The Dog Bay Line, a Silurian suture key to deciphering Appalachian accretionary history, was first recognized in Newfoundland. It marks where the Ordovician Tetagouch–Exploits ensimatic back-arc basin (TEB), which had opened within the leading peri-Gondwanan Gander terrane, finally closed. Here, we extrapolate this suture into New England, placing it between the Liberty–Orrington–Miramichi inliers (LOM) and the Merrimack–Fredericton trough (MFT). Southeastward, marine strata of the MFT overlie the TEB passive margin, exposed in the Ganderian St. Croix block, and display southeast-vergent structures transected by Acadian cleavage. They structurally underlie southeast-vergent thrusts at the base of the LOM. Northwestward, the LOM, Central Maine – Matapedia trough (CMMT), and Lower Silurian igneous rocks record elements of the upper plate trench–arc system, respectively, a subduction complex, forearc basin, and arc. The CMMT forearc received detritus both from the northwesterly arc region, and also from the Early Silurian-exhumed subduction complex. Minimal contrast in Silurian turbidites near the line may be due to sediment bypassing the subduction complex, and (or) a common provenance when the complex emerged above sea level. Salinic unconformities in the upper plate (arc–trench) reflect episodes of shortening, within an overall extensional setting that resulted in thinned, weakened lithosphere, and also final uplift accompanying latest Silurian slab breakoff. Silurian strata of the Coastal Volcanic Belt document a separate arc system built on Ganderia’s trailing edge, where northwest-directed subduction of a narrow seaway led to latest Silurian collision with buoyant, strong lithosphere of Avalonia’s passive margin, and the onset of Acadian typically dextral-oblique, northwest-vergent deformation.
The Earth’s unique continental crust, a highly complex mosaic of orogens, each in turn highly complex, contains the vast majority of the planet’s precious geologic record. The Paleozoic Appalachian–Caledonide orogen in particular has been fertile ground for several key advances in understanding Earth history and process, notably the rock cycle (Hutton 1795), first clues to ancient continental margins (Hall 1859), supercontinent cycle (Wilson 1966), terrane concept (Coney et al. 1972), and first plate tectonic map of an entire orogen (Williams 1978). The Appalachian orogen grew at the expense of the Iapetus Ocean, located between Laurentia and Gondwana, by episodic accretion of outboard arcs and ribbon continents. Somewhat paradoxically, whereas one of the earliest accretionary events, the Taconic collision between Laurentia and a peri-Laurentian arc, was elucidated decades ago (e.g., Rowley and Kidd 1981), only recently have the detailed relationships among its terranes, subsequent accretionary events, and Middle Paleozoic orogenesis been clarified (van Staal et al. 2009).
Purpose and scope
Our purpose here is to extrapolate a subtle but important Silurian suture, first recognized in Newfoundland and named the Dog Bay Line (Williams et al. 1993), into the New England region. This suture marks the former site of the wide Tetagouche–Exploits back-arc basin (TEB) that developed during Ordovician time within the leading peri-Gondwanan terrane known as Ganderia (van Staal 1994; van Staal et al. 1998). It was first recognized on the basis of contrasting Lower Silurian strata that are well exposed on the north-central coast of Newfoundland. West of the line, marine to subaerial clastic sediments are linked to Laurentia; east of the line, marine sediments are instead linked to the Gander margin (Pollock et al. 2007).
In the following, we first summarize the significant stratigraphic, structural, and contact relationships southeast of the line, in the Merrimack–Fredericton trough (MFT) and Coastal Volcanic Belt (CVB), and to the northwest, in the Liberty–Orrington–Miramichi inliers (LOM) and Central Maine – Matapedia trough (CMMT), including correlative Upper Silurian cover and Lower Silurian arc rocks that mostly lie farther to the northwest. We then reconstruct the tectonic configurations through the Silurian and suggest that subduction of the TEB beneath Laurentia during the Early Silurian led to a mid to Late Silurian collision. The arc–trench system built on Laurentia’s southeastern margin was juxtaposed with a passive margin built on the main block of Ganderian rocks exposed in the St. Croix Belt (SCB) near the Gulf of Maine coast. This Silurian collision was the principal cause of the Salinic orogeny (Dunning et al. 1990). It also set the stage for the kinematically distinct Acadian orogeny that spread northwestward across the Salinic-deformed region during the latest Silurian through Middle Devonian. Prior to summarizing the tectonic evolution, we discuss differences between our proposed model and the plethora of existing models (reviewed by Tremblay and Pinet 2005).
The region of interest lies between rocks in the Laurentian realm (Hibbard et al. 2006) affected by Taconic orogenesis (Taconic orogen on Fig. 1) and the Avalon terrane that crops out in coastal New Brunswick southeast of the Caledonia Fault and in southeastern New England southeast of the Bloody Bluff Fault. It is approximately centred on the less well known and complex Gander terrane of medial New England. The Avalon terrane, which defines the paleogeographic element known as Avalonia, hosts a Lower Paleozoic mud-rich continental platform sequence (O’Brien et al. 1996) that is distinct from the Gander sequence (van Staal et al. 1996). Ganderia was the first and Avalonia the second of the peri-Gondwanan terranes to be accreted to Laurentia. Rocks along the coast of Maine, formerly grouped with Avalon (Williams 1978), are now included with the Gander terrane of central New England (van Staal and Fyffe 1995; Schultz et al. 2008; Fyffe et al. 2009).
Toward the northwest, the region straddles the main Iapetan suture of Late Ordovician age (∼450 Ma), known as the Red Indian Line (RIL) that separates peri-Laurentian and peri-Gondwanan rocks (Williams et al. 1988; Figs. 1, 2). Largely concealed in New Brunswick and in northern Maine, in northwestern Maine the RIL is now well located between the peri-Gondwanan Dead River Formation and peri-Laurentian Chain Lakes complex in northwestern Maine; the associated Dunnage and Hurricane Mountain mélanges probably formed on the Gondwanan side, which is supported by the Ganderian affinities of the inherited zircons (van Staal et al. 2010; Christopher Gerbi, personal communication, 2010). The RIL divides the Dunnage Zone of Williams (1979), representing vestiges of the Iapetus Ocean, into the peri-Laurention Notre Dame Subzone and the peri-Gondwanan Exploits Subzone. Lower to Middle Ordovician volcanic sequences of the Exploits Subzone, built on extended Neoproterozoic to Lower Cambrian Gander continental basement, constitute the Popelogan–Victoria arc in Canada and part of the Bronson Hill arc in New England (van Staal et al. 1998; Dorais et al. 2011).
Silurian strata of the medial northern Appalachian orogen extend across the terrane boundaries defined on contrasts in the Ordovician and older rocks (Hibbard et al. 2006). They are locally continuous across the Red Indian Line, indicating that the leading peri-Gondwanan arc had collided with Laurentia by the end of the Ordovician. Laurentian detritus and fossils present in the overlying Upper Ordovician sedimentary rocks in Maine, New Brunswick, and Newfoundland confirm this relationship (Roy 1980; Fyffe and Fricker 1987; van Staal 1994; van Staal et al. 1998; Wilson et al. 2004). The Silurian strata crop out in four main belts (Figs. 1 and 2). From northwest to southeast, these are (1) the margins of the Connecticut Valley – Gaspe trough (CVG), including a number of plutons within the Boundary Mountains – Bronson Hill inlier; (2) the Central Maine – Matapedia trough (CMMT), including some strata that onlap the adjacent inliers; (3) the Merrimack–Fredericton trough (MFT); and (4) the Coastal Volcanic Belt (CVB). Later, we interpret these belts as, respectively: (1) a short-lived Early Silurian arc built on the margin of Laurentia; (2) the associated forearc; (3) a foredeep developed on the descending plate; and (4) a separate and independent arc built contemporaneously on Ganderia’s trailing edge above the descending Avalon plate to the southeast.
The Maine – New Brunswick region includes the type localities of the Salinic and Acadian orogenies. The former is documented by multiple Early to Late Silurian unconformities in belts 1 and 2 (e.g., Boucot 1962, 1969; Wilson et al. 2004). The Acadian orogeny was defined on the basis of an unconformity in eastern Maine separating Upper Devonian Perry Formation from Upper Silurian Eastport Formation (Miller and Fyffe 2002; McLaughlin et al. 2003). Similar relationships are observed in northern Maine and northern New Brunswick, but here the Acadian orogeny is younger and entirely restricted to the Early to Middle Devonian (Bradley et al. 2000). The two orogenic events have been related, respectively, to the docking of Ganderia and the docking of Avalonia (van Staal et al. 2004, 2009).
Proposed location of the Dog Bay – Liberty Line
In New Brunswick, the Dog Bay – Liberty Line has been placed along the Bamford Brook fault between the Miramichi inlier and Fredericton Trough (Fig. 1; van Staal 2007; van Staal et al. 2009). Near the New Brunswick – Maine border, the Codyville fault occurs in this position; it was seismically imaged as a crust-penetrating, shallow to moderately northwest-dipping structure and is dextrally offset along the steeply inclined Norumbega Fault (Ludman et al. 1990, 1993; Ludman 1991). Farther southwest, the Liberty–Orrington and Boothbay thrusts place rocks of the Liberty–Orrington inlier, an along-strike correlative of the Miramichi (West et al. 2003), over the Bucksport Formation of the Fredericton trough (Tucker et al. 2001). In southwestern Maine, the line lies along the northwestern margin of the Merrimack trough, possibly at the base of the Gonic complex and correlative highly sheared rocks (Thompson et al. 2004) or it is concealed beneath the Berwick Formation (Hussey et al. 2010). In New Hampshire and Massachusetts, it continues in this position between the Merrimack trough and Neoproterozoic Massabesic Gneiss Complex that lies to the northwest (Dorais et al. 2001).
Geologic relationships southeast of the Dog Bay – Liberty Line
Marine strata of the Merrimack–Fredericton trough overlie a block of the northwest-facing Gander passive margin, exposed in the adjacent St. Croix Belt, and they structurally underlie remnants of the southeast-facing Brunswick subduction complex (van Staal et al. 2008), exposed in the hangingwall of the Bamford Brook fault system. Hence, they were deposited in a foredeep tectonic setting. Farther southeast in the Gander terrrane, Silurian rocks of the Coastal Volcanic Belt rest on an angular unconformity that truncates the contact between the St. Croix and Ellsworth belts.
The Merrimack–Fredericton trough (MFT, Fig. 1) comprises the Silurian Kingsclear Group of New Brunswick (Fyffe and Riva 2001; West et al. 1992), including in ascending order the Digdeguash and Flume Ridge Formations (Figs. 1, 2, location 13); the Appleton Ridge and Bucksport Formations of eastern through mid-coastal Maine (Figs. 1, 2, location 14); and the Merrimack Group of southwestern Maine, southeastern New Hampshire, Massachusetts, and Connecticut (Figs. 1, 2, location 15; Wintsch et al. 2007). The latter includes the Eliot and Kittery Formations. All of these sequences display trends of increasing grain size and carbonate content upward. The basal Digdeguash, Appleton Ridge, and Eliot Formations consist of thinly bedded, variably calcareous pelite-rich turbidites. The overlying Flume Ridge, Bucksport, and Kittery Formations comprise calcareous sandstones and shales, locally metamorphosed to calc-silicate as in the aureole of the Lucerne Granite east of Bangor, Maine. Paleocurrent features in the Kittery indicate a source to the southeast (Hussey et al. 1984; 2010). The Berwick Formation has been removed from the Merrimack Group, as it more strongly resembles strata of the Central Maine trough and is everywhere in fault contact with the Eliot (Hussey et al. 2007). Included within the Appleton Ridge Formation, the Ghent Member (previously Ghent Phyllite) consists of distinctive dark pelites (West et al. 2000).
Silurian (Salinic) folds in the Fredericton trough of New Brunswick (Fig. 1, location 13) verge to the southeast (Park and Whitehead 2003). Younger folds are related to northwest-vergent thrusting and dextral shear. In Maine, east of Bangor (Fig. 1, location 14), the Bucksport Formation dips steeply southeast, more steeply than the cleavage, suggesting the beds are overturned and top to the northwest; as in New Brunswick, this deformation overprints early tight folds (Stephen Pollock, personal communication, 2005). At Ogunquit, Maine, along Marginal Way (Fig. 1, near location 15), the Kittery Formation displays northwest-vergent folds with, apparently, a single, southeast-dipping axial-plane cleavage. These are believed to correlate with the F2 Acadian folds described in New Brunswick. Nearby, at Bald Head, these F2 folds overprint F1 isoclines with southeast vergence (Figs. 3e and 3f).
Contact relationships (northwest)
In New Brunswick and adjacent Maine (Fig. 1, between locations 13 and 14), the Bamford Brook – Codyville fault system on the northwestern margin of the Fredericton trough dips to the northwest, and juxtaposes relatively high-grade metamorphic rocks of the Miramichi inlier in its hanging wall with low-greenschist or sub-greenschist facies rocks in its footwall. In addition, the hanging wall rocks were unroofed during the Silurian (Ludman et al. 1993) with detritus in the Kingsclear Group (Fyffe 1995) clearly establishing a link between these two tectonic elements. These relationships, in combination with the F1 southeast-vergent folds, suggest that the Dog Bay Line in New Brunswick was a major southeast-vergent thrust during the Silurian.
In Maine, between Penobscot Bay and Casco Bay (Fig. 1, southeast of location 14), Tucker et al. (2001) presented evidence that the Liberty–Orrington thrust also verged southeast during the Silurian. Strata of the Fredericton trough top northwest toward the Liberty–Orrington thrust at the base of the Liberty–Orrington thrust sheet (Tucker et al. 2001). Stitching plutons are at least as old as 418 Ma but probably at least 422 Ma (Tucker et al. 2001). East of Casco Bay (Figs. 1, 4), several klippen of Ordovician rocks on top of Bucksport Formation (Osberg et al. 1985) further suggest a regionally significant thrust, known locally as the Boothbay thrust (Hussey et al. 2010), which cuts up-section and verges to the southeast. At the northeast end of the Liberty–Orrington inlier, the Liberty–Orrington thrust is broadly folded, becoming north–south where it is truncated by the Norumbega Fault. This relationship is consistent with an early age of juxtaposition, as is a shared set of fabrics (Kaszuba and Simpson 1989). Along the southwest-striking section to the southwest, well-documented, Norumbega-related late dextral shear features (e.g., Short 2000; West et al. 2000) obliterate evidence of the early movement history.
Farther southwest (Fig. 1, location 15), the Merrimack Trough is bounded on the northwest by the late Nonesuch River and other faults, locally marked by metamorphosed phyllonites (Thompson et al. 2004), that juxtapose the Elliot Formation of the MFT with the Berwick Formation (Hussey et al. 2010).
Contact relationships (southeast)
In Maine, the Sennebec Pond Fault (Fig. 1) juxtaposes the Fredericton trough with the St. Croix belt to the southeast. The movement history on this late, post-metamorphic fault is controversial. It has been portrayed both as a northwest-vergent thrust (Stewart et al. 1993; West et al. 1995; Hibbard et al. 2006) and a southeast-side-down normal fault (Tucker et al. 2001). In Massachusetts, the Clinton–Newbury Fault (Fig. 1) juxtaposes the Merrimack trough with the Nashoba terrane, a belt of Lower Paleozoic volcanic strata that was intensely deformed and metamorphosed during the Silurian (Hepburn et al. 1995).
In New Brunswick, the primary stratigraphic relationship between the Fredericton trough and southeasterly St. Croix belt is preserved around the margin of a Fredericton outlier (near location 16 on Fig. 1). Here, Fredericton trough sediments disconformably overlie the pre-Silurian Cookson Group, in descending order, the Kendall Mountain, Woodland, and Calais Formations (Fyffe 1995; Fyffe and Riva 2001; Park and Whitehead 2003). We attribute this disconformity to the passage of a foreland-propagating flexure caused by the load of southeast-directed thrust sheets in the Brunswick subduction complex (van Staal et al. 2008).
St. Croix and Ellsworth belts
In the St. Croix belt (Figs. 1, 2), the Kendall Mountain Formation of easternmost Maine contains orthoquartzites interbedded with Caradocian graptolitic black shales (Fig. 2, column 14; Fyffe and Riva 1990). It is interpreted as part of the southeastern passive margin of the Tetagouche basin (Fig. 5a). Caradocian quartzites are also present at the top of the Harts Neck Formation in western Penobscot Bay (Hussey et al. 2010). The Tremadoc Calais Formation correlates with the Penobscot Formation, which on the west side of Penobscot Bay overlies the Megunticook Formation and Battie Quartzite, also of continental affinity. Neoproterozoic rocks (e.g., Seven Hundred Island Formation, western Penobscot Bay), based on field relationships (Johnson 2001; Johnson and McLeod 1996), detrital zircon studies (van Staal et al. 2004; Fyffe et al. 2009), and isotopic data (Whalen et al. 1994; Samson et al. 2000; McLaughlin et al. 2003), have been interpreted as Ganderian continental basement.
The St. Croix and related rocks were deformed post-Caradoc but prior to Late Silurian time (West et al. 1995; Tucker et al. 2001). Extremely stretched Caradocian brachiopods in the Benner Hill Formation illustrate the severity of this regional deformation. Isoclinal folds in the Camden Hills have northwest vergence (Berry 1987).
Structurally above the St. Croix belt, the Middle Cambrian Ellsworth Schist (Figs. 1, 2, location 17) may have accumulated in an extensional setting (Reusch 1995; Reusch and Rust 2001; Schulz et al. 2008), possibly related to the rifting of Ganderia from Gondwana. The Ellsworth and correlative New River belts were juxtaposed with the St. Croix belt post-Tremadoc, as Ellsworth rocks structurally overlie the Penobscot Formation in the Lord’s Cove window along the Bagaduce River of eastern Penobscot Bay (Reusch 2002). A Lower Ordovician (478 Ma) stitching pluton across the Annidale – New River contact (Johnson et al. 2009) and, in Penobscot Bay, a quartz pebble conglomerate of suspected mid-Ordovician age (Eusden et al. 1996a) suggest pre-Caradoc crustal thickening prior to development of a passive margin setting during the Caradoc.
Coastal Volcanic Belt
Rocks of the Coastal Volcanic Belt (CVB; Figs. 1, 2, locations 16–17) overlie an angular unconformity on the southeastern flank of the St. Croix–Ellsworth block. Related plutonic rocks intrude as far northwest as the Liberty–Orrington inlier (Tucker et al. 2001). These stitching plutons indicate a pre-Late Silurian (423–421 Ma) age for thrusting of Central Maine rocks onto the Fredericton trough along the Dog Bay – Liberty Line.
In New Brunswick, the Llandovery Oak Bay Formation (Miller and Fyffe 2002) overlies the St. Croix belt via a faulted unconformity (Figs. 1, 2, location16). Spectacular conglomerates contain a great variety of igneous rocks, as well as pebbles derived from the underlying Cookson Group, that most likely records the rapid unroofing of pre-Late Ordovician and older magmatic arc(s). Detrital zircon ages match most of the Ganderian basement ages; and Neoproterozoic through Cambrian ages typical of the Ellsworth belt confirm that the CVB is an overstep sequence (Fyffe et al. 2001, 2009).
In Penobscot Bay (Figs. 1, 2, location 17), the correlative Ames Knob Formation (Smith et al. 1907) has a relatively thin basal conglomerate that contains a greater fraction of sedimentary materials. Foliated black shale, present as a clast in this conglomerate (Fig. 3G), resembles the Penobscot and Calais Formations of the Cookson Group. The Coastal Volcanic Belt is only moderately deformed; beds generally dip southeast less steeply than cleavage (Reusch 2000).
Geologic relationships northwest of the Dog Bay – Liberty Line
Northwest of the Dog Bay – Liberty Line, the Brunswick subduction complex (LOM on Fig. 1) underlies the Central Maine – Matapedia trough. The Upper Ordovician – Lower Silurian of the trough contains detritus from an arc that lay to the northwest as well as detritus from the subduction complex. It is interpreted as a forearc basin. Upper Silurian strata of the trough correlate with thin shelf units deposited with angular unconformity on the medial Ordovician inliers to the northwest, and locally nonconformably on Upper Ordovician – Lower Silurian arc plutons.
Liberty–Orrington and Miramichi inliers
Between the site of the proposed Dog Bay – Liberty Line and Central Maine – Matapedia trough, diverse Ordovician and older rocks crop out in the Miramichi and Liberty–Orrington inliers of New Brunswick and Maine, possibly the Gonic complex of southwestern Maine and southeastern New Hampshire, and the Massabesic Gneiss Complex of New Hampshire and Massachusetts (Fig. 1).
In New Brunswick (Figs. 1, 2, location 11), the Miramichi inlier hosts a quartz-rich siliciclastic sequence (Miramichi and Woodstock groups) and an overlying volcanic-rich sequence (Bathurst Supergroup). The former represents a distal, extended part of the Cambrian –Early Ordovician Gander margin, and the latter parts of a Middle Ordovician arc–back-arc system. Together, they have been assembled into a southeast-vergent thrust stack, the Brunswick subduction complex, of Late Ordovician through Early Silurian age (van Staal 1994).
A similar twofold division probably exists in the Liberty–Orrington inlier (Figs. 1, 2, location 12). The lower part is scantily represented by the Copeland Formation, which contains mature quartzites and crops out at the northeast end of the inlier. The upper part comprises Middle Ordovician volcanic and related rocks of the Falmouth–Brunswick sequence and Casco Bay Group (West et al. 2003; Hussey et al. 2010), including, near its top, black shales (Diamond Island Formation) and ribbon limestones (Spurwink Limestone) of Caradoc age. Near the basal thrust, on the east shore of Bailey Island in eastern Casco Bay, a lens of soapstone suggests an early imbrication as in the Brunswick subduction complex (van Staal et al. 2008).
In southwestern Maine and adjacent New Hampshire (Fig. 1, northwest of location 15), the undated Gonic complex (Thompson et al. 2004), locally known as the Phyllonite at Church Road (Hussey et al. 2010), underlies strata of Central Maine. It is a highly sheared dark pelite full of dismembered quartz veins (Fig. 3d). Evidence of both sinistral and dextral shear is present (Thompson et al. 2004), suggesting a complicated movement history, as observed elsewhere along the line (Williams et al. 1993). Farther southwest, the Neoproterozoic Massabesic Gneiss Complex must represent basement of the Central Maine trough.
Central Maine – Matapedia trough
The boundaries of the Central Maine – Matapedia trough (CMMT) are somewhat arbitrarily defined. The northwestern boundary is taken to be the base of the Silurian Greenvale Cove Formation, which rests on the Upper Ordovician – Lower Silurian Quimby Formation (443 Ma, Moench and Aleinikoff 2003). The southeastern boundary is taken to be the southeastern limit of the Vassalboro Group (Marvinney et al. 2010), where it is typically faulted against Ordovician strata of the Liberty–Orrington inlier. In New Brunswick, Silurian strata onlap the Brunswick subduction complex, exhumed in part during the Early Silurian. Locally, Ordovician inliers are absent, and the CMMT is in direct, presumably fault contact with the MFT. The trough extends southwestward into Connecticut and northeastward beneath Chaleur Bay.
Stratigraphy (northwestern margin)
An important section of the CMMT is exposed along Route 4 between Rangeley and Phillips, Maine (Figs. 1, 2, location 6). Here, Silurian strata conformably overlie Ashgill–Llandovery turbidites of the Quimby Formation (Moench and Pankiwskyj 1988). On the north limb of the Mountain Pond syncline, shales and laminated calcareous sandstones of the thin Greenvale Cove Formation are followed upward by spectacular conglomerates of the Rangeley Formation (Fig. 3b). The Rangeley is the same age as igneous rocks to the northwest, and received detritus from them. Clasts of volcanic and plutonic rocks record the unroofing of an Upper Ordovician – Lower Silurian (Attean, 443 Ma) and (or) Lower Ordovician (Skinner, 472 Ma; Gerbi et al. 2006) magmatic arc. Several brachiopod occurrences indicate a late Llandovery C3–C5 age (Moench and Pankiwskyj 1988). The overlying Perry Mountain Formation consists of relatively clean quartz-rich turbidites and very sparse volcanic rocks (Moench and Pankiwskyj 1988). The Smalls Falls Formation contains abundant black shale, recording an abrupt change to anoxic conditions. Graptolites found several tens of kilometres to the east are early Ludlow (Moench and Pankiwskyj 1988). Both the Perry Mountain Formation and the Smalls Falls Formation are recognized southeastward to the Currier Hill syncline in Skowhegan.
A sequence almost identical to the one southeast of Rangeley is present along the west margin of the Aroostook anticlinorium (Figs. 1, 2, location 9). The Frenchville Formation contains conglomerates like those of the Rangeley, and displays a similar direction of sedimentary transport from a northwesterly source (Roy 1980).
Uppermost Silurian through Lower Devonian strata constitute part of a northwest-migrating clastic wedge deposited in an Acadian-related foreland basin (Bradley et al. 2000; Bradley and Tucker 2002). The unfossiliferous Madrid Formation consists of calcareous sandstone turbidites. The thick Carrabassett Formation, basal formation of the Seboomook Group, is dominated by black, nonrusty shales and lesser sandstone turbidites.
Stratigraphy (medial part)
In western Maine, the medial part of the Central Maine trough, exposed in the Athens thrust sheet (Tucker et al. 2001), is dominated by the Sangerville and Anasagunticook Formations (Figs. 1, 2, location 7). The presence of thinly bedded (ribbon) limestones and calcareous shales in these units, as well as in the nearby Waterville Formation to the southeast, distinguishes this part of the trough from its northwestern and southeastern margins. The Sangerville Formation, a heterogeneous unit of dominantly weakly calcareous, lithic sandstone turbidites and lesser ribbon limestones, black shales, and granule to pebble conglomerates, contains Wenlock graptolites (Bradley et al. 2000). Near Skowhegan, it underlies the Perry Mountain Formation and therefore correlates with the Rangeley Formation.
Thinly bedded limestones of this medial part of the Central Maine trough suggest correlation with the graptolitic Carys Mills Formation of Ashgill through early Llandovery age present in northeastern Maine (Figs. 1, 2, location 9; Pavlides et al. 1964). The source of carbonate sediment in the Carys Mills was likely the Anticosti platform at the edge of Laurentia (Wilson et al. 2004). The overlying Smyrna Mills Formation of the Perham Group (Roy 1980; Fyffe 1982) contains red shales in its lower part (Llandovery to Wenlock), which may correlate with the Waterville (see below), and black shales in its upper part (Wenlock to Ludlow), which correlate with the Smalls Falls Formation (Fyffe 1982). The Smyrna Mills locally hosts economically significant manganese-rich banded iron formations.
In New Brunswick (Figs. 1, 2, location 10), the Matapedia trough overlaps the Gander, Dunnage, and Humber Zones (Williams 1979). Its strata were therefore deposited after the Popologan arc collided with Laurentia. Above a minor disconformity exposed in the Popologan inlier, the Upper Ordovician Grog Brook Formation of siliciclastic turbidites records erosion of the arc–arc collision zone (Wilson et al. 2004; Roy 1980; Neuman 1994; Zagorevski et al. 2008; van Staal et al. 2009). Deep-marine carbonate-shale beds of the Ashgill through Llandovery Matapedia Group (Carys Mills Formation in Maine) also document a link with the Anticosti platform of the Laurentian continental margin. These strata grade northwestward into proximal terrigenous facies of the Frenchville Formation.
Stratigraphy (southeastern margin)
In Maine, the southeastern margin of the Central Maine trough (Figs. 1, 2, location 8) hosts the Vassalboro Group (Marvinney et al. 2010), exposed in the Lake Messalonskee thrust sheet (Tucker et al. 2001). The Vassalboro Group comprises, in descending stratigraphic order, the Mayflower Hill, Waterville, and Hutchins Corner Formations. The Mayflower Hill and Hutchins Corner Formations are very similar lithologically, both dominated by thin to very thickly bedded, immature, slightly calcareous lithic sandstones to semi-mature quartz-rich sandstones and shales. The Waterville Formation of Llandovery age (Orr and Pickerill 1995) consists of interbedded siltstones and shales, and it includes a central limestone member and an uppermost rusty shale member. The Waterville correlates with the fine-grained Anasagunticook Formation (Ludman and Osberg 1987; Osberg et al. 1985). The Mayflower Hill Formation correlates with the Sangerville Formation (Marvinney et al. 2010). Sparse maroon shales within the Vassalboro (Ludman and Griffin 1974) suggest a general correlation with the Llandovery–Wenlock part of the Smyrna Mills Formation.
The Vassalboro Group is most likely in thrust contact to the southeast with metamorphosed and deformed Ordovician rocks of the Liberty–Orrington inlier (Marvinney et al. 2010), although both erosional (Osberg 1988) and conformable (Tucker et al. 2001) relationships have been suggested, the latter based on the map pattern of parallel pre-Caradoc marine volcanic rocks, Caradoc shales, and Hutchins Corner turbidites. Near Bangor, rare sole marks suggest an eastern provenance (S. Pollock, personal communication, 2005) and Hussey et al. (2010) presented evidence from southern Maine for a southeastern provenance. While geochemical data display a Grenville (Laurentian) signature (Dorais and Wintsch 2001), these data are not inconsistent with a southeastern provenance as the sediment may have been recycled through the Brunswick subduction complex.
In New Brunswick (Figs. 1, 2, location 10), carbonate facies of the Matapedia Group pass upward into the lower part of the Chaleur Group, which is capped by Upper Silurian volcanic rocks (Wilson et al. 2004). Along the CMMT’s southeastern margin, sedimentary facies become increasingly proximal toward the southeast, where Lower Silurian strata onlap the Miramichi inlier. At least two internal angular unconformities record a complex deformation history, including folding and uplift, related to the Salinic orogeny (van Staal et al. 2009). The timing of erosion matches the ages of the two clastic pulses recorded in the Vassalboro Group (Fig. 2, see stratigraphic columns for locations 8 and 10).
Within the CMMT of west-central Maine through New Brunswick, the most prominent structures are tight to isoclinal, subhorizontal, northeast-trending folds with steeply dipping axial planes and axial plane cleavage (F2 folds of most workers; for example, the Bear Hill syncline (Moench and Pankiwskyj 1988)). Many of the best exposures consist of dominantly southeast-topping sections (e.g, Rangeley–Phillips, Bald Mountain between Weld and Wilton, Skowhegan (Fig. 3c), Carmel, Bangor, Orono). Consistent with this observation, Griffin reported that northwest-topping limbs are commonly sheared and thinned. The folded Lake Messalonskee and Athens thrusts, both cryptic and inferred on stratigraphic grounds, are assumed to verge northwest (Tucker et al. 2001) as are the thrusts shown on cross sections by Solar and Brown (2001). Large isoclinal F1 folds in central Maine, inferred on the basis of younging–cleavage relationships, are likewise assumed to verge northwest (Osberg 1988; Tucker et al. 2001). Evidence of contemporaneous and (or) late subvertical dextral shear is also variably present across the region. All of these features have been related to the Early Devonian Acadian orogeny.
Silurian-age structures are present in central and northern New Brunswick. Rast et al. (1980) documented pre-Early Devonian Salinic recumbent folds in the Carys Mills Formation, and pre-Late Silurian folds and broken formation have been identified in northern New Brunswick (van Staal and de Roo 1995; Dimitrov et al. 2004; van Staal et al. 2009).
In New England, early structures of possible Silurian age have recently been recognized. In Carmel, west of Bangor (Fig. 1), Vassalboro Group turbidites have been deformed into an early sinistral duplex. Near Farmington, southeast-topping beds have been sinistrally extended (Reusch et al. 2010; Fig. 5c).
Southwestward, Acadian deformation, affecting strata presumed to be Lower Devonian (Littleton Formation and correlatives), is intense. In the Presidential Range of New Hampshire, Eusden et al. (1996b) documented large southeast-vergent nappes (D1). In southeastern New Hampshire and adjacent Maine, the Rochester-Lebanon antiformal synclinorium is also southeast-vergent (Thompson et al. 2004). Enigmatic disturbed horizons in the Lower Silurian Rangeley Formation of the Presidential Range may be either olistostromes formed contemporaneously with early southeast-vergent thrusts (Eusden et al. 1996b) or artifacts of anatexis (Solar and Brown 2001).
The Central Maine – Matapedia trough displays a variety of conformable, unconformable, and faulted contact relationships around its periphery. In multiple locations, Silurian strata of the trough conformably overlie Ordovician marine strata. Thus, deposition was continuous through the interval of Taconic orogeny, which impacted rocks to the northwest of the Red Indian Line. Strata along the southeastern margin of the CMMT locally overlie highly sheared rocks (e.g., Gonic complex, Fig. 3d). Within the CMMT, Silurian strata underlie, in most places conformably, the Madrid Formation and Seboomook Group (unit 32a of Hibbard et al. 2006), interpreted as deposits of an Acadian-related foreland basin with southeastern provenance (Bradley et al. 2000). An exception occurs in southern New Hampshire, where the Lower Devonian Goose grits contain clasts of the Smalls Falls Formation (Eusden and Lyons 1993).
Upper Silurian cover on medial Ordovician inliers
In western New Hampshire (Figs. 1, 2, location 3), the Silurian Clough Quartzite and overlying Fitch Formations constitute a thin shelf facies on the Bronson Hill inlier. In particular, the Clough, of late Llandovery or Wenlock age, rests with angular unconformity on the Ordovician Ammonoosuc Volcanics and correlates with the Perry Mountain Formation of the CMMT. This relationship confirms that the CMMT, along its northwestern margin, was fed from the northwest.
Just 60 km northeast of Rangeley, Maine, The Forks Formation, considered Late Silurian (Marvinney 1984), contains limestone and volcanic rocks (Figs. 1, 2, location 5). It unconformably overlies the locally strongly deformed Dead River Formation (Moench and Aleinikoff 2003). This unconformity might be related to “late Taconic” deformation associated with collision between the Notre Dame and Popologan arcs (van Staal et al. 2007; Zagorevski et al. 2008); or, it could be Salinic.
West of Katahdin (Figs. 1, 2, location 4), calcareous shallow marine beds of the Ripogenus Formation, which contain brachiopods as old as late Llandovery, unconformably overlie Ordovician pillow lavas (Bradley et al. 2000). They underlie the West Branch Volcanics and Frost Pond Shale of latest Silurian to Early Devonian age (Schoonmaker et al. 2005). The Ripogenus resembles the coeval Weir Formation of the Chaleur Group in northern New Brunswick (Walker and McCutcheon 1994), which unconformably overlies Middle Ordovician to lower-most Silurian rocks of the Fournier Group (van Staal et al. 2009). Field relationships and geochemical studies indicate that the underlying lavas form part of the Tetagouche back-arc basin (Winchester and van Staal 1994), and hence the unconformity is associated with the Salinic orogeny and accretion of these rocks to Laurentia.
In the Moose River area (Figs. 1, 2, location 2), the Upper Silurian Hardwood Mountain Formation (Boucot 1961, 1969) nonconformably overlies the Upper Ordovician – Lower Silurian Attean Pluton (Fig. 3a), and is unconformably overlain by the Hobbstown Formation and Seboomook Group (Boucot and Heath 1969). The Salinic disturbance, representing a break in the stratigraphic record extending from Middle Ludlow to Lower Gedinnian (Lockhovian), was defined on the basis of the higher unconformity (Boucot 1962). In the Presque Isle area (Figs. 1, 2, location 9), the Salinic is represented by a probable disconformity separating the Perham and Lower Devonian Dockendorff Groups (Boucot et al. 1964).
In the region north of Katahdin (Fig. 1, north of location 4), Hibbard (1994) mapped pre-Acadian sinistral, southeast-vergent oblique reverse faults that place Ordovician over Silurian strata. Richter and Roy (1974) documented prehnite– and pumpellyite–actinolite facies metamorphism in Lower Silurian rocks of northern Maine, which appears to be absent in the overlying Early Devonian rocks.
Along the international border between New Hampshire and Quebec (Fig. 1, west of location 2), the Pridoli–Pragian Frontenac Formation (Lavoie and Asselin 2004) consists of arkosic turbidites infested with gabbro sills and dikes, indicating Silurian–Devonian extension in this area.
Lower Silurian Quimby – Attean – Point aux Trembles (QAP) arc
The Point aux Trembles Formation in Quebec (David and Gariépy 1990; Figs. 1, 2, location 1) and correlatives in the Chaleur Group of New Brunswick (Wilson et al. 2008) comprise volcanic rocks of Llandovery C3–C4 age (436–431 Ma) and arc affinity. In Quebec, these strata overlie rocks accreted to Laurentia during the Taconic orogeny and are buried beneath a thick sequence of Devonian turbidites (Seboomook Group and correlatives). In New England, igneous rocks of Late Ordovician through Early Silurian age and arc affinity are present in the Quimby Formation (conformable on black shales of the Partridge Formation, in turn conformable on the Ammonoosuc Volcanics) and related Highlandcroft and Oliverian Plutonic Suites (Moench and Aleinikoff 2003; Dorais et al. 2008). Bimodal volcanic rocks of the Quimby are 443 ± 4 Ma, and the plutons range from 456 to 435 Ma. The volcanic rocks and post-445 Ma plutons correlate with the third and final phase of the Laurentian Notre Dame arc in Newfoundland (Whalen et al. 2006; van Staal et al. 2007) and closure of the Tetagouche back-arc basin.
Southeast of the Dog Bay – Liberty Line, the Merrimack–Fredericton trough strata record a foredeep setting, developed on the southeastern passive margin (Cookson Group) of the Tetagouche Basin. Whereas several inliers hosting Ganderian rocks occur northwest of the Dog Bay – Liberty Line, the principal block of Ganderian crust, referred to as the St. Croix–Ellsworth block, lies southeast of the Fredericton trough. The correlation of St. Croix and Ellsworth rocks with the Gander Zone in adjacent New Brunswick and Newfoundland, as opposed to the Avalon Zone, is based on contrasts with both the Cambro–Ordovician Avalon platform sequence and subtle but distinct differences in their Neoproterozoic basements (e.g., van Staal et al. 2004; Fyffe et al. 2009).
Northwest of the Dog Bay – Liberty Line, Lower Silurian rocks record elements of a short-lived southeast-facing arc-trench system built on the active margin of composite Laurentia above a northwest-dipping subduction zone. Arc rocks are present in the Connecticut Valley – Gaspe trough along its northwestern (Point aux Trembles Formation) and southeastern (e.g., Chaleur Group, Attean and related plutons, and Quimby Formation) margins. The Central Maine – Matapedia Trough occupies a forearc position. The Central Maine and Fredericton troughs are, over large distances, separated by Cambro-Ordovician rocks of the Liberty–Orrington and Miramichi inliers, which constitute exposed parts of the Brunswick subduction complex (van Staal 1994; van Staal et al. 2008).
Where Ordovician inliers are absent immediately northwest of the DBL, a long-standing debate centres around the relationships between Vassalboro Group strata of the CMMT and adjacent strata of the Fredericton trough. Both are calcareous turbidites, which frustrates delineation of the boundary between the troughs. In southern Maine, the Berwick Formation was recently removed from the Merrimack Group partly because of its resemblance to the Vassalboro (Hussey et al. 2007). East of Bangor, the Central Maine strata (Vassalboro) have also been compared to the Bucksport Formation. Recent work in this area, however, suggests that the Vassalboro and Bucksport Formations may not be equivalent. The Bucksport is more calcareous and carries an extra deformation (S. Pollock, personal communication, 2005), implying that the Vassalboro may be younger than the presumed Early Silurian Bucksport, or accumulated in a different tectonic setting.
An alternative explanation for the lithologic resemblance of turbidites in the Fredericton and southeastern Central Maine troughs invokes the modern analog of Timor, located in Indonesia. Timor is a large island situated on the outer forearc or top of the subduction complex. It has been recently uplifted after the leading edge of Australia’s passive continental margin slid into the Java Trench (Audley-Charles et al. 1990); the uplifted forearc block supplied sediments to both adjacent syncollisional basins, including the still active “A-subduction”-related trench. We suggest that such a mechanism can also explain the similarity of sandstones in the Fredericton trough (foredeep) and southeastern margin of the Central Maine trough (distal forearc). Forearc sediments can also bypass the forearc basin if there is no outer arc high (Dickinson and Seeley 1979).
Comparison of the Vassalboro and Chaleur Groups (Fig. 2, columns 8 and 10) suggests that erosional hiatuses in the Chaleur Group, marked by angular unconformities, may be represented by clastic pulses in the Vassalboro Group (Mayflower Hill and Hutchins Corner Formations). In addition, the Waterville Formation, stratigraphically below the Mayflower Hill and above the Hutchins Corner, contains a ribbon limestone member in its central part, which may correspond with an erosion minimum in the Chaleur area.
The igneous rocks present in the Pridoli – Lower Devonian Frontenac Formation of the Connecticut Valley – Gaspe trough, and by implication the associated extension, can be tied to slab break-off at the end of the Salinic collision (Whalen et al. 2006; Wilson et al. 2008). Hence, the observed Late Silurian uplift in northern Maine, where Boucot (1962, 1969) first recognized the Salinic disturbance, is probably also related to these final stages of the Salinic orogenic cycle. Charlton (1991) showed how slab breakoff in the Banda arc-Australia collision on Timor also led to adjacent regions of uplift and extension.
Silurian strata of the Coastal Volcanic Belt are probably unrelated to the tectonic system described above. Instead, they are linked to a separate arc built upon Ganderia’s trailing edge related to northwest-dipping subduction of the Avalon plate that lay farther to the southeast (Barr et al. 2002; van Staal et al. 2009).
Contrasts across the Dog Bay – Liberty Line
In general, Upper Ordovician strata display the greatest contrasts across the Dog Bay – Liberty Line, interpreted to reflect a wide Tetagouche–Exploits basin, and these contrasts decrease up-section, becoming nonexistent in Middle to Upper Devonian strata.
Middle to Upper Devonian terrestrial sediments deposited in post-orogenic basins (e.g., Mapleton and Perry Formations) are similar, as is expected for the post-Acadian cover sequence. Lower Devonian strata are only present to the northwest of the line, as rocks previously thought to be Lower Devonian southeast of the line are now known to be Upper Silurian.
In the Central Maine trough, Ludlow black shales are present from its northwestern margin southeastward through its medial part. Farther southeast, Ludlow plutons (424–419 Ma) intrude the Coastal Volcanic Belt, where they are comagmatic with volcanic rocks; they also intrude the St. Croix–Ellsworth, Fredericton, and Liberty–Orrington–Miramichi belts, and stitch the Dog Bay Line (Tucker et al. 2001). If these post-collisional plutons were associated with uplift, whether because of crustal thickening or slab breakoff, then a shoreline must have existed between them and the depocenter of the Smalls Falls Formation. The upper part of the Vassalboro Group, or Ludlow equivalent, may represent a sand-rich facies deposited between the uplifted southeastern source area and the Smalls Falls side of the basin, which was probably still fed from the northwest at this time (Bradley et al. 2000).
The relative maturity of the Clough Quartzite and Perry Mountain Formation suggested a passive margin setting for these sediments (e.g., Bradley 1983) but a passive setting is difficult to reconcile with the occurrence, albeit minor, of volcanic rocks (Moench and Pankiwskyj 1988). Alternatively, they accumulated in a syncollisional forearc position with respect to older arc-affinity rocks of the Point aux Tremble and Quimby Formations and Attean and related plutons (QPA on Fig. 5b), which correlate with the third and final phase of the Notre Dame arc elsewhere (van Staal and de Roo 1995; Wilson et al. 2004, 2008; Whalen et al. 2006; NDA on Fig. 5a refers to an earlier phase of the Notre Dame arc). Hence, a more likely scenario is that arc detritus from the extinct and unroofed northwesterly arc(s) matured and accumulated in a relatively stable setting.
Lower Silurian strata northwest of the Dog Bay Line reveal a transition from northwesterly arc through proximal forearc to distal forearc (e.g., slope basins). Igneous-rich Rangeley and Frenchville conglomerates are the same age as a variety of arc-affinity igneous rocks to the northwest. Grain size becomes finer in sediments toward the southeast, at least within the northwestern margin through medial part of the CMMT.
Along the Dog Bay Line, calcareous strata may have been derived from a subduction complex fringed with reefs (Dickinson and Seeley 1979; Audley-Charles et al. 1990). If the Vassalboro Group is equivalent to the Berwick Formation, as has been hypothesized based on lithologic similarities, then these strata may reflect a ponded or sloped forearc (Dickinson and Seeley 1979) in which some of the sediment (Berwick) bypassed the slope basins above the subduction complex to be deposited into the trench.
Paleocurrent data from the Kittery Formation of the Merrimack trough indicate an eastern provenance (Hussey et al. 2010), possibly from Ganderian basement uplifted on extensional faults in a foreland setting.
The most significant contrast across the Dog Bay – Liberty Line is displayed within the pre-Silurian basements of the Central Maine and Fredericton troughs. Southeastward, the Fredericton trough overlies Caradoc quartzites (Kendall Mountain and Harts Neck Formations), interpreted as a passive margin sequence. To the northwest, the Central Maine trough overlies a variety of marine volcanic rocks, formed within the Arenig–Caradoc Tetagouche back-arc basin and subsequently tectonically assembled into the Brunswick subduction complex during the Late Ordovician – Early Silurian closure of this basin. Llandeilo–Caradoc strata display the continent–ocean polarity best. In the Saint Croix (subFredericton trough) basement, orthoquartzites indicate proximity to a slowly subsiding continental source. To the northwest, across the width of the Exploits Subzone, black shales and cherts indicate pelagic oceanic conditions far removed from any exposed continental source area.
Comparisons with Newfoundland
Southeast of the Dog Bay – Liberty Line, Merrimack–Fredericton trough strata are equivalent to the Indian Islands Group in Newfoundland (Table 1). The Ghent member (black phyllite) of the Appleton Ridge Formation may be a New England example of the widespread Caradoc shales that are present all across Newfoundland’s Exploits Subzone, including the Indian Islands belt. The Coastal Volcanic Belt correlates with the La Poile Group of southern Newfoundland.
Along the northwestern margin of the Central Maine trough, the Quimby through Rangeley Formations correlate, respectively, with the Sansom graywackes and Goldson conglomerates of the Badger Group, north–central Newfoundland (Table 1; Williams et al. 1993). Possible olistostromes in the Rangeley Formation may correlate with Silurian mélanges of north–central Newfoundland interpreted to have formed along southeast-vergent, syn-sedimentary thrusts (Reusch 1987) in equivalent rocks immediately southeast of the Red Indian Line. In Newfoundland, tightly folded rocks of the Llandovery Badger Group are unconformably overlain by gently folded rocks of the Wenlock Botwood Group (van der Pluijm et al. 1993). The Upper Silurian volcanic rocks present within and to the northwest of the Central Maine – Matapedia trough (e.g., West Branch, The Forks), and (or) the sparse volcanic rocks in the Perry Mountain Formation (Moench and Pankiwskyj 1988), may share a common genesis with the lower volcanic part of the Botwood Group (Williams et al. 1993). Middle to Upper Silurian siliciclastic strata (e.g., Clough Quartzite, Perry Mountain Formation) may correlate with the upper sedimentary part of the Botwood Group but include a greater proportion of marine strata. Moench and Pankiwskyj (1988) show an unconformity near the top of the Rangeley Formation that is in approximately the same position as the Badger–Botwood unconformity.
Differences with previous models
The tectonic model presented here agrees closely with the scenario of Hussey et al. (2010), differing principally in our equating their Boothbay thrust with the actual Salinic suture. Tucker et al. (2001) and Tremblay and Pinet (2005) proposed widespread Silurian extension, which they attribute to post-Taconic orogenic collapse, whereas we attribute the extension to an upper plate setting above a retreating Silurian subduction zone, the supporting evidence having been presented previously (van Staal and de Roo 1995; van Staal et al. 2008). Many workers have viewed pre-Acadian deformation in the region as Taconic, but the effects of the main Taconic Laurentia-arc collision (Taconic 2 of van Staal et al. 2009) are restricted to northwest of the Red Indian Line. Subsequent orogenesis related to arc–arc collision along this Late Ordovician suture (Taconic 3 of van Staal et al. 2009), which likely caused the Taconic clastic wedge to reach its maximum westward extent across Laurentia in the latest Ordovician (Rodgers 1971), spatially overlapped Salinic orogenesis. However, southeastward of the Red Indian Line, conformable Ordovician through Silurian sequences are common, thus confirming the absence of penetrative Taconic orogenesis here. Perhaps the major difference between several previous models and the current one involves the polarity of Middle Paleozoic subduction. For example, Bradley (1983), Robinson et al. (1998), and Moench and Aleinikoff (2003) all invoked southeast-dipping subduction to account for the Coastal Volcanic Belt, whereas van Staal et al. (2009) present multiple lines of evidence indicating that the passive Gander and Avalon margins subducted northwestward beneath an active Laurentian margin. Northwest-vergent Acadian deformation occurred in a back-arc setting analogous to that of the eastern Andes.
Beginning in the Early Ordovician, the Tetagouche back-arc basin opened and became a wide marginal oceanic basin. Late Ordovician (Caradoc) black shales and cherts blanketed the back-arc basin, while orthoquartzites accumulated on its southeastern passive margin (Fig. 5a). After the late Caradoc collision between the Ammonoosuc–Popologan arc (APA) and the Notre Dame arc (NDA) on the leading edge of Laurentia, spreading stopped and the basin was inverted (van Staal et al. 1998; Zagorevski et al. 2008).
The arc–arc collision forced subduction to step to the southeast (Mueller and Phillips 1991) and consume oceanic lithosphere of the Tetagouche–Exploits basin during the latest Ordovician through Early Silurian (Fig. 5b). This lithosphere was likely as much as 15–20 million years old, and therefore sufficiently dense, in part, to subduct. The Brunswick subduction complex was initiated and grew during this interval. Detritus from the Quimby–Point aux Trembles arc (QPA) spilled into the forearc region (Central Maine – Matapedia trough) from the northwest. The great width of the forearc basin suggests a shallowly dipping slab, which is to be expected in view of the youth and relatively high buoyancy of most of the Tetagouche oceanic lithosphere. Llandoverian conglomerates accumulated along the northwest margin of the CMMT in New England (Rangeley Formation) and in Newfoundland (Badger Group), where they are linked to southeast-vergent thrusts (Reusch 1987). The upper plate experienced episodes of shortening, recorded by angular unconformities, and also extension (Wilson et al. 2004; Tremblay and Pinet 2005) induced by slab rollback (Royden 1993) and (or) stepping back of the subduction zone following the accretion of buoyant blocks distributed within the Tetagouche–Exploits basin (van Staal et al. 2008).
By the end of the Llandovery, the Tetagouche–Exploits basin was closed (Fig. 5c) and the Quimby – Point aux Trembles arc (QPA) became extinct. Mature quartz-rich sandstones of the Clough and Perry Mountain formations were deposited as the QPA shut down and the medial New England basement cooled and subsided. The Fredericton Trough fits the model of a foredeep (trench to foreland basin) developed on the northwest-facing passive margin that was subducted beneath the Liberty–Orrington–Miramichi inlier (remnants of the Brunswick subduction complex) to the northwest. The uplifted and unroofed Brunswick subduction complex (van Staal et al. 2008) shed immature sands in all directions (e.g., Berwick Formation). Possibly, late Llandovery Silurian conglomerates at the base of the Coastal Volcanic Belt mark normal faults formed in the forebulge region of the descending slab, although an alternative explanation invokes back-arc basin related extension associated with subduction of the Avalonian plate.
The region of collision, centered on the Liberty–Orrington–Miramichi, Fredericton, and St. Croix belts, was intruded during the Late Silurian by the Pocomoonshine and related plutons (Fig. 5d). Significantly, Late Silurian plutons are rare or absent in the Central Maine trough (Tucker et al. 2001), where black shales were deposited (Smalls Falls Formation). A mechanism for the black shale deposition invokes isolation of the forearc basin as a result of collision-induced uplift of the adjacent, partly subjacent subduction/collision complex (Dickinson and Seeley 1979). The Central Maine – Matapedia forearc experienced a general shallowing, due to a combination of basin infilling and uplift. In this area, the Salinic orogeny is manifested in several ways, ranging from angular unconformities (van Staal et al. 2009) to sinistral-oblique thrusts (Hibbard 1994). Minor Late Silurian within-plate, bimodal volcanism (West Branch and The Forks Formations and equivalents in the Chaleur Group), approximately coeval with the coastal magmatism, is related to slab break-off and asthenospheric upwelling beneath the uplifted forearc block. A location within the back-arc region of the Coastal Volcanic Belt may also have contributed to high heat flow.
The entire region, including the Central Maine and Fredericton troughs, became part of a northwest-migrating fold-and-thrust belt by Early Devonian time (Acadian orogeny) (Fig. 5e). We prefer a back-arc thrust setting above a northwest-dipping Avalon slab based on the relative positions of back-arc and arc rocks in the Coastal Volcanic Belt, high-pressure rocks in the forearc (White et al. 2006), northwest-dipping seismic reflections (van der Velden et al. 2004), and evidence for a passive margin on Avalon rather than an active one (van Staal 2005; Waldron et al. 1996).
Following intrusion of Middle Devonian plutons (e.g., Lucerne), minor dextral faulting affected the region. The Norumbega Fault dextrally offsets the Dog Bay – Liberty Line several tens of kilometres (Fig. 1). It seems highly likely that the location of the Norumbega Fault was largely controlled by the earlier history of tectonism along the Dog Bay – Liberty Line. Most Devonian plutons in Maine are undeformed, but some are strongly deformed in the vicinity of the line (Fig. 1). Other sutures, such as the Baie Verte Line and Red Indian Line, commonly display a prolonged history of movement that continues long after the suturing event (Zagorevski et al. 2007). Finally, minor vertical offsets of the Moho across the Norumbega Fault have been attributed to Atlantic rifting during the Mesozoic.
Framework for future studies
The present study provides a framework for several avenues of future research. Continued analysis of Central Maine – Matapedia sediments and stratigraphic relationships is warranted to evaluate and refine the proposed forearc basin model. In particular, very little is known about their compositional variability in space and time, and comparisons can be made with modern examples (e.g., Dickinson 1982). The fossil-rich Upper Silurian sedimentary rocks that crop out northwest of the CMMT in northern Maine are ripe to be evaluated in terms of slab breakoff-related uplift. Southeast of the DBL, early structures in the MFT, where not obscured by later dextral shear, warrant scrutiny. Finally, the timing and mechanisms of deformations in the St. Croix–Ellsworth block are currently not well constrained but are highly relevant to understanding the role of these rocks during the critical Tremadoc–Ludlow interval.
The first author is thankful for a sabbatical leave during Fall 2010. Several of the observations were made while funded by a National Science Foundation grant administered by the Maine Geological Survey during 1992–1993. D. Reusch is indebted to Dr. Harold Williams for support during the spring of 1981 of a project that involved compiling a Silurian lithofacies map, based on stratigraphic columns from more than 50 locations, whose objective was to decipher the region’s tectonic evolution following the Taconic orogeny. Alex Zagorevski, Dyk Eusden, and Alain Tremblay provided constructive reviews. This is Geological Survey of Canada publication 20100257.
- Received October 7, 2010.
- Accepted April 22, 2011.
- Published on the NRC Research Press Web site at http://cjes.nrc.ca on November 24, 2011.
- Published by NRC Research Press