Hartetal2004_FelsicVolcanicVMS_

8/12/2019 Hartetal2004_FelsicVolcanicVMS_

http://slidepdf.com/reader/full/hartetal2004felsicvolcanicvms 1/12

0361-0128/01/3444/1003-11 $6.00 1003

Introduction

A preferential association of geochemically distinctive, sub-aqueous felsic volcanic rocks with volcanogenic massive sul-fide (VMS) deposits was first highlighted in the Archean Su-perior province of the Canadian Shield by Thurston (1981)and Campbell et al. (1982). Subsequent studies have shownthat all VMS deposits in the Superior province, includingKidd Creek and the deposits in the Noranda and SturgeonLake camps, are associated with geochemically distinctiverhyodacites, rhyolites, and high silica rhyolites (e.g., Lesher etal., 1986; Barrie et al., 1993).

These rocks have been classified as calc-alkaline and tholei-itic felsic volcanic rocks by Campbell et al. (1982), felsic vol-canic groups FII and FIII by Lesher et al. (1986), groups I,II, and III rhyolites by Barrie et al. (1993), and transitionaland tholeiitic rhyolites by Barrett and MacLean (1994; Table1). These classifications have been a useful area selection toolin the exploration for VMS deposits in Archean and Protero-zoic volcanic successions, and Lentz (1998) has shown thatthese geochemically distinctive felsic volcanic rocks are asso-ciated with many Phanerozoic VMS deposits. Because of common usage, we have retained the classification of Lesheret al. (1986). FI felsic volcanic rocks are characterized by steep REE patterns with weakly negative to moderately pos-itive Eu anomalies, high Zr/Y, and low abundances of high

TRACE ELEMENT GEOCHEMISTRY AND PETROGENESIS OF FELSIC VOLCANIC ROCKS ASSOCIATED WITH VOLCANOGENIC MASSIVE Cu-Zn-Pb SULFIDE DEPOSITS

T. R. HART,†

Precambrian Geoscience Section, Ontario Geological Survey, 933 Ramsey Lake Road, Sudbury, Ontario P3E 6B5, Canada, and

Mineral Exploration Research Centre, Department of Earth Sciences, Laurentian University, 933 Ramsey Lake Road,Sudbury, Ontario P3E 6B5, Canada

H. L. GIBSON, AND C. M. LESHER

Mineral Exploration Research Centre, Department of Earth Sciences, Laurentian University, 933 Ramsey Lake Road,Sudbury, Ontario P3E 6B5, Canada

Abstract

Volcanogenic massive Cu-Zn-(Pb) sulfide (VMS) deposits occur primarily in subaqueous rift-related envi-ronments (e.g., oceanic, fore-arc, arc, back-arc, continental margin, or continental), are hosted primarily by bi-modal, mafic-felsic volcanic successions, and are typically associated with felsic volcanic rocks with specific geo-chemical characteristics. FI alkalic dacites and rhyodacites, despite being abundant in the rock record, aretypically barren. Some FII calc-alkalic rhyodacites and rhyolites host VMS deposits, but most are barren. FIII

tholeiitic and FIV depleted rhyolites and high silica rhyolites are much less abundant in the rock record butcommonly host VMS deposits, regardless of age, and FIII rhyolites appear to host the largest deposits.Most petrogenetic models proposed for the formation of FII and FIII-FIV felsic volcanic rocks link felsic

magma genesis to fractionation processes in high-level magma chambers now represented by associated subvolcanic intrusions, where the magma is also interpreted to have supplied the heat and/or metals required togenerate and sustain the VMS-forming convective hydrothermal system. However, the relatively constant com-positions of FII and FIII-FIV felsic volcanic rocks within individual areas, the high eruptive temperatures (ator above liquidus) of FIII rhyolites, and the bimodality of VMS-hosting volcanic successions indicate that frac-tional crystallization within subvolcanic intrusions could not have generated or significantly modified the com-positions of FII and FIII-FIV magmas. This, coupled with detailed geological, geochemical, and geochrono-logical studies indicates that many of these subvolcanic intrusions were emplaced in multiple phases and thatthe later, most voluminous phases often cut ore-associated, hydrothermally altered rocks.

A reassessment of the physical conditions responsible for producing the geochemistry of ore-associated FIIand FIII-FIV felsic volcanic rocks and a review of the compositions of felsic volcanic rocks associated with VMSdeposits that range in age from Mesoarchean to Cenozoic provide important constraints on models for VMS-

associated felsic volcanic rocks and their relationship to mineralization. The compositions of felsic volcanicrocks may be explained by low to moderate degrees of partial melting of mafic sources at a range of depths within rift environments where the mineralogy and composition of the source regions, modes, and degrees of partial melting, pressure and temperature of melting, and, to a lesser extent, subsequent fractionationprocesses, account for the compositional variations from FI through FII to FIII-FIV. Long-lived, enhancedheat flow and structural permeability of rift environments that allows partial melting to form some FII rhyo-lites at midcrustal levels (10–15 km) and FIII-FIV rhyolites at shallow crustal levels (<10 km), both within thezone of brittle fracture permeability, are essential to sustain the high-temperature convective hydrothermal sys-tems that are required to form large VMS deposits and camps. Rift environments contain long-lived, thermal,magmatic, and structural corridors that focus magma ascent, heat flow, high-temperature convective hy-drothermal systems, and emplacement of subvolcanic intrusions that are favorable environments for the for-mation of VMS deposits and FII and FIII-FIV felsic volcanic rocks.

©2004 by Economic Geology Vol. 99, pp. 1003–1013

† Corresponding author: e-mail, [email protected]



field strength elements (HFSE; e.g., HREE, Y, Zr, Hf). FIIfelsic volcanic rocks are characterized by gently sloping REEpatterns with variable Eu anomalies, moderate Zr/Y, and in-termediate abundances of HFSE. FIII felsic volcanic rocksare rhyolites and high silica rhyolites characterized by rela-tively flat REE patterns. FIII rhyolites may be subdividedinto two types. FIIIa rhyolites exhibit variable negative Eu

anomalies, low Zr/Y, and intermediate abundances of HFSelements. FIIIb rhyolites exhibit pronounced negative Euanomalies, low Zr/Y, and high abundances of HFSE. FIV fel-sic volcanic rocks are rhyolites and high silica rhyolites char-acterized by flat to slightly LREE-depleted REE patterns andlow REE and HFSE abundances. Our compilation of Mesoarchean to Cenozoic VMS deposits indicates that de-spite their abundance no VMS deposits are hosted by FI fel-sic volcanic rocks, however, FI felsic volcanic rocks may bepresent in the same volcanic succession. Some FII rhyo-dacites and rhyolites host VMS deposits, but most are barren.FIII and FIV high silica rhyolites are much less abundant inthe rock record but commonly host VMS deposits, regardlessof their age, and FIII rhyolites appear to host many of thelargest deposits.

Lesher et al. (1986) and Lentz (1998) suggested that FIIfelsic volcanic rocks host most Phanerozoic and Proterozoic

VMS deposits, whereas most Archean VMS deposits arehosted by FIII rhyolites. This change in felsic volcanic rockcomposition over geologic time suggests a change in the pet-rogenetic process(es) by which felsic volcanic rocks haveformed, although most other features of VMS mineralizationremain unchanged. As a result, there has been some reluc-tance to use felsic volcanic rock classification as an explo-ration tool in younger volcanic successions, a situation furthercomplicated by the wide variety of petrogenetic models andtectonic environments in which different authors consider

FII, FIIIa, and FIIIb felsic volcanic rocks to have formed(Table 1).

Although a variety of petrogenetic models have been pro-posed for the formation of FII and FIII-FIV felsic volcanicrocks (Table 1), most link felsic magma genesis to fractiona-tion processes in high-level magma chambers, where themagma is also interpreted to have supplied the heat and/ormetals required to generate and sustain the ore-forming VMShydrothermal system (e.g., Campbell et al., 1981; Franklin etal., 1981). This argument has been supported by the presenceof high-level, comagmatic, subvolcanic intrusions such as theFlavrian pluton in the Noranda area and the Beidelman Bay pluton in the Sturgeon Lake area, which have felsic phases

that are geochemically equivalent to FIII and FII felsic vol-canic rocks (e.g., Goldie, 1976; Campbell et al., 1981; Mortonet al., 1991; Paradis et al., 1993).

Detailed geological, geochemical, and geochronologicalstudies have shown that these intrusions are composite, sill-like plutons that intrude the bases of their volcanic edificesand, in some instances, underlie large, VMS-hosting synvol-canic subsidence structures, such as the Noranda cauldronand Sturgeon Lake caldera (Goldie et al., 1979; Gibson, 1990;Morton et al., 1991; Galley, 2003). They were emplaced assill-dike swarms into hydrothermally altered rocks whereeach intrusive phase may have been accompanied by hy-drothermal alteration, and subsequent intrusive phases often

crosscut that alteration (e.g., Gibson and Watkinson, 1990;Galley et al., 2000; Galley, 2003). For example, at SturgeonLake the most voluminous intrusive phase of the BeidlemanBay pluton cuts hydrothermally altered footwall rocks, but itis not clear whether the alteration predates or is syngenetic

with the VMS ores (Galley et al., 2000; Galley, 2003). How-ever, as we will argue below, FII felsic volcanic rocks that host

VMS deposits at Sturgeon Lake must have originated atdepths (>10 km) greater than the level of the Beidleman Bay subvolcanic intrusion (<2 km), thus negating this intrusion asthe source of the ore-hosting FII felsic volcanic rocks. Inother areas, the most voluminous intrusive phase clearly cutsore-related discordant alteration zones (e.g., Noranda:Goldie, 1976; Gibson and Watkinson, 1990; Galley, 2003;Snow Lake: Galley, 2003), indicating that, at least in theseareas, the most voluminous phase of the intrusion postdates

VMS ore formation. The occurrence of altered and VMS-mineralized strata kilometers below some subvolcanic intru-sions, for example below the Flavrian pluton at Noranda, hasbeen interpreted to indicate that the intrusions were em-placed into a rifted tectonic environment, referred to as a“thermal corridor” by Galley (2003), which focused long-lived, high-temperature convective hydrothermal systems(Cathles, 1981; Parry and Hutchinson, 1981; Gibson et al.,1983; Hannington et al., 2003).

This raises the possibility that the high-level (<2-km) subvolcanic intrusions in these areas were not responsible for thegeneration of the ore-associated FII and FIII-FIV rhyolitesand were not the sole heat engines that drove the ore-form-ing, convective hydrothermal systems. Thermal models by Cathles (1981), Cathles et al. (1997), and Barrie et al. (1999)showed that the period of time that an intrusion is able to sus-tain high-temperature hydrothermal convection is primarily afunction of the mass of the intrusion, the temperature of the

magma, and the temperature and permeability of the hostrocks. For example, larger, higher temperature mafic or ul-tramafic intrusions emplaced at deeper crustal levels (10–18km), where the country rocks are characterized by lower per-meability and higher wall-rock temperatures, favor the gen-eration of long-lived, high-temperature, convective hy-drothermal systems (e.g., 5 × 105–1 × 106 yr; Cathles et al.,1997). In contrast, the smaller pre- and synore phases of somehigh-level subvolcanic intrusions may cool too quickly to sus-tain long-lived, convective hydrothermal systems but, de-pending on their mass and temperature, may be capable of generating high-temperature convective systems of smaller

volume and shorter duration (e.g., <1× 105 yr; Cathles et al.,

1997). We have also considered the possibility that thesmaller, pre-ore phases of the intrusions may have been dy-namic magma chambers that processed much greateramounts of magma and therefore generated much greateramounts of heat than is evident from the mass of preservedrocks, but the felsic-dominated intrusions show no evidenceof fractional accumulation and contain no cyclic units or otherevidence for a replenished system (Galley, 2003). Instead,they are composed of multiple intrusive phases separated by fine-grained intrusive contacts (Goldie, 1976; Galley, 2003).Therefore, although high-level subvolcanic intrusions mostlikely contributed heat to the hydrothermal systems andcertainly identify long-lived thermal-magmatic-structural

SCIENTIFIC COMMUNICATIONS 1005

0361-0128/98/000/000-00 $6.00 1005



anomalies of FII, FIII, and FIV felsic volcanic rocks are in-terpreted to indicate equilibration with a plagioclase residualphase (Fig. 1). Also significant is that at these high tempera-tures a rhyolitic melt has a less polymerized structure and iscapable of accommodating greater amounts of HFSE, a diag-nostic characteristic of FIII and FII felsic volcanic rocks. Fur-thermore, an FIII or FII melt would also be close to H2O sat-

uration, which, accompanied by the higher temperatures, would result in a lower viscosity and facilitate melt separationfrom restite and melt ascent (e.g., Huppert and Sparks,1988). The petrogenesis of FIV magmas is less well con-strained, as fewer analyses are available to characterize theirgeochemistry, but the lower overall abundances of REE andHFSE suggest formation under P-T conditions similar to theFIII magmas and derivation from a more depleted source. Infact, there should be a continuum in the trace element geo-chemistry of the derivative rhyolitic to dacitic magmas, asmelts are generated and extracted across a range of depths

and source compositions. This is illustrated in Figure 2, where the data for felsic rocks from various VMS depositsoverlap the fields for FI, FII, FIII, and FIV felsic volcanicrocks that are derived from data in this study and from Lesheret al. (1986). The compositions within a field vary not only

with differences in the P-T conditions of melting, but also with differences in the compositions of the source. In the

model proposed here, no variations in tectonic setting or ge-ologic age are required, only differences in the depths of melting and compositions of the crust owing to differences inthe degree of rifting and crustal evolution. The geochemistry of FI, FII, and FIII-IV felsic volcanic rocks can be explainedby variations in the mineralogy and composition of the crust,mode and P-T conditions of melting, and, to a lesser extent,by subsequent fractionation processes (Fig. 1).

An independent check on the validity of the temperatureand pressure conditions using the mineralogy, mineral chem-istry, and petrology of FI, FII, and FIII-IV rhyolites and


0361-0128/98/000/000-00 $6.00 1007

0 50 100 150 200

1

10

100

Archean

FIIIb

FI

FIIIa FIV

FII

Ybcn

( L a / Y b )

c n

Mesozoic -Cenozoic

FIIIb

FI

FIIIa

FIV

FII

( L a / Y b )

c n

Ybcn

0 50 100 150 200

1

10

100

Proterozoic

FIIIb

FI

FIIIa

FIV

FII

0 50 100 150 200

1

10

100

( L a / Y b )

c n

Ybcn

0 50 100 150 200

1

10

100

( L a / Y b )

c n

Ybcn

Paleozoic

FIIIb

FI

FIIIa

FIV

FII

FIG. 2. Chondrite-normalized plots, showing the shift with age from predominantly FIII to FII felsic volcanic rocks asso-

ciated with VMS deposits, except for FIII rhyolites hosting younger, larger tonnage deposits (e.g. Kidd Creek, Neves Corvo,United Verde, Eskay Creek). Normalizing factors from Nakamura (1974). Mesozoic deposits: s = Hokuroko district (Dudaset al., 1983), u = Eskay Creek (Barrett and Sherlock, 1996a), ✚ = Seneca (McKinley et al., 1996), e = Murgul (Schneider etal., 1988), ∇ = Shasta district (Bence and Taylor, 1985; LaPierre et al., 1985). Paleozoic deposits: s= Iberian pyrite belt (Mit- javila et al., 1997), q = Neves Corvo (Grimes et al., 1998), v= Thalanga (Stolz, 1995), = Kutcho Creek (Barrett et al., 1996),n = Benambra (Stolz et al., 1997), x = Myra Falls (Barrett and Sherlock, 1996b; Robinson et al., 1996), ✲ = Brunswick 12—inverted triangle right half filled (Lentz and Goodfellow, 1992), x= Heath Steele (Lentz and Wilson, 1997), u = Chester (vanStaal et al., 1995), = Tulsequah Chief (Sebert and Barrett, 1996). Proterozoic deposits: u = Garpenburg (Allen et al., 1996;Kumpulainen et al., 1996), s= Flin Flon (Syme, 1998), ✚ = Deri (Tiwary and Deb, 1997), e = Boliden (Weihed et al., 1996),n = Stirling (Dostal et al., 1992), x = United Verde (Gustin, 1988, 1990). Precambrian deposits: e = Kidd Creek (Campbell etal., 1984; Muirhead and Hutchinson, 1999; Hart, 2001), s = Manitouwadge (Geco, Willroy, Willecho, Big Nama; Schandl etal., 1995),∇ = Sturgeon Lake (Campbell et al., 1984; Hart, 2001), = Selbaie (Barrie and Krogh, 1996), ✳ = Sulphur Spring(Vearncombe and Kerrich, 1999), q = Scuddles (Whitford and Ashley, 1992), x= South Bay (Thurston and Fryer, 1983), u =Noranda (Corbet: Barrett et al., 1993; Aldermac: Barrett et al., 1991; Mobrun: Barrett et al., 1992; Horne and Quemont:MacLean and Hoy, 1991), x = Kamiskotia (Hart, 1984; Barrie and Pattison, 1999).

f

f



dacites (e.g., Streck and Grunder, 1997) is hampered by themetamorphism and alteration associated with VMS deposits.However, the zircon geothermometer used by Barrie (1995),

which is based on the zircon solubility model of Watson andHarrison (1983), provides a means to determine the temper-ature of rhyolitic flows that are saturated in zircon. Using thisgeothermometer, Barrie (1995) calculated that the FIIIb rhy-

olites at Kidd Creek formed at temperatures of 840º to 940ºC,consistent with temperatures on the edge of the field pro-posed for FIII magma production in Figure 1. Watson andHarrison (1983) raised the possibility that a source containing>100 ppm Zr might be saturated in zircon and retain residualzircon at all degrees of partial melting. However, residual zir-con may not be a present if the melt is close to H 2O satura-tion, as inferred for FIII rhyolites. Watson and Harrison(1983) noted that the ability of a melt to dissolve zircon in-creases with increasing H2O concentrations up to a limit of about 2 percent H2O, at which point the solubility of zircon isindependent of the H2O concentration. In addition, recentresearch suggests that the high F and Cl contents of some rift-related magmas may increase the solubility of Zr and theHFSE observed in the high-temperature (815º–1,150ºC) FIIfelsic volcanic rocks of the Parana basin (Kirstein et al., 2001).Nevertheless, it appears that FII and FIII rhyolites formedand erupted at high temperatures and that they contained sig-nificant amounts of volatiles.

The high eruption temperature and volatile content inter-preted for FIII rhyolites are also reflected in their flow mor-phology and textures. FIII rhyolites at Noranda form low re-lief, broad, extensive lobe-hyaloclastite plateaus that couldonly be constructed by a relatively low viscosity magma (Gib-son, 1990; Gibson et al., 1999). These flows are often amyg-daloidal and locally pumiceous along lobe margins, indicatinga relatively high volatile content. Moreover, FIII rhyolites at

Noranda and Kidd Creek are characterized by low phenocrystor microphenocryst contents and have spherulitic ground-masses, textures that are indicative of eruption at or above liq-uidus temperatures, which is consistent with their inferredhigher temperature of formation (Gibson, 1990).

A potential concern is the physical extraction of magma at very low degrees of partial melting. However, Wolf and Rapp(1994) and Sawyer (2001) illustrated that melt interconnec-tivity may be achieved and that magma may be extractedunder the high-temperature conditions and low degrees of melting proposed for the formation of FII and FIII-IV felsic

volcanic rocks in Figure 1 (Beard and Lofgren, 1991; Wolf and Wyllie, 1995; Bea, 1996).

Thus, the relatively constant compositions of FII and FIII-IV felsic volcanic rocks within individual areas, the high erup-tive temperatures (at or above liguidous), the bimodality of

VMS-hosting volcanic successions, and the probable lack of alarge magma chamber in the region of melting (e.g., Petfordet al., 2000) indicate that fractional crystallization within sub-

volcanic intrusions or in the source could not have generatedor significantly modified the compositions of FII-FIII-IV magmas. Rather, FII and FIII-IV felsic volcanic rocks andtheir associated subvolcanic intrusions are most likely theproducts of similar partial melting processes operating underthe same, relatively low pressure and high temperature con-ditions. Therefore, the questions are not if FII and FIII-IV

felsic volcanic rocks can form through high-temperature par-tial melting processes but where the high heat-flow environ-ment develops and whether there is a connection betweenthis environment and VMS metallogenesis.

The VMS-Felsic Volcanic Rock-SubvolcanicIntrusion Petrogenetic Connection

A large number of factors, including source(s) of metals andsulfur, water depth (phase separation), volcanic environment(flows versus volcaniclastic rocks), depositional processes, andsynvolcanic structures (block faulting and cross-strata hy-drothermal permeability) influence the formation and preser-

vation of VMS deposits. Importantly, there must be a heatsource that is large enough, hot enough, and at an appropri-ate level within the crust (depending on mass, temperature,and permeability) to generate and sustain a long-lived, high-temperature convective hydrothermal system, and there mustbe sufficient permeability to permit hydrothermal circulationand discharge (Cathles et al., 1997; Barrie et al., 1999). Thesefactors are common to almost all rift environments, explain-ing why VMS deposits occur in such a wide variety of rift-related extensional, subaqueous volcanic environments, in-cluding oceanic, back-arc, fore-arc, mature arc, continentalmargin, and continental rifts (e.g., Sillitoe, 1982; Lentz, 1998;Franklin et al., 1998).

In rift environments there is a thinning of the crust and as-sociated enhanced structural permeability that allows ascentof mafic to ultramafic magmas to high crustal levels, pondingand resulting partial melting of the crust to generate rhyoliticmelts. Figure 3 illustrates the approximate conditions for theformation of FI dacite to rhyolite deep in the crust (>30 kmand not restricted to rift environments), FII rhyolite to daciteat higher levels (30–10 km), and FIII rhyolite to high silicarhyolite high in the crust (<15 km). The trace element geo-

chemistry of the different felsic magmas is determined by theresidual mineral phases, wherein FI magmas equilibrate withgarnet-bearing residua, FII magmas equilibrate with amphi-bole-plagioclase–bearing residua, and FIII-IV magmas equi-librate with a plagioclase dominant, garnet- and amphibole-free residua (Fig. 1). A continuum in the trace elementgeochemistry of the derivative felsic magmas that are gener-ated and extracted across a range of depths and crust compo-sitions is expected. There is no tectonic or geologic age infer-ence in this model, only differences in the thickness andcomposition of the crust due to differences in the degree of rifting, crustal evolution, and depth of partial melting.

The conceptual model in Figure 3 also highlights two fun-

damental features critical to the formation of VMS deposits.First, there must be a heat source of sufficient mass andtemperature and at an appropriate position in the crust(Cathles et al., 1997). The heat required for partial meltingprocesses responsible for the generation of some FII and allFIII-FIV felsic volcanic rocks is a result of ultramafic to maficintrusions emplaced at shallow levels in the crust and withinthe zone of brittle fracture permeability (<10–15 km; e.g.,Sibson, 2002). Fracture permeability governs the depth of hydrothermal fluid circulation, and fracturing is estimated tooccur at depths of up to 12 km in cratonic environments and8 km in midocean ridge settings (Nur and Wader, 1990; Tay-lor, 1990; McClain et al., 1993; Barrie et al., 1999). Second,

1008 SCIENTIFIC COMMUNICATIONS

0361-0128/98/000/000-00 $6.00 1008



the extensional environment of rifts allows ascending magmasto follow preexisting pathways into upper crustal magmachambers (subvolcanic intrusions), which results in enhancedheat flow and permeability in long-lived, reactivated synvol-canic structures An important consequence of this model isthe possibility of fluid flow through deeper sections of thecrust, thus increasing the volume of rock (and metals) avail-able to leaching via convecting high-temperature hydrother-mal fluids. The model also allows for a direct contribution of metals from a magmatic hydrothermal fluid exsolved frommagmas emplaced at various levels within the rift or frommagmas derived from, or causing, partial melting (e.g.,

Tomkins and Mavrogenes, 2003). In fact there are many linesof evidence (isotopic, trace metal signatures, metal balance,fluid inclusions, and metal-rich fluids in melt inclusions) thatsuggest a magmatic component, especially for the formationof large high-grade VMS deposits, is required (Stanton, 1991;de Ronde, 1995; Yang and Scott, 1996, 2002; Hannington etal., 1999a). Rifting is also characterized by block faulting thatmay aid in the preservation of VMS deposits once formed.

As discussed above and shown in Figures 1 and 3, it is alsopossible to form FII felsic volcanic rocks at depths between15 and 30 km or below the depth of fracture permeability, ex-plaining, in part, why some FII felsic volcanic rocks are notassociated with VMS deposits. At this time, we are not able to

geochemically distinguish between FII felsic volcanic rocksgenerated at depths of 15 to 30 km and those that formed atdepths of 10 to 15 km, which are more likely to be associated

with VMS deposits. This is particularly important because FIIfelsic volcanic rocks are more abundant than FIII and FIV rhyolites. However, FII rhyolites associated with VMS de-posits occur in bimodal volcanic successions (e.g., Bathurst:Lentz and Goodfellow, 1992; Rio Tinto: Mitjavila et al., 1997;Sturgeon Lake: Hart, 2001), whereas barren FII rhyolites aretypically associated with continuous, differentiated volcanicsuccessions ( e.g., Buchanan et al., 2002; Moghazi, 2003). Thelow proportion to near absence of basalt in some VMS-host-

ing FII (e.g., Bathurst: Lentz and Goodfellow, 1992) and FIII(e.g., Neves Corvo: Grimes et al., 1998) felsic bimodal vol-canic successions indicates that basalts derived from mafic orultramafic magmas responsible for partial melting may notconstitute a significant proportion of the erupted material.

Our compilation indicates that FIII high silica rhyolites arenot restricted to the Archean and that they are associated withmany of the large-tonnage and higher grade VMS deposits(e.g., Kidd Creek, Neves Corvo, United Verde). However, thecompositions of felsic volcanic rocks associated with VMS de-posits display a general decrease in the abundances of com-patible and moderately incompatible trace elements (e.g., Y,HREE), with only a slight increase in the ratios of highly to


0361-0128/98/000/000-00 $6.00 1009

C r u s t a l D e p t h

( k m

)

40 -

- 0.0

60 -

- 2.0

P r e s s u r e

( G p a )

(Not to scale)

Ductile Limit of fracture permeability

0 -

20 -

- 1.0

Mineralization Potential Low High

FII

Barrie et al. (1993)

Barrett and MacLean (1994)

Lesher et al. (1986)

IV III I

Calc-Alkaline

FIIIb

Transitional Tholeiitic

II

FIIIaFI

Stable Minerals Garnet + pyroxene+/- amphibole

Plagioclase + quartz+/- pyroxene

Amphibole +plagioclase +/-

10 -La

Melt

LuSource

100 -

10 -La Lu

100 -

10 -La Lu

100 -TypicalREEPatterns

Brittle

Gar net out

A m p h i b o l e i n

Crust

Mantle

VMS

Garne t ou t


C r u s t

M a n t l e

Crus t

Man tle

Lim i t o f f r a c t u re p e rm e a b ili ty

VMS

G a r ne t o u t


-

-

-

FIG. 3. Conceptual petrogenetic model for the formation of FII and FIII-IV felsic volcanic rocks by partial melting atprogressively shallower crustal depths in a rift environment. Combined high heat flow and an extensional-rift environmentallow low-pressure, higher temperature crustal melting within the zone of brittle fracture permeability and promote convective seawater fluid flow. The complex arrangement of magma chambers depicted for FII felsic volcanic centers corre-sponds to the fact that FII felsic volcanic rocks forming below the maximum depth of convective fluid flow are barren, whereas those forming above this depth may be mineralized.



moderately incompatible elements (e.g., La/Yb, Zr/Y, Ti/Zr)over geologic time, such that Archean VMS deposits morecommonly occur in FIII rhyolites and Phanerozoic VMS de-posits more commonly occur in FII felsic volcanic rocks (Fig.2). This shift to dominantly FII compositions over time hasbeen noted by other researchers and may be the result of anumber of factors including a general thickening of the crust,

resulting in deeper and therefore higher pressure melting, achange to more felsic crustal compositions, and, perhaps, dif-ferent tectonic processes (e.g., Lesher et al., 1986; Lentz,1998; Piercey et al., 2003). However, there are fewer studiedexamples of felsic volcanic rocks associated with younger VMSdeposits, so more data are required to better define this trend.Regardless, the occurrence of FIII rhyolites in some Phanero-zoic VMS-hosting volcanic successions indicates that the ig-neous processes responsible for the formation of FII and FIIIfelsic volcanic rocks have not changed appreciably over time.

A logical extension of the generation of FIII and FII felsic volcanic rocks by partial melting, as presented in Figures 1 and3, is that felsic volcanic rocks of these compositions should beobserved in subaerial rift environments, if the physical condi-tions for melting are appropriate. In this regard, it is significantto note that FIII rhyolites occur in the Columbia River system(Snake River rhyolites: Streck and Grunder, 1997), the RioGrande rift (Bandelier tuff: Valles caldera; Self and Sykes,1996), the Mid-Atlantic Ridge in Iceland (e.g., Sigurdsson andSparks, 1981), the Midcontinent rift (Lightfoot et al., 1999),and the Parana basin (Garland et al., 1995). Although VMS de-posits do not form in subaerial rifts, this environment is favor-able for the formation of epithermal precious metal deposits.For example, the Valles caldera, localized along the RioGrande rift, hosts an active, intracaldera incipient epithermal,precious metal hydrothermal system. This suggests a broaderrelationship between the two types of mineralization as pro-

posed by other authors (e.g., Hannington et al., 1999b).

Summary and Conclusions

The geochemistry of FII-FIII-FIV felsic volcanic rocks isinfluenced by the mineralogy and composition of the source,the P and T of the primary melting process, and, to a lesserextent, any subsequent fractionation processes. The high-temperature and low-pressure conditions required to gener-ate FII, FIII, and FIV felsic volcanic rocks and their comag-matic subvolcanic intrusions characterize rift environmentsthat have anomalous, high heat flow due to upwelling of hotmantle, a thinner crust, and ascent of magma through thecrust. Deep structures, produced during rifting, channel the

flow of successive batches of ascending magma from the man-tle and/or lower crust to surface or into high-level subvolcaniccomposite intrusions that may occur at various crustal levels.A byproduct of rifting is enhanced fracture permeability that,coupled with high temperatures, results in an increase in theefficiency of heat transfer through the development of alarge, high-temperature convective hydrothermal system es-sential for the formation of VMS deposits.

There are a number of applications of these findings in ex-ploration for VMS deposits:

1. VMS deposits are generally hosted by bimodal, mafic-felsic volcanic successions with FII, FIII, and FIV felsic

volcanic rocks that display minimal geochemical variation andnot by FI felsic volcanic rocks.

2. There is a preferential association of VMS deposits withFII, FIII, and FIV felsic volcanic rocks that occurs in a vari-ety of rift-related extensional, subaqueous volcanic environ-ments, including oceanic, back-arc, fore-arc, mature arc, con-tinental margin, and continental paleorifts.

3. FIII rhyolites, regardless of age, tend to host many of the larger tonnage and higher grade VMS deposits (e.g., KiddCreek, Neves Corvo, United Verde, Eskay Creek) and may represent preferred exploration targets.

4. As noted by Lesher et al. (1986), some VMS-hostingsuccessions (e.g., Sturgeon Lake, Confederation Lake) con-tain FI felsic volcanic rocks as well as FII or FIII felsic vol-canic rocks. Therefore, careful, systematic sampling is criticalin utilizing trace element geochemistry as a tool in mineral ex-ploration (Lesher et al., 1986).

5. Because felsic volcanic rocks of favorable geochemistry may form in environments in which VMS deposits may notform or be preserved, lithogeochemistry cannot be used inisolation as an exploration tool. Other factors, including an as-sessment of tectonic setting, volcanic environment (including

water depth), and the presence of a subvolcanic intrusion alsomust be considered. Recognition of FII, FIII, or FIV rhyo-lites does not target VMS deposits, only environments favor-able for their formation.

Acknowledgments

We are very grateful to two Economic Geology reviewersPaul Spry and Wallace MacLean for their critical reviews of the original manuscript, which resulted in many improve-ments. We are especially grateful to Ian Campbell for his crit-ical and insightful comments on several versions of the paper,

which considerably improved the presentation and discus-

sion. We have also benefited from discussions with many col-leagues, particularly Tucker Barrie, Larry Cathles, PaulDavis, Michael Easton, Jim Franklin, Al Galley, Tony Green,Mark Hannington, Jack Parker, Steve Piercey, and PhilThurston. This research has been supported by grants fromthe Natural Sciences and Engineering Research Council of Canada (HLG and CML) and has been done with the supportof the Ontario Ministry of Northern Development and Mines(Ontario Geological Survey).

December 6, 2002; March 24, 2004

REFERENCES

Allen, R.L., Lundstrom, I., Ripa, M., Simeonov, A., and Christofferson, H.,1997, Facies analysis of a 1.9 Ga, continental margin, back-arc, felsic caldera

province with diverse Zn-Pb-Ag-(Cu-Au) sulfide and Fe oxide deposits,Bergslagen region, Sweden: ECONOMIC GEOLOGY, v. 92, p. 979–1008.

Barrett, T.J., and MacLean, W.H., 1994, Chemostratigraphy and hydrother-mal alteration in exploration for VHMS deposits in greenstone belts and younger volcanic rocks: Geological Association of Canada Short CourseNotes 11, p. 433–467.

——1999, Volcanic sequences, lithogeochemistry, and hydrothermal alter-ation in some bimodal volcanic-associated massive sulfide systems: Reviewsin Economic Geology, v. 8, p. 101–113.

Barrett, T.J., and Sherlock, R.L., 1996a, Geology, lithogeochemistry and vol-canic setting of the Eskay Creek Au-Ag-Cu-Zn deposit, northwesternBritish Columbia: Exploration and Mining Geology, v. 5, p. 339–368.

——1996b, Volcanic stratigraphy, lithogeochemistry, and seafloor setting of the H-W massive sulfide deposit, Myra Falls, Vancouver Island, British Co-lumbia: Exploration and Mining Geology, v. 5, p. 421–458.


0361-0128/98/000/000-00 $6.00 1010



Kumpulainen, R.A., Mansfeld, J., Sundblad, K., Neymark, L., and Bergman,T., 1996, Stratigraphy, age, and Sm-Nd isotope systematics of the country rocks to Zn-Pb sulfide deposits, Ammeberg district, Sweden: ECONOMIC

GEOLOGY, v. 91, p. 1009–1021.LaPierre, H., Cabanis, B., Coulon, C., Brouxel, M., and Albarede, F., 1985,

Geodynamic setting of early Devonian Kuroko-type sulfide deposits in theeastern Klamath Mountains (northern California) inferred by the petrolog-ical and geochemical characteristics of the associated island-arc volcanicrocks: ECONOMIC GEOLOGY, v. 80, p. 2100–2113.

Lentz, D.R., 1998, Petrogenetic evolution of felsic volcanic sequences asso-ciated with Phanerozoic volcanic-hosted massive sulfide systems: The roleof extensional geodynamics: Ore Geology Reviews, v. 12, p. 289–327.

Lentz, D., and Goodfellow, W., 1992, Re-evaluation of the petrochemistry of felsic volcanic and volcaniclastic rocks near the Brunswick no. 6 and 12massive sulfide deposits, Bathurst mining camp, New Brunswick: Geologi-cal Survey of Canada Paper 92-1E, p. 343–350.

Lentz, D., and Wilson, R.A., 1997, Chemostratigraphic analysis of the vol-canic and sedimentary rocks in the Heath Steele B-B5 zone area, Bathurstcamp, New Brunswick: Stratigraphic and structural implications: Geologi-cal Survey of Canada Current Research 1997-D, p. 21–33.

Lesher, C.M., Goodwin, A.M., Campbell, I.H., and Gorton, M.P., 1986,Trace-element geochemistry of ore-associated and barren, felsic metavol-canic rocks in the Superior province, Canada: Canadian Journal of EarthSciences, v. 23, p. 222–237.

Lightfoot, P.C., Sage, R.P., Doherty, W., Naldrett, A.J., and Sutcliffe, R.H.,

1999, Mineral potential of Proterozoic Keweenawan intrusions: Implica-tions of major and trace element geochemical data from bimodal mafic andfelsic volcanic sequences of Mamainse Point and the Black Bay Peninsula,Ontario: Ontario Geological Survey Open File Report 5998, 91 p.

MacLean, W.H., and Hoy, L.D., 1991, Geochemistry of hydrothermally al-tered rocks at the Horne mine, Noranda, Quebec: ECONOMIC GEOLOGY, v.86, p. 506–528.

McClain, J.S., Begnaud, M.L., Wright, M.A., Fondrk, J., and von Damm,G.K., 1993, Seismicity and tremor in a submarine hydrothermal field: Thenorthern Juan de Fuca Ridge: Geophysical Research Letters, v. 20, p.1883–1886.

McKinley, S.D., Barrett, T.J., and Thompson, J.F.H., 1996, Volcanic stratig-raphy and lithogeochemistry at the Seneca Zn-Cu-Pb prospect, southwest-ern British Columbia: Exploration and Mining Geology, v. 5, p. 399–420.

Mitjavila, J.M., Marti, J., and Soriano, C., 1997, Magmatic evolution and tec-tonic setting of the Iberian pyrite belt volcanism: Journal of Petrology, v. 38,

p. 728–755.Moghazi, A.M., 2003, Geochemistry and petrogenesis of a high-K calc-alka-line Dokhan volcanic suite, South Safaga area, Egypt: The role of late Neo-proterozoic crustal extension: Precambrian Research, v. 125, p. 161–178.

Morton, R.L., Walker, J.S., Hudak, G.J., and Franklin, J.M., 1991, The early development of an Archean submarine complex with emphasis on the Mat-tabi ash-flow tuff and its relationship to the Mattabi massive sulfide deposit:ECONOMIC GEOLOGY, v. 86, p. 1002–1011.

Muirhead, E.M., and Hutchinson, R.W., 1999, Mass change profiles in thefootwall of the Kidd Creek orebody: ECONOMIC GEOLOGY MONOGRAPH 10,p. 297–308.

Nakamura, N, 1974, Determination of REE, Ba, Fe, Mg, Na, and K in car-bonaceous and ordinary chondrites: Geochimica et Cosmochimica Acta, v.38, p. 757–775.

Nur, A.M., and Wader, L., 1990. Time-dependant hydraulics of the earth’scrust, in Geophysics Study Committee, eds., The role of fluids in crustalprocesses: Studies in geophysics: National Research Council: Washington,

D.C., National Academy Press, p. 113–127.Oskarsson, N., Sigvaldason, G.G., and Steinthorsson, S., 1982, A dynamic

model of rift zone petrogenesis and the regional petrology of Iceland: Jour-nal of Geology, v. 23, p. 28–74.

Paradis, S., Taylor, B.E., Watkinson, D.H., and Jonasson, I.R., 1993, Oxygenisotope zonation in the northern Noranda district, Quebec: Evidence forhydrothermal fluid flow: ECONOMIC GEOLOGY, v. 88, p. 1512–1525.

Parry, S., and Hutchinson, R.W., 1981, Origin of a complex alteration assem-blage, Four Corners Cu-Zn prospect, Quebec, Canada: ECONOMIC GEOL-OGY, v. 76, p. 1186–1201.

Petford, N., Cruden, A.R., McCaffrey, K.J.W., and Vigneresse, J.-L., 2000,Granite magma formation, transport and emplacement in the Earth’s crust:Nature, v. 408, p. 669–673.

Piercey, S.J., Mortensen, J.K., and Creaser, R.A., 2003, Neodynium isotopegeochemistry of felsic volcanic and intrusive rocks from the Yukon-Tanana

terrane in the Finlayson Lake region, Yukon, Canada: Canadian Journal of Earth Sciences, v. 40, p. 77–97.

Prior, G.J., Gibson, H.L., Watkinson, D.H., Cook, R.E., and Hannington,M.D., 1999a, Rare earth and high field strength element geochemistry of the Kidd Creek rhyolites, Abitibi greenstone belt, Canada: Evidence forArchean felsic volcanism and massive sulfide ore formation in an Icelandic-style rift environment: ECONOMIC GEOLOGY MONOGRAPH 10, p. 457–484.

Prior, G.J., Gibson, H.L., Watkinson, D.H., Cousens, B.L., Cook, R.E., andHannington, M.D., 1999b, Sm-Nd isotope study of rhyolites from the KiddCreek mine area, Abitibi subprovince, Canada: ECONOMIC GEOLOGY

MONOGRAPH 10, p. 485–496.Rapp, R.P., 1995, Amphibole-out phase boundary in partially melted

metabasalt, its control over liquid fraction and composition, and sourcepermeability: Journal of Geophysical Research, v. 100, p. 15601–15610.

Robinson, M., Godwin, C.I., and Stanley, C.R., 1996, Geology, lithogeo-chemistry, and alteration of the Battle volcanogenic massive sulfide zone,Buttle Lake mining camp, Vancouver Island, British Columbia: ECONOMIC

GEOLOGY, v. 91, p. 527–548.Sawyer, E.W., 2001, Melt segregation in the continental crust: Distribution

and movement of melt in anatectic rocks: Journal of Metamorphic Geology, v. 19, p. 291–309.

Schandl, E.S., Gorton, M.P., and Wasteneys, H.A., 1995, Rare earth elementgeochemistry of the metamorphosed volcanogenic massive sulfide depositsof the Manitouwadge mining camp, Superior province, Canada: A potentialexploration tool?: ECONOMIC GEOLOGY, v. 90, p. 1217–1236.

Schneider, H.-J., Ozgur, N., and Palacios, C.M., 1988, Relationship betweenalteration, rare earth element distribution, and mineralization of the Mur-gul copper deposit, northwestern Turkey: ECONOMIC GEOLOGY, v. 83, p.1238–1246.

Sebert, C., and Barrett, T.J., 1996, Stratigraphy, alteration, and mineraliza-tion at the Tulsequah Chief massive sulfide deposit, northwestern BritishColumbia: Exploration and Mining Geology, v. 5, p. 281–308.

Self, S., and Sykes, M.L., 1996, Part I: Field guide to the Bandelier tuff and Valles caldera: New Mexico Bureau of Mines and Mineral Resources Bul-letin 134, p. 7–55.

Sigurdsson, H., 1977, Generation of Icelandic rhyolites by melting of pla-giogranites in the oceanic layer: Nature, v. 269, p. 25–28.

Sigurdsson H., and Sparks, R.S.J., 1981, Petrology of rhyolite and mixedmagma ejecta from the 1875 eruption of Askja, Iceland: Journal of Petrol-ogy, v. 22, p. 41–84.

Sillitoe, R.H., 1982, Extensional habitats of rhyolite-hosted massive sulfide

deposits: Geology, v. 109, p. 403–407.Spulber, S.D., and Rutherford, M.J., 1983, The origin of rhyolite and pla-giogranite in oceanic crust: An experimental study: Journal of Petrology, v.24, p. 1–25.

Stanton, R.L., 1991, Understanding volcanogenic massive sulfide deposits—past, present, and future: ECONOMIC GEOLOGY MONOGRAPH 8, p. 82–95.

Stolz, A.J., 1995, Geochemistry of the Mount Windsor volcanics: Implica-tions for the tectonic setting of Cambro-Ordovician volcanic-hosted mas-sive sulfide mineralization in northeastern Australia: ECONOMIC GEOLOGY, v. 90, p. 1080–1097.

Stolz, A.J., Davies, G.R., and Allen, R.L., 1997, The importance of differenttypes of magmatism in VHMS mineralization: Evidence from the geo-chemistry of the host volcanic rocks to the Benambra massive sulfide de-posits, Victoria, Australia: Mineralogy and Petrology, v. 59, p. 251–286.

Streck, M.J., and Grunder, A.L., 1997, Compositional gradients and gaps inhigh-silica rhyolites of the Rattlesnake tuff, Oregon: Journal of Petrology, v.38, p. 133–163.

Syme, E.C., 1998, Ore-associated and barren rhyolites in the Central FlinFlon belt: Case study of the Flin Flon mine sequence: Manitoba Energy and Mines Geological Services Open File Report OF98-9, 26 p.

Taylor, H.P., Jr., 1990, Oxygen and hydrogen isotope constraints on the deepcirculation of surface waters into zones of hydrothermal metamorphismand melting, in Geophysics Study Committee, eds., The role of fluids incrustal processes: Studies in Geophysics, National Research Council, Washington, D.C., National Academy Press, p. 72–95.

Thurston, P.C., 1981, Economic evaluation of Archean felsic volcanic rocksusing REE geochemistry: Geological Society of Australia Special Publica-tion, v. 7, p. 439–450.

Thurston, P.C., and Fryer, B.J., 1983, The geochemistry of repetitive cyclical volcanism from basalt through rhyolite in the Uchi-Confederation green-stone belt, Canada: Contributions to Mineralogy and Petrology, v. 83, p.204–226.


0361-0128/98/000/000-00 $6.00 1012



Tiwary, A., and Deb, M., 1997, Geochemistry of hydrothermal alteration atthe Deri massive sulfide deposit, Sirohi district, Rajasthan, NW India: Jour-nal of Geochemical Exploration, v. 59, p. 99–121.

Tomkins, A.G., and Mavrogenes, J.A., 2003, Generation of metal-rich felsicmagmas during crustal anatexis: Geology, v. 31, p. 765–768.

van Staal, C.R., Fyffe, L.R., Langton, J.P., and McCutcheon, S.R., 1995, TheOrdovician Tetagouche Group, Bathurst camp, northern New Brunswick,Canada: History, tectonic setting and distribution of massive-sulfide de-posits: Exploration and Mining Geology, v. 4, p. 153–173.

Vearncombe, S.E., and Kerrich, R., 1999, Geochemistry and geodynamic set-ting of volcanic and plutonic rocks associated with Early Archean vol-canogenic massive sulfide mineralization, Pilbara craton: Precambrian Re-search, v. 98, p. 243–270.

Watson, E.B., and Harrison, T.M., 1983, Zircon saturation revisited: Tem-perature and composition effects in a variety of crustal types: Earth andPlanetary Science Letters, v. 64, p. 295–304.

Weihed, J.B., Bergstrom, U., Billstrom, K., and Weihed, P., 1996, Geology,tectonic setting, and origin of the Paleoproterozoic Boliden Au-Cu-As de-posit, Skellefte district, northern Sweden: ECONOMIC GEOLOGY, v. 91, p.1073–1097.

Whitford, D.J., and Ashley, P.M., 1992, The Scuddles volcanic-hosted mas-sive sulfide deposits, Western Australia: Geochemistry of the host rocks andevaluation of lithogeochemistry for exploration: ECONOMIC GEOLOGY, v. 87,p. 873–888.

Wolf, M.B., and Wyllie, P.J., 1995, Liquid segregation parameters from am-phibolite dehydration melting experiments: Journal of Geophysical Re-search, v. 100, p. 15611–15621.

Wyllie, P.J., and Wolf, M.B., 1993, Amphibolite dehydration-melting: Sortingout the solidus: Geological Society of London Special Publication 72, p.405–416.

Yang, K., and Scott, S.D., 1996, Possible contribution of a metal-rich mag-matic fluid to a seafloor hydrothermal system: Nature, v. 383, p. 420–423.

——2002, Magmatic degassing of volatiles and ore metals into a hydrother-mal system on the modern sea floor of the eastern Manus back-arc basin, western Pacific: ECONOMIC GEOLOGY, v. 97, p. 1079–1100.


0361-0128/98/000/000-00 $6.00 1013



0361-0128/98/000/000-00 $6.00 1014

Hartetal2004_FelsicVolcanicVMS_

Documents

Transcript of Hartetal2004_FelsicVolcanicVMS_