U-Pb detrital zircon analysis of sedimentary rocks of the southeastern New England Avalon terrane in the U.S. Appalachians: Evidence for a separate crustal block

Avalonia comprises an ~20–200-km-wide zone of subterranes that define the eastern flank of the northern Appalachians between Newfoundland, Canada, and southeastern New England in the United States. These terranes are considered to represent the remnants of a microcontinent that rifted from supercontinent Gondwana in the Ordovician (e.g., Nance et al., 2008; Pollock et al., 2012) and accreted to Laurentia during the latest Silurian to Middle Devonian Acadian orogeny (e.g., Skehan and Rast, 1990; Nance et al., 2008; van Staal et al., 2009; Hatcher, 2010). The southeastern New England Avalon terrane in southeastern Massachusetts, Rhode Island, and southeastern Connecticut is comparable with Avalonian terranes in Newfoundland, Nova Scotia, and New Brunswick, Canada (Fig. 1), and collectively these are called Avalonia, or West Avalonia to distinguish it from East Avalonia in Europe (Williams and Hatcher, 1982, 1983; Rast and Skehan, 1993; van Staal, 2005; Hibbard et al., 2007). In the northern part of the southeastern New England Avalon terrane (north of the blue dashed line in Fig. 1C), correlations with Avalonian terranes in Canada are based on rock types and ages and U-Pb detrital zircon signatures of metasedimentary rocks (Hepburn et al., 2008; Thompson and Bowring, 2000; Thompson et al., 2012, 2014). However, south of that line, the correlation has primarily been based on rock types and ages from Ediacaran granitoid rocks, and the true nature and origins of those rocks are uncertain.

The southeasternmost tip of Massachusetts and adjacent offshore regions southeast of the Nauset magnetic anomaly (Fig. 1A) have previously been interpreted as the Gondwanan-derived Meguma terrane, based on seismic reflection, gravity and magnetic surveys, and K-Ar dates (e.g., Hutchinson et al., 1988; Stewart et al., 1993; van Staal et al., 2009; White and Barr, 2010). The Meguma terrane departed from Gondwana in the early Silurian (MacDonald et al., 2002) and accreted to Laurentia during the Middle Devonian to Earliest Carboniferous (e.g., van Staal et al., 2009; White and Barr, 2010). The terrane is exposed in Nova Scotia, Canada (Fig. 1A), but not in southeastern Massachusetts. The presence of an Ediacaran granite in southeastern Massachusetts, southeast of the Nauset magnetic anomaly (Leo et al., 1993; CC2 in Fig. 1C) is not consistent with the Meguma terrane in Nova Scotia, which does not contain Precambrian rocks. The area southeast of the Nauset magnetic anomaly is more likely to be part of the Avalon terrane, or alternatively part of a northwest African crustal block, as is present below the Georges Bank, 143 km east-southeast of Nantucket, Massachusetts, near the eastern continental boundary of North America (Fig. 1A; Kuiper et al., 2017). Therefore, the nature and origin of this southeastern extremity of Massachusetts is as uncertain as the remainder of the Avalon terrane south of the blue dashed line in Figure 1C.

In this contribution, we present detrital zircon U-Pb dates from seven metasedimentary rocks of the Avalon terrane of southeastern New England in the area south of the blue dashed line in Figure 1, obtained by laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) analysis of zircon. These dates, together with other isotopic studies of Avalonia, are used to place constraints on models for the Paleozoic evolution of the southeastern New England Avalon terrane.

The southeastern New England Avalon terrane is bounded to the northwest by the Honey Hill–Lake Char–Bloody Bluff fault system (Fig. 1C) and is characterized by tectonostratigraphic relationships between Proterozoic to Permian rocks that are distinct from those in the rest of New England. The Avalon terrane of southeastern New England consists of Ediacaran granitoid rocks and Cambrian sedimentary and volcanic rocks (Hepburn et al., 1993; Thompson et al., 1996, 2007, 2010b; cf. Zartman and Naylor, 1984) and Ediacaran to Devonian alkaline plutons (Hermes and Zartman, 1985, 1992; Thompson et al., 2018) (Figs. 1C–1D). In the Late Devonian to Carboniferous, graben sediments of the Narragansett and Norfolk basins (Fig. 1C) were deposited (e.g., Mosher, 1983; Skehan et al., 1986; Thompson and Hermes, 2003; Murray et al., 2004, and references therein). Metamorphism throughout much of the Avalon terrane in southeastern New England is no higher than greenschist facies, but locally reached the amphibolite facies (Grew and Day, 1972; Goldsmith, 1991c; Wintsch et al., 1992, 2014; Fetherston et al., 2014). Much, if not all, of the ductile deformation and metamorphism in the southern part of the Avalon terrane of southeastern New England may be attributed to the Alleghanian orogeny (e.g., Wintsch et al., 2014). The effects of the Alleghanian orogeny decrease to the north and are minimal north of the blue dashed line in Figure 1C (Severson, 2020).

As proposed by O’Hara and Gromet (1985) and refined later (Gromet, 1989; Hermes et al., 1994, 1998; Gromet et al., 1998), the southeastern New England Avalon terrane may be subdivided into the Hope Valley and Esmond-Dedham subterranes. The Hope Valley subterrane in the western Avalon terrane (Fig. 1C) consists of primarily ca. 0.6 Ga leucocratic metaigneous rocks that have, without exception, undergone at least one episode of deformation under upper amphibolite facies metamorphic conditions, resulting in annealed textures and the absence of any preserved, primary features. Where dated (Gromet et al., 1998; Wintsch et al., 1998a, 1998b, and references therein), structures and age relationships indicate pervasive Alleghanian dynamic recrystallization and annealing, in which earlier fabric elements (if ever present) have been completely overprinted.

By contrast, within the Esmond-Dedham subterrane (Fig. 1C), the intensities of deformation and metamorphism are much more varied, and relict primary structures are preserved in many places. Based on age relationships, this subterrane may be divided into three lithotectonic groups. The first group consists of (1) metamorphic rocks of unknown, but presumed Neoproterozoic age, of which the Blackstone Group is the most distinctive; (2) variably metamorphosed ca. 0.6 Ga granitoid rocks plus subordinate volcanic, volcaniclastic, and clastic sedimentary rocks; and (3) unconformably overlying Cambrian shallow marine sedimentary rocks that contain distinctive Acado-Baltic fauna unlike fossil assemblages of comparable age found elsewhere in New England. This group is often referred to as the “Avalonian basement” of southeastern New England. The second group consists of (1) latest Precambrian to late Devonian bimodal, alkalic igneous rocks; and (2) Late Devonian to Late Pennsylvanian molasse deposits, of which those of the Narragansett Basin are the most significant. The third group consists of Permian granitoid rocks (e.g., the Narragansett Pier Granite; Fig. 1C) that were emplaced during the waning stages of the Alleghanian orogeny. It is worth noting that although the regions east and south of the Narragansett Basin are generally considered to be a continuation of the Esmond-Dedham subterrane, the rock types, metamorphic grade, and structures in the southern area around New Bedford (Fig. 1C) compare more favorably to the gneissic rocks of the Hope Valley subterrane and Narragansett Pier Granite in southern Rhode Island and southeastern Connecticut.

Based on detrital zircon signatures of metasedimentary rocks, inherited zircon in plutonic and volcanic rocks, and Nd model ages, Avalonia is interpreted as having a 1.3–1.0 Ga basement (Nance and Murphy, 1994, 1996; Keppie et al., 2003; Thompson and Bowring, 2000; Thompson et al., 2010b, 2012; Murphy et al., 2013; Henderson et al., 2016). In Avalonian rocks in Canada, previous detrital zircon analyses yielded a predominant ca. 750–500 Ma population, and scattered ages up to Archean, suggesting an Amazonian origin (Keppie et al., 1998; Barr et al., 2003, 2012, 2019; Murphy et al., 2004; Pollock et al., 2009; Satkoski et al., 2010; Henderson et al., 2016). Inherited zircon in igneous rocks in Avalonia yielded ca. 2.5–2.7, ca. 2.2–2.3, ca. 1.4–1.5, and ca. 1.0–1.1 Ga ages (e.g., Zartman and Hermes, 1987; Bevier and Barr, 1990; Bevier et al., 1993; Barr et al.,1994; Samson, 1994; Nance and Murphy, 1996).

Analyzed inherited zircon in igneous rocks in the Avalon terrane of southeastern New England is rare. In the northern Avalon terrane in Massachusetts, immediately west of the Boston Basin (Fig. 1C), the Ediacaran Dedham Granite has one inherited zircon with a ca. 1.34 Ga ²⁰⁷Pb/²⁰⁶Pb date (Thompson et al., 2010b; Fig. 1C), and an Ediacaran rhyolite porphyry in the Lynn Volcanics yielded one inherited zircon with a ca. 1.62 Ga ²⁰⁷Pb/²⁰⁶Pb date (Thompson et al., 2007; Fig. 1C). The Permian Westerly granite in southeastern Rhode Island (WG in Fig. 1C; Zartman and Hermes, 1987) yielded highly discordant zircon with Archean upper intercept ages. All other analyzed Ediacaran to Devonian igneous rocks in the Avalon terrane of southeastern New England do not show inheritance that is significantly older than the age of crystallization (Zartman and Naylor, 1984; Hermes and Zartman, 1985, 1992; Zartman et al., 1988; Thompson and Hermes, 2003; Thompson et al., 1996, 2010a, 2010b; Ellison et al., 2020).

The general detrital zircon signature of the southeastern New England Avalon terrane based on previous analyses northwest of the blue dashed line in Figure 1 consists of minor ca. 3.1–2.4 Ga and ca. 2.2–1.6 Ga and larger ca. 1.6–1.0 Ga age populations, with a large peak at ca. 1.2 Ga, and some samples also display a ca. 0.6 Ga population. Details are described below. Three detrital zircon samples from Ediacaran sandstone of the Roxbury Conglomerate and a quartzite clast in the Squantum diamictite in the Boston Basin (Fig. 1C) show abundant ca. 610–580 Ma and ca. 2.1–1.0 Ga zircon and few ca. 2.8–2.5, ca. 3.0, and ca. 3.2 Ga grains based on isotope dilution–thermal ionization mass spectrometry (ID-TIMS) methods (Thompson and Bowring, 2000; Thompson et al., 2014). Five samples of the Westboro Formation quartzite in the northwestern Avalon terrane (Fig. 1C) were analyzed by U-Pb zircon ID-TIMS (Thompson and Bowring, 2000) and LA-ICP-MS (Hepburn et al., 2008; Severson et al., this volume) methods and yielded ca. 3.0–2.7, ca. 2.5, and ca. 2.1–1.0 Ga zircon age populations. The samples of Hepburn et al. (2008) and Severson et al. (this volume) also yielded ca. 0.6 Ga zircon. Magee and Gromet (1996) previously conducted ²⁰⁷Pb/²⁰⁶Pb single-grain direct evaporation analysis on 14 zircon grains of a sample of Blackstone Group quartzite in Cumberland, Rhode Island (Fig. 1C). Six grains yielded dates between ca. 1230 and ca. 1170 Ma, two were ca. 1340 Ma, another two ca. 1530 Ma, and other single grains yielded 1770, 2060, 2170 and 2850 Ma dates. Karabinos and Gromet (1993) conducted ²⁰⁷Pb/²⁰⁶Pb single-grain direct evaporation analysis on 13 zircon grains from a Plainfield Formation quartzite sample from the “upper” quartzite unit of the Plainfield Formation in the East Killingly quadrangle in northeastern Connecticut (Fig. 1C). One grain is ca. 570 Ma, and the other analyses scatter between ca. 2745 and ca. 1200 Ma. Severson et al. (this volume) conducted U-Pb zircon LA-ICP-MS analysis on two Plainfield Formation quartzite samples ca. 4 km southeast of Putnam, Connecticut, near the location of Karabinos and Gromet (1993), and near Webster, Massachusetts (Fig. 1C), which yielded 2.8–2.4 and 2.1–1.0 Ga populations and minor scattered Neoproterozoic grains.

Detrital zircon samples of the Avalon terrane of Nova Scotia and Newfoundland, Canada, show similar age distributions as the ones described here (see Discussion; Severson et al., this volume; cf. Fig. 2). Therefore, the northwestern part of the southeastern New England Avalon terrane (northwest of the blue dashed line in Fig. 1) can be correlated with those based on existing data and is not further tested here. Major periods of volcanism, sedimentation, and plutonism in the Neoproterozoic and Paleozoic are summarized in Figure 2, and known ages of Proterozoic and Archean detrital, magmatic, and inherited zircon are summarized in Figure 3.

The Meguma terrane of Nova Scotia (Fig. 1A) consists of early Cambrian to Early Ordovician turbiditic metasandstone and slate, locally overlain unconformably by a thin sequence of early Silurian to Early Devonian slate, quartzite, and volcanic rocks (Schenk, 1997; Waldron et al., 2009, 2011; White and Barr, 2010, 2012; Pothier et al., 2015). The Meguma terrane departed from Gondwana in the early Silurian as indicated by ca. 440 Ma within-plate rift volcanism in the White Rock Formation of southern Nova Scotia (MacDonald et al., 2002; Fig. 2). Early to Middle Devonian deformation under greenschist to amphibolite facies conditions was probably related to accretion of the terrane to Laurentia. The terrane is intruded by granitic plutons that are mainly Middle to Late Devonian (White and Barr, 2010; White et al., 2012).

The detrital zircon in various units of the Meguma terrane, including the oldest Cambrian unit, has a strong West African signature with large ca. 2.2–1.9 Ga and ca. 700–500 Ma populations, but yields some ca. 1.5–0.9 Ga ages within the West African age gap (Krogh and Keppie, 1990; Waldron et al., 2009; Pothier et al., 2015; Kuiper et al., 2017; Fig. 3). The ca. 1.5–0.9 Ga zircon grains are likely to be derived from Amazonia. An origin near West Africa on the Amazonian side has, therefore, been suggested (Waldron et al., 2009).

Kuiper et al. (2017) analyzed detrital zircon in metasedimentary rocks from the bottom of the COST No. G-1 core in the Georges Bank (Fig. 1A), in order to test whether these rocks were part of the Meguma terrane, the Avalon terrane, or any other crustal material. This is the only well in offshore Massachusetts that was drilled deep enough to penetrate basement below the overlying Jurassic and younger rocks. The detrital signature was not consistent with those of the Meguma or Avalon terranes, but with the Paleoproterozoic Taghdout quartzite of the Anti-Atlas of Morocco, which is part of a Paleoproterozoic passive margin sequence that exists directly on top of the Archean–Paleoproterozoic West African Craton (e.g., Abati et al., 2010). Thus, some of the rocks below the Georges Bank are part of northwest African crustal block. Because the Anti-Atlas was adjacent to Massachusetts at the time of Pangea, the simplest interpretation is that the crustal block remained behind after the breakup of Pangea (Kuiper et al., 2017). Alternatively, it may have arrived earlier as the other Gondwanan terranes described herein.

The area southeast of the Nauset anomaly, including basement below the Georges Bank (Fig. 1A), has been interpreted as the Meguma terrane, based on seismic, gravity, and magnetic data (Stewart et al., 1993). A felsic pluton (Franklin batholith in Fig. 1) has a flat magnetic field and a large negative Bouguer gravity anomaly, similar to the South Mountain batholith in the Meguma terrane of Nova Scotia (Haworth and MacIntyre, 1975; Hutchinson et al., 1988). Hutchinson et al.’s (1988) data and the presence of graphitic slate, schist, and phyllite with 540–450 Ma whole-rock K-Ar dates (Steinkraus, 1980) in the COST No. G-1 well in the Georges Bank (Fig. 1A) led Stewart et al. (1993) to interpret the area around it as possible Meguma terrane. Stewart et al.’s (1993) interpretation of the Nauset anomaly as the Avalon-Meguma boundary is carried over in most modern publications and maps (e.g., Hibbard et al., 2006; Barr et al., 2011; van Staal et al., 2020).

A piece of information that was not available to Hutchinson et al. (1988) and perhaps Stewart et al. (1993), and has been overlooked in later publications (e.g., Hibbard et al., 2006; Barr et al., 2011; Kuiper et al., 2017), is that zircon from granite from core ~7 km southeast of the Nauset anomaly on Cape Cod (CC2 in Fig. 1C) yielded a 584 +9/–8 Ma U-Pb ID-TIMS age (Leo et al., 1993). The age is older than the oldest Cambrian unit exposed in the Meguma terrane of Nova Scotia (Schenk, 1997; White et al., 2012). Leo et al. (1993) interpreted the core granite as part of the Avalon terrane. It is possible, however, that the granite is West African. It could be basement to the Meguma terrane, though no such basement is exposed in Nova Scotia. Alternatively, it could be part of the northwest African block of Kuiper et al. (2017).

The possibility exists that parts of other Gondwanan terranes (Ganderia, Cadomia, and Carolinia) are exposed in onshore and offshore southeastern New England. While unlikely, for reasons outlined below, for completeness, characteristics of these terranes are summarized in Figures 2 and 3 and briefly described here.

Rocks to the northwest of the Avalon terrane are part of Ganderia or have Ganderian affinity. The Putnam-Nashoba terrane (Fig. 1) is a strongly deformed Cambrian–Ordovician arc-backarc complex composed of metavolcanic, metasedimentary, and volcanogenic metasedimentary rocks (Bell and Alvord, 1976; Goldsmith, 1991b, 1991c; Hepburn et al., 1995; Kay et al., 2017; Fig. 1). Its Ganderian affinity is based on U-Pb detrital zircon ages and Sm-Nd isotopic compositions of the Nashoba terrane (Loan, 2011; Kay et al., 2017). In general, the detrital zircon signature of Ganderia shows a large ca. 750–500 Ma population and minor age populations distributed between the Archean and Neoproterozoic, indicating an Amazonian affinity (Fig. 3; Barr et al., 2003; Fyffe et al., 2009; Pollock et al., 2007; Loan, 2011). Ganderia rifted from the Amazonian part of Gondwana in the Cambrian and accreted to Laurentia in the Late Ordovician and Silurian (e.g., Nance et al., 2008; Pollock et al., 2012; van Staal et al., 2009; Hatcher, 2010; Kay et al., 2017). The Nashoba terrane is intruded by Silurian to early Carboniferous granitic and intermediate composition plutons (Zartman and Naylor, 1984; Hepburn et al., 1995; Acaster and Bickford, 1999; Dabrowski, 2014). Ediacaran and Pennsylvanian–Permian gneiss and granite in the Lyme dome of southern Connecticut (Fig. 1) also have Ganderian affinity, based on Nd and Pb isotope compositions (e.g., Wintsch et al., 1990, 1992; Walsh et al., 2007). The Lyme dome is exposed within rocks interpreted as the Avalon terrane as described above. Because Ganderia exists northwest of the Avalon terrane in southeastern New England, it is unlikely that it also exists southeast of it.

Cadomia in the broadest sense is a group of terranes including Armorica, the central and northern Iberian Massif, the Bohemian Massif and French Massif Central, and the Ossa Morena and Saxo-Thuringian zones, all exposed in southern and eastern parts of Europe. It has a variety of Neoproterozoic and Paleozoic rocks (Fig. 2; Samson et al., 2003; Linnemann et al., 2004; Nance et al., 2008). It has some exposed Paleoproterozoic (ca. 2.0 Ga) basement (Samson et al., 2003). Detrital zircon yielded Archean, Paleoproterozoic, and Neoproterozoic ages, but Mesoproterozoic grains are rare or absent (Fig. 3; Linnemann et al., 2004, 2007, 2008; Samson et al., 2005; Sláma et al., 2008; Pereira et al., 2010, 2011, 2012; Miller et al., 2001; Cambeses et al., 2017), except ca. 1.0 Ga detrital zircon in the central and northern Iberian Massif (e.g., Gutiérrez-Alonso et al., 2015; Shaw et al., 2014). Inherited zircon in igneous rocks has been recorded from various part of Cadomia, including 2150–1750 and 850–500 Ma zircon from the Bohemian Massif (Linnemann et al., 2004), ca. 2.2 Ga from Armorica (Miller et al., 2001), and ca. 2.4 Ga from southwest Iberia (Pereira et al., 2011) (Fig. 3). Based on these data, Cadomia is interpreted to have a ca. 2 Ga basement (e.g., Nance and Murphy, 1994, 1996; Keppie et al., 2003; Samson et al., 2003, 2005; Bea et al., 2010; Murphy et al., 2013; Linnemann et al., 2014; Garfunkel, 2015; Henderson et al., 2016). The characteristics of Cadomia are consistent with a West African affinity. Based on a variety of data, paleogeographic reconstructions consistently show that Cadomia rifted off the African part of Gondwana in the early to middle Paleozoic, after which it moved northward (present-day coordinates) to become part of Europe (Torsvik and Cocks, 2004; Linnemann et al., 2007, 2008; Stampfli et al., 2013; Domeier and Torsvik, 2014). It never was south of Newfoundland (present-day coordinates) until after the breakup of Pangea (Torsvik and Cocks, 2004; Domeier and Torsvik, 2014) and it is therefore unlikely that southeastern New England contains a fragment of Cadomia.

Carolinia is a collection of Neoproterozoic to early Paleozoic magmatic arc and sedimentary terranes stretching from central Virginia to Alabama. It generally consists of ca. 670–610 Ma arc rocks overlain by younger clastic sedimentary rocks, and ca. 560–<532 Ma arc magmatic and intra-arc sedimentary rocks (Fig. 2; Pollock et al., 2012, and references therein). Inherited zircon in volcanic and volcaniclastic rocks yielded 1229–965 Ma dates (Fig. 3; Mueller et al., 1994). Detrital zircons in clastic sedimentary rocks from Carolinia reveal clusters of ages at 2.6–2.2, 1.8–1.5, 1.3–0.8, and 0.65–0.6 Ga (Fig. 3; Pollock et al., 2010). Based on detrital and inherited zircon ages, and on Nd isotopic data, Carolinia is interpreted as having an Amazonian affinity (Pollock et al., 2012; Samson et al., 1995; Wortman et al., 2000). Carolinia rifted from Gondwana in the early Cambrian and accreted to Laurentia in the Late Ordovician to early Silurian (Pollock et al., 2012). Carolinia thus has a similar origin as Avalonia (cf. Figs. 2 and 3) but a different Paleozoic geological history. Furthermore, based on paleomagnetic data, Carolinia is interpreted as having traveled at low latitudes, making it unlikely that parts were stranded in southeastern New England. Also, at the time of accretion of Carolinia at southern latitudes, Ganderia accreted to Laurentia in New England, and was followed by late Silurian accretion of the Avalon terrane to southeastern New England. It is thus highly unlikely that a piece of Carolinia now exists outboard of the Avalon terrane.

Seven samples were collected in order to investigate the various parts of the southeastern New England Avalon terrane south of the dashed blue line in Figure 1C. Five samples were collected from the Esmond-Dedham subterrane, and include the oldest (Neoproterozoic Blackstone Group) and youngest (Cambrian North Attleboro Formation) strata. Two samples of the Plainfield Formation were collected from the Hope Valley subterrane. Quartzites of the Blackstone Group and Plainfield Formation may be correlative (see below) and are therefore described together in the following section.

The Blackstone Group as shown on the bedrock geologic map of Rhode Island (Hermes et al., 1994; Fig. 1C) represents a diverse suite of northwest-trending metaigneous and metasedimentary rocks that primarily occur in northeastern Rhode Island in the vicinity of the Blackstone River (Quinn, 1971; Quinn et al., 1949). It consists of the following map units (Hermes et al., 1994, and references therein): (1) greenstone (formerly called the Hunting Hill greenstone by Quinn, 1971), amphibolite, serpentinite; (2) quartzite (formerly called the Quinnville quartzite by Quinn, 1971) and marble; (3) epidote and biotite schist (formerly called the Sneech Pond schist and Mussey Brook schist by Quinn (1971); and (4) undifferentiated rock. Although there are no radiometric or fossil dates from the Blackstone Group it is presumed to be the oldest sequence in Rhode Island, because the 599 ± 2 Ma Esmond Granite (Thompson et al., 2010b) truncates an early schistosity (Dreier, 1985) and contains roof pendants of the Blackstone Group. Subsequent Permian metamorphism and deformation related to the Alleghanian orogeny overprint these features, making it difficult to determine the extent to which they are evidence of widespread Neoproterozoic “Avalonian” tectonism that affected the southeastern New England Avalon terrane (Dreier, 1985; Bailey et al., 1989; cf. van Staal et al., 2020).

Attempts to define a coherent stratigraphy within the Blackstone Group have proven difficult, because mafic and ultramafic volcanic rocks are tectonically interleaved with mature continental sedimentary rocks. However, based on mapping and structural analysis, Dreier (1985) identified a stratigraphy with the Hunting Hill Greenstone at the base and the Quinnville quartzite at the top. As described by Dreier (1985) and Bailey et al. (1989), the Quinnville quartzite occurs as lenses and layers of quartz arenite interbedded with mudstone, and all units of the Blackstone Group display evidence of soft-sediment deformation. They consider that at least in the Blackstone River Valley region, where the Blackstone Group is best exposed, the tectonostratigraphic relationships represent an olistrostrome that formed as part of the early stages of the Neoproterozoic Avalonian orogeny (Dreier, 1985; Bailey et al., 1989).

The Plainfield Formation in eastern Connecticut and western Rhode Island and adjacent Massachusetts consists of quartzite and phyllite (Goldsmith, 1991a). The Plainfield Formation is overlain by a suite of largely mafic metavolcanic rocks of the Waterford Group (Goldsmith, 1980). Based on lithological, stratigraphic, and structural relationships, the Plainfield Formation has been interpreted as equivalent to the Westboro Formation in Massachusetts and quartzite and schist in the Blackstone Group (Goldsmith, 1991a; Fig. 1C). Two quartzite samples from the Blackstone Group and two from the Plainfield Formation were analyzed.

The Newport Group as shown on the bedrock geologic map of Rhode Island (Hermes et al., 1994) represents a suite of Late Proterozoic metasedimentary and metavolcanics rocks in southeastern Rhode Island. These multiply deformed greenschist facies rocks are truncated by the 595 ± 12 Ma Newport (currently known as the Lily Pond; Fig. 1D) Granite (Rb-Sr whole-rock date; Smith, 1978). The Newport Group consists of three formations: the Price Neck Formation, the Newport Neck Formation, and the Fort Adams Formation (Fig. 1D). Although sedimentary structures such as graded bedding allow younging directions to be determined within formations, a lack of outcrops or faults prevents age relationships among formations to be established. Based on the presence of a variety of soft-sediment deformation features in the Newport Neck Formation and olistoliths in the Fort Adams Formation, the Newport Group is considered to have preserved a record of Late Proterozoic rifting of a continental margin (Bailey et al., 1989). One sample from the Price Neck Formation and one from the Newport Neck Formation samples were collected from shoreline exposures.

At Hoppin Hill (in North Attleboro, Massachusetts; see location of sample 320 in Fig. 1C), massive to cross-bedded quartzite nonconformably overlies Ediacaran Dedham Granite, and in turn is overlain by fossiliferous red shale with shell hash limestones at several horizons. Shaler et al. (1899) described early Cambrian fauna from here, and since then additional studies documented the presence of a siliciclastic stratigraphy with affinities to Cambrian rocks of the Avalon platform. Detailed maps showing measured stratigraphic sections and descriptions of the paleontology and paleoenvironmental setting for this site are given in Anstey (1979) and Landing (1988, 1996). The units seen at this locale were called the Hoppin Formation on the Massachusetts Bedrock Geologic Map (Zen, ed., 1983), but subsequently the quartzite was named the North Attleboro Formation and the shale plus limestone were referred to as the Weymouth Formation (Landing, 1988, 1996). Further discussion of the Cambrian section at Hoppin Hill, in terms of its depositional setting and relationship to other contemporaneous Avalonian successions, are given in Landing (1988, 1996). One sample of North Attleboro Formation quartzite was analyzed.

Zircon grains were separated from ~10–17-kg samples using standard techniques at the Colorado School of Mines or Boise State University and subsequently further processed at Boise State University. Grains were annealed at 900 °C for 60 h in a muffle furnace and mounted in epoxy and polished until their centers were exposed. Cathodoluminescence (CL) images were obtained with a JEOL JSM-300 scanning electron microscope and Gatan MiniCL. Zircon was analyzed by LA-ICP-MS as detailed in Appendix A.

The northern sample of Blackstone Group quartzite was collected from a quarry in Cumberland, Rhode Island (Fig. 1). The sample is a somewhat variegated tan and brown quartzite (Fig. 4). It is interbedded with phyllite, and the contacts are gradual. In thin section (Fig. 5) the quartzite shows 97% quartz, 2% feldspar, and minor biotite and chlorite along fractures. Feldspar shows sericitic alteration. Quartz and feldspar grains are equant and 0.1–0.2 mm and internally undeformed. Most zircon grains are rounded to subrounded, with a few subangular grains. Grains have various degrees of CL brightness, and some grains have secondary growth rims. Most grains have oscillatory zoning, and some have sector zoning. Some grains have darker or lighter cores. Of the 135 concordant analyses, the youngest population and interpreted maximum depositional age are 1053 ± 19 Ma (mean square of weighted deviates [MSWD] = 1.4, probability of fit [pof] = 0.23, n = 4; Table S1¹), and two younger grains are 1002 ± 18 Ma and 952 ± 38 Ma. The main peak is at ca. 1.22 Ga and older peaks are up to ca. 3.07 Ga with a small age gap at ca. 2.4–2.3 Ga (Fig. 6).

This sample was taken from an outcrop at Winman Junior High School in Warwick, Rhode Island (Fig. 1), from an outlier that has been correlated with the main body of the Blackstone Group (Hermes et al., 1994). It is a tan-gray quartzite (Fig. 4). Bedding is faintly visible. The thin section (Fig. 5) shows 97% quartz, 1% feldspar, and 1% biotite and chlorite. Grains are 0.1–0.5 mm and show a weak foliation. Most zircon grains are rounded to subrounded, with a few subangular grains. Grains have various degrees of CL brightness. Most grains show oscillatory zoning, and some sector zoning. Some grains show overgrowths, and some have darker or lighter cores. Of the 135 concordant analyses, the youngest population is 997 ± 12 Ma (MSWD = 0.27, pof = 0.98, n = 10; Table S1), which is interpreted as the maximum depositional age, and one younger grain is 949 ± 25 Ma. The largest age population is ca. 1.22 Ga (Fig. 6), and there is a spread of populations up to ca. 3.17 Ga, except ca. 2.6–2.3 Ga.

The western sample of Plainfield Formation quartzite was taken from a roadcut in Groton, Connecticut (Fig. 1), and is a tan-gray quartzite (Fig. 4) with some quartz veins, which were avoided during sample collection. A thin section (Fig. 5) reveals 91% quartz, 7% feldspar, and 1% biotite with minor muscovite and chlorite. Biotite shows a weak foliation, and 0.3–1 mm quartz grains show some subgrains and undulose extinction, but are equant, suggesting static recrystallization. Most zircon grains are rounded to subrounded, with a few subangular grains. Grains have various degrees of CL brightness, and some grains have secondary growth rims. Most grains have oscillatory zoning, and some have sector zoning. Some grains have darker or lighter cores. Of the 131 concordant analyses, the youngest population is 1002 ± 9 Ma (MSWD = 0.51, pof = 0.85, n = 9; Table S1 [see ^{footnote 1}]) and forms the maximum depositional age. The main peaks are ca. 1.01, 1.23, and 1.54 Ga (Fig. 6). The oldest grains are ca. 2.7 and ca. 3.0 Ga.

This sample of Plainfield Formation quartzite was collected from an outcrop in the forest in Hopkinton, Rhode Island (Fig. 1), and is a tan-gray quartzite (Fig. 4) with a few quartz veins, which were avoided during sampling. It contains 89% quartz, 8% feldspar, 2% biotite, and 1% muscovite and chlorite. Grain boundaries show orange staining, and feldspar shows sericitic alteration. Quartz grains are 0.5–2 mm and show inclusions of other minerals. Biotite shows a foliation, while quartz is mostly undeformed, suggesting static recrystallization. Most zircon grains are rounded to subrounded, with a few subangular grains. Grains have various degrees of CL brightness, and some grains have overgrowths. Most grains have oscillatory zoning, and a few have sector zoning. A few grains have darker or lighter cores. Of 133 concordant analyses, the youngest population is 973 ± 15 Ma (MSWD = 1.7, pof = 0.14, n = 6; Table S1), interpreted as the maximum depositional age. Peaks are at ca. 1.02, 1.17, 1.23, and 1.52 Ga. The oldest grain is ca. 3.11 Ga (Fig. 6).

This sample was collected from a shoreline exposure at Ragged Point, near the Castle Hill Inn in Newport, Rhode Island (Fig. 1). It is a tan-gray siltstone collected from multiple coarsest-grained layers (Fig. 4). The thin section (Fig. 5) shows <3 mm calcite, quartz, feldspar, opaques, and minor epidote and muscovite in a fine-grained matrix. It is weakly foliated. Zircon grains are rounded to subrounded, with a few euhedral, faceted grains. They have various degrees of CL brightness, and most grains have either sector or oscillatory zoning. Some grains have overgrowths, and a few grains have cores, which were not analyzed. Of the 145 concordant analyses, the youngest population and maximum depositional age are 614 ± 8 Ma (MSWD = 0.18, pof = 0.99, n = 8; Table S1). Peaks are at ca. 647 and ca. 745 Ma (Fig. 6). Two grains are ca. 1.54–1.43 Ga, and the oldest four grains are ca. 2.12–1.98 Ga.

The sample is a tan-gray phyllite (Fig. 4), collected from the shoreline at Sachuest Point in Middletown, Rhode Island (Fig. 1). The unit is generally fine grained with a few conglomeritic layers, and the sample was taken from a mix of both. In thin section (Fig. 5), the conglomeratic phyllite shows <3 mm grains of quartz, feldspar, pyrite, muscovite, and minor calcite. Most zircon grains are subangular to subrounded, with a few angular grains and a few rounded grains. They have various degrees of CL brightness, and most grains show either sector or oscillatory zoning. Some grains have overgrowths, and a few grains have cores. Of the 153 concordant analyses, the youngest population is 613 ± 6 Ma (MSWD = 0.31, pof = 0.95, n = 8; Table S1), which is the maximum depositional age. Peaks are at ca. 643 and ca. 733 Ma (Fig. 6); smaller populations are up to ca. 2.20 Ga, and four older grains are between ca. 2.61 and ca. 3.13 Ga. One concordant grain is 412 ± 9 Ma (Table S1). It is unclear whether this is a result of contamination and therefore is excluded from figures and interpretation.

The North Attleboro Formation quartzite, or quartzitic arenite, was sampled from the western edge of the southern Hoppin Hill Reservoir, North Attleboro, Massachusetts (Fig. 1), when the reservoir was low enough to access the outcrops of North Attleboro Formation quartzite. The sample is from massive quartz arenites at the base of the Cambrian section, immediately east of an arkosic sandstone/phyllite unit, presumably close to the unconformity with the overlying Narragansett basin deposits. It has mm- to cm-scale elongate clasts of quartzite and gray phyllite within an otherwise quartz-rich, white-tan quartzitic arenite (Fig. 4). The sample was taken from loose rocks that are still within their original layer configuration. The matrix in thin section (Fig. 5) shows 93%–98% quartz, 1%–5% feldspar, and 0%–5% fine grained matrix, mostly in clasts. Feldspar shows strong sericitic alteration. Grains are generally 0.5–2 mm, undeformed and equant. Most zircon grains are rounded, eight grains are more euhedral and faceted, and 21 grains had a blue tint in transmitted light. Grains showed various degrees of CL brightness, and most grains have either oscillatory or sector zoning. A few grains showed overgrowth or cores. Of the 122 concordant analyses, the youngest age population and maximum age of deposition are 601 ± 6 Ma (MSWD = 0.75, pof = 0.61, n = 7; Table S1 [see ^{footnote 1}]). The biggest peaks are at ca. 603 Ma and ca. 1.21 Ga (Fig. 6), with others between ca. 2.98 Ga and ca. 987 Ma. Of the eight more faceted grains, six fall within the youngest population, and two fall in the larger mid- to late-Proterozoic population. Only two rounded grains are younger than 940 Ma. No blue tinted grains are younger than 950 Ma, but otherwise show a similar date distribution to the other grains in the sample.

Based on the youngest detrital zircon populations, depositional ages are Tonian or younger for the Blackstone Group and Plainfield Formation quartzites and Ediacaran or younger for the other samples. The Blackstone Group is intruded by Ediacaran granite, including the 606.3 ± 1.2 Ma Milford Granite and 599 ± 2 Ma Esmond Granite (Thompson et al., 2010b). The Blackstone Group quartzite, and by inference Plainfield Formation quartzite, are therefore no younger than Ediacaran. The North Attleboro Formation quartzite is Cambrian based on stratigraphic position and fossil evidence (Shaler et al., 1899; Anstey, 1979; Landing, 1988, 1996). While the Plainfield Formation quartzite lies within the Hope Valley subterrane and the Blackstone Group and North Attleboro Formation quartzites in the Esmond-Dedham subterrane, and the North Attleboro Formation quartzite is younger than the Blackstone Group quartzite samples, they all have similar detrital zircon age populations. These are consistent with those of published Avalonian samples north of the dashed blue line in Figure 1C, and in Canada (Fig. 6; cf. Severson et al., this volume). The quartzites are, therefore, discussed together in the section below. The two Newport Group samples (Newport Neck and Price Neck Formations) yielded detrital zircon age populations that are consistent between the two samples but different from the Plainfield Formation, Blackstone Group and North Attleboro Formation quartzites, and are discussed in a subsequent section.

The Blackstone Group, Plainfield Formation, and North Attleboro Formation quartzite samples all show minor detrital zircon age populations of ca. 3.2–2.6 and ca. 2.2–1.6 Ga and major populations between ca. 1.6 and 0.9 Ga, with the largest peak at ca. 1.2 Ga. Sample 328 of the Blackstone Group quartzite also shows a small ca. 2.5 Ga population, but otherwise the five samples have a gap between ca. 2.6 and ca. 2.2 Ga. The North Attleboro Formation quartzite has a ca. 0.6 Ga population that is absent in the Blackstone Group and Plainfield Formation quartzite samples. The cumulative probability diagrams look generally similar for all five samples (Fig. 7). The southern Blackstone Group quartzite shows stronger resemblance with the two Plainfield Formation quartzites, while the northern Blackstone Group quartzite shows stronger resemblance with the North Attleboro Formation quartzite. The difference between the two Blackstone Group quartzites may indicate that sources were somewhat, but not significantly, different for the two quartzites of the same unit. Zircon in the North Attleboro Formation quartzite may have been derived from the northern Blackstone Group quartzite. If the Blackstone Group and Plainfield Formation quartzites are Tonian, then their cumulative probability plots are consistent with a convergent margin setting, where the ages of the youngest detrital zircon populations are typically very close to the ages of deposition (Cawood et al., 2012). If they are Ediacaran, then the ages of the youngest detrital zircon populations are significantly older than the ages of deposition, which is more consistent with a collisional or passive margin setting (Cawood et al., 2012). The cumulative probability plot for the North Attleboro Formation quartzite is consistent with a collisional or passive margin setting (Cawood et al., 2012).

The zircon age populations of all five samples are similar to those of other samples of the southeastern New England Avalon terrane northwest of the dashed blue line in Figure 1C and described above, which generally contain small populations between ca. 3.1 and ca. 2.4 Ga and between ca. 2.2 and ca. 1.6 Ga, and larger populations between ca. 1.6 and ca. 1.0 Ga, with a large peak at ca. 1.2 Ga (Fig. 6H; Severson et al., this volume). These samples have similar detrital signatures as Avalonian (meta)sedimentary rocks in Canada (Fig. 6H; Keppie et al., 1998; Barr et al., 2003, 2019; Murphy et al., 2004; Pollock et al., 2009; Satkoski et al., 2010; Henderson et al., 2016; compiled by Severson et al., this volume). It is unclear whether the Plainfield and Blackstone Group quartzites analyzed in this study may be as old as Tonian, or whether they are Ediacaran as some of the other Plainfield Formation and Westboro Formation quartzites of the northern and western parts of the southeastern New England Avalon terrane (Hepburn et al., 2008; Thompson et al., 2012; Severson et al., this volume), or a combination. Based on detrital zircon ages, the North Attleboro Formation quartzite is younger than 601 ± 6 Ma, consistent with the better constrained Cambrian age based on stratigraphic and paleontological arguments (Shaler et al., 1899; Anstey, 1979; Landing, 1988, 1996). Based on detrital zircon age populations, all five samples are interpreted as part of the southeastern New England Avalon terrane.

The Folly River and Gamble Brook formations of the Bass River Block in the eastern Cobequid Highlands previously yielded maximum depositional ages of ca. 1 Ga (White et al., 2019b, cf. Keppie et al., 1998; Barr et al., 2003; Henderson et al., 2016). Recent results indicate that the Gamble Brook Formation is younger than 945 ± 12 Ma (White et al., this volume). It is intruded by plutons of the 622–605 Ma Bass River suite (e.g., Keppie et al., 1998; Murphy, 2002; Beresford, 2014; cf. van Staal et al., 2020). It may be an equivalent to the Blackstone Group and Plainfield Formation quartzites analyzed in this study, which have a similar potential age range and detrital zircon signature.

The Price Neck and Newport Neck formations detrital zircon signatures are different from the general Avalonian signature described above. They have main peaks at ca. 647–643 and ca. 745–733 Ma and minor populations up to ca. 3.1 Ga, especially in the Price Neck Formation. Their cumulative probability plots (Cawood et al., 2012) are indicative of a convergent margin setting. In the Price Neck Formation, the smaller ca. >745 Ma populations resemble the ca. 3.2–2.6 and ca. 2.2–0.9 Ga populations in the Avalonian Blackstone Group, Plainfield Formation, Westboro Formation, and North Attleboro Formation quartzites described above, suggesting an Avalonian (Baltican-Amazonian) affinity. The ca. 647–643 Ma population is also consistent with Avalonia. The large ca. 745–733 Ma peak, however, is unusual, and the source for these zircon grains is enigmatic.

While ca. 745–733 Ma source rocks have not been recognized in southeastern New England, arc igneous rocks of that age have been documented in the Avalon terrane of Nova Scotia, (Doig et al., 1993; White et al., 2019a, 2019b, 2020a, 2020b; Vaccaro et al., 2020; White et al., this volume). The Mount Ephraim block of the eastern Cobequid Highlands (Fig. 1A) includes ca. 800–770 Ma quartzofeldspathic, semi-pelitic, and pelitic gneiss and schist of the Mount Thom Formation, ca. 750 Ma dacitic tuff of the Dalhousie Mountain Formation, 780(?) to 730 Ma gabbroic/dioritic to granitic plutons of the Mount Ephraim suite, the intermediate to felsic ca. 734 Ma Economy River orthogneiss, and ca. 740–730 Ma dioritic rocks of the Six Mile Brook pluton (Doig et al., 1993; Vaccaro et al., 2020; White et al., 2020a and this volume). A ca. 745 Ma rhyolite porphyry exists in the Jeffers block of the eastern Cobequid Highlands (White et al., 2020a). The 650–643 Ma Hammondvale Metamorphic Suite (Fig. 1A) of the Avalonian Caledonia terrane in New Brunswick, which has been interpreted as an accretionary complex or forearc sequence (White et al., 2001; Satkoski et al., 2010), also contains ca. 770–630 Ma zircon (Barr et al., 2012).

The ca. 730 Ma Hawke Hills Tuff (Fig. 1A) of the southern Conception Bay of eastern Newfoundland (Israel, 1998; O’Brien et al., 2001) is interpreted as a possible equivalent to the similar-aged arc rocks in Nova Scotia (van Staal et al., 2020). On the Burin peninsula of southeastern Newfoundland, primitive submarine mafic volcanic rocks, oceanic gabbro, ultramafic rocks, minor intermediate and felsic igneous rocks, and associated clastic and carbonate olistostromes comprise the ca. 760 Ma fault-bounded Burin Group (Fig. 1A; Krogh et al., 1988; O’Brien et al., 1996; O’Driscoll et al., 2001; Murphy et al., 2008). U-Pb zircon ID-TIMS dates of the complex are from a 763 ± 2 Ma gabbroic sill (Krogh et al., 1988) and a 764.5 ± 2.1 Ma gabbro elsewhere in the group (Murphy et al., 2008).

The only detrital zircon signature of samples reported in the literature from eastern North America, Baltica, West Africa, and Amazonia that has a large ca. 740 Ma peak like this is sample MS04 of Bea et al. (2020), from a metasedimentary rock in the Oulad Dlim Massif in southern Morocco. Similar to the Price Neck and Newport Neck formations, it yielded large ca. 650 and ca. 740 Ma populations and scattered Meso- and Paleoproterozoic analyses. The ca. 745–733 Ma population in sample MS04, and perhaps in the Price Neck and Newport Neck formations, may represent the Pan-African I orogeny in the Anti-Atlas region of Morocco (D’Lemos et al., 2006; Hefferan et al., 2014; Triantafyllou et al., 2016, 2018; Soulaimani and Hefferan, 2017) because no other orogenies of that age are known to exist in West Africa. If correct, the ca. 647–643 Ma population would represent the Pan-African II orogeny (D’Lemos et al., 2006; Hefferan et al., 2014; Soulaimani and Hefferan, 2017). Previously, the Burin Group has been correlated with the ca. 770–700 Ma Bou Azzer ophiolite complex in the Anti-Atlas of Morocco (Krogh et al., 1988; O’Driscoll et al., 2001; cf. D’Lemos et al., 2006; El Hadi et al., 2010). These potential West African affinities for parts of what is currently interpreted as Avalonia are further discussed in the following section.

Potential origins and/or sediment source areas for Avalonia may have included Amazonia, Baltica, and/or West Africa, and are further discussed below. Igneous rocks and detrital zircon in (meta)sedimentary rocks with ca. 760–730 Ma ages are becoming increasingly recognized along the southeastern margin of Avalonia, including the Burin Group, Mount Ephraim Block, Hammondvale Metamorphic Suite, and Price Neck and Newport Neck Formations (see above). All of these domains have previously been interpreted as part of Avalonia. If correct, any interpretations for origins and sources for Avalonian rocks would have to include these domains as well. Alternatively, some or all of these domains may not be part of Avalonia. Below we briefly discuss interpreted Avalonian affinities and then discuss implications for the domains with ca. 760–730 Ma igneous rocks and/or detrital zircon populations.

Detrital zircon analyses of Avalonian rocks generally yield a predominant ca. 750–500 Ma population, and scattered ages up to Archean, suggesting an Amazonian origin (Keppie et al., 1998; Barr et al., 2003, 2012, 2019; Murphy et al., 2004; Pollock et al., 2009; Satkoski et al., 2010; Severson et al., this volume; Fig. 3). Inherited zircon in igneous rocks yielded ca. 2.7–2.5, ca. 2.3–2.2, ca. 1.5–1.4, and ca. 1.1–1.0 Ga dates (e.g., Zartman and Hermes, 1987; Bevier and Barr, 1990; Bevier et al., 1993; Barr et al., 1994; Samson, 1994; Nance and Murphy, 1996; Fig. 3). Based on detrital zircon signatures of metasedimentary rocks, inherited zircon in plutonic and volcanic rocks, and Nd model ages, Avalonia is interpreted as having a 1.3–1.0 Ga basement (Nance and Murphy, 1994, 1996; Keppie et al., 2003; Thompson and Bowring, 2000; Thompson et al., 2010b, 2012; Murphy et al., 2013; Henderson et al., 2016), which is consistent with an Amazonian origin, but not indicative of it.

A possible northwest African affinity has been suggested as an alternative based on paleomagnetic data for the Avalon terrane of Newfoundland by McNamara et al. (2001) and New England by Thompson et al. (2007, 2010a). However, Mesoproterozoic sources are not typical for northwest Africa (Fig. 3). In the Anti-Atlas, no ca. 1.2–1.0 Ga detrital zircon ages have been reported in Ediacaran to early Cambrian metasedimentary rocks (e.g., Abati et al., 2010; Avigad et al., 2012), except one detrital zircon in an Ediacaran rock (Bouougri et al., 2020). Rare ca. 1 Ga zircon grains do exist in middle and late Cambrian and Ordovician rocks of the Western Meseta domains, Anti-Atlas, High Atlas, and Sehoul Block (Fig. 1B; Avigad et al., 2012; Letsch et al., 2018; Pérez-Cáceres et al., 2017; Perez et al., 2019), and larger ca. 1 Ga populations exist in Ordovician rocks of the Eastern Meseta and Middle Atlas (Ghienne et al., 2018; Accotto et al., 2019, this volume). These may have been derived from distant sources, such as the Arabian-Nubian Shield and/or Saharan Metacraton of northeast Africa (Avigad et al., 2012; Accotto et al., 2019, this volume; Perez et al., 2019). No Mesoproterozoic or older inherited grains have been reported in Ediacaran to early Cambrian igneous rocks in the Anti-Atlas (e.g., Thomas et al., 2002; Inglis et al., 2004, 2005; Gasquet et al., 2005; D’Lemos et al., 2006; Walsh et al., 2002, 2012; Toummite et al., 2013; Bouougri et al., 2020) and Meseta domains (Letsch et al., 2018; Ouabid et al., 2017), except in the El Jadida horst (Fig. 1B) of the Western Meseta Coastal Block (El Haibi et al., 2020). Ediacaran igneous rocks in the Anti-Atlas region of West Africa only show rare ca. 2.1–2.0 Ga inherited grains (Thomas et al., 2002). Therefore, the presence of large Mesoproterozoic detrital zircon populations in Avalonian (meta)sedimentary rocks and the suggestion for a 1.3–1.0 Ga basement (see above) are probably inconsistent with a northwest African origin for Avalonia.

In Mauritania in West Africa, Neoproterozoic to Ordovician rocks (Bradley et al., 2015) yielded detrital zircon U-Pb age populations similar to those in the southeastern New England Avalon terrane. Because potential sediment source regions for these rocks in Africa would not provide the Mesoproterozoic populations, Bradley et al. (2015) interpreted their source to be on other continents, such as Baltica and/or Amazonia, with initial sediment transport having occurred during the existence of supercontinent Rodinia. Similar zircon age populations are also present in four metasedimentary rocks in the Oulad Dlim Massif in southern Morocco (Bea et al., 2020). Based on detrital zircon populations, it is possible that Avalonia was derived from the Mauritanian and adjacent southern Moroccan part of West Africa. However, the interpreted 1.3–1.0 Ga basement for Avalonia does not suggest a West African origin (see above; cf. Bradley et al., 2015). Thus, while it is possible that Avalonia was derived from West Africa, a Baltican/Amazonian origin may be more likely.

A Baltican or mixed Baltican/Amazonian affinity for Avalonia has been suggested based on paleomagnetic data, a general absence of 900–800 Ma zircon in the detrital zircon record, and Lu-Hf data (Thompson et al., 2012; Henderson et al., 2016; Severson et al., this volume). van Staal et al. (2020) built a tectonic model based on that interpretation, also including the ca. 760–730 Ma rocks described above, where Avalonia formed as Tonian passive margin on a Baltica-derived ribbon in the Mirovoi Ocean. In the model, formation of olistostromes and deformation and metamorphism by obduction of an oceanic terrane onto the ribbon prior to ca. 750 Ma were followed by formation of a 750–730 Ma arc. While this model is possible, there is no evidence for such an arc. Severson et al. (this volume) suggest that a mixed Baltican/Amazonian source was created at the time of the existence of supercontinent Rodinia, where Baltican crust was transferred to Amazonia, or vice versa, to create a mixed source terrain that later formed Avalonia. While the origin of Avalonia can be interpreted multiple ways and needs further investigation, based on the data currently available, this may be the simplest solution.

As outlined above, Baltica has no significant sources for the ca. 745–733 Ma zircon population present in the Price Neck and Newport Neck formations (Figs. 2 and 3). At that time, 800–730 Ma marine sediments including carbonates and iron formations (Melezhik et al., 2015) and 800–550 Ma siliciclastic and carbonate rift deposits (Pease et al., 2008) were deposited. Compiled U-Pb geochronological data recording magmatic events for the Sveconorwegian and Caledonian belts of Baltica (Bingen and Solli, 2009) show 1805–1130 and 1060–914 Ma zircon only in the Precambrian record. They show that Caledonian and Sveconorwegian belts are poor in Cryogenian to Ediacaran magmatism. The only significant exceptions are the Seve and Kalak Nappes in Arctic Norway (cf. Kirkland et al., 2006), which contain minor magmatic suites of 876–825, 711–680, and 571–523 Ma (Bingen and Solli, 2009) and metamorphic rims on zircon as a result of the Snøfjord tectonic event (Kirkland et al., 2006).

Amazonia does have ca. 760–700 Ma arc igneous rocks (e.g., Babinski et al., 1996; Saalmann et al., 2006; Heilbron et al., 2020), but the abundance relative to zircon sources of other ages is not as high as shown in the Price Neck and Newport Neck formations detrital zircon record. The abundance of ca. 745–733 Ma zircon, combined with the paucity of Mesoproterozoic zircon in the Price Neck and Newport Neck formations that are common for Amazonia and Baltica, and other Avalonian (meta)sedimentary rocks, suggests that the Newport Group may not have a Baltican/Amazonian affinity and a different origin than the Avalonian rocks to the northwest (including the North Attleboro Formation, Blackstone Group, and Plainfield Formation quartzites). It is possible that the Newport Group samples represent a stratigraphic interval during which ca. 745–733 Ma detrital zircon was supplied by sources distinct from those of the older or coeval Blackstone Group and Plainfield Formation and the younger North Attleboro Formation, which do have consistent sources. However, this may be unlikely, because it would require changes in stratigraphic provenance such that the source for the ca. 745–733 Ma zircon contributed only to the Newport Group, but not to the Blackstone Group, Plainfield Formation, and North Attleboro Formation. Furthermore, such a source is not known to exist currently in the vicinity of the Newport Group. Alternatively, the Newport Group may instead represent a different terrane or crustal block, and not a part of Avalonia. Because the nature and boundaries of this crustal block remains unclear, we refer to it as the Newport Block.

The scarcity of Mesoproterozoic zircon in the Newport Block may suggest a northwest African origin. The ca. 770–700 Ma Bou Azzer and Tasriwine ophiolites of the Anti-Atlas in Morocco (e.g., Samson et al., 2004; El Hadi et al., 2010; Hefferan et al., 2014; Soulaimani and Hefferan, 2017), and/or the possible correlative Burin Group in Newfoundland (Krogh et al., 1988; O’Driscoll et al., 2001; see above) are unlikely source rocks for zircon in the Price Neck and Newport Neck formations, because ophiolites are poor in zircon. However, in the northeastern Anti-Atlas, the ca. 770–750 Ma Tachakoucht-Bou Azzer and Tichibanine-Ben Lgrad oceanic island arcs, and arcs formed during the subsequent 750–700 Pan-African I or Iriri-Tichibanine orogeny, are viable sources (D’Lemos et al., 2006; Hefferan et al., 2014; Triantafyllou et al., 2016, 2018; Soulaimani and Hefferan, 2017). No other major orogenies of that age are known to exist in West Africa. Arc-related orthogneiss in the northwest African-derived Cadomian Armorican massif of France yielded ca. 750 Ma ages (Egal et al., 1996; Samson et al., 2003) and may have been part of the same orogen.

The model of ca. 770–750 Ma ophiolite and oceanic arc formation and subsequent 750–730 Ma orogenesis and obduction in the Anti-Atlas (D’Lemos et al., 2006; Hefferan et al., 2014; Triantafyllou et al., 2016, 2018; Soulaimani and Hefferan, 2017) is similar to van Staal et al.’s (2020) model of ophiolite and arc obduction prior to ca. 750 Ma followed by formation of a 750–730 Ma arc in the vicinity of Baltica. Because evidence for source rocks exists in the Anti-Atlas of Morocco, but not in Baltica, we suggest that the Newport Block (and perhaps associated domains, see below) has a northwest African origin.

Based on data currently available, the extent of the Newport Block is unclear. It may be as small as the southern part of Aquidneck Island, Rhode Island (Fig. 1D), containing the Price Neck and Newport Neck formations of the Newport Group, or it may extend farther on land. Previously, based on rock types, all Ediacaran to Cambrian metasedimentary rocks of both the Conanicut and Newport Groups (Fig. 1D) southeast of the Beaverhead shear zone, a northeast-trending shear zone along the southeastern Narragansett Basin (Figs. 1C–1D), have been considered to be a separate and potentially non-Avalonian terrane, called the “Bulgarmarsh” terrane (Skehan et al., 1978; Rast and Skehan, 1990; Skehan and Rast, 1990; Carter and Mosher, 2013). We attempted to conduct U-Pb detrital zircon geochronology on samples from the Conanicut Group in order to test its origin. Unfortunately, these samples yielded no zircon large enough for analysis. However, the Newport Block may include all rocks southeast of the Beaverhead shear zone. In fact, it may extend as far north as the dashed black line in Figure 1C, but no farther west than exposures of the North Attleboro Formation, Blackstone Group, and Plainfield Formation quartzites. The maximum possible extent of the Newport Block is outlined in Figure 1C.

In addition to the Newport Block, the Burin Group, Mount Ephraim Block, and possibly Hawke Hill Tuff and Hammondvale Metamorphic Suite (marked red in Fig. 1A) may have northwest African affinities, while other parts of Avalonia may have Amazonian and/or Baltican origins. There are various ways in which these northwest African blocks may have arrived. The northwest African crustal block containing metasedimentary rocks below the Georges Bank in offshore Massachusetts is interpreted as having remained behind after the breakup of Pangea (Kuiper et al., 2017; see above; Fig. 8A). It is possible that other northwest African blocks arrived the same way. If correct, then the Pangean suture zone lies on land, northwest of the Newport Block, and perhaps northwest of the Burin Group. It is more difficult to explain the Mount Ephraim Block, Hawke Hill Tuff, and Hammondvale Metamorphic Suite in the same way because they are located more inland.

It is alternatively possible that a piece or pieces of northwest Africa had rifted earlier from the African continent and accreted to Laurentia as a terrane in a manner similar to (Fig. 8B), and perhaps as part of (Fig. 8C), Avalonia. Rocks with the same detrital signature as those below the Georges Bank have been found in Ganderian rocks near the interpreted northwestern margin of Avalonia in North Islesboro in the Penobscot Bay of southern Maine (Reusch et al., 2018; Fig. 1A) and on Grand Manan Island in southern New Brunswick (Barr et al., 2019; Fig. 1A). Their origins are not clear, but they are too far on the continent to have been stranded as a result of formation and breakup of Pangea. They may reflect a northwest African block that had amalgamated with Laurentia prior to the formation of Pangea, perhaps as a terrane (Fig. 8B) or part of a composite terrane (Fig. 8C). Likewise, the Newport Block, Burin Group, Mount Ephraim Block, Hawke Hill Tuff, and/or Hammondvale Metamorphic Suite may reflect northwest African terranes that amalgamated with Laurentia prior to the formation of Pangea. They may have arrived as part of Avalonia, or perhaps as separate terranes, or a combination. If those crustal blocks arrived as part of Avalonia, they must have amalgamated with Avalonia at an earlier time. If they arrived separately, they are truly a separate terrane or terranes. In southeastern New England it is also possible that the Newport Block, perhaps including the metasedimentary rocks from the COST No. G-1 well in the Georges Bank (Fig. 1; Kuiper et al., 2017), represent Proterozoic basement to the Cambrian to Devonian rocks of the Meguma terrane exposed in Nova Scotia.

Regardless of which of the above interpretations is correct, it is clear that the southeastern New England Avalon terrane is truly a composite terrane or terrain, with at least two domains with different origins. Based on detrital zircon results, the boundary does not lie between the previously interpreted Hope Valley and Esmond-Dedham subterranes, but between the Newport Group and rocks to the northwest. This boundary may be (1) a terrane boundary, where the Newport Group and associated rocks arrived separately from the Avalon terrane; (2) a subterrane, where they amalgamated with Avalonia prior to accretion to Laurentia; or (3) the Pangean suture zone. It is thus possible that the Newport Block is not truly part of the Avalon terrane, but a separate terrane or crustal block.

In the southeastern New England Avalon terrane, the two Plainfield Formation quartzite samples of the Hope Valley subterrane and two Blackstone Group quartzite samples from the Esmond-Dedham subterrane all have Tonian youngest detrital zircon age populations and similar detrital zircon age populations. Therefore, they may all represent the same unit and based on detrital zircon analysis, there is no need to separate between the Hope Valley and Esmond-Dedham subterranes. The North Attleboro Formation quartzite has the same detrital zircon signature as the Plainfield Formation and Blackstone Group quartzites, and an additional Ediacaran population, all consistent with an Avalonian detrital zircon signature of published data from Westboro Formation quartzite and Plainfield Formation samples northwest of the blue dashed line in Figure 1C and of (meta)sedimentary rocks in the Canadian Avalon terranes. These are all interpreted as Avalon terrane with an Amazonian/Baltican affinity.

The Price Neck and Newport Neck formations of the Newport Group in southeastern Rhode Island display a different detrital zircon signature, which is most consistent with a northwest African affinity, and different from the Avalon terrane to the northwest. These rocks, referred to as the Newport Block, may (1) be part of a composite Avalon terrane, where these northwest African rocks may have amalgamated with the Amazonian/Baltican part of the Avalon terrane prior to mid-Paleozoic amalgamation with Laurentia; (2) have arrived as a separate terrane after accretion of the Avalon terrane; or (3) have arrived during the formation of Pangea and been stranded after the breakup of Pangea, as has been proposed for rocks of the Georges Bank.

We thank Chris Hepburn, Allison Severson, Francis Macdonald, Abder Soulaimani, Cristina Accotto, and Andrea Mills for fruitful and interesting discussions of various aspects of the geology, and Chris Hepburn for commenting on an earlier version of this manuscript. Reviews by Francis Macdonald, David Jesús Martínez Poyatos, and volume editor Margaret Thompson helped us improve this manuscript significantly. This work was supported by NSF grant EAR-1827429 and is part of International Geoscience Programme project IGCP683. Gold Open Access publishing was partially funded by a Mines Open Access Mini Grant Award.

Zircon was analyzed by LA-ICP-MS using a ThermoElectron X-Series II quadrupole ICP-MS and New Wave Research UP-213 Nd:YAG UV (213 nm) laser ablation system. In-house analytical protocols, standard materials, and data reduction software were used for acquisition and calibration of U-Pb dates and a suite of high field strength elements (HFSE) and rare earth elements (REE). Zircon was ablated with a laser spot of 25 µm wide using fluence and pulse rates of 5 J/cm² and 5 Hz, respectively, during a 45-second analysis (15 s gas blank, 30 s ablation) that excavated a pit ~25 µm deep. Ablated material was carried by a 1.2 L/min He gas stream to the nebulizer flow of the plasma. Dwell times were 5 ms for Si and Zr, 200 ms for ⁴⁹Ti and ²⁰⁷Pb, 80 ms for ²⁰⁶Pb, 40 ms for ²⁰²Hg, ²⁰⁴Pb, ²⁰⁸Pb, ²³²Th, and ²³⁸U, and 10 ms for all other HFSE and REE. Background count rates for each analyte were obtained prior to each spot analysis and subtracted from the raw count rate for each analyte. Ablations pits that appear to have intersected glass or mineral inclusions were identified based on Ti and P. U-Pb dates from these analyses are considered valid if the U-Pb ratios appear to have been unaffected by the inclusions. Analyses that appear contaminated by common Pb were rejected based on mass 204 being above baseline. For concentration calculations, background-subtracted count rates for each analyte were internally normalized to ²⁹Si and calibrated with respect to NIST SRM-610 and -612 glasses as the primary standards. Temperature was calculated from the Ti-in-zircon thermometer (Watson et al., 2006). Because there are no constraints on the activity of TiO₂, an average value in crustal rocks of 0.8 was used.

Data were collected in five experiments in May 2019, seven experiments in November 2019, and one experiment in January 2021. For U-Pb and ²⁰⁷Pb/²⁰⁶Pb dates, instrumental fractionation of the background-subtracted ratios was corrected, and dates were calibrated with respect to interspersed measurements of zircon standards and reference materials. The primary standard Plešovice zircon (Sláma et al., 2008) was used to monitor time-dependent instrumental fractionation based on two analyses for every 10 analyses of unknown zircon. A secondary correction to the ²⁰⁶Pb/²³⁸U dates was made based on results from the zircon standards Seiland (531 Ma; see Appendix B), FC1 (1098 Ma, Paces and Miller, 1993), and 91500 (Wiedenbeck et al., 1995), which were treated as unknowns and measured once for every 10 analyses of unknown zircon. Two secondary reference zircons were analyzed during each experiment. These results showed a linear age bias of several percent that is related to the ²⁰⁶Pb count rate. The secondary correction is thought to mitigate matrix-dependent variations due to contrasting compositions and ablation characteristics between the Plešovice zircon and other standards (and unknowns).

Radiogenic isotope ratio and age error propagation for all analyses includes uncertainty contributions from counting statistics and background subtraction. Weighted mean dates are calculated from equivalent dates (i.e., pof of >0.05) by first using Isoplot 3.0 (Ludwig, 2003), using errors on individual dates that do not include a standard calibration uncertainty, and then a standard calibration uncertainty is propagated into the error on the weighted mean date. This uncertainty is the local standard deviation of the polynomial fit to the interspersed primary standard measurements versus time for the time-dependent, relatively larger ²⁰⁶Pb/²³⁸U fractionation factor and the standard error of the mean of the consistently time-invariant and smaller ²⁰⁷Pb/²⁰⁶Pb fractionation factor. These uncertainties are 0.3%–0.7% (2σ) for ²⁰⁶Pb/²³⁸U and 0.1%–0.6% (2σ) for ²⁰⁷Pb/²⁰⁶Pb. Age interpretations are based on ²⁰⁷Pb/²⁰⁶Pb dates for analyses with ²⁰⁷Pb/²⁰⁶Pb and ²⁰⁶Pb/²³⁸U dates >800 Ma. Otherwise, interpretations are based on ²⁰⁶Pb/²³⁸U dates. Analyses with >15% positive or negative discordance are not considered below but are included in the Wetherill Concordia diagrams in Table S1. Errors on the dates are at 2σ. Probability density plots and weighted means were created using Isoplot version 4.15 (Ludwig, 2012). Maximum depositional age estimates for detrital zircon samples are reported using the youngest single concordant zircon grain for all samples, and the (preferred) weighted mean of the youngest population of dates that overlap at the 2σ uncertainty level, where possible (Dickinson and Gehrels, 2009).

U-Pb dates were obtained by the CA-TIMS method from analyses composed of single fragments of Seiland zircon (Table S2 [see ^{footnote 1}]), modified after Mattinson (2005). Seiland zircon was obtained from J. Hanchar at Memorial University of Newfoundland during an inter-laboratory geochronology comparison.

Fragments were placed in a muffle furnace at 900 °C for 60 h in quartz beakers and then put into 3 ml Teflon PFA beakers and loaded into 300 μl Teflon PFA microcapsules. Fifteen microcapsules were placed in a large-capacity Parr vessel and the zircon partially dissolved in 120 μl of 29 M HF for 12 h at 180 °C. Zircon was returned to 3 ml Teflon PFA beakers, HF was removed, and zircon was immersed in 3.5 M HNO₃, ultrasonically cleaned for an hour, and fluxed on a hotplate at 80 °C for an hour. The HNO₃ was removed, and zircon was rinsed twice in ultrapure H₂O before being reloaded into the 300 μl Teflon PFA microcapsules (rinsed and fluxed in 6 M HCl during sonication and washing of the zircon) and spiked with the EARTHTIME mixed ²³³U-²³⁵U-²⁰⁵Pb tracer solution (ET535). Zircon was dissolved in Parr vessels in 120 μl of 29 M HF with a trace of 3.5 M HNO₃ at 220 °C for 48 h, dried to fluorides, and re-dissolved in 6 M HCl at 180 °C overnight. U and Pb were separated from the zircon matrix using an HCl-based anion-exchange chromatographic procedure (Krogh, 1973), eluted together, and dried with 2 µl of 0.05 N H₃PO₄.

Pb and U were loaded on a single outgassed Re filament in 5 µl of a silica-gel/phosphoric acid mixture (Gerstenberger and Haase, 1997), and U and Pb isotopic measurements made on a GV Isoprobe-T multicollector TIMS equipped with an ion-counting Daly detector. Pb isotopes were measured by peak-jumping all isotopes on the Daly detector for 160 cycles and corrected for 0.18 ± 0.03%/a.m.u. (1 σ) mass fractionation. Transitory isobaric interferences due to high-molecular weight organics, particularly on ²⁰⁴Pb and ²⁰⁷Pb, disappeared within ~30 cycles, while ionization efficiency averaged 10⁴ cps/pg of each Pb isotope. Linearity (to ≥1.4 × 10⁶ cps) and the associated deadtime correction of the Daly detector were determined by analysis of NBS982. Uranium was analyzed as UO₂⁺ ions in static Faraday mode on 10¹² ohm resistors for 300 cycles and corrected for isobaric interference of ²³³U¹⁸O¹⁶O on ²³⁵U¹⁶O¹⁶O with an ¹⁸O/¹⁶O of 0.00206. Ionization efficiency averaged 20 mV/ng of each U isotope. U mass fractionation was corrected using the known ²³³U/²³⁵U ratio of the ET535 tracer solution.

CA-TIMS U-Pb dates and uncertainties were calculated using the algorithms of Schmitz and Schoene (2007), ET535 tracer solution (Condon et al., 2015) with calibration of ²³⁵U/²⁰⁵Pb = 100.233, ²³³U/²³⁵U = 0.99506, and ²⁰⁵Pb/²⁰⁴Pb = 11268, and U decay constants recommended by Jaffey et al. (1971) and ²³⁸U/²³⁵U of 137.818 (Hiess et al., 2012). ²⁰⁶Pb/²³⁸U ratios and dates were corrected for initial ²³⁰Th disequilibrium using D_Th/U = 0.2 ± 0.1 (2σ) and the algorithms of Crowley et al. (2007), resulting in an increase in the ²⁰⁶Pb/²³⁸U dates of ca. 0.09 Ma. All common Pb in analyses was attributed to laboratory blank and subtracted based on the measured laboratory Pb isotopic composition and associated uncertainty. U blanks are estimated at 0.013 pg.

A weighted mean ²⁰⁶Pb/²³⁸U date is calculated from equivalent dates (pof >0.05) using Isoplot 3.0 (Ludwig, 2003). Errors on weighted mean dates are given as ± x / y / z, where x is the internal error based on analytical uncertainties only, including counting statistics, subtraction of tracer solution, and blank and initial common Pb subtraction, y includes the tracer calibration uncertainty propagated in quadrature, and z includes the ²³⁸U decay constant uncertainty propagated in quadrature. Internal error should be considered when comparing our date with ²⁰⁶Pb/²³⁸U dates from other laboratories that used the same tracer solution or a tracer solution that was cross-calibrated using EARTHTIME gravimetric standards. Error including the uncertainty in the tracer calibration should be considered when comparing our date with those derived from other geochronological methods using the U-Pb decay scheme (e.g., LA-ICP-MS). Error including uncertainties in the tracer calibration and ²³⁸U decay constant (Jaffey et al., 1971) should be considered when comparing our date with those derived from other decay schemes (e.g., ⁴⁰Ar/³⁹Ar, ¹⁸⁷Re-¹⁸⁷Os). Errors are at 2σ.

Six zircon fragments from Seiland yield equivalent ²⁰⁶Pb/²³⁸U dates with a weighted mean of 530.71 ± 0.14 / 0.21 / 0.59 Ma (MSWD = 0.4, pof = 0.82). This is the interpreted igneous crystallization age.