Synchronous Periadriatic magmatism in the Western and Central Alps in the absence of slab breakoff

Periadriatic Alpine magmatism has long been attributed to slab breakoff after Adria–Europe continental collision, but this interpretation is challenged by geophysical data suggesting the existence of a continuous slab. Here, we shed light on this issue based on a comprehensive dataset of zircon U–Pb ages and Hf isotopic compositions from the main western Periadriatic intrusives (from Traversella to Adamello). Our zircon U–Pb data provide the first evidence of Eocene magmatism in the Western Alps (42–41 Ma in Traversella), and demonstrate that magmatism started synchronously in different segments of the Alpine belt, when subduction was still active. Zircon U–Pb ages define younging trends perpendicular to the strike of the European slab, suggesting a progressive Eocene–Oligocene slab steepening. We propose that slab steepening enhanced the corner flow. This process was more effective near the torn edge of the European slab, and triggered Periadriatic magmatism in the absence of slab breakoff.

According to Davies and von Blanckenburg (1995), slab breakoff magmatism would be induced by the passive asthenosphere upwelling along the breakoff gap. Such magmatism should exhibit a mantle parentage, should be extremely localized, its trace should be nearly linear, its duration very short, and intrusions may display symmetrically younging trends with distance away from the breakoff gap . Only part of these features are observed in the Alpine region: the Periadriatic plutons are indeed clustered along the Insubric Fault (Figure 1a,b), which probably favoured magma ascent (Rosenberg, 2004), but the widespread Periadriatic dykes (Bergomi, Zanchetta, & Tunesi, 2015;D'Adda et al., 2011) form a much wider belt parallel to the European slab ( Figure 1b).
The location and age of magmatism may also reflect parameters such as the distance from the slab, its angle and the polarity and rate of subduction. These parameters can vary during subduction thus modifying the time, location and geochemistry of F I G U R E 1 (a) Relationships between tectonic structure and Periadriatic magmatism in the Western and Central Alps (slab structure after Zhao et al., 2016). GF = Giudicarie Fault; VVP = Venetian Volcanic Province. (b) Summary of weighed mean 206 Pb-238 U zircon ages in the western Periadriatic intrusives (see Figures DR1-DR5 for detailed age maps and Concordia diagrams). S1-S49 = samples analysed in this work. Literature U-Pb ages (lozenges) compiled from: Berger et al., 2012;Bergomi et al., 2015;Broderick et al., 2015;D'Adda et al., 2011;Gianola et al., 2014;Hansmann & Oberli, 1991;Liati et al., 2000;Mayer et al., 2003;Romer et al., 1996;Samperton et al., 2015;Schaltegger et al., 2009;Schoene et al., 2012;Stipp et al., 2004;Tiepolo et al., 2011Tiepolo et al., , 2014von Blanckenburg, 1992. Green dots = Periadriatic dykes (after Bergomi et al., 2015;Bistacchi & Massironi, 2000;Kapferer, Mercolli, Berger, Ovtcharova, & Fügenschuh, 2012;Malusà, Philippot, Zattin, & Martin, 2006;Rosenberg, 2004). Dashed blue line = trace of the European slab according to the teleseismic tomography model of Zhao et al., 2016(150 km depth slice, after Salimbeni et al., 2018. (c) Migration of Periadriatic magmatism in three steps  and relationships with the European slab as outlined by seismic tomography (Zhao et al., 2016). The distribution of Periadriatic dykes (shaded blue area) forms a belt parallel to the European slab that gets progressively wider from the Western to the Central Alps. The southern boundary of this envelope is near-parallel to the alignment of Periadriatic intrusives emplaced at 43-40 Ma (sTR, sRC, nRC, CA), whereas the northern boundary is near-parallel to the strike of the slab (dashed blue line) and to the trend defined by intrusives emplaced at 32-30 Ma (BI and BG). White arrows = younging trends defined by zircon U-Pb ages [Colour figure can be viewed at wileyonlinelibrary.com] the magmatism Mullen, Paquette, Tepper, & McCallum, 2018). Notably, only part of the Periadriatic plutons have been dated by modern techniques so far (e.g. Samperton et al., 2015;Tiepolo, Tribuzio, & Langone, 2011), and a full coverage of Hf isotopic analyses is still missing for most of these plutons (e.g. Broderick et al., 2015;Schoene et al., 2012;Tiepolo, Tribuzio, Ji, Wu, & Lustrino, 2014). In this work, we provide the first self-consistent dataset of zircon U-Pb ages and Hf isotopic compositions from the main Periadriatic intrusives of the Western and Central Alps. Age and isotopic trends resulting from our analyses are discussed within the framework of available geodynamic constraints for the Alpine region, shedding new light on the complex slab-mantle interactions during the latest stages of Alpine evolution.

| G EOLOG I C BACKG ROUND
The European Alps are the result of Cretaceous-to-Palaeogene oblique subduction of the Alpine Tethys and adjoining European palaeomargin beneath Adria (Handy et al., 2010;Malusà et al., 2015) (Figure 2a-c). In the Western Alps, Alpine subduction was active until the late Eocene, as attested by (U)HP rocks that reached the eclogitic peak at ~35 Ma (Rubatto & Hermann, 2001) and were rapidly exhumed during upper-plate divergent motion by 32 Ma, corresponding to the age of the stratigraphic cover of the Voltri massif (Liao et al., 2018;Malusà, Faccenna, Garzanti, & Polino, 2011;Quaranta, Piazza, & Vannucci, 2009). In the Eastern Alps, Alpine subduction was active until the early Oligocene, as attested by eclogites of the Tauern Window that reached their pressure peak at ~31 Ma (Glodny, Ring, Kühn, Gleissner, & Franz, 2005) and then experienced crustal shortening (Rosenberg et al., 2018).

encased into
Sesia-Lanzo metamorphic rocks to the NW of the Insubric Fault, similarly to the Biella pluton (BI) that includes monzonites, syenites, granitoids and leucogranites dated at 31-30 Ma (Romer, Schärer, & Steck, 1996). On the opposite side of the fault, the nearby Miagliano tonalite (MI) was emplaced at ~33 Ma within lower crustal rocks of the Ivrea-Verbano Zone (Berger, Thomsen, F I G U R E 2 (a-c) Palinspastic reconstruction of the Adria-Europe plate boundary zone in three steps (after Malusà et al., 2011Malusà et al., , 2015Zanchetta et al., 2012Zanchetta et al., , 2015; purple arrows show the relative plate motion, numbers = ages in Ma (Dewey, Helman, Turco, Hutton, & Knott, 1989); note that Alpine subduction was oblique to the European passive margin, and the inception of continental subduction migrated progressively from the Western to the Central Alps. Colour codes as in Figure 1. (d) Present-day relationships between the Alpine (European) and Dinaric (Adriatic) slabs as outlined by the high-resolution teleseismic P wave tomography model of Zhao et al. (2016)  Ovtcharova, . In the Central Alps, the Bregaglia pluton (BG) includes granodiorites and tonalites with minor gabbros and diorites, emplaced around 32-30 Ma into Austroalpine and Penninic units (Samperton et al., 2015), whereas the Novate leucogranite (NV) was dated at ~24 Ma (Liati, Gebauer, & Fanning, 2000). Farther east, the composite Adamello batholith was emplaced into South Alpine units between 43 and 32 Ma (Broderick et al., 2015;Mayer et al., 2003), forming distinct magmatic units now exposed between the Insubric and Giudicarie faults (Figure 1b). In the early Oligocene, Periadriatic plutons were also intruded in the Eastern Alps (e.g. Rensen, Rieserferner and Karawanken plutons, see Bergomi et al., 2015 for a review of available age constraints).
At the transition between the Central and Eastern Alps, the slab structure is particularly complex due to the onset of Dinaric subduction in the middle Eocene (Carminati, Lustrino, & Doglioni, 2012) (Figure 2c). Some authors (Handy, Ustaszewski, & Kissling, 2015;Schmid, Scharf, Handy, & Rosenberg, 2013), based on the Lippitsch, Kissling, and Ansorge (2003) F I G U R E 3 Single zircon 206 Pb-238 U dates and Hf isotope compositions from the main Periadriatic intrusions of the Western and Central Alps. The field for the Bregaglia mafic rocks is after Tiepolo et al. (2014); literature data from the Re di Castello unit (small grey triangles) are from Schoene et al. (2012) and Broderick et al. (2015). The inset on the top-left summarizes the main trends discussed in the text. Note the systematic ε Hf (t) decrease from east to west in the mafic end-members (RCm, BGm and sTR) [Colour figure can be viewed at wileyonlinelibrary.com]

| ME THODS
We performed zircon U-Pb dating by LA-ICP-MS and in situ Hf isotope analyses on 49 rock samples (see locations in Figure 1b). A detailed description of the analytical procedures, detailed age maps, raw U-Pb and Hf isotope data and whole-rock major and trace element datasets are provided in the Data S1.

| RE SULTS
Zircon U-Pb ages and Hf isotopic compositions are summarized in Zircons from the Biella monzonitic, syenitoid and granitoid complexes (S7-S11) yielded ages at 31-30 Ma, whereas those from the Miagliano tonalite (S12) are around 33 Ma. Both results are consistent with literature ages Romer et al., 1996).

| D ISCUSS I ON
Our results can be interpreted in the light of the slab structure outlined by the tomography models of Zhao et al. (2016) and Sun et al. (2019), with particular emphasis on the attitude of the European slab and the role of the slab edge formed by vertical tearing. As shown in Figure 1c, the middle Eocene zircon U-Pb ages   (Figure 1b,c).
The post-Oligocene dextral movements along the Insubric Fault (Malusà, Anfinson, Dafov, & Stockli, 2016;Schmid, Aebli, Heller, & Zingg, 1989) may have an impact on the above age trends. The Insubric Fault is an inherited Permian structure (Muttoni et al., 2003) lying at ~30° relative to the strike of the European slab (Figure 1c). Post-Oligocene movements along this fault are minor in the Western Alps (Bistacchi & Massironi, 2000;Malusà, Polino, & Zattin, 2009), but estimates of right-lateral slip in the Central Alps range from 10 to 20 km (Garzanti & Malusà, 2008) or ~30 km (Müller et al., 2001) to >100 km (Schmid & Kissling, 2000;Schmid, Kissling, Diehl, van Hinsbergen, & Molli, 2017). In this latter case, both the original distance between the Bregaglia and Adamello plutons and the migration of magmatism inferred from Figure 1c would have been much larger. However, for very large fault offsets, some plutons would be located too far from any slabs, inconsistent with magma generation in a supra-slab environment as constrained by trace element compositions.
The incompatible trace element compositions of the Adamello, Bregaglia and Traversella mafic rocks are in fact supportive of a subduction related origin, and in particular of mantle sources fluxed by slab-derived components (Tiepolo et al., 2014). As shown in Figure 3, the Hf isotopic ratios of these mafic end-members systematically decrease from east to west, from the Adamello batholith (RCm in Figure 3) through the Bregaglia pluton (BGm in Figure 3) to southern Traversella (sTR in Figure 3). The mafic rocks from the Bregaglia pluton and the Adamello batholith were demonstrated to be almost unaffected by shallow level crustal contamination, and reflect the primary Hf isotopic signature of the mantle wedge (Tiepolo et al., 2014). Diorites from Traversella were suggested to register limited crustal contamination (De Lummen & Vander Auwera, 1990), an hypothesis in line with the absence, in diorites, of zircon inheritance (Table DR4), and with trace element compositions resembling those of uncontaminated Bregaglia and Adamello mafic rocks (Table DR6).
However, the chemical features of the Traversella diorites contrast with the extremely low ε Hf (t) values, that are similar to those observed for the Novate leucogranite. This ε Hf (t) signature requires the addition of 176 Hf-depleted crustal material to the mantle source during subduction. Notably, the higher potassium contents at almost comparable SiO 2 in mafic rocks, and the higher concentrations of Ba and Th (Table DR6 and  Their distribution even better marks the progressive steepening of the European slab, and gets progressively wider towards the Central Alps, where slab steepening was more pronounced (Figure 1c).