cooling at all levels in the orogenic crust and upper lithospheric mantle around the Ronda peridotite is most easily explained in terms of conductive heating from asthenospheric mantle emplaced at about 67.5 km depth. Radiogenic heating cannot reproduce the observed shape and timing of the P–T paths, in particular the very late timing and shallow depth at which the thermal peak was attained. The lack of igneous intrusions of significant size in the area eliminates magmatic heat transfer as a mechanism, although melt percolation may have contributed to the very high peak temperature reached within the peridotite massif itself. Similar conclusions were reached from a study of crustal rocks retrieved from ODP Site 976 in the Alboran Sea by Platt et al. (1998), who attributed the strong syn-exhumational heating to delamination of the lithosphere at the base of the orogenic crust. They pointed out that this could be achieved by convective removal of lithosphere (Houseman & Molnar 1997) assuming a wet olivine rheology, if Moho temperatures reached 750 8C. In the Ronda area, the preservation of lithospheric mantle within the system indicates that lithospheric removal did not take place along the base of the crust, so delamination in the sense of Bird (1979) may not be the most appropriate concept to describe this process. Our estimated duration for the exhumation process in the Ronda area (5 Ma) is significantly less than the 9 Ma estimate for the Alboran Sea drill core (Platt et al. 1998), and the similar estimate for the Carratraca area based solely on the zircon U–Pb age (Platt & Whitehouse 1999). In both cases the estimates were based on more limited information than we now have available: no U–Pb zircon ages could be obtained from the Alboran Sea core, and we now have considerable additional evidence both on the timing of zircon growth in the Carratraca area and on the cooling ages from all the levels in the crustal section. There may be some genuine differences in the detailed exhumation history of different parts of the Alboran Domain, however; this is suggested by the slightly younger cooling ages (18–20 Ma) obtained from the Alboran Sea core and other areas around the Alboran Sea (Platt et al. 2003b). Early Miocene exhumation in the Alboran Domain was achieved primarily by tectonic processes rather than erosion. This is demonstrated, first, by the fact that the original overburden is still preserved in the Carratraca area in a highly attenuated state: it was thinned, not stripped off. Second, the later stages of exhumation coincide with the initiation of sedimentation and rapid subsidence in the Alboran Sea basin: the surface of the Alboran Domain now lies 7 km or more beneath sea level in places (Comas et al. 1999), and the Alboran Sea core itself was recovered from 2 km below sea level. Sediments of late Aquita-
E X H U M AT I O N O F T H E RO N DA P E R I D OT I T E
673
Fig. 13. Schematic illustrations of postulated exhumation history for Ronda peridotite and its crustal envelope. All drawings are to scale, with north or NE to left. Dashed pattern, mantle peridotite; grey, crustal rocks. Faults active in each stage are shown with bold line, and with grey dash– dot line in previous stage. (a) Situation at 25 Ma, after crustal thickening and removal of lithospheric mantle below 67.5 km depth. Letters and symbols show positions at this stage of rocks now seen at the surface: M1, M2, M3, extensional allochthons in the Malaguide Complex near Malaga (see text); A, F, G, andalusite schist, fibrolite schist, and garnet gneiss zones in Carratraca area; P, mylonitic margin of Ronda peridotite. Isotherms shown before conductive heating. (b) Ductile extensional detachment D1 brings Ronda peridotite close to base of orogenic crust, creates peridotite mylonites under high-P conditions. (c) Situation at 21 Ma: vertical thinning of lithospheric column by one-third, lesser degree of thinning to north (Sierra Alpujata region; see text). Isotherms have moved up relative to rock positions as a result of conductive heating. (d) Extensional detachment D2 brings lower-crustal rocks of Alpujata region beneath Ronda peridotite at high T and low P. It should be noted that stages (a)–(c) may have overlapped in time. Scale in (d) is twice that in (a)–(c). (e) Situation at 20 Ma. Continued thinning by slip on extensional detachments. D3 places peridotites beneath garnet gneisses at high T and low P, reactivating D1. D4 defines present Malaguide–Alpujarride boundary. Scale in (e) is twice that in (d). (f) Peridotite and crustal envelope are tilted and thrust onto Iberian margin (external thrust belt), producing present configuration.
nian to Burdigalian (early Miocene) age locally overlie metamorphic rocks in the vicinity of the Carratraca massif (Sanz de Galdeano et al. 1993), and make up much of the fill in the West Alboran Basin (Comas et al. 1999). Rapid extensional exhumation therefore closely followed removal of sub-orogenic lithospheric mantle, which suggests a causal link between the two (Platt & Vissers 1989). Any account of the exhumation history of the Ronda peridotite must address the question of its present position as a relatively thin slice interleaved with crustal rocks. The peridotite sheet overlies a sequence of metamorphosed crustal rocks in the Sierra Alpujata that are similar in terms of protolith type to other metamorphic units in the Internal Betics, and strikingly similar in terms of metamorphic grade and evolution to the garnet gneiss and migmatite overlying the peridotite. Some of the migmatites have a remarkable fragmental texture, consisting of pebble- to boulder-sized fragments of graphitic mica schist in a cordierite granite matrix. Cordierite-bearing leucogranite dykes also intrude the peridotite complexes. These lines of evidence suggest that the peridotite was emplaced tectonically above the crustal
sequence under high-temperature conditions consistent with significant partial melting of crustal rocks (Tubı´a et al. 1997). Radiometric ages from these rocks are very similar to those from the metamorphic sequence above the peridotite: zircons in metabasic rocks have rims that yield early Miocene U–Pb ages (Sa´nchez-Rodrı´guez & Gebauer 2000), and Ar/Ar ages on muscovite and biotite cluster around 19 Ma (Sosson et al. 1998). The peridotite slice therefore appears to have been emplaced above crustal rocks at almost exactly the same time as it was being exhumed by high rates of stretching in the crustal rocks above it. One possibility is that the contact below the peridotite is the first contractional structure to have formed in the process that finally led to the emplacement of the Alboran Domain onto the Iberian continental margin, forming the Miocene external thrust belt. The rocks beneath the peridotite have closer affinities to the Alboran Domain itself than to the Iberian margin, however. We therefore explore the alternative possibility here that the basal contact of the peridotite sheet is in fact an extensional ductile shear zone that formed during the overall exhumational history of the region.
674
J. P. P L AT T E T A L .
Rapid extensional exhumation during the earliest Miocene can be demonstrated throughout much of the Alboran Domain (Platt et al. 2003b), but the precise timing of cooling varies by as much as 3 Ma from place to place. The Ronda peridotite and its overlying metamorphic envelope were already largely exhumed by 20 Ma, at which time the rocks beneath were still above the closure temperature for muscovite (Sosson et al. 1998). We therefore propose that an extensional detachment, formed as part of the overall exhumational process in the early Miocene, emplaced the peridotite and its cover laterally onto a closely related sequence of metamorphic rocks that had been exhumed slightly more slowly, and hence was at a lower level in the crust, and still at high temperature. A scheme for the exhumation history of the Ronda peridotite and its crustal envelope that incorporates this idea is shown in Figure 13. The situation at the start of the exhumational process at 25 Ma, after crustal thickening and removal of lithospheric mantle below 67.5 km depth, is shown in Figure 13a. Five stages of exhumation are then shown. An early ductile detachment (D1) is postulated that penetrated the upper mantle, creating the peridotite mylonites seen along the upper surface of the Ronda massif (Fig. 13b). Transient cooling associated with this displacement moved the peridotites temporarily into the garnet stability field. Vertical thinning of the whole lithospheric column then took place: by one-third in the Carratraca region, but less in a region originally north or NE of the Ronda peridotite, now exposed in the Sierra Alpujata (Fig. 13c). This process was accompanied by conductive heating of the whole lithospheric column. A second extensional detachment (D2) then brought the lower-crustal rocks of the Alpujata region beneath the Ronda peridotite (Fig. 13d). The formation of cordierite-bearing fragmental migmatites associated with this contact indicates that it formed at high temperature but relatively low pressure. By 20 Ma the sequence had been reduced to its present thickness by slip on a series of extensional detachments (Fig. 13e). D3 emplaced the peridotite beneath the garnet gneiss at high T and low P, producing the late high-T, low-P mylonitic fabric in the garnet gneiss, and reactivating D1. D4 defines the present Malaguide–Alpujarride boundary. The final stage in the exhumation process involved the thrusting of the peridotite and its crustal envelope onto the Iberian margin at the start of deformation in the external thrust belt, producing the present configuration (Fig. 13f). Final erosional exhumation was largely complete by around 17 Ma, and the sequence is overlain by Burdigalian marine sediments (Boulin et al. 1973). This work was supported by grant GR3/10828 and studentship GT4/92/ 242/G from the Natural Environmental Research Council of Great Britain (to J.P.P. and T.W.A., respectively). The Nordic geological ion-microprobe facility (NordSIM) is jointly funded by Denmark, Norway and Finland; M.J.W. acknowledges support of the Swedish NFR via a research fellowship. L.L. is funded by the Royal Society and acknowledges the London University Central Research Fund for a fieldwork grant. We thank K. Johnson for help in the field and for some of the fission-track analyses. We are grateful to T. Tubı´a, J. Selverstone, S. Harley and K. Hodges for their careful and thought-provoking reviews of various versions of this paper.
References Amato, J.M., Johnson, C.M., Baumgartner, L.P. & Beard, B.L. 1999. Rapid exhumation of the Zermatt–Saas ophiolite deduced from high-precision Sm– Nd and Rb–Sr geochronology. Earth and Planetary Science Letters, 171, 425–438. Andersen, T.B., Osmundsen, P.T. & Jolivet, L. 1994. Deep crustal fabrics and a
model for the extensional collapse of the southwest Norwegian Caledonides. Journal of Structural Geology, 16, 1191–1204. Andriessen, P.A.M. & Zeck, H.P. 1996. Fission-track constraints on timing of Alpine nappe emplacement and rates of cooling and exhumation, Torrox area, Betic Cordilleras, S. Spain. Chemical Geology, 131, 199–206. Argles, T.W., Platt, J.P. & Waters, D.J. 1999. Attenuation and excision of a crustal section during extensional exhumation: the Carratraca Massif, Betic Cordillera, Southern Spain. Journal of the Geological Society, London, 156, 149–162. Azan˜o´n, J.M. & Alonso-Chaves, F.M. 1996. Alpine tectono-metamorphic evolution of the Tejeda Unit, an extensionally dismembered Alpujarride nappe (western Betics). Comptes Rendus de l’Acade´mie des Sciences, 322, 47–54. Azan˜o´n, J.-M. & Goffe´, B. 1997. Ferro- and magnesiocarpholite assemblages as record of the high-P, low-T metamorphism in the Central Alpujarrides, Betic Cordillera (SE Spain). European Journal of Mineralogy, 9, 1035–1051. Balanya´, J.C., Azan˜o´n, J.M., Sa´nchez-Go´mez, M. & Garcı´a-Duen˜as, V. 1993. Pervasive ductile extension, isothermal decompression, and thinning of the Jubrique unit in the Palaeogene (Alpujarride Complex, western Betics Spain). Comptes Rendus de l’Acade´mie des Sciences, 316, 1595–1601. Balanya´, J.C., Garcı´a-Duen˜as, V., Azan˜o´n, J.M. & Sa´nchez-Go´mez, M. 1997. Alternating contractional and extensional events in the Alpujarride nappes of the Alboran Domain (Betics, Gibraltar arc). Tectonics, 16, 226–238. Barranco, L.M., Ansorge, J. & Banda, E. 1990. Seismic refraction constraints on the geometry of the Ronda peridotitic massif (Betic Cordilleras, Spain). Tectonophysics, 184, 379–392. Bhattacharya, A., Krishnakumar, K.R., Raith, M. & Sen, S.K. 1991. An improved set of a–X parameters for Fe–Mg–Ca garnets and refinements of the orthopyroxene–garnet thermometer and the orthopyroxene–garnet– plagioclase–quartz barometer. Journal of Petrology, 32, 629–656. Bird, P. 1979. Continental delamination and the Colorado plateau. Journal of Geophysical Research, 84, 7561–7571. Blanco, M.J. & Spakman, W. 1993. The P-wave velocity structure of the mantle below the Iberian Peninsula: evidence for subducted lithosphere below southern Spain. Tectonophysics, 221, 13–34. Blichert-Toft, J., Albare`de, F. & Kornprobst, J. 1999. Lu–Hf isotope systematics of garnet pyroxenites from Beni Bousera, Morocco: implications for basalt origin. Science, 283, 1303–1306. Boulin, J., Bourgois, J. & Chauve, P. et al. 1973. Age mioce`ne infe´rieure de la formation de la Vin˜uela, discordante sur les nappes internes be´tiques. Comptes Rendus de l’Acade´mie des Sciences, 276, 1245–1248. Bouybaoue`ne, M., Michard, A. & Goffe´, B. 1998. High-pressure granulites on top of the Beni Bousera peridotites, Rif belt, Morocco: a record of an ancient thickened crust in the Alboran domain. Bulletin de la Societe´ Ge´ologique de France, 169, 153–162. Buick, I.S. & Holland, T.J.B. 1989. The P–T–t path associated with crustal extension, Naxon, Cyclades, Greece. In: Daly, J.S., Cliff, R.A. & Yardley, B.W.D. (eds) Evolution of Metamorphic Belts. Geological Society, London, Special Publications, 43, 365–370. Burov, E., Jolivet, L., Le Pourhiet, L. & Poliakov, A. 2001. A thermomechanical model of exhumation of high pressure (HP) and ultra-high pressure (UHP) metamorphic rocks in alpine-type collision belts. Tectonophysics, 342, 113–136. Calvert, A., Sandvol, E. & Seber, D. et al. 2000. Geodynamic evolution of the lithosphere and upper mantle beneath the Alboran region of the western Mediterranean: constraints from travel time tomography. Journal of Geophysical Research, 105, 10871–10898. Carminati, E., Wortel, M.J.R., Spakman, W. & Sabadini, R. 1998. The role of slab detachment processes in the opening of the western–central Mediterranean basins: some geological and geophysical evidence. Earth and Planetary Science Letters, 160, 651–665. Chemenda, A.I., Mattauer, M., Malavieille, J. & Bokun, A.N. 1995. A mechanism for syn-collisional rock exhumation and associated normal faulting; results from physical modelling. Earth and Planetary Science Letters, 132, 225–232. Cloos, M. 1982. Flow melanges: numerical modelling and geologic constraints on their origin in the Franciscan subduction complex, California. Geological Society of America Bulletin, 93, 330–345. Comas, M.C., Garcı´a-Duen˜as, V. & Jurado, M.J. 1992. Neogene tectonic evolution of the Alboran Basin from MCS data. GeoMarine Letters, 12, 157–164. Comas, M.C., Platt, J.P., Soto, J.I. & Watts, A.B. 1999. The origin and tectonic history of the Alboran Basin: insights from Leg 161 results. In: Zahn, R., Comas, M.C. & Klaus, A. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 161. Ocean Drilling Program, College Station, TX, 555–579. Davies, G.R., Nixon, P.H., Pearson, D.H. & Obata, M. 1993. Tectonic implications of graphitized diamonds from the Ronda peridotite massif, southern Spain. Geology, 21, 471–474. Docherty, C. & Banda, E. 1995. Evidence for eastward migration of the Alboran
E X H U M AT I O N O F T H E RO N DA P E R I D OT I T E Sea based on regional subsidence analysis: a case for basin formation by delamination of the subcrustal lithosphere? Tectonics, 14, 804–818. England, P.C. & Thompson, A.B. 1984. Pressure–temperature–time paths of regional metamorphism I. Heat transfer during the evolution of regions of thickened continental crust. Journal of Petrology, 25, 894–928. Ernst, W.G. 1988. Tectonic history of subduction zones inferred from retrograde blueschist P–T paths. Geology, 16, 1081–1085. Fitzsimons, I.C.W. & Harley, S.L. 1994. The influence of retrograde cation exchange on granulite P–T estimates and a convergence technique for the recovery of peak metamorphic conditions. Journal of Petrology, 35, 543–576. Fraser, G., Ellis, D. & Eggins, S. 1997. Zirconium abundance in granulite-facies minerals, with implications for zircon geochronology in high-grade rocks. Geology, 25, 607–610. Galbraith, R.F. & Laslett, G.M. 1993. Statistical model for mixed fission track ages. Nuclear Tracks and Radiation Measurements, 21, 459–470. Ganguly, J., Cheng, W. & Tirone, M. 1996. Thermodynamics of aluminosilicate garnet solid solution: new experimental data, an optimized model, and thermometric applications. Contributions to Mineralogy and Petrology, 126, 137–151. Garcı´a-Casco, A. & Torres-Rolda´n, R.L. 1996. Disequilibrium induced by fast decompression in St–Bt–Grt–Ky–Sil–And metapelites from the Betic Belt (Southern Spain). Journal of Petrology, 37, 1207–1239. Gebauer, D., Schertl, H.-P., Brix, M. & Schreyer, W. 1997. 35 Ma old ultrahigh-pressure metamorphism and evidence for very rapid exhumation in the Dora Maira Massif, Western Alps. Lithos, 41, 5–24. Goffe´, B., Michard, A. & Garcı´a-Duen˜as, V. et al. 1989. First evidence of high-pressure, low-temperature metamorphism in the Alpujarride nappes, Betic Cordilleras (S.E. Spain). European Journal of Mineralogy, 1, 139–142. Go´mez Pugnaire, M.T. & Camara, F. 1990. La asociacio´n de alta presio´n distena þ talco þ fengita coexistente con escapolita en metapelitas de origen evaporı´tico (Complejo Nevado–Fila´bride, Cordilleras Be´ticas). Revista de Sociedad Geologico de Espan˜a, 3, 373–384. Go´mez-Pugnaire, M.T. & Ferna´ndez-Soler, J.M. 1987. High-pressure metamorphism in metabasites from the Betic Cordilleras (S.E. Spain) and its evolution during the Alpine orogeny. Contributions to Mineralogy and Petrology, 95, 231–244. Graham, C.M. & Powell, R. 1984. A garnet–hornblende geothermometer: calibration, testing, and application to the Pelona Schist, Southern California. Journal of Metamorphic Geology, 2, 13–31. Hanne, D., White, N. & Lonergan, L. 2003. Subsidence analyses from the Betic Cordillera, southeast Spain. Basin Research, 15, 1–21. Harley, S.L. 1984. An experimental study of the partitioning of Fe and Mg between garnet and orthopyroxene. Contributions to Mineralogy and Petrology, 86, 359–373. Holland, T.J.B. 1980. The reaction albite ¼ jadeite þ quartz determined experimentally in the range 600–1200 8C. American Mineralogist, 65, 129–134. Holland, T.J.B. 1983. The experimental determination of activities in disordered and short-range ordered jadeitic pyroxenes. Contributions to Mineralogy and Petrology, 82, 214–220. Holland, T.J.B. & Blundy, J. 1994. Non-ideal interactions in calcic amphiboles and their bearing on amphibole–plagioclase thermometry. Contributions to Mineralogy and Petrology, 116, 433–447. Holland, T.J.B. & Powell, R. 1992. Plagioclase feldspars: activity–composition relations based upon Darken’s quadratic formalism and Landau theory. American Mineralogist, 77, 53–61. Houseman, G.A. & Molnar, P. 1997. Gravitational (Rayleigh–Taylor) instability of a layer with non-linear viscosity and convective thinning of continental lithosphere. Geophysical Journal International, 128, 125–150. Huerta, A.D., Royden, L.H. & Hodges, K.V. 1999. The effects of accretion, erosion, and radiogenic heat on the metamorphic evolution of collisional orogens. Journal of Metamorphic Geology, 17, 349–366. Johnson, T.E., Hudson, N.F.C. & Droop, G.T.R. 2001. Partial melting in the Inzie Head gneisses: the role of water and a petrogenetic grid in KFMASH applicable to anatectic pelitic migmatites. Journal of Metamorphic Geology, 19, 99–118. Kohn, M.J. & Spear, F.S. 1990. Two new geobarometers for garnet amphibolites, with applications to southeastern Vermont. American Mineralogist, 75, 89–96. Kretz, P. 1983. Symbols for rock-forming minerals. American Mineralogist, 68, 277–279. Krogh Ravna, E. 2000. The garnet–clinopyroxene Fe2þ –Mg geothermometer: an updated calibration. Journal of Metamorphic Geology, 18, 211–219. Lenoir, X., Garrido, C.J., Bodinier, J.-L., Dautria, J.-M. & Gervilla, F. 2001. The recrystallization front of the Ronda peridotite: evidence for melting and thermal erosion of subcontinental lithospheric mantle beneath the Alboran Basin. Journal of Petrology, 42, 141–158. Lonergan, L. & White, N. 1997. Origin of the Betic–Rif mountain belt. Tectonics, 16, 504–522.
675
Loomis, T.P. 1972a. Contact metamorphism of pelitic rock by the Ronda ultramafic intrusion, southern Spain. Geological Society of America Bulletin, 83, 2449–2474. Loomis, T.P. 1972b. Diapiric emplacement of the Ronda high-temperaure ultramafic intrusion, southern Spain. Geological Society of America Bulletin, 83, 2475–2496. Michard, A., Goffe´, B., Bouybaoue`ne, M.L. & Saddiqi, O. 1997. Late Hercynian–Mesozoic thinning in the Alboran Domain: metamorphic data from the northern Rif, Morocco. Terra Nova, 9, 171–174. Monie´, P., Galindo-Zaldı´var, J., Gonza´lez-Lodeiro, F., Goffe´, B. & Jabaloy, A. 1991. 40 Ar/39 Ar geochronology of Alpine tectonism in the Betic Cordilleras (southern Spain). Journal of the Geological Society, London, 148, 289–297. Monie´, P., Torres-Rolda´n, R.-L. & Garcı´a-Casco, A. 1994. Cooling and exhumation of the western Betic Cordilleras, 40 Ar/39 Ar thermochronological constraints on a collapsed terrane. Tectonophysics, 238, 353–379. Obata, M. 1980. The Ronda peridotite: garnet, spinel, and plagioclase lherzolite facies and the P–T trajectories of a high-temperature mantle intrusion. Journal of Petrology, 21, 533–572. O’Brien, P.J., Zotov, N., Law, R., Khan, M.A. & Jan, M.Q. 2001. Coesite in Himalayan eclogite and implications for models of India–Asia collision. Geology, 29, 435–438. Pearson, D.G., Davies, G.R., Nixon, P.H. & Milledge, H.J. 1989. Graphitized diamonds from a peridotite massif in Morocco and implications for anomalous diamond occurrences. Nature, 335, 60–63. Platt, J.P. 1986. Dynamics of orogenic wedges and the uplift of high-pressure metamorphic rocks. Geological Society of America Bulletin, 97, 1037–1053. Platt, J.P. & England, P.C. 1994. Convective removal of lithosphere beneath mountain belts: thermal and mechanical consequences. American Journal of Science, 294, 307–336. Platt, J.P. & Vissers, R.L.M. 1989. Extensional collapse of thickened continental lithosphere: a working hypothesis for the Alboran Sea and the Gibraltar arc. Geology, 17, 540–543. Platt, J.P. & Whitehouse, M.J. 1999. Early Miocene high-temperature metamorphism and rapid exhumation in the Betic Cordillera (Spain): evidence from U–Pb zircon ages. Earth and Planetary Science Letters, 171, 591–605. Platt, J.P., Soto, J.-I., Whitehouse, M.J., Hurford, A.J. & Kelley, S.P. 1998. Thermal evolution, rate of exhumation, and tectonic significance of metamorphic rocks from the floor of the Alboran extensional basin, western Mediterranean. Tectonics, 17, 671–689. Platt, J.P., Allerton, S., Kirker, A., Mandeville, C., Mayfield, A., Platzman, E.S. & Rimi, A. 2003a. The ultimate arc: differential displacement, oroclinal bending, and vertical axis rotation in the formation of the External Betic–Rif arc. Tectonics, 22, 10.1029/2001TC001321. Platt, J.P., Whitehouse, M.J., Kelley, S.P., Carter, A. & Hollick, L. 2003b. Simultaneous extension and exhumation across a collapsing back-arc basin: implications for the causes of late-orogenic extension. Geology, 31, 259–262. Platt, J.P., Kelley, S.P., Carter, A. & Orozco, M. 2004. Timing of tectonic events in the Alpuja´rride Complex, Betic Cordillera, S. Spain. Journal of the Geological Society, London, in press. Platzman, E., Platt, J.P., Kelley, S.P. & Allerton, S. 2000. Large clockwise rotations in an extensional allochthon, Alboran Domain (southern Spain). Journal of the Geological Society, London, 157, 1187–1197. Puga, E., Nieto, J.M., Diaz de Federico, A., Bodinier, J.L. & Morten, L. 1999. Petrology and metamorphic evolution of ultramafic rocks and dolerite dykes of the Betic Ophiolite Association (Mulhacen Complex, SE Spain): evidence of eo-Alpine subduction following an ocean-floor metasomatic process. Lithos, 49, 23–56. Reisberg, L. & Lorand, J.-P. 1995. Longevity of sub-continental mantle lithosphere from osmium isotope systematics in orogenic peridotite massifs. Nature, 376, 159–162. Roberts, M.P. & Finger, F. 1997. Do U–Pb zircon ages from granulites reflect peak metamorphic conditions? Geology, 25, 319–322. Ruppel, C. & Hodges, K.V. 1994. Pressure–temperature–time paths from twodimensional thermal models. Tectonics, 13, 17–44. Ruppel, C., Royden, L. & Hodges, K.V. 1988. Thermal modelling of extensional tectonics: application to pressure–temperature–time histories of metamorphic rocks. Tectonics, 7, 947–957. Sa´nchez-Rodrı´guez, L. & Gebauer, D. 2000. Mesozoic formation of pyroxenites and gabbros in the Ronda area (southern Spain), followed by Early Miocene subduction metamorphism and emplacement into the middle crust: U–Pb sensitive high-resolution ion microprobe dating of zircon. Tectonophysics, 316, 19–44. Sa´nchez-Vizcaı´no, L.V., Rubatto, D., Go´mez-Pugnaire, M.T., Trommsdorff, V. & Mu¨ntener, O. 2001. Middle Miocene high-pressure metamorphism and fast exhumation of the Nevado–Fila´bride Complex, SE Spain. Terra Nova, 13, 327–332. Sanz de Galdeano, C., Serrano, F., Lopez-Garrido, A.C. & Martı´n-Perez,
676
J. P. P L AT T E T A L .
M. 1993. Palaeogeography of the Late Aquitanian–Early Burdigalian basin in the western Betic Internal Zone. Geobios, 26, 43–55. Seber, D., Barazangi, M., Ibenbrahim, A. & Demnati, A. 1996. Geophysical evidence for lithospheric delamination beneath the Alboran Sea and Rif– Betic mountains. Nature, 379, 785–791. Sosson, M., Morillon, A.C., Bourgois, J., Feraud, G., Poupeau, G. & SaintMarc, P. 1998. Late exhumation stages of the Alpujarride Complex (western Betic Cordilleras, Spain); new thermochronological and structural data on Los Reales and Ojen nappes. Tectonophysics, 285, 253–273. Spear, F.S. & Kohn, M.J. 1996. Trace element zoning in garnet as a monitor of crustal melting. Geology, 24, 1099–1102. Torne´, M., Banda, E., Garcı´a-Duen˜as, V. & Balanya´, J.C. 1992. Mantle– lithosphere bodies in the Alboran crustal domain (Ronda peridotites, Betic– Rif orogenic belt). Earth and Planetary Science Letters, 110, 163–171. Torres-Rolda´n, R.L. 1981. Plurifacial metamorphic evolution of the Sierra Bermeja peridotite aureole (Southern Spain). Estudios Geologicos, 37, 115– 133. Tubı´a, J.M. & Gil-Ibarguchi, J.L. 1991. Eclogites of the Ojen nappe: a record of subduction in the Alpuja´rride complex (Betic Cordilleras, southern Spain). Journal of the Geological Society, London, 148, 801–804. Tubı´a, J.M., Navarro-Vila, F. & Cuevas, J. 1993. The Mala´guide–Los Reales Nappe: an example of crustal thinning related to the emplacement of the Ronda peridotites (Betic Cordillera). Physics of the Earth and Planetary Interiors, 78, 343–354. Tubı´a, J.M., Cuevas, J. & Gil-Ibarguchi, J.-I. 1997. Sequential development of
the metamorphic aureole beneath the Ronda peridotites and its bearing on the tectonic evolution of the Betic Cordillera. Tectonophysics, 279, 227–252. Van der Wal, D. 1993. Deformation processes in mantle peridotites. PhD Thesis, University of Utrecht. Van der Wal, D. & Bodinier, J.L. 1996. Origin of the recrystallisation front in the Ronda peridotite by km-scale pervasive porous melt flow. Contributions to Mineralogy and Petrology, 122, 387–405. Van der Wal, D. & Vissers, R.L.M. 1993. Uplift and emplacement of upper mantle rocks in the western Mediterranean. Geology, 21, 1119–1122. Van der Wal, D. & Vissers, R.L.M. 1996. Structural petrology of the Ronda peridotite, SW Spain: deformation history. Journal of Petrology, 37, 23–43. Vissers, R.L.M., Platt, J.P. & Van der Wal, D. 1995. Late orogenic extension of the Betic Cordillera and the Alboran Domain: a lithospheric view. Tectonics, 14, 786–803. Waters, D.J. 1991. Hercynite–quartz granulites: phase relations, and implications for crustal processes. European Journal of Mineralogy, 3, 367–386. Whitehouse, M.J. & Platt, J.P. 2003. Dating high-grade metamorphism— constraints from rare-earth elements in zircon and garnet. Contributions to Mineralogy and Petrology, 145, 61–74. Zeck, H.P., Monie´, P., Villa, I.M. & Hansen, B.T. 1992. Very high rates of cooling and uplift in the Alpine belt of the Betic Cordilleras, southern Spain. Geology, 20, 79–82. Zindler, A., Staudigel, H., Hart, S.R., Endres, R. & Goldstein, S. 1983. Nd and Sm isotopic study of a mafic layer from Ronda ultramafic complex. Nature, 304, 226.
Received 1 October 2002; revised typescript accepted 5 May 2003. Scientific editing by Mike Cosca
