Magmatism during late Ordovician-early Silurian accretion of the Caledonides of Arctic Scandinavia: the Halti–Guolasjávri area revisited

: Magmatism during late


Introduction
During final closure of the Iapetus Ocean and initial collision of the Laurentia and Baltica continents, marginal basins with mafic igneous activity developed along the subducting Baltoscandian margin (cf.Slagstad & Kirkland 2018 for summary).In northern parts of the Scandinavian Caledonides, the evidence of this late Ordovician -early Silurian magmatism was considered restricted to terranes derived from the Iapetus Ocean and the Laurentian margin until ages between 435-450 Ma (Vaasjoki & Sipilä 2001;Andréasson et al. 2003) were obtained from the Halti Igneous Complex (HIC).The complex intruded continental sandstones carried by the Corrovarre Nappe (CN), considered derived from Baltoscandian rift basins.Various orogenic models, such as subduction flip, slab roll-back, and ridge subduction, have been proposed for the short-lived magmatism in the CN and coeval magmatism in overlying nappes of oceanic origin (Andréasson et al. 2003;Kirkland et al. 2005Kirkland et al. , 2007b;;Corfu et al. 2006Corfu et al. , 2011;;Slagstad & Kirkland 2018).The CN has by tradition been treated as the uppermost nappe of the Kalak Nappe Complex (KNC; Zwaan 1988;Zwaan & van Roermund 1990; http://geo.ngu.no/kart/berggrunn/).Lack of consensus concerning the affinity (Baltica or Laurentia?) of the KNC (Kirkland et al. 2007a, Corfu et al. 2014 for review), and the proposed relocations of the orogenic (Scandian) suture (Slagstad & Kirkland 2018) add fuel to the debate.
The CN is overlain by the Vaddas Nappe (VN; Lindahl et al. 2005), considered bipartite with a lower part composed of quartzite, meta-arkose, marbles and black schists hosting metabasites, and an upper part of conglomerates and marbles passing upwards into turbidites hosting volcanics, gabbro bodies and sulphide ores.The boundary between these parts of the nappe, at Lake Guolasjávri with conglomerates and marbles preserving late Ordovician-Silurian fossils (Binns & Gayer 1980), has been considered by some geologists as a major regional unconformity, by others of none, or of only local significance (cf.Binns 1989).
In the Halti-Guolasjávri area (Fig. 2A), the pristine rocks of the HIC and the fossiliferous limestone, well preserved pillow lavas, pyroclastics and turbidites of the VN provide excellent opportunities to investigate further the enigmatic age relations.Based on published and own documentation from the area, our revisit includes reexamination of the tectonostraticgraphic settings of units and of the debated unconformity below the fossils.New age data for HIC magmatism are presented, but now from anatectic granitic dykes associated with the intrusions.Aware of the need for geochronological and geochemical confirmation, we add our new insights and data to a previously proposed model for the late Ordovician-early Silurian extensional magmatism by arc-continent collision, but now within a contiguous Laurentia-Baltica scenario.Geographic coordinates used in the text refer to the WGS84 (decimals) system.

Regional geological setting
The Caledonides of northernmost Norway are dominated by the Kalak Nappe Complex (KNC) of continental affinity (Fig. 2A) and the overlying Reisa Nappe Complex of mainly oceanic origin (Corfu et al. 2014;Corfu & Andersen 2022).Isolated volcano-sedimentary and magmatic successions (Magerøy Nappe) occur on the islands of Magerøy and Sørøy and on the Porsanger pensinsula.The lower KNC is mainly composed of quartzo-feldspathic sandstones, variously intruded by dolerite dykes and minor granites.The highest components of the complex, Seiland, Gjesvaer and Corrovarre nappes, are of higher metamorphic grade and include igneous suites.The Sørøy-Seiland Nappe is dominated by the Seiland Igneous Province (SIP) ranging in composition from alkaline gabbro to nepheline syenite.
The Gjesvaer Nappe is dominated by migmatites.The KNC overrides other thrust sheets, also dominated by siliciclastic successions, the Laksefjord and Gaissa nappes, and all these allochthons overlie a thin Cambrian-Ediacaran sedimentary veneer and the Precambrian crystalline basement of Baltica.

Figure 2. (A)
The map is compiled from Lindahl et al. (2005) and Zwaan (1988).The legend is compiled with modifications from Lindahl et al. (2005) and Zwaan (1984).Numbers of units  refer to the map compiled by Zwaan (1984).(B) The map is compiled with modifications from Zwaan (1984), Lehtovaara (1995), Sipilä (1991), http://geo.ngu.no/kart/berggrunn/, and our own reconnaissance.White dots are localities of sampled granitic dykes (1-2, cf. Fig. 4-5;Andréasson et al. 2013), plagiogranite in gabbro (3;Andréasson et al. 2003) and gabbroic and granitic dykes (4-5;Vaasjoki & Sipilä 2001).Note that the blue colour in the section refers to the blue and green colour in the map, and the light blue colour in the section to the light blue and violet colour in the map.

Baltoscandian rift basins and dyke swarms
Evidences of magmatism during different stages of the evolution of the Baltoscandian Margin, from proximal rift basins to the continent-ocean transition (COT), are preserved as swarms of mafic dykes extending from Egersund in southernmost Norway to the fjords of the Barents Sea, the Baltoscandian Dyke Swarms (BDS; Andréasson 1994;Kumpulainen et al. 2021) also referred to as the Scandinavian Dyke Complex (Tegner et al. 2019).At Egersund, the dyke swarm is found in the Precambrian basement (Bingen et al. 1988); elsewhere, the swarms are carried by the Seve Nappe Complex (SNC) and Särv Nappe and equivalent nappes (Andréasson et al. 1998;2005;Gee et al. 2020).Recent studies (Tegner et al. 2019;Kjøll et al. 2019aKjøll et al. , b, 2020) ) have considerably improved our knowledge about the magmatectonic and structural evolution of the BDS as part of the Central Iapetus Magmatic Province (CIMP; Ernst & Bell 2010).The concept of a preserved, belt-long system of c. 600 Ma old rift magmatism proved useful also for reconstructions of the colliding margin.Some nappes of the SNC carry mafic dykes and lavas metamorphosed under high to ultrahigh pressures during subduction of rift basins and basement plinths.Some rift basins and the outermost margin and continent-ocean transition (COT) were decoupled from the diving margin and escaped subduction and now occur in the Sarek, Abisko and Indre Troms mountains (Andréasson 1986;Kathol 1989;Stølen 1989).
In North Norway, high-grade units of the Kalak Nappe Complex (KNC) have been treated as equivalent to the SNC.Andréasson (1996) grouped all terranes characterised by inferred Ediacaran rift basin fill and magmatism into the 'Seve Superterrane', later (Andréasson et al. 1998) proposed to include also the KNC ('Seve-Kalak Superterrane').Subsequent work soon disclosed such wholesale correlation of Seve and Kalak units as too simplistic (Andréasson et al. 2018).However, the correlation between a Seve terrane in the Kebnekaise Mts. (Fig. 1) and a Kalak unit in Finnmark based on c. 845 Ma granites hosted by psammitic and migmatitic gneisses proposed by Paulsson & Andréasson 2002 (cf.also Andréasson et al. 2018) is maintained here.Underlying the KNC, the Laksefjord nappes (Fig. 1) include dolerite-intruded arkosic sandstones.Dolerite chemistry (Gayer et al. 1985) compares to that of the BDS and detrital-zircon age distributions of the arkosic sandstones (Gee et al. 2017) compare to those of the BDS host rocks.
The belt-long BDS concept was widely accepted until the KNC and SNC were declared exotic in origin (Kirkland et al. 2007a, Corfu et al. 2006, 2007, 2011;Corfu in Andréasson et al. 2013;Slagstad & Kirkland 2018).Corfu et al. (2014) reviewed unsolved problems of the origin of the SNC and the Seve-Kalak relations.Recently, Slagstad et al (2020, their Fig. 11) lumped together the SNC and Laurentian margin rift magmatic provinces and placed Särv dyke swarms separated from the CIMP by an upside-down Baltica, resulting in a scenario where the basement-hosted Egersund dyke swarm and Fen and Alnön carbonatites remain as the only evidence of Baltoscandian rift magmatism.In the present study, we treat the Corrovarri Nappe and Seve COT units in Troms county as Baltocandian in origin (cf.also Kjøll et al. 2019a, Fig. 11).

Corrovarri Nappe
At Guolasjávri (Figs. 2B, 8A), the CN is dominated by the Raisduoddar-Halti complex of troctolite and dunite and the syngenetic Ridnitsohkka swarm of gabbro dykes and sills (Sipilä 1991;Lehtovaara 1995), in the following together referred to as Halti Igneous Complex (HIC).Along the lower western slope of Halti (Fig. 2B), the floor thrust of the CN developed below the gabbroic body and is defined by flaser or amphibolitic gabbro overlying mylonitic psammitic or granodioritic gneiss and locally garnetiferous granite of the Nabar Nappe (KNC).Bøe (1976; cf. also http://geo.ngu.no/kart/berggrunn/) referred to the foliated amphibolite at the base of the gabbro body as ʻkyanite-bearing amphiboliteʾ.We observed coarse aggregates of kyanite in the necks of stretched amphibolite, which could represent growth during vertical thinning following nappe stacking.In places, the amphibolite passes into greenschist.East-south-east-oriented isoclinal folds and stretching lineation developed in the mylonitic psammite below the gabbro.Blastomylonitic quartz-feldspathic gneiss with biotitegarnet fabric was sampled (G86-89) at the base of the Ridnitsohkka dyke swarm, above gneiss of the Nabar Nappe.To the west of Lake Guolasjávri (outside the map area of Fig. 2), the CN pinches out and a thin garnetiferous phyllonite separates the VN from the Precambrian granodioritic gneiss of the Nabar Nappe.On the southern and eastern slopes of Halti, the floor thrust beneath the gabbro dyke swarm is defined by blastomylonitic meta-arkose.
The roof thrust of the CN is exposed at the Goddejohka stream at the western end of Lake Guolasjávri (Fig. 2B), where bodies of metabasite and gabbro occur close below the thrust.East of the lake, c. 300 m of meadow separate outcrops of folded meta-arkose, meta-basite and palingenetic granite of the Ridnitsohkka dyke swarm from rusty quartzite and black schist of the Vaddas Nappe (Fig. 4).
Two samples of the granite were dated by ID-TIMS on zircon and rutile at the University of Oslo.Zircon grains were selected under a binocular microscope, air abraded, and spiked with a mixed 202-205Pb-235U spike.Other details of the procedure are given in Corfu (2004).Analytical data are presented in Table 1.Sample H01:18 was collected from granitic dykes folded with gabbroic dykes and host-rock meta-arkose (Fig. 4A-B).The sample contains euhedral to variably resorbed zircon prisms.
Three zircon analyses are concordant but show a slight spread along the Concordia curve (Fig. 5A), which is assumed to be due to Pb loss.The oldest analysis is thus considered to most closely approximate the time of crystallisation yielding 441.2 ±1.8 Ma.The age is consistent with the less precise analysis of coexisting rutile.The alternative interpretation is that the oldest analysis could be affected by some inheritance, in which case the real age would be somewhat lower around 436 Ma.
More analyses would be needed to sort out this question.Sample H01:19 is a 10 cm-thick granite dyke cutting a gabbro dyke (Fig. 4C).The zircon population is dominated by prisms locally surrounding cores.
Two analyses of abraded zircon tips without cores yield overlapping concordant data indicating an age of 436.7 ± 2.5 Ma (Fig. 5B).
2,4) weight and concentrations are known to better than 10%, except for those near and below the ca. 1 ug limit of resolution of the balance.
3) Th/U model ratio inferred from 208/206 ratio and age of sample.4) Pbc = total common Pb in sample (initial +blank).5) raw data corrected for fractionation and blank.6) corrected for fractionation, spike, blank and initial common Pb (based on Stacey and Kramers, 1975); error calculated by propagating the main sources of uncertainty.

Vaddas Nappe
Lindahl et al. ( 2005) described the Vaddas Nappe as a "right way up stratigraphic and magmatic succession" with an estimated thickness of <1000 m for the lower, shallow-water sequence (Kvaenangen Group), and >1600 m for the upper, volcano-sedimentary succession (Oksfjord Group).
The nappe wedges out rapidly towards southeast and northwest; locally on the coastal islands (e. g. Uløya, Fig. 2A), the VN is only a couple of hundred metres thick (http://geo.ngu.no/kart/berggrunn/).
The stratigraphic subdivision and names used by Lindahl et al. (2005) in the VN type area are retained here (Fig. 2A), with the exception of the basal formation (Gaeirajávri Formation) of the Kvaenangen Group which we, for reasons given below, suggest be treated as a member of the overlying Ciččenvárri Formation.

Ciččenvárri Formation
The formation takes its name from Ciččenvárri hill 20 km WNW Lake Guolasjávri ("C" in Fig. 1).
Here, the at least 600 m-thick formation consists of quartzite cut by a dense swarm of dolerite dykes (Fig. 6).Over a distance of 50 m, we counted 10 dykes, 1-3 m wide; locally the sedimentary screen is only half a metre wide.Sheeted dykes occur.The dolerite is ophitic and plagioclase porphyritic.
The quartzite is white, but weathers brown due to disseminated sulphides.Meta-arkose and graphitic schist are subordinate.From the slopes of Ciččenvárri and the Abmelăsvággi valley adjacent in the north, Padget (1955, p. 34-35) described the Big Limestones Series.This occurs below the quartzite as three gently dipping, <30 m-thick sheets of grey marble, intercalated by garnet-mica schist (Fig. 6B).
Dolerite dykes intruded, but due to the ductility of the marble, the dykes now appear as bands or twisted lenses; ophitic texture is only locally preserved.At the bottom of the valley, blastomylonitic mica schist overlies granodioritic gneisses of the Nabar Nappe.The same succession of intercalated mica schist and marble occurs also at higher level within the quartzite and meta-arkose; and on the northwestern corner of the Ciččenvárri hill it is overlain by schists of the Oksfjord Group (Zwaan 1988).
At Lake Guolasjávri, the Ciččenvárri formation is strongly reduced in thickness, and telescoped.Rusty quartzite, foliated grey marble and graphitic mica schist alternate at the base; all with bands or boudins of metabasite, some of which preserve ophitic textures (Fig. 7A-C).According to our guide (K.B.Zwaan pers. comm. 1988), the quartzite is typical of the Ciččenvárri Formation and is treated here as corresponding to the Riehppejohka Quartzite member of Lindahl et al. (2005).We interpret the marble as equivalent to the Big Limestones Series.The schist carries porphyroblasts of kyanite and staurolite and garben sheaves of hornblende (Fig. 7D-E).Away from the floor thrust of the Vaddas Nappe, the schist becomes phyllitic, and mafic dykes preserve their intrusive shape and chilled margins (Fig. 7F).We interpret this schist as the Oksfjorddalen Schist Member (Lindahl et al. (2005).
The matrix is mostly carbonaceous and schistose (muscovite and brown or green biotite) carrying amphibole and locally rich in garnets (Fig. 8C).In samples with pelitic matrix (G86-33), conspicuous aggregates of skeletal garnet growing along quartz grain boundaries indicate postkinematic growth at low metamorphic grade.The marble is grey with brownish-yellow weathering and is commonly schistose.In places, the carbonaceous schist passes into conglomerate with garben schist as matrix and with pebbles of quartzite.The succession is telescoped, but treated here as equivalent to the Jiehkkejohka Marble Member and the Rássevággi Schist and Conglomerate Member of Lindahl et al. (2005).Mafic dykes cutting the succession were not observed.The polymict conglomerate with intercalations of metabasite (Fig. 9F) may correspond to the Frukosthaugen Conglomeratic Marble member and the overlying Loftani Greenstone Member.(Binns & Gayer 1980, Nature 284, 53-55).(E) Marble with sparse pebbles of quartzite overlain by volcanics (V) of the Loftani Greenstone Member.

Akkejávri and Ankerlia formations
The Akkejávri Formation takes its name from the lake close to the Ruvdnaruvza ridge (Fig. 2B).
The name Goddejávri Calc-Biotite Schist is derived from a lake 5 km to the west of Lake Guolasjávri, (outside the map of Fig. 2B); the unit corresponds to the basal zone of the "Lower Brown Series" of Padget (1955).In the present area, it overlies the volcanics of the Loftani Greenstone Member; elsewhere, it may occur as lenses also within the greenstone (Lindahl et al. 2005).
The rock is not pervasively schistose (Fig. 10A).Thin-sections display syn-postkinematic, idioblastic porphyroblasts of actinolite and rare, small syn-postkinematic garnets in a fine-grained matrix of quartz and brown biotite with strong preferred orientation (G86-32).Zwaan (1988) mapped another belt of "thin-banded meta-greywacke with calcite" overlying the Goddejávri Calc-Biotite Schist, in turn overlain by Ankerlia metagreywackes (Figs.2A &10B).Padget (1955) and Vokes (1957 a. o. p. 53) decribed gradational contacts and intercalations of the schist and the metagreywacke and interpreted them as sedimentary facies changes.However, Padget (1955) described the same change towards the overlying Kåfjord Nappe ("Upper Brown Series") suggesting that the transition into schists may be due to increasing strain towards the upper and lower contacts of the formation (cf. also Faber et al. 2019, p. 122).

Status of the Corrovarre Nappe
In a first report on the Silurian age of the HIC (Andréasson et al. 2003, their Fig. 2), a thrust separates the complex from the underlying KNC.However, in the introduction of the paper, the authors referred to existing maps according to which the HIC is located within the KNC.
This ambiguity invited to the misinterpretation that the HIC should have intruded the entire KNC, even the basement (Kirkland et al. 2007b).Following Zwaan (1988; cf. also http://geo.ngu.no/kart/berggrunn/) and our own reconnaissance, we treat in the present study the HIC as hosted by the CN, but question the affinity of the CN to the KNC.Zwaan (1984;1988) mapped the CN from the type area and 60 km southwards to Halti (Fig. 2A).Although the belt is narrow and locally pinches out, field evidences such as identical lithofacies and structural geology and the spectacular features of intrusive relations in both areas support correlation.Recent U-Pb zircon dating of the dolerites and granites at the CN type locality (Fig. 2A) yielded 606 Ma for the dolerite (Kjøll et al. 2019a) and 610 Ma and 613 Ma (Andréasson et al. 2013;Kjøll et al. 2019a) for the granite.When this is written, the 445-435 Ma old gabbroic dykes and palingenic granite have been reported only from Halti.We infer that additional sampling will eliminate differences of geochronological records from the two areas.If so, there is reason to regard at least some of the anatectic features at Corrovarri as related to Llandoverian rather than Ediacaran magmatism.According to our observations at Corrovarre hill (69,837778°/21,650000°), there are distinct dykes of granite cutting mafic dykes, which in turn cut the dated granite.
Zwaan & van Roermund (1990) interpreted mafic dykes and host rocks at Corrovarre as belonging to the BDS (cf.also Roberts (1990), which is supported by the 606 Ma age of dykes and by detrital zircon provenance studies of their host rocks.However, based on Sm-Nd and Rb-Sr ages (c.580 Ma) of the dykes, they correlated the CN with also the Seiland Igneous Province of the uppermost KNC (Fig. 1), which is accordingly not supported by the 606 Ma age.Gee et al. ( 2017) compared the P-T evolution of the CN as decribed by Zwaan & van Roermund (1990) with that of the wall-rock paragneisses of SIP intrusions on the Øksfjord Peninsula, some 25 km to the north of Corrovarri, as recorded by Elvevold et al. (1994) and Li (2013).However, Kjøll et al. (2019a; cf. also 2020) demonstrated that the metamorphic history of the CN differs significantly from that of the high-grade rocks of the KNC.The andalusitebearing wall-rock of the Ridnitsohkka gabbro dykes reported in our present study indicates ambient pressures <2 kbar and depth <8 km.Against this background and based on inferences presented below, we move the CN from proximal KNC settings to the continent-ocean transition of the SNC.
According to our observations on Halti hill and surroundings, the Ridnitsohkka gabbro dykes do not cut the floor thrust of the CN, nor did we observe any gabbroic dykes and partial melting of wall-rock in the underlying Nabar and Nalganas nappes.However, Lindahl et al. (2005, p. 38-39) reported that gabbroic and related amphibolitic bodies within the Nabar Nappe (Fig. 2A: Oappis and Beabcegealhaldi) seemed to cut the floor thrust of the CN and discussed if the two nappes could have amalgamated during an early event corresponding to the late Cambrian−Early Ordovician collision known from the Seve Nappe Complex.To our present knowledge, there are no published ages of the Oappis and Beabcegealhaldi gabbros.Without such data and a closer description of the field geological setting of the gabbros, we do not address this aspect further.
Based on the 434 Ma age of the gabbro dykes and the proximity (10 km) to the Guolasjávri Formation with early Silurian fossils, Vaasjoki & Sipilä (2001, p. 252) interpreted the thrust below the HIC as an extension of the "Stordalen thrust underlying the Guolas sequence (Binns 1989)".However, the Stordalen thrust, as introduced by Binns & Gayer (1980; Fig. 3) is not located at the base of the fossiliferous limestone but at the base of the lower limestone (Big limestone series of Padget 1955) of the Kvaenangen Group.Their interpretation would nevertheless imply that the HIC is a klippe of the Vaddas Nappe, a view adopted by Andréasson et al. (2003), but modified after additional field studies (Fig. 4 in Andréasson et al. 2013).
North of Vaddas (Fig. 1), a sheet of granite (Raphesvárri Metagranite) occurs, intruded by mafic dykes and with rafts of partially melted, dyke-intruded meta-arkose.Pearson (1971) placed the granite in the core of a recumbent fold (Kvaenangen Fold).Lindahl (1974) interpreted the granite as formed by partial melting of an arkosic variant of the upper quartzite member (Skarddalen Quartzite Mb) of the Ciččenvárri Formation, and the granite was included in the VN on maps.However, Lindahl et al. (2005) reinterpreted the Skarddalen Quartzite as a meta-arkose of the KNC.Corfu et al. (2007)  The bipartite character of the Vaddas Nappe.
In several recent studies, the nappe has been divided into a ʻlower Vaddas nappeʾ and an ʻupper Vaddas nappeʾ (a.o.Faber et al. 2019) strictly speaking implying two different nappes, or regarded as "hybridic" (Corfu et al. 2007).Lindahl et al. (2006, p. 39) inferred that the rocks of the lower part of the Kvaenangen Group ʻcan be much olderʾ than those of the overlying formations.Here, we propose an explanation for the bipartite character of the Vaddas Nappe.
The dyke swarm on Ciččenvárri hill (Fig. 6) differs strongly from the Corrovarri and Ridnitschokkha dyke swarms, but bears striking resemblance to dense mafic dyke swarms of continent-ocean transition affinity underlying the Lower Köli Nappe in Norrbotten (Kathol 1989;Andréasson et al. 1992Andréasson et al. , 2018;;Kjøll 2020).Stølen (1994a,b;1997) mapped these dyke complexes of the SNC northwards into Indre Troms; they are well preserved as far north as Njunis (SE Målselv; M in Fig. 1) and strongly thinned and deformed as far north as Tjuotmer (T in Fig. 1).The Rohkunborri Nappe consists of up of 65-70% mafic dykes.Host rocks include quartzite, fine-grained psammite and arkose, all stained by sulphide impregnation.Marble, dolomite, graphitic and psammitic schists occur on several levels in the succession.In Indre Troms, carbonates predominate wall-rocks suggesting a shallow-marine and carbonate platform realm for the margin.According to Lindahl et al. (2005), the typically rusty quartzite unit with graphitic mica schist (i.e. Riehppejohka Quartzite Mb) is found widely distributed in the northern Troms region .... and towards the south as far as Signaldalen (S in Fig. 1) (cf.also Binns 1967, Zwaan 1988).Farther south it is not found as a continuous marker horizon, but it occurs in the same suite of rocks as far as Målselv in southern Troms (M in Fig. 1).Aware of the need for geochronological and geochemical confirmation of our reinterpretation, we interpret the Ciččenvárri formation at Guolasjávri as a telescoped extension from Indre Troms of the Rohkunborri Nappe of the uppermost Seve Nappe Complex.
The marble at the base of Reisa Nappe Complex at Guolasjvari and elsewhere in Troms has been correlated with the Falkenes marble of the Sørøy succession on Sørøya in Finnmark (a.o.Binns 1989).
An age of deposition of 760-710 Ma (Slagstad et al. 2006) is compatible with the Cryogenian carbonates of the BDS basins, cut by the COT dyke swarms; in Troms carried by the Rohkunborri Nappe of the SNC (Stølen 1994a,b;1997).The garnet-staurolite-kyanite blastesis on foliation planes of Ciččenvárri schists (Fig. 7D-E) indicates intermediate pressure amphibolite-facies metamorphism during thrusting on top of the CN.

The unconformity
The marbles and schists of the Kvaenangen Group at Lake Guolasjávri correlated with the Falkenes and Åjord groups of the Sørøy succession of the KNC were considered as a Cambrian, continuous depositional sequence (Sturt et al 1978).After the discovery of Upper Ordovician fossils in the Guolasjávri Formation (Binns and Gayer (1980;Binns 1989), Ramsay et al. (1985) introduced an unconformity at the base of the formation at Frukosthaugen (F in Fig. 2B), where conglomerates carry quartzite pebbles which "preserve an internal fabric oblique to the pebble shape fabric and the foliation of the matrix and associated rocks, implying derivation from an already metamorphosed source" (p.178).This older metamorphic event was supposed to be the Finnmarkian phase (c.540-490 Ma) of the Caledonian orogeny in Scandinavia as introduced by Sturt et al. (1978).Based on way-up evidence from pillow shapes of the lavas and clast content, Binns & Gayer (1980 p. 53) interpreted the succession at Lake Guolasjavri to be in "normal stratigraphic order", and objected to their interpretation, a.o.by pointing out that the fossiliferous marble at Lake Guolasjávri underlies the conglomerate package, in right-way-up succession.The concept of a Finnmarkian tectonic and geochronological event has not been consolidated by subsequent work (Corfu et al. 2006, p. 451 for summary;Corfu et al. 2011).
The Sørøy Group is no longer considered a continuous sequence, and the Falkenes marble (on Sørøy) was deposited before 710 Ma; moreover, it has been correlated with marbles of inferred Laurentian ancestry in north-central Norway (Slagstad et al. 2006).
Rejecting the Finnmarkian connotation of the unconformity proposed by Ramsay et al. (1985), Lindahl et al. (2005 p. 14, 17, 21) interpreted the boundary between the Kvaenangen and Oksfjord groups as "a major change in sedimentation from shallow-water to a deeper marine environment with contemporaneous volcanic activity".They described the contacts of the Guolasjávri Formation with under-and overlying formations as "well defined, displaying a stratigraphic hiatus of unknown importance and duration, but mostly appears as a nonangular unconformity" (p.14).At Lake Guolasjávri, parageneses of the metavolcanics and metagreywackes of the Oksfjord Group above the fossiliferous marble indicate greenschist to epidote-amphibolite-facies metamorphism.In contrast, the garnet-kyanite-staurolite-bearing assemblage of the schists of the Čiččenvárri Formation indicate intermediate pressure, amphibolite-facies conditions during pervasive deformation of the formation.Thus, the latter tectonometamorphism must have occurred prior to juxtaposition with the overlying Oksfjord Group.The marbles and schists of the Čiččenvárri Formation are cut by mafic dykes.
The overlying conglomerates, marbles and pillow lavas are, according to our observations, not cut by any dykes, nor are they carrying any bands or lenses of metabasite.The mafic dykes within the marble "around the lake" reported by Lindahl et al. (2005 p. 14) appear to be restricted to the Ciččenvárri Formation.
Metamorphic breaks and truncated dyke swarms may also indicate tectonic juxtaposition.
At Guolasjávri, mylonites are absent, but there is an increase in schistosity of the metagreywackes towards the base, and stretching of conglomerate pebbles with the ESE orientation typical of Scandian thrusts.However, the juxtaposition of formations of shallow-and deep-water deposition at Guolasjávri is due to extensive tectonic telescoping of the Oksfjord and upper part of Kvaenangen groups, localised to the soft limestones and schists of the Guolasjávri and Ahkejavri formations.We are inclined to interpret the contact between the Čiččenvárri and Guolasjávri formations as an angular unconformity, reworked during Scandian accretion, but contrary to previous interpretations (Ramsay et al. 1985) on top of the SNC instead of the KNC.

The late Ordovician-early Silurian magmatism and the Halti enigma
Previous models Andréasson et al. (2003)  Clearly, the various models reflect lack of consensus concerning the orogenic situation prior to the late Ordovician-early Silurian magmatism.

Subduction and exhumation of the Baltoscandian margin
The concept of early Palaeozoic subduction of the Baltoscandian margin is based mainly on the eclogite-bearing nappes of the SNC, and most tectonic models describe wholesale subduction of the SNC.However, studies in the eighties demonstrated that the COT nappes in the Sarek, Abisko and Indre Troms mountains had escaped HP-UHP metamorphism (Andréasson 1986;Kathol 1989;Stølen 1989).
The dyke swarms and their host rocks in these nappes display excellently preserved intrusive structures and contact-metamorphic and other parageneses corresponding to depth between 9-16 km Kjøll et al. (2019a,b).Lacking sedimentary screens but carrying ultramafic bodies, the COT of the Kebnekaise Mts.
(Kebne Dyke Complex) likely represents the outermost of the COT nappes, overridden by the Lower Köli Nappe.Andréasson et al. (2018) interpreted a U-Pb age of 487 ± 7 Ma (MSWD = 0.9) of a titanite fabric of a sheared plagiogranitic dyke of the complex as evidence of exhumation following a pressure increase suggested by preserved garnet coronas on plagioclase of deformed dolerites.Based on PTtmodelling and dating, Baird et al. (2022) inferred exhumation at c. 480 Ma of the partially subducted complex.In the host rocks of the CN (this study) and the Sarektjåkkå Nappe (Kjøll et al. 2019a), andalusite rimmed by sillimanite and garnet may indicate a pressure increase at the onset of subduction, before detachment (Fig. 11A).Fassmer et al. (2021) discussed if the upper UHP Seve Nappe ("Tsäkkok lens"), which includes marble and eclogitised metabasites preserving pillow structures (Kullerud et al. 1990) could represent the outermost continental margin.However, the lithofacies and structural history of this unit differs strongly from that of the COT nappes; moreover, pillow basalts may form also in proximal rift basins.In Norrbotten, nappes derived from the COT dyke swarm are overlain by the Lower Köli Nappe (Kulling 1964(Kulling , 1982;;Andréasson et al. 2018, Andréasson 2020), which is interpreted as derived from an early Ordovician to early Silurian island arc outboard the subducting Baltica (Stephens & Gee 1985).
It is likely that a slice of this arc added to the accretionary prism at an early stage of subduction (cf.also Barnes et al. 2019).
Age and tectonic setting of the late Ordovican-early Silurian magmatism The Kågen gabbro hosted by the VN Nappe exposed on the island of Kågen 60 km north of Halti (Fig. 2A) intruded at 439 Ma (Faber et al. 2019).The authors refer to the presence of marble and conglomerate and infer that the c. 100 m-thick tectonostratigrapthic succession represents the ("upper") VN.They recorded a depth of intrusion of 26-34 km and proposed "either a very deep basin or thickened continental crust facilitated by initial collision" (p.140).On the slope below the sample location, the rocks of the extremely thinned lowest formation of the nappe include quartzite, quartz-feldspathic quartzite, mica schist with layers of graphitic schist and metabasite lenses (http://geo.ngu.no/kart/berggrunn/), i. e. the same lithology found better preserved at Steinsvik, 15 km to the southeast (S in Fig. 2A).There, we observed marble, boudins of metabasite and a brown schist, which according to our guide (D.M. Ramsay, pers.comm., July 2003) carries garnet, staurolite and kyanite, i. e. clearly an occurrence of the Čiččenvárri Formation which, in accordance with our reinterpretation, represents the COT of the SNC.Corfu et al. (2006) dated the gabbro and an associated granite of the Honningsvåg Igneous Complex on the island of Magerøya to 438 Ma and peraluminous granites to 435-436 Ma.

Revival of early models
Scandian accretion was a complex series of events including in addition to normal and out-ofsequence thrusting also late extension superimposed on the nappe stack.Differential movements between nappes during exhumation and subsequent accretion resulted, at Halti-Guolasjávri, in a tectonostratigraphy where Silurian mafic dykes are separated from overlying Silurian pillow lavas by a nappe carrying a dolerite dyke swarm of presumably Ediacaran age.From the island Uløya (Fig. 2A),
Halti ---granite sheet above main belt of dolerites

Figure 5 .
Figure 5. A-B.Concordia diagrams showing the U-Pb zircon data for the granitic dykes shown in Figs. 4 A-C.Analyst: Fernando Corfu, University of Oslo.

Figure 6
Figure 6.(A) Dolerite dyke swarm with brown screens of the Riehppejohka Quartzite (Ciččenvárri Fm.) on the northern slope of Ciččenvárri Hill.Estimated length of section is c. 300 m.Arrow: location of close-up figure C. (B) Two levels of the Big Limestone Series (Padget 1955) outcropping on the lower, northern slope of Abmelašvaggi valley, below the dyke swarm.Brown rock between and above the limestones is garnet-mica schist.View towards ENE.Coordinates of the waterfall in lower left corner: N69,41858°/E20.66788°.(C) Close-up of dykes and host rock screens (at arrow in figure A).

Fig. 8 .
Fig. 8. Guolasjávri Formation (Ruvnaruvža ridge).(A)View towards southeast across Lake Guolasjávri.The Rássevággi Schist and Conglomerate Member (K.B.Zwaan,  pers.comm., 1986)  is exposed in the ridge at the shore.The Jiehkkejohka Marble Member is exposed at the bend of the road (in D) and in outcrops in the meadow in the foreground (outside the figure).(B) Polymict conglomerate with carbonaceous matrix and stretched pebbles of quartzite, marble and subordinate metabasite.(C) Pebbles of quartzite in carbonaceous matrix rich in garnet.(D) Schistose grey marble at the fossil locality(Binns & Gayer 1980, Nature 284, 53-55).(E) Marble with sparse pebbles of quartzite overlain by volcanics (V) of the Loftani Greenstone Member.
published an age of 602 ± 5 Ma for the Raphesvárri Metagranite.Since Corfu obtained nearly the same age (610 ± 1) for a palingenetic granite within the CN adjacent (<2 km) in the east(Andréasson et al. 2013), we treat in this study the granite, mafic dykes and meta-arkosic host rock at Raphesvárri as a slice of the CN, formed by local imbrication of the contact between the VN and CN during Scandian nappe emplacement(Gee et al. 2017, Fig. 5).Lindahl et al. (2005 p. 7) reported such tectonic intercalations along the contact between the CN and the Kvaenangen Group in the Lankavarri valley south of Vaddas.
interpreted the 445-435 Ma ages of dykes in the CN at Halti as evidence of short-lived marginal-basin magmatism related to E-dipping subduction caused by a flip in subduction polarity induced by the docking of continents.The model gained meagre support; the main objections being the absence of the Halti dykes within the KNC, and the lack of time for a change of polarity before westward Scandian subduction.Kirkland et al. (2005 p. 1000, 2007b, Fig. 25)  proposed that the formation of the 438 Ma age old granite in the VN equivalent to the Hellefjord Schist on the Porsanger peninsula was "located proximal to Laurentia and within a rifting back-arc basin setting associated with slab rollback").WhileCorfu et al. (2006) discussed if the Honningsvåg Igneous Complex could represent an event of ridge subduction at a ridge-trench intersection located close to the Laurentian margin,Corfu et al. (2011 p. 436)  preferred "an arc or back-arc setting associated with extension or transtensional tectonics".Slagstad & Kirkland (2018) applied slab rollback, but now after docking of the continents.Their model is based on the concept that the Baltoscandian margin remained passive until Early Silurian docking of the continents, thus neglecting the influence on the geothermal profile and mantle heat flow of the 50 million years of subduction, HP-UHP metamorphism and multistage exhumation of the margin documented from Seve eclogite provinces.As a consequence of the proposal that "ca.438-434 Ma magmatic rocks only exist in the units above the suture" (p.7), the authors place the (Scandian) suture between Laurentia and Baltica at the base of the CN, at variance with recent geochronological results supporting affinity of the CN to the Baltoscandian Dyke Swarm (cf.above under Baltoscandian rift basins and dyke swarms).Thermomechanical numerical models of docking continents (e.g.,Sizova et al. 2019) demonstrate how the upper crust of the subducting plate decouples and protrudes into the overriding plate, resulting in an erratic suture, but also implying that not only the upper plate, but also the lower one will be exposed to magmatism generated by slab rollback or breakoff.Thus, one would expect to find evidence of the late Ordovician-early Silurian magmatism also in the SNC, which is not the case, with the possible exception of the HIC proposed in the present study.As treated above, the geochemical signature of the magmatism is of back-arc basin affinity and different from orogen-scale synexhumation magmatism triggered by decompression hydration and partial melting of accreted continental crust.Ziemniak et al. (2019) applied slab rollback but occurring prior to docking of the continental crusts.Their model takes into account the subduction-exhumation event prior to the magmatism, adding the eclogite-bearing Tromsø Nappe (TN in Fig.1) to the subducting slab.Considered of Laurentian affinity, the Tromsø Nappe has striking similarities to eclogite-bearing Seve nappes in Norrbotten, and out-of-sequence thrusting of the SNC, as proposed by Janak et al. (2012) is preferred here.Faber et al. (2019) referred to an anticlockwise PT path obtained from the Reisa Nappe Complex models featuring the VN (Reisa Nappe Complex) in back-arc position above a subducted SNC, which was separated from Baltica by an ocean.

Figure
Figure11Adescribes the subduction of the Baltoscandian margin and an island arc(Stephens & Gee 1985;Dallmeyer & Gee 1986;Grimmer & Greiling 2012); burial of the lower continental crust and partial decoupling and imbrication of the upper continental crust, mainly composed of rift basins.The leading edge of the margin (Seve dyke swarms of the COT and the carbonate platform) is decoupled during initial subduction.An accretionary prism and a foreland basin developed(Gee 2020, Fig. 23.5;Greiling & Garfunkel 2007).Slices of the imbricated margin were subducted to different depths and metamorphosed at varying high to ultrahigh pressures; exhumation of some slices began.

Figure
Figure11Bis essentially the scenario proposed byDunning & Pedersen (1988, see inset figure), with the early ophiolites and the Taconic allochthon accreted to Laurentia, but as different from the inset figure situation, Baltica has arrived.Stephens (2020 p. 570 and Fig. 22:14a)  called attention to faunal indications(Cocks & Fortey 1990) that, by the end of the Ordovician, the "continents of Baltica (Seve Nappe Complex and forthcoming lower thrust sheets) and Laurentia (forthcoming uppermost thrust sheets), and the intermediate volcanosedimentary successions deposited earlier in an oceanic setting (forthcoming Köli Nappe Complex), had attained a contiguous spatial relationship".Pedersen   et al. (1988, inset figure)  included an arc (with question mark) for ʻcontinental margin magmatismʾ.The Váddášgáissát gabbro and pillow lavas and the Sulitjelma gabbro have geohemical similarities with Solund-Stavfjord magmatism, for which the spreading basin was constructed.The peraluminous granites and adamellite of the Honningsvåg Igneous Complex of the Magerøy Nappe could represent such an arc.As an explanation of the enigmatic magmatism in continental sandstones at Halti, we speculate that this magma rose from mantle which filled the space created by decoupling and exhumation of UPH slices.A similar development has been described from slab break-off with uprise of asthenospheric mantle, a popular(Davies & von Blankenburg 1995; cf.however Garzanti et al.   2018;Niu 2017) hypothesis for magmatism and heating related to exhumation of UHP crust.However, based on numerical modelling,Yamato et al. (2008, p. 73; cf. also Sizova et al. 2019 andNiu 2017) concluded that syn-convergent exhumation at the rear of the accretionary wedge is a transient process (~10 Myr) largely controlled by buoyancy forces in the depth interval of 100-35 km, and by erosion at shallower depths, without significant impact from slab break-off"….(which)…"isnot needed for exhumation and does not impact on the exhumation rates".With regard to the magmatism, the depleted mantle signature of the HIC indicates that deep mantle was not involved.

Faber
et al. (2019)  reported a titanite age of c. 432 Ma and a pressure of 12-13 kbar from a shear zone in migmatitic gneiss of the CN (http://geo.ngu.no/kart/berggrunn/)located c. 30 m below the base of the Vaddas Nappe, which is c. 100-150 m thick here.As concluded by these authors, the anti-clockwise PT-path is consistent with stacking of a nappe complex (i.e. Reisa Nappe Complex).However, a depth of 43-46 km almost corresponds to eclogite depth, which has not been recorded so far from Scandian thrusting of nappes equivalent to the Lower Köli Nappe elsewhere in Troms and Norrbotten.On the other hand, the rapid thinning and burial of the nappes westwards from Guolasjávri to Uløya, a distance of 70 km, resembles the development along the Jämtland -Western Gneiss Region traverse.ConclusionsThe CN is reinterpreted as a slice of the Baltoscandian margin, decoupled and accreted with other continent-ocean transition nappes of the Seve Nappe Complex at an early stage of subduction of the margin.The overlying "lower part "of the Vaddas Nappe consists of the Čiččenvárri Formation which, in its type area, is composed of a very dense dolerite dyke swarm with screens of quartzite, marble and subordinate black schist.Aware of the need for geochronological and geochemical confirmation, we correlate this unit with the Rohkunborri Nappe (SNC) of Indre Troms, considered derived from the outermost margin and carbonate platform.It accreted to the CN at intermediate-P conditions.The contact with the overlying telescoped succession of conglomerates, marbles, volcanics and turbidites of the Vaddas Nappe is tentatively interpreted as an unconformity, tectonised and recrystallised during greenschist to lower amphibolite-facies, Scandian accretion.To the classical scenario of back-arc spreading outboard Laurentia during closure of the Iapetus Ocean, we add the arrival of Baltica, composed of a prism of decoupled slices of the COT and exhumed or exhuming HP-UHP nappes.The spreading centre of the basin provides a plausible setting for gabbros and pillow lavas of the Vaddas Nappe, and the shores for a Seve-Köli unconformity.Less dramatic than our first interpretation (subduction flip) of the magmatectonic setting of the Halti Igneous Complex, the model proposed here depends strongly on the results of future PTt studies, in order to relate the proposed shallow syn-exhumation magmatism to the metamorphic evolution of the HP-UHP nappes of the SNC, and to understand the order of subsequent stacking of the nappes.

Table 6 )
Pedersen et al. (1991)the Váddášgáissát gabbro with dykes and its amphibolitic variety and the Loftani pillow lavas.Ternary relations of incompatible elements indicate E-MORB and back-arc tholeiite affinities and in multielement content diagrams (primitive mantle normalised), the lines of the VN rocks run above N-MORB, on the levels of E-MORB and back-arc tholeiites.The chemistry of the cumulate sequence of the HIC indicates mildly depleted mantle source (initial ɛ Nd 5.7; Matsson 1994).Sipilä(1991)reported LREE enrichment of the Ridnitsohkka gabbro dykes.The depth of the andalusite at the Halti gabbro dyke (G86-94) indicates c. 2 kbar if T = 600° (less if higher T) i.e., <c.7.5 km.Of other coeval complexes in the Arctic Caledonides, chemical data are (to our present knowledge) available for only the Sulitjelma Gabbro, which according toPedersen et al. (1991)has E-MORB affinity and was interpreted as back-arc spreading magmatism.