1 The Creignish Hills and North Mountain areas of western Cape Breton Island (Fig. 1) consist mostly of Neoproterozoic metasedimentary and plutonic rocks typical of the Bras d’Or terrane of Barr and Raeside (1989) and Raeside and Barr (1990). These areas have been mapped and studied by numerous workers over the years (e.g., Kelly 1967; Milligan 1970; White et al. 1990; Campbell 1990; Justino 1991; Lynch and Brisson 1996; Keppie et al. 1998a; White et al. 2003; Wessel et al. 2005), and the resulting interpretations are varied, hampered by complex field relations and limited age constraints. White and Boehner (2008) compiled and evaluated existing data to produce a revised 1:50 000-scale geological map of the Whycocomagh NTS sheet, and Swanton et al. (2010) extended the mapping to the northeast to include metamorphic and plutonic rocks in the Lewis Mountain-Aberdeen Ridge area (Fig. 2).

2 The purpose of this paper is to provide an overview of the pre-Devonian geology based on these maps, and to present a new detrital zircon age spectrum from a metasedimentary unit in the Creignish Hills which better constrains the age of the metamorphic units in the area. We also take the opportunity to publish the full database for U-Pb zircon ages from the area reported only in preliminary form by White et al. (2003), and to present previously unpublished total fusion ⁴⁰Ar/³⁹Ar ages for muscovite from a metasedimentary unit in the Creignish Hills. In addition, we re-examine chemical data from Neoproterozoic plutons in these areas in the light of the current age information, and propose revised names for some of the rock units in the area based on all of the currently available data.

3 The Bras d’Or terrane of central Cape Breton Island is characterized by low- to high-grade metamorphic rocks intruded by abundant plutons of mainly Late Neoproterozoic age (Fig. 1). Barr and Raeside (1989) and Raeside and Barr (1990) documented tectonostratigraphic differences between these characteristic components of the Bras d’Or terrane and rocks of similar age that characterize the Avalonian Mira terrane of southeastern Cape Breton Island. Those differences have been supported by a number of subsequent studies (e.g., Barr et al. 1995, 1998; Potter et al. 2008a, b), and led to the inclusion of the Bras d’Or terrane, along with similar areas in central Newfoundland and southern New Brunswick, in Ganderia by Hibbard et al. (2006). In contrast, a number of other studies have focused on age similarities between rock units in the Bras d’Or and Mira terranes, and have produced alternative models based on the interpretation that these areas were not separate in the Neoproterozoic (e.g., Keppie and Dostal 1998; Keppie et al. 1990, 1998a, 2000).

4 The boundary between the Bras d’Or and Mira terranes of Barr and Raeside (1989) is placed in Bras d’Or Lake between North Mountain and Sporting Mountain (Fig. 1). It extends to the northeast through the Boisdale Hills where it coincides with a Carboniferous fault system (MacIntosh Brook – Georges River). From there it has been postulated to extend across the Cabot Strait to the south coast of Newfoundland (Barr et al. 1998, 2014a; Rogers et al. 2006). To the southwest, the terrane boundary is suggested to be offset at the Strait of Canso by the Canso fault and inferred to follow the western margin of the Creignish Hills (King 2002; Barr et al. 2012).

5 The northern boundary of the Bras d’Or terrane is assumed to be located north of the Creignish Hills beneath Carboniferous rocks. This interpretation is based on the occurrence southwest of Lake Ainslie of rhyolite and granite similar to those in the Lake Ainslie-Gillanders Mountain area of the Aspy terrane (Barr and Jamieson 1991). To the northeast, the boundary is inferred to follow high-strain zones through the central Cape Breton Highlands (Fig. 1). However, the relationship between the Bras d’Or and Aspy terranes is uncertain; available evidence suggests that they are linked in a complex basement/cover relationship, and that Bras d’Or terrane “basement” is likely present under Aspy terrane (e.g., Lin 1993, 1995, 2001; Lin et al. 2007; Price et al. 1999).

6 Traditionally, most metamorphic rocks in the Bras d’Or terrane were termed the George River Series or Group (e.g., Milligan 1970; Keppie 1979). Raeside and Barr (1990) subdivided these rocks into two assemblages based on metamorphic grade: mainly lower grade units collectively termed the George River metamorphic suite and mainly higher grade and typically gneissic units termed the Bras d’Or metamorphic suite. They also proposed local names for components of both of these metamorphic suites in different areas of the terrane because correlations among them could not be verified. Keppie (2000) termed the higher grade units the Bras d’Or Gneiss, rather than metamorphic suite, but the latter term is retained here as not all components of the unit are gneissic. Rocks of the George River and Bras d’Or metamorphic suites are everywhere separated from one another by known or inferred faults, or by plutonic units, and hence their relationship is enigmatic. Similarities in rock types suggest that they are the same rocks with different metamorphic histories (e.g., Barr et al. 2013) but even if so, their distribution and geological relations are difficult to explain, as exemplified by the map patterns in Figures 1 and 2.

7 The lower grade (mainly greenschist facies) metamorphic rocks in the Creignish Hills have been named the Blues Brook Formation (Raeside and Barr 1990; Campbell 1990). A compilation of previous work combined with their own mapping led White and Boehner (2008) to divide the Blues Brook Formation into 5 unnamed members, based on dominant rock types (Fig. 2): (1) slate interbedded with minor metasandstone, metasiltstone, and metacarbonate rocks (member sl); (2) metacarbonate rock interbedded with minor metasandstone, metasiltstone, slate, and quartzite (member ca); (3) mainly quartzite interbedded with minor metacarbonate rocks (member qz); (4) a volcanic unit consisting mainly of metamorphosed andesitic to basaltic lithic and lithic crystal tuff with minor basalt flows (member vt); and (5) metasandstone and metasiltstone interbedded with minor slate, metacarbonate rocks, quartzite, and rare basaltic lithic tuff (member ss). Similarities in rock types and their map distribution suggest that the members were originally a single stratigraphic succession, although younging directions are poorly preserved and lateral equivalency is possible.

8 In the extension of the Creignish Hills into the Lewis Mountain – Aberdeen Ridge area to the northeast, Swanton (2010) and Swanton et al. (2010) also recognized the Blues Brook Formation, including a small area of massive white quartzite (member qz; Fig. 2). However, the remaining rocks in the formation in that area could not be readily assigned to any of the other members of White and Boehner (2008) and hence two additional members were identified: member ps, mainly pelitic schist interbedded with subordinate metacarbonate rocks, amphibolite, and quartzite and member qf, quartzofeldspathic schist with quartzite, metaconglomerate, and minor amphibolite (Fig. 2). To the east along the shore of Bras d’Or Lake, Swanton et al. (2010) identified a separate formation, Aberdeen Ridge, which consists of muscovite-chlorite quartzite with minor amphibolite.

9 In the North Mountain area to the south (Figs. 1, 2), the lower grade metamorphic rocks have been named the Malagawatch Formation (Raeside and Barr 1990; Justino 1991). The Malagawatch Formation has not yet been subdivided into members but consists of an assemblage of rock types similar to those of the Blues Brook Formation, including metasiltstone, slate, calcitic and dolomitic marble, calc-silicate rocks, minor quartzite, and mafic metavolcanic rocks (Justino 1991).

10 Rocks of the Blues Brook Formation are regionally metamorphosed to greenschist facies and grade into lower amphibolite facies in the Whycocomagh Mountain area (Armitage 1989; Campbell 1990; Swanton 2010; Swanton et al. (2010). The metamorphic mineral assemblage in most of the pelitic rocks consists of chlorite, muscovite, biotite, and albitic plagioclase whereas in the Whycocomagh Mountain area the grade is higher and garnet and andalusite are present. Most of the metacarbonate rocks are monomineralic (calcite or dolomite) but locally contain talc, muscovite, phlogopite, garnet, and minor tremolite. In the higher metamorphic grade calc-silicate rocks diopside and forsterite are present (Armitage 1989). The mafic metavolcanic rocks have been metamorphosed to assemblages of actinolite, chlorite, and albitic plagioclase; stilpnomelane has also been reported (Campbell 1990). Contact metamorphic effects are locally well developed around some of the plutonic units and most evident in the pelitic rocks where a hornfelsic texture is preserved containing randomly oriented biotite, cordierite, and rare andalusite.

11 The protolith age of these metamorphic units is constrained to greater than ca. 586 Ma by dating of mainly undeformed cross-cutting plutons, as described below. However, they have generally been assumed to be much older than 586 Ma, perhaps Mesoproterozoic, based on the presence of stromatolite-like structures in the Malagawatch Formation, similar to those in the Green Head Group of southern New Brunswick which are of inferred Early Proterozoic age (Hofmann 1974; White and Barr 1996; White et al. 2007). This interpretation has been supported by detrital zircon populations that are mainly Archean and Mesoproterozoic in quartzite samples from both areas (Keppie et al. 1998a; Barr et al. 2003). However, a sample of schist from the Blues Brook Formation (Fig. 2) yielded 2 euhedral zircon grains with slightly discordant ²⁰⁶Pb/²³⁸U ages of 637 ± 3 Ma and 638 ± 2 Ma (Keppie et al. 1998a). Keppie et al.(1998a) interpreted the schist to have formed from a felsic volcanogenic unit, and the ages to be the age of the igneous protolith. If so, these dates are a maximum age for the Blues Brook Formation and, by inference, other units in the George River metamorphic suite. A younger ⁴⁰Ar/³⁹Ar cooling age of 455 ± 1 Ma was reported for muscovite from the same area (Keppie et al. 2000). Campbell (1990) obtained older muscovite ages of ca. 625 Ma from elsewhere in the same unit, but those data are unpublished and of uncertain reliability.

12 Keppie (1993), Keppie et al. (1998a), and Wessel et al. (2005) interpreted the presence of a folded shear zone between the low- and high-grade rocks in the Skye Mountain area of the Creignish Hills. They postulated that this shear zone repeatedly intersects the present erosion surface, implying that it is a low-angle tectonic boundary in which the high-grade rocks occur in structural domes. That interpretation was not adopted by White and Boehner (2008), who divided the high-grade metamorphic rocks in the Creignish Hills into two units, one at the northeastern tip for which they used the name Skye Mountain metamorphic suite, and the other, which they termed the Melford Formation, in two areas to the southwest: a fault-bounded block across the trend of the Creignish Hills in the northeast, and a smaller area near Melford in the central part of the Creignish Hills (Fig. 2). This distinction was made because the rocks in the latter two areas are not generally gneissic and differ in rock types from those in the area at the northeastern tip. We here introduce the name Chuggin Road instead of Skye Mountain for the rocks at the northeastern tip of the Creignish Hills because that area is not geographically located on Skye Mountain, and furthermore, the name Skye Mountain has long been used for plutonic rocks in the Skye Mountain area (see below). The area of high-grade rocks shown by Armitage (1989), Raeside and Barr (1990), and White and Boehner (2008) on Whycocomagh Mountain was reassigned by Swanton (2010) and Swanton et al. (2010) to the Blues Brook Formation as noted above, based on rock types and metamorphic grade. In the North Mountain area, rocks of the Bras d’Or metamorphic suite are known as the Lime Hill gneissic complex and occur in a large area in the south-central part of North Mountain (Fig. 2).

13 As described by Raeside (1990), Raeside and Barr (1990), Campbell (1990), and Sangster et al. (1990) the high-grade metamorphic rocks in both the Chuggin Road and Lime Hill units include biotite, bioitite-cordierite, and sillimanite-bearing paragneiss, migmatitic paragneiss, marble, quartzite, amphibolite, and tonalitic orthogneiss. In contrast, the Melford Formation has somewhat different rock types, including biotite, biotite-cordierite, sillimanite, and garnet-bearing schist, marble, quartzite, and granitic orthogneiss (White and Boehner 2008).

14 Keppie et al. (1998a) provided age constraints on rocks in the Melford Formation of White and Boehner (2008). A muscovite-biotite ‘paragneiss’ from the southern block yielded mainly mid-Proterozoic ages, but the youngest grain is ca. 977 Ma (Keppie et al. 1998a). A pelitic paragneiss inferred to have a volcanogenic protolith in the fault-bounded block at Skye Mountain yielded zircon grains with ²⁰⁶Pb/²³⁸U ages between 688 Ma and 694 Ma (Keppie et al. 1998a). These ages were interpreted to be the best estimate of igneous crystallization ages for the inferred volcanogenic protolith. A monazite grain interpreted to be of metamorphic origin in the same sample yielded a concordant age of 552 ± 8 Ma. Farther north and also in the Melford Formation of White and Boehner (2008), a foliated granitic sheet, considered to be syntectonic, yielded two concordant monazite grains with ages of 551±1 Ma, interpreted to be igneous crystallization ages (Keppie et al. 1998a). Two zircon grains yielded discordant but similar ages of ca. 553 Ma. These ages provide a minimum age for the Melford Formation. In contrast to these Precambrian ages, ⁴⁰Ar/³⁹Ar ages from muscovite in the Melford Formation are younger: 441, 449, 461, 473, and 485 Ma (Campbell 1990; Dallmeyer and Keppie 1993; Keppie et al. 1998a). Similar ⁴⁰Ar/³⁹Ar ages were reported for muscovite (484, 524 Ma) and biotite (485 Ma) from the area of Melford Formation to the south near Melford (Campbell 1990; Keppie et al. 2000) and muscovite from the Melford pluton (Keppie et al. 2000). An igneous crystallization age of ca. 561 Ma reported by White et al. (2003) for orthogneiss in the Chuggin Road metamorphic suite is described below.

15 In the case of the Lime Hill gneissic complex, ages are constrained by a concordant U-Pb monazite date from paragneiss at 549±2 Ma, and syn- and post-tectonic plutons that yielded discordant U-Pb (zircon) data with lower intercepts of 539 ± 1 Ma and 545–540 Ma (Sangster et al. 1990). An additional constraint is provided by a concordant monazite age of 553 ± 1 Ma from an anatectic granite sheet in pelitic paragneiss, interpreted to represent the time of peak metamorphism (Keppie et al. 1998a). Keppie et al. (1998a) reported that another monazite grain and rounded zircon grains from the same sheet yielded older but discordant ages.

16 Dioritic to granitic plutons intruded both the low- and high-grade metamorphic units in both the Creignish Hills and North Mountain areas (Fig. 2). White et al. (1990) assigned most of the plutonic rocks in the Creignish Hills area to the Creignish Hills pluton, and subdivided the pluton into five units based on modal mineralogy; texture, and lithogeochemistry: tonalite-diorite, granodiorite-tonalite, granodiorite-monzogranite, coarse-grained monzogranite, and fine-grained monzogranite. This work was done in the early days of geochronological work in Cape Breton Island, and the authors were much influenced by an Rb-Sr isochron using a range of rock types which indicated an age of about 446 Ma. However, ⁴⁰Ar/³⁹Ar cooling ages of ca. 540 Ma from hornblende in the tonalite – diorite unit indicated a much older minimum age for that body. Furthermore, the granitic components of the pluton had petrological similarities to the Kellys Mountain and Cape Smokey plutons which had yielded U-Pb (zircon) ages of 498 and 493 Ma, respectively. Overall, it was concluded based on the available data that the rocks are (questionably) of Ordovician age. Two small plutons in the northeastern part of the belt were together named the Skye Mountain pluton, described as separate bodies of gabbro and granodiorite, and assigned an age anywhere between Precambrian and Carboniferous (White et al. 1990). These plutons were reported to differ from any of the components of the Creignish Hills pluton based on descriptions provided by earlier workers but were not part of the study by White et al. (1990).

17 Knowledge of the ages of these plutons was much improved as a result of work by Keppie et al. (2000), who reported U-Pb (zircon) ages of 553 ± 2 for a sample from the largest tonalite-diorite body of the Creignish Hills pluton. This age is consistent with the previously reported ca. 544 Ma hornblende cooling age from that body (Keppie et al. 1990). However, a smaller body of tonalite (River Denys pluton of Keppie et al. 2000) yielded a younger U-Pb age of 540±3 Ma, although ⁴⁰Ar/³⁹Ar ages for amphibole from the same sample and another sample in the same unit are older (ca. 550 and 551 Ma; Keppie et al. 1990). Keppie et al. (2000) concluded that, overall, these 540–550 Ma ages “post-date closely the time of intrusion”, and that the River Denys Pluton is part of the same magmatic event as the 553 Ma diorite of the Creignish Hills pluton. The age of the Creignish Hills pluton was further constrained by a ca. 553 Ma age for the main granitic unit reported by White et al. (2003) and presented below.

18 In contrast to these Neoproterozoic ages, a sample reported to be from the gabbro-diorite body of the Skye Mountain pluton yielded a Silurian age of 438 ± 2 Ma (Keppie et al. 1998b, 2000). A sample from the separate granitic body to the west yielded only old (ca. 737 Ma) discordant zircon grains, but Keppie et al. (2000) inferred a Silurian age, based on muscovite plateau ages in the surrounding host rocks, especially an age of 441 Ma from what they interpreted to be the contact metamorphic aureole adjacent to the granite. Keppie et al. (2000) and Wessel et al. (2005) showed the Skye Mountain granite and gabbro-diorite as contiguous bodies on their maps, whereas previous maps (White et al. 1990; Horton 1994) showed them as separate bodies. Horton (1994) interpreted them to be unrelated based on their very different petrographic and chemical characteristics. He considered the foliated granitic sheet in the Melford Formation north of the pluton to be related to the granite and hence its ca. 551 Ma age to be the age of the pluton.

19 After a re-evaluation of outcrop locations and additional mapping in the area, White and Boenher (2008) confirmed that the two plutons are separated by a belt of rocks of the Melford Formation. They included the western (granitic) pluton in the Creignish Hills pluton, to which it is petrographically similar, and inferred a similar ca. 550 Ma age. Hence, to avoid further confusion, the name Skye Mountain is retained here for only the gabbroic-dioritic body and a new name, Skye Mountain Road, is introduced for the western body, named after the road on which it is best exposed.

20 The small granitic body termed the Melford pluton by Keppie et al. (2000) yielded an older U-Pb (zircon) age of 586 ± 2 Ma, although this two-point upper intercept age is of limited reliability. The pluton intruded gneissic rocks assigned by Keppie et al. (2000) to the Bras d’Or Gneiss (Melford Formation in this study), and was interpreted to be late syn-tectonic based on the presence of a weak foliation parallel to that in its host rocks. White and Boehner (2008) showed that the pluton is larger than suggested by Keppie et al. (2000), and that it is faulted on its southwestern and northwestern edges against rocks of the Blues Brook Formation. They included the pluton in the granitic unit of the Creignish Hills pluton, in spite of its apparently somewhat older age and the occurrence of muscovite for which Keppie et al. (2000) reported a cooling age of 472 Ma. Muscovite is not a typical component of the Creignish Hills granite but has also been reported in part of the Skye Mountain Road pluton (Horton 1994).

21 Plutons form most of the North Mountain area, and were studied in detail by Justino (1991) and Justino and Barr (1994). Those authors divided the plutonic rocks into the Big Brook and Marble Mountain plutons, of mainly granodiorite composition, and the West Bay Pluton, composed of coarse-grained monzogranite. The latter pluton is cut off by a fault according to the map by Giles et al. (2010) but likely extends south of the fault with minor offset, based on the presence of outcrops of similar granite in the Sugar Camp gypsum quarry (Fig. 2). Small tonalitic to dioritic bodies (Mill Brook quartz diorite and related plutons) occur along the margins of these large plutons. Age constraints are provided by ⁴⁰Ar/³⁹Ar hornblende cooling ages of 555 ± 5 Ma apparently from the Marble Mountain biotite granodiorite unit (based on approximate location shown by Keppie et al. 1990) and 550 ± 5 Ma and 545 ± 6 Ma on two separate bodies of the Mill Brook quartz diorite (Keppie et al. 1990) and the ca. 553 Ma age for the West Bay Granite reported by White et al. (2003) and presented below.

41 The U-Pb ages presented here for detrital zircon in the Blues Brook Formation of the Creignish Hills confirm the hints of Late Neoproterozoic ages reported by Keppie et al. (1998a) and show without doubt that the depositional age of the unit is no greater than about 600 Ma. Although it is possible that some components of the formation are much older, such as the quartzite in which the youngest detrital zircon age obtained is ca. 1000 Ma (Keppie et al. 1998a), similarities in rock types and contiguous field relations suggest that this is not the case. Based on similarities in rock types and previously published detrital zircon ages of ca. 690 Ma (Keppie et al. 1998a), it is likely that the high-grade metasedimentary rocks of the Melford Formation and Chuggin Road complex are of the same age, and that they represent the same or stratigraphically equivalent units at higher metamorphic grade, rather than unrelated units. The minimum ages of both the low- and high-grade units in the Creignish Hills and nearby North Mountain are provided by cross-cutting syn- and post-tectonic plutons with ages of mainly 565–550 Ma, indicating that sediments were deposited, regionally metamorphosed and deformed, and intruded by plutons in less than 40–50 million years. The assemblage of pelitic, psammitic, and carbonate rocks indicates that a passive margin in a tropical climate was quickly changed to an active Andean-type continental margin in which voluminous calc-alkaline dioritic to granitic plutons were emplaced. This sedimentary and tectonic history is characteristic of the Bras d’Or terrane and is shared by its likely correlative, the Brookville terrane in southern New Brunswick (Barr et al. 2014b).