Volcanic-associated massive sulfide deposits: Processes and examples in modern and ancient settings

Abstract

Volcanic-associated massive sulfide deposits (VMS) are predominantly stratiform accumulations of sulfide minerals that precipitate from hydrothermal fluids at or below the seafloor, in a wide range of ancient and modern geological settings (Figs. 1, 2). They occur within volcano-sedimentary stratigraphic successions, and are commonly coeval and coincident with volcanic rocks. As a class, they represent a significant source of the world's Cu, Zn, Pb, Au, and Ag ores, with Co, Sn, Ba, S, Se, Mn, Cd, In, Bi, Te, Ga and Ge as co- or by-products.

The understanding of ancient, land-based VMS deposits has been heavily influenced by the discovery and study of active, metal-precipitating hydrothermal vents on the seafloor. During the last three decades, excellent descriptions of seafloor sulfides and related vent fluids and hydrothermal plumes have provided modern analogs for the land-based VMS deposits (Rona, 1988; Rona and Scott, 1993; Hannington et al., 1995). Conversely, the geology and mineralogy of land-based deposits have provided insight into the plumbing systems and sulfide mineral paragenesis of sulfide deposits relevant to seafloor hydrothermal systems. This volume capitalizes on the complementary nature of ancient, land-based VMS deposits and active, metal-precipitating hydrothermal systems on the seafloor, much as the Reviews in Economic Geology Volume 2 (Berger and Bethke, editors, 1985) did with epithermal deposits and active, subaerial geothermal systems, and draws equally from land-based- and seafloor VMS research.

This volume attempts to provide a balanced view of VMS systems, with descriptions of the processes involved in VMS formation and of important examples representing a variety of VMS deposits and districts, in modern and ancient settings. It is not meant to be a comprehensive review; rather, it presents a spectrum of current ideas based on research since the benchmark paper of Franklin et al. (1981). The contributions are divided into two parts. In Part I, reviews of the most significant geological, physical and chemical processes involved in the formation of land-based and seafloor VMS deposits are presented. These include: the volcanology of subaqueous settings and the relationship between volcanology and VMS systems by Gibson et al.; structural aspects of magmatism and hydrothermal circulation in ocean floor and ophiolitic settings by Harper; the relationship between magma chemistry and hydrothermal venting, with emphasis on the thickened oceanic crust in the Galapagos area by Perfit et al., and more generally in bimodal volcanic settings by Barrett and MacLean; hydrothermal alteration of the oceanic crust by Alt; fluid-rock interactions in VMS systems as recorded by stable isotope systematics by Huston; the metal transport capabilities of hydrothermal fluids by Seyfried et al.; precious metal enrichment associations and processes in VMS systems by Hannington et al.; and heat and fluid flow in VMS systems by Barrie et al.

In Part II, descriptions of land-based and seafloor deposits or districts are given, within the context of the processes described in Part I. They are arranged in an order from primitive, mafic-dominant systems to evolved, felsic and sedimentary rock-dominant systems, similar to the order in the classification scheme described below. Deposits in mafic-dominant, ophiolitic settings are described by Galley and Koski. Many of these deposits are believed to represent supra-subduction zone tectonic settings, and they provide the closest comparison to mid-ocean ridge hydrothermal vent fields. A review of the giant Kidd Creek deposit of the Late Archean Abitibi subprovince by Barrie et al. provides a view of a bimodal-mafic, primitive arc VMS setting. The largest Besshi-type, or mafic-siliciclastic type VMS deposit in the world at Windy Craggy, British Columbia is described by Peter and Scott. In a broad sense, modern analogs for deposits like Windy Craggy are described by Goodfellow and Zierenberg in their paper on massive sulfide-forming hydrothermal systems in sediment-covered oceanic spreading centers. Felsic volcanic and sedimentary-influenced VMS systems are found in more mature volcanic or continental arc settings. These types are represented in this volume by the precious metal-rich Eskay Creek deposit, described by Roth et al., and the incredibly prolific Iberian Pyrite Belt of Portugal and Spain, described by Carvalho et al.

Full text

Volcanic-associated massive sulfide deposits (VMS) are
predominantly stratiform accumulations of sulfide miner-
als that precipitate from hydrothermal fluids at or below
the sea floor, in a wide range of ancient and modern geo-
logical settings (Figs. 1, 2). They occur within volcano-
sedimentary stratigraphic successions, and are commonly
coeval and coincident with volcanic rocks. As a class, they
represent a significant source of the world's Cu, Zn, Pb,
Au, and Ag ores, with Co, Sn, Ba, S, Se, Mn, Cd, In, Bi, Te,
Ga, and Ge as co- or by-products.
The understanding of ancient, land-based VMS deposits
has been heavily influenced by the discovery and study of
active, metal-precipitating hydrothermal vents on the sea
floor. During the last three decades, excellent descriptions
of sea-floor sulfides and related vent fluids and hydrother-
mal plumes have provided modern analogs for the land-
based VMS deposits (Rona, 1988; Rona and Scott, 1993;
Hannington et al., 1995). Conversely, the geology and
mineralogy of land-based deposits have provided insight
into the plumbing systems and sulfide mineral paragenesis
of sulfide deposits relevant to sea-floor hydrothermal sys-
tems. This volume capitalizes on the complementary na-
ture of ancient, land-based VMS deposits and active,
metal-precipitating hydrothermal systems on the sea floor,
much as the Reviews in Economic Geology Volume 2 (Berger
and Bethke, eds., 1985) did with epithermal deposits and
active, subaerial geothermal systems, and draws equally
from land-based and sea-floor VMS research.
This volume attempts to provide a balanced view of VMS
systems, with descriptions of the processes involved in
VMS formation and of important examples representing a
variety of VMS deposits and districts, in modern and an-
cient settings. It is not meant to be a comprehensive re-
view; rather, it presents a spectrum of current ideas based
on research since the benchmark paper of Franklin et al.
(1981). The contributions are divided into two parts. In
Part I, reviews of the most significant geological, physical,
and chemical processes involved in the formation of land-
based and sea-floor VMS deposits are presented. These in-
clude: the volcanology of subaqueous settings and the re-
lationship between volcanology and VMS systems by
Gibson et al. (1999); structural aspects of magmatism and
hydrothermal circulation in ocean floor and ophiolitic set-
tings by Harper (1999); the relationship between magma
chemistry and hydrothermal venting, with emphasis on
1
Chapter 1
Classification of Volcanic-Associated Massive Sulfide Deposits
Based on Host-Rock Composition
C. T. BARRIE* AND M. D. HANNINGTON
Geological Survey of Canada, 601 Booth St., Ottawa, Ontario, Canada K1A 0E8
ARCHEAN
AR/PTAR/MZ
PROTER.
PALEOZOIC
PROT/MZMESOZOIC
CENOZOIC
1
2
3
456
7
8
910 11
12
13-16
17-20
21
22
23
24
25
26
28
27
29
31
30
32
33
35
34
1 Windy Craggy, NW British Columbia: 297 MT, Mesozoic M-S;
2 N. Cordillera, Canada (including Eskay Creek): 100 MT, L. Paleozoic, M,
B-M, B-F;
3 Jerome, Arizona, USA: 30 MT, E. Proterozoic, B-M;
4 Slave Province, Canada: 50 MT, L. Archean, B-M, B-F;
5 Flin Flon-Snow Lake, Man. and Sask., Canada: 160 MT, E. Proterozoic, B-
M, M;
6 Ladysmith-Rhinelander, Wisconsin, USA: 80 MT, E. Proterozoic, B-M;
7 Abitibi-Superior, Canada (incl. Kidd Creek): 500 MT, L. Archean, B-M, B-
F.
8 Ducktown, Tennessee, USA: 180 MT, L. Proterozoic, M-S;
9 Bathurst, New Brunswick, Canada: 250 MT, E. Paleozoic, B-S;
10 Buchans-Victoria Lake, Newfoundland: 20 MT, E. Paleozoic, B-F;
11 Iberian Pyrite Belt, Portugal and Spain: 1000+ MT, E. Paleozoic, B-S;
12 Rouez, Bretagne, France: 100 MT, M. Proterozoic, M-S;
13 Trondheim, Norway: 40 MT, E. Paleozoic, B-M, M-S, B-F;
14 Skellefte (including. Boliden), Sweden: 75 MT E. Proterozoic, B-M, B-F;
15 Rana-Grong-Sulitjelma, Sweden, Sweden: 80 MT, E. Paleozoic, B-M, M-S;
16 Outokumpu-Pyhasalmi, Finland: 60 Mt, E. Proterozoic, B-M, M-S;
17 Central Urals, Russia: 100+ MT, E. Paleozoic, B-F, B-M;
18 Buribai-southern Urals, Russia and Kazikstan: 100+MT, E. Paleozoic, B-F,
B-M;
19 Turkey (including Murgul): 170 MT, Paleozoic and Mesozoic, B-M, B-F, M;
20 Zyryanowsk, Kazakstan: 500 MT, Paleozoic, B-F, B-M;
21 Troodos, Cyprus: 35 MT, Mesozoic, M;
22 Presika, Otjahasi, Matchless, S. Africa-Namibia: 140 MT, M. Proterozoic,
M-S;
23 Gai-Uchali, Kazakstan: 100+ MT, Paleozoic, B-F, B-M;
24 Yidan, China: ?, Mesozoic, B-F;
25 Qilian, China: ?, E. Paleozoic, B-M, B-F;
26 Kang Dian, China: 500 MT, L. Proterozoic, B-M;
27 Hongtouchan, China: ?, L. Archean, B-M;
28 Hokuroku, Japan: 100 MT, Miocene, B-F;
29 Besshi, Japan: 230 MT, L. Paleozoic and Mesozoic, M-S;
30 Big Stubby, Mons Cupri, Whim Creek, W. Australia: 15 MT, M.+L.
Archean, B-F;
31 Scuddles-Golden Grove, W. Australia: 45 MT, L. Archean, B-M;
32 Philippines: 60 MT, Cenozoic, B-F, M, B-S;
33 Mt. Morgan, Queensland, Australia: 50 MT, E. Paleozoic, B-M;
34 Benambra-Woodlawn, Victoria-NSW, Australia: 35 MT, E. Paleozoic, B-F;
35 Mt. Read, Tasmania, Australia: 150 MT, E. Paleozoic, B-F.
FIG. 1. Location of major VMS districts of the world, with total tonnages
in million tonnes (MT). M: Mafic; B-M: Bimodal Mafic; M-S: Mafic Silici-
clastic; B-F: Bimodal-Felsic; B-S: Bimodal-Siliciclastic.
*Alternate address: Barrie & Associates, 23 Euclid Avenue, Ottawa,
Ontario, Canada K1S 2W2.

the thickened oceanic crust in the Galapagos area by Per-
fit et al.(1999), and more generally in bimodal volcanic
settings by Barrett and MacLean (1999); hydrothermal al-
teration of the oceanic crust by Alt (1999); fluid-rock in-
teractions in VMS systems as recorded by stable isotope
systematics by Huston (1999); the metal transport capabil-
ities of hydrothermal fluids by Seyfried et al. (1999); pre-
cious metal enrichment associations and processes in VMS
systems by Hannington et al. (1999); and heat and fluid
flow in VMS systems by Barrie et al. (1999a).
In Part II, descriptions of land-based and sea-floor de-
posits or districts are given, within the context of the
processes described in Part I. They are arranged in an
order from primitive, mafic-dominant systems to evolved,
felsic and sedimentary rock-dominant systems, similar to
the order in the classification scheme described below.
Deposits in mafic-dominant, ophiolitic settings are de-
scribed by Galley and Koski (1999). Many of these de-
posits are believed to represent supra-subduction zone
tectonic settings, and they provide the closest compari-
son to mid-ocean ridge hydrothermal vent fields. A re-
view of the giant Kidd Creek deposit of the Late Archean
Abitibi subprovince by Barrie et al. (1999b) provides a
view of a bimodal-mafic, primitive arc VMS setting. The
largest Besshi-type, or mafic-siliciclastic type VMS deposit
in the world at Windy Craggy, British Columbia is de-
scribed by Peter and Scott (1999). In a broad sense, mod-
ern analogs for deposits like Windy Craggy are described
by Goodfellow and Zierenberg (1999) in their paper on
massive sulfide-forming hydrothermal systems in sedi-
ment-covered oceanic spreading centers. Felsic volcanic
and sedimentary-influenced VMS systems are found in
more mature volcanic or continental arc settings. These
types are represented in this volume by the precious
metal-rich Eskay Creek deposit, described by Roth et al.
(1999), and the incredibly prolific Iberian pyrite belt of
Portugal and Spain, described by Carvalho et al. (1999).
The Classification Scheme
For the purposes of this volume, a simple, five-fold clas-
sification of VMS deposits is proposed that encompasses
most of the known ancient and active VMS settings. The
classification draws from a comprehensive database se-
lected from Mosier et al. (1983), the Geological Survey of
Canada VMS database for Canada (Franklin, 1993), the
published literature, annual reports of major mining com-
panies, and our own files. The database includes 878 de-
posits, of which 811 have proper geological control and ac-
curate grade and tonnage information (mined, mineable,
and geological reserves, and drill indicated resources). The
deposits in the database represent ~60 percent of the total
subaerial continental areas; data from deposits in ex-Soviet
bloc countries, China, and Antarctica are not included.
The five-fold classification is based on host rock compo-
sition, with emphasis on the pre-alteration composition of
coeval, or nearly coeval (within 3–4 m.y.) volcanic host
rocks. Rocks up to ~3 km into the stratigraphic footwall, ~1
km into the stratigraphic hanging wall and up to 5 km
2BARRIE AND HANNINGTON
FIG. 2. Location of sea-floor hydrothermal vent sites and massive sulfide deposits. Modified after Hannington et al., 1995.
70°
60°
30°
0°
30°
60°
70°
70°
60°
30°
0°
30°
60°
70°
90° 120° 150° 180° 150° 120° 90° 60° 30° 0° 30° 60° 90°
90° 120° 150° 180° 150° 120° 90° 60° 30° 0° 30° 60° 90°

along strike are considered. Previous studies have used the
base metal content (e.g., Hutchinson, 1973; Solomon, 1976;
Franklin et al., 1981; Large, 1992), tectonic setting (e.g.,
Sawkins, 1976; Hutchinson, 1980), host rock textures
(Morton and Franklin, 1987; see Gibson et al., 1999), or
host rock lithology (e.g., volcanic, volcano-sedimentary,
sedimentary divisions: Sangster and Scott, 1976) as the
principal criteria for classifying VMS sulfide deposits.
Given that most of the metals in the majority of VMS de-
posits are derived from leaching of a footwall substrate
(Large, 1992) which is predominantly volcanic, and that
the composition of the volcanic substrate commonly re-
flects the gross tectonic setting, it is not surprising that
there is broad agreement and overlap among these classi-
fication schemes and with the one presented here. In this
classification, similar VMS deposits group together reason-
ably well regardless of their age, and many of the ambigu-
ities in classification based solely on metal content (e.g.,
“Cu-Zn” deposits may include felsic- and mafic-dominant
successions), tectonic setting (e.g., deposits in metamor-
phic terrane with unclear origins) or age (e.g., many
“Archean Cu-Zn” deposits similar to Phanerozoic deposits)
are avoided.
From the most primitive to the most evolved in a chem-
ical sense, the five host rock compositions considered are:
mafic, bimodal-mafic, mafic-siliciclastic, bimodal-felsic, and
bimodal-siliciclastic. The average grade and tonnage of de-
posits in these groups are given in Table 1, and the aver-
age grade and tonnage for each type divided by time pe-
riod is given in Table 2. These data are presented in a
variety of bar graphs and ternary plots in Figures 3–6.
Mafic Type
The mafic type is defined by two principal criteria: a pre-
dominantly (e.g., >75%) mafic host rock stratigraphic suc-
cession, and rare or absent (<1%) felsic volcanic rocks.
The host rocks commonly have minor (<10%) siliciclastic
or ultramafic rocks, or both. The mafic type encompasses
CLASSIFICATION OF VMS DEPOSITS BASED ON HOST-ROCK COMPOSITION 3
TABLE 1. Total and Average Grade and Tonnage for VMS Types, Excluding China and ex-Soviet Block Countries
TYPE Total Tonnage1Total Cu1Total Pb1Total Zn1Total Au1Total Ag1
nin billion tonnes in million tonnes in million tonnes in million tonnes in tonnes 102in tonnes 103
Mafic 62 0.18 3.7 0.04 1.3 2.31 2.6
Bimodal-mafic 284 1.45 24.3 2.0 44.3 12.91 38.2
Mafic-siliciclastic 113 1.24 16.2 0.6 9.7 4.03 9.2
Bimodal-felsic 255 1.29 7.1 13.2 54.2 14.18 120.0
Bimodal-siliciclastic 97 2.50 21.5 24.0 55.1 4.11 60.0
Total 811 6.66
(878)2(6.93)2
Average size Average Cu Average Pb Average Zn Average Au Average Ag
in million tonnes grade in wt % grade in wt % grade in wt % grade in g/t grade in g/t
Mafic 2.8 2.04 0.10 1.82 2.56 20.0
Bimodal-mafic 5.1 1.88 0.75 4.22 1.52 36.5
Mafic-siliciclastic 11.0 1.74 1.83 2.43 0.84 19.8
Bimodal-felsic 5.2 1.44 1.64 5.63 2.06 92.8
Bimodal-siliciclastic 23.7 1.10 1.84 4.16 1.13 84.4
Number Number Number Number Number
of deposits of deposits of deposits of deposits of deposits
>100 MMT 50–100 MMT 20–50 MMT 10–20 MMT 5–10 MMT
Mafic 0 0 3 1 7
Bimodal-mafic 1 6 9 16 20
Mafic-siliciclastic 3 1 10 7 10
Bimodal-felsic 0 3 12 19 29
Bimodal-siliciclastic 9 4 5 6 11
Number Number Number Number
of deposits of deposits of deposits of deposits
in situ value4in situ value4in situ value4in situ value4
>$1010 5–10 $1091–5 $1090.5–1 $109
Mafic 05052
Bimodal-mafic 1551628
Mafic-siliciclastic 1511010
Bimodal-felsic 0524236
Bimodal-siliciclastic 2510 16 9
1Grade and tonnage for combined mined and mineable reserves and resources
2Includes deposits with limited information
3Several small deposits with reported high Au grades disproportionately bias this value
4In US $, with 1 lb. Cu = $1.10, 1 lb. Zn = $0.60, 1 lb. Pb = $0.30, 1 oz. Au = $350, 1 oz. Ag = $5.00; excludes other metals
5Kidd Creek: $24.6 109, Brunswick #12: $22.1 109; Neves Corvo deposits: $16.1 109; Windy Craggy: $10.8 109

ophiolitic settings (Galley and Koski), and the examples
are found almost exclusively in Phanerozoic rocks (Fig. 4).
Modern analogs are found in ocean ridge, advanced back-
arc rift, and supra-subduction zone nascent arc settings.
The basaltic host rocks are predominantly tholeiitic, and
locally boninitic. Pre-Phanerozoic examples include the
Potter mine in the Late Archean Kidd-Munro assemblage
of the western, Abitibi subprovince, Ontario, Canada, and
the Coronation mine and nearby deposits in deformed ul-
tramafic-gabbroic tholeiite sequence in the Early Protero-
zoic Amisk Group in northeast Saskatchewan, Canada.
Mafic VMS are fewer in number (n = 62), smaller (average
of 2.8 MT: Fig. 3), and on average, they are Cu-rich and Pb-
poor in comparison to all other deposit types (Figs. 4, 6a).
Bimodal-Mafic Type
The bimodal-mafic type is defined as having >50 percent
mafic rocks and >3 percent felsic rocks in the host strati-
graphic succession, with subordinate siliciclastic rocks. Most
have a ratio of mafic/felsic volcanic rocks of 3:1 or greater,
but felsic rocks are commonly the immediate host rocks.
They predominate in Late Archean and Early Proterozoic
rocks (Table 2; Fig. 4). In broad terms, the composition of
the host rocks reflects primitive volcanic arc, or rifted prim-
itive volcanic arc settings. The mafic volcanic rocks are gen-
erally basaltic and tholeiitic, although they may be transi-
tional to calc-alkalic; felsic volcanic rocks are commonly high
silica rhyolites or transitional with calc-alkalic rhyolites (see
Barrett and MacLean). Classic examples are the deposits of
the Noranda district, Quebec (Gibson and Watkinson,
1990), the Flin Flon deposit in the Flin Flon-Snow Lake belt
of Manitoba-Saskatchewan (Syme and Bailes, 1993), and the
United Verde mine of the Jerome district in Arizona
(Gustin, 1990; Fig. 1). Kidd Creek also falls into this cate-
gory, but is atypical due to its immense size and its predomi-
nantly ultramafic footwall stratigraphic succession (Barrie et
al., see also Economic Geology Monograph 10, Hannington
and Barrie, eds., in press). Bimodal-mafic VMS are the most
common of the VMS types (n = 286), and they have higher
average Cu content than all but the mafic VMS type.
Mafic-Siliciclastic Type
The mafic-siliciclastic VMS type has subequal proportions
of mafic volcanic or intrusive rocks and turbiditic siliciclastic
4BARRIE AND HANNINGTON
TABLE 2. Grade and Tonnage for VMS Types by Time Periods
Average Average Average Average Average Average
Total tonnes Tonnes Cu grade Pb grade Zn grade Au grade Ag grade
nin MT in MT in wt % in wt % in wt % in g/t in g/t
MAFIC
Archean 1 1.5 1.5 (1.5)1 (4.15)
Early Proterozoic 3 1.9 0.6 (4.83) (0.34) (1.72) (5.23)
Middle and Late Proterozoic 0 0.0
Early Phanerozoic 23 60.0 2.6 1.77 (0.05) 2.86 (3.02) (18.0)
Late Phanerozoic 35 115.9 3.3 2.00 (0.10) (1.13) (1.74) (25.2)
BIMODAL-MAFIC
Archean 1212606.7 0.5 1.66 0.42 5.04 1.32 38.6
Early Proterozoic 73 410.2 5.6 2.20 0.98 4.32 1.47 28.7
Middle and Late Proterozoic 17 24.5 1.4 2.06 (0.97) 2.64 (1.42) (37.9)
Early Phanerozoic 54 278.8 5.2 1.93 (0.35) 3.02 2.40 44.4
Late Phanerozoic 19 130.6 6.9 1.74 (0.43) 2.54 (1.60) 28.4
MAFIC-SILICICLASTIC
Archean 2 1.4 (0.7) (1.37) (1.46) (42.5)
Early Proterozoic 7 159.8 (22.8) (2.38) (0.01) (1.27) (0.49) (25.7)
Middle and Late Proterozoic 16 307.4 19.2 1.68 (2.91) (2.44) (0.51) (17.4)
Early Phanerozoic 25 256.3 10.3 1.46 (1.73) 4.21 0.80 (33.2)
Late Phanerozoic 63 519.4 8.2 1.81 (0.02) 0.80 1.00 (12.4)
BIMODAL-FELSIC
Archean 24 170.2 7.1 1.09 1.23 6.23 0.83 125.2
Early Proterozoic 42 222.9 5.3 1.05 0.72 4.45 1.65 49.3
Middle and Late Proterozoic 14 68.0 4.9 1.53 0.85 4.07 1.47 109.2
Early Phanerozoic 82 375.0 4.6 1.53 2.50 6.69 2.63 85.8
Late Phanerozoic 93 472.6 5.1 1.64 1.52 5.29 2.04 115.7
BIMODAL-SILICICLASTIC
Archean 2 0.6 0.3 (1.23) (1.67) (4.60) (1.36) (37.7)
Early Proterozoic 9 24.6 2.7 (1.60) (1.82) (5.45) (1.09) (63.2)
Middle and Late Proterozoic 4 13.3 3.3 (1.15) (1.61) (5.28) 0.97 (57.1)
Early Phanerozoic 75 2451.1 32.7 0.93 1.74 3.83 0.76 54.8
Late Phanerozoic 7 14.9 2.1 (2.06) (2.13) (4.48) (2.85) (238.3)
1Grades in parentheses for averages based on less than 10 values
2Values in bold highlight data appreciably higher than other grade-tonnage data

rocks; felsic volcanic rocks are minor or absent. There may
be significant amounts of carbonate within the siliciclastic
rocks, but the siliciclastic component always predominates
(Slack, 1993). They are principally of Middle Proterozoic
age and younger, and they are commonly complexly de-
formed. The Besshi deposits of Japan and the Windy
Craggy deposit of British Columbia, Canada (Peter and
Scott), are type examples on land. The rifted continental
margin in the Guaymas basin of the Gulf of California, the
sedimented oceanic rift of Middle Valley and the Escanaba
trough in the northeast Pacific Ocean (Goodfellow and
Zierenberg, 1999), and the Atlantis II deeps of the Red
Sea (Zierenberg, 1990) provide three distinct tectonic set-
tings as analogs for the land-based deposits. Mafic-silici-
clastic VMS deposits are less numerous (n = 113) than
most of the other types, but their average tonnage (11.0
MT) is second only to the bimodal-siliciclastic VMS type.
Bimodal-Felsic Type
The fourth subdivision is the bimodal-felsic VMS type.
This type is defined by having >50 percent felsic volcanic
rocks, and <15 percent siliciclastic rocks in the host strati-
graphic succession, with mafic volcanic and intrusive
rocks accounting for the bulk of the remainder. Bimodal-
felsic VMS deposits have a similar age distribution as the
bimodal-mafic deposits, but they are most abundant in
the Phanerozoic (Table 2; Fig. 4). Almost by definition,
they are found in more compositionally mature volcanic
arcs, or rifted volcanic arc settings than the bimodal-mafic
types. The felsic host rocks are principally calc-alkalic,
although transitional, high-silica rhyolite to calc-alkalic
compositions are common. Similarly, mafic rocks are calc-
alkalic, or transitional tholeiitic to calc-alkalic volcanic
rocks (see Barrett and MacLean, 1999). Classic examples
are the Miocene deposits of the Hokuroku district, Japan
(Ohmoto and Skinner, 1983), the Rosebery deposits of the
Cambrian Mt. Read district, Tasmania (Green et al., 1981;
Large, 1992), and the Late Archean Izok Lake deposit,
Northwest Territories, Canada (Morrison and Balint,
1993). Bimodal-felsic deposits are the second-most nu-
merous, and on average contain the most Zn and Ag of
the five deposit types (Figs. 3, 4). They also commonly
contain barite.
The data for bimodal-felsic types indicate clear tem-
poral trends in terms of relative base metal contents
(Fig. 6b). Collectively, the Archean bimodal-felsic de-
posits are predominantly Zn-rich, whereas the relative
CLASSIFICATION OF VMS DEPOSITS BASED ON HOST-ROCK COMPOSITION 5
FIG. 3. Histograms for total tonnage, average tonnage and number of de-
posits for VMS types in database. The bimodal-siliciclastic type clearly has
the highest total and average tonnage, whereas the bimodal-mafic and bi-
modal-felsic types are the most numerous.
FIG. 4. Histogram of average metal contents for VMS types. Legend same
as in Fig. 3. See Table 1 for values. The mafic type has the highest aver-
age Cu content and the lowest average Pb content, whereas the opposite
is true for the bimodal-siliciclastic type.
Cu
wt. %
Zn
wt. %
Pb
wt. %
Ag
g/t
Au
g/t

proportions of Pb and Cu increase through time. The
average gold contents broadly increase through geolog-
ical time also. These trends in metal content through
time may reflect subtle differences in source rocks, with
younger volcanic arc systems relatively enriched in com-
parison to their Early Proterozoic and Late Archean
counterparts. Radiogenic decay accounts for an increase
of ~30 percent Pb in the crust (over 4.5 Ga), and Pb is
also expected to become enriched in crustal reservoirs
through time because it partitions strongly into the melt
during mantle partial melting. Cu and Au enrichment in
source rocks can be explained by a variety of magmatic
and hydrothermal processes (e.g., Candela and Hol-
land, 1986; Urabe, 1987; Hedenquist and Lowenstern,
1994).
Bimodal-Siliciclastic Type
The fifth type is termed bimodal-siliciclastic, and has ap-
proximately equal proportions of volcanic and siliciclastic
rocks. Felsic volcanic rocks are generally more abundant
than mafic ones. The vast majority of bimodal siliciclastic
deposits are Phanerozoic, principally in the Iberian Pyrite
Belt of Portugal and Spain or in the Bathurst camp of New
Brunswick, Canada. The felsic host rocks are generally
calc-alkalic, and in some cases it can be argued that they
were derived by partial melting of sedimentary sources,
consistent with a continental arc, or rifted continental arc
setting (see Carvalho et al., 1999). Mafic rocks are gener-
ally tholeiitic, but both the Bathurst district and the Iber-
ian pyrite belt have mildly alkaline basalts high in the
stratigraphic sections (alkaline rocks are rare in VMS host
rocks of any type). The bimodal-siliciclastic VMS deposits
represent the greatest tonnage of the VMS types (2.50 bil-
lion tonnes), and they have the largest average deposit size
6BARRIE AND HANNINGTON
20
50
20
50
3.0 2.0 1.0
100
150
200
250
300
350
40
60
100
150
200
250
40
60
80
100
120
140
BIMODAL-SILICICLASTIC
Age, in Ga
5
10
15
20
25
MAFIC
MAFIC-SILICICLASTIC
BIMODAL-MAFIC
BIMODAL-FELSIC
Tonnage, in MT
P
M
Mu
MS
D
SR
WC
KC
AE SM
M
MC
G
H
R
C
ML
S
NC
RT
LZ
FIG. 5. Tonnage vs. age for VMS types. The major periods of VMS forma-
tion were in the Late Archean at ~2730–2700 Ma, in the Early Protero-
zoic at ~1890–1870 Ma, in the Early Paleozoic and more broadly through
the Mesozoic and Cenozoic. The largest examples are labeled as follows:
Mafic: AE: Anayatak-Ergani, Turkey; SM: Sirrt Madenkoy, Turkey; M:
Mavrovouni, Cyprus. Bimodal-Mafic: KC: Kidd Creek, Ontario, Canada;
G: Geco, Ontario, Canada; H: Horne, Quebec, Canada; R: Ruttan, Man-
itoba; C: Crandon, Wisconsin, USA; ML: Mount Lyell, Tasmania, Aus-
tralia. Mafic-Siliciclastic: S: Saladipura, Rajasthan, India; D: Ducktown,
Tennessee, USA; R: Rouez, Bretagne, France; WC: Windy Craggy, British
Columbia, Canada. Bimodal-Felsic: MC: Mons Cupri, Western Australia;
S: Selbaie, Quebec, Canada; P: Pyhasalmi, Finland; M: Mt. Morgan,
Queensland, Australia; Mu: Murgul, Turkey; MS: Matsumine-Shakanai,
Hokuroku, Japan. Bimodal-Siliciclastic: NC: Neves Corvo, Portugal; RT:
Rio Tinto, Spain; LZ: La Zarza, Spain.
Cu
Pb Zn
MAFIC
BIM.-MAF.
MAF.-SIL.
BIM.-FEL.
BIM.-SIL.
a.
Cu
BIMODAL-FELSIC
Pb Zn
L. PHAN.
E. PHAN.
M+L PROT.
E.PROT.
ARCH.
b.
FIG. 6. Ternary diagrams of base metal contents in VMS deposits. a. The
five VMS types. The proportion of Pb with respect to Cu and Zn increases
from the mafic type to the bimodal-siliciclastic type, whereas the propor-
tion of Cu with respect to Pb and Zn tends to decrease. b. Bimodal-felsic
VMS types during five time periods, excluding the Middle Archean, Pb-
rich Mons Cupri deposit. The proportion of Pb and Cu with respect to
Zn tends to increase through time.

(23.7 MT, Fig. 3). They have on average the lowest Cu con-
tent and the highest Pb content of the five VMS deposit
types (Fig. 4).
Host-Rock Composition and VMS Metal Content:
The Use of Primitive Mantle-Normalized Plots
Host-rock compositions may influence the metal con-
tent in VMS deposits (Franklin et al., 1981), and this is
clear in the classification scheme presented here. Mafic
rocks contain ferromagnesian minerals and minor mag-
matic sulfide (immiscible sulfide-oxide solid solution) that
are preferentially enriched in Cu; the Cu is available to hy-
drothermal fluids when the crystal lattice of the host min-
eral is destroyed during hydrothermal alteration. Simi-
larly, felsic rocks contain feldspars that are preferentially
enriched in Pb and Ba, and siliciclastic rocks contain
feldspars and clays enriched in Pb, Ag and Zn. Gold en-
richment in host rocks can be due to partial melting of a
residual mantle that has retained Au-enriched magmatic
sulfide to form boninites (Hamlyn et al., 1985), suppres-
sion of oxide and sulfide fractionation that sequester
gold in alkaline, high fO2magmatic systems leading to
(McInnes and Cameron, 1994), and a variety of hydro-
thermal processes (Hannington et al.).
The base and precious metal contents for the VMS de-
posit type averages are presented on primitive mantle-
normalized diagrams for comparison in Figure 7, and in Fig-
ure 8, a comparison is made between deposits and possible
source rocks. The primitive mantle composition is used for
two reasons: the metal values are reasonably well estab-
lished (Table 3), and it provides a reasonable comparison
for crustal reservoir source rocks as well as the deposits, so
both potential source rocks and deposits can be compared
from district to district. The ordering of the elements—Pb,
Ag, Au, Zn and Cu—corresponds to the degree of incom-
patibility in a source rock during magmatic processes (as-
suming the presence of trace immiscible sulfide-oxide
solid solution), with Pb representing the most incompati-
ble metal. The ordering also corresponds to the relative
enrichment (enriched to the left) of metals in most VMS
deposits compared to the primitive mantle and to the rel-
ative solubility of different metals in modified seawater
with increasing temperatures. A more rigorous compari-
son would require that the metal contents for the deposit
averages be normalized to 100 percent sulfide, assuming
that the Cu, Zn and Pb are in chalcopyrite, sphalerite and
galena, respectively, with the remaining Fe and S distrib-
uted between pyrite, pyrrhotite and magnetite according
to the mineralogy. This approach, which has proven suc-
cessful with magmatic sulfide ores (Naldrett and Duke,
1980), would alleviate some of the uncertainty in compar-
ing semi-massive and massive ores, but requires knowledge
of the bulk Fe and S contents of the deposits and the
mode of iron sulfide and iron oxide minerals, which usu-
ally are not reported. It would also be desirable to include
other VMS metal co- and by-products (e.g., Hg, As, Sb, Ba,
Co, Sn, S, Se, Mn, Cd, In, Bi, etc.). In this way, variable
source rock or magmatic contributions, or the distinctive
physical properties of different hydrothermal fluids, could
be fingerprinted, even within a single deposit, but more
information is needed before such a comprehensive ap-
proach can be taken.
As expected, the mafic VMS-type averages have the flat-
test patterns when normalized to the primitive mantle.
They also have a tendency toward a relative enrichment in
Au in comparison to Ag and Zn. This could be due to: (1)
supergene processes that cause preferential enrichment in
Au in comparison to the other elements; (2) the recovery
CLASSIFICATION OF VMS DEPOSITS BASED ON HOST-ROCK COMPOSITION 7
TABLE 3. Metal and Elemental Concentrations in Chondritic, Mantle, and Crustal Reservoirs
Metal N-MORB Primitive Mantle Chondrite (C1 type) Bulk Continental Crust9
Ba (ppm) 13.87106.049102.4100250.00
Au (ppb) 0.87201.39000 15210.00 3.0.0
Ag (ppb) 3030.00 1950.000 18030.00 80.00
Pb (ppm) 0.35800.175102.47408.0.0
Cd (ppb) 13090.00 4090.000 64070.00 98.00
Sn (ppm) 1.35000.124001.62402.5.0
Zn (ppm) 8480.00 5090.000 30070.00 80.00
Cu (ppm) 7080.00 2810.000 10810.00 75.00
Co (ppm) 47.110010410.000 5160.00 29.00
Ni (ppm) 149.51002,08010.000 10,50010.00 105.00
As (ppm) 190.00 10050.000 1,50070.00 1.0.0
Sb (ppm) 0.01500.005500.16500.2.0
Mo (ppm) 0.31500.063500.92501.0.0
Bi (ppb) 790.00 1090.000 11070.00 60.00
Ga (ppm) 1770.00 470.000 1070.00 18.00
S (wt.%) 0.1010 0.02510 5.4010 ND
Se (ppm) 0.16700.047001910.00 0.05
Mn (ppm) 1,00070.00 1,01070.000 1,70070.00 1,400.00
1Hofmann, 1988
2Hamlyn et al., 1985
3Keays and Scott, 1976
4Sun and McDonough, 1989, Joachum et al., 1993
5Wolf and Anders, 1980
6Palme et al., 1981
7Sun, 1982
8At Mg number = 70: Doe, 1995
9Taylor and McClennan, 1985. Values listed under N-MORB are for
average oceanic crust
10 McDonough and Sun, 1995

of metals for deposits mined historically (e.g., Zn not re-
covered from Cu ores in many Troodos deposits), or (3) a
relative enrichment in the source rocks by magmatic
processes (e.g., boninites relatively enriched in Au: Ham-
lyn et al., 1985). It is noted that there are few accurate gold
grades reported for the mafic VMS types, so the apparent
Au enrichment should be considered with caution. All of
the other types have “negative” slopes for their primitive-
mantle normalized patterns, with the steepest slopes for
the felsic-influenced VMS types. Systematic changes in the
patterns through geological time are lacking for all but the
bimodal-felsic VMS deposits. Younger bimodal-felsic de-
posits have higher Cu, Pb and Au contents.
In Figure 8, a comparison is made between the mafic
type and the bimodal-siliciclastic deposits that represent
the most primitive and the most evolved VMS types, re-
spectively. The relatively flat pattern for the mafic VMS
type at ~103primitive mantle values is broadly parallel
to the pattern for N-MORB. Similarly, the steep negative
slope for the bimodal-siliciclastic VMS type from 105to ~5
102primitive values broadly parallels the relatively
steep, negative slope of the bulk continental continental
crust. Both deposit types exhibit metal enrichments of
3–13 102over their potential host rock compositions,
with the greatest enrichment for Pb. The patterns in Fig-
ure 8 are consistent with the leaching and transport of
metals by fluids that are undersaturated with respect to all
of the metals, as is generally observed at vent sites on the
modern sea floor (Hannington et al., 1995). Detailed as-
pects of metal transport are discussed in Seyfried et al.,
1999.
8BARRIE AND HANNINGTON
MAFIC
BIMODAL-
SILICICLASTIC
BIMODAL-FELSIC
BIMODAL-MAFIC
MAFIC-SILICICLASTIC
Pb Ag Au Zn Cu
102
103
104
10
102
103
104
105
103
105
104
105
103
104
103
104
5
FIG. 7. Primitive mantle-normalized metal values for average of VMS
types by age period. Archean averages: diamonds; Early Proterozoic av-
erages: squares; Middle and Late Proterozoic averages: triangles; Early
Phanerozoic (Paleozoic) averages: x’s; Late Phanerozoic averages: aster-
isks. Primitive mantle values given in Table 3.
EARLY PHAN.MAFIC VMS AVE.
N-MORB
102
103
104
10
1
Pb Ag Au Zn Cu
E. PHAN. BIMODAL-
SILICICLASTIC VMS AVE.
PM-NORMALIZED PM-NORMALIZED
102
103
104
105
10 1
Pb Ag Au Zn Cu
ENRICHMENT BY
HYDROTHERMAL PROCESSES
ENRICHMENT BY
HYDROTHERMAL PROCESSES
ENRICHMENT BY
MAGMATIC PROCESSES
ENRICHMENT
BY CRUSTAL-
BUILDING PROCESSES
BULK CONTINENTAL CRUST
EARLY PHAN.
MAFIC VMS AVE.
N-MORB
FIG. 8. Primitive mantle-normalized metal content plots for VMS deposit
type averages and possible source rocks. Metal contents and normalizing
values are given in Tables 2 and 3. a. Early Phanerozoic mafic VMS and
N-MORB plots. In broad terms, there is a 500–1,000enrichment for all
of the metals in the deposit average in comparison to N-MORB. The
average gold content for the VMS average is based on relatively few
deposits and may be anomalously high. b. Early Phanerozoic bimodal-
siliciclastic VMS and bulk continental crust plots. Bimodal-siliciclastic
deposits have higher Pb and Ag contents and lower Cu contents than
mafic VMS, and their sloped pattern parallels that of the bulk continen-
tal crust.
a
b

Deposit Size: Host-Rock Permeability,
Duration of Heat Source
Among the most important controlling factors on the
size of a VMS deposit are the permeability of the host-rock
stratigraphic succession and the duration of the magmatic
heat source. The presence of a significant siliciclastic com-
ponent to the host stratigraphic succession favors large
VMS deposits, as the largest deposits are either mafic-sili-
ciclastic or bimodal-siliciclastic (Fig. 3). This is perhaps
not surprising if a continuum with sedimentary-exhalative
deposits (SEDEX) is considered. The typical SEDEX de-
posit is 41.3 MT (n = 62: Lydon, 1996) and is larger than
siliciclastic-poor VMS systems by a factor of 8–15, but
larger than siliciclastic-rich VMS systems by a factor of only
2–4 (see Table 1). Turbidites are less permeable than vol-
canic rocks, and in the absence of abundant faulting, a
turbidite-rich setting can effectively insulate a hydrother-
mal cell and its heat source from rapid advective cooling,
thus allowing for a longer-lived hydrothermal system, and
relatively efficient, subsea-floor metal deposition (Good-
fellow and Zierenberg).
On a local scale, large deposits are also favored by
porous and relatively permeable epiclastic or autoclastic
breccias in the area of metal deposition. Epiclastic rocks
may provide a favorable location for sulfide-after-silicate
replacement in the subsurface, leading to a high deposi-
tional efficiency. This is true at the giant Kidd Creek and
Horne deposits in the Abitibi subprovince, for example
(Hannington et al., in press; Kerr and Gibson, 1993).
High level (within 3 km in the footwall), synvolcanic in-
trusions are reasonable sources of heat, and they can drive
convection of metal-precipitating hydrothermal fluids
through the adjacent rocks (Campbell et al., 1981). They
may also provide some metals to the hydrothermal system
(e.g., Large et al., 1996). Such intrusions are present in the
stratigraphic footwall in ~75 percent of the mafic VMS types,
~50 percent of the bimodal-mafic and bimodal-felsic VMS
types, but are much less common in the mafic-siliciclastic
and bimodal-siliciclastic types (excluding relatively thin
sills). Tectonic imbrication and other structural complexi-
ties can account for the lack of preservation of high level
synvolcanic intrusions in some cases, but it would appear
that such intrusions were not present at the time of for-
mation for as many as one-third of all VMS deposits. The
heat sources for hydrothermal convection in these cases
may have been deeper in the crust. Larger, hotter and
longer-lived magmatic heat sources lead to larger deposits.
Relationships between large, hot heat sources, host-rock
permeability, and the size of VMS deposits are explored
quantitatively using two-dimensional finite element heat
and fluid flow modeling by Cathles et al. (1997; see Barrie
et al., 1999).
Tectonic Setting and
VMS Deposits through Geologic Time
A spectrum of tectonic settings are recognized for VMS
deposits. They include: oceanic ridges (e.g., the TAG
hydrothermal field and associated sea-floor massive sulfide
deposits (Rona et al., 1993), thickened oceanic crust, e.g.,
the Galapagos area (Perfit et al.), sedimented oceanic
ridges and sedimented continental margin rifts (Middle
Valley and Escanaba trough; Guaymas basin, respectively:
Goodfellow and Zierenberg, 1999), and a variety of rifted
arc settings, including nascent arcs (most ophiolites:
Galley and Koski, 1999), primitive volcanic arcs (many
Archean and Early Proterozoic deposits), mature volcanic
arcs (e.g., Hokuroku district, Japan: Ohmoto and Skinner,
1983), and continental arcs (Iberian Pyrite Belt: de Car-
valho, 1999).
The vast majority of VMS deposits have at least a minor
amount of mafic volcanic rocks in their host stratigraphic
succession. As most mafic rocks are derived from the
upper mantle, it is implicit that, in the broadest sense,
heat derived from the upper mantle is fundamentally re-
sponsible for the thermal anomalies in the crust that lead
to VMS mineralization. A corollary to this is that there are
no known VMS deposits related to anorogenic magmatism
driven principally by radiogenic heat production (e.g.,
many S-type and A-type, minimum melt granitic systems).
The most prolific periods of VMS mineralization in
terms of the number of deposits represented in the rock
record are in the Late Archean (2750–2700 Ma), the Early
Proterozoic (1900–1800 Ma), the early Phanerozoic (500–
450 Ma) and two periods in the late Phanerozoic (390–
250 Ma and 30–0 Ma: Fig. 5). Bimodal-mafic types are
most abundant in the (Late) Archean and Early Protero-
zoic, whereas the bimodal-felsic and bimodal-siliciclastic
types are more abundant in the Phanerozoic. This is con-
sistent with a decrease in the global heat flux through time
that favors the generation of primitive, mafic arcs in ear-
lier times, and more evolved, felsic arcs in the Phanero-
zoic. That the mafic types are found almost exclusively
within Phanerozoic rocks reflects the scarcity of ophiolitic
sections in earlier times. Titley (1993) examined the for-
mation of strata-bound ore deposits, including VMS de-
posits, through Wilson tectonic cycles during the Protero-
zoic and Phanerozoic. He noted that an abundance of
arc-related (including mafic, bimodal-mafic, and bimodal-
felsic) VMS deposits in the mid-Cambrian to Silurian and
in the late Cretaceous corresponded to: (1) periods of high
sea level where the depth of seawater is hundreds of meters
above the average continental shelf edge, (2) oceanic
anoxic events (see also Eastoe and Gustin, 1996), (3) open
stages in the Wilson Cycle, with fragmentation and dis-
persal of the continents due to “craton heating,” and (4)
abundant volcanic arc and back-arc tectonic settings.
There are relatively few VMS deposits of any type in the
Middle and Late Proterozoic (Hutchinson, 1980), a time
represented by limited arc magmatism and stable cratonic
environments (Windley, 1977).
In summary, the deposits described in this volume
represent many tectonic settings that have occurred
through Earth’s history, and they illustrate key processes
that have been responsible for VMS formation in a range
of environments.
CLASSIFICATION OF VMS DEPOSITS BASED ON HOST-ROCK COMPOSITION 9

Acknowledgments
We thank the three organizations that sponsored the short
course that lead to this volume: the Society of Economic Ge-
ologists, the Mineral Deposits Division of the Geological As-
sociation of Canada, and the Geological Association of
Canada. We also thank the Geological Survey of Canada and
the Chief Geologist’s office for support for this publication.
Mike Lesher, the former Series Editor for Reviews in Economic
Geology, invited us to consider overseeing this volume, and
we are grateful to him for providing us with this opportunity.
This contribution has benefited from discussions with Ian
Jonasson and John Lydon, and from comments by Wayne
Goodfellow. GSC Contribution No. 1997092.
REFERENCES
Alt, Jeffrey C., 1999, Hydrothermal alteration and mineralization of
oceanic crust: Mineralogy, geochemistry, and processes: Reviews in Eco-
nomic Geology, v. 8, p. 131–153.
Barrett, Timothy J., and MacLean, Wallace H., 1999, Volcanic sequences,
lithogeochemistry, and hydrothermal alteration in some bimodal vol-
canic-associated massive sulfide systems: Reviews in Economic Geology,
v. 8, p. 101–132.
Barrie, C.T., Cathles, L.M., Erendi, A., Schwaiger, H., and Murray, C.,
1999, Heat and fluid flow in volcanic-associated massive sulfide-forming
hydrothermal systems: Reviews in Economic Geology, v. 8, p. 197–215.
Barrie, C.T., Hannington, M.D., and Bleeker, W. , 1999, The Giant Kidd
Creek volcanic-associated massive sulfide deposit, Abitibi subprovince,
Canada, Reviews in Economic Geology, v.8, p. 241–253.
Berger, B.R., and Bethke, P.M., eds., 1985, Geology and geochemistry of
epithermal systems: Reviews in Economic Geology, v. 2, 298 p.
Campbell, I.H., Franklin, J.M., Gorton, M.P., and Hart, T.R., 1981, The
role of subvolcanic sills in the generation of massive sulfide deposits:
Economic Geology, v. 76, p. 2248–2253.
Candela, P.A., and Holland, H.D., 1986, A mass transfer model for cop-
per and molybdenum in magmatic hydrothermal systems: The origin
of porphyry-type ore deposits: Economic Geology, v. 81, p. 1–19.
Carvalho, D., Barriga, F.J.A.S., and Munha J., 1999, Bimodal-siliciclastic
systemsCthe case ofhe Iberian Pyrite belt, 1999, Reviews in Economic
Geology, v. 8, p. 369–402., D., Barriga, F.J.A.S., and Munha J., 1999, Bi-
modal-siliciclastic systems—the case of the Iberian Pyrite belt, 1999, Re-
views in Economic Geology, v. 8, p. 369–402.
Cathles, L.M., Erendi, A.H.J., Thayer, J.B., and Barrie, C.T., 1997, How
long can a hydrothermal system be sustained by a single intrusion
event? Economic Geology, v. 92, p. 766–771.
Doe, B.R., 1995, Zinc, copper, and lead geochemistry of oceanic igneous
rocks: Ridges, islands and arcs: International Geology Reviews, v. 37, p.
379–420.
Eastoe, C.J., and Gustin, M.M., 1996, Volcanogenic massive sulfide de-
posits and anoxia in the Phanerozoic oceans: Ore Geology Reviews, v.
10, p. 179–197.
Franklin, J.M., 1993, Volcanic-associated massive sulphide deposits, in
Kirkham, R.V., Sinclair, W.D., Thorpe, R.I., and Duke, J.M., eds., Min-
eral deposit modeling: Geological Association of Canada Paper 40, p.
315–334.
Franklin, J.M., Sangster, D.M., and Lydon, J.W., 1981, Volcanic-associated
massive sulfide deposits: Economic Geology 75th Anniversary Volume,
p. 485–627.
Galley, Alan G., and Koski, Randolf A., 1999, Setting and characteristics
of ophiolite-hosted volcanogenic massive sulfide deposits: Reviews in
Economic Geology, v. 8, p. 221–246.
Gibson, H., 1999, Submarine volcanic processes, deposits, and environ-
ments favorable for the location of volcanic-associated massive sulfide
deposits: Reviews in Economic Geology, v. 8, p. 13–52.
Gibson, H.L., and Watkinson, D.H., 1990, Volcanogenic massive sulphide
deposits of the Noranda cauldron and shield volcano, Quebec: The
Northwestern Quebec Polymetallic Belt: A Summary of 60 Years of
Mining Exploration, Canadian Institute of Mining and Metallurgy Spe-
cial Volume 43, p. 119–132.
Goodfellow, W.D., and Zierenberg, R.A., 1999, Genesis of massive sulfide
deposits at sediment-covered spreading centers: Reviews in Economic
Geology, v. 8, p. 297–324.
Green, G.R., Solomon, M., and Walshe, J.L, 1981, The formation of the
volcanic-hosted massive sulfide ore deposit at Rosebery, Tasmania: Eco-
nomic Geology, v.76, p. 304–338.
Gustin, M.S., 1990, Stratigraphy and alteration of the host rocks, United
Verde massive sulfide deposit, Jerome, Arizona: Economic Geology, v.
85, p. 29–49.
Hamlyn, P., Keays, R.R., Cameron, W.E., Crawford, A.J., and Waldron,
H.M. 1985, Precious metals in magnesian low-Ti lavas: Implications for
metallogenesis and sulfur saturation in primary magmas: Geochimica
et Cosmochimica Acta, v. 49, p. 1797–1811.
Hannington, M.D., and Barrie, C.T., 1998, The giant Kidd Creek vol-
canogenic massive sulfide deposit, Western Abitibi Subprovince, Canada,
preface and introduction, Economic Geology Monograph 10, in press.
Hannington, M.D., Jonasson, I.R., Herzig, P.M., and Petersen, S., 1995,
Physical and chemical processes of seafloor mineralization at mid-
ocean ridges, in Humphris, S.E., Zierenberg, R.A., Mullineaux, L.S.,
and Thomson, R.E., eds., Seafloor hydrothermal systems: Physical,
chemical, biological and geological interactions: Geophysical Mono-
graph 91, American Geophysical Union, p. 115–157.
Hannington, M.D., Bleeker, W., and Kjarsgaard, I., 1998, Sulfide miner-
alogy, geochemistry, and ore genesis of the Kidd Creek deposit, part I:
North, central and south orebodies: Economic Geology Monograph
10, in press.
Hannington, M.D., Poulsen, K.H., Thompson, J.F.H., and Sillitoe, R.H.,
1999, Volcanogenic gold in the massive sulfide environment, Reviews in
Economic Geology, v. 8, p. 324–356.
Harper, G.D, 1999, Structural styles of hydrothermal discharge in ophio-
lite/sea-floor systems: Reviews in Economic Geology, v. 8, p. 53–74.
Hedenquist, J.W., and Lowenstern, J.B., 1994, The role of magmas in the
formation of hydrothermal ore deposits: Nature, v. 370, p. 519–527.
Hofmann, A.W., 1988, Chemical differentiation of the Earth: The rela-
tionship between mantle, continental crust and oceanic crust: Earth
and Planetary Science Letters, v. 90, p. 297–314.
Huston, David L., 1999, Stable isotopes and their significance for under-
standing the genesis of volcanic-hosted massive sulfide deposits: A re-
view: Reviews in Economic Geology, v. 8, p. 157–180.
Hutchinson, R.W., 1973, Volcanogenic sulfide deposits and their metal-
logenic significance: Economic Geology, v. 68, p. 1223–1246.
——1980, Massive base metal sulphide deposits as guides to tectonic evo-
lution, in Strangway, D.W., ed., Geological Association of Canada Spe-
cial Paper 20, p. 659–684.
Joachum, K.P., Hofmann, A.W., and Seufert, H.M., 1993, Tin in mantle-
derived rocks: Constraints on Earth evolution: Geochimica et Cos-
mochimica Acta, v. 57, p. 3585–3595.
Keays, R.R., and Scott, R.B., 1976, Precious metals in ocean-ridge basalts:
Implications for basalts as source rocks for gold mineralization: Eco-
nomic Geology, v. 71, p. 705–720.
Kerr, D.J., and Gibson, H.L., 1993, A comparison between the volcanol-
ogy and geochemistry of volcanic successions hosting the Horne mine
deposit and smaller intra-caldron deposits of the mine sequence: Eco-
nomic Geology, v. 88, p. 1419–1443.
Large, R.R., 1992, Australian volcanic-hosted massive sulfide deposits:
Features, styles and genetic models: Economic Geology, v. 87, p. 471–510.
Large, R.R., Doyle, M., Raymond, O., Cooke, D., Jones, A., and Heasman,
L., 1996, Evaluation of the role of Cambrian granites in the genesis of
world class VMHS deposits in Tasmania: Ore Geology Reviews, v. 10, p.
215–230.
Lydon, J.W., 1996, Sedimentary exhalative sulphides (SEDEX), in Eck-
strand, O.R., Sinclair, W.D., and Thorpe, R.I., eds., Geology of Cana-
dian mineral deposit types: Geological Survey of Canada, Geology of
Canada, no. 8, p. 130–152.
McDonough, W.F., and Sun, S.S., 1995, The composition of the Earth:
Chemical Geology, v. 120, p. 223–253.
McInnes, B., and Cameron, E., 1994, Carbonated alkaline hybridizing
melts from a sub-arc environment: Mantle wedge samples from the
Tabar-Feni-Lihir-Tanga-Feni Arc, Papua New Guinea: Earth and Plane-
tary Science Letters, v. 122, p. 125–141.
Morrison, I.R., and Balint, F., 1993, Geology of the Izok massive sulphide
deposits, Northwest Territories, Canada, in Matthew, I.G., ed., World
10 BARRIE AND HANNINGTON

Zinc ’93: Proceedings of the International Symposium on Zinc, Aus-
tralasian Institute of Mining and Metallurgy, Parkville, Australia, July,
1993, p. 161–170.
Morton, R.L., and Franklin, J.M., 1987, Two-fold classification of Archean
volcanic-associated massive sulfide deposits: Economic Geology, v. 82,
p. 1057–1063.
Mosier, D.L., Singer, D.A., and Salem, B.B., 1983, Geologic and grade-
tonnage information on volcanic-hosted copper-zinc-lead massive sul-
fide deposits: U.S. Geological Survey Open File Report 83-89, 78 p.
Naldrett, A.J., and Duke, J.M., 1980, Platinum metals in magmatic sulfide
ores: Science, v. 208, p. 1417–1424.
Ohmoto, H., and Skinner, B.J., 1983, eds., The Kuroko and related vol-
canogenic massive sulfide deposits: Economic Geology Monograph 5,
604 p.
O’Neill, H.S.C., Dingwell, D.B., Bosisov, A., Spettel, B., and Palme, H.,
1995, Experimental petrochemistry of some highly siderophile ele-
ments at high temperatures, and some implications for core formation
and the mantle’s early history: Chemical Geology, v. 120, p. 255–273.
Palme, H., Suess, H.E., and Zeh, H.D., 1981, Abundances of the elements
in the solar system, in Hellwege, K.H., ed., Landolt-Bornstein, Group
VI: Astronomy, astrophysics, extension and supplement 2, subvolume a:
Berlin, Springer, p. 257–272.
Perfit, Michael R., Ridley, W. Ian, and Jonasson, Ian R., 1999, Geologic,
petrologic, and geochemical relationships between magmatism and
massive sulfide mineralization along the eastern Galapagos spreading
center: Reviews in Economic Geology, v. 8, p. 75–100.
Peter, Jan M., and Scott, Steven D., 1999, Windy Craggy, Northwestern
British Columbia: The world’s largest Besshi-type deposit: Reviews in
Economic Geology, v. 8, p. 261–296.
Rona, P.A., 1988, Hydrothermal mineralization at oceanic ridges: Cana-
dian Mineralogist, v. 26, p. 431–466.
Rona, P.A., and Scott, S.D., 1993, A special issue on sea-floor hydrother-
mal mineralization: New perspectives–Preface: Economic Geology, v.
88, p. 1933–1976.
Rona, P.A., Hannington, M.D., Raman, C.V., Thompson, G., Tivey, M.K.,
Humphris, S. E., Lalou, C., and Petersen, S., 1993, Active and relict sea-
floor hydrothermal mineralization at the TAG hydrothermal field, Mid-
Atlantic Ridge: Economic Geology, v. 88, p. 1989–2017.
Roth, Tina, Thompson, John F.H., and Barrett, Timothy J., 1999, The
precious metal-rich Eskay Creek deposit, Northwestern British Colum-
bia: Reviews in Economic Geology, v. 8, p. 357–374.
Sangster, D., and Scott, S.D., 1976, Precambrian strata-bound massive Cu-
Zn-Pb sulfide ores of North America, in Wolf, K.H., ed., Handbook of
strata-bound and stratiform ore deposits: Amsterdam, Elsevier, p.
129–222.
Sawkins, F.J., 1976, 1976, Massive sulphide deposits in relation to geotec-
tonics: Geological Association of Canada Special Paper 14, p. 221-240.
Seyfried, W.E., Jr., Ding, Kang, Berndt, Michael E., and Chen, Xian, 1999,
Experimental and theoretical controls on the composition of mid-
ocean ridge hydrothermal fluids, Reviews in Economic Geology, v. 8, p.
180–200.
Slack, J.F., 1993, Descriptive and grade-tonnage models for Besshi-type
massive sulphide deposits, in Kirkham, R.V., Sinclair, W.D., Thorpe,
R.I., and Duke, J.M., eds., Mineral deposit modeling: Geological Asso-
ciation of Canada Paper 40, p. 343–372.
Solomon, M., 1976, Volcanic massive sulphide deposits and their host
rocks: A review and an explanation, in Wolf, K.H., ed., Handbook of
strata-bound and stratiform ore deposits, II: Regional studies and spe-
cific deposits: Amstedam, Elsevier, p. 1307–1328.
Sun, S.S., 1982, Chemical composition and origin of the earth’s primitive
mantle: Geochimica et Cosmochimica Acta, v. 46, p. 179–192.
Sun, S.S., and McDonough, W.F., 1989, Chemical and isotopic systemat-
ics of oceanic basalts: Implications for mantle composition and
processes, in Saunders, A.D., and Norry, M.J., eds., Magmatism in the
ocean basins: Geological Society of America Special Publication 42, p.
313–345.
Syme, E.C., and Bailes, A.H., 1993, Stratigraphic and tectonic setting of
Early Proterozoic volcanogenic massive sulfide deposits, Flin Flon,
Manitoba: Economic Geology, v. 88, p. 566–589.
Taylor, S.R., and McLennan, S.M., 1985, The continental crust: Its com-
position and evolution: Oxford, Blackwell Scientific Publications, 312
p.
Titley, S.R., 1993, Relationship of stratabound ores with tectonic cycles
of the Phanerozoic and Proterozoic: Precambrian Research, v. 61, p.
295–322.
Urabe, T., 1987, The effect of pressure on the partitioning ratios of lead
and zinc between vapor and rhyolite melts: Economic Geology, v. 82, p.
1049–1052.
Windley, B.F., 1977, The evolving continents: New York, Wiley, 420 p.
Wolf, R., and Anders, E., 1980, Moon and Earth: Compositional differ-
ences inferred from siderophiles, volatiles, and alkalis in basalts:
Geochimica et Cosmochimica Acta, v. 44, p. 2111–2124.
Zierenberg, R.A., 1990, Deposition of metalliferous sediment beneath a
brine pool in the Atlantis II Deep, Red Sea, in McMurry, G.R., ed.,
Gorda Ridge: A seafloor spreading center in the United States Exclu-
sive Economic Zone: New York, Springer-Verlag, p. 131–142.
CLASSIFICATION OF VMS DEPOSITS BASED ON HOST-ROCK COMPOSITION 11