Volume 25, Issue 3
Free Access

Magnetic and structural constraints for the noncylindrical evolution of a continental forebulge (Hyblea, Italy)

Andrea Billi

Andrea Billi

Dipartimento di Scienze Geologiche, Università “Roma Tre”, Roma, Italy

Search for more papers by this author
Massimiliano Porreca

Massimiliano Porreca

Dipartimento di Scienze Geologiche, Università “Roma Tre”, Roma, Italy

Search for more papers by this author
Claudio Faccenna

Claudio Faccenna

Dipartimento di Scienze Geologiche, Università “Roma Tre”, Roma, Italy

Search for more papers by this author
Massimo Mattei

Massimo Mattei

Dipartimento di Scienze Geologiche, Università “Roma Tre”, Roma, Italy

Search for more papers by this author
First published: 01 June 2006
Citations: 20


[1] The Hyblean Plateau in southeastern Sicily, Italy, consists of an isolated and elevated forebulge structure formed since the early Miocene time by bending the foreland lithosphere below the advancing Maghrebian thrust-fold belt. The Hyblean forebulge is presently located in front of an orogenic recess and partly surrounded by two orogenic salients. We analyzed magnetic (anisotropy of magnetic susceptibility) and structural (joints, faults, slickenside lineations, and bed attitudes) fabric data from Neogene carbonate rocks exposed atop the Hyblean forebulge. Results showed unidirectional fabrics developed in response to a NW-SE extension at the end the of early Miocene time and beginning of middle Miocene time, and duple-to-multiple fabrics developed in response to major NW-SE and NE-SW extensions from the Langhian time onward. We interpreted these results as the evidence for the growth of a doubly plunging forebulge due to the presence of foreland crustal heterogeneities, which enhanced differential retreating processes of the foreland along the subduction zone and the subsequent formation of orogenic salients and recesses. A semianalytical model shows that the observed brittle deformations potentially involved a significant thickness of the carbonate sedimentary cover, thereby reducing the lithosphere flexural rigidity and enhancing its bending aptitude.

1. Introduction

[2] The Hyblean Plateau (or Hyblea) in southeastern Sicily, Italy, is the northernmost edge of the continental African foreland in the western Mediterranean (Figure 1) [Patacca et al., 1979]. Toward the northwest, the Hyblean foreland is thrust by the Maghrebian thrust-fold belt, which grew within the Africa-Eurasia suture belt as a south and SE verging accretionary wedge during Neogene-Quaternary times [Caire, 1970]. The Hyblean Plateau is presently a structurally elevated area representing the modern peripheral bulge (or forebulge) [Pedley and Grasso, 1992]. Isopachs of Miocene-Pliocene marine deposits in the Hyblean Plateau and adjacent areas show that the Hyblean forebulge developed since at least the mid-Miocene time with a faint ENE trending hinge zone [Grasso and Pedley, 1990], parallel to the Maghrebian thrust front and perpendicular to the tectonic transport of the Maghrebian belt in the southeastern Sicily [Grasso et al., 1995]. Lower Pleistocene shallow water sediments at elevation in excess of 600 m above sea level (asl) [Schmincke et al., 1997] suggest that the flexural uplift might have endured until the mid-Quaternary time at least. Although the subsurface stratigraphic data suggest that the Hyblean forebulge grew principally as an ENE trending foreland monocline [Grasso and Pedley, 1990], tectonic data [Barrier, 1992; Sirovich and Pettenati, 1999] as well as computer-aided simulations of the collisional process in Sicily [Ben-Avraham et al., 1995] suggest a noncylindrical tectonic evolution for this foreland region. In particular, the structural architecture of Hyblea is characterized by multiple trends of extensional and strike-slip faults (Figure 2), whose kinematics is only in part consistent with a NNW-SSE extension as expected on the outer arc of an ENE trending cylindrical forebulge [Turcotte and Schubert, 1982]. Moreover, fault plane solutions of shallow earthquakes registered in the Hyblean region between 1977 and 2002 (Figure 1b) show both strike-slip and extensional kinematics [Goes et al., 2004]. The strike-slip solutions are consistent with the kinematics of the regional faults in this area [Grasso et al., 2000], whereas the extensional solutions are mostly compatible with a NE-SW maximum stretching axis [see also Musumeci et al., 2005], which is apparently discordant with an ENE trending cylindrical forebulge. Also, in southeastern Sicily, the strain tensor calculated from the GPS velocity data is characterized by a horizontal NW-SE stretching axis [Hollenstein et al., 2003], whereas the strain deduced from borehole breakout data is characterized by a NNW trending maximum contractional axis (i.e., SHmax [Ragg et al., 1999]).

Details are in the caption following the image
(a) Schematic tectonic map of the Mediterranean region displaying major thrust-fold and strike-slip belts. (b) Schematic tectonic map of Sicily and adjacent areas. Abbreviations are ext. faults, extensional faults; s.-s. faults, strike-slip faults. The rectangular box shows the location for Figure 2. Fault plane solutions of seven representative shallow earthquakes (≤25 km deep) in the Hyblean region are represented as beach balls. Fault plane solution 1 is from Amato et al. [1995]. Fault plane solutions 2, 3, 4, 5, 6, and 7 are from Goes et al. [2004]. (c) Schematic geological profile across the Hyblean forebulge, the foredeep, and the front of the Maghrebian fold-thrust belt [Carbone et al., 1982]. See the profile track in Figure 1b.
Details are in the caption following the image
(a) Schematic geological map of the Hyblean foreland [after Pedley, 1981; Grasso and Lentini, 1982, 1984]. See location in Figure 1. Abbreviations are ext. faults, extensional faults; s.-s. faults, strike-slip faults; Fm., formation. Two asterisks indicate the approximate locations of the Gela (offshore) and Ragusa (onshore) oil fields in fractured carbonate rocks at depths <6 km. (b) Schematic summary of the Miocene-Pliocene stratigraphy (i.e., referring only to the analyzed formations) of the Hyblean Plateau (modified after Pedley [1981] and Grasso and Lentini [1982] from the data from Romeo and Sciuto [1987] and Courme and Mascle [1988]).

[3] The aim of this paper is to reconstruct the Neogene-Quaternary tectonic evolution of the Hyblean forebulge, with particular focus on the flexural geometry and kinematics. For this purpose, we analyzed AMS (i.e., anisotropy of magnetic susceptibility) and structural data collected from rock exposures atop the Hyblean Plateau. These data led to the understanding of the noncylindrical evolution of the foreland flexure. The potential repercussions of the observed flexure-related deformations on the bending aptitude of continental forelands are analyzed and discussed.

2. Geological Setting

[4] In the western Mediterranean area (Figure 1), the foreland compartment that includes the Hyblean Plateau and the Maltese Islands has crustal properties different from those of the surrounding foreland regions [Ben-Avraham et al., 1995]. Toward the east, Hyblea is offshore delimited by the Malta Escarpment, a fault-controlled submarine slope [Grasso et al., 1990], which separates the thick and continental Hyblean-Maltese crust from the thinned oceanic Ionian Sea basin [Cernobori et al., 1996]. Toward the west and southwest, the Hyblean-Maltese region is bounded by the Pelagian Sea block, an offshore foreland compartment characterized by a continental crust less thick than that occurring in the adjacent Hyblean-Maltese compartment [Ben-Avraham and Grasso, 1991]. This crustal heterogeneity strongly influenced the Africa-Eurasia convergent processes in Sicily and surrounding regions [Ben-Avraham et al., 1995]. As a result, salients and recesses developed along the front of the Maghrebian belt. In particular, a recess developed in front of the thick and buoyant Hyblean block that resisted subduction, whereas two salients developed in correspondence of the adjacent Pelagian and Ionian foreland compartments (Figure 1), where lithosphere properties eased subduction and thrust migration toward the south and southeast. Evidence of foreland crustal heterogeneity in Sicily and in its offshore areas is also provided by the heterogeneous flexural behavior of the foreland lithosphere as shown by the heterogeneous dip values of the foreland monocline beneath the Maghrebian belt [Mariotti and Doglioni, 2000].

[5] The Hyblean Plateau (Figure 2) consists of a succession of ∼6 km [Agocs, 1959] of Mesozoic-Cenozoic carbonate rocks with interleaved marls and mafic volcanic levels [Patacca et al., 1979]. The exposed rocks are mostly carbonates and marls of Oligocene to Miocene age [Grasso and Lentini, 1982]. These rocks are in general subhorizontal (i.e., commonly dipping <5°). Three main families of faults affect the Hyblean foreland (Figure 2): (1) a set of NE striking extensional faults truncates the Hyblean Plateau toward the northwest and constitutes the tectonic boundary between the forebulge and the foredeep domains [Monaco et al., 2003]. The activity of this set of faults occurred during late Miocene–Quaternary times [Grasso et al., 1990; Monaco et al., 2003]; (2) toward the west and southwest, a set of NNE striking dextral strike-slip faults (Scicli Fault system) cuts through the Hyblean foreland. The activity of the Scicli Fault system is dated back to the early Pleistocene time and its cumulated strike-slip displacement is assessed as not exceeding 200–300 m over a fault length of more than 50 km [Grasso and Reuther, 1988]; and (3) toward the east, a set of NW and NNW striking extensional faults downthrows the Hyblean Plateau toward the eastern offshore area and marks the transition with the Ionian Sea oceanic basin [Grasso, 1993; Bianca et al., 1999]. A down-to-the-east throw of more than 2 km is assessed along the Malta Escarpment. The throw increases toward the north where the Ionian plate bends and plunges beneath the Calabria [Grasso, 1993; Argnani and Bonazzi, 2005]. The extensional faulting in the eastern Hyblean Plateau is mostly Quaternary. On seismic cross sections, some of these faults bear evidence of recent displacements and are seismically active [Bianca et al., 1999].

3. Methods and Results

3.1. Magnetic Data

[6] In 16 sampling sites (Table 1) located in the topographically highest sector of the Hyblean Plateau, we collected a total of 165 oriented cores from Neogene gray marls and marly limestones interleaved in the Hyblean succession of pelagic-to-platform limestones [Grasso, 1999]. By referring to the stratigraphic subdivisions of Pedley [1981] and Grasso and Lentini [1982] (i.e., as reviewed by Spezzaferri et al. [2001]), as mapped by Grasso and Lentini [1984] and Grasso [1999], we sampled rock exposures from the following formations (Table 1 and Figure 2b): (1) the upper part of the Ragusa Formation (i.e., Irminio Member, lower-middle Miocene), (2) the Tellaro Formation (middle-upper Miocene), and (3) the Palazzolo Formation (upper Miocene, including Messinian according to Romeo and Sciuto [1987], to lower Pliocene according to Courme and Mascle [1988] and Courme [1991]). Each sampling site consisted in a rock exposure a few square meters in size characterized by weak tectonic deformations. No faults and exposure-scale folds were observed on the exposures selected for the core sampling. The number of cores at each site varied between 8 and 12 (Table 1). From each core, one to four standard cylindrical (25 mm diameter × 22 mm high) specimens were obtained for the measurement of the low-field anisotropy of magnetic susceptibility (AMS).

Table 1. Geographical and Geological Data for the AMS Sample Sites
Site Latitude N Longitude E Lithology Formation Age Number of Cores
PL01 37°03′01″ 14°52′27″ marly limestone Tellaro middle-upper Miocene 12
PL02 37°03′18″ 14°49′26″ marly limestone Ragusa Lower Miocene 10
PL03 37°03′02″ 14°50′49″ marly limestone Tellaro middle-upper Miocene 11
PL04 37°02′38″ 14°51′16″ marly limestone Ragusa lower Miocene 10
PL05 36°58′54″ 14°52′46″ marly limestone Ragusa lower Miocene 12
PL06 36°59′06″ 14°50′13″ marly limestone Ragusa. lower Miocene 9
PL07 37°02′37″ 14°47′46″ marly limestone Ragusa. lower Miocene 11
PL08 36°58′55″ 14°49′29″ marly limestone Ragusa lower Miocene 11
PL09 37°02′42″ 14°53′18″ marly limestone Palazzolo upper Miocene–lower Pliocene 10
PL10 37°04′45″ 14°53′15″ marly limestone Palazzolo upper Miocene–lower Pliocene 12
PL11 37°06′42″ 14°56′48″ marly limestone Palazzolo upper Miocene–lower Pliocene 9
PL12 37°06′27″ 14°55′02″ marly limestone Palazzolo upper Miocene–lower Pliocene 10
PL13 37°05′14″ 14°46′18″ marly limestone Ragusa lower Miocene 9
PL14 37°08′42″ 14°45′14″ marly limestone Tellaro middle-upper Miocene 8
PL15 37°05′43″ 14°47′28″ marly limestone Palazzolo upper Miocene–lower Pliocene 10
PL16 37°04′04″ 14°54′10″ marly limestone Palazzolo upper Miocene–lower Pliocene 11

[7] AMS of rocks can be defined as a second rank symmetric tensor geometrically represented by an ellipsoid with the principal axes Kmax > Kint > Kmin. The shape of the AMS ellipsoids and the degree of anisotropy can provide information on the fabric of rocks in relation to their sedimentary [e.g., Clark, 1970] and tectonic [e.g., Oertel, 1983] histories either in contractional [e.g., Kissel et al., 1986] or extensional [e.g., Mattei et al., 1997] environments. In the case of a magnetic fabric induced by tectonic processes, the maximum axis of susceptibility (Kmax) commonly becomes perpendicular to the direction of maximum shortening in contractional environments [e.g., Borradaile and Tarling, 1981], whereas it aligns along the direction of maximum stretching in extensional environments [e.g., Mattei et al., 1997]. In extensional environments, it is well documented that the AMS of rocks can be imparted during the very early stages of deformation [e.g., Cifelli et al., 2004].

[8] The shape of the AMS ellipsoids and the degree of anisotropy are described by a set of parameters and associated statistical limits explained and listed in Table 2. Values for the shape parameter (T) show that the analyzed specimens are characterized by an oblate AMS ellipsoid; however, the anisotropy degree (P′) is low and included between 1.006 and 1.022 (Table 2). These low P′ values are typical of weakly deformed sedimentary rocks.

Table 2. List of the AMS Parameters for Each Sample Sitea
Site Km (SD) L (SD) F (SD) P′ (SD) T (SD) D, I, (E1–2) Kmax D, I, (E2–3) Kmin Magnetic Lineation
PL01 39.8 (3.2) 1.001 (0.001) 1.004 (0.001) 1.006 (0.001) 0.481 (0.246) 95.1, 5.3, (23) 224.6, 81.7, (6) well-defined
PL02 17.3 (3.4) 1.002 (0.001) 1.012 (0.002) 1.015 (0.002) 0.767 (0.075) 311.5, 2.7, (23) 82.0, 85.8, (5) well-defined
PL03 18.5 (0.6) 1.003 (0.001) 1.014 (0.002) 1.018 (0.002) 0.669 (0.153) 47.9, 5.5, (24) 224.8, 84.5, (3) well-defined
PL04 13.7 (1.4) 1.004 (0.003) 1.008 (0.002) 1.012 (0.003) 0.314 (0.241) 337.6, 3.4, (51) 91.8, 81.9, (15) undefined
PL05 9.4 (1.0) 1.004 (0.002) 1.009 (0.002) 1.013 (0.003) 0.347 (0.209) 340.5, 1.6, (50) 244.7, 74.3, (8) undefined
PL06 21.7 (1.0) 1.002 (0.001) 1.010 (0.002) 1.014 (0.001) 0.684 (0.128) 3.8, 2.2, (12) 222.0, 87.1, (7) well-defined
PL07 23.3 (1.4) 1.002 (0.001) 1.016 (0.002) 1.019 (0.002) 0.797 (0.098) 159.8, 3.9, (40) 271.7, 79.7, (10) poorly defined
PL08 15.5 (1.0) 1.004 (0.002) 1.011 (0.005) 1.016 (0.005) 0.466 (0.299) 11.1, 3.7, (17) 254.4, 81.8, (6) well-defined
PL09 34.7 (6.3) 1.003 (0.002) 1.012 (0.003) 1.017 (0.003) 0.538 (0.258) 293.1, 1.4, (20) 96.8, 88.5, (5) well-defined
PL10 24.8 (1.3) 1.004 (0.002) 1.009 (0.002) 1.013 (0.002) 0.406 (0.198) 304.1, 3.3, (53) 119.9, 86.7, (12) undefined
PL11 16.8 (4.9) 1.004 (0.001) 1.013 (0.005) 1.018 (0.005) 0.442 (0.255) 111.2, 2.6, (15) 296.6, 87.4, (14) well-defined
PL12 19.0 (0.7) 1.003 (0.001) 1.017 (0.003) 1.022 (0.003) 0.721 (0.128) 111.8, 1.8, (58) 230.6, 86.2, (9) undefined
PL13 26.6 (4.6) 1.002 (0.001) 1.009 (0.001) 1.012 (0.001) 0.580 (0.180) 153.9, 5.1, (31) 264.1, 75.6, (10) poorly defined
PL14 37.6 (3.7) 1.002 (0.002) 1.010 (0.003) 1.014 (0.002) 0.629 (0.335) 284.6, 1.4, (35) 88.0, 88.5, (12) poorly defined
PL15 21.8 (1.8) 1.003 (0.001) 1.010 (0.002) 1.014 (0.003) 0.542 (0.108) 37.3, 4.2, (17) 290.4, 75.9, (5) well-defined
PL16 33.2 (4.2) 1.001 (0.001) 1.006 (0.001) 1.008 (0.001) 0.667 (0.160) 169.9, 3.7, (51) 281.0, 79.7, (7) undefined
  • a Km (in 10−6 SI) is the bulk susceptibility and provides a relative estimation of the amount of magnetic minerals in the sample. Km = (Kmax + Kint + Kmin)/3. L = Kmax/Kint. F = Kint/Kmin. P′ is the corrected anisotropy degree. P′ = exp {2[(η1 − ηm)2 + (η2 − ηm)2 + (η3 − ηm)2]}1/2, where η1 = ln Kmax, η2 = ln Kint, η3 = ln Kmin, and ηm = (η1 η2 η3)1/3. T is the shape parameter of the AMS ellipsoid. T = (2 η2 − η1 − η3)/(η1 − η3). T varies between −1 (i.e., perfectly prolate ellipsoid, Kmax ≫ Kint and Kmin) and +1 (i.e., perfectly oblate ellipsoid, Kmax and Kint ≫ Kmin). T = 0 corresponds to a triaxial (or neutral) ellipsoid [Jelinek, 1981; Hrouda, 1982; Tarling and Hrouda, 1993]. D and I are azimuth and inclination, respectively, in degrees for Kmax and Kmin. E1–2 is the confidence angle in degrees for the Kmax axis in the Kmax-Kint plane. E2–3 is the confidence angle in degrees for the Kmin axis in the Kint-Kmin plane. SD is the standard deviation. The parameter values correspond to the arithmetic means.

[9] By using the Jelinek's [1977] statistics, we determined the average AMS value for each site (Table 2) from the AMS values measured in the site specimens. The magnetic lineation (Figure 3 and Table 2) is well defined in eight sites (i.e., where the E1–2 confidence angle of Kmax in the Kmax-Kint plane is ≤30°) and poorly defined in three sites (i.e., where the E1–2 angle is >30° and ≤40°). The remaining five sites have an oblate fabric with virtually no lineation (the E1–2 angle is ≥50°). Kmin is commonly perpendicular to bedding and nearly vertical (Figure 3). Consequently, Kmax and Kint are nearly horizontal. The distribution of the entire population of the Kmax azimuths is characterized by a diffuse pattern with major clusters along the following trends (Figure 4a): WNW-ESE, NW-SE, N-S, and NE-SW. By sorting the Kmax azimuths by the stratigraphic formation and hence by age (see Table 1), different patterns arise (Figures 4b, 4c, and 4d): predominant north trending and NW trending Kmax azimuths for the Ragusa Formation (lower Miocene); predominant WNW trending and NE trending Kmax azimuths for the Tellaro Formation (middle-upper Miocene); and a diffuse pattern of Kmax azimuths with major clusters along the WNW-ESE and NE-SW trends for the Palazzolo Formation (upper Miocene–lower Pliocene).

Details are in the caption following the image
Location of AMS sample sites and related lower hemisphere equal-area plots of AMS data and statistics. Note that plots are only those referring to the sites with the E1–2 confidence angle of Kmax in the Kmax-Kint plane ≤40° (see Table 2).
Details are in the caption following the image
Rose diagrams showing histograms of the Kmax azimuths as measured in single specimens. (a) Total data; (b) data from the Palazzolo Formation (upper Miocene–lower Pliocene time); (c) data from the Tellaro Formation (middle-upper Miocene time); and (d) data from the Ragusa Formation (lower Miocene time).

[10] To compare the magnetic and structural fabrics site by site, in each AMS sampling site, we collected the attitude of joints affecting the competent carbonate beds. The site-by-site comparison between the AMS and the joint data (Figure 5) shows that the azimuth of the mean Kmax and the mean azimuth of joints are about parallel in one site (Figure 5a) and about perpendicular in six sites (Figures 5c, 5d, 5e, 5g, 5j, and 5k). In the sites where two distinct sets of joints are present (Figures 5b, 5h, and 5i), one of the joint sets strikes perpendicularly to the azimuth of the mean Kmax. In particular, we observed that in these three sites (PL02, PL11, and PL13), the azimuth of the mean Kmax trends NW-SE and the joint sets strike preferentially NE-SW and NW-SE. We also observed that in two of these three sites (i.e., PL02 and PL13 sites in the Ragusa Formation) the NW striking set of joints abuts against the NE striking one, showing that the NW striking set is younger than the NE striking one [e.g., Pollard and Aydin, 1988]. Eventually, in one site, the azimuth of the mean Kmax and the mean azimuth of joints intersect at an angle of ∼48° (Figure 5f).

Details are in the caption following the image
(a)–(k) Site-by-site comparison between AMS and joint attitude data. (left) Lower hemisphere equal–area plots of AMS data and statistics. (right) Contours (lower hemisphere equal-area plots) to joint poles sampled in the corresponding AMS sites. Note that the AMS plots are only those referring to the sites with the E1–2 confidence angle of Kmax in the Kmax-Kint plane ≤40° (Table 2).

[11] In the above discussed specimens, the natural remanent magnetization (NRM) at room temperature was also measured to eventually assess block rotations in the Hyblean forebulge. This analysis provided very low values for the intensity of magnetization (less than 4 × 10−5 A/m), this result hindering any further paleomagnetic determination; however, it is known from previous paleomagnetic analyses that no significant rotations with respect to Africa occurred in the Hyblean forebulge since the late Cretaceous time [Grasso et al., 1983; Besse et al., 1984].

3.2. Structural Data

[12] We collected data about the geometry, kinematics, and crosscutting relationships of joints, faults, and carbonate strata in 129 measurement sites distributed over the Hyblean Plateau (Figure 6). The studied exposures locate within the following formations: the Palazzolo Formation, the Tellaro Formation, the Monti Climiti Formation, and the Ragusa Formation (Figure 2b). To separate the effect of the regional tectonics (e.g., the lithosphere flexure and the associated flexural uplift) from local processes (e.g., fault-related deformations), we analyzed the rock structural fabric in sites characterized by near-horizontal strata and located far away from the regional faults (e.g., the Scicli Fault system, Figure 2) and their associated structures as mapped by Grasso and Lentini [1982] and Grasso [1999] (Figure 6).

Details are in the caption following the image
Location of selected structural sampling sites and related plots showing contours to joint poles (lower hemisphere equal-area plots).

[13] Faults observed in the visited exposures are high-angle (≥55°), low-displacement (throw <10 cm), extensional and oblique-extensional structures striking preferentially NE-SW and NNW-SSE. These faults are usually isolated and generate null or very limited downbending of strata.

[14] Joints (Figure 7) are the most ubiquitous tectonic features in the Hyblean Plateau. The observed joints consist of subvertical surfaces with opening displacement in the 0.001–0.1 m range. Joint spacing is between a minimum of ∼0.01 m and a maximum of ∼3 m as depending on the bed thickness. We recognized two major joint sets whose preferential strikes are NE-SW and NW-SE. Both these sets are usually present in the visited exposures (e.g., Figures 6 and 7) and in all the analyzed formations. The two sets of joints intersect perpendicularly or at high angles (“T” intersections [Pollard and Aydin, 1988]) according to two types of crosscutting relationships: (1) a first type of relationship, in which the NW striking set abuts against the NE striking one (Figure 7b), and (2) a second type of relationship, in which the NW striking and the NE striking sets mutually abut (Figures 7c and 7d). We observed the first type of crosscutting relationship in the Ragusa Formation and, occasionally, in the Monti Climiti Formation (i.e., in the lower and middle portion of this formation), whereas we observed the second type of crosscutting relationship in the Monti Climiti (i.e., upper portion), Tellaro, and Palazzolo formations, and, sporadically, in the Ragusa Formation and in the lower-middle portion of the Monti Climiti Formation. Joints and bedding surfaces concur to form a network of interconnected mechanical discontinuities. These discontinuities transformed the pristinely continuous carbonate strata into an assemblage of quasi-regular polyhedral fragments (i.e., nearly orthorhombic and parallelepiped-like, Figures 7e and 7f). The linear size of these fragments is commonly in the 0.1–1 m range (Figure 7), but they may occasionally be a few centimeters in size (i.e., 2–4 cm, Figure 7f).

Details are in the caption following the image
Photographs of jointed carbonates from the Hyblean Plateau. See Figure 6 for exposure locations. (a) Plumose structures over a joint surface, showing the dilational origin of the joint [e.g., Pollard and Aydin, 1988]. (b) Photograph (map view) showing that the NW striking joints abut against (i.e., postdate) the NE striking ones. Note the “T” intersections. (c) Photograph (map view) showing complex crosscutting relationships between the NW striking and the NE striking sets of joints, which mutually abut (i.e., synchronous growths). (d) Line drawing of fractures after Figure 7c. (e) and (f) Photographs (map views) showing surfaces of fractured carbonate strata from the Palazzolo Formation. Note that the pristinely continuous carbonate bed is entirely transformed into an assemblage of rock polyhedral (i.e., rectangular in cross-sectional view) fragments. In this site, the fragments are at their minimum observed size.

[15] Although in each site we discerned major sets of systematic fractures (i.e., joints and faults), the synoptic diagrams of the collected structural data (Figure 8) show that the observed structures are weakly systematic at the scale of the Hyblean foreland. In particular, azimuths of both joints (Figure 8a), strata (Figure 8b), faults (Figure 8c), and fault slickenside lineations (or slickenlines) (Figure 8d) are dispersed except for clusters about the following trends: the NW-SE and NE-SW trends for the joint azimuths (Figure 8a); the WNW-ESE and NE-SW trends for the carbonate bed azimuths (Figure 8b); the NE-SW and NNW-SSE trends for the fault azimuths (Figure 8c); and the NE-SW and NW-SE trends for the fault slickenline azimuths (Figure 8d).

Details are in the caption following the image
(a) Lower hemisphere equal-area plot showing contours to poles of all the sampled joints. Black indicates the maximum density. Shadings indicate major ranges of joint azimuths and show the weakly systematic nature of the sampled joints at the scale of the Hyblean Plateau. (b) Rose diagram of all sampled azimuths of strata. Note the weakly systematic pattern. (c) Lower hemisphere equal-area plot showing contours to poles of all sampled faults (i.e., extensional faults). Black indicates the maximum density. Shadings indicate major ranges of fault azimuths. (d) Lower hemisphere equal-area plot showing contours to all sampled fault slickenside lineations (extensional). Black indicates the maximum density. Shadings indicate major ranges of fault heave azimuths and show the weakly systematic distribution of data.

4. Discussion

4.1. Development of Magnetic and Structural Fabrics

[16] We henceforth present a model to test whether the Hyblean flexure is at the origin of a stress field compatible with the development of the observed magnetic and structural fabrics. By assuming a simple elastic model of flexure, in which the foreland lithosphere is characterized by a constant flexural rigidity and is approximated to a semi-infinite thin elastic plate subject to a linear end load, the flexure-related fiber stress (σxx) atop the forebulge is
equation image
where Mb is the flexural moment in the forebulge and Te is the effective elastic thickness of the lithosphere [Turcotte and Schubert, 1982]; σxx is the stress that develops, when the lithosphere bends, as perpendicular to the flexure axis and parallel to the flexed fibers (or layers). σxx is tensile in the flexure outer arc (i.e., in the forebulge). Mb is
equation image
where D is the flexural rigidity of the plate, wb is the vertical upward deflection of the forebulge, and (xbx0) is the horizontal half width of the forebulge.

[17] Cogan et al. [1989] developed the flexural model for the Hyblean foreland along four NW-SE cross sections and found the following flexural parameters: D = 1021 N m, Te = 8000 m, wb = 1000 m, (xbx0) = 30,000 m. By using these parameters for solving equation (1) via equation (2), we obtained an estimate of about −128 MPa (i.e., tensile stress) for the theoretical fiber stress (σxx) atop the Hyblean forebulge. Such a tensile stress is far greater in module than the tensile strength that commonly characterizes carbonate rocks under ambient conditions (i.e., usually between ∼4 and ∼20 MPa [e.g., Paterson, 1978]), and hence it is compatible with the generation of the NE striking set of joints observed atop the Hyblean forebulge. Because it is demonstrated that an extension-related AMS of rocks can develop even in the absence of evident brittle deformations [e.g., Cifelli et al., 2004], the theoretical fiber stress computed for the Hyblean forebulge is even more so compatible with the development of an extension-related AMS of rocks with NW trending Kmax axes.

[18] Analogously to the NE striking joints and the NW trending magnetic lineations, we interpret the occurrence of NW striking joints and NE trending magnetic lineations over the entire Hyblean forebulge as the result of a secondary bending of the Hyblean foreland about a NW trending axis (i.e., perpendicular to the major flexural axis). The attitude of strata with preferential NW-SE and NE-SW dip azimuths (Figure 8b) provides evidence for the doubly plunging shape of the Hyblean forebulge. Such doubly plunging shape is also consistent with the fault slip data (Figure 8d), which indicate the NE-SW and NW-SE trends as the major directions of extension.

[19] The joint crosscutting relationships and the age of the jointed rocks allow to reconstruct schematically the temporal evolution of the joint network. The NE striking set of joints nucleated and partially grew during the Miocene time by the mid-Langhian time (i.e., the transition between the Ragusa Formation and the Tellaro Formation, Figure 2b). From the mid-Langhian time onward, the NW striking set of joints nucleated and grew simultaneously with the NE striking ones until at least the Pliocene time (i.e., corresponding to the top of the Palazzolo Formation, Figure 2b). The temporal evolution of the structural fabric is consistent with that of the magnetic fabric. Whereas, in fact, the Kmax axes trend prevailingly NNW-SSE in the Ragusa Formation (Figure 4d), in the Palazzolo and Tellaro formations, the Kmax axes trend prevailingly NE-SW and WNW-ESE (Figures 4b and 4c). This fabric can be explained by an early, NNW-SSE, stretching tectonics (i.e., involving the Ragusa Formation, lower-middle Miocene time), and a subsequent doubly verging one with the NE-SW and WNW-ESE major stretching axes (i.e., involving the Tellaro and Palazzolo formations, middle Miocene–Pliocene time). The dependence of the measured magnetic fabrics on the extensional tectonics is demonstrated by the site-by-site comparison between magnetic and structural fabrics (Figure 5). This analysis shows that the Kmax axis is usually perpendicular to the average strike of joints and hence parallel to their opening displacement. Where two sets of joints occur, the Kmax axis is perpendicular to the early set as inferred from the joint crosscutting relationships, confirming that the AMS of rocks is strongly dependent on the early rock strain [e.g., Cifelli et al., 2004]. The parallelism between the Kmax axis and the average joint azimuth in the PL01 site (i.e., Tellaro Formation, Figure 5a) can be explained by two deformational phases undergone by the studied rocks. During the first phase, rocks underwent an extensional tectonics with a WNW trending stretching axis. This process imparted a tectonic magnetic fabric in the analyzed rocks of the Tellaro Formation, but it was not sufficiently effective (at least locally) to generate a set of joints. The WNW striking joints were generated in response to a NNE trending tensile stress during a subsequent tectonic phase, which imparted no further magnetic anisotropy in the analyzed rocks. An analogous polyphase tectonics can be hypothesized also for the PL08 site (Figure 5f), where we found a NNE trending Kmax axis and a NW trending set of joints.

4.2. Evolution of the Hyblean Forebulge

[20] The results obtained from the analysis of magnetic and structural fabrics can be integrated in the following evolutionary model (Figure 9) for the development and growth of the Hyblean forebulge within the evolution of the SE migrating Africa-Eurasia subduction zone in the western Mediterranean [Faccenna et al., 2004].

Details are in the caption following the image
Evolutionary model for the Hyblean forebulge within the framework of the evolution of the Africa-Eurasia subduction zone in the western Mediterranean (Figures 9a, 9b, and 9c are modified after Faccenna et al. [2004]). (a) Model at more than 23 Ma (Paleogene time). (b) Model between 23 and 16 Ma (early Miocene time). (c) Model at the present time. (d) Three-dimensional cartoon schematically showing the doubly plunging shape of the Hyblean forebulge at the present time (fold drawing is modified after Tavani et al. [2006]). The observed joint network and magnetic lineations are drawn atop the proper formations.

[21] By ∼23 Ma (Paleogene time), the Africa-Eurasia subduction zone extended from southern Iberia to the Ligurian region across the modern central Tyrrhenian Sea (Figure 9a). At this stage, the load exerted on the lithosphere by the growing orogenic wedge and that exerted by the slab pull were probably too far away from Hyblea to produce its deformation and flexure.

[22] At the end of early Miocene time and beginning of middle Miocene time (∼15–16 Ma, Langhian-Burdigalian time), the subduction zone was located in the modern southern Tyrrhenian Sea (Figure 9b). At that time, the orogenic load and that exerted by the slab pull were probably located sufficiently near to induce extensional deformations and the initiation of the Hyblean flexure about a NE trending axis. Such early cylindrical process of flexure generated unidirectional magnetic and structural fabrics in the rocks located atop the growing forebulge (i.e., lower Miocene sediments, Ragusa and Monti Climiti formations). The flexural uplift of Hyblea is proved, since at least the Tortonian time (∼11 Ma), by the deposition around the modern forebulge of shallow marine sediments with typical syntectonic fan patterns [Grasso and Pedley, 1990; Pedley and Grasso, 1992].

[23] From ∼11 Ma (Tortonian time) onward (Figure 9c), the subduction zone has kept migrating toward the south and southeast with different slab retreating velocities as controlled by the heterogeneous lithosphere properties [Mariotti and Doglioni, 2000]. The most evident result of such heterogeneous retreating process is that since the middle Miocene time, the Hyblean foreland, which resisted subduction, has been progressively surrounded by advancing orogenic salients (Figure 9c). The growing salients have progressively exerted additional loads on the southwestern and northeastern sides of the Hyblean forebulge, thus producing a secondary flexure of the foreland about a NW trending axis. Such noncylindrical evolution of the forebulge generated the observed orthogonal fabrics. Strike-slip and extensional faulting across the Hyblean forebulge (i.e., along the Malta Escarpment and the Scicli Fault system [Grasso and Reuther, 1988; Grasso, 1993; Argnani and Bonazzi, 2005]) have probably accommodated part of the differential retreating kinematics of the foreland monocline and part of the extensional strain induced by the doubly plunging flexure. The doubly plunging flexure may also explain the apparently contradictory strain fields derived from earthquake (Figure 1b) [Goes et al., 2004], GPS velocity [Hollenstein et al., 2003], and borehole breakout [Ragg et al., 1999] data.

4.3. Implications for the Lithosphere Aptitude for Bending

[24] We henceforth present a semianalytical model to estimate whether the flexure-related jointing may have involved a significant thickness of the Hyblean sedimentary cover, thereby increasing the lithosphere vulnerability to bending.

[25] In the Hyblean foreland, the large difference between the carbonate strength and the theoretical, flexure-related, fiber stress (i.e., −128 MPa) implies that part of this stress must have been relieved by some inelastic deformation, including the fracturing of the brittle portion of the lithosphere, consistently with the strength of rocks under confining pressures. Among other factors, in fact, the strength of rocks in the subsurface depends on the confining pressure. The overburden-related confining pressure (pL) is
equation image
where ρ is the rock density, g is the gravitational body force, and z is the depth [Turcotte and Schubert, 1982]. By considering the experimental data provided by Schmidt and Huddle [1977] on the fracture toughness of the Indiana Limestone with increasing confining pressures (Table 3), we assessed ∼19 MPa and ∼95 MPa for the lower and upper bounds, respectively, of the tensile strength of limestones at a confining pressure of 62.1 MPa (see Appendix A), which is the highest confining pressure reached in the Schmidt and Huddle's [1977] experiments. By considering an average density (ρ) of 2700 kg m3 for the carbonate rocks, from equation (3), a confining pressure (pL) of 62.1 MPa corresponds to a depth (z) of ∼2350 m. By comparing the theoretical fiber stress estimated for the Hyblean forebulge (−128 MPa) with the estimated tensile strength of carbonate rocks at a depth of 2350 m (between ∼19 and ∼95 MPa), we infer that the Hyblean carbonate succession should be affected by joints up to a depth of at least ∼2350 m. Moreover, by assuming that the exponential law that relates the fracture toughness of Indiana Limestone with the confining pressure data [Schmidt and Huddle, 1977] can be extended to confining pressures greater than 62.1 MPa (see Appendix A), from equations (3) and (A1) (see equation (A1) in Appendix A), we assessed ∼2793 m and ∼5210 m for the lower and upper bounds, respectively, of the depth corresponding to 128 MPa of limestone tensile strength (i.e., equal to the module of the fiber stress estimated for the Hyblean forebulge). Even considering the tendency of the flexure-related tensile fiber stress to reduce in module to zero with depth (i.e., approaching the neutral midplane), from the above discussed estimates, we infer that the magnitude of the flexure-related fiber stress in the Hyblean forebulge is large enough to produce jointing of the carbonate strata up to a depth of a few thousands of meters at least. The occurrence of the Gela and Ragusa oil fields (Figure 2a) hosted in fractured Triassic-Liassic carbonates of the Hyblean foreland (i.e., lying at the base of the Hyblean carbonate succession, Mattavelli et al. [1993]) supports the hypothesis of flexure-related joints at depths of a few thousands of meters. The above estimates for the depth of jointed carbonates are oversimplifications that do not consider factors such as the pore pressure [e.g., Handin et al., 1963], the thermal regime [e.g., Steckler and ten Brink, 1986], the strain rate [Atkinson, 1987], and the fact that early rock deformations perturb the stress field and introduce heterogeneities that may influence subsequent deformations [e.g., Behn et al., 2002]. Moreover, the presented model is based on the rheology of the Indiana Limestone, whose relevance for the Hyblean carbonates is not tested. The point we wish to make, however, is that the flexure of continental foreland plates may significantly reduce the elastic properties of crustal rocks (see the rock deformational state in Figure 7), thus abating the lithosphere effective elastic thickness and increasing the vulnerability of the lithosphere to bending. Rock jointing may partially explain the low value of Te for the Hyblean foreland (i.e., ∼8 km [Cogan et al., 1989]). By comparing the estimated thickness of jointed carbonates (∼3–5 km) with the actual effective elastic thickness for the Hyblean forebulge (∼8 km), we obtain a hypothetical abatement of the effective elastic thickness caused by rock jointing between ∼25% and ∼40%.
Table 3. List of Fracture Toughness and Confining Pressure Data From Experiments by Schmidt and Huddle [1977] on Indiana Limestonea
Confining Pressure, MPa Fracture Toughness, MPa m1/2 Estimated tensile Strength, MPa
Lower Bound Upper Bound
0.0 0.93 4.00 20.00
6.9 1.00 4.75 23.77
20.7 1.55 6.72 33.58
34.5 2.10 9.49 47.43
48.3 3.05 13.40 66.99
62.1 4.20 18.92 94.62
  • a Relative estimates for the lower and upper bounds of tensile strength are listed.

5. Conclusions

[26] 1. The observed multidirectional magnetic and structural fabrics atop the Hyblean forebulge are consistent with a progressive doubly plunging flexure of this region during the Neogene-Quaternary.

[27] 2. A computational analysis, based on the general rheological properties of carbonate rocks with increasing confining pressures, demonstrates that the flexure-related tensile fiber stress is large enough in module to produce rock jointing in the Hyblean forebulge at depths on the order of 3–5 km.

[28] 3. The fragmentation status of the Hyblean carbonate strata, as observed atop the Hyblean forebulge and inferred at depth, justifies, at least partially, the low value for the effective elastic thickness of the Hyblean lithosphere. It follows that the massive fracturing process associated to the lithosphere flexure may have significantly enhanced the bending aptitude of the Hyblean foreland.


[30] A.B. thanks C. Monaco and L. Tortorici for an insightful discussion on the tectonics of the Hyblean foreland at the beginning of the fieldwork, and P. Favali, R. Funiciello, P. Meredith, C. Poloni, F. Salvini, F. Storti, and M. Tiberti for perceptive suggestions. F. Salvini is also thanked for kindly providing the Daisy 3.0 software. P. Casero is thanked for useful discussions and for indications regarding the stratigraphy of Hyblea. The Editor, Associate Editors, and reviewers of Tectonics provided insightful comments and kind assistance. This work was funded by a GNDT Project whose coordinator was L. Beranzoli.

    Appendix A:: Estimating the Tensile Strength of Limestone Under Confining Pressures

    [29] Schmidt and Huddle [1977] provided a set of experimental data about the fracture toughness of Indiana Limestone under increasing confining pressures (Table 3). These data are well fitted by the following exponential law:
    equation image
    where x is the confining pressure, y is the fracture toughness and e is the Neper's number. M and N are 0.0250272 and 0.895932, respectively. The statistical limits for the best fit equation are as follows: regression sum of squares = 1.81411; residual sum of squares = 0.00672394; coefficient of determination = 0.996307; and residual mean square = 0.00168099. In equation (A1), N is the approximation of the fracture toughness under ambient conditions. The fracture toughness (KIc, for mode I fractures) of a linearly elastic medium is equal to the critical value (i.e., at the fracture propagation) of the stress intensity factor (KI), which quantifies the intensity of the stress singularity at the fracture tip [Atkinson, 1987]. Unlike the tensile strength, the fracture toughness of a material can be measured under confining pressures [e.g., Schmidt and Huddle, 1977]. The Griffith's criterion, as explicated by Irwin [1958], relates the strength and the relative stress intensity factor through material and fracture parameters, which depends, under specific boundary conditions, upon the fracture tip geometry and the loading geometry [e.g., Atkinson, 1987]. A simplified form for the Griffith's criterion is:
    equation image
    where Kc and σc are the fracture toughness and the strength, respectively, and β and c are material constants related to the fracture tip geometry and the loading geometry. Equation (A2) shows that for a given material (e.g., limestone), the fracture toughness and the strength are proportional. It follows that an exponential law similar to that expressed by equation (A1) is presumably applicable to the tensile strength of limestone with increasing confining pressures. Hence the tensile strength of limestone with increasing confining pressure can be estimated through equation (A1) (Table 3) by substituting the fracture toughness with the tensile strength (y), and giving N the value of 4 MPa, when computing the strength lower bound, and the value of 20 MPa, when computing the strength upper bound. The tensile strength of limestone under ambient conditions is, in fact, between a minimum of ∼4 MPa and a maximum of ∼20 MPa [Paterson, 1978].