Structural mapping in central Newfoundland has identified seven distinct phases of deformation (D1 to D7), the most significant of which are D1, D2, and D4. D1 involved the formation of a Middle and Late Ordovician southsoutheast-directed thrust belt and concomitant development of mylonite and phyllonite. A Late Ordovician to Early Silurian D2 thrust and fold belt overprints D1 mylonitic deformation and is the most distinctive deformation event in the study area. Late Silurian to Devonian D4 is responsible for folds and north-northwest-directed dextral thrust and reverse faults that overprint D1 to D3 structures. D4 structures in central Newfoundland include the Exploits–Gander boundary. Subsequent deformation is generally of local significance only. The arc–back-arc complexes making up the various terranes in central Newfoundland are predominantly juxtaposed along D1 shear zones, which include the Red Indian Line. Our data indicate that terrane boundaries initiated during D1 may have protracted deformation histories spanning several deformation events. This has important implications for the interpretation of terrane boundaries in Newfoundland, as D1 terrane boundaries may be interpreted as D2 or D4 shear zones depending on the intensity of overprinting or reactivation. The deformation history proposed in this paper corresponds closely to that of established Appalachian orogenic cycles. D1 is correlated with the Ordovician Taconic orogeny and involved accretion of arc–back-arc complexes to the Laurentian margin. D2 and D4 are correlated with the Ordovician–Silurian Salinic and Silurian–Devonian Acadian orogenies, which involved the subsequent accretion of the Ganderia and Avalonia microcontinents to the Laurentian margin, respectively.
The Appalachian–Caledonian orogen is a classic example of a long-lived (ca. 200 Ma) Paleozoic accretionary orogen formed as a result of the successive closure of Iapetus, Tornquist, and Rheic oceans (Williams and Hatcher 1983; van Staal et al. 1998; van Staal 2005). The Appalachian deformation along the Laurentian margin can be described in terms of five orogenic episodes: the Early to Middle Ordovician Taconic, Early to Late Silurian Salinic, latest Silurian to Early Devonian Acadian, Middle Devonian to Early Carboniferous Neoacadian, and Carboniferous to Permian Alleghenian orogenies (van Staal 2005). These orogenies resulted from the successive arrival of the Dashwoods, Ganderia, Avalonia, and Meguma microcontinents and Gondwana continent at the Laurentian margin. Reconstruction of the assembly and dispersal of these terranes along the Laurentian margin requires understanding of structures associated with the Appalachian orogenic episodes, their kinematic significance, and absolute age.
This study is based on regional field investigations (i.e., Lissenberg et al. 2005b; Rogers et al. 2005b; van Staal et al. 2005c; van Staal et al. 2005b; van Staal et al. 2005a) and addresses the structural history of two composite terranes, the peri-Laurentian Annieopsquotch accretionary tract (e.g., van Staal et al. 1998; Zagorevski et al. 2006) and the peri-Gondwanan Victoria Lake Supergroup (e.g., Evans and Kean 2002; Rogers and van Staal 2002) along the Red Indian Line, the main Iapetus suture zone (Fig. 1). The interpretations presented herein are supported by geochronology (e.g., Lissenberg et al. 2005a; see following text), geochemistry (e.g., Lissenberg et al. 2005a; Zagorevski et al. 2006), and both shallow and crustal-scale seismic reflection surveys (i.e., Thurlow et al. 1992).
The deformation history of the central mobile belt of the Newfoundland Appalachians is herein described in terms of seven phases of deformation (D1 to D7). The different generations of structures are primarily recognized by overprinting relationships, as well as style, orientation, and (or) metamorphic assemblages. The deformation history is correlated with the well-established orogenic cycles in the Appalachians, specifically the Taconic, Salinic, and Acadian orogenies (van Staal et al. 2005c), and emphasizes the importance of the Middle to Late Ordovician Taconic orogeny. Support for Ordovician and Silurian southeast-directed thrust and fold belts, as well as identification of significant thrust and strike-slip deformation during the Ordovician and Silurian, has important implications for the assembly of terranes along the composite Laurentian margin. Significantly, the absence and (or) excision of tectonostratigraphic units and juxtaposition of unrelated terranes, as hypothesized by Elders (1987) for example, may be related to strike-slip deformation. The proposed deformation history has significant implications for the regional tectonic framework. Specifically, the apparent disparity between the structural history inferred from the seismic studies, which suggest the presence of a southeastdirected accretionary tract (e.g., Thurlow et al. 1992; van der Velden et al. 2004), and that from surface mapping, which documents a northeast-directed thrust belt (e.g., Lafrance and Williams 1992; Kerr et al. 1996), can be addressed.
Regional ductile deformation in central Newfoundland has been traditionally interpreted as Taconic or Acadian (e.g., Dean and Strong 1977; Kean and Jayasinghe 1980; Karlstrom et al. 1982; Hibbard 1983; Kean 1983). Regional investigations of the Early to Middle Ordovician Taconic deformation in the Dunnage Zone are rare (e.g., Dean and Strong 1977; Hibbard 1983; Szybinski 1995) or restricted to mine scale (e.g., Calon and Green 1987; Gaboury et al. 1996), although significant progress has been made using geochronology (Lissenberg et al. 2005a) and seismic reflection surveys (Thurlow et al. 1992; van der Velden et al. 2004).
Most of the tectonic studies have presented evidence for Silurian–Devonian deformation (e.g., Karlstrom et al. 1982; Arnott et al. 1985; van der Pluijm 1986; Dunning et al. 1990; Elliott et al. 1991; Lafrance and Williams 1992; Williams et al. 1993; O’Brien 2003), initially ascribed to the Acadian orogeny (e.g., Pajari et al. 1974). However, the apparent age span (>40 million years) does not adhere to the original strictly Late Silurian to Early Devonian definition of the Acadian orogeny (Bradley 1983; Robinson et al. 1998; van Staal 2005). To rectify this problem, Dunning et al. (1990) proposed a correlation of the Early to Late Silurian orogenesis with the Salinic disturbance recognized by Boucot (1962) in New England. The Salinic orogeny in western Newfoundland involved the development of a northwest-directed thrust and fold belt in the Humber Zone (Fig. 1; e.g., Cawood et al. 1994) and a southeast-directed thrust and fold belt in the eastern Notre Dame subzone and adjacent Exploits subzone (e.g., Dean and Strong 1977; Lafrance and Williams 1992; Szybinski 1995; O’Brien 2003). The Acadian deformation resulted in northeast-trending steeply inclined folds and steeply dipping faults in the Notre Dame Bay area (Fig. 1). The faults comprise both dextral transcurrent shear zones and northwest-directed thrusts (van der Pluijm 1986; Elliott et al. 1991; Lafrance and Williams 1992; O’Brien 2003). The northwest-directed thrust panels within the Notre Dame subzone are locally overturned (e.g., Dean and Strong 1977; Kerr et al. 1996), which can be ascribed to superimposition on an earlier southeast-directed thrust stack formed during the Early to Late Silurian (e.g., van der Pluijm 1986).
Cambrian and Ordovician rocks
The stratigraphic relationships in central Newfoundland have been recently redefined (i.e., Lissenberg et al. 2005a, 2005b; Rogers et al. 2005b; van Staal et al. 2005c; van Staal et al. 2005b; van Staal et al. 2005a; Rogers et al. 2006; Zagorevski et al. 2006, 2007a) and are only briefly summarized here. Central Newfoundland Appalachians are underlain by two major composite terranes, the peri-Laurentian Annieopsquotch accretionary tract (van Staal et al. 1998; Zagorevski et al. 2006) and the peri-Gondwanan Victoria Lake Supergroup (Evans and Kean 2002; Rogers and van Staal 2002), which are juxtaposed along the Red Indian Line (Figs. 1⇓–3). Both of these consist of multiple fault-bounded tectonostratigraphic units.
The peri-Laurentian Annieopsquotch accretionary tract lies to the west of the Red Indian Line and comprises the Annieopsquotch ophiolite belt (ca. 480 Ma), Lloyds River Ophiolite Complex (ca. 473; Zagorevski et al. 2006), continental bimodal volcanic arc rocks of the Buchans Group (ca. 473 Ma; Zagorevski et al. 2006), and continental bimodal arc and back-arc volcanic and epiclastic rocks of the Red Indian Lake Group (ca. 465–460 Ma; Zagorevski et al. 2006). The Annieopsquotch ophiolite belt, Lloyds River Ophiolite Complex, and Buchans Group were stitched by synkinematic intrusive rocks of the Otter Pond Complex by 468 Ma (Lissenberg et al. 2005a). The Annieopsquotch ophiolite belt and its northwestern boundary are intruded by diorite to granodiorite of the synkinematic Pierre’s Pond Suite (462– 459 Ma; Lissenberg et al. 2005a).
The Victoria Lake Supergroup lies to the east of the Red Indian Line and comprises the volcano-sedimentary rocks of the Wigwam Brook Group (ca. 453 Ma; Zagorevski et al. in press a) that unconformably overlie the bimodal tuffaceous ensialic arc rocks and pillow basalt of the Lower Ordovician Pats Pond Group (ca. 487 Ma; Zagorevski et al. 2007a), the volcano-sedimentary back-arc rocks of the Sutherlands Pond Group (ca. 462–457 Ma; Dunning et al. 1987; Rogers et al. 2005b; Zagorevski et al. 2007b), and the volcaniclastic ensialic arc rocks of the Tulks Group, which is the easternmost tectonostratigraphic unit in the study area. Regionally, the Ashgill to Llandovery marine turbidites of the Badger Group conformably overlie the Caradoc cover to the Victoria Lake Supergroup and its correlatives (Williams et al. 1993). The Badger Group has been interpreted to represent fore-arc siliciclastic rocks deposited during the closure of the Tetagouche–Exploits back-arc basin (van Staal 1994; van Staal et al. 1998; Valverde-Vaquero et al. 2006).
The Dunnage Zone underwent extensive Silurian magmatism related to the terminal stages of the Notre Dame arc and slab break-off (Fig. 3; 440 Ma to 427 Ma; Whalen et al. 2006). The Boogie Lake (ca. 435 Ma; Dunning et al. 1990) and Puddle Pond (430–427 Ma; Whalen et al. 2006) intrusive suites were emplaced in part syntectonically along the boundary of the Annieopsquotch ophiolite belt and Lloyds River Ophiolite Complex (i.e., Otter Brook shear zone, Figs. 2, 3; Lissenberg et al. 2005a). Related dykes syntectonically intrude the Sutherlands Pond Group in proximity to the Red Indian Line (ca. 432 Ma; see following text). Coeval continental red beds (ca. 430–427 Ma; Chandler et al. 1987; Dunning et al. 1990) that unconformably overlie the Notre Dame subzone and portions of the Exploits subzone (Fig. 3; Williams et al. 1993) are variably deformed throughout the study area and are commonly spatially associated with terrane boundaries (Zagorevski and van Staal 2002). The youngest Silurian unit in the study area, the Rogersons Lake conglomerate (Kean 1983), is mylonitized in the footwall of the Victoria Lake shear zone (Fig. 3). The presence of sparse red bed cover and plutonic rocks in the study area is crucial as it allows the separation of structures formed during Ordovician and Silurian deformation.
The study area displays a relatively simple macroscopic structure with internally folded structural panels bounded by curvilinear shear zones (Fig. 3; e.g., van Staal et al. 2005a). Despite the apparent macroscopic simplicity, seven phases of deformation (D1 to D7) have been recognized based on overprinting relationships, as well as available age and stratigraphic constraints. The thickness of individual structural panels varies between 0 and 5 km, indicating partial to total excision of tectonostratigraphic units along strike. For ease of discussion, the study area has been subdivided into four structural domains (King George IV, Pats Pond, Tulks Valley, and Red Indian Lake; Fig. 3). Structural data from the Red Indian Lake domain were compiled from Rogers et al. (2005a; 2005b).
The earliest deformation (D1) recorded in Cambrian– Ordovician rocks is represented by a schistosity, cleavage, or differentiated layering (S1), which was subsequently folded by upright open to isoclinal F2 folds in the study area (Figs. 4⇓–6). S1 is commonly transposed to S2 crenulation cleavage resulting in S1-2 composite foliation that forms the regional penetrative foliation; however, S1 can generally be identified and separated from the D2 overprint. In areas of low D1 strain, S1 is at shallow angle or subparallel to compositional layering (S0; Figs. 5a–5c). S1 is axial planar to very rare, tight to isoclinal recumbent F1 folds of S0 and quartz veins (Fig. 5d).
In general, D1 strain is heterogeneous and localized into metre- to kilometre-wide high-strain zones that form the boundaries of the various tectonostratigraphic units. Highstrain zones in the competent ophiolitic complexes frequently contain centimetre- to decimetre-scale metabasite boudins enveloped by belts of S1 mylonite. In sedimentdominated units, D1 strain is strongly partitioned into shale, resulting in boudinage and dismemberment of the more competent sandstone or tuff layers within the incompetent shale matrix and development of broken formation or mélange (e.g., Victoria Delta Fault; see fig. 12 in Thurlow et al. 1992). Within high-strain zones, S1 frequently contains a strongly developed elongation lineation (L1, Fig. 4) defined by minerals, lapilli, pebbles, dismembered porphyroclasts, and veins. The variability in the pitch of the lineations can generally be attributed to F2 and F4 folding, although variation due to possibly transpressional character of D1 shear zones cannot be ruled out. Rarely developed S–C fabrics and asymmetric porphyroclasts observed on surfaces parallel to L1 and perpendicular to S1 suggest south-southeast-directed motion (Figs. 5f, 5g).
In Ordovician rocks, D2 structures are characterized by well-developed, southwest-trending, steeply northwest-dipping, spaced S2 crenulation cleavage or schistosity (Figs. 5a–5e, 6). S2 is axial-planar to open to isoclinal, shallowly to moderately plunging upright F2 folds of S0 and S1 (Figs. 5a–5e, 6). Regionally, F2 folds are asymmetric and locally slightly overturned to the southeast with the northwest-dipping limbs longer than the southeast-dipping limbs. In addition, F2 folds are non-cylindrical on the regional scale (Fig. 4). For example, F2 folds in rocks straddling the Red Indian Line plunge moderately to the southwest in the Pats Pond domain and moderately to the northeast in the northeast portion of the Red Indian Lake domain (Figs. 3, 4; Lissenberg et al. 2005b; van Staal et al. 2005a).
Along northwest-dipping sections of D1 shear zones, mesoscopic transposition of S1 into S2 is commonly indicated by the presence of isoclinal to rootless intrafolial F2 folds of S1 and boudins containing S1 differentiated layering (Fig. 5e). This is accompanied by development of S–C fabric that suggests sinistral oblique reverse shear with the northwest side up. D2 strain was, therefore, localized along portions of D1 shear zones, which were reactivated as steep southeast-directed sinistral oblique reverse faults. Boudinage and segmentation of epidote veins and porphyroclasts in greenschist-facies metabasites suggest that the rocks in the composite D1-2 shear zones generally accommodated large strains with extensions >100%. The predominant deformation mechanisms were solution transfer, evidenced by presence of truncated porphyroclasts, mica-rich foliation surfaces, and neocrystallization in pressure shadows; and crystal plastic deformation evidenced by development of monocrystalline quartz ribbons, undulose feldspar and quartz, and kinking of mica porphyroblasts. In layered or laminated rocks, D2 strain was preferentially partitioned into phyllosilicate-rich septae, evidenced by extensive dynamic recrystallization of muscovite and chlorite into very fine-grained laminae.
D2 fabrics in the Ordovician tectonites are continuous with the fabric in the Silurian plutons and locally in the unconformably overlying Silurian supracrustal rocks (Fig. 3). The S2 foliation is the first fabric that is developed in the Silurian rocks, although it is not regionally penetrative in all of the Silurian rocks. Red beds and volcanic rocks around King George IV Lake, for example, generally lack macroscopic cleavage, although they are tilted into subvertical attitudes in many places. Local overturning of the red beds is attributed to later deformation. In the King George IV domain, a thin lenticular band (0–20 m) of red beds, which unconformably overlies the Lloyds River Ophiolite Complex, represents a tight syncline immediately situated in the footwall of the Otter Brook shear zone (Zagorevski and van Staal 2002). Bedding in the red beds is subparallel to the S1-2 composite fabric in the Otter Brook shear zone, suggesting that the red beds were overthrust by the Annieopsquotch ophiolite belt and progressively steepened during D2 and later deformation. Although D2 deformation has resulted in utilization of the Otter Brook shear zone as a D2 oblique reverse fault, this shear zone clearly experienced an earlier phase of deformation, indicated by F2-folded S1 mylonite and phyllonite (Fig. 5e). Additionally, the ca. 468 Ma Otter Pond Complex, which intrudes along the Otter Brook shear zone, contains xenoliths of D1 tectonite, indicating early Ordovician deformation (Lissenberg et al. 2005a). Therefore, Otter Brook shear zone is probably best described as a composite D1-2 shear zone (Fig. 7).
The best developed regional D2 fabrics in the Silurian rocks occur in the Puddle Pond Intrusive Suite in the Star Brook area (Tulks Valley Domain, Fig. 3; Whalen et al. 2006). Along the Otter Brook shear zone, the Puddle Pond Intrusive Suite has attained a very strong northeast-trending compositional layering and foliation that, although forming the first fabric in these rocks, are correlated with the regional S2 (Fig. 6). S2 compositional layering is locally manifested on a metre-scale as interlayering between tonalite and diorite sheets (Fig. 8a). On the microscopic scale, S2 is a differentiated layering with quartzofeldspathic and mica-rich layers transected by C and C′ shear bands (Fig. 8a inset). S2 is associated with a downdip to northerly raking L2 mineral lineation (Fig. 6). The attitude of the S–C fabrics and pitch of the lineation suggest sinistral oblique west-side-up motion along this portion of the Otter Brook shear zone during D2, consistent with an inverted metamorphic gradient in the footwall.
Evidence for D2 reactivation of D1 shear zones other than the Otter Brook shear zone has also been observed in the Wood Lake shear zone, Red Indian Line, Victoria Delta fault, and Barren Pond Fault. Significantly, the mélange zone that marks the Victoria Delta fault contains a suite of syntectonic dykes (see subsequent sections), which cut S1 but are F2-folded and dismembered in the mélange zone (see fig. 12 in Thurlow et al. 1992), suggesting reactivation of the Victoria Delta Fault, probably during D2.
In the footwall of the Otter Brook shear zone in the Star Brook area, a shallow to moderately northwest-dipping spaced zonal crenulation cleavage is locally developed in Ordovician rocks (Figs. 5b, 5c, 6). This crenulation cleavage is subparallel to S2 in the adjacent Puddle Pond Intrusive Suite; however, it clearly cuts across the composite S1-2 in this area and is designated as S3. The geometry and asymmetry of the F3 mesoscopic and microscopic open crenulations and associated cleavage (Fig. 5b) can be interpreted as shear bands related to southeast-directed reverse shear. Similarly oriented, shallowly dipping crenulation cleavages and related open folds of composite S1–2 occur throughout the Cambrian– Ordovician rocks, although they are particularly well developed in the mechanically weak phyllonite associated with the D1-2 shear zones. Since these are folded by F4 folds on the regional scale (Fig. 6), they are tentatively correlated with S3.
The intensity of D4 deformation is highest in the King George IV domain (Fig. 3), where a steep to moderately southeast-dipping crenulation cleavage (S4) and variably plunging F4 folds of composite S1-2 fabrics are developed (Figs. 5e, 9). Discrimination between S4 and composite S1-2 can be tenuous as both are steep and have subparallel strike. Wherever overprinting relationships cannot be established, they are differentiated on the basis of orientation, as S1-2 is generally northwest dipping, whereas S4 is predominantly southeast dipping (Figs. 6, 9).
The Silurian Rogersons Lake conglomerate (King George IV domain, Fig. 3; Kean 1983) lacks D1–D3 fabrics and displays the best D4 fabrics in the study area. S4 is a spaced composite cleavage typified by excellent development of S–C fabric and a strong downdip to northeast-plunging L4 elongation lineation defined by stretched pebbles and cobbles (Fig. 8b). The asymmetry of S–C fabric, strain shadows around pebbles (Fig. 8b), and the plunge of L4 indicate northwest-directed dextral reverse shear. These have the same kinematic significance as the fabric along the Victoria Lake shear zone (Fig. 3; Valverde-Vaquero and van Staal 2001), which truncates all of the D1-2 shear zones in the King George IV domain (Valverde-Vaquero and van Staal 2002; van Staal et al. 2005b; van Staal et al. 2005a). Accordingly, the Victoria Lake shear zone is interpreted as a D4 shear zone.
To the northwest, the red beds in the footwall of the Otter Brook shear zone (Zagorevski and van Staal 2002) are folded and exhibit a steeply to moderately southwest-dipping cleavage oriented similarly to S4 in Rogersons Lake conglomerate (Fig. 5e). The hanging-wall red beds and volcanic rocks (S0) are folded on the regional scale without significant cleavage development. Limited data on red beds suggests that these folds are shallowly plunging and overturned to the northwest (Fig. 9). Similarly, the Otter Brook shear zone is also folded by shallowly plunging, northwest-verging F4 folds (Figs. 5e, 7), resulting in a generally southeast-dipping composite S1–2 in the King George IV domain (Fig. 4). The F4 folds are intimately associated with narrow (1 cm to 1 m) brittle–ductile northwest-directed thrust faults, which potentially mark the latest stages of D4 in this area (see fig. 2f in Zagorevski and van Staal 2002).
The effects of D4 deformation are less evident outside the King George IV domain, and development of S4 cleavage is not penetrative. Here, S4 is largely manifested as a late, steep crenulation cleavage that cuts the composite S1-2 foliation and is associated with variably plunging asymmetric F4 folds. The small angle between the composite S1-2 and S4 structures has locally resulted in the reactivation of S1-2 surfaces during D4 (Fig. 5b). In these cases, the presence of F4 folds of shallowly dipping S3 definitively indicates D4 deformation. In the absence of well-developed S3, D4 partially to completely unfolded F3 folds (Fig. 10), obscuring evidence of D3. In the Pats Pond and Tulks Valley domains, the effects of D4 are locally indicated by dextral boudinage of quartz veins in D1-2 shear zones under low-grade conditions, suggesting local reactivation of D1-2 shear zones as steep dextral faults.
D1–D4 structures are overprinted by a heterogeneously developed northwest-striking, steeply to moderately dipping S5 crenulation cleavage and moderately to steeply plunging F5 folds of S1 to S4 (Fig. 9). Development of S5 is most intense in the area at the boundary of Pats Pond and King George IV domains, where finely spaced S5 is axial-planar to steeply plunging F5 folds of composite S1-2. In the Tulks Valley area, S5 is spaced on a metre scale and is manifested as northwest-striking kink bands. Shallowly plunging recumbent F6 crenulations and box folds with locally developed axial-planar spaced S6 crenulation cleavage and conjugate kink bands are almost ubiquitously developed in all units and lithologies, although intensity appears to decrease toward the northwest. The establishment of conclusive overprinting relationships with S5 is difficult due to heterogeneous development of S5 and S6; however, several outcrops suggest that S5 is folded by F6. The latest phase of deformation (D7) in the study area is represented by late east-trending brittle faults with up to 700 m apparent dextral displacement on the surface (Oneschuk et al. 2001; Oneschuk et al. 2002; Lissenberg et al. 2005b; Rogers et al. 2005b; van Staal et al. 2005c; van Staal et al. 2005b; van Staal et al. 2005a). The faults are not exposed and no small-scale structures have been positively related to them, although they are most likely steep structures. Accordingly, their kinematic significance is ambiguous at this time.
To constrain the age of D2 deformation, two U–Pb geochronology determinations were conducted at the Geological Survey of Canada (Ottawa, Ontario) utilizing thermal ionization mass spectrometry (TIMS). U–Pb TIMS analytical methods utilized in this study are outlined in Parrish et al. (1987). Heavy mineral concentrates were prepared using standard crushing, grinding, Wilfley™ table, and heavy liquid techniques. Mineral separates were sorted by magnetic susceptibility using a Frantz™ isodynamic separator. Multigrain zircon fractions were very strongly air abraded and multigrain titanite fractions were lightly air abraded following the method of Krogh (1982). Treatment of analytical errors follows Roddick et al. (1987), with errors on the ages reported at the 2σ level (Table 1). U–Pb TIMS concordia diagrams are presented in Fig. 11. Concordia ages (Ludwig 1998) and errors quoted in the text are at 2σ with decay constant errors included. Additional information on the analytical procedures is provided in Zagorevski (2006).
Puddle Pond Intrusive Suite (VL02A-241)
At Star Brook (Fig. 3), Puddle Pond Intrusive Suite diorite (431.6 ± 4 Ma, 40Ar/39Ar hornblende; Whalen et al. 2006) is interlayered on metre scale with Puddle Pond tonalite (427 ± 1 Ma zircon; Whalen et al. 2006) and intruded into the trace of the Otter Pond shear zone. Both the diorite and tonalite are strongly sheared (Fig. 8a). Shear-sense indicators indicate south-southeast-directed reverse shear at amphibolitefacies metamorphic conditions; these rocks are progressively overprinted by greenschist-facies metamorphic assemblages to the east (footwall), suggesting hot D2 emplacement of plutons over D1 tectonites in the footwall. Puddle Pond diorite was sampled to constrain the duration of amphibolite-facies D2 deformation. The sample yielded abundant clear, uniformly light brown-orange fragments of euhedral titanite crystals (average 310 by 220 μm) with rare fractures and opaque inclusions. Three multigrain titanite fractions were analyzed and all three analyses overlap each other and the concordia (Fig. 11). A concordia age calculated using all three analyses is 426.0 ± 1.4 Ma (mean square of weighted deviates (MSWD) of concordance and equivalence = 0.29, probability = 0.92). As the closure temperature of titanite is >500 °C (Frost et al. 2000), this age is interpreted to represent cooling of the diorite below amphibolite-facies conditions. Since the deformation outlasted amphibolite-facies conditions as indicated by greenschist overprint, D2 at Star Lake Dam must have continued after 426 Ma.
Felsic dyke (RAX00-902)
A suite of feldspar porphyritic felsic dykes intrudes black shale mélange and broken formation of the Victoria Delta Fault (Thurlow et al. 1992), a few hundred metres south of the Red Indian Line that is here hidden beneath Red Indian Lake (Fig. 3). These dykes crosscut the existing scaly mélange fabric (see fig. 12 of Thurlow et al. 1992), which is parallel to the composite S1−2 foliation in the adjacent volcanic and sedimentary rocks. The felsic dykes are also dismembered within the mélange zone and some parts contain a weak foliation parallel to S1-2. These relationships suggest syn-D2 or inter-tectonic (D1 to D2) emplacement of the dyke into the mélange. The dyke sample contained a moderate amount of zircon and four multigrain fractions were analyzed (Fig. 11). Two fractions (C1 and E1) overlap each other and concordia. A concordia age using these two analyses is calculated to be 432.4 ± 0.8 Ma (MSWD of concordance and equivalence = 0.19, probability = 0.90). Fraction B1 is interpreted to have undergone minor Pb loss and fraction C2 contains an inherited component, probably of Ordovician age. The age of 432 ± 1.4 Ma is interpreted to be the crystallization age of the felsic dyke. The timing of D2 deformation is thus syn- to post- 432 ± 1.4 Ma.
The central Newfoundland Appalachians preserve a complex collage of tectonostratigraphic units that were formed and structurally juxtaposed during the closure of the Iapetus Ocean. Four main phases of ductile deformation (D1–D4) were identified, followed by several minor brittle–ductile episodes (D5–D7). Significantly, the less complex structural history of the unconformable Silurian cover in the study area and heterogeneous overprint of D1 by D2–D7 structures has allowed the separation of Ordovician D1 deformation and elucidation of its kinematic significance, which allows us to address long-standing issues in the assembly of the Dunnage Zone, principally the timing of accretion and the vergence of D1 structures. To this end, reconstruction of D1 tectonostratigraphy and constraints on the age of structures resulted in the schematic block diagram depicting D1–D4 evolution presented in Fig. 12.
D1 strain is most intense along the boundaries of the structural panels containing the tectonostratigraphic units (i.e., the Otter Brook shear zone, Wood Lake shear zone, Red Indian Line, Victoria Delta Fault, and Barren Pond Fault, Figs. 2, 3). Although these zones generally contain a composite S1–2 fabric, the predominance of mylonitic S1 combined with overprinting by F2 folds on small (Figs. 5d, 5e) and regional scale indicate that these boundaries are principally D1 shear zones that were reoriented and locally reactivated during D2 deformation, rather than D2 shear zones.
Preservation of shallow-dipping mylonitic S1 in the hinges of upright F2 folds (Fig. 5d) and moderately north-northwest-dipping enveloping surfaces of asymmetric F2 folds suggests that S1 was dipping shallowly northwest prior to folding. Although this geometry could also indicate location on the northwest-dipping limb of a regional-scale upright fold, the lack of repetition of tectonostratigraphic units to the southeast strongly suggests this is not the case. More likely the original orientation of the enveloping surface of S1 controlled subsequent asymmetry of upright F2 folds, such that its enveloping surface also had a shallow northwest dip (Fig. 12a).
Localization of D2 strain along D1 shear zones and local reactivation complicates the interpretation of the direction of transport during D1. In the Tulks Valley domain, for example, L1 extension lineations in the Barren Pond Fault are commonly reutilized as F2 fold axes, whereas other shearsense indicators, such as S–C fabrics, are likely to have formed during D2, complicating elucidation of the D1 transport direction. In the Pats Pond domain, the upright F2 folds plunge to the southwest, and orientation of S2 is oblique to the very well-developed shallowly south-plunging to horizontal L1 stretching lineation in west-dipping S1. The sparsely preserved D1 shear-sense indicators suggest sinistral strike-slip to southeast-directed normal oblique motion (Figs. 5f, 5g). If the effects of F2 folding are removed and the surface is restored to a northwest dip, L1 lineation would plunge to the north and have a moderate rake suggesting south-southeast-directed sinistral oblique motion.
The consistent southeast younging of the structural panels in the Annieopsquotch accretionary tract (Fig. 2), subparallelism of the S1 fabrics with S0 compositional layering, and map-scale low-angle cut-offs of tectonostratigraphic units are best explained by originally shallow northwest-dipping D1 shear zones. Combining this evidence with the consistent old over young relationships across D1 shear zones suggests that D1 shear zones most likely represent southeast-directed thrusts formed during accretion of the elements of the Annieopsquotch accretionary tract and Victoria Lake Supergroup to the Laurentian margin (Fig. 12b).
Age of D1 deformation and local correlations
Geochronological, isotopic, and stratigraphic data indicate that D1 thrusting was not a single short-lived accretionary episode but was protracted. Early D1 thrusts, such as the Otter Brook and possibly Mink Lake shear zones (Figs. 2, 3), juxtaposed Annieopsquotch ophiolite belt, Lloyds River Ophiolite Complex and Buchans Group by ca. 468 Ma (Lissenberg et al. 2005a). Outside the immediate study area, shear zones that share the age and characteristics with early D1 thrusts include the Hungry Mountain thrust (Figs. 2, 3; Thurlow 1981; Calon and Green 1987; Thurlow and Swanson 1987) and Lloyds River Fault (Figs. 2, 3; Lissenberg et al. 2005a).
The Hungry Mountain thrust (Thurlow 1981; Calon and Green 1987; Thurlow and Swanson 1987) is a probable correlative of the Otter Brook shear zone and juxtaposes the plutonic Hungry Mountain Complex (463 ± 4; 467 ± 8; and 469 ± 1 Ma; Whalen et al. 1997; van Staal et al. 2007) with the structurally underlying epidote–amphibolite-facies ophiolite (i.e., Harry’s River metabasite; e.g., Calon and Green 1987) and sub-greenschist-facies Buchans Group (Thurlow 1981; Calon and Green 1987; Thurlow and Swanson 1987). The development of the Buchans thrust stack is thought to have occurred synchronously with the Hungry Mountain thrust (e.g., Airport Thrust; Calon and Green 1987; Thurlow and Swanson 1987; Thurlow et al. 1992) and hence, during D1. Hot emplacement of the Hungry Mountain Complex is suggested by an inverted metamorphic gradient in the footwall (Thurlow 1981), thus the thrusting must have occurred prior to the cooling of the youngest pluton (463 ± 4 Ma; Thurlow 1981; van Staal et al. 2007). The Lloyds River Fault, which forms the boundary between the Annieopsquotch ophiolite belt and Dashwoods subzone (Figs. 2, 3), accommodated sinistral south-southeast-directed underthrusting of the Annieopsquotch ophiolite belt from 470 to 458 Ma beneath the tonalites of the Notre Dame arc (Fig. 3; Lissenberg et al. 2005a; Lissenberg and van Staal 2006) under intermediate pressure but anomalously high temperatures (780 °C and 0.6 GPa; Lissenberg and van Staal 2006), suggesting correlation with the early D1 tectonic setting in Buchans.
D1 shear zones to the east of the Lloyds River Ophiolite Complex (i.e., Wood Lake shear zone, Red Indian Line, Victoria Delta Fault, and Barren Pond shear zone, Fig. 3) involve rocks that are too young (<460 Ma) to be correlatives of the Middle Ordovician phase D1 thrusting (Fig. 2). Hence these shear zones are interpreted to be Late Ordovician D1 thrusts (Fig. 12b). The almost ubiquitous presence of Caradoc (Late Ordovician) black shale mélange along segments of late D1 thrusts (e.g., Victoria Delta Fault; Thurlow et al. 1992; and Red Indian Line; Rogers and van Staal 2002) and involvement of tectonostratigraphic units as young as 453 ± 4 Ma (Wigwam Brook Group; Zagorevski 2006) suggest that the juxtaposition of the Annieopsquotch accretionary tract and Victoria Lake Supergroup along the Red Indian Line and initiation of D1 in the Victoria Lake Supergroup occurred in the late Caradoc (i.e., 455–450 Ma; McKerrow and van Staal 2000), consistent with the lack of involvement of any identifiable Ashgill age (latest Late Ordovician) rocks in these D1 shear zones.
Correlatives of the late D1 shear zones can be readily identified in the Buchans area (Fig. 3). For example, the Powerline Fault, an Ordovician thrust locally marked by black shale mélange (i.e., Wiley’s River Fault; Thurlow et al. 1992), terminates the early D1 antiformal thrust stack in Buchans and juxtaposes portions of the Buchans Group with the structurally underlying Red Indian Lake Group. The minimum age of D1 deformation is provided by the unconformably overlying Silurian strata, which contain D1 tectonite clasts (Kean and Jayasinghe 1980; Kean 1983), Silurian crosscutting plutons containing xenoliths of D1 tectonites (e.g., Boogie Lake; Zagorevski and van Staal 2002; and Lloyds River Granite; C.J. Lissenberg, personal communication, 2005), and Silurian plutons that stitch shear zones (e.g., Halfway Mountain Granodiorite; Whalen et al. 1987; Whalen et al. 2006). These relationships suggest that D1 deformation was finished by ca. 435 Ma at the latest, the age of the oldest crosscutting pluton in the study area (Fig. 3; Boogie Lake Intrusive Suite; Dunning et al. 1990). However, as no record of ca. 450– 435 Ma deformation has been reported in central Newfoundland, the span of D1 was probably ca. 470 to 450 Ma.
Age of D2 deformation
The Boogie Lake Intrusive Suite (ca. ; Dunning et al. 1990) provides a maximum age of D2 deformation. This suite is locally weakly foliated (S2), and satellite dykes are folded by F2, suggesting pre- to syn-D2 emplacement and cooling. Similarly, Silurian dykes (432.4 ± 0.8 Ma; see previous sections) suggest syn-D2 emplacement, indicating initiation of D2 prior to ca. 432 Ma.
Puddle Pond Intrusive Suite diorite and tonalite (431 ± 4 Ma 40Ar/39Ar and 427 ± 1 Ma U/Pb zircon, respectively; Whalen et al. 2006) at Star Brook provide constraints on the duration of D2. Both tonalite and diorite are deformed at amphibolite (hanging wall) to greenschist (footwall) facies. The cooling of titanite below its closure temperature (>500 °C: Frost et al. 2001) followed by lower grade overprint with the same kinematic significance indicates that D2 continued after 426 ± 1 Ma (U/Pb titanite; see previous sections).
The plutonic rocks in the study area constrain D2 deformation to have started at least by ca. 435 Ma and finished sometime after ca. 426 Ma. This age range of deformation implies that all of the Silurian supracrustal rocks (ca. 430– 427 Ma; Chandler et al. 1987; Dunning et al. 1990) were deposited syntectonically. In the study area, Silurian red beds (ca. 429 Ma; Dunning et al. 1990) are intimately associated with the footwall of the D2 reactivated Otter Brook shear zone, suggesting that the red beds were probably deposited syntectonically at the toe of the southeast-vergent thrust stack (Fig. 12c; cf. Szybinski 1995). If correct, the latter implies that most of the thrust stack had breached sea level during the Early Silurian and was being eroded, considering the regional distribution of the red beds.
Ages of D3–D7 deformation
There are no direct constraints on the timing of D3 deformation. S3 clearly postdates F2 upright folding. However, in the Star Brook area, S3 is nearly parallel with S2 in the Puddle Pond Intrusive Suite Pluton, suggesting that S3 may be kinematically related to and a late manifestation of D2. The D3 deformation is thus best constrained as post- 426 ± 1 Ma.
Similarly to D3, the age of D4 is poorly constrained at present. Since F4 folds clearly overprint composite S1-2 fabric, the start of D4 postdates 426 ± 1 Ma (see earlier in the text). To the southeast of the study area, the age of D4 deformation is constrained by the Meelpaeg subzone synkinematic migmatite (417 ± 2 Ma; Currie et al. 1992; van Staal et al. 2005c), which immediately predates the final juxtaposition of the Meelpaeg subzone amphibolite-facies tectonites with the Exploits subzone greenschist-facies tectonites along the Victoria Lake shear zone (Fig. 12d; Valverde-Vaquero and van Staal 2001).
Regional constraints on D1 and D2
Interpretation of significance and location of the terrane boundaries in the correlative tracts at Notre Dame Bay has been controversial (e.g., Dean and Strong 1977; Arnott et al. 1985; van der Pluijm 1986; Elliott et al. 1991; Lafrance and Williams 1992; Dec and Swinden 1994; Szybinski 1995; Kerr et al. 1996; Dec et al. 1997). Polyphase deformation has resulted in interpretations of terrane boundaries ranging from north-directed thrusts, south-directed thrusts, or strikeslip faults of regional significance to relatively minor faults (see Dean and Strong 1977 for discussion, also Lafrance and Williams 1992; Dec and Swinden 1994; Kerr et al. 1996). The earliest deformation in the Annieopsquotch accretionary tract of the western Notre Dame Bay area is recorded by the synkinematic Coopers Cove Pluton (465 ± 2.5 Ma U/Pb zircon; Szybinski 1995), probably correlative to D1 as presented here. However, since some of the terrane boundaries in Notre Dame Bay involve Silurian red beds, they are generally interpreted as wholly Silurian structures, formed during southeast-directed thrusting and folding (Lafrance and Williams 1992; Szybinski 1995). The minimum age of Silurian deformation in Notre Dame Bay is provided by Ludlow felsic dikes (422 ± 2 Ma; Elliott et al. 1991) that cut across Silurian structures in the Botwood Group (Williams et al. 1993), as well as by the Stony Lake volcanic rocks (Ma; Dunning et al. 1990), which unconformably overlie deformed rocks of the Botwood Group. The age and polarity of these structures suggest a strong correlation to D2–D3 and place a minimum age limit on the timing (i.e., 422 Ma). Subsequent deformation in Notre Dame Bay was accommodated by a northwest-directed thrust and fold belt (Szybinski 1995; Kerr et al. 1996) and (or) dextral shear zones and folds (Lafrance and Williams 1992). The maximum age of this deformation is constrained by crosscutting dykes of the Loon Bay Suite (408 ± 2 Ma: Elliott et al. 1991). The age and vergence of the structures suggest that this phase of deformation is correlative to D4. Hence, D4 is constrained to ca. 422 to 408 Ma.
Application of the structural history to regional geology
Positive correlation of D2 to D4 structures from central Newfoundland to Notre Dame Bay yields significant promise for establishing a regionally consistent structural framework. However, correlation of D1 is difficult as no terrane boundaries could be demonstrably related to this period of orogenesis in Notre Dame Bay despite identification of either Middle Ordovician deformation (i.e., D1; Szybinski 1995) or the assembly of the fault-bounded Roberts Arm terranes prior to the Silurian indicated by unconformable Springdale Group cover (e.g., Kerr et al. 1996). We postulate that this discrepancy can be reconciled by considering the evidence for Silurian and younger D2– D4 reactivation of the D1 terrane-bounding faults (Fig. 7). In addition, earlier formed mylonites marking terrane boundaries can be buried below northwest-directed thrusts, as is the case along portions of the Otter Brook shear zone (Fig. 7; Zagorevski and van Staal 2002).
These points are discussed by considering the Lobster Cove – Chanceport (LCC) fault in Notre Dame Bay, which separates the composite Cambrian – Early Ordovician oceanic Lushs Bight terrane (van Staal et al. 2007) from the Annieopsquotch accretionary tract (Fig. 13; e.g., Dean and Strong 1977; Lafrance and Williams 1992; Szybinski 1995). The LCC fault records a period of Silurian south-southeastdirected thrusting, which emplaced Lushs Bight Group oceanic rocks over the Roberts Arm Group and its unconformable Silurian Springdale Group cover (Fig. 13; Dean and Strong 1977; Lafrance and Williams 1992; Szybinski 1995). Subsequent deformation steepened and locally overturned the thrust (Dean and Strong 1977) and was followed by imbrication of Roberts Arm Group, Springdale Group, and adjacent ophiolitic rocks along north-directed thrusts (Kerr et al. 1996). On Sunday Cove Island, the LCC fault is marked by a locally extensive phyllonite zone, which emplace greenschist-facies metamorphic tectonites of the Lushs Bight Group over the Roberts Arm Group. In comparison, the Springdale Group, which is situated in the LCC’s footwall, is only weakly deformed, locally overturned, and virtually unmetamorphosed. This contrast in metamorphic grade and structure suggests that the LCC fault was active in the Ordovician and reactivated during D2 southeast-directed reverse faulting following pre-Springdale exhumation of D1 tectonites. Proposed reactivation of Ordovician D1 LCC and its subsequent reactivation during Silurian D2 would be analogous to the Otter Brook shear zone (Fig. 7).
Tectonic significance of the structures
D1 in the Annieopsquotch accretionary tract has been constrained by geochronology and tectonostratigraphic relationships to involve several discrete or continuous thrusting events from Middle to Late Ordovician (Lissenberg et al. 2005a). The enveloping surface of asymmetrical F2 folds of S1 and old-over-young relationships indicate juxtaposition of the tectonostratigraphic units along shallow northwest-dipping thrusts (Figs. 12a, 12b). Shallow (Thurlow et al. 1992) and crustal-scale Lithoprobe seismic reflection surveys (van der Velden et al. 2004) support our interpretation of the original orientation and kinematic significance of the D1 thrusts. Thurlow et al. (1992) have identified many of the D1 thrusts in the Buchans area, where they dip between 0° and 45° to the northwest. Crustal-scale reflectors (>25 km depth) that correspond to the surface trace of the Lloyds River Fault, Otter Brook shear zone, and the Red Indian Line dip approximately 45°–55° to the northwest on the reinterpreted seismic reflection profiles of van der Velden et al. (2004). D1 thrusts are continuous along the entire length of the mapping area (∼150 km) and likely continue to the northeast to the Roberts Arm Group in Notre Dame Bay (Bostock 1988).
The regional and crustal scale continuity of the D1 structures, as well as the character of metamorphism ranging from low-grade (e.g., pumpellyite-bearing assemblages) to amphibolite-facies conditions (Lissenberg and van Staal 2006; Zagorevski 2006), indicate that D1 structures formed during Middle to Late Ordovician assembly of the Annieopsquotch accretionary tract after Early Ordovician initiation of west-directed subduction. The latter was initiated as a result of closure of the Humber Seaway and the Laurentia – Notre Dame arc collision (Waldron and van Staal 2001; van Staal et al. 2007), which significantly slowed down or stopped convergence between Laurentia and the Notre Dame arc and its Dashwoods infrastructure. This phase of deformation (early D1; Fig. 12a, 14a) involves strictly peri-Laurentian tectonic elements and is, therefore, Taconic (van Staal et al. 2007).
The late Caradoc (Late Ordovician) arrival of the peri-Gondwanan Victoria Arc at the Laurentian margin closed the main Iapetan oceanic tract in a Molucca sea-type collision (Fig. 14b; van Staal et al. 1998; Zagorevski et al. in 2007b) and juxtaposed the peri-Laurentian and peri-Gondwanan terranes along the Red Indian Line (late D1; Figs. 12b, 14c). This terminal phase of D1 led to continuation of southsoutheast-directed thrusting that accommodated underthrusting and underplating of the Victoria arc beneath the Notre Dame arc and Annieopsquotch accretionary tract (Fig. 14c; van der Velden et al. 2004), marked by uplift in the Notre Dame subzone and syntectonic Late Ordovician to Early Silurian marine sedimentation in the Exploits subzone (lower part of Badger Group, Williams et al. 1993). Subsequent to the collision, subduction stepped back into the Exploits– Tetagouche back-arc basin (Fig. 14c; van Staal et al. 1998). Involvement of both peri-Laurentian and peri-Gondwanan terranes during late D1 indicates the end of the Taconic orogeny and start of the Salinic convergence at ca. 450 Ma.
The presence of the Silurian overlap sequence in the Annieopsquotch accretionary tract and western part of the Exploits subzone indicates that this portion of the Dunnage Zone was assembled prior to the Silurian. Development of a Silurian south-southeast-directed D2 thrust and fold belt (ca. 435 to 422 Ma: see earlier in the text), intrusion of syntectonic plutons (see earlier in the text; Dunning et al. 1990), local mélange formation (e.g., Joey’s Cove Mélange; Lafrance and Williams 1992), Silurian unconformities (Williams 1995), and reactivation of D1 shear zones (e.g., Red Indian Line and Otter Brook shear zone; see earlier in the text) represent products of the Salinic orogeny (D2– D3). Salinic F2 folds steepened and reactivated portions of D1 shear zones as south-southeast-directed sinistral oblique reverse faults and thrusts (Fig. 12c). The Taconic and Salinic orogenies in the Annieopsquotch accretionary tract thus were kinematically similar, which is not surprising, since they are both related to accretionary events caused by west-directed subduction beneath Laurentia. The Salinic orogeny, because of the Ashgill–Wenlock (latest Ordovician to early Silurian) closure of the Exploits–Tetagouche back-arc basin (Fig. 14d; van Staal et al. 1998; Valverde-Vaquero et al. 2006), led to the accretion of Ganderia to the composite Laurentian margin along the Dog Bay Line suture zone (e.g., Dunning et al. 1990; Williams et al. 1993; van Staal et al. 1998; van Staal 2005).
During the early stages of the Salinic accretion, subduction had initiated outboard in the narrow seaway that separated Ganderia and Avalonia (van Staal 2005; Rogers et al. 2006; Valverde-Vaquero et al. 2006). The latest Silurian accretion of Avalonia is responsible for the start of the Acadian orogeny (ca. 422 Ma to 408 Ma; see earlier in the text) marked by the development of northwest-directed F4 folds and D4 shear zones, such as the Victoria Lake shear zone (e.g., Valverde-Vaquero and van Staal 2002; van der Velden et al. 2004), and reactivation of D1-2 shear zones as dextral faults. Acadian D4 is the last significant phase of regional ductile deformation in the study area (Fig. 12d), and was followed by the formation of the regionally insignificant D5–D7 structures.
Mapping in central Newfoundland has allowed the identification of distinct structures related to the Taconic, Salinic, and Acadian orogenies. The time span of the Taconic structures (D1: ca. 470 to 450 Ma) indicates that it was a prolonged event, involving multiple accretionary episodes, marked by formation of south-southeast-directed sinistral oblique thrusts. Despite having similar kinematic significance, the Salinic (D2– D3: 435 to 422 Ma) structures can be separated from Taconic utilizing syntectonic Silurian rocks and overprinting relationships. The start of the Acadian orogeny (D4: 422 to 408 Ma) is marked by reversal of polarity of structures and initiation of northwest-directed thrust and fold belt. Application of the structural model developed for central Newfoundland to the Notre Dame Bay region yields significant promise for establishing a regionally consistent structural framework. Salinic and Acadian deformation are easily comparable between the two areas; however, the identification of Taconic structures in Notre Dame Bay is difficult at present. Future studies should concentrate on the identification of Ordovician D1 structures as these constrain the first-order architecture of the Dunnage Zone and exert primary control on the distribution of tectonostratigraphic units, including those hosting significant mineralization.
This work was undertaken as part of the first author’s Ph.D. thesis at the University of Ottawa (Ottawa, Ontario, Canada) supported by Geological Survey of Canada Targeted Geoscience Initiative, and funding from Natural Sciences and Engineering Research Council, Ontario Graduate Scholarship, and Strategic Areas of Development programs. Early reviews from K. Hattori, S. Carr, K. Benn, S. Barr, and M. Sanborn-Barrie and formal reviews by B. O’Brien, B. Lafrance, and N. Pinet are gratefully appreciated.
- Received March 5, 2007.
- Accepted June 19, 2007.
- Published on the NRC Research Press Web site at http://cjes.nrc.ca on November 20, 2007.
- © 2007 NRC Canada