Geology of the High Sillaro Valley (Northern Apennines of Italy)

We map the High Sillaro Valley, in the Northern Apennines of Italy. Here the Ligurian wedge overthrust the foredeep deposits of the Marnoso Arenacea formation – MA. The overthrusting occurred through a shear zone with components derived from the frontal part of the Ligurian wedge and interpreted as a tectonic mélange (Sestola-Vidiciatico Tectonic unit SVU) here subdivided into two different subunits structurally independent one from the other. In the area, the Visignano chaotic body (VIS) constitutes an intercalation within the Serravallian-Tortonian MA and derives, as the SVU, from the frontal part of the Ligurian wedge. We performed a detailed investigation of the components of the VIS that helps to better define the architecture of the frontal part of the Ligurian wedge and to constrain the timing of the deformation phases affecting this portion of the Northern Apennines during the middle-late Miocene. ARTICLE HISTORY Received 30 March 2020 Revised 28 September 2020 Accepted 2 October 2020

The chaotic body has been considered as part of the Sestola-Vidiciatico Tectonic Unit (SVU) Vannucchi et al., 2008 and references therein) and defined as a tectonic mélange composing the shear zone between the Ligurian wedge and the foreland deposits.
Along the Sillaro Line, the Visignano chaotic body (VIS) (Figure 2, Main Map), formerly known as the 'Visignano olistostrome' (Landuzzi, 2006 and references therein), represents another large intercalation ∼200-300 m thick, interposed within the upper Serravallian-lower Tortonian MA (Martelli, 2014). We clearly define the components of the VIS and compare them with the overthrusting Ligurian Units and the SVU. This is fundamental to better define the composition of the frontal part of the Ligurian wedge and to constrain the timing of the deformation of the area.
This new geological map focused on i) the distinction of the SVU in two different subunits and careful characterization of their boundaries; ii) the definition of the components of the VIS chaotic unit on the base of a detailed field mapping and new biostratigraphic data; iii) the correlation between extrabasinal elements inside the underthrusting units and the components of the overthrusting unit. This allows to: i) better define the architecture of the frontal part of the Ligurian wedge during the middle-late Miocene and ii) constrain the timing of the deformation phases affecting this portion of the Northern Apennines in that period.

Methods
The geological map was produced from observations collected at the 1:10.000 scale and accompanying stratigraphic, structural and biostratigraphic analyses. The mutual relationships between tectonic contacts, the geometry of the different stratigraphic units, and the stratigraphic unconformities of different rank allowed the reconstruction of the tectono-stratigraphic evolution of the mapped area. The results are documented in the attached 1:25.000 scale map (see Main Map), using the topographic maps at 1:10.000 and 1:5.000 provided by the Tuscany and Emilia Romagna administrations respectively. The distinction among different chaotic units (i.e. sedimentary mélange, tectonic mélange, broken formation) followed specific criteria defined in the literature (e.g. Bettelli et al., 2004;Bettelli & Panini, 1989;Festa et al., 2019).

Regional setting
The Northern Apennines (Figure 1) are the result of the convergence and collision between the European and Adria plates (Vai & Martini, 2001 and references therein). After the middle Eocene, an already formed oceanic submarine accretionary prism constituted by the Ligurian Units (Principi & Treves, 1984) progressively overrode the subducting continental Adria plate and its sedimentary cover (Tuscan and Umbro-Marchean units). Since the middle Eocene to late Miocene, the Ligurian accretionary prism has been unconformably overlain by the forearc slope sediments of the Epiligurian succession (Ricci Lucchi, 1986;Ricci Lucchi & Ori, 1985).
At least from early to middle Miocene, the relative displacement between the Ligurian accretionary prism and the Adria plate sedimentary cover was accommodated by deformation localized within a 100-500-mthick shear zone. This zone now crops out as a tectonic mélange, the SVU, in the southeastern part on the Emilia Apennines. The more coherent Subligurian Units represent the same shear zone to the northwest (Vannucchi et al., 2008;. At present, the shear zone represented by the SVU is cut by several folds and thrusts but the age of its deactivation is not well constrained. Vannucchi et al., 2008 suggest a Serravallian or younger deactivation of the shear zone on the base of the youngest age of the slices that were tectonically recycled within the SVU, but our careful mapping suggests that these slices lie outside the tectonic mélange (see discussion) allowing for an older de-activation of the shear zone.

Data
The study area (Figure 2, Main Map) is located along the Sillaro Valley. Here the Ligurian Units overthrust the Umbrian-Marchean Units through the tectonic   Figure 2; Bettelli, Boccaletti et al., 2002;. The differences between these tectonic units are mainly lithostratigraphic and probably related to different depositional environments within the Ligurian ocean (see legend of the Main Map), while their deformation style is similar throughout the whole sequence, even if strongly influenced by the lithology (Bettelli & Vannucchi, 2003;Vannucchi & Bettelli, 2002).

Sestola-Vidiciatico Tectonic Unit
The SVU is a tectonic mélange (Vannucchi et al., 2008 and reference therein) made by m-to-km-size tectonic slices, separated by mechanical contacts. Composition and areal distribution allow the subdivision of the SVU into two different subunits (Bettelli & Panini, 1992), structurally independent one from the other: the Sestola-Vidiciatico tectonic subunit A (SVA) (Chaotic unit A in Bettelli & Panini, 1992) and the Sestola-Vidiciatico tectonic subunit B (SVB) (Chaotic unit B or Carpinaccio unit in Panini, 1992 and. The SVB crops out along the Sillaro Line, from Sasso di Castro to Sassoleone and Gesso area ( Figure  2). It is mainly formed by large slices of strongly deformed pelagic shales intercalated with marls and/ or calcareous or siliciclastic turbidites of Late Cretaceous to?middle Miocene age (Bettelli, Boccaletti et al., 2002;Panini et al., 2002). This subunit includes tectonic slices pertaining to the Ligurian Units. The mapped units include the Val Samoggia varicolored shale (AVS, Bettelli & Panini, 1992) (Figure 3(a)), the Sillano formation (SIL) (Figure 3(e)), including also blackish shale alternated to siliceous calcilutite (SILa) and deep-sea carbonates to hybrid turbidites alternating with shales (SILb) (Nirta et al., 2007), the Mt. Morello formation (MLL, Bortolotti et al., 2009) (Figure 3(c), Figure 4(a)) and siliciclastic and calcareous turbidites (HLM). Subordinate amounts of debris flow deposits (BAP) of uncertain age (Figure 3(b)), mainly deriving from the reworking of the AVS, are also present. The new fieldwork and biostratigraphic analyses revealed that the SVB contains slices of marls, shales and peculiar biocalcarenites and arenitic turbidites (MSB) ranging in age from late Eocene/early Oligocene to?early Miocene ( Table 1). Some of these latter slices were already distinguished in Benini et al. (2009) andFarabegoli (2009) who correlated them partly to the Epiligurian succession or the Subligurian Units, respectively.
The structural position, general composition and internal structure of the SVB are similar to the SVU cropping out from Suviana (BO) to the Val Secchia (Remitti et al., 2007;Vannucchi et al., 2008). However, the composition of the SVB differs from the composition of the SVU elsewhere, given the presence of slices of AVS, large slices of MLL and the peculiar composition and age of the MSB components.
The SVA crops out in the internal, SW area going from the Giogo Pass to Peglio ( Figure 2). It consists of slices of thick submarine debris flow deposits (BAP) (Figure 3(b)) and large slices of pelagic shales intercalated with marls and/or calcareous turbidites, of Upper Cretaceous to lower Eocene age (SIL and HLM in the Main Map;   (Figure  3(e)). Moreover, the SVA contains slices of marls, shale and arenitic turbidites (FMC) ranging in age from the late Eocene to early Langhian (middle Miocene) (Figure 3(d)). Age, stratigraphy, sedimentology and lithological composition of the FMC exotic blocks suggest a correlation with the Fiumalbo shales, Marmoreto marls, Civago marls, Baigno marls or Suviana sandstones (Benini et al., 2014;Bettelli, Boccaletti et al., 2002) elsewhere representing portions of the lithostratigraphic units included in the SVU. Therefore, the SVA shares a similar composition with the SVU cropping out elsewhere along the Northern Apennines excluding the SVB.

Umbrian-Marchean Units
The Marnosa Arenacea formation (MA) is a turbidite succession, >4000 m thick, deposited in the foredeep basin developed in front of the NE-advancing Ligurian wedge (Ricci Lucchi, 1975). Following Landuzzi (2006) and Tinterri and Tagliaferri (2015), the Casaglia (CAS) and the Visignano (VIS) chaotic bodies (Bruni, 1973;De Jager, 1979;Lucente et al., 2002;Lucente & Pini, 2003;Ogata et al., 2016; represent lithostratigraphic markers, allowing the subdivision of the MA succession in MA1 (deposited before CAS), MA2 (deposited between CAS and VIS events), MA3 (deposited after VIS) (Main Map). Both CAS and VIS are composed by a main body with extra-basinal (i.e. Ligurian) blocks and by an associated slump of Miocene marls and thin-bedded turbidites ( Figure 3(f and h)). The main body of CAS crops out SW of the mapped area whereas in the Sillaro area is only represented by the coeval slumped layers (Benini et al., 2014).
The  Table 1; see discussion below) locally unconformably cover this chaotic assemblage. These components suggest that the extra-basinal material is sourced (or shared the source) from both the SVA (as testified by the presence of VIS FMC ) and the SVB (as testified by the presence of VIS AVS ).
Considering as a marker the top of the CAS and the basal surface of the VIS, the MA2 thickness decreases abruptly moving toward NE (Main Map, Figure 5), showing sharp angular unconformities and the interruption of the CAS itself, that reappears only a few km northeastward. As suggested by several previous works (Ricci Lucchi, 1986;Roveri et al., 2002;Tagliaferri et al., 2018;Tinterri & Tagliaferri, 2015) it is likely that at the time of emplacement of the VIS within the MA basin, the MA was affected by strong synsedimentary propagation of a blind thrust and growth of the related Coniale anticline (Figure 2; 5; 6). The axial trace of this fold could be followed at the surface only in its western part (Castelnuovo-Visignano area) whereas eastward it evolves into a thrust surface (Figure 2; 5; 6). Therefore, the VIS was probably emplaced on a structural/topographic high where turbiditic sediments were replaced by pelitic sediments (Main Map and Figure 5) as testified by the presence, below the VIS, of marly sediments and thin-bedded turbidites of late Serravallian-Tortonian age (MA2b) and drastic pinch-out of the MA2 sandstone strata (Tinterri & Tagliaferri, 2015; Table 1).
The basal surface of MA3 above VIS is an unconformity on top of different blocks, including blocks of VIS AVS (Figure 3(g) and Figure 4(c)). The overturned stratigraphic contact between VIS AVS and MA3 is preserved NW of Mt. Porrara (Main Map; Figure 3(g)). The age (early Serravallian) of some of the marls inside the VIS (VISa) (see Table 1) and the structural position of some of the blocks of VISa, surrounded by VIS AVS (Figure 4(f), Main Map) suggest that their deposition has begun before the emplacement of the VIS in its current position. However, near Castelvecchio, the deposition of the MA3 occurs in continuity atop and parallel to the marly or thinbedded turbidite sediments of VISa (Castelvecchio marls, Landuzzi, 2006;Tinterri & Tagliaferri, 2015) without apparent discontinuity. To sum up, at least some of the marls and thin-bedded turbidites (VISa and VISb) can be considered belonging to the VIS, sedimented on the foredeep inner lower slope on top of the very frontal part of the Ligurian wedge and subsequently displaced, together with their original chaotic substratum, in the foredeep basin during the late Serravallian-early? Tortonian.
From Poggio Belmonte to Mt. Belvedere, marls of Serravallian age (SER) crop out at the boundary between the SVA and SVB and the MA3 (Figure 4 (b). They have been previously considered part of the SVU (Vannucchi et al., 2008), but the present mapping suggests that they can be better described  as tectonic slices of more internal derivation (maybe VISa or marls belonging to MA3), dragged along a major top-to-NE low angle fault.

Structural setting
In the area, several major faults crop out (see tectonic scheme beside the Main Map). A major thrust fault, at present dipping toward NW (probably due to a later rotation), juxtaposes the Ligurian Units (MOH and LEO) with the SVB; another fault thrusts the MOH over the LEO. The SVA thrusts over the MA including part of the VIS with slip direction towards NE. This latter thrust is cut by another NNE-trending major fault (red line in the Main Map), at present dipping toward the Po plain, putting the SVB on top of the SVA in the south-western part and directly over the MA3 in the northeastern part, with slip direction still toward NE. Along the fault, several slices of SER are dragged and displaced. In the SE part of the map a major thrust, known as the Mt. Castellaccio (Tinterri & Tagliaferri, 2015) or S. Benedetto in Alpe thrust (Benini et al., 2014)) cuts through the MA from SE to NW for more than 30 km, becoming progressively blind towards NW (see Figures 5 and 6). The fault is associated with an overturned NE verging anticlinal and synclinal folding the MA2, MA3 and intercalated VIS successions. The fold axis plunges toward NW becoming progressively more open and gentler ( Figure 6). Several NE-and SE-trending high angle faults represent the most recent structures. In the western part of the map a major high angle fault, likely with right lateral cinematic, (Idice fault system, Bettelli & Panini, 1992) cuts the boundary between the MOH and the LEO Units.
Inside the SVA and SVB, all the contacts between different lithologies are mechanical and they anneal the pre-existing sedimentary relationships. These contacts may derive from the gravitational or tectonic rework at the front of the Ligurian wedge subsequently deformed inside the shear zone between the Ligurian wedge and the Umbro-Marchean Units.
Similarly, inside the VIS most of the contacts are mechanical (Figure 4(e and f)). They may be possibly inherited from the structuration of the frontal part of the prism from which the VIS components derive, or due to the emplacement of the VIS, even if some of the stratigraphic contacts at the base of VISa and VISb are preserved.

Discussion
The presented map gives several constraints on the interplay, during the early-middle Miocene, between the overthrusting of the Ligurian wedge atop the Umbro-Marchean Units, the sedimentation inside the foredeep basin (on the inner and outer slopes), the deformation of the foredeep deposits and, at the end, the exhumation phase.

Considerations on the composition of SVA, SVB, VIS and frontal part of the Ligurian wedge
Following the interpretation by Remitti et al. (2007) and Vannucchi et al. (2008), the SVU resulted from the underthrusting during the Miocene by frontal tectonic erosion of a proto-mélange forming the frontal part of the overriding plate. The proto-mélange was made by a former accretionary prism (made by Ligurian Units) and by its sedimentary cover. Given its composition, the proto-mélange of the SVA was made by already deformed Ligurian Units (SIL and HLM) and their wedge-toe sedimentary cover (BAP and FMC), while for the SVB the proto-mélange also included AVS and MLL blocks and its wedge-toe sedimentary cover (BAP and MSB) (Figure 7(a)). MSB probably originated from slope deposits, unconformably sedimented, in a paleogeographic domain intermediate between the Epiligurian domain and the foredeep inner lower-slope.
The age of the youngest slope deposits involved inside the SVU constrains the age of the underthrusting of the proto-mélange (closure of the deposition atop the frontal prism), therefore the SVB should have been underthrusted from the early Miocene while the SVA from the middle Miocene.
The VIS derived from the frontal part of the overriding plate and, since it contains blocks of both the SVA and SVB, at the age of its emplacement (late Serravallian/Tortonian) the frontal part of the wedge should have been formed by both the SVA (including its peculiar FMC blocks) and SVB components (including its peculiar AVS blocks) and the overtaking of the SVA by the SVB should have occurred later on (Figure 7(a)).

Considerations on the emplacement of VIS
In literature, the VIS (e.g. Landuzzi, 2006) is considered a mass transport deposits (MTD) deriving from the frontal part of the Ligurian wedge. However, the current data are alternatively compatible with a tectonic emplacement of the frontal part of the Ligurian wedge atop the already tectonically uplifted MA2 (Figure 7(b)). After this emplacement, the fault at the base of the VIS was deactivated and probably replaced by a more internal thrust moving the SVA over the internal part of the VIS, while the external part was sealed by the deposition of MA3. In turn, the SVB and the Ligurian wedge overthrust the MA3 (Figure 7(b)).

Deformation history
At the time of the VIS emplacement, as already claimed by various authors (Ricci Lucchi, 1986;Tagliaferri et al., 2018;Tinterri & Tagliaferri, 2015), the MA was already deformed by the Coniale anticline (Figure 7(b)). This implies that the deformation front of the overthrusting Ligurian wedge cut through the sedimentary cover of the lower plate, differently from what was suggested by the underthrusting of almost undeformed older foredeep deposits (Vannucchi et al., 2008). This probably preludes to a change in the dynamics of the Northern Apennines.
At the same time, as testified by the VIS components, deformed Ligurian Units and their sedimentary cover, pertaining both to the SVA and SVB, composed the toe of the Ligurian wedge. This is true either if we consider the VIS as an MTD deriving from the toe or the toe itself.
However, after the emplacement, the VIS itself is overthrust by the SVA and sealed, in its most external part, by the sedimentation of the MA3. Then the whole sequence is involved in the folding of the Coniale anticline and cut by the Mt. Castellaccio thrust Figure 7(c).
The Coniale anticline, overturned in the eastern sector, becomes progressively a very gentle open fold in the western sector until disappearing below the Ligurian Units ( Figure 5).
Then a major fault brought the SVB on top of the SVA and the MA (Figure 7(d)). The meaning of this fault (base of the SVB) is still questionable. It is a low angle fault, at present tilted toward the north, bringing older material on top of younger material traditionally mapped as a thrust at least in the northern segment (Farabegoli, 2009). Northeast of the mapped area the fault ends in a series of imbricated thrusts bringing the Ligurian and Epiligurian units on top of the late Miocene MA and the overlying late Miocene-Pliocene succession (Farabegoli, 2009;Pini, 1999). However, SW of the mapped area (Futa Pass area), the Ligurian Units are directly superposed on the Tuscan Units (i.e. without the SVU that should have been sandwiched between the two) possibly by the same fault. Moreover, between Futa Pass and Sasso di Castro both the SVB and the Ligurian Units show wedge shape geometry with an abrupt decrease in thickness towards SW. Therefore, in some cases, this shear zone seems to  display the characteristic of low angle normal fault driving tectonic elision. However, the nature of this contact is still uncertain.
The Coniale anticline closure towards NW does not allow to infer whether the base of the SVB is folded or not, thus if the fault at the base of the SVB was active after or before the Mt. Castellaccio thrust.
Considering the SER slices as not being part of the SVU implies that the SVU mélange did not incorporate material up to Serravallian, as suggested before. Since Serravallian, then, the mechanism of frontal tectonic erosion suggested by Vannucchi et al. (2008) was likely no longer active and a change in the deformation mechanisms affecting the shallowest portion of the chain occurred. This is coherent with the activation of deeper décollement actively deforming the foredeep deposits (e.g. the Mt. Castellaccio thrust) Several vertical faults, cutting all the pre-existing low angle major faults, record the last stage of deformation in the area, likely allowing the different rates of exhumation of the different sectors and generally accommodating the small amounts of lateral displacement.

Conclusions
The new data suggest that: (i) before early Serravallian the components feeding both the SVA and SVB were not completely underthrust and they formed the frontal part of the Ligurian wedge; (ii) after the late Serravallian, the SVU likely no longer represented a major active shear zone; (iii) after the emplacement of the VIS, SVB and the overlying Ligurian wedge were progressively detached from the underlying SVA by a major shallow-dipping fault characterized by top to the NE kinematics.

Software
The geological map was first digitized in GIS environment using qGIS (2.18 version) and then edited along with map insertions using Adobe Illustrator CC.