The Dunnage Zone of the Newfoundland Appalachians hosts diverse Cambrian–Ordovician volcanogenic massive sulfide (VMS) deposits. The peri-Laurentian Notre Dame Subzone contains Cu–Zn–Au mafic and bimodal mafic deposits in ∼501–485 Ma ophiolitic rocks and Zn–Pb–Cu–(Au–Ag) deposits in ∼471–465 Ma bimodal rifted continental arc sequences (e.g., Buchans). The peri-Gondwanan rocks of the Exploits Subzone host Zn–Pb–Cu–(Au–Ag) bimodal felsic, felsic siliciclastic, and Zn–Ag–Au hybrid bimodal felsic deposits in the ∼513–486 Ma Victoria Lake supergroup; Cu–Zn bimodal felsic to bimodal mafic deposits of the ∼486 Ma Wild Bight Group; and Cu–(Au) mafic siliciclastic deposits of the ∼466 Ma Great Burnt Lake/South Pond belt. Regardless of age or stratigraphic hosts, all VMS deposits are associated with specific magmatic assemblages and extensional tectonism (i.e., rifting). Gold-enriched deposits of the Rambler-Ming district are associated with felsic rocks that formed via slab melting and subsequent melt-mantle wedge interaction, which likely enhanced precious metal enrichment in these deposits. Whereas many deposits exhaled on the seafloor, some deposits formed via subseafloor replacement of host units or as re-sedimented sulfides generated in sediment-gravity flows. Metals in the deposits were derived from leaching of underlying footwall rocks; however, Au–Ag- and epithermal suite element-enriched deposits show evidence for metal contributions from magmatic hydrothermal fluids. Sulfur in deposits was derived predominantly from leaching of H2S from underlying footwall rocks and from thermochemical sulfate reduction of seawater sulfate, with lesser input from bacteria-derived H2S and magmatic-hydrothermal-derived H2S. Despite recent research advances and historic mining, numerous questions remain unresolved and provide opportunities for future study.

Volcanogenic massive sulfide (VMS) deposits are important sources of Cu, Zn, Pb, Au, Ag, and other metals for the Canadian and global economy (Franklin et al. 2005). They occur in extensional settings (e.g., mid-ocean ridges, back-arc basins, and rifted arcs) and have been forming from the Archean to present (Hannington et al. 2005; Piercey 2011; Ross and Mercier-Langevin 2014). Appalachian VMS deposits have been important contributors to the economy of much of eastern Canada, including the world class deposits of the Bathurst Mining District (Goodfellow 2007) and numerous districts in central Newfoundland (Swinden and Kean 1988). In the Dunnage Zone of central Newfoundland, VMS deposits were important sources of metal prior to Confederation with Canada (Martin 1983) and are still important contributors to the economy, as well as the foci of continued exploration and development.

In addition to economic significance, VMS deposits in the Newfoundland Dunnage Zone have contributed greatly to our global understanding of VMS deposits, including the roles of (1) footwall domains in influencing the metals present in VMS (e.g., Swinden and Thorpe 1984), (2) lithogeochemistry and applications in tectonostratigraphic research (e.g., Swinden 1991), (3) thrust imbrication and its importance in deposit reconstructions and exploration (e.g., Thurlow 2010), (4) the role of metamorphism in creating metamorphosed alteration assemblages (e.g., Upadhyay and Smitheringale 1972), (5) sulfur isotopes behaviour during sulfide metamorphism (e.g., Bachinski 1977; Cloutier et al. 2015), (6) exhalation and subseafloor replacement (e.g., Squires et al. 1991; Piercey et al. 2014), and (7) magmatic fluids/volatiles in promoting precious metal and epithermal suite (e.g., Au–Ag–Hg–As–Sb–Bi–Te–Se–Sn–W) element enrichments in VMS deposits (e.g., Santaguida and Hannington 1996; Brueckner et al. 2014b, 2016; Gill et al. 2016; Pilote et al. 2016).

In this paper, we review the geological setting and styles of VMS deposits in the Dunnage Zone of central Newfoundland and discuss recent key advances in our understanding of the genesis of these deposits and potential avenues for future research.


VMS deposits are the products of submarine hydrothermal systems that form in extensional (rifted) tectonic regimes via the circulation of seawater through crust driven by heat from underlying magmatic rocks (Fig. 1; Franklin et al. 1981, 2005; Lydon 1988). As seawater penetrates the underlying crust through the recharge zone, it heats up and reacts with host rocks, resulting in the leaching of metals (e.g., Zn, Pb, and Cu, among others) and reduced sulfur (H2S) in the reaction zone; the metals then bond with Cl from seawater and are subsequently transported predominantly as metal chloride complexes (Lydon 1988; Galley 1993; Skirrow and Franklin 1994; Jowitt et al. 2012). If sulfate (forumlaSO42) is present in the downwelling seawater reaction zones and reductants are present in the wall rock (e.g., Fe and organic C), sulfate will undergo thermochemical sulfate reduction (TSR) to generate H2S (Shanks et al. 1981; Shanks and Seyfried 1987). These hot, metal chloride- and H2S-bearing fluids are generally buoyant and will move upwards from depth along synvolcanic (and synsedimentary) faults, will conductively cool, react with wall rock to form hydrothermal alteration zones (e.g., sericite–chlorite–quartz assemblages), and will mix with seawater at or near the seafloor, resulting in the precipitation of sulfides (Fig. 1; Lydon 1988; Large 1992). In some hydrothermal systems where magmatic fluids are important, precious metals and epithermal suite elements (e.g., Au–Ag–Hg–As–Sb–Bi–Te–Se–Sn–W) are discharged from underlying subvolcanic magma chambers into the overlying hydrothermal system, which upon cooling and mixing with seawater may result in precipitation of deposits enriched in these metals and having distinctive argillic/advanced argillic or adularia–sericite alteration assemblages (e.g., Sillitoe et al. 1996; Hannington et al. 1999).

Deposit classification

Although a generalized model for VMS deposits has been presented above, there are distinctive sub-types of VMS deposits and the most widely accepted classification of these sub-types is the tectonostratigraphic classification of Barrie and Hannington (1999) and subsequent revisions thereof (e.g., Franklin et al. 2005; Galley et al. 2007; Piercey et al. 2015). Minor variants on this classification have been proposed with expansions to deposits associated with specific rock types (e.g., ultramafic VMS classification of Patten et al. 2022), but these are not discussed herein as there are no bona fide examples of these in the Newfoundland Appalachians.

Deposits in the Newfoundland Appalachians are classified into the following groups (modified after Galley et al. 2007): (1) mafic, (2) bimodal mafic, (3) mafic siliciclastic, (4) bimodal felsic, (5) felsic siliciclastic, and (6) hybrid bimodal felsic (Fig. 2).

Mafic (also referred to as Cyprus-type) deposits are those hosted predominantly in mafic volcanic rocks and sheeted dikes within forearc or back-arc ophiolites (Fig. 2). These deposits are predominantly Cu-rich, but some have produced Zn and Au, whereas others contain enrichments in Co and Ni. In the Newfoundland Appalachians, deposits in the Lushs Bight, Betts Cove, and Bay of Islands ophiolites are examples of this deposit type (Fig. 3; Hibbard 1983; Swinden and Kean 1988; Kean et al. 1995).

Bimodal mafic (also referred to as Noranda-type) deposits are hosted in primitive arc belts dominated by mafic volcanic and intrusive rocks but where the deposits are specifically hosted by felsic volcanic rocks (Fig. 2). They typically occur in rifted primitive arcs and, in some cases, forearcs and are polymetallic with a Zn–Cu–(Au–Ag) metal signature. In the Newfoundland Appalachians, the Cu–Au-bearing Rambler and Ming deposits in the Baie Verte belt are examples of this VMS deposit type (Fig. 3; Brueckner et al. 2014b, 2016).

Mafic siliciclastic (also referred to as Besshi-type) deposits are hosted in belts that contain mafic volcanic rocks, volcaniclastic and intrusive rocks, and abundant sedimentary rocks, including argillite, siltstone, and sandstone (Fig. 2). Felsic volcanic rocks, if present, only comprise a minor component of the stratigraphy. These deposits are typically associated with back-arc (and forearc) environments and are Cu–Zn-rich. In the Newfoundland Appalachians, the Great Burnt Lake and South Pond deposits are examples of this VMS deposit type (Fig. 3; Swinden 1988a; Barry et al. 2022).

Bimodal felsic (also referred to as Kuroko-type) deposits are hosted in belts of predominantly felsic volcanic and volcaniclastic rocks with lesser mafic rocks and are interpreted to have formed in mature rifted continental arcs and back-arc basins (Fig. 2). These deposits are polymetallic with Zn–Pb–Cu–(Au–Ag), and in the Newfoundland Dunnage Zone, deposits of the Buchans district (Thurlow 2010), and some within the Tally Pond and Tulks belts of the Victoria Lake supergroup (Squires and Moore 2004; Hinchey 2011), are examples of this VMS deposit type (Fig. 3).

Felsic siliciclastic (also referred to as Bathurst-type) deposits are those hosted in siliciclastic sedimentary and felsic volcaniclastic-rich belts that contain abundant felsic volcaniclastic rocks, felsic intrusive and volcanic rocks, black shales and other siliciclastic sedimentary rocks, iron formations, and local mafic volcanic and intrusive rocks (Fig. 2). These deposits are interpreted to have formed in rifted continental margin arcs with high rates of sedimentation and are polymetallic with Zn–Pb–Cu–(Au–Ag) metal signatures. In the Newfoundland Appalachians, deposits in the southern part of the Tulks belt of the Victoria Lake supergroup, including the Boomerang and Tulks East deposits (Hinchey 2011), are examples of this VMS deposit type (Fig. 3).

Hybrid bimodal felsic deposits were defined by Galley et al. (2007) as those with geological features like bimodal felsic deposits but also with features similar to epithermal Au deposits, including argillic to advanced argillic alteration (±adularia–sericite alteration), epithermal suite element enrichments (e.g., Au–Ag–Hg–As–Sb–Bi–Te–Se–Sn–W), and complex sulfide and sulfosalt ore mineral assemblages (Fig. 2). These deposits are interpreted to have formed in relatively shallow water (<1500 m below sea level) within rifted arc settings like those deposits actively forming in the Manus Basin or Kermadec arc and have evidence for both VMS and magmatic-hydrothermal (epithermal) fluid inputs (e.g., Binns and Scott 1993; Sillitoe et al. 1996; de Ronde et al. 2005). In the Newfoundland Appalachians, the Bobby’s Pond and Daniels Pond deposits in the northern part of the Tulks belt in the Victoria Lake supergroup (Hinchey 2011; Sparkes and Hinchey 2023) show some characteristics of this VMS deposit type (Fig. 3). Although a hybrid bimodal mafic classification was not advanced by Galley et al. (2007), the Au–Ag-sulfosalt-rich deposits of the Rambler-Ming district are potential examples of a hybrid bimodal-mafic deposit (i.e., bimodal mafic—Au-rich) having features like bimodal mafic VMS deposits but also displaying geological and mineralogical features of intermediate sulfidation epithermal Au–Ag deposits (Fig. 3; Brueckner et al. 2014b,2016).

Grade–tonnage data

Grade–tonnage data for VMS deposits from central Newfoundland are presented in Figs. 4 and 5 and Table S1 in the Data Repository. These data are compiled from previous studies, including Franklin et al. (2005), Galley et al. (2007), Huston et al. (2015), and Piercey et al. (2015). Further updates include recent National Instrument (NI)-43-101 resource data obtained from company technical reports. Grade–tonnage data without recent NI-43-101 resources are considered historic resources and reserves. All data are presented as geological (total) resources, are in metric tonnes, and include both resources and reserves where relevant.

Base metal abundances in the deposits are shown in grade–tonnage space in Fig. 4. The highest-grade deposits are predominantly bimodal felsic and felsic-siliciclastic deposits; however, this result is partly influenced by deposits from the Buchans mining district, which are some of the highest-grade VMS deposits on the planet (e.g., Piercey et al. 2015). The highest tonnage deposits include those from every type of VMS deposit (Fig. 4A). In general, the mafic deposits are Cu- and Cu–Zn-dominated, whereas all other deposit types associated with felsic rocks are polymetallic with Zn–Cu–Pb, albeit those of the bimodal felsic, felsic siliciclastic, and hybrid bimodal felsic types contain greater Zn and Pb enrichment relative to Cu (Fig. 5A).

Gold and Ag are generally not enriched in most VMS deposits in central Newfoundland (Figs. 4B, 5B–5C). The deposits in the Rambler-Ming district in Baie Verte, specifically the Rambler Main deposit and ancillary lenses, are auriferous, whereas the Rendell-Jackman prospect has high Au grades but is low tonnage (Figs. 4B, 5B–5C). Other deposits do have precious metal-enriched zones and Au–Ag as ancillary metals, including Ming (e.g., 1806 zone; Brueckner et al. 2014b), Lemarchant (e.g., Gill et al. 2016), and Boomerang and Daniels Pond (e.g., Hinchey 2011); however, the geological resources are not considered Au-rich or auriferious (see Mercier-Langevin et al. 2011).

Descriptions of deposits in this paper refer primarily to those that have published resources and we will not focus here on small showings or occurrences, unless they provide critical insight into processes or the VMS metallogeny of the central Newfoundland Appalachians.

Regional geological setting

The Newfoundland Appalachians are divided into four tectonostratigraphic zones (Williams 1979), from west to east: Humber, Dunnage, Gander, and Avalon (Fig. 3). The Dunnage Zone hosts most VMS deposits in central Newfoundland and separates rocks of the Humber Zone from those zones of Gondwanan affinity—Gander, Avalon, and Meguma (van Staal and Barr 2012). The Dunnage Zone is subsequently subdivided into the peri-Laurentian Notre Dame Subzone and the peri-Gondwanan Exploits Subzone, which are juxtaposed along the Beothuk Lake Line (=Red Indian Line; Williams et al. 1988; Fig. 3). The Dunnage Zone comprises arc, back-arc, and ophiolitic rocks built on variable basements and contains vestiges of the Iapetus Ocean. The peri-Laurentian portion of the Dunnage Zone, west of the Beothuk Lake Line, consists of ophiolitic sequences in the Lushs Bight (∼505 Ma) and Baie Verte (∼489–485 Ma) oceanic tracts and rocks of the Annieopsquotch accretionary tract and Buchans-Roberts Arm belt (∼480–460 Ma; Fig. 3; Dunning and Krogh 1985; Swinden et al. 1997; Zagorevski et al. 2006; Skulski et al. 2010). The peri-Gondwanan portion of the Dunnage Zone east of the Beothuk Lake Line consists of remnants of the Penobscot (∼515–485 Ma) and Popelogan-Victoria (∼475–455 Ma) arc and back-arc systems that formed proximal to Ganderia (Fig. 3; MacLachlan and Dunning 1998; van Staal et al. 1998; Rogers et al. 2006; Zagorevski et al. 2007). These terranes were built atop Neoproterozoic to Early Cambrian (∼620–530 Ma) basement rocks that have subduction-related affinities (Barr et al. 2003; Rogers et al. 2006; Zagorevski et al. 2007, 2010). Peri-Laurentian rocks of the Dunnage Zone were obducted during the Taconic orogeny when rocks of the Lushs Bight Group were emplaced onto the Dashwoods microcontinent, a rifted fragment of Laurentian basement, during Taconic 1 obduction (∼500–493 Ma). Then, these rocks and the ophiolitic rocks of the Baie Verte oceanic tract were emplaced onto the Humber margin during Taconic 2 obduction at ∼475–460 Ma (e.g., Bedard 1999; Waldron and van Staal 2001; van Staal 2007; Staal and Barr 2012; Castonguay et al. 2014). Rocks of the peri-Gondwanan Exploits Subzone were emplaced upon the Gander margin during the Penobscot orogeny (∼486–475 Ma; e.g., Colman-Sadd et al. 1992; van Staal 2007; Zagorevski et al. 2009; Fyffe et al. 2012; Johnson et al. 2012; van Staal and Barr 2012). The Notre Dame and Exploits subzone rocks were accreted to each another along the Beothuk Lake line in the last stages of the Taconic orogeny during Taconic 3 (∼455–450 Ma) arc–arc collision (van Staal and Barr 2012).

Rocks and VMS deposits of the Dunnage Zone were also influenced by younger tectonothermal events related to (1) protracted magmatism within the Notre Dame arc (∼488–435 Ma), (2) syn- to post-Taconic deformation (Whalen et al. 1997; van Staal et al. 2007; van Staal and Barr 2012), (3) ∼453–430 Ma Salinic (Dunning et al. 1990; van Staal 1994; van Staal et al. 1998; van Staal et al. 2003), (4) ∼420–400 Ma Acadian (e.g., Robinson et al. 1998; Bradley et al. 2000; Castonguay and Tremblay 2003; van Staal 2007; van Staal et al. 2009, 2014), and (5) ∼400–350 Ma Neoacadian orogenesis (Murphy et al. 1999; van Staal 2007; van Staal and Barr 2012).

Outlined below are the main VMS belts within the central Newfoundland Appalachians, subdivided into peri-Laurentian/Notre Dame Subzone-hosted VMS belts and the peri-Gondwanan Exploits Subzone-hosted VMS belts (Fig. 3).

Peri-Laurentian VMS belts

Springdale belt

The Springdale belt contains VMS deposits hosted in ∼510–501 Ma ophiolitic rocks of the Lushs Bight Group (Kean et al. 1995) that form part of the broader Lushs Bight oceanic tract (Fig. S1; van Staal 2007). The Lushs Bight Group contains pillow lavas, sheeted dikes and mafic intrusive rocks, ultramafic rocks, red cherts, and rarer felsic volcaniclastic rocks (Kean et al. 1995). The main deposits within the belt include the mafic-type Little Deer, Whalesback, and Little Bay deposits (all past producers), which are hosted within arc tholeiitic and boninitic pillow lavas and are interpreted to have formed within forearc assemblages associated with subduction zone initiation (Fig. S1; Kean et al. 1995; Swinden et al. 1997; van Staal 2007). The Miles Cove mine, Rendell-Jackman, Whalesback, and Little Bay deposits are dominated by stringer vein, breccia, and locally massive to semi-massive pyrite–chalcopyrite–pyrrhotite assemblages with lesser sphalerite, cobaltite, and secondary, deformation-related Ag–Bi–Hg tellurides and arsenides (Figs. S1 and S6; Kean et al. 1995; Toman 2013; Cloutier et al. 2015). Kanehira and Baschinski (1968) also noted in the Whalesback deposit the occurrence of pentlandite, mackinawite, magnetite, cubanite, galena, and ilmenite, with minor supergene marcasite, covellite, and goethite. In the above deposits, mineralization is hosted in variably chlorite altered basalts and is commonly deformed into sulfide-bearing chlorite-(sericite) schists and locally quartz-chlorite-(carbonate) schists (Fig. 6; Kean et al. 1995; Toman 2013; Cloutier et al. 2015). The main Cu-rich assemblages in the deposits are interpreted to have formed from high temperature (T > 300 °C), reduced, acidic fluids, typical of ophiolite-hosted mineralization globally (Toman 2013; Cloutier et al. 2015).

Baie Verte belt

The Baie Verte belt contains VMS deposits within the ∼489–487 Ma ophiolitic rocks of the Betts Cove Complex and correlative rocks across the peninsula (e.g., Hibbard 1983). The rocks of the Baie Verte oceanic tract are interpreted to have formed within a forearc setting during subduction zone initiation within the Humber seaway (Bedard 1999; van Staal 2007). The two main types of deposits in the belt are mafic deposits hosted within the Betts Cove Complex and Au-enriched bimodal mafic deposits hosted within the Pacquet Complex in the Rambler-Ming district (Figs. S2 and S7). In the Betts Cove Complex, VMS deposits are hosted by boninitic to island arc tholeiitic rocks of the Betts Head Formation and occur at two stratigraphic positions: within mafic volcanic/pillow lavas just above the sheeted dike complex (Betts Cove) or higher in the stratigraphy within mafic volcanic/pillow lavas (Tilt Cove; Fig. 7; Strong and Saunders 1988; Sangster et al. 2007). These are overlain by the Nugget Pond horizon, which includes red chert with hematite and magnetite that is interpreted to be an exhalative iron formation (e.g., Mueller et al. 2021b); locally, this unit also hosts Silurian–Devonian orogenic Au mineralization. In the past-producing Betts Cove and Tilt Cove deposits, mineralization is dominated by Cu–Zn–Au with lesser Ag and Pb and comprised predominantly pyrite and chalcopyrite with lesser sphalerite, and trace galena, gold, and Ag- and Pb-tellurides (Fig. 7; Upadhyay and Strong 1973; Strong and Saunders 1988; Sangster et al. 2007). Deposits vary from semi-massive to disseminated sulfides and are locally deformed resulting in chlorite–sulfide-rich shear zones (Sangster et al. 2007). In some locales, the disseminated sulfides and chlorite-altered rocks form pipe-like bodies interpreted to be VMS-related feeder/stringer vein upflow zones (Sangster et al. 2007). In some areas proximal to the VMS deposits, shear zones cut underlying ultramafic rocks and host Ni–As-rich sulfides; however, these are interpreted to be unrelated to (and postdate) the VMS mineralization (Papezik 1964; Sangster et al. 2007).

The Cu–Au deposits of the Rambler-Ming district are within a bimodal mafic sequence composed of boninitic to arc tholeiitic rocks of the Mount Misery Formation, which are overlain by dacitic to rhyolitic rocks of the Rambler Rhyolite formation that hosts the Ming, Rambler, and Rambler East deposits (Figs. S2 and S7). These bimodal-mafic deposits occur within a package of coherent and volcaniclastic dacitic rocks with the sulfide lenses overlain by lapilli tuffs and tuff breccias that contain fragments of the underlying host rocks and massive sulfides (Coates 1990; Pilote et al. 2017); the latter sulfide fragments are locally Au-rich (Pilote et al. 2016). The Rambler Rhyolite formation and VMS deposits are regionally overlain by the Goldenville horizon (Fig. 7), an iron formation interpreted to be equivalent to the Nugget Pond horizon in the Betts Cove region (Skulski et al. 2009) and to represent low-temperature hydrothermal discharge related to the Rambler-Ming deposits (Mueller et al. 2021a). The Rambler-Ming deposits contain both massive sulfide and stringer sulfides (Fig. 7). In the past-producing Rambler Main mine, the deposit consists of banded to massive pyrite–chalcopyrite–sphalerite with rarer galena and Au–Ag–Te-bearing phases, which are underlain by stringer pyrite–chalcopyrite–(sphalerite) in green mica (fuchsite)-sericite schists (Coates 1990; Weick 1993). The producing Ming mine contains numerous massive sulfide zones with variable grades and metal assemblages, which are underlain by an intensely chlorite-altered footwall zone with a stringer-type chalcopyrite–pyrite–pyrrhotite assemblage (Fig. 7; Brueckner et al. 2016). Significant mineralogical and metal diversity exists in the massive sulfide zones, which are dominated by pyrite–chalcopyrite with lesser sphalerite and galena. These zones also contain a plethora of sulfosalts, tellurides, Ag–Au–Hg–Bi phases, and Fe- and Sn-oxides; these complex assemblages are interpreted to be reflective of normal VMS metal leaching processes and magmatic-hydrothermal input into the Ming hydrothermal system (Fig. 7; Brueckner et al. 2014b, 2016). The Ming deposit also displays typical VMS-related hydrothermal alteration, including sericite proximal to the massive sulfide mineralization and intense chlorite–(quartz) alteration in the footwall zones; however, this deposit also contains distinctive green mica (fuchsite?) like that present at the Rambler Main deposit (Fig. 7; Pilote et al. 2020).

Buchans-Roberts Arm belt

The Buchans-Roberts Arm belt extends from Buchans in the south to Pilleys Island in the north and contains past-producing VMS deposits in the Buchans, Gullbridge-Lake Bond, and Pilley’s Island regions (Fig. S3).

The Buchans district is host to numerous VMS deposits that have some of the highest grades globally with exceptional Zn–Pb grades (Figs. 4–5 and Table S1; e.g., Piercey et al. 2015). The district produced from 1928 to 1984 (Thurlow 2010) and has been the subject of numerous comprehensive studies (e.g., Swanson et al. 1981, and references therein; Kirkham 1987, and references therein), recent research (e.g., van Hees et al. 2012), and more recent exploration, discovery, and deposit expansion (e.g., Lundberg Zone deposit of Buchans Resources Ltd.). The deposits are bimodal-felsic and hosted in calc-alkalic sequences, with most mineralization occurring in the felsic-dominated Buchans River Formation and less in the underlying basalts of the Ski Hill Formation (Fig. 8). Rocks of the district are pervasively imbricated by thrust faults (Fig. 8) with most VMS deposits having at least one faulted contact; these thrust faults are interpreted to be reactivated synvolcanic faults that nucleated on phyllosilicate-rich alteration zones during basin inversion associated with Taconic thrusting (Thurlow and Swanson 1987; Thurlow 2010).

Past-producing deposits of the district are of three main types: in situ (e.g., Lucky Strike), stringer/stockwork deposits (e.g., Lundberg zone), and transported (e.g., MacLean; Fig. 8). The in situ deposits (i.e., formed within and proximal to volcanic hosts in interpreted exhalative chimneys/mounds) are the least well-understood deposits because most were mined out by the 1960s and before comprehensive research was undertaken in the district (Thurlow 2010). These deposits contain massive sulfides of varying textures that are underlain by stockwork mineralization with associated chlorite–sericite alteration zones and are variably deformed due to thrust faults (Thurlow 2010).

Transported ore deposits consist of unsorted, matrix-supported conglomeratic ores that contain clasts of both volcanic rocks, as well as sphalerite–galena (black ore), chalcopyrite (yellow ore), and barite within a matrix of ash and fine-grained barite (Fig. 8). These ores are interpreted to have been sulfide–volcaniclastic debris flows that were channelized within extensional basins during rifting (Thurlow and Swanson 1981; Binney 1987).

Stockwork mineralization is hosted predominantly in basalts of the Ski Hill Formation and partly in felsic rocks of part of the Buchans River Formation and consists of stringer-style base metal sulfides, barite, and locally quartz, commonly exhibiting zonation in alteration from chlorite and quartz proximal to the stockwork and sericite distal from the stockwork (Thurlow 2010; van Hees et al. 2012). Locally, there is evidence of bladed barite (Fig. 8) and quartz and illite alteration that has been interpreted to reflect magmatic-hydrothermal fluid input in these stringer systems (Thurlow 2010), which has also been observed elsewhere in the Buchans-Roberts Arm belt (Sparkes 2022; Sparkes and Hinchey 2023).

The mineralization in the Buchans district has remarkably low pyrite contents and is dominated by sphalerite, galena, chalcopyrite, and barite with lesser tetrahedrite–tennantite that display excellent textural preservation and have features like those found on the modern seafloor, including relict framboids and chimney structures, atoll textures, and evidence for zone refining and dissolution–reprecipitation processes (Fig. 8).

The Gullbridge-Lake Bond district is hosted within a structurally imbricated, ∼470 Ma bimodal volcanic sequence that contains arc tholeiitic to back-arc basin basalts and lesser calc-alkalic felsic to intermediate composition tuffaceous volcanic rocks (Sparkes et al. 2021). The two main deposits are the bimodal-mafic Gullbridge and Lake Bond deposits (O’Brien 2007; Fig. 9A). The Gullbridge deposit contains Cu-rich stockwork vein and disseminated mineralization that contains pyrite, pyrrhotite, chalcopyrite, and magnetite within a metamorphosed VMS alteration zone dominated by cordierite‒andalusite‒anthophyllite‒iron chlorite (Upadhyay and Smitheringale 1972; Sparkes 2020; Fig. 9B). The Lake Bond deposit consists of a Zn-rich zone of stockwork-style mineralization dominated by pyrite, sphalerite, lesser chalcopyrite, and trace galena, hosted by ∼470 Ma chlorite–epidote-altered island arc tholeiitic mafic rocks (Fig. 9C; Swinden 1988b).

The Pilley’s Island district is within the northern extremity of the Buchans-Roberts Arm belt in Notre Dame Bay (Fig. S3). Like elsewhere in this belt, the Roberts Arm Group in this region is thrust imbricated and comprises panels of felsic volcanic and volcaniclastic rocks intercalated with panels of basaltic rocks (Thurlow 1996; McKinley 2013). All VMS deposits in the region are hosted within the Spencer’s Dock, Old Mine, and Bull Road thrust panels (Fig. 10). Only the Old Mine deposit has a historic resource; other deposits (Jane’s Cove, Rowsell’s Cove, and Spencer’s Dock) within the Spencer’s Dock panel are pyrite-dominated and lack formally reported resources (Thurlow 1996). The Old Mine and 3B panels contain showings, some of which have high grades but are of limited areal extent. The historic, past-producing Old Mine is found within the Old Mine panel and is hosted by multiple rhyolite/dacite flow units and volcaniclastic rocks. It contains chalcopyrite–pyrite lenses underlain by stockwork mineralization associated with sericite (and lesser chlorite) alteration (Tuach 1988, 1990; Santaguida et al. 1992; McKinley 2013). The mineralization is interpreted to have formed from exhalative processes with deposition of the sulfide on the seafloor and having been fed by an underlying stockwork feeder system; these deposits have numerous textures similar to sulfides that have formed on the modern seafloor (Fig. 10; Tuach 1988, 1990; Santaguida et al. 1992; McKinley 2013). The Old Mine panel also hosts the 3B and Bumblebee Bight showings. The 3B showing consists of lenses of pyrite and chalcopyrite with underlying stringer mineralization hosted in rocks like the Old Mine stratigraphy (Santaguida et al. 1992), whereas the Bumblebee Bight showing comprises pyrite–chalcopyrite stringers within pillow lavas and felsic volcaniclastic rocks (Fig. 10). The Bull Road showing is hosted in the Bull Road panel and consists of massive sulfide clasts within a polylithic breccia containing both sulfide clasts and volcanic fragments; locally, clasts can be traced over 200 m and some have Zn–Pb–Cu grades and sulfide textures (Fig. 10) like the transported ores in the Buchans district (Tuach 1988, 1990). In the Spencer’s Dock panel, three pyrite-dominated deposits are hosted by dacitic rocks in three sub-panels (Thurlow 1996). Volcanic rocks in this panel are remarkably well preserved and dominated by flow-banded dacite and rhyolite that contain vesicles, amygdules, and perlitic cracks (Fig. 10; Thurlow 1996). Deposits in the Spencer’s Dock panel are associated with extensive sericite alteration and show evidence for formation via subseafloor replacement of the felsic flows (Thurlow 1996).

Peri-Gondwanan VMS belts

Tally Pond belt

The Tally Pond belt contains VMS deposits hosted by the bimodal ∼513–509 Ma Tally Pond group (Table S1; Fig. 14). The Tally Pond group has been subdivided into the felsic-dominated Bindons Pond formation and the mafic-dominated Lake Ambrose formation, which collectively are interpreted to represent a rifted arc that was built on ∼565 Ma juvenile arc crust of Ganderian affinity (Dunning et al. 1991; Rogers et al. 2006).

The deposits are subdivided into the past-producing Cu–Zn Duck Pond and Boundary deposits (Table S1) and the Zn–Pb–(Ag–Au–Cu) Lemarchant deposit (Squires et al. 2001; McNicoll et al. 2010; Piercey et al. 2014; Gill et al. 2016; Cloutier et al. 2017). The Duck Pond deposit is hosted by ∼509 Ma felsic rocks of the Bindons Pond formation within a mineralized block that is overthrust by a bimodal sequence of basalts and rhyolitic rocks that host minor hydrothermal mudstones but lack significant mineralization (Fig. S4). The Duck Pond deposit consists of numerous lenses of pyrite, chalcopyrite, sphalerite, and lesser galena and sulfosalts, with the bulk of all mined mineralization coming from the Upper Duck lens. The lenses are hosted within jigsaw-fit rhyolite breccias and to a lesser extent massive, blocky rhyolite (Figs. 11 and 12). Individual sub-lenses within the Upper Duck lens show zone refining with earlier pyrite–sphalerite assemblages overprinted and recrystallized by chalcopyrite-rich assemblages, which are associated with sericite and chlorite-(dolomite) alteration, respectively (Fig. 13).

The Boundary deposit is located ∼5 km north of the Duck Pond deposit and consists of two mineralized zones, the North and South zones, offset by the Wagner fault. The deposit is hosted by a footwall lapilli tuff and lapillistone and a hanging wall of flow-banded rhyolite (Figs. 11 and 12; Squires et al. 2001; Piercey et al. 2014). The mineralized zones consist of pyrite, sphalerite, chalcopyrite, and lesser galena and exhibits features indicative of zone refining. The mineralized zones are also spatially associated with both pipe-like and laterally extensive blankets of footwall chlorite–(dolomite) and sericite–quartz alteration and a hanging wall dominated by quartz–sericite alteration (Fig. 12). The ores at both Duck Pond and Boundary display features that indicate formation occurred predominantly by subseafloor replacement of volcaniclastic and volcanic rocks, with much lesser exhalative mineralization (Squires et al. 2001; Piercey et al. 2014).

The Lemarchant deposit is located south of the Duck Pond deposit and is hosted in a thrust-imbricated package of bimodal volcanic rocks (Gill et al. 2016; Cloutier et al. 2017). The deposit contains two zones, the Main Zone and Northwest Zone, with most sulfide mineralization found in the Main Zone. The deposit consists of stratiform sulfide–barite that is at the boundary between footwall blocky rhyolite, rhyolite breccias, and hanging wall basaltic rocks (Figs. 11 and 12; Gill et al. 2016; Cloutier et al. 2017). The mineralization is mineralogically complex, consisting of massive sulfide, barite, and stringer sulfides that contain abundant sulfides, oxides, precious metal phases, and sulfosalts that show evidence of formation from both exhalative processes, as well as potential fluid boiling and input from magmatic-hydrothermal fluids and deposition in relatively shallow water (∼1500 mbsl; Fig. 12; Gill et al. 2016).

Long Lake belt

The Long Lake belt contains the Long Lake VMS deposit and several other occurrences within the ∼514–506 Ma volcanic and volcaniclastic rocks of the Long Lake Group of the Victoria Lake supergroup (Fig. 13; Evans and Kean 2002; Hinchey and McNicoll 2016). The basalt, andesite, and rhyolite host rocks have partially overlapping U–Pb ages with the Tally Pond group. Near the Long Lake deposit, the rocks are bimodal with felsic-(to intermediate) compositions predominating over mafic rocks. Felsic rocks within the group can be broadly divided into two packages: (1) ∼506 Ma quartz ± feldspar phyric, felsic to intermediate rocks and medium- to coarse-grained volcaniclastic rocks—the upper stratigraphy and (2) ∼514 Ma white to pink, aphyric to quartz ± feldspar porphyritic, magnetite bearing, massive rhyolite, and fine-grained magnetite-bearing felsic tuff—the lower stratigraphy (Fig. 13; Hinchey and McNicoll 2016). Both packages contain fine-grained tuff, volcanogenic siltstone, and graphitic shale and are associated with hydrothermal alteration associated and disseminated to massive sulfides. Mafic rocks in the group are dominated by mafic tuff, pillow basalt, and breccia. The bimodal felsic Long Lake deposit is hosted within the lower stratigraphic units. The group has been subjected to polyphase deformation resulting in fold-thrusted structures that have overprinted and modified stratigraphy and mineralized horizons (Hinchey 2014). The Long Lake group is interpreted to have formed within a continental arc environment with the lower VMS package having geochemical signatures indicative of formation in a rifted continental arc (Hinchey and McNicoll 2016), like other parts of the Penobscot–Victoria arc, including the Tally Pond group to the east (Rogers et al. 2006; Zagorevski et al. 2010; Piercey et al. 2014).

The Long Lake deposit is hosted by a sequence of felsic and mafic volcanic rocks with minor cherty sedimentary rocks that contain narrow (cm to m scale) intervals of barite-rich massive sulfide composed of sphalerite, chalcopyrite, galena, and pyrite (Fig. 14). The host rocks have experienced pervasive and intense sericite, pyrite, chlorite, carbonate, and quartz alteration. Both the host rocks as well as the massive sulfides display evidence of recrystallization related to post-VMS structural overprinting (Hinchey 2014).

Tulks belt

The ∼498–491 Ma Tulks volcanic belt outcrops as a 65 km long by 8 km wide belt of rocks trending from the northeast to southwest parallel to the Beothuk Lake Line (Fig. 15). This volcano-sedimentary belt is dominated by felsic compositions with lesser mafic volcanic/volcaniclastic and sedimentary rocks. The predominant lithologies are quartz–feldspar porphyritic felsic to intermediate composition volcaniclastic rocks, massive rhyolite, and felsic to intermediate tuffs (with variable crystals and lapilli) and minor subvolcanic porphyritic intrusions. Mafic volcanic rocks are subordinate and consist of fine tuff, lapilli tuff, breccia, local pillow basalt, massive flows, and hypabyssal intrusions. Black shale, argillite, and greywacke are also locally abundant, especially in the southern portion of the belt (Hinchey 2011).

The Tulks volcanic belt hosts five VMS deposits, from south to north: Boomerang/Domino, Tulks Hill, Tulks East, Daniels Pond, and Bobby’s Pond (Fig. 15). The deposits are hosted by felsic volcanic and volcaniclastic rocks and sedimentary rocks. They are interpreted to have had varying emplacement styles of formation, including exhalative deposition (e.g., Daniels Pond) and subseafloor replacement (e.g., Boomerang, Tulks East, Tulks Hill; Kean and Evans 1986; Hinchey 2011, and references therein). The deposits are interpreted to have formed within a continental back-arc or arc rift environment as it transitioned from normal continental arc magmatism (Hinchey 2011).

The deposits in the belt include felsic–siliciclastic, bimodal-felsic, and hybrid bimodal-felsic deposit types. There is a regional shift in volcanic environment and interpreted tectonic setting from the southern to northern part of the belt. Deposits in the southern part of the belt (Boomerang/Domino, Tulks Hill, and Tulks East) are bimodal-felsic to felsic–siliciclastic deposits that are Zn-rich and polymetallic (Zn–Pb–Cu–Ag–Au) and have variable sericite–chlorite–quartz–pyrite alteration. In some deposits, siliciclastic sedimentary rocks are more abundant, particularly proximal to the mineralized zones (e.g., Boomerang deposit; Fig. 14). Of these deposits, Boomerang deposit displays textural evidence for subseafloor replacement-style mineralization whereby sulfides contain relict textures and crystals of the original host, which is interpreted to reflect replacement of permeable volcanic and sedimentary facies by sulfide during ore formation (Hinchey 2011, and references therein; English 2019).

In the northern part of the Tulks belt, the VMS deposits are bimodal felsic to hybrid bimodal felsic, polymetallic and hosted by felsic to intermediate volcanic successions. The host rocks commonly display jigsaw-fit breccia textures and vent-proximal rhyolite facies (Fig. 14; Gibson et al. 1999). In addition to sericite, quartz, chlorite and pyrite alteration zones, the Bobby’s Pond and Daniels Pond deposits also have minor aluminous argillic alteration with paragonite, montmorillonite ± kaolinite (Fig. 14), assemblages that are common of VMS deposits that have been influenced from low-pH magmatic-hydrothermal fluids (Sillitoe et al. 1996; Huston and Kamprad 2001; Dubé et al. 2007). Situated geographically between the Daniels Pond and the Bobby’s Pond VMS deposits are the Bobby’s Pond native sulfur occurrence (Fig. 14) and the North Pond VMS occurrence. These occurrences contain massive to laminated sulfide (predominantly pyrite) hosted within felsic volcanic rocks with zones of advanced argillic, aluminous alteration (e.g., pyrophyllite, dickite, and alunite), native sulfur, topaz, orpiment, and possible stibnite (Fig. 14), which are flanked by paragonite- and kaolinite-bearing zones. The Bobby’s Pond native sulfur occurrence also displays vuggy textures, in addition to having laminated exhalative massive sulfide horizons. When considered together, the local examples of acidic aluminous argillic alteration at the Bobby’s Pond and Daniels Pond VMS deposits, this hybrid epithermal-VMS environment occurs over a strike length of 8 km (Fig. 15). The deposits in the northern Tulks belt are interpreted to have formed in relatively shallow water (<1500 mbsl) relative to the deep water, sediment-rich distal environment suggested for the southern Tulks belt (Hinchey 2011; Sparkes and Hinchey 2023).

Great Burnt Lake/South Pond belt

The Great Burnt Lake volcanic belt contains VMS deposits within the Ordovician (∼466 Ma) Cold Spring Pond Formation (Fig. 16) (Swinden 1988a). The Cold Spring Pond Formation is composed of volcano-sedimentary rocks dominated by siliciclastic sedimentary rocks (volcaniclastic arkose, argillite, and greywacke) with subordinate (∼30%) mafic volcanic rocks, including pillow lavas and massive basalt and felsic volcanic rocks (Swinden 1988a). The formation is in fault contact on all sides with rocks of the Baie D’Espoir Group of the Exploits Subgroup and the Pipestone Pond ophiolite complex to the east and rocks of the Meelpaeg Subzone of the Gander Zone to the west (Fig. 17).

Mafic–siliciclastic-type VMS deposits in the belt occur within sequences of interlayered chlorite quartz ± carbonate-altered basalt flows and mafic tuffs and fine-grained, finely laminated biotite-rich argillite to siltstone (Fig. 16) inferred to have formed in a back-arc basin environment (Swinden 1988a; Barry et al. 2022). Sulfide zones are close to lithological contacts. Four main zones of mineralization are recognized, including the Great Burnt copper deposit, the South Pond copper deposit, the South Pond copper–gold zone and the End Zone copper prospect (Fig. 16); the first two have formal calculated resources (Barry et al. 2022). The Great Burnt copper deposit comprises stratabound zones of pyrrhotite–chalcopyrite ± sphalerite, galena, and pyrite as stringers, blebs, and massive sulfide (Fig. 17); more recent drilling has also documented Au-enriched (over 1 g/t over 4 m; Putrich and Barry 2015) pyrrhotite–chalcopyrite zones down dip from the main deposit. The South Pond copper deposit occurs approximately 10 km to the north of the Great Burnt deposit and is hosted within chlorite–quartz-altered mafic tuffaceous rocks interbedded with fine-grained volcaniclastic rocks (Figs. 16 and 18).

Although very similar to the Great Burnt and South Pond copper deposits, the South Pond Cu–Au zone has predominant gold-bearing sulfides disseminated to semi-massive pyrrhotite, chalcopyrite, and pyrite hosted within quartz-altered and sheared mafic tuffaceous rocks interbedded with fine-grained metasedimentary rocks (Figs. 17 and 18). Gold grades range from 100 to 3000 ppb; values up to 13.37 g/t gold over 1.17 m have been drilled (Denoyers 1991; Barry et al. 2022) with higher gold grades correlating with increased sulfide content. However, it is uncertain whether the gold enrichment is syngenetic or epigenetic, given the intense foliation and shearing, local brecciation, and increased quartz alteration (Fig. 17; e.g., Desnoyers 1991).

Point Leamington belt

The Point Leamington belt is within Cambrian volcanic and volcaniclastic rocks of the Wild Bight Group (Fig. S5; Swinden 1988b; MacLachlan and Dunning 1998; MacLachlan et al. 2001). Sulfide deposits are hosted predominantly in the Glover’s Harbour Formation, which is composed of a >486 Ma sequence of pillowed and brecciated mafic rocks, lesser quartz–feldspar phyric felsic flows and domes, interbedded breccias, lapilli tuffs, green and red chert, and argillite (MacLachlan and Dunning 1998; MacLachlan et al. 2001). Compositionally, mafic and felsic rocks of this formation have island arc tholeiitic and low-K, high-silica rhyolite affinities, respectively, and are interpreted to record formation within a peri-Gondwanan rifted primitive arc (MacLachlan and Dunning 1998; MacLachlan et al. 2001).

The Point Leamington VMS deposit is contained in the southwestern part of the Wild Bight Group within a variably thrust-imbricated and folded package of volcanic and volcaniclastic rocks (Fig. S5; Swinden 1988b; Walker and Collins 1988; Stone et al. 2021). The deposit has a deep mafic footwall that is overlain by felsic volcanic and volcaniclastic rocks that form the immediate stratigraphic footwall to the sulfide zone (Walker and Collins 1988; Stone et al. 2021). These shallow footwall felsic rocks are variably quartz- and feldspar-bearing and altered to a chlorite–sericite–pyrite alteration assemblage proximal to mineralization (Fig. 18). Overlying the sulfide zone is a marker unit of chert–argillite that occurs in most locales at the contact between hanging wall and footwall rocks and is overlain by a hanging wall of mafic to intermediate volcanic and volcaniclastic rocks (Stone et al. 2021). Mineralized zones consist of pyrite-dominated massive sulfide and fine-grained, pyrite-clast bearing sulfides that contain much lesser sphalerite, chalcopyrite, galena, and other trace sulfide/sulfosalt phases (Fig. 18; Walker and Collins 1988; Stone et al. 2021). The mineralized lenses are typically lower grade along their upper contacts but increase in grade both inwards and down dip, with higher Zn–Au-enrichments occurring at the base of the lenses (Stone et al. 2021). There are also possibly multiple lenses, as some sulfide lenses are intercalated with altered felsic volcaniclastic rocks (Stone et al. 2021). Underlying the massive sulfides are stringer sulfides dominated by pyrite with much lesser sphalerite, arsenopyrite, and chalcopyrite (Walker and Collins 1988; Stone et al. 2021).

The Lockport deposit is located proximal to Glover’s Harbour in the northern part of the Wild Bight Group (Fig. S5). This deposit has seen past production in the 1880s and sporadic exploration to the present (Swinden 1984; Barbour and Churchill 2001). The deposit is hosted by variably deformed and altered pillowed to amygdaloidal basaltic flows, andesites, and hyaloclastite, in which mineralization occurs predominantly as stringers of pyrite–chalcopyrite associated with quartz–sericite–chlorite altered host rocks (Fig. 18; Swinden 1984; Swinden 1988b; Barbour and Churchill 2001). Felsic rocks are not immediate hosts to sulfide zones; they are found along strike within the mineralized package and regionally and are interpreted to be equivalent to felsic host rocks at Point Leamington (Swinden 1984, 1988b; Barbour and Churchill 2001). Swinden (1984) noted that there is no exhalative mineralization preserved in situ and on surface in the Lockport deposit; however, variably Zn-rich boulders occur within the waste pits in the region, suggesting that past mining may have mined Zn-rich massive sulfide. Numerous variably base metal-rich sections were also intersected in drilling in the late 1990s and early 2000s (Barbour and Churchill 2001).

Tectonics, magmatism, and VMS formation

Volcanogenic massive sulfide deposits, including Appalachian VMS deposits, are associated with specific tectonic settings, tectonostratigraphy, and magmatic processes throughout geological time (e.g., Piercey 2011). For example, Swinden (1991) utilized trace element geochemistry of mafic rocks and noted that the most prospective sequences in Newfoundland are those associated with rifted arcs, be they intra-oceanic (e.g., suprasubduction zone ophiolite environments) or continental (e.g., Buchans-Roberts Arm belt). Recent work has expanded on these original contributions and provided both regional and deposit-scale relationships that illustrate the complexity of magmatism and volcanism associated with VMS mineralization.

In the case of mafic-dominated deposits, available geochemical and isotopic data suggest that boninitic to arc tholeiitic rocks are important hosts and they are interpreted to have formed in forearc sequences, at least for those associated with the Lushs Bight Group and the Betts Cove ophiolite (van Staal 2007; Bedard et al. 1999). This boninite-low-Ti tholeiite association is also present in the bimodal mafic Point Leamington belt where mafic rocks are dominated by island arc tholeiites and lesser boninites (MacLachlan and Dunning 1998). Sulfide deposits in this district, however, are associated with tholeiitic felsic rocks that formed during arc rifting due to upwelling of basaltic magmas and remelting of hydrated mafic arc crust (MacLachlan and Dunning 1998), similar to models proposed for modern primitive arc rift environments (e.g., Schmitt and Vazquez 2006; Shukuno et al. 2006; Tamura et al. 2009). Remelting of crust during rifting was also an important process in the bimodal felsic environments of the Dunnage Zone. For example, in the Tally Pond, Tulks, and Buchans-Roberts Arm belts, arc rifting and remelting of variably evolved continental crust was important in generating the VMS-hosting felsic rocks in these belts (Rogers et al. 2006; Zagorevski et al. 2006; Piercey et al. 2014; McKinley 2013).

The relationships outlined above between VMS deposits and specific geochemical/petrological clans and their interpreted tectonic settings is not at all surprising. For example, the boninite-low Ti ± tholeiitic rhyolite association is common in many primitive arc VMS environments (e.g., Piercey 2011). These rocks are generally attributed to subduction initiation and forearc rifting, which result in extensive spreading and extension (e.g., Stern and Bloomer 1992) and creation of the crustal permeability needed for VMS hydrothermal fluid flow (e.g., Bedard et al. 1999). The forearc extension in these environments also likely resulted in upwelling of hot, asthenosphere-derived magmas that drove both crustal melting to form VMS-associated rhyolites and also fluid circulation for forming both mafic and bimodal mafic VMS deposits within these settings (e.g., Piercey 2011). Upwelling of basalt and underplating of continental crust were additionally important in the formation of VMS deposits in felsic-dominated belts. The upwelling of basalt during continental arc rifting and back-arc formation, and melting of continental crustal basement, is interpreted as the formation mechanism for VMS-associated felsic rocks in the Tally Pond, Long Lake, Tulks, and Buchans-Roberts Arm belts (Rogers et al. 2006; Zagorevski et al. 2006; Hinchey 2011, 2014; McKinley 2013; Piercey et al. 2014; Hinchey and McNicoll 2016). Rifting and upwelling of mafic magma were also likely critical for driving hydrothermal circulation and fluid focusing within these continental VMS environments (e.g., Piercey 2011).

In contrast to the models outlined above, the dacitic rocks that host the Au–(Cu)-rich deposits of the Rambler-Ming district in the Baie Verte belt have relatively unique geochemical and isotopic signatures and magmatic history. These rocks have extremely depleted Y and heavy rare earth element signatures, coupled with distinctly high La/Yb, Zr/Sm, εNdt–εHft, and positive ΔεHf values, which have been attributed to these felsic rocks having been direct melts from the subducted oceanic slab (Pilote and Piercey 2018; Piercey and Pilote 2021). Previous work has suggested that gold enrichment, as well as other epithermal suite elements in the Rambler-Ming deposits were due to magmatic-hydrothermal contributions (Brueckner et al. 2014b, 2016).

Slab melting may provide a possible mechanism to explain this magmatic-hydrothermal association and gold enrichment. Numerous workers have shown that slab melting generates magmas that are siliceous, oxidized, sulfide-undersaturated, and SO4-rich (Fig. 19; e.g., Mungall 2002). Further, the geochemical signatures of the Rambler felsic rocks suggest that garnet was a stable residual phase within the slab during magma genesis (e.g., Defant and Drummond 1990), which likely resulted in magmas that were rich in ferric iron, as garnet preferentially retains ferrous iron (Lee and Tang 2020). Such melts are sulfide-undersaturated and will not fractionate magmatic sulfide as they transit the crust, thus allowing for a build-up in both Au and Cu in the melt during magma fractionation and subsequent emplacement at higher crustal levels. Models for slab melts also suggest that they undergo melt-rock reaction with the mantle wedge en route to emplacement within the crust (Defant and Drummond 1990). Given the sulfur-undersaturated nature of slab melts, any potential interaction with sulfides in the mantle wedge, or with Au–Cu-rich boninitic rocks in the stratigraphy (e.g., Keays 1987), would have resulted in dissolution of sulfides from these rocks (Kerr and Leitch 2005; Yang 2012), potentially further increasing the metal abundance within the slab-derived melts (Fig. 19). If such a melt exsolved and contributed magmatic-hydrothermal fluids to the Rambler-Ming VMS system, this mechanism could explain the enrichment of Au and epithermal suite elements (Brueckner et al. 2014b, 2016) and alteration types (Pilote et al. 2020) found in this district (Fig. 19). The above processes may also explain the association of similar fractionated FI-type (Lesher et al. 1986; Hart et al. 2004) felsic rocks associated with Au-rich VMS deposits in Archean greenstone belts (e.g., Mercier-Langevin et al. 2007; Gaboury and Pearson 2008).

Although there are general understandings of the relationships between magmatism and VMS mineralization in the NL Appalachians, fundamental questions remain unanswered. For example, the general knowledge regarding tectonics and magmatic relationships to VMS is accepted, but there are very poor controls on the camp- and deposit-scale chemostratigraphy and petrochemical assemblages (e.g., Piercey 2011) associated with VMS deposits in many districts. Furthermore, how specific magmatic clans, their petrogenesis, and underlying basement and(or) footwall sequences in a given district control metal assemblages and abundances in each VMS belt is poorly understood. In the Buchans-Roberts Arm belt, for example, there are similarities in magmatic products and interpreted basement/footwall sequences, but in one part of the belt are some of the highest-grade VMS deposits in the world (e.g., Buchans), whereas in the other end of the belt there are large but generally pyrite-bearing deposits with low grades (e.g., Pilleys Island)—why this is the case is not well understood. A related problem involves minimal data and understanding of the bulk rock isotopic signatures (particularly Hf isotopes) and mineral-scale chemical and isotopic data for zircon or other resistate phases in VMS-related and barren rocks in Newfoundland and globally. These data can provide constraints on potential basement compositions, the role of juvenile basalt underplating (Manor et al. 2022), as well as the physicochemical conditions of magma formation (e.g., Ferry and Watson 2007; Dilles et al. 2015) and their potential role in sourcing and(or) concentrating metals in VMS environments. These topics are potentially fruitful avenues for future research, e.g., on the relationship of magmatism to VMS deposit genesis in the Dunnage Zone of the Newfoundland Appalachians.

Emplacement mechanisms

Genetic models for ancient VMS deposits are based on our understanding of their modern analogues, seafloor massive sulfide (SMS) deposits, which invoke fluids “exhaling” into the ocean forming sulfide/sulfate chimneys and accumulations of massive sulfide on the seafloor (e.g., Franklin et al. 2005; Hannington 2014). Although numerous VMS deposits are interpreted to have formed in this manner, there are also many deposits that show evidence for subseafloor replacement by either infilling permeable/porous material or via dissolution and replacement of existing rock units (e.g., Doyle and Allen 2003; Piercey 2015). In addition, some deposits record deposition from debris flows via the collapse of chimneys and subsequent mass flow sedimentation (e.g., Thurlow and Swanson 1981; Binney 1987; Roth et al. 1999).

Numerous VMS deposits within the Dunnage Zone have features indicative of seafloor exhalative VMS activity, including mound-like barite chimneys (Thurlow 2010; Lajoie et al. 2020), and microscopic evidence, including colloform textures and micro-chimneys (Gill et al. 2016). These features are interpreted to record seawater and VMS vent fluid mixing and subsequent deposition of sulfates (Ohmoto 1996) and(or) sulfides (Lydon 1988) on the seafloor. Some of these exhalative deposits and occurrences are also spatially associated with laterally extensive hydrothermal mudstone and iron formations that have geological, mineralogical, geochemical, and isotopic signatures indicative of venting and deposition on the seafloor (Hinchey 2011; Lode et al. 2015; Piercey et al. 2018).

Transported sulfides also are interpreted to have formed on the seafloor, albeit due to gravity-induced redeposition of exhalative (±replacement) mineralization. In the Buchans district, transported ore bodies were attributed to the collapse of sulfide deposits and transportation as sediment-gravity flows (e.g., sulfide-rich turbidites; Thurlow and Swanson 1981; Binney 1987). In other cases, the transported deposits contain bedded to laminated sulfides, even in deposits interpreted to have formed primarily by subseafloor replacement (Piercey et al. 2014). In both cases, it is likely that basin extension played a role in the instability of sulfide chimneys/deposits and in their subsequent transport and deposition. Further, given the presence in many deposits of textures indicative of dissolution, reprecipitation, and zone refining (e.g., Eldridge et al. 1983), it is also likely that these exhalative deposits experienced chimney collapse due to retrograde solubility of anhydrite (Blounot and Dickson 1969), coupled with subsequent re-deposition of sulfide either proximal to the vent (e.g., Piercey et al. 2014) or distal within sediment-gravity flows (e.g., Thurlow and Swanson 1981; Binney 1987).

Recently, numerous workers have shown that subseafloor replacement is an important deposit-forming mechanism. Squires et al. (1991, 2001) suggested that many of the VMS ores in the Duck Pond deposit formed via replacement. Similarly, Thurlow (1996) proposed that ores in the Spencer’s Dock area of the Pilley’s Island district formed by partial replacement of perlitic cracks and glass in felsic flows. Hinchey (2011) illustrated that in the Tulks belt, some volcaniclastic- and sediment-rich VMS deposits (e.g., Boomerang, Tulks East) have lithofacies, alteration styles, and mineralized zones indicative of subseafloor replacement. Piercey et al. (2014) suggested that the Boundary deposit formed due to permeability contrasts between relatively impermeable, hanging wall flow banded rhyolites and unconsolidated and porous footwall lapilli tuff, resulting in the lateral replacement of pore spaces in volcaniclastic rocks by sulfide (Fig. 20). Observations and relationships from the Boundary and Duck Pond deposits were also critical for creating a revised global model for subseafloor replacement processes in VMS deposits (Piercey 2015).

Despite the recognition of variable emplacement mechanisms for VMS formation in the Newfoundland Dunnage Zone, it is still uncertain how such mechanisms control grades and tonnages and the metal assemblages preserved in these deposits. Further, in many VMS deposits and belts the nature of emplacement of sulfide deposits is not fully understood, but this is required for refinement and enhancement of predictive mineral exploration models for VMS deposits.

Sources of metals, fluids, and processes of metal enrichment

Insights into the origin of metals, fluids, and sulfur in VMS deposits have come from studies of ancient VMS deposits and from seafloor systems where the mineralizing fluids can be sampled directly. Research on modern seafloor vent fluids demonstrates significant diversity of compositions, but many of these fluids are dominated by modified seawater (e.g., Von Damm 1990; German and Seyfried 2014). Importantly, however, it has been demonstrated that some hydrothermal vent fields contain input from magmatic fluids/volatiles (e.g., de Ronde et al. 2011). This diversity of fluid compositions also results in contrasting metal and alteration assemblages (Hannington et al. 2005). In ancient VMS environments, fluid inclusion studies are very limited owing to preservation and resetting/overprinting by post-VMS orogenic events (e.g., Lajoie et al. 2020). Thus, a direct understanding of VMS fluid characteristics is poor in the Dunnage Zone; hence, it is inferred predominantly from metal assemblages and textures preserved in the sulfides and sulfates, from hydrothermal alteration assemblages, and to a lesser extent from S isotope systematics. For example, in many bimodal felsic deposits (e.g., Duck Pond and Boundary), there is well-developed metal zoning with low-temperature (<300 °C) Zn–Pb–sericite–(quartz) assemblages and higher temperature (>300 °C) Cu-chlorite assemblages (e.g., Piercey et al. 2014; Buschette and Piercey 2016), whereas some ophiolite-hosted mafic deposits display high-temperature, reduced chalcopyrite–pyrrhotite-stringer and chlorite-dominated assemblages (e.g., Cloutier et al. 2015), features typically found in zone-refined, seawater-dominated SMS on the modern seafloor (e.g., Haymon 1983; Graham et al. 1988). From this indirect evidence, it is reasonable to assume that hydrothermally modified seawater was a component of the hydrothermal fluids and sulfur in most VMS deposits in the Dunnage Zone.

The metals in VMS were in part derived from footwall rocks that underlie VMS deposits, be they crystalline basement and(or) footwall stratigraphic sequences (e.g., footwall volcanic and(or) sedimentary rocks). For example, empirical studies have shown that regional scale leaching of footwall rocks can provide sufficient Cu, Zn, Fe, and Pb to potentially explain the metals present in many VMS deposits (e.g., Gibson et al. 1983; Galley 1993; Skirrow and Franklin 1994; Jowitt et al. 2012). Furthermore, the metal assemblages of VMS deposits in Newfoundland are interpreted to be in part impacted by the footwall units that underlie these deposits. For example, ophiolite-hosted mafic VMS deposits are notable for being Cu–Zn-rich and Pb-poor, compared to more bimodal-felsic and felsic–siliciclastic deposits that are more Zn–Pb–(Cu)-rich and partially to fully underlain by continental crust and crustal-derived volcanic and sedimentary rocks (e.g., Swinden and Kean 1988). The sourcing of VMS metals from footwall sequences is also partly supported by Pb isotope data. For example, Swinden and Thorpe (1984) illustrated that there was distinct provincialism in Pb isotopic signatures of galena in VMS deposits in the Dunnage Zone that reflected their underlying footwall and(or) basement domains, an interpretation that has been replicated and supported by more recent studies (e.g., Gill et al. 2019).

Although there is evidence to suggest leaching of footwall rocks is important in generating the metals in some deposits, other deposits with exceptional metal endowment and(or) magmatic-hydrothermal metal and alteration assemblages cannot be explained solely by leaching. Mass-balance modelling of basement leaching for Pb by Sawkins and Kowalik (1981) in the Buchans district illustrated that basement leaching models cannot account for all the Pb in the Buchans VMS deposits, thereby requiring additional input, likely from magmatic-hydrothermal fluids. A similar case may be made for the exceptional Zn grades found in these deposits (e.g., Piercey et al. 2015). In other deposits, magmatic input has been used to explain the high concentrations in Au and Ag, as well as of other epithermal suite/magmatic-hydrothermal elements (Hg–As–Sb–Bi–Te–Se–Sn–W), present in some VMS deposits. For example, in the Ming deposit, Brueckner et al. (2014b, 2016) demonstrated that enrichments in Au–Ag and epithermal-suite elements were due to inputs from magmatic-hydrothermal fluids that also led to the formation of distinctive green mica (fuchsite) alteration and quartz alteration (Fig. 19; Pilote et al. 2020). In the Lemarchant deposit, features including the sulfosalt-rich, high fO2fS2 ore mineral assemblages (with low-Fe sphalerite), Au–Ag-epithermal-suite element enrichments, illite–K–feldspar alteration zones, and textures like metal-rich bladed barite and carbonate have all been interpreted to record low- to intermediate-sulfidation magmatic-hydrothermal inputs into the VMS hydrothermal system (Gill et al. 2016, 2019; Cloutier et al. 2017; Lajoie et al. 2020; Fig. 21).

In other deposits, aluminous alteration assemblages have been invoked to reflect input from magmatic-hydrothermal fluids. For example, in the Tulks belt, aluminous alteration is present in the Bobby’s Pond deposit and accompanied by precious metal enrichment at the Daniels Pond deposits, which collectively attribute the influence from acidic fluids derived from magmatic-hydrothermal activity (Hinchey 2011; Sparkes and Hinchey 2023). The nearby Bobby’s Pond sulfur occurrence and the associated alunite–topaz–pyrophyllite–orpiment–stibnite alteration associated with the occurrence have features like those of high sulfidation epithermal Au deposits interpreted to be derived from magmatic-hydrothermal fluids (e.g., Hedenquist and Arribas 2022), albeit in a subaqueous VMS setting (Hinchey 2011). Similarly, advanced argillic alteration zones recognized in the Buchans-Roberts Arm belt at the Mary March prospect are also suggestive of magmatic-hydrothermal input in a broadly VMS environment (Sparkes 2022; Sparkes and Hinchey 2023).

The increased recognition of magmatic-hydrothermal influences in some VMS deposits in the Dunnage Zone has been a major advancement in our understanding of VMS deposits in Newfoundland over the past decade. Despite this, there are still significant knowledge gaps, including what ultimately controls grade and tonnage in VMS deposits in Newfoundland—is it the source of metals, efficiency of deposition at the seafloor–vent interface, extent of zone refining, duration of magmatic activity and associated hydrothermal systems, permeability/porosity of hosting sequences, and(or) magmatic-hydrothermal fluid input? Moreover, although there is a reasonable understanding of the causes of enrichment of some critical metals (e.g., Cu and Zn), the distributions and causes of enrichments in critical metals, like Co, Ni, Te, As, and Sb in VMS deposits in Newfoundland (e.g., Papezik 1964; Toman 2013) and globally (Monecke et al. 2016; Grant et al. 2018), is incompletely understood. Given the predicted increased need for such elements for electrification of the economy, having a better understanding of critical metal siting and enrichment processes in Newfoundland VMS deposits (and others globally) is essential and a potentially fruitful area for future research.

Diverse sources of sulfur

Most sulfur in VMS deposits comes from leaching of igneous sulfur from the underlying footwall rocks and subsequently TSR of seawater sulfate (e.g., Huston 1999). Less commonly, sulfur is derived from the bacteriogenic reduction of seawater sulfate (e.g., Lode et al. 2017) or rarely from magmatic-hydrothermal fluids by disproportionation of magmatic SO2 (e.g., Herzig et al. 1998). There have been remarkably few studies of sulfur isotopes in VMS deposits of the Dunnage Zone until recently. Classic work by Bachinski (1977) and Kowalik et al. (1981) demonstrated that sulfur in mafic and bimodal felsic deposits in the Betts Cove (δ34S = +5.5‰ to +23‰) and Lushs Bight (δ34S = +3.5‰ to +7.4‰) complexes and Buchans district (δ34S = +2.9 to +8.7), respectively, had δ34S values consistent with sulfur derivation by TSR of (Cambrian)-Ordovician seawater sulfate. In other VMS showings in the Buchans-Roberts Arm belt, there are indicators of TSR-derived S but also potentially S leached from footwall igneous rocks (δ34S = +2.5‰ to +6.5‰; Waldie et al. 1991; Santaguida et al. 1992).

All the S isotope studies cited above employed conventional methods using mineral separates, which generally consist of impure mineral mixtures, particularly for fine-grained materials. Most recent S isotope work on VMS deposits in the Dunnage Zone has utilized in situ methods (e.g., secondary ion mass spectrometry), which allows for greater absolute numbers of isotope determinations and importantly micro-scale paragenetic and textural control on S-isotope signatures of the sulfide minerals. Brueckner et al. (2015) illustrated that there was both igneous sulfur and TSR-derived sulfur in the Ming VMS deposit and that the sulfides there with higher temperature origins have greater igneous sulfur contributions. These authors also indicated some of the lower δ34S values may be from magmatic volatiles; however, it was not possible to delineate this source from leached igneous sulfur on the basis of S isotopes alone. Cloutier et al. (2015) demonstrated similar TSR-igneous sulfur sourcing in the mafic Whalesback deposit and that the S-isotope signatures in sulfides were unchanged during metamorphism and reflected original syngenetic VMS values, corroborating the classic work of Bachinski (1977). In the Lemarchant deposit, Gill et al. (2019) illustrated shifts in S-isotope signatures with paragenesis, by which early low-temperature (150–250 °C) Zn–Pb–sulfosalt-rich mineralization produced relatively low δ34S signatures and was dominated by igneous sulfur, whereas higher-temperature (>300 °C) Cu-rich assemblages had higher δ34S values and were dominated by TSR-derived S. They also argued that some of the lower δ34S values (δ34S < 0‰) in the Zn–Pb–sulfosalt-rich assemblages at Lemarchant may be due to input from disproportionated magmatic SO2. Lode et al. (2017) also report extreme diversity in sulfur sources in hydrothermal exhalites/mudstones metres from massive sulfide in the Lemarchant deposit. Sulfide assemblages within these mudstones have extreme δ34S variability (−38.8‰ to +14.4‰), which involved sulfur derived from bacterial reduction of seawater sulfate, microbial sulfide oxidation, and microbial disproportionation of intermediate sulfur compounds, as well as sulfur derivation from TSR and igneous sulfur; contributions from the latter sources increase in proximity to the deposit (e.g., within metres; Fig. 21).

The S-isotope studies of the VMS deposits of the Dunnage Zone illustrate the complex and diverse origins of sulfur in these deposits; however, detailed paragenetically and texturally controlled samples are restricted to only a few deposits and more research is needed to evaluate deposits with different metal assemblages, alteration types, and emplacement mechanisms (e.g., biogenic S is likely to be important in some replacement systems; Piercey 2015). Nevertheless, these recent studies demonstrate that the complexity of S-isotope signatures observed in ancient VMS in the Newfoundland Dunnage Zone is like that documented in other VMS belts globally and in modern SMS environments (Shanks and Seyfried 1987; Herzig et al. 1998; Nozaki et al. 2021).

The Dunnage Zone of the Newfoundland Appalachians hosts diverse styles of Cambrian-Ordovician VMS deposits in both the peri-Laurentian Notre Dame Subzone and the peri-Gondwanan Exploits Subzone. In the Notre Dame Subzone, ophiolitic rocks of the ∼510–501 Ma Lushs Bight oceanic tract and ∼488–485 Ma Baie Verte oceanic tract host Cu–(Zn)-rich mafic-type deposits within mafic rocks in ophiolite sequences. Bimodal mafic, Au-rich VMS deposits are hosted by ∼488–485 Ma dacitic rocks in the Baie Verte oceanic tract, whereas bimodal felsic deposits are within felsic-dominated sequences in the Buchans-Roberts Arm belt, including some of the highest-grade Zn–Pb–Cu–(Au–Ag) ore bodies ever mined on Earth. In the Exploits Subzone, Cambrian (∼513–509 Ma) bimodal felsic deposits are hosted in the Tally Pond and Long Lake belts, which are stratigraphically overlain by felsic-dominated VMS deposits in the ∼498–491 Ma Tulks belt. The Tulks belt has remarkable metallogenic diversity, with the southern end of the belt having polymetallic felsic siliciclastic deposits, whereas the northern part of the belt contains hybrid bimodal felsic deposits with some attributes typical of VMS deposits (e.g., massive Zn-rich sulfides and felsic host rocks) and other attributes that are hallmarks of epithermal Au deposits (e.g., precious metal- and epithermal-suite element enrichment and aluminous alteration). In the northern extremity of the Exploits Subzone, the Wild Bight Group hosts the ∼486 Ma Point Leamington belt that contains Cu–Zn-bearing bimodal felsic to bimodal mafic deposits. The youngest VMS mineralization in the Exploits Subzone consists of the Cu–Au mafic-siliciclastic deposits in the Great Burnt Lake/South Pond belt.

The VMS deposits of the Dunnage Zone in Newfoundland are associated with rift episodes and mineralization in extensional tectonic regimes (e.g., forearc rift, arc rift, and back-arc basins) with coeval magmatism that is interpreted to have driven hydrothermal circulation. Some Au-enriched deposits of the Rambler-Ming district are also associated with slab-melt-related felsic rocks. Slab melting, subsequent melt-rock interaction and magmatic-hydrothermal fluid/volatile exsolution, was likely important in causing Au enrichment in this district and possibly in other Au-rich VMS deposits globally.

Many of the VMS deposits formed by exhalative processes and deposition on the seafloor (e.g., Lemarchant and some in situ deposits in Buchans). Parts of some deposits formed via re-sedimentation and transportation of sulfide as sediment gravity flows (e.g., transported deposits in Buchans) or by subseafloor replacement of volcanic (e.g., some Jane’s Cove deposit in Pilley’s Island), volcaniclastic (e.g., Duck Pond and Boundary deposits), and(or) sedimentary rocks (e.g., Boomerang deposit) beneath the seafloor.

Metals within the VMS deposits were likely derived mainly by the leaching of underlying footwall rocks, but some have features indicative of derivation from magmatic–hydrothermal fluids, particularly deposits enriched in precious metals (e.g., Au–Ag) and epithermal-suite elements (e.g., Hg–As–Sb–Bi–Te–Se–Sn–W).

Sulfur in most of VMS deposits was sourced from either leached igneous sulfur in footwall rocks followed by TSR of seawater sulfate. A subset of deposits show evidence of influence from bacterial sulfur-derived BSR of seawater sulfate with the sulfur in some deposits potentially derived by disproportionation of magmatic SO2.

Despite a long history of mining, exploration, and research on VMS deposits in the Newfoundland Appalachians, numerous questions remain unresolved, thus providing important opportunities for future research and resource exploration and development.

There have been numerous graduate students and post-doctoral fellows in the Piercey research group who have contributed significantly to our understanding of VMS in the central Newfoundland Appalachians, including Alicia Escribano, Stefanie Bruckner, Mike Buschette, Jonathan Cloutier, Michelle English, Shannon Gill, Regan Jacobson, Kyra Kennedy, Marie-Eve Lajoie, Stefanie Lode, Conor McKinley, Taylor Mugford, Jean-Luc Pilote, and Helena Toman. We also build on the work of numerous geologists who have contributed immensely to our understanding of VMS deposits in Newfoundland and would like to recognize their contributions, including Terry Brace, Peter Dimmell, Greg Dunning, David Evans, George Jenner, Baxter Kean, Paul Moore, Brian O’Brien, Larry Pilgrim, Frank Santaguida, Tom Skulski, Kerry Sparkes, Gerry Squires, Peter Tallman, Scott Swinden, Geoff Thurlow, and Lawrence Winter. Frank Santaguida is especially thanked for donating Piercey his extensive collection of samples from VMS deposits in the Notre Dame Bay area, which have become a valuable resource for research (and teaching!) and have contributed significantly to this manuscript. Piercey’s VMS research has been supported by numerous partners in industry (Altius Minerals, Aur Resources, Canadian Zinc, Champion Iron, Cornerstone Resources, Messina Minerals, Paragon, Rambler Metals, and Teck) and government (Geological Survey of Canada, Geological Survey of Newfoundland and Labrador), and by the Natural Sciences and Engineering Research Council of Canada (NSERC) through both the Discovery Grant and Collaborative Research and Development grant programs. We thank John Slack and Jan Peter for formal reviews and Associate Editor Michelle DeWolfe for editorial comments—their suggestions have greatly improved this manuscript. Brendan Murphy and Sally Pehrsson are thanked for the invitation to contribute this manuscript (and gentle nudges to get it finished!).

All data presented in the paper are available in the original manuscripts cited or can be obtained from the authors upon request. The data used for constructing Figs. 4 and 5 can be found in Table S1 and are also available from publicly available sources cited in the table.

Conceptualization: SJP, JGH, GWS

Data curation: SJP, JGH, GWS

Formal analysis: SJP, JGH, GWS

Funding acquisition: SJP, JGH, GWS

Investigation: SJP, JGH, GWS

Methodology: SJP, JGH, GWS

Project administration: SJP, JGH, GWS

Resources: SJP, JGH, GWS

Validation: SJP, JGH, GWS

Visualization: SJP, JGH, GWS

Writing – original draft: SJP, JGH, GWS

Writing – review & editing: SJP, JGH, GWS

During construction of this manuscript, SJP was supported by an NSERC Discovery Grant. Previous research conducted by the authors and reviewed in this article was funded by various sources as outlined in the Acknowledgements.

Supplementary data are available with the article at

This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.

Competing Interests

The authors declare there are no competing interests.

Supplementary data