The eastern edge of the Appalachian orogen is composed of a collection of Neoproterozoic – early Paleozoic domains, Avalonia, Carolinia, Ganderia, Meguma, and Suwannee, which are exotic to North America. Differences in the geological histories of these peri-Gondwanan domains indicate that each separated independently from Gondwana, opening the Rheic Ocean in their wake. Cambrian departure of Ganderia and Carolina was followed by the Ordovician separation of Avalonia and Silurian separation of Meguma. After separation in the early Paleozoic, these domains constituted the borderline between the expanding Rheic Ocean and contracting Iapetus Ocean. They were transferred to Laurentia by early Silurian closure of Iapetus and Devonian–Carboniferous closure of the Rheic Ocean during the assembly of Gondwana and Laurentia into Pangaea. The first domain to arrive at Laurentia was Carolinia, which accreted in the Middle Ordovician during the Cherokee orogeny. Salinic accretion of Ganderia occurred shortly thereafter and was followed by the Acadian accretion of Avalonia. The Acadian orogeny was immediately followed by Middle Devonian – Early Carboniferous accretion of Meguma and possibly Suwannee which led to the Fammenian orogeny. The episodicity of orogeny suggests that the present location of these domains parallels their order of accretion. However, each of these crustal blocks was translated along strike by large-scale Late Devonian – Carboniferous dextral strike–slip motion. The breakup of Pangaea occurred outboard of the Paleozoic collision zones that accreted Carolinia, Ganderia, Avalonia, Meguma, and Suwannee to Laurentia, leaving these terranes appended to North America during the Mesozoic opening of the Atlantic.
The peri-Gondwanan realm is a collection of terranes along the eastern flank of the Appalachian–Caledonian orogen (Hibbard et al. 2006b). First recognized on the Avalon Peninsula in eastern Newfoundland in the mid 1960s (Williams 1964), these domains have been recognized as being far-traveled terranes that record a geologic and tectonic history that predates inception of the Iapetus Ocean and are therefore exotic to native Laurentian rocks elsewhere in the orogen (O’Brien et al. 1983; Williams 1984). Along-strike comparison of these exotic sequences from Florida northward to Scotland (Fig. 1), has identified several diverse crustal elements that are considered to have formed along the northern margin of Gondwana in the Neoproterozoic and early Paleozoic (Hibbard et al. 2007b; Murphy et al. 2006 and references therein). Each of these domains (Avalonia, Carolinia, Ganderia, Meguma, and Suwannee), collectively termed the peri-Gondwanan realm, were separated from Gondwana in the early Paleozoic by opening of the Rheic Ocean where they constituted the boundary between the expanding Rheic Ocean and contracting Iapetus Ocean (Murphy et al. 2010). After separation, these terranes were transferred to Laurentia during the Early Silurian closure of Iapetus as well as during the Devonian–Carboniferous closure of the Rheic. Ultimate closure of the Rheic Ocean produced the climactic Variscan–Alleghanian–Ouachita orogeny during the assembly of the supercontinent Pangaea (Murphy and Nance 2003; Hibbard et al. 2010; Nance et al. 2010).
During the past few years, there has been a dichotomy of thought with reference to the paleogeographic history and lithotectonic integrity of these Appalachian domains that formed along the northern Gondwanan margin. Traditionally, these domains were grouped together and interpreted to belong to a single tectonic terrane that faced an open ocean that predates Iapetus (Rast and Skehan 1983; Hatcher 1989). Subsequent researchers, however, have shown that some of these terranes have very distinct depositional and tectonothermal histories and were therefore most likely tectonically decoupled forming individual and separate crustal blocks at some time in the Paleozoic (Barr and Raeside 1989; O’Brien et al. 1996; Hibbard et al. 2007a; van Staal et al. 2009).
In recent years, this controversy has been simmering and remains unresolved — do these domains represent a collection of independent crustal blocks that traveled separate or did they travel together as a single coherent tectonic entity? In this paper we compile and summarize the relevant geological data pertinent to the paleogeography of these lithotectonic components to determine the relationships between them. A major problem in constructing accurate early Paleozoic paleogeographic reconstructions is that critical relationships on relative motions may not be preserved due to the relatively short lifespan of oceans that document past relative plate motions (Dewey 1969). Our approach examines the detailed geological record of these exotic peri-Gondwanan terranes to track their tectonic travel path from a source location to their present position (e.g., Hibbard et al. 2005). This type of paleogeographic analysis typically relies on comparison of the lithologic, magmatic, and tectonothermal histories to distinguish rock affiliations that were formed in a common tectonic setting during a finite time span. Therefore, any independent crustal block should show faunal provinciality, paleomagnetic, isotopic, geochronologic, and (or) sedimentologic evidence of having experienced temporally unique depositional, magmatic, metamorphic, and deformational events. The implications of these data and resolution of these problems are fundamental to understanding the development and geometry of the Rheic Ocean, formation of the Appalachian–Caledonian orogen and global paleogeography.
It is fitting to examine the paleogeography of the Appalachian peri-Gondwanan realm in this special issue dedicated to E.R. Ward Neale. In a seminal 1963 paper, Neale demonstrated that significant lateral transport occurred in the allochthonous rocks of western Newfoundland thereby refuting the fixist conceptions of mountain building (Rodgers and Neale 1963). Neale’s identification of klippe in Newfoundland before the wide acceptance of continental drift and plate tectonics was a critical observation that led to the recognition (Williams 1964) of the two-sided geologic nature of the Appalachian orogen.
Definition and extent
Carolinia is composed of a collection of Neoproterozoic – early Paleozoic magmatic arc and sedimentary terranes exposed for more than 600 km in an elongate northeast–southwest-trending belt from central Virginia to Alabama (Fig. 2). Although the geological evolution is complex and only partially known, a simplified history of Carolinia, condensed from all constituent terranes, is subdivided into a tripartite succession consisting of (i) an old arc phase (ca. 670 Ma) containing the oldest dated rocks which are preserved in the fault-bounded Roanoke Rapids complex; (ii) an episode of younger, juvenile arc magmatism, including the Hyco arc, that formed in an open oceanic setting between 630 and 610 Ma (Samson et al. 1995; Wortman et al. 2000) and are overlain by younger clastic sedimentary rocks; and (iii) a mature phase of arc magmatism and related intraarc sedimentation that formed from ca. 560 to at least ca. 532 Ma and is preserved in the Albemarle magmatic arc (Hibbard et al. 2002). Carolinia is defined as including all exposed, proven peri-Gondwanan rocks of the southern Appalachians that lie east of the central Piedmont shear zone and extend to the Atlantic coastal plain. Along its northern margin Carolinia surrounds on three sides and is in tectonic contact with the ca. 1 Ga Goochland terrane, which has been interpreted as either a structural window of Laurentian basement or an exotic terrane (Hibbard et al. 2002).
The Neoproterozoic tectonomagmatic history is characterized by multiple composite magmatic arc terranes that formed in an open ocean prior to the opening of Iapetus. Coeval calc-alkaline igneous activity, present along the entire northwestern margin of Gondwana, extending from South America through Atlantic Canada to West Africa and central Europe, places Carolinia along this margin. A peri-Gondwanan setting is also indicated by Middle Cambrian rocks of the upper Asbill Pond Formation which contain trilobite genera (Ptychagnostus atavus) that have strong affinities with cool-water faunas of Armorica and Avalonia, but little affinity with warmwater, shelf faunas of Laurentia (Samson et al. 1990). In many cases, notably species of Paradoxides, these fauna are conspecific with those from contemporaneous rocks in other peri-Gondwanan terranes.
Although faunal and geological data indicate that Carolinia was positioned along the margin of west Gondwana, the absence of reliable paleomagnetic data and a more robust faunal record precludes using these methods to determine its source craton in Gondwana. However, characterization of the basement provides a potential paleogeographical linkage between Carolinia and its former cratonic source. The age and nature of the unexposed basement beneath Carolinia is inferred from isotopic and geochronological data. Volcanic and volcaniclastic rocks of the Hyco arc yield ages of 965–1229 Ma from inherited zircons (Müeller et al. 1994) and whole-rock samples have predominantly positive εNd values (+2 to +7) and depleted mantle model ages in the range 0.8 to 1.1 Ga (Samson et al. 1950; Wortman et al. 2000). Detrital zircons in clastic sedimentary rocks from Carolinia reveal clusters of ages at 0.6–0.65, 0.8–1.3, 1.5–1.8, and 2.2–2.6 Ga (Pollock et al. 2010). The combined Sm–Nd and U–Pb database suggest that Carolinia developed upon Mesoproterozoic basement that contained a sedimentary–volcaniclastic cover derived in part from adjacent Paleoproterozoic–Archean highlands. A temporally and isotopically similar basement is present in the Tocantins Province of central Brazil (Nance and Murphy 1996; Pimentel and Fuck 1992), which suggests that Carolina formed proximal to the Amazon craton and its peripheral orogenic belts on the margin of west Gondwana.
Time of rifting
The absence of useful biostratigraphic index fossils, reliable paleomagnetic data, and stratigraphic control preclude an accurate time of rifting of Carolinia from Gondwana. The maximum age of departure is constrained by the presence of Middle Cambrian trilobites in the Asbill Pond formation (Samson et al. 1995), which places Carolinia at high latitudes along the margin of Gondwana. The minimum age of departure is constrained by a Late Ordovician (ca. 455 Ma) tectonothermal event responsible for upright folding and greenschist facies metamorphism in the Albemarle magmatic arc (e.g., Offield et al. 1995) that has been interpreted (see the following) to be related to the accretion of Carolinia to eastern Laurentia.
Geological, geochemical, and isotopic data (Hibbard et al. 2002; Pollock and Hibbard 2010) suggest that the Neoproterozoic – Early Cambrian evolution of Carolinia was the result of arc magmatism related to subduction along the margin of west Gondwana. This magmatism is represented by the rocks of the Neoproterozoic Hyco magmatic arc and is unconformably overlain by a thick succession of marine volcaniclastic sedimentary rocks of the Ediacaran – earliest Cambrian Albemarle magmatic arc (Pollock et al. 2010). Mafic rocks of the Stony Mountain intrusive suite (SMIS) intrude all units of the Albemarle arc. Field relations indicate that the SMIS represents the final phase of magmatism following the eruption and deposition of the Albemarle Group, yet the suite predates Late Ordovician regional metamorphism and deformation.
The SMIS has a geochemical and isotopic signature that suggests juvenile magmatism (εNd +2 to +3) derived from hydrous partial melting of a mixed lithospheric–asthenospheric mantle source beneath an active island-arc (Pollock and Hibbard 2010). Rocks of the SMIS are comparable to MORB-like to OIB-type enriched rocks from the modern Lau Island and Sumisu Rift in the Izu–Ogasawara arc and are interpreted to have formed within an evolving Early Cambrian island arc–back-arc rift–basin system in Carolinia.
The recognition of an Early Cambrian arc rift to back-arc basin setting for Carolinia is significant because rifting of the arc may have been in response to, or induced, the initial opening of the Rheic Ocean between the Carolinian microcontinent and Gondwana. The rapid Paleozoic opening of the Rheic Ocean was probably achieved by hinge retreat of the subducting oceanic lithosphere and was synchronous with a polarity flip and the initiation of subduction beneath Laurentian elements along the northern margin of Iapetus. The increasing slab pull forces in Iapetus together with slab roll back beneath the arc in the Rheic Ocean were responsible for the transfer of Carolinia from Gondwana longitudinally into and across the contracting Iapetus Ocean. This model is analogous to the movement of the Cimmerian microcontinent across PaleoTethys and opening of Neo-Tethys in the Cenozoic (Stampfli and Borel 2002; Murphy et al. 2010). Alternatively, Keppie et al. (2003) proposed a scenario in which the transition from arc to rift geodynamic setting in Carolinia occurred as a result of ridge–trench collision in which the trench was replaced by a transform fault as the ridge was overridden, leading to a switch from arc-related to rift-related magmatism. As the ridge-trench-transform triple point migrated, the complex interplay between extensional forces caused the triple point to become unstable leading to oblique extension along short spreading centers and the separation of Carolinia in a manner similar to the modern opening of the Sea of Cortez.
Travel path and time of accretion
Carolinia contains Paradoxides and other Middle Cambrian trilobites (Samson et al. 1990) not known from early Paleozoic Laurentia, which places Carolinia along the northern margin of Gondwana. There are no available paleomagnetic data that can be used to determine an accurate travel path for Carolina. However, two independent paleomagnetic studies (Vick et al. 1987; Noel et al. 1988) of the Albemarle arc, place Carolinia at low southerly latitudes (22°S) near the eastern Laurentian margin in the Late Ordovician (455–450 Ma). Taking into account the errors involved in paleomagnetism, these data do not precisely constrain the timing of Carolinia–Laurentia collision. The nature of this accretion is further obscured because the collision zone between Carolina and Laurentia is buried beneath the central Piedmont shear zone, a late Paleozoic thrust fault that telescoped Carolinia over the original (and thus, now buried) suture zone (Hibbard et al. 1998; Wortman et al. 1998).
There are, however, numerous lines of evidence from across the southern Appalachians that suggest a Late Ordovician – Early Silurian docking of Carolinia: (i) in Laurentia the Early Silurian Cherokee unconformity is considered to reflect loading of the Laurentian margin due to the accretion of an exotic terrane (Dorsch et al. 1994); (ii) the voluminous Late Ordovician – Early Silurian plutons in the Piedmont zone (which formed the leading edge of Laurentia in the Paleozoic) have a geochemical signature consistent with formation in a supra-subduction zone setting; and (iii) 40Ar/39Ar cooling ages (ca. 455–444 Ma) from micas indicate that Carolinia was affected by Middle to Late Ordovician tectonism and metamorphism (Noel et al. 1988; Offield et al. 1995; Ayuso et al. 1997). In addition, 40Ar/39Ar cooling ages from amphibole, ca. 430–425 Ma (Sutter et al. 1983), record Late Ordovician to Silurian uplift of Carolinia. Collectively, these features are attributed to the Late Ordovician sinistral transpressive docking of Carolinia; thus, subduction was to the northwest beneath Laurentia implying that Carolinia was situated on the lower plate (Hibbard 2000). The accretionary event is termed the Cherokee orogeny (Hibbard et al. 2010) after the expression of this tectonic event in Laurentia, the Cherokee unconformity (Dennison and Head 1975; Hibbard 2000).
Definition and extent
Ganderia is defined (van Staal et al. 1996, 1998) on the basis of a distinct sequence of Lower Cambrian to Tremadoc clastic sedimentary rocks that extends from the type area in eastern Newfoundland to the Miramichi Highlands of New Brunswick and southwest into Maine and New England (Fig. 3), and northwest across the Atlantic Ocean to Ireland and Great Britain (Fig. 4). These rocks are dominantly continent-derived arenites that are considered to represent the distal part of a passive margin near Gondwana (Williams 1964; van Staal et al. 1996). On the basis of field relationships, detrital zircon signatures (Fyffe et al. 2009) and early Paleozoic tectonic overprint (O’Brien et al. 1993), these Paleozoic sedimentary rocks (e.g., Gander, Miramichi, and Cookson groups) are interpreted to represent the cover to a sequence of Middle Cambrian arc–back-arc systems that were active along the northern edge of Ganderia (e.g., Victoria Lake Supergroup, Balmoral Group, and Ellsworth–Castine formations). Spatial relationships and isotopic signatures (Barr and White 1996; Rogers et al. 2006; Schultz et al. 2008) suggest that these arc-related sequences were built on a basement of Neoproterozoic arc-related volcanic and plutonic rocks that extends from central and southern Newfoundland (Cinq Cerf – Grey River) west to Nova Scotia (Bras D’Or) and south to New Brunswick (Brookville). Collectively, all of these rocks are overlain by a series of allochthonous Upper Cambrian back-arc ophiolites and associated volcanic rocks that were emplaced before 474 Ma during the Penobscot orogeny (Colman-Sadd et al. 1992; Johnson et al. 2009).
Included in this definition of Ganderia are the classic Gander zone and Exploits subzone of the Canadian Appalachians, the Pelham dome, Massesbesic gneiss, Nashoba terrane and Lunksoos arch in New England, Brookville and Bras d’Or belts in Maritime Canada, the Hertimage Flexure of Newfoundland, the Rosslare–Leinster terranes and Bray and Cahore groups in Ireland, and the Manx and Skiddaw groups and Lakesman terrane in Great Britain (Fig. 4) (van Staal et al. 1998; Hibbard et al. 2006b). The northwest boundary of Ganderia is the Red Indian Line, the main Iapetan suture that separates peri-Laurentian and peri-Gondwanan sequences and delineates the northwest limit of exotic terranes in the orogen. This boundary extends from New England into Maritime Canada and Newfoundland and across the Atlantic Ocean into Ireland and Great Britain where its location is marked by Leadhills Line and Southern Uplands fault, respectively. The southern boundary of Ganderia is located along the contact with Avalonia and is marked by the Bloody Bluff – Lake Char – Honeyhill faults in New England, Caledonia – Clover Hill faults in Maritime Canada, Georges River and Dover–Hermitage faults in Newfoundland and the Menai Strait fault in Great Britain.
The presence of Cambrian trilobites of the genera Kootenia and Baliella in a limestone block of the Dunnage Melange shows that Ganderia formed at high latitudes along the Gondwanan margin (Kay and Eldredge 1968). The use of faunal and lithological data, however, to elucidate the source craton of Ganderia is limited. The Early Cambrian platform sequence that dominates Ganderia represents the outer shelf to slope of a passive margin and has a notoriously sparse fossil population. The inboard shelf portion is not preserved, thus preventing correlation with cover sequences elsewhere in Gondwana. Consequently, investigation into the source area of Ganderia in Gondwana are best constrained by isotopic and detrital mineral data.
The Sm–Nd isotopic composition of Neoproterozoic Ganderian plutonic rocks is characterized by εNd values of +1.6 to –4.1 and depleted mantle model ages of 1.0 to 1.3 Ga (Whalen 1993; Kerr et al. 1995; Hibbard et al. 2007a and references therein). In contrast, the clastic supracrustal rocks from the type area of Ganderia have strongly negative values (εNd between –7 to –8 for t = 470 Ma; Fryer et al. 1992) indicating an inverted Sm–Nd crustal residence structure with a moderately isotopically evolved basement overlain by a highly isotopically evolved cover. Detrital minerals in the Ganderian cover in Newfoundland, Maine and New Brunswick are dominated by ca. 2.7–2.4, 1.7–1.2, and ca. 1.0 Ga detritus (O’Neill 1991; Reusch et al. 2004; Fyffe et al. 2009). The combined Sm–Nd and U–Pb dataset suggests that Ganderia developed upon ca. 1.3 Ga crystalline basement and lay adjacent to a source region comprising Archean, Paleo-, Meso- and Neo-proterozoic components. Rocks of these ages are present along the Amazonian margin of Gondwana and correlate with the Sunsás–Aguapeí (0.9–1.2 Ga), Rondônia–San Ignacio (1.2–1.4 Ga), Rio Negro–Jurena (1.5–1.75 Ga), and Trans-Amazonian (1.9–2.2 Ga) orogenic belts located along the periphery of the Archean Amazon craton (Bettencourt et al. 1999; Dall’Agnol et al. 1999; Tassinari and Macambira 1999). These data suggest that, prior to separation, Ganderia formed the part of Gondwana adjacent to the Amazonian craton.
Time of rifting and (or) drifting
Bimodal volcanic rocks from the Middle Cambrian Ellsworth terrane in Maine (Schultz et al. 2008) are characterized by magmatism with mantle components similar to modern OIB, N-MORB, and E-MORB sources that are not influenced by subducting oceanic crust. Contemporaneous volcanic rocks of the Ellsworth terrane also occur in New Brunswick in the 513 Ma New River belt (Johnson and McLeod 1996), in Nova Scotia in the 509 Ma Bourinot volcanic belt (White et al. 1994), and in Newfoundland as the 513–509 Ma Victoria Lake Supergroup (Dunning et al. 1991; Rogers et al. 2006). These sequences have been interpreted as remnants of an arc–back-arc system, the Penobscot complex (van Staal et al. 1998; Zagorevski et al. 2010), which was active ca. 513 to 485 Ma and was in part accompanied by the development of a back-arc basin (Zagorevski et al. 2007). Hibbard et al. (2007) interpreted the Ellsworth terrane as having formed as an extensional back-arc basin behind the Penobscot arc. Analogous relationships have also been observed in the Irish and British Caledonides (van Staal et al. 1998). This tectonic setting implies that the opening of the Rheic Ocean was related to subduction along an active plate margin of Gondwana with slab rollback leading to the opening of a back-–arc basin in a manner analogous to the present-day Okinawa Trough in the western Pacific (Shinjo et al. 1999).
Alternatively, Schultz et al. (2008) note the geochemical and isotopic similarity between the Ellsworth terrane and modern Cenozoic protooceanic rifts like the Gulf of California and Gulf of Aden rift systems. In their model, the Ellsworth terrane evolved inboard from the Ganderian margin in response to extensional forces controlled by regional plate motions in a manner similar to that of the recent Red Sea (Bosworth et al. 2005) or Tasman Sea (Sdrolias et al. 2003), i.e., rifting in response to slip vector changes and oblique extension as part of a strike-slip regime along the Ganderian margin (e.g., Keppie et al. 2003). Although the actual mechanism of riftto-drift of Ganderia remains uncertain, the combined geologic, geochemical, and isotopic data indicated that the oceanic rift related to the separation of Ganderia from the northern margin of Gondwana occurred during the development of the Middle to Late Cambrian Penobscot arc–back-arc complex. Anorogenic Late Cambrian – Early Tremadoc bimodal plutons that cut deformed Neoproterozoic Ganderian plutons in the Hermitage Flexure and Bras d’Or terrane (Dunning et al. 1990a, b) are potentially related to the terminal stages of rifting.
Travel path and time of accretion
The lack of biogeographically useful faunas in the rocks of the classic Gander sequence, precludes their use in paleogeographic reconstruction; thus, the only useful fossils are found in overstep sequence sedimentary rocks. Late Tremadoc–Arenig rocks in central Newfoundland contain cyclopygid trilobites and the trilobite genus Anamitella (Williams et al. 1992; Boyce et al. 1993), which are of Gondwanan affinity. Brachiopod and graptolite assemblages of similar age in Newfoundland and the Grangegeeth terrane in Ireland, specimens of Aulograptus cf. cucullus and Azygograptus, both of which occur in high paleolatitude settings (Harper and Parkes 1989; Williams et al. 1995) confirm this affinity. Therefore, the Early Ordovicican fauna of Ganderia have Gondwanan affinities, suggesting that this terrane was then situated in a cool water region at relatively high latitudes along the margin of Gondwana.
Llanvirn – early Llandeilo conodonts, trilobites and brachipods in the Exploits subzone of Newfoundland (Dean 1971; Neuman 1984), Central and Grangegeeth terranes in Ireland (Harper and Parkes 1989; Owen et al. 1992) and Southern Uplands of Scotland (Owen and Clarkson 1992) have strong Laurentian affinities and record the transition into warmer water environments. The occurrence of the late Caradocian graptolite Geniculograptus typicalus (Williams 1993) in the Ganderian overlap sequence in central Newfoundland, confirms this warming trend as it is restricted to the eastern margin of Laurentia. Therefore, the timing of eradication of the Early Ordovician Gondwanan faunal signature and its replacement by Laurentia affinities in the Middle Ordovician reflects the transfer of Ganderia from a cool water, high latitude environment to warmer, lower latitude conditions. This movement of Ganderia corresponds with the closure of the main tract of the Iapetus Ocean, which was simultaneously subducting on both the peri-Laurentian and peri-Gondwanan margins, and is recorded by Early to Middle Ordovician obduction of ophiolites on both Laurentia and Ganderia.
During the Silurian there was no distinct faunal provincialism along the active margin of Iapetus. Graptolites (Rastrites peregrinus and Atavograptus; Williams 1993), trilobites (Symphysops and Cyclopyge; Boyce et al. 1991) and bivalve (Cuneamy arata; Boyce et al. 1993) fossils in Ganderia are all similar to adjacent Early Silurian faunas deposited in shallow marine settings along the Laurentian margin. This cosmopolitan faunal assemblage supports the idea that the accretion of Ganderia reduced Iapetus to a narrow ocean basin, thereby hindering faunal interchange with Gondwana and allowing interchange with Laurentia.
Paleomagnetic data from interbedded volcanic and sedimentary rocks of the Bourinot Group on Cape Breton Island indicate that Ganderia was situated at a paleolatitude of 49°S ± 11° during the Middle Cambrian (Johnson and van der Voo 1985). A study by Liss et al. (1993) of Llanvirn volcanic rocks (470–466 Ma Tetagouche Group), which overlie the Gander passive margin sequence in New Brunswick, indicate that several different volcanic flows yielded a stable primary magnetic remnance corresponding to an average paleolatitude of 52°S +21°/–16°. However, considering (i) its large analytical error, and (ii) that faunal data indicate Ganderia was consistently moving north away from Gondwana after Cambrian separation to low latitudes (20°S) in the Caradoc, we consider that calculation of this average is erroneous and places Ganderia at too high a paleolatitude.
At the onset of the Early Silurian, Ganderia had moved to low latitudes, ca. 10°S, as recorded in the overlap sequences in Newfoundland and Ireland. Lava flows from the Llandovery–Wenlock Lawrencton Formation and sedimentary rocks from the overlying Wigwam Formation of the Botwood Group in Newfoundland yield paleolatitudes of 14°S and 8.5°S ± 5°, respectively (Buchan and Hodych 1992; Smethurst and McEnroe 2003). A comparable paleolatitiude of 11° ± 5° is reported from the Wenlock (ca. 425 Ma) Mill Cove Formation of the Dingle Peninsula in Ireland (Mac Niocaill 2000). Collectively these data indicate that by the onset of the Late Silurian, Ganderia was situated along the Laurentian margin and that the width of Iapetus had narrowed to be less than the limits of paleomagnetic resolution.
The accretion of Ganderia to Laurentia was the cause of the Salinic orogeny, the timing of which is constrained by geological, sedimentological, isotopic and geochronological data (van Staal et al. 2009). The main tract of Iapetus that was situated between Laurentia and Ganderia was subducting on both margins leading to accelerated convergence between these terranes (Zagorevski et al. 2008). Accretion was initiated in the Caradoc, indicated by cessation of magmatism and uplift of the Victoria and Popelogan arcs in Ganderia (Zagorevski et al. 2010). This convergence was likely terminated by the Late Ordovician collision of Ganderia with Laurentia in a Molucca Sea-type arc–arc collision which juxtaposed peri-Laurentian and peri-Gondwanan terranes along the suture zone, the Red Indian Line (Zagorevski et al. 2007). The earliest sedimentological linkage is provided by Laurentian-derived detritus in late Katian–Hirnantian conglomerates of the Badger Group that overlie Early to Middle Ordovician volcanic rocks in Ganderia (Nelson 1981; Pollock et al. 2007).
Following collision, the only remaining Iapetus oceanic lithosphere was located in the ca. 1000 km-wide Tetagouche–Exploits back-arc basin in Ganderia. Continued sinistral–oblique convergence between Ganderia and Laurentia caused a stepping back of the subduction zone into the Tetagouche–Exploits basin, the northern part (i.e., accretion-modified Notre Dame arc) of which formed the active Laurentian margin. Closure of the Tetagouche–Exploits basin was accomplished by Hirnantian–Wenlock northwest subduction of lithosphere beneath Laurentia. Silurian sedimentary rocks on both sides of this subduction zone in Newfoundland remained distinct and had different source areas until the Wenlock (Pollock et al. 2007), when the basin closed along the Dog Bay Line (Williams et al. 1993). Geochronological constraints (West et al. 1992, 2002; Tucker et al. 2001; Reusch and van Staal 2011) on the age of structures indicate that closure was approximately coeval in New Brunswick (Bamford Brook fault system) and Maine (Liberty–Orrington fault). In Ireland, coeval basin closure occurred along the Navan–Silvermines fault system, which separates Llanvirn arc-related volcanic rocks of the Grangegeeth Terrane that contain Caradocian Laurentian–Baltic brachiopods from the Bellewstown Terrane, which formed near the outboard margin of Ganderia (van Staal et al. 1998).
Ocean closure along the Dog Bay Line was coincident with the development of a south–southeast-directed thrust and fold belt (ca. 435 to 422 Ma: van Staal 1994; Zagorevski et al. 2007), intrusion of syntectonic plutons (Dunning et al. 1990b; Zagorevski et al. 2007), local mélange formation (Lafrance and Williams 1992), high pressure metamorphism (van Staal et al. 2008), breakdown of faunal provinciality (Williams et al. 1995), disruption in sediment provenance diversity (Pollock et al. 2007), and Silurian unconformities (Williams 1995)that represent products of the terminal phase of Salinic orogeny. Collectively these data indicate that final accretion of Ganderia to Laurentia and closure of the Tetagouche–Exploits back-arc basin, and by implication the Iapetus Ocean, was complete by the Late Silurian (ca. 421 Ma).
Definition and extent
Avalon was initially recognized as including Precambrian rocks overlain by Lower Paleozoic shelf deposits in eastern Newfoundland (Williams 1964). These sequences were termed the “Avalon platform” by Kay and Colbert (1965). In the classic synthesis of the northern Appalachian orogen, Williams (1979) defined the “Avalon zone” as including all exotic rocks that extend from eastern Newfoundland southwest to Maritime Canada into the eastern Massachusetts – Rhode Island area of New England. With the introduction of the terrane concept of Coney et al. (1980), the Avalon zone was recognized as a far-travelled, fault-bounded block and renamed the “Avalon terrane” (Williams and Hatcher 1983; O’Brien et al. 1983). Numerous researchers have since suggested separate terrane status for many of these areas and viewed Avalonia as an amalgamation of multiple terranes termed the “Avalonian plate” (Rast and Skehan 1983), “Avalon composite terrane” (Keppie 1985), “Avalon superterrane” (Gibbons 1990), “East Avalonia” (Pickering et al. 1988), “West Avalonia” (Nance and Murphy 1994), and “Avalon sensu lato” (O’Brien et al. 1996). The grouping of these areas into a common terrane, however, was done to emphasize their similarities even though the affinity of many of these areas had not been explicitly demonstrated and portions of Avalonia assigned to the composite terrane by some (Nance et al. 2002) have been excluded and assigned to separate terranes by others (Barr et al. 2003). The plethora of names for these regions and the connotations of using this varied nomenclature clearly indicate the need to herein define the term “Avalonia”.
Avalonia refers to a single crustal block defined by well-preserved Neoproterozoic clastic sedimentary, low-metamorphic-grade volcanic and plutonic rocks overlain by an Ediacaran – Early Ordovician cover of fine-grained, undeformed siliciclastic rocks (Landing 1996). Some researchers (e.g., Keppie 1985; Thompson et al. 1996) regard Neoproterozoic subduction-related plutonic and volcanic rocks that record tectonothermal and depositional events between ca. 750–545 Ma as the defining characteristic of Avalonian terranes. The rocks that lack an early Paleozoic platformal cover sequence are not part of the Avalonia definition.
Avalonia is the largest block of the peri-Gondwanan realm. In Newfoundland, Avalonia extends from its type area of the Avalon Peninsula, encompassing all rocks that lie southeast of the Dover–Hermitage faults (Avalon sensu stricto, O’Brien et al. 1996) including the islands of St. Pierre and Miquelon, France (Rabu et al. 1986) and the continental shelf. In Maritime Canada, Avalonia is bounded to the northwest by the Georges River and Caledonia – Clover Hill faults and includes the Mira terrane (Barr and Raeside 1989), Antigonish Highlands and Cobequid Mountains of Nova Scotia (Williams 1979), and the Caledonia terrane of New Brunswick (Barr and White 1996). In New England, Avalonia is bounded to the northwest by the Bloody Bluff – Lake Char – Honeyhill fault systems encompassing the “Boston Avalon” area of Massachusetts and Rhode Island. Avalonia extends to the eastern side of the Atlantic in the Caledonides and occurs south of the Menai Strait fault in southeastern England, Wales, and the Channel Islands (Nance et al. 2002). Faunal homogeneity (Cocks and Torsvik 2002), similar paleomagnetic data (Trench et al. 1992), and a lithostratigraphically correlative cover sequence preclude the possibility that these Avalonian terranes were separate and indicate that Avalonia was a once-continuous, unified early Paleozoic microcontinent.
Geological, lithostratigraphic, isotopic, faunal, and paleomagnetic data indicate that Avalonia records a Neoproterozoic tectonomagmatic history that predates opening of the Iapetus Ocean and originated along an active margin Gondwana (Williams and Hatcher 1983; O’Brien et al. 1983; Murphy et al. 1999; Nance et al. 2002). However, previously proposed connections to West Africa (O’Brien et al. 1983; Landing 1996; McNamara et al. 2001; Thompson et al. 2007) are inconsistent with available geological data.
The hallmark of Avalonia is the development of an Ediacaran – Early Ordovician shallow-marine, shale-dominated clastic platformal sedimentary succession containing diagnostic Avalonian trilobite fauna (Walcott 1891; Theokritoff 1979). Paleomagnetic data (Trench et al. 1992; McNamara et al. 2001; Hamilton and Murphy 2004; Thompson et al. 2010) place this Avalonian platform at intermediate to high paleolatitudes, ca. 40°–60°S, whereas stratigraphic data (Landing 1996) suggest that deposition occurred on an epeirogenically active platform in a cool-water environment. In contrast, the lowest part of the Pan African cover sequence in Morocco is dominated by a warm-water, archaeocyathan-bearing carbonate lithofacies that contains a distinct low latitude trilobite fauna. These contrasts in cover sequence lithofacies, paleolatitude, and family-level differences in trilobite fossils indicate that the Neoproterozoic – early Paleozoic history of Avalonia is unrelated to that in Morocco and suggest that Avalonia and West Africa were not in close proximity during this time.
Avalonia’s Neoproterozoic evolution is characterized by largely juvenile arc-related volcanism and plutonism and associated volcanic arc–rift sedimentary sequences (Nance and Murphy 1994; O’Brien et al. 1996). U–Pb ages of detrital zircons from these rocks in Newfoundland and Maritime Canada (Barr et al. 2003; Murphy et al. 2004a; Pollock et al. 2009) broadly match the age signatures of rocks found along both the West African and Amazonian margins of Gondwana. The presence of Meosproterozoic – early Neoproterozoic zircons in Avalonia is, however, incompatible with a West African source. Because West Africa is characterized by the absence of 680–660 Ma PanAfrican magmatism and contains no record of tectonothermal events between 0.7–2.0 Ga (Rocci et al. 1991), it appears Avalonia contains detritus from a fundamentally different region and it is very unlikely that Avalonia was in close proximity to West Africa in the Neoproterozoic. Instead, Avalonia was proximal to a continent with significant Mesoproterozoic crust, which was most likely Amazonia (Keppie 1993). An Avalonia–Amazonia connection is also supported by Nd isotope data; Avalonian crust-derived felsic igneous rocks are characterized by positive εNd values (t = 610) and depleted-mantle model ages between 0.75–1.1 Ga (Nance and Murphy 1996; Hibbard et al. 2007a; Murphy et al. 2008) that indicate a juvenile ca. 1 Ga basement beneath Avalonia.
Time of rifting
Current paleogeographic reconstructions for the Neoproterozoic – early Paleozoic place Laurentia along the equator during the opening of the Iapetus Ocean (e.g., Hoffman 1991; Murphy et al. 2006; Nance et al. 2008). At this time, Gondwana was being assembled at high latitudes in the southern hemisphere. Faunal, paleomagnetic, and stratigraphic data place Avalonia at high paleolatitudes near the West African margin of Gondwana and at considerable latitudinal distance from contemporary Laurentia. There is, however, controversy over the exact time of rifting and separation of Avalonia. Some researchers favor a Neoproterozoic–Cambrian separation influenced by faunal and paleomagnetic evidence (McKerrow et al. 1991; Trench et al. 1992; Landing 1996, 2005). A Neoproterozoic or Early Cambrian separation of Avalonia from Gondwana is, however, inconsistent with geological (O’Brien et al. 1996), paleomagnetic (van Staal et al. 1998), isotopic, and detrital mineral data (Hibbard et al. 2007a) as it relies on a prerifting connection to West Africa and the interpretation of Avalonia–Ganderia–Meguma as a single tectonic entity. However, evidence for Neoproterozoic – early Paleozoic lateral transport of Avalonia along the margin of Gondwana has been documented. Evolved Nd isotopic values and changing detrital zircon provenance from Cambrian clastic sedimentary and metasedimentary rocks in New Brunswick (Satkoski et al. 2010) suggest tectonic slivering and dextral strike–slip displacement of Avalonia along the Gondwanan margin. This transcurrent motion is also coeval with pervasive depletion in oxygen isotope composition in igneous rocks (Potter et al. 2008) and the termination of subduction zone magmatism and transition to an intracontinental wrench-related tectonic setting (Nance et al. 2002; Keppie et al. 2003).
Paleomagnetic, faunal, isotopic, and stratigraphic data indicate that separation of Avalonia occurred in the Early Ordovician. Paleomagnetic data from Ediacaran rocks of the Musgravetown Group in eastern Newfoundland (24° ± 5°S, Pisarevsky et al. 2011) and the Caldecote volcanics in England (27° ± 6°S, Vizan et al. 2003) indicate that Avalonia was situated at low latitudes. It moved to reside at high southerly latitudes (60°–54°) near Gondwana during the Cambrian (van der Voo and Johnson 1985; Hodych and Buchan 1998; Mac Niocaill 2000). During the Floian, Avalonia moved rapidly north to intermediate (41 ± 8°S) latitudes indicating increasing separation from Gondwana concurrent with a decrease in the separation between Avalonia and Laurentia (Johnson and Van der Voo 1990; Hodych et al. 2004; Hamilton and Murphy 2004). Comparable faunal assemblages support Gondwanan linkages for Avalonia until the Tremadoc (Cocks et al. 1997; Fortey and Cocks 2003); after which, the faunal provinciality indicates changing affinities from Gondwanan to genera with Laurentian ancestors.
The separation of Avalonia was followed by widespread Floian subsidence recorded in the Stiperstone Quartzite in England and the Redmans Formation in Newfoundland, interpreted to reflect the rift–drift transition of the Rheic Ocean (Prigmore et al. 1997; van Staal et al. 1998; Murphy et al. 2006; Pollock et al. 2009). U–Pb detrital zircon data from the Redmans Formation contrast significantly with that from the underlying platformal units and records the separation and departure of Avalonia from Gondwana during the Floian (Pollock et al. 2009). Evidence for Early Ordovician separation is also provided by Middle Cambrian to Middle Ordovician minor bimodal rift volcanic rocks in New England, Newfoundland, Wales, and England (Kokelaar 1988; van Staal et al. 1998; Thompson et al. 2010). These rocks may mark the propagation of an active, intraoceanic spreading ridge inboard to the edge of the Gondwanan margin and record the continued opening of the Rheic Ocean (Keppie et al. 2003; Sánchez-García et al. 2003).
Travel path and time of accretion
There are several lines of evidence that support the Ordovician–Silurian movement of Avalonia from Gondwana to Laurentia. Paleomagnetic data from the 490 Ma Treffgarne volcanics in Wales and 489 Ma Nahant igneous rocks in New England indicate that Avalonia was situated at ca. 65°–62°S before separation from Gondwana (Trench et al. 1992; Thompson et al. 2010). After its Early Ordovician departure, Avalonia moved across Iapetus to 44°–41°S as shown by the paleomagnetic results from the Llanvirn Builth Wells volcanics in Wales (Trench et al. 1991) and 460 Ma Dunn Point volcanics in Nova Scotia (Hamilton and Murphy 2004). By the earliest Silurian Avalonia had reached an intermediate latitude of 32° ± 8° (441 Ma Cape St. Mary’s sills; Hodych and Buchan 1994), yet it was still a considerable distance (ca. 2000 km) from the leading edge of composite Laurentia, now represented by Ganderia at 14°S, which was situated proximal to the equator (Smethurst and McEnroe 2003). These data imply about 6 cm per year for the latitudinal component of the divergence between Gondwana and Avalonia during the Ordovician (Mac Niocaill et al. 1997; Hamilton and Murphy 2004), suggesting that motion was accommodated by a combination of slab pull in Iapetus and ridge push in the Rheic Ocean.
The movement of Avalonia is also well characterized by changing affinities of faunas during its transit across Iapetus. The Cambrian to Early Ordovician cold-water Paradoxides trilobite and low-diversity shelly fossils (e.g mollusks and hyoliths) in Avalonia (Fletcher et al. 2005) are similar, and in many cases identical, to those found along the margins of Gondwana and consistent with a position adjacent to the South Pole (Fortey and Cocks 1992). This indicates that there was strong faunal interconnection between Avalonia and the neighbouring continent of Gondwana at that time. These faunas are in complete contrast to the warm-water benthic Olenellus trilobite and Schizambon brachiopod faunas of Laurentia and separation by the wide (ca. 5000 km) intervening Iapetus Ocean provides a mechanism for limiting faunal dispersal between the continents (Cocks and Fortey 1982). Throughout the Ordovician, however, faunal interchange became progressively more marked, as trilobite and brachiopod genera that had originated in Laurentia became more common in Avalonia as the separation between the two decreased and the latter moved north into a warm-water environment (Pickering et al. 1992; Lees et al. 2002; Cocks and Fortey 2009).
By the Early Silurian, the progressive migration of genera across the closing Iapetus Ocean meant there were few significant differences between most of the faunas of Avalonia and Laurentia. In contrast, the development of faunal distinctiveness between Avalonian and Gondwanan shelly faunas during this interval suggests the presence of significant faunal barriers, most likely the widening Rheic Ocean that had developed between the two landmasses (Cocks and Fortey 1982). Although brachiopod and trilobite faunal distinctions had disappeared across Iapetus, most Silurian ostracod faunas of Avalonia were substantially different from those in Laurentia up to the Early Devonian (Berdan 1990). This disparity suggests that, brachiopods and trilobites were able to cross the narrowing Iapetus Ocean, but the ocean was wide enough to act as an effective barrier to young ostracods (Cocks et al. 1997). By the Early Devonian, faunal similarity in shallow water and nonmarine environment genera in both Avalonia and Laurentia (Williams et al. 1995) seems to indicate the closure of the narrow seaway and juxtapositioning of Avalonia and Ganderia that formed the leading edge of Laurentia after the Salinic orogeny.
The accretion of Avalonia to Ganderia and, by implication Laurentia, is interpreted to be the cause of the Acadian orogeny (van Staal et al. 2009). After the Salinic accretionary events, Avalonia was still separated from Ganderia, which was the active margin of Laurentia, by a narrow oceanic tract (van Staal 2007). Closure of this basin occurred during the Katian–Pridoli and is recorded in a collection of syntectonc, arc-related rocks that extend along the southeastern margin of Ganderia from Maine and New Brunswick (Coastal volcanic, Kingston, and Mascarene belts; Fyffe et al. 1999; Barr et al. 2002) to Nova Scotia (Skye Mountain granite; Keppie et al. 2000) and Newfoundland (Burgeo Batholith; Kerr et al. 1995). These sequences comprise bimodal volcanic and consanguineous plutonic rocks that have compositions typical of arc and back-arc settings indicating that Avalonia was situated on the lower plate and subduction was to the northwest beneath Ganderia (Dunning et al. 1990b; Valverde-Vaquero et al. 2003). Hence, the northwest-dipping subduction zone beneath composite Laurentia jumped eastward behind the accreted Kingston–La Poile arc in Ganderia. This setting is also supported by the Pridolian Stonehouse Formation of the Arisaig Group in Antigonish Highlands of Nova Scotia, which records the transition from a shallow shelf to a foreland basin sequence. This change is interpreted to represent accelerated subsidence and foundering of the Avalonian platform due to tectonic loading of Avalonia’s passive margin (Waldron et al. 1996). Late Silurian – Early Devonian magmatic rocks that intrude deformed arc and back-arc sequences in Ganderia, however, exhibit more within-plate characteristics reflecting a switch in tectonic regime to a back-arc setting (van Wagoner et al. 2002). Collectively, these data support a model in which the accretion of Avalonia changed from initial northwest subduction of oceanic crust beneath Ganderia to later dextral-oblique convergence after continental collision.
Magmatism associated with the Acadian orogeny is temporally distributed throughout Ganderia and constrains the age of accretion of Avalonia. The tectonic inversion of the Mascarene and La Poile basins thath are intruded by 423–421 Ma plutons (Utopia and Bocabec plutons; Fyffe et al. 1999) and 421–417 Ma plutons (Otter Point granite and Hawks Nest Porphyry; O’Brien et al. 1991), respectively, provide the best estimate for the onset of Avalonia–Ganderia collision. Ganderia contains abundant syn- to late tectonic, collision-related Late Silurian – Early Devonian granite intrusions (Dunning et al. 1990a, b; Valverde-Vaquero et al. 2003) that are coeval with cleavage development and dextral shear along the Avalonia–Ganderia boundary zone (Holdsworth 1994; Dallmeyer and Nance 1994; d'Lemos et al. 1997) and hence postdate the onset of Avalonia–Laurentia collision. Collision also initiated the development of Early Devonian (post 417 ± 2 Ma) northwest-directed shear zones and fold belts in Ganderia (Zagorevski et al. 2007). The final late Early Devonian pulse of Acadian magmatism (ca. 400–395 Ma) has been ascribed to detachment of the oceanic lithosphere attached to Avalonia (van Staal et al. 2009).
The faunal and paleomagnetic evidence discussed in the previous text indicates that Avalonia was docked to Laurentia by the Early Devonian. The earliest lithological linkage is established by a sampling of Ganderian lithologies in synorogenic sediments that unconformably overlie Neoproterzoic rocks of Avalonia in southern Newfoundland. Conglomerates and sandstones of the Early Devonian Cinq Isles and Pool’s Cove formations contain metamorphic and plutonic detritus derived from Ganderia (Williams 1971) and reflect uplift and transcurrent faulting related to collision. These two formations and the Avalon–Ganderia boundary and Dover–Hermitage Bay fault, are cut by a classic stitching pluton, the ca. 370 Ma Ackley Granite batholith (Tuach et al. 1986), which provides a minimum age of accretion.
Definition and extent
Meguma, originally defined as a distinct crustal block as the Meguma zone (Williams 1979), is restricted on land to southern Nova Scotia (Fig. 3). Its regional extent, however, is much larger as it extends beneath the Scotian Shelf and the Gulf of Maine to the south, and its easterly extension underlies the southern part of the Grand Banks of Newfoundland (Keen and Haworth 1985; Pe-Piper and Jansa 1999). Its total area is estimated at 200 000 km2 making Meguma the second largest exotic terrane in the Appalachian orogen.
Meguma is defined by a thick (>10 km) Cambrian to Early Ordovician turbiditic sandstone–shale succession of the Meguma Supergroup that represents a shoaling upward sequence of deep-sea fan complexes deposited on the continental rise and (or) slope to outer shelf of a passive margin (Waldron 1992; White 2008). Chemical and petrographic data from the Goldenville and Halifax groups (White and Barr 2010) demonstrate that the source eroded to produce the strata was not compositionally mature and included abundant intermediate and felsic igneous rocks. White and Barr (2010) interpret the depositional environment to be a passive continental margin in a Cambrian – Early Ordovician rift along the Amazonian margin of Gondwana. The Meguma Supergroup is disconformably overlain by a siliciclastic succession, the Annapolis belt, that has Upper Ordovician – Lower Silurian volcanic rocks at its base (White Rock Group) and fossiliferous Silurian to Lower Devonian shallow marine sedimentary rocks at its top (Torbrook Group) (Schenk 1997).
Meguma is separated from Avalonia to the north by the Cobequid–Chedabucto and Minas faults, which are interpreted to be a major dextral transcurrent fault system that had repeated episodes of strike–slip movement in the late Paleozoic (Murphy and Keppie 1998; Murphy et al. 2011b). The southern boundary of Meguma lies offshore and is unexposed; however, on the basis of geophysical analyses (e.g., gravity, magnetic, seismic) and isotope geochemistry and geochronology it is generally inferred to be the Maine–Rhard fracture (Lefort and Haworth 1981; Pe-Piper and Jansa 1999)
Stratigraphic and petrographic analyses of the Meguma Supergroup suggest that its source area was a low-lying, deeply eroded, stable continental terrane. Sediment dispersal patterns and a great volume of mature, fine-grained, siliciclastic sedimentary rocks indicate an easterly provenance of continental dimensions (Schenk 1997). The lower part of the Meguma Supergroup contains trilobites, including Paradoxides, and other Middle Cambrian Acado – Baltic genera and trace fossils (Pratt and Waldron 1991; Gingras et al. 2011) that are distinct from those of Laurentia and suggest an original location along the margins of Gondwana. U–Pb dating of detrital zircons from the Middle Cambrian Goldenville Group show dominantly late Neoproterozoic (0.75 to 0.54 Ga) distributions with lesser Proterozoic and Archean populations (Waldron et al. 2009). εNd values are strongly correlated with detrital zircon data and display a change from relatively juvenile sources (εNd +0.75) in older samples, to more evolved sources (εNd –6 to –8) at stratigraphically higher levels. The combined data set are consistent with sediment derivation from local sources in the Trans-Saharan and Rokelide orogenic belts along the West African continental margin (Krogh and Keppie 1990; White and Barr 2010). However, the data are also permissive of an origin along other parts of Gondwana such as the northern African margin adjacent to the Arabian Shield (Stoeser and Stacey 1988; Stern 1994). A position along the passive margin adjacent to Gondwanan Africa is the most probable location as the lithology, dispersal patterns, and paleoenvironments of Meguma mimic those of northern Mali, Mauritania, southern Morocco, and Algeria (Schenk 1971, 1997). Moreover, the regional unconformities on the West African craton that mark times of significant cratonic stripping and transport of detritus toward the continental margin are of sufficient time and extent to produce the enormous quantities of sediment now in Meguma (Deynoux 1980).
Time of rifting
Direct evidence for the separation of Meguma from Gondwana is not recorded in the paleomagnetic or faunal history. However, Late Ordovician to Early Silurian rift-related volcanic rocks of the Annapolis belt are considered by some to mark the onset of Meguma’s rifting from Gondwana (van Staal 2007). The within-plate alkalic chemical signature from mafic metavolcanic rocks and minor intermediate and felsic rocks of the White Rock Group in Nova Scotia is consistent with a continental extensional anorogenic setting (MacDonald et al. 2002). This extension is interpreted to be related to rifting of Meguma from Gondwana, which is constrained to the Llandovery by 442 ± 4 Ma and 438 +3/–2 Ma zircon dates from the upper part the formation (Keppie and Krogh 2000; MacDonald et al. 2002). The White Rock Group is unconformably overlain by a sequence of Ludlovian siliciclastic rocks of the Fales River and Tremont formations that record rapid deposition on a submerging platform (Schenk 1997) and probably reflect the rift–drift transition. The Early Silurian separation of Meguma was preceded by the Cambrian departure of Ganderia and Carolina and Early Ordovician departure of Avalonia, implying that the rifting and separation of terranes from Gondwana, and also by implication the opening of the Rheic Ocean, took place diachronously.
At present, the drift history of Meguma is poorly constrained due to the paucity of reliable paleomagnetic data and geographically diagnostic fossils. Sequence stratigraphy of the cratonic source area and Annapolis belt are compatible and interpreted by Schenk (1997) to indicate a high-latitude or even near-polar setting for Meguma during the Late Ordovician and Early Silurian. Correlation of the Lower Ordovician Fales River Formation with the Armorican quartzite of Cadomia implies that Meguma was located ca. 60°S, within a marginal, fine-grained siliciclastic belt offshore northern Morocco and Iberia (Cocks and Fortey 1982). Mature quartz arenites of the Deep Hollow Formation correspond stratigraphically and sedimentologically to West African glaciogenic sedimentary rocks of the Abtelli Group (Schenk 1997), which were deposited in response to a eustatic drop in sea-level because of Saharan (ca. 460 to 430 Ma) glaciation (Schenk 1972). A high-latitude peri-African setting is also supported by black slate of the Tremont Formation, which contains the graptolites Monograptus nilssoni and M. colonus (Bouyx et al. 1985) that are characteristic of the Ludlow in North Africa (Berry and Boucot 1973).
Travel path and time of accretion
Unlike the abundant and varied fauna of Avalonia and Ganderia, a limited number of fossil localities are known from Meguma. The Middle Cambrian trilobite assemblages, Paradoxides (sl.), Agraulos, Dorypyge, and Ellipsocephalidae, from the upper portion of the Goldenville Formation are all of Acado–Baltic affinity and identical to Middle Cambrian successions in Avalonia, Europe, the Middle East, and West Africa (Pratt and Waldron 1991). Ludlow–Pridolian graptolites (Monograptus nilssoni and colonus), crinoids (Scyphocrinites), mollusks (Modiolopsis and Orthoceras) and brachiopods (Camarotechia and Delthyris) from the Tremont Formation are comparable with those in West Africa and Europe and are of Rhenish–Bohemian affinity (Bouyx et al. 1985). The invasion of Rhenish fauna into Meguma suggests that during the Late Silurian, Meguma had separated from Gondwana and was close to the southern marginal areas of Baltica in the Rheic Ocean.
Early Devonian brachipods, trilobites, and crinoids in Meguma are typically cosmopolitan and belong to the “old world realm”. Lochovian brachiopods (Schizophoria runegatensis, leptostrophia and index meristella renaudae), and trilobites (digonus acuminatus) from the Torbrook Formation demonstrate strong linkages with coeval sequences from both the Rhenish and north Gondwanan domains (Boucot 1960; Bouyx et al. 1997). Although most Emsian taxa are typically Rhenish and Gondwanan forms, minor shelly fossils appear to have Laurentian affinities as they are similar to those found in the eastern United States and Canada. Specifically, these include brachiopods (Schizophoria provulvaria and Acrospirifer sp.) of the Appalachian faunal province (Johnson and Dasch 1972), which suggest a change in biogeographic affinity. These mixed Early Devonian fauna indicate that a previous barrier to reproductive communication responsible for maintaining the integrity of the old-world realm and Appalachian province in the Lochovian had been breached so as to permit faunal integration. The migration of Meguma to Laurentia because of the closure of the intervening Rheic Ocean is interpreted to be the mechanism responsible.
The collision of Meguma to Laurentia is the cause of the Famennian orogeny throughout the northern Appalachians (van Staal et al. 2009; Hibbard et al. 2010). This tectonothermal event has also been dubbed the “Neo-Acadian” or “Quaboagian” and was originally defined on the basis of Middle Devonian to Early Carboniferous deformation and metamorphism in southern New England (Robinson et al. 1998, 2007). These tectonic events correlate chronologically with magmatism, orogenesis and deposition of the Maritimes Basin in Meguma (Hicks et al. 1999; Keppie et al. 2002).
Collision was initiated in the Middle–Late Devonian as recorded by granitoid and subordinate bimodal plutons that are distributed across the entire orogen. Plutonism in New England and offshore Nova Scotia was synchronous with granulite–facies metamorphism and ductile deformation (Robinson et al. 1998) and has been attributed to break-off of the oceanic lithosphere attached to the downgoing Meguma plate and mantle upwelling of asthenospheric material following accretion (van Staal 2007, 2009), subduction beneath Meguma of Rheic Ocean lithosphere outboard of Meguma (Moran et al. 2007), and ridge subduction at a low angle to the margin (Murphy and Keppie 2005). Further time constraints are provided by subsidence and formation of the Maritimes Basin throughout Atlantic Canada. Deposition started in the Middle–Late Devonian with interbedded volcanic and marine clastic sedimentary rocks (Dunning et al. 2002; Force and Barr 2006) that were laid down in a series of polycyclic basins with repeated subsidence and inversion of fault-bounded depocenters associated with strike–slip faults (Gibling et al. 2008)
Accretion of Meguma to Avalonia, i.e., the active margin of Laurentia, is interpreted to be dextral oblique strike–slip and largely accommodated by the Cobequid–Chedabucto fault system (Murphy and Keppie 1998). Meguma has no apparent Middle Paleozoic or older sedimentological linkage with Avalonia (or Laurentia); although Upper Ordovician – Lower Devonian strata of Meguma contain zircon populations similar to those in coeval strata of Avalonia and suggest that both Meguma and Avalonia resided along the northern margin of the Rheic Ocean at that time (Murphy et al. 2004b). The minimum age of accretion is constrained by Carboniferous terrestrial sedimentary rocks (e.g., Maritimes Basin) that extend across the entire exposed orogen as an unconformable overlap sequence (Hibbard et al. 2006b). Thermal evidence for Devonian thickening of Meguma crust indicates that Meguma is allochthonous over Avalonia (Waldron et al. 1988) and U–Pb dating of basement xenoliths suggests that Avalonian crust is present beneath Meguma at depth (Greenough et al. 1999). Alternatively, northwest-dipping reflectors at upper-mantle depths suggest that subduction of oceanic lithosphere between Meguma and Avalonia was to the northwest beneath Laurentia (Keen et al. 1991). Murphy et al. (1999) proposed that the dip of this subduction zone was very shallow because of interaction with a mantle plume. These contrasting tectonic regimes are resolved in a model where Meguma was initially situated on a shallowly dipping subducting plate and then transferred to the overriding Laurentian plate as a result of tectonic wedging (van Staal 2007).
Definition and extent
The Tallahassee–Suwannee terrane, herein termed Suwannee, was originally recognized as an exotic terrane by Williams and Hatcher (1983) on the basis of magnetic and gravity data. Its rocks are entirely buried beneath Cretaceous rocks of the Atlantic coastal plain (Fig. 5) and accessible only by exploration boreholes. It comprises a collection a Neoproterozoic, intermediate to felsic volcanic rocks of the North Florida volcanic series, plutonic rocks of the Osceola granite complex, and a suite of high-grade metamorphic lithologies (St. Lucie Metamorphic Complex) (Chowns and Williams 1983; Dallmeyer 1989). These rocks are unconformably overlain by an undeformed Ordovician – Middle Devonian clastic sedimentary sequence, the Suwannee basin of King (1961). The unexposed nature of this sedimentary basin means that its thickness is not well constrained as boreholes in Florida and Georgia have only penetrated this sedimentary sequence to a depth of 600 m (Thomas et al. 1989). However, seismic refraction and gravity anomalies indicate a sedimentary thickness of ca. 3 km and a total volcanic–plutonic and sedimentary thickness of 6 km (Nelson et al. 1985).
Suwannee underlies all of the pre-Middle Jurassic rocks of the Atlantic coastal plain in Florida, southern Georgia and southeastern Alabama. Consortium for Continental Reflection Profiling (COCORP) multichannel seismic reflection data suggest that Suwannee’s northern boundary with Carolinia is the Brunswick magnetic anomaly (BMA), an 80 km-wide zone characterized by an abrupt change in seismic, gravityand magnetic properties (Lizarralde et al. 1994). Suwannee extends eastward underneath the Gulf of Mexico where it is truncated by the Bahamas Fracture Zone, a sharply defined northwest-trending structure that represents a transform fault related to Mesozoic opening of the Atlantic (Thomas 2006).
Unlike the abundant and stratigraphically varied fauna of other accreted Appalachian terranes, the only fossiliferous rocks from Suwannee are present in drill cores from exploration boreholes. Quartzitic sandstones and shales from Florida and Alabama contain abundant ichnofossils including Skolithos and inarticulate brachipods that constrain the age of the sequence to the Late Cambrian – Early Ordovician, but are not biogeographically diagnostic (Pojeta et al. 1976). Other boreholes in southestern Alabama, however, contain Lower Ordovician brachiopod and graptolite assemblages including Didymograptus deflexus, which is unknown in native Laurentian rocks, but is diagnostically Gondwanan (Pojeta et al. 1976). A Gondwanan affinity for Suwannee is also supported by the occurrence of the Dapingian trilobite Colpocoryphe exsul, which has been reported from Eastern Europe and North Africa (Whittington and Hughes 1972). Silurian–Devonian pelecypod fossils from four wells in Florida and Georgia confirm the affinity of these faunas with the widespread Pridoli–Lochkovian faunas of central Europe, Nova Scotia, North Africa, and South America (Pojeta et al. 1976).
Zircons from calc-alkaline rocks of the North Florida volcanic series, dated at 552 ± 8 Ma (Heatherington et al. 1996), are coeval and interpreted as consanguineous with plutonic rocks of the ca. 551 Ma Osceola granite. Trace element data (LFSE enriched REE patterns, low abundances of HFSE) and evolved whole-rock εNd values (+1.1 to –4.1) suggest that these Neoproterozoic sequences were built on a continental lithosphere above an active subduction zone (Heatherington et al. 1996). Depleted mantle model ages from 1.0 to 1.6 Ga, U–Pb ages of inherited zircons (1.0 to 1.2 Ga) and Pb isotope signatures (Mueller et al. 1994) suggest the involvement of Mesoproterozoic crustal material which therefore may underlie portions of Suwannee.
Crustal material that fulfills the isotopic and geochronologic requirements of Suwannee basement is present in the 1.0–1.6 Ga Borborema Province of South America that, in pre-Mesozoic reconstructions, lies between the West African – Sao Luis and São Francisco – Congo cratons (Murphy et al. 2004a). Detrital zircon populations from Paleozoic sedimentary sequences (Mueller et al. 1994), however, are comparable to those in Meguma and suggest the presence of detritus from the PanAfrican and Eburnian orogens in West Africa. A West African connection is also supported by 40Ar/39Ar post-magmatic cooling ages of 530−525 Ma and lithological comparisons of the St. Lucie Metamorphic Complex, which is interpreted to represent an extension of the Rokelide Orogen (Dallmeyer 1989). Collectively these data place Suwannee along the margin of Gondwana in a location proximal to both the Trans–Amazonian and West African cratons.
Travel path and time of accretion
There are scant available data that track Suwannee’s Paleozoic movement across the Rheic Ocean. The only constraint on its Paleozoic position is from a single paleomagnetic study of Ordovician–Silurian? sedimentary rocks that indicate Suwannee was at 49°S at this time (Opdyke et al. 1987). This result is consistent with Silurian–Devonian fossils containing high-latitude trilobite and acritarch fauna of Gondwanan affinity that have generally been correlated with the Bove Basin of West Africa (Whittington and Hughes 1974; Cramer and Diez 1974). Unlike other accreted terranes of the Appalachians, there is no direct evidence in the geological record that Suwannee rifted from Gondwana prior to its accretion to Laurentia. The lack of exposure and limited dataset, however, do not preclude an earlier rifting event.
The accretion of Suwannee to Laurentia has traditionally been ascribed to the late Paleozoic Alleghanian orogeny (Woodward 1957; Secor et al. 1986). The suture zone between Suwannee and rocks of the Appalachian orogen is not exposed at the surface anywhere along its entire length. Its location is constrained by stratigraphic and structural relationships in exposed portions of adjacent terranes and it is inferred to coincide with an intense, long-wavelength magnetic low of the BMA (Nelson et al. 1985). However, the pre-Mesozoic subcrop position of the suture does not everywhere coincide with the BMA (Secor et al. 1986). Because this boundary trends oblique to major structures and truncates other terrane boundaries in the orogen, the collision of Suwannee is interpreted by most workers (e.g., Dallmeyer 1989) to be a manifestation of the Carboniferous–Permian Alleghanian orogeny. This boundary however, may be related to post-accretional dextral transcurrent motion that structurally truncated and translated accreted peri-Gondwanan terranes (Hibbard and Waldron 2009).
An older episode of uplift, erosion and cooling in metasedimentary rocks spatially associated with Suwannee is indicated by K–Ar and 40Ar/39Ar ages ranging from 374 to 341 Ma (Scholle 1979; Dallmeyer 1991). As noted by Hibbard et al. (2010), these ages are coeval with high-grade metamorphism and intense ductile deformation present in inboard terranes of the orogen (Dennis and Wright 1997; Trupe et al. 2003). These variations in 40Ar/39Ar mineral cooling ages may be related to differential uplift and rapid post–metamorphic cooling along different portions of the orogen. As the age of this collision overlaps chronologically with the collision of Meguma in the northern Appalachians, it implies that Suwannee interacted with Laurentia during an orogen-wide Famennian event.
Time of abandonment of Appalachian peri-Gondwanan blocks
It has long been recognized that old suture zones and accreted domains of basement weakness are susceptible to rifting and subsequent ocean opening (Burke and Dewey 1973; Williams 1984; Murphy et al. 2006). The breakup of Pangaea occurred outboard of the Paleozoic collision zones that accreted Carolinia, Ganderia, Avalonia, Meguma, and Suwannee to Laurentia, leaving these terranes appended to North America during the Mesozoic opening of the Atlantic Ocean.
Local rifting of Pangaea began at the Triassic–Jurassic boundary, ca. 199 ± 8 Ma, with the eruption of areally extensive low-Ti tholeiitic flood basalts and basaltic andesites, and intrusion of diabase dykes, which were contemporaneous with magmatism of the Central Atlantic Magmatic Province (Heatherington and Mueller 2003). Initial subsidence during rifting was followed by thermal cooling of the North American lithosphere characterized by attenuated continental crust, which led to the formation of several large Mesozoic–Cenozoic sedimentary basins seaward of the East Coast Magnetic Anomaly (ECMA). The ECMA is generally thought to represent the boundary between oceanic and continental crust (Nelson et al. 1985). These basins, e.g., Bahamas Basin, Carolina Trough, Baltimore Canyon Trough, Georges Bank, Scotian, and Grand Banks basins, are comprised of Early Jurassic nonmarine lacustrine deltaic clastic rocks overlain by widespread salt layers deposited under hypersaline conditions in an arid to semi-arid paleoclimate (Miall et al. 2008). Most basins record a change in sedimentation to marine deposits as the evaporites are overlain by extensive carbonate and clastic sequences (Poag 1991) indicating progressively deeper marine conditions. Subsidence curves suggest almost continuous sedimentation, indicative of cooling and thermal contraction of the lithosphere (Harry and Sawyer 1992).
The rapid subsidence in the Early and Middle Jurassic is interpreted to mark the rift-to-drift transition and the onset of seafloor spreading. The first oceanic lithosphere adjacent to Suwannee began to form outboard of the Rheic Ocean suture zones in the Bathonian (ca. 166 Ma) and spreading prograded from south to north along the ECMA to the Grand Banks of Newfoundland (Janney and Castillo 2001). A pronounced unconformity separates Jurassic sediments and overlying Lower Cretaceous sediments and is associated with the breakup of the stretched continental crust and separation of the North American plate from Pangaea (Libby-French 1984). This breakup unconformity marks the birth of the Atlantic Ocean, the opening of which left Appalachian peri-Gondwanan crustal blocks stranded from the margin of Gondwana and affixed to North America.
The Late Neoproterozoic – Early Cambrian is of much interest as Earth was characterized by dramatic worldwide changes in environmental conditions including accelerated plate motion that may have triggered the rapid appearance and diversification of animal phyla, the development of multiple glaciation events at equatorial latitudes, and profound extremes in global sea-level, ocean chemistry and climate (Dalziel 1997; Hoffman and Schrag 2002; Narbonne and Gehling 2003). These changes are interpreted to be related to widespread rifting and orogenic activity that accompanied supercontinent assembly and dispersion. Geological evidence (Hoffman 1991; Dalziel 1997) suggests the breakup of Rodinia and opening of the Iapetus Ocean was broadly contemporaneous with Pan-African subduction-related and collisional events that resulted in the assembly of Gondwana in the Late Neoproterozoic – Early Cambrian.
Current paleogeographic reconstructions for the Neoproterozoic – early Paleozoic place Laurentia along the equator during opening of the Iapetus Ocean (Hodych et al. 2004). The results of our compilation are consistent with the popular consensus (e.g., Nance et al. 2010) that the domains of the peri-Gondwanan realm of the Appalachian orogen shared a common Gondwanan heritage at high paleolatitudes near the margin of Gondwana and at considerable latitudinal distance from contemporary Laurentia. The closure of the Iapetus and Rheic oceans was responsible for the formation of the Appalachian–Caledonian orogen and assembly of Gondwana and Laurentia into the Pangaea supercontinent in the Carboniferous–Permian. Thus, sometime between the Cambrian and Devonian the Gondwanan margin rifted to spawn a collection of microcontinent terranes that opened the Rheic Ocean in their wake.
Neoproterozoic – Early Cambrian
The available lithotectonic, isotopic, geochronologic, and paleomagnetic data are compatible with Avalonia, Ganderia, and Carolinia forming adjacent to the Amazonian craton (Nance and Murphy 1996; van Staal et al. 1996; Hibbard et al. 2007b). These domains were in all probability proximal to each other as they share a common Neoproterozoic tectonomagmatic history dominated by voluminous subduction-related magmatism and accretionary events that were built upon older Mesoproterozoic continental crustal substrates. In the latest Neoproterozoic, however, Avalonia records the termination of arc magmatism and transition without collision to a transform regime whereas Ganderia and Carolina both record suprasubduction zone magmatism that continues into the Cambrian overlain by a sequence of Cambrian deepwater clastic sedimentary rocks. These contrasts in the tectonic evolution are interpreted to indicate that Ganderia and Carolinia formed along an active convergent plate margin on either an independent crustal block or adjacent segment of Gondwana.
The initial rifting of Carolinia followed cessation of Early Cambrian arc volcanism accompanied by arc extension and subsequent rifting, forming an ensimatic back-arc marginal basin between the arc and Gondwana. Although the actual mechanism of the rift-to-drift transition is uncertain, it is probable that arc rifting may have induced the initial opening of the Rheic Ocean between the rifted Carolinian microcontinent and cratonic northern Gondwana.
Separation of Ganderia from Gondwana in the Middle Cambrian (Fig. 6) was probably facilitated by slab rollback of a subducting oceanic plate, which was broadly synchronous with initiation of subduction and ophiolite obduction (Lushs Bight oceanic tract, van Staal 2007) in the peri-Laurentian realm of Iapetus. The increasing slab pull forces in Iapetus together with trench rollback beneath the Penobscot arc may have aided in the transfer of Ganderia from Gondwana longitudinally into and across the contracting Iapetus Ocean. An alternate scenario, however, is equally plausible in which a transition occurred by truncation by a strike–slip fault as a result of ridge–trench collision; the trench was replaced by a transform as the ridge was overridden and led to a switch from arc to rift magmatism (Murphy et al. 1999; Keppie et al. 2003; Schultz et al. 2008).
In contrast, during the Early Cambrian Avalonia occupied a stable platform marked by deposition of a shale-rich platformal sedimentary succession (Landing 1996). Although, geochronologic and isotopic data suggest strike–slip displacement of Avalonia along the Gondwanan margin (Gutierrez–Alonso et al. 2005; Potter et al. 2008; Satkoski et al. 2010), paleomagnetic and geochronologic data imply that Avalonia was still juxtaposed with the northern margin of Gondwana (Cocks and Torsvik 2002; Pollock et al. 2009).
Early Ordovician – Early Silurian
The separation of Avalonia during the Early Ordovician is attributed to either (i) subduction along an active plate margin of Gondwana with hinge retreat leading to the opening of a back-arc basin that spread through to Avalonia (van Staal et al. 1998), in a manner analogous to the Miocene rifting of the Ryukyu arc and opening of the present-day Okinawa Trough (Shinjo et al. 1999); (ii) propagation of an active, intraoceanic spreading ridge inboard of the edge of the Gondwanan margin in a tectonic setting similar to the rifting of the Jan Mayen and Seychelles microcontinents (Müller et al. 2001); or (iii) a switch from transform motion to spreading along the Gondwanan margin in response to subduction and slab pull in Iapetus (Nance et al. 2002). Early Ordovician separation of Avalonia is consistent with most paleogeographical reconstructions that place the Paleozoic Avalonian margin of Gondwana in the Iapetus Ocean (e.g., Cocks and Torsvik 2002; Stampfli and Borel 2002; Murphy et al. 2010). The separation of Avalonia from Gondwana was contemporaneous with closure of the Taconic seaway and formation of the Notre Dame arc, collision of the Dashwoods microcontinent during the Taconic orogeny and thrusting of the Annieopsquotch ophiolite belt beneath the composite the Laurentian margin (van Staal et al. 2007). During this time, the Penobscot orogeny records arc magmatism accompanied by rapid rollback, arc rifting and obduction of Penobscot ophiolites on the opposite Ganderian margin of Iapetus (Colman-Sadd et al. 1992; Zagorevski et al. 2010). Thus, the simultaneous subduction on both margins of Iapetus may have resulted from a major plate boundary reorganization that contributed to the partial breakup and lateral displacement of Avalonia from Gondwana as a ribbon continent.
While Avalonia was separating from Gondwana in the Early Ordovician, Carolinia and Ganderia were making their transit across Iapetus. Early to Middle Ordovician arc–back-arc volcanism in Ganderia is interpreted as having formed an elongate magmatic arc system that trended oblique to Iapetus (Zagorevski et al. 2010). Thus, the main tract of Iapetus that was situated between Ganderia and Laurentia was rapidly being closed by two outwardly dipping western Pacific-type subuction zones on each margin that led to accelerated convergence. Although Carolinia is not associated with an Ordovician peri-Gondwanan oceanic tract, Carolina and Ganderia share similar lithotectonic histories (Hibbard et al. 2007a) and presumably occupied the same peri-Gondwanan plate margin. Paleomagnetic data indicate that they moved rapidly from Gondwana to a location along the Laurentian margin by the Sandbian (Vick et al. 1987; Noel et al. 1988) indicating a paleolatitudinal component of drift of about 8–10 cm/year and suggesting the role of slab pull in their transit. This rapid closure of Iapetus reflects the northward migration of the peri-Gondwanan realm culminating in the Late Ordovician arrival of both Carolinia and Ganderia in a lower plate configuration at the eastern Laurentian margin. Carolinia docked first, at ca. 455 Ma, during the sinistral transpressive Cherokee orogeny (Hibbard 2000; Hibbard et al. 2010).
The accretion of Ganderia (450–421 Ma) occurred shortly after the arrival of Carolinia and was accomplished by a diachronous two-stage collision. The active leading edge of Ganderia, the Popelogan–Victoria arc system was sutured along the Red Indian Line to the active margin of Laurentia in the Late Ordovician. Following collision, stepping back of the west-directed subduction zone in Ganderia led to sinistral oblique convergence between the trailing passive margin of Ganderia and composite Laurentia, which closed the Tetagouche–Exploits basin resulting in the complete closure of Iapetus and docking of Ganderia during the Salinic orogeny (Fig. 7) (Pollock et al. 2007; van Staal et al. 2009).
The Salinic orogeny appears to have been contemporaneous with rifting of Meguma from Gondwana. This continental marginal rifting and spreading was characterized by intracontinental rifts, eruption of volcanic sequences, and transform extension along large-scale strike–slip faults and shear zones (Schenk 1997). The mechanisms responsible for the onset of Meguma rifting are unclear but most likely related to rifting above a thermal anomaly in the underlying mantle. However, the absence of a hotspot trail across Meguma, coupled with Sm–Nd isotopes that indicate magma derivation from underlying Proterozoic subcontinental lithospheric mantle during the Ordovician (Murphy et al. 2011a), suggest that rifting was not the product of a long-lived, deep-seated mantle plume. Rather, the volcanism appears to owe more to partial melt formation and extraction from a source region within the upper mantle (MacDonald et al. 2002). Nd isotope, geochronologic, and lithogeochemical evidence suggest that the voluminous magmas formed over a relatively short time interval with magmas derived from depleted mantle sources in part modified by crustal contamination (Clarke et al. 1993; Keppie and Krogh 2000). In this model, decompression melting during continental breakup formed due to dynamic asthenospheric upwelling during passive rifting as a result of a thermal anomaly in the upper mantle that was subsequently dissipated by the separation of Meguma (White et al. 1987). We speculate that this thermal anomaly was formed from the large-scale interaction between the Iapetus, Rheic, and the Gondwanan plates. The fast movement northward of the Rheic oceanic plate, relative to the Gondwana plate, may have caused an extensional stress field and intense continental intraplate deformation. Extension along pre-existing faults would trigger asthenospheric upwelling, leading to the volcanism related to rifting and continued stretching causing the continental lithosphere to break; ultimately forming a new oceanic lithosphere between Meguma and Gondwana.
In the latest Silurian, following accretion of Ganderia–Carolinia and rifting of Meguma, the composite Laurentian margin was likely the site of collisional orogensis resulting from the accretion of Avalonia (421–400 Ma) during the Acadian orogeny. Docking was marked by (i) termination and structural inversion of the Silurian arc–back-arc system on the active Laurentian margin; (ii) tectonic loading of Avalonia’s west-facing passive margin; and (iii) kinematic reversal, from deformation with a sinistral component to one that was dominantly dextral (Nance et al. 1992; Hibbard 1994). These tectonic events correlate chronologically with Ludlow–Emsian plutonic rocks that intruded into arc and back-arc sequences in Ganderia (which was formed the active margin of eastern Laurentia) and are interpreted to be related to northwest subduction of the oceanic lithosphere between Avalonia and composite Laurentia (van Staal et al. 2009).
The Acadian orogeny was followed by Middle Devonian – Early Carboniferous accretion of Meguma (395–350 Ma), which led to the Fammenian orogeny. The docking of Meguma was accommodated by wedging of the leading edge of Laurentia and dextral transpression on the Minas–Cobequid–Chedabucto fault zones (van Staal 2007; Murphy et al. 2011b). The Fammenian orogeny was characterized by the absence of any accompanying arc magmatism, probably due to flat-slab subduction of oceanic lithosphere that was attached to the Meguma plate (van Staal et al. 2009).
The late Paleozoic accretion of Suwannee has traditionally been correlated with the Carboniferous–Permian–Alleghanian orogeny. There is no evidence that Suwannee rifted from Gondwana prior to accretion; however, considering the limited data set for this subsurface block, it is possible that Suwannee rifted and separated from Gondwana during the Devonian. The latter scenario implies an earlier interaction between Suwannee and the Laurentian margin than previously thought, and is supported by Late Devonian to Mississippian tectonothermal events in Suwannee and the southern Appalachians that are coeval with Fammenian orogenesis (375–340 Ma) in the northern Appalachians (Hibbard et al. 2010). Comparable faunal assemblages (Bouyx et al. 1997) and similarities in detrital mineral compositions (Mueller et al. 1994; Waldron et al. 2009) between Suwannee and Meguma suggest that both crustal blocks were situated relatively close to the margin of West Africa where they constituted the leading interface of Gondwana that generated an orogen-wide Famennian event.
Convergence was probably achieved by southwest subduction of Rheic Ocean crust beneath Suwannee as indicated by seismic reflection data (Nelson et al. 1985) and distribution of mafic and ultramafic rocks along the suture in southern Alabama (Chowns and Williams 1983). Collision of Suwannee probably involved significant counter-clockwise rotation of Gondwana to allow convergence of the northwest African margin with Laurentia resulting in diachronous deformation, metamorphism, and plutonism along the composite Laurentian margin (Hibbard et al. 2010). The lack of evidence for significant shortening deformation along the Suwannee suture zone coupled with the age of structures in the Appalachian foreland (Secor et al. 1986), suggest a highly oblique, dextral transpressional boundary culminating in the collision of Gondwana and Laurentia to form the supercontinent Pangaea. The diachroneity of orogeny, presence of over-step basins, and continuity of structural trends throughout the Appalachians suggest that the present location of domains of the peri-Gondwanan realm parallels their order of accretion. Each of these domains, however, may have been translated alongstrike of the orogen as documented by the significant, orogen-scale Late Devonian–Carboniferous dextral strike–slip motion that rearranged major components of the orogen (Hibbard and Waldron 2009).
The data and interpretations outlined in the previous text provide the basis for a paleogeographic model with a collage of accreted crustal domains that offers new information on their paleogeographic settings during formation adjacent to Gondwana, travel in the Rheic Ocean, and incorporation in the Appalachian orogen. Although we note similarities among rocks and structures in some of these domains (e.g., Ediacaran suprasubduction zone magmatism in Avalonia, Carolinia, and Ganderia) we are impressed by their differences. These differences are difficult to reconcile with a model that portrays the Appalachian peri-Gondwanan elements as having formed a single crustal block. As such, the major implication of our analysis is that it emphasizes the disparate paleogeographic histories in Carolinia, Ganderia, Avalonia, Meguma, and Suwannee, which we interpret to reflect first-order differences in their isotopic, faunal, paleomagnetic, and lithotectonic evolution. Our model rationalizes relationships between the different peri-Gondwanan elements by interpreting the Appalachian peri-Gondwanan realm as a collection of at least five early Paleozoic independent lithotectonic crustal fragments that diachronously rifted from the Gondwanan margin and were subsequently welded together along collisional orogenic belts in the Appalachian–Caledonian orogen.
Implicit in this model of independent peri-Gondwanan microcontinents is that the Rheic must have been a complex ocean with numerous spreading centres. The development of a series of crustal fragments along the Gondwanan margin during the early Paleozoic means that extension of the continental lithosphere was dispersed over multiple rift systems that originally extended into several intraoceanic spreading centres in Iapetus. Initial spreading of the Rheic Ocean progressed episodically along the Amazonian margin to the West African margin. The axis of first spreading parallels the Gondwanan margin and has a preference for zones of orogenesis along Neoproterozoic suture zones between previously accreted terranes related to the assembly of Gondwana (Murphy et al. 2006). Thus, younger zones of rifting reactivated preexisting zones of weakness or sutures that penetrated the lithosphere. These features of the Rheic Ocean are analogous to those of the modern North Atlantic as the Mesozoic axis of spreading in the latter propagated from south to north and follows the zone of preexisting boundaries related to Paleozoic closure of the Rheic Ocean and assembly of Pangaea (Williams 1984).
The rapid motion of the peri-Gondwanan domains may have been driven by a combination of ridge push from the mid-Rheic ridge and slab pull from a subduction zone within or bounding Iapetus. The Rheic Ocean probably achieved its maximum width by the Early Silurian, coincident with accretion of Ganderia and final closure of Iapetus. The width of the Rheic is difficult to assess, due to the limited resolution provided by paleomagnetic data (Mac Niocaill and Smethurst 1994). However, rifting of Carolinia in the Early Cambrian provides ample time for the formation of a 3000 km-wide ocean assuming a conservative spreading rate of 3–4 cm per year.
Whether Carolinia and Ganderia occupied the same lithospheric plate margin during their transit across the Rheic, or were separated as distinct crustal blocks is uncertain. Although both domains share a common Neoproterozoic subduction-dominated history and have kinematically and spatially similar Middle Ordovician docking styles, Ganderia records the evolution of a Middle Cambrian – Early Ordovician arc–back-arc system that is not preserved in Carolinia. The absence of a Cambrian arc in Carolina may be an artifact of exposure; it may lie structurally buried beneath the Central Piedmont shear zone. These disparities, however, can be resolved in a model whereby Carolinia and Ganderia formed as an elongate arc system (possibly separated in the Middle Cambrian by a transform boundary) that spanned across Iapetus to the Amazonian margin of Gondwana. We note, however, that the lack of arc-volcanic rocks does not automatically imply the absence or termination of subducting oceanic lithosphere; extensive parts of some arcs (e.g., Andes) contain significant time periods that have no record of magmatism (Dewey and Lamb 1992).
It is evident that the oceanic separation between Ganderia and Laurentia, which was present in the Ordovician, was terminated by the Early Silurian docking of Ganderia. The late Caradoc arrival of the Victoria arc at the Laurentian margin, however, implies a diachronous closure of Iapetus and that deformation along the Dog Bay – Liberty Line suture zone was not the first interaction between Laurentia and Ganderia.
In contrast to the paleogeographic similarities between Ganderia and Carolinia, the differences between Ganderia and Avalonia preclude their correlation as either a single domain or as distinct and juxtaposed terranes. Although though Ganderia and Avalonia have similar arc-dominated Neoproterozoic basement rocks, they have very distinct magmatic, stratigraphic, isotopic and tectonic histories (Hibbard et al. 2007a; van Staal et al. 2009) and were most likely tectonically separated, forming independent and distinct crustal blocks in the early Paleozoic. There is no evidence of any pre-Devonian linkage between Ganderia and Avalonia.
We have used combined constraints and data from stratigraphic, faunal, isotopic, paleomagnetic, and geochronologic studies to produce a paleogeographic model that provides new insights into the provenance, global travel paths and collision histories of the Appalachian peri-Gondwanan realm. However, many important problems remain unanswered:
Are the individual peri-Gondwanan elements described here the product of amalgamation of numerous crustal fragments that were welded together during the Neoproterozoic assembly of Gondwana? In Avalonia for example, evidence is circumstantial at best, as relationships between dispersed relics of protracted Neoproterozoic tectonothermal and depositional sequences are not well known and unequivocal basement has not been identified.
Do additional peri-Gondwanan terranes lie unexposed elsewhere in the orogen? One obvious candidate is the continental Goochland domain (Farrar 1984). However, insufficient data exist to classify its ca. 1 Ga (Owens and Tucker 2003) crustal gneissic rocks. They may be exposed as a structural window of Laurentia or it is equally possible that Goochland represent an exotic (Gondwanan?) accreted microcontinent. Additional peri-Gondwanan terranes may also be present in the southern Appalachian orogen that lie either inboard of Carolinia and are presently buried beneath Alleghanian thrust sheets, or outboard of Carolinia and buried beneath sediments of the Atlantic coastal plain (Horton et al. 2011).
What is the nature and mode of accretion between Meguma and Avalonia? Was there a significant oceanic tract that lay between the two domains? At present, there is no evidence for a Devonian volcanic arc in Avalonia that would imply the presence of subducting oceanic crust attached to Meguma.
What is the relationship between Meguma and Suwannee? Although both domains were accreted to Laurentia in the Late Devonian–Carboniferous? evidence indicating that Suwannee is a direct correlative of Meguma is absent.
Did Suwannee separate from Gondwana prior to its collision with Laurentia or did it form part of the Gondwanan margin during Alleghanian orogenesis? Such questions demand further analysis and synthesis from several disciplines to establish a consistent paleogeographic framework for the Appalachian peri-Gondwanan realm.
Since Williams (1964) recognized that the eastern flank of the Appalachian orogen comprises a collection of exotic crustal domains that predate Paleozoic orogensis during Appalachian–Caledonian mountain building, numerous workers have attempted to elucidate their provenance and detail their lithotectonic and paleogeographic evolution.
Compilation of pertinent faunal, paleomagnetic, structural, and isotopic data indicates that the Appalachian peri-Gondwanan realm consists of several along-strike segments that record diverse lithotectonic histories. We interpret these variations to reflect important differences in source area and paleogeographic track and suggest that the traditional view of a single unified Appalachian peri-Gondwanan microcontinent is too simplistic. We conclude that at least five independent crustal blocks, Carolinia, Avalonia, Ganderia, Meguma, and Suwannee, were spawned from the margin of Gondwana in the early Paleozoic leading to the formation of the Rheic Ocean. Each of these domains behaved as relatively stable terranes during their transit across the Rheic and were sutured along early Paleozoic collision zones resulting from subduction of oceanic lithosphere and accretion to Laurentia.
The overall tectonic evolution of the Paleozoic circumIapetus and Rheic region is dominated by continent separation and ocean formation along the northern Gondwanan margin that was contemporaneous with convergence and accretion along Laurentia. Carolinia and Ganderia separated from Gondwana first in the Middle Cambrian at which time subduction was initiated in Iapetus. Initial opening of the Rheic Ocean was followed by the dispersal of Avalonia in the Early Ordovician, which coincided with ophiolite obduction (Penobscot orogeny) in Iapetus and collision of the Dashwoods continent to Laurentia (Taconic orogeny). The accretion of Ganderia to Laurentia in the early Silurian during the Salinic orogeny was coeval with rifting of Meguma from Gondwana. Suwannee, however, appears to have remained attached to Gondwana and thus did not separate until after its accretion to Laurentia during Alleghanian orogenesis. Thus, parts of Gondwana experienced episodic rifting and distension of the Rheic Ocean, while subduction and accretion were active in Iapetus and Laurentia. It is not entirely fortuitous that the opening of the Rheic Ocean was contemporaneous with the destruction of the Iapetus Ocean, suggesting that patterns for rifting and opening of the Rheic Ocean were set by a complex interplay of mantle processes and the large-scale geodynamic interaction between lithospheric tectonic plates in Iapetus and Gondwana.
All of the authors are grateful for the opportunity to contribute to this special volume in honour of E.R. Ward Neale. JCP and JPH owe a debt of gratitude to Ward for building the geology department at Memorial University of Newfoundland into a worldclass centre for tectonic and Appalachian research; JPH's graduate years at Memorial benefited from Ward being at the helm of the department and the vigorous geoscience department that Ward assembled persisted into JCP's undergraduate years at Memorial. Helpful and thorough reviews by Damian Nance, Brendan Murphy, and an anonymous reviewer are gratefully acknowledged. Our Appalachian research has been supported by grants from the National Science Foundation (USA) and Natural Sciences and Engineering Research Council of Canada.
- Received April 12, 2011.
- Accepted June 13, 2011.
- Published on the NRC Research Press Web site at http://cjes.nrc.ca on November 15, 2011.
- Published by NRC Research Press