Distinct 18O depletion is characteristic of a large majority of the 620–550 Ma felsic igneous rocks of Avalonia in the northern Appalachian orogen. Neoproterozoic rocks in the Boston Avalon terrane have the lowest δ18OWR values (≥–3.1‰), followed by the Mira terrane in Cape Breton Island and the Caledonia terrane in New Brunswick (≥–1.2‰), the Avalon terrane in Newfoundland (≥+2.8‰), and the Antigonish Highlands in Nova Scotia (≥+5.3‰). In contrast, this depletion of 18O is observed in very few of the Paleozoic felsic igneous rocks from these Avalonian terranes, and also in very few of the Neoproterozoic and Paleozoic felsic igneous rocks from the inboard Ganderian terranes. The low-18O character of the Neoproterozoic igneous rocks is related to regional pervasive, post-magmatic alteration by predominantly meteoric-hydrothermal fluids (δ18OH2O ∼–6‰ to –4‰) at 200–450 °C. The alteration likely occurred during late Neoproterozoic transtensional extension of Avalonia. Large-scale fluid infiltration and circulation within the Avalonian crust accompanied this extension with development of pull-apart sedimentary basins and extension-related magmatism that were the prelude to Cambrian submergence of Avalonia. This regional 18O depletion provides a geochemical fingerprint by which Avalonia can be distinguished from other peri-Gondwanan terranes. These data suggest that Avalonia existed as a composite terrane on the Gondwanan margin in the Neoproterozoic, separate from Ganderia.
The comparative tectonostratigraphic histories and relationships among the peri-Gondwanan composite terranes comprising Avalonia and Ganderia in the northern Appalachian orogen have been the subject of intense debate (e.g., Hibbard et al. 2007, and references therein). Originally, Williams (1979) defined Avalonia (then known as the Avalon zone) in the northern Appalachian orogen to include the Boston area of southeastern New England, coastal Maine, southern New Brunswick, most of Prince Edward Island, northern mainland Nova Scotia, all of Cape Breton Island, and eastern Newfoundland (Fig. 1, inset). Subsequent fieldwork, geochronology, and other petrological, isotopic, and geophysical data have led to the identification of distinct tectonostratigraphic packages in Cape Breton Island and southern New Brunswick (Fig. 1) that display Ganderian rather than Avalonian affinities (e.g., Barr and Raeside 1989; Barr et al. 1998; Barr and White 1999; Samson et al. 2000). As a result, a number of studies have redefined Avalonia as a smaller area, and Ganderia has been expanded (Fig. 1; Hibbard et al. 2006). However, the assignation of a number of these tectonostratigraphic packages to Ganderia or Avalonia and the timing of juxtaposition of these terranes remain controversial (e.g., Keppie et al. 1991; Dallmeyer and Nance 1992; van Staal et al. 1996; Johnson and McLeod 1996; Keppie et al. 2000). The relationship of Ganderia and West Avalonia of the northern Appalachians to Carolinia in the southern Appalachian orogen and East Avalonia in Europe also remains problematic (e.g., Nance and Murphy 1996; McIlroy and Horák 2006; Hibbard et al. 2007).
Previous isotopic studies in Newfoundland (Fryer et al. 1992), Cape Breton Island (Ayuso et al. 1996), and New Brunswick (Whalen et al. 1994, 1996) noted unusual oxygen-isotope compositions in some igneous rocks from Avalonia, distinct from those in the inboard Ganderian terranes. A recent detailed oxygen-isotope study (Potter et al. 2008) confirmed that Neoproterozoic igneous rocks of the Avalonian Mira and Caledonia terranes of Cape Breton Island and southern New Brunswick, respectively, have been affected by pervasive 18O depletion. This alteration is absent from the more inboard Ganderian Bras d’Or, Aspy, Brookville, Kingston, and New River terranes (Potter et al. 2008), all once thought to be part of the Avalon zone (Williams 1979; Keppie et al. 1991). Potter et al. (2008) suggested that the 18O depletion resulted from pervasive post-magmatic hydrothermal alteration of the Avalonian crust, most likely during 560–550 Ma extension-related faulting.
In this paper, we report new whole-rock oxygen-isotope data for Neoproterozoic and Paleozoic felsic igneous units from Avalonian and Ganderian terranes in Newfoundland, northern mainland Nova Scotia, and the Boston area of Massachusetts, together with a compilation of earlier data. These data, representing a total of 290 granitoid and felsic volcanic samples, are then used to provide a better understanding of the tectonomagmatic and juxtaposition histories of these terranes and possibly other peri-Gondwanan “suspect” terranes.
The whole-rock oxygen-isotope results (δ18OWR) reported here were measured at the Laboratory for Stable Isotope Science (LSIS) at the University of Western Ontario. The oxygen-isotope analyses were performed following the method of Clayton and Mayeda (1963), as modified by Borthwick and Harmon (1982). Whole-rock powders and silicate mineral standards were reacted in Ni-reaction vessels overnight at 550 °C with ClF3 to release O2. The O2 was then converted to CO2 over red-hot graphite and the resulting CO2 analyzed using an Optima dual-inlet mass spectrometer. Average δ18O values (versus VSMOW (Vienna standard mean ocean water)) obtained for NBS-28 (quartz), NBS-30 (biotite), and KGa-1 (kaolinite) standards were +9.6‰ ± 0.1‰, +5.0‰ ± 0.2‰, and +21.1‰ ± 0.4‰ over the period of the study, which compares well with accepted values of +9.6‰, +5.1‰, and +21.5‰, respectively. Reproducibility of the δ18OWR values of samples was generally better than ± 0.2‰.
Figure 2 illustrates all available oxygen-isotope data for felsic igneous units from Avalonia and Ganderia. These data include new analyses for 66 samples from Newfoundland, northern mainland Nova Scotia, and the Boston area, and previously reported data for Cape Breton Island and southern New Brunswick (Whalen et al. 1994, 1996; Ayuso et al. 1996; Potter et al. 2008) and Newfoundland (Fryer et al. 1992). Table 1 provides a compilation of oxygen-isotope data for Newfoundland, northern mainland Nova Scotia, and the Boston area within Massachusetts. A complete compilation of oxygen-isotope data for New Brunswick and Cape Breton Island is provided in Potter et al. (2008).
A large population of samples from the Avalonian ca. 620 Ma felsic units (Fig. 2a) shows a shift to δ18O values much lower (as low as –1.2‰) than the range typical of “normal” granitic rocks (~+7‰ to +10‰; e.g., Taylor 1968), with the majority having δ18OWR values of +2‰ to +8‰. The Avalonian ca. 600 Ma felsic units from northern mainland Nova Scotia have a narrow range of δ18OWR values (+5.3‰ to +8.1‰; Fig. 2b), with the majority of samples showing only a small shift downwards from “normal” granitic rock δ18O values. The ca. 600 Ma felsic units from the Boston area are much more depleted in 18O; the majority of δ18OWR values are <+5‰, and some are as low as –3.1‰ (Fig. 2b). The Avalonian ca. 570–560 Ma felsic units from Newfoundland, Cape Breton Island, and southern New Brunswick also show distinct 18O depletion (Fig. 2c), with a large proportion of samples having δ18OWR values of +2‰ to +4‰ (Fig. 2c). In marked contrast to the Neoproterozoic units, the majority of δ18OWR values for Paleozoic felsic units in Avalonia fall within the range of δ18OWR values typically obtained for “normal” granitic rocks, with the exception of five samples with low δ18OWR values and two samples with high δ18OWR values (Fig. 2d).
The majority of δ18OWR values for Neoproterozoic felsic units in the Ganderian terranes of Newfoundland, Cape Breton Island, and New Brunswick lie within the range of “normal” granitic rocks, with only three samples having anomalously low δ18OWR values (Fig. 2e). Ganderian Paleozoic felsic units exhibit the same pattern (Fig. 2f).
Neoproterozoic felsic igneous rocks from Avalonia are characterized by pervasive 18O depletion. This effect is most strongly demonstrated by samples from the Boston area, which have some of the lowest δ18OWR values (≥–3.1‰), followed by samples from Cape Breton Island and New Brunswick (≥–1.2‰), Newfoundland (≥+2.8‰), and then northern mainland Nova Scotia (≥+5.3‰). Rocks with such low-18O compositions are not characteristic of Paleozoic units from Avalonia (Fig. 2d) or Neoproterozoic and Paleozoic units from the inboard Ganderian terranes (Figs. 2e, 2f). This pattern suggests that regional 18O depletion occurred before the earliest Paleozoic magmatic activity (∼490 Ma; Table 1) and affected all of the Neoproterozoic igneous rocks in Avalonia up to ∼550 Ma (Table 1). Such 18O depletion did not affect the rocks of Ganderia.
The Neoproterozoic rocks of Avalonia exhibit classic characteristics of postmagmatic interaction with hydrothermal fluids at ∼200–450 °C (see Potter et al. 2008). These traits include
(1) an abundance of recrystallized, turbid feldspars replaced by sericite, epidote, and hematite;
(2) ubiquitous replacement of primary mafic phases by chlorite, epidote, and titanite; and
(3) numerous generations of veins consisting of quartz–epidote, epidote–chlorite–quartz, quartz–carbonate, and chlorite–white mica–carbonate–zeolite minerals.
Potter et al. (2008) suggested that the hydrothermal fluids were predominantly meteoric in origin. Higher temperature alteration associated with the formation of epidote (∼350–450 °C) produced 18O depletion in the majority of samples (i.e., δ18O = –3‰ to +3‰); alteration at lower temperatures, associated with formation of pale green chlorite, carbonate, and zeolite (∼200 °C), caused 18O enrichment (+11‰ to +14‰) in a few samples (Figs. 2a, 2c). The distribution of 18O depletion is broadly ubiquitous with little evidence of any prominent and (or) localized zonation (see Potter et al. 2008).
To investigate how pervasive hydrothermal alteration could occur on such a large scale, we have examined the tectonic and magmatic history of individual terranes that compose Avalonia. A simplified geological history for each terrane is illustrated in Fig. 3. The Avalonian terranes all have similar Neoproterozoic history. Formation of mature, subduction-related island arcs resulted in voluminous magmatism in each terrane at various times between 620 and 570 Ma (Fig. 3; e.g., Thompson et al. 1996; Barr and White 1996; O’Brien et al. 1996; Murphy 2006). This regime was then replaced by extension-related bimodal and in some places peralkaline magmatism. Initiation of the transform extensional regime along the Gondwanan margin arose from the termination of subduction caused by rift-trench collision (Fig. 3; e.g., Landing 2004).
Extension occurred diachronously across the terranes with voluminous peralkaline magmatism in Newfoundland starting at ca. 590 Ma (e.g., O’Brien et al. 1996) and bimodal magmatism in Cape Breton Island and New Brunswick occurring later at ca. 560–550 Ma (Fig. 3; e.g., Barr and White 1996). Peralkaline Neoproterozoic magmatism was limited in the Boston Avalon (the Dartmouth pluton, dated at 595 Ma; Hermes and Zartman 1992) and in the Avalon of northern mainland Nova Scotia (the Georgeville pluton in the Antigonish Highlands, 580 Ma; Murphy et al. 1998). Subduction-related magmatism waned in the Boston area at ∼590 Ma. It was replaced by deposition of the Boston Bay Group in the rapidly forming sedimentary rift basins that reflected a transition to a rift-related environment some time between 590 and 550 Ma (Fig. 3; Thompson et al. 1996). In the Antigonish Highlands, subduction-related volcanism ceased at ca. 600 Ma with transition from a continental subaerial island arc to a submarine back-arc basin, associated with deposition of deep marine turbidites and extrusion of continental tholeiitic basalts (Fig. 3; Murphy 2006). By 540 Ma, a stable marine platform had developed on Avalonia, and sediments containing Acado-Baltic or Avalonian fauna were deposited (Fig. 3; e.g., Landing 1996, 2004).
The Paleozoic histories of the individual terranes show some divergence (Fig. 3). The Avalon and Mira terranes of Newfoundland and Cape Breton Island show little sign of Paleozoic magmatic activity until the Devonian, at which time there was emplacement of within-plate granitoid rocks related to the docking and accretion of Avalonia and Meguma to Laurentia and Ganderia (Barr and Macdonald 1992; O’Brien et al. 1996). In northern mainland Nova Scotia, the Antigonish Highlands record a distinctly different Paleozoic history, with Ordovician extension-related volcanism and deposition of Silurian – Early Devonian transgressive sedimentary sequences (Murphy et al. 1996; Hamilton and Murphy 2004). Within-plate Devonian–Carboniferous granitic–gabbroic plutonism occurred during the juxtaposition of northern mainland Nova Scotia with the Meguma terrane (Doig et al. 1996; Piper and Pe-Piper 2001; Murphy and Keppie 2005). In the Caledonia terrane of southern New Brunswick, Paleozoic magmatism is largely absent with only minor extension-related volcanism of possible Silurian and Devonian ages (Barr and White 1999). The Boston area experienced Late Devonian bimodal magmatism related to rifting and strike-slip motion (Thompson and Hermes 2003) that might have been linked to Meguma docking in more northerly terranes. The Boston area also experienced widespread Silurian–Devonian plutonism (Hermes and Zartman 1985; Hepburn et al. 1993; Acaster and Bickford 1999).
From this history, we conclude that regional influx of hydrothermal fluids most likely occurred during transform faulting and extension within these terranes in the late Neoproterozoic (Figs. 3, 4a). Extension-related magmatism at this time provided the heat source needed to drive hydrothermal fluid circulation. Opening of rift–wrench basins, with associated faulting, provided conduits for infiltration of fluids deep into the Avalonian crust (Fig. 4b). For the most part, these hydrothermal fluids were meteoric in origin, reflective of the collapse of the subaerial island arc before final submergence and cooling by ca. 540 Ma (Fig. 3). A possible exception is provided by the Antigonish Highlands of northern mainland Nova Scotia, where Neoproterozoic granitic rocks have undergone only limited 18O depletion (Fig. 2b). The Antigonish Highlands have a somewhat different history than the other Appalachian Avalonian terranes. By 620–600 Ma, this terrane had developed into a submarine back-arc basin (Murphy 2006). Hence, the δ18OWR values of the samples from the Antigonish Highlands may indicate
1. limited hydrothermal alteration of the crust within this submarine back-arc setting and (or)
2. interaction with seawater-derived hydrothermal fluids (0‰) at ∼250 °C (Cole et al. 1992) rather than the meteoric water (δ18O = –6‰ to –4‰; Potter et al. 2008) that affected the other Avalonian terranes.
The transform–extensional regime that made possible large-scale infiltration and circulation of meteoric water in the Avalonian crust could be similar to that observed in California and Mexico today. The cessation of subduction and the transition to a strike-slip regime at ∼16–12 Ma in these latter areas has led to the formation of extensive geothermal fields dominated by meteoric water and significant oxygen-isotope exchange between the heated water and the host volcanic rocks (e.g., Donnelly-Nolan et al. 1993; Portugal et al. 2000; Sherlock 2005; Camprubí et al. 2008).
At the time of transtension in Avalonia, subduction and arc-related magmatism were still ongoing in Ganderia (Fig. 4a; Barr et al. 1998; Hibbard et al. 2007). This arc-related magmatism continued well into the Cambrian (∼510 Ma; White and Barr 1996) before the formation of a passive margin in the Middle Cambrian (Hibbard et al. 2007). The large-scale extensional magmatism and opening of rift-related basins observed in Avalonia are absent in Ganderia. This tectonomagmatic history, together with the distinctly different Nd-isotope basement signatures (e.g., Barr and Hegner 1992; Samson et al. 2000) and the absence of widespread 18O depletion of Neoproterozoic rocks in Ganderia, suggest that Avalonia and Ganderia were separate micro-continents, with Ganderia situated somewhere farther along the Gondwanan margin (Fig. 4a). The widespread 18O depletion of the Avalonian crust from Newfoundland to Boston further suggests that Avalonia was a coherent composite arc terrane on the Gondwanan margin in the Neoproterozoic.
This composite terrane may also have included those parts of Avalonia now situated in the United Kingdom and possibly Carolinia in the southern Appalachian orogen of North America. The presence or absence of regionally extensive 18O depletion in Neoproterozoic igneous rocks from these terranes could provide insight into their relationship to and juxtaposition history with Avalonia. We are unaware of oxygen-isotope data for the East Avalonian terranes in the United Kingdom. However, there is some indication that Carolinia may have been affected in the Neoproterozoic by processes similar to those responsible for the alteration of Avalonia. Feiss et al. (1993) and Feiss (1982) described intense, localized 18O depletion and enrichment associated with alteration and gold mineralization in the Neoproterozoic Piedmont volcanogenic massive sulphide deposits of the Carolinian terrane, as well as regional, post-depositional alteration of Neoproterozoic rocks in the Carolina Slate Belt. Klein and Criss (1988) reported δ18O values as low as 0‰ for several Neoproterozoic granitic plutons and associated volcanic rocks located throughout the Carolina Slate Belt, including the Pilot Mountain porphyry system. These results also suggest large-scale hydrothermal interaction with meteoric water. Klein et al. (2007) also described 18O depletion and enrichment associated with the Russell Gold Deposit. In contrast, as is the case for Paleozoic intrusive rocks from Avalonia, the 325–300 Ma granitic plutons in Carolinia have δ18OWR values within the “low-normal” range for igneous rocks (+6‰ to +8‰; Wenner 1981).
These results could be interpreted to suggest that 18O depletion of Carolinia occurred as it did in Avalonia and that there is a linkage between the two domains. However, in the studies of Carolinia, 18O depletion (and (or) enrichment) is always linked to localized synmagmatic hydrothermal systems that produced porphyry-style and high-sulphidation epithermal deposits in the Neoproterozoic igneous and sedimentary rocks of the Slate Belt (Klein et al. 2007). Because of their similar Neoproterozoic–Cambrian magmatic and tectonothermal histories, it has been suggested recently that Carolinia is more closely associated with Ganderia than Avalonia (Hibbard et al. 2007). Arc-related magmatism continued into the Cambrian in Carolinia (up to ∼535 Ma), associated with arc–arc collision with the Charlotte terrane. This collision produced intense localized metamorphism, unlike the arc–transform transition in Avalonia (Hibbard et al. 2007). Whether the Neoproterozoic igneous rocks in Carolinia experienced regional 18O depletion in a similar fashion to Avalonia, therefore, still remains to be determined.
Avalonian Neoproterozoic igneous rocks of eastern Newfoundland, Cape Breton Island, southern New Brunswick, the Boston area (Massachusetts), and perhaps northern mainland Nova Scotia have experienced pervasive hydrothermal alteration that produced 18O depletion in the majority of igneous units. This 18O depletion likely occurred during transtension of Avalonia between 600 and 550 Ma, and was associated with the development of pull-apart basins and extension-related magmatism. Hydrothermal fluids infiltrated the Avalonian crust at this time before its subsequent submergence at ca. 545–540 Ma. The distribution and scale of this 18O depletion in Neoproterozoic igneous rocks of Avalonia potentially provides a geochemical fingerprint that is useful in distinguishing components of Avalonia from other peri-Gondwanan terranes.
The authors are grateful to Andy Kerr for providing additional Newfoundland samples. We thank Kim Law for superb technical assistance in the laboratory. We also thank Jim Hibbard, Brendan Murphy, and D. Lavoie for helpful reviews that improved the final version of this manuscript. This study was funded by the Natural Sciences and Engineering Research Council of Canada grants to FJL (A4230) and SMB (A7387) and the Canada Foundation for Innovation infrastructure grant to FJL.
Laboratory for Stable Isotope Science (LSIS), The University of Western Ontario, Contribution 236.
- Received January 10, 2008.
- Accepted June 16, 2008.
- Published on the NRC Research Press Web site at http://cjes.nrc.ca on August 26, 2008.