Structures and metamorphism at Brygfjell-Simafjell within the Rodinqsfjallet Nappe Complex, Nordland, Norway

Structures and metamorphism at Brygfjell-Simafjell within the Rodinqsfjallet Nappe Complex, Nordland, Norway

BJ0RGEBRATIU

Brattli,B. 1996: Structuresand metamorph ism at Brygfje ll-Simafjell wit hintheRbdin gsfjallet Napp e Complex, Nordland,Norway.Nor.geol.unders.Bull.431,19-32.

St ruct uresandmetamo rphicfeaturesare describedfrompartsof the areaof1:50,000 map-sheetKorgen.The area invest igated is situatedwit hin theRodinqsfjalletNapp e Complexinthe Caledo nidesof Nordland.The rocksare divi- ded intotwounits,an uppe r unitnamedtheBrygfjellGroupconsist ingofaschist- marble complex,andan underly- ing unit,theBjorn skolten Group, composedmainly ofheterogeneousgneisses.Therockshave been affectedby fourphases (F₁,F2,F3and F4)of folding. Thefirstdeform at ionproducedisoclin alfolds witha penet rative schistosi- ty,51'These st ructuresare deformed bytighttonear isoclinalF2fold s, wit h NW-plunging axesandwit h near-hori- zontalaxial planes.TheF₂folds control thelarge-scale outcro p pattern ofthearea.In approp riaterocks a poorly develop ed schistosity,52'isinplaces definedbythegrowthofwhite mica.The third deformat ionphaseproduced opensynforms and ant iforms of parallel type wit hvert ical axialplanes and WNW-ESEtrendingaxes,while thefourth phase isrepresentedby agent lebucklingoftherocksalong N-S axes.Folds ofthelasttw ophasesdonotcarry any axialplanarcleavageorschistosi ty.

The met amo rphic peak occurred dur ing thefirst phaseoffold ing,andisest imatedtoc.680°Cand7.5-8.5kbar by use ofthegarnet-biotitetherm ometer,and observed reaction-isog rad sindiff erentrocks.Thehig hestgarnet-biotite temperat ures,whichareingood agreementwit h thetemperatu reint ervalforthe observed reaction-isogr ads,are derivedfor garnetcores toget her withinclusions of biot itein the coreareas.Calculationsusing garnet rimstogeth er wit h matrixbiot itegive temperatu res which areabout 100°Clowerthanthemaximum.Itisprop osedthatthis could representthetemperature conditions during theF₂folding.

A secondor retr ogrademineralassociat ionisrepresent ed bythechlorit isat ionofgarnet,bioti teand amphibolein appropriate rocksand a sericit isat io n/saussuritisationofplagiocla seand AI-silicates. It isimpossible torelatethe secondary mineralsto anyspecificstructuralelements.However,textural relationshipsindicatethatthereplace- mentprocesseshavetakenplaceover alo ng periodof time,probablyfrompre-F2to post-F4'

Bjorge Brattli, Departm ent of Geolog y and Mineral ResourcesEngin eering, Nor wegian University ofScience and Techn ology,7034Trondheim,Norway.

Introduction

The purposeof thisinvesti gationisto describeand quan- tify the metamorphism in part of the highest Caledonian nappe unit in Nordland and to propose a possiblerelati- onship between the metamorphic evolution and the structures.The study areais situatedin the Rodinq sfjallet Nappe Complex (RNC)(Kulling 1955),c.10 km north of the Bleikvassli sulphide mine in the mountain s of Brygfjell -Tverrfjell- Sima-fjel l.The RNC, together with the Helgeland Nappe Complex (HNC) (Ram berg 1967, Gustavson1975),formsthe UppermostAllochthon of the north-central Caledonides.To the east the HNc/RNC is underlainby the Seve-Kolinappes(Fig.1).

The Uppermost Allochthon is composedof med ium- to high-grade and low-grade metasup racrustal rocks intruded bylarge massifsof gabbroicto graniticcompo- sition, (Gustavson 1975, 1978, 1982,Gjelle 1978, Terud - bakken & Brattli 1985, Brattli & Terudbakken 1987, Gustavson & Gjelle 1991, 1992).The allochthon forms a large synformal st ruct ure betweeneastern and western antiformal culm inations of Precambrian crystalline rocks (Fig.1).Duringthe CaledonianorogenythePrecambrian rocks were deformed together wit h the rocks of the UppermostAllochthon.Theindivid ual nappe complexes have been divided into various tectonostratigraphical

units and the boundariesbetween these unitsgenerally represent thrustcontacts (Ram berg 1967, Stephens et al.

1985, Brattli & Terudbakken 1987).The age and prove- nance of the Uppermost Allochthon is poorly known,but it is regarded as a majortectonic unit which hasbeen transporteda great distance.Riis&Ramberg(1979), Cribb (1981) and Brattli et al.(1982) have presented isotopic dating evidence that certaingneissunits within the RNC are of Precambrian age. The lower age limits of these rocks remain somewhat controversial.Skauli et al.(1992) have reported a metamorphicage of464±22 Ma for the peakmetamorphismat Bleikvassli.Age determinat ionsof intr usive rocks in the RNC are sparse. Rb-Sr whole-rock age determination studieson foliation-discordant,grani- tic dykes near Umbukta inthe RNC(Claesson 1979)indi- cate an age of 447±7 Ma. Farther north in the Beiarn Nappe Complex (BNC)(Rut land & Nicholson 1965),which is consideredto be asub-unit of the RNC,an age of 440

±30 Ma hasbeen reported on a post-tectonic (post the mainfold episodes, F1 and F2)granite (Ter udbakken&

Brattli 1985). Rb-Sr whole-rock age determinationson some gneisses (Harefj ell gneisses) within the BNC have given agesof 415±26 and 470± 59 Ma(Cribb 1981).These rockswere previously interpreted by Rutland (1959) as metamo rphosed supracrustals,but are now believed to be plutonic rocks (A.Solli pers.comm.1996) which intru-

20 BjetqeBrarrli NGU-BULL431.1996

/]~

Uppermost Allochthon

Upper Allochthon

o

I

N

Fig.1.Prin cipal recronosr ra rigraphicunit s of partsof Nordland,north-centralNorwegianCaledonides,based upon maps publishedbyrheGeological Survey ofNorway (Norges geologiske undersokelse).The shadedrectan glemar kstheareacoveredbythis investigatio n.

HNC

Fig.2.Schematic tectonostratiqrapb icand iithostra tiqrap ni csuccession far the investi- gated area.Aobrevotions,HNC=Helgeland NappeComplex,RNC=ROdingsfjiilletNappe Comptex,S-K-N= Seve-K61iNappes Gamet-mica gneiss

Garnet-kyanite- muscovitegneiss Feldspar-mica gneiss Mica gneiss Ouartziticgneiss Graniticgneiss Garnet-micaschist

Quartz-feldspargneiss Amphibolite

Marble

Marble

Undifferentiated } ...

gneissesandschists Micagneisswith calcite

.

I I

I,

The^{BrygfjellGroup}

!

The Bje rn- skoiten Group RNC

S-K-N Uppermost

Allochthon

NGU-BULL431,1996 Bj(')rgeBtauli 21

D •

• ^§

D D

Quartz-feldspargneiss

Amphibolite Marble

Undifferentiatedgneissesandschists Micagneisswithcalcite

,./

20

o

Foliation(5 1)withangleof dip indicated

FoldaxisF2withplunge indicated

2km

N

t

Fig.3.Sim p li fiedgeologicalmap ofthe Brygfj ell- Simafjellareawithin the RodingfjalletNappeComplex,(see alsoGustavso netal.1990).Insetmap inthetop left corner showsthedivision into groups.The symbolsmarkthe localities of thesamplestakenforthe electron microprobe investigation(....=^sample23.8,

....=^sample13.8,'=esample 3.8).

- 22 Bj0rgeBratt li

ded between the F₁ and F₂ fold episodes.These ages, together with those from the RNC provide an upper limit for the main deformation and metamorphism in the Uppermost Alloc hthonofthe Nordland Caledo nide s.

Theinvest igated area covers a

c.

southern most part of the 1:50,000 map- sheet Korgen (1927 11) and the northernmost part of the map-sheet Resvatnet(1926 I).The rocks have been subdivided into two unit s.The upper unit is named the Brygfjell Group and consists of a schist-m arble association, whil e the underlying unit,theBj o m skolte nGroup,consists mainly of heterogeneousgneisses(Figs. 2 and 3).The Brygfje ll Group can be correlated with the Anders Larsa Group in the Bleikvassli area (Ram berg 1967) while the Bjornskolte n Group is correlated wit h the Kongsfje llet Group in the same area.A simplified geological map is presented in Fig.3.A prelim inary version of the bedrock geology of map-sheet Korgen was published in 1990 (Gustavson et al. 1990).

In order to study the metamorphic conditions in the rocks,both index minerals(isog rads)and mineral reacti- ons(reaction-isograds) havebeenexam ined.A quantit a- tiveevaluat ion of the temperatureand pressureconditi- ons under the main metamorphism has been made by using the calibration for the partition ingofiron and mag- nesium between biot ite and garnet in pelit ic rocks, in combinat ion wit h react ion-isogr ads for vario us minera l reactions.

GU-BULL431. 1996

Fig.4.Rootless intrafoli alF1fold in micaschist.The fig ureillustrates how lim bs becomethinn edandtransposedintothe planeoffoliation.

IIS_l

I

I \-

r ,

\

\

a b

'---'lmm

Deformation

Studying the metamorphic histo ry in anareareq ui resa thorough integrationwit h struct uralfieldworkaswell as studies of microstructures in oriented samples.In the fol- lowing section the different Caledonian st ruct ures are described,asthey are observed inthe field and under the microscope.

F olds

The rocks of the area are multipl ydeformed.Structures ascribedto four phases of fold ing(F₁, F₂,F₃,andF4) have been found inbothgroups.

F

folds

Flattening seems to have been veryinten se during this deformationevent,affectingthe initi allydeveloped folds to ult imately produce isoclinal folds and aprono unced axial plane foli ati on,51'Thus,in many casesit isdiffic ult to recognise F₁folds.However,mappi ng of layers which act asmarkers hasshow n that it ispossibleto follow Iit- hologicallaye rsaroundF₁fold closures.AtSlegda,a large F₁fold hasbeen recognised by mapping the amp hiboli- teswhich const it utethinbandswithin the marble(Fig.3).

In many outcrops the layers seem to be parallel,but care-

Fig.5.Microscopicfoldsobservedinthin-sect ion s. (a) Parallel-orient ed mica outlinesapossibleF1foldclosure.Note tha tall mineralsareparalleltothe 51foliation.(b) F}foldsin micaschist.Growt hof whitemicadefinesapoor' Iydevelop edschistosity,51'in rocksofappropriatecomposi tion.

ful mapping hasshown that they commo nly represent thelimbs of large-scaleisocli nalF₁folds.Despite a pro- no un ced transposition ofth e hin g eareas,defined by a penetr ativ e preferred mineralorientation parallelto the axial plane of the F₁ fold,the layers can still be traced around theclosures.

Mappableminor F₁foldsare seldom seen inoutc rop.

How ever, in rocks of approp riate composition, small bandsandlenses(5-10 cmlong)of quartzcan berecogni- sed as intrafolial and root-less int rafo lialfolds,indicating isoclin al folding. The limbs have become thinned and modifie d/transposedintothe plane offoliation,ultimate- lycausing the limbsto disappear(Fig. 4).

Clearly defined microscopic F₁ folds have not been observed,despite studiesof thin-sect ionsfrom differ ent parts of large-scaleF₁struct ures.How ever,Fig.5a illustra- teshowparallel-orientedmica mayoutline F₁folds.

F

folds

The second foldepisodeisrepresente d bylarge,recum -

NGU-BULL431,1996

a)

b)

c)

Fig.6.Differentstyles ofF₂foldsdevelop ed asaconseque nceof the rock- typ es that arefoldedand thedifference inrheologybetweenthelayers.(a) Class1Cfolds. (b)Para llelfolds. (c) Sim ilarfolds.The visua l classification is based on Ramsay (1967).

Bj0fg eBtattli 23

bent,tightto isoclinal folds.The large-scale F₂ foldscan easily be traced throughout the area,especially by follo- wing the marbles, but also by mappingthe amphibo lites and quartz-feldsparlayers.The style of the F₂structuresis apparently quite variable (Fig.3).However,this is mainly a consequence of the horizonta lorientation of the axial planes. In areas where the axial planes cut the surface (t opog raphy) at right angles,as in the northern hillside of Gronnfjellet,the folds appear as close to tightstructures.

Otherwise, where the topography is smooth and the axial plane subparallel to thesurface,the folds appear as close to open structures.The S,foliation is folded around the F₂ closures.Statistical distribution of the S, poles indicates that the mean axial direction of the F₂ folds plunges moderately towards WNW (Fig.7a).This is in fairly good agreement with the most prevalent lineation orientation in the area.

Minor F₂folds exposed in outcrop are very common.

The fold style varies depending on the relationship to lar- ger folds,the intensity of the deformation and the cha- racter of the alternating lit hologies that are folded. F₂ minorfolds are mainly tight to isocli nal. Some are clearly parasit icto larger foldswith a long and a short limb.In rocks with alternating layersof mica schist and quartz- feldspar gneiss of equal thickness,the fold styleis that of class 1C(Ramsay 1967) (Fig. 6a).In other areas where competent layers dominate,the folds appear as parallel folds(Fig. 6b).In caseswit h little or no competence diffe- rence between the layers,the folds are of a similartype (Fig. 6c).Measured fold axes plunge moderately(c.20⁰⁾ towards WNW and the axialplanes dip gently(5-10°)bet- ween NW and WNW.

Microscopic F₂ folds have been observed in several thin-sections.Usually the minerals are folded around the F₂closures.In some cases a second generation of white mica has been observed parallel to the axial plane of the F₂folds (Fig.5b).Quartz and epidote have also recrystalli- sed in the hinge area of the folds.

F

folds

The F₃folds are easilyseparated from the earlier structu- res on the basisof their style and the orientation of their axial planes. Usually they appear as parallel folds and form large, open synformal and antiformal structures.

Plots of measured F₂ axial planes give a1t3 axis which plunges moderately (c. 20⁰⁾ towards WNW (Fig. 7b).

Because of interference between the synforms and anti- forms and the topography (E-W oriented valleys), the foldsmay appear largerthan theyin fact arein reality .

MinorF₃folds in outcrop have been observed in only a few places.The folds are open(120°),of parallel type and have symmetrical limbs (10-15 m long) and vertical axial planes.

F₄folds

A gentle buckling of the rocks along N-S axes has produ- ced open folds which are characterised by a small ampli- tude to wavelength ratio.The folds are only recognised in

24 ^{Bj0rgeBrartli} GU-BULL 431, 1996

N

areaswith asmooth topography andwhere therocksare verywellexposed.

a

N

Contourspr Iperc entarea

D 0.J-0.6%

D ^0.6-0.9%

D 0.9-1.8%

~1,8-2,7%

D ~.7-4,2%

~ 4,2-6,3%

IilJ ^6.3-7.8%

- 7.8 -10.50/,

. 10.5-15%

. .>15%

Plana r structures Se/SS

In allunit s most of the rocksexhibitseveraltypesof pla- nar structures.Genuine primarystructures probably do not exist as therockshavebeen modified and transfor- med by metamorphic and deformatio n processes.

However,the Iitholog icalbanding represented by alter- nating rock-types mustbe recog nised as a primaryfeatu- re,andishereabbreviated asSo'In some rocks a planar structu re defi ned by a layering isvisible in hand speci- men.In mica-rich rocks thiscan be describedas a gneissic layering of alternatin g light and darkminerals,whi lein the marbles itisrep resente d by an alternating grain-size distribution.These st ruct uresseem to beparallelto the Iithologicallayering So'andare probablymore a resultof secondary transposition and metamorphic processes thana primary feature. In orderto separate thesefrom otherplanar stru ctures,they have beendesignatedSS.

5/

The most prevalent and penetrative planar structure foundthroughoutall unit sisthe 51foliation.In mica-rich rocks itisdefined by adomainal stru ctureof lens-likeand disc-like quartz and feldspar grains surrounded by a parallelarrangem entof mica. In maficrocks(am phibolite, cale-silicate rocks, et c.) it is defi ned by a parallel align - ment of lamellarandprismat ic chainsilicates(am phibo- les/pyroxene).Thefoliation is consider ed to beparallelor sub-parallelto the axialsurface ofthe F₁foldsandhence to the50155structures,except in the hing e areasof the F]

folds where the foliat ion int ersects theIit hologi cal laye- ring, So' In out crop, how ever, all stru ctures arealigned also in the foldclosure area,because of theintense defor- mation and transposit io nduring the F₁folding episode (seeabove).

Altogether 329S1poleshavebeen plott edand contou- red(Fig.7a).The contours do not defineanunambig uous great circle and TC2 axis.Actually it ispossible to draw many great circlesthrough the contoured area.In doing so,the majori ty of theTCaxes fall within a restricted area (stippled)in the fourt hquadrant. Thepoo rly definedloca- tion fortheTC axisispresumab lydueto the factthat F₂ and F₃axesare subparallel(seeabove)whichhasresulted in a broad spreadof the 51poles. Thus,the sti ppledarea representsboththe F₂andthe F₃axes.

Inorder to makeamore precisedecisiononthelocati- on and orientation of the statisticalF₃axis,43F₂axial pla- nes havebeenplotted (Fig.7b).The const ructedTC3axis plungesat 220towards290⁰, whichis within thestipp led area for theTCaxesobtainedfromthe51poles(Fig.7a).

52

Anot herschistosity,52'is observedinsomesamples.The

Contourspr Ipercentarea

D 2-5%

fD 5-7%

7-11%

. 11-13%

. 13 -16%

+ N

, /

c b

Fig.7.Statistica l representati onof 5and L measurements.Symbolsused: - -=construct edn circle. (x=const ructedfold axis.Stipp led area represents F}andF_Jaxes.0

=

NGU-BULL 431 ,1996

structure is poorly developed and rare in outcrop,but can be recognised in thin -section .Usually it is defined by the growth of white mica parallel to the axial planes of F2 folds(Fig Sb).In appropriaterocks it may also be defined asan incipient crenulation cleavage in connection with the F2episode.

Lineations

In the studied area several types of linearstructures have been observed. Only the most prevalent will be described here.

A fairly common (alt ho ug h not penetrative) linear structure is a mineral shape orientation.It is defined by a parallel alignment of grains with an elongate habit, like amphibole,and it is especially common in rocks contai- ning such minerals.In most cases the lineation seems to be parallel to the F2fold axes.In the granitic gneiss at 5imafjell another kind of mineral lineation is developed as a result of extremeelongation of feldspar. In some pla- ces the long axesof the elongated minerals(augen) have been measured to over one metre.This lineation is also oriented mainly parallelto the F2axial trend.

In rocks where an 52schistosity (see above)is develo- ped, a lineation is also defined by the intersection line between 52and 51'

In Fig.7c,the measured linear structures (63 F2axis lineationsand 68 unrelated linear structures)have been plotted. The stereographic presentation shows that the trend and plunge of the linearstructures vary to a certain degree, but they mostly fall within the fourth quadrant of the stereogram. The spread is probably due to the fact that the Iineations have been affected and'di st u rbed ' by later fold structures.

Petrography

In metamorphic rocks, characteristic minerals and mine- ral assemblages commonly provide a means of assessing the intensity and nature of the metamorphism and the relationship between the metamorphic history and the structures. The following section is concerned only with rocks and mineral assemblages which turned out to be of special interest for the evaluation of the metamorphic condit ions.Other rock-types will not be describedin this context. The nomenclature is based on quantitative mineralogicalcomposit ion,and is after Winkler(1979).

Pelitic rocks(m etap elites)

The metapelitic rocks are dominatedby mica schi st s and micagneisses.Basedon field st udi esandmodal ana lyses (fi ft een samples), itis possible to separate between seve- ral types of schists/g neisseswhere the mostim po rt ant in this connection are;garnet-micaschist(m o st commonin

Bj0rge Bratt/i 25

the Brygfjell Group, see Figs.2and 3),garnet-micagneiss and garnet-kyanite-muscovite gneiss (rest rict ed to the Bjernskolten Group,see Figs.2 and 3).All samplescon- tain quartz,plagioclase, biotite,white mica,epidote/cli- nozoisiteand garnet.Kyanite may appear in the garnet- mica schist, but is more common in the garnet-mica gneisswhere it makes up to 5%of the mode(in certain narrow zones in the garnet-kyanite-muscovite gneissit makes up as much as30%).5taurolite is present in the gneisses as inclusio nsin garnet together withquartz,bio- tite and white mica,but has never been observed in the matrix.Calcite has been found in only two samples of the garnet-mica schist. Chlorite is rather rare,but is recogni- sed in a few samples as an alteration product after garnet and biotite .The mineral occurs either along grain boun- dariesor cracks (garn et). In biotite it may be present as 'sand w ich' layers coherently disposed between relict lay- ers of biotite. The mineral assemblages in the different rocks are summarised below.

Garnet-mica schist:Oz+ PIAn36- 40(core)+ Bt+Ms+ Ep/Czo + Gt±Ky±Cal±Chi

Garnet-mica gneiss:Oz+ PIAn36(core)+ Bt + Ms + Ep/Czo + Gt + Ky±5t±Hbl ±Chi

Garnet-kyanite -muscovitegneiss:Oz+ PI

+

±Bt±5t±Chi

Amphibolite

Plag ioclaseand hornblende make up the bulk mineralo- gy of the amphibolites. All other minerals are modally subordinate.The mineral assemblage,based on studies of five thin-sections, is listed below. Epidote may appear in small amounts« 1%),but mostly it is lacking.

Amphibolite: Hbl + PIAn 32-36(core)+ Bt + Oz±Gt (±Ep)± Chi

Siliceous carbonate rocks

Impure (siliceous) Iimestones/dolomitesexist as a transi- tion zone along the border between the marbles and the metapelites.Rocks consisting of quartz,dolomite and eit- her calcite or magnesite remain unaffected at very low- grade metamorphism. At higher temperatures some minerals start to react and at high grade numerousreacti- ons occur between the minerals.Based on thin-section studies (8 samples), it is possible to separate between three types of mineralassociations(react io ns).

Dolomitic limestone, (ty pe A):Tic + Cal+Tr+Dol Siliceous dolomitic marble, (ty pe B):Tr + Cal+ Dol+Oz Siliceous dolomiticmarble,(t ype C):Tr + Di+Cal+ Dol+ Oz.

26 BjorgeBtattli

Re lat ion between mi neral g rowth and structures

Itis possib leto distingu ishbetwee n twostages of mine- ral growth.The mainmineral assemblagedefinesthe 51 foliat ion. In the metapelit es 51isoutlined by minerals such as kyanite, biot ite, garnet, muscovite, plagioclase andquartz, whil ein the amph ibo lite s the foliat ion is defi- ned by hornblende, plagioclase, biot ite and in som e cases garnet.In the siliceouscarbonate rockthe text ural features are not sowell developed as in the metapelites.

However, the mineral associat ions (types A, B and C) listed above are considered to be syntectoni c,Le. they have had their principal growt h period in connection with themainfold episode,probably duringthe formati- on ofthe penet rat iveplanarfoliation.

The main mineral assemb lages are usually folded around microscopic F2folds. However, some minerals mayshowa zonation where theoute rmost rims cut the foliation. Onesuch mineral is garnet. Inmetapelitesand amphibolites the garnets have inclusion -rich cores (see above), in some cases wit h snow ballstructur e(5_{i) ,} follo- wedby massiveidi oblasti crims. The 5_iseemsto be paral- lel with the 51(or 5_e^), indicating that the main grow th must have beensyntectoni cwith respectto F_1•The rims usually cut the mainfoliat ion 51'and must therefore be post-tecto nic to F₁. Inappropriate rocksa lat erphase of white mica is partly wrapped around and has partly grow ninto therims ofthe garnet .This mica(w hich must beseparated from the randomly distributed muscovite, see below) probably belongs in time to the muscovite which appearsto grow parallel to the52schistosity,indi- catingthat therims arepartly pre-tectoni c toF2and part- ly post-tecton icto F2.

Other minerals which are zoned are plagioclase and epidote. Plagioclase typ icallyexists withagradualzonat i- on ranging from andesine (An36-4o)in the coresto oligo- ciase(An26-30)inthe rims.Theplagiocl aseis usually seri.ci- tised/ saussurit ised,especially inthe cores. Togetherwith quartz,the mineral forms lensesand bands(part icularly in the gneisses)parallel to the 51'Itis reasonable to assu- me that the plagioclasecore is in equilibr ium wit h the mainmineral assemblage defining the 51foliat ion, while the rim isin equilibriumwith the later formed retrograde mineralassemblage(see below).Epidote appears asidi o- morphic grains, wit h interference colour ranging from third order(cores)to low second order and even middle first order (rims).The opposite zonation has also been observed.Epidotemayalsooccurasneedles inconnect i- onwith the saussuritisation of the plagioclase(t his phase should beseparatedin timefrom the idi omorphic grains in the matrix).The epidote israndomly distributed relati- ve tothe mainfoliation,51'but atleastsome of the grains seem to be folded around microscopic F2 folds. The growth must therefo re have takenplaceoveralong peri- od of time,probably from post-F.to post-F.and maybe eveninto the F3phase(see below).

Except for a local redistributionof quartz locatedin the

GU-BULL 431.1996

hinge areas ofthe F2folds andthe presence of whitemica parallelto theaxial plane of the F2folds,there is no speci- fic mineralgrowth which canclearlybe asccribed to the F2deformat ion. Themineralsformedduring the F_l defor- mation were,with small modification (zoned minerals), also stableunder the F2deforma tion.The zonation indi - cates that the growth has probably taken place under varyin g PT conditions with in the regional metamorphic belt. However,as most of themineralsbelong ing to the assemblage actually define the main foliation,51' their princip al growth period must have been during the F₁ deformation.

The second or retrogrademineralassociation is repre- sentedbychloritisation of garnet,biotiteand amphibole inappropriate rocks and asericitisation/saussuritisation ofplagioclaseandAI-silicates.Randomlydistributedmus- covite andepidote also probablybelongto the retrogra- deassemblage.

Thereplacement processeshave probably taken place overa long period of timeinvolving manyreactionsand variousstagesof alteration.It is impossi ble to relate the secondary or retrogrademineral assemblagesspecifically to anyone of the structuralelement s.How ever,asthe chlorit isation has affectedthe massiveidi oblasti crimsof the garnet,theprocessmust havetaken placeafterthe formationof the rims,and therefore after theF2phase.

Metamorphic condit ions

Theinvestigated metapelit espresent severalinteresting mineralrelationshipsthat need to bediscussed.Kyanite is the stable alum inosilicateinthe rocks.It is foundas a matrix mineral,usually parallel orsub-parallelto the 51 foliat ion. The occurrenceof the mineralindicatesthatthe pressure andtemperat ure are aboveoralongthe ky-line (Fig.9)Themost important mineralassociat ionisfound in the garnet- kyanite-muscovite gneiss, represented by the occurrence of staurolite,quartz,kyanite and garnet.

Accordingto Winkler(1979) staurolitestarts to decompo- sein quartz-bearing rocksatabout670°C.The'coexisten- ce'of the minerals hasthereforebeen extensively studied in thin-section.It isimportant to note that staurolite is found only as inclusionsin garnet together wit h quartz and never as a matrix mineral. Kyanite,howeve r,is the dominant alumino-silicatein thematrixand isalsopre- sent as verysmall inclusions,together with stauroliteand quartz, in the garnet. These observat ions lead to the conclusion thattheremustbe atleast two local medium- grade sub-assemblagesin the rocks; onesealed in the garnet and probably represented by the equilibr ium reaction(Bucher&Frey1994,p.202):

(1) staurolite + quartz

=

and onein the matrixrepresented bythecoexistenceof garnetand kyanite.

NGU-BULL 431,1996 Bj01geBtottl! 27

+

XC0 2 ^~

Fig.8.Schematic TX_C02diagram showing import antequili bria and invariant points in the siliceouscarbonatesystem(aft erEggert

& Kerrick1981).Abbreviations:Col

=

Oi= diopside,Dol=dolomite,Fo =iorsteri- re,Qz=quartz,Tic =talk,Tr = tremolite.

(4) 3 Dol+4 Qz+H₂₀=Tic+3 Cal+3CO₂

800 700

600 500

400

2-t-~.---.----,r---,.----r---r---r--.---.----l

300 4 12

10 'I:'CIi oD~ ' - '

<l) 8

...

=

enen

<l)

...

0..

6

explainwhy quartzis apparentlylacking andwhy tremo- liteis present witharim of dolomite.

In order to facilitate the equilibrium reacti ons, the rocks must containtalc. The mineralisstableover a wide range of physicalconditions(P,T, XC0₂)and couldbe for- med as aresult of an early reaction betwee n dolomite and quartz(Winkler 1979);see alsoFig. 8:

Theequation(1)defines a linein a PTdiagram (Fig. 9).

In the quartz-saturated matrix all staurolite is replaced by garnet and kyanite(see eq.1).The two-mineralassembla- ge (Grt (aim)+ Ky)is diagnostic for the initi al stages of upper amphibolitefacies.

In the amphibo litesthe main mine ralassemblagecon- sists of andesine,green hornblende,biotite±garnet.This is a typical amphibolite-facies association,stable at a tem- perature around 600°C. According to Bucher & Frey (1994) c1inopyroxene appears at higher temperatures, with the diopside-hedenbergite seriesas the dominant cpx at around 650°C.In this case no such minerals have been observed.

The mineral association in the siliceous carbonate rocks also exhibits some very interesting relationships.

Mostof the minerals have taken part in one or more reac- tions. Basedon thin-section studies,the following trans- format ions between the mineralsare observed:

In type A:Reaction between tremolite,dolomite, calcite and talc

In type B: Dolomite^1&tremolite In type C:Tremolite~diopside Comments on reactionA

The most usual association is tremolite, dolomite, talc and small amounts of calcite. Quartz is not observed in connection with this association.In caseswhere calcite is in contact with tremolite,the latter appears with a rim of newly formed dolomite.A possible formation sequence could be after the following equations(Winkler1979);see also Fig.8:

(2) 5 Tic+6 Cal+4 Qz= 3 Tr+6 CO₂+2H₂₀ (3) 2 Tic+3 Cal

=

At the time when all quartz has been consumed,talc,cal- cite and tremolite are the stable minerals(eq.2). At slight- ly higher temperatures,talc and calcite will react to form more tremolite and dolomite (eq.3). The reactions may

Temperature(0C)

Fig.9.PT-diagramshowingsomeimport ant equilibriumreactions.Shaded field:coexistenceof staurolite+quartz.(1)75St+312Qz =100 Aim+575 Ky+150Hp(2)75 St+3 12 Qz = 100Alm+575 Si/ +150Hp(3) 8Cid+10 Ky=2St+3Qz+4Hp .(4) 8Cid+10 And=2St+3Qz+4Hp . (5) Constructed curve based on the position ofthe invariantassem blage(B) (siliceo us limestone)at2,4,6,8kbar(see also Eggert&Kerrick1981).The positions of reactions(1),(2),(3),(4)areafterBucher &Frey (1994).

28 Bjl2JfgeBrartli

The conclusion must be that in dolomi t e limeston e of type A, there existsa localequilib rium assemb lage repre- sented by equation(3) and observedin thin- sectionasan associat ion of Tic+Cal+Tr+Dol.Lack of quartz and an 'unfavourable' mole fraction of XC0₂ are probably the cont rolling fact orspreven ti ng theformati on of diopside (Fig.8).

Com ment son reaction B

The princ ip almine ral associati on consistsof quartz,tre- molite, calcit e and dolomite. Studies of thin-sections show that tremo lite is growing from dolomi te as this decom po ses.As calcit ealso seemsto for minconnection wit h the repla ceme nt ofdolomi t e,thefoll owi ngreacti on isproposed(Winkler1979); seealso Fig. 8:

(5) 5Dol+8 Qz+H₂₀=Tr+3Cal+7 CO₂

The equilibrium equ at ionrepr esents a prograde evoluti- on which will continuetoproduce tremolite as the tem- perature increases.Thisis a typicalmiddle amphibo lite- facies associat ions which repr esentstemperatures and pressuresofc.600°Cand 5-6 kbar(Bucher &Frey 1994).

Commentson reactionC

In thistype of rock,diopside makesup as much as 60 mod al% and consti t utes the main mineral. Depending on the exist ing reactan tsin the rockand the physical con- ditions(T, X( 0₂₎,the formationofdiopsidemaytakepla- ceafter the follo wing reaction s(Winkler 1979);see also Fig.8.

(6) Dol+2 Qz

=

(7) Tr+3Cal=Dol+4 Di+CO₂+H₂0 (8) Tr+3Cal+2Qz

=

(9) From thereacti onin theinvariant pointB:Qz+Tr+Di +Dol+Cal

St ud ies ofthin-sectionshave shown thatallthe minera ls in reaction(9) are present,whichind icat esthat the rock has reached theinvariant pointB,probablyalong reacti- on(5)in Fig. 8. Format ionof diopside accordin g toreacti- ons(6),(7) or (8)is not possib leas long asthe invariant mineral assemblage is present. As anexam p le,diopside formsfrom reaction (8)only if reaction(5)(seeabove)has consum ed all quartzat so melow er temperatu res, which isnottrueinthiscase.

In astudy of dolomitic limeston ecarried out by Eggert

& Kerrick (1981),the temperatu re int erval forthe eq uili-

brium react ionsand the po sition ofthe invar iant points have been established asa funct io n of POuid and XC0₂.

Accor ding to the expe rimentalcalib rat ion s, it ispossib le to constructcurvesfor minimum and maximum conditi- ons ofXC0₂•Ofall the possibleequilibriumreactions abo- ve, reaction (9) refl ects the high est temperature and is thereforethe most int erestin g in this connection.Based on the position of the invariant point B at 2,4,6 and 8 kbar(see Fig .8 in Eggert&Kerrick1981),a curvehasbeen

GU-BULL431.1996

constructed (line 5in Fig .9).Becauseofsom euncert ainti - es in thelocationsof the inv ariant points,the curvedefi - nesa broad line in the pT-diagram.However,using the infor mation from the limestone in combinat ion wit h the informa tion from the garnet-kyanite -mica gneiss, it is possibleto concludethat the temperature and pressu re must have beenequ al to or greater than c.675°C and 7.7-7.9kbar(the PTestimat eisbasedonthe int ersecti on point between line 5 andline1inFig. 9).

Geothermoba rometry

In order to make the determination ofthe maximum PT condi tionsasprecise as possible,a geothe rmobarometric study has been carried out on biot it e and garnet and combined with theresults from thepetrographicinvesti- gation.

Using thermobarometrytodeterm in ethe tempera ture and pressureat which the rock was metamorphosed is basedon the assump tionthat theminerals are form ed at equilibrium. Howe ver , it isquestionable whet her or not the minerals in a rock ever achieved equilibrium.

According to Winkler(1979),itis possible to finddifferent generati ons ofsub-assemblages in a rock, hopefu lly each in a state of local equilibr ium. The succeedi ng work is basedon these assumptio ns.

Analytical me thods and pro cedures

The most wid ely used geoth ermometers for met amor- phicrocks involve theexchange of Fe²+andMginthefol- lowing minerals;olivine,garn et,clino pyroxene,orthopy- roxene, biotite, pheng ite,chlorite and hornblende.The compositio n of the rocks(or the presence of minera ls) will therefore place restri ctionson the typeof thermome- terthat can be used.In thisparticular case, garnet and biotite are common minerals in the metapelites.When the mine ralsare in mutualcontact,it is anticipatedthat they are in equil ib rium and can be used asa thermome- ter. The garnet-biotite exchange thermomet er existsin manyversions and itisusedin awide varietyofrocks and over a broad range ofmetamorphic condit io ns.Howeve r, because of insufficient knowledge of how im purities affect the Fe²+-Mg exchange between theminerals(see, for inst ance,Bhattacharyaet al.1992 andSpear 1993)itis difficult to choose betweenthe manyversions. For that reason the Ferry &Spear(1978) originalthermomet er is used in this work.

T(0C)=(2089+9.56P(kbar))/ (0.782 -InKo)-273

Theauthors sug gest ed that thethermomete r shoul d be restrictedtousagefor garnet s lowin Ca and Mn,with(Ca +Mn)/ (Ca+Mn +Fe+Mg)~0.2andwit h biotit eslowin Al^v;andTi, wit h(AI^V;+Mn)l (AI^Vi+Ti+Fe+Mg)_~0.15.

As the Kovalueswill vary as a function ofinhomoge-

NGU-BULL431, 1996 Bj0fge Brarr/i 29

neit ies and zonation effects in the minerals; it is impor- ^{Tabl e1.}Examples of electr on probe microanalysesof garnet.

tant to have a strategyfor selecting the'rig ht'Kovalues.

Sam ple 1.3,13.8 1.4,1 3.8 3.6,23.8 1.5,3.8 3.2,23.8 2.6,3.8 One way is to choo se the analyseswhich provide the _{rim rim rim core core core} lowest Kovalues(the Ko-min .method').This meth odwill

give the highest temperat ure and hence informati on ^Si02 38.40 38.57 40.05 35.87 40.12 36.62 about the metamorph ic peak.In orderto bereasonably ^{Ti02 0.01 0.28 0.07}

AI20 3 21.77 21.73 22.38 21.23 22.22 22.56 surethat theselect ed Kovalue really represent sthehig- _{Fe203 31.45 31.28 29.33} 29.42 28.86 29.22 hesttem perat ure, thechoice must be based on a number ^{MnO 0.38 0.53 1.27 1.01 1.90 1.90} of analyses. An interesting aspect of the'Ko-min.meth od' ^{MgO 4.92 4.77 5.35 5.35 5.39 5.21} is that if the same philosophyis used for the highest Ko ^{CaO 5.84 5.81 6.19 5.42 6.19 6.16}

Na20 0.05 0.03 0.08

value('Ko-max. met hod'),thisshould provideinformati- _K20 on about the lowest equilibriumtemperature for the coe-

xisti ng minerals. ^{Total 102.83 102.71 104.96 98.34 104.74 101.82}

The metapelitesin the invest igat ed area contain poiki- _{Formu la(based on12oxygens)}

loblastic garnet wit h inclusions of mica, quartz and in _Si 2.952 2.968 3.001 2.869 3.009 3.250 somecases staurolite.The micaismainly muscovite,but ^{AIW 0.048 0.032 0.131}

biotite may also occur.The garnetis usually represented ^{Ti 0.001 0.013 0.005}

AI" 1.919 1.934 1.971 1.865 1.963 1.170 byanouteridioblastic rim which,incontrasttothe core, _{Fe" 0.128 0.166 0.014 0.266 0.030 0.332} is mostly free of inclusions.The rim seems partlyto cut Fe" 1.885 1.84 2 1.822 1.696 1.776 1.818 the main bioti te schistosit y. In order to infer a possible ^{Mn 0.023 0.032 0.08 1 0.067 0.117 0.140} metamorph icevolution overtime,the poikiloblasticgar- ^{Mg 0.566 0.544 0.594 0.638 0.603 0.688}

Ca 0.479 0.480 0.495 0.465 0.495 0.586

nets(w iththe rims)were chosen for the analysis.It should _{Na 0.007 0.005 0.005}

also be stressedthat the minerals (garnet and biotite) K

should lookasfresh aspossible (no sign of retrograde

Total 8.000 8.000 8.000 8.000 8.000

chlorite associated wit h the grains). It was decided to ^8.000

Ca+Mn

concent rate the microprobe analyses on two types of _Ca+Mn+Fe+Mg

mineral relationship: ^{0.19 0.16 0.19 0.20 0.20 0.20}

1.Core-areasof garnet togetherwith inclusions ofbiotite _Table2.Examples ofelectro nprobemicroanalyses ofbiot ite.

located within that area(t he bioti te inclusionsare very

small). Sample 1.10,13.8 1.2,13.8 3.4,23.8 1.1, 3.8 3.3,23.8 1.9,3.8

mat rix mat rix matrix incl usion inclusion inclusion

2.Outermost inclusion-free rim of the garnet together _Si0

2 39.48 37.16 38.67 38.69 38.75 37.76

with biotitein the matrix in contactwit h the rim. Ti02 1.62 1.80 1.53 1.57 1.96 1.42 AI20 3 18.17 18.36 18.40 17.73 18.65 17.96 After thin-sectionstudies(30samples from the metapeli- ^{Fe203 15.50 16.83 16.42 16.10 16.08 15.87} tes in the area),three different rock sam ples,two from ^MnO_MgO ^0.02_13.04 ^0.03_{13.30 13.29 13.37 13.36 13.21} thegarnet-m ica schist/gn eiss and one from the garnet- CaO

kyanite-muscovite gneiss,were chosenfor theinvest iga- ^{Na20 0.25 0.31 0.20 0.23 0.42 0.39} tion(Fig 3).Theselected samplesall displayedthe'desi- ^{K20 8.30 9.00 9.28 9.12 9.44 8.42} red' biotite-garnet association described above.In each _{Tot al 96.40 96.87 67.82 96.87 98.70 95.07} sampleseveralgarnet-biotiteassociationswere analysed.

Altogether, 12 pairsof garnet(core composition)-inclusi-

Formula (basedon22oxygens)

on biotite and 16 coexisting pairs of garnet(rim composi-

Si 5.744 5.470 5.613 5.657 5.569 5.611

tion)-matrix bioti tewereanalysedaccording to the follo- _Alw 2.256 2.530 2.387 2.343 2.431 2.389 wingprocedure.In analysingthe biotite in the matrixthe Ti 0.175 0.194 0.165 0.175 0.207 0.159 electron beam was directed towards the border zone of ^{AI" 0.851 0.650 0.746 0.709 0.729 0.716} the grains whichwerein contact with the rim of the gar- ^{Fe 1.877 2.060 1.984 1.964 1.934 1.962}

Mn 0.003 0.003

net.In calculating the Kovalues,alliron in the biot ite is _{Mg 2.819} 2.916 2.871 2.903 2.857 2.975 initially regarded to bein the ferrous state.Examplesof Ca

electr onmicroprobeanalyses are givenin Tables1 and2. ^{Na 0.070 0.088 0.052 0.070 0.104 0.107}

K 1.536 1.696 1.723 1.702 1.726 1.605

Results and discussions

Alvl+Mn

It seems to be possible to differentiate between two AIVI+Mn+Fe+Mg

0.17 0.15 0.13 0.15 0.16 0.15

'populatio ns' or groups of Kovalues;one with 'low' Ko valuesrepresented bythe garnet-corearea

+

30 Bj0rgeBrauli NGU-BULL431,1996

700 800 600

400 500

2+-....L...----,-- -,-- ...L....,c-L-- rL.--,L---'r---,--i 300

12

betwe en85 and 95%(of total iro n),with a meanof 90%

(see,for instance; Deer et al. 1962, 1992).According to thisresult,the iron conten t in the bioti teshas beenrecal- culated to a fixedferric-ferrousratioof 1:9(i.e.Fe³+/Fe²+= 1:9). Indoing so,the Kocurvesmoveto the left in thePT diagram,indicating temperaturesof,respectively,680°C and 588°C at 6 kbar.Theadjusted Kocurvesare shownas broadlinesin Fig.1O.

If the adjusted Komin.line(garnet coretogetherwith inclusions of biotit e in the core area)iscombined with reaction 5in Fig.9,theposit ionof the intersection poin t of the two curves falls within the stability field for the two-mineral assemblagealm+ky(Fig11).The assemblage isdiagnostic fo r the early stagesof upperamp hibolite fadesin the peliticrocks.Usingtheposit ionof theinter- sectionpointas a quant itative measure ofthePTconditi- ons,oneobtainsa temperatureofc.680°Canda pressu- re of 7.5-8.5 kbar (Fig.11).The lackof staurolite in the matrixofthegneiss alsosuggeststhatthe temperatureis higher thanc.670QC(seeabove). It isbelieved that the tempera t ure and pressurerepresents the met amorphic conditionsduring the F₁fold ing and thedevelopme ntof themainschistosity,51'

The second PT condition represented by Ko values fromthe garnet-rimtoget herwit h matrixbiotiteis harder todefi ne;or expre ssedinanot herway,there are no cur- ves that intersectwith theKoline,defining a point ora restricte d areain the PTdiag ram for thetemperature and pressure.However,the Komax.line gives atemperature that is about 100QCbelow the maximum tem perature obta ined from the Komin. line(Fig.10).TheKomax.met- hod issupposed to give informat ion about the lowest

Temperature

Fig.11.Combinationoftheinforma tio n given inFigs.9and10.Seetextfor furtherexplanation.

10

,.-.,

s:

a

~ '-.../

Cl) 8

!::::

'"Oil 0...~

6 800

700 600

500 400

2+-...£..,.--- ,- --.-- -r---.-"--.'-- ...,...--',,---.---1 300

6

Fig.10.Kolines using the qamet-biotitecalibrationofFerry& Spear(7978).

The lowestKovalues are from the garnet cores together withinclusionsof biotite inthegarnetcorearea.The highest Kovalues are from the garne t rims togetherwithmatrixbiotitein contactwiththegarnet.Thethick lines showtheposition s inthePT diagram after adjustingforferriciron inthe biotit e(seetext).

Temperature(0C)

10 12

of biotite,and onewith 'high' Kovaluesrepresented by the garnet- rim + matrix biotite.The Kovaluesfrom the firstgroup varybetwee n 4.35 and 3.92,whilethosefrom the second group range from 5.0to 4.43.Using the 'Ko min. method'on the low-value group and the 'Ko -max.

method'on the high-valuegroup givesa maximumtem- perat ureof 716°Cat 6kbar andaminimumtemp erature of624°Cat 6 kbar.TheKostabilit ycurvesareshow n in Fig.10.

Thethermo met er probably givessom ewhat too high temperatu res. Rocks of pelit ic compo sit io n will start to melt at about 650-700°C if saturated with water (Bucher

& Frey 1994). In this study no migmatite format ion has

been observed either in outc rop or in thin-section.

Howeve r,the meltin gtem perat ureis str on glydepend ent on the aH₂0.As the H₂0 pressure decreases,the temp- eratureforincipient part ial melting will rise.Ifthe fluid pressure in the rocks is low er than unity, no meltin g occurs despite the hightemperature .Onthe otherhand, in calculating the Kovalues,alliron in the biotite isconsi- dered to bein the ferrous state.Inmost casesthis isnot true, since biotite usually has acertainamount offerric ironinthe tetrahed ralposit io n.Using too hig haquanti ty of ferrou siron in thethermo meterwill givea too high calculated equilib rium temperature. In orderto makea reasonab leadjustment of the ferric-ferrou s rati o in the analysed biotite s, pub lished biot ite analyses of corres- ponding metamorphicrocks havebeen st udied.The gre- at majorityof biot iteshave a ferrous content that varies