Geology and Genesis of the Giant Gorevskoe Pb-Zn-Ag Deposit, Krasnoyarsk Territory, Russia


 The Gorevskoe Pb-Zn-Ag mine is currently the largest producer of Pb and Zn in Russia, exploiting one of the largest sediment-hosted Pb-Zn deposits worldwide. Despite its size and economic importance, the Gorevskoe deposit remains poorly understood. It is located on the western margin of the Siberian craton within the Yenisei Ridge, a Neoproterozoic orogenic belt. Mineralization consists of three tabular orebodies that are in turn composed of multiple stacked stratiform to strata-bound lenses of galena-pyrrhotite-sphalerite-rich massive sulfide ore, hosted in organic-rich marine metalimestones and calcareous slates of Stenian to Tonian age (1,020 ± 70 Ma). Extensive Fe-Mg-Mn-carbonate alteration haloes surround the ore lenses. The Pb isotope signature of the deposit is consistent with derivation of Pb, and probably all associated metals, from an evolved crustal source at the time of formation of the host rocks. The sulfur isotope compositions of pyrrhotite, sphalerite, galena, arsenopyrite, and pyrite (δ34S = 16.0–20.4‰) do not vary considerably across the deposit and are within the range reported for contemporaneous seawater, indicating complete reduction of marine sulfate as the main source of sulfur.
 The available geologic and geochemical data indicate that the Gorevskoe deposit belongs to the sediment-hosted massive sulfide (SHMS) class of Zn-Pb deposits, with an affinity to Selwyn-type deposits. Hydrothermal mineralization appears to be temporally related to rifting and distal mafic volcanism in a passive margin setting. Geologic relationships suggest that the orebodies formed in a diagenetic environment. Furthermore, the predominance of primary pyrrhotite over pyrite as the major iron sulfide, the presence of abundant siderite, and the relatively homogeneous sulfur isotope signature of the ores indicate highly reducing conditions during ore formation. They also constrain the character of the metal-bearing fluid to be similarly reducing, and of moderate temperature (200°–300°C). Gorevskoe may thus be regarded as one of the world’s largest Selwyn-type SHMS deposits.


Introduction
With pre-mining reserves and resources of 106.43 million metric tons (Mt) of ore at 6.14% Pb, 1.82% Zn, and 49 g/t Ag (Makarov et al., 2014), the Gorevskoe deposit is one of the largest Pb-Zn-Ag deposits worldwide. It is located on the Angara River, 30 km upstream of its confluence with the Yenisei River. Because the greater part of the orebodies is situated underneath the riverbed, its exploitation required the Angara River to be diverted, and the 215-m-deep open pit to be protected by a large dam (Fig. 1).
Despite its size and economic importance, very little has been published on the geology and genesis of the Gorevskoe deposit in the western scientific literature. Smirnov and Gorzhevski (1977) and Seltmann et al. (2010) only provided very broad descriptions. As a result, Gorevskoe has been variably considered as either a sediment-hosted massive sulfide (SHMS) deposit (Leach et al., 2005;Lobanov and Nekos, 2017) or a Mississippi Valley-type (MVT) deposit (Leach et al., 2010). This uncertainty is also reflected in the Russian literature where epigenetic (Sherman, 1968;Okhapkin, 1981; The Gorevskoe deposit was discovered in 1956 during a 1:200,000-scale geologic survey led by Yu. A. Glazyrin (Sherman et al., 1963). Its discovery was greatly facilitated by the contemporaneous completion of the Irkutsk Hydroelectric Power Plant near Lake Baikal. The flow of the Angara River was stopped to fill up the reservoir, significantly lowering the water level downstream. This enabled the discovery of outcrops of Pb-Zn mineralization in the riverbed (Strimzha, 2017). The deposit was subsequently explored below thin younger cover by drilling and various geophysical methods (Sherman et al., 1963). Most papers published in the Russian (Soviet) literature are based on the results of this early exploration campaign (e.g., Sherman, 1968;Distanov, 1977;Ponomarev and Akimtsev, 1981;Brovkov et al., 1985;Kuznetsov et al., 1990).
Mine development began in 1975, and by 1984 the mine had reached a production capacity of 0.2 Mt of ore per year. In 1991, construction of the first ore processing plant was completed, and from 1993 through 2008 the Gorevskoe mine produced and processed 0.4 Mt of ore annually.
Since 2008, production has gradually been increased to 2.5 Mt of ore per year (1.8 Mt of Pb-rich ore and 0.7 Mt of Pb-Zn-rich ore in 2017), making the Gorevsk Ore Mining and Processing Enterprise the largest active Pb-Zn-Ag mine in Russia. Construction of a new protective dam was recently completed to develop the open pit to an overall depth of 435 m below surface and allow the mine to continue operations for the foreseeable future (Fig. 1).
This contribution provides the first comprehensive Englishlanguage description of the Gorevskoe deposit. Its major aim is to develop a better understanding of the geotectonic context and genesis of the deposit based on a reinterpretation of the existing literature complemented by new observations.

Regional Geology
The Gorevskoe deposit is located on the western margin of the Siberian craton within the Central Angara terrane of the Yenisei Ridge, a Late Proterozoic orogenic belt (Fig. 2). It is by far the largest of more than 300 known occurrences of polymetallic sulfide mineralization in the Yenisei Ridge (Ponomarev et al., 1991b), and the only one that is currently mined.
The Siberian craton consists of several Archean to Early Proterozoic terranes that amalgamated during the Late Paleoproterozoic (Rosen and Turkina, 2007;Glebovitsky et al., 2008). In the Neoproterozoic, it was probably part of the Rodinia supercontinent ( Fig. 3; Li et al., 2008;Evans, 2009), which was assembled at ~1.1 Ga and subsequently broke up at ~750 to 700 Ma (Li et al., 2008). The Siberian craton is thought to have formed a northeastern to eastern promontory of Rodinia, being directly or indirectly connected to Laurentia Fig. 2. Location of Gorevskoe and other prominent stratiform and strata-bound Pb-Zn-Ag deposits along the borders of the Siberian craton. Modified after Distanov (1977), Lobanov and Nekos (2017). with its present southern margin (Li et al., 2008;Ernst et al., 2016;Fig. 3).
Up to the Early Neoproterozoic, the current western margin of the Siberian craton, then mostly on its northern side (Fig. 3), was not tectonically active. However, around 950 Ma the transformation to a subduction zone occurred (Priyatkina et al., 2018;Fig. 3). Subsequent rifting between the Siberian craton and northern Laurentia during the break-up of Rodinia at ∼800 to 700 Ma resulted in their separation, so that from about 720 Ma the Siberian craton evolved as an independent landmass (Pisarevsky et al., 2013;Pavlov et al., 2015). Subduction along the current western margin of Siberia in the Neoproterozoic led to the formation of an accretionary orogenic belt that is now referred to as the Yenisei Ridge (Vernikovsky et al., 2003;Vernikovskaya and Vernikovsky, 2006;Kuzmichev and Sklyarov, 2016).
The Yenisei Ridge forms part of the western margin of the Siberian craton, extending for almost 700 km along the Yenisei River in a northwest-southeast direction (Fig. 4). It is subdivided into five lithotectonic terranes: Angara-Kan, East Angara, Central Angara, Isakovka, and Predivinsk (Vernikovsky et al., 2003). The oldest domains of the Yenisei Ridge are located in the Angara-Kan terrane preserving Paleoproterozoic crust of the Siberian craton basement, and the East and Central Angara terranes marked by a passive margin sequence of Meso-to Early Neoproterozoic age, which probably formed on the Siberian continental slope (Vernikovsky et al., 2009). The Isakovka and Predivinsk terranes, on the other hand, comprise Neoproterozoic volcano-sedimentary successions related to an island arc and ophiolite complex (Vernikovsky et al., 2003(Vernikovsky et al., , 2009.
The five terranes are separated from each other by large, steeply dipping regional thrust faults (Angara,Priyenisei,Ishimba,Tatarka and,Ankinov,Fig. 4) that primarily strike toward the northwest. These high-angle faults are commonly accompanied by higher order splays and collisional structures of smaller blocks (Egorov, 2004). The nearly E-W-striking Angara fault divides the Yenisei Ridge into the South-Yenisei and Transangaria segments. The Priyenisei and Tatarka-Ishimba fault systems that bound the Central Angara terrane to the east and west, respectively, are the largest thrust zones in the region (Vernikovsky et al., 2003(Vernikovsky et al., , 2011.

Local Geology
The Gorevskoe deposit is located near the southern margin of the Central Angara terrane close to the junction between the Angara and Priyenisei regional faults (Figs. 4,A1).

Lithostratigraphy
The host rocks of the Gorevskoe deposit are part of a thick Meso-to Early Neoproterozoic metasedimentary succession, comprising four major units, all of which occur in the immediate surroundings of the deposit: the Sukhopit, Tungusik, Kirgiteisk, and Shirokinsk Groups (Makarov et al., 2014;Figs. 5B, 6). The entire succession consists predominantly of deformed and metamorphosed marine siliciclastic and carbonate-rich rocks, with minor intercalated volcaniclastic rocks (Fig. 5B). The metamorphic grade of the rocks hosting the Gorevskoe deposit ranges from zeolite to lower greenschist facies (App. A, Fig. A2). Thin sediments of mid-Paleozoic to Cenozoic age overlie the Meso-to Neoproterozoic rocks (Fig. 5A).
Near the deposit, the Sukhopit Group (Postelnikov, 1980) consists of a ∼1.7-km-thick succession of mostly gray to greenish-gray phyllites and slates with minor interbeds of yellowish-gray carbonate-rich rocks and basaltic metavolcanics. The dark-gray to black carbon-bearing, locally calcareous slates of the Tungusik Group unconformably overlie the Sukhopit Group with a maximum thickness of ~0.7 km. Thus, the combined thickness of these two groups in the Gorevskoe area is much lower than elsewhere in the Central Angara terrane (8-10 km; Zuev et al., 2009).
The Kirgiteisk Group, which unconformably overlies the Tungusik Group consists of mixed siliciclastic and carbonaterich metasediments. Basal multicolored (bluish, brown, red, green) quartzitic, locally carbon-bearing slates of the Udorongsk Formation pass into pinkish-gray to brown conglomeratic metalimestones with associated lenses of metasandstones and then into minor black to light-gray carbon-bearing slate and massive metalimestones of the Lower Stepanovsk Formation. Increasingly finer grained, mostly siliciclastic rocks in the Middle and Upper Stepanovsk Formation occur above this carbonate-rich unit. Although dark-gray slates and greenishgray metasiltstones dominate the Middle Stepanovsk Formation, the Upper Stepanovsk Formation is dominated by greenish-to brownish-gray lithologies. Overall, the metasediments of the Kirgiteisk Group attain a thickness of ~ 2 km in the deposit area.  Li et al. (2008), Cawood et al. (2016), Priyatkina et al. (2018). Note that subduction around the northern and western sides of the Siberian craton is thought to have been initiated around this time. Until 950 Ma, these were passive margins.
The Shirokinsk Group, which hosts the Gorevskoe deposit and many other polymetallic sulfide occurrences in the region, is separated from the Kirgiteisk Group by another unconformity. It is subdivided into the older, carbonatedominated Gorevsk Formation (1,020 ± 70 Ma, Pb-Pb; Kuznetsov et al., 2019) and the younger, predominantly siliciclastic Sukhokhrebtinsk Formation (Fig. 5B). Regional mapping at the 1:200,000 and 1:100,000 scales (Tselykovsky, 2002;Makarov et al., 2014) shows that the Gorevsk Formation with a total thickness of >2 km can in turn be subdivided into three units.
The Lower Gorevsk Formation, which unconformably overlies the Kirgiteisk Group, is characterized by a basal metaconglomerate containing abundant fragments of metamorphosed siliciclastic rocks derived from the older metasedimentary basement of the Central Angara and the East Angara terranes.  Vernikovsky et al. (2016) and Vernikovsky and Vernikovskaya (2006). A more detailed intermediate-scale geologic map is provided in Appendix 1 (Fig. A1 This is overlain by gray to black carbonaceous metamarlstones and calcareous slates. The Middle Gorevsk Formation consists of abundant lightto dark-gray metamarlstones, dolomitic metalimestones, and calcareous slates, many of which are carbon-bearing (Zuev et al., 2009). The dark-gray laminated metalimestones and calcareous slates of this unit host sulfide mineralization at Gorevskoe and Rudakovskoe .
The Upper Gorevsk Formation comprises crossbedded and laminated dolomitic metalimestone and metadolostone, intercalated with thin basaltic tuff horizons. The laminated dolomitic metalimestones host the Kartichnoe massive sulfide occurrence .
The Sukhokhrebtinsk Formation conformably overlies the Gorevsk Formation. It is the youngest member of the Meso-to Neoproterozoic succession in the region and is markedly different in lithologic composition from the ore-bearing Gorevsk Formation. Carbonate rocks are virtually absent. Dark-gray to black carbonaceous slate, variably colored metasandstone, metasiltstone, and mafic metavolcanic and metavolcaniclastic rocks define the Lower Sukhokhrebtinsk Formation. The Upper Sukhokhrebtinsk Formation has a similar composition but lacks metavolcanic and metavolcaniclastic components.

Igneous rocks
In addition to the metavolcanic and metavolcaniclastic rocks occurring in the Upper Gorevsk and Lower Sukhokhrebtinsk Formations (Fig. 5B), swarms of doleritic dikes and sills are present throughout the Proterozoic lithologies hosting the Gorevskoe deposit (Figs. 6-8; Kuznetsov et al., 1990). They generally form steeply dipping lens-like bodies up to hundreds of meters in lateral extent and 0.1 to 20 m in width, with a predominantly northwest strike. Butan (1997) reported the major element composition of the least altered dolerites near Gorevskoe to be mostly basaltic (tholeiitic, with minor picrobasalts and trachytes, App. A, Fig. A3). It is likely that most of these dikes and sills are related to the Stepanovsk Igneous Complex occurring about 20 km to the north of the deposit Tselykovsky, 2002;Zuev et al., 2009).
Based on geologic observations in the open pit and from drill cores, Strimzha (2017) suggested that there are dikes of pre-, syn-and postmineralization age. However, this observation must be considered with caution due to the strong deformational overprint of the deposit. Sherman (1971) cited a K-Ar age of ~915 Ma for some of the least altered dolerite sills near the deposit. This age falls outside the age range of the Gorevsk Formation (1,020 ± 70 Ma; Kuznetsov et al., 2019) and therefore suggests that the dolerites could be substantially younger than the mafic volcanic and volcaniclastic rocks present in the Upper Gorevsk and Lower Sukhokhrebtinsk Formations. However, Sherman (1971) does not report any uncertainties for the ~915 Ma age, and the available evidence on the relative age of the dolerites and host rocks therefore remains equivocal.

Structural setting
The Gorevskoe deposit is located in the northeast limb of the NW-striking (305°-310°) Gorevsk syncline, which forms part of the major regional fold system (Fig. 7). The axial plane of this structure dips steeply toward the southwest (75°-85°). Minor later folds overprint the Gorevsk syncline (Makarov et al., 2014). Local structures in the deposit are further characterized by a complex mosaic array of smaller fault blocks created by the intersection of the two major regional fault systems ( Fig. 7): (A) the NW-trending thrust and normal faults, and (B) the N-to NE-trending mixed thrust and transform faults that offset faults of group A (Makarov et al., 2014, Fig. A1).
Two major normal faults bound the mineralization toward the southwest and northeast (Fig. 8). The spacing between these two structures varies from 150 m in the northwest part of the deposit to 350 m in the southeast part. Although the total downthrow in the hanging wall of the southwest fault is >800 m (Fig. 8), the total amount of vertical movement on the northeast fault is not well constrained.
Located between two major normal faults, the deposit is strongly deformed. The intensity of this deformation increases toward the northwest as the spacing between the bounding faults decreases. Brecciation, cataclasis, mylonitization, and deformation-induced foliation and boudins at different scales are commonly observed. Secondary faulting and folding are also widely developed. The secondary faults are mostly bedding parallel, dipping steeply toward the southwest (70°-80°) and are characterized by estimated along-dip displacements of 20 to 50 m (Makarov et al., 2014).
One of the largest group B faults displaces the Main and Western orebodies on its east side from the Northwestern orebody on its west side (Fig. 8). The NE-striking mixed thrust and transform faults of group B are generally steeply dipping (70°-80°) and show along-strike displacements of 50 to 120 m and downdip displacements of 50 to 70 m (Makarov et al., 2014). They are accompanied by up to 70-m-wide zones of deformation and low-amplitude en echelon fault arrays (Makarov et al., 2014). Cataclasis-dominated fault cores, 4 to 15 m wide, are common and generally consist of millimeter-to centimtersize angular fragments of shale, quartz, carbonates, and sulfides. Even though these breccia bodies have previously been described as hydrothermal-explosive breccias by Strimzha (2017), they postdate the formation of massive sulfide ore.

Mineralization
The polymetallic sulfide mineralization at Gorevskoe occurs as three tabular strata-bound bodies in a 10-to 300-m-wide and more than 2,200-m-long zone that strikes to the northnorthwest and dips steeply to the southwest (70°-85°). Beyond the limits of the Gorevskoe deposit, minor occurrences of polymetallic sulfide mineralization exist at Kartichnoe and Rudakovskoe (Figs. 6, 7;Makarov et al., 2014).
The three ore lenses are referred to as the Main (or Central), Western, and Northwestern orebodies (Sherman et al., 1963;Fig. 8) and consist of stacked subparallel lenses of wellmineralized material separated by weakly mineralized host rocks. The contacts between well-and poorly mineralized material are sharp to gradational and the grade distribution of Pb and Zn within individual ore lenses is irregular to patchy (Fig. 9).
The mining operation distinguishes between two major ore types: Pb-and Pb-Zn-rich ores. The distinction is made because these types are treated separately during beneficiation.  Makarov et al. (2014). Stage and period boundaries in the Proterozoic rocks were changed compared to Makarov et al. (2014) to be consistent with the most recent published age for the Gorevsk Formation (Kuznetsov et al., 2019). Note, however, that the exact positions of these boundaries are still not well constrained, as indicated by the broken lines. Their mean compositions are listed in Table 1. The Pb-rich ores make up about 64% of current reserves, while the Pb-Znrich ores constitute the balance. The Main orebody is 60 to 90 m wide ( Fig. 8) and extends to a depth of 1,200 m below surface. Below this depth, the orebody splits into a series of thin, stacked ore lenses with an overall thickness of 24.5 m that wedge out at depth. The Main orebody consists of Pb-rich (77%) and Pb-Zn-rich (23%) massive to semimassive sulfide ores and contains ~73% of the total ore reserves of the deposit (Makarov et al., 2014).
The Western orebody is located 30 to 100 m to the southwest of the Main orebody and is separated from it by units of hydrothermally altered host rocks (Fig. 8). It has a sheet-like appearance with a thickness of 1 to 43 m (avg ~18 m) and a strike length of 950 m at surface. It parallels the western outlines of the Main orebody with which it merges at depth (Fig. 8).
Although it only hosts about 9% of the total ore reserves, it contains relatively less Pb-rich (36%) and more Pb-Zn-rich ore (64%) than the Main orebody (Makarov et al., 2014).
The Northwestern orebody is located underneath the Angara River. At surface, the orebody appears as a group of   Sherman et al. (1963) and Makarov et al. (2014). Note that the footwall and hanging wall of the deposit are defined in both stratigraphic and structural terms (i.e., footwall is stratigraphically lower).  8). It contains both Pb-rich (29%) and Pb-Zn-rich ore (71%) and constitutes ~18% of the total reserves of the deposit (Makarov et al., 2014).

Alteration lithologies
Extensive sideritization, dolomitization, and silicification of the metalimestones and calcareous slates of the Middle Gorevsk Formation are associated with the massive sulfide ore lenses (Fig. 8). The alteration is systematically zoned away from the sulfide lenses, with the thickness of individual zones varying from 2 to 80 m. In most locations, this alteration has obliterated all primary sedimentary textures (Makarov et al., 2014).
Quartz-siderite and quartz-only assemblages dominate in direct contact with the sulfide ores. This is followed outward by siderite-only assemblages that account for 75 to 80% of the altered rock volume. With increasing distance from the sulfide lenses, first ankerite and then dolomite predominate as alteration products. The altered rocks generally have sharp contacts with the unaltered limestones ( Fig. 10B, C). Quartzcarbonate veins and lenses, oriented parallel to foliation are abundant in the ankerite-and dolomite-rich alteration zones (Fig. 10D). All altered lithologies are characterized by finegrained granoblastic textures of carbonates and contain a similar assemblage of accessory minerals (biotite, sericite, chlorite, tourmaline) to the unaltered limestones (Sherman et al., 1963).
Low-grade sulfide mineralization is abundant, especially in the siderite-only and siderite-quartz alteration zones of the hanging wall. It occurs mainly as fine disseminations, stringers, veinlets or irregular clusters of galena, sphalerite, and pyrrhotite ( Fig. 11C-F). Quartz-carbonate-sulfide veins locally crosscut the alteration lithologies. These are generally between 1 and 100 cm wide and postdate the major deformation and metamorphism (Figs. 10F, 11A, D). In the alteration halo, galena, sphalerite, and pyrrhotite are generally associated with milky quartz and sparry calcite (Figs. 12,13).

Mineralization styles
Semimassive to massive sulfide ores dominate. They are generally fine-grained and show strong evidence of extensive remobilization and deformation. In most locations, chaotic folding and brecciation have resulted in the complete destruction of primary textures. However, in some hand specimens compositional banding with a probable pre-metamorphic origin is still recognizable (Sherman et al., 1963).
Based on the degree and style of deformation, three major textural ore-types are distinguished: brecciated, folded, and banded ores (Sherman et al., 1963). However, any specific specimen usually shows features of several of these types, and to some degree the exact classification depends on the scale of observation. In particular, brecciation textures generally occur at the scale of larger blocks and do not appear at the scale of individual hand specimens. Even though the brecciated and folded ores are present across the deposit, their occurrence is rarely linked to any obvious large-scale geological structures.
Brecciated ores are widespread and occur in all orebodies. They are characterized by subangular to rounded (Fig. 12A, B) undeformed fragments of shale, silicified limestone, and quartz-carbonate rocks (with a predominance of quartz over carbonates) hosted in a uniformly fine-grained quartz-carbonate-sulfide matrix (5-50 vol %). Boundaries between the clasts and matrix are generally sharp, with some evidence for partial replacement or abrasion of clasts.
In cases where isolated clasts of quartz or quartz-rich hostrock occur in a sulfide-only matrix, they are generally well rounded, showing strong signs of abrasion (Fig. 12C, D). This feature is referred to as "ball-texture" by Russian authors (Kovalev, 1984;Kuznetsova, 2007). It indicates large degrees of plastic deformation in the sulfide matrix. This is also indicated by the contours of alternating layers of sulfides (e.g.,    Table 3 for mineral abbreviations. Chl = chlorite. Clasts of fine-grained limestones with pyrite mineralization are cemented by galena-pyrrhotite-quartz assemblage. C, D. "Ball-textured" quartz clasts in highly deformed massive sulfide ores. E, F. Folded ores: dense networks of galena-pyrrhotite veinlets within complexly deformed and brecciated host rock consisting of limestone (light-gray), shale (dark-gray), and megaquartz (white), note boudinage of quartz in (E). All samples taken from Main orebody. Mineral abbreviations as in Figure 11 and Table 3, Lst = limestone. pyrrhotite and galena) that produce flow bands and record the injection of sulfide-rich matrix into clasts of the more competent host-rocks (Fig. 12C, upper left).
The folded ores are characterized by a predominance of plastic deformation in the host-rocks. While the limestones and quartz-rich lithologies mostly behaved in a relatively brittle manner during deformation, preferentially forming brecciated textures and boudins, the calcareous shales deformed more plastically (Fig. 12E, F). In these ores, sulfides mostly occur as networks of veinlets within the host rock, or as matrix between larger host-rock fragments (Fig. 12E, F).
The banded ores are characterized by coarse, mostly bedding-or foliation-parallel layers (<0.1-5 cm) of fine-to medium-grained sulfides in the host rock (Fig. 13A, B). Boudinaging of the host rock layers has generally resulted in small-scale fragmentation, giving them a granular appearance. The total sulfide content of the banded ores usually varies between 50 and 70%.
Overall, the textural variability of the Gorevskoe ores is best explained by the interplay of differences in the rheology of the major ore constituents, variations in their relative proportions, and overall strain rates during deformation. Plastic-ity decreases in the order sulfides > shales >> limestones > quartz, resulting in the generally plastic behavior of the sulfides in all ore types, while the behavior of the other components depended on the relative proportions of different rock types. For instance, shales behaved plastically in shale-quartzdominated lithologies (folded ores) but behaved in a brittle manner in the sulfide-dominated ores (banded and brecciated ores). Depending on their composition, ores that have been affected by greater strain rates would generally be expected to show greater degrees of foliation and smaller grain sizes (Spry, 1969).

Ore mineralogy and metal zonation
The sulfidic ores are characterized by a relatively uniform mineral assemblage that does not vary substantially with mineralization style and spatial position in the deposit (Makarov et al., 2014). The mineralogy is summarized in Table 2 and illustrated by selected micrographs in Figure 14. The total sulfide content mostly varies between 20 and 25 vol % (semimassive ores) and 50 to 70 vol % (massive ores). Galena, sphalerite, and pyrrhotite are the dominant sulfides, with a generally greater abundance of galena than sphalerite (3:1-5:1). Pyrite   Table 3, Msq = muscovite. is generally absent, or only present in minor amounts, being most abundant in the peripheral parts of the Main and Western orebodies, as well as in the Northwestern orebody. Several other minor and accessory sulfide minerals such as tennantite, boulangerite, and pyrargyrite are also present (cf. Table 2). While these minerals are not important in volumetric terms, they are major hosts for various trace elements (e.g., Ag).
The main gangue minerals are carbonates (10-50 vol %) and quartz (20-60 vol %). Chlorite, sericite, and biotite commonly occur as minor constituents, while tourmaline, rutile, and to a lesser degree ilmenite and garnet occur as accessories.
Magnetite is rare in the Main and Western orebodies, except for isolated occurrences as disseminated grains and aggregates which are mostly related to the alteration haloes of some dolerite dikes (Makarov et al., 2014) and the metamorphic overprint (Kuznetsov et al., 1990). It is, however, more abundant in the Northwestern orebody, where it commonly occurs together with siderite and pyrite (in addition to pyrrhotite), and rarely forms separate quartz-magnetite-pyrite ± pyrrhotite ± siderite horizons (Kuznetsov et al., 1990;Ponomarev et al., 1991a;Makarov et al., 2014).
It is worth noting that in addition to the dominant sulfide ores, small volumes of oxidized ores (Figs. 8, 10E, 0.64% of the total resources, Makarov et al., 2014) are present in the shallower parts of the deposit as irregular ochre-colored bodies. Mineralogically, these bodies are dominated by iron oxides, cerussite and smithsonite replacing the sulfides.
Based on the relative proportions of the major ore minerals, four mineralogical types of sulfide ores have been distinguished (Ponomarev et al., 1991b): galena, pyrrhotite-galena, pyrrhotite, and galena-pyrrhotite-sphalerite ores. The classification of these types is somewhat arbitrary since there are no clear boundaries between them, and all possible transitional varieties exist within the mine. Nevertheless, they are useful for describing the compositional variability of the ores across the deposit.
While the distribution of ore types within individual orebodies is highly irregular, resulting in similarly irregular distribution patterns of the metal grades (Fig. 9), several broad compositional trends exist across the deposit. In general, the ratio of galena to sphalerite within the ores decreases from the footwall towards the hanging wall of the deposit, as well as updip within the ore zone (Kuznetsov et al., 1991). The relative abundance of pyrrhotite shows a similar trend to sphalerite. Thus, the footwall and lower portions of the Main Orebody are almost entirely composed of galena ores. The central parts of the Main Orebody are mostly composed of galena and pyrrhotite-galena ores, while galena-pyrrhotite-sphalerite ores predominate in the Western and Northwestern orebodies and occur in the hanging wall of the Main Orebody (Kuznetsov et al., 1991). In contrast to sphalerite, the abundance of pyrrhotite also increases down-dip, resulting in large masses of pyrrhotite ores in the lower horizons of the orebodies.

Geochemistry of Ores and Host Rocks
The following subsections give a brief overview of available data on the chemical and isotopic composition of the ores, as reported in the Russian literature and company reports (Sherman et al., 1963;Grinenko et al., 1984;Kuznetsov et al., 1990;Ponomarev et al., 1991a;Makarov et al., 2014).

Chemical composition
In conjunction with variations in the absolute and relative abundances of the different ore and gangue minerals, the sulfide ores at Gorevskoe show considerable variability in their elemental composition. This is illustrated by the data presented in Table 3, which are derived from the studies of Sherman et al. (1963) and Makarov et al. (2014). Note, however, that the means and ranges of values cited in this table refer to sample sets collected for research purposes. Therefore, reported values for Cd and Ag differ slightly from those cited in Table 1, which are based on the current block model of the entire deposit. Table 3 also includes information on the major host mineral(s) of each element where this is known.
The overall composition of the deposit is unusually Pb-rich and Cu-poor compared to typical SHMS and Mississippi Valley-type (MVT) Pb-Zn deposits (Figs. 15B, C). With a value of 3.4, the Pb:Zn ratio of the Gorevskoe ores is higher than 90% of all known SHMS deposits. In fact, the Pb-Zn-Cu composition of the ores is most similar to the sandstone-hosted Pb subgroup of MVT deposits, which typically have a Pb:Zn ratio > 1 (Fig. 15D).
At 49 g/t, the mean Ag content of the Gorevskoe ores is moderate, being somewhat higher than the median value for SHMS deposits (42 g/t; cf. data in Singer et al., 2009). Silver is mostly hosted by galena and a suite of silver minerals (Table 3), including sternbergite, argentite, and pyrargyrite (cf. Table 2).
The concentrations of other valuable (Ga, Ge, In) and deleterious (As, Bi, Cd, Sb, Se, Te, Tl) trace elements in the ores also fall within the general ranges expected for Pb-Zn ores (cf. Feiser, 1966;Schwarz-Schampera and Herzig, 2002;Cook et al., 2009;Frenzel et al., 2016). Unfortunately, a more detailed comparison is not possible at present since databases with representative ore compositions similar to those provided in Singer et al. (2009) for Pb, Zn, Cu, and Ag do not exist for these elements. Nevertheless, an interesting feature of the ores at Gorevskoe is the reported importance of quartz as a host of Ge, accounting for 85% of the total Ge content (Table 3). This is unusual, since sphalerite is generally considered to be the most important host for this element in Pb-Zn ores (e.g., Bernstein, 1985;Frenzel et al., 2014).

Sulfur isotopes
Sulfur isotope studies of mineral concentrates from various ore types have yielded remarkably uniform δ 34 S values be-  (Grinenko et al., 1984; Table 4), with an overall median value of +19 ‰, which falls within the range of Late Mesoproterozoic to Early Neoproterozoic seawater sulfate (δ 34 S = 10 -24 ‰; Chu et al., 2007;Guo et al., 2015;Fakhraee et al., 2019). Within reported analytical uncertainties, the observed sulfur isotope values appear to be mostly independent of mineral, mineral association, and ore type (Grinenko et al., 1984). However, a trend towards isotopically heavier values is observed in the peripheral parts of the orebodies (Grinenko et al., 1984). Sulfides in the syn-to postmetamorphic veins have systematically lighter sulfur isotopes than their counterparts in the semimassive to massive ores, yielding mean δ 34 S values between +11.9 and +16.6 ‰ (Table 4). Sulfur isotope analyses were also obtained from unmineralized hostrocks of the Gorevsk Formation, as well as isolated samples of crosscutting dolerite dikes. These gave δ 34 S values of +13.0 (±4.2) ‰ and +2.7 (±1.0) ‰, respectively (Grinenko et al., 1984). Table 5 summarizes the results of carbon and oxygen isotopic analyses of carbonate minerals in unaltered and altered host rocks within and around the Gorevskoe orebodies. The mean δ 13 C values of most lithologies fall within the range of -2 to +1 ‰, typical of Proterozoic marine carbonate rocks (Schidlowski et al., 1975;Chu et al., 2007). Only siderite from the siderite-quartz altered rocks has a significantly lighter isotopic signature with a mean δ 13 C value of −7.1 (±1.4) ‰.

Carbon and oxygen isotopes
Oxygen isotope values show two groupings. While the unaltered limestone proximal and distal to the sulfide ores, as well as siderite from the siderite-quartz zone show lighter mean δ 18 O values of around +16 to +17 ‰, dolomite and ankerite from the more distal alteration zones show values of around +20 ‰. These differences are statistically significant. However, all measured δ 18 O values are still within the range expected for Proterozoic marine carbonates (Schidlowski et al., 1975).

Lead isotopes
There are no major differences between the 206 Pb/ 204 Pb and 207 Pb/ 204 Pb ratios of the ores and host rocks reported by Grinenko et al. (1984) and Kuznetsov et al. (1990) (Table 6).
Only the 208 Pb/ 204 Pb ratio is somewhat higher in the ores. In general, the Pb isotope composition of the ore is relatively homogeneous and significant lateral or vertical changes in lead isotopes are not observed within the deposit (Kuznetsov et al., 1990). Based on a comparison with different models for the evolution of lead isotopes in the continental crust, the measured signatures permit a derivation of the lead from a crustal source at the time of formation of the host rocks at 1,020 ± 70 Ma (cf. Appendix A).

Discussion
Tectonic setting and depositional environment of the host rocks to ore Given their age, the Meso-to Early Neoproterozoic sedimentary rocks around the Gorevskoe deposit were probably laid down along the passive northern margin of the Siberian Craton, up to and perhaps overlapping with, its transformation into a subduction zone at ~950 Ma (Likhanov et al., 2014;Priyatkina et al., 2018;Kuznetsov et al., 2019). Based on a comparison with modern and Proterozoic sedimentary environments (Reading, 1996;Grotzinger and James, 2000), we further interpret the Kirgiteisk and Shirokinsk Groups, including the immediate host rocks of the mineralization, to record deposition in a carbonate ramp environment (Wright and Burchette, 1996). Lithological variations within the two groups are likely due to changes in relative sea-level and sediment input. Fine-grained, commonly carbonaceous, siliciclastic rocks, marlstones and laminated limestones record the deep-water environments of the outer ramp, with terrigenous material derived either from aeolian input, or through long-shore transport (Wright and Burchette, 1996). Massive and crossbedded limestones of the Upper Gorevsk Formation record mid to inner ramp conditions, and therefore document a shallower water depositional environment. The siliciclasticdominated units (siltstones, sandstones) of the Upper Stepanovsk and Sukhokhrebtinsk Formations probably represent transient increases in terrigenous input to the basin. In the Sukhokhrebtinsk Formation, this appears to be related to rifting as indicated by the presence of contemporaneous mafic volcanic and volcaniclastic rocks. Extension associated with incipient subduction prior to ~950 Ma is the most likely cause of this rifting event (cf. Priyatkina et al., 2018).

Relative and absolute ages of mineralization and host rocks
The presence of extensive Fe-Mg-Mn-carbonate alteration around the sulfide lenses strongly suggests that ore-formation Table 3. Chemical Composition of Gorevskoe Ores (Sherman et al., 1963;Makarov et al., 2014) Sherman et al. (1963) Notes: Host mineral column left blank where unknown: -= number not provided in original reference 1 Mineral abbreviations: ank = ankerite, aspy = arsenopyrite, boul = boulangerite, bour = bournonite, cb = carbonates, cc = calcite, cer = cerussite, cpy = chalcopyrite, dol = dolomite, gn = galena, ilm = ilmenite, mag = magnetite, po = pyrrhotite, py = pyrite, qtz = quartz, ru = rutile, sid = siderite, smi = smithsonite, sph = sphalerite, tetr = tetrahedrite, tn = tennantite; note that cer and smi only occur in oxidized ores and are therefore not included in Table 2 2 Traces, generally 1 µg/g or less occurred in a diagenetic environment. It is inconsistent with exhalative ore-formation since this would have resulted in extensive alteration of the footwall only. Similarly, it is inconsistent with epigenetic ore-formation, since the low permeability of the fine-grained host-rocks after lithification would not have allowed widespread penetration and circulation of the ore-forming fluids and the associated alteration. We further note that a diagenetic age is consistent with the formation of most well-preserved SHMS deposits elsewhere in the world (Cooke et al., 2000;Leach et al., 2005;Magnall et al. 2016Magnall et al. , 2018, with which Gorevskoe shares many geological similarities. Accepting that the ores formed diagenetically, they should have approximately the same absolute age as the host rocks, i.e., 1,020 ± 70 Ma (Kuznetsov et al., 2019). This just overlaps with the previously published Pb model age of 850 ± 100 Ma (Ponomarev et al., 1991a) for the deposit. However, we note that Pb model ages generally have large uncertainties, as demonstrated by a re-evaluation of the Pb isotope data available for Gorevskoe in Appendix A using different models for the evolution of the continental crust (section A4). This showed that the uncertainty of ±100 Ma cited by Ponomarev et al. (1991a) is probably understated by at least a factor two. Therefore, we prefer the more reliable direct Pb-Pb age of Kuznetsov et al. (2019) for the deposit.

Relationship to magmatic activity
The presence of mafic volcanic and volcaniclastic rocks in the Upper Gorevsk and Lower Sukhokhrebtinsk Formations suggests that distal magmatic activity coincided with ore formation. However, while the dolerite sills and dikes in the immediate surroundings of the deposit have been described as showing evidence for pre, syn-and post-mineralization emplacement (Strimzha, 2017), such a close association is not unequivocal.
Nevertheless, the available data point towards a temporal association between distal mafic magmatism and mineralization. It is interesting to note in this context that an association with magmatism is a feature of some sediment-hosted Pb-Zn deposits (Leach et al., 2005;Emsbo et al., 2016). For example, temporal associations with mafic rocks like those associated with the Gorevskoe deposit have been described for several Selwyn-type deposits (Cooke et al., 2000), including Rammelsberg (Large and Walcher, 1999) and Sullivan (Lydon, 2004). A broad temporal association with distal mafic magmatism has also been described for the Zn-Pb deposits of the Irish Midlands (Wilkinson and Hitzman, 2015). Generally, these associations are thought to reflect the extensional tectonic setting in which the deposits formed rather than a direct genetic link to magmatic activity (Leach et al., 2005;Emsbo et al., 2016). That is, crustal thinning caused both magma generation and the regionally elevated geothermal gradients necessary for fluid circulation and ore formation. This also seems to be the case at Gorevskoe. The relatively small volume of magmatic rocks present in the vicinity of the deposit makes it unlikely that they could have acted as a major heat source.

Conditions of ore formation
Despite the strong deformational overprint, the available mineralogical, geologic, and geochemical data allow us to derive some important constraints on the physico-chemical conditions of ore formation at Gorevskoe.
First, the alteration halo of the deposit does not contain any calc-silicates or other silicate minerals such as andradite garnet or pyroxenes typical of high-temperature (300°-500°C) carbonate replacement deposits (Meinert et al., 2005). Instead, it has a simple overall mineralogy consisting mostly of quartz and various Fe-Mg-Ca-carbonates. This is similar to many classic SHMS deposits, such as the Tom and Jason deposits (Cooke et al., 2000;Magnall et al., 2016), Rammelsberg (Large and Walcher, 1999), Lady Loretta (Large and McGoldrick, 1998), Century (Broadbent et al., 1998), and McArthur River . It points toward a low to moderate temperature of ore formation (100°-300°C; cf. Cooke et al., 2000;Magnall et al., 2016). The chemical composition of the ores, with a predominance of Ge and Ga over In, is also typical of low-to moderate-temperature Pb-Zn ores (cf. Frenzel et al., 2016).
Second, the abundance of pyrrhotite and siderite, as well as the relative scarcity of pyrite in most of the deposit indicates Ankerite (peripheral to ore bodies) 4 -2.0 (±0.9) +19.5 (±0.7) Siderite in sideritequartz alteration zone 9 -7.1 (±1.4) +15.9 (±3.1) Notes: PDB = Pee Dee Belemnite (Brand et al., 2014); SMOW = Standard Mean Ocean Water (Brand et al., 2014)  that ore formation occurred under low fO 2 and low fS 2 conditions (Berner, 1964;Toulmin and Barton, 1964;Cooke et al., 2000;Magnall et al., 2016). The coexistence of pyrite and pyrrhotite, as well as magnetite, in some ores allows us to constrain the maximum values for both fO 2 and fS 2 , assuming equilibrium between these minerals during ore formation. This assumption is discussed further below. The values represent maximum estimates, because many parts of the deposit only contain pyrrhotite (and siderite) without pyrite or magnetite and must therefore have formed at lower f S 2 and fO 2 .
Using the equations of Kishima (1989) for the pyrite-pyrrhotite-magnetite buffer which fixes both fS 2 and fO 2 , and assuming a temperature of 250°C and pressure of 250 bars, the maximum log fO 2 value we derive for ore formation is -38.1, while the maximum log fS 2 value is -13.4. We chose these P-T conditions to be able to compare the estimated fO 2 value with the phase diagrams of Cooke et al. (2000) and Magnall et al. (2016) that are commonly used in the description of SHMS systems (Fig. 16). These diagrams show the stability regions of different Fe minerals in equilibrium with a typical 250°C basinal brine, as well as solubility contours for Pb, Zn, and Ba.
As Figure 16 shows, pyrrhotite(-siderite)-dominant assemblages are only stable at log fO 2 < -37 in this model system. This is in good agreement with the maximum fO 2 value estimated from the pyrite-pyrrhotite-magnetite buffer above and identifies the conditions at Gorevskoe as highly reducing in the sense of Cooke et al. (2000), i.e., very far into the field where reduced sulfur species predominate in the fluid.
Ore formation under low fO 2 conditions is also supported by the relatively homogeneous sulfur isotope composition of the ores across the deposit ( Table 4). As Ohmoto (1972) showed, appreciable sulfur isotope fractionation between fluid and sulfide minerals, which could cause strong spatial fractionation, is only expected to occur above log fO 2 values of -38 to -35 at 250°C, depending on pH.
Finally, we note that the pyrrhotite stability field contracts with decreasing temperature and falls outside of the Pb-Zn transport window below ~200°C (cf. Cooke et al., 2000, figs. 3, 4). This provides a lower limit on the formation temperature of the deposit.
These inferences rely on the interpretation of pyrrhotite and associated Fe minerals as primary minerals. While pyrrhotite is often a product of metamorphic processes in massive sulfide deposits (Craig and Vokes, 1993), the low metamorphic grade and comparative lack of preserved pyrite-rich  Kuznetsov (1990) Kuznetsov et al. (1990) (Grinenko et al., 1984) Fig. 16. Log fO 2 -pH diagrams relevant to Selwyn-type ore-forming systems, showing the stability fields for Fe sulfides, oxides, and siderite, as well as solubility contours for Pb, Zn, and Ba under the conditions proposed in Cooke et al. (2000). Panel (A) is adapted from the original diagram in Cooke et al. (2000), whereas panel (B) is adapted from Magnall et al. (2016) with higher total carbon but otherwise the same parameters, showing the appearance of stable siderite between the pyrite and pyrrhotite stability fields. Pressure for both diagrams is 250 bars. Ore-forming conditions at Gorevskoe are constrained by the mineralogy of its ores (pyrrhotite-siderite) to lie within the solid red boxes. Possible conditions in the northwest orebody, where primary magnetite may occur in addition to pyrrhotite, siderite, and pyrite, are also indicated. The most likely chemistry of the corresponding ore-forming fluids is also indicated. It must clearly have been of Selwyn type. Note that Ba contours in (B) were modified from Cooke et al. (2000) to reflect higher total carbon values.

A) B)
ores at Gorevskoe make such a scenario unlikely. Extensive pyrite-pyrrhotite conversion is only observed in deposits that experienced much higher metamorphic grades than Gorevskoe (amphibolite and granulite facies; Craig and Vokes, 1993). Pyrite in massive sulfide deposits metamorphosed to upper zeolite or lower greenschist facies, like Gorevskoe, does not usually convert to pyrrhotite (e.g. Neves Corvo, Relvas et al., 2006;Frenzel et al., 2019;Rammelsberg, Large and Walcher, 1999;Century, Broadbent et al., 1998). Therefore, the abundance of pyrrhotite at Gorevskoe very likely reflects a primary feature of the mineralization. The same is probably true for the quartz-magnetite-pyrite horizons reported from the Northwestern orebody, even though their exact mode of formation is not well constrained (Makarov et al., 2014). So far, they have only been reported from drill core. For disseminated pyrite and magnetite, a primary origin is unclear. However, we note that the assumption of equilibrium between pyrite, pyrrhotite, and magnetite for the ore assemblage constrains the maximum fO 2 value. If pyrite and magnetite do not reflect primary features of the mineralization but are instead due to secondary overprints, then the f O 2 value must necessarily be lower than this maximum.
Note in this context that abundant primary pyrrhotite also occurs in several other massive sulfide deposits (e.g., Nicholas-Denys (VHMS/SHMS) in the Bathurst mining camp, Canada (Deakin et al., 2015), and Draa Safaar (VHMS) in Morocco (Marcoux et al., 2008;Moreno et al., 2008)). These examples appear to have formed under similar conditions as Gorevskoe (moderate temperature, low fO 2 , and sulfur-deficient conditions). Figure 16 indicates that the ore fluids were reduced brines in the field labeled "Selwyn-type ore fluids" (cf. Cooke et al., 2000). Fluids from the field "oxidized brines," while able to carry enough Pb and Zn, would not be able to reach the indicated ore-forming environment without precipitating their metal load.

Character of the metal-bearing fluid
Accepting these arguments, the ore fluid associated with Gorevskoe appears to resemble a Selwyn-type ore-fluid as proposed by Cooke et al. (2000). An alternative, a McArthur River-type fluid would be oxidizing and should have formed at a lower temperature (~150°C; Cooke et al., 2000) than the available evidence permits. We also note that the nature of the underlying basin fill is in agreement with a Selwyn-type setting rather than a McArthur River-type one (cf. App A, sec. A3). Furthermore, the fluid must have been relatively sulfurpoor, since it would not otherwise have been able to transport sufficient Pb and Zn (cf. Emsbo et al., 1999;Emsbo, 2000).
Despite the constraints on the reducing character and moderate temperature of the ore fluid, Gorevskoe shows two important characteristics that are not typical for formation from a Selwyn-type fluid: (1) the extensive Fe-Mg-Mn-carbonate alteration haloes surrounding the ore lenses, and (2) the apparent absence of barite and/or witherite. Both would be more typical of a McArthur River-type deposit (Cooke et al., 2000). However, they may also be explained by the specific features of the Gorevskoe deposit.
First, the Fe-Mg-Mn-carbonate halo can be explained by ore formation in a carbonate-rich host rock rather than the si-liciclastic, carbonate-poor host lithologies generally associated with Selwyn-type deposits (cf. Cooke et al., 2000). As shown by Magnall et al. (2016) (cf. Fig. 16B), even a slight increase in the availability of carbonate in the ore-forming environment can stabilize siderite under reducing conditions, a feature not reported by Cooke et al. (2000).
Second, the absence of Ba minerals can be explained by a lack of Ba in the ore-forming fluids. As Figure 16 shows, the probable pH-fO 2 field for ore-forming conditions at Gorevskoe crosses the Ba solubility contours. While Ba would have been mobile in the ore-forming environment at log fO 2 values below -39 to -40 (see Fig. 16), the pyrite-magnetite-quartz assemblages in the Northwestern orebody, if primary, can only have formed beyond the solubility limits of Ba. Therefore, if appreciable Ba had been present in the ore-forming fluids it should have precipitated in association with the Northwestern orebody.
Following the arguments of Emsbo (2000), the absence of Ba is atypical of a reducing SHMS ore fluid. However, we note that recent work on the Tom and Jason deposits, both classic examples of Selwyn-type mineralization (Cooke et al., 2000), has shown that the formation of Ba minerals preceded Pb-Zn mineralization, representing a distinct diagenetic event (Magnall et al., 2020). This suggests that the Pb-Zn mineralizing fluids in these deposits may also have been poor in Ba. The absence of significant Ba at Gorevskoe could then be due to a lack of diagenetic Ba fixation from seawater during the diagenesis of the host rocks, rather than the ore fluid having an atypical composition.
Thus, there is no contradiction between these apparently atypical features and the formation of the deposit by a Selwyntype fluid. Particularly the presence of Fe-Mg-Mn-carbonate alteration appears to be inconclusive evidence for oxidized ore fluids as previously suggested (cf. Cooke et al., 2000). Instead, it can be produced from oxidizing and reducing fluids. Cooke et al. (2000) identified several possibilities to cause metal precipitation from a Selwyn-type fluid: temperature decrease (± reduction), pH increase, and the addition of (reduced) sulfur by fluid mixing. All of these are likely to occur at a typical trap site (Cooke et al., 2000;Magnall et al., 2016). However, by far the most efficient process for the precipitation of Pb and Zn is the addition of reduced sulfur (Cooke et al., 2000;Magnall et al., 2016).

Precipitation mechanism
At Gorevskoe, all these processes probably operated in concert to precipitate Pb-Zn mineralization. Buffering by carbonates in the host rock could have caused an increase in fluid pH, while reaction with organic material could have caused an increase in fS 2 . Significant cooling of the ore fluids would also have occurred due to proximity of the site of ore formation to the sediment-water interface (cf. Magnall et al., 2016). Finally, the advection of seawater into the ore-forming system combined with thermochemical sulfur reduction, aided by the oxidation of organic matter in the host rocks, may have introduced additional sulfur (cf. Magnall et al., 2016).

Sources of sulfur and metals
Lead and sulfur isotope compositions provide some constraints on the sources of Pb and sulfur. In particular, the Pb isotope composition of the deposit is compatible with derivation from an evolved crustal source at the time of formation of the host rocks (App. A), whereas sulfur isotope compositions are indistinguishable from Stenian-Tonian boundary seawater (Chu et al., 2007;Guo et al., 2015). Therefore, the most likely scenario is that Pb, as well as the other metals, were leached from the continental crust underlying the deposit, while sulfur was derived from contemporaneous seawater.
However, the deposit is unusually Pb-rich and Cu-poor when compared to typical SHMS deposits, with which Gorevskoe shares many other geologic characteristics (Table 7). The global median Pb/Zn ratio for SHMS deposits is 0.44 (Fig.  15B), only slightly higher than the crustal average of 0.20 to 0.25 (Rudnick and Gao, 2003). This compares with a value of 3.4 for Gorevskoe.
The unusual Pb enrichment at Gorevskoe could reflect either a source signature or fractionation of Pb from Zn during leaching, transport, and/or precipitation of the metals by the ore-forming fluids. Since the hydrothermal fluid system that formed the deposit likely leached a very large volume of crustal rocks (cf. Leach et al., 2005), the mean composition of the source region of the metals would not be expected to be considerably different from average continental crust (cf. Stacey and Kramers, 1975;Leach et al., 2005). Therefore, it is likely that fractionation played a significant role in generating the high Pb/Zn ratio. This fractionation most likely occurred during leaching and/or precipitation of the metals. Yardley (2005) showed that elevated Pb/Zn ratios relative to the crustal average may occur in low-to moderate-temperature crustal fluids. Leaching experiments have also shown that Pb can be more easily mobilized than Zn from most sedimentary rocks (Long and Angino, 1982;Lydon, 2015). However, the observed enrichments in these cases are only slight and cannot account for the high Pb/Zn ratio seen at Gorevskoe. Therefore, it is likely that the Pb/Zn ratio of the ore fluids was increased further by preferential precipitation of galena during ore formation. Figure 16 shows that galena is less soluble than sphalerite and should therefore precipitate first when the fluid enters the ore-forming environment (as indicated by the positions of the solubility contours for Pb and Zn). Evidence for this is preserved by the metal zonation of the Gorevskoe deposit, with Pb/Zn ratios decreasing stratigraphically upward. If indeed precipitation of the metal load from the ore-forming fluid was incomplete, the lower solubility and preferential precipitation of galena should have resulted in the formation of a Pb-enriched deposit.
Similar fractionation processes are thought to have operated in the formation of sandstone-hosted Pb deposits, which Table 7. Classification of Gorevskoe According to Criteria in Cooke et al. (2000) and Leach et al. (2005)   where only SHMS is listed, this means the subtype indication is not unambiguous 2 Feature affected by metamorphism and deformation; may have been altered significantly in the process 3 Statistics of median values for individual deposits as summarized in Leach et al. (2005) Notes: Mineral abbreviations are identical to those used in Table 3 and Figure 1 show similarly high Pb/Zn ratios as Gorevskoe ( Fig. 15D; Bjørlykke and Sangster, 1981;Bjørlykke and Thorpe, 1982). We also note that most Broken Hill-type deposits show Pb/Zn ratios >1 (Spry et al., 2009). However, the origins of this class of deposits remain enigmatic since all known examples have been affected by high-grade metamorphic overprints (Large, 2003;Emsbo et al., 2016).

Classification
Drawing on the empirical classification scheme introduced by Leach et al. (2005) for sediment-hosted Pb-Zn deposits, Table 7 summarizes the major features of Gorevskoe and indicates whether they favor its classification as MVT, SHMS, or both. With reference to the work of Cooke et al. (2000), the SHMS category is further subdivided into McArthur River and Selwyn-type deposits.
Within the SHMS class, Gorevskoe shows ambiguous characteristics that point to both subtypes. While its association with mafic magmatism and the setting in a rifted continental margin would be more typical of a Selwyn-type deposit, the carbonate-rich nature of the host rocks, apparent absence of barite, and extensive Fe-Mg-Mn-carbonate alteration halo point toward an affinity with McArthur River-type deposits.
However, we note that the key distinguishing characteristic between Selwyn-and McArthur River-type deposits according to Cooke et al. (2000) is the nature of the ore fluids. There is strong evidence for Gorevskoe having formed from reduced, sulfur-poor ore fluids with close affinities to the Selwyn-type fluids defined by Cooke et al. (2000). Thus, Gorevskoe should be included with the Selwyn-type deposits, rather than those of McArthur River-type.

Summary and Conclusions
This contribution provides the first English-language account of the geologic, geochemical, and mineralogical data available on the giant Gorevskoe Pb-Zn deposit. Overall, the available evidence suggests that Gorevskoe is an SHMS deposit with affinities to Selwyn-type mineralization that formed during the rifting of a passive margin sequence. Its subseafloor, diagenetic formation probably occurred at low to moderate temperatures in a highly reducing, sulfur-deficient environment that was accompanied by extensive Fe-Mg-Mn-carbonate alteration of the host limestones. The metals were likely derived from a crustal source, while sulfur appears to have been derived from contemporaneous seawater. Sulfide precipitation from the reduced ore fluid was probably caused by a combination of cooling, pH increase, reduction, and possibly addition of sulfur from advected seawater.
Several important features of the deposit are somewhat atypical of classic Selwyn-type SHMS deposits. These are (1) the extensive Fe-Mg-Mn-carbonate alteration halo surrounding the ore lenses, (2) the abundance of primary pyrrhotite, and (3) the absence of barite. However, these features are consistent with a reducing, carbonate-rich, ore-forming environment that was produced from a reducing, Ba-poor, ore fluid with a temperature of 200° to 300°C.
Gorevskoe has a Pb/Zn ratio (3.4/1) that is higher than 90% of all other SHMS deposits. However, the origin of this feature is not well constrained. It could reflect either an enriched source or, more probably, fractionation of Pb from Zn during metal leaching, transport, and precipitation.