The Emplacement and Origin of Granite
© Copyright, 1995 by R.A.Kanen, All Rights Reserved
Granitic magma is a general term used to describe magma that is similar in composition to granite; that is, containing greater than 10% of quartz. Outcrop of plutonic granite on the earth's surface requires some kind of erosion to expose the buried granite. Granites may take the form of batholiths; sills and sheets; swarms of plutonic intrusions or migmatite complexes. They form the major part of surface exposure of continental crust. Plutonic rocks also exist in the oceanic crust; however, by geophysical methods and drilling, these have been determined to predominantly be of basic or ultrabasic composition. Granites are associated with volcanic areas, continental shields and orogenic belts. To explain their emplacement, it is first necessary to obtain an understanding of their origin. The study of outcrops, geophysical surveys and of its extrusive equivalent, rhyolite, are some of the original methods used. In general, two feasible theories resulted. One, known as the magmatic theory, states that granite is derived by the crystal fractionation of magma. The second, known as the granitization theory states that granite is formed "in situ" by ultrametamorphism. There is evidence to support both theories and current thinking is that magma forms from both processes; in many instances, from a combination of the two.
The magmatic theory involves the use of the Bowen Reaction Series. Thus, if crystal fractionation of a magma of tholeiitic basalt composition were to occur, one of its end products would be granite. In many places, emplacement of granite plutons is synchronous to volcanic eruptions. They commonly form ring complexes around 10 km in diameter with volcanic remnants that have subsided into the couldron as central blocks. This has occurred in the Permian Oslo Graben Province (Carmichael et al., 1974). They also occur as stocks and plutonic intrusions near volcanic centers which consist of granodiorite in an andesitic volcanic province. The former are generally alkali granites in a rhyolitic province. Such plutons typically show sharp contacts; a lack of deformation in the country rock, chilled margins and contact aureoles. All of these phenomena suggest that granite was emplaced as a liquidus magma. Chemically, there is similarity in the composition of many granite plutons to their extrusive associates, the andesite-dacite-rhyolite series of rocks. This suggests there is some kind of relationship between the emplacement of granite plutons and volcanism. Some of the theories that have been postulated include that granite is a surface expression of magma derived from a deep seated batholith and that plutons are derived from the same source as the extrusive but along separate supply routes. Gilluly (1963) has found that volcanism in the Western United States has been more prolific, continuous and diversified than granite plutonism. He found that at some areas and at certain periods, volcanic products were mainly andesitic, at other times and another place, they were mainly basaltic and so on. All types were very prolific. In terms of plutonic rocks, quartz diorites, granodiorites and quartz monzonites were dominant. He concluded plutonism must depend on processes whose time and scale is of essentially a different order of magnitude from those of volcanism and tectonism. He also suggests that the volcanic magmas have been derived from the oceanic crust and plutonic magmas from more siliceous materials.
An interesting question in magmatic evolution of granite batholiths is how is the immense amount of country rock removed to make room for the batholith? Various mechanisms have been proposed: massive explosions removing the country rock; formation of ring dykes and couldron subsidence; lifting of the country rock and subsequent erosion; and digesting of the country rock as the magma rises. Not one of these would adequately explain its removal on its own but rather it is more likely that all of these occur at some time or other.
The granitization theory explains the origin of granite by the process of ultrametamorphism or anatexis. Anatexis is defined as the melting of pre-existing rock to give granite. The "crux" of the granitization theory is migmatites. Migmatites consist of two components: one light coloured granitic component, called neosome, and a dark metamorphic component called paleosome. Both components have been ultra-mixed. In the granitization theory it was thought that migmatites were rocks in the process of becoming granite. Thus, the neosome, granitic component, was anatexic component; in which case, it may have formed by partial melting of the rock in place and segregation of the melt from the solid, or migration of the melt from its source of origin and intrusion into the host rock. The resultant migmatite which formed by the first process is called a venite, the migmatite formed by the second process is an orterite. Mehnert (1963) reinforces this idea. He has mapped granodioritic masses up to 10km in diameter with a homogenous center and increasingly heterogeneous material towards the outer border. There is no distinct contact of the granodiorite with the country rock, but rather grading into a migmatite zone and finally a gneiss containing eyes of oligoclase and potash feldspar.. He interprets this as meta-greywacke which has been completely fused in the center, granodiorite zone, and partly fused in the migmatite zone. He proposes the outer, gneissic zone, is metamorphic in origin, but due to metasomatism rather than fusion of a melt. In more recent years, it has been recognized that migmatites can form in various other ways. These are granitization, by ion exchange and diffusion, especially with K+ and Na+; mobilization and injection of granitic material from depth, an; and alkali metasomatism, using aqueous pore solutions as a medium. Thus, granites can form both by magmatism and granitization or a combination of the two.
In modern times, it has been shown that the relationship between regional metamorphism and the formation of granite (ultrametamorphism) is more complex than originally thought. Autron et. al.(1970) shows that the formation of large volumes of granite have formed during periods where there has been no metamorphism; for example, the Lower Silurian Caledonian Granites. However, crustal thickening was occurring and it is thought that this takes place as a consequence of crustal shortening by collision and under plating by subduction. The local thickening of the crust provides a sufficient rise in temperature at depth for crustal melting to occur. This may rise in several different ways to produce granite. Experimentally, it has been proven that granite can form by remelting of the crust; however, it has also been shown that the temperatures and water concentrations at depth might not be high enough for remelting to occur (Brown and Hennessy, 1978) unless magma from the mantle influenced in some way.
When this theory is applied to subduction zone areas, remelting of the underplated crust will not occur because of a lack of heat and volatiles. In terms of a collision type situation between two plates, it is thought sufficient water is available from the dehydration of micas during metamorphism of the metasedimentary crustal wedges for remelting to occur. Lochenbruch(1968) suggests that although this may occur, it is impossible for granitic melts of such origin to be homogenous. However, other authors have argued that homogenization can occur fairly easily (e.g. Talbot, 1971).
An interesting problem in an anatexic origin is explaining how the granite melt separates from its associated solid material, especially in pseudo-plastic, or Bingham bodies. Weertman (1971) proposed shear melting as an important part of the mechanism.
Various mechanisms to explain how magma rises through the crust have been discussed and proven acceptable. These include gas fluxion, tectonic squeezing, expansion on melting, seismic pumping and diapirism. All of them are related to the moving of magma down a pressure gradient. Opposing the upward driving force of these mechanisms is drag. Drag increases rapidly as the magma cools, especially when it collects large amounts of xenoliths. Thus, different parts of the magma body ma move upwards at different rates. This provides another explanation for the lack of compositional homogeneity and difference in internal contacts.
Magmas make their way up through the crust via major lineaments, such as fault zones. This is very evident in the Andes and the Coastal Batholith of Peru (Pitcher and Bussel, 1977). Leake (1978) suggested that not only do lineaments provide a route for magmas but also generate them where deep faults cause large pressure reductions and shear zones, where segregation of crystal mushes and Bingham Bodies could occur. It is thought that formation of massive granite bodies via lineaments is restricted to fractured continental plate edges, where massive lineaments are present. In areas which have not undergone large scale faulting, intrusions are more dispersed.
Estimating the depth at which a magma is emplaced is very difficult. However, a very rough approximation could probably be obtained by the fluid inclusions in minerals; the composition of the minerals; for example the aluminum content of hornblende and the ordering of alkali feldspars. However, this is restricted, as feldspars may have reacted to weathering and metamorphism. Furthermore, in the formation of depth models, adjacent intrusions are inferred to be syndepositional in origin and to have a vertical continuity, for example, Buddingtons epi-, meso- and kato-zones. Another depth model recognized by Eskola(1938) and Read (1950) proposes a granite series showing different features at different depths. The first zone, zone of differential anatexis, is the lowest level in the crust. Some granitic magma is formed in situ and mobilization begins, venites are also formed. The zone of injection, or potash metasomatism, is the zone in which the crystal mush and anterites form. Portions of the magma become more liquid and rise to its upper surface. Metasomatism also occurs, indicated by large k-feldspar crystals. In the upper zone, regional metamorphism precedes or accompanies emplacement of the magma. Many veins and sharp contacts indicate the fluid nature of the granite. Fyfe(1970) envisages a tear dropped shaped magma melt with tail inverted rising upwards by diapirism through a cooler, more solid country rock which may have left on it an imprint, in the form of migmatites. "..some migmatite zones might represent a region through which a drop [evolving pluton] passed rather than a region where melting started"
According to Pitcher (1979) granite magma may be emplaced forcefully or permissively into the upper crust. During forceful intrusions, the pushing aside and updoming of country rock occurs. In some cases, they have been found to their way through the crust, especially where there is little overburden. Here, there is little or no evidence of strain in the immediate country rock. In deeper environments, radial distension occurs. There is often little evidence of drag occurring but radial shortening as indicated by strain markers, clasts in the country rock and xenoliths in the intrusive. Pitcher (1979) says, "that in examples of expanding diapirs in Donegal the superimposed deformation aureoles extending almost exactly as far as the thermal recystallization aureoles, indicating that plastic deformation depended upon the pre-heating of the envelope." He goes on to say, " the succession of metasediments entering the aureole of the Ardara multiple diapir [at Donegal] is thinned to at least 10% of the original thickness, and although the exact deformational path which leads to this result cannot be deduced from the presently observed fabric (Dixon, 1975), Holder (1978) showed that these are incremental increases in strain in the aureole rocks of the Ardara Pluton which correlate both with the several phases of metamorphism and the intrusion of magmatic pulses." In other words, evidence for the occurrence of radial distension is very strong, forceful intrusion (forceful punching) is on a much weaker footing. It may also explain why there is no evidence of the envelope surrounding the intrusion, flowing around and behind it. Furthermore, new pulses of magma into the pluton, expanding and contracting it, suggest plutons are continuous tubes connected to the source, not individual drops (Pitcher, 1979).
Mechanisms for the permissive intrusion of granitic magmas include subsidence, couldron upheaval and stoping. These are characteristic of shallow seated, sub volcanic or aureole type granite plutons. The upward thrust combined with thermal expansion upon cooling can still result in updoming and fracturing of the overlying material. Fracturing of the country rock to allow intrusion of magma is thought to be mainly due to thermal expansion. However, new fractures may also form from the seismic shocks which accompany the emplacement of magma. Hydraulic fracturing could extend the initial fractures. Myers (1975) contends "…that gas penetration along micro-breccia zones entrains the fragments, forming a proto-tuffite, so opening a pathway for the advancing gas charged magmas". This allows easy penetration of the country rock and the prising (stoping) off of xenoliths. A lot of the supportive evidence for this process is in the form of ring dykes containing large amounts of xenoliths. Widening of the channels by gas clearing and gas entrainment may also occur in a vertical sense, thereby removing country rock up or down. Such a mechanism is necessary to explain the removal of the country rock, especially since ring dykes are vertical or dipping inward.
Removal of country rock by subsidence is a proven fact. In some places the degree of subsidence can be measured. Nested or stacked intrusions have been explained by the subsidence of a central block in a ring dyke and infilling of the space left above it by pulses of magma. A widely accepted theory is that they form by magma being pumped into a pluton which has a solid outer surface, but still crystallizing inner core.
Whether the couldron or diapir will form depends upon the contrast in ductility between the country rock and the magma. Pitcher (1979) says it is wrong to assume that the ductility contrast decreases with depth. It depends on other factors also, such as, deformation and metamorphism of the crust. As deformation increases, ductility decreases, therefore, the early intrusions are more likely to be diapiric and the older intrusions in the form of a couldron. However, diapirs have also formed at the end of a sequence, therefore, low viscosity, liquid magmas may be injected to form couldrons and higher viscosity, and crystal mushes may form diapirs.
In terms of classifying granitic magmas according to their origin, Pitchers (1979) classification scheme is probably best. In terms of orogenic environments, he recognized a crustal S-type produced by anatexis and which is compositionally restricted; and a mantle I-type derived from a basic igneous source and which is compositionally expanded. To differentiate between the two, measurement of the Sr87/Sr86 ratio is done. Those that are higher than 0.7060 tend to be S-types, those lower, I-types.
The origin of emplaced granitic magma is diverse and mechanisms used to explain its emplacement varied. As techniques for studying granite become more refined, a greater understanding of them will result.
Carmichael, Turner and Veerhoogen, "Igneous Petrology", McGraw Hill, 1974
Gilluly, I, "The Tectonic Evolution of the United States", Q.J. Geol. Soc. London, v.119, 1963
Pitcher, W.S., "The Nature, Ascent and Emplacement of Granitic Magmas", J. Geol. Soc. of London, v.136, pg 627-662, 1979
Read, H.H.,"The Granite Controversy", Murby, 1957.