The Gaspé Belt in the Restigouche area comprises three successions separated by a Late Silurian (Salinic) disconformity and an Early Devonian angular unconformity. The lower, Upper Ordovician to Lower Silurian sequence consists of siliciclastic turbidites of the Boland Brook and Whites Brook formations (Grog Brook Group), overlain by calcareous turbidites of the Pabos and White Head formations (Matapédia Group), and slope and shelf deposits of the Upsalquitch and Limestone Point formations (lower Chaleurs Group). Above the Salinic disconformity, the upper Chaleurs Group and the Dalhousie Group record a transgressive–regressive cycle. The former comprises Pridolian carbonate rocks of the West Point Formation and overlying Pridolian to Lochkovian sedimentary rocks of the Indian Point Formation. The Chaleurs Group is conformably overlain by Lochkovian to early Emsian subaerial volcanic rocks of the Dalhousie Group (Val d’Amour Formation), which is unconformably overlain by alluvial–lacustrine deposits of the late Emsian Campbellton Formation. Acadian orogenesis began during the Emsian and is characterized by open to closed folding, heterogenous cleavage development, and reverse and strike-slip faults. The Salinic orogeny is manifested in extensional block faulting, within-plate volcanism, and uplift and deep erosion of Early Silurian strata. Early Devonian high-level intrusion of the Matapédia Group, White Head clasts in Indian Point conglomerate, and thermal maturation data all indicate an extended period of Late Silurian – Early Devonian uplift in parts of the Restigouche area. Thermal maturities of West Point and Indian Point strata are within the oil and condensate windows and suggest potential for hydrocarbons in the study area.
The geology of the northern Appalachians is a record of the opening and closing of the early Paleozoic Iapetus Ocean and the collision and accretion of island arcs and microcontinents that populated the Iapetus Ocean between Laurentia and Gondwana (e.g., van Staal et al. 1998). The order and timing of terrane accretion and deformational events are in large part established by stratigraphic, paleontological, and structural studies of successor basins that link major elements of the orogen. The rocks underlying the Restigouche area constitute part of the Gaspé Belt (Bourque et al. 1995), also referred to as the Matapédia Cover Sequence (Fyffe and Fricker 1987; van Staal and de Roo 1995). The Gaspé Belt is a Late Ordovician to Middle Devonian (i.e., post-Taconian to broadly syn-Acadian) successor basin that oversteps the margins of two major zones of deformed Cambrian to Middle Ordovician rocks, namely the Humber Zone (Laurentian margin) to the northwest and Dunnage Zone (Iapetan oceanic tract) to the southeast (van Staal and de Roo 1995; Malo and Bourque 1993) (Fig. 1). Evidence that the Gaspé Belt is mainly underlain by rocks of Dunnage affinity is provided by numerous inliers of pre-Late Ordovician volcanic and sedimentary rocks in Maine, New Brunswick, and the Gaspé Peninsula (Fig. 1).
Previous workers have explained the Silurian–Devonian geologic history of the Gaspé Belt according to different sedimentological, structural, and tectono-magmatic models. For example, in the Gaspé Peninsula, stratigraphic and sedimentary facies analyses have led to an interpretation of basin evolution in terms of several transgressive–regressive cycles (Malo and Bourque 1993; Bourque et al. 1995, 2000, 2001; Malo 2001). In Maine, sedimentological studies (Bradley and Hanson 1998, 2002) and geochronological data from pre-, syn-, and post-tectonic plutons (Bradley et al. 2000; Bradley and Tucker 2002; Robinson et al. 1998) have been used to trace the Silurian–Devonian northwestward migration of the Acadian orogenic wedge and foreland basin. Kirkwood (1995), Malo and Kirkwood (1995), and Malo et al. (1995), based on progressive strain studies of deformed rocks and kinematic analyses of faults, have interpreted basin evolution in terms of the deformational history of the Gaspé Peninsula. Hibbard (1994) resolved structural evidence from northern Maine and elsewhere in the northern Appalachians into a two-phase kinematic pattern consisting of Silurian north–south shortening and sinistral shear, followed by Devonian (“classical” Acadian) dextral slip and northeast-trending folds and cleavage related to northwest–southeast shortening. Several models have been invoked to explain prolonged Gaspé Belt volcanic activity, beginning ca. 427 Ma (middle Wenlockian) with the Bryant Point Formation in northern New Brunswick (Walker and McCutcheon 1995; Meyers et al. 1998) and persisting to at least 407 Ma (early Emsian) based on concordant U–Pb ages for the Traveler Rhyolite in Maine (Rankin and Tucker 1995; Bradley et al. 2000) and the upper part of the Dalhousie Group in northern New Brunswick (this paper). These tectono-magmatic models variously involve delamination and asthenospheric upwelling (van Staal and de Roo 1995), transpressional rifting (Dostal et al. 1989, 1993; Keppie and Dostal 1994), and subduction (Bradley et al. 2000; Bradley and Tucker 2002).
In New Brunswick, results of regional mapping programs in different parts of the Gaspé Belt have been reported by Greiner (1967), Hamilton-Smith (1970), St. Peter (1978a, 1978b, 1979, 1982), Irrinki and Crouse (1986), Irrinki (1990), Wilson (1990, 2000a, 2002), and Walker and McCutcheon (1995). Except for areas proximal to the northern Miramichi Highlands (e.g., van Staal and de Roo 1995), however, there has been little attempt to understand the stratigraphy or structure of the belt within the larger context of orogenic evolution. Resolution of these problems has been hindered in New Brunswick by scanty radiometric data (especially from volcanic rocks), substandard volcanic lithogeochemical data, lack of detailed geological mapping, imprecise fossil ages, and very limited structural analysis. A treatment of all of these issues is beyond the scope of this paper, but some progress, based on the results of recent mapping programs in the northern and northwestern parts of the province (Wilson 2000a, 2002; Carroll 2003), is reported herein. The objective of this paper is to outline the stratigraphic framework of the middle Paleozoic succession in northernmost New Brunswick and, within this framework, combine sedimentological, palynological, structural, and thermal maturation data to interpret successor basin evolution. The Restigouche area (Figs. 1, 2) is underlain by rock units that span the entire temporal range of the Gaspé Belt, by a small inlier of Ordovician Dunnage basement, and by local post-orogenic (Carboniferous) terrestrial redbeds. It is transected by several faults that reveal a complex history of crustal movement and juxtapose domains of contrasting structural and thermal history. In addition, two erosional unconformities and two angular unconformities interrupt the Late Ordovician to Carboniferous stratigraphic succession. The Restigouche area is therefore of considerable interest when making comparisons with the other parts of the Gaspé Belt and a key location for understanding the evolution of the belt as a whole.
Regional stratigraphic relationships in four areas of northern New Brunswick and southeastern Gaspé Peninsula are summarized in Fig. 3. The oldest rocks in the Restigouche area are mafic volcanic rocks and overlying shale and chert of the Middle to Late Ordovician Balmoral Group, which is exposed in the Popelogan Inlier (Fig. 2). The Balmoral Group constitutes part of the Popelogan–Victoria arc (van Staal et al. 1998) and comprises subduction-related picritic to andesitic flows and pyroclastic rocks of the Goulette Brook Formation (Wilson 2003) and overlying dark grey slates and cherts of the Popelogan Formation. The latter contains early Caradocian graptolites of the Nemograptus gracilis zone (J. Riva, in Irrinki 1990) and is therefore coeval with Late Ordovician cherts that are widespread on the Gondwanan margin of Iapetus (e.g., van Staal and Fyffe 1991 and references therein). Late Ordovician (Caradocian) collision between the Popelogan-Victoria arc and Laurentia is inferred from a Caradocian hiatus between the Balmoral Group and overlying rocks of the Gaspé Belt successor basin (van Staal 1994; van Staal et al. 1998).
The Gaspé Belt is commonly regarded as comprising three structural zones, namely, from northwest to southeast, the Connecticut Valley – Gaspé Synclinorium, Aroostook– Percé Anticlinorium, and Chaleur Bay Synclinorium (Rodgers 1970) (Fig. 1). The rocks in the Restigouche area overstep the boundary between the latter two zones and occur in the hinge area of the Restigouche Syncline, a structural subdivision of the Chaleur Bay Synclinorium at the head of Chaleur Bay (Fig. 1). The Aroostook–Percé Anticlinorium is host to the oldest rocks in the Gaspé Belt, namely Upper Ordovician to Lower Silurian deep-water turbidite deposits that are broadly divided into a lower siliciclastic assemblage and an upper carbonate assemblage. The siliciclastic rocks have been assigned to the Garin Formation in Quebec (Malo 1988), the Madawaska Lake Formation in Maine (Roy and Mencher 1976), and the Grog Brook Group in northern New Brunswick (St. Peter 1978a), whereas the carbonate rocks are assigned to the Matapédia Group in New Brunswick and Gaspé (St. Peter 1978a; Lespérance et al. 1987) and the Carys Mills Formation in Maine (Pavlides 1968).
The Chaleur Bay Synclinorium consists of Lower Silurian to Lower Devonian rocks of the Chaleurs Group and Lower Devonian rocks of the Dalhousie Group and Campbellton Formation. On the southeastern margin of the Chaleur Bay Synclinorium, the Chaleurs Group unconformably overlies the Middle to Upper Ordovician Fournier Group in the Miramichi Highlands and Elmtree Inlier (Alcock 1935; Helmstaedt 1971; Walker et al. 1993; Walker and McCutcheon 1995). Where not faulted, the contact between strata assigned to the Aroostook–Percé Anticlinorium and Chaleur Bay
Synclinorium is conformable (e.g., St. Peter 1978a; Bourque et al. 1995), and Matapédia rocks crop out at several locations within the latter; hence, there is no stratigraphic basis for the distinction between the two zones. The complex history of the Chaleur Bay Synclinorium is expressed by (i) locally abrupt lateral and vertical facies changes related to differential uplift and eustatic sea level changes (Bourque 2001), (ii) Wenlockian–Ludlovian and Lochkovian to Emsian intraplate magmatic activity, and (iii) Late Silurian tectonism (e.g., Malo and Bourque 1993; van Staal and de Roo 1995; Malo and Kirkwood 1995). The latter period of tectonic activity has been referred to as the Salinic disturbance (Boucot et al. 1964), Salinian orogeny (Cawood et al. 1995), and Salinic orogeny (Dunning et al. 1990); the latter term is used herein.
In the Restigouche area, uplift associated with the Salinic orogeny has produced a widespread erosional unconformity (referred to herein as the Salinic disconformity) that separates the lower and upper parts of the Chaleurs Group. The upper part of the Chaleurs Group is conformably overlain by volcanic and minor sedimentary rocks of the Dalhousie Group, which in turn is unconformably overlain by the Campbellton Formation, following a brief Emsian hiatus (Fig. 3). Coarse-grained, flat-lying, Carboniferous terrestrial redbeds of the Bonaventure Formation unconformably overlie the Chaleurs and Dalhousie groups and the Campbellton Formation.
Gaspé Belt stratigraphy
Grog Brook Group
The Grog Brook Group (St. Peter 1978a) comprises a thick series of mainly siliciclastic turbidites that crop out in the central part of the Aroostook–Percé Anticlinorium. Sedimentary structures and bedforms, such as full and partial Bouma sequences, graded bedding, flute casts, and other sole markings (e.g., St. Peter 1978a; Wilson 1990; Carroll 2003), are typical of deep-water facies. The Grog Brook Group is divided into the Boland Brook Formation and conformably overlying Whites Brook Formation (Wilson 2002). A Late Ordovician age for the Grog Brook Group is indicated by collections of graptolites, brachiopods, bryozoans, and corals from scattered locations in northern New Brunswick (St. Peter 1978a). In the Restigouche area, chitinozoan microfaunas from some of the oldest exposed parts of the Boland Brook Formation, and from the Whites Brook Formation, have been assigned to the Cyathochitina vaurealensis and Hercochitina crickmayi zones of Richmondian (early to middle Ashgillian) age; these faunas permit a correlation with the Vauréal Formation on Anticosti Island and with the lower part of the White Head Formation at Percé (Martin 1980; Achab 1989). In contrast, the age of correlative rocks of the Garin Formation in the Gaspé Peninsula ranges into the Caradocian, suggesting regional diachroneity in the transition from the Garin Formation – Grog Brook Group to the Matapédia Group (Fig. 3). For example, graptolites from the upper part of the Garin Formation have been assigned a middle to late Caradocian age (upper Climacograptus spiniferus to Paraclimacograptus manitoulinensis zones) (Riva and Malo 1988).
Boland Brook Formation
The Boland Brook Formation mainly consists of thin-bedded noncalcareous siltstone, mudstone, fine-grained sandstone, and minor polymictic conglomerate (Wilson 2002). Bed thickness typically ranges from 4 to 15 cm, although some beds of fine- to medium-grained sandstone are up to 50 cm. The oldest exposed rocks in the Restigouche area are thin-to thick-bedded (up to about 1 m), fine- to coarse-grained, cross-laminated sandstone and polymictic conglomerate. Boland Brook conglomerates contain lithologically diverse, rounded to subangular clasts of felsic and mafic volcanic rock, fine-grained sedimentary rock, chert, quartz, feldspar, minor calcite, and accessory zircon, in a mudstone or siltstone matrix. Lithic clasts, with few exceptions, are unfoliated. At outcrop scale, conglomerate beds display tabular morphology; the larger scale geometry is unknown and may be channel-form. Sandstones are typically immature feldspathic lithic arenites and wackes, containing poorly to moderately sorted, subrounded to angular grains of quartz, feldspar, rock fragments, minor calcite, and abundant accessory zircon. Beds of weakly to moderately calcareous siltstone and sandstone become more common in the upper part of the unit, where it grades into the overlying Whites Brook Formation. The base of the Boland Brook Formation is not exposed in the Restigouche area, but near Oxford Brook, 40 km south of Squaw Cap Mountain (Fig. 2), coarse-grained clastic rocks similar to those in the lower Boland Brook (“Pat Brook beds” of Carroll 2000) overlie green cherts correlated with the Popelogan Formation. The thickness of the exposed part of the Boland Brook Formation has been estimated at 1600 m along Upsalquitch River (Wilson 2002) (Fig. 2).
Whites Brook Formation
The Whites Brook Formation predominantly consists of thin- to thick-bedded (∼6 cm to >1 m), medium- to coarsegrained, typically calcareous sandstone, grit, and minor conglomerate, with thin (2–8 cm) interbeds of dark grey noncalcareous shale or mudstone (Wilson 2002). In thin section, coarser grained lithotypes in the Whites Brook and Boland Brook formations are very similar, except for a much greater abundance of carbonate in the former. Lithic, feldspathic, calcareous sandstones and conglomerates contain angular to subrounded clasts of unfoliated felsic and mafic volcanic rock, quartz, feldspar, calcite, fine-grained sedimentary rock, chlorite, and accessory zircon. The Whites Brook Formation gradually thins out from southwest to northeast. The greatest thickness (up to 4000 m) is present at the type section on Whites Brook 35 km southwest of Squaw Cap Mountain (Fig. 2), but at Upsalquitch River (Fig. 2) the maximum thickness is about 500 m. Locally, the Whites Brook Formation is absent and the Boland Brook Formation is overlain by the Pabos Formation (see next section).
Regionally, the Matapédia Group shows remarkable lithological homogeneity. Nevertheless, in the Gaspé Peninsula, subtle but consistent lithologic variations have allowed the recognition of formations and members (Lespérance et al. 1987). Hence, in the Percé area, the Matapédia Group has been divided into the Ashgillian Pabos Formation and conformably overlying Ashgillian to Llandoverian White Head Formation; the latter includes, in order of younging, the Burmingham, Côte de la Surprise, L’Irlande, and Des Jean members (Skidmore and Lespérance 1981; Lespérance et al. 1987; Malo 1988). The Pabos and White Head formations have been recognized in New Brunswick (Wilson 2000a, 2002; Carroll 2003), although the White Head members have not. However, intervals of noncalcareous shale sandwiched between thick sequences of thin-bedded, deep-water lime mudstone (calcilutite) and calcareous shale have been reported in northwestern New Brunswick (Hamilton-Smith 1970; St. Peter 1978a; Wilson 2002) and northeastern Maine (Pavlides 1968) and may correlate with the upper Ashgillian Côte de la Surprise Member.
The Pabos Formation is a mainly terrigenous unit that is transitional between underlying siliciclastic rocks of the Whites Brook Formation and overlying calcareous rocks of the White Head Formation. It consists mainly of thin-bedded calcareous siltstone interbedded with lesser calcilutite and fine-grained calcareous sandstone. In places, the lower part of the Pabos Formation contains abundant calcareous to noncalcareous, parallel-, cross-, or convolute-laminated sandstone in 10–50 cm beds, intercalated with varying proportions of calcareous siltstone or mudstone. The sandstone beds resemble those in the Whites Brook Formation and are similarly interpreted as turbidite deposits. However, mudstones in the Pabos Formation are distinctly calcareous, whereas those in the Whites Brook Formation are not. Petrographically, the sandstones comprise poorly sorted, poorly rounded, grain-supported clasts of quartz, feldspar, calcite, abundant lithic fragments (felsic and mafic volcanic, felsic intrusive, and fine-grained sedimentary rocks), and accessory zircon. In the Restigouche area, the Pabos Formation reaches a maximum thickness of about 650 m along the Restigouche River west of the Sellarsville Fault (Fig. 2).
Nowlan (1983a) has reported conodonts of probable late Ashgillian (Gamachian) age from a section of interbedded turbiditic sandstone and calcareous siltstone on the Restigouche River east of the Sellarsville Fault (Fig. 2). These rocks were considered by St. Peter (1978a) to belong to the Grog Brook Group but were reassigned to the Pabos Formation by Wilson (2002) because of their calcareous nature. Graptolites collected from the same section have been identified as diagnostic of the Dicellograptus complanatus Zone of middle Ashgillian age (Riva and Malo 1988).
White Head Formation
The most widespread lithotype in the White Head Formation is a medium to dark grey, very fine grained calcilutite, regularly interbedded with calcareous shale; minor fine-grained calcarenite and noncalcareous shale or siltstone are also commonly reported (Pavlides 1968; Ayrton et al. 1969; Hamilton-Smith 1970; Roy and Mencher 1976; St. Peter 1978a; Stringer and Pickerill 1980; Lespérance et al. 1987; Pickerill et al. 1987; Malo 1988; Wilson 1990, 2000a). Sedimentary structures, bedforms, and trace fossils generally support deposition as turbid flows in a deep-water setting (Ayrton et al. 1969; St. Peter 1978a; Malo 1988), although (Stringer and Pickerill 1980) propose a shallower water, slope environment. In the Restigouche area, most of the White Head section consists of thin-bedded silty calcilutite with abundant laminae of calcareous siltstone and no interbedded calcareous shale. The terrigenous silt component is particularly abundant in the Chessers Brook Syncline, west of the Sellarsville Fault (Fig. 2). These rocks correlate with the Burmingham Member of the type area, as they immediately overlie the Pabos Formation and therefore constitute the oldest part of the White Head section. The upper part of the White Head Formation, equivalent to the L’Irlande and (or) Des Jean members, is well exposed east of the McKenzie Gulch Fault, where it consists of interbedded calcilutite and calcareous siltstone, passing upward to cryptically bedded calcareous shale or shaly calcilutite near the gradational contact with the overlying Upsalquitch Formation (see next section).
In northern New Brunswick the White Head Formation is conformably and gradationally overlain by the Upsalquitch Formation (Chaleurs Group) (St. Peter 1978a; Wilson 2000a). Near Squaw Cap Mountain (Fig. 2), however, a thinned White Head section is disconformably overlain by the Indian Point Formation (Chaleurs Group), and the Upsalquitch Formation is absent. In the southwestern part of the study area the White Head Formation is juxtaposed against the Boland Brook Formation along the McKenzie Gulch Fault (Fig. 2). The total thickness of the White Head Formation cannot be estimated for the Restigouche area, as the top of the unit is absent west of the McKenzie Gulch Fault, and the base is unexposed to the east of the fault. The thickness of the White Head Formation is estimated at 1200 m in the Chessers Brook Syncline (Wilson 2002) and 2400 m on the east limb of the Wheeler’s Gulch Anticline (Fig. 2). On the flanks of the Popelogan Inlier, however, the thickness of the White Head Formation is locally greatly reduced and ranges from 150 m on the east limb to about 1200 m on the west limb.
Brachiopods, trilobites, conodonts, and graptolites from a number of locations in Quebec, Maine, and New Brunswick indicate that the White Head Formation ranges from Ashgillian to Llandoverian (Lespérance et al. 1987; Malo 1988; Nowlan 1983b; Pavlides 1968; Rickards and Riva 1981; Hamilton-Smith 1970; St. Peter 1978a). Chitinozoan microfaunas have been obtained from two samples of the White Head Formation, just below the disconformable contact with the overlying Indian Point Formation near Squaw Cap Mountain (Fig. 2). These faunas belong to the H. crickmayi Zone, indicating a late Richmondian (middle Ashgillian) age and confirming that the upper (Llandoverian) part of the White Head Formation has been eroded in that area. At the Popelogan Inlier, limestone lenses in a thin unit of calcareous grit that overlies the Balmoral Group and underlies typical White Head strata contain diverse conodont faunas that have been assigned a latest Ashgillian age (Nowlan 1988; Wilson 2000a).
In the eastern part of the Chaleur Bay Synclinorium in New Brunswick (east of the study area), the Chaleurs Group has been divided into several formations, including two dominantly volcanic units (Walker and McCutcheon 1995). In the Restigouche area, the Chaleurs Group comprises upper and lower sequences of sedimentary rock, with an intervening (Salinic) disconformity (Fig. 3). The lower sequence consists of the Upsalquitch Formation, which conformably overlies the White Head Formation, and the much thinner, commonly eroded, Limestone Point Formation. The upper sequence comprises the West Point Formation and overlying and laterally equivalent rocks of the Indian Point Formation.
The Upsalquitch Formation typically consists of thin-bedded, bioturbated, slightly micaceous, calcareous siltstone and finegrained sandstone, with minor calcilutite and fine-grained calcarenite. Fine-grained calcareous sandstone occurs either as thin, irregular or discontinuous, high-angle cross-stratified beds and lenses intercalated with darker coloured siltstone, or as uniform, thin (3–10 cm), hummocky cross- or parallel-laminated beds (St. Peter 1978a; Lee and Noble 1977; Wilson 2000a). Sedimentary structures include slumps, sole markings, local graded bedding, and parallel-, current-ripple-, or convolute-laminated intervals of rhythmically alternating strongly and weakly calcareous laminae, indicating deposition on a slope as small-scale turbid flows (cf. Lee and Noble 1977). East of the Popelogan Inlier the Upsalquitch Formation is dominated by thin-bedded, micaceous, noncalcareous, finegrained, feldspathic to arkosic sandstone (Wilson 2000a). The Upsalquitch Formation gradationally overlies the White Head Formation and is either gradationally overlain by the Limestone Point Formation or disconformably overlain by the Indian Point or Val d’Amour formations. In the Charlo – Upsalquitch Forks area, 25–30 km southeast of Squaw Cap Mountain, the thickness of the Upsalquitch Formation has been estimated at 1500–1600 m (Lee and Noble 1977; Wilson 2000a). In the Restigouche study area, the thickness of a northwest-dipping Upsalquitch sequence between the White Head inlier at Blair Athol and the contact with the overlying Limestone Point Formation south of Val d’Amour (Fig. 2) is ∼ 3200 m.
Reported ages of fossil assemblages from the Upsalquitch Formation range from Llandoverian C3 to early Wenlockian (Lee and Noble 1977; St. Peter 1978a; Irrinki 1990; Wilson 2000a). In the Restigouche area, the association of Eisenackitina dolioliformis and Conochitina sp. 6 Asselin et al. (1989) in one sample is indicative of a Telychian (late Llandoverian) age (Fig. 3). This association is also present in the Early Silurian Anse Cascon Formation in the Gaspé Peninsula (Asselin et al. 1989). At one of the Restigouche locations, just south of St. Arthur (Fig. 2), the Upsalquitch Formation contains an impoverished suite of spores and acritarchs, including the taxa Laevolancis (undifferentiated) and Ambitisporites avitus. This suite of terrestrial spores within marine strata is consistent with terrestrial palynomorph assemblages that first appear in the Upper Member of the Ross Brook Formation in Nova Scotia (Beck and Strother 2001). The co-occurrence of chitinozoans and spores in this area clearly places these beds in the late Llandoverian (Telychian).
Limestone Point Formation
Near the Black Lake Fault (Fig. 2), the Upsalquitch Formation is conformably and gradationally overlain by the Limestone Point Formation, which is composed of thin- to medium-bedded, massive to parallel-laminated calcareous sandstone and minor highly fossiliferous limestone. The Limestone Point Formation is only sporadically exposed in the study area. For example, it is absent northwest of the Squaw Cap Fault, whereas southeast of the fault the unit is discontinuous and varies considerably in thickness (Fig. 2). The observed distribution may be attributed to Late Silurian uplift and erosion, creating an irregular paleosurface of promontories and depressions (Fig. 4a). At the type section in the Nigadoo River Syncline (Fig. 1) the Limestone Point Formation comprises calcareous sandstones that underlie the La Vieille Formation (Noble 1976). Walker and McCutcheon (1995), on the other hand, considered the Limestone Point Formation a member of the La Vieille Formation. Brachiopods and conodonts from the type area indicate a late Llandovery to early or middle Wenlock age for the Limestone Point Formation (Noble 1976; Lee and Noble 1977; Noble and Howells 1979; Nowlan 1983b). In the Restigouche area, chitinozoans from an outcrop near Val d’Amour support a Telychian (late Llandoverian) age, and conodont assemblages in several collections confirm a late Llandoverian – early Wenlockian range (Fig. 3). Conodont elements include some rare or uncommon species found in the Jupiter and Chicotte formations on Anticosti Island (Uyeno and Barnes 1983).
West Point Formation
The West Point Formation comprises white to pale grey, typically coral-rich limestone, thin-bedded, fossiliferous calcarenite and calcilutite, and thin- to thick-bedded, light grey, fossiliferous, moderately calcareous, fine- to coarsegrained sandstone. The West Point Formation lies on the Salinic unconformity and has an uneven distribution, probably because of the sporadic occurrence of limestone as pinnacle reefs (Fig. 5). It is mainly exposed between the Sellarsville and Sellarsville East faults, although scattered outcrops are found farther east. In most cases, West Point rocks occur in the core of anticlines (Fig. 2). Southwest of Flatlands (Fig. 2), the West Point is represented by a sequence of coral-rich mudstone, mudstone containing abundant limestone clasts, limestone breccia, and pale grey, massive limestone. Massive limestone, and fragments in the limestone breccia, locally exhibits concentric laminar structures that suggest an origin as algal bioherms. The breccia consists of angular to subrounded fragments of very fine grained limestone in a matrix of dark grey calcareous mudstone or fine-grained calcarenite containing abundant fossil debris. The mudstone– limestone association resembles descriptions of the Anse à la Loutre member of the Indian Point Formation in the Gaspé Peninsula (Bourque et al. 1986). The Anse à la Loutre member is equivalent to the Anse Beebe facies (West Point Formation) of Bourque and Lachambre (1980); it is interpreted as a basinal facies deposited adjacent to (platformal) algal reefs of the Anse à la Barbe member (West Point Formation) and grades from proximal limestone-rich reef talus to distal mudstone (Bourque et al. 1986).
Brachiopods and corals identified in previous collections of fossils in the Flatlands – Glen Levit area indicate a Late Silurian age for the West Point Formation (Wilson 2002 and references therein). An isolated outcrop of limestone just above the Salinic unconformity 3 km south of Glencoe (Fig. 2) has yielded the conodont Ozarkodina remscheidensis eosteinhornensis, indicating a late Ludlovian to Pridolian age, whereas at Glen Levit the West Point Formation contains the early Pridolian to Lochkovian conodont Ozarkodina remscheidensis remscheidensis.
Indian Point Formation
The Indian Point Formation is a lithologically diverse unit consisting mainly of calcareous mudstone and fine-grained sandstone gradational to calcilutite and calcarenite, respectively. Minor lithotypes include biostromal limestone, limestone conglomerate, coarse-grained sandstone, polymictic conglomerate, and mafic volcanic rocks. The Indian Point conformably overlies the West Point Formation between the Sellarsville and Sellarsville East faults, but elsewhere it typically disconformably overlies either the White Head, Upsalquitch, or Limestone Point formations, depending on the depth of Salinic erosion. The Indian Point Formation can be divided into four facies associations that are, in general, restricted to specific stratigraphic intervals within the formation (Fig. 5). From oldest to youngest, these include a sandstone– conglomerate facies (a), a thin-bedded calcareous siltstone – sandstone facies (b), a medium- to thick-bedded calcareous mudstone facies (c), and a calcarenite–conglomerate–limestone facies (d).
The sandstone–conglomerate facies (a) is found only southeast of the Squaw Cap Fault, where it locally forms the base of the Indian Point Formation and disconformably overlies the Limestone Point or Upsalquitch formations. It is composed of (i) medium- to thick-bedded, light grey, massive to parallel-laminated, variably calcareous, fine- to coarsegrained quartzose sandstone; (ii) light grey to pinkish grey polymictic conglomerate containing clasts of limestone, finegrained sandstone, fossils (mainly corals), and mafic and felsic volcanic rock; and (iii) minor thin-bedded grey calcareous mudstone. Felsic volcanic clasts in the polymictic conglomerate include porphyritic, devitrified rhyolite and welded vitric-crystal tuff resembling subaerial volcanic rocks in the Wenlockian–Ludlovian (423 ± 3 Ma; Walker et al. 1993) Benjamin Formation, which crops out just to the east of the Restigouche study area. Thick-bedded sandstone typically contains thin fossiliferous bands (brachiopods, corals) and in places hosts carbonized plant matter. The sporadic exposure of facies (a) is likely related to its deposition in channels on the Late Silurian erosional paleosurface.
The most common lithological association in the Indian Point Formation is a thin-bedded siltstone facies (b), which consists of 1–6 cm beds of light to medium grey, moderately to strongly calcareous, locally fossiliferous siltstone and minor fine-grained, calcareous, commonly parallel- or cross-laminated sandstone in beds up to 40 cm. A second fine-grained association, the calcareous mudstone facies (c), is found only northwest of the Squaw Cap Fault. It comprises medium to dark grey or greenish grey, medium- to thick-bedded (up to roughly 1 m), locally fossiliferous, moderately to strongly calcareous mudstone and local thin beds of light grey calcareous siltstone and fine-grained, parallel-laminated sandstone. In places, alternating layers of dark grey mudstone and light grey, fine-grained sandstone form sequences of thin-bedded turbidites.
East of the Sellarsville East Fault, fine-grained facies associations (b) and (c) are overlain by the coarser grained, calcarenite–conglomerate–limestone facies (d) (Fig. 5). The latter comprises light grey, medium- to thick-bedded (up to ∼ 1 m), locally parallel-laminated calcarenite (gradational to calcareous sandstone), limestone conglomerate, biostromal limestone, and minor light grey argillaceous calcilutite and noncalcareous fine-grained sandstone. Biostromal limestones occur in beds from 20 cm to >1 m and are typically light grey to light pinkish grey bioclastic wackestones. Two or more types of fossils are commonly present, including brachiopods, bryozoans, corals, trilobites, crinoids, and ostracodes, set in a micritic matrix containing minor quartz silt and clay. Limestone conglomerate is present only near Glen Levit, between the Sellarsville East and Squaw Cap faults (Fig. 2), where it is interbedded with fine-grained, light grey, fossiliferous calcarenite in a sequence at least 150 m thick. The conglomerate is monomictic, consisting of very well rounded, unfoliated, clast-supported pebbles and cobbles of pale grey calcilutite in a fine-grained calcareous matrix containing some fossil debris. The calcilutite cobbles are unlike the limestone clasts present in facies (a). Late Ordovician (Gamachian) conodont elements recovered from the cobbles confirm a source in the White Head Formation, as proposed by Alcock (1935). Alcock (p. 51) interprets the conglomerate as a basal Devonian conglomerate between Silurian “limestones” to the west and Lower Devonian sedimentary and volcanic rocks to the east and equates it with limestone- and volcanic-cobble conglomerate directly overlying the Silurian (Salinic) disconformity in the southern Gaspé Peninsula. The Salinic disconformity lies well below the limestone conglomerate at Glen Levit, however, and the basal (post-Salinic) conglomerate in the Gaspé Peninsula should instead be correlated with conglomerates present in facies (a). The limestone conglomerate and associated carbonate rocks of facies (d) are interpreted here as part of a shallowing-upward (regressive) sequence, transitional between deeper water Indian Point facies (b) and (c) and overlying, largely subaerial volcanic rocks of the Val d’Amour Formation (see next section).
Spores in the Indian Point Formation are mainly Late Silurian – Early Devonian forms, with a Lochkovian age most likely. Emphanisporites epicautus, Emphanisporites micrornatus, Chelinohilates lornensis, and Chelinohilates erraticus (?) show affinity with the early Lochkovian E. micrornatus –Streelispora newportensis Assemblage Zone of Richardson and McGregor (1986). Acritarch taxa pointing to an early Lochkovian age include Ammonidium cornuatum, Polyedryxium pharaonis, Cymatiosphaera cornifera (?), Ozotobrachion pulvinus, and Ozotobrachion paliodigitalis. The latter species, along with the spore taxon E. micrornatus, have also been identified in the Jacquet River Formation (lower part of Dalhousie Group; Fig. 3) to the east of the Restigouche study area (E.T. Burden, unpublished data, 2000). The Indian Point Formation is intruded by the Squaw Cap Felsite, which has yielded a middle Lochkovian U–Pb (zircon) age of 415.0 ± 0.5 Ma (V. McNicoll, written communication, 2001) (Fig. 2) and provides an upper age limit for the unit.
The thickness of the Indian Point Formation is estimated at 500 m between the Sellarsville and Sellarsville East faults and up to 1100 m southeast of the Squaw Cap Fault (Wilson 2002). The Indian Point Formation is truncated by the Black Lake Fault in the Val d’Amour area (Fig. 2) and has not been observed farther to the east, where the Upsalquitch Formation is disconformably overlain by the Dalhousie Group (Fig. 2). The abrupt thinning of the Indian Point Formation between Val d’Amour and Dalhousie suggests that a disconformity may occur at the base of the Dalhousie Group; however, similar spore assemblages in the Indian Point Formation and lower part of the Dalhousie Group (i.e., micrornatus–newportensis Assemblage Zone; see next section) support a conformable relationship. Furthermore, a disconformity has not been reported at the base of the Dalhousie volcanic sequence in the southern Gaspé Peninsula (e.g., Bourque and Lachambre 1980). Instead, the absence of the Indian Point Formation east of Val d’Amour is more likely a result of nondeposition (see Structural geology section).
Research within the Dalhousie Group has a long history beginning with Clarke (1909), who provided a detailed account of the stratigraphy and paleontology of the “Dalhousie Formation” at Dalhousie, establishing it as one of the two best representative sections of the Lower Devonian in North America. Dalhousie Group stratigraphy was first defined by Greiner (1967) and revised by Walker and McCutcheon (1995), who recognized five formations in the Charlo – Jacquet River area (Figs. 1, 3). In the Restigouche area, the Dalhousie Group is represented only by volcanic and minor sedimentary rocks of the Val d’Amour Formation (new name).
Val d’Amour Formation
The Val d’Amour Formation comprises a complexly interbedded sequence of mafic, intermediate, and felsic effusive and pyroclastic rocks, fine-grained to very coarse grained volcaniclastic rocks, locally interbedded fine-grained sedimentary rocks, and subvolcanic plugs and domes. The abundance of intermediate volcanic rocks distinguishes the Val d’Amour Formation from the bimodal suites that dominate Siluro-Devonian volcanic belts elsewhere in the northern Appalachians (e.g., Keppie and Dostal 1994; Dostal et al. 1989; Hon et al. 1992; Seaman et al. 1999). On the other hand, bimodal suites are relatively uncommon in the Gaspé Peninsula (Laurent and Bélanger 1984; Dostal et al. 1993). The Val d’Amour Formation forms a thick monoclinal sequence that dips consistently to the north. Average dips, based on measurements on interbedded sedimentary rocks, bedded tuffs, and volcanic flow tops, increase from 40° at the base to 70° near the top, allowing total thickness to be estimated at about 6100 m.
From bottom to top, the Val d’Amour Formation records a general transition from mafic to intermediate to felsic compositions. The lower part of the unit consists mainly of basaltic to andesitic, massive to amygdaloidal, locally scoriaceous flows, and thin-bedded to very thick bedded mafic ash and lapilli tuffs. South of Campbellton, in the Val d’Amour area (Fig. 2), the upper part of the mafic sequence is succeeded by massive to flow-layered, dacitic to andesitic effusive rocks and coarse-grained intermediate lithic tuffs and tuff– breccias. Minor volcaniclastic conglomerate and felsic lithic tuff and tuff–breccia are associated with the intermediate flows, which constitute the dominant Val d’Amour lithotype in the eastern part of the study area, especially near Dalhousie. At Campbellton, the upper part of the Val d’Amour Formation consists mainly of pink to maroon flow-layered rhyolite–rhyodacite, with minor basalt and volcaniclastic rocks. A concordant U–Pb (zircon) age of 407.4 ± 0.8 Ma (early Emsian) was calculated from a sample of this rhyolite (S. Kamo, written communication, 2000) (Fig. 2). Most mafic to intermediate flows are massive, with amygdaloidal to scoriaceous horizons marking the location of flow tops; pillow basalts and related hyaloclastites typical of subaqueous emplacement have been observed only in one mafic flow unit on the coast at Dalhousie. Thin- to thick-bedded mafic ash and lapilli tuffs observed at Val d’Amour and Dalhousie resemble the products of subaerial to shallow-water maar or tuff ring eruptions (cf. Fisher and Schmincke 1984).
Sedimentary rocks locally interbedded with the volcanic rocks include calcareous mudstones and fine-grained sandstones and volcaniclastic rocks ranging from medium- to coarse-grained, arkosic lithic sandstones to very thick bedded volcanic boulder conglomerates of probable debris-flow origin. Shallow-water mudstones and sandstones generally form thin intercalations a few metres thick but range to about 40 m. One such section near Sugar Loaf Mountain (Fig. 2) includes a thin bed of coal, the presence of which is consistent with subaerial emplacement of the volcanic rocks. More than 50 species of terrestrial spores have been identified in sedimentary strata intercalated with the volcanic rocks. In general, all productive strata are thought to be freshwater in origin, again confirming a subaerial to shallow-water environment. The very few marine acritarchs identified in one sample from near the bottom of the formation are thought to be reworked from older Silurian strata.
Palynomorphs recovered from a sample collected near the base of the formation are very dark in colour, indicating that they are thermally mature. Taxa include relatively small specimens of Apiculiretusispora sp., Dictyotriletes minor, and Emphanisporites sp. Larger taxa include specimens of Chelinospora opegyrus? and a variety of hilate and inaperturate taxa, such as Cymbohilates magnus and C. lornensis. The age of these beds is not well constrained beyond it being a part of the Lochkovian micrornatus–newportensis Assemblage Zone (Richardson and McGregor 1986). Strata from the Midland Valley of Scotland containing a broadly similar assemblage of fossils are considered late early to early late Lochkovian (Wellman 1993). Spores from beds stratigraphically higher in the formation, one of which lies just below the dated rhyolite, are well preserved and relatively light in colour (thermally immature). Taxa include Dibolisporites quebecensis, Dictyotriletes emsienensis, Dictyotriletes favosus, Emphanisporites rotatus var. robustus, Retusotriletes maculatus, Retusotriletes ocellatus, and Verrucosisporites polygonalis. These strata are considered to lie within the upper Pragian through earliest Emsian polygonalis–emsiensis Assemblage Zone (Richardson and McGregor 1986; Streel et al. 1987), consistent with the early Emsian age of the overlying rhyolite.
Clear differences exist in fossil preservation and assemblage zone composition for samples from the bottom and top of the Val d’Amour Formation. The sample from the lower part of the Val d’Amour Formation has much in common with those collected from the top of the Indian Point Formation and with samples from the Dalhousie Group in the Jacquet River area (E.T. Burden, unpublished data, 2000). In contrast, samples from the top of the Val d’Amour Formation have no known counterparts in New Brunswick. These assemblages do compare well with illustrations provided for taxa from the Pragian Shiphead Formation of the northeastern Gaspé Peninsula (McGregor and Owens 1966) and the Stooping River Formation of northern Ontario (McGregor and Camfield 1976). The absence of floras of the intervening breconensis– zavallatus Assemblage Zone (Richardson and McGregor 1986; Streel et al. 1987) may simply reflect nonexposure of sedimentary rocks in the volcanic-dominated middle part of the Val d’Amour Formation; however, the possibility of a late Lochkovian or early Pragian hiatus in sedimentation or volcanism cannot be discounted.
The Campbellton Formation is a coarsening-upward alluvial– lacustrine sequence comprising calcareous to noncalcareous, fine- to medium-grained sandstone and siltstone locally containing abundant plant fossils, grading upward to very coarse grained arkosic sandstone and pebble to cobble conglomerate (see also Rust et al. 1989; Gamba 1990). Conglomerates include medium- to thick-bedded polymictic pebble conglomerate and very thick bedded volcanic cobble–boulder conglomerate. Minor lithotypes include red siltstone, coal, and carbonaceous mudstone.
The lower and upper parts of the Campbellton Formation are equivalent to the Lagarde and Pirate Cove formations (Dineley and Williams 1968), respectively, in southern Gaspé Peninsula (Fig. 3). A spore-based late Emsian age has been reported for the Lagarde Formation (Bourque et al. 1995; Malo and Bourque 1993) and the lower part of the Campbellton Formation (Gamba 1990), although Andrews et al. (1975) preferred an early Middle Devonian (Eifelian) age. The contact with underlying, early Emsian volcanic rocks of the Val d’Amour Formation is not exposed. Dineley and Williams (1968) state, however, that the Lagarde Formation is unconformable on the Restigouche Volcanics (equivalent to Val d’Amour Formation) in southern Gaspé. The average (northerly) dip of Campbellton strata ranges from 16° to 45°, significantly less than dips observed in the Val d’Amour Formation, supporting an angular discordance between the two units. The thickness of the Campbellton Formation is estimated at 500 m at Point La Nim (Fig. 2). West of Point La Nim the Campbellton Formation is unconformably overlain by nearly flat-lying, coarse-grained terrestrial redbeds of the Carboniferous Bonaventure Formation.
Folds and cleavage
In contrast to polyphase deformation and intense greenschist facies metamorphism south of the Rocky Brook – Millstream Fault (i.e., in the Miramichi Highlands; Fig. 1), most rocks in the Restigouche area, including those in the Popelogan Inlier (where zeolites are preserved in mafic volcanic rocks), are relatively weakly deformed and have experienced very low grade regional metamorphism (Wilson 2000a, 2003). Folds typically strike northeast, are slightly asymmetrical (axial planes inclined steeply to the east), and plunge gently or moderately to the northeast or southwest (Wilson 2000a, 2002). All units were affected by only one period of compressive deformation (the Acadian Orogeny), so variations in fold styles are attributed to rheological differences and different depths of burial. For example, mudstones, siltstones, and calcilutites in the Grog Brook, Matapédia, and lower Chaleurs groups are typically thrown into close folds whose limbs dip moderately to steeply northwest and southeast, whereas carbonates and sandstones in the upper Chaleurs Group exhibit open folds with shallow to moderate dips ranging from west to north to east. The effects of Late Silurian (Salinic) deformation are more subtle. In the eastern Gaspé Peninsula, local northwest-trending folds without axial planar cleavage, synsedimentary block faulting, and within-plate volcanism are attributed to Salinic extensional tectonics (Malo and Kirkwood 1995; Malo 2001; Bourque 2001). In the Restigouche area, indirect evidence exists for weak deformation predating the main period of (Acadian) folding. For example, early folding around northwest-trending axes may be responsible for doubly plunging Acadian folds, of which the Popelogan Anticline is the best example. In the Kedgwick area 30 km southwest of Squaw Cap Mountain, the distribution of Grog Brook and Matapédia rocks in the Aroostook–Percé Anticlinorium provides clearer evidence of overprinting of early, northwest-trending macroscale folds by Acadian folds with penetrative, northeast-trending, axial planar cleavage (Carroll 2003). However, there is no evidence in northern New Brunswick of Late Silurian sinistral faults or shear zones such as those documented in the Millimagassett Lake area of north-central Maine (the “early Acadian” deformation of Hibbard 1994).
The intensity of cleavage development varies considerably. In the Grog Brook and Matapédia groups and Upsalquitch Formation, a moderately well developed penetrative cleavage is the norm, except between the Sellarsville and McKenzie Gulch faults (Fig. 2). Here, on the limbs of the Coxs Brook Anticline (Fig. 2), cleavage in Grog Brook and Matapédia rocks is poorly developed. The Coxs Brook Anticline can be traced in younger rocks north of the Squaw Cap Fault, where only a local fracture cleavage is observed in the West Point and Indian Point formations. This is in sharp contrast with the penetrative fabric developed in the Matapédia Group west of the Sellarsville Fault (i.e., in the Chessers Brook Syncline; Fig. 2). Very low grade metamorphism and near absence of cleavage also characterize rocks in the core of, and east of, the Popelogan Anticline (Wilson 2000a, 2003). Thus, the Restigouche area encompasses four domains of alternately weak and moderate deformation: west of the Sellarsville Fault (moderate); Coxs Brook Anticline – Restigouche Syncline (moderate to absent); between the McKenzie Gulch – Black Lake faults and Popelogan Inlier (moderate); and within and east of the Popelogan Inlier (very weak). In much of the Restigouche study area, therefore, crustal shortening is accommodated primarily by folding rather than by cleavage development. No estimate of the total amount of shortening has as yet been attempted.
The Restigouche area is located between two major dextral transcurrent faults, the Restigouche – Grand Pabos Fault to the north and the Rocky Brook – Millstream Fault to the southeast (Fig. 1), and is transected by several faults that affect all units except the Bonaventure and Campbellton formations. The Restigouche area lies along a north–south-trending “hinge” zone where the orientation of the Rocky Brook – Millstream and Restigouche – Grand Pabos faults, as well as the Catamaran Brook Fault farther south (Fig. 1), changes from dominantly north (mainly vertical displacement) to dominantly east (mainly lateral displacement). Very complex architectures involving differential movement of fault blocks can result at such flexures or restraining bends (e.g., Aksu et al. 2000), possibly expressed as positive or negative flower structures depending on whether the local tectonic regime is transtensional or transpressional. In the Gaspé Peninsula, some faults show a complex history of dextral strike-slip, thrust, and normal movements because of repeated reactivation in response to varying stress regimes (Lavoie 1992), and it seems likely that the same has occurred in the Restigouche area. Nevertheless, considerable uncertainty is involved in interpreting the history of movement along most faults, with the information presently available.
An exception to this uncertainty is the Sellarsville Fault, which is well exposed on the south shore of Restigouche River, near the confluence with Upsalquitch River (Fig. 2). The Sellarsville Fault is an east-verging reverse fault that dips about 45° to the west and juxtaposes the Whites Brook Formation (hanging wall) against the Pabos Formation (footwall) (Figs. 2, 4b). In the Gaspé Peninsula, it has been interpreted as a post-Middle Devonian structure synchronous with initiation of basement strike-slip faulting (Malo and Bourque 1993; Malo and Kirkwood 1995). The Sellarsville East Fault, a splay of the Sellarsville Fault, has experienced similar reverse displacement and forms a tectonic contact between the Indian Point and Val d’Amour formations at Flatlands (Fig. 2). It therefore postdates at least the oldest (Lochkovian) rocks of the Val d’Amour Formation. Clearly, the latest motion on the Sellarsville Fault also postdated development of Acadian folds, as the Chessers Brook Syncline is truncated by the fault (Fig. 2).
The McKenzie Gulch Fault, which terminates just southeast of Squaw Cap Mountain (Fig. 2), is an important regional structure that continues to the southwest at least as far as the New Brunswick – Maine border (Fig. 1). South of Squaw Cap Mountain, the McKenzie Gulch Fault juxtaposes Grog Brook rocks to the west against Matapédia rocks to the east, indicating significant vertical offset (Fig. 2). A slight discordance between the orientations of the fault and those of major fold axes (Fig. 2) suggests that the timing and sense of the most recent displacement are the same as those of the Sellarsville Fault, i.e., it may be a Middle Devonian reverse fault (Fig. 4b). Indirect evidence suggests, however, that initial movement on the McKenzie Gulch, as well as some other faults in the area, occurred during the Late Silurian. For example, during the Late Silurian to Early Devonian, the “Squaw Cap block,” as we refer to the area between the Sellarsville and McKenzie Gulch – Black Lake faults (Figs. 4a, 4b), occupied an elevated (subaerial) position with respect to adjacent rocks to the southeast. This is based on deep erosion of pre-Salinic strata (Fig. 4a) and high-level intrusion of the Matapédia Group by the Squaw Cap Felsite (∼415 Ma) and a similar intrusion 3 km northwest of Squaw Cap Mountain (Fig. 2). Furthermore, the Indian Point Formation roughly doubles in thickness on the southeast side of the Squaw Cap Fault, implying that the latter was a synsedimentary (growth) fault in the Late Silurian to Early Devonian. Movement on the Squaw Cap and McKenzie Gulch – Black Lake faults was therefore coeval with extensional, synsedimentary Salinic faulting in the eastern Gaspé Peninsula (Malo and Kirkwood 1995; Malo 2001; Bourque 2001). A half-graben structure developed during Salinic extensional tectonism (Fig. 4a), like those described by Bourque (2001) for the eastern Gaspé Peninsula, may explain preferential development of West Point reefs adjacent to the Sellarsville Fault (Fig. 2) and easterly thickening of the Indian Point Formation. It could also account for nondeposition of the Indian Point Formation beyond an inferred graben-bounding fault east of Val d’Amour.
Whereas the Sellarsville and McKenzie Gulch faults were reactivated as Acadian reverse faults, other faults were reactivated as late Acadian strike-slip faults (Fig. 4b). The Sellarsville Fault and Coxs Brook Anticline are offset dextrally by the Squaw Cap Fault, and dextral offsets of the Coxs Brook Anticline and the Limestone Point Formation occur along the Black Lake Fault (Fig. 2). Vertical displacement is evident on the Indian Lake and Arsenault Brook faults (Fig. 2), although the timing of these offsets (Salinic or Acadian), and whether they were normal or reverse, is unclear. The chronological sequence implied by the aforementioned relationships consists of Late Silurian (extensional) normal or block faulting, followed by late Early to Middle Devonian folding, reverse faulting, and Middle Devonian dextral strike-slip motion contemporaneous with the Restigouche – Grand Pabos and Rocky Brook – Millstream faults (Figs. 4a, 4b). This sequence of events agrees well with the faulting history described for the Gaspé Peninsula (e.g., Malo 2001).
Sixty-four analyses of Late Ordovician to Early Devonian sedimentary rocks were carried out to establish levels of thermal maturation in the Restigouche area, including 36 vitrinite reflectance (Ro) analyses, 13 colour alteration index (CAI) analyses from conodonts, and 16 thermal alteration index (TAI) analyses from spores. Two of the CAI values (from the Pabos Formation) were previously reported by Nowlan (1983a). Vitrinite reflectance is estimated from the reflectance of zooclasts (graptolites, chitinozoans, and scolecodonts), various macerals of the vitrinite group, and solid bitumen fragments according to the method described in Bertrand and Malo (2001). Colour alteration and thermal alteration values were normalized to equivalent Ro, using the correlation charts of Williams et al. (1998) and Bustin et al. (1985), to facilitate contouring of the data (Fig. 6). Most of the reflectance values in the Squaw Cap block (between the Sellarsville and McKenzie Gulch – Black Lake faults) fall in the oil window (Ro up to 1.4) or condensate zone (Ro = 1.4–2.0; Williams et al. 1998; Bustin et al. 1985). Surprisingly, this includes several samples from the Boland Brook, Pabos, and White Head formations along or near Upsalquitch River south of Squaw Cap Mountain (Figs. 2, 6). Specifically, Ro in the Squaw Cap block ranges from 0.7% to 1.5% in Indian Point and West Point rocks, 0.6% (just below the Salinic disconformity) to 1.1% in the White Head Formation, and 1.1% to 2.5% in the Boland Brook Formation, Ro is 1.3% in the Pabos Formation and 2.5% in one sample of Whites Brook shale near Squaw Cap Mountain. Thermal alteration studies carried out on spores from a suite of Indian Point samples have revealed that most have a TAI of 2.9–3.1 (gas condensate to dry gas), although one sample, collected 3 km west of Glen Levit, has a TAI of <2.3 with vivid yellow fluorescence; this rock lies well within the oil window. Colour alteration indices of conodonts recovered from the Limestone Point, Indian Point, and Val d’Amour formations range from 1.5 to 2.0. Thermal maturity values therefore suggest potential for gas-productive strata in the Squaw Cap block.
A small number of samples from the Val d’Amour and Campbellton formations were examined for spore TAI and fluorescence. Spores from two Campbellton Formation samples are black and pitted, indicating oxidation or thermal overprinting and carbonization. Val d’Amour samples show mixed thermal maturation characteristics. Spores in a sample from near the base of the formation are dark brown in colour (TAI = 3.1) and indicate that the strata lie in the gas condensate to dry gas zone. Overlying strata from the upper part of the Val d’Amour Formation contain relatively light coloured spores (TAI < 2.7) that fluoresce with yellow through dull amber colours. These rocks still lie within the oil window. A reflectance value of 0.4% in a second sample from the upper Val d’Amour Formation confirms a low thermal maturity.
Data are sparse from the areas west of the Sellarsville Fault and east of the McKenzie Gulch – Black Lake Fault. Nevertheless, it is clear that these areas are characterized by higher thermal maturities (Fig. 6). Reflectance, CAI, and TAI values west of the Sellarsville Fault are 4.2%–4.5%, 4.0–5.0 (Nowlan 1983a; Nowlan and Barnes 1987), and 3.1, respectively. East of the Squaw Cap block (but close to the bounding faults) Ro values are 1.7%–1.9% in the Limestone Point Formation, 2.0% in the Upsalquitch Formation, and 1.5%–2.7% in the White Head Formation. Together, these data argue for shallow burial depth and prolonged uplift of the Squaw Cap block. In contrast, a considerable thickness of Late Silurian – Early Devonian sediments probably accumulated in the area west of the Sellarsville Fault, which presumably remained deeply buried until exhumation following Acadian reverse faulting.
Reflectance, TAI, and CAI values of 0.7%, 2.3, and 1.5–2.0, respectively, in the West Point and Indian Point formations correspond to paleotemperatures of about 80–100 °C (e.g., Williams et al. 1998). Assuming a geothermal gradient of 25–30 °C/km, a burial depth of 2.5–3.5 km for the Indian Point Formation is most likely. The thickness of original Carboniferous to Permian sediment cover removed by erosion, based on apatite fission-track analysis, is estimated by Ryan et al. (1991) to be up to 3 km and by Ryan and Zentilli (1993) to be up to 4 km, which compares well with burial depths based on thermal maturity.
Sedimentary history and paleogeography
Early Ashgillian to late Llandoverian sedimentary rocks in the Restigouche area are subdivided into a lower siliciclastic sequence (Grog Brook Group), a middle carbonate sequence (Matapédia Group), and an upper siliciclastic sequence (Upsalquitch Formation). Together, these units define a shallowing-upward trend reflecting gradual basin infilling and eustatic sea-level fall (the R1 regressive phase of Malo and Bourque 1993; Bourque et al. 2000; Bourque 2001). The regressive sequence culminated in extensive carbonate platform development, represented in the Restigouche area by shelf sandstones and limestones of the Limestone Point Formation. The total thickness of late Caradocian to late Llandoverian deep-water sediments ranged from 8 to 12 km, but probably locally exceeded this figure (e.g., St. Peter 1978a; Malo 1988; Wilson 2002).
Sedimentary structures and bedforms (Bouma sequences, sole markings, etc.), trace fossil assemblages, and general paucity of body fossils in the Grog Brook and Matapédia groups are consistent with deep-water turbidite facies (e.g., St. Peter 1978a; Pickerill 1980; Malo 1988; Riva and Malo 1988). The Whites Brook Formation has been interpreted to record deposition by turbidity currents in a major submarine channel system, compared to a distal (or overbank) submarine fan environment for most of the Boland Brook Formation (Wilson 2002). The Grog Brook Group and correlative units in northeastern Maine and the Gaspé Peninsula contain abundant feldspathic sandstone, lithic wacke, and volcanic detritus (Roy and Mencher 1976; Roy 1980; St. Peter 1978a; Hamilton-Smith 1971; Riva and Malo 1988; Wilson 1990, 2002), implying a provenance, at least in part, from emergent areas of the Popelogan Arc (Fig. 7a). The age of most source material must postdate Taconic deformation because virtually all lithic clasts in Grog Brook conglomerates are unfoliated. The bulk of the clastic material is quartzose silt and fine sand derived from the north or northeast, based on paleocurrent measurements (St. Peter 1978a). A great thickness of Grog Brook sedimentary rocks was deposited in the Aroostook–Percé Anticlinorium (St. Peter 1978a; Wilson 2002), but in the eastern part of the Chaleur Bay Synclinorium, overlying the Balmoral Group, the Grog Brook is either absent (on the flanks of the Popelogan Anticline) or relatively thin (e.g., the “Pat Brook beds” of Carroll 2000) (Fig. 4a). Instead, the oldest Gaspé Belt rocks at the Popelogan Inlier are latest Ashgillian calcareous grits (Wilson 2000a), demonstrating that Dunnage basement in that area was emergent during most of the Late Ordovician while deep-water sedimentation occurred farther west (Fig. 7a).
Deposition of mainly fine-grained calcareous rocks of the Matapédia Group reflects a change from mixed, distal(?) siliciclastic and proximal volcanic sources to a distal carbonate source. Several analyses of sedimentary structures and lithofacies (e.g., Ayrton et al. 1969; St. Peter 1978a; Malo 1988) support deposition as turbidites in a deep basinal environment. In contrast, Stringer and Pickerill (1980) present a number of observations that argue for deposition as contourites and hemipelagic deposits with periodic introduction of turbidites in a slope or continental rise environment. This proposal, suggesting somewhat shallower water deposition, is consistent with gradual infilling of the Gaspé deep-water basin during Ashgillian and Llandoverian time.
The provenance of the Matapédia lime mudstones, based on facies zonation, paleocurrent analysis, and similarity in conodont fauna, is generally considered to be located in contemporaneous carbonates of the peri-Laurentian Anticosti platform (Nowlan 1981; Lespérance et al. 1987; Malo 1988). The fossiliferous nature of the White Head Formation in the eastern Gaspé Peninsula (Lespérance et al. 1987; Malo 1988) supports a transition from relatively shallow water in the direction of the presumed source area to deeper water farther southwest. Hence, a northeastern provenance is indicated for both the Grog Brook and Matapédia groups. Specifically, their probable lithological counterpart on the Anticosti Platform comprises mixed carbonate and siliciclastic rocks of the Ashgillian Vauréal Formation (Petryk 1981). As discussed in a previous section, a correlation between the Boland Brook and Vauréal formations is demonstrated by chitinozoan micro-faunas contained in the two units. Furthermore, chitinozoans in the Upsalquitch Formation, and conodonts in the Limestone Point Formation, are also found in the Jupiter and Chicotte formations on Anticosti Island, suggesting continued input from that area.
Turbidite sedimentation continued during deposition of the Upsalquitch Formation (Chaleurs Group) (Lee and Noble 1977; St. Peter 1978a; Kim et al. 2000; Wilson 2002). Bioturbation, trace fossils (Kim et al. 2000), and shelly fossils are all locally abundant, especially in comparison to the underlying White Head Formation, and are taken to reflect deposition in shallower, more oxygenated water. The dominantly siliciclastic character of the Upsalquitch Formation signals a change in provenance compared to the White Head Formation; specifically, chemical signatures demonstrate an increasing input from a more mature, continental source (Wilson 2000a).
In the Limestone Point Formation, basal shell-lag deposits in some limestone beds, and parallel-laminated intervals overlain by hummocky cross- or convolute-stratification at upper bed boundaries, imply that it represents a storm-dominated mid- to outer-shelf sequence. Local reddish colouration in some calcareous sandstones and siltstones at the top of the unit probably resulted from subaerial oxidation (groundwater percolation?) following Late Silurian uplift; the red colour is evidently not primary because the thickness of affected sedimentary rocks varies greatly over even short distances along strike. The Squaw Cap block was deeply dissected by erosion, which penetrated well into the White Head Formation, removing the upper part of that unit, and all of the Chaleurs Group. An irregular, Late Silurian erosional paleosurface in a block-faulted landscape explains local pinch-outs of the Upsalquitch and Limestone Point formations. Similarly, conglomerates and coarse-grained sandstones in the lower part of the Indian Point Formation were probably deposited in paleotopographic depressions during post-Salinic sea-level transgression (Fig. 5).
Above the Salinic disconformity, a transgressive episode is indicated by the transition from West Point reef facies to deeper water facies (b) and (c) of the Indian Point Formation, corresponding to the T2 transgressive phase of Malo and Bourque (1993). Lithological associations and bedforms in Indian Point facies (b) and (c) indicate deposition in a marine shelf environment below wave base, and palynological evidence points to a fully marine setting (i.e., a predominance of marine taxa), with varying degrees of input from terrestrial sources. However, local beds of conglomerate and limestone, abundance of fossils, local plant debris, and scarcity of turbidite bedforms argue against a deep-water setting. Marine sedimentation was relatively short-lived, as a Lochkovian–Emsian regression began with Indian Point facies (d) and continued with subaerial volcanism of the Val d’Amour Formation. This event correlates with, but appears to predate, the Emsian–Eifelian regression documented in the Gaspé Peninsula (Malo 2001; Bourque 2001). A subaerial to shallow-water environment in the Restigouche area is supported by terrestrial spores and sedimentary and volcanic facies of the Val d’Amour Formation. Volcanism spanned the interval from ca. 415 Ma (i.e., the age of the Squaw Cap Felsite) to ca. 407 Ma, the age obtained from rhyolite near the top of the Val d’Amour Formation. Differences in spore assemblages associated with the bottom and top of the Val d’Amour Formation (inception and waning of volcanic activity, respectively) may be linked to the pre- and post-volcanic landscapes, i.e., older spores represent a flora that predates volcanism, whereas by the early Emsian a subaerial edifice had been constructed and volcanism had abated sufficiently to allow a vegetated landscape to develop. Above an Emsian hiatus, the coarsening-upward clastics of the Campbellton Formation reflect terrestrial deposition in lacustrine and alluvial fan environments (Rust et al. 1989; Dineley and Williams 1968), on what was presumably a dissected, mountainous volcanic landscape.
From the Lochkovian to Emsian, sedimentation in the Gaspé Belt occurred within a broad tract that encompassed the Upper Gaspé Limestones and the Fortin Group in the Gaspé Peninsula (carbonate deep shelf facies and siliciclastic turbidite facies, respectively, of Malo 2001) and sandstone and siltstone turbidites of the Wapske Formation (Tobique Group), which were deposited in an outer shelf or slope environment in the southern part of the Chaleur Bay Synclinorium (Boucot and Wilson 1994; Wilson 1990; Pickerill 1991). A shallow-water to subaerial setting for the upper Indian Point and Val d’Amour formations implies that the Restigouche area constituted a relative topographic high flanked to the northwest and southeast by deeper water. The geometry of the Early Devonian basin may therefore be envisioned as a compartmented trough with a central, at least partially emergent ridge (Fig. 7b). This ridge is a likely source of the large clasts of White Head calcilutite that are found in limestone conglomerate near the top of the Indian Point Formation.
Summary of Gaspé Belt evolution in New Brunswick
Caradocian to mid-Wenlockian (ca. 450–430 Ma)
The sedimentary, magmatic, and deformational histories of different parts of the Gaspé Belt in New Brunswick are summarized in Fig. 8. Gaspé Belt evolution can be considered to begin with Caradocian uplift caused by collision of the Popelogan–Victoria arc with Laurentia, which is reflected in the Late Ordovician unconformity at the top of the Balmoral Group (van Staal et al. 1991, 1998; van Staal 1994). Following collision, continued convergence between Avalonia and Laurentia was enabled by northwest-directed subduction of the Tetagouche–Exploits back-arc basin (van Staal and de Roo 1995; van Staal et al. 1998) (Fig. 7a). Extension and subsidence leading to basin formation are interpreted as a response to this subduction (van Staal et al. 2003). Late Caradocian to Wenlockian sedimentation therefore occurred in a fore-arc setting with respect to back-arc subduction and Early Silurian arc volcanism in the Lac Témiscouata area of Quebec (David and Gariépy 1990) (Fig. 7a). Late Ordovician – Early Silurian fore-arc basin infilling may simply reflect increasing sediment thickness but is also related to Wenlockian eustatic sea-level regression (Bourque 2001) and may be linked to uplift of the adjacent Brunswick subduction complex, which was emergent by the late Llandoverian (van Staal et al. 2003). The Popelogan Inlier was emergent for most of the Late Ordovician; the implications of this are addressed in the Discussion.
Mid-Wenlockian to Pridolian: Salinic orogeny (430–420 Ma)
Closure of the Tetagouche–Exploits back-arc basin coincided with sinistral oblique collision of the Cabot promontory of Avalon with the Newfoundland promontory of Laurentia in the Late Silurian, culminating in the Salinic orogeny (Dunning et al. 1990; Lin et al. 1994; Cawood et al. 1995). A middle Wenlockian to late Ludlovian hiatus in the sedimentary record defines the disconformable contact between the lower and upper parts of the Chaleurs Group and is the most obvious manifestation of the Salinic orogeny in northern New Brunswick (Figs. 3, 8). The impact of the Salinic orogeny varied markedly across the northern Appalachians. In the northern Miramichi Highlands, Salinic tectonism consisted of climactic D2 deformation related to sinistral transpression (van Staal 1994; van Staal and de Roo 1995). In the Quebec Reentrant (Fig. 9), most of the Gaspé Belt experienced only normal faulting and differential uplift in the Late Silurian, although Silurian folds have been documented in parts of the Gaspé Peninsula and Maine (Malo and Kirkwood 1995; Malo et al. 1995; Malo 2001; Hibbard 1994). In the Restigouche area, the effects of Salinic deformation were relatively modest and characterized primarily by uplift, block faulting, and possibly open folding (without cleavage) around northwest-trending axes. In the Kedgwick area to the southwest, fold interference patterns in the Aroostook–Percé Anticlinorium provide clearer evidence of early, probably Salinic folding (Carroll 2003). Bimodal, within-plate volcanism ranging from Wenlockian to Emsian (Fig. 8) indicates that Salinic tectonism occurred within an overall extensional regime (Dostal et al. 1989, 1993; Keppie and Dostal 1994). The causes of uplift associated with the Salinic orogeny are undoubtedly rooted in the thermal anomaly associated with magmatic activity; this subject will be explored further (see Discussion section).
In the Charlo – Jacquet River area (Fig. 1), within-plate, mainly subaerial volcanic rocks of the middle Wenlockian to Ludlovian Bryant Point and Benjamin formations (Chaleurs Group) are disconformably overlain by a thin unit of polymictic, fossiliferous conglomerate at the base of the Mitchell Settlement Formation, which is the basal unit of the Dalhousie Group section in that area (Walker and McCutcheon 1995; Wilson 2000b) (Fig. 3). At Charlo – Jacquet River, therefore, rocks immediately below the Salinic disconformity are younger than those in the Restigouche area, where no record of Silurian volcanism exists (Figs. 3, 8); nevertheless, both areas were emergent in the Late Silurian. In contrast, near Port Daniel in the eastern Gaspé Peninsula (the type area of the Indian Point Formation), a Late Silurian unconformity is absent and the Indian Point Formation forms the top of a conformable Chaleurs Group section (e.g., Bourque et al. 2000). Similarly, Silurian sedimentation was more or less continuous in the Nigadoo River Syncline of northern New Brunswick (Fig. 1), although a very brief Wenlockian hiatus is suggested by clasts of the late Llandoverian to early Wenlockian La Vieille Formation in middle Wenlockian to middle Ludlovian conglomerates of the Simpsons Field Formation (Walker and McCutcheon 1995).
Pridolian to Pragian (420–410 Ma)
In the Late Silurian and Early Devonian, the Gaspé Belt was the site of a foreland basin developed in front of a northwest-migrating Acadian orogenic wedge (Malo 2001; Bradley et al. 2000) (Fig. 7b). Deposition of the Indian Point Formation was, at least in part, contemporaneous with D3 extensional collapse that followed rapid Late Silurian uplift in the Miramichi Highlands (de Roo and van Staal 1994; van Staal and de Roo 1995). As proposed by van Staal and de Roo (1995), extensional collapse in the Early Devonian is likely responsible for the Lochkovian (T2) transgression of Malo and Bourque (1993) and Bourque et al. (2000). Nevertheless, continued uplift in the Squaw Cap block is demonstrated by intrusion of the Pabos and White Head formations by Lochkovian hypabyssal felsic intrusive rocks (Squaw Cap Felsite) and the presence of White Head cobbles in the upper part of the Indian Point Formation. In addition, poorly developed cleavage and low thermal maturity of sedimentary rocks in the Squaw Cap block support shallow burial before the onset of Acadian deformation. For example, a significant decrease in burial depth east of the Sellarsville Fault is indicated by a decrease in conodont CAI in the Pabos Formation, from a CAI of 4–5 on the west side (Nowlan 1983a; Nowlan and Barnes 1987) to 1–2 on the east side (Nowlan 1983a and this study). Low thermal maturities in the Squaw Cap block are consistent with an illite crystallinity transition to higher values (representing lower metamorphic grade) to the east of a line coinciding with the Sellarsville Fault (Duba and Williams-Jones 1983; Hesse and Dalton 1991).
Magmatic activity in the Restigouche area consisted of high-level intrusion of the Squaw Cap Felsite and its sister intrusion to the northwest and initiation of subaerial to shallow-water Val d’Amour volcanism (Figs. 7b, 8). A middle Lochkovian spore-indicated age for the lower part of the Val d’Amour Formation demonstrates that early Val d’Amour volcanism was roughly coeval with intrusion of the Squaw Cap Felsite (415.0 ± 0.5 Ma). In the southern and eastern parts of the Chaleur Bay Synclinorium in New Brunswick, Early Devonian volcanic activity was mainly subaqueous. In the Tobique Group, south of the Rocky Brook – Millstream Fault (Fig. 1), bimodal, within-plate volcanic rocks are interbedded with deep shelf to outer slope sedimentary rocks (Dostal et al. 1989; Wilson 1992; Pickerill 1991). A sample of rhyolite from this area has yielded a U–Pb (zircon) age of 412.5 ± 2.0 Ma (M.L. Bevier, written communication, 1990) (Figs. 8, 9). At Charlo – Jacquet River, east of the study area (Fig. 1), the Dalhousie Group comprises felsic to mafic volcanic rocks interbedded with fossiliferous marine sedimentary rocks. Deposition of these rocks was interpreted by Walker and McCutcheon (1995) to have occurred in a foredeep formed in response to crustal loading by the Acadian orogenic wedge.
Emsian to Eifelian (410–390 Ma)
Late Early to Middle Devonian transpressive convergence of Laurentia and Gondwana is manifested in the development of folds, cleavage, southeast-verging reverse faults, dextral strike-slip faults, and continued uplift in the Restigouche study area. These events are coeval with D4 dextral transpression in the northern Miramichi Highlands (van Staal and de Roo 1995; de Roo and van Staal 1994) and coincide with the R3 regressive phase of Malo and Bourque (1993). Contrasting intensities of deformation in the Restigouche area suggest that the Squaw Cap block and Popelogan Inlier remained in relatively elevated crustal positions compared to areas west of the Sellarsville Fault and between the Popelogan Inlier and McKenzie Gulch – Black Lake faults. A middle Emsian angular unconformity is inferred from differences in average bedding attitudes between the Val d’Amour and Campbellton formations and is supported by an early Emsian concordant U–Pb age from the top of the Val d’Amour Formation, compared with a spore-indicated late Emsian age for the overlying Campbellton Formation. Together, these circumstances imply that Acadian orogenesis began in the Restigouche area around middle Emsian time, which is compatible with proposed late Emsian deformation of the Fortin Group farther northwest (Bourque et al. 2001).
The Balmoral Group forms a window of Dunnage Zone basement exposed in the core of the Popelogan Anticline. Following collision of the Popelogan Arc with composite Laurentia in the Caradocian, the Popelogan Inlier experienced uplift, as demonstrated by a sedimentary hiatus that lasted until the latest Ashgillian. During this time, the uplifted area shed detritus that was deposited in the young fore-arc trough to the west. Subsequently, the Balmoral Group was buried by sedimentary rocks of the White Head and Upsalquitch formations. The total thickness of these units in the Popelogan area could not have been great, however, as metamorphism never exceeded zeolite facies. Similarly, weak deformation in the Balmoral Group is best explained if these rocks remained shallowly buried during Acadian orogenesis.
Models of middle Paleozoic tectonic evolution in northern New Brunswick should attempt to explain modest deformation and metamorphism in the Popelogan Inlier. Clues to possible resolution of this problem are found in detailed studies of geodynamic conditions in fore-arc and foreland settings at active continental margins. First, the presence and location of an emergent area on the eastern margin of the Gaspé Belt fore-arc basin conform to the position of the “structural high” in the ridged fore-arc model of Dickinson and Seely (1979) (Fig. 7a). This ridge (i.e., Popelogan Inlier) would have been located on the inboard (fore-arc) side of the structurally highest nappe in the Brunswick subduction complex, i.e., the earliest nappe to be incorporated into the subduction complex in the Late Ordovician. The inboard location must have been sufficiently removed from the underlying sole of the nappe to escape the effects of shear exerted by the underthrust, subducting plate. The “Popelogan nappe” gradually subsided during prograde accretion of successive thrust slices at the trench (Fig. 7a) (cf. Ingersoll 1988) and by the latest Ashgillian was disconformably overlain by the White Head Formation. Interestingly, uplift of the fore-arc structural high has been linked to the formation of restricted basins within the fore-arc region, where organic-rich muds accumulate under anoxic conditions (Dickinson and Seely 1979). This may account for the abundant carbonaceous shale present in the Ritchie Brook Member, which forms the top of the Boland Brook Formation near Kedgwick, southwest of the study area (Carroll 2000, 2003).
Second, a possible explanation for a deformational “shadow” in the Popelogan area arises from similarities between the Gaspé Belt and basins formed adjacent to the “retreating subduction boundaries” of Royden (1993). Fore arcs associated with retreating subduction boundaries are characterized by low-grade to low metamorphism, little or no involvement of basement rocks in crustal shortening, and a protracted history of flysch deposition in deep basins (Royden 1993), all of which apply to the Gaspé Belt. At such boundaries, the rate of subduction (in this case, of Tetagouche–Exploits back-arc crust) exceeds the overall rate of convergence, placing all of the supra-subduction region, including the fore arc, under extension (cf. van Staal and de Roo 1995).
Following the Salinic orogeny, sedimentation in the Gaspé Belt occurred in a foreland setting with respect to convergence of the Avalonian (Acadian) orogenic wedge (Fig. 7b). Robinson et al. (1998), Bradley et al. (2000), and Bradley and Tucker (2002) have proposed that the Late Silurian – Middle Devonian evolution of the belt can be understood in terms of a northwest-propagating, forebulge – foredeep – orogenic wedge “sine-wave.” The migration of the emergent orogenic wedge that presumably supplied at least some of the detritus to the foreland basin (Bradley and Hanson 2002; Malo 2001) cannot be traced in New Brunswick with the same degree of confidence as in Maine because of poor fossil and geochronological control, but it is possible to estimate the position of the deformation front at two points in time. At the base of the Lochkovian (418 Ma) it must have been located near the western margin of the Miramichi Highlands (Fig. 9), as the Fredericton Trough was deformed and uplifted by this time (van Staal and de Roo 1995), and several post-tectonic plutons in the Miramichi Highlands are latest Silurian to earliest Devonian in age (Whalen 1993; Bevier and Whalen 1990). In the middle Emsian (ca. 400 Ma), the position of the deformation front is constrained in the northern part of the Chaleur Bay Synclinorium by a 401 ± 1 Ma U–Pb (zircon) age from the post-tectonic Jerry Ferguson Porphyry (McCutcheon and Bevier 1990) (Figs. 8, 9), and by the unconformity between early Emsian volcanic rocks of the Val d’Amour Formation and late Emsian molasse of the Campbellton Formation, which reflects close proximity to the emerging orogenic wedge. This inferred position of the middle Emsian deformation front fits well with the reported position in northern Maine at that time (Bradley et al. 2000) (Fig. 9).
Several hypotheses have been put forward to explain widespread Late Silurian – Early Devonian uplift and within-plate volcanic activity in the Gaspé Belt. Dostal et al. (1989, 1993) and Keppie and Dostal (1994) concluded that Siluro-Devonian volcanic rocks in the northern Appalachians were erupted in a northwest–southeast-trending rift zone that developed during dextral transpression in the Quebec Reentrant (Fig. 9). The Salinic “disturbance,” according to this model, was produced by thermal uplift associated with rifting. On the other hand, van Staal and de Roo (1995) and Cawood et al. (1995) present evidence in support of a delamination mechanism to explain uplift caused by rebound of delaminated continental crust in the Late Silurian, followed by magmatism related to asthenospheric upwelling, and extensional collapse of the elevated orogen in the Early Devonian. Bradley et al. (2000) and Bradley and Tucker (2002) invoke various models involving subduction or “double subduction” under the Central Maine Basin to explain volcanic activity in the Picataquis Belt in northern Maine and the Coastal Volcanic Belt in southern Maine and New Brunswick.
The presence of sedimentary basins and the subsidence they imply are important considerations in comparing and evaluating these models. Postcollisional delamination of a dense lithospheric root (e.g., Kay and Kay 1993; Sacks and Secor 1990) is an attractive mechanism for explaining orogen-wide volcanism, rapid uplift, and subsequent extensional collapse of the northern Miramichi Highlands. Furthermore, separation of the downgoing slab would have been necessary to allow for continued convergence of the Avalonian and Laurentian continental plates in the Devonian (Fig. 7b). Most Early Devonian sedimentation and volcanism in the Chaleur Bay Synclinorium clearly occurred in a marine setting, however, suggesting subsidence (or T2 transgression of Malo and Bourque 1993) rather than uplift. Conversely, the rifting model of Keppie and Dostal (1994), which more readily accommodates subsidence, does not illustrate a clear spatial relationship between rifting at the Quebec Reentrant – St. Lawrence Promontory margin and the broad distribution of Siluro-Devonian volcanic rocks between western Maine and northeastern Gaspé. Moreover, the switch from sinistral to dextral transpression, invoked by Keppie and Dostal to explain rift-related volcanism in the Silurian, coincides with the Salinic orogeny (van Staal and de Roo 1995; Cawood et al. 1995; Dunning et al. 1990), whereas, as we have seen, volcanic rocks of the Bryant Point and Benjamin formations (Chaleurs Group) underlie the Salinic unconformity. It is intriguing to note that the “retreating subduction boundary” paradigm discussed previously is reported to be especially common at reentrants along irregular plate margins (Royden 1993). The model implies that, during and after promontory– promontory collision, convergence between the opposing continental masses would have ceased, at least temporarily, while subduction of dense lower crustal lithosphere continued in the Quebec Reentrant. Therefore, subduction rate would have exceeded convergence rate, placing the overriding plate in the reentrant under tension. A detailed study of the geochemistry, petrogenesis, and mantle–crustal sources of Silurian and Devonian volcanic rocks, including how the character of volcanism may have changed through time, may shed more light on geodynamic settings, and which interpretation or combination thereof is more plausible. Such a study, however, is dependent on a critical mass of high-quality lithogeochemical and isotopic data.
Reviews by C. St. Peter, T. Chowns, and journal Associate Editor Jisuo Jin significantly improved the manuscript and are gratefully acknowledged. The principal author has benefitted from stimulating discussions and field trips with D. Brisebois, D. Lavoie, and M. Malo in the Gaspé Belt of New Brunswick and Québec. S. McCutcheon acted as sounding board for some of the ideas presented herein and reviewed an early draft of the manuscript. Phil Evans assisted with the figures. This is a contribution to the “Appalachian Forelands and St. Lawrence Platform” NATMAP (National Geoscience Mapping Program) project.
↵1 This article is one of a selection of papers published in this Special Issue on Eastern Canada Silurian–Devonian Gaspé Belt NATMAP Project.
- Received July 1, 2003.
- Accepted February 4, 2004.
- Published on the NRC Research Press Web site at http://cjes.nrc.ca on May 18, 2004.
- © 2004 NRC Canada