Out-of-Sequence Thrusting at Orogenic Scale: Assembly of Allochthons along the Western Flank of the Southernmost Appalachian Orogen, USA Open Access

Appalachian thrust faults were among the first “overthrust” structures identified in orogenic belts [1]. As such, studies of these thrust faults have been incorporated in many of the seminal works addressing the mechanics of thrusting [2], the geometry and sequencing of thrust faults [3], and attempts to classify thrust systems in ancient and modern orogens [4, 5]. In the southern and central Appalachians, the latest Paleozoic collision of Laurentia with northwest Gondwana (western South America and Africa) completed the assembly of Pangea [6-10] and created the craton-directed thrusts and structures described herein.

Understanding the kinematic and mechanical development of thin-skinned foreland fold-and-thrust belts (FFTB) has been a fundamental challenge for geologists since these important structural elements were first identified and studied [2]. In modern contractional orogens, thrust belts commonly have a prism or wedge shape cross section, with a topographic surface inclined toward the foreland, a zone of imbrication at the foreland toe of the wedge, and a hinterland-dipping décollement [11, 12]. Coupling field observations, balanced cross sections, and modeling (e.g. critical taper theory), our understanding of mechanically enigmatic tectonic features in foreland and hinterland thrust systems has significantly increased over the past several decades [11-14]. Critical taper theory, especially, has become important for understanding the large-scale mechanical and kinematic evolution of active FFTBs and accretionary prisms [11, 12, 15, 16]. In the critical taper model, rocks driven by compressional tectonic forces in a developing orogenic wedge deform internally, subsequently steepening the foreland-facing orogenic front as a function of Mohr-Coulomb failure [12]. Over time the orogenic wedge evolves toward a critical taper geometry, constrained by the foreland sloping topographic surface and hinterland dipping basal décollement. Upon reaching the critical taper angle, frictional forces at the base of the wedge are balanced by gravitational and compressive forces [16], with tectonically driven deformation accommodated by stable sliding along the décollement [12, 13, 17, 18]. Once the orogenic wedge has reached critical taper, forward propagating thrust faults develop at the toe of the wedge, with progressively younger thrusts forming “in-sequence” (IS) from the hinterland toward the foreland while critical taper is maintained. The thrust system then grows as new ramps cut into the foreland, while older thrusts are abandoned and ride “piggy back” above the décollement. For many decades, IS thrust models have been applied to both the Appalachian FFTB and hinterland [3, 8, 19-26]. However, in the Appalachians and other orogens, thrust faults have been described that formed in a reverse sequence, from the foreland to the hinterland [27]. These out-of-sequence (OOS) or break-back thrusts [26] have often been interpreted as local anomalies or exceptions within an overall forward propagating thrust belt [27-37], receiving less emphasis in the literature. Some workers, however, have suggested a more prominent role for OOS thrusts in the Appalachians. For example, Faill [38, 39] argued that deformation in the north central Appalachians progressed from the foreland toward the hinterland. By integrating structural, stratigraphic, and thermal maturity data in the Alabama FFTB, Pashin [40] argued that OOS thrusts progressed from the most distal foreland toward the hinterland over a distance >60 km. The importance of OOS thrusts has also been recognized from other orogenic belts (e.g. Alps), where the interplay between surficial and tectonic processes is thought to have shifted the progression of thrust faults from IS to OOS [41].

In the southernmost Appalachians of Alabama (AL) and Georgia (GA), Laurentian platform rocks (foreland) as well as allochthons composed predominantly of Lower Paleozoic, Laurentian-affinity metasedimentary and metaigneous rocks (Blue Ridge), were emplaced along the orogen’s northwest flank (Figure 1) during the Late Paleozoic Alleghanian orogeny [37, 42-44]. Large-scale, northwest translation of both previously deformed and metamorphosed Blue Ridge rocks (hinterland), as well as unmetamorphosed to anchimetamorphic rocks of the FFTB, occurred primarily after the Mississippian (ca. 330 Ma) thermal-metamorphic peak [41]. Rocks of the southern Appalachian FFTB make up classic thin-skinned thrust sheets involving subhorizontal displacement and associated deformation of stratified rocks forming sedimentary “cover” above unaffected crystalline “basement” which were translated across a series of connected flats that generally followed mechanically weak strata, and ramps that cut upward through more competent strata [11, 45]. Hinterland thrust sheets southeast of the FFTB (Figure 1) might intuitively be considered thick-skinned thrust sheets because they incorporate only metamorphic rocks. However, these hinterland “crystalline” thrust sheets differ in many respects from thick-skinned thrust sheets defined in traditional classification schemes [5, 45].

In the traditional thick-skinned thrust model, thrusts cut through the underlying crystalline basement at depth, ramping up into overlying sedimentary cover, and carry both basement and cover rocks toward the foreland. Both thin- and thick-skinned thrust models inherently draw a distinction between crystalline basement and stratified cover as significantly influencing the kinematic evolution of the thrust systems [5, 46]. We differentiate between (1) classic FFTB thin-skinned thrust sheets, (2) crystalline thin-skinned thrust sheets of the orogenic hinterland, and (3) classic thick-skinned thrust sheets containing both older crystalline basement and younger, generally unmetamorphosed cover sequences. In traditional models, thrust mechanics are attributed to the presence or absence of favorably oriented weak zones (e.g. bedding) extending for long distances in stratified rocks that are more spatially restricted in crystalline basement [5, 11, 45]. In the Appalachians, crystalline basement is generally synonymous with Mesoproterozoic (i.e. Grenville) deep crustal metamorphic rocks predating the Appalachian orogenic cycle. Younger, metamorphosed stratified rocks of the western hinterland formed between the Grenville and Alleghanian orogenies and were deposited as cover above the older Grenville basement, but never served as basement to the stratified sequences involved in FFTB thin-skinned thrusting. Thus, the hinterland crystalline thrust sheets described herein are thin-skinned in the sense that they formed within stratigraphic cover sequences above the older Grenville basement and have similar structural thicknesses and lateral dimensions typical of thin-skinned thrust sheets. However, they are bounded by thrust faults that behaved mechanically similar to thick-skinned thrusts because the faulted hinterland rocks were internally strong and mechanically homogenous bodies. In addition, these crystalline thrust sheets were bound by thrusts that commonly cut both up and down stratigraphic section in the direction of displacement and did not follow internal mechanical anisotropies (e.g. bedding, foliation, older faults) for significant distances because those anisotropies were folded prior to later faulting, rendering them unsuitably oriented for exploitation by younger faults.

The complex architecture of these crystalline thin-skinned thrust sheets, which includes deformed and metamorphosed rocks of both the Laurentian-margin platform and an extensive fringing marginal basin (Wedowee-Emuckfaw-Dahlonega back-arc (WEDB) basin of [44]), directly results from the telescoping of a broad segment of the Laurentian plate during late Paleozoic orogenesis and OOS thrusting that placed higher grade metamorphic rocks on lower grade ones, cutting obliquely through metamorphic isograds, fabrics, and lithostratigraphy.

In the following, structural relationships from field studies, geologic maps, and cross sections are used to determine the relative timing and overprinting of at least six regional, cross-cutting hinterland contractional deformation phases and their associated regional structures, all of which formed OOS relative to FFTB deformation. Here, we present the first detailed documentation of widespread, significant displacement, OOS thrusting and deformation across a >150 km wide, >300 km-long segment of Laurentian margin cover rocks along the west flank of the southern Appalachian orogen (Figure 2) that occurred during the assembly of Pangea. We conclude that most of the current thrust architecture along the west flank of the orogen in AL and GA was assembled by OOS thrusting during the Alleghanian orogeny, likely as a result of changes from critical to subcritical taper of the foreland-facing orogenic front during the latest Mississippian-earliest Pennsylvanian. Finally, we argue that the transition from critical to subcritical taper and IS to OOS thrusting is coeval with a climatic transition that caused erosion rates to exceed uplift rates along the orogenic front and led to deposition of the thick, Upper Carboniferous Appalachian basin clastic wedge. As such, the dominant OOS foreland and hinterland thrust architecture was formed by interactions between climate and the dynamics of the orogenic system.

Each of the thrust sheets discussed herein (Figures 1 and 2) contains rocks that can be palinspastically linked either to the Paleozoic Laurentian shelf or to a fringing marginal (i.e. back-arc) basin on the seaward edge of the Laurentian plate. Foreland stratigraphy in this region includes the 4–5 km thick southeast Laurentian passive margin cover sequence (Lower Cambrian Chilhowee Group to Cambrian-Lower Ordovician Knox Group) and overlying Middle Ordovician to Pennsylvanian clastic wedge units (Figure 3).

In the hinterland, the structurall a belt (Figures 1-4) contains correlative, shallow-water shelf sequences of the Lower Cambrian to Lower Ordovician outer Laurentian trailing margin (Kahatchee Mountain and Sylacauga Marble Groups), an ~2400 m thick sequence of Silurian(?)-Devonian turbiditic rocks (Lay Dam Formation), and younger Devonian-earliest Mississippian (?) units of the Talladega Group [47-52]. Hinterland thrust sheets southeast of the Talladega belt and northwest of the Brevard fault zone (eastern Blue Ridge, Figures 1-3) contain rocks of the thick (up to ~10 km), extensive, now-dismembered, WEDB Ordovician back-arc basin. These back-arc sequences occur in three separate thrust sheets (Hillabee sheet, Pumpkinvine Creek-Canton sheet, and Ashland-Wedowee-Emuckfaw-New Georgia sheet), with each containing Ordovician bimodal metavolcanic rocks. In addition, the latter thrust sheet also contains thick sequences of turbiditic metasedimentary rocks intruded by numerous Ordovician, Silurian, and Middle Paleozoic, semiconcordant, granitic orthogneiss plutons [53-61]. To the southeast, these back-arc basin sequences are paired with a structurally overlying volcanic/plutonic arc allochthon, the Dadeville complex (Figure 1) [62].

Characteristics of metasedimentary and metaigneous rocks within the hinterland thrust sheets indicate they formed on the distal Laurentian shelf and in a flanking Ordovician-Silurian, Laurentian plate back-arc basin oceanward of the continental margin hinge zone and above subducting Iapetus oceanic lithosphere. Observations linking these Ordovician back-arc basin sequences to the Laurentian margin include the following:

1. Isotopic analyses of detrital zircons from metasedimentary rocks within the back-arc basin units [53, 63-67] demonstrate that the basin contains a significant population of Mesoproterozoic zircons, ranging in age from 1.2 to 0.9 Ga. Such ages are typical of Laurentian crustal materials formed during the Grenville orogeny [68] and most likely represent sediment derived from the adjacent Laurentian continental margin or exposed rifted-margin Proterozoic basement blocks and/or their cover strata along that margin [53]. Abundant Ordovician-Silurian detrital zircons, coupled with U-Pb zircon ages of intercalated metavolcanic rocks and intrusive granitoids, restricts the depositional age of these rocks between Ordovician and Silurian [53, 61, 63].

2. Neodymium (Nd) model ages from WEDB metasedimentary units [66] range from 0.943 to 1.119 Ga, suggestive of Grenville and Granite-Rhyolite Province source rocks [69, 70].

3. Involvement of Mesoproterozoic crust in the generation of the Hillabee Greenstone and Pumpkinvine Creek Formation (Dahlonega Gold belt) (Figures 3 and 4) felsic metavolcanic rocks is evident in their Hf, Nd, and Sr isotopic compositions [66, 71].

4. Lu-Hf analyses, initial ƐHf values, common Pb ratios, TNd DM distributions in zircon grains, and xenocrystic zircon from the ca. 460–430 Ma Zana Granite and Kowaliga Gneiss that intrude the Ashland-Wedowee-Emuckfaw-New Georgia sheet in AL (Figure 2) indicate their source magmas were strongly influenced by Grenville-aged continental crust [59, 61].

   Metamorphism of the frontal hinterland allochthon (Talladega belt) occurred under variable conditions, increasing from lower greenschist facies (subbiotite) in AL (Figure 4(a)), to biotite-grade at the AL-GA border, and garnet grade in the southern part of the Mulberry Rock recess (Figure 4(b)). Metamorphic timing is bracketed by fossiliferous units of Middle Devonian to Early Mississippian(?) age [47, 49] and Early Mississippian (Visean) ⁴⁰Ar/³⁹Ar white mica ages of ca. 325–335 Ma [72, 73] (Figure 3). The age of metamorphism in the overlying eastern Blue Ridge allochthon is constrained by U-Pb zircon ages of pre- to syn-metamorphic plutons which are as young as 335 Ma, and Sm-Nd ages of garnet at ~331–320 Ma [58, 60]. Shortly after these hinterland allochthons reached peak metamorphic conditions (Visean or Serpukhovian; Middle to Late Mississippian, Alleghanian orogeny), sediments of an Early Pennsylvanian clastic wedge (Straven-Pottsville clastic wedge [74, 75]) were derived from uplifts on the southeast (Figure 3). A second Early Pennsylvanian clastic wedge, the Pennington-Lee clastic wedge [74], was shed into the GA and AL foreland from orogenic highlands on the northeast. Importantly, these >3 km thick clastic wedges are the first indicators of foreland sedimentation derived from Alleghanian orogenic uplifts along the southeasternmost Laurentian margin. Depositional constraints from these clastic wedges, coupled with (U-Th)/He ages and thermal modeling of zircon in Paleozoic foreland strata, indicate that the metamorphic allochthons discussed below did not begin to shed detritus into the foreland until the Mississippian, with the main pulse of Alleghanian clastic wedge deposition occurring during the Pennsylvanian [76, 77].

The Hillabee Greenstone in the Talladega belt frontal metamorphic allochthon is a >2.6 km thick, Middle Ordovician (470-468 Ma) bimodal metavolcanic assemblage with minor metasedimentary rocks that formed within the WEDB back-arc basin on the seaward edge of the Laurentian plate [53, 71, 78-81] (Figures 3 and 4). The unit extends along the southeast flank (structural top) of the Talladega belt for >230 km, from inliers within Cretaceous rocks of the Gulf Coastal Plain in AL, to near Buchanan, GA (Figure 4) [71, 79, 82]. Hillabee protoliths consist of tholeiitic basaltic rocks intermixed with lesser (~25%) calc-alkaline dacitic ash flows up to 150 m thick [79, 81]. The base of the unit is a premetamorphic fault (Hillabee thrust; Figures 2,and 4) that emplaced the Ordovician Hillabee Greenstone atop younger rocks (Middle Devonian to earliest Mississippian?) of the Laurentian outer margin shelf [71, 73, 83]. The metavolcanic complex and its basal thrust lie at the structural top of the younger Talladega belt allochthon within imbricate footwall slices of the significantly younger (Alleghanian), postmetamorphic OOS Hollins Line footwall thrust duplex system (see following) (Figures 1, 2 and 4) [71, 83]. The Hillabee thrust represents the first pulse of Appalachian contractional deformation observed within southeastern Laurentian platform margin cover sequences and displaced the Hillabee allochthon relative to its immediate footwall a minimum of 18 km northwestward (displacement vector B, Figure 4(a)) onto Laurentian shelf rocks prior to subsequent greenschist facies metamorphism at depths of 12–15 km.

Emplacement of the Hillabee allochthon along a fault trajectory that followed undeformed stratigraphy in both the hanging and footwall is evident from numerous observations, indicating that the Hillabee thrust had a flat-on-flat geometry over the vast majority of its 230 km strike-length [52, 71, 83]. Compositional layering in laminated mafic phyllite, stratabound massive sulfide zones, and tabular metadacite sheets in the hanging wall of the thrust are not only concordant within the Hillabee allochthon but also parallel Devonian to Mississippian(?) units of the uppermost Talladega Group (Erin Slate-Jemison Chert) in the footwall (Figures 4 and 5) [71, 82-85]. These concordant relationships between rocks of the hanging and footwall of the Hillabee thrust are observed both along and across strike within duplex horses of the younger Hollins Line thrust system (Figure 4), which palinspastically extended tens of kilometers southeastward [71, 79, 83, 86]. In addition, the Hillabee thrust in both the duplex’s parautochthon and imbricate horses maintains a nearly constant distance above the base of underlying Talladega Group units and is never in contact with units stratigraphically beneath them (e.g., Sylacauga Marble) [53, 71, 83]. Emplacement of the Hillabee allochthon, therefore, was stratigraphically controlled by the mechanical character of units within the footwall (upper Talladega Group) and hanging wall (lower Hillabee Greenstone), with rocks of the allochthon emplaced atop the Laurentian shelf as a large, continuous intact sheet of volcanic rocks. There is no evidence that either the Hillabee allochthon or the Talladega Group parautochthon were dismembered or significantly deformed prior to later Alleghanian Hollins Line thrust duplexing [71, 83, 87](Figure 4). Rare exposures of the Hillabee thrust at the boundary between the Hillabee Greenstone and underlying Talladega Group (Jemison Chert-Erin Slate) along its >230 km strike length consist of a “transition” zone up to 100 meters thick of alternating muscovite phyllite, chlorite-sericite phyllite and greenstone lithologies likely representing the lowermost units of the Hillabee allochthon. This suggests the Hillabee thrust exploited mechanically weaker sedimentary sequences at the base of the more competent and dominant basalt lithologies during pre-metamorphic emplacement of the allochthon.

Units in both the footwall and hanging wall of the Hillabee thrust experienced the same peak greenschist facies metamorphic conditions and contain coplanar and colinear metamorphic fabric and deformational features, indicating a premetamorphic origin for the fault [71, 83, 87]. The pervasive greenschist facies overprinting of the thrust apparently obscured any fault zone fabrics/materials along it [88]. The thrust’s extensive flat-on-flat geometry indicates the Hillabee allochthon was emplaced as a thin-skinned thrust sheet that transported rocks of the Hillabee back-arc basin (WEDB) from their palinspastic location southeast of the Laurentian shelf, across the continental margin hinge zone and onto and across Laurentian shelf sequences (Talladega Group) without significantly deforming hanging or footwall stratigraphy.

The youngest paleontologically dated Talladega Group units (Jemison Chert- Erin Slate) in contact with the Hillabee thrust (Figure 4) are at least uppermost Devonian (Famennian) and possibly lowermost Mississippian (Tournasian) in age [47, 49, 83, 85], which means emplacement of the Hillabee Greenstone and movement along the Hillabee thrust can be no older than ~372 Ma (Figure 3). Thrust movement must have ceased prior to Mississippian metamorphism (ca. 330 Ma) of the amalgamated Talladega belt [72, 73], as metamorphic fabrics across the fault are coplanar and fault-related rocks are absent in the rare cases where the fault is exposed. These temporal constraints indicate that the Hillabee allochthon was emplaced between 372 and 330 Ma (Famennian to Visean), most likely during the Early to Mid-Mississippian, prior to metamorphism of the Talladega Group and Hillabee allochthon [71, 73, 87, 88] (Figure 3). Because the Hillabee Greenstone and metasedimentary rocks of the Talladega belt share a common metamorphic and deformational history, with no evidence for earlier thermal or deformational events, the Ordovician Hillabee Greenstone (~468–472 Ma) must have remained tectonically undisturbed at or below greenschist facies following eruption of the mafic/felsic volcanic protoliths in a continental margin back-arc setting, until its emplacement atop Laurentian shelf rocks (ca. 372–330 Ma), that is for 102–140 m.y. (Figure 3). This indicates that the Iapetus-facing, southeastern Laurentian continental margin along the former Alabama continental promontory remained open throughout the period of the Taconic orogeny and did not undergo significant deformation until the Mississippian, consistent with structural analysis and retrodeformation of the sub-Lay Dam Formation unconformity (Figures 3 and 4) at the base of the Talladega Group [89]. Because the Talladega belt can be palinspastically restored to at least the present location of the Pine Mountain belt (Figure 1) [37, 51, 90], rocks of the Talladega Group structurally beneath the Hillabee allochthon must have formed at or beyond the Paleozoic Laurentian outer continental margin. The Hillabee’s exact stratigraphic position within the WEDB is unknown because of its tectonic base and top, but its stratigraphic level is considered to be similar to that of the Pumpkinvine Creek Formation to the northeast in GA [57, 91] (Figure 6).

Although the Hillabee thrust represents the first pulse of unequivocal contractional deformation observed within southeastern Laurentian platform margin cover sequences, some workers [92, 93] have suggested the presence of premetamorphic faults (e.g. Greenbrier-Rabbit Creek and Dunn Creek fault systems) in the western Blue Ridge of North Carolina (NC) and Tennessee (TN). Other workers [94], however, have called into question both the timing and magnitude of displacement along these supposed premetamorphic faults, suggesting they postdate the major prograde and retrograde mineral assemblages. Premetamorphic thrusts other than the Hillabee thrust may exist in the southern Appalachian hinterland, but significant debate over the age of the major fabric forming event in the western Blue Ridge of GA and NC ultimately affects interpretation of any proposed premetamorphic fault. While many workers have attributed peak metamorphic conditions of the broad Blue Ridge-Piedmont “mega thrust” sheet to Ordovician Taconian orogenesis (~470–440 Ma) [95-101], recent studies present both paleontological [102] and U/Pb zircon data [103, 104] identifying post-Ordovician units within the western Blue Ridge Greenbrier thrust sheet of GA that carry the main regional isograd-forming metamorphic overprint. In addition, other isotopic systems (garnet Sm-Nd ages and muscovite Ar-Ar cooling ages) in the region suggest that the dominant dynamothermal metamorphic event in the southern Blue Ridge of AL and GA was associated with Alleghanian (Upper Mississippian)

orogenesis [72, 105, 106]. In an analysis of the Greenbrier fault, Clemons and Moecher [94] recognized that a lack of premetamorphic folding as well as the absence of a metamorphic foliation that could be attributed to Ordovician tectonism in the Great Smoky Mountains region required any “Taconic” metamorphic isograds to be largely static in nature, an observation noted earlier by Hadley and Goldsmith [107].

It is possible, therefore, that Carboniferous (Alleghanian) metamorphism in the Blue Ridge overprints an earlier Taconic, generally static metamorphic event. But, whereas isotopic and paleontologic constraints from the Talladega belt allow for unequivocal interpretation of the Hillabee thrust as a premetamorphic fault, interpretation of the age of faulting for other purported “Taconic” Blue Ridge fault systems will require resolving the age of the peak fabric forming metamorphic event in that region. The Hillabee thrust, therefore, marks the earliest, unequivocal interval of thrust faulting along the west flank of the orogen.

During the time interval in which the Hillabee thrust was active, foreland rocks immediately northwest of the Talladega belt hinterland show no evidence for coeval latest Devonian-Mississippian deformation (i.e. thrust imbrication, folding, or angular unconformities). However, well data from the Wiley dome, a subcircular structure with ~120 m of structural relief in the most distal AL foreland, 75 km northwest of the current location of the Hillabee thrust (Figure 2), indicates deformation there began during the Late Mississippian [40]. Subsurface data suggest the shallow Wiley dome is cored by a complex, asymmetric structure interpreted as a breakthrough fault-propagation fold above a thrust fault that soled in the underlying Cambrian Conasauga Formation and duplexed younger Devonian-Mississippian units. Because this structure is truncated by an angular unconformity at the base of the Mississippian-Pennsylvanian Parkwood Formation, the dome and underlying thrust fault must have been active during Serpukhovian (Upper Mississippian) to Bashkirian (Lower Pennsylvanian) times [40]. Deformation associated with the Wiley dome follows the time interval during which the Hillabee thrust was active, but is coeval with the age of dynamothermal metamorphism in the Talladega belt and eastern Blue Ridge hinterland [60, 72, 73]. Thus, the Wiley dome in the most distal foreland must have developed IS following displacement on the Hillabee thrust at the hinterland’s northwest flank and during the time of peak metamorphism in the southernmost Appalachian Talladega belt and Ashland-Wedowee-Emuckfaw belt. This interpretation is consistent with thermal modeling of (U-Th)/He data [77] from detrital zircon in AL foreland units, which cooled through He closure temperatures (200°C) prior to 350 Ma (Lower Mississippian) and must have been uplifted during that time frame. Timing constraints on the Wiley dome and Hillabee thrust coupled with He closure ages from detrital zircon in younger foreland units are consistent with a forward propagating orogenic front and the development of IS structures that migrated from the hinterland to the distal foreland (Wiley dome) during the Late Devonian and/or Mississippian. Collectively, age constraints for the Hillabee thrust and Wiley dome temporally bracket the most southeastern (hinterland) and northwestern (foreland) phases of IS deformation. The earliest possible movement along the Hillabee thrust can be no older than ca. 372 Ma, but possibly as young as ca. 330 Ma, while the oldest possible age for initiation of the Wiley dome structure is Serpukhovian (~327 Ma). Thus, movement along the Hillabee thrust and deformation of the Wiley dome must be separated by an interval of no less than 3 m.y., but possibly as much as 45 m.y. Although there is no definitive evidence that rocks between the older Hillabee thrust and younger Wiley dome structure (Figure 2) were involved in thrust faulting during this time interval, it is likely that at least some of the intervening foreland rocks experienced IS deformation that was subsequently obscured by younger OOS deformation (see below).

Although the current across-strike distance between the Hillabee thrust and Wiley dome is ~75 km (Figure 2), younger OOS thrusting would have significantly shortened the distance between them. Thus, 75 km is a minimum distance between the two structures at the time of their formation. Palinspastic reconstruction of the AL FFTB [37] indicates the leading edge of the hinterland (Talladega-Cartersville fault, see below) and the most distal edge of deformed foreland rocks (Sequatchie anticline, see below) were separated by >200 km prior to Alleghanian shortening. Thus, during the Mississippian, the Hillabee thrust and Wiley dome were likely separated by a distance of at least 100–150 km, developing over an interval of a few million to a few tens of millions of years. Similar relationships are observed in the FFTB of central Mexico, where illite Ar/Ar dating of thrust faults and shear zones documents progressive IS deformation from hinterland to foreland across a >150 km-wide zone over a 40 m.y. time interval [108]. The evolution of thrust faults in the Himalayan orogenic belt also occurred over a similar time frame, with IS faults developing between the Tethyan thrust belt and Subhimalayan thrust system over the past ca. 50 m.y [109].

In this section, we describe the development of a series of OOS structures between the distal foreland Alleghanian deformation front and the western Inner Piedmont in the hinterland (Figures 1 and 2). The relative timing of each of these structures has been described by many geologists, who noted map and subsurface (cross section) relationships demonstrating the footwalls of these thrusts were deformed prior to arrival of the overlying thrust sheet. While these structures developed during an overall interval of OOS deformation that swept from northwest to southeast across the foreland, we also note isolated intervals of simultaneous or alternating IS deformation within the FFTB. We cannot rule out the possibility that some of these well documented OOS thrusts initially formed as IS thrusts that were reactivated following the initial wave of foreland propagating deformation, with younger OOS thrusting obscuring their original IS character.

Northeast of and structurally higher than the Wiley dome, the asymmetric Sequatchie anticline becomes the northwesternmost structure of the FFTB, extending for ~400 km from central AL to northeast TN (Figure 2). The anticline has a steep northwestern limb cut by a thrust fault along much of its length [110, 111]. At its southwestern terminus, ~20 km east of the Wiley dome, this structure forms a broad, plunging detachment fold developed in synorogenic sedimentary rock, with growth strata indicating initiation of the anticline by the Early Pennsylvanian (late Bashkirian, ca. 320 Ma). Because this postdates the Late Mississippian Wiley dome, the structurally higher, thrust faulted Sequatchie anticline must have developed as an OOS structure relative to the Wiley dome [40].

Southeast of the Sequatchie anticline, the southeast verging Murphrees Valley anticline lies in the hanging wall of the northwest dipping Straight Mountain fault [37] (Figure 2). Thomas and Neathery [112] noted that differences in thickness of the Parkwood Formation northwest and southeast of the Murphrees Valley anticline suggest that formation of the anticline and displacement along the Straight Mountain back thrust must have occurred during the latest Mississippian, immediately after development of the sub-Parkwood Formation unconformity to the northwest. Accordingly, both structures are OOS relative to the Wiley dome, but older than the major sequence of thrust faulting at higher structural levels in the more proximal FFTB to the southeast.

Approximately 17 km southeast of the Wiley dome, the steep forelimb of the structurally higher, asymmetric Blue Creek anticline and the along-strike Opossum Valley thrust (Figure 2) were structurally detached in the Cambrian-Ordovician section and secondarily in the Mississippian-Pennsylvanian Parkwood Formation [111]. Using thermal maturity data from this metallurgical coal province to constrain the relationship between folding and thermal events, Pashin [40] demonstrated that the thermal maturation is postkinematic, indicating the Blue Creek anticline is younger (latest Bashkirian to early Moscovian, ~315 Ma) than both the Wiley dome and the southwest terminus of the Sequatchie anticline. To the southeast, the Jones Valley thrust (Figure 2) cuts the trailing limb of the Blue Creek anticline, postdating its formation. As such, the Jones Valley thrust is OOS relative to deformation in the Blue Creek and Opossum Valley thrust sheets (Figure 7(a)).

The Birmingham anticlinorium, a foreland structure cored by large ductile duplexes in the Conasauga Shale [113], lies northeast and along strike of the Blue Creek anticline at the northwest flank of the Cahaba synclinorium, but is cut by the Jones Valley and Opossum Valley thrust faults (Figure 2). The Pennsylvanian Pottsville Formation northwest of the Birmingham anticlinorium is primarily marine, whereas terrestrial rocks dominate the unit southeast of the Cahaba synclinorium. These sedimentological differences indicate the Birmingham anticlinorium generated enough topographic relief by the Lower-Middle Pennsylvanian (latest Bashkirian to early Moscovian, ~315 Ma) to isolate the adjacent Cahaba synclinorium from most of the marine transgressions on the northwest flank of the anticlinorium [40]. Therefore, the Birmingham anticlinorium and Cahaba synclinorium formed in the same time frame as the along-strike Blue Creek anticline and must be OOS relative to structures on its northwestern flank (e.g. Sequatchie anticline Figure 2).

Northeast of the Jones Valley fault, the Big Canoe Valley thrust cuts major folds like the Chandler Mountain syncline along the trailing margin of the Straight Mountain thrust sheet and thus is OOS relative to the older Straight Mountain thrust to the northwest [114]. The subhorizontal Rome thrust, the first thrust to be recognized in the southern Appalachians [115], lies structurally above the Cahaba synclinorium-equivalent Lookout Mountain syncline in the Wills Valley thrust sheet (Figure 2). The Rome thrust truncates the older Big Canoe Valley thrust (Figure 2), as well as folds in that thrust’s footwall, and is therefore considered to be OOS relative to both the underlying Wills Valley and Big Canoe Valley thrust sheets [37, 116]. The footwall Big Canoe Valley OOS folds are coaxial with the folded Rome fault surface, but the fault-truncated footwall beds are more tightly folded than the fault surface itself. This suggests the footwall folds may have been tightened by later IS deformation related to subsequent movement on the Wills Valley thrust, which folded the Rome thrust sheet along with its footwall [37].

The structurally higher Helena fault, marking the southeastern boundary of the Cahaba synclinorium, cuts older, more external faults, including the Rome thrust, along a lateral ramp at the northeast end of the synclinorium (Figures 2 and 7(d)). Thomas and Osborne [25] and Thomas and Bayona [37] identified the Helena thrust as a “rare” OOS foreland thrust, but data presented here indicate that it is one of many OOS thrusts in this foreland segment. Thermal maturity data from the footwall of the Helena thrust indicate fault propagation during Moscovian time (~312 Ma), consistent with its interpretation as an OOS thrust relative to the more northwestward structures described above [40].

The southeastern and most interior foreland structure to contain a significant section of the Pennsylvanian Pottsville Formation is the Coosa synclinorium in the hanging wall of the Helena thrust (Figure 2). Southeast of this structure, synorogenic sedimentary units have been eroded and only preorogenic rocks are present. Thus, the relative age of thrusts must be based upon cross-cutting relationships rather than constrained by the tectono-stratigraphy (Figures 2 and 7). The Dry Creek-Yellowleaf thrust (Figure 2) truncates the southeastern flank of the Coosa synclinorium and is considered to be OOS relative to the structurally lower Helena thrust sheet to the northwest [25, 117]. South of the Coosa synclinorium and west of the AL-GA state line, the western Coosa fault (Figure 2) also truncates the structurally underlying Helena fault and is OOS relative to it and the structurally lower Rome thrust [118].

In the southwestern FFTB of AL, in the hanging wall of the Dry Creek-Yellowleaf thrust sheet, the Coosa deformed belt contains folds that are truncated by the younger, structurally overlying OOS Pell City thrust, which is in turn cut by the structurally higher OOS Wilsonville fault in the southwest (Figures 2 and 4(a)) [25, 37, 116], as well as the eastern Coosa fault (Figure 2). The eastern Coosa fault also cuts the structurally lower western Coosa fault [118]. The Pell City and Wilsonville thrust sheets lie in the immediate footwall of the Talladega-Cartersville fault at the structural base of the Talladega belt hinterland in the southwestern FFTB and are cut by that structurally higher fault. Northeast of the Wilsonville thrust sheet, allochthons containing the oldest foreland stratigraphy (Lower Cambrian Chilhowee Group-Rome Formation) represent the structurally highest foreland thrust sheets in the region. These include the Choccolocco Mountain, Indian Mountain, and Johnson Mountain thrust sheets, which are bound at their base, respectively, by the Jacksonville, Indian Mountain, and Johnson Mountain thrusts (Figures 2 and 4(b)). The OOS Jacksonville fault cuts both the Pell City and eastern Coosa faults [112, 118, 119] indicating that it is the youngest thrust in this section of the FFTB (Figures 2 and 7(c)). Pelitic rocks in the eastern Coosa, Choccolocco Mountain, Indian Mountain, and Johnson Mountain thrust sheets each contain an anchizone cleavage that is axial planar to folds cut by the basal thrusts. This cleavage, which is penetrative into Mississippian rocks in the eastern Coosa thrust sheet, likely formed during earliest IS deformation [120].

Thus, the southwestern FFTB in AL and GA contains at least thirteen examples of major thrust faults that propagated in a backward-breaking, OOS sense, from the edge of the Appalachian Plateau (Wiley dome) to the most proximal parts of the foreland. These OOS structures form the major structural architecture of the thrust belt in this region, similar to that outlined by Faill [38, 39] for the north central Appalachian FFTB of Pennsylvania.

McCalley [121] was the first to map a thrust fault (Talladega-Cartersville fault) along the northwest flank of the Talladega belt, between it and the structurally underlying rocks of the unmetamorphosed to anchimetamorphosed FFTB (Figures 1, 2 and 4). In GA and AL, the postmetamorphic Talladega-Cartersville fault marks the foreland-hinterland boundary along a ~240 km long exposed segment of the orogen, continuing to the southwest beneath the Gulf Coastal Plain for another 200 km [122]. The Talladega belt allochthon in AL contains a complete section of the Lower Cambrian to Lower Ordovician Laurentian trailing margin shelf sequence, equivalent to that in the underlying foreland thrust sheets. In addition, the allochthon contains younger Silurian(?) to earliest Mississippian(?) units [47-49] (Figures 3 and 4) that served as footwall stratigraphy during emplacement of the older, premetamorphic Hillabee allochthon, all of which were metamorphosed to greenschist facies during the Mississippian (ca. 330 Ma) [52]. Northeast of a large, oblique thrust ramp at the latitude of Cartersville, GA (Figures 1 and 4(b)), the frontal hinterland fault system is the Great Smoky fault, which carries rocks of the western Blue Ridge in its hanging wall [123]. This part of the Talladega-western Blue Ridge allochthon contains isoclinally folded, Mesoproterozic Grenville basement massifs, as well as cover sequences of the Neoproterozoic Ocoee Supergroup and younger Paleozoic rocks. Southwest of the oblique ramp, the Talladega-Cartersville fault rises into basal Cambrian and younger units in the Talladega belt [123].

Palinspastic restoration and cross-sectional balancing of stratigraphy within the FFTB, beneath the Talladega-Cartersville fault, place the Talladega belt at least 190 km southeast of its current position, astride or southeast of the present position of Laurentian basement and cover rocks within the Pine Mountain window (Figure 1) [37, 90]. Based on overlap distances in half windows, the minimum horizontal component of net slip of the Talladega-Cartersville fault relative to its immediate footwall is 21–24 km (Figure 4(b)) [43]. This displacement, coupled with that of thrust sheets in the underlying FFTB, indicates that the cover stratigraphy in AL and GA was shortened by ~53% during late Paleozoic, Alleghanian telescoping of the Laurentian margin [37, 90]. To the northeast, in TN, the FFTB was shortened as much as ~70% [124]. The Talladega-Cartersville fault has a large stratigraphic throw (5-7 km), locally placing the Lower Cambrian Chilhowee Group above the Upper Mississippian Floyd Shale, and is likely a latest Pennsylvanian or Permian thrust [42, 43] (Figure 2). Over much of the region the fault dips ~10° southeastward, is marked by an abrupt change in metamorphic grade, and contains a well-developed shear band cleavage in immediately overlying hanging wall rocks with a top-to-the-northwest shear sense [125].

Because the Talladega-Cartersville fault postdates significant map-scale body deformation in both hanging and footwalls, its trajectory is not typical of foreland thrusts. Its ~240 km long exposed trace is marked by truncation of both hanging and footwall mesoscopic fabrics, stratigraphy, and map-scale folds [43, 44, 116, 123, 126]. For example, in the Rockmart, GA area, between the Tennessee salient and Alabama recess and within the southern Appalachian foreland’s deepest southeastward structural indentation (Figures 4(b) and 6), the Talladega-Cartersville fault cuts obliquely northwestward across its footwall for at least 24 km, truncating foreland stratigraphy and structures in the hanging wall of the eastern Coosa fault [120] (Figures 1 and 6). This foreland stratigraphy includes the Cambrian-Ordovician Knox Group, overlying Middle Ordovician rocks of the Blount clastic wedge (Rockmart Slate), the Devonian Frog Mountain Sandstone, and Mississippian Fort Payne Chert and Floyd Shale [120, 127] (Figures 3 and 6). These strata were folded into a regional-scale synclinorium within the eastern Coosa thrust sheet and were deformed by higher-order map-scale, doubly plunging, northwest-overturned folds with wavelengths of ~4 km. Pelitic rocks within the foreland strata contain a pervasive axial planar, subgreenschist (anchizone) slaty cleavage (Figure 8). Palinspastic restoration by cross-sectional balancing reveals fold-related, northwest-southeast apparent shortening within these easternmost foreland rocks of ~30%. Southwestward from GA into AL, the Talladega-Cartersville fault continues to cut obliquely through underlying foreland stratigraphy and structures along its trace, including the aforementioned folds, their axial planar cleavage, and metamorphic isograds (Figures 4(b), 6 and 8), progressively passing structurally downward through several major foreland thrust sheets toward the southwest [119]. These include the Choccolocco Mountain-Indian Mountain thrust complexes near the AL-GA border (Figures 2 and 4(b)) and the Wilsonville and Pell City thrust sheets west of Childersburg, AL (Figure 2 and 4(a)) [119]. Seismic interpretation of the Talladega-Cartersville fault also indicates that it truncates the Jacksonville fault in the subsurface [37, 128] (Figure 7(d)). All of these relationships demonstrate that the Talladega-Cartersville fault is OOS relative to all of the major thrusts along the foreland interior.

A similar relationship to that found in the Rockmart area occurs within the southwestern part of the Pell City thrust sheet near Harpersville, AL (Figure 8(a)), where the Talladega-Cartersville fault decapitates a set of map-scale, north-northwest-trending, west-verging, upright, subhorizontal, tight to isoclinal, doubly plunging folds of the Harpersville generation that deform Lower to Middle Ordovician rocks [43, 116, 129]. These decapitated folds have steep axial surfaces striking ~N25° to N30°W and record bulk regional shortening (47%–52%) within the Pell City thrust sheet prior to the emplacement of the Talladega-Cartersville fault [116].

Two regional salients that formed as a result of later cross-folding of the Talladega-Cartersville fault demonstrate some of the fundamental differences between the trajectories of typical thin-skinned foreland thrust faults and those associated with emplacement of thin-skinned hinterland crystalline thrust sheets. One salient extends from beneath the Gulf Coastal Plain to southeast of the city of Talladega, AL, and the other extends northeastward from Choccolocco Mountain to east of the AL-GA border (Figure 4). Within these two regional salients, rocks of the Talladega belt extend 10–15 km farther to the northwest than along other segments of the thrust. As such, the immediate hanging wall of the thrust is at a significantly lower stratigraphic level within the Talladega belt than outside of the salients, exposing up to 4.7 km of the Lower Cambrian Kahatchee Mountain Group. Between these salients, as well as northeast of the northern salient (Figure 4), the thrust cuts up section to near the base of, or into the Silurian(?)-Lower Devonian Lay Dam Formation, placing these younger rocks above older (Cambrian and Ordovician) underlying foreland rocks. These relationships indicate that the thrust cuts to higher stratigraphic levels in its hanging wall toward the southeast (down-dip) and requires that the thrust dip at a shallower angle than the overlying regional stratigraphy and generally coplanar slaty cleavage. This relationship is demonstrated southeast of Talladega, AL, near Ironaton, where a cross-fold plunging 31° S63°E indents the thrust’s trace ~1 km (Figure 4(a)). Using bedding cutoffs along the small recess, Lim [125] calculated a fault dip of ~7°, whereas the bedding and coplanar slaty cleavage dip ~30°. In a similar analysis, Erickson [130] calculated a fault dip of 13° in the area east of Columbiana, AL (Figure 4(a)).

An oblique down-dip view of the fault’s trajectory can be observed immediately east of the AL-GA state line, where the fault’s trace trends northwest, nearly parallel to the transport direction (Figure 4(b)). As elsewhere, it can be demonstrated that the fault cuts to significantly higher stratigraphic levels (up section) in the down-dip direction of the hanging wall, which indicates that the trajectory of the fault was not controlled by stratigraphy within the lower greenschist facies Talladega belt. This regional characteristic, in which the basal fault of the Talladega-Cartersville thrust sheet cuts down-section in the direction of displacement, is one of two unique relationships that Morley [27] cited for recognizing OOS thrusts and contrasts significantly with the behavior of archetypal thin-skinned thrusts in FFTBs.

In addition to its obliquity relative to the internal stratigraphy of the Talladega belt, the Talladega-Cartersville fault also cuts obliquely through the regional metamorphic isograds. The lowest metamorphic grade rocks are in the southwest, where a number of the units preserve distinctive fossil assemblages [43]. The biotite isograd crosses obliquely through the belt at the AL-GA state line and to the northeast the garnet isograd crosses the belt in the southern part of the Mulberry Rock recess (Figure 4(b)). Some segments along the northwest flank of the allochthon, for example northeast of Sylacauga, AL [131], and near the AL-GA state line [132,133], have been broken into footwall imbricate fans, repeating the structurally lower part of the Talladega belt stratigraphy in several thrust slices. Parts of these imbricate slices have been isolated as klippen above foreland rocks (e.g. Sleeping Giants range, Figure 4(a)) [131, 134].

Regional, three-dimensional views of the Talladega-Cartersville fault documenting its low dip result from folding of the thrust by two regional fold phases that produced major recesses (half-windows) along the thrust’s trace (Figure 4). The oldest phase, the Jemison-Columbiana generation, occurs in the southwestern part of the Talladega belt and underlying FFTB (Figure 4(a)) [43, 135]. These folds are gently doubly plunging (< 30°) to the NE-SW, with steeply dipping (NW and SE) curved axial surfaces trending 30° to 40° NE, and gentle to open (interlimb angles between 154° to 115°) with amplitudes of ~1.5 km and wavelengths up to ~5 km. Folding was accommodated by flexural slip along bedding and coplanar slaty cleavage. Fold orientation varies because they fold previously deformed, nonparallel surfaces above and below the basal thrust and are themselves refolded by a later regional fold phase (see following).

While the evidence presented herein demonstrates that the predominant thrust architecture of the foreland and northwest flank of the hinterland was largely the result of Pennsylvanian-Permian OOS deformation following an interval of latest Devonian(?)-Mississippian IS deformation, this does not preclude the possibility of an intervening period(s) of IS deformation during the overall OOS episode. For example, the Kelly Mountain antiform of the Jemison-Columbiana generation folds the foreland Pell City and Wilsonville thrust sheets, as well as the overlying hinterland Talladega thrust sheet (Figure 4(a)). Brewer [117] noted that the Talladega-Cartersville fault cuts across stratigraphic contacts on both limbs of this antiform in its footwall, suggesting that at least some of the structural relief on the antiform predates emplacement of the overlying Talladega, Pell City, and Wilsonville thrust sheets, and supports the interpretation that these three thrusts are OOS relative to older movement on the deeper Yellowleaf fault. It is clear, however, that the Talladega, Wilsonville, and Pell City thrust sheets experienced younger deformation coaxial with development of the antiform, which was responsible for much of its structural relief. This suggests the presence of a period of forward-breaking movement on blind imbricate thrusts carrying the Pell City, Wilsonville, and Talladega thrust sheets piggy-back during a period of younger, IS movement on the Yellowleaf (Eden) fault (Figure 4(a)), but after emplacement of the latter three thrust sheets [117, 130]. This structural sequence is similar to the scenario described above in which footwall folds were subsequently tightened by later IS deformation that folded the Rome thrust sheet along with its footwall. Importantly, the Jemison-Columbiana fold set does not deform the overlying Hollins Line fault system (figures 1, 2 and 4(a)) southeast of the Talladega belt (see following) [43, 135]. While the metamorphic fabrics and associated structures within the Talladega belt represent allochthonous deformation, the Jemison-Columbiana deformation was imprinted upon the allochthon following its emplacement along the Talladega-Cartersville fault.

In eastern AL, the Alleghanian Hollins Line fault is the boundary between greenschist facies rocks of the Talladega belt footwall on the northwest and upper amphibolite facies (kyanite-sillimanite grade) rocks of the Ashland-Wedowee-Emuckfaw belt hanging wall on the southeast (Figures 1, 2 and 4). The fault’s ~200 km long trace which forms the western-eastern Blue Ridge boundary in this segment of the orogen marks a significant topographic, structural, stratigraphic, and metamorphic discontinuity. The Ashland Supergroup in the immediate hanging wall is a distinctive sequence not seen elsewhere in the eastern Blue Ridge and includes graphitic quartzite, garnetite, roscoelite-bearing (vanadium muscovite) schist, and amphibolite [42]. The Ashland Supergroup represents either the late Neoproterozoic to Early Ordovician Laurentian outer margin slope-and-rise-sequence [53, 63, 71, 136-138] or the oldest exposed units within the WEDB back-arc basin [139].

The Hollins Line fault [71, 83, 135, 140-144] is the roof thrust of a postmetamorphic, dextral, transpressional footwall duplex fault system (Hollins Line fault system) (Figures 1, 4, 5 and 9). Imbricate splays connect the roof and floor thrusts and define a hinterland dipping duplex [3] up to 5 km in width. The roof thrust (Hollins Line fault) separates lower greenschist facies footwall rocks from the overlying upper amphibolite facies eastern Blue Ridge. Dextrally offset, footwall thrust duplex horses vary in length from 2 to 100 km (Figure 9) [3, 145], are typically stacked in an en échelon right sense, and contain the Hillabee Greenstone, premetamorphic Hillabee thrust, and rocks of the upper Talladega Group (Figures 4 and 9). Above the Talladega belt parautochthon and floor thrust, imbricate splays cut the metamorphic fabrics, but separate rocks of the same metamorphic grade [71, 83, 146-150] (Figures 5 and 9). Most of the dextrally offset duplex horses were detached near the base of the Devonian–earliest Mississippian(?) Jemison Chert/Erin Slate (Figure 3) and displaced the overlying Hillabee thrust and Hillabee Greenstone passively forward atop these younger units (Figures 5 and 9). Below the floor thrust, the parautochthon locally contains the Hillabee Greenstone and underlying Hillabee thrust. Rocks of the eastern Blue Ridge allochthon are absent below the roof thrust. Stratigraphy and metamorphic fabrics of the uppermost Talladega Group and Hillabee Greenstone in the Talladega belt footwall, as well as similar features in the structurally overlying Ashland Supergroup, are truncated by the Hollins Line system faults.

Although the Hillabee Greenstone exerted the dominant mechanical control on the footwall trajectory of the roof thrust, the splay thrusts most commonly cut obliquely through the Hillabee and sole in mechanically weak metapelites of the structurally underlying Jemison-Erin interval, which make up >56% of the rocks immediately above the floor thrust. Only locally does the floor thrust cut below the Jemison-Erin interval into the stratigraphically lower Butting Ram-Cheaha Quartzite (Figures 5 and 9), and only at one locality does the floor thrust cut beneath the latter unit into the Lay Dam Formation (Figure 5(a)). For most of their lengths, the imbricate splays joining the floor and roof thrusts parallel the internal stratigraphy of the horses. Once these imbricate faults join the floor thrust, stratigraphy within both the horses and the parautochthon are parallel to the floor thrust for long distances along strike (Figures 4, 5, and,9). For detailed geologic maps showing structural fabric measurements and topographic relationships along the entire Hollins Line thrust system, see [83].

At several locations within the parautochthon immediately below the floor thrust, and within the imbricate horses, postmetamorphic, map-scale antiform-synform pairs (half wavelength >1.5 km) have the same dextral shear sense as the stacked, en-échelon duplex imbricate slices (Figures 5 and 9), suggesting these folds are related to transpressional deformation within the Hollins Line fault’s footwall during eastern Blue Ridge emplacement, but before footwall imbrication of the upper Talladega Group and Hillabee Greenstone. These earlier folds were subsequently decapitated by the floor and/or roof thrusts (Figures 5 and 9), and in addition to the regional en échelon-right map pattern of the imbricate slices, indicate the Hollins Line fault system resulted from oblique dextral thrust motion [52, 71, 83, 148-150]. None of the thrusts in the Hollins Line duplex system (roof, imbricates, or floor thrust) followed the trajectory of the older, premetamorphic Hillabee thrust. In contrast, these duplex system thrusts cut obliquely across this earlier structure and units in its hanging and foot walls, indicating the earlier premetamorphic thrust was not a zone of postmetamorphic mechanical weakness when the Hollins Line system was active. Because of the earlier emplacement of the Hillabee allochthon, the Hollins Line footwall imbricate faults commonly place younger rocks (Devonian Butting Ram-Cheaha Quartzite and Jemison Chert-Erin Slate) structurally above older rocks (Ordovician Hillabee Greenstone).

In its hanging wall, the Hollins Line roof thrust cuts obliquely stratigraphically downward toward the northwest through >12 km of Ashland Supergroup lithostratigraphy (Figure 4(a)), indicating that rocks of the hanging wall stratigraphy exerted no mechanical control over the trajectory of the roof thrust. As noted previously, one of the important characteristics of OOS thrust faults is the observation that they often cut stratigraphically down section in the direction of displacement [27]. Unlike the mechanical isotropy of rocks above the roof thrust, the mechanical character of the Hillabee Greenstone appears to have exhibited significant control over the stratigraphic level of the roof thrust’s trajectory in the immediate Talladega belt footwall. As such, along ~82% of the trace of the Hollins Line roof thrust it abuts rocks of the Hillabee Greenstone in the footwall, with rocks of the Jemison Chert-Erin Slate making up the remaining ~18% of the structural top of the Talladega belt [52]. On average, the Hollins Line roof thrust dips 15°–20° to the southeast and juxtaposes mica schists, quartzites, and amphibolites of the Ashland Supergroup against structurally underlying chlorite-actinolite phyllites or greenstones of the Hillabee Greenstone. The contrast in weathering resistivity between the easily weathered greenschist facies mafic rocks in the footwall and the more resistant amphibolite facies mica schists and quartzites in the hanging wall commonly results in a distinct topographic escarpment along the Hollins Line roof thrust [52, 83].

Measured offsets along segments of the Hollins Line roof thrust deformed by major cross-folds of the Millerville-Childersburg generation (see following) indicate a >19 km minimum horizontal net slip component (Figure 4(b), displacement vector C). However, because the roof thrust juxtaposes rocks of significantly different metamorphic grade, with no evidence of stratigraphic continuity across it, the net slip on this fault must be greater than that of thrust faults to the northwest. As such, the Hollins Line roof thrust likely accommodates the most significant displacement among all faults northwest of its position within the hinterland.

Several features indicate that imbrication of the Talladega belt footwall was relatively late in the kinematic history of the Hollins Line fault system. First, correlation of stratigraphy, including the Jemison Chert-Erin Slate, Butting Ram-Cheaha Quartzite, and lower Hillabee metadacite sheet, between the imbricate horses as well as in the parautochthon, signifies limited displacement across the floor and splay thrusts. Second, the absence of a metamorphic break between the footwall duplex horses, but a significant metamorphic break across the roof thrust, suggests significant displacement occurred along the roof thrust before development of the footwall duplex. Finally, the decapitated dextral map-scale folds associated with the duplex indicate that rocks of the Talladega belt footwall recorded some degree of ductile deformation associated with the emplacement of the eastern Blue Ridge allochthon before brittle imbrication of the footwall by the imbricate and floor thrusts. Collectively, this indicates that the most significant displacement (>19 km) across the Hollins Line fault system occurred along the roof thrust, with imbrication of the Talladega belt footwall developing in the latest stages or even after the eastern Blue Ridge allochthon was emplaced near or at its current structural level. Mica schists (“button schist”) in the immediate hanging wall of the Hollins Line roof thrust, as well as within floor and imbricate thrusts, commonly exhibit asymmetric S-C fabrics that typically develop at angles of 15°–25° to the plane of bulk flow within a shear zone [151] and yield a top-to-the-northwest or west-northwest slip sense, consistent with distinct deflections of planar fabrics across the roof thrust [52, 83, 144].

Two regional folds of the Jemison-Columbiana generation (see previous), the Jemison synform and the Kelly Mountain antiform [43], fold the Talladega-Cartersville fault, the earlier Hillabee thrust, units within the Talladega belt, and rocks of three underlying foreland thrust sheets (Yellowleaf, Pell City, and Wilsonville) (Figure 4(a)). However, relationships near the Gulf Coastal Plain onlap indicate that these folds are decapitated by the Hollins Line fault system. For example, at the southwestern exposed extent of the thrust belt, the Hollins Line fault decapitates the Kelly Mountain antiform (Figure 4(a)). West of the Gulf Coastal Plain onlap, isolated stratigraphic inliers within surrounding Cretaceous sediments expose the Hillabee Greenstone and Jemison Chert. The positions of these exposures (Figure 4(a)) as well as data outlined in [85] demonstrate that the Hillabee thrust and underlying Talladega-Cartersville thrust were folded by regional folds of the Jemison-Columbiana generation, while the trace of the Hollins Line fault system is not deflected by this fold phase, implying the latter fault system is both younger than and OOS relative to the underlying Talladega-Cartersville fault. Consequently, the eastern Blue Ridge allochthon above the Hollins Line fault system must have been emplaced after the underlying Talladega belt had been thrust onto the foreland and folded by the earlier Jemison-Columbiana generation of regional folding (Figure 4).

A second regional cross-fold set, the Millerville-Childersburg generation, folds the Talladega-Cartersville fault and structurally underlying Wilsonville and Pell City thrust sheets in the proximal foreland, but is cut by younger forward-breaking displacement along the Yellowleaf fault in the central foreland [43] (Figure 4(a)). Unlike the earlier Jemison-Columbiana fold generation, these younger structures also fold the Hollins Line duplex and the overlying eastern Blue Ridge allochthon (Figure 4(a)) [43]. This indicates that the two regional cross-fold generations were separated by a major period of thrusting along the Hollins Line system. Axial traces of the second regional set of flexural-slip cross-folds trend generally west to northwest, at a high angle to the earlier Jemison-Columbiana generation. Millerville-Childersburg folds have amplitudes of several kilometers and wavelengths of ~16 km (Figure 4(a)), are gentle (interlimb angles 133°–143°), and plunge gently to moderately (20°-40°) to the southeast [43, 144]. Because folds of both the Jemison-Columbiana and Millerville-Childersburg generations affect nonparallel surfaces in separate thrust sheets (e.g. bedding) as well as intervening thrust faults, the folds’ axial traces commonly change trend at fault boundaries. In addition, the younger Millerville-Childersburg fold phase created interference patterns with the earlier Jemison-Columbiana structures and produced complex map patterns for both stratigraphic units and the trace of the earlier Talladega-Cartersville fault [43, 135] (Figure 4(a)). Superposition of the two regional fold phases resulted in Ramsay type-1 basin-and-dome interference patterns [152] in the traces of stratigraphic contacts and the Talladega-Cartersville fault. Because the fold axes of the two generations are nearly orthogonal, the slip line of the second generation is nearly parallel to the axial surfaces of the earlier one [43]. Where Jemison-Columbiana generation folds are crossed by folds of the Millerville-Childersburg generation, the earlier folds reverse plunge directions and the axial traces of the latter folds coincide with lines of axial culminations and depressions in the earlier folds (Figure 4(a)). This fold interference forms half-windows in the trace of the Talladega-Cartersville fault and the overlying Talladega belt. The axial culmination in the Kelly Mountain antiform (Figure 4(a)) is interpreted as an eyelid window [130], although this interpretation is open to debate [153]. As with the earlier Jemison-Columbiana generation of folds, mesoscopic, parasitic, postmetamorphic folds with a secondary spaced axial planar crenulation cleavage are also commonly associated with the map-scale Millerville-Childersburg generation folds.

The nearly linear trace (N47°E to N60°E) of the Allatoona thrust (Figures 1, 2 and 4) extends for at least 280 km, primarily as an internal eastern Blue Ridge fault, but locally as the eastern-western Blue Ridge boundary [44]. As such, it marks a significant change in the regional structural architecture of the orogen. Northwest of the fault’s trace, distinctive orogenic curvature of several thrust sheets is common at a variety of scales [123], with earlier faults folded by regional folds creating significant strike changes in their traces. Between the Allatoona fault and frontal Brevard zone fault (Abanda-Rosman fault), the eastern Blue Ridge forms a highly linear belt with a nearly constant width (28-38 km) for >400 km (Figures 1 and 2).

The Allatoona fault cuts structurally downward through tectonostratigraphy in its footwall northeastward from the towns of Goodwater to Hightower, AL, through the entire Ashland Supergroup, decapitating map-scale folds like those of the Millerville-Childersburg generation in both its hanging and footwall, and cutting the Hollins Line and Hillabee thrusts at the towns of Millerville and Hightower [44] (Figure 4). Northeast toward the town of Cartersville, GA, the fault excises nearly the entire western Blue Ridge/Talladega allochthon in its footwall and decapitates the regional Cartersville antiform in the foreland [44] (Figures 2 and 4). This antiform extends for ~110 km from the southeast edge of the Appalachian Plateau (the northwest margin of the foreland deformation front), southeastward to the Allatoona fault, crossing the entire width of deformed Laurentian platform rocks. These relationships indicate that the Cartersville antiform is younger than the ~N80°W Millerville-Childersburg fold set, which is itself cut by late movement on the Yellowleaf fault in the central foreland (Figure 4(a)).

From Cartersville to Dahlonega, GA, the Allatoona fault cuts down section in its footwall into the Neoproterozoic Great Smoky Group in the western Blue Ridge, becoming an internal eastern Blue Ridge fault again west of Dahlonega (Figure 2). Southwest of Dahlonega, the Jasper antiform folds the Talladega-Cartersville fault equivalent Great Smoky fault, the Murphy synclinorium in the western Blue Ridge, and the “ancestral” Allatoona and Hayesville faults in the eastern and central Blue Ridge [44], creating the regional “S” curve along this segment of the orogen [154]. West of Dahlonega, the Allatoona fault cuts the Jasper antiform, and the “ancestral” Allatoona and Etowah River thrust faults [44, 155] (Figure 2).

In its hanging wall, from Goodwater, AL, to approximately 45 km northeast of the AL/GA border at the Mulberry Rock recess, the Allatoona fault cuts up section through the 6–8 km thick Wedowee Group, completely excising the unit at the recess (Figure 4(b)). This recess, a breached eyelid window through the Allatoona and underlying and overlying thrust faults, exposes rocks in the core and flanks of an open, north-south antiform that indents the western/eastern Blue Ridge boundary ~20 km (Figure 4(b)). Exposures in the window demonstrate that the Allatoona fault has a very low regional dip (<20°), with a minimum of ~17 km horizontal net slip relative to its immediate footwall (Figure 4(b)) [44]. Along the window’s southern rim and northwest flank, the Allatoona fault cuts up through the older Burnt Hickory Ridge fault to form the structural base of the Wedowee Group to the southwest. However, along the window’s northeast margin, the Allatoona fault forms the structural base of the Dahlonega gold belt, a distinctive sequence extending northeastward into NC (Figures 2 and 4(b)). With the possible exception of the Abanda-Rosman fault along the northwest border of the Brevard zone, the above relationships demonstrate that the Allatoona fault is the youngest OOS thrust in the entire region.

Although the Allatoona fault displaced the Dahlonega gold belt 10’s of kilometers northwestward and is younger than the Hayesville fault, northeast of Dahlonega the northwest boundary of the gold belt with the central Blue Ridge allochthon has been previously interpreted as the Hayesville rather than the Allatoona fault [65, 155, 156]. Our work does not extend northeast of Dahlonega, but indicates that the Allatoona fault is OOS relative to all of the structures previously mentioned. Consequently, we consider the fault that forms the southeast boundary between the central Blue Ridge and Dahlonega gold belt (eastern Blue Ridge) northeast of Dahlonega (Figure 2) to be an unresolved issue in need of further study.

Southeast of the Allatoona fault in AL and GA, the linear trace (~N50°E) of the Abanda fault marks the structural base of the Brevard zone (Figure 1), lying in the same structural position as the equivalent Rosman fault in South Carolina and NC. Brittle cataclastic deformation along the northwest-directed, reverse slip Rosman fault postdates earlier phases of plastic deformation that produced Brevard zone mylonites [157]. O’Hara and Becker [158] obtained a Rb-Sr muscovite model age of 302.2 Ma from mylonitic Henderson Gneiss near Rosman, NC, which they considered to more accurately represent the time of Alleghanian retrograde Brevard zone mylonitization than an earlier ca. 273 Ma Rb-Sr age at the same locale [159]. Deformation along the Rosman fault postdates this latest Pennsylvanian mylonitization event [157]. Hatcher [160] considered the Rosman fault to be OOS relative to the NW-directed Blue Ridge-Piedmont megathrust sheet and is also likely OOS relative to the Allatoona fault, which would make the Abanda-Rosman fault the youngest and most southeastern OOS thrust fault along the west flank of the southernmost segment of the orogen (Figure 10)

Unlike the OOS hinterland thrusts described above, the oldest thrust in the region, the Lower to Middle Mississippian Hillabee thrust, emplaced a sheet of Ordovician volcanic rocks onto Laurentian shelf rocks (Figures 2 and 10) prior to metamorphism. The youngest Laurentian shelf rocks immediately beneath the Hillabee thrust include the Jemison Chert and facies equivalent Erin Slate. Age constraints for the Jemison Chert include a well-documented Middle-Upper Devonian faunal assemblage [47]. A Periastron reticulatum specimen, the petiole of a leaf recovered from the upper Erin Slate [49], suggests a Tournaisian age for the unit. However, based on work by Beck [161] and the presence of Veryhachium, a marine acritarch characteristic of Devonian and older rocks collected from the same outcrop as the rare Periastron [162], the Erin Slate could be as old as Devonian [52]. As such, we bracket the age of emplacement of the Hillabee allochthon between the latest Devonian and Late Mississippian, that is after deposition of the youngest sedimentary rocks (Famennian) but prior to metamorphism of the Talladega belt. ⁴⁰Ar/³⁹Ar white mica age constraints indicate the amalgamated Talladega-Hillabee allochthons were metamorphosed together at 334–321 Ma (Serpukhovian) [71-73] (Figures 2 and 10).

Using a combination of deformed and undeformed strata in the younger IS Wiley dome, Pashin [40] constrained formation of the domal uplift and related thrust faults to near the Mississippian-Pennsylvanian boundary (ca. 323 Ma). It is conceivable that the Murphrees Valley anticline and Straight Mountain fault developed during this same time period [112], regardless of whether or not they formed slightly IS or OOS relative to the Wiley dome. Pashin [40] used similar arguments (e.g. depositional age, relative relationships between deformed and undeformed strata, presence of growth strata, thermal maturity data) to argue for Bashkirian (323-315 Ma) development of the OOS Sequatchie anticline, with the younger Blue Creek, Birmingham, Cahaba, Helena, and Coosa structures (Figure 2) initiating during the Moscovian (315-307 Ma) (Figure 3).

While many of the aforementioned foreland structures have relatively restrictive age constraints for their initiation, other more proximal foreland and hinterland OOS thrusts do not. Many of these latter structures deform Mississippian and Pennsylvanian rocks, or earlier structures that deform those rocks. Other than these cross-cutting relationships, there are few absolute age determinations with which to directly delimit the timing of emplacement of the proximal foreland and crystalline hinterland thrust sheets (Figure 2). Paleomagnetic studies and the timing of illite recrystallization within these unmetamorphosed rocks have been used in attempts constrain the broad timing of thrust propagation in the FTTP [70]. Episodic basin-wide subsurface expulsion of hot saline solutions (connate brines) between 303 and 272 is interpreted to have been driven by enhanced topographic recharge and/or migration of metamorphic brines during loading of the foreland by hinterland thrust sheets [163-165]. These fluids resulted in precipitation of new magnetic domains, remagnetization of foreland sedimentary rocks [166], and recrystallization of illite [167]. Hnat and van der Pluijm [168] obtained similar ⁴⁰Ar/³⁹Ar illite ages (clay gouge) to those in the Elliott and Aronson [167] study from the Great Smoky, Knoxville, Copper Creek, and St. Clair thrusts in the Appalachian foreland of TN and Virginia and attributed them to fault motion at ca. 280–276 Ma (Early Permian). Hatcher et al. [8] suggested a somewhat broader age range (286-266 Ma) for Alleghanian thrusting of post-Carboniferous rocks. We regard all of the crystalline thrust sheets described above, with the exception of the older, premetamorphic Hillabee thrust, to have formed during approximately the same time span as regional fluid flow into the adjacent FFTB (Figure 2), purportedly between 303 and 266 Ma (Late Pennsylvanian to middle Permian).

Importantly, each of the hinterland thrust sheets that developed during this time frame were metamorphosed and deformed prior to their emplacement in the latest Pennsylvanian-Permian (Figures 10(c)–10(j)). The youngest plutons that contain the peak amphibolite facies metamorphic fabrics (defined by biotite, muscovite, epidote, and garnet) in the eastern Blue Ridge thrust sheet of AL and GA above the Allatoona fault yield U/Pb zircon ages of ca. 335 Ma (i.e. Visean) [56, 60]. Estimates of Alleghanian thrusting between 303 and 266 Ma [8, 166, 167] imply that these crystalline thrust sheets were metamorphosed and polydeformed 70–20 m.y. before emplacement. Thus, it is clear that emplacement of these hinterland allochthons occurred significantly after the thermal-metamorphic peak, as their bounding faults cut and telescope metamorphic isograds and do not parallel the isotherms as is common in overthrust sheets of crystalline rock [169]. Goldberg and Dallmeyer [170] obtained ⁴⁰Ar/ ³⁹Ar plateau ages and Rb-Sr white mica ages ranging from ca. 323 to 336 Ma in mylonite zones separating several internal Blue Ridge thrust sheets in NC, which is within the time frame of the IS Hillabee thrust and FFTB Wiley dome formation. As such, this Late Mississippian, internal Blue Ridge mylonitization could represent an early interval of IS crystalline thrust sheet assembly at the onset of the Alleghanian orogeny, prior to the younger OOS thrusting described here.

Although the Alleghanian crystalline thrust sheets described herein verged away (northwestward) from the collisional front southeast of the Brevard zone, telescoping metamorphic isograds and placing higher grade rocks on lower grade ones, there is no evidence in footwall mineral assemblages to suggest that regional downward heating occurred as a result of the emplacement of overlying “hot” thrust sheets. Thus, the overlying higher metamorphic grade thrust sheets must have cooled enough by the time of their emplacement that they could not conduct appreciable heat into footwall rocks, with retrograde metamorphism of hanging wall rocks in these thrust sheets limited to the immediate vicinity of the fault zones.

At typical orogenic cooling rates (~10°C/m.y) [171], crystalline thrust sheets of the hinterland should have been uplifted well above the brittle-ductile transition zone (BDTZ) and the realm of crystal plastic processes after peak Alleghanian metamorphism (ca. 330 Ma), but prior to thrusting (303-266 Ma). As a result, each postmetamorphic, hinterland thrust sheet was emplaced as a mechanically strong, relatively cold brittle slab during the latest Pennsylvanian or Permian. As an example, rocks of the Talladega belt cooled below muscovite closure temperatures of 350° ± 50°C between 334 and 321 Ma [72, 73]. Assuming that muscovite-bearing rocks in this belt reached their blocking temperature (~350° C) at ca. 330 Ma, and the temperature of the BDTZ in this cratonal setting was ~250–400°C, then rocks of this allochthon would have cooled to a temperature below that of the BDTZ (ΔT°<100°C) in less than 30 m.y. (i.e. the time gap between peak metamorphism and maximum age of foreland thrusting). Such a cooling/exhumation rate (~3.3°C/m.y.) is well within the lower range of that determined for other orogenic belts during uplift and thrusting [171-173]. This minimum cooling rate would have brought rocks of the allochthon well above the BDTZ by the time of thrusting, as indicated by the postmetamorphic nature of the Talladega-Cartersville and Hollins Line fault systems. A similar analysis can be made for the eastern Blue Ridge allochthons above the Hollins Line roof thrust. Rocks of the Wedowee and Emuckfaw Groups reached a peak metamorphic temperature of ~540°C at 331–320 Ma [60]. Cooling to a temperature below the BDTZ (ΔT° 140–290°C) over a 20–30 m.y. time frame requires rates between 14.5° and 4.7°C/m.y., well within the cooling rates documented for other orogenic belts during uplift and thrusting [171-173].

Major dextral shear zones within the southeastern Appalachian Piedmont suggest the Alleghanian collision of Gondwana and Laurentia was highly oblique, with early dextral translational deformation occurring in the southeastern parts of the hinterland [174-176]. By the latest Pennsylvanian to Early Permian, however, northwest-vergent, craton-directed, crystalline thrust sheets were propagating OOS toward and onto the southern Appalachian foreland [8, 160, 177], with the energy required for their emplacement supplied by tectonic collision [178]. The age of this collision-driven thrusting in the foreland and hinterland has been broadly interpreted as contemporaneous with deposition of a proximal synorogenic clastic wedge sequence, the Pottsville Formation (Figure 3). Interpreted as the earliest Alleghanian clastic wedge to form on the southeastern Laurentian shelf [74, 179, 180], the Pottsville Formation consists of a thick sequence (>3 km) of synorogenic sedimentary rocks (shale, mudstone, siltstone, sandstone, extra and intraformational conglomerate, and coal) derived from proximal uplifts along the southeast flank of the basin in response to thrust loading. Some workers have argued for a primary orogenic source in the Ouachita orogen, west-southwest of the basin [181]. However, other workers [182-184] noted that many of the sedimentary clasts within the upper Pottsville Formation of AL are similar to rocks in the western Blue Ridge/Talladega belt, suggesting a proximal orogenic highland on the basin’s southeast flank. Recent studies confirm that high grade metamorphic clasts in the Pottsville are inconsistent with a source in the low-grade Ouachita orogen and that detrital muscovite Ar-Ar ages are more consistent with a source from the southern Appalachian Blue Ridge and Piedmont terranes [180, 184]. Detrital muscovite ⁴⁰Ar/ ³⁹Ar ages as young as 310 Ma in the lower Pottsville Formation and a ca. 295 Ma metatonalite clast from the upper part of this unit constrain deposition of this synorogenic clastic wedge to between Early Pennsylvanian and earliest Permian [184] (Figure 3). This implies that deposition of this clastic wedge occurred after metamorphism of the adjacent Blue Ridge and the IS thrust faults (e.g. Hillabee thrust) but coeval with initiation of the major phase of OOS thrusting within the foreland [166, 167].

Although the thicknesses of the Alleghanian allochthons under discussion range from a few to >10 km, their areal extents (>6000² km) demonstrate that they behaved as “thin-skinned” thrust sheets whose basal thrusts soled primarily above the underlying Mesoproterozoic, Laurentian (i.e. Grenville) basement. In contrast to the classical detachment zones in the FFTB or the earlier Hillabee allochthon, where subhorizontal, mechanically weak rocks could localize and propagate the detachment, these crystalline thrust sheets were internally strong, brittle slabs of moderately dipping, intact cover sequences that had been previously metamorphosed and polydeformed during earlier Alleghanian orogenic pulses. As such, these crystalline thrust sheets were deformed within the same orogenic cycle and are not, therefore, traditional “basement.” Their compressive strengths must have been sufficiently high enough (i.e. high coefficients of internal friction) to allow for the propagation of major low-angle shear fractures into allochthon-bounding thrust faults. Those cratonward-propagating thrust faults cut obliquely through earlier fabrics and lithostratigraphy over distances of tens of kilometers in the direction of displacement and laterally across wide areas hundreds of kilometers in length. These crystalline thrust sheet-bounding faults were generally unable to follow typical ramp-flat trajectories common to foreland thrust sheets because earlier polydeformation of mechanically weak zones significantly limited the spatial extent over which these surfaces could remain planar. The low-angle trajectories of these hinterland crystalline thrusts were essentially unaffected by obvious internal mechanical anisotropies like compositional layering, metamorphic fabrics, or older thrusts boundaries. In their hanging walls, the thrusts cut obliquely at angles between 20° and 40° through moderately southeast dipping lithostratigraphy and metamorphic fabrics. They were only affected by the regional stress fields at the time of their formation, as if the stress fields were acting on relatively isotropic materials with few apparent internal mechanical anisotropies.

Only in the footwall of the Hollins Line thrust did compositional layering seem to have a significant influence on the trajectory of the fault. After propagating northwestward toward the craton for a distance >20 km, cutting obliquely through internal stratigraphy and fabrics in both its hanging and footwalls, the thrust soled at the structural top of or within Hillabee Greenstone metavolcanic rocks in the footwall (Talladega belt), where the stratigraphy and metamorphic grade significantly differed from rocks in the hanging wall. The Hollins Line roof thrust is predominantly restricted to the Hillabee Greenstone in its footwall along a strike length >250 km, and across strike for >20 km, with the footwall metavolcanic sequence crudely approaching the geometry of a footwall flat (Figure 4). This geometry suggests that despite the earlier greenschist facies metamorphism and deformation in the Talladega belt, the Hillabee metavolcanic sequence formed a regionally extensive, semi-planar surface along and across the footwall of the far-traveled, eastern Blue Ridge allochthon.

For each of the crystalline thrust sheets discussed above, we see no evidence that any significant internal deformation within the thrust sheets themselves was related to their propagation or emplacement, that is thrust-related meso- and microfabrics appear to have been concentrated mainly within or near the fault zones themselves. The intensity of the earlier, prethrusting deformation appears uniform throughout each thrust sheet and is not concentrated in those structural sections adjacent to bounding thrust faults. The thermal regime of the relatively low temperature fault-zone fabrics at the base of the crystalline thrust sheets, where only growth of secondary sericite and chlorite is observed, likely represents the up-dip equivalent of more ductile, down-dip fabrics that are not observed at current structural levels.

In general, the kinematic and mechanical modeling of modern collisional orogens suggests that they developed a critically tapered wedge geometry, with faults in these thrust belts propagating in a forward-breaking sequence (i.e. IS) from the hinterland to the foreland over time [3, 185, 186]. Similar ideas related to critical taper theory have been applied to thrust belts in ancient orogens like the Appalachians. However, because many of the key variables involved (e.g. dip and strength of the basal decollement; slope, erosion rate, and flexural subsidence rate of the paleotopographic surface [12, 18, 24, 187]) cannot be accurately known for these orogens, caution should be taken in such application [188]. For example, it has been suggested that the composite Appalachian Blue Ridge-Piedmont crystalline mega thrust sheet, the frontal allochthons of which are addressed here drove the foreland deformation in front of it [8, 188]. In this interpretation, the Blue Ridge-Piedmont composite thrust sheet must have detached at the BDTZ near the edge of the Laurentian plate, ramping upward to the northwest toward the craton along an extensive low-angle décollement at the base of the deforming orogen. In this model, the thrust fault at the base of the composite thrust sheet then ramped upward into rift and then drift-facies sedimentary rocks (e.g. Lower Cambrian Chilhowee Group), eventually propagating into higher and higher detachments onto the platform sequences of the foreland. To the southeast, within the frontal metamorphic allochthon (western Blue Ridge) along part of the Tennessee salient [74], the detachment is, as expected, at a deeper stratigraphic level, within the Grenville basement and its Neoproterozoic cover (Ocoee Supergroup). However, in the Talladega belt to the southwest (i.e. Alabama recess), the detachment is at a similar stratigraphic level as that of the eastern foreland thrusts, within Chilhowee Group-equivalent units (Kahatchee Mountain Group), and locally within rocks of the Silurian-Devonian Talladega Group [51]. This relationship emphasizes fundamental differences in the mechanical response to the Alleghanian collisional events between the Alabama recess and the Tennessee salient [123].

The western Blue Ridge-Talladega belt allochthon and all thrust sheets to its northwest contain Laurentian rifted-margin and/or more cratonward trailing-margin shelf sequences formed on relatively thick continental crust. However, once the Hillabee allochthon and thrust sheets of the eastern Blue Ridge appear in the thrust stack (Figure 10), the décollement generally switched to a much higher stratigraphic level in Ordovician units. The bounding faults of these crystalline thrust sheets did not propagate along the basal detachment of the deforming platform wedge, in contrast to expectations for internally brittle slabs of intact crust (composite basement) that detach within the BDTZ (i.e. “Type C megathrusts” of [189]). These eastern Blue Ridge thrust sheets are interpreted to have dismembered an extensive, Laurentian-margin fringing Middle Ordovician back-arc basin (WEDB) that formed in a palinspastic position southeast (oceanward) of the continental margin hinge zone on thinned Laurentian continental crust (Figure 10(a)). The deep-water sequences of this marginal basin likely formed at lower elevations than the older sequences northwest of the continental hinge zone [53] and are now incorporated into major thrusts sheets that have undergone tens of kilometers of displacement along thrust surfaces with very low regional dips. In this region, none of the thrusts, with the exception of the Great Smoky fault, contain evidence that they incorporated crystalline Proterozoic basement rocks as classic “thick-skinned” thrust sheets. It is likely, however, that during the late Paleozoic, the western Blue Ridge external Grenville basement massifs were already allochthonous blocks that had become detached along extensional listric faults during earlier Neoproterozoic rifting [190, 191] or during subsequent Ordovican back-arc extension [53, 63]. This implies that during later compressional events, the external massifs may have been reactivated and inverted, becoming incorporated into thrust sheets which soled near the top of autochthonous basement.

Most of the foreland thrusts and all of the hinterland thrusts described here, with the exception of the Hillabee thrust, formed as OOS faults (Figure 10) that propagated through previously deformed and/or metamorphosed sequences. Other than the Hillabee thrust, hinterland thrusts postdate major faulting and associated deformation in most of the more cratonward FFTB (Figures 2 and 10). There is no evidence, given their trajectories, that these crystalline thrusts involved reactivation of earlier thrusts, as may have been the case in some of the true thin-skinned thrust sheets of the FFTB. Nevertheless, such OOS thrusts are not uncommon in orogenic belts and might be considered part of a normal deformation sequence [27]. For example, analysis of thrust kinematics and timing constraints from the Subalpine German Molasse Basin in the northern Alps revealed the presence of OOS thrust faults that were likely related to changes in the critical taper of the orogenic wedge as a result of increased erosion along the orogenic front [41]. In that analysis, Lower Miocene shortening associated with IS thrusts led to critical taper [12], with the Alpine orogenic wedge propagating into the foreland. With increasing erosion in the northern Alps between 10 and 5 Ma [192-194], accompanied by a decreasing slope along the orogenic front, thrusting moved back into the hinterland, with OOS thrusts increasing the orogenic slope in an attempt to regain critical taper [41].

Central Pangea, along with the ongoing southern Appalachian orogen, moved northward during the Mississippian Period through dry latitudes, but had drifted into wet latitudes of the paleointertropical convergence zone by Early Pennsylvanian [195]. During the Pennsylvanian and Early Permian, this was a region of significant peat accumulation (i.e. coal measures), as well as laterite/bauxite formation. These and other paleoclimate indicators suggest that during this time, the southern Appalachians recorded significant precipitation (>100 cm/yr) in excess of evapotranspiration [196]. In fact, paleosol studies from eastern North America suggest mean annual precipitation of ~130 cm/yr [195, 197] during the Early Pennsylvanian, similar to precipitation levels in modern tropical-subtropical climates. Kahmann and Driese [198] used Indonesia as a modern analogue for the Appalachians and the transition from monsoonal wet-dry conditions during the latest Mississippian (Chesterian) to the “ever-wet” conditions of the Pennsylvanian coal swamps. Changes in mean annual precipitation estimates from paleosol proxies [198] support interpretation of a significant climate shift from the latest Mississippian to earliest Pennsylvanian, with humid, “ever-wet” conditions and abundant rainfall presumably resulting in accelerated erosion of the rising Appalachians from Mississippian to Pennsylvanian times.

Assuming that development of the Hillabee thrust and Wiley dome correspond with an initial interval of IS thrusting (see previous), the southernmost Appalachian FFTB orogenic front must have been at the critical taper angle during much of the Mississippian. The unconformity within the Wiley dome at the base of the Mississippian-Pennsylvanian Parkwood Formation [40] indicates exposure of that structure at the surface during the transition from IS to OOS thrusting described herein. Timing of the sub-Parkwood unconformity is coeval with a climatic shift from the drier Mississippian to wetter Pennsylvanian. We argue that this latest Mississippian-earliest Pennsylvanian climatic transition resulted in erosion rates that exceeded uplift rates along the Appalachian orogenic front and subsequent deposition of the thick (>3 km) Pottsville Formation clastic wedge (Figure 3). Increased erosion along the critically tapered orogenic surface then reduced the topographic slope, bringing the orogenic front into subcritical taper, and resulted in development of OOS thrust faults from the distal foreland backward into the hinterland. Miocene erosion along the Alpine orogenic front and subsequent OOS thrusting in that system [41], therefore, is an excellent analogue for the Pennsylvanian transition from IS to OOS thrusting in the southernmost Appalachians, with both systems serving as valuable models for the evolution of IS and OOS thrust faults in modern and ancient orogens.

Using structural characteristics, absolute and relative timing constraints, and the tectonic history of lithotectonic belts within the southern Appalachian orogen, a number of conclusions can be drawn with respect to development of the overall thrust architecture.

1. Using structural, stratigraphic, and thermal maturity data, it can be shown that most of the structural architecture of the FFTB in an ~300 km long, 90 km wide segment of the southernmost Appalachians in AL and GA resulted from thin-skinned thrust sheets being assembled in an OOS fashion.

2. This OOS thrust progression continues for at least another 60 km into the thrust sheets of the metamorphic hinterland in the Blue Ridge/Talladega belt. We cite evidence for as many as seventeen major thrust faults between the Appalachian Plateau on the northwest and the Brevard zone on the southeast that formed in a backward-breaking sense relative to structures in their footwall (Figure 10). These thrusts account for the vast majority of documented tectonic shortening along this segment of the orogen.

3. Once OOS thrusting stepped rearward into the hinterland crystalline thrust sheets, the geometry of the thrust trajectories changed significantly. Unlike the classic thin-skinned thrust sheets of the foreland, the younger hinterland faults that emplaced the crystalline terranes of the Talladega belt and eastern Blue Ridge were not significantly influenced by primary or secondary planar fabrics. Only in the case of the Hollins Line roof thrust did a stratified sequence (Hillabee Greenstone) significantly affect the trajectory of the fault along its footwall. However, these late-stage crystalline thrust sheet-bounding faults clearly soled above the level of underlying Mesoproterozoic basement and dismembered stratigraphic sequences that were metamorphosed in an earlier pulse of the same Alleghanian orogenic cycle. Thus, the latest thrusts (Talladega-Cartersville, Hollins Line fault system, Allatoona, and Abanda-Rosman faults) can neither be considered archetypal thin-skinned thrusts, basement involved thin-skinned thrusts, or thick-skinned thrusts [5]. Instead, they might be considered “thin-skinned crystalline thrusts” and likely represent an intermediate form between faults that form in highly anisotropic rocks at shallow depths (thin-skinned thrusts) and faults that sole in the deep crust (Moho?), translating significant amounts of relatively isotropic basement toward the foreland in the process.

4. The hinterland thrusts generally dip at low angles (20° or less) and cut obliquely through moderately southeast-dipping lithostratigraphy, metamorphic isograds, and planar metamorphic fabrics, causing them to commonly cut down-section in the direction of displacement. Development of each of the lower two hinterland thrusts (Talladega-Cartersville and Hollins Line) was followed by a period of regional cross folding likely related to oblique or lateral ramps at depth (Figures 10(h) and 10(i)). These folds were in turn decapitated by the next structurally higher OOS thrust.

5. The next structurally highest thrust, the Allatoona fault, cuts a regional cross antiform (Cartersville antiform) that extends northwestward for ~110 km to the southeast edge of the Appalachian Plateau and crosses essentially the entire width of deformed Laurentian platform rocks, making this OOS thrust (along with possibly the Abanda-Rosman fault to the southeast) the youngest fault in the regional kinematic sequence (Figures 1 and 10(j)).

6. The transition from IS to OOS thrust faulting takes place at or near the Mississippian-Pennsylvanian boundary and is coincident with a well-recognized climatic transition from monsoonal to ever-wet conditions in this segment of Pangea that likely drove erosion rates to exceed uplift rates, reduced the slope of the topographic front, drove the orogen from critical to subcritical taper, and is temporally associated with deposition of the synorogenic Pottsville Formation clastic wedge.

All data utilized in this manuscript have been previously published in a wide variety of journal articles, theses, and dissertations, or are included in the manuscript text and figures.

The authors declare that they have no conflicts of interest.

Fumding sources include: National Science Foundation grants EAR-0309284 and EAR-1220540, U. S. Geological Survey grants 01HQAG0160, 02HQAG0100, O3HQAG0100, O4HQAG093,05HQA0107, O7HQAG0139G09AC00336, G12AC20240, G13AC00202, G13AC00202, 1434-HQ-96-AG-01538.

Research related to this study has been supported by the U. S. Geological Survey, National Cooperative Geologic Mapping Program under assistance Awards, including Nos. 01HQAG0160, 02HQAG0100, O3HQAG0100, O4HQAG093,05HQA0107, O7HQAG0139G09AC00336, G12AC20240, G13AC00202, G13AC00202, 1434-HQ-96-AG-01538, and the National Science Foundation (EAR-0309284 and EAR-1220540). We thank Bob Hatcher for a very helpful review of an earlier draft of this manuscript, and two anonymous reviewers for helpful comments.