The role of the fluid phase during regional metamorphism and deformation

Evidence from rock microstructures, mass transfer and isotopic exchange indicates that substantial quantities of aqueous fluids are involved in low- and medium-grade regional metamorphism. Similar conclusions are drawn from many retrograde environments, whereas high-grade metamorphic fluids may be melt dominated. The mobile fluids play essential roles in metamorphic reactions, mass transport and deformation processes. These processes are linked by the mechanical consequences of metamorphic fluid pressures (P_f) generally being greater than or equal to the minimum principal compressive stress. Under such conditions metamorphic porosity comprises grain boundary tubules and bubbles together with continuously generated (and healed) microfractures. Deformation results in significant interconnected porosity and hence enhanced permeability. Lithologically and structurally controlled permeability variations may cause effective fluid channelling.
Simple Rayleigh-Darcy modelling of a uniformly permeable, crustal slab shows that convective instability of metamorphic fluid is expected at the permeabilities suggested for the high P_f metamorphic conditions. Complex, large-scale convective cells operating in overpressured, but capped systems may provide a satisfactory explanation for the large fluid/rock ratios and extensive mass transport demonstrated for many low- and medium-grade metamorphic environments. Such large-scale fluid circulation may have important consequences for heat transfer in and the thermal evolution of metamorphic belts.     

Successive overprinting granulite facies metamorphic events in the Anmatjira Range, central Australia

The Anmatjira Range and adjacent Reynolds Range, central Australia, comprise early Proterozoic metasediments and othogneisses that were affected by three, and possibly four, temporally distinct metamorphic events, M_1–4, and deformation events, D_1–4, in the period 1820–1590 Ma. The north-western portion of the range, around Mt Stafford, preserves the effects of ±1820 Ma M₁-D₁, and shows a spectacular lateral transition from muscovite + quartz-bearing schists to interlayered andalusite-bearing migmatites and two-pyroxene granofelses that reflect extremely low-pressure granulite facies conditions, over a distance of less than 10 km. Orthopyroxene + cordierite + garnet + K-feldspar + quartz-bearing gneisses occur at the highest grade, implying peak conditions of ±750°C and 2.5 ± 0.6 kbar. An anticlockwise P–T path for M₁ is inferred from syn- to late-D₁ sillimanite overprinting andalusite, petrogenetic grid considerations and quantitative estimates of metamorphic conditions for inferred overprinting assemblages. The effects of M₁ have been variably overprinted to the south-east by a c. 1760 Ma M₂–D₂ event. Much of the central Anmatjira Range, around Ingellina Gap, comprises orthogneiss, deformed during D₂, and metapelites that have M₁ andalusite and K-feldspar overprinted by M₂ sillimanite and muscovite. The south-eastern portion of the range, around Mt Weldon, comprises metasediments and orthogneisses that were completely recrystallized during M₂–D₂, with metapelitic gneisses characterized by spinel + sillimanite + K-feldspar + quartz-bearing assemblages that suggest peak M₂ conditions of >750°C and 5.5 ± 1 kbar. Overprinting parageneses in metapelitic gneisses imply that D₂ occurred during essentially isobaric cooling. A third granulite facies event, M₃, affected rocks in the Reynolds Range, immediately to the south of the Anmatjira Range, at c. 1730 Ma. A possible fourth event, M₄, with a minimum age of c. 1590 My affected both Ranges, but resulted in only minor overprinting of M_1–3 assemblages. The superimposed effects of M_1–4, mapped for the entire Anmatjira–Reynolds Range area, indicate that only minor or no dislocation of the regional geology occurred during any of the metamorphic and accompanying folding, events. Although the immediate cause of each of the metamorphic events involved advection, the ultimate causes were external to the metasediments and most probably external to the crust.     

Growth and concentration of fibrous sillimanite related to heterogeneous deformation in K-feldspar-sillimanite metapelites

Abstract In foliated K-feldspar-sillimanite metapelites, fibrous sillimanite is commonly concentrated in folia that anastomose between lenticular pods of coarser-grained aggregates rich in quartz, feldspar and biotite, with or without garnet, cordierite and residual andalusite. Many of the folia appear to be limbs of crenulation microfolds. The sillimanite concentrations may be due largely to the ability of fibrous sillimanite aggregates to undergo strong non-coaxial deformation by grain-boundary sliding ('fibre sliding';) without appreciable build-up of dislocations, whereas other minerals are unstable in these zones and so concentrate in lower-strain interfolial zones. Initiation, and especially concentration of the sillimanite in folia, may be assisted by fluid flow and local base-cation leaching, whereby minerals unstable in zones of strong non-coaxial strain are dissolved and removed from these zones.     

Timing of porphyroblast growth in the Fleur de Lys Supergroup, Newfoundland

Abstract In the Fleur de Lys Supergroup, western Newfoundland, inclusion trails in garnet and albite porphyroblasts indicate that porphyroblasts overgrew a crenulation foliation, without rotation, probably during the deformation event that produced the crenulations. Further deformation of the matrix resulted in strong re-orientation and retrograde metamorphism of the matrix foliation, which is consequently highly oblique to the crenulation foliation preserved in the porphyroblasts. The resulting matrix foliation locally preserves relics of the early crenulations, and also has itself been crenulated later in places. Thus the porphyroblasts grew before the later stages of deformation, rather than during the final stage, as had been suggested previously. The new interpretation is consistent with available ⁴⁰Ar/³⁹Ar cooling ages which indicate a late Ordovician-early Silurian metamorphic peak, rather than the Devonian peak suggested by previous workers. The inclusion patterns and microprobe data indicate normal outward growth of garnet porphyroblasts from a central nucleus, rather than as a series of veins as proposed by de Wit (1976a, b). However, the observations presented here support growth of porphyroblasts without rotation, which is implied by the de Wit model.     

Reaction relationships during retrograde metamorphism at Olary, South Australia

Abstract In metapelitic schists of the north-eastern Weekeroo Inliers, Olary Block, Willyama Supergroup, South Australia, syn-S1 and syn-S2 assemblages involving staurolite, garnet, biotite and another mineral, most probably cordierite, were overgrown by large syn-S3 andalusite porphyroblasts, owing to isobaric heating from metamorphic conditions that existed during the development of S2. Conditions during the development of S3 probably just reached the andalusite—sillimanite transition. During the development of S4, at somewhat lower temperatures than those that accompanied the development of S3, the following reaction occurred:
staurolite + chlorite + muscovite ± biotite + andalusite + quartz + H₂O.
The amount of retrogression is controlled primarily by the amount of H₂O added by infiltration. As the syn-S3 matrix assemblage was stable during the development of S4, but the andalusite porphyroblasts were no longer stable with the matrix when H₂O was added, the retrogression is focused in and around the porphyroblasts. With enough H₂O available, and if quartz was consumed before biotite in a porphyroblast, then the following reaction occurred:
staurolite + chlorite + muscovite + corundum ± biotite + andalusite + H₂O.
This reaction allowed corundum inclusions in the andalusite to grow, regardless of the presence of quartz in the matrix assemblage.     

Growth and deformation of porphyroblasts in the Foothills terrane, central Sierra Nevada, California: negotiating a microstructural minefield

Abstract The main porphyroblastic minerals in schists and phyllites of the Foothills terrane, Western Metamorphic Belt, central Sierra Nevada, California, are cordierite and andalusite (mostly chiastolite). Less commonly, biotite, muscovite, chlorite, garnet or staurolite are also present as porphyroblasts. The variety of porphyroblast and matrix microstructures in these rocks makes them suitable for testing three modern hypotheses on growth and deformation of porphyroblasts: (1) porphyroblast growth is always syndeformational; (2) porphyroblasts nucleate only in low-strain, largely coaxially deformed, quartz-rich (Q) domains of a crenulation foliation and are dissolved in active high-strain, non-coaxially deformed, mica-rich (M) domains, the spacing between which limits the size of the porphyroblasts; and (3) porphyroblasts generally do not rotate, with respect to geographical coordinates, during deformation, provided they do not deform internally, so that they may be used as reliable indicators of the orientation of former regional structural surfaces, even on the scale of orogenic belts. Some porphyroblast–matrix relationships in the Foothills terrane are inconsistent with hypotheses 1 and 2, and others are equivocal. For example, in many rocks it cannot be determined whether the porphyroblasts grew where the strain had already been partitioned into M and Q domains, whether the porphyroblasts caused this partitioning, or both. Although most porphyroblasts appear to be syndeformational, as predicted by hypothesis 1, observations that do not support the general application of hypotheses 1 and 2 to rocks of the Foothills terrane include: (a) lack of residual crenulations in many strain-shadows and alternative explanations where they are present; (b) absence of porphyroblasts smaller than the distance between nearest mica-rich domains; (c) nucleation of crenulations on existing porphyroblasts, rather than nucleation of porphyroblasts between existing crenulations; (d) presence of micaceous ‘arcs’in an undifferentiated matrix against some porphyroblasts, suggesting static growth; (e) absence of crenulations in porphyroblastic rocks showing sedimentary bedding; and (f) porphyroblasts with very small, random inclusions, which are probably pre-deformational. Similarly, porphyroblasts that have overgrown sets of crenulations and porphyroblasts with micaceous ‘arcs’are probably post-deformational, at least on the scale of a large thin section and probably over much larger areas, judging from mesoscopic structural evidence. Some porphyroblasts in rocks of the Foothills terrane do not appear to have rotated, with respect to geographical coordinates, during matrix deformation, in accordance with hypothesis 3, at least on the scale of a large thin section. However, other porphyroblasts evidently have rotated. In some instances, this appears to be due to mutual interference, but many apparently rotational porphyroblasts are too far apart to have interfered with each other, which indicates that the rotation was associated with deformation of the matrix. The occurrence of planar bedding surfaces adjacent to porphyroblasts about which bedding and/or foliation surfaces are folded suggests rotation of the porphyroblasts during non-coaxial flow parallel to bedding, rather than crenulation of the matrix foliation around static porphyroblasts. It appears that porphyroblasts may rotate during deformation if the matrix is relatively homogeneous, so that the strain is effectively non-coaxial. This may occur after homogenization of a matrix in response to the strongest degree of crenulation folding, whereas the same porphyroblasts may have been inhibited from rotating previously, when strain accumulation was partitioned in the matrix.     

Metamorphism and deformation of mafic and felsic rocks in a magma transfer zone,Stewart Island,New Zealand

In the mingled mafic/felsic Halfmoon Pluton at The Neck, Stewart Island (part of the Median Batholith of New Zealand) some hornblende gabbros and diorites retain magmatic structures, whereas others show evidence of major changes in grain and inclusion shapes, and still others are amphibolite‐facies granofelses with few or no igneous relicts. These mafic to intermediate magmas crystallized in felsic magma relatively quickly, with the result that most deformation occurred at subsolidus conditions. It is suggested that mafic‐intermediate rocks with predominantly igneous microstructures spent less time in the magmatic system. The metamorphism of the mafic rocks appears to be ‘autometamorphic’, in the sense that elevated temperatures were maintained by magmatic heat during subsolidus cooling. Elevated temperatures were maintained because of repeated sheet injection and subconcordant dyke injection of hot basaltic and composite mafic‐felsic magmas, into a dominantly transtensional, km‐scale, outboard‐migrating, magmatic shear zone that operated semi‐continuously for between c. 140 and c. 130 Ma. Complete cooling occurred only when the system evolved to transpressional and the locus of magmatism migrated inboard (southward) between c. 130 and c. 120 Ma, associated with solid‐state mylonitic deformation. Intermingled granitic rocks escaped metamorphism, because they remained magmatic to lower temperatures, and experienced shorter and lower‐temperature subsolidus cooling intervals. However, the felsic rocks underwent relatively high‐temperature solid‐state deformation, as indicated by myrmekite replacing K‐feldspar and chess‐board subgrain patterns in quartz; locally they developed felsic mylonites. The felsic rocks were deformed in the solid state because of their high proportion of relatively weak minerals (quartz and biotite), whereas the mafic rocks mostly escaped subsolidus deformation, except in local high‐strain zones of hornblende‐plagioclase schist, because of their high proportion of relatively strong minerals (hornblende and plagioclase). We suggest that such contrasting microstructural features are diagnostic of long‐lived syntectonic magma transfer zones, and contrast with the more typical complex, batholith‐scale magma chambers of magmatic arcs.     

Problems with inferring P–T–t paths in low-P granulite facies rocks

Microstructural evidence commonly is used to infer metamorphic reactions, which are used to infer pressure–temperature–time ( P–T–t ) paths. However, this approach in low- P /high- T  (LPHT) granulite facies metamorphic terranes has two main problems. (1) Microstructural evidence may be inconclusive, so that reactions cannot be inferred with confidence. In particular, relative timing of mineral growth inferred from inclusions, moulding relationships and foliation–porphyroblast relationships is commonly ambiguous or invalid. The most reliable indicators of metamorphic reactions are partial pseudomorphs and corona structures, especially if symplectic intergrowths (indicating simultaneous growth of two or more minerals) are involved. (2) Even reactions that can be inferred with confidence do not indicate unique P–T  trends, owing to P–T  slopes of reaction curves. Where successive reactions can be shown to have occurred in the same rock, a line or curve joining reaction-curve intersections gives an apparent single-event path. However, isotopic evidence is needed to prove that polymetamorphism (involving more complex paths making fortuitous intersections with the apparent single-event path) did not occur. Although these problems are well known, their importance is not always emphasized in metamorphic investigations.
The difficulties are illustrated by published work on P–T–t paths for Proterozoic LPHT granulite facies rocks of central Australia and Antarctica. Recent work in Antarctica has shown that P–T–t paths may be episodic and more complex than the simple, single-event paths commonly inferred from microstructural evidence alone.     

Sillimanite and andalusite produced by base-cation leaching and contact metamorphism of felsic igneous rocks

Abstract Contact metamorphism adjacent to a porphyritic quartz-monzodiorite at Kentucky, New South Wales, Australia has produced hornfelses in porphyritic leucogranite at a peak temperature of about 650–700° C and a maximum confining pressure of about 2 kbar (200 MPa). A gradation appears to exist from normal slightly peraluminous to modified strongly peraluminous metagranite hornfelses, which have also been enriched in sulphur. The strongly peraluminous hornfelses, containing cordierite, andalusite, sillimanite, biotite, pyrite and pyrrhotite, retain residual porphyritic igneous microstructures. These rocks appear to have been formed by leaching of base cations, during and possibly just before the contact metamorphism. Folia of fibrous sillimanite anastomose between lenticular grains of quartz and feldspar and truncate igneous zoning in plagioclase grains, suggesting that cation leaching and solution transfer occurred during growth of the sillimanite. Fibrous sillimanite also grew in grain boundaries of polygonal aggregates formed by the contact metamorphism. Therefore, at least some of the cation leaching appears to have occurred at the highest metamorphic grade. Metasandstones that are locally strongly peraluminous adjacent to the monzodiorite stock also, have probably undergone similar leaching.