Fault-related fold kinematics recorded by terrestrial growth strata, Sant Llorenç de Morunys, Pyrenees Mountains, NE Spain
James H. Carrigan
a,*, David J. Anastasio
a, Kenneth P. Kodama
a, Josep M. Par es
baEarth and Environmental Sciences Department, Lehigh University, 1 West Packer Ave, Bethlehem, PA, 18015, USA
bCentro National de Investigacion sobre la Evolucion Humana, Paseo Sierra de Atapuerca, 09002, Burgos, Spain
a r t i c l e i n f o
Article history:
Received 29 December 2015 Received in revised form 20 August 2016
Accepted 4 September 2016 Available online 8 September 2016
Keywords:
Fault-related-folding Foreland basin Cyclostratigraphy Magnetostratigraphy Pyrenees
a b s t r a c t
Foreland basin growth strata are ideal recorders of deformation rates and kinematics in tectonically active regions. This study develops a high-resolution chronostratigraphic age model to determine folding rates in the Eocene-Oligocene terrestrial growth strata of the Berga Conglomerate Group, NE Spain. The Berga Conglomerate Group was sampled for rock magnetic, magnetostratigraphic, and magnetic sus- ceptibility (c) cyclostratigraphy analyses. Analysis of rock magnetic measurements indicate a mixed mineral assemblage with both paramagnetic and ferromagnetic minerals. A new magnetic reversal stratigraphy constrains the time frame of folding and is in agreement with previous interpretations. Time series analysis ofcvariations show statistically significant power at expected orbital frequencies and provides precession-scale (20 kyr) temporal resolution. Strain measurements including anisotropy of magnetic susceptibility (AMS) fabrics and bedding plane strain worm burrow distortion are consistent withfixed hinge,flexural folding kinematics. Fault-related folding was modeled usingccyclostratigraphy timing and strain measurement kinematic constraints. The onset of folding was at 33.85 Ma and the end of deformation is less constrained but is younger than 31.06 Ma. Deformation and sediment accumu- lation rates are unsteady at 20 kyr time scales but appear artificially steady at polarity chron time scales.
©2016 Elsevier Ltd. All rights reserved.
1. Introduction
Terrestrial growth deposits in foreland basins provide a prox- imal record of tectonically active regions and are ideal for recov- ering fault slip at intermediate time scales. Because growth strata record both the interaction between fold evolution and changes in depositional environment (Fig. 1) separation of orogenic and sedi- mentary processes is important for geologic interpretations.
Despite their widespread occurrence, terrestrial systems are typi- cally devoid of biostratigraphic control which makes high- resolution timing of deformation difficult. Terrestrial deposition occurs in high energy systems, oxidizing environments, and with depositional hiatuses, all of which make fossil preservation un- likely. Magnetic polarity chronology has been used in foreland basins throughout the world (e.g.Burbank et al., 1992; Abdul Aziz et al., 2000; Fang et al., 2003; Homke et al., 2004) when other chronologic methods were unfeasible. Rock magnetic cyclo- stratigraphy coupled with magnetic polarity chronology provides a
robust method to date growth strata which can be used to constrain slip on buried and emergent faults. We combine kinematic con- straints and high-resolution timing to recover orogenic fold evo- lution at 104e105yr time scales.
Rock magnetic properties have been used as a proxy for high- resolution chronostratigraphy in many kinds of studies related to climate, deposition, and deformation (e.g.Latta et al., 2006; Spahn et al., 2013; Gunderson et al., 2013, 2014; Kodama and Hinnov, 2015). Cyclostratigraphy provided temporal resolution on the or- der of 104105 years in clastic and carbonate marine (e.g. Latta et al., 2006; Kodama et al., 2010; Hinnov et al., 2013; Gunderson et al., 2014) and terrestrialfluvial (Nador et al., 2003) deposits.
Foreland basins are ideal sites for cyclostratigraphy as they have sufficiently high sediment accumulation rates and ample accom- modation space to recover high frequency orbital signals. In this study, absolute time was determined by correlation with the global magnetic reversal chronology ofGradstein et al. (2012). Absolute time allows for calibration of a cyclostratigraphic section to Earth's known orbital parameters to remove the effect of variable sedimentation.
In order to determine kinematic controls on fault-related folding, oriented cores were analyzed using anisotropy of
*Corresponding author.
E-mail address:[email protected](J.H. Carrigan).
Contents lists available atScienceDirect
Journal of Structural Geology
j o u r n a l h o m e p a g e :w w w . e l s e v i e r . c o m / l o c a t e / j s g
http://dx.doi.org/10.1016/j.jsg.2016.09.003 0191-8141/©2016 Elsevier Ltd. All rights reserved.
magnetic susceptibility (AMS) and geologic strain analysis. The kinematics of the Sant Llorenç de Morunys structure were previ- ously debated (Ford et al., 1997; Suppe et al., 1997; Alonso et al., 2011) but never definitively resolved. The mechanisms proposed for the formation of the Sant Llorenç structure are: fold limb rotation (Ford et al., 1997), kink-band migration (Suppe et al., 1997), and a hybrid folding model with both limb rotation and hinge migration (Alonso et al., 2011). These models predict a different deformationfield for various structural positions that are testable with strain data. Fixed hinge folding (e.g.Ford et al., 1997) predicts higher strain in the hinge, whereas a migrating hinge model (e.g.
Suppe et al., 1997) predicts different limb strain histories as a function of distance from the hinge. In this paper we combine our finite strain and growth strata chronology to generate a consistent fault-propagation kinematic model for the Sant Llorenç de Morunys structure.
2. Geologic setting
This study focuses on the proximal Ebro foreland basin and the Pyrenees in northeastern Spain (Fig. 2a). The Pyrenean mountain front is typical of contractional orogens with multiple thrust faults and folds such as the Vallfogona thrust sheet and its leading edge imbricate. Fault-related folding created deformed syntectonic conglomerates during shortening (Puigdefabregas et al., 1992) in the proximal foreland basin. The basin deposits record a typical orogenic sequence of marine to terrestrial deposition as the Vall- fogona thrust sheet was emplaced (Puigdefabregas et al., 1992;
Williams et al., 1998). Previous work (Riba, 1976; Ford et al., 1997;
Suppe et al., 1997; Alonso et al., 2011) generally agrees that the folding is related to the propagation of a blind thrust but these
studies lack absolute time constraints on the initiation of thrusting and variations of deformation rate through time.
Originally defined bySole-Sugra~nes and Mascare~nas (1970)the Berga Conglomerate Group was described in detail byRiba (1976) for exposures along the Rio Cardener. Just south of the town of Sant Llorenç de Morunys, the Berga Conglomerate Group lies conformably on the marine marls and marly limestones assigned to the Bartonian-Lower Priabonian (Sole-Sugra~nes and Mascare~nas, 1970), which are laterally equivalent to the Igualada Formation further south. The Berga Conglomerate Group consists of con- glomerates, sandstones and siltstones of alluvial, fluvial, and lacustrine origins. Changes in drainage location, drainage area, and tectonic rates all influence the depositional environment through time. A stratigraphic column presented inFig. 2b shows changes in litho-facies through the sampled section primarily along road C- 462. Formation descriptions for this study are based on 1) presence of conglomerate, 2) bed thickness, and 3) facies association. For this study the Camps de Val-Llonga Formation consists of predomi- nantly sand and siltstones with occasional conglomerates not exceeding 50 m in maximum thickness, and the El Castell Forma- tion is characterized by two massive,>75 m, conglomerate beds separated by thin sandstones and siltstones. Other formations of the scheme ofFord et al. (1997)andWilliams et al. (1998)are used in this study without modification.
3. Methods
3.1. Magnetic mineralogy characterization
Unoriented samples were processed to determine the magnetic mineralogy of the Berga Conglomerate Group. Samples were used Fig. 1.Schematic block diagram showing the depositional environment during the formation of the Sant Llorenç de Morunys structure. The depositional environment is controlled by long term climate variations as well as the changing position of axial and longitudinal drainage of the region.
to identify relative contributions to magnetic remanence of different coercivity phases based on Isothermal Remanence Magnetization (IRM) acquisition modeling (Kruiver et al., 2001),
and relative influence of paramagnetic minerals incbased cyclo- stratigraphy by room and low temperature susceptibility analysis.
Seven samples, chosen for variations in grain size and Fig. 2.A) Geologic map of Sant Llorenç de Morunys structure and surrounding area. Inset map shows the location of the sampled region on the Iberian Peninsula. Map units are described in Section2and the line of cross section shown inFig. 6is indicated. B) Stratigraphic column for the Berga Conglomerate Formation sampled in this study. I) Formation abbreviations fromFig. 2a legend II) Lithologic variations follow the facies designations of (Barrier et al., 2010). III) Cyclostratigraphic and magnetostratigraphic sampling interval (Fig. 8). IV) Bedding inclination indicating growth and pregrowth sections. The grey region shows 2svariability in bedding orientation.
Fig. 2.(continued).
coloration, underwent detailed IRM acquisition up to 5 T in an IM- 10-30 impulse magnetizer (ASC Scientific, USA) at Lehigh Univer- sity. Each sample was crushed and sieved for grain sizes<2.5f. All samples were placed in afield free room for at least a week to allow any viscous magnetization obtained during processing to decay and to reduce exposure to any strong external magnetic field. IRMs were applied in 23 steps which were chosen to be as close as possible to uniformly spaced on a log scale. The coercivity components were modeled following Kruiver et al. (2001)'s software.
Room and low temperature bulk susceptibility was conducted for 10 additional specimens to determine the relative influence of paramagnetic, diamagnetic, and ferromagnetic minerals. Specimen cwas measured at room temperature (21C) using a KLY-3S sus- ceptibility meter (AGICO, Czech Republic). For low temperature measurements specimens were immersed in liquid nitrogen for 2 min to fully cool the sample to196C. Specimens were quickly removed from the liquid nitrogen and immediately measured in an AGICO KLY-3S susceptibility meter while still cold.
3.2. Oriented sampling strategy
Oriented cores were collected for magnetic reversal stratigraphy and anisotropy of magnetic susceptibility (AMS) analyses. Samples were collected using a water-cooled drill, nonmagnetic 25 mm drill bits, and orienting equipment. All sites had3 oriented cores and
>5 specimens in order to ensure statistical significance of our re- sults (Fisher, 1953).
3.2.1. Magnetic reversal stratigraphy
Samples were collected for the magnetic reversal stratigraphy during multiplefield seasons and sample spacing was refined after a preliminary magnetostratigraphy was constructed. Preliminary results suggested a minimum polarity interval of 290 kyr and an average deposition rate of 0.25 m/kyr. Therefore, a sampling in- terval of 15 m was selected to allow for multiple sites in the shortest polarity interval while maintaining enough spacing to make
>1.5 km sampling feasible. In total, 119 sites were analyzed.
Paleomagnetic measurements were made on a 755-4 K super- conducting rock magnetometer (2G Enterprises, USA) at CENIEH, Burgos, Spain. Thermal demagnetization was conducted with a TD- 48 SC Thermal Demagnetizer (ASC Scientific, USA) in up to 15 temperature steps from 150 C to 660 C. Alternating field demagnetization was used on a small number of samples using an in-line 170 mT AF (2G Enterprises) sample demagnetizer. Charac- teristic remanence magnetizations (ChRM) were completed for all specimens using principal component analysis (Kischvink, 1980).
Virtual Geomagnetic Pole (VGP) latitudes were used to determine local magnetostratigraphy.
3.2.2. Anisotropy of magnetic susceptibility (AMS) and geologic strain analysis
AMS was used as a proxy for rock strain to establish fold kine- matics (e.g. Hrouda, 1982, 1991; Borradaile, 1988; Pueyo-Morer et al., 1997). Samples for AMS analysis were collected from similar lithologies from all fold structural positions and consistency of measurements allowed us to consolidate results in domains from the northern steeply overturned limb of the fold, the hinge region of the fold, or the upright, shallowly dipping limb of the fold to the south. Samples were collected from siltstone for most locations (20/
24 sites) and where obtained in sandstone wherefiner lithologies were not present. Cores were analyzed in an AGICO KLY-3S sus- ceptibility meter at room temperature (21C) at Lehigh University, to determine the degree of anisotropy using 15 different orienta- tions to construct an anisotropy tensor. Using this method provides
the orientation and magnitude of the greatest, intermediate, and least principal susceptibility axes,k1,k2andk3. In total 241 speci- mens were analyzed (Supplemental Material 1) using methods based onTarling and Hrouda (1993). Only siltstone lithologies were compared (201 specimens) for detailed analysis to remove the ef- fect of differing grainsize.
To evaluate the AMS data with more traditional strain mea- surement, the aspect ratio of deformed Skolitos worm burrows were measured in bedding planes at sites where they were abun- dant (Fig. 2a). Because Skolitostubes are statistically circular in cross section and normal to bedding in their undeformed state (Hallam and Swett, 1966) they provide a good measure of bedding plane distortion. Kinematic constraints determined from AMS and geologic strain data were then used in forward models that allowed us to convert fold limb tilt to fault slip assuming the subsurfacial geometry of the fault-propagation fold.
3.3. Unoriented sampling strategy
Rock magnetic cyclostratigraphy samples span 800 m from pre- growth strata at the base to a large conglomerate interval near the top of the section (Fig. 2b). Rock magnetic samples had a spacing of 50e100 cm to ensure the 4e5 kyr spacing needed for accurate precession index (20 kyr) recovery. Sample size varied from 1 cm cores to large hand samples. Hand samples were crushed and sieved to<2.5fused for analysis and care was taken to prevent any crushing of framework clasts. The crushed sediment was packed into pre-weighed 8 cm3 nonmagnetic containers and stabilized with sodium silicate to prevent movement of sediment during analysis. The samples were analyzed for volume normalized bulkc at room temperature (21C) using an AGIGO KLY-3S susceptibility meter at Lehigh University. Magnetic susceptibility was chosen to develop a rock magnetic stratigraphic record because of prior success using this parameter in chronologic studies (e.g.Rafini and Mercier, 2002; Kodama et al., 2010; Gunderson et al., 2013; Spahn et al., 2013).
Any statistical outliers >5 s from the mean were discarded before further analysis, this represented<1% of samples. To further reduce inter-sample variability data were log10normalized before time series analysis. Gaps larger than 5 m (z20 kyr) werefilled using single spectrum analysis (SSA) utilizing the SSA-MTM toolkit (Ghil et al., 2002; Kondrashov and Ghil, 2006) to eliminate gap artifacts. The SSA toolkit fills gaps in the data series by using dominant observed frequencies, thus avoiding the introduction of artificial cycles. In total, 13% of thecdata is derived from SSA gap filling. The entire data series was used to better constrain the low frequency signals. After gapfilling, the data series was resampled at equal intervals using a linear interpolation every 15 cm.
The data series were analyzed using the multi-taper method (MTM) (Thomson, 1982) spectral analysis with 2pprolate multi- tapers to identify significant periodicities. The data series were tied to absolute time using the magnetic reversal chronology and chron boundary ages (Gradstein et al., 2012) and the astrochron package for R (Meyers, 2014). After the data series was correlated to time it was resampled in uniform spacing every 1 kyr. Power spectra were calculated using the SSA-MTM toolkit (Ghil et al., 2002). We iden- tified significant frequencies in the power spectra at the 90%, 95%, 99% confidence interval above a log fit robust red noise model following Mann and Lees (1996). After Milankovitch-scale fre- quencies were observed in the data, a detailed astrochronology was constructed by tuning the Gaussian-filtered time series to the theoretical obliquity signal (41 kyr) (Laskar et al., 2004) in thec data series.
4. Results
4.1. Magnetic mineralogy
IRM acquisition modeling shows either two or three ferromag- netic components in all specimens (Supplemental Material 2). One component has a low coercivity (~30 mT) while the two high coercivity components are identified around 1 T and>3 T, but are poorly determined. The low coercivity component ranges from 10 mT to 60 mT and becomes saturated around 300 mT. The in- termediate coercivity component ranges from 631 mT to 1584.9 mT and averages around 1 T; this component was fully saturated. The highest coercivity component ranges from 3162.3 mT to 5011.9 mT and was not fully saturated. All three ferromagnetic components contribute to theccyclostratigraphy.
Results from low temperature c experiments are graphed in Fig. 3a. All samples showed an increase in susceptibility at liquid nitrogen (77 K) temperatures between 170 and 260%. Relative fre- quency of room temperature bulkcof all AMS specimens is shown in Fig. 3b. Frequency distribution of bulk c shows a clustering around 1.1105(SI). In order to determine the relative strength of magnetite to paramagnetic minerals in bulk susceptibility mea- surements calculations following Pares (2004) were completed.
The average saturation magnetic intensity was 6.7 108 A/m which suggests a magnetite concentration of 0.4104% assuming a magnetite grain size of 1mm (O'Reilly, 1984). This corresponds to a volume susceptibility of 2106or≪1% of the average bulk sus- ceptibility of the samples in this study. Magnetite is used in this calculation to assess the relative strength of ferromagnetic vs paramagnetic material because it has the highest volume suscep- tibility of any ferromagnetic material by orders of magnitude, 1 SI vs 1103SI for goethite (O'Reilly, 1984; Dekkers, 1989). Because of the large difference in magnetite susceptibility compared to other ferromagnetic minerals it is the dominant component even when in very low abundance. Therefore, ferromagnetic minerals identi- fied through IRM acquisition modeling had a small contribution to the lowfieldcmeasurements.
4.2. Local magnetic reversal stratigraphy
Thermal demagnetization isolated sample ChRM directions better than the alternatingfield method and was selected for all of the samples. Thermal demagnetization was applied until intensities were very low and remanence became erratic due to spontaneous magnetization. Most specimens show an overprint, often parallel to
the modernfield direction (Fig. 4a) Typically, ChRM directions were obtained above 300C. ChRM directions were compared before and after bedding correction (Fig. 4b). In geographic coordinates two clusters are observed: south and downwards and north and up- wards, respectively. After bedding correction, directions are ori- ented north and down or south and up, displaying both normal and reverse polarities, respectively. There was a considerable scatter in directions likely due to incomplete removal of the secondary component, and coarse grain size of the majority of sampled ho- rizons. After calculating ChRM directions for individual specimens, site mean directions were computed and the corresponding VGP latitudes were used to establish the local magnetostratigraphy of the Berga Conglomerate Group (Fig. 5).
Data were classified into two groups based on angular distance to the expected pole. VGP angular departures45of the paleo- magnetic north pole or its antipode were designated normal or reversed, respectively. VGP angular distances>45of the pole (<3%
of specimens) were of indeterminate polarity. Two or more consecutive sites of the same polarity were interpreted to consti- tute a magnetozone in the local magnetostratigraphy; however, one site with a different polarity than its neighbors was designated a sub-magnetozone. Following this, twelve magnetozones were identified, N1-N6, R1-R3, and I1-I3 (Fig. 5). The border between magnetozones was chosen as the mid-point between any two opposite polarities or 15 m from thefirst sample when entering or exiting an indeterminate region.
4.3. Finite strain distributions and fold structural position
The location of AMS sites are shown in map view (Fig. 2a), cross section (Fig. 6), and UTM coordinates are provided inSupplemental Material 1. All samples show low levels of anisotropy, with anisotropy degree (P0) (Jelinek, 1981) varying from 1.02 to 1.09 (Fig. 7a). The low level of anisotropy observed in the dataset is pervasive and extends from north to south for>10 km from the thrust front. A majority of sites display an oblate fabric, but both limbs show a nearly plane strain character. The hinge is more oblate than either limb, but is still within 1serror limits of the mean of both limbs. DeformedSkolitos tubes in siltstone bedding planes show an ellipticity of 1.28 ± 0.16 (harmonic mean ± 1s) and 1.28±0.15 in the steep fold limb and 1.29±0.21 in the hinge zone.
Skolitosburrows in sandstone bedding planes show an ellipticity of 1.29±0.20 and 1.37±0.13 on the steep fold limb, and 1.39±0.17 in the hinge zone, respectively. TheSkolitostubes record the same strains within 1serrors in both sandstone and siltstone beds. The
Fig. 3.Magnetic characterization of the Berga Conglomerate Formation. A) Susceptibility temperature dependence of 10 specimens. Each specimen had a volume normalized bulk susceptibility at room (~21C) and liquid nitrogen temperatures (196C). B) Frequency distribution of bulk susceptibility of all siltstone AMS specimens.
harmonic mean was chosen because it most closely represents the tectonic strain in clastic rocks especially for low values of strain (Lisle, 1977).
Orientation analysis of the three principal axes of susceptibility were constructed for each region to better understand the geom- etry of the strain (Figs. 6 and 7). Both limbs recordk1, the maximum susceptibility axis, parallel to the strike of bedding;k2, the inter- mediate susceptibility axis, in the bedding plane and oriented nearly N-S; and k3, the minimum susceptibility axis, nearly orthogonal to the bedding plane. In the overturned limb k3 is rotated clockwise from the pole to the bedding plane, whereas in the shallow limbk3 is rotated counterclockwise (Fig. 7b). In the hinge zonek1is oriented parallel to the strike of bedding and thek2
and k3axes produce a girdle oriented nearly N-S. The angle be- tween k3 and the pole to bedding varies with structural location (Fig. 7b). Hinge sites show the greatest difference betweenk3and the pole to bedding with a mean difference of 63, while steep and shallow limbs show a mean difference of 12and 10, respectively.
Hinge specimens show a moderate transecting cleavage in siltstone beds. In sites containing cleavage,k2is oriented parallel to the strike of foliation, which is the same as the strike of bedding.
4.4. ccyclostratigraphy
Bulk susceptibility values range from 4.13105to 1.96102 S.I. and show multi-hierarchical amplitude and frequency Fig. 4.A) Vector end point diagrams for four paleomagnetic samples plotted in geographic coordinates, followingZijderveld (1967). The data points represent a particular demagnetization step projected onto the horizontal (solid dots) and the vertical planes (open squares). Numbers indicate temperature at that point in degrees Celsius except for specimen LA013-3B where they indicate the peak alternatingfield strength in mT. B) Equal area projection of Characteristic Remanence Magnetization (ChRM) directions in geographic coordinates and after 100% unfolding. Sites projected unto the lower hemisphere are indicated byfilled circles and sites projected unto the upper hemisphere are indicated by open circles.
variability throughout the data series (Fig. 8). Amplitude variations incare most likely related to changes in lithology. High amplitude variations coincide with conglomerate facies especially in the more
poorly stratified conglomerates. Lower amplitude variations inc were recovered from the sandy and silty beds. Frequency variations ranged from 2 to 100 m and were also related to lithologic Fig. 5.Virtual geomagnetic pole (VGP) latitudes and correlation to the geomagnetic polarity time scale GPTS2012 (Gradstein et al., 2012). Sites that have an opposite polarity from surrounding sites are indicated with a half bar of the corresponding polarity. Solid circles are site means and open circles are individual specimens. Dashed lines show correlations with strong confidence and dotted lines show plausible correlations.
variations with short-scale frequency variations occurring on the same scale as sandyfluvial channels and silty overbank couplets while long-scale variations relate to alluvial tofluvial sequences.
Power spectra obtained from time correlated MTM spectral analysis shows significant power in frequencies associated with Milanko- vitch scale cycles. Specifically, the precession index band and obliquity (41 kyr) are significant to the 99% confidence level and long eccentricity (400 kyr) is significant to the 90% confidence level as compared to the robust red noise model (Mann and Lees, 1996).
To reduce the effects of high frequency variable sedimentation rates on stratigraphic ages thecdata series was correlated to the theoretical obliquity model of Laskar et al. (2004). This was accomplished by applying a Gaussianfilter centered on 0.02439 cycles/kyr (41 kyr cycle with cutoffs of 0.021739 and 0.027778 cy- cles/kyr) to the data series. The filtered data series was then correlated to the theoretical obliquity model (Laskar et al., 2004) by visually matching identified maxima of thefiltered series to max- ima of the theoretical model. Obliquity was chosen because of its high significance,>99%, above the robust red noise model. The tuned data series shows similar significant frequencies as the un- tuned data while improving the astronomical signal and decreasing the power in non-Milankovitch cycles (Fig. 9) suggest- ing that signals recovered from c data are global astronomical variations. The significant peak at 20 kyr (precession) was used to provide high-resolution timing for structural modeling.
5. Discussion
5.1. Magnetic contribution toc
IRM acquisition modeling suggests that multiple ferromagnetic minerals are present in Berga Conglomerate Group samples (Fig. 3).
These are interpreted as detrital magnetite for the low coercivity phase with a mean coercivity of ~30 mT. The Berga Conglomerate Group was deposited influvial to alluvial systems making biogenic magnetite unlikely. Iron sulfides are also unlikely in a highly
oxidizing environment. The higher coercivity components are consistent with the coercivities of hematite and goethite. Thermal demagnetization supports this interpretation with a loss of rema- nence around 120 C and another high temperature >650 C component (Fig. 4b) (Lowrie, 1990).
Magnetic susceptibility is controlled by ferromagnetic, para- magnetic, and diamagnetic minerals. In terrestrial sediments, paramagnetic grains such as clay particles very often dominate the totalcsignal (Pares and van der Pluijm, 2002). Low temperature analysis suggests that paramagnetic contributions were large in comparison to the ferromagnetic contribution to susceptibility because of an average increase in susceptibility of 230%. Ferro- magnetic minerals do not show a temperature dependence so a large increase in susceptibility at low temperatures is indicative of high concentrations of paramagnetic minerals (Pares et al., 1999;
Tauxe, 2010). Diamagnetic contributions must also be relatively small as all specimens display a positive response to the applied field. Susceptibility of paramagnetic minerals such as phyllosili- cates show an inverse temperature relationship (Richter and van der Pluijm, 1994; Pares and van der Pluijm, 2002) following the Curie-Weiss Law:
c
¼ CTTC (1)
where C is the material specific Curie constant (units of m3K/kg), T is the temperature at which the measurement is taken (units of K), and TC is the paramagnetic Curie Temperature (units of K).
Substituting typical values (Beausoleil et al., 1983) for these pa- rameters, C¼5e17106m3K/kg; T¼77 K (liquid nitrogen); and TCz0 K (Dunlop and Ozdemir, 1997), gives a 3.8 times increase in susceptibility at liquid nitrogen temperatures as compared to room temperature (300 K) in pure paramagnetic phyllosilicates. The observed 2.3 time increase at low temperatures implies than on average 61% of the observed susceptibility is due to paramagnetic minerals.
Fig. 6.Cross section showing AMS data grouped into steep limb (SLM 4, 5, 8, 9, 10, 17, LA, LC, and LD), hinge (SLM 3, 6, 11, 12, 13, 18, 19, 20, and 21), and upright limb (SLM 1, 2, 7, 14, 15, and 16) to show variations in fabric across the main syncline. Principal axisk1(squares),k2(triangles),k3(circles), bedding plane (solid line) and cleavage plane (dashed line) are plotted in geographic coordinates on equal area, southern hemisphere projections for each region.
5.2. Correlation of local magnetic reversal stratigraphy to the GPTS
Just north of the studied section, the Berga Conglomerate Group lies conformably on marine marls and marly limestones which are correlated to the Igualada Formation. The Igualada Formation and its lateral equivalents were studied by a number of authors (e.g.
Taberner et al., 1999; Costa et al., 2010; Rodríguez-Pinto et al., 2012), and were shown to have a Bartonian age (middle-upper Eocene) based on multi-species biostratigraphy and magneto- stratigraphy (middle-upper Eocene) (Sole-Sugra~nes and Fig. 7.A) Flinn diagram (Flinn, 1962) of siltstone AMS specimens by structural position.
Dark symbols give the average strain at each structural position with±1serror bars. B) Box and whiskers plot of the angle betweenk3and pole the bedding as a function of structural position. Steep and shallow limb positions show an opposite sense of rotation suggesting someflexural slip folding. C) Kinematic model of the main syncline from fabric analysis. AMS fabrics suggest mainlyflexural slip with someflexuralflow during syncline formation. Thek1axis is parallel to the cleavage plane in the hinge where cleavage is present.
Fig. 8.Volume normalizedcdata series as a function of stratigraphic distance. Grey regions indicate where tunnel gaps in the original data werefilled by SSAfitting with 100 m averaging (Ghil et al., 2002).
Mascare~nas, 1970; Puigdefabregas et al., 1992) or chron C17. Our local magnetic polarity stratigraphy begins with a poorly resolved normal polarity chron which shows a conspicuous long reverse chron (R3) in the mid-upper section. The GPTS (Gradstein et al., 2012) shows that after chron C17 (middle-late Eocene), thetermi- nus post quemof the Berga Conglomerate Group thefirst prominent reverse chron is interpreted as C13r, which is followed by an even longer reverse chron (C12r). The most reasonable interpretation is
that normal magnetozone N5 corresponds to Chron C13n, and magnetozones R2 and R3 as Chrons C13r and C12r, respectively.
Continuing the one-to-one correlation downsection of magneto- zones to chrons, normal magnetozones N4 and N1 correspond to Chrons C15n and C16n respectively. The overall local magneto- stratigraphy would then encompass Chrons C16n to C12n, rendering the age of the Berga Conglomerate Group as late Eocene- early Oligocene (ca. 36 to 31 Ma) in agreement with previous Epoch Fig. 9.A) MTM spectral analysis of the GPTS2012 correlatedcdata series. Expected astronomical frequencies are shown by the black regions. Some significant spectral peaks are identified. The obliquity centered Gaussianfilter is shown as the grey curve. B) MTM spectral analysis of the obliquity tunedcdata series. Peaks without expected Milankovitch periodicities lose significance after tuning to theoretical obliquity. Both A) and B) have an inset to better display expected precessional frequencies.
interpretations (e.g.Keller et al., 1994; Ford et al., 1997; Verges et al., 2002; Barrier et al., 2010). Polarity reversal boundaries do not correlate to changes in lithologic or rock magnetic properties and therefore are related to long term polarity changes.
Knowing absolute ages produced from the magnetic reversal stratigraphy allows for the calculation of the average sediment accumulation rates for entire section as well as for each fully con- strained reversal chron. The average sediment accumulation rate for the 1691 m sampled section is 27.8 cm/kyr, and rates for each chron are as follows: C15r: 27.4 cm/kyr (115 m); C15n: 32.8 cm/kyr (100 m); C13r: 18.6 cm/kyr (245 m); C13n: 27.8 cm/kyr (125 m);
and C12r: 28.2 cm/kyr (685 m). The bounding magnetochrons C16n and C12n are poorly constrained so sediment accumulation rates have not been determined for them.
The principle unconformity (senso Riba, 1967) occurs within C12r so the calculated sediment accumulation rate for C12r is a maximum rate assuming no non-depositional hiatus. Because beds on either side of the unconformity have a reversed polarity we assumed that the unconformity remains within the same magne- tochron. If the unconformity instead were to end in reversed chron C11r and not C12r the minimum duration of the unconformity would be the length of C12n, or 440 kyr. A large depositional hiatus or erosional removal of material during C12r is unreasonable due to the section's proximal position and the aggradational nature of the basin. Therefore, the unconformity is likely to remain within C12r and to be relatively short in duration. For all calculations the sediment accumulation rate for C12r was taken to be the maximum rate of 28.2 cm/kyr.
5.3. Anisotropy of magnetic susceptibility (AMS) fabrics
Magnetic mineralogy results suggest AMS fabrics are dominated by paramagnetic clays and the fabrics found in the siltstone and sandstone beds of the Berga Conglomerate Group are similar to those found infine grained formations (e.g.Borradaile, 1991; Pares et al., 1999; Pares, 2004; Soto et al., 2009). The recovered magnetic fabrics were interpreted to record multiple stages of deformation.
The earliest is a compactional fabric that is pervasive throughout the region; a compactional AMS fabric is common in sedimentary rocks (Kodama, 2012). This fabric was observed in samples collected>10 km apart, including nearlyflat-lying exposures to the south. This fabric would have formed as the phyllosilicates settled with the minimal principal axis nearly orthogonal to bedding and was enhanced by sedimentary compaction (Sintubin, 1994; Pares, 2004) during burial. This would have produced an oblate fabric withk1andk2randomly oriented within bedding.
The next fabric to be recorded was the layer parallel shortening (LPS) fabric observed in the fold limbs and oriented E-W. It is recorded by thek1axis which is perpendicular to the N-S short- ening directionFig. 6. Thek1axis orientation is interpreted as an intersection lineation of phyllosilicate grains (Pares et al., 1999;
Tauxe, 2010). As deformation progressed the originally oblate fab- ric became triaxial withk3remaining perpendicular to bedding,k1
reorients to be perpendicular to the shortening direction, andk2
reoriented to be parallel to the N-S shortening direction. The last fabric to form is only present in cleaved hinge samples of the syncline (Fig. 6). In these samples, thek3axis is oriented at a high angle to the pole to bedding while defining a girdle which includes thek1andk2axes and is perpendicular to the shortening direction (Fig. 6). This last fabric is expressed by cleavage infine grained li- thologies and mesoscopic faulting in coarser beds. The antithetic rotation recorded by thek3axes in the fold limb specimens (Fig. 7) provide evidence for layer shear consistent withflexural folding as the bedding planes became mechanically active.
5.4. Kinematic constraints
The competing kinematic models of folding at Sant Llorenç de Morunys (Ford et al., 1997; Suppe et al., 1997; Alonso et al., 2011) each predict unique strain histories. Ourfinite strain data provides an objective method to evaluate fold kinematics.Suppe et al. (1997) advocates a migrating hinge model which would produce differing rock strain as a function of structural position about the fold. As rocks move through the hinge region, sites on the upright limb should record strains reflecting the transition from flat limb to upright limb. Afixed hinge kinematic model (Ford et al., 1997) predicts a different strain distribution. Limbs in afixed hinge fold retain their structural position through time. Thus, limbs and hinge region would experience independent deformational histories. Our strain data shows no evidence of a significant difference in theflat lying and upright limbs except for the degree of bedding rotation (Fig. 6). The distribution of rock strain is interpreted as evidence for fixed hinge kinematics as suggested byFord et al. (1997). Geologic strain measurements fromSkolitostubes in bedding did not provide definitive differences between the limbs and hinge within mea- surement errors.
5.5. ccyclostratigraphy tuning to magnetic reversal chronology
A robust cyclostratigraphy, with significant spectral peaks at expected Milankovitch periods, based on c variations has been developed for the Berga Conglomerate Group (Fig. 10). This cyclo- stratigraphy spans from 31.6 Ma within the largest conglomerate in the El Castell Formation to 34.4 Ma at the boundary between the El Bastets and Sobirana Formations a total stratigraphic distance of 653.4 m. Low temperaturec measurements indicate that para- magnetic minerals dominate the mixed magnetic mineralogy in controllingc variations. Because of the terrestrial nature of the Berga Conglomerate Group the paramagnetic minerals are inter- preted to be clays formed by the weathering of source material prior to and during sediment transport.
The magnetic reversal stratigraphy provided broad sediment accumulation rates for correlation of spectral peaks to absolute time. Tuning thecdata series to theoretical obliquity (Laskar et al., 2004) for the Eocene - Oligocene allows corrections forfiner vari- ations in sediment accumulation rates. Obliquity tuning was selected because of a strong spectral significance,>99% confidence above a robust red noise model. Tuning to theoretical obliquity reduced noise caused by sediment accumulation variations (Fig. 9).
It also enhanced precession scale frequency by eliminating some peaks and enhancing the signal at the 25.7 and 20 kyr frequencies.
This cyclostratigraphy provides 20 kyr resolution (precession scale) for structural reconstruction of deformation caused by the fault propagated fold.
5.6. Climate encoding
Encoding mechanisms are problematic to determine incbased cyclostratigraphy because c can be carried by ferromagnetic, paramagnetic, and diamagnetic minerals. Magnetic mineral char- acterization of the Berga Conglomerate Group suggests that varia- tions in paramagnetic clays dominate thecsignal. We suggest that the source of these paramagnetic clays is not authigenesis but rather detrital clay sourced from the Pyrenees. It is unlikely that high-energy terrestrial fluvial and alluvial environments were producing this clay in place. Variations in paramagnetic material abundance may be driven by monsoonal strength as suggested by Gunderson et al. (2012). This would also cause higher concentra- tions of detrital magnetite concurrently further enhancing a wet/
dry climate signal.
5.7. High-resolution sedimentation and folding rates
Kinematic controls on the fold growth and therefore fault slip can be inferred from growth strata geometries (Suppe et al., 1992).
The 20 kyr resolution we obtained was used to calculate both sediment accumulation and folding rates to examine their influ- ence on growth strata formation. Sediment accumulation rates at 20 kyr and magnetic polarity chron scales are shown in Fig. 11.
Cyclostratigraphy based accumulation rates vary from 0.10 to 0.37 m/kyr and average 0.24 m/kyr. Both the chron and preces- sional frequency averaged resolution predict similar accumulation rates in sites younger than 33.25 Ma. However, the two methods result in different rates in older rocks with the chron averaged rates overestimating the sediment accumulation rate. Folding history at both 20 kyr and magnetic polarity chron scales are shown inFig. 12.
Cyclostratigraphy based folding rates vary from 0 to 100/Myr.
Chron averaged timing also overestimate the folding rate and are artificially steady. The 20 kyr resolution determined from magnetic cyclostratigraphy recovers a much more unsteady deformational history; however, bedding orientation uncertainty exacerbated in channelized deposits limits the resolution of bedding tilt changes to 3. Our folding rate determination is limited by bedding inclination uncertainty rather than timing. Therefore, in order to reduce un- certainty, the folding rate was averaged over 100 kyr provide the most certain measure of folding rate variability. This 100 kyr record displays two periods of slow tectonic activity from 33.4 to 34.4 Ma and from 32.6 to 33.1 Ma punctuated by periods of more rapid folding.
These results illustrate the difference between magnetic polarity-averaged and high-resolution cyclostratigraphy rates for understanding deposition and deformation. Higher temporal
resolution recovers more variability in both sediment accumulation and deformation rates suggesting that longer timescale records do not recover the complete variance of intrinsic behavior. The cyclostratigraphic histories display periods of tectonic quiescence during orogenesis which were not recovered in longer time scale Fig. 10.Correlation ofcdata series to theoretical obliquity model ofLaskar et al. (2004). A) Thecdata series (grey) wasfiltered at obliquity, 41 kyr, (black) to allow for correlation to the theoretical model. Correlations betweenfiltered data and theoretical obliquity (B) are dashed for clarity.
Fig. 11.Sediment accumulation correlated to precessional (solid) and magnetochron (dashed) time scales. Accumulation rates are determined by using the slope of the curve. Error bars show absolute timing uncertainties and stratigraphic location precision.
reconstructions. Because sediment accumulation rates were more steady than deformation rates we suggest that the structural basin built by deformation was not sediment limited. This implies a large sediment source area to feed the growing basin.
5.8. Structural modeling
The Fault-Fold v6 software package from Richard Allmendinger (Allmendinger, 1998; Hardy and Allmendinger, 2011) was used to convert fold growth rates to fault propagation rates. This software package allows for forward and inverse modeling of fold growth assuming general trishear faulting (Erslev, 1991). Hardy and Ford
(1997) successfully modeled the general geometry of the Sant Llorenç de Morunys structure with a thrust fault that steps up from a horizontal detachment, a ramp angle of 15, a trishear angle of 5, and a propagation to slip ratio of 1.61. Combining their results with our high-resolution chronology in the growth stratigraphy provides better control on the duration of the principal unconformity. Re- sults of our model run with a 100 kyr depositional hiatus is shown inFig. 13. Modeled fault slip rates used in the creation of this model varied from 115 to 265 m/Myr. The minimum duration of the principle unconformity can be estimated by the amount of time needed for the synclinal axial plane to migrate upsection as seen in field observations. A 100 kyr depositional hiatus is sufficiently long Fig. 12.Percent folding as a function of time at precessional (solid) and magnetochron (dashed) scales. The right axis shows 100 kyr (dotted) folding rate averaged from precessional timing. Error bars for magnetochron folding are displayed. Inset shows precessional scale folding rates for a“constant”rate at 100 kyr scale. Higher resolution timing recovers both greater variability in fold rate and greater % uncertainty.
Fig. 13.Results of an individual Fault-Fold modeling run. Model ran for extent of magnetostratigraphic time constraint, 5.67 Myr. Pregrowth strata are solid lines, growth strata are dotted (cyclostratigraphy) and dashed (magnetochron). Sediment was added based on predicted sediment accumulation rate fromFig. 11.
to cause the recognized shift in the axial plane. This coincides with a time, z31.6 Ma, when more rapid uplift outpaced sediment accumulation. Because sediment accumulation was nearly con- stant, fault slip rates had the major control on the location of the unconformities in the Berga Conglomerate Group section.
6. Conclusions
A high-resolution time scale based on Milankovitch orbital frequencies was recovered from c in terrestrial growth strata.
Magnetic susceptibility based cyclostratigraphy is an objective and robust method of chronology in terrestrial systems that have suf- ficient sediment accumulation and accommodation space such as in foreland basins. New magnetostratigraphy confirm an Eocene- Oligocene age for the Sant Llorenç de Morunys structure and the Berga Conglomerate Group (e.g. Sole-Sugra~nes and Mascare~nas, 1970; Puigdefabregas et al., 1992). Rock magnetic analyses iden- tify three ferromagnetic minerals which contribute to magneto- stratigraphy remanence: magnetite, hematite, and goethite.
Magnetic susceptibility variations are dominated by paramagnetic clay contributions and have a mixed magnetic signal. Strain data support a synclinal fixed hinge and flexural folding kinematics associated with blind thrust faulting ahead of a propagating thrust fault tip. Correlating thecdata series to magnetostratigraphic time and tuning to theoretical obliquity (Laskar et al., 2004) removes the effect of variable sediment accumulation and improves signals at precessional and long eccentricity time scales. The cyclo- stratigraphy of the Berga Conglomerate Group allows for timing of deformation and sediment accumulation at 20 kyr time resolution.
We have recovered the onset of deformation of the Sant Llorenç structure at 33.85 Ma and model the depositional hiatus for the principal unconformity to be 100 kyr. Our age model suggests little correlation between sediment accumulation and deformation rates. This suggests that the structural basin built by deformation was not sediment limited and sediment was supplied from a large drainage area. Sediment accumulation rates are nearly constant for 3 Myr before the principal unconformity. Folding rates were more variable and appear to be separated into long periods of quiescence punctuated by rapid faulting and folding. Model derived fault slip rates vary from 115 to 265 m/Myr. Our results suggest highly var- iable deformation rates. High-resolution timing revealed more unsteadiness in deformation than recovered at magnetic polarity time scales. This speaks to the unsteadiness of fault related folding and how sufficient resolution is needed to recover high frequency variability in fault slip. Rock magnetic cyclostratigraphy is a useful tool for providing the time resolution needed to examine changes in deformation rates.
Acknowledgments
This work was conducted as part of the M.S. Thesis of J. Carrigan at Lehigh University. We would like to thank Marisa Repasch and Claudia Alvarez forfield assistance, as well as Josep Cases for his logistical support. Partial funding for this project was provided to James Carrigan from the AAPG L. Austin Weeks Memorial Grant and the Earth and Environmental Science Department Palmer Research Fund, Lehigh University. We would like to thank M. A. Mamtani and an anonymous reviewer for their help improving this manuscript.
Appendix A. Supplementary data
Supplementary data related to this article can be found athttp://
dx.doi.org/10.1016/j.jsg.2016.09.003.
References
Abdul Aziz, H., Hilgen, F., Krijgsman, W., Sanz, E., Calvo, J.P., 2000. Astronomical forcing of sedimentary cycles in the middle to late Miocene continental Cala- tayud basin (NE Spain). Earth Planet. Sci. Lett. 177, 9e22.http://dx.doi.org/
10.1016/S0012-821X(00)00035-2.
Allmendinger, R.W., 1998. Inverse and forward numerical modeling of trishear fault-propagation folds. Tectonics 17 (4), 640e656.
Alonso, J.L., Colombo, F., Riba, O., 2011. Folding mechanisms in a fault-propagation fold inferred from the analysis of unconformity angles: the Sant Llorenç growth structure (Pyrenees, Spain). In: McClay, K., Shaw, J., Suppe, J. (Eds.), Thrust Fault- related Folding, vol. 94. AAPG Memoir, pp. 137e151.http://dx.doi.org/10.1306/
13251336M943430.
Barrier, L., Proust, J.-N., Nalpas, T., Robin, C., Guillocheau, F., 2010. Control of alluvial sedimentation at foreland-basin active margins: a case study from the north- eastern Ebro Basin (southeastern Pyrenees, Spain). J. Sediment. Res. 80, 728e749.http://dx.doi.org/10.2110/jsr.2010.069.
Beausoleil, N., Lavallee, P., Yelon, A., Ballet, O., Coey, J., 1983. Magnetite properties of biotite micas. J. Appl. Phys. 54 (2), 906e915.
Borradaile, G.J., 1988. Magnetic susceptibility, petrofabrics, and strain. Tectono- physics 156, 1e20.
Borradaile, G.J., 1991. Correlation of strain with anisotropy of magnetic susceptibility (AMS). Pure Appl. Geophys. 135 (1), 15e29.
Burbank, D.W., Puigdefabregas, C., Mu~noz, J.A., 1992. The chronology of the Eocene tectonic and stratigraphic development of the eastern Pyrenean foreland basin, northeast Spain. Geol. Soc. Am. Bull. 104 (9), 1101e1120.http://dx.doi.org/
10.1130/0016-7606(1992)104<1101.
Costa, E., Garces, M., Lopez-Blanco, M., Beamud, E., Gomez-Paccard, M., Larrasoa~na, J.C., 2010. Closing and continentalization of the South Pyrenean foreland basins(NE Spain): magnetochronological constraints. Basin Res. 22, 904e917.http://dx.doi.org/10.1111/j.1365-2117.2009.00452.x.
Dekkers, M.J., 1989. Magnetic properties of natural goethite-I. Grain size depen- dence some low- and high-field related rockmagnetic parameters measured at room temperature. Geophys. J. Int. 97, 323e340. http://dx.doi.org/10.1111/
j.1365-246X.1989.tb00504.x.
Dunlop, D., Ozdemir, 1997. Rock Magnetism: Fundamentals and Frontiers. Cam- bridge University Press.
Erslev, E.A., 1991. Trishear fault-propagation folding. Geology 19 (6), 617e620.
Fang, X., Garzoine, C., Van der Voo, R., Li, J., Fan, M., 2003. Flexural subsidence by 29 Ma on the NE edge of Tibet from the magnetostraitgraphy of Linxia basin, China. Earth Planet. Sci. Lett. 210, 545e560.
Fisher, R., 1953. Dispersion on a sphere. In: Proceedings of the Royal Society of London a: Mathematical, Physical and Engineering Sciences, vol. 217, pp. 295e305.
Flinn, D., 1962. On folding during three-dimensional progressive deformation. Q. J.
Geol. Soc. 118, 385e428.
Ford, M., Williams, E.A., Artoni, A., Verges, J., Hardy, S., 1997. Progressive evolution of a fault-related fold pair from growth strata geometries, Sant Llorenç de Morunys, SE Pyrenees. J. Struct. Geol. 19 (96), 413e441. http://dx.doi.org/
10.1016/S0191-8141(96)00116-2.
Ghil, M., Allen, M.R., Dettinger, M.D., Ide, K., Kondrashov, D., Mann, M.E., Robertson, A.W., Saunders, A., Tian, Y., Varadi, F., Yiou, P., 2002. Advanced spectral methods for climate time series. Rev. Geophys. 40, 3.1e3.41.http://
dx.doi.org/10.1029/2001RG000092.
Gradstein, F.M., Ogg, J.G., Schmitz, M., Ogg, G. (Eds.), 2012. The Geologic Time Scale 2012 2-Volume Set. Elsevier.
Gunderson, K.L., Kodama, K.P., Anastasio, D.J., Pazzaglia, F.J., 2012. Rock-magnetic cyclostratigraphy for the Late Pliocene-early Pleistocene Stirone section, Northern Apennine mountain front, Italy. Geol. Soc. Lond. Spec. Publ. 373 (1), 309e323.http://dx.doi.org/10.1144/SP373.8.
Gunderson, K.L., Anastasio, D.J., Pazzaglia, F.J., Picotti, V., 2013. Fault slip rate vari- ability on 104-105 yr timescales for the Salsomaggiore blind thrust fault, Northern Apennines, Italy. Tectonophysics 608, 356e365. http://dx.doi.org/
10.1016/j.tecto.2013.09.018.
Gunderson, K.L., Pazzaglia, F.J., Picotti, V., Anastasio, D.A., Kodama, K.P., Rittenour, T., Frankel, K.F., Ponza, A., Berti, C., Negri, A., Sabbatini, A., 2014. Unraveling tec- tonic and climatic controls on synorogenic growth strata (Northern Apennines, Italy). Geol. Soc. Am. Bull. 126 (3), 532e552.http://dx.doi.org/10.1130/B30902.1.
Hallam, A.N., Swett, K.E., 1966. Trace fossils from the lower Cambrian pipe rock of the north-west highlands. Scott. J. Geol. 2 (1), 101e107.
Hardy, S., Allmendinger, R.W., 2011. Trishear: A Review of Kinematics, Mechanics, and Applications (AAPG Special Volumes).
Hardy, S., Ford, M., 1997. Numerical modeling of trishear fault propagation folding.
Tectonics 16 (5), 841e854.
Hinnov, L.A., Kodama, K.P., Anastasio, D.J., Elrick, M., Latta, D.K., 2013. Global Milankovitch cycles recorded in rock magnetism of the shallow marine lower Cretaceous Cupido Formation, northeastern Mexico. Geol. Soc. Lond. Spec. Publ.
373 (1), 325e340.
Homke, S., Verges, J., Graces, M., Emami, H., Karpuz, R., 2004. Magnetostratigraphy of Miocene-Pliocene Zagros foreland deposits in the front of the Push-e Kush Arc (Lurestan Province, Iran). Earth Planet. Sci. Lett. 177, 397e410. http://
dx.doi.org/10.1016/j.epsl.2004.07.002.
Hrouda, F., 1982. Magnetic anisotropy of rocks and its application in geology and geophysics. Geophys. Surv. 5, 37e82.
Hrouda, F., 1991. Models of magnetic anisotropy variations in sedimentary thrust sheets. Tectonophysics 185, 203e210.
Jelinek, V., 1981. Characterization of the magnetic fabric of rocks. Tectonophysics 79, T63eT67.http://dx.doi.org/10.1016/0040-1951(81)90110-4.
Keller, P., Lowrie, W., Gehring, A.U., 1994. Palaeomagnetic evidence for post- thrusting tectonic rotation in the southeast Pyrenees, Spain. Tectonophysics 239, 29e42.http://dx.doi.org/10.1016/0040-1951(94)90105-8.
Kischvink, J.L., 1980. The least-squares line and plane and the analysis of paleo- magnetic data. Geophys. J. R. Astron. Soc. 62, 699e718.
Kodama, K.P., Anastasio, D.J., Newton, M.L., Pares, J.M., Hinnov, L.A., 2010. High- resolution rock magnetic cyclostratigraphy in an Eoceneflysch, Spanish Pyr- enees. Geochem. Geophys. Geosystems 11 (June), 1e22. http://dx.doi.org/
10.1029/2010GC003069.
Kodama, K.P., 2012. Paleomagnetism of Sedimentary Rocks: Process and Interpre- tation. Wiley-Blackwell.
Kodama, K., Hinnov, L., 2015. Rock Magnetic Cyclostratigraphy. John Wiley&Sons.
Kondrashov, D., Ghil, M., 2006. Spatio-temporal filling of missing points in geophysical data sets. Nonlinear Process. Geophys. 13, 151e159. http://
dx.doi.org/10.5194/npg-13-151-2006.
Kruiver, P.P., Dekkers, M.J., Heslop, D., 2001. Quantification of magnetic coercivity components by the analysis of acquisition curves of isothermal remanent magnetisation. Earth Planet. Sci. Lett. 189, 269e276.http://dx.doi.org/10.1016/
S0012-821X(01)00367-3.
Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A., Levrard, B., 2004. A long- term numerical solution for the insolation quantities of the Earth. Astron.
Astrophy. 428 (1), 261e285.
Latta, D.K., Anastasio, D.J., Hinnov, L.A., Elrick, M., Kodama, K.P., 2006. Magnetic record of Milankovitch rhythms in lithologically noncyclic marine carbonates.
Geology 34, 29e32.http://dx.doi.org/10.1130/G21918.1.
Lisle, R., 1977. Clastic grain shape and orientation in relation to cleavage from the Aberystwyth Grits, Wales. Tectonophysics 39, 381e395.
Lowrie, W., 1990. Identification of ferromagnetic minerals in a rock by coercivity and unblocking temperature properties. Geophys. Res. Lett. 17 (2), 159e162.
Mann, M.E., Lees, J.M., 1996. Robust estimation of background noise and signal detection in climatic time series. Clim. Change 33, 409e445.
Meyers, S., 2014. Astrochron: An R Package for Astrochronology (Version 0.3.2).
Nador, A., Lantos, M., Toth-Makk, A., Thamo-Bozso, E., 2003. Milankovitch-scale multi-proxy records fromfluvial sediments of the last 2.6 Ma, Pannonian Basin, Hungary. Quat. Sci. Rev. 22, 2157e2175. http://dx.doi.org/10.1016/S0277- 3791(03)00134-3.
O'Reilly, W., 1984. Rock and Mineral Magnetism. Springer.
Pares, J.M., 2004. How deformed are weakly deformed mudrocks? Insights from magnetic anisotropy. Geol. Soc. Lond. Spec. Publ. 238 (1), 191e203.
Pares, J.M., van der Pluijm, B.A., 2002. Phyllosilicate fabric characterization by low- temperature anisotropy of magnetic susceptibility (LT-AMS). Geophys. Res. Lett.
29 (24), 68e71.
Pares, J.M., van der Pluijm, B.A., Dinares-Turell, J., 1999. Evolution of magnetic fabrics during incipient deformation of mudrocks (Pyrenees, northern Spain).
Tectonophysics 307 (1), 1e14.
Puigdefabregas, C., Mu~noz, J.A., Verges, J., 1992. Thrusting and foreland basin evo- lution in the southern Pyrenees. In: McClay, K.R. (Ed.), Thrust Tectonics.
Springer, London, pp. 247e254.http://dx.doi.org/10.1007/978-94-011-3066-0_
9.
Pueyo-Morer, E.L., Millan-Garrido, H., Pocoví-Juan, A., Pares, J.M., 1997. Determi- nation of the folding mechanism by AMS data; study of the relation between
shortening and magnetic anisotropy in the Pico del Aguila Anticline (southern Pyrenees). Phys. Chem. Earth 22 (1), 195e201.http://dx.doi.org/10.1016/S0079- 1946(97)00102-X.
Rafini, S., Mercier, E., 2002. Forward modelling of foreland basins progressive un- conformities. Sediment. Geol. 146, 75e89. http://dx.doi.org/10.1016/S0037- 0738(01)00167-1.
Riba, O., 1967. Resultados de un estudio sobre el Terciario continental de la parte este de la depresion central catalana. Acta Geol. Hisp. 2, 1e6.
Riba, O., 1976. Syntectonic unconformities of the Alto Cardener, Spanish Pyrenees: a genetic interpretation. Sediment. Geol. 15, 213e233.http://dx.doi.org/10.1016/
0037-0738(76)90017-8.
Richter, C., van der Pluijm, B.A., 1994. Separation of paramagnetic and ferrimagnetic susceptibilities using low temperature magnetic susceptibilities and compari- son with highfield methods. Phys. Earth Planet. Inter. 82 (2), 113e123.
Rodríguez-Pinto, A., Pueyo, E.L., Serra-Kiel, J., Samso, J.M., Barnolas, A., Pocoví, A., 2012. Lutetian magnetostratigraphy calibration of larger foraminifera zonation (SBZ) in the Southern Pyrenees: the Isuela section. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 333e334, 107e120.http://dx.doi.org/10.1016/j.palaeo.2012.03.012.
Sintubin, M., 1994. Clay fabrics in relation to the burial history of shales. Sedi- mentology 41, 1161e1169.
Sole-Sugra~nes, L., Mascare~nas, P., 1970. Sobre las formaciones Ager y Baga, del Eoceno del Cadí (Perpirineo oriental) y de unos pretendidos olistolitos del mismo. Acta Geol. Hisp. 4, 97e101.
Soto, R., Larrasoa~na, J.C., Arlegui, L.E., Beamud, E., Oliva-Urica, B., Simon, J.L., 2009.
Reliability of magnetic fabric of weekly deformed mudrocks as a paleostress indicator in compressive settings. J. Struct. Geol. 31, 512e522.http://dx.doi.org/
10.1016/j.jsg.2009.03.006.
Spahn, Z.P., Kodama, K.P., Preto, N., 2013. High-resolution estimate for the deposi- tional duration of the Triassic Latemar Platform: a new magnetostratigraphy and magnetic susceptibility cyclostratigraphy from basinal sediments at Rio Sacuz, Italy. Geochem. Geophys. Geosystems 14 (4), 1245e1257. http://
dx.doi.org/10.1002/ggge.20094.
Suppe, J., Chou, G.T., Hook, S.C., 1992. Rates of folding and faulting determined from growth strata. In: McClay, K.R. (Ed.), Thrust Tectonics. Springer, London, pp. 105e121.http://dx.doi.org/10.1007/978-94-011-3066-0_9.
Suppe, J., Sabat, F., Mu~noz, J.A., Poblet, J., Roca, E., Verges, J., 1997. Bed-by-bed fold growth by kink-band migration: Sant Llorenç de Morunys, eastern Pyrenees.
J. Struct. Geol. 19 (96), 443e461. http://dx.doi.org/10.1016/S0191-8141(96) 00103-4.
Taberner, C., Dinares-Turell, J., Gimenez, J., Docherty, C., 1999. Basinfill architecture and evolution from magnetostratigraphic cross-basin correlations in the southeastern Pyrenean foreland basin. Geol. Soc. Am. Bull. 111, 1155e1174.
http://dx.doi.org/10.1029/1999PA000491.
Tarling, D., Hrouda, F., 1993. Magnetic Anisotropy of Rocks. Springer Science&
Business Media.
Tauxe, L., 2010. Essentials of Paleomagnetism. Univ. of California Press.
Thomson, D.J., 1982. Spectrum estimation and harmonic analysis. Proc. IEEE 70, 1055e1096.
Verges, J., Fernandez, M., Martínez, A., 2002. The Pyrenean orogen: pre-, syn-, and post-collisional evolution. J. Virtual Explor. 55e74.
Williams, E.A., Ford, M., Verges, J., Artoni, A., 1998. Alluvial gravel sedimentation in a contractional growth fold setting, Sant Llorenç de Morunys, southeastern Pyr- enees. Geol. Soc. Lond. Spec. Publ. 134, 69e106.
Zijderveld, J.D.A., 1967. AC demagnetization of rocks: analysis of results. Methods Paleomagn. 3, 254.