Devonian ages from 4°Ar/39Ar dating of plagioclase in dolerite dykes, eastern Varanger Peninsula, North Norway

PHILlPG.GUISE&DAVIDROB ERTS NGU-BULL 440,20 0 2 - PAGE 27

Devonian ages from 4°Ar/39Ar dating of plagioclase in dolerite dykes, eastern Varanger Peninsula, North Norway

PHILlPG.GUISE&DAVIDROBERTS'

Guise,P.G.&Robert s,D.2002 : Devo ni an agesfrom"'Arj"Ar datingof plag ioclase in dol erite dykes,easternVaranger Pen in sula,NorthNorway.Norgesgeologiskeundersokelse 440, 27-37.

Gas-releasespe ctra derivedfro m the analysisof plagi ocl asefrom threegeog raphi callydistinctbut geoch emi cally com pa rabledol eritedykes from theeasternpartofVarange r Penin sula,northernNorway,showsimilar features and favou r aninter p retati o n thatthe dyk es wereintrudedin LateDevoniantime atarou nd370Ma.Theseparticula r dykeshadpreviou sly yielde dfairlysimilar,K-Arwho le-rock ages.Asoneofthedykeshad earlie r beentraced into the Troll fjorden-Ko m ag elv aFaultZonewit h the aid ofaprot on- m agn et om et er,this wou ldindi cat ethat allsig nifi cant, displacivemoveme ntalo ng thismajorfaultzonehadceased bylatestDevonian time.Thedyke agesrep ortedhere fit into aknownpatternof Mid DevoniantoEarlyCarbon iferou sriftin gandsporadi c mafic magmatismreported fromadjacent partsofKolaPen in sula andneighbouringareasalo ngthenorth eastern marg inoftheFen noscand ian Shield.

P.G.Guise,Departmentof Earth Sciences,University ofLeeds,LeedsLS29JT, England,UK(Present address:-56StationRoad, Tadcastet,North Yorkshire,LeedsLS249JR,England,UK).

D.Roberts,Norgesgeologiske undersekeise,7491Trondheim,Norway.

• -Corresponding author:E-mail--david.roberts@ngu.no

(1992). However,isotop icdating st udieson thedykes of the peninsulahavebeenfew.

Introduction

Inthe Caledonidesof Scandinavia,many of thelithostrati- graphic alsuccessions in diversethrustsheets areintr uded by mafic dykes.Occurring either in isolation or,in a few places, in swarms, dykes of this type generally relate to import ant phases of crustalextension and/or rift magma- tismor,less commonly,to late-stage emplacements along joi nts orfaultsin compressionalregimes.Subtle variations in chemical compositionalsoallow for fairassessmentsto be made of the likely palaeot ecton ic settings of these hypabyssalrocks.Inthefew cases whereisotopi c agesare available,these serveas abonus in helping usto define the local geolog icalhistory wit h greaterprecision.

On Varanger Peninsula in Finnmark, in the extrem e northeastof Norway (Fig. 1),theonly sign ofigneous activity is provided by dolerit edykesthat cut Riphean to Vendian Iithostratigr aphic successio nsexposed oneit her side of the maj or, WNW-ESE trend ing, Trollfjord en-Komagelva Fault Zone (TKFZ). The geology of this peninsula is now well known through thesystematic mappingand stratigraph ic and sedim entolog icalstudies of Sied lecka &Siedlec ki (1967) and Siedlecki (1980),summari sed in Siedlecka & Roberts

Fig.1. (a)Outlin e mapshowingthelocat io nof theRybachiandSred ni Peninsulasinrelat iontoVaran gerPenin sula.BSR- Barents Sea Region;

TVR -Tanafjorden-Varangerfjo rd en Reg io n;TKFZ-Tro llfj orden-Komag- elvaFaultZone.(b)Much sim plifie d geologicalmapofVarangerPenin- sula, show ingthelocation s ofthest udied dyke s atFinnvika, Kom agn es andStoreEkkeroya.Separatesimplified legendsare shownfortheTVR and BSR.Majorfaults areindicated by co ntin uousthicklines.

A

B

VarangerPeninsula

TVR

[]]I]]VestertanaGroup

~TanafjordenGroup

G:illvadso Group

c::=::::::J^{BarentsSea Group}

Bar ents Sea Rb hlP . I Kildin Island

, ac ,

oo'"': ~:/ r

sokm Ml,lrTTlll.,.. k

f!..S.B

ISSSSJ^Tanahorn Nappe ..IL..1:L

~LmkvikfjelletGroup

L=:J^{BarentsSea Group}

NGU-BULL440,2002 -PAGE28

AK-Arwho le-rock investigationof dolerit e/metad ol erit e dykesfromdifferent areas on the pen in su la byBeckin sale et al.(1975)is the only detailed work published so far. Other investig ation s restricte d to parti cula r dykes are those of Robert s etal.(1995)and Roberts &Walker(1997).The rehave also been studiesdealing with palaeomagnetic dating of certa indykes(Knut sen1995,Tor svik et al.1995).

Inthiscontribution,we presentthe resultsofa4°Arf'9Ar investigation ofplag io clase separatedfrom two prominent dolerite dykes from the southeastern part of Varanger Peninsu la,south of theTKFZ,oneofwhich has earlieryielded dispar ate K-Ar and palaeomagnetic ages - Late Devonian and Vendia n,respec tively.The4fjAr/³⁹Ar analytical data, in ou r view,help to resolvethesedifferen ces.We alsopresentpla- gioclase analytical data from another dyke,sampled from no rt h oftheTKFZ.Allth ree dykes have given similarplateau ag es.

Geological setting

Ourknowledge of theNeoprot erozo ic-EarlyPalaeozoi c geo- logicalevolution ofVarang erPen insul aderives fromdiverse, detailed investiga ti ons of lithostratigraphy,tectonic struc- ture, low -gr ad e metam orphi sm, microp alaeonto log y and rem ote sensing application s.Specific acco unts or reviews incl ud e those of Siedlecka&Sied lecki(1967,1971),Ban ks et al.(1971),Rob ert s (1972), Sied lecka (1975),Johnson et al.

(1978), Picke ring (1981),Taylo r & Picker ing (1981), Vid al (198 1),Ed w ard s(1984),Bevins etal.(1986),Riceet al.(1989), Karpuz et al. (1993), Rice (1994), Rice & Reiz (1994) and Laajoki(200 2). In addit ion,theeffectsofTima nian defo rm a- tio n in the easternmost parts of Varanger Peninsu la has beendiscu ssed by Roberts(1995,1996).

TheTrollfjord en-Ko m agelva Fault Zone provides a nat-

2

PHILl PG.GUISE& DAVID ROBERT S

ural, st ruct uraldivi de,separati ng the peni nsulainto anort h- eastern Barents Sea Region (BSR) and a so uthwest ern Tanafjorden- Varangerfjo rden Region (TVR).The establi shed, formallit hostrat igra ph iesfo rthetw o regionsare presented in Fig .2.Although detail sdo not concern ushere,theTVR hasbeen terme d apericratonic sedimentationdomain and theBSR abasinaldom ain(Siedlecka & Roberts 1995),with referencetothedeveloping,north east ern passivemarg in of Balt ica in Rip hean-Vend ian time (Olovyani sh ni kov et al.

2000, Rob erts & Sied lecka 2002).Importantly, the critica l stratigraph ic relat ionshi p betweenthetw odomains, involv- ing an unconfo rm it y where the Ekkeroya Fo rm atio n lies direct ly upon a ste eper dipping Bat sfj o rd Formation,has beendescribedby Rice(1994)fromone small area in west- ern Varanger Peninsula.Th is unconformity has also been documented byRobert s & Karpuz(1995).

Dolerite dykesare particularl y common in cert ain parts of the Barent sSea Region,especi ally in northwestern and central areas. Onthe cont rary,dy kes are extremely rare in the TVR.Based on theirK-Ar resul ts, Beckinsale et al.(1975) distinguished two principa lgroups ofdy ke ages: (A)c.360 Ma, and (B) c. 650 Ma (both ages recalcula ted after Dalrympl e1979);and ath ird group(C)of stronglycleaved dykeswit h questionable'ag es'of945to>1900Ma.Dykes of age-grou p sBand Care restri cted to the rocks of the BSR, whereas thefar lesscommongroup Adykes occuron either sideoftheTro llfjord en -Komage lvaFau ltZo ne.Later,unpub - lished work by Beckinsale (pers. comm. 1977) tended to favour an agecloser to550- 560 Ma forthe dykes ofGro u p B.

Palaeomagnet icstudiesby Knutsen(1995)also support eda Vendian ag e.An att em pt to provide betterageconstraints for the dykes of Group A by applicat ion of the 4fjAr1'9Ar met hod to pyroxen es(Robertset al.1995)wasnotsuccess-

3

Age Uthostratigraphic units and theirthicknesses Age Lithostratigraphicunitsandtheir thicknesses Age Lithostratigraphic units and their thicknesses

E Formation Member

E Formation Member Formation Member

io Grascaten upper

'"

<0

~ 280m Lower

'"

~

e Slae,snale200m .... 0 _{1200 m}

Vagge80m 5

'"

0.. 0 _0.. Kistedalro j'" Slqrergarosneset

:::l Gamesfjejet280-300m 0 :::l w ·

0 a: 710-735 m Slack ooanzue10-35 m ..,0 _210m

a: 0 11.-

<.:J Ferrugmoussandstone130m

c:

Z <.:J 5'" Styret

<t Z 'k' member62m Z Sandstone and shale200m z ><

w w Dakkovarre <t ^Z <t

s

I 0

273-350m 'Imember46 m

iE

0- a: rmember35m :::l Ooartzneaod sIlalelOOm z SanOfJordro

a' ⁰ ::;; ::;; ^w

a:

u:

w < 0 w ^{MaSSM?beddedquartane}

0- Z r>tangroes205-255rr <.:J DuoIbasgalssa :xJOm ^Z l'yvJofjellet

0.. f' is

s

:::l Granneset130-200m 500-520 m Thin-bedded200-220quartzitem ^I0.. 0..:::l 1500 m

Ekkeraya 15-1 90m BreMka 800m a' _{g E} Batsfprd Sk"",ka1100-1 300m

* V8 _{I -} 0.. 1.:00-1800m AmlJOI<ka300m

GoIneselva 50-135m :::l Manndrapselva190m <.:J8

g E <0

0..'

StapP09joode* VI ^Wo Hestman600-1300m

:::lE Paddeby25-120m <.:J'" Imerelva 275m ^1/)-

~g ^{Z <t'" 505-545m 1/) '} Basnrenngro God,ela490-1450m

Andersby25-40m ^~ z~ ""0

<.:J0l 0 <t ,

~~ 2500-3500 m Seglodden100-350 m

Q ' z ...." ^{Ulleva:net40-80 m}

I/)~ Fugleberget125m w a:_

< N<Eflngselva 500-1200m

> ^WM Mortensnes 10-60 m

~~ ....

-

KJubbnasro 50m ^I/)_W NjOOrg 200-400 m Kongsfjord * V4 ^~reset Flsfprde1

Veinesbotn 300m > _{>3500m 2000m 1000-1500m}

Sma~jord2·50m

Fig.2 . Lithostratigra phicsuccessions,VarangerPeninsula,show ing thelocati onsof theinvestiga tedsamplesofdolerite dykes,Vl,V4 and VB.Columns 1and2arefromthe Tanafj orden-Varangerfj ord en Region,southwestofthe TKFZ(Fig.1I.andcolum n 3from theBarent s Sea Region. northeast ofthe faultzone.

PHILlP G.GUISE& DAVID ROBERTS

ful,thoughthe datasetdid showa slight biastoward sa pos- sibleDevono-Carb on iferou s age.

TheGroupC dykes,generall yterm ed metadolerites, are part icu larly com monin the Kongsfj o rd dist rict (Fig. 1).They carry a penetra tive cleavage which is also axial plana rto ab u nda nt ENE-WSW-t rendi ng folds in the country rocks (Roberts 1972, Rice & Reiz 1994), and many of thedykes are boudinag ed.One of these Kongsfjo rd dykeshasbeen dat ed by the Sm-Nd methodtoaround 550 Ma(d ata attributedto B.5undvollinAndersen&Sundvoll,1995).

Group B dyk es are most profu se in the Batsfjord area (Fig.1).They are very weakly cleaved,a cleavage which is agai n paralle ltothe axialsurfacesof open totig htfold sin the ho st meta sed im entary rocks. The dykes,cleavag e and fold axes all trend ap proxi mately ENE-WSW. The Group A dolerites are comparat ively fresh and either unmet amor- phosed or very slig ht ly alte red. They generall y trend bet w een N-S and NE-SW.

Geo ch em ically,the Group Cmet ado lerites carry signa - turesquiteclose totho se ofabyssal tholeiites,though som e- what tran sitionalto acontin ental margin reg im e (Roberts 1975).The Group A dolerites,on the ot her hand, have chem- ical features more akin to those of continenta l tholeiites developed in a plate-marginalrather than continental inte- rior situation.Someunpublished geochemical data do exist for dykes of Group B(D.Ro be rts,in prep.),indicating that they are of transition al,oceanic/co nt inenta l tholeiit echarac- ter.

In addition to the above,it should be mentioned that

NG U-BULL 440,2002 - PAGE29

just one dol erite dyke sampled from the easte rn side of Batsfjorden has provided indicati on s of a co nside rab ly youngerage.Basedon prel iminarySm-Nd and Rb-Sranalyt- ical data, this part icula r dy ke may possibl y have been empla ced du ring the Ju rassic period (B.5u nd vo ll, writ ten comm u nication1991).

A few dolerite dykes also occu ron the nearby Rybachi and Sredni Pen in sulasin NW Russia (Fig.1a), cutting litho- st rat ig rap hicalsuccessio ns thatare comparabletothose on Varang erPen in sula (Polkan ov 1935, Sin itsin 1967,Bekk er et al. 197 6,Lyu btsov et al.2000);andsimilardykes are also pre- senton the adjacentmainland of the KolaPeni nsu la cutting Arch aean and Palaeop ro t erozoic gneisses (Fieand t 1912, Hausen 1932,Fed ot ov &Amelin 1998).Som e of these Kol a dykeshave yielde d Vendian ages,whe reasot hers arelat est Devoni an (Juve et al. 1995, Robert s & Onstott 1995, V.

Negrutsa, pers. com m .1991) toEarlyCarb on ifero us (Fed oto v

&Amelin 1998). Inonecase, on Rybach i,there is conf licting evi dencefrom 4°Arf39Arand palaeom agn et icdat ing of one partic u larswarm of mafic dykes,where eitherVendian or Devon ian ages have been sug geste d(cf.Torsviket al.1995, Roberts &Onstott 1995).

The investigated dolerite dykes

Thedykesinvestigated in this argon-d ati ng st udyoccur in theTVR near Komagnes (sam p le V1) and on the island of StoreEkkerey a(V8), and intheBSR close to the small bay Finnvika (V4)(Fig.1).Theirlocati o ns in the lit hostra ti gr aphi- cal successions are shown in Fig.2. Alt hou gh Beckin sale et

Fig.3.(a)The doleritedykejust westofKomagnes,cuttingthin-bedded shalesand mudstonesof theInnerelvaMember of the Sta ppogiedde Formation;photolookingalmostduenorth. (b)Foreshoreexposureof thedolerite dyke just northwest of Finnvika,cutting low-grade,tur- biditic sedimentary rocksof the Kongsfjo rd Formation,BarentsSea Group.

NGU-BULL440,20 0 2 -PA G E30

al. (1975) did not give any precise sampling localities, it seems reasonably certain that all three dykes analysed for thepresent studybelongtotheir GroupAclassification,i.e., with K-Ar whole-rockages of around 360 Ma.With reference to the sample numbers Vl,V4 and V8,these appear to corre- spond to samp les R12,R51 and MlO of Beckinsale et al.

(1975;cf.their fig.l andtabl e1).

Field relationships

Komagnes dyke

Thisc. 2.5m-thickdyke is easy to detect in the old,raised cliff Giviaida,north of the mainroad, c. 1 km west of the promon- tor y Komagn es (Fig. 3). on 1:50,000 map-sheet 2435 I1 Ekkerey (4-NORedition,grid- reference 038 910). The dyke trends c. N-Sand dips at75°east,and cuts through flat-lying, thin-bedd ed, blue-green shales and reddish-grey mud- stones of the Innere lva Member of the Late Vend ian, Stap pogied de For mat ion,the highest partoftheVestert ana Group (Siedlecka & Roberts1992) (Fig. 2). Another, similar, 40 cm-thickdolerite dyke is presentin the same raised cliff c.150 mwest of the main dyke,but thisisnot considered further here.

Closer inspection reveals that the Komagnes dykein the cliff-face really consists of two parallel dykes with a thin screen of hornfelsed sedi me nt in betw een. Tow ard sthe foreshore,and in the intertidalzone,thedyke(or composite dyke)is nearerto verticaland even west-dipping ,and splits intoseveral thinn er dykes wit hoffshootswhich locally curve intoa bedding-parallelorientatio n. Adj acenttothe dyke,or dykes,a near-vertical,widelyspaced fracture cleavage ispre- sent in the hornfelsedshales.The Innerelva Member other- wiseshows littl eeffectsoftectonicdeformation.Thereis a good compactional fabric in the shales, however, and this burial diagenetic event has been dated (Rb-Sr onillit e sub - fract ions)to c. 560 Ma(Gorokhov et al. 2001).

In thin-sect ions of the dyke, a uniform mineralogy is dominated by plagioclaseandc1inopyroxene wit h ophiticto subophitic texture. The plagioclase is andesine to lab rado rite and locally shows oscillatory zoning,and the c1inopyroxene shows the optical properties of pigeonite (Robert s 1975).Accessory minerals are (tit ano)magnetit e, apatite,rare calcite and traces of interst it ial or pyroxene- margin al chlo rite.Thefeld spar shows onlyincipien t stages of sericitisation.

Finnvika dyke

This 9-9.5 m-thick dyke is located c.l km north w est of Fin nvika on the northern coast ofVaranger Peninsula(Fig. 1);

on 1:50,000 map-sheet 2436 11 Syltefjord (4-NOR edition, grid -reference 108 235).The dyke trend s between NE-SW and ENE-WSW and here dips at c. 75° southeas t,cutting through thick-bedded,immat ure sandstones and inte rca- lated, cleaved shaly unit s of the low-grade, turbid itic Kongsfjord Formation, thelow est stra t ig raphic unit ofthe Barents Sea Group (Figs. 2 & 3).

PHILlP G.GUISE& DAVID ROBERTS

The dykecontac t saresub parallelto bedding,butinthe foreshore exposuresthe dykelocallytransectsbedd ingat a low angle in aleft-steppi ng sense befor eresuming a bed- ding-parallelattitude.Cleavageinthe pelite svariesbetween verticaland east-south eastdip orientations.Thedyke shows no evidence of deformation ormetamorph ism,other than small offsetsalong WNW-ESE fault s,and clearly postdates the pervasive cleavage and some related, small-scale uprightfolds.In this same generalareaand sameformation, Taylor & Pickering (1981) reported a Rb-Sr whole -rock isochron age of 520±47 Ma on cleavedmudstones,inter- preted to date the folding and associate d axial-surface cleavage.

Petrographically,the Finnvikadyke is sim ilarto theone from Komagnes,withplag ioclase and c1inopyroxeneshow - ing subophit ic texture;and the feldspar looks to be quite fresh.

Store Ekkeroya dyke

Anear-vertical,N-S-trendingdolerite dyke reachingup to 16 minthicknesscutsthroughgentlyNE-d ipping strata of the Ekkereya Formation,thehigh estunit of theVadseGroup,in the southweste rnpartoftheislandof Store Ekkercya(Figs. 1

&2).The sampling localityis on 1:50,000 map-sheet243511 Ekke rey(4-NORedition)at grid reference890 774.The dyke is particularly prom ine nt in the cliffs at Flagan,a nature reserve and bird colony.TheEkkereyaFormation herecon- sists of medium-bedded sandston e and subordinat econ- glomerate with intercalations of siltstone and mudston e (Siedlecka& Robert s 1992),and is devoid of cleavage or folds.

Inthin-section, the centralpart of theStoreEkkereya dyke is coarser grained than the other two dolerite dykes and varies in texture from ophitic to locally glome ropor- phyr itic , wit h the plagioclase lat hs occurring in scatte red clustersup to 3.5-4mm across. Both thesefeldspar clusters and the normal,individuallaths show variable degreesof sericitisation . Nearer thedyke margins,grainsize iscompa r- able to that in the Komagnes and Finn vika dykesamples, andsericitisationis more advanced. Themin eralparagenesis isotherwisethe sameas for the Komagnes and Finnvika dykes.

4 °ArJ39Ar dating

Analytical procedure

The plagiocl ase separateswere prepared at theGeological Surveyof Norway(NGU),Trondheim,and the sam plesirradi- ated at Rise Nuclear Reactor,Roskilde,Denma rk. Fullproce- duraldetailsarepresent edin anAppendix.The fastneutron dosewas monitored by Leeds biot it estandard Tinto,409.2 Ma(Rexetal. 1986)and hornblende Hb3gr, 1072 Ma(Turner et al. 1971). Fluxvariat io n over the packagelengt hwasof the order of 3%.The interferencecorrect io n factor s used were;(40/39)K= 0.048,(36/ 39)Ca= 0.38 and (37/39)Ca = 1492.

PH/LIP G. GU/SE &DAVIDROBERTS NGU-BUL L 440,20 02 - PAGE 31

Table 1. "'Arl"Aranalyticaldata.

Temp 39Arl( )7Arc~ 18Aro

-c

~A r%At m 19Arl( 40Ar

Age Error %"Ar,

{ Ma I

V1 Plagioclase

boxheightsare 1(J

470, - -- - - - - -- - - - - - - -- - - ,

660 0.20 3.1 0.008

755 0.53 11.7 O.OOg 830 0.68 16.6 0.008 885 0.55 12.5 0.008

930 0.39 7.7 0.006

980 0.32 4.3 0.005

1010 0.24 2.4 0.006 1060 0.32 4.0 0.006 1165 0.67 10.3 0.015 1300 0.46 8.8 0.012

32 44 49 45 39 27 20 25 31 38

52.47 58.6 43.87 40.7 43.82 23.4 43.72 14.4 44.00 35.9 44.33 39.5 44.77 22.2 44.86 17.5 49.64 15.7 52.45 46.1

442.7 16.6 4.5 377.1 3.3 12.2 376.7 4.8 15.6 376.0 3.7 12.7 378.2 5.2 9.0 380.7 10.4 7.2 384.1 9.8 5.5 384.8 6.1 7.3 421.4 5.2 15.3 442.5 5.3 10.5

450

430

Age410

(Ma)

390

Platea uage=371.6:l:1.8Ma(10') MSW D=0.15,probabdlty=O.99

Include s 69.7% ofthe39Ar

V1 Plagioclase,run 2461 weight=0.05322g,Jvalue=0.00530±0.5% Tot al gasage 395±3Ma (weig ht%K=0.22, ''''Ar=37.7x10'cm3g')

370

80 100

Model1 on 7 points Age=377,2 t84 Ma

MSWD=0.16 40136interceI 296t1

40 60

Cumulative39Ar Percent

20

0,0014 0.0016

V1 Plagioclase

350+ - - - + - - - - + - - - <- - - -1-- - - - 1

o

a

5.8 17.9 10.9 6.4 5.8 7.0 7.5 13.2 22.2 3.4 2.8 0.4 1.0 2.0 1.4 1.7 2.3 0.8 0.4 2.8 367.4 368.9 372.1 369.9 367.1 368.8 369.4 365.9 368.2 394.9 37.3

9.6 9.3 14.0 12.2 17.3 16.5 8.9 7.0 32.7 42.61 42.81 43.22 42.94 42.58 42.79 42.87 42.43 42.72 46.17 2.3 6.3 6.0 4.1 2.7 1.6 1.6 1.9 3.0 12.2 0.04

om

0.04 0.03 0.03 0.03 0.04 0.06 0.12 0.03 1.5 1.8 4.7 15.0 2.9 8.6 1.7 3.5 1.5 2.1 1.8 1.5 2.0 1.6 3.5 3.3 5.9 8.7 0.9 5.5 630

760 830 865 905 955 1000 1050 1160 1300

0.0012

V4Plagioclase,run2462 weight=0.06140g,Jvalue=0.00530±0.5% Totalgas age 369±2Ma (weight%K=1.2,''''Ar=185x10"cm3g')

0.022 0.020

0,016 0.0 18 39Ar

r

0,014 0,0006

0.0002+ - - - + - - - - 1 - - - + - - - + - - " - - - - 1 0.0 12

0,0004 0,0008

V1 Plagioclase

b

4.2 2.1 0.8 5.5 0.7 8.2 0.6 8.7 0.5 8.7 0.3 13.3 0.2 15.0 0.2 16.0 0.4 15.3 0.6 7.2 308.7

367.8 384.2 378.0 375.7 373.8 374.6 375.1 438.8 491.4 35.22 80.0 42.66 29.2 44.78 9.0 43.98 14.2 43.69 9.2 43.44 8.0 43.54 10.1 43.60 12.7 51.96 20.9 59.08 22.0 0.14 0.8

0.17 2.0 0.10 3.8 0.13 2.5 0.11 1.1 0.17 0.8 0.22 0.5 0.28 0.6 0.66 1.5 0.43 3.9 1.4 0.5

3.6 3.7 5.4 10.3 5.7 7.1 5.7 3.2 8.7 3.4 9.8 2.5 10.5 3.3 10.0 7.3 4.7 9.1 555

680 750 825 870 920 975 1025 1120 1300

V8Plagioclase, run2460 weight=0.08236g,Jvalue=0.00530±0.5%

Total gas age 393± 2Ma (weight%K=2.1 ,''''Ar=366x10'cm3g') ^{60,--- - - -- -- - -} - -- - - -- - - , Errorsare1(1. °«iArJvolumeof Radiogenic«iAr,gasvolumescorrected to STP.

50

Isotopic analyses were performed withamodifiedMS10 mass spectromete r; measured at mospheric 4°Ar/'6Ar was 287.8±0.2and sensitivity 1.1 x10.⁷cm'V '.The J-valueun- certainty is includedin the errors for the totalgas ages but the individual step ages have analytical errors only.All errors are quoted at the 1-sigmalevel. The analytical data are pre- sentedin Table 1.Age spectraand age correlation plotswere produced using'Isoplot/Ex'(Ludw ig 2000).

40

- CalK

30

-

20

Fig.4. Plagioclase samp leVl:(a)4OAr/"Ar age spectru m;(b) isotope correlation plot;{c)CalKplot.

experim ent lieon the excess argon sideof the line ar array, evidence forthecont ribut ionofolderagesof these ste ps.

20 40 60 80 100

Cumulative39Ar Percent

10+---+----+----t---+--~

o

c

Results and interpretation

Vl plagioclase:The agespectrumshowssome disturbance with older ages at both low- and high-temperature,gas release steps.A plateauof 377.6±1.8Mais defined by 70%

of the gasreleased(Fig.4).The isotope correlation plotgives a good lineartrendfortheplateausteps and yield sthesame age. The intercept4°Ar/'6Ar of 296is close to the accepted at mosphericargon value of 295.5.The lasttwo steps of the

NGU-BU LL 44 0 , 20 0 2 -PAGE32 PH ILlPG.GU ISE& DAVIDROBERT S

36Art^OAr0.0008

Fig.5. Plagioclase sampleV4:(al"'Ar/^l9Ar age spectrum;(b)isotope correlationplot;CalKplot.

in V1 and V8may relat e to acquisition ofradiogenicargon from the countr yrocksduring magm aascent;orfrom fluids

100

100 0.024 0.022 0.020

UoOeIton9poI'U AQe·36e~t 35"'"

~1I'lte<CeOl291.3!83

USWO· 018.

0.018 39A rt^OAr 0.016

Plateauage=-368 571:0 23Ma(10) MSWO=0 65.probablhty=O 73

Includes 966%of!he"AI

~ ~ 60 60

Cu m u lativ e39ArPerc ent

20 40 60 60

Cum u la t iv e39A r Perce n t 0.014

0~--_+_---+_--_--- __1

o

0.0016" ' - - - ,

0.0012

0.0004

V4Plagio clas e

0.0000+ - - - < - - - + - - - _ - - > - - - - + - - - ' - - - 1 0.012

V4Plagioc las e

box heights are 10:

400

12

14. . -- - - - - - - - - - - - - - - - --,

10

370

EJf::==~:::::::J===-i

360

J e-

390~

Age (Ma)380

a

CalK

b

c

Variation sintheCalK ratioshowno correlation withthe age spectrum. This indicates that all the slight composi- tional variat ions present in the plagioclases are therefore givingthe same age.Itsho uldbenotedthat theKcontent of this particular plagioclase is much lowerthan that of the oth er tw osamples, resultinginthe largererrors on agesfor individual steps and higher CalK values.This may suggest thehidden presenceof subtledifferencesin the mineralogy of thisparticularseparate,compared wit hthatof V4 and V8.

We interpret the 378±2 Maplateau age(Fig.4) as closely correspo nding to the actualcrystallis ationage of this dyke.

V4 plagioclase: The age spect rumofV4shows a well- define dplateauat 369Ma consist ing ofmorethan 95%of thegas released(Fig.5).This ageisfurth er confirmed by the isotope correlationplot which has agood lineartrend with

the4°Ar1'6Ar intercept at 297, again close tothe atmospheric

argo nvalue. In this sample,thereis no indication ofexcess argon being present. This spect rum conforms to that expected of a sample that underwentrapid coolingand has sinceremained thermally undisturbed.The variation in the CalKrat ios ofthe steps isnot reflec tedinthe agespectrum . The369±0.23Maplateaulook stoprovid e asgood an age as oneislikelyto getfromthistechniqu e, and this is sup- ported bythe inverse isochron plot.Accordingly, we inter- pret thisto represent the crystallisation age of the Finvika dyke.

VBplagioclase:This samp leshowsadisturb ed spectrum with evidence ofargon loss in the low-temperature steps and thepresence of excessargon inthe high-tem perature ste ps(Fig.6).Thedistu rbanceis such thatthereis no plateau as defined by the criteria of Isoplot/Ex (Ludwig 2000).

However,a fitwas forced through the steps shown in Fig.6 to givean ageof 375±1 Ma(95% confidence)wit h just over 50%Of thegasreleased.The isotope correlation plotagain givesa lineartrend,but confirms the presence of excess argon in the two high est temperat ure steps. A lin e fitted throughthe first eight stepsofthe experimentgives an age of377 Maand an4OArj36Ar intercept of 278 with large uncer- taint y.This value is lower than the accepted at mos pheric argo n value, whichwould be expectedwhen there has been loss of radiogenic4OAr.Variat ion in theCalK ratio is approxi- matelymirroredinthe age spectrum,indicati ng that slightly diffe ringcomp osit ionalvariati on s could also be contribut- ingtothedisturbanc eofthe age spect rum.

We int erpret the weighted mean age of c. 375 Ma as a likely approximation tothe intr usiveage of thisdyke.

Of the three samp les,the age of V4 at 369 Ma can be accorded the hig hest confid ence.The othertwo samples,V1 and V8,both show varying degreesof disturbance in their spectra .It is noteworthy that the V4 samplewas collected from the very centreof the 9 m-thickFinnvika dyke.Sample V1,on the ot her hand,wastakenc.80cmin from the margin of the 2.5 m-th ick Komagnes dyke; and V8approximately 1met re in from themargin ofthe 16 m-thick StoreEkkeroya dyke.Itistherefore possiblethatthe disturbancesrecorded

PH/LIPG.GUISE&DAV/D ROBERTS ^{NG U-BULL 440,2002 - PAGE}

33 1

box heightsare 20

occurrenceofa significantevent- and mostly likelyone of dykeintrusion- ataround370Ma.

340

CalK Age

(Ma)380f---'~3== _____j

Fig. 6. Plagioclasesam pleV8:(a),oAr/^lOAragespect rum; (b) isotope correlation plot;(clCalKplot.

Discussion

Based on all published dati ng,bot h isot opic and palaeo- magn eti c,of maficdykescutt ing Neop rot erozoicrocksfrom theVaranger-North Kolasegmentof thenorthernmargin of theFennoscandian Shield, there appear tobetwo principal ages of dolerite dyke int rusion, namely Vendi an and Devoni an. Thedykesinvestigated byusbelongtoGroup A of Beckin sale et al.(1975), i.e.,those which provided Late Devoni an, K-Arwho le-rockages.One of thesedolerit edykes, theV1 dykefrom Komag nes,has also been analysed palaeo- magn eti cally, and in this caseprovid ed aVend ian agewith noindicati on ofany Devon ian magnetic resettin g (Torsviket al. 1995).The palaeomagneti c datafromthe Komag nesdyke arealmost ide ntica l to thosederivedfrom adolerit edyke from the Sredn i Penin sulathat,in this case,hasprovid ed a Vend ian age bybot h K-Ar and4°Ar1'9Aranalytica l meth ods.

Notwith standin g thisevide nt conflict of result sbet ween thepalaeom agnet ic and isotopic dating ofthe Komagnes dyke,theplagioclase age spect ra report edhere from these three,widely separated dykes are mutually comparableand, togeth er, favour an interpr etation that the dykes were emplacedand crystallised ataround370 Ma.This inter pre- tation thus supports the earlier K-Ar datin g st udy of Beckin sale etal.(1975),wherethesevery samedykesyield ed concordantmaximum agesofaround363±10 Ma (recalcu- lated follow in g Dalrymple 1979).On current Phanerozoic tim e-scales,theDevon ian-Carboniferou s boundaryisplaced at eit her362(Tuckeret al. 1998) or 355 Ma (Remaneet al.

2000),and theVarangerdykes analysed inthis investi gation would thusfall in the Famenni an stag eof LateDevonian time.

Inawiderperspect iveof theFenn oscandian Shieldand EastEurop eanCraton,the easte rnand northernmargin s of Baltica werecharacteri sed by amajor episodeof rifti ngin Mid Devoni an to Early Carboniferoustime (Ziegler 1988, Johansen et al.1993,Nikishin etal. 1996).Rift basinforma- tion along a mainly NW-SE trend occurred beneath the PechoraBasin and easte rn Barents Sea (Fig.7), mimi cking the st ruct ural trendin thepre-Palaeozoicbasement.In the Kola Penin sula region and westernpart s of the BarentsSea, amore NE-SWrifting trend is evide nt (Nikishinetal. 1996, Gudlaugsson et al. 1998,Wilson et al. 1999),and ageneral doming of the Kola-Whit e Sea area occurred in Late Devon iantime (Fig.7).In theTiman-Pechorarift syste m,Late Devoni an basaltic volcanism was widespread, and the Kontozero graben on Kola Peninsula is well known for its alkalin eand kimb erlitic magm atismdurin g theperiod 380 to 360 Ma(Kramm etal. 1993).ln weste rn part s of Kola,Pb- Zn veinmineralisation sof LateDevonian ageare com mo nly associatedwith N-Sto NE-SWtrendin gdolerit edykes (Juve et al. 1995).

An allied topi c isthefact that U-Pb zircon data from

100 0,026

=

0,022 80

0,018 Model2 on 8 poonts Age:378 8t94Ma 40/36inle ,cept: 2 78t60 MSWO =4 1.Probab ihty=O.OOO

0,014 39Art^OAr

- V8 Plagioclase

0,010

40 60

Cumulat ive39Ar Percent

Plateauage = 375 ± 1Ma(95% conf.)] MSW D"5.6.probablhty=O.OOl

Includes529% oflhe)!iAr

20

20 40 60 80

Cumulative39Ar Percent

0,006

2 -

0+-- - -+-- - - -+-- - --+- - - -;-- ----1

o

0,000+----<---_---.----+---+-'~-__1

0,002

260" -- - --+-- - - -+-- --<-- -- -->-- - ---1

o 100

0,003

0,002

0,001 420 460 500

c b a

36Art^OAr

circulating along or closetothedykemargins.Whatever the case,the three,fairly similar ages do tend to confirm the

NGU-BUL L440,2002 -PAGE 34 PHILlP G.GUISE& DA VIDROBERTS

[SS]

o

"-''---LJ oceaniccrust,with spreading axis

~

o

§

~

Fig.? Simplifiedsketch-map ofBalti cashowi ng themain Late Devoni an-Early Carboniferousriftsbasinsand otherfeatures.Modified from Nikishinet al.(1996). KD-KolaDome;KG-Kontozero graben;PB- Pechora Basin; SBAB- Sakmarianback-arcbasin;TH-TimanHigh;VD-Vyatka Dome.

Palaeoproterozoic and Archaean rocksfrom thisnorthern Fennoscandiandomaincommonly show Devonian to Early Carboniferous lower intercepts on concordia (e.g., Levchenko et al. 1995,Larson & Tullborg 1998).A similar, Devonian,lower intercept date has been reported from a dole rite dyke with an interpretedVendian age(U-Pb, zircon) from near Hamningbergin the Barents Sea Region (Roberts

&Walker 1997).This recurrent feature has beeninterpret ed by Larson&Tullborg(1998)torelateto the thermal effects of a Devonian foreland basin,with a sediment cover ~3 km thick (now removed),arising from rapid erosion of the Caledonian mou ntain chain. Alt hough supp ort for this notion of a Devonian sedimentary blanket comes from ongo ing fission-tr ack stu dies (G.Murrell,pers.comm.2002), Nikishin et al. (1996) have stated that theFenno scandian Shield,except for it srift basins,"mayhaveremained emer- gent throughout the Devonian". Wheth er or not a thick, Devonian ,foreland basin sedi mentary cover exist ed inthis shieldarea,thereisnow suff icientevidence fromthisnorth- ernFennoscandianregion that there aremafic dykesofbot h Devon ian and Vend ian age, andthatthe latt er mayor may notshowDevonian overprints.

Ret urni ng to Varanger Peninsula, and the Komagnes dyke inparticular,Herrevold (1993) and Karpuzet al.(1993) have reportedthat the dykecould be tracedinland wit h the help of a hand borne proton-magnetometerbeneath a thin Quaternary cover into the trase of the Trollfjorden- Komagelva Fault Zone,without change of strike,until its magnetic signature eventually faded and disappeared.No strike-slip offset of the magnetic anomaly could be detected.Accepting that the dyke is almost certainly of Devonian age,then this would indicate that all majorst rike- slip movement s along the fault zone had ceased by Late

Devoni an time.A simil ar suggestion was also made by Beckin saleetal. (1975), at that time based on the general, though sporadic occurr ence of their Group A dykes on eithersideof theTKFZ.

Conclus ions

Age spectraderivedfrom the analysisof plagioclasesepa- rates from three,geochemically similardoleritedykesfrom the eastern part of the Varanger Peninsula, north ern Norway, show comparab le featuresand leadto aninterpre- tation that thedykes wereint rudedin LateDevoniantime,at c.370 Ma.These particular dykes had previously yielded quit e simil ar,K-Arwhole-rockages. AninferredVendianage for oneof the dykes,at Komagnes,based on palaeomag- net ic data,should now be dismissed. Thisparticular dyke has beentraced beneath thin superficial deposits,by mag- net ometer, directl y int o the Trollfjorden-Komagelva Fault Zone,wit ho ut any visiblestrike-slipoffset of the magn etic anoma ly,thus indicatingthat allmajor displacivemovement alongthis fault zone had ceasedbylatestDevoniantime.

Acknowledgements

We are grateful toElizabethEide andAnnaSiedlecka for theirpertinent and most useful comments on an early version of the manuscript.

Const ructive comme nts and valuable suggestions by the reviewers, MartinTimmerman andStephenLippard,were helpful inimproving the final manuscript.Irene Lundquistassistedin draftingthree of the fig- ures.