Downloaded from gsabulletin.gsapubs.org on 30 September 2009 Possible causes of arc development in the Apennines, central Italy Andrea Billi1,† and Mara Monica Tiberti2 1Dipartimento di Scienze Geologiche, Università “Roma Tre,” Largo S.L. Murialdo, 00146 Rome, Italy 2Istituto Nazionale di Geofi sica e Vulcanologia, Via di Vigna Murata 605, 00143 Rome, Italy ABSTRACT INTRODUCTION beneath the Adriatic and Ionian Seas; Fig. 1) is mostly the result of noncylindrical rollback of In central Italy, the geometry, kine matics, Arcuate belts are among the most ubiquitous a subducting segmented lithosphere (Royden and tectonic evolution of the late Neogene but also enigmatic and debated structures within et al., 1987; Faccenna et al., 2004; Rosenbaum Umbrian Arc, which is one of the main orogenic settings (e.g., Marshak, 2004; Suss- and Lister, 2004). In contrast, the origin of sev- thrusts of the northern Apennines, have man and Weil, 2004). In a recent classifi cation eral curved thrusts within the greater northern long been studied. Documented evidence of curved orogens, Weil and Sussman (2004) and southern arcs is still unexplained. for orogenic curvature includes vertical- recognized primary arcs, progressive arcs, and Paleomagnetic studies of the Umbrian Arc axis rotations along both limbs of the arc oroclines. Primary arcs, which are nonrotational provide confl icting interpretations on the de- and a posi tive orocline test along the entire curves, adopt their curvature during the initial velopment of mountain belt curvature, such arc. The cause of the curvature is, however, phase of deformation and experience no appre- as (1) oroclinal bending of an originally linear still unexplained. In this work, we focus ciable tightening and vertical-axis rotations dur- orogen (Channell et al., 1978; Eldredge et al., our attention on the southern portion of ing subsequent deformation. In contrast, both 1985; Muttoni et al., 1998), and (2) an arc with the Umbrian Arc, the so-called Olevano- oroclines and progressive arcs are rotational fold-axis trends that have no relationship to Antrodoco thrust. We analyze, in particular, curves (e.g., Weil, 2006). Progressive arcs either vertical-axis rotations recorded by paleomag- gravity and seismic-refl ection data and con- acquire their curvature progressively throughout netic declinations (Hirt and Lowrie, 1988). Re- sider available paleomagnetic, stratigraphic, their deformation history (i.e., thrust rotations cent paleomagnetic data from the Umbrian Arc structural, and topographic evidence from accommodate continuous along-strike varia- (Fig. 1C) conclusively demonstrate sec ondary the central Apennines to infer spatial extent, tions in shortening; Sussman et al., 2004) or orogenic curvature (Speranza et al., 1997; Mattei attitude, and surface effects of a midcrustal acquire a portion of their curvature during a sub- et al., 1998). In particular, evidence was pro- anticlinorium imaged in the CROP-11 deep sequent deformation phase. Oroclines acquire vided for a positive orocline test along the en- seismic profi le. The anticlinorium has hori- their curvature in a two-step process consist- tire Umbrian Arc, the curvature of which was zontal dimensions of ~50 by 30 km, and ing of (1) the formation of a linear orogen and acquired by simultaneous, and opposite-sense, it is located right beneath the Olevano- (2) the bending of that orogen to form an arc. vertical-axis rotations of the arc’s limbs mostly Antrodoco thrust. Stratigraphic, structural, It is relatively simple to distinguish among after Messinian time (Mattei et al., 1995, 1998; and topographic evidence suggests that the oroclines, progressive arcs, and primary arcs Speranza et al., 1997; Muttoni et al., 1998). The anticlinorium produced a surface uplift dur- when the appropriate methods of surface inves- cause for such rotations is, however, still unclear. ing its growth in early Pliocene times. We tigation can be used (e.g., paleomagnetic, struc- Most authors (Eldredge et al., 1985; Calamita propose an evolutionary model in which, tural, and stratigraphic analyses to understand and Deiana, 1988; Ghisetti and Vezzani, 1997) during late Neogene time, the Olevano- the temporal relationship between thrusting have hypothesized that the main cause for oro- Antrodoco thrust developed in an out-of- and vertical-axis rotations; Weil and Sussman, genic curvature of the southern limb of the sequence fashion and underwent ~16° of 2004). In contrast, understanding the cause of Umbrian Arc is connected with contrasting clockwise rotation when the thrust ran into curvature is usually diffi cult because of, among mechanical competence of the involved rocks and was then raised and folded by the grow- other reasons, the paucity of subsurface data. (i.e., Latium and Sabina carbonates; Fig. 1B). ing antic linorium (late Messinian–early Plio- In this paper, we address the problem of the According to this model, stiff carbonate rocks cene time). This new model suggests a causal cause of curvature for the case of the southern in the central Apennines (i.e., Latium platform link between midcrustal folding and surfi cial portion of the Umbrian Arc (i.e., the so-called carbonates) restrained the advancement of the orogenic curvature that is consistent with Olevano-Antrodoco thrust) in the Apennine arc’s southern limb, thus causing a displacement several available data sets from the northern fold-and-thrust belt, Italy. This belt includes gradient along the northern Apennine thrusts and and central Apennines; more evidence is, two main orogenic arcs, namely, the northern their subsequent curvature. however, needed to fully test our hypothesis. and southern arcs (Fig. 1). These arcs are dif- The Umbrian Arc intersects with the south- Additionally, due to the occurrence of mid- ferent in size, shape, shortening, and involved ern arc thrusts in the central Apennines, where crustal basement-involved thrusts in other rocks, and they include a set of major and minor the CROP-11 deep seismic profi le highlights orogens, this model may be a viable mecha- curved thrusts (Royden et al., 1987; Ghisetti and the presence of a thick midcrustal anticlinorium nism for arc formation elsewhere. Vezzani, 1997; Macedo and Marshak, 1999). (Billi et al., 2006). This structure is located be- It is widely accepted that the development of neath the Olevano-Antrodoco thrust, which is †E-mail: [email protected] the greater external arcs (i.e., presently buried the southern limb of the Umbrian Arc (Fig. 1C). GSA Bulletin; September/October 2009; v. 121; no. 9/10; p. 1409–1420; doi: 10.1130/B26335.1; 7 fi gures; Data Repository item 2009093. For permission to copy, cont act [email protected] 1409 © 2009 Geological Society of America Downloaded from gsabulletin.gsapubs.org on 30 September 2009 Billi and Tiberti A 10°E 15°E N Alps B Adriatic Sea 50 km northern arc 250 km AAA 44°N nnoorrtthheerrnn ooouuuttteeedddrrr rrrAAAiiiaaapppttteeeiiicccnnn nnnfffoooiiinnnrrreeelllaaannndddN CnnooorrttrhhseeiPcrrnnao AAPppleeanninnniinneessUmAbrrcian DDiinnaarriiddeess 45°N Tuscany AAppeennnniinneess eee fffrrrooonnnttt Figure 1RCome ssoouutthhAAddrriiAAaattppiiccuu llffiiooaarreellaanndd Umbrian Sardinia eerrnn 40°N Arc Tyrrhenian AApp 43°N backarc basin eenn Elba Is. nnii nn Olevano- Roveto eess Sicily Antrodoco Valley Ionian Tyrrhenian thrust Sea Sea Gran Tell volcanic and plutonic Sasso Malta rocks (Neogene-Quaternary) thrust Maiella southern postorogenic basinal sedimentary Latium tAhtrlausst 10°E 15°E arc rocks (Neogene-Quaternary) foredeep sedimentary rocks Fucino (Pliocene) Rome foredeep sedimentary rocks (Messinian) Tuscan succession Simbruini Marsica Cervarola flysch (Mesozoic-Paleogene) thrust (Oligocene-Miocene) basement rocks thrust Epi-Ligurian units (Triassic-Paleozoic) (Eocene-Miocene) main thrusts Ligurian units (Mesozoic-Paleogene) outer Apennines Volsci Sabina carbonates front thrust (Mesozoic-Miocene) tilt-corrected paleomagnetic Latium carbonates (Mesozoic-Miocene) 11°E Figure 1C 13°E declinations ((CCB)) Figure 5 40 km OO TTiibbee lleevvaa GGrraann ffoorr Adriatic Sea rr nn SS ee vvaalllleeyyW oo--AAnnttrrooddoo RVaolvleeytoFucinaaossssoo tthh..ddeeeeppMMaaiieellllaa ffoorreellaanndd cc MM tt oo tthh SSii E basin oorrrroo hh.. N hhiinntteerrllaanndd ROME .. mmbbrruuiinnii tthhVV..aalllleellooMMnnaaggrraass iittcchhaa.. tthh..nnee tthh.. VV Tyrrhenian Sea oollssccii tt hh.. volcanic deposits synorogenic deposits carbonate rocks (Pleistocene) (Tortonian-Pliocene) ((TTrriiaass-sMici-oMcieoncee)ne) marine and continental clastic Molise pelagic basin thrusts deposits (Pliocene-Quaternary) (Oligocene-Miocene) normal faults conglomerates on Simbruini th. carbonate and arenaceous (Messinian) turbidites (Cretaceous-Eocene) W E CROP-11 profile 1410 Geological Society of America Bulletin, September/October 2009 Downloaded from gsabulletin.gsapubs.org on 30 September 2009 Arc development in the Apennines Figure 1. (A) Digital elevation model of Italy and surrounding areas. Main Cenozoic fold-and- mum of ~22 km in the Tyrrhenian side of the thrust belts are shown. The northern and southern arcs of the Apennine fold-and-thrust belt belt to a maximum of almost 50 km in the axial are indicated. The main curved structure of the northern arc is the Umbrian Arc (Eldredge sector (Barchi et al., 1998; Cassinis et al., et al., 1985). The study area is located in the central Apennines and includes the N-S–trending 2003; Billi et al., 2006; Mele et al., 2006; southern limb of the Umbrian Arc, also known as the Olevano-Antrodoco thrust. (B) Geo- Di Luzio et al., 2009). logical map of the northern Apennines. The Cervarola fl ysch mainly consists of sandstones and shales. Epi-Ligurian units are mainly sandstones and marls derived from a pristine, Central Apennines c ontinent-ocean margin. Ligurian units are ophiolites and sedimentary and low-grade meta- morphic rocks derived from a pristine, oceanic basin. The Sabina carbonaceous sequence Major thrust sheets in the central Apennines includes transitional carbonates originally located between the Latium platform carbonates (i.e., Volsci, Simbruini, Marsica, Morrone, Gran (central Italy) and the Umbrian pelagic carbonates (northern Apennines). The Sabina se- Sasso, and Maiella thrust sheets; Fig. 1C) are quence includes a thick sequence of platform limestones (Calcare Massiccio Formation) at mostly NW-striking, NE-verging structures the base of the succession. The Latium carbonates are mostly platform limestones and dolo- (Parotto and Praturlon, 1975; Vezzani and stones. The Tuscan succession consists of a pile of carbonates, marls, shales, evaporites, and Ghisetti , 1993). Dimensions of the exposed sandstones deposited in different environments from Mesozoic to Paleogene times. (C) Geo- portion of these thrust sheets are between ~40 logical map of the central Apennines; “th.” stands for thrust sheet. Arrows are tilt-corrected and 120 km along-strike and ~20–30 km across- paleomagnetic declinations (Mattei et al., 1998). The CROP-11 seismic profi le (W–E) cuts strike. Most of these structures are thrust sys- across the southern limb (i.e., the Olevano-Antrodoco thrust) of the Umbrian Arc. tems consisting of several imbricate major and minor thrusts. For instance, the Marsica thrust system includes several thrusts, of which the westernmost Vallelonga thrust is a NW-striking , Our hypothesis is that the anticlinorium and deformation (Malinverno and Ryan, 1986; Roy- NE-verging structure located to the west of the the associated crustal thickening caused a den et al., 1987; Dewey et al., 1989; Faccenna Fucino Basin (Fig. 1C). We refer to the sub- significant surface uplift, possibly consti- et al., 2004; Rosenbaum and Lister, 2004). surface prolongation of this structure in our tuting an obstacle to the migration of the The Apennines are characterized by major analysis of the CROP-11 profi le to infer the age Olevano-Antrodoco thrust. We used gravi- NW-striking thrust sheets generally dipping of midcrustal deformation (Fig. 3). metric and seismic-refl ection data to determine toward the southwest with gentle angles and a Toward the west, the Olevano-Antrodoco attitude and spatial extent of the anticlinorium vergence toward the northeast (Fig. 1). Thrust thrust is the N-trending southern limb of the and its geometric relationship with the Olevano- imbrication occurred mostly in a forelandward Umbrian Arc (Fig. 1B). Geometries, kine- Antrodoco thrust. We then combined our re- piggyback sequence with some out-of-sequence matic indicators, and stratigraphic relationships sults with available paleomagnetic, strati- or backward thrusting episodes (e.g., Ghisetti along and over this fault show its contractional graphic, structural, and topographic data to and Vezzani, 1997; Cavinato and DeCelles, nature and reverse displacements (Salvini understand the infl uence of the anticlinorium 1999; Patacca et al., 2008) (Fig. 2). The thrust- and Vittori, 1982; Corrado, 1995; Ghisetti and on the development of the southern limb of ing style of the Apennine belt has been for Vezzani, 1997). In addition to reverse displace- the Umbrian Arc. Based on this evidence, we years the subject of contrasting interpreta- ments, however, some kinematic indicators on argue that our hypothesis of a causal relation- tions, including, in particular, thin-skinned and the O levano-Antrodoco thrust zone and strati- ship between midcrustal folding and orogenic thick-skinned styles (e.g., Ghisetti et al., 1993; graphic evidence show that this structure ac- curvature in the central Apennines is viable; Mazzoli et al., 2000). Because of the paucity of commodated younger right-lateral strike-slip more evidence is, however, needed to fully test subsurface data, in most sectors of the Apennine displacements during late Neogene time, possi- our hypothesis. Additionally, due to the occur- chain, it is still unclear if thin-skinned or thick- bly during early Pliocene time (Castellarin et al., rence of midcrustal basement-involved thrusts skinned thrusting is the most appropriate model 1978; Salvini and Vittori, 1982). in other orogens, this model may be a viable for structural style (e.g., Mazzoli et al., 2008; The Olevano-Antrodoco thrust marks an mechanism for arc formation elsewhere. Steckler et al., 2008). important surface lithologic transition. In the Normal faults and associated extensional hanging wall, Mesozoic transitional carbonates GEOLOGICAL SETTING basins of Miocene–Pleistocene age are wide- and marls are the dominant lithology (Sabina spread in the Tyrrhenian side of the Apennines transitional carbonates in Fig. 1B), with some Regional Setting and also in the axial sector of the fold-and-thrust exceptions, such as the Rocca di Cave shelf belt (Fig. 1B) (Malinverno and Ryan, 1986; (Accordi and Carbone, 1986). In contrast, the Within the framework of Alpine-Himalayan Barchi et al., 1998; Jolivet et al., 1998; Cavi- footwall mostly consists of Mesozoic platform orogenesis, the Apennine fold-and-thrust belt nato et al., 2002). Through time, the locus of carbonates (Latium platform carbonates in developed mostly during Neogene time as a extension has progressively migrated toward the Fig. 1B; Parotto and Praturlon, 1975; Accordi consequence of tectonic convergence between east (Fig. 2), parallel but west of the eastward- and Carbone, 1986). The thickness of the Sabina the European and African (i.e., Nubia) plates migrating locus of contractional deformation and Latium carbonates is ~3 and 5 km, respec- (Fig. 1). The parallel migration of the trench and (Malinverno and Ryan, 1986; Patacca et al., tively (Parotto and Praturlon, 1975; Accordi orogenic wedge toward the east and southeast 1992). The lag time between the onset of thrust- and Carbone, 1986). The contrasting thick- occurred concurrently with westward and north- ing and initial extension at any given locality ness and rigidity between the soft marly rocks westward subduction of oceanic lithosphere in the central Apennines is ~2–4 Ma (Fig. 2) (Sabina carbonates) to the west of the Olevano- beneath the European plate and with the pro- (Cavinato and DeCelles, 1999). Antrodoco thrust and hard platform carbon- gressive involvement of the Adriatic (African Seismic data across the Apennines show ates (Latium carbonates) to the east have been affi nity) continental margin with contractional that the crust thickness increases from a mini- considered the main cause for formation of the Geological Society of America Bulletin, September/October 2009 1411 Downloaded from gsabulletin.gsapubs.org on 30 September 2009 Billi and Tiberti Figure 2. Time-distance dia- gram, including thrusting, out-of-sequence thrusting, and postorogenic basin sedimenta- tion in the central Apennines (modifi ed after Cavinato and DeCelles, 1999). The diagram refers to a SW-NE transect across the central Apennines, including all main thrusts and extensional basins. The dia- gram shows that the central Apennines mainly evolved by in-sequence thrusting. Some late out-of-sequence thrusts also occurred. In each locality, the onset of extensional basin sedimentation is the temporal upper limit for contractional tectonics. The inferred age for the midcrustal anticlinorium (Fig. 3B) is shown. Umbrian Arc (e.g., Calamita and Deiana, 1988; Synchronous thrusting also occurred (e.g., temporally and spatially successive foredeeps Ghisetti and Vezzani, 1997). However, it should Marsica and Morrone thrusts). Out-of-sequence and thrust-top basins (Fig. 1C), indicates that be noted that the Sabina carbonates are bound, thrusting is documented for the Gran Sasso and the late M essinian–early Pliocene activity of the at their base, by the thick and rigid Calcare Mas- Olevano-Antrodoco thrusts (Cipollari et al., 1993; Olevano-Antrodoco thrust occurred out-of- siccio Formation (i.e., Jurassic platform carbon- Ghisetti and Vezzani, 1997; Satolli et al., 2005). sequence (Cipollari and Cosentino, 1995; ates), which controls the deformation pattern In particular, during late Messinian–early Plio- Mattei et al., 1995). Some authors infer addi- of the entire succession (Coward et al., 1999). cene time, the Olevano-Antrodoco thrust was tional activity of the Olevano-Antrodoco thrust From exploratory well data, it is known that the emplaced over late Messinian, siliciclastic, during early-middle Messinian time (Cipollari Calcare Massiccio Formation is ~0.8 km thick foredeep deposits (i.e., fl ysch) exposed to the and Cosentino, 1992; Cipollari et al., 1993; (Anelli et al., 1994). north of the Simbruini thrust (Cipollari et al., Cavinato and DeCelles, 1999). In contrast, 1993; Cipollari, 1995). The age of thrusting based on the age of foredeep deposits presently Thrust Timing and Vertical-Axis Rotations (Fig. 2), along with the truncation relation- exposed to the east and west of the Olevano- ship with earlier adjacent thrusts, and with Antrodoco thrust, Mattei et al. (1995) argue that In the central Apennines, detailed strati- graphic analyses of syntectonic sedimentary rocks that fi ll temporally and spatially succes- sive foredeeps and thrust-top basins (Patacca Figure 3. (A) Central segment of the CROP-11 seismic-refl ection profi le. See the related et al., 1992; Cipollari and Cosentino, 1995; track (W-E) in Figure 1C. TWTT is two-way traveltime. The entire CROP-11 profi le from Patacca and Scandone, 2001) constrain the the Tyrrhenian Sea to the Adriatic Sea and the related acquisition and processing param- timing of thrust evolution (e.g., Cavinato and eters are available in Billi et al. (2006). (B) Line drawing and interpretation of the CROP-11 DeCelles, 1999; Cosentino et al., 2003). The profi le displayed in A. The midcrustal anticlinorium (i.e., indicated as “hanging wall”) in- cessation of thrusting in each locality is marked volves refl ections between ~7 and 9 s (i.e., the midcrustal shear zone) and the topographic by the onset of continental sedimentation driven surface in the footwall of the Olevano-Antrodoco thrust. (C) Enlargement (left) and related by postorogenic extensional tectonics. Figure 2 interpretation (right) of the sector of the CROP-11 profi le, including the near-surface por- presents a synoptic diagram of results from pre- tion of the Olevano-Antrodoco thrust, which is interpreted as a shallow, low-angle, reverse vious studies showing that the central Apennines structure. (D) Gravity cross section along the segment of the CROP-11 profi le shown in A. mostly grew by in-sequence thrusting between The shaded area is the effect (i.e., negative) induced by the tectonic duplication (i.e., the late Tortonian (Volsci thrust) and early Plio- anticlinorium) shown in the CROP-11 profi le. Without this structure, the gravity cross sec- cene (Gran Sasso and Maiella thrusts) times. tion would run along the upper (dotted) line. 1412 Geological Society of America Bulletin, September/October 2009 Downloaded from gsabulletin.gsapubs.org on 30 September 2009 Arc development in the Apennines Geological Society of America Bulletin, September/October 2009 1413 Downloaded from gsabulletin.gsapubs.org on 30 September 2009 Billi and Tiberti the Olevano-Antrodoco thrust emplaced no ear- the seismic image. The weak, ramp-fl at geom- were removed during the stripping-off proce- lier than late Messinian time. etry (consisting of near-horizontal fl ats and low- dure, are known from ~60 logs of hydrocarbon In and around the southern sector of the Um- angle ramps) of the Olevano-Antrodoco thrust wells drilled in the central Apennines, and from brian Arc (Fig. 1C), three main paleomagnetic seems only partially parallel to the geometry several previously published studies on the sub- domains are recognized (Mattei et al., 1995, of the underlying anticlinorium. In particular, surface geology of this region (e.g., Bally et al., 1998): (1) the Sabina region (i.e., the Olevano- the hinge between the crest and the backlimb 1988; Anelli et al., 1994; Butler et al., 2004). In Antrodoco thrust sheet), which rotated clock- of the anticlinorium is not coincident with particular, density data used in this paper are wise by ~16° (95% confidence half-angle, the hinge between the ramp and fl at segments mainly after Mostardini and Merlini (1986). α = 11.5°) after early-middle Miocene time of the O levano-Antrodoco thrust. Moreover, These data were integrated with other published 95 (Mattei et al., 1995) but has shown no sig- the interl imb angle of the anticlinorium is information (see Table DR11). nifi cant rotations since middle Pliocene time smaller than the angle between the fl at and The Tyrrhenian and Adriatic domains are (Sagnotti et al., 1994); (2) the Roveto Val- ramp segments of the Olevano-Antrodoco characterized by gravity highs, whereas a rela- ley, which rotated counterclockwise by ~28° thrust. These geometric relationships sug- tive gravity low occurs in the axial sector of (α = 10.5°) during post-Messinian times; and gest that the Olevano-Antrodoco thrust the Apennine fold-and-thrust belt (Fig. 4B). The 95 (3) the Tuscan-Latium (i.e., Tyrrhenian side postdates the anticlinorium; however, some gravity low is ascribed to the regional deepening of the Apennines) neoautochthonous basins parall elism between the anticlinorium crest and of both the Moho and the top of the crystalline (N eogene-Quaternary), which have been af- the O levano-Antrodoco thrust main fl at may basement (Tiberti and Orlando, 2006). fected by nonrotational deformation (Sagnotti represent evidence that the Olevano-Antrodoco In the Olevano-Antrodoco and Simbruini et al., 1994; Mattei et al., 1998). thrust was affected by some folding connected thrust areas (Fig. 1C), a second-order gravity with anticlinorium growth. low affects the Bouguer and regional grav- CROP-11 SEISMIC-REFLECTION To the west of and beneath the Fucino Basin, ity anomalies (Figs. 3D, 4A, and 4B). This PROFILE some shallow refl ections (shallower than 2 s second-order gravity low consists of a negative TWTT) located east of the midcrustal anticlino- variation of ~10 mGal (Fig. 3D), and it suggests The CROP-11 deep seismic-refl ection pro- rium are parallel to its forelimb (Fig. 3B). Such the presence of a relatively low-density body fi le was planned and acquired across the central a geometric relationship suggests that these in the middle crust. Provided that the effects of Apennines (Fig. 1C) to image the crustal-scale shallow refl ections were involved in the anti- all the shallower bodies have been properly re- tectonic architecture of the junction between clinorium-related folding and, therefore, that the moved, the 40 km wavelength of the 10 mGal the northern (i.e., Umbrian Arc) and southern anticlinorium postdates the shallow thrust sheets gravity low is consistent with a source depth be- orogenic arcs. Parameters of acquisition and located immediately to the east. The loca tion tween ~10 and 20 km, since, at greater depth, a processing of the CROP-11 profi le were provided of the E-dipping shallow refl ections suggests density contrast of about ±100 kg/m3 would in a previous paper (Billi et al., 2006). Time-to- that they represent the eastward subsurface pro- affect an area broader than 40 km. depth conversion of the CROP-11 seismic profi le longa tion of the Marsica thrust sheet (i.e., the To defi ne the areal extent of the 10 mGal grav- was not attempted because detailed velocities of bedding panels forming the Vallelonga thrust ity low in the Olevano-Antrodoco and Sim bruini P-waves in the study area were not appropriately sheet; Fig. 3B), the age of which is late Mes- thrusts area, the tips (i.e., the lateral closures) of known at the time of data processing. sinian to early Pliocene (Figs. 1C and 2). The the second-order gravity low were searched in 70 The central segment of the CROP-11 profi le overall E-dipping attitude (i.e., by ~20°) of the gravity cross sections arranged on a grid covering (Fig. 3) shows the core of the orogenic wedge, exposed portion of the Vallelonga thrust sheet the study area, including the Olevano-Antrodoco where strong refl ections occur between ~5 and (Servizio Geologico d’Italia, 1967; Vezzani and and the Simbruini thrust sheets (Fig. 1C). One of 8–9 s two-way traveltime (TWTT). These re- Ghisetti, 1993) supports the hypothesis of a link- these cross sections (i.e., the one coincident with fl ections outline a wide and thick anticlinorium age between the exposed Vallelonga thrust sheet the CROP-11 profi le) is shown in Figure 3D. that is interpreted as a fold developed above and the E-dipping shallow refl ections imaged in The obtained tips were then plotted on the map a middle to lower crust shear zone occurring Figure 3B to the west of and beneath the Fucino of Figure 5 and joined with a closed line (i.e., between ~7 and 9 s TWTT with variable dip Basin (see also Patacca et al., 2008). the dashed line encompassing the shaded area angles. From seismic-refraction data (Cassinis in Fig. 5). The resulting area is roughly elliptical et al., 2003), we obtain that the shear zone is as ANALYSIS OF GRAVITY DATA and elongated to the N-S, and it is ~50 km long deep as ~20–22 km (i.e., corresponding to ~9 s by 30 km wide. Its N-S–trending long axis ap- TWTT). The anticlinorium is characterized by We analyzed the regional gravity data of proximately coincides with the surface trace of two hinge zones imaged by two sets of upward- the central Apennines to determine the grav- the Olevano-Antrodoco thrust. convex refl ections (Fig. 3B). The vertical com- ity signature of the midcrustal anticlinorium Assuming that the second-order gravity low ponent of displacement on the basal shear zone, (Fig. 3B) and to infer its areal (i.e., map-view) imaged in the gravity cross section (Fig. 3D) as estimated on the CROP-11 profi le, is ~2 s extent (Fig. 4). The gravity data set was ob- and in the maps of the Bouguer and regional TWTT, corresponding to ~4–5 km. tained through a stripping-off procedure (sensu gravity anomalies (Figs. 4A and 4B) is related In the CROP-11 profi le, the near-surface por- Hammer, 1963), which consisted of removing to the anticlinorium (Fig. 3B), we applied again tion of the Olevano-Antrodoco thrust is imaged the effect of all geological bodies located in the stripping-off procedure, which consisted, as a low-angle, shallow structure dipping toward the upper crust from the Bouguer anomaly data this time, of calculating and removing from the the west and resting above the crest and back- (Carrozzo et al., 1991; Fig. 4A). The stripping- limb of the midcrustal anticlinorium (Figs. 3B off procedure applied to obtain the map shown 1GSA Data Repository item 2009093, Includes a table with rock densities used for the stripping-off pro- and 3C). The geometric relationship between the in Figure 4B is thoroughly explained in Tiberti cedure and the color version of Figure 4, is available Olevano-Antrodoco thrust and the underlying et al. (2005). The geometry and density of the at http://www.geosociety.org/pubs/ft2009.htm or by anticlinorium appears as not straightforward in shallow bodies, the gravity effects of which request to [email protected]. 1414 Geological Society of America Bulletin, September/October 2009 Downloaded from gsabulletin.gsapubs.org on 30 September 2009 Arc development in the Apennines Bouguer anomaly map (Fig. 4A) the effect of a geological body with the same geometric and geologic characteristics (i.e., location, depth, and shape) as the anticlinorium. The areal ex- tent of the structure was inferred approximately from the seismic section (Fig. 3B) and from the anomalous shape of the gravity isolines (see in- set in Fig. 4B), and the maximum overall thick- ness of the structure was fi xed at ~8 km from the seismic profi le (Fig. 3B), dropping progres- sively toward the lateral closures of the structure as drawn in Figure 3B. By an iterative proce- dure, we found that the second-order gravity low could be best compensated by assigning a density of 2570 kg/m3 to the rocks forming the anticlinorium. Such a density, for rocks lying at the depth of the midcrustal anticlinorium, is Reduction density: 2670 kg/m3 consistent, for instance, with low-grade meta- morphic rocks such as argillites or some kinds of phyllites. The occurrence of fl uids within these rocks may have reduced their density and increased their seismic refl ectivity. The reli- ability of the data used to model the anticlino- rium is shown by the result of the stripping-off procedure displayed in Figure 4C, where the second-order gravity low is almost completely absent (i.e., compare insets in Figs. 4A, 4B, and 4C) and the main gravity anomalies become ap- proximately linear and aligned with the NW-SE regional structural trend. DISCUSSION The rotational origin of the Olevano- Antrodoco thrust is demonstrated by a posi- tive orocline test and by paleomagnetic data Reduction density: 2670 kg/m3 (Fig. 1B), which show a clockwise rotation of ~16° between early-middle Miocene and middle Pliocene times (Sagnotti et al., 1994; Mattei et al., 1995, 1998; Speranza et al., 1997). Paleomagnetic measurements are from seven sites on the Prenestini Mountains, which form the hanging wall of the Olevano-Antrodoco thrust. We think that more paleomagnetic data are necessary to better constrain the rotation of the Olevano-Antrodoco thrust sheet and reduce the error connected with past measurements (Mattei et al., 1995). Analysis of the CROP-11 seismic image (Fig. 3A) shows the presence of a thick dome structure related to folding in the middle crust (i.e., the midcrustal anticlinorium) right be- neath the Olevano-Antrodoco thrust (Fig. 3B). For the part visible in the CROP-11 profi le, the Reduction density: 2670 kg/m3 Olevano-Antrodoco thrust is imaged as a shal- low thin-skinned thrust sheet, the basal thrust of which emerges above the crest region of the Figure 4. Gravity maps of central Italy. (A) Bouguer anomaly map. (B) Regional gravity anticlinorium (Figs. 3C and 5). Due to the lack map. (C) Regional gravity map after removal (stripping-off procedure) of the modeled mid- of appropriate subsurface data, the subsurface crustal anticlinorium shown in Figure 3B. The color version of this fi gure is available in the geometry of the Olevano-Antrodoco thrust was GSA Data Repository (Fig. DR1 [see text footnote 1]). Geological Society of America Bulletin, September/October 2009 1415 Downloaded from gsabulletin.gsapubs.org on 30 September 2009 Billi and Tiberti 12°30′E 13° 13°30′ 14° ′0 2 ° 2 4 Figure 5. Schematic tectonic map of central Italy (location of this fi gure is indicated in Fig. 1C); “th.” stands for thrust sheet. The shaded area represents the gravity anomaly induced by tectonic duplication associated with the anticlinorium imaged in the CROP-11 profile (Fig. 3B). This area is interpreted ° to represent a N-trending midcrustal anti- 2 4 clinorium related to displacement on a basal shear zone lying at the middle to lower crust level (Fig. 3B). The shaded area is drawn by joining points representing the projection on the map of the lateral tips of the gravity N anomaly connected with the anticlinorium ′0 and observed on 70 gravity cross sections 0 ° 1 (see one of these cross sections in Fig. 3D). 4 As such, this area represents the approxi- mate areal extent (i.e., where the gravity signature induced by the tectonic duplica- tion is suffi ciently marked to be detected) of the midcrustal anticlinorium. Paleomag- netic data are from Sagnotti et al. (1994) and Mattei et al. (1995). mostly unknown before the acquisition of the geometric relationship between the anticlino- thermore, the geometric relationship between CROP-11 profi le. We were able to compensate rium and the exposed or shallow thrusts, the the Olevano-Antrodoco thrust and the anticlino- for the gravity anomalies observed in both cross ages of which are known from previous stud- rium (Fig. 3B) is consistent with the inferred section (Fig. 3D) and map (Fig. 4) views with ies (Fig. 2). In particular, in the CROP-11 image age of anticlinorium growth (i.e., early Pliocene a geological body similar to that observed in (Fig. 3B), the Vallelonga thrust sheet is parallel time). The Olevano-Antrodoco thrust, the age of the CROP-11 profi le at the midcrustal level and to the forelimb of the underlying anticlinorium. which is late Messinian–early Pliocene (Fig. 2), characterized by a horizontal dimension of ~50 This relationship suggests that the Vallelonga seems as partly involved in the midcrustal fold- by 30 km. The long axis of this structure trends thrust, which has an age of late Messinian–very ing. The hypothesized tectonic interaction be- approximately N-S (Fig. 5). By combining early Pliocene (i.e., see the age of the Marsica tween the Olevano-Antrodoco thrust and the these results with the seismic-refl ection image, thrust in Fig. 2), was involved in the growth of underlying anticlinorium likely took place dur- we interpret the near-elliptical geological body the anticlinorium. It follows that the age of the ing the latest phase of thrust activity (i.e., during detected by the gravity analysis as a N-trending anticlinorium should be early Pliocene (Fig. 2). early Pliocene time). anticlinorium related to contractional displace- In middle Pliocene time, contractional deforma- To verify whether the Olevano-Antrodoco ment on a basal shear zone lying at a middle to tion was mostly inactive across the presently thrust is folded, we analyzed the elevation lower crust level (Figs. 3 and 5). exposed portion of the central Apennines, and pattern of the emerging thrust (A-B and C-D The timing of anticlinorium development can normal faulting was already active at least in the cross sections in Fig. 6B). The longitudinal be inferred, at least in part, by analyzing the inner and axial sectors of the belt (Fig. 2). Fur- topographic profi le of the Olevano-Antrodoco 1416 Geological Society of America Bulletin, September/October 2009 Downloaded from gsabulletin.gsapubs.org on 30 September 2009 Arc development in the Apennines A D ttGG hh rr rraa uu nn ss OO tt SS ll aa ee ss vv ss aa oo nn oo -- AA B nn trtr C oo dd oo MM oo cc rr oo Fucino basin oo nn t t nn hh ee ruru tthhrr ss uu tt sstt VV MM Rome aallll aarr SSii eell ssii mm oo cc 42°N Albani Hills A bbrruuiinnii tthhrruusstt RRoovveettoo nnVVggaaaallll eetthhrruussttaa tthhrruusstt yy Volcano VV ooll ss N ccii tthh rr uu sstt Tyrrhenian Sea 10 km 12°E 13°E marine and Pliocene- main Messinian Tortonian- continental Plio- Pleistocene Pliocene conglomerates Pliocene Quaternary deposits volcanic cover conglomerates flysch B Umbrian Latium Triassic main main tracks of Meso-Cenozoic Mesozoic- carbonates normal thrusts topographic carbonates Cenozoic faults A cross sections carbonates B 2000 A B C D 1500 m) 1000 ht ( g 500 ei h 5 km 0 m S N S N Figure 6. (A) Geological map of central Apennines. The tracks of topographic cross sections shown in B are displayed. The fl ysch deposits (i.e., mainly sandstones with shale intercalations) are undifferentiated and indicated as Tortonian-Pliocene in age. Because of the north- eastward progression of the fold-and-thrust belt, fl ysch deposits are younger toward the northeast. In the Roveto Valley, fl ysch deposits are Messinian in age and are overlain by the Olevano-Antrodoco thrust, the latest age of which is, therefore, post-Messinian (i.e., early Pliocene) (Cipollari and Cosentino, 2002). (B) Projection of the altitude of the emergence of the Olevano-Antrodoco thrust on the A–B and C–D tracks. The vertical scale is greatly exaggerated. Geological Society of America Bulletin, September/October 2009 1417