Timing and kinematics of deformation in the northwest Adirondack Lowlands, New York State: Implications for terrane relationships in the southern Grenville Province

Multiple orogenic events constructed the ca. 1700–980 Ma Grenville Province (e.g., Wynne-Edwards, 1972; Davidson, 1995; Rivers, 2008; McLelland et al., 2010a), and not all events affected all parts of the province equally (Fig. 1A). Although terminology and details vary among sources, the main orogenic events that constructed the province are: the Elzevirian orogeny (ca. 1245–1220 Ma), the Shawinigan orogeny (ca. 1190–1140 Ma), emplacement of anorthosite-mangerite-charnockite-granite magmas (AMCG suite; ca. 1160–1140), and the Ottawan (ca. 1090–1050 Ma) and Rigolet (ca. 1010–980 Ma) phases of the Grenvillian orogeny (Rivers, 2008; McLelland et al., 2010a).

Like many orogens, the Grenville Province is subdivided into crustal blocks with a unique set of characteristics, such as similar lithologies that have common metamorphic, deformation, and plutonic histories. These crustal blocks, typically referred to as “terranes,” produce the architecture of the orogen. Some terranes are subdivided into domains, in which the geology of the crustal block of the domain is not unique enough to suggest a different terrane designation. Separating these terranes, and likely the domains, are shear zones that juxtapose crustal blocks with different geology. The term exotic terrane refers to a specific kind of terrane that originated far from its current position within an orogen, such as an island arc. The terrane-bounding shear zone for exotic terranes is therefore a suture, but not all terrane-bounding shear zones are sutures.

Detailed study of the deformation within terranes and the bounding shear zones can shed light on the processes of accretionary tectonics, regional metamorphism, plutonism, and orogenic collapse, and why these processes repeatedly occur across the orogen, often only affecting some terranes but not others. The Grenville Province is particularly important for tectonic study because its age and erosion level expose originally deep crustal rocks, whereas in actively developing mountain belts, these types of rocks and the deep crustal processes at work within the mountain belt typically can only be inferred.

The Adirondacks (Fig. 1A), composed of the Adirondack Highlands and Adirondack Lowlands (Fig. 1B), form an outlier of the Grenville Province (e.g., McLelland et al., 2010a). The Adirondack Lowlands, situated between the Frontenac terrane, of the main Grenville Province, and the Adirondacks Highlands, are one region in the Grenville Province where the timing and kinematics of deformation(s) are not well constrained (Fig. 1B). Nor has it been clear which orogenic event(s) are responsible for the multiple folding phases and numerous brittle, brittle-ductile, and ductile shear zones that dissect the Adirondack Lowlands.

This work integrates detailed structural analysis with U-Pb zircon geochronology to understand the timing and significance of Lowlands deformation, metamorphism, and plutonism, with respect to specific orogenic events. Further, the results will contribute to a better understanding of the relationship between the Lowlands and the adjacent Frontenac terrane, in particular, whether they are domains of the same terrane or are different terranes—exotic or otherwise.

The rocks of the Lowlands are quite varied, but they can be categorized into one of a handful of identifiable units. Although the origin of many of these units is contested, the generally accepted stratigraphy or tectonostratigraphy has been described by Carl et al. (1990) and consists of the following units from lowest to highest: Hyde School Gneiss, Lower Marble, Popple Hill Gneiss, and Upper Marble (Fig. 2).

The Hyde School Gneiss, an alkaskitic to tonalitic gneiss, originally interpreted to be rhyolitic ash-flow tuffs forming the basal stratigraphic unit of the Lowlands (Carl and Van Diver, 1975), is argued by McLelland et al. (1992) to be plutonic in origin, and if so, has an intrusive age of 1172 ± 5 Ma (Wasteneys et al., 1999). The Lower Marble consists of graphite, phlogopite, and brown tourmaline-bearing calcite marbles with calcsilicate layers. North of the Beaver Creek fault (Fig. 2), the Lower Marble tends to be less rich in carbonate marbles and more rich in other rock types including quartzite. This part of the Lower Marble is termed the Black Lake member of the Lower Marble (Carl et al., 1990; Hudson, 1994). The Popple Hill Gneiss is a locally migmatitic gray quartz-biotite-feldspar gneiss interlayered with amphibolites and is interpreted by some to be a metadacite (Carl, 1988; Miller et al., 2010), but as a metapelite by others (Heumann et al., 2006; Chiarenzelli et al., 2010). Leucosome formation within the Popple Hill Gneiss occurred at 1170–1160 Ma (Heumann et al., 2006). The Upper Marble consists of predominately dolomitic marble and contains domical features interpreted to be preserved stromatolites (Carl et al., 1990).

Rocks of undisputed intrusive igneous origin that intrude the above metasediments and possible metavolcanics are: the Antwerp-Rossie suite, the Hermon Gneiss, the Rockport Granite, and AMCG suite rocks (Fig. 2). The Antwerp-Rossie suite, mostly a coarse-grained granodioritic gneiss, dated at 1203 ± 14 Ma, is considered to be calc-alkaline plutons formed by subduction prior to Shawinigan deformation (Chiarenzelli et al., 2010). The Hermon Gneiss is a porphyritic granitic gneiss (Buddington, 1939), may be related to the Antwerp-Rossie suite (Carl and deLorraine, 1997; Chiarenzelli et al., 2010), and has an intrusive age of 1182 ± 8 Ma (Heumann et al., 2006). Rockport Granite is typically a coarse-grained to pegmatitic pink leucogranite (Carl and deLorraine, 1997). Timing of intrusion of the Rockport Granite has been constrained to 1173 ± 4 Ma (van Breemen and Davidson, 1988) and 1172 ± 5 Ma (Wasteneys et al., 1999). The AMCG suite rocks are relatively scarce in the Lowlands with the Diana syenite (e.g., Hargraves, 1969) and the Edwardsville syenite (Buddington, 1934; McLelland et al., 1993) being the best-known examples and have intrusive ages of 1164 ± 11 Ma (Hamilton et al., 2004) and 1164 ± 4 Ma (McLelland et al., 1993), respectively (Fig. 2). The AMCG suite developed during mild postorogenic extension where ponded gabbroic mantle melts at the crust-mantle boundary could crystallize anorthosite and related mafic intrusions (A) and mangerite, charnockite, and granite melts (MCG) developed through crustal melting (e.g., McLelland et al., 2004; McLelland et al., 2010b; Regan et al., 2011).

Metamorphic conditions of the Lowlands are generally considered to be upper-amphibolite facies. Kitchen and Valley (1995) used carbon-isotope thermometry to show that metamorphic temperatures varied somewhat across the Lowlands. In the center of the Lowlands, temperatures were ∼640 ± 30 °C and increased to the southeast to ∼680 ± 20 °C, as well as to the northwest to 670 ± 25 °C. Metamorphic pressures were ∼6.5–7.0 kbar throughout the Lowlands (Bohlen et al., 1985).

It has long been recognized that the Adirondack Lowlands have been multiply folded (e.g., Brown, 1936; Brown and Engel, 1956; Foose and Brown, 1976; Foose and Carl, 1977; Wiener, 1983; Wiener et al., 1984). Most identify three to four phases of folding. Some argue that these were discrete events or pulses of deformation; others suggest that some of these phases represent “continuous” folding resulting from a single stress system.

The earliest folding phase is not everywhere recognized but is described by Wiener (1983) as isoclinal, northeast trending, and overturned to the southeast, with axes plunging moderately to steeply north-northwest. This earliest phase of folding is characterized by an axial-plane compositional foliation that Weiner described as comprising the regional foliation. Lineation associated with this folding event is parallel to fold axes.

Most workers are in general agreement regarding the characteristics of phases two and three in the four-phase folding model (or phases one and two in the three-phase model; Foose and Carl, 1977). Based on the four-phase model, phase two is tight to isoclinal with axial planes that strike northeast and dip northwest. These folds have an axial planar foliation and an axis-parallel mineral lineation. Second-phase fold axial orientations are similar to, but are more variable than those of the first-phase folds. Third-phase folds are open to tight, have fanned axial planes that strike northeast and dip northwest, and fold axes are variable but similar to second-phase fold axes. Type 3, hook-style interference patterns (Ramsay, 1967), are recognized as forming between phase one and two and phase two and three folds (Wiener, 1983).

Fourth-phase folds (or third-phase in the three-phase model) are upright and open, with steep, northwest-striking axial planes, and Wiener (1983) identified an axial-planar cleavage, microfaults, and mineral and intersection lineation parallel to the axes. The interference of these “cross” folds with the previous phases is thought (e.g., Foose and Carl, 1977) to create a Type 1, dome-and-basin interference pattern (Ramsay, 1967).

These northwest-trending cross folds (F₃ of Foose and Carl, 1977; F₄ of Wiener et al., 1984) are enigmatic. What appear to be domes of the Hyde School Gneiss (Fig. 2) seem to attest to their existence, but no intervening basins, with circular map patterns, have been described within the Lowlands. Wiener (1983) identified an axial-planar structure apparently associated with these cross folds. However, this could be a local phenomenon associated with proximity to the Carthage-Colton mylonite zone (CCMZ; Figs. 1B and 2), the boundary between the Adirondack Lowlands and Highlands (Geraghty et al., 1981; Mezger et al., 1992; Streepey et al., 2001; Baird and MacDonald, 2004; Johnson et al., 2004; Selleck et al., 2005) where Wiener's (1983) detailed work was focused. However, Tewkesbury (2010, personal commun.) states that no axial-planar structures or parasitic structures are associated with this last phase of folding. Further, the cross folds have been difficult to explain because they traverse across the Carthage-Colton mylonite zone and are reportedly present in both the Highlands and Lowlands (Wiener et al., 1984), though evidence for them has not been found along the Carthage-Colton mylonite zone in central portions of the Diana syenite (Baird and MacDonald, 2004). Tewkesbury (1993) suggested that the apparent dome-and-basin pattern in the Lowlands could have been created by sheath folding, causing highly curved hinges as the result of regional ductile shear during the second-phase of the four-phase model.

The Lowlands are dissected by numerous northeast-southwest–trending, steeply dipping faults and ductile shear zones (Fig. 2), including the Noname ductile shear zone (see below), Pleasant Lake fault, Beaver Creek fault, and Hailsboro ductile deformation zone (deLorraine and Carl, 1993). Some have argued that at least some of these zones create a tectonostratigraphy and possibly juxtapose originally different terranes (Wiener et al., 1984; McLelland, 1991; Grant, 1993). However, detailed mapping by deLorraine and Carl (e.g., 1993) suggests that, with the possible exception of the Black Lake fault (Fig. 1B), the stratigraphy is traceable across the fault-bounded panels and/or domains of the Lowlands (deLorraine and Carl, 1993; Carl and deLorraine, 1997).

Many of these faults have both ductile and brittle characteristics and likely have a history of reactivation. Brown (1989) argued that early fault movement was dominantly strike-slip and was characterized by granite and pegmatite intrusions. Tewkesbury (1993) advocated early regional shear along a deep, subhorizontal shear zone (likely distinct from the above-named faults), and also cited evidence for late dextral shear along the steeply dipping, northeast-striking faults.

U-Pb sphene (closure temperature of ∼600–650 °C) ages of 1156–1103 Ma within the Lowlands (Mezger et al., 1991; Mezger et al., 1993), combined with ⁴⁰Ar-³⁹Ar for hornblende (closure temperature of ∼500 °C) and biotite (closure temperature of ∼300 °C) of generally 1100–1000 Ma and 1000–900 Ma, respectively (Dahl et al., 2004), indicate cooling from peak metamorphism associated with the Shawinigan orogeny (ca. 1170–1130 Ma; Mezger et al., 1992). However, there is scattered evidence that the Lowlands underwent some metamorphism during the Ottawan and perhaps Rigolet orogenies (e.g., Streepey et al., 2003; Dahl et al., 2005; Hudson et al., 2008; Selleck, 2008). Yet, Lowlands temperatures must not have risen above the U-Pb closure temperature of sphene, and the ⁴⁰Ar-³⁹Ar hornblende ages are not reset either (Streepey et al., 2000; Dahl et al., 2004).

Timing of deformation in the Lowlands has never been directly studied or constrained. Both the ca. 1203 Ma Antwerp-Rossie suite (Chiarenzelli et al., 2010) and the ca. 1182 Ma Hermon Gneiss (Heumann et al., 2006) possess deformational fabrics, and paragneiss xenoliths found in the Antwerp-Rossie suite also have gneissic layering (Carl and deLorraine, 1997), thus suggesting some deformation occurred at or prior to ca. 1203 Ma as well as at or after ca. 1182 Ma. Heumann et al. (2006) indicate that some deformation is associated with leucosome formation in the Popple Hill Gneiss during the Shawinigan orogeny at ca. 1180–1160 Ma. Also, if the Hyde School Gneiss is indeed plutonic in origin, it is believed to be synkinematic in nature, so its intrusive age of ca. 1172 Ma likewise constrains the timing of some deformation in the Lowlands (McLelland et al., 1992). The ca. 1164 Ma (McLelland et al., 1993) Edwardsville syenite is reported to contain a deformational fabric (Buddington, 1934) indicating the timing of later Shawinigan deformation or perhaps deformation of the Ottawan or Rigolet orogenic phases. Further, Wiener (1983) states the ca. 1164 Ma (Hamilton et al., 2004) Diana syenite intruded during F₂ (four-phase model), indicating that at least F₂ resulted from the Shawinigan orogeny. Therefore, there appears to at least be a strong Shawinigan deformational signature in the Lowlands. However, the timing of deformation with respect to specific structures and kinematics is hitherto unknown.

The thermochonology constraints on Lowlands temperature conditions, and what is known regarding the timing of deformation, do not necessarily exclude the presence of significant deformation resulting from Ottawan or Rigolet orogenesis. All plutons for which there are reliable dates are associated with the Shawinigan orogeny, and all are deformed, so lack of Ottawan and Rigolet orogeny deformation cannot be conclusively confirmed. Further, Ottawan-aged deformation resulting from compressional and extensional stresses has been reported along the Carthage-Colton mylonite zone (Streepey et al., 2001; Johnson et al., 2004; Selleck et al., 2005). Also, it is thought that the Sylvia Lake Syncline within the Upper Marble is a relatively recent structure and is tentatively considered to be a result of Ottawan deformation (W. deLorraine, 2010, personal commun.).

The Adirondack Highlands, in contrast to the Lowlands, are dominated by metaigneous rocks, although metasedimentary rocks of similar character to the Lowlands are found in the Highlands. Oldest recognized metaigneous rocks in the Highlands are the 1350–1250 Ma arc affinity tonalites and granodiorites of the Dysart–Mount Holly suite (Fig. 1B) found in the southern and eastern Highlands (McLelland et al., 1996; McLelland et al., 2010a). Though scarce in the Lowlands, the ca. 1160–1155 Ma AMCG rocks are widespread (McLelland et al., 2004; McLelland et al., 2010b; Regan et al., 2011). Other metamorphosed intrusives found in the Highlands include a 1190–1100 Ma A-type hornblende granite called the Hawkeye granite suite (McLelland et al., 2001), and the mostly postorogenic 1060–1037 Ma aplitic Lyon Mountain Granite (McLelland et al., 2001; Selleck et al., 2005).

Metamorphism in the Highlands is generally considered to be granulite facies in a fluid-absent environment (Valley et al., 1990). Carbon-isotope thermometry reveals metamorphic temperatures are highest in the central portions of the Highlands at ∼780 °C and decrease outwardly to ∼670 °C (Kitchen and Valley, 1995). Metamorphic pressures within the Highlands range from ∼6.5 to 8.0 kbar (Bohlen et al., 1985).

Multiple phases of folding are the dominant deformation style in the Highlands with fold geometries tending to vary across the large area of the terrane (Wiener et al., 1984). The earliest recognized F₁ folds can only be identified in hinges of F₂ folds where an earlier fabric and fold set can be observed to be folded by a younger fold set. The F₂ folds tend to be isoclinal, have a strong axial-planar foliation, and are large scale, being observed in map patterns. The F₃ folds are open and upright, coaxial with F₂ folds, and have a strong hinge-line–parallel mineral lineation developed due to the intersection of axial-planar foliations of F₃ and all prior folding phases. The F₄ folds are thought to extend across the Carthage-Colton mylonite zone (Wiener et al., 1984) and like in the Lowlands, are open, upright, and northwest trending. In the Highlands, F₄ folds have an axial-planar foliation defined by mineral alignment. An F₅ fold generation is very similar in character to F₄ folds including development of an axial-planar foliation, but trends more north-south and can be very tight. The F₄ and F₅ folds develop a Type 1 dome-and-basin fold-interference pattern on a Type 3 fold-interference pattern developed from interference of F₂ and F₃ folds (Wiener et al., 1984). Shear zones, of the nature described in the Lowlands, generally have not been identified in the Highlands, but two east-west–trending Ottawan to Rigolet-aged sinistral ductile shear zones occur across the southern half of the Highlands (Gates et al., 2004).

The Highlands have a significantly different cooling history than that of the Lowlands as high-grade Ottawan-aged metamorphism was imprinted on the Highlands. Evidence for this comes from U-Pb sphene ages ranging from 1050 to 982 Ma (Mezger et al., 1993; Streepey et al., 2000), with ⁴⁰Ar-³⁹Ar hornblende ages of ca. 950 Ma and ⁴⁰Ar-³⁹Ar biotite ages of ca. 940–890 Ma (Streepey et al., 2000).

Deformation in the Highlands has been attributed both to Shawinigan and Ottawan orogenesis. Chiarenzelli et al. (2011a) report strong deformational fabrics in a diopside-bearing quartzite unit truncated by a ca. 1172 Ma granite indicating some deformation occurred in the Highlands at or before ca. 1172 Ma. U-Pb zircon and monazite ages from deformed leucosomes in metapelites demonstrate that two generations of leucosome exist and deformation occurred at 1180–1160 Ma and ca. 1050 Ma, corresponding to Shawinigan and Ottawan orogenesis (Heumann et al., 2006; Bickford et al., 2008). The Hawkeye granite suite (1190–1100 Ma) and AMCG rocks (ca. 1160–1155 Ma) commonly have a tectonic fabric attributed to Ottawan deformation. The Lyon Mountain Granite (1060–1037 Ma) is weakly deformed to undeformed, and its intrusion marks the end of Ottawan deformation (McLelland et al., 2001; McLelland et al., 2004).

In the south-central Highlands, Kusky and Loring (2001) used dikes with various relationships to structures in combination with U-Pb zircon geochronology to constrain the timing of F₁, F₂, and F₃ folding. They found that F₁ occurred between ca. 1172 Ma and ca. 1165 Ma and is attributed to Shawinigan deformation, whereas F₂ and F₃ occurred between ca. 1165 Ma and ca. 1052 Ma. Kusky and Loring (2001) assign F₂ and F₃ to Ottawan deformation; however, it should be noted that if F₂ and/or F₃ occurred at or close to ca. 1165 Ma, they both could be Shawinigan in age.

Like the Adirondacks, metasedimentary rocks are common throughout the Frontenac terrane, including marbles, quartzites, and pelitic gneisses (e.g., Corfu and Easton, 1997; Davidson and van Breemen, 2000). Two plutonic suites occur within the terrane; the first is the 1180–1150 Ma “A”-type syenitic-monzonitic-granitic Frontenac suite, of which the Rockport Granite is a member. The second is the 1090–1065 Ma “A”-type monzonitic to dioritic Skootamatta suite (Davidson and van Breemen, 2000).

Recent carbon-isotope thermometry applied across the Frontenac terrane reveals granulite-facies conditions of 853 ± 75 °C (Tortorello and Peck, 2010), which is higher than past estimates of 650–700 °C for the terrane (Streepey et al., 1997). Streepey et al. (1997) summarized the pressure conditions across the terrane as 7–8 kbar in the eastern half of the terrane, lowering to 6–7 kbar in the western half.

Two coaxial northeast-plunging fold sets dominate the structure of the Frontenac terrane. Wynne-Edwards (1967) suggests that deformation with a significant component of sinistral strike-slip shear was able to produce the sets of coaxial folds. Additionally, stereographic analysis by Wynne-Edwards (1967) suggests a set of cross folds. The Frontenac terrane has a few northeast-trending shear zones (Wynne-Edwards, 1967), including the Perth Road mylonite zone in the central Frontenac terrane (Cosca et al., 1992; Fowlow and Bjornerud, 1994; Davidson and van Breemen, 2000).

U-Pb sphene ages in the Frontenac terrane are 1178–1157 Ma (Mezger et al., 1993), while ⁴⁰Ar-³⁹Ar hornblende ages are 1125–1104 Ma hornblende (Cosca et al., 1992). These ages are interpreted to be generated by cooling from peak metamorphic conditions at 1175–1150 Ma (Mezger et al., 1993) during the Shawinigan orogeny. However, Corriveau and van Breemen (1994), utilizing both zircon and monazite data, suggest metamorphism occurred at 1190–1180 Ma, which would still be appropriate for the Shawinigan orogeny.

In the eastern part of the Frontenac terrane (Fig. 1B), older members of the Frontenac suite (e.g., ca. 1172 Rockport Granite; Wasteneys et al., 1999) are deformed, while younger members (ca. 1164 Ma) are unfoliated, and therefore considered to be undeformed, post-tectonic intrusions (Davidson and van Breemen, 2000). This is supported by the presence of the ca. 1160 Ma undeformed Kingston dike swarm (Davidson, 1995). However, in the centrally located Perth Road shear zone, an 1166 ± 3 Ma monzonite (Marcantonio et al., 1990) is deformed (Fowlow and Bjornerud, 1994). In the western part of the terrane, plutons as young as ca. 1157 Ma are considered synkinematic, possibly associated with the Maberly shear zone. These data support a strong Shawinigan (ca. 1165 Ma) deformational signature for the eastern half of the Frontenac terrane, but in the western portion, deformation history is more complex.

Currently, there is little doubt that the Adirondack Lowlands and Highlands are different terranes given the distinct difference in their plutonic suites and the lack of significant Ottawan-aged metamorphism in the Lowlands. The history of the Carthage-Colton mylonite zone remains somewhat enigmatic, but is thought to either facilitate the Lowlands’ escape from significant Ottawan orogenesis by thrusting the Lowlands over the Highlands, or the Lowlands and Highlands were laterally separate prior to Ottawan orogenesis. They were placed in their current juxtaposition along the Carthage-Colton mylonite zone (e.g., Mezger et al., 1992; Streepey et al., 2001; Johnson et al., 2004; Selleck et al., 2005). Although the crustal blocks are different, some aspects of their geology, e.g., paragneiss protoliths and specific folding phases, may be corollary (Wiener et al., 1984), but additional detailed work beyond the scope of this study is needed to confirm this.

In general, the Adirondack Lowlands and Frontenac terranes appear to be quite comparable, and many (e.g., Carr et al., 2000; Davidson and van Breemen, 2000; Hanmer et al., 2000; Rivers and Corrigan, 2000; Rivers, 2008) consider the Lowlands to be part of the Frontenac terrane, or to be domains within the same terrane. This is supported by the existence of similar metasedimentary rocks and the fact that Rockport Granite is in both regions (Carl and deLorraine, 1997). Both regions are deformed by a very similar folding and shear zone style. Further, the conventional knowledge has been that only Shawinigan deformation and metamorphism affected the Lowlands and eastern Frontenac terrane.

However, the differences in their geology are clear suggesting that they may in fact be distinct terranes. The Kingston dike swarm does not extend to the southeast over the St. Lawrence River (Davidson, 1995), while certain distinctive Adirondack Lowland rock units—the Antwerp-Rossie suite, the Hermon Gneiss, and the Hyde School Gneiss—do not extend to the northwest over the St. Lawrence River (Chiarenzelli et al., 2010). Other differences include the peak metamorphic conditions being ∼1 kbar and 180 °C higher in the Frontenac terrane. But despite having a higher metamorphic temperature, U-Pb sphene ages are ∼30–40 m.y. older on average in the Frontenac terrane than in the Lowlands indicating much more rapid cooling of the Frontenac terrane. However, Dahl et al. (2004) point out that ⁴⁰Ar-³⁹Ar hornblende and biotite ages get older west across the Lowlands, which is attributed to relatively late Lowlands tilting of ∼9°. Sphene data across the Lowlands generally follow the trend of ⁴⁰Ar-³⁹Ar hornblende and biotite ages, such that the westernmost U-Pb sphene ages in the Lowlands are perhaps only 10 m.y. older in the Frontenac terrane.

Another line of evidence suggesting the regions are different terranes is that from the Edwardsville syenite (Fig. 2), or southeast of it, and moving northwest into the Frontenac terrane, the pluton zircon δ¹⁸O is ∼2.5‰ higher (Peck et al., 2004). Also, the Antwerp-Rossie suite plutons adjacent to the St. Lawrence River (Fig. 1) has the oldest Nd model ages and lowest ε_Nd values (Chiarenzelli et al., 2010). These data indicate a geochemical shift in the western portions of the Lowlands.

But if the Frontenac terrane and Adirondack Lowlands are indeed different terranes as suggested by, most notably, a difference in plutonic suites, difference in metamorphic temperatures, pluton zircon δ¹⁸O, and Nd isotope systematics, then a shear zone must be at the boundary between them and be located somewhere just southeast of the St. Lawrence River. It has been proposed that the Black Lake fault, also referred to as the Black Lake or Creek fault/lineament/shear zone (Davidson, 1995; Wallach, 2002; Peck et al., 2004; McLelland et al., 2010a), could be the terrane-bounding shear zone. It is a post-Paleozoic brittle deformation zone running along Black Creek and up the northwest side of Black Lake with unknown Grenvillian significance (Figs. 2 and 3). This structure should not be confused with a structure defined in Wong et al. (2011), termed the Black Lake shear zone, which is described as a broad, diffuse, heterogeneous ductile deformation zone, likely slightly oblique to the Black Lake fault and perhaps located just northwest of it. Wong et al. (2011) suggest that this feature could be the terrane-bounding shear zone, and in fact characterized it as a suture. The Black Lake fault and Wong et al.’s (2011) Black Lake shear zone, when appropriate, will be referred to collectively as the Black Lake structures, given their uncertain tectonic significance (Figs. 1 and 2).

This study is focused on the area immediately southeast of the Black Lake fault (Figs. 2 and 3), an area whose structures are typical of the Lowlands. This contribution aims to improve our knowledge of Lowlands deformation in terms of style, timing, and kinematics, thus adding detail to the comparison with adjacent regions and producing a better understanding of the role of the tectonics events responsible for their geology.

In the northwest Lowlands, within the area of focus (Fig. 3), a migmatitic unit predominates northwest of the Grass Creek fault and consists of quartz-biotite ± clinopyroxene ± garnet ± sillimanite nodular gneiss with ubiquitous coarse-grained granitic leucosomes that are usually tourmaline bearing. Pure to impure quartzite layers are common within the migmatite unit, as well as less common calc-silicate with marble layers. This migmatitic unit is thought to be a metapelite to metapsammite within the Black Lake member of the Lower Marble (Carl et al., 1990; Carl and deLorraine, 1997). Southeast of the Grass Creek fault (Fig. 3), the lithology is predominately a calcite-graphite ± clinopyroxene ± sphene granoblastic marble gneiss and is also likely part of the Black Lake member of the Lower Marble (W. deLorraine, 2010, personal commun.).

The Antwerp-Rossie suite found throughout the area (Chiarenzelli et al., 2010) is typically a fine- to medium-grained, biotite granodiorite–tonalite–quartz diorite. It is weakly foliated or massive and in places contains a net-like granitic vein network. The Split Rock Road body of the Antwerp-Rossie suite (Fig. 3) has an intrusive date of 1203 ± 14 Ma and is reported in Chiarenzelli et al. (2010). Various bodies of the pink, coarse-grained to pegmatitic ± tourmaline Rockport Granite also occur as dikes and sills scattered throughout the migmatite unit, with some occurrences in the marble unit.

Four major structures occur in the study area (Fig. 3). The Alamgin antiform is a prominent ∼2-km-wide fold obvious on topographic maps and aerial photography. Merging with the eastern limb of the Alamgin antiform is the ∼0.5–2-km-wide, northeast-striking, subvertical Noname ductile shear zone. The Noname ductile shear zone is a name modification of this structure identified by Guzowski (1979) and is characterized by consistently oriented, finely laminated, high-strain fabrics and a strong rodding and/or mineral lineation defined by layer boudinage, biotite alignment, and sillimanite nodule alignment. The Noname ductile shear zone extends to the southwest to the brittle and/or ductile Grass Creek fault (Guzowski, 1979), which lies along the contact between the biotite migmatite unit and the marble unit (Fig. 3).

The Black Lake fault, an ∼0.5–1-km-wide brittle structure, borders the northwest part of the field area. No ductile strain gradient has been observed associated with the feature. However, cataclasis, which increases to the northwest with proximity to the Black Lake fault, may have obscured ductile structures associated with the fault. Although brittle structures may have played a role in Grenville orogenesis, they experienced post-Paleozoic reactivation (Wallach, 2002) and are beyond the scope of this study. Therefore, we focus here on the ductile structures.

There is some evidence for an early deformation event (Pre-D₁) that predates the intrusion of the Antwerp-Rossie suite. Carl and deLorraine (1997) described gneissic layering within paragneiss xenoliths in the Antwerp-Rossie suite. In addition, deLorraine has recognized a very early foliation in the southwestern portion of the Lowlands (W. deLorraine, 2010, personal commun.) and our own observations of rare isoclinal intrafolial folds of thin layer-parallel granitic leucosomes (Fig. 4A) and of quartz-rich layers in marble (Fig. 4B) could be evidence of deformation prior to D₁ and likely equivalent to Wiener's (1983) earliest phase of folding. However, what is here defined as D₁ (see below) appears to have been sufficiently penetrative that most evidence of a previous deformation has been rotated and/or eradicated such that it cannot be convincingly distinguished from D₁ structures as defined here.

Gneissic layering (S₁; Fig. 4A) is defined predominately by banding of quartz-rich and quartz-poor layers and the alignment of many granitic leucosomes, commonly boudinaged parallel to S₁ or folded with S₁ parallel to the axial plane (Figs. 4A, 4C, and 4D). Biotite and elongate sillimanite nodules, if present, are also parallel to the gneissic layering. The compositional banding, with respect to the quartz content, is thought to reflect bedding compositional variations in the protoliths. D₁ is interpreted to have rotated bedding (S₀ = S₁) into parallelism with the axial plane of F₁ folded granitic leucosomes (Fig. 4C). Some granitic leucosomes cut across S₁ gneissic layering and are essentially undeformed (Fig. 4E), and tourmaline-bearing granitic material fills tension gash or boudin-neck–like structures oriented perpendicular to S₁ (Fig. 4F). These field relations suggest that leucosome formation occurred throughout and after D₁ deformation. Presence of melt, likely derived in situ (see Heumann et al., 2006), and sillimanite corroborate the regional metamorphic grade being approximately upper-amphibolite facies and was accompanied by deformation.

The S₁ gneissic layering and F₁ folds are folded by a second deformation event, D₂. The F₂ folds resulting from D₂ deformation are distinguished from F₁ folds by their axial planar features with F₂ folds have a variably developed weak schistosity, S₂, parallel to their axial plane (Figs. 4A and 5A). S₂ is defined exclusively by the alignment of biotite oblique to the features defining S₁. The F₂ folds are generally upright and relatively open or exist as small-scale crenulations of S₁. The F₁ and F₂ folds are essentially coaxial, thus producing a Type 3 fold-interference pattern (Ramsay, 1967) with hinge lines and fold axes of both F₁ and F₂ folds moderately plunging to the southwest (Fig. 6A). Ideal Type 3 fold-interference patterns generally are not observed in the migmatite gneiss, but they are well developed in the marble (Fig. 5B). Where both S₁ and S₂ are observed (only in the biotite migmatite), a weak mineral lineation defined by the alignment of biotite, exists parallel to the intersection of the two fabrics (Fig. 6A).

The Alamgin antiform is defined by a systematic curving of S₁ (Fig. 3). The axial plane of the Alamgin antiform is ∼213, 79 NW in the northern part of the fold (Fig. 6B) and is ∼186, 66 NW in the southern part of the fold (Fig. 6C). The fold axis varies as well: 67→244 in the northern part of the fold, to 34→200 in the southern part of the fold. In an Antwerp-Rossie suite sill, defining the northern part of the fold (Fig. 3), a strong, hinge-line–parallel mineral lineation is developed.

It should be noted that although S₂ has a fanned orientation, S₂ is generally subparallel to the estimated axial plane of the Alamgin antiform (Fig. 6B). Also, asymmetries of small-scale F₂ folds are consistent with them being parasitic folds to the large-scale Alamgin antiform (Fig. 5C). In the southern hinge area, crenulations of S₁ are much better developed than in surrounding areas where S₂ cannot be identified. Therefore, we interpret the Alamgin antiform to be a large-scale F₂ fold.

The Noname ductile shear zone (Figs. 2 and 3) is defined by high-strain, finely laminated straight gneisses, with a subvertical foliation on average measuring 036, 78 SE, with a well-developed mineral rodding, or stretching lineation on average measuring 61→198 (Figs. 5D and 6D). The high-strain nature of the shear zone is best demonstrated by the strong lineation and very planar gneissic layering, with gneissic layering-parallel leucosomes being boudinaged parallel to foliation in a manner indicating extension both parallel and perpendicular to lineation. Outside the shear zone, lineation is at best poorly developed, and the foliation is not as planar or consistent in orientation within an outcrop or across the study area. The east limb of the Alamgin antiform merges with the Noname ductile shear zone (Fig. 3) as demonstrated by rotation of S₁, development of a strong lineation, and the leucosome becoming more strongly boundinaged as the Noname ductile shear zone is encountered.

The folding pattern observed in the northwestern section of the Lowlands is most consistent with the F₁ and F₂ fold patterns identified by Foose and Carl (1977) and others. Evidence for an earlier phase of folding (F₁ of Wiener, 1983) is rare and equivocal. No northwest-southeast–trending cross folding, which is thought to create a Type 3 fold-interference pattern—F₃ of e.g., Foose and Carl (1977) or F₄ of e.g., Weiner (1983), or associated foliation were recognized in this study. Foose and Carl (1977) and Tewksbury (1993) both describe only two foliations (S₁ and S₂) in the Lowlands, similar to this work.

The fabric within the Noname ductile shear zone is very consistent in orientation, and the Noname ductile shear zone is unfolded (Fig. 6D). It would not be possible to create the Alamgin antiform without also folding the Noname ductile shear zone if it preceded the Alamgin antiform. Therefore, the Noname ductile shear zone is the last significant ductile structure formed in the area (Fig. 3). However, in a few locations, the high-strain Noname ductile shear zone fabrics are cut by a weak schistosity, defined by the alignment of biotite like S₂, and consistent in orientation with S₂ found elsewhere (Fig. 5D). This seemingly contradictory relationship is considered to indicate that during development of F₂ folds and S₂ fabric, the Noname ductile shear zone was active, but during the later stages of D₂, the Noname ductile shear zone largely was inactive and was locally overprinted by S₂.

Planes parallel to lineation and perpendicular to foliation within the Noname ductile shear zone rarely displayed asymmetrical features that could indicate kinematics. When such features were identified, within the same outcrop and from outcrop to outcrop, indicators were inconsistent. However, convincing sinistral kinematic indicators have been observed on subhorizontal faces in the Noname ductile shear zone. These include tension gashes, typically filled with granitic material, oblique to the main Noname ductile shear zone fabric in a manner consistent with sinistral shear (Fig. 7A). Some gashes are sigmoidal with “s” asymmetry also consistent with sinistral shear (Fig. 7A). Other kinematic indicators include oblique-sinistral extensional ductile shear bands (Fig. 7B) and oblique fabrics (S₁?) rotated into parallelism with the Noname ductile shear zone fabric (Fig. 7C). In one location, elongate granitic blebs, likely melt at the time of shearing, align to form a fabric oblique to the Noname ductile shear zone fabric (subparallel to the shear plane). The blebs arranged themselves parallel to the plane of extension in sinistral shear (Fig. 7D) consistent with the manner in which melt segregates in oblique planes due to shear in the experiments of Holtzman and Kohlstedt (2007). Lastly, small “s” asymmetric folds consistent with sinistral shear have also been observed in the Noname ductile shear zone (Figs. 7E and 7F).

The subvertical foliation and subvertical lineation, with consistent kinematic indicators observed in subhorizontal planes, coupled with extension both parallel and perpendicular to lineation, strongly suggest that the Noname ductile shear zone formed through transpression (Sanderson and Marchini, 1984). In transpression, the lineation indicates a significant amount of vertical stretch subperpendicular to mainly strike-slip motion. Near-vertical lineations, such as those associated with the Noname ductile shear zone, have been described in other strike-slip shear zones (Hudleston et al., 1988; Tikoff and Greene, 1997; Williams and Vernon, 2001). Also consistent with this model is the uniform metamorphic grade across the Lowlands (Bohlen et al., 1985; Kitchen and Valley, 1995), which would be disrupted if significant vertical displacements occurred along shear zones.

A schematic kinematic model explaining the formation of the Alamgin antiform and Noname ductile shear zone is shown in Figure 8. The model demonstrates that the Alamgin antiform developed as a result of the same sinistral transpression that developed the Noname ductile shear zone—consistent with the structural relationships. In this way, the Alamgin antiform can be viewed as a shear-zone propagation fold or a strike-slip nappe. This model suggests that development of D₂ structures (S₂, F₂, and Noname ductile shear zone) was synchronous, but the model alone does not explain the origin of D₁ structures.

To constrain the timing of deformation and metamorphism, four samples were collected for zircon geochronology. Approximately 10 kg–sized samples were collected directly from outcrop at the locations indicated in Figure 3 (Table 1). Zircon separation and concentration were conducted at the University of Minnesota. Sample crushing was accomplished by a jaw crusher and disk mill, and pulverized samples were washed using a Gemeni table. The heavy-mineral fraction was magnetically separated with a Frantz isodynamic separator with the least magnetic fraction being further separated by density via methylene iodide (specific gravity of 3.32), which yielded a very pure zircon sample.

Sensitive high-resolution ion microprobe–reverse geometry (SHRIMP-RG) analysis was carried out at the Stanford U.S. Geological Survey (USGS) Micro Analysis Center (SUMAC). Grain-mount preparation and SHRIMP-RG analysis protocols are outlined below, and additional details can be found in e.g., Barth and Wooden (2006), Premo et al. (2008), appendix of Chiarenzelli et al., (2010), and the references therein for each.

Zircon grains were handpicked to select the clearest inclusion-free grains. Picked zircons were mounted in epoxy and polished to uncover zircon interiors. Analysis locations were chosen based on patterns observed in cathodoluminescence (CL) images such that the area analyzed was very likely to be free of inclusions and likely to be of uniform age. For the granitic samples (see below), the focus was any noninherited core-oscillatory zoned regions, which would most likely provide the age of magmatic crystallization. For the quartzite (see below), the focus was any dark, generally featureless rims indicative of metamorphic overgrowths on detrital grains.

The SHRIMP-RG dates ∼20-μm-diameter spots on individual grains by ablating the sample with an O₂⁻ beam. Ablated sample ions are analyzed isotopically by the mass spectrometer portion of the instrument. During data collection, the 1200 Ma SUMAC VP-10 standard was analyzed after every 3–5 analysis spots to ensure instrument calibration, which was within 0.71% at 2σ. The SHRIMP-RG data were reduced by SQUID (Ludwig, 2001); data analysis and plot generation were done by Isoplot (Ludwig, 2003).

Sample 12a (syn-D₁). Sample 12a is of a tourmaline-bearing leucogranite (Fig. 3). It was collected from an ∼1-m-wide, low-angle, with respect to S₁, dike with an apophysis folded into F₁ folds (Fig. 4D). Thus, this particular dike is considered to be pre- to early-D₁. Numerous observations of nearly identical tourmaline-bearing granites clearly intruding during D₁ (e.g., Figs. 4E and 4F) suggest that the crystallization age of sample 12a is that of D₁.

Sample 12a's zircons are subhedral, elongate, with dipyramidal terminations. Grains ranged in length from 100 to 400 μm and are clear to brown in color. The clearest colorless to slightly colored grains were handpicked for dating. The CL images (Fig. 9) of sample 12a's zircons revealed the likely existence of inherited cores in many grains as indicated by a differing CL response to that of the rim. These were usually avoided, and oscillatory zoned rims (tips) were the focus of SHRIMP-RG analysis. Based on grain shape and the oscillatory zoning, it is interpreted that zircon rims grew from magma, and their age represents crystallization of the intrusion.

Analyses revealed Pb loss is a significant issue with these zircons and analyses were discordant. Despite this, six analyses (5%–54% discordance) from rims define a regression line with lower and upper intercepts of 298 ± 33 Ma and 1160 ± 9 Ma, respectively (Fig. 9; Table 1). Circa 1160 Ma is interpreted to be the age of crystallization of the host tourmaline-bearing leucogranite.

Sample 18c (late D₁ to early D₂). Sample 18c was collected immediately adjacent to the Black Lake fault (Fig. 3). This outcrop, presumed to be Rockport Granite based on location and field character, has xenolithic blocks of the migmatite that contain S₁ as indicated by leucosomal-defined gneissic layering and the alignment of biotite (Fig. 10A). Careful inspection of the contact of the granite with a large xenolithic block of the migmatite country rock revealed that the granite and the country rock were deformed by F₂ folds, as defined by the S₁ gneissic layering being folded into decimeter-scale open folds typical of F₂ (Figs. 10B and 10C). Therefore, sample 18c's intrusion can be bracketed between late-D₁ to early-D₂.

Zircons separated from sample 18c are very similar to those separated from sample 12a with respect to size, shape, and color. Zircon of sample 18c differed only in that pyramidal terminations were most common. Like sample 12a, the clearest colorless to slightly colored grains were handpicked for dating. Sample 18c's zircons display very dark and convoluted patterns in CL owing to their likely metamict state as a result of high U content (Table 1; Fig. 11), although in patches, oscillatory zoning was observed. Obvious inherited cores are scarce. The presence of oscillatory zoning along with the elongate subhedral crystal form is interpreted to indicate that the zircons grew from magma and therefore their age represents the time of crystallization of the pluton.

One concordant and three discordant (41%–83%) analyses well define a regression line with lower and upper intercepts of 233 ± 17 Ma and 1170 ± 15 Ma, respectively (Fig. 11). The upper intercept of ca. 1170 Ma is interpreted as the age of the crystallization of this Rockport Granite pluton, and this age is consistent with other ages obtained for this unit: 1173 ± 4 Ma (van Breemen and Davidson, 1988) and 1172 ± 5 Ma (Wasteneys et al., 1999).

Sample WP2 (late D₂). In the Noname ductile shear zone, not all leucosome is highly deformed, but some leucosome forms irregular blob-like masses cutting Noname ductile shear zone fabric, are boudinage-neck filling (Fig. 12A), and tension-gash filling (Fig. 7A). Taken together, this indicates melt was present throughout Noname ductile shear zone shearing. One irregular granite body within the Noname ductile shear zone was large enough to allow collection for geochronologic studies (Fig. 12B). The pluton is unfoliated, but small veinlet offshoots from the main pluton are folded and boudinaged indicating that the granite is slightly deformed (Fig. 12C). Sample WP2 was collected from this granite and is considered to have intruded late in Noname ductile shear zone deformation, and hence, late during D₂.

Sample WP2’s zircons are large, typically 200–500 μm or larger, reflecting the grain size of the host rock. As a result, there were many broken portions of larger grains. Whole grains are euhedral and equant to stubby in shape. Color ranged from clear to brown with the clearest, colorless to slightly colored grains handpicked for dating. Oscillatory zoning is obvious in all grains from CL images of sample WP2 (Fig. 13). Most grains display zoning indicative of growth from a single event. However, some inherited cores appear to be present, but analyses from the suspected cores did not result in a different age from the rest of the grain or other grains (Table 1). The euhedral character and oscillatory zoning leads to the interpretation that zircon growth is coincident with magmatic crystallization of the pluton.

All 11 analyses appear to be part of the same age population with most being concordant or near concordant (−3% to 4% discordance). Three analyses were culled from final age calculation because of significant (12%) discordance compared to the other analyses, or slightly older ²⁰⁷Pb/²⁰⁶Pb ages (36–48 m.y.). A regression line through the remaining data has lower and upper intercepts of –1179 ± 2800 Ma and 1156 ± 8 Ma, respectively. The lower intercept provides a geologically impossible age that also includes the origin when the error is considered. This suggests that the data are best summarized as a ²⁰⁷Pb/²⁰⁶Pb weighted mean average (Fig. 13, inset) and produces a best estimate of WP2 granite magmatic crystallization of 1158 ± 12 Ma.

Sample 6c (quartzite). Sample 6c was initially collected because the outcrop is mapped as the same kind of granite as sample 18c (Buddington, 1934); however, the field character of this rock is much more fine grained and quartz rich than the granite of sample 18c (Fig. 14). A thin section revealed that the rock is ∼95% quartz with minor amounts of plagioclase, muscovite, and pyrite (determined during mineral-separation process). Both zircon and apatite lack crystal face development and are, with rare exceptions, subequant to columnar in shape. Therefore, sample 6c is considered to be a quartzite, which is unsurprising given the common occurrences of quartzite in the area.

Sample 6c's zircon grains are small, ranging in size from 75 to 200 μm in length. Zircon shape is subequant to columnar with, at best, subhedral crystal face development. Color ranged from clear to pink to brown; clear to pink grains were handpicked for dating. In CL, the zircons of sample 6c have rounded, detrital cores with a variable CL response. Around the cores are thin, dark featureless rims interpreted as metamorphic overgrowths due to lack of zoning (Fig. 15). Though rims could be identified on most grains, their size precluded analysis because the primary beam often could not be placed to avoid the detrital core or a significant amount of epoxy.

Only five rim analyses were possible from all the grains mounted (Table 1). Analyses ranged from concordant to somewhat reverse discordant (–5% to 0%). A regression line fit four of the five analyses well providing lower and upper intercepts of 179 ± 700 Ma and 1169 ± 20 Ma, respectively. The fifth analysis was discarded because of its lack of fit to this line. Because the lower intercept includes the origin within error, the ²⁰⁷Pb/²⁰⁶Pb weighted mean average of 1173 ± 10 Ma from the four analyses is the best estimate of the timing of rim growth and metamorphism of the quartzite.

Figure 16 summarizes the age constraints of D₁, D₂, plus the associated metamorphism, and indicates that all events occurred during the maximum interval of ca. 1185–1145 Ma. Although the constraint on metamorphism places it early in the ca. 1185–1145 Ma interval, the obvious presence of melt during D₂ shearing along the Noname ductile shear zone suggests that metamorphism was wholly coeval with both deformation phases (e.g., Figs. 4E, 4F, 7A, 7D, and 12A). In contrast, field relations clearly distinguish between D₁ and D₂’s relative timing, but the absolute ages of D₁ and D₂ cannot be distinguished given the large overlap of crystallization age error bars on the intrusions that constrain the timing of deformation. Based on the overlap of error bars (Fig. 16), it could be argued that the events occurred during a much narrower window of time: ca 1170–1163 Ma, though the precision of the data precludes concluding this with confidence.

F₁ and F₂ folds are coaxial resulting in a Type 3 interference pattern. It is possible for various fold-interference patterns, such as the Type 3 patterns, to be produced through progressive deformation or synchronously (e.g., Wynne-Edwards, 1963; Forbes et al., 2004). Therefore, we suggest that D₁ and D₂ either occurred as separate, but temporally close events, or are the result of progressive deformation. Regardless of the precise relationship between the deformation phases, the D₂ sinistral transpression, in addition to producing the large-scale Alamgin antiform, produced the Noname ductile shear zone or reactivated a prior existing shear zone.

The timing of all events places them well within the established time frame for the Shawinigan orogeny. No ductile deformation associated with the Elzevirian orogeny (with the possible exception of the pre-D₁ discussed above) or Ottawan phase of the Grenvillian orogeny was detected and supports prior conclusions that the majority of the Adirondack Lowlands escaped deformation during the Ottawan and Rigolet phases of Grenvillian orogeny.

Although Wiener (1983) identified additional fold phases, that work was done immediately adjacent to the Carthage-Colton mylonite zone, so deformation associated with this zone may complicate the structures there. Otherwise, the folding events defined in this work are identical in character to the coaxial folding events described across the entire Lowlands in numerous studies. Therefore, we suggest that the two deformation phases observed throughout the Lowlands occurred at ca. 1185–1145 Ma. The penetrative D₁ established the regional gneissic layering, and D₂ formed the large-scale regional folds and sliced the Lowlands into panels or reactivated previously existing shear zones, creating blocks of folded rock separated by anastomosing shear zones (deLorraine and Carl, 1993).

Results of kinematics analyses of other Lowlands shear zones produce similar results as does this research. Marble mylonites from the Beaver Creek fault zone (Fig. 2) reveal oblique-sinistral kinematics on a steeply northwest-dipping structure (Colony et al., 2010). In addition, preliminary analysis of ductile kinematic indicators in high-strain marble gneisses in an area where the Beaver Creek fault zone and the Pleasant Lake fault zone appear to merge (Fig. 2) indicates sinistral strike-slip kinematics on a near-vertical ductile shear zone. Hudson (1994) studied the Hailesboro ductile deformation zone located at the base of the Popple Hill Gneiss (Fig. 2) and found sinistral strike-slip kinematic indicators thought to be of Shawinigan age. Lastly, kinematic results from the Carthage-Colton mylonite zone, possibly of late-Shawinigan age, reveal oblique-sinistral thrusting (Baird, 2008). These data further support that a common deformation style occurred throughout the Adirondack Lowlands.

No evidence for cross folds was found in this work. The sheath fold model of Tewksbury (1993) could explain the circular map patterns of e.g., the Hyde School Gneiss, but given the steepness of the structures in the Lowlands, this would mean that significant subvertical simple shear must have occurred. As mentioned above, no evidence for significant regional tilting or crustal-scale differential motion has been revealed. Therefore, the plate tectonic driving force behind such a deformation is difficult to envision. Alternatively, the domes of the Hyde School Gneiss could be just that—domes, produced when the Hyde School Gneiss bodies intruded syntectonically (McLelland et al., 1992). However, the solid-state folding of the rock bodies is well documented (Foose and Carl, 1977), and the folding and associated foliations are part of the dome structure of the bodies, indicating that significant tectonic reworking of the bodies was partially involved in forming the domes. It is possible to produce essentially the same domical structure through some combination of undulation amplification and differential vertical extrusion as a result of the pure shear component of the proposed D₂ transpression that deformed the Lowlands. This removes the requirement for a subsequent northwest-trending fold set to produce the dome pattern of the Hyde School Gneiss.

The Shawinigan orogeny 1190–1140 Ma (Corrigan, 1995; Rivers, 1997) is generally accepted to have resulted from accretionary tectonics, but interpretations of the origin of the metasediments in the Frontenac terrane and Adirondacks vary. Carr et al. (2000) suggest that the metasediments of the Frontenac terrane and Adirondacks represent contiguous platform sediments deformed and metamorphosed during the Shawinigan orogeny when the Elzevir terrane (Fig. 1B), and additional crustal blocks to the west, collided with the Frontenac terrane and Adirondacks offshore from Laurentia. Others interpret the metasediments of the Frontenac terrane and Adirondacks as being deposited in a backarc basin that was closed and accreted to Laurentia during the Shawinigan orogeny (Hanmer et al., 2000; Rivers and Corrigan, 2000; Hynes and Rivers, 2010). Further developing the backarc basin interpretation for the Frontenac terrane and Adirondacks, McLelland et al. (2010a) and Chiarenzelli et al. (2010) interpret the Shawinigan orogeny to have resulted from the reattachment of a rifted Laurentian slice, Adirondis (Dysart–Mount Holly suite; Gower, 1996; Fig. 1B), to Laurentia and the already accreted Frontenac terrane, itself an additional rifted slice of the Laurentian margin. This model has the Lowlands and portions of the Highlands metasediments originally being deposited in what has been termed the Trans-Adirondack backarc basin (TAB; Fig. 1B) between the rifted Laurentian basement slices of the Frontenac terrane and the southern Adirondack Highlands (Chiarenzelli et al., 2010). In this scenario, the Black Lake fault approximately demarcates the transition from pre-Grevillian Laurentian basement underlying the metasediments of the Frontenac terrane to oceanic crust underlying the metasediments of the Trans-Adirondack backarc basin (Chiarenzelli et al., 2010; Chiarenzelli et al., 2011b).

This paper demonstrates that the deformation across the Lowlands resulted from an oblique convergence or collision to the current structural grain of the Frontenac terrane and the Lowlands, resulting in sinistral strike-slip and transpressional shear zones. Following the model of Chiarenzelli et al. (2010), as it incorporates much new data pertinent to reconstructing the Shawinigan orogeny, we suggest Adirondis collided approximately northward with respect to modern-day plate configurations (Fig. 17).

Our proposed sinistral transpressional model is consistent with Shawinigan-aged deformation in the Lac-Saint-Jean area (Fig. 1A). There, the synkinematic 1157 ± 3 Ma AMCG suite Lac-Saint-Jean anorthosite and associated comagmatic dikes were deformed by sinistral strike-slip motion in a north-northeast–oriented shear zone (Hervet et al., 1994; Higgins and van Breemen, 1996; Higgins et al., 2002). Further, the work of Wynne-Edwards (1967) indicated that sinistral strike-slip motions were responsible for the coaxial folding of the Frontenac terrane and correlates well with this work.

McLelland et al. (2010a) suggest that Shawinigan tectonism in the Lowlands was largely restricted to ca. 1180–1160 Ma, with the younger limit on this age range constrained by the undeformed Kingston dike swarm in the Frontenac terrane (Davidson, 1995). This work supports this. However, our data also indicate that significant deformation and metamorphism (as demonstrated by the presence of melt) may have lasted as late as ca. 1158–1145 Ma. This suggests that Shawinigan deformation in the Adirondack Lowlands outlasted deformation in the Frontenac terrane by at least a few to perhaps 15 m.y. and further characterizes the differences between the Frontenac terrane and Lowlands. It should be noted that the diachronous nature of the deformation may be gradational between the two terranes, or may be discrete. Additional evidence that the geologic difference between the Frontenac terrane and the Lowlands is discretely different comes from the large metamorphic temperature difference occurring across the vicinity of the Black Lake structures. This suggests that a terrane-bounding shear zone or perhaps a suture does juxtapose the Frontenac terrane and Adirondack Lowlands. Candidates for the boundary include the Black Lake fault (Davidson, 1995; Peck et al., 2004; McLelland et al., 2010a; Chiarenzelli et al., 2010) or proposed Black Lake shear zone (Wong et al., 2011).

Significant differences between the Frontenac terrane and Adirondack Lowlands exist, but these do not prove a suture juxtaposes them. As Hanmer et al. (2000) point out, many of the proposed sutures in the Grenville Province do not have compelling evidence supporting that they are indeed sutures. In the case of the Frontenac terrane and Adirondack Lowlands, differences in metamorphism and cooling histories could be accounted for by normal faulting allowing different crustal profiles to be exposed. Differences in lithology between the terranes, supported by differences in carbon-isotope ratios between the regions (Kitchen and Valley, 1995; Tortorello and Peck, 2010), also do not necessitate a suture but could be due to sedimentary facies transitions or by adjacent, but independent sedimentary basins (Chiarenzelli et al., 2010). The geochemical differences (Nd systematics and δ¹⁸O data) are compelling but also do not prove the existence of a suture as the Nd systematics have been interpreted as resulting from a transition of the metasedimentary rocks being underlain by Laurentian crust (Frontenac terrane) to oceanic crust (Lowlands, Chiarenzelli et al., 2010; Chiarenzelli et al., 2011b). The shift in δ¹⁸O only loosely corresponds with any proposed boundary between the two crustal blocks and was originally interpreted to be caused by partial melting of subducted high δ¹⁸O sediments and oceanic crust, placed mostly under the Frontenac terrane (Peck et al., 2004). Given the outlined difference in the geology of the Frontenac terrane and the Adirondack Lowlands, we feel it is appropriate to consider the two crustal blocks as separate terranes; however, there is not compelling evidence to suggest that a suture exists between them. Further work is needed to clarify the nature of the boundary between these terranes.

1. The Adirondack Lowlands underwent two penetrative deformation phases (D₁ and D₂) between ∼1185 and 1145 Ma during the Shawinigan orogeny as determined by U-Pb zircon geochronology of granitic plutons with various relationships to the deformation. This deformation was responsible for developing two coaxial fold sets plus regional-scale ductile shear zones. The F₁ folds are outcrop-scale with the regional gneissic layering being axial-planar, whereas F₂ folds have a weak axial-planar schistosity and range from outcrop-scale to km-scale. The F₂ folds developed synchronously with the regional-scale ductile shear zones.

2. D₁, D₂, and the Shawinigan orogeny were accompanied by upper amphibolite to granulite-facies metamorphism, as indicated by the presence of partial melt during all deformation, metamorphic rims on detrital zircons in a quartzite, and the presence of sillimanite.

3. D₂, and likely D₁, resulted from sinistral transpression caused by the collision of Adirondis, part of the Adirondack Highlands, northward into the accreted Frontenac terrane. Deformation potentially resulting from this sinistral transpression can be found in the Frontenac terrane and Lac-Saint-Jean area as well.

4. The circular outcrop patterns of the e.g., Hyde School Gneiss in the Adirondack Lowlands, were produced by undulation amplification and/or vertical extrusion during D₂ sinistral transpression, thus removing the need for a set of cross folds to form a Type 1 fold-interference pattern to produce such outcrop patterns.

5. Timing of deformation in the Adirondack Lowlands may have continued later than that in the Frontenac terrane by possibly a few to perhaps 15 m.y.

6. Although the Frontenac terrane and Adirondack Lowlands have similarities, the differences are significant enough to confirm they are different terranes. The location of the boundary between them remains poorly defined. However, it is unlikely, given the current data, that the Frontenac terrane and Adirondack Lowlands were once far-separated crustal blocks; therefore a suture does not exist between them.

Jim McLelland is thanked for directing us to this particular field area and for his years of excellent Adirondack research. Bruce Selleck is thanked for providing the opportunity to date some of the samples by the SHRIMP-RG at Stanford. Some SHRIMP-RG samples were mounted and imaged by Jade Star Lackey. Joe Wooden, Ariel Strickland, Bruce Selleck, and Martin Wong are all thanked for assistance in SHRIMP-RG operation. GBB thanks Annia Fayon for her willingness to help with mineral separation, as well as Bjorn Batdorf for accommodations during mineral separation work. Ken Ludwig is acknowledged for use of Isoplot/EX 3.00, and Richard Allmendinger's Stereowin produced the stereonets. Funding for this research was from University of Northern Colorado startup funds to GBB, the 2008 Adirondack Lowlands Keck Project, and a St. Lawrence University research grant to CHS. Bill deLorraine, John Bursnall, and Bruce Selleck provided stimulating conversation in regards to this project. Detailed and thoughtful comments from two anonymous reviewers greatly focused and strengthened this contribution.