An extended episode of early Mesoproterozoic metamorphic fluid flow in the Reynolds Range,central Australia*

ABSTRACT The products of metamorphic fluid flow are preserved in zones within the marbles and metamorphosed semipelites of the Upper Calcsilicate Unit in the granulite portion of the Late Palaeoproterozoic Reynolds Range Group, northern Arunta Block, central Australia. The zones of retrogression, characterized by minerals such as wollastonite, grossular and clinohumite, local resetting of oxygen isotopic compositions and local major element metasomatism, were channelways for water-rich fluids derived from granulite facies metapelites. U–Th–Pb isotopic ages measured by the SHRIMP ion microprobe on zircon and monazite from a granulite facies semipelite, an early semiconcordant aluminous quartz-rich fluid-flow segregation and a late discordant quartz-rich segregation record some of the extended thermal history of the area. Zircon cores from the semipelite show its likely protolith to be an igneous rock 1812 ± 11 Ma old, itself derived from a source containing zircon as old as 2.2 Ga. Low-Th/U overgrowths on the zircon grew during granulite facies metamorphism at 1594 ± 6 Ma. Monazite cooled to its blocking temperature at 1576 ± 8 Ma. Zircon cores from the semiconcordant segregation are dominantly >2.3 Ga old, indicating that the source of the fluids was not the particular metamorphosed semipelite studied. Two generations of low-Th/U overgrowths on the zircon give indistinguishable ages for the older and younger of 1589 ± 8 and 1582 ± 8 Ma, respectively. The monazite age is the same, 1576 ± 12 Ma. Zircon from the late discordant segregation gave 1568 ± 4 Ma. Fluid flow occurred for at least 18 ± 3 (σ) Ma and ended 26 ± 3 (σ) Ma after the peak of metamorphism, suggesting a very slow cooling rate of ～3°C Ma^–1. The last regional high-grade metamorphism in the Reynolds Range occurred at ～1.6 Ga, not ～1.78 Ga as previously thought. The high-grade event at ～1.78 Ga is a separate event that affected only the basement to the Reynolds Range Group.     

The generation of quartz-normative melts and corundum-bearing restites by crustal anatexis: petrogenetic modelling based on an example from the Lewisian of North-West Scotland

Abstract Rare layers of an aluminous, muscovite-rich rock from the Lewisian Complex at Stoer, North-West Scotland, display evidence which suggests that the rock has undergone local partial melting to form quartz-bearing veins and a corundum-bearing restite. The assemblages observed in these rocks match those predicted by modelling in the system KAlO₂-NaAlO₂-Al₂O₃-SiO₂-H₂O (KNASH) where certain bulk compositions melt peritectically to give corundum-bearing restites and quartz-normative melts. Study of the model system shows that the observed parageneses could have formed from a range of bulk compositions with a variety of possible values of a H₂O which could have been internally or externally buffered. The KNASH petrogenetic grid, together with another in the system CaO-Na₂O-FeO-Al₂O₃-SiO₂-H₂O (CNFASH), allows the P–T path of the rocks to be delineated and an estimate to be made of the conditions at the peak of metamorphism as > 11 kbar and 900-925°C. This estimate is in agreement with P–T estimates using thermobarometric methods on adjacent lithologies: The activity of H₂O in the system throughout metamorphism is calculated to have been >0.3.     

Stable isotopic evidence for fluid infiltration during contact metamorphism in a multiply-metamorphosed terrane: the Reynolds Range, Arunta Block, central Australia

The Lander Rock Beds form the local basement of the Reynolds Range in the Arunta Inlier of central Australia. These dominantly quartzose and pelitic lithologies underwent low-grade ( c.   400  °C) regional metamorphism prior to contact metamorphism ( c.   2.5  kbar) around S-type megacrystic granitoids at 1820–1800  Ma. The Lander Rock Beds are overlain by metasediments of the Reynolds Range Group, which were subsequently intruded by granitoids at c. 1780  Ma. Regional metamorphism at 1590–1580  Ma produced grades varying from greenschist (400  °C at 4–5  kbar) to granulite (750–800  °C at 4–5  kbar) from north-west to south-east along the length of the Reynolds Range. Oxygen isotope ratios of the Lander Rock Beds were reset from 13.4±0.8 to as low as 6.7 adjacent to the contacts of the larger plutons, and to 10.3±1.1 around the smaller plutons. Biotite in all the major rock types found in the aureoles has δD values between −52 and −69, probably reflecting resetting by a cooling igneous+metamorphic fluid near the plutons. Sapphirine-bearing and other Mg- and Al-rich rock types have low δ¹⁸O values (4.0±0.7). The precursors to these rocks were probably low-temperature ( c. 200  °C) diagenetic–hydrothermal deposits of Mg-rich chlorite, analogous to those in Proterozoic stratiform precious metal and uranium deposits that form by the infiltration of basin brines or seawater. As in the overlying Reynolds Range Group, regional metamorphism involved little fluid–rock interaction and isotopic resetting.     

Determining the direction of contact metamorphic fluid flow: an assessment of mineralogical and stable isotope criteria

ABSTRACT One-dimensional fluid advection-dispersion models predict differences in the patterns of mineralogical and oxygen isotope resetting during up- and down-temperature metamorphic fluid flow that may, in theory, be used to determine the fluid flow direction with respect to the palaeotemperature gradient. Under equilibrium conditions, down-temperature fluid flow is predicted to produce sharp reaction fronts that separate rocks with isobarically divariant mineral assemblages. In contrast, up-temperature fluid flow may produce extensive zones of isobarically univariant mineral assemblages without sharp reaction fronts. However, during contact metamorphism, mineral reaction rates are probably relatively slow compared with fluid velocities and distended reaction fronts may also form during down-temperature fluid flow. In addition, uncertainties in the timing of fluid flow with respect to the thermal peak of metamorphism and the increase in the variance of mineral assemblages due to solid solutions introduce uncertainties in determining fluid flow directions. Equilibrium down-temperature flow of magmatic fluids in contact aureoles is also predicted to produce sharp δ¹⁸O fronts, whereas up-temperature flow of fluids derived by metamorphic devolatilization may produce gradational δ¹⁸O vs. distance profiles. However, if fluids are channelled, significant kinematic dispersion occurs, or isotopic equilibrium is not maintained, the patterns of isotopic resetting may be difficult to interpret. The one-dimensional models provide a framework in which to study fluid-rock interaction; however, when some of the complexities inherent in fluid flow systems are taken into account, they may not uniquely distinguish between up- and down-temperature fluid flow. It is probably not possible to determine the fluid flow direction using any single criterion and a range of data is required.     

The gabbro–eclogite transformation: an oxygen isotope and petrographic study of west Alpine ophiolites

Both magmatic and eclogitic parageneses are preserved in the gabbros of western Alpine ophiolites. Samples with relic magmatic mineralogies display partial transformation to eclogitic assemblages along cracks and grain boundaries. Gabbros with eclogitic mineralogies contain zoned pseudomorphs after olivine, comprising talc-rich cores with kyanite, Mg-chloritoid and omphacite in outer cores and garnet rims. The compositional zonation of these olivine pseudomorphs closely parallels that shown by olivines in hydrothermally altered ocean-floor gabbros. The eclogitic gabbros are hydrous, containing paragonite, zoisite and other water-bearing minerals, and it has been suggested that water was introduced during high-pressure metamorphism. However, the similarity of olivine alteration patterns to those of ocean-floor gabbros suggests that hydration and local metasomatism leading to the stability of aluminous minerals in olivine sites occurred during hydrothermal alteration prior to subduction. Oxygen-isotope systematics are consistent with this proposal: Alpine gabbros with magmatic relics have a mean δ¹⁸O value of 5.7±0.7, similar to that of unaltered oceanic crust, whereas eclogitic gabbros have a mean δ¹⁸O value of 4.8±0.9.This statistically significant difference is consistent with the eclogitic samples having undergone high-temperature ocean-floor alteration. The preservation of magmatic and hydrothermal δ¹⁸O values in ocean-floor gabbros that have been metamorphosed at 2–2.5 GPa (60–75 km) implies that the deeper levels of ocean crust have not experienced pervasive fluid flow during subduction or subsequent exhumation. Magmatic assemblages were preserved despite an overstep of eclogitization reactions by at least 0.6–1.1 GPa implying that equilibrium was not attained in undeformed parts of the system because of slow diffusion in water-deficient rock volumes.     

Fluid migration and vein formation during deformation and greenschist facies metamorphism at Ormiston Gorge, central Australia

During the Alice Springs Orogeny, deformation at Ormiston Gorge, central Australia, occurred under lower- to middle-greenschist facies conditions. Dolomites of the Bitter Springs Formation and quartzites. metagreywackes, and metapelites of the Heavitree Quartzite contain abundant early-, syn-, and post-tectonic veins. However, though vein densities locally approach 15%, the distribution of veins and the oxygen isotope geochemistry of wallrocks and veins suggest that fluid movement was on a local scale. The Heavitree Quartzite contains quartz veins that, even along the main thrust plane, have similar δ¹⁸O values (13.5–16.9%o) to those of their wallrocks (13.6–16.9%o), with Δ¹⁸O(vein-wallrock) values of -0.6 to 0.4%o. In contrast, the Bitter Springs Formation contains predominantly dolomite veins that have δ¹⁸O values of 23.4 to 27.7%o. These differences are observed even at the boundary between the Heavitree and Bitter Springs rocks, implying that significant fluid exchange between these rocks has not occurred, or that fluid flow was channelled through areas outside those sampled for this study. By contrast with the Heavitree Quartzite, δ¹⁸O values of wallrocks in individual samples of the Bitter Springs Formation are significantly higher (23.3–29.1%o) than those of the veins, with δ¹⁸O(vein-wallrock) values up to -4%o (average of -2.1%o). These systematic differences in δ¹⁸O values most likely result from oxygen isotope fractionation caused by fluid immiscibility or disequilibrium dissolution. Smaller differences in δ¹³C values between some dolomite veins and wallrocks [δ¹³C(vein-wallrock) up to -1.9%o, average of -0.5%o] are also explained by these processes. This study indicates that large volumes of veins may be produced by repeated fracturing and fluid migration within particular rock units, without involving large volumes of externally derived fluids.     

Phonolitic Diatremes within the Dunedin Volcano, South Island, New Zealand

The Port Chalmers Breccia is a vent-filling, clastic volcanicunit exposed within the Miocene Dunedin Volcano of South Island,New Zealand. Clasts (up to in excess of 1 m but generally <20cm) are supported in ash and fine lapilli of phonolitic (ne-benmoreiteor tephro-phonolite) composition and the dominant clast type(55 to almost 100%) is also phonolitic. Less abundant lithologiesinclude ne-normative basalt (basanite), hawaiite, mugeariteand trachyandesite, syenites and microsyenites, coarse-grainedmafic (gabbros) and ultramafic rocks (pyroxenites, hornblendites),schists and sediments. The breccias were emplaced as diatremesassociated with localized, but highly explosive, eruptive eventsin which mantle-derived CO₂ was an important component. Thesyenitic and ultramafic clasts could represent intrusive suitesproduced by crystal fractionation acting on parental ne-benmoreitemagmas that may themselves have been derived by crystal fractionationfrom basanitic precursors. An alternative variation on thismodel is that the parental ne-benmoreites were generated throughpartial melting of an alkalic igneous underplate. Sr, Nd andPb isotopic compositions are strikingly similar to those ofintraplate igneous rocks, ranging in age from 100 to less than10 Ma, from elsewhere in the South Island, and New Zealand'ssub-Antarctic islands, the south Tasman Sea and the Ross Searegion. This regional, HIMU-influenced, isotopic signature isbelieved to be derived from within the lithospheric mantle. KEY WORDS: phonolite; diatreme; nepheline syenite; Dunedin Volcano; alkalic rocks; fractional crystallization     

Partial Melting of Metapelitic Rocks Beneath the Bushveld Complex, South Africa

Metapelitic rocks in the aureole beneath the Bushveld Complexpreserve evidence for both high- and low-a_H2O anatexis. Theaureole is characterized by an inverted thermal structure inwhich suprasolidus rocks potentially interacted with an H₂O-richvolatile phase derived from underlying, dehydrating rocks. Atlower grade (T < 700°C) the rocks contain fibrolite matsand seams that record local redistribution of volatiles. Incongruentreactions consuming biotite produced small quantities (<1mol %) of liquid and peritectic cordierite that remained trappedwithin the mesosome. Larger volumes of melt (3–4%), preservedas coarse-grained discordant leucosomes, were produced by congruentmelting following a structurally focused influx of H₂O. Subhorizontalvolatile-phase flow was concentrated within thin (     

Evidence for pre-regional metamorphic fluid infiltration of the Lower Calcsilicate Unit, Reynolds Range Group (central Australia)

Grandite garnet-rich calcsilicate rocks from the Lower Calcsilicate Unit of the regionally metamorphosed Reynolds Range Group (central Australia) crop out along a strike-parallel section in which a transition zone from M2₂ amphibolite to granulite facies rocks is exposed. Across this transition the grandite-rich layers do not show systematic changes in mineral assemblages, compositions and modes, or stable isotope compositions. These layers are deformed by F2₂ folds that are associated with the peak of regional low-pressure/high-temperature metamorphism. Therefore, the grandite-rich layers appear to pre-date regional metamorphism and to have acted as closed chemical systems during prograde M2₂ metamorphism. Mineral assemblages in the grandite-rich layers are consistent with their formation through the infiltration of oxidized, water-rich fluids (X_co2 < 0.1–0.3; log f_o2 -16 to -14). The stable isotope values of calcite (Δ¹³C=-4.2 to -0.8%₀ PDB; Δ¹⁸O = 10.5–14.0%⁰ V-SMOW) and bulk-silicate fractions (Δ¹⁸O = 6.1 to 10.8%) of the grandite-rich layers are most consistent with the infiltrating fluid being from a magmatic source. It is most likely that fluid infiltration occurred during the pre-M2₂ contact metamorphism (M2₁) that affected much of the Reynolds Range Group. The preservation of these assemblages is probably due to their high variance and little pervasive fluid-rock interaction having occurred during M2₂. The clinopyroxene- and feldspar-rich calcsilicate rocks that host the grandite-rich layers contain poikiloblastic grandite garnet that formed during prograde M2₂ metamorphism. Thin marbles that locally occur with the grandite-rich layers contain a third garnet generation that is post- or late M2₂. This grossular-rich garnet occurs in coronas around calcite, plagioclase, clinopyroxene, wollastonite and scapolite. These coronas are consistent with cooling and/or compression. However, because the marble assemblages are themselves overprinted by M2₁ grandite-rich layers the development of coronal garnet does not reflect a continuous P-T-t path. Rather, it more probably reflects the partial re-equilibration of M2₁ contact metamorphic assemblages to post-M2₂ conditions.