R. G. Park and J. Tarney write:

Droop et al. (1999) make a welcome addition to the Lewisian literature in providing the first rigorous published estimates of PT conditions at the Laxfordian meta-morphic peak in the Loch Maree Group (530–630°C, 6.5 kbar, using various methods including calcite-dolomite and garnet-biotite geothermometers and garnet-plagioclase geobarometer). In their discussion of the implications of the relatively high pressure-temperature estimates, they note an apparent contradiction with the tectonic model for the early Laxfordian deformation presented by Coward and Park (1987) and Park et al. (1987). The 1987 tectonic model for the Gairloch area resulted from the recognition that the strong early Laxfordian deformation was associated with a WNW-ESE movement direction implying oblique slip on gently inclined planes, with both strike-slip and dip-slip components. The reason for choosing the dextral/normal-slip option for these movements then was because of the observation that certain of the more prominent shear zones exhibited dextral, NE-down movement-sense criteria. However subsequent more detailed work on the Gairloch shear zones (Lei and Park 1993) showed that the movement history on the shear zones was more complex than first thought, that both sinistral/normal and dextral/reverse movement criteria existed for the early shear zones, but that the main early movements involved overthrusting to WNW, i.e. compressional, in agreement with Droop et al.'s conclusions.

Droop et al.'s estimates of the metamorphic conditions provide timely additional evidence in support of new work by ourselves, together with J.N. Connelly, recently accepted for publication (Park et al. in press), in which we reinterpret the Loch Maree Group (LMG) as an accretionary complex consisting of a stack of slices of oceanic material, together with continentally derived clastics, underthrusting continental basement to the ESE. In this model, the lower parts of the complex would be subjected to the higher pressures, and it is interesting to note that, while Droop et al.'s assertion that the Gairloch outcrops of the LMG present lower-pressure assemblages is true for the main part of the outcrop, it is not true for the southwestern parts, where garnet coexists with hornblende and andesine in the amphibolites (although the assemblages are commonly retrogressed). Furthermore, staurolite is found in the thin band of schists NE of the Aundrary amphibolite and the marble bands 1.5 km further to the NE again contain diopside rather than tremolite as is the case within the main Gairloch outcrop. It would appear therefore that the Gairloch outcrop juxtaposes different metamorphic assemblages; as is very commonly observed in subduction-accretion complexes (Barr et al. 1999).

Our model envisages the earliest movements being compressional, i.e. overthrusting to the WNW (underthrusting to ESE), followed by localized extension, accompanied by retrogression, on certain discrete shear zones in response to tectonic overthickening of the crust. It would be interesting to know therefore whether Droop et al. have evidence of an extensional sense of movement in the shear zones of their ‘late D1’ event at Loch Maree or whether only compressional movement-sense criteria were found.

G. T. R. Droop, L. A. D. Fernandes and S. Shaw reply:

We are interested to learn that Park and Tarney have constructed a new compressional tectonic model for the Loch Maree Group (LMG), and we are pleased to hear that they consider that it is supported by our conclusions on Laxfordian peak-metamorphic PT conditions (530–630°C, 6.5 kbar) and that it meets the implicit requirement for tectonic burial of the LMG to mid-crustal depths.

Park and Tarney are quite correct to point out that minor structures indicative of early Laxfordian (i.e. pre/syn-metamorphic) compressional tectonics within the Gairloch shear zones have indeed been described in the literature, and we acknowledge that evidence for early compression in the Gairloch - Loch Maree region is not merely circumstantial. Lei and Park (1993) concluded that the dominant sense of movement during ‘D2’ (terminology of Park et al. 1987) was sinistral top-to-WNW shear on a major gently NE-dipping shear zone. Provided that original dip of this shear zone was to the NE, thrusting is implied. If this structure can be assumed to be representative of the Laxfordian crust at that time (i.e. symptomatic of a crustal-scale stack of WNW-directed nappes), no discrepancy would exist between the kinematic data and the tectonic requirements of our metamorphic results.

Without the benefit of having seen it, it is difficult to assess exactly how well the new tectonic model of Park and Tarney agrees with our metamorphic data. Our data do, however, permit some general points to be made concerning tectonic setting:

  • the Laxfordian metamorphism of the LMG was of Barrovian type, i.e. of medium-P/T facies series. As such, it is extremely unlikely to have developed purely within an active subduction zone. Being areas of extremely low mantle heat flow, active subduction zones are characterized by metamorphism of high-P/T facies series, manifested as blueschists and eclogites (e.g. Ernst 1977). No such rocks have been found in the LMG. The geothermal gradients implied by our PT data (c. 20–30°Ckm−1) are much higher than those typical of subduction-zone metamorphism (c. 5–8°Ckm−1, e.g. Fry and Barnicoat 1987; Okay and Kelley 1994). Overprinting of high-P mineral assemblages by Barrovian ones is to be expected, however, if a subduction-accretion complex is removed from the site of active subduction and incorporated into a collisional orogen to undergo ‘thermal relaxation’, as is thought to have happened in the European Alps (e.g. Piatt 1986). In that case, one might expect to find local relics of high-P mineral assemblages such as partially amphibolitised eclogite remnants or pseudomorphs after lawsonite. We have not seen any such relics in the LMG; it would be very interesting to know if they exist.

  • different peak-metamorphic conditions seem to have prevailed in different parts the LMG. We were able to detect a significant difference in the temperatures recorded by those parts of the Gairloch and Loch Maree outcrops we sampled. However, we would like to emphasise that we found no significant difference in the metamorphic pressures; within error, both sets of samples gave identical pressures, viz. 6.5 ±1.5 kbar. Until we have a larger set of PT data, it is probably premature to read very much into this result. Never-theless the PT differences between the two outcrops are curious and potentially important. The two groups of rocks appear to have reached different peak temperatures at the same depth (c. 24 km), i.e. appear to have been metamorphosed under somewhat different mean geothermal gradients (22+7/−4 °Ckm−1 at Gairloch, 26+9/−5 °Ckm−1 at Loch Maree) with the higher-grade rocks recording a higher gradient. Such behaviour is predicted for rocks in tectonically thickened crust, but only if the buried rocks have been held at depth for long enough (tens of millions of years?) to heat up, and only if exhumation was sufficiently slow for there to have been a period when structurally deep rocks continued to heat up while high-level rocks cooled, leading to a PT array with a shallow dP/dT slope (England and Richardson 1977). ‘Flat’ PT arrays have been documented in orogenic belts formed by continental collision, such as the Eastern Alps, in which the amphibolite-facies metamorphism significantly post-dated, and was probably a consequence of, tectonic thickening (Droop 1985). Again, our data seem to point to an extended period of heating of the LMG within a ‘mature’ collisional orogen. Information on the PTt paths taken by individual LMG samples could be used to test this idea.

Park and Tarney suggest that there may also be evidence for metamorphic PT variations within the Gairloch outcrop of the LMG. Our sample set, from localities close to Gairloch itself, is too small to show any spatial variations that may exist. The coexistence of garnet, hornblende and andesine in amphibolites from the SW part of the outcrop is suggestive of higher grade conditions there than at Gairloch (though this assemblage occurs in our garnet-hornblende-epidote-schist sample Lms9a). Without knowing details of the mineral assemblages, it is impossible to assess whether or not the diopside and staurolite occurrences mentioned by Park and Tarney imply higher-grade conditions NE of the Aundrary Amphibolite than at Gairloch. Tremolite and diopside need not imply different conditions; they can coexist in stable equilibrium within a single rock. Also, carbonate-silicate reactions in marbles are notoriously sensitive to variations in fluid composition (e.g. Slaughter et al. 1975), so that it is possible for different tremolite-and diopside-bearing assemblages to exist at the same P and T if they have equilibrated with metamorphic fluids of different H2O/CO2 ratio. Likewise, the occurrence of staurolite, on its own, is insufficient evidence for PT variations, as the PT conditions deduced for Gairloch overlap the low-T stability limit of the assemblage staurolite + quartz (Rao and Johannes 1979; Powell and Holland 1990). If, the staurolite coexists with biotite (as in sample L21 from Loch Maree) then the argument for PT variation is much stronger. Clearly, there is a need for a detailed metamorphic survey of the Gairloch outcrop. If PT variations are confirmed in the Gairloch outcrop (either as smoothly varying metamorphic zones, or as abrupt, late-shear-zone-controlled discontinuities), they would be more likely (for reasons stated above), to reflect differential sampling of a post-thrusting thermal overprint than a series of independently metamorphosed thrust slices assembled sequentially in an accretionary complex.

As regards kinematic indicators in the Loch Maree area, we found none that could be unequivocally ascribed to the annealed prograde fabric (‘D1’ of Droop et al. 1999; ‘D1’ of Fernandes 1987). All the kinematic indicators we observed were associated with the intense mylonitic fabric (‘DM’ of Droop et al. 1999; ‘late D1’ of Fernandes 1987) and all that could be interpreted with confidence indicate top-to-NW movement. Whether these reflect compression or extension depends on a series of factors such as the original attitude of the mylonitic foliation. The complex geometry produced by several sets of superimposed folds has hindered determination of the sense of tectonic flow in these rocks, so that we have had to work with the assumption that the original dip was to the SE, which is the principal direction of plunge of the DM lineations. To explain the fact that DM movements were retrograde with respect to the peak of amphibolite-facies metamorphism, one would have to postulate that DM was immediately preceded either by a quiescent period of erosion and cooling or by a phase of extensional low-angle normal faulting at a level structurally above the rocks of the Loch Maree outcrop.