6.2 Geological setting

Leucogranites are widespread within the upper structural levels of the HHCS in the whole Zanskar area. The abundance of these intrusions increases from north-west Zanskar towards the south-eastern parts of this area. In the upper Doda, Suru, Kishtwar and Warwan regions, the occurrence of leucogranites is limited to disseminated dikes and sills (Honegger, 1983; Herren, 1987; Kündig, 1988). The claim of Searle and Fryer (1986) that 30-50% of the granites in this region are of leucogranitic type was thus quite overoptimistic. From the region of the Haptal Tokpo valley, south of Padum, and towards the south-east, the leucogranites start to appear as frequent concordant intrusions with a size that varies between 5 mm and 600 meters (Herren, 1987). The largest amounts of leucogranites in the whole Zanskar area are undoubtedly occurring in the south-westernmost part of the HHCS, which is to say in the region covered by the present work.

In this area, the leucogranite do indeed represent a large part of the HHCS (fig. 6.9) and they occur as a more or less continuous belt of plutons of variable size from the Gianbul valley to the Kamirup . It is in the latter valley that the leucogranites can be seen to show on the surface for the last time, for they have not been mentioned to occur more to the south-east. It is however very likely that they are still present at an unexposed level and that they might be quite close to the surface under the high-grade metamorphic outcrop south of Sarchu.

South-east Zanskar, and especially the Gianbul valley, represents probably one of the most interesting regions of the Himalaya for the study of the leucogranites as it is possible to follow these intrusions all the way from their probable source region up to their summit along this single valley.

The leucogranites in the Gianbul valley form an intrusion complex that can be divided into four zones from base to top (fig. 5.8).

• The first zone is the probable production zone of the leucogranitic melts and is represented by the high-grade migmatites (fig. 6.2) that we have described in chapter 5. The restites in the migmatitic zone still contain significant quantities of both plagioclase and alkali feldspar which suggest that the production of melt occurred through vapour-absent melting of muscovite (Harris et al., 1993).

• The second zone corresponds to the feeder dikes which are rooted into the migmatitic zone and ascend more or less vertically through the gneisses of the HHCS over a distance of about one kilometre (fig 6.3). As this zone is exposed in cliffs which are of very difficult access, we could not get a close view of the relation between the dikes and the surrounding country-rock gneisses.

• The third zone is formed by an approximately one kilometre thick belt of massive leucogranitic plutons fed by the underlying dikes. The lower contact of these plutons is very irregular. One of these massive plutons is beautifully exposed in the northern ridge of the Gumburanjun mountain (fig. 6.4). Even if this outcrop of leucogranites is relatively small compared to those of the Gianbul valley, we will use this name for the whole set of leucogranitic bodies of south-east Zanskar as they are most likely all connected to each other and because this name was already used in the literature (Gaetani et al. 1985, Ferrara et al. 1991; Dèzes et al. 1999). Metric to decametric blocks of country rocks are preserved as xenoliths both at the base and at the top of the plutons (fig. 6.5).

  These rocks have a fine grained equigranular texture and are composed of a large amount of quartz (60-70%), with biotite (20-30%), albite (10-15%) and minor K-feldspar (3-5%) and tourmaline (2-3%). The amount of the rafts decreases gradually towards the centre of the plutons. These blocks have an angular shape and show a well marked foliation which is continuous from block to block and parallel with the main foliation of the country rocks (fig. 6.4). It thus appears that these xenoliths have preserved their original orientation and were not tilted within the plutons as would be expected if they were surrounded by large volumes of leucogranite in a liquid state. This is a strange feature which we tentatively attribute to the fact that the leucogranitic intrusions do not form plutons in the usual sense of the term but that they are the result of a very high concentration of dikes and sills which intruded as continuous pulses and aggregated as massive bodies of leucogranites where the individual veins are indistinguishable from one another. This interpretation is supported by the fact that in the regions outside of the plutons it is often impossible to establish a chronological relation between two dikes, as they seem to be perfectly welded together at their point of junction. Another interesting feature of the xenoliths is that they seem to be affected by some kind of leaching process. One can indeed observe that these rocks, which have no particular reason to represent a different protolith than the rest of the HHCS rocks, do show a texture and a mineralogical composition that is quite different from the usual metapelites or orthogneisses, in the sense that they are almost completely depleted in muscovite and have a fine grained equigranular texture (~ hornfelsic texture). That these rocks also initially contained muscovite is testified by the fact that we could observe, in thin section, a single perfectly rounded muscovite grain preserved inside a quartz crystal. Thus, it seems more than likely that these xenoliths reached metamorphic conditions sufficient for the breakdown of muscovite and that they also represent a source for leucogranitic melts. The overall aspect of these leucogranitic intrusion containing angular rafts of country rocks resembles that of agmatitic migmatites (Mehnert, 1968). The xenoliths should thus be considered as restites.

• The fourth zone represents the structurally uppermost levels of the intrusion complex. Above the roof of the leucogranitic plutons, a gradually decreasing array of leucogranitic veins intrudes the country rock gneisses. These veins ascend vertically for a distance of about 100 meters before most of them are reoriented parallel to the Zanskar Shear Zone (fig. 6.8) and often strongly boudinaged by extensional movements associated to the shear zone (fig. 6.6) Several dikes do however penetrate the ZSZ without being affected by extensional deformation (fig. 6.7), which clearly indicates that these dikes must have intruded the shear zone after ductile movements have ceased. The upper boundary of the leucogranitic intrusion coincides with the kyanite zone and they were never seen above the 10-50 meter thick horizon of calcsilicate rocks, which is nearly always present at this structural level in the whole studied area.

  Both aplitic and pegmatitic dikes are observed towards the top of the intrusion complex and especially in the uppermost zone. The pegmatites (up to 10 meters in thickness) contain quartz, K-feldspar, plagioclase, muscovite, tourmaline, garnet, and beryl. The individual size of these minerals reaches up to 10 cm (yes, even the beryl!). In one sample from the kyanite zone, a large (12 cm) K-feldspar crystal incorporated pre-existing centimetre sized kyanite blades. Aplites, in the absence of biotite and tourmaline are pristine white with red flecks of euhedral garnets. Many of the pegmatitic dikes seem to be late as they are mostly undeformed by the ZSZ.

  The association of leucogranites with pegmatites does not imply hydrous (aH2O=1) conditions during melting. Clemens (1984) estimated that a granitic melt produced by muscovite breakdown would be close to saturation, having ~ 10 wt. % dissolved water at 5 kbar (assuming pure OH on the hydroxyl site in micas). Crystallization of such melts can liberate a large proportion of volatiles and consequently leads to the formation of migmatites without external fluids.