post-collisional strongly peraluminous granites.pdf

Ž . Lithos 45 1998 29–44 Post-collisional strongly peraluminous granites Paul J. Sylvester ),1 Research School of Earth Sciences, The Australian National UniÕersity, Canberra, ACT 0200, Australia Received 28 January 1998; accepted 25 June 1998 Abstract Ž . Strongly peraluminous SP granites have formed as a result of post-collisional processes in various orogens. In ‘high-pressure’ collisions such as the European Alps and Himalayas, post-collisio
of 16
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Ž . Lithos 45 1998 29–44 Post-collisional strongly peraluminous granites Paul J. Sylvester  ) ,1  Research School of Earth Sciences, The Australian National Uni Õ ersity, Canberra, ACT 0200, Australia Received 28 January 1998; accepted 25 June 1998 Abstract Ž . Strongly peraluminous SP granites have formed as a result of post-collisional processes in various orogens. In‘high-pressure’ collisions such as the European Alps and Himalayas, post-collisional exhumation of overthickened crust Ž . ) 50 km , heated by radiogenic decay of K, U and Th during syn-collisional thickening, produced small- to moderate- Ž . volume, cool  - 875 8 C SP granite melts with high Al O  r TiO ratios. In ‘high-temperature’ collisions such as the 2 3 2 Ž . Ž . Hercynides and Lachlan Fold Belt LFB , there was less syn-collisional crustal thickening  F 50 km . Crustal anatexis wasrelated to post-collisional lithospheric delamination and upwelling of hot asthenosphere, forming large-volume, hot Ž . Ž . G 875 8 C SP granite melts with low Al O  r TiO ratios. Both clay-rich, plagioclase-poor  - 5% pelitic rocks and 2 3 2 Ž . clay-poor, plagioclase-rich  ) 25% psammitic rocks have been partially melted in high-pressure and high-temperature Ž . collisional orogens, with the pelite-derived SP granites tending to have lower CaO r Na O ratios  - 0.3 than their 2 psammite-derived counterparts. The predominance of pelite-derived SP granites in the Himalayas and psammite-derived SPgranites in the LFB suggests that mature continental platforms made up more of the accreted crust in the Himalayan collisionthan in the LFB.  q 1998 Elsevier Science B.V. All rights reserved. Keywords:  Granites; Plate collision; Suture zones; Orogeny; Anatexis 1. Introduction Subduction of oceanic lithosphere has invariablyled to closure of ocean basins, collisions betweenisland-arcs and continental blocks, and the formationof granite magmatism along the resulting sutures Ž Pitcher, 1983; Pearce et al., 1984; Harris et al., . 1986 . This collisional tectonic setting is distinct ) Tel.:  q 709-737-4736; Fax:  q 709-737-2589; 1 Present address: Department of Earth Sciences, MemorialUniversity of Newfoundland, St. John’s, NF, Canada A1B 3X5. from the preceding one of oceanic plate subduction Ž . and commonly involves 1 an early syn-collisionalstage of thrusting and folding, resulting in crustal Ž . thickening; and 2 a later post-collisional stage of strike–slip and extensional faulting, reflecting ad- justment of the accreted blocks to waning compres-sional stresses and leading to final stabilization.Collision-related granites exhibit a variety of compositions, commonly divided into calc-alkaline Ž . Ž . Pitcher, 1983 , alkaline Sylvester, 1989 and, theparticular concern of this paper, strongly peralumi-nous varieties. Peraluminous granites have ratios of  Ž . molecular Al O r CaO q Na O q K O or ‘A r 2 3 2 2 0024-4937 r 98 r $ - see front matter q  1998 Elsevier Science B.V. All rights reserved. Ž . PII: S0024-4937 98 00024-3  ( )P.J. Syl Õ ester  r  Lithos 45 1998 29–44 30 Ž . CNK’ exceeding 1, but strongly peraluminous SPgranites possess A r CNK ratios  G 1.1. This is re-flected in a mineralogy that includes highly-aluminous primary phases such as muscovite, Ž cordierite, garnet, tourmaline and andalusite Miller, . 1985 . Many are leucogranites, by definition contain-ing  - 5% mafic minerals, but others are granites Ž . Ž s.s.  , granodiorites and even some tonalites Chap- . pell and White, 1992 . SiO concentrations are gen- 2 erally  ) 67 wt.%, initial  87 Sr r 86 Sr ) 0.706, initial 18 Ž ´   - y 2 and  d   O ) q 9.5‰ Le Fort et al., 1987; Nd . White and Chappell, 1988 .Collision-related SP granites appear to have srci-nated through a diverse set of processes. Some areexposed along thrust sheets or shear zones and aredeformed; others crop out as completely post-tectonic Ž . plutons Searle et al., 1997; Finger et al., 1997 .Some may have formed from partial melts of quartz- Ž . ofeldspathic meta-igneous orthogneiss crustal rocks Ž . Miller, 1985 or by reaction between basaltic melts Ž . and crustal rocks Patino Douce, 1995 . However, ˇ many others contain metasedimentary enclaves Ž . White and Chappell, 1988 or are associated with Ž . migmatitic paragneiss Le Fort et al., 1987 , suggest-ing an origin wholly or dominantly from partialmelts of metasedimentary rocks in the crust. Both‘mature’ argillaceous-rich pelitic rocks, principally Ž . meta-shales Searle et al., 1997 , and ‘immature’quartzofeldspathic-rich psammitic rocks, principally Ž . meta-greywackes White and Chappell, 1988 , seemto have been involved. Anatexis may have resulted Ž from decompression of over-thickened crust Le Fort . et al., 1987 , fluid focusing along crustal shear zones Ž . Strong and Hanmer, 1981 , or advection r conduc-tion of mantle-derived heat into the lower crust Ž . Wickham and Oxburgh, 1987 .In part because of the diversity of SP granitemagmatism, there is considerable uncertainty aboutits relationship to collisional tectonism. Are mostcollision-related SP granites the products of syn-col-lisional crustal thickening or are they related tosubsequent post-collisional events? Were all colli-sion-related SP granite generated at similar tempera-tures in the crust or were certain collisional orogenscharacterized by ‘hotter’ granite magmas than oth-ers? What would hot SP granite magmas indicateabout the evolution of the lithosphere in collisionzones? Do metasedimentary-derived SP granites havedominantly pelitic sources in certain collision-relatedorogens and psammitic sources in others? If so, whatdoes this suggest about the nature of the crustalblocks involved in the collisions?This paper addresses these questions and presentsa general framework in which collision-related SPgranites can be understood. In essence it is arguedthat there is a spectrum of collisional orogens, whichis reflected in the diversity of collision-related SPgranites. In ‘high-pressure’ collisions, the crust wasthickened to more than  ; 50 km by thrust stacking Ž . and shortening. ‘Cool’  ; 875 8 C or less SP granitemelts formed primarily by a two-step process involv- Ž . ing 1 syn-collisional, in situ radiogenic heating; Ž . and 2 post-collisional decompression melting alongsome of the deep-rooted thrusts. In ‘high-tempera-ture’ collisions, the crust remained comparatively Ž . Ž . thin  F 50 km , and ‘hot’  G 875 8 C SP granitemagmatism was primarily the product of a post-colli-sional, mantle-derived heat source. This model isdeveloped by examining both the geologic character-istics and chemical compositions of collision-relatedSP granites and, in particular, their CaO r Na O and 2 Al O r TiO ratios. 2 3 2 2. Timing of SP granites: syn-collisional or post-collisional? Ž . Ž . Pitcher 1983 , Pearce et al. 1984 and Harris et Ž . al. 1986 suggested that collision-related SP granitesformed during the early syn-collisional phase of crustal shortening and stacking, and contrasted themwith a later post-collisional phase of potassium-richcalc-alkaline or ‘I-type Caledonian’ granites, formedduring uplift, extension and strike–slip faulting.However, it is now apparent that the large majorityof collision-related SP granites are best described as‘post-collisional’, in that they were emplaced  after the climax of crustal thickening . This is certainlytrue of the voluminous and widespread 340–300 MaSP granites of the Hercynian orogeny of Europe:almost all post-date an early collision-related Ž . medium-pressure Barrovian metamorphic event andare instead associated with later high-temperature r low-pressure regional metamorphism and extensional Ž and strike–slip fault movements Strong and Han- . mer, 1981; Wickham, 1987; Finger et al., 1997 . The  ( )P.J. Syl Õ ester  r  Lithos 45 1998 29–44  31 SP granites are spatially and temporally associated Ž with calc-alkaline granites Emmermann, 1977; Frasl . and Finger, 1991; Pamic et al., 1996 that wouldundoubtedly be classified as post-collisional usingthe Pitcher–Pearce–Harris scheme.Collision-related SP granites in the European Alpsalso appear to be post-collisional. Following colli-sion-related high-pressure regional metamorphism at45–35 Ma, a trivial amount of SP granite formedbetween 33–25 Ma, along with moderate volumes of  Ž calc-alkaline granite Visona and Zirpoli, 1984; von . Blanckenburg, 1992; Bellieni et al., 1996 and minorextension-related shoshonitic r ultrapotassic magma- Ž . tism Venturelli et al., 1984 . The granites wereemplaced along the Periadriatic–Insubric strike–slipfault system during post-folding north–south trans-gression and east–west extension, commonly de- Ž scribed as ‘post-collisional’ e.g., Altherr et al., . 1995 .In the Caledonian orogeny of Britain, the main440–390 Ma phase of plutonism is the type exampleof the post-collisional calc-alkaline granite suite of  Ž . Pitcher 1983 , but this event also involved the em-placement of numerous SP granites, particularly south Ž of the Iapetus suture Hall, 1972; Harmon et al., . 1984; O’Brien et al., 1985; Sweetman, 1987 . Thereare some SP granites that formed before the mainphase of plutonism, during or just after the climax of  ; 480 Ma Barrovian metamorphism of the GrampianOrogeny. However, as this tectonothermal event oc-curred before final closure of the Iapetus Ocean, it isunclear how these granites relate to the Caledonian Ž . collision Harmon et al., 1984 .Even in the High Himalaya, moderate volumes of so-called syn-collisional SP granites were emplaced Ž . between 24–14 Ma Scharer et al., 1986 , probably ¨  just after the climax of high-pressure metamorphismand at the beginning of a period of rapid uplift and Ž . exhumation Searle et al., 1997 . At just about the Ž . same time  ; 20 Ma–present , small volumes of shoshonitic r ultrapotassic volcanics and SP granites Ž . McKenna and Walker, 1990 began forming to thenorth on the Tibetan Plateau, during extension-drivencollapse that is widely described as ‘post-collisional’ Ž . e.g., Turner et al., 1996 . Thus, Himalayan SPgranites formed very late in the period of collision-related crustal thickening, some 30 million years or Ž . more after the collision began Scharer et al., 1986 , ¨ and it is likely their appearance marked the start of post-collisional events in this orogen. 3. High-pressure vs. high-temperature collisionalorogens The well-studied Himalayan orogen is often usedas a template for collisional orogenesis; however ithas been long known that there is significant diver- Ž . sity among collisional belts. Zwart 1967 introducedthe concept of a ‘duality of orogens’, referring inpart to the higher pressure regional metamorphism of the Alps as compared to that in the Hercynides. Thisconcept remains relevant, and helps understand thediversity of post-collisional SP granites.In the Alps and Himalayas, widespread high-pres-sure metamorphism preceded SP granite plutonism Ž . Scharer et al., 1986; von Blanckenburg, 1992 , sug- ¨ gesting that syn-collisional crustal thickening wasextreme in these orogens. For instance, the presenceof pyrope–coesite rocks in the Alps suggests subduc-tion of sialic crust to depths of   ; 130 km during the Ž Alpine collision, at least locally Gebauer et al., . 1997 . Moreover, seismic data indicate that the crust Ž beneath Tibet is now  ; 70 km thick Hirn et al., . 1984 , twice the thickness of ‘normal’ continentalcrust. Post-collisional exhumation of the overthick-ened crust, after an ‘incubation’ period of in situ Ž radiogenic decay of heat-producing elements K, U, . Th at depth, probably led to the production of smallto moderate volumes of SP granites in these belts Ž . Le Fort et al., 1987; Searle et al., 1997 . Modelling Ž of this process of anatexis Zen, 1988; Thompson . and Connolly, 1995 indicates that the extent of melting would have increased with the length of theincubation period, and decreased with the exhuma-tion rate of the thickened crust. This may explainwhy SP granites are much rarer in the Alps than inthe Himalayas. Although both collisions began at ; 50 Ma, exhumation of high-pressure Alpine meta-morphic rocks occurred somewhat earlier than in the Ž Himalayas  ; 35 vs. 25 Ma; Scharer et al., 1986; ¨ . von Blanckenburg, 1992 and, at least locally, oper- Ž ated at remarkably rapid rates 20 km r Ma; Gebauer . et al., 1997 .Compared to the Alpine and Himalayan colli-sions, crustal thickening during the Hercynian colli-  ( )P.J. Syl Õ ester  r  Lithos 45 1998 29–44 32 sion was probably modest, as reflected in the early,syn-collisional regional metamorphism being more Ž . generally of a medium-pressure Barrovian type Ž . e.g., Finger et al., 1997 . Estimates of maximumcrustal thicknesses produced during collisions in an-cient orogens are necessarily model dependent but Ž . the balanced cross sections of Behrmann et al. 1991suggest  ; 50 km for the Hercynides. Moreover,large volumes of Hercynian granites are associatedwith a widespread post-collisional high-tempera-ture r low pressure metamorphic event that is largelyabsent in the Alpine and Himalayan collisions. Insitu radiogenic heating will not typically produceextensive high-temperature r low pressure metamor- Ž phism and melting in  - 50-km thick crust Thomp- . son and Connolly, 1995 . Instead, the high-tempera-ture metamorphism and SP granite genesis probablyreflect a large mantle heat source at the base of the Ž crust caused by asthenospheric upwelling Wickham . and Oxburgh, 1987; Thompson and Connolly, 1995 Ž following slab break-off Davies and von Blancken- . Ž burg, 1995 and lithospheric delamination Black and . Ž Liegeois, 1993 . Modelling e.g., Davies and von ´ . Blanckenburg, 1995 indicates that if the astheno-sphere was able to ascend to depths of 50 km or less,in the wake of lithosphere thinned by delamination,mantle melting would have been possible and basalticmagmatism injected into the crust could have led towidespread melting and post-collisional granite gen-esis. Unlike in the Alps and Himalayas, HercynianSP granites are associated with large volumes of contemporaneous post-collisional calc-alkaline gran-ites, which exhibit evidence for interaction withwidespread mantle-derived magmas that may have Ž been derived from the asthenosphere Rottura et al., . 1991; Pamic et al., 1996; Finger et al., 1997 .In the Alps and Himalayas, the main evidence forpost-collisional mantle melting is small-volumeshoshonitic r ultrapotassic rocks, which probably re-flect small-degrees of melting of the mantle litho- Ž . sphere Venturelli et al., 1984; Turner et al., 1996 .Post-collisional calc-alkaline granites are rare or ab-sent in the Himalayas and form only small to moder- Ž . ate volumes in the Alps Bellieni et al., 1996 .Because syn-collisional lithospheric thickening wasmuch more substantial in the Alps and Himalayas, itis likely that post-collisional lithospheric delamina-tion was insufficient to allow upwelling astheno-sphere to rise to melting depths. Asthenosphericmantle heat would have been transferred to the crustand mantle lithosphere by conduction, which haslong lag times compared to advective heat transfer Ž . Davies and von Blanckenburg, 1995 . Conductiveheating from the mantle may have assisted SP gran-ite genesis in these orogens but in situ crustal radio-genic heating and decompression melting duringcrustal uplift would have dominated.These observations suggest that a distinction may Ž . be made between 1 ‘high-pressure’ collisions such Ž . as the Alps and Himalayas Fig. 1A , where small-to moderate-volume ‘cool’ post-collisional SP gran-ites formed by decompression melting of overthick- Ž . ened crust  ) 50 km heated by the in situ decay of  Ž . K, U and Th; and 2 ‘high-temperature’ collisions Ž . such as the Hercynides Fig. 1B , in which large-volume ‘hot’ post-collisional SP granites formed bymantle-derived heating of normally thickened crust Ž . F 50 km after lithospheric delamination. The Cale-donian orogeny of Britain may represent an interme-diate case, as it lacks both widespread syn-collisional Fig. 1. Schematic representation of the generation of post-colli- Ž . Ž . sional SP granites in A thick, high-pressure orogens, and Bthin, high-temperature orogens. Not drawn to scale. Intermediatecases would be likely where crustal thicknesses on opposing sidesof the suture differed significantly.
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks