Exhumation of continental margin rocks from mantle depths to orogenic foreland: example from the Seve Nappe Complex of the central Scandinavian Caledonides

Abstract

The diamond-bearing Seve Nappe Complex (SNC) in the Scandinavian Caledonides records subduction of continental margin rocks to (ultra)high-pressure conditions at mantle depths, and exhumation thereafter from beneath the hinterland to the Earth’s surface in the foreland. Structural data of the Upper, Middle, and Lower SNC in central Jämtland, Sweden demonstrate a triclinic bulk deformation during the exhumation of the still ductile SNC from crustal levels where migmatites formed. ⁴⁰Ar/³⁹Ar data from the Upper SNC constrain the timing of cooling through 450‒300 °C to be ~ 418 to 416 Ma. In combination with a review of the published pressure–temperature–time data and regional geology of the central Scandinavian Caledonides, four stages of exhumation of the central Jämtland SNC are summarized: (1) buoyancy-driven exhumation during ~ 455 to 433 Ma from ultrahigh-pressure depths to granulite-facies depths triggered by tectonic under-pressure; (2) tectonic exhumation in ~ 433 to 418 Ma at lower- to mid-crustal levels resulted from accretion of the Lower Köli Nappes onto Baltica; (3) eduction of the Western Gneiss Region lithosphere and piggyback transport of the SNC in ~ 418 to 375 Ma from hinterland to foreland, coupled with extensional faulting at mid- to upper-crustal levels; and (4) gravitational collapse- and erosion-driven exhumation following the end of the Scandian Orogeny at ~ 375 Ma. This multi-stage exhumation transported the SNC for > 100 km vertically from mantle depths to the Earth’s surface and > 350 km horizontally from the Caledonian hinterland to the foreland. This contribution provides a typical example of the complex exhumation of deeply subducted continental rocks in orogenic belts.

Introduction

Recognition of coesite and diamond inclusions in garnet, zircon, and/or clinopyroxene (Chopin 1984; Smith 1984; Sobolev and Shatsky 1990) in orogenic belts indicates subduction of crustal rocks to ultrahigh-pressure (UHP) conditions. Exhumation of UHP rocks from depths of > 100 km to the Earth’s surface demonstrates an extraordinary tectonic process, but is not yet well understood (e.g., Guillot et al. 2009; Hacker and Gerya 2013; Warren 2013). The Seve Nappe Complex (SNC) in the Middle Allochthon of the Scandinavian Caledonides (Fig. 1) is one of the well-known crustal terranes subducted to high-pressure (HP) and UHP conditions. Regional occurrences of eclogites, garnet pyroxenite, garnet peridotites (e.g., Brueckner et al. 2004; Janák et al. 2013; Majka et al. 2014a; Gilio et al. 2015; Klonowska et al. 2016; Bukała et al. 2018), and metamorphic microdiamonds (Majka et al. 2014b; Klonowska et al. 2017; Janák et al. 2017) in the SNC from southern Norrbotten to central Jämtland, Sweden (Fig. 1A) suggests a large HP/UHP metamorphic belt (> 400 km north–south) along the orogenic foreland. The SNC is considered to be derived from the Baltoscandian continental outer margin and continent-ocean transition zone, based on its metasedimentary and mafic magma-rich protoliths (Andréasson 1994; Andréasson et al. 1998; Jakob et al. 2019), the latter of early Ediacaran (ca. 600 Ma) age (Svenningsen 2001; Gee et al. 2013). Considering its current position in the Caledonian foreland above the low-grade sedimentary rocks of the Baltoscandian foreland, the SNC documents a prolonged history from the extension of the Baltica continental margin, through deep subduction beneath Cambrian to earliest Ordovician oceanic crust, to exhumation and accretion along the Baltoscandian margin, followed by transport from the hinterland to the foreland (e.g., Gee 1978; Gee et al. 2020).

Fig. 1

Simplified Geological maps of the central Scandinavian Caledonides (A, B) (Solli and Nordgulen 2008) and central Jämtland (C) showing the study areas. The contours of ⁴⁰Ar/³⁹Ar age are from Walsh et al. (2013). Other muscovite ⁴⁰Ar/³⁹Ar ages are from Dallmeyer et al. (1985), Dallmeyer (1990), and Hacker and Gans (2005). Regional stretching lineations are from Seranne (1992), Hacker and Gans (2005), and Hacker et al. (2010). N Norrbotten, VJ southern Västerbotten and northern Jämtland, J central Jämtland, COSC-1 Collisional Orogeny in the Scandinavian Caledonides scientific drilling hole (Gee et al. 2010)

Abundant pressure–temperature–time (P–T–t) data became available recently for the SNC, related to Ordovician subduction during Iapetus closure, with conditions of up to 2.3‒4.2 GPa in ~ 483 to 455 Ma (Stephens and Van Roermund 1984; Andréasson et al. 1985; van Roermund 1985, 1989; Albrecht 2000; Brueckner et al. 2004; Brueckner and van Roermund 2007; Janák et al. 2013, 2017; Klonowska et al. 2014, 2016, 2017; Majka et al. 2014b; Andersson 2015; Fassmer et al. 2017; Bukała et al. 2018; Walczak et al. 2019, 2022). In central Jämtland, the HP to UHP metamorphism was followed by a high-temperature (850‒860 °C), lower-pressure event resulting in migmatization at ~ 445 to 435 Ma (Gromet et al. 1996; Majka et al. 2012; Ladenberger et al. 2014; Klonowska et al. 2017; Walczak et al. 2019, 2022). These data provide important constraints on the exhumation of the SNC from HP/UHP to migmatization conditions. The exhumation thereafter through crustal levels, including eastward transport of the allochthons for several hundred kilometers onto the Baltoscandian platform, however, remains elusive.

This contribution reports new structural and ⁴⁰Ar/³⁹Ar data associated with the exhumation of the central Jämtland SNC following the ~ 445 to 435 Ma migmatization event, and synthesizes literature data providing orogenic scale constraints on the exhumation of the SNC. The goals are to provide a comprehensive view of the exhumation of the SNC in the central Scandinavian Caledonides (Fig. 1B) from mantle depths beneath the Caledonian hinterland to the Earth’s surface above the foreland, and to provide insights into the general processes of exhumation of deeply subducted continental rocks in orogenic belts.

Tectonostratigraphy of the central Scandinavian Caledonides

The central Scandinavian Caledonides is characterized by large thrust sheets of continental or oceanic origins, all emplaced eastwards onto the Baltoscandian platform (Gee 1975, 1978) of the Baltic Shield that largely consists of the Paleoproterozoic Svecokarelian orogen and Late Paleoproterozoic Transscandinavian Igneous Belt (e.g., Gorbatschev 1985). Four groups of nappes have been defined (Gee et al. 1985, with subsequent minor modification), the Lower, Middle, Upper, and Uppermost allochthons. The Lower has been inferred to be derived from the Baltoscandian platform and inner margin, the Middle from the outer margin and continent-ocean transition zone, the Upper from the Iapetus Ocean, and the Uppermost from the Laurentian continental margin. Descriptions of these allochthons and interpretations of their origins have been widely discussed, with numerous contributions during the last forty years, from Gee and Sturt (1985) to those in Corfu et al. (2014) and the subsequent discovery of widespread UHP. In the study area of the central Scandinavian Caledonides (Fig. 1B, Gee 1975), the Uppermost Allochthon (Roberts et al. 2007 and references therein), representing Laurentian margin, is not preserved. The Upper Allochthon (i.e., the Köli Nappes) is mostly composed of greenschist-facies assemblages (Beckholmen 1982; Sjöström 1984) including Early Ordovician (and perhaps late Cambrian) oceanic crust, for examples the Handöl (Bergman 1993) and Raudfjellet (Nilsson et al. 2005) ophiolites, island-arc/forearc/backarc lithologies, and deep-water sediments (Gee et al. 2014). Structurally below these Iapetus-related nappes is the SNC (included in the Middle Allochthon, Gee et al. 2008), representing the outer and outmost magma-rich margin of western Baltica (Andréasson 1994; Gee et al. 2013; Andersen et al. 2022). The underlying Jämtland Supergroup of the Lower Allochthon comprises Proterozoic basement-derived thrust sheets and overlying Neoproterozoic rifted margin, Ediacaran‒Cambrian passive margin, and Ordovician‒Silurian foreland basin successions (Gee 1975). These allochthons are separated by a major décollement from footwall Cambrian alum shale on the top of the underlying autochthon, the latter dominated by Precambrian crystalline basement rocks commonly with overlying thin quartzites (Gee et al. 2014). In western Sweden, the metamorphic grade of the allochthons increases upwards from low greenschist-facies in the Lower Allochthon to amphibolite-, granulite-, and eclogite-facies in the top of the Middle Allochthon (Sjöström 1984). The latter contrasts markedly with the greenschist-facies and, locally lowest amphibolite-facies Köli Nappes of the Upper Allochthon (Bergman 1993). The Scandinavian nappe pile is up to several kilometers thick and vast in areal distribution (up to 100 s km wide and > 1000 km long) with a general decrease in thickness towards the hinterland in western Norway, particularly notable in the Middle Allochthon (Gee et al. 2010, 2013).

The Middle Allochthon, the focus of this study, was assembled by thrusting of nappes with distinctive protoliths and metamorphic grades. From top to bottom, the high-grade SNC overlies the Särv Nappes that consist of upper greenschist-facies metamorphosed thick feldspathic sandstones and subordinate carbonates and tillites, all intruded by rift-related dolerite dykes (Strömberg 1961; Arnbom 1980; Be’eri-Shlevi et al. 2011; Andersen et al. 2012) of Ediacaran age (ca. 600 Ma, Kjøll et al. 2019). The Särv Nappes were thrust as a mainly rigid duplex (Gilotti and Kumpulainen 1986) along a basal mylonite zone onto a basement-derived allochthon, the Tännäs Auguen Gneiss Nappe (Röshoff 1978); farther north this footwall comprises feldspathic sandstones and minor conglomerates of the Offerdal Nappe (Gee et al. 2014). The initial stacking of these Middle allochthons may have occurred in the Ordovician, providing the source for the westerly derived, Middle Ordovician turbidites of the foreland basin (Karis in Karis and Strömberg 1998). The Middle allochthons in the Åre and Offerdal synforms of central Jämtland overlie the Llandovery turbidites and conglomerates of the Ӓnge group (Kulling 1933; Gee et al. 2014), which constrains the timing of Scandian emplacement of the overlying allochthons to be post ca. 433 Ma.

The distance of thrusting of the Middle Allochthon, from hinterland to foreland, was inferred to be over 100 km by Törnebohm (1888). Subsequent studies provided evidence of much greater distances of > 350 km (e.g., Gee 1975, 1978; Nystuen et al. 2008; Gee et al. 2017, 2020) or of > 1200 km if the pre-Caledonian margin of Baltica was hyperextended (e.g., Andersen et al. 2012, 2022; Jakob et al. 2019). The major lines of evidence supporting long-distance thrusting of these nappes include: (1) the Archean‒Proterozoic Baltica basement of the autochthon-paraautochthon extends from the Caledonian front in Sweden to the Norwegian coast (Gee et al. 2010); (2) the magnetic anomalies (e.g., Dyrelius 1985) related to the autochthonous basement can likewise be followed from the autochthon to the west coast; (3) the Baltoscandian platform-cover sediments (e.g., Andersson et al. 1985) represented by Ediacaran‒Cambrian quartzites and Cambrian alum shales (organic-rich; high uranium, vanadium, molybdenum, and nickel) crop out from the Caledonian front to the hinterland windows in Tömmerås and Trollheimen; (4) the Särv Nappes dolerite dyke-swarms are not present in the underlying allochthons east of Trondheimsfjord; and there, only in the underlying augen gneisses; (5) subduction-related SNC with felsic intrusions has recorded much higher P–T conditions (e.g., Klonowska et al. 2017) than any underlying nappe in the hinterland to the east of the Western Gneiss Region, where the metamorphism of the Lower allochthons reaches UHP conditions and is of Early Devonian age. Thus, the dyke-intruded thrust sheets in the Middle Allochthon (SNC and Särv Nappes) and overlying Upper and Uppermost allochthons, must have been transported more than 350 km towards the foreland onto the Baltoscandian platform, with initial displacements during Ordovician accretion and subsequent thrusting during Scandian continent–continent collision.

Seve Nappe Complex in central Jämtland, Sweden

Regional framework

The SNC in western Sweden is dominated by psammitic and pelitic schists and gneisses with subordinate marbles, calc-silicate gneisses, amphibolitized dolerites and gabbros, and minor eclogites and peridotites (e.g., Zachrisson 1973; Ladenberger et al. 2014). The similarity in protolith of the Lower SNC and the Särv Nappes from Jämtland to Norrbotten (Fig. 1A), with similar aged mafic dykes and a dominance of 1800‒900 Ma detrital zircons in the psammitic gneisses and schists (Gee et al. 2014), are consistent with the Neoproterozoic rift setting along the Baltica margin (Kjøll et al. 2019). The youngest detrital zircons of 726 ± 20 Ma and 811 ± 23 Ma (2σ) from two metapsammite samples of the SNC, respectively (Kirkland et al. 2011), suggest that some of the protoliths were deposited during the Cryogenian or early Ediacaran (Gee et al. 2013). Likewise, some of the SNC paragneisses share detrital zircon age-spectra with some of the Neoproterozoic rift strata of the Särv Nappes (Be’eri-Shlevin et al. 2011) that were deposited during the Neoproterozoic rifting (Kumpulainen 1980). Therefore, the depositional age of the metasedimentary protoliths of the SNC is inferred to be broadly coeval with the Neoproterozoic continent rifting that led to the opening of the Iapetus Ocean (e.g., Andersen et al. 2012; Jakob et al. 2019; Gee et al. 2020).

HP/UHP metamorphism is recorded in the Lower and Middle SNC at several locations along the orogen. In southern Norrbotten (Fig. 1A), HP (1.5–3.1 GPa, 610–780 °C) eclogite, garnet pyroxenite, and associated paragneiss of the Vaimok Lens (equivalent of the Lower SNC) (Stephens and Van Roermund 1984; Andréasson et al. 1985; van Roermund 1989; Albrecht 2000; Bukała et al. 2018) yielded a monazite rim Th–U–Pb age of 498 ± 10 Ma (Barnes et al. 2019), titanite U–Pb ages of ca. 500–475 Ma (Essex et al. 1997), and zircon U–Pb ages of 482 ± 1 Ma (Root and Corfu 2012) and ca. 480–475 Ma (Barnes et al. 2019). About 300 km to the south in southern Västerbotten and northern Jämtland (Fig. 1A), HP/UHP (2.3–4.0 GPa, 750–960 °C) eclogite, garnet pyroxenite, garnet peridotite, and microdiamond are documented in the Middle and Lower SNC (van Roermund 1985; 1989; Brueckner et al. 2004; Janák et al. 2013, 2017; Majka et al. 2014a; Klonowska et al. 2016). The timing of HP/UHP metamorphism in this area is constrained to be ca. 460 Ma based on a weighted mean of mineral Sm–Nd isochron ages of 458 ± 5 Ma (Brueckner and van Roermund 2007), zircon rim U–Pb ages of ca. 460 Ma (Andersson 2015) and 459 ± 3 Ma (Fassmer et al. 2017), and a garnet-whole rock Lu–Hf age of 459 ± 1 Ma (Fassmer et al. 2017). About 250 km further to the south in central Jämtland (Fig. 1A), UHP metamorphism (4.1–4.2 GPa, 830–840 °C) is reported in the Middle SNC from kyanite–garnet migmatitic paragneiss of Åreskutan (Fig. 1B), which is corroborated by occurrences of metamorphic microdiamond at Åreskutan and Tväråklumparna (Fig. 1C) (Majka et al. 2014b; Klonowska et al. 2014, 2017). The timing of UHP in central Jämtland is interpreted to be 483 ± 4 Ma (Walczak et al. 2022) on the basis of U–Pb data from metamorphic zircon overgrowths in the diamond-bearing gneiss at Tväråklumparna (Majka et al. 2014b). The HP condition was likely maintained to 455 ± 11 Ma based on a monazite U-Th-Pb age at Åreskutan (Majka et al. 2012). Migmatization following the HP/UHP metamorphism occurred widely in the Middle SNC and is characterized by granulite-facies assemblages in rocks of pelitic composition (Gee et al. 2013). This high-temperature (850–860 °C) melting event has been constrained to be ca. 445–435 Ma by zircon and monazite U–Pb ages (Gromet et al. 1996; Ladenberger et al. 2014; Klonowska et al. 2017; Walczak et al. 2019, 2022).

The formation of HP/UHP parageneses in the SNC is considered to result from subduction of the SNC beneath the Iapetus oceanic lithosphere that is partially preserved in the Köli Nappes of the Upper Allochthon (e.g., Majka et al. 2014b). The plutonic and volcanic island-arc assemblages of the Upper, Middle, and Lower Köli Nappes range in age from late Cambrian to Late Ordovician, which suggests that subduction in the Iapetus Ocean was active throughout the Ordovician (Gee 2015). In the Upper Köli Nappes, late Cambrian to Early Ordovician ophiolites are overlain by Early to Middle Ordovician limestones with fossils of North American affinity (Bruton et al. 1985), which indicates that these oceanic crust fragments were obducted before or during the Early Ordovician, and that the obduction of this oceanic crust was proximal to the Laurentian margin. Above the obducted Lower Köli oceanic crust (e.g., the Vågåmo ophiolite, Sturt et al. 1991), detrital serpentinites were deposited and included a fauna of Early–Middle Ordovician age with mixed Baltica-North American affinities (Harper et al. 2008). Above the detrital serpentinites in the succession, Late Ordovician quartzites contain Baltica-derived detrital zircons (Gee et al. 2014). Therefore, the Lower Köli ophiolite was obducted proximal to the Baltic margin. Following emplacement of ophiolites and/or plutons in the Upper, Middle, and Lower Köli, turbidites, shales, conglomerates, sandstones, limestones, and/or volcaniclastics were deposited and some of them are as young as early Silurian in the Lower and Middle Köli (Gee 2015). This suggests that the Middle Köli was not yet stacked on the Lower Köli and the latter was not yet emplaced onto Baltica until early Silurian. Since the Lower Köli was proximal to the Baltic margin in the Late Ordovician (Gee et al. 2014), the formation of HP/UHP parageneses in the SNC was probably due to subduction of the SNC beneath an oceanic terrane represented by the Lower Köli Nappes. This interpretation is consistent with the condition that the Handöl ophiolite in Sweden (Bergman 1993) and the Raudfjellet ophiolite in Norway (Nilsson et al. 2005) at the base of the Lower Köli Nappes are directly underlain by the high-grade SNC.

In central Jämtland, the SNC is subdivided into three units, i.e., Upper, Middle, and Lower SNC. The Upper SNC is represented by highly strained kyanite-bearing garnet-mica schists and amphibolites (Sjöström 1984; Bergman 1992). The Middle SNC (e.g., Åreskutan Nappe) is characterized by granulite-facies migmatites and migmatitic gneisses, kyanite- and sillimanite-bearing paragneisses, leucogranite, and metamorphosed mafic dykes (e.g., Arnbom 1980; Ladenberger et al. 2014). The Lower SNC is largely composed of amphibolite-facies metapsammites, mica schist, calc-silicate gneisses, marbles, amphibolites, and locally metadolerites and ultramafites, which are locally overprinted by greenschist-facies retrogression (e.g., Gee et al. 2014; Klonowska et al. 2017).

Lithologies of the Seve Nappe Complex near Lake Ånn

Detailed mapping was conducted by the current study in two areas (Fig. 1C) to the south of Lake Ånn in central Jämtland between the diamond-bearing outcrops at Tväråklumparna and Åreskutan. The western area (Fig. 2A) was chosen for exposures of the Upper and Middle SNC, and the eastern area (Fig. 2B) for the Middle and Lower SNC.

Fig. 2

Geologic map of the western (A) and eastern study area (B) with representative structural data. See map locations in Fig. 1C and detailed maps in Figure S1

In general, the Upper and Middle SNC are compositionally similar and continuous; the major difference lies in the degree of strain accumulated, i.e., distinctively higher strain in the Upper SNC indicated by more continuous foliation and finer-grained porphyroclasts and/or porphyroblasts. The Upper SNC is composed of protomylonitic‒mylonitic (kyanite-)sillimanite-garnet-biotite quartzofeldspathic gneiss, garnet amphibolite, calc-silicate gneiss, and a mylonite‒ultramylonite roof shear zone (80‒120 m thick) of (kyanite-)garnet-muscovite-biotite quartzofeldspathic gneiss/schist immediately beneath the Köli Nappes (Fig. 2A). Synkinematic muscovite flakes (up to 3 cm) characteristically define the foliation of the roof shear zone. The Middle SNC comprises (kyanite-) sillimanite-garnet-biotite quartzofeldspathic gneiss and migmatite along with garnet amphibolite and calc-silicate gneiss (Fig. 2). Rocks at the base of the Middle SNC are deformed to protomylonites and mylonites above a thrust (Fig. 2B). Immediately underlying the basal thrust of the Middle SNC, poorly exposed rocks of the Lower SNC (amphibolite, muscovite-metapsammite, and calc-silicate gneiss) exhibit mylonite‒ultramylonite fabrics (Fig. 2B). Phyllite and schist also occur in the Lower SNC further east according to regional geologic maps (e.g., Sjöstrand 1999).

The migmatite of the Middle SNC features leucosomes containing coarse-grained garnet in a quartzofeldspathic matrix and paleosomes of mostly biotite, finer-grained garnet, and sillimanite (Fig. 3A). Some sillimanite crystals are > 7 cm long. Two phases of garnet growth are recognized in the leucosome; dark-rose garnets (grt I) are overgrown by a much more common, light-rose phase (grt II). The two generations of garnet form core-and-mantle, clasts-in-matrix, or complete replacement structures (Fig. 3B). Large (~ 12 by 2 cm) kyanite grains showing stable contacts to the dark-rose garnets (Fig. 3C,D), kyanite overgrown by sillimanite (Fig. 3E), and radial fractures in garnet around quartz inclusions suggesting former presence of coesite (Fig. 3F) are collectively consistent with the interpretation of (U)HP metamorphism overprinted by a high-temperature event in the SNC. The growth of sillimanite due to the high-temperature event was polyphase, which is indicated by sillimanite of different generations in migmatite of the Middle SNC (Fig. 3G).

Fig. 3

Representative lithologies of the Seve Nappe Complex. A Migmatite with leucosome containing coarse-grained garnet and sillimanite. B Two generations of rose garnet (grt-I: dark and older; grt-II: light and younger). C Large kyanite (a closer look in D) in equilibrium with the dark-rose garnets. E Thin-section of kyanite-sillimanite-bearing gneiss. F Microphotograph showing quartz inclusions in garnet surrounded by radial fractures suggesting breakdown of former coesites. G Microfolds defined by sillimanite and biotite in the paleosome of a migmatite. Note that the sillimanite outlined by the box is strained and show transition to biotite in the same grain whereas the sillimanite at upper left is strain-free and has sharp contact to biotite, which suggests two generations of sillimanite. bt biotite, grt garnet, ky kyanite, ms muscovite, qtz quartz, sil sillimanite, PPL plane polarized light, CPL cross polarized light

Biotite-muscovite pegmatitic dykes and pods are widespread in the mapped SNC. Most show no deformation as indicated by euhedral micas up to 6 cm across, whereas some pegmatites are strained into lens-shaped blocks containing euhedral minerals. Amphibolitized mafic dykes are also common and extensively stretched or variously strained. In short, the Upper, Middle, and Lower SNC in the study area are compositionally consistent with passive margin sequences intruded by mafic dykes that have collectively been overprinted by high-grade metamorphism, partial melting, and pervasive poly-deformation that is described below.

Ductile structures of the Seve Nappe Complex near Lake Ånn

Generations of fabrics

Three generations of ductile fabrics are distinguished in the current study based on crosscutting relationships. For example, blocks of pre-migmatization elliptical and planar fabrics are enclosed in migmatite that in turn is overprinted by ductile shear fabrics. The most distinctive fabric is the alternating mafic and felsic layers in the Middle SNC migmatite, including planar, discordant (Fig. 4A), and folded foliations (in both mesoscale and microscale, Fig. 4B). The mesoscale folds are tight to isoclinal and commonly rootless (Fig. 4B). Within the migmatite, garnet-amphibolite lenses and sporadic blocks of paragneiss paleosome contain discordant foliations and rare kyanite. Development of a partially melted zone in some of the blocks (Fig. 4C) suggests that these predate migmatization and that their fabrics are older than those in the migmatite. Elliptically and randomly orientated foliations of the paragneiss blocks (Fig. 4D) further indicate the existence of earlier fabrics. The migmatitic fabrics are commonly modified by ductile shear (Fig. 4E) resulting in gneissic, schistic, or mylonitic fabrics (Fig. 4F). This modification might have obliterated many original migmatite fabrics formed during the high-temperature event at about 440 Ma in the Middle SNC (e.g., Ladenberger et al. 2014).

Fig. 4

Typical fabrics of the Seve Nappe Complex (SNC). A Planar and discordant migmatitic foliations. B Folded migmatitic foliations in the mesoscale (left) and microscale (right). C Isolated block containing discordant foliations and a partially melted zone surrounded by migmatite. D Isolated blocks of elliptically and randomly orientated foliations enclosed in migmatite. E Ductile shear fabrics in the Middle SNC. F Ultramylonitic fabrics in the Upper SNC. bt biotite, ms muscovite, qtz quartz, sil sillimanite, PPL plane polarized light

S-L fabrics

The foliation and lineation associated with the ductile shear following the migmatization of the SNC are characterized by crystal-plastic deformation indicative of relatively high-temperature conditions. Elevated temperatures are demonstrated by the occurrence of gneissic fabrics, microstructures in quartz and feldspar, and mineralogy of sheared rocks.

In the Upper SNC, dynamically recrystallized quartz shows lobate and amoeboid grain boundaries (Fig. 5A), typical of grain boundary migration (Passchier and Trouw 2005). Quartz ribbons enclosing garnets and defining the mylonitic foliation have strain-free grains showing polygonal and straight boundaries with interfacial angles of ~ 120° (Fig. 5B). The latter suggests grain boundary area reduction for decreasing energy during and/or following high-temperature deformation (Passchier and Trouw, 2005). In the shear zones within the Middle SNC, quartz is generally recrystallized by grain boundary migration (Fig. 5C) and some feldspar twins are bended (Fig. 5D). In the shear zones of the Upper and Middle SNC, stable (i.e., no retrogression) amphibole and garnet occur along the ductile lineation and new sillimanite was crystallized along the mylonitic foliation. These lines of evidence collectively suggest high-temperature conditions during ductile shear of the SNC after the migmatization event.

Fig. 5

Microphotographs showing microstructures of quartz and feldspar. A Lobate and amoeboid grain boundaries of quartz in the Upper Seve Nappe Complex (SNC). B Quartz ribbons showing strain-free grains with polygonal and straight boundaries of ~ 120° interfacial angles in the Upper SNC. C Quartz recrystallized by grain boundary migration in the Middle SNC. D Bended feldspar twins in the Middle SNC. am = amphibole, bt biotite, fsp feldspar, grt garnet, ms muscovite, qtz quartz, CPL cross polarized light

The foliation is generally defined by (re)crystallized mica and quartz in schist and alternating mafic and felsic layers in migmatite and gneiss. The lineation is commonly characterized by ductile ridge-in-groove lineation (Lin et al. 2007) defined by quartz, feldspar, biotite, and/or amphibole along with preferred orientation of newly grown muscovite, quartz, and/or sillimanite.

In general, from the top of the Upper SNC to the interior of the Middle SNC, foliations get shallower and are less consistently oriented (Fig. 6a‒6c, stereoplots i). In the roof shear zone of the Upper SNC, the mylonitic foliations strike E-W and dip moderately to the north with limited variation along and across the length of the unit (Fig. 6a, stereoplot i). In the remaining part of the mapped Upper SNC, the foliations (gneissosity overprinted by mylonitic foliations) strike mostly E–W, locally NW–SE and NE–SW, with a moderate to shallow dip to the north (Fig. 6b, stereoplot i). In the Middle SNC, migmatitic layers and gneissosity are overprinted at several structural levels by mylonitic foliations that are subhorizontal and related to shear zones. Outside of these shear zones, steep migmatitic foliations of various attitudes are common (Fig. 6c, d, stereoplots i). Foliations in the Lower SNC are defined by schistosity and gneissosity overprinted by mylonitic foliations, and dip shallowly to the west or WSW (Fig. 6e, stereoplots i). These fabrics are concordant with the mylonitic foliations of the Middle SNC immediately above (Fig. 6d, stereoplots i).

Fig. 6

Structural data of foliations and lineations from various structural levels of the Seve Nappe Complex (SNC), equal-area lower hemisphere projection

The lineations (Fig. 6a‒e, stereoplots ii) are shallowly-plunging in gently folded rocks and generally subhorizontal in planar and isoclinally folded zones. The lineation largely trends between E–W and ESE–WNW.

Kinematics of deformation

The post-migmatization ductile shear of the SNC rocks was strongly partitioned into rheologically different lithologies as shown by distinctive fabrics of adjacent rock units (Fig. 7A). Ductile ridge-in-groove lineations and sheath folds with fold axes parallel to the lineation are commonly well-developed (Fig. 7B,C). The sense of shear is revealed by kinematic indicators appearing on the plane perpendicular to the foliation and parallel to the lineation. Top-to-ESE and top-to-WNW shear both occur but the former is predominant. In the mesoscale, widespread sigma-porphyroclasts and S-C structures in garnet amphibolites (Fig. 7D) and migmatitic gneisses (Fig. 7E), sigma-porphyroclasts of felsic melt pods in schists (Fig. 7F), and pervasive S-C structures in metapsammites (Fig. 7G) all indicate a dominant top-to-ESE shear. Top-to-WNW shear indicators occur at a few localities (Fig. 2A) where they are well-developed as sigma-porphyroclasts, S-C structures, and asymmetric blocks (Fig. 7H, I).

Fig. 7

Outcrop structures on lineation-parallel planes associated with the post-migmatization ductile deformation of the Seve Nappe Complex. A Strain partitioning between calc-silicate gneiss and sillimanite-garnet-biotite quartzofeldspathic gneiss (sil-grt-bt-qfp gneiss). B Well-developed ductile ridge-in-groove lineations trending E–W in sillimanite-garnet-biotite quartzofeldspathic gneiss. C Sheath folds with axes parallel to the lineation. D–G Asymmetric porphyroclasts/blocks, S-C structures, and sigma-porphyroclasts showing top-to-ESE shear (to the right in photographs) on the plane perpendicular to the foliation and parallel to the lineation. H–I Asymmetric blocks, S-C structures, and sigma-porphyroclasts showing top-to-WNW shear (to the left in the photographs) on the plane perpendicular to the foliation and parallel to the lineation

In the microscale, the pervasive top-to-ESE shear is recorded by mica fish, S-C structures, delta-porphyroclasts, sigma-porphyroclasts, and shear bands (Fig. 8A–C). The top-to-WNW shear is documented mostly by S-C structures and asymmetric porphyroclasts (e.g., Fig. 8D,E).

Fig. 8

Thin-section kinematic indicators on the plane perpendicular to the foliation and parallel to the lineation (the right side of each photo points to the east). A–C S-C structures of synkinematic sillimanite, delta-porphyroclast, mica fish, and shear band showing top-to-ESE shear. D–E S-C structures and sigma-porphyroclasts showing top-to-WNW shear and coexistence of top-to-ESE and top-to-WNW shear in the microscale. bt biotite, fsp feldspar, grt garnet, ms muscovite, qtz quartz, sil sillimanite, PPL plane polarized light, CPL cross polarized light

Coexistence of top-to-ESE and top-to-WNW shear has been observed in multiple samples. In a thin-section of sillimanite-garnet-biotite quartzofeldspathic gneiss from the lower part of the Upper SNC, biotite and quartz are arranged in S-C structures with opposite senses of shear at a distance of ~ 2 mm (Fig. 8D). In another thin-section of garnet-muscovite-biotite quartzofeldspathic schist from the roof shear zone of the Upper SNC, biotite and garnet form a winged porphyroclast indicating top-to-WNW shear and, along the adjacent foliations, biotite, sillimanite, and muscovite show a S-C structure of top-to-ESE shear (Fig. 8E).

In short, the kinematic indicators preserved in the SNC show a dominant top-to-ESE shear locally accompanied by a top-to-WNW shear. The fabrics displaying opposite kinematics are concordant without overprinting and/or crosscutting relationships and are broadly similar in metamorphic grade.

Structures that accommodated the mass transport along the shear direction probably include boudins and faults in both mesoscale and microscale along with intergranular and intragranular shear in the microscale. The most common boudins in the SNC are amphibolite blocks formed by flattening and stretching of mafic dykes. An estimate of a minimum amount of stretch of 315% was obtained from a representative boudinaged dyke (Fig. 9A). Foliation-parallel sliding combined with foliation-oblique thrusting along a fault at the bottom of the Middle SNC (Fig. 9B) suggests that foreland-directed faults have contributed to the mass transport.

Fig. 9

A Boudinaged mafic dyke along the lineation providing an estimate of the minimum amount of stretch (about 315%). B Top-to-ESE thrust that is partly oblique and partly parallel to the foliation at the base of the Middle Seve Nappe Complex in the eastern area, Fig. 2B

Structures in lineation-normal plane

Structures exposed in the surface perpendicular to lineation help reveal the three-dimensional structures associated with the ductile shear of the SNC. On this plane, some of the sheath folds are flattened (Fig. 10A) and symmetric blocks of felsic melt show a typical boudinage geometry (Fig. 10B). In the same plane, asymmetric blocks (Fig. 10C) and shear bands in the Middle SNC immediately beneath the Upper SNC in the western area (Fig. 2A) show top-to-NNE shear, whereas S-C structures (Fig. 10D) in the Middle SNC immediately above the Lower SNC in the eastern area (Fig. 2B) demonstrate top-to-SSW shear. Gentle, large (100 s m-scale, Fig. 10E) and tight, small (cm-scale, Fig. 10F) folds are recorded in the lineation-normal plane as well. These folds have steep axial surfaces striking parallel to the trend of the lineation.

Fig. 10

Outcrop structures on lineation-normal planes in the Seve Nappe Complex (SNC), looking WNW. A Flattened sheath folds. B Boudinaged felsic melt with a symmetric geometry. C Asymmetric blocks of quartzofeldspathic gneiss in the Middle SNC showing a top-to-NNE (to the right in the photograph) non-coaxial shear. D S-C structures of quartzofeldspathic gneiss in the bottom of the Middle SNC showing a top-to-SSW (to the left in the photograph) non-coaxial shear. E 100 s m-scale gentle folds, stereoplot showing gneissic foliation (line) and lineation (dot). F Centimeter-scale folds

⁴⁰Ar/³⁹Ar thermochronology

Samples and analytical methods

Muscovite, biotite, and potassium-feldspar from two rocks were analyzed for ⁴⁰Ar/³⁹Ar geochronology. The sample (MCH-1) from the roof shear zone of the SNC is a garnet-muscovite-biotite quartzofeldspathic mylonite containing minor muscovite quartzofeldspathic melt (Fig. 11A-i). The latter occurs as ~ 1 to 25 cm thick synkinematic melt patches and 1‒10 cm thick layers, forming S-C structures and/or sigma-type asymmetric kinematic indicators indicative of top-to-ESE shear. Biotites from the mylonitic gneiss (Fig. 11A-ii) and muscovites from the synkinematic melt (Fig. 11A-iii) were picked for ⁴⁰Ar/³⁹Ar analysis. The sample (MCH-23) from the interior of the Upper SNC is a thin (1‒5 cm), coarse-grained, synkinematic felsic melt that is composed of garnet, tourmaline, muscovite, potassium-feldspar, plagioclase, and quartz along with minor biotite and sillimanite (Fig. 11B-i). Some muscovites were variably strained and some potassium-feldspar was partially replaced along grain boundaries by myrmekite (Fig. 11B-ii, -iii); such grains were avoided during picking.

Fig. 11

Outcrops, thin-sections, and ⁴⁰Ar/³⁹Ar age spectra (1σ errors) of samples MCH-1 and MCH-23. The gray boxes in the age spectra indicate the steps used for calculation of integrated or plateau ages. bt biotite, grt garnet, kfs potassium-feldspar, ms muscovite, qtz quartz, sil sillimanite, tur tourmaline, PPL plane polarized light, CPL cross polarized light

Rocks were crushed, sieved, washed, and hand-picked for pure biotite, muscovite, and feldspar mineral separates before being analyzed at the Geochronology laboratory at the University of Alaska, Fairbanks. Samples and standards were irradiated in the McMaster Nuclear Reactor at McMaster University in Hamilton, Ontario, Canada. The mineral MMhb-1 (Samson and Alexander 1987) with an age of 523.5 Ma (Renne et al. 1994) was used to monitor neutron flux. The monitor minerals and unknown samples were heated in an ultra-high vacuum extraction line using a 6-W argon-ion laser following the technique described in York et al. (1981) and Benowitz et al. (2014). Samples were analyzed in a VG-3600 mass spectrometer at the Geophysical Institute, University of Alaska, Fairbanks. The argon isotopes measured were corrected for system blank and mass discrimination, as well as calcium, potassium and chlorine interference reactions following procedures outlined in McDougall and Harrison (1999). Correction factors for nucleogenic interferences during irradiation were determined from irradiated CaF₂ and K₂SO₄. Mass discrimination was monitored by running calibrated air shots. The mass discrimination during these experiments was 1.3% per mass unit. While doing our experiments, calibration measurements were made on a weekly to monthly basis to check for changes in mass discrimination with no significant variation seen during these intervals. Each of the mica separates has been heated in eight steps and the potassium-feldspar in eleven steps.

A summary of the multi-grain step-heating ⁴⁰Ar/³⁹Ar results and age spectra are given in Fig. 11C, with all ages quoted to the ± 1 sigma level and calculated using the constants of Renne et al. (2010). The spectrum provides a plateau age if three or more consecutive gas fractions represent at least 50% of the total gas released and are within two standard deviations of each other (Mean Square Weighted Deviation ≤ 2.5). Otherwise, an integrated age is calculated for those consecutive steps and is based on weighted average of the data with error at the 95% confidence level. The complete step-heating data are tabulated in Table S1.

Results

The ⁴⁰Ar/³⁹Ar data of the muscovite and biotite separates are characterized by flat age spectra showing no signs of argon loss (Fig. 11C). For the MCH-1 muscovite, 99.3% of ³⁹Ar was released in the last two of the eight steps during the heating experiment. These two steps give concordant ages within 1σ error and yield an integrated age of 418.1 ± 1.7 Ma, statistically indistinguishable from the total gas age of 418.8 ± 1.8 Ma (Fig. 11C-i). The MCH-1 biotite progressively released ³⁹Ar after the first step of the heating experiment. The seven steps (including 99.0% of ³⁹Ar released) give concordant ages within 1σ error and yield a plateau age of 415.7 ± 1.4 Ma, nearly identical with the total gas age of 415.6 ± 1.4 Ma (Fig. 11C-ii). The MCH-23 muscovite also released ³⁹Ar progressively after the second step with concordant ages within 1σ error. The six steps of 99.4% of ³⁹Ar released give a plateau age of 416.7 ± 1.4 Ma, identical to the total gas age of 416.7 ± 1.4 Ma (Fig. 11C-iii).

The potassium-feldspar from sample MCH-23 gave a staircase age-spectrum (Fig. 11C-iv). After the first three steps, a total amount of 98.6% ³⁹Ar was progressively released in eight steps. The ⁴⁰Ar/³⁹Ar ratio steadily increases during heating and the individual steps yield apparent ⁴⁰Ar/³⁹Ar ages ranging from ~ 340 to ~ 400 Ma. Three continuous steps (gray boxes in Fig. 11C-iv) including 23.5% ³⁹Ar released are concordant within 1σ error and yield an integrated age of 390.0 ± 2.0 Ma. The minimum age of ~ 340 Ma is defined by the fourth step including 25.1% ³⁹Ar released.

Discussion

Exhumation of the central Jämtland Seve Nappe Complex following migmatization and prior to extensional collapse

Exhumation of the central Jämtland SNC from (U)HP conditions to crustal levels where migmatites formed could be generally defined by the P–T–t data published by previous works. The extensional collapse of the Scandinavian Caledonides is understood as a part of the Devonian orogen-scale, hinterland-directed normal-faulting that likely started between ~ 408 and ~ 402 Ma in southern Norway (Fossen and Dunlap 1998; Fossen 2010). The new data reported in the current study aim to constrain the exhumation of the central Jämtland SNC between these two periods, i.e., exhumation from crustal levels where migmatites formed to crustal levels where extensional faulting takes over the deformation.

Nature of deformation during the exhumation

The post-migmatization ductile shear fabrics of the Upper and Middle SNC mapped in this study are (1) structurally concordant without overprinting and/or crosscutting relationships, (2) broadly similar in metamorphic grade, (3) related to the foreland-directed transport of the SNC, and (4) are cut by several normal faults or extensional shear zones (Sjöström et al. 1996). Therefore, the ductile shear fabrics of the Upper–Middle SNC are broadly coeval and likely record the bulk deformation path after the migmatization of the SNC and prior to the extensional collapse of the orogen.

The maximum principal strain axis, X, parallel to the X-axis of the finite strain ellipsoid, is revealed by the dominant ductile ridge-in-groove lineations. The lineations generally trend ESE-WNW and plunge shallowly (< 20°) to subhorizontally (Figs. 2A and 6). Structures formed along this direction are asymmetric on the plane perpendicular to the foliation and parallel to the lineation (e.g., sigma- and delta-porphyroclasts, S-C structures, mica fish, shear bands, asymmetric folds, Figs. 7 and 8), which indicates that the general strain component along the X-axis can be described as a non-coaxial shear path. Migmatites and migmatitic gneisses are locally preserved between shear zones in the Middle SNC (e.g., the outcrop in Fig. 12), which suggests that the X-axis strain was partitioned.

Fig. 12

Three-dimensional diagram summarizing the structures and kinematics of the post-migmatization ductile deformation of the Seve Nappe Complex (SNC) in the study area. Note the opposite senses of non-coaxial shear on the lineation-parallel plane and secondary non-coaxial shear on the lineation-normal plane. See Figs. 7 and 8 for examples of structures on the lineation-parallel plane and Fig. 10 for those on the lineation-normal plane. The picture shows one of the internal shear zones in the Middle SNC

The minimum principal strain axis, Z, is constrained to be subvertical on the basis of boudinage structures such as the boudinaged mafic dyke exposed on the plane perpendicular to the foliation and parallel to the lineation (Fig. 9A), as well as the boudinaged quartzofeldspathic layer exposed on the plane perpendicular to the lineation (Fig. 10B). The flattened sheath folds on the lineation-normal plane (Fig. 10A) also support a subvertical shortening along the Z-axis. These structures show that the general strain component along the Z-axis was characteristic of vertical flattening.

The strain component along the intermediate principal strain axis, Y, is revealed by structures exposed on the lineation-normal plane. The asymmetric blocks (Fig. 10C), shear bands, and S-C structures (Fig. 10D) found on this plane show lateral flow of ductile rocks along the foliation, top-to-NNE at the top and top-to-SSW at the bottom of the Middle SNC. This suggests that the general strain component perpendicular to the lineation and parallel to the foliation could be described as a non-coaxial shear path. The various scales of folds on the lineation-normal plane with axial surfaces parallel to the lineation (Fig. 10E,F) might indicate local convergent flow of ductile rocks along ESE-WNW. An alternative interpretation, especially for the 100 s m-scale gentle folds (Fig. 10E), is that the variation of thickness of the underlying rocks caused perturbation of the overlying ductile flow.

In summary, the bulk deformation path of the Upper–Middle SNC in central Jämtland is consistent with a triclinic or general shear (Jiang and Williams 1998; Lin et al. 1998; 2007), which is defined by a dominant non-coaxial shear along ESE-WNW, a subordinate non-coaxial shear along NNE-SSW, and a vertical flattening component (Fig. 12). Because the microstructures of quartz and feldspar (Fig. 5) and formation of stable garnet/amphibole and synkinematic sillimanite (e.g., Fig. 8B) indicate relatively high temperatures during the ductile shear, this deformation path describes the nature of the ductile flow due to the exhumation of the SNC through mid-crustal levels. This is supported by the P–T conditions of 0.8‒1 GPa and ~ 600 to 650 °C for the formation of mylonitic fabrics of the lower SNC in the COSC-1 borehole (Fig. 1B, Giuntoli et al. 2018, 2020). Previous studies have described this deformation as a non-coaxial progressive flow to the east or southeast combined with concurrent vertical ductile thinning (Bergman and Sjöstrom 1997; Bender et al. 2018). The current study supports these conclusions and provides additional constraints on the deformation, which are: (1) the bulk deformation path is triclinic; (2) the dominant top-to-ESE non-coaxial shear along the foliation had various flow rates at different structural levels. The second constraint is based on the kinematically contrasted shear zones of various scales at different structural levels (Fig. 2A), i.e., structurally concordant, top-to-ESE and top-to-WNW shear fabrics from the Upper SNC to the Middle SNC (Figs. 7, 8).

Timing of exhumation and cooling

Stable amphibole/garnet and newly formed sillimanite in mylonite, and relatively high temperature microstructures including grain boundary migration recrystallization of quartz (Fig. 5A,C), grain boundary area reduction of quartz (Fig. 5B), and ductile deformation of plagioclase (Fig. 5D) jointly indicate that the deformation in the shear zones of the Upper and Middle SNC occurred at temperatures (likely > 500 °C, Passchier and Trouw 2005) higher than the argon closure temperature intervals of muscovite (~ 450 to 350 °C), biotite (~ 380 to 300 °C), and potassium-feldspar (~ 350 to 150 °C) in crustal rocks (e.g., McDougall and Harrison 1999; Foster et al. 2010). This interpretation is consistent with the amphibolite-facies metamorphism recorded by parageneses in the mylonitic Upper SNC (Sjöström 1984). The muscovite and biotite ages of sample MCH-1 (418.1 ± 1.7 Ma, 415.7 ± 1.4 Ma, Fig. 11C-i, -ii, respectively) and the muscovite age of sample MCH-23 (416.7 ± 1.4 Ma, Fig. 11C-iii), therefore, are interpreted to reflect the timing of mica cooling in the Upper SNC through temperatures of ~ 450 to 300 °C. The ⁴⁰Ar/³⁹Ar age spectrum of the potassium-feldspar from sample MCH-23 reveal further cooling of the Upper SNC possibly through temperatures of ~ 180 to 150 °C by 340 Ma (the minimum age in Fig. 11C-iv). These cooling data indicate that the argon isotopic systems in micas and potassium-feldspar were not disrupted after 418‒416 Ma and 340 Ma, respectively. This suggests that there was no pervasive deformation or significant thermal event in the SNC following 416 Ma, which is supported by the metamorphic patterns of rocks from either side of the Seve‒Köli boundary indicating no post-metamorphic movements (Bergman 1992).

In conjunction with the Llandovery turbidites and conglomerates of the Ӓnge group underlying the Middle allochthons in central Jämtland (Kulling 1933), the timing of exhumation of the SNC following the migmatization is bracketed between 433 and 418 Ma. This timing is well supported by recent geochronology data from the SNC, including the multi-mineral Rb–Sr isochron ages of mylonitic rocks from outcrops across the SNC in central Jämtland (433–423 Ma) (Bender et al. 2019) and the Lower SNC in the COSC-1 borehole (~ 429 Ma) (Fig. 1B, Glodny et al. 2017), the Th-U–Pb ages (424 ± 6, 423 ± 13 Ma) of fluid-assisted grown monazites in the Middle SNC at Åreskutan (Majka et al. 2012), the U–Pb ages of 439 ± 4 Ma of zircon overgrowths from the diamond-bearing gneiss at Tväråklumparna (Walczak et al. 2022), and the U–Pb ages of synkinematic titanite (429 ± 20 Ma, 417 ± 9 Ma) in the Lower SNC mylonite of the COSC-1 borehole (Giuntoli et al. 2020). Thereafter, the SNC underwent cooling through ~ 450 to 300 °C at about 418‒416 Ma and ~ 180 to 150 °C possibly at about 340 Ma without any pervasive deformation, except for extensional collapse along several normal fault zones (e.g., Sjöström et al. 1996).

Exhumation of the central Jämtland Seve Nappe Complex from mantle depths beneath the hinterland to the Earth’s surface in the foreland

The regional geology and published P–T–t data of the central Jämtland SNC along with the new structure and ⁴⁰Ar/³⁹Ar data of the current study provide insights into a complete view of the exhumation of the SNC. The following discussion aims to constrain the exhumation of the central Jämtland SNC from depths of (U)HP conditions beneath the hinterland to the Earth’s surface above the foreland basin. The total exhumation of the central Jämtland SNC includes > 100 km vertical and > 350 km horizontal displacements. With the hyperextension model of the Baltica margin (e.g., Andersen et al. 2012, 2022; Jakob et al. 2019), the horizontal displacement of the SNC would be > 1200 km. The data defining the exhumation of the SNC are summarized in Fig. 13 and the geologic context of the central Scandinavian Caledonides for each stage of the exhumation is shown in Fig. 14.

Fig. 13

Summary of the P–T–t data and sources defining the four stages exhumation of the Seve Nappe Complex in central Jämtland along a schematic exhumation trajectory from mantle depths beneath the hinterland to the Earth’s surface in the foreland. WGR Western Gneiss Region

Fig. 14

Subduction and exhumation of the Seve Nappe Complex (SNC) in the geologic context of central Scandinavian Caledonides. Data sources: (1) Majka et al. (2014b), (2) Ladenberger et al. (2014), (3) Brueckner et al. (2004), (4) Beckman et al. (2014), (5) Spengler et al. (2009), (6) Young (2017), (7) Gilotti et al. (2008), (8) Augland et al. (2010), (9) Cuthbert et al. (2000), (10) Brueckner et al. (2010), (11) Krogh et al. (2011). WGR Western Gneiss Region

A variety of mechanisms have been proposed for exhuming rocks from mantle and crustal depths, such as subduction channel flow (e.g., Gerya et al. 2002), thrusting of crustal slivers (Chemenda et al. 1995), extrusion (e.g., Thompson et al. 1997; Ring and Glodny 2010), buoyancy (e.g., Warren 2013), eduction (i.e., reversal of subduction, Andersen et al. 1991), piggyback translation (Brueckner and Cuthbert 2013), tectonic under-pressure by delamination, slab rollback, or microplate rotation (e.g., Hacker and Gerya 2013; Majka et al. 2014b), trans-mantle diapirs (e.g., Hacker and Gerya 2013), surface processes, and extensional detachment faulting (e.g., Ma et al. 2019, 2021). Exhumation of HP/UHP rocks can be generally described as a fast process controlled by density contrast followed by a slow process that is generally influenced by tectonic and surface processes (e.g., Yamato et al. 2008). Exhumation of the SNC from (U)HP conditions to crustal levels where migmatites formed was likely rapid such that some of the HP/UHP mineral assemblages can survive from decompressional melting (e.g., Thompson et al. 1997; Yamato et al. 2008) and form the eclogite, garnet pyroxenite, garnet peridotite, and diamond-bearing gneiss currently preserved in the SNC. Prior to the subduction of the central Jämtland SNC (Fig. 14A), its protoliths occurred as passive margin sequences overlying the westernmost Baltica and intruded by mafic dykes. During the Ordovician, the SNC and the underlying continental lithosphere were probably subducted to UHP and HP conditions (Fig. 14B) under an oceanic terrane represented by the Lower Köli Nappes.

The depths from which the central Jämtland SNC was exhumed are constrained by P–T conditions of peak metamorphism (4.1–4.2 GPa and 830–840 °C) based on garnet + phengite + kyanite + rutile assemblages of paragneisses from Åreskutan (Fig. 1B), which was verified by the identification of diamond inclusions in garnet of the same rocks (Klonowska et al. 2017). About 10 km to the west of the study area at Tväråklumparna (Fig. 1C), metamorphic diamonds were also discovered as inclusions in garnet (Majka et al. 2014b). Zircon overgrowth and monazite from these diamond-bearing gneisses have been dated and provide records of UHP and HP metamorphism at 483 ± 4 Ma (Walczak et al. 2022) and 455 ± 11 Ma (Majka et al. 2012), respectively. The SNC of the current study, located between these diamond-discovery sites and exposing the same tectonostratigraphic units, most likely underwent a similar metamorphism P–T–t trajectory. Two features in support for this interpretation are (1) the occurrence of radial fractures in garnet around quartz inclusions (Fig. 3F) suggesting transformation of former coesite, and (2) the dark-rose garnets (grt I in Fig. 3C) in equilibrium with kyanite (Fig. 3C,D) where the former probably correspond to the Ca- and Mg-rich garnets formed at UHP conditions in the Middle SNC of Åreskutan (Klonowska et al. 2017).

First stage exhumation: upward flow from UHP depths due to buoyancy

On the basis of literature data, it is inferred that the first stage exhumation of the central Jämtland SNC was initiated following the HP metamorphism, possibly at or slightly after 455 Ma. The first stage exhumation was likely characterized by a nearly isothermal decompression of the SNC rocks from depths of (U)HP conditions (≥ 100 km, Klonowska et al. 2017) to depths of granulite-facies (≈ 40 km for a granitic composition of the overlying continental crust) in ~ 455 to 433 Ma (Figs. 14, 15C).

The exhumation of the UHP rocks during the decompression was probably dominantly driven by buoyancy, because at temperatures of 800–900 °C, a quartzofeldspathic continental crust with minor mafic/ultramafic rocks, such as the SNC, would be positively buoyant to the mantle (Hacker and Gerya 2013). In general, mechanisms that could exhume rocks for > 100 km vertically would be likely controlled by (1) positive buoyancy of the exhuming rocks, (2) availability of space for storing the exhumed rocks at shallower depths and for accommodating the increased volume of the partially melted exhuming rocks during isothermal decompression, and (3) local tectonic forces to trigger the upward flow of the exhuming rocks (e.g., Warren 2013). To promote the exhumation of the SNC, detachment of a lithospheric mantle fragment above the subduction zone, as proposed by Majka et al. (2014b), would provide both a local tectonic force (i.e., tectonic under-pressure, Majka et al. 2014b) and space at shallower depths (Fig. 14C). The rupture of this mantle fragment probably resulted from contractional stresses during the arc-continent convergence (Fig. 14B) that resulted in the deposition of the mid-Ordovician turbidites in the foreland basin (Gee et al. 2014). Alternatively, progressive hydration of the hanging-wall mantle wedge could widen the subduction channel and trigger the onset of forced return flow for exhuming UHP rocks (Gerya et al. 2002). However, there is no evidence for such hydrated mantle rocks in the central Jämtland SNC. The fabrics of the blocks enclosed in the migmatitic gneisses (Fig. 4C,D, and S1-B-o) were potentially developed during this period of the exhumation. Positive buoyancy of the bulk SNC rocks probably played an important role in this stage (Warren 2013). Near the end of the first stage exhumation (Fig. 14C), UHP metamorphism was recorded in the Western Gneiss Region (WGR) beneath the hinterland by the 434 ± 3 Ma Flemsøy peridotite (Spengler et al. 2009).

Second stage exhumation: tectonic flow at crustal levels resulted from accretion of the Lower Köli Nappes

The second stage exhumation of the central Jämtland SNC started at depths corresponding to 1.0–1.1 GPa and 850–860 °C following the migmatization at 440–435 Ma (Fig. 13). Since the SNC underwent decompression melting during the first stage exhumation, the density contrast between the SNC and surrounding crust during the second stage exhumation was probably minimal. Therefore, the exhumation of the SNC thereafter was likely not controlled by buoyancy.

The second stage exhumation was characterized by ductile shear of the migmatized SNC, likely associated with foreland-directed thrusting of the Lower Köli Nappes (discussed below) during 433‒418 Ma (Fig. 14D). Based on variations of metamorphic grade, this stage of the exhumation of the SNC in lower granulite- to amphibolite-facies conditions was spatially confined between the underlying Särv Nappes and the overlying Köli Nappes. This is demonstrated by an upwards increase in metamorphic grade of the greenschist-facies Särv Nappes to garnet-mica schists and amphibolites, and a local downwards increase in grade of the Köli Nappes deviating from the general greenschist- to lower amphibolite-facies, reflecting downwards and upwards heat fluxes, respectively, from the intervening SNC (Sjöström 1984; Bergman 1992; Bergman and Sjöström 1997).

The structures and kinematics of the ductile deformation revealed in the current study (Fig. 12) provide key insights into the mechanism of the second stage exhumation. The evidence for top-to-WNW shear (Figs. 7H, I, 8D–F) that was concordant with the dominant top-to-ESE shear is important because coexistence of opposite kinematics across different structural levels indicates various rates of ductile flow extruding toward the foreland. Therefore, extrusion tectonics (e.g., Thompson et al. 1997), should have contributed to the exhumation of the SNC at certain structural levels in addition to the dominant foreland-directed non-coaxial shear (Fig. 12). This interpretation is consistent with the spatial relationship that the SNC was confined by the Särv and Köli Nappes, which provides proper boundary conditions for extruding. An extrusion wedge bounded by a normal fault (top-to-NW) at the top and a coeval thrust (top-to-SE) at the base (e.g., Ring and Glodny 2010) was proposed for the exhumation of the entire Upper and Middle SNC in the area at the Västerbotten-Jämtland border (Fig. 1A) (Grimmer et al. 2015). A key premise for this model was the top-to-NW slip along the roof shear zone. The kinematics of the roof shear zone in the current study, however, is dominated by top-to-ESE shear. An extrusion wedge as a solitary mechanism for exhuming the whole Upper and Middle SNC (Ring and Glodny 2010), therefore, is not tenable for the SNC in central Jämtland. Bender et al. (2018) also showed that evidence for extrusion of the entire SNC is lacking in Jämtland. The extensive top-to-ESE slip suggests that the non-coaxial shear of the SNC had a gradient of displacement that progressively decreased toward lower structural levels. Rocks at the top of the SNC, therefore, should record the largest ESE-ward displacement. The mylonitic‒ultramylonitic fabrics of the roof shear zone containing the highest degree of strain support this interpretation. This gradient of ductile flow could be interpreted by two alternative processes: (1) the overriding Lower Köli Nappes were thrusting toward the ESE, which induced passive flow of the underlying SNC rocks; or (2) the SNC was extruding toward the ESE with highest flow rates at the top and progressively decreased rates toward lower structural levels. The latter process is not common because the highest rate of particle movement during extrusion is expected to be in the middle of the extrusion channel (Thompson et al. 1997). We, therefore, consider that the ESE-ward thrusting of the overlying Lower Köli Nappes due to the initial collision between Laurentia and Baltica since ~ 440 Ma (Spengler et al. 2009) was the primary cause for the non-coaxial flow of the central Jämtland SNC. Extrusion likely contributed subordinately at some structural levels for exhuming the SNC as shown by the sparse outcrops displaying opposite sense of shearing. The extrusion was probably later given to thrusting due to cooling of the SNC (e.g., Gee et al. 2013). This interpretation requires that at least the lowermost Köli Nappes had been emplaced upon the central Jämtland SNC prior to or during the second stage exhumation of the SNC, which is supported by the following evidence from the Köli-Seve boundary.

Peak metamorphism during thrusting of the Lower Köli Nappes in central Jämtland is indicated by the shared amphibolite-facies mylonites along the basal shear zone of the Köli Nappes and the roof shear zone of the SNC (Bergman 1992). Also, P–T paths demonstrate metamorphic convergence across the boundary between the Lower Köli Nappes and the Upper SNC (Bergman 1992). The accretion of the Lower Köli Nappes onto the SNC, therefore, should be broadly coeval with the second stage exhumation of the SNC.

Third stage exhumation: piggy-back transport due to eduction of the Western Gneiss Region lithosphere

The third stage exhumation of the central Jämtland SNC (Fig. 14E) was likely expediated by eduction of the WGR lithosphere (Fig. 1B) beneath the hinterland (Andersen et al. 1991). The WGR locally underwent decompressional melting at about 395 Ma (Fig. 14E) (Krogh et al. 2011), which followed the widespread (U)HP metamorphism in the WGR in 415‒400 Ma (Kylander-Clark et al. 2007; Krogh et al. 2011) and initiation of foreland-directed eclogite-facies shear zones in 415‒412 Ma (Young 2017). The decompressional melting was likely caused by eduction of the Baltica lithosphere possibly triggered by detaching of the eclogitized continental lithosphere outboard of the WGR. Available muscovite cooling ages of the SNC and Köli Nappes in central Jämtland vary between 425–410 Ma (Fig. 1B) (Dallmeyer et al. 1985), similar to those in the eastern Trondheim Nappe Complex (Fig. 1B). This suggests that, prior to ~ 410 Ma, the Upper and Middle Allochthons in these regions were tectonically active and cooled through argon closure in muscovite diachronously at different structural levels and locations. To the west, after ~ 410 Ma, the western Trondheim Nappe Complex and WGR cooled through argon closure in muscovite systematically, i.e., progressively younging westward until ~ 380 Ma along the Norwegian west coast (Fig. 1B) (Walsh et al. 2013). This systematic, post-410 Ma cooling is one of the key pieces of evidence supporting eduction of a coherent slab as the main mechanism for exhuming the (U)HP WGR (Andersen et al. 1991; Young et al. 2007). Other evidence such as westward decrease in titanite ages and increase in regional pressure gradient, and E-W exhumational fabrics are discussed by earlier studies (e.g., Young et al. 2007; Kylander-Clark et al. 2008; Hacker et al. 2010; Spencer et al. 2013). Therefore, the third stage exhumation of the central Jämtland SNC was likely characterized by piggyback transport of the SNC and the surrounding nappes atop Baltica toward the foreland (Brueckner and Cuthbert 2013). ⁴⁰Ar/³⁹Ar isotopes of mica and potassium-feldspar could record the cooling ages during this stage because it is expected no significant deformation to reset the ⁴⁰Ar/³⁹Ar system with the piggyback model (Brueckner and Cuthbert 2013). The mica cooling ages of 418–416 Ma obtained from the SNC roof shear zone in the current study are interpreted to document cooling through 450–300 °C at the early phase of the third stage exhumation. The top-to-W and top-to-E extensional shear zones truncating the tectonostratigraphy with displacements on the order of a few kilometers or tens of kilometers along some of the larger zones in the WGR (e.g., Fossen and Dunlap 1998) and in central Jämtland (e.g., Sjöström et al. 1996; Sjöstrand 1999) likely formed in this time at mid- to upper-crustal levels, which contributed to vertical thinning of the allochthon pile and additional exhumation of the SNC.

At the Laurentian side, (U)HP rocks of 410‒390 Ma now exposed in eastern Greenland (Gilotti et al. 2008; Augland et al. 2010) might be exhumed during this stage as the WGR was being educted. The ending of the third stage and beginning of the fourth stage exhumation of the SNC (Fig. 13) is marked by the completion of the Scandian Orogeny at ~ 375 Ma in the central Scandinavian Caledonides (Tucker et al. 2004). The eduction of the WGR lithosphere probably was the major contributor for the > 350 km horizontal transport of the central Jämtland SNC into the cratonic interior without pervasive deformation (Brueckner and Cuthbert 2013).

Final stage exhumation: uplift due to gravitational collapse and erosion

The final stage exhumation of the central Jämtland SNC was likely driven by gravitational collapse and surface processes (Fig. 14F), as the Scandian contractional tectonics ceased at ~ 375 Ma (Tucker et al. 2004). The ⁴⁰Ar/³⁹Ar data of potassium-feldspar (Fig. 11C-iv) in the current study appear to have captured a moment of cooling through ~ 180 to 150 °C at about 340 Ma during this stage. Eventually, surface processes became the dominant mechanism for exhuming the SNC until it was exposed at the Earth’s surface.

Concluding remarks

This study integrates newly acquired structure and ⁴⁰Ar/³⁹Ar data from the central Jämtland SNC into the published P–T–t dataset and regional geology, which forms the framework of our understanding of the exhumation of the central Jämtland SNC from (U)HP depths beneath the Caledonian hinterland to the Earth’s surface in the foreland. The following summarizes the main conclusions.

• 1. Structural data from the post-migmatization, pre-normal-faulting fabrics of the central Jämtland SNC demonstrates that the bulk deformation was triclinic, which is defined by a predominantly top-to-ESE non-coaxial shear along subhorizontal foliations, a subordinate non-coaxial shear along the foliation and orthogonal to the WNW-ESE lineation, and a subvertical flattening component (Fig. 12). An extrusion component is also documented at several structural levels. This deformation is characteristic of the second stage exhumation of the central Jämtland SNC.

• 2. ⁴⁰Ar/³⁹Ar data from the SNC roof shear zone and the Upper SNC reveal cooling through ~ 450 to 300 °C at about 418‒416 Ma and through ~ 180 to 150 °C at about 340 Ma (Fig. 11). These data indicate that the second stage exhumation occurred prior to ~ 418 Ma and provide insights into the cooling history of the third and final stages exhumation of the central Jämtland SNC.

• 3. In combination with published P–T–t data, four stages of exhumation (Fig. 13) of the central Jämtland SNC are summarized: (1) buoyancy-driven exhumation triggered by tectonic under-pressure, characterized by isothermal decompression and upward flow of the SNC from (U)HP depths to granulite-facies depths (Fig. 14C); (2) tectonic exhumation at crustal levels resulted from accretion of the Lower Köli Nappes, characterized by dominant foreland-directed thrusting, vertical thinning, and minor extrusion (Fig. 14D); (3) eduction of the WGR lithosphere at depths and extensional faulting at mid- to upper-crustal levels, largely characterized by > 350 km piggyback transport of the central Jämtland SNC from the hinterland to the foreland (from E to F in Fig. 14); and (4) gravitational collapse- and erosion-driven exhumation uplifting the SNC to the Earth’s surface (after F in Fig. 14).

• 4. The first three stages of the exhumation involved simultaneous combination of vertical and horizontal displacements, i.e., toward the Earth’s surface and the Caledonian foreland, respectively. The first stage was likely dominated by vertical movement while the third stage by horizontal movement. The latter probably made the greatest contribution to the > 350 km west-to-east transport of the central Jämtland SNC, which might be considered as a combination of relative displacements along fault zones at various levels above the Baltic basement and the absolute distance traveled by Baltica itself during the eduction of the WGR lithosphere.

This study provides insights into the general processes of exhumation of deeply subducted continental rocks in orogenic belts using the SNC in central Jämtland as an example. Future work of the SNC may reveal a more inclusive model incorporating (i) diachronous exhumation of the SNC along the Scandinavian Caledonides and (ii) progressive exhumation of the SNC from and to different depths across the orogen between the foreland and hinterland.