Saskatchewan Geological Society Special Publication Number 14: MINEXPO'96 SYMPOSIUM - Advances in Saskatchewan Geology and Mineral Exploration, Proceedings of a Symposium, Saskatoon, Saskatchewan 21 - 22 November, 1996. Editors: K.E. Ashton and C.T. Harper, 1999

New Ideas on the Classification, Age, Interpretation, and Tectonic History of the Precambrian Shield in Saskatchewan, Pages 44-54.

The Anglo-Rouyn Deposit, La Ronge, Saskatchewan: A Besshi-type Massive Sulphide Deposit

R. Gwilym Roberts
Departrnent of Geological Sciences, University of Waterloo, Waterloo, ON N2L 3Gl

R.O. Maxeiner
Saskatchewan Geological Survey, Saskatchewan Energy and Mines, 2101 Scarth Street, Regina, SK S4P 3V7

A. Legault
Department of Geology, University of Regina, Regina, SK S4S 0A2

The Anglo-Rouyn deposit is on the northwest shore of Lac La Ronge, approximately 35 km northeast of La Ronge, Saskatchewan. Between 1966 and 1972, 1.7 million tonnes of ore with an average grade of 1.7 percent Cu, 1.2 g/t Au, and 5.5 g/t Ag were extracted from three orebodies of chalcopyrite, pyrite, pyrrhotite, magnetite, and trace amounts of sphalerite. The property consists of: 1) a northeast-trending, northwest-facing sequence of metabasaltic flows and tuffs, accompanied by greywacke-type metasediments; 2) overlying footwall rocks of calc-silicate gneiss, hornblende-plagioclase gneiss and laterally equivalent, gneissic magnetite iron formation; and 3) the ore zone rocks, which consist of the host quartz-hornblende-plagioclase gneiss and quartz-plagioclase gneiss in the hanging wall.

The rocks of the area have undergone three episodes of deformation. The first of these is the most significant, since it established the structural architecture of the region and coincided with the regional, amphibolite fades metamorphism. The units of the mine area are on the northwest limb of an antiform that closes to the southeast. Both regional- and property-scale structural evidence indicate that the units face, as well as dip, to the northwest. The Jepson Lake granite, a regional-scale, sill-like body, was intruded into the hanging-wall units before or during deformation.

Major and trace element whole rock geochemistry was used in association with the geological data to determine the protoliths of the gneissic rocks associated with the ore. The stratigraphic succession in the mine area records the history of a volcano-sedimentary package in a volcanic arc environment as it evolved from mafic volcanism and siliceous greywacke-type sedimentation into a basin in which both clastic and exhalative, chemical sediments were deposited. The exhalative component in the footwall rocks changed from carbonate-silica in the sediments overlying the mafic volcanics to iron oxide-carbonate in the sediments underlying the ore zone. The clastic component is similar in each unit and was derived from the underlying mafic volcanics, probably by a combination of near- surface, submarine alteration and sedimentary processes. The cupriferous sulphides of the Anglo-Rouyn ore immediately overlie the magnetite iron formation, and the maximum development of sulphides in the basin appears to be at the fringes of the iron oxide-rich sediment (the magnetite iron formation). The units of the ore zone (quartz- hornblende-plagioclase gneiss and quartz-plagioclase gneiss) signify a stage in the development of the volcanic- sedimentary basin when exhalative sedimentary processes dominated over clastic sedimentation. The rocks are believed to be isochemically metamorphosed, mixed clastic-exhalative sediments.

The Anglo-Rouyn is an exhalative deposit and several characteristics, listed below, point to its classification with Besshi-type massive sulphide deposits: 1) the metal content of the sulphide ore is Cu, ±Zn, Au, low Pb; 2) the association of the ore body with mafic volcanic rocks and greywacke sediments in a probable back-arc setting; 3) the sheet-like form of the sulphide units, and the absence of any evidence offocused discharge sites; and 4) the close association of the ore with a succession of exhalative-clastic sediments, in which silica-carbonate exhalite is widely dispersed and the sulphide ore is associated with the more restricted iron oxide exhalite (magnetite iron formation).

The Anglo-Rouyn Mine, a past producer of copper and gold, is located approximately 35 km northeast of La Ronge, on Williams Peninsula, on the northwest shore of Lac la Ronge. Between 1966 and 1972, Rio Algom Mines extracted 1.7 million t of ore from three orebodies with an average grade of 1.7 percent Cu, 1.2 g/t Au (0.03 oz/ton), and 5.5 g/t Ag. The mine development involved two open pits and an underground operation which was developed on five levels and serviced by a 244 m shaft. The history of the discovery and development of the deposit is summarized by Roberts and Krey (1995).

The deposit is in Paleoproterozoic rocks of the Nut Bay Belt (Figure 1), a subdivision of the La Ronge Domain, which is part of a collage of Paleoproterozoic accreted terrains that make up the Reindeer Zone (Stauffer, 1984; Lewry et al., 1990) of the Trans-Hudson Orogen (Hoffman, 1990). The Nut Bay Belt is described by Lewry and Slimmon (1985) as comprising foliated granodiorite gneisses intercalated with subordinate supracrustals, including hornblende gneiss and amphibolite of volcanic origin, and lesser metasediments. Mafic and ultramafic intrusive rocks are minor components.

Figure 1 - Tectonostratigraphic subdivisions of the La Ronge Domain, and the location of the Anglo- Rouyn deposit.

The genesis of the deposit has been the subject of debate in the literature (Coombe, 1991, p67), primarily because the deposit has been metamorphosed to amphibolite facies, which tends to obscure the origin of the host rocks, and their relationships to the deposit. Roberts and Krey (1995) concluded that the deposit is volcanogenic, probably related to mafic volcanism. This conclusion is supported by the recognition of basaltic flows in the rocks underlying the immediate footwall of the ore (Maxeiner, 1994).

This paper summarizes the geology of the Anglo- Rouyn deposit, presents an interpretation of geochemical data from the enigmatic host rocks and provides a genetic model for its formation.

The rocks underlying the mine area trend northeast and dip at moderate to steep angles to the northwest (Figure 2). From southeast to northwest, the sequence consists of basaltic metavolcanic rocks with associated units of siliceous metagreywacke, followed by the footwall rocks to the deposit, which comprise calc-silicate gneiss, hornblende-plagioclase gneiss, and magnetite iron formation. The stratabound copper-iron sulphides comprising the ore overlie the iron formation and are hosted by quartz-hornblende-plagioclase gneiss. A unit of quartz-plagioclase gneiss forms the hanging wall to the deposit. The Jepson Lake granite, a regional-scale, sill-like body, was intruded into the hanging wall before, or during, deformation.

Figure 2 - A) Geological map of the Anglo-Rouyn area; B) longitudinal section through the ore body; and C) cross section through the ore body at the shaft.

Metavolcanic and Associated Sedimentary Rocks

The oldest rocks of the sequence in the mine area are metabasaltic flows and tuffs intercalated with psammitic metagreywackes (Figure 2). The metavolcanic rocks of the area (Maxeiner, 1994), previously mapped as amphibolites by Forsythe (1971) and Roberts and Krey (1992), are generally dark to pale green, fine grained, and consist of hornblende (>45 percent), plagioclase, epidote, and accessory biotite, sphene, and calcite. The metagreywackes vary from gray, biotite-garnet, and in part sillimanite- bearing rocks, to greenish gray, hornblende-biotite- bearing varieties.

Andesitic metavolcanics occur in the overlying footwall rocks and comprise tuffs, distinguished by their compositional layering, and flows (or intrusive rocks) which are not layered, but weakly foliated, with a strong plagioclase mineral lineation. Hornblende and epidote constitute approximately 30 percent of the andesitic rocks.

The Footwall Units

Mappable units in the footwall rocks include: 1) hornblende-plagioclase gneiss, 2) magnetite iron formation, and 3) calc-silicate gneiss. A continuous unit of calc-silicate gneiss overlies the metabasalt and underlies the hornblende-plagioclase and magnetite iron formation units. In the southwest part of the property, hornblende-plagioclase gneiss forms the immediate footwall to the ore zone. Northeast of the Main Pit, hornblende-plagioclase gneiss passes along strike into magnetite iron formation. The magnetite iron formation increases in thickness northeastwards from approximately one metre at the Main Pit, to form the lateral equivalent of the hornblende-plagioclase gneiss unit (Figure 2).

Hornblende-plagioclase gneiss consists principally of hornblende and andesine-labradorite with lesser amounts of quartz, diopside, epidote, calcite, sphene, and biotite. The gneissic fabric is defined by alternating hornblende- and feldspar-rich layers.

The magnetite iron formation has the mineralogy and gneissic fabric of hornblende-plagioclase gneiss with the addition of disseminated magnetite in the mafic layers. Magnetite locally forms layers up to several centimetres thick, which are best developed in the northeast part of the property where the unit is thickest. The magnetite is generally associated with increased amounts of calcite and epidote, each of which may characterize layers. Quartz or recrystallized chert layers are not a feature of the unit. The iron formation can be traced with a magnetometer, and the weathered surfaces are generally distinguished by secondary, red iron oxides and trace amounts of malachite.

Calc-silicate gneiss is characterized by the presence of calcite, scapolite, almandine garnet, diopside and epidote. The scapolite has high calcium (17 to 18 percent) and low sodium (3 to 4 percent). The calc- silicate minerals are generally concentrated in quartzofeldspathic layers of otherwise typical hornblende-plagioclase gneiss. The carbonate-rich quartzofeldspathic layers may be laminated, and the recessive weathering of the calcite gives the outcrops a characteristic ribbed weathered surface. Based on field evidence a few kilometres along strike and to the northeast of the open pits, Maxeiner (1994) interpreted the footwall units as a succession of reworked volcanics, calcareous sediments, lean iron formation, and intercalated tuffaceous horizons.

The Ore Zone Units

The silicate units of the ore zone are: quartz- hornblende-plagioclase gneiss, which contains the sulphide ore; and quartz-plagioclase gneiss, which forms the hanging wall to the ore, although each may contain subordinate units of the other. Northeast of Pit "C", the quartz-hornblende-plagioclase gneiss was not found in outcrop, and the ore zone units are represented by quartz-plagioclase gneiss.

The mineralogy of the quartz-hornblende-plagioclase gneiss is similar to that of the footwall hornblende- plagioclase gneiss, with the principal exception of a marked increase in quartz and, to a lesser extent, biotite content. Scapolite and calcite, in association with plagioclase, sphene, epidote, diopside, and grossular garnet, locally constitute several modal percent of the unit.

The quartz-plagioclase gneiss can be divided into two principal assemblages: quartz-plagioclase, with up to 8 percent modal biotite and disseminated calcite; and quartz-plagioclase, with up to 20 percent combined hornblende, epidote, diopside, and sphene, with trace amounts of scapolite, apatite, and grossular garnet. Pyrite is common in the hornblende- and epidote-rich units but is rarely associated with the biotite-rich quartz-plagioclase gneiss. In some sections of the unit, the quartz content may increase to more than 80 percent. The quartz-rich units frequently have an opalescent sheen, typical of chert. Layering and fine laminations occur in the unit, particularly in biotite- rich and quartz-rich lithologies.

According to Forsythe (1971), the deposit consists of a number of "ruler-shaped" sulphide lenses distributed over a strike length of 3000 m (Figure 2). The lenses are grouped into three orebodies, "A", "B", and "C". The entire deposit lies approximately on a single plane, but in detail, the four lenses that constitute the "A" ore body lie on two planes within a 10 m width (Figures 2B and C). The largest lens of the deposit attains a thickness of 15 m, has a down-dip dimension of 60 m and a down-plunge length of 1750 m. The lenses of sulphide ore occur in quartz-hornblende-plagioclase gneiss immediately overlying iron formation. The ore minerals in order of abundance are: chalcopyrite, pyrite, pyrrhotite, and magnetite, but include trace amounts of sphalerite, molybdenite, and skutterudite. In addition, Coombe (1991) makes reference to galena, pentlandite, and trace amounts of chalcocite and bornite.

Intrusive Rocks

The Jepson Lake pluton is a concordant regional-scale granitic intrusion consisting of a central zone of megacrystic potassium feldspar granite and a marginal zone of strongly foliated granodiorite. In the mine area, the pluton intrudes the quartz-plagioclase gneiss of the ore zone. Irregular pegmatites and narrow, vertical granitic dykes striking between 160° and 180° intrude the gneissic rocks.

The rocks of the area have undergone three periods of deformation, the first of which was coeval with amphibolite facies metamorphism. The S₁-L₁ fabric developed during D₁ is defined by the preferred orientation of the amphibolite facies minerals and by the gneissic fabric of the rocks. The S₁ foliation strikes approximately parallel to the rock unit contacts and dips to the northwest at 45° to 85°. The mineral lineation plunges 0° to 20° to the southwest south of the main open pit, and to the northeast north of the open pit. The gneissic layering of all the footwall and ore zone units is the combined result of metamorphic segregation and transposition of an earlier layering into S₁. The early compositional layering predates deformation, and it is assumed to be bedding. The axes of intrafolial folds associated with the transposition structures plunge parallel to the L₁ lineation.

In vertical sections exposed at the Main Pit and Pit "C", the S₁ foliation is oblique to S₀, where S₀ is defined by the contacts of subunits in the ore zone. The steeper dip of the S₁, foliation is evidence that the ore zone is on the north limb of an antiform that closes to the southeast. In his regional study, Maxeiner (1994) notes that "The lithological similarity between rocks from Cochrane and Ore bays on the southeastern side of the Jepson Lake sill to some of the rocks of the Sulphide Lake package is striking." Since the rocks of the "Sulphide Lake package" northwest of the Jepson Lake sill young consistently to the southeast (Sibbald,1986), it suggests that the Jepson Lake sill is at the closure of a syncline, and the units of the Anglo- Rouyn area face northwest.

The second deformational event (D₂) is characterized by the development of shear zones subparallel to the earlier gneissic layering. Gneissic layering above the D₂ shear zones is folded into mesoscopic, open to isoclinal folds with the development of spaced cleavage. A lineation, which may take the form of rodding, is developed by the intersection of the shear zones and spaced cleavage with the folded gneissic layering. In many exposures, folding is not evident and the deformation is represented only by shearing. Imbricate structures rooted in the shear zones give the structure a characteristically broken or fractured appearance. The overturned nature of the folds indicates a dip-slip reverse movement on the shear zones. The displacement on the structures is small, probably of the order of a few metres. The most extensive development of the D₂ structures is in the calc-silicate gneiss, and sulphide ore.

The D₃ deformation is represented by folds and associated faults and veins. F₃ folds occur as open to closed, isolated folds or groups of folds in the gneissic layering. The D₂ shear structures are also folded by the F₃ folds. The folds are formed by a combination of buckling and flexural slip, and die out rapidly when traced along their axial surfaces. The axes of the folds plunge 30° to 60° northwards, and axial surfaces dip 60° to 85° towards the east. Quartz veins and minor faults are developed approximately parallel to the axial surfaces of F₃ folds.

The effects of the three deformational events are found in the ore and in other sulphide-rich units. The deformation structures in the ore include: 1) F₁ folds defined by siliceous (principally chert), pyrite- and magnetite-rich layers; 2) sulphide and silicate layers transposed into the S₁ foliation; 3) chalcopyrite and pyrite remobilized into the pull-apart structures associated with D₁ boudins; 4) F₂ folds of the S₁ foliation in the ore; 5) duchbewegung texture, in which the ductile sulphides, pyrrhotite, and chalcopyrite enclose partially rounded clasts, probably formed by D₂ deformation of silicates, pyrite, and magnetite; and 6) chalcopyrite with subordinate iron sulphides in quartz veins associated with F₃ folds. At the megascopic scale, the sulphide ore lenses are coaxial to the L₁ lineation (Figure 2).

In order to classify the ore body, it was necessary to establish or confirm the protoliths of the gneissic and metavolcanic rocks associated with the deposit. The ore and host rocks were analysed for major and trace elements by XRF at Geochemical Laboratories, McGill University, and by XRF and ICP at Activation Laboratories, Ancaster, Ontario. Gold, silver, and REE analyses were determined by neutron activation at Activation Laboratories. The analyses of sixty-five samples are provided by Roberts and Krey (1995) and are summarized in Table 1.

Table 1 - Average chemical composition of units in the Anglo-Rouyn area; (9) Average chemical composition of metasedimentary unit, Ljusnarsberg-Stalldalen (Parr, 1988); (10) Chemical composition of greywacke, Beatty Formation, Abitibi greenstone belt (Feng and Kerrich, 1990); Oxides and LOI in weight percent; elemental values in parts per million; total iron is expressed as Fe₂O₃.

Metavolcanic Rocks

The igneous origin of the units identified in the field as basaltic and andesitic flows and tuffs, is confirmed by their major element composition (Table 1), and by correlations of Cr, Ni, and Co with MgO, (Pearson correlation coefficients, r=0.95, 0.79, and 0.74 respectively) (Figures 3A and B), and Ti with V (r=0.56). Samples of the flows and tuffs plot in the subalkaline basalt and andesite fields of the SiO₂ vs. Zr/TiO₂ discrimination diagram of Winchester and Floyd (1977) (Figure 3C).

Figure 3 - Chemical variation diagrams for Anglo-Rouyn volcanic rocks. The covariance of MgO with Cr (A) and with Ni (B) confirms the magmatic character of units mapped as volcanic. (C) The classification of the volcanic rocks according to Winchester and Floyd (1977); basalt (II), filled circles; basalt (I), crosses; andesite, filled triangles; and tuff, filled square.

The major and trace element relationships from nine samples of metabasalt indicate the presence of two magma types (Table 1): basalt (I) is characterized by comparatively high TiO₂ values, averaging 0.95 wt % and basalt (II) which contains an average of only 0.40 wt % TiO₂. The distinction between the two basalt magma types is also shown by the REE plots (Figures 4A and 4B). Basalt (I) has a comparatively flat pattern, whereas the plots for basalt (II) are comparatively steep with relatively higher enrichment of LREE and depletion of the HREE. This latter pattern is duplicated by the REE plots for andesite, but with higher values for the REE. Although basalts (I) and (II) are clearly distinguished by their geochemistry, the existence of two magma types was not apparent in the field.

Figure 4 - Diagrammatic representation of volcanic rock chemistry: A) and B) are chondrite normalized rare earth element plots of the volcanic rocks showing the distinction of andesite (area 1), basalt (II) (area 2), and basalt (I) (area 3); and C) a Ti vs. Cr plot of the volcanic rocks and tuffs with fields for modern arc and ocean floor basalts from Pearce (1975). Symbols as in Figure 3.

The Ti-Cr discrimination plot of Pearce (1975) indicates the emplacement of the volcanic rocks in an arc environment (Figure 4C). Their association with greywacke-type sediments is also compatible with an arc environment.

Footwall Rocks

The hornblende-plagioclase gneiss occurs in the iron formation and in the calc-silicate gneiss as small units ranging from a few centimetres to a metre thick. Calc-silicate gneiss is characterized by the addition of calcite, quartz, and scapolite; the iron formation by the addition of magnetite, calcite, scapolite, and minor sulphides (pyrite, pyrrhotite, and chalcopyrite). The field and petrographic evidence suggests that these two units consist of mixtures of exhalative material, magnetite, calcite, and quartz, and a clastic material which is the protolith of the hornblende-plagioclase gneiss.

The major and trace element content of the hornblende-plagioclase gneiss is comparable to that of a basic igneous rock (Table 1), but differs in its widely variable iron (6.22 to 19.16 wt %), consistently higher CaO (12.5 to 17.6 wt %) and generally lower MgO (1.12 to 4.78 wt %) contents. The trace elements most characteristic of basic igneous rocks, Cr, Ni, and Co, are present in equivalent amounts: 198 to 310 ppm, 70 to 143 ppm, and 14 to 50 ppm respectively. Compared to greywackes from greenstone belts (Table 1), the hornblende-plagioclase gneiss is notably lower in SiO₂ and higher in CaO, but the MgO contents are comparable.

The higher and strongly variable iron and calcium contents in the footwall rocks, as compared to basalt, can be ascribed to hydrothermal exhalative processes affecting a predominantly volcanogenic environment. Comparatively low MgO values of the hornblende- plagioclase gneiss imply that the original ferromagnesian minerals of the volcanic protoliths have been affected by alteration processes. The correlation of MgO with Cr, Ni, and Co, compared to that in the volcanic rocks, is shown in Table 2. There is a strong correlation of MgO with Cr and Ni in the iron formation, a moderate correlation with Ni, and a negative correlation with Cr in the hornblende-plagioclase gneiss. Co has only a weak negative correlation with MgO in the hornblende-plagioclase gneiss and the iron formation, and has only weak or negative correlation of these elements in the calc- silicate gneiss (Figures 5A and B).

Table 2 - Correlation coefficients of MgO with Cr, Ni, and Co in basalt (II) and the footwall units.

Figure 5 - Chemical variation diagrams for the footwall rocks: plots A) and B) illustrate the relationship of MgO to Cr and Ni in the footwall units; C) illustrates the positive Al₂O₃-SiO₂ correlation for hornblende-plagioclase gneiss and iron formation; and D) illustrates the decoupling of Al₂O₃ with SiO₂ in the calc-silicate gneiss unit. Symbols: hornblende-plagioclase gneiss, filled diamond; iron formation, cross; and calc-silicate gneiss, open circles.

In the hornblende-plagioclase and iron formation units of the footwall, Si02 correlates strongly with Al₂O₃ (r=0.94) (Figure 5C). This contrasts with the moderate correlation of 0.44 for SiO₂ and Al₂O₃ in the volcanic rocks. This is interpreted to signify that, in hornblende-plagioclase gneiss and iron formation, silica is controlled principally by secondary alumino-silicate minerals. If significant amounts of silica, in the form of detrital quartz or hydrothermal chert, had been contributed to these units, the correlation between SiO₂ and Al₂O₃ would have been weak or strongly negative due to the dilution effect. Greywackes, for example, may show such a dilution effect due to the addition of detrital quartz (Camiré el al., 1993; La Flèche and Camiré, 1996; Feng and Kerrich, 1990). The lack of correlation in the calc-silicate unit (Figure 5D) is probably due to the addition of exhalative silica.

In alteration processes, the high-field strength elements, which include Zr, Y and Ti, are generally considered immobile (Barrett and MacLean, 1994). Taylor and McLennan (1985) and Bhatia and Crook (1986) suggest that, because of their relatively low mobility in sedimentary processes and short residence time in seawater, these elements in sedimentary rocks reflect the signature of their parent material. The relative abundances of Ti, Zr, and Y in the footwall rocks (Figures 6A and B), are consistent with the assumption, from field and petrographic data, that the three footwall units have a common clastic component, and furthermore, that it may have been derived from alteration of underlying mafic material of basalt (II) composition. The linear trend in the Zr-Y plot for the footwall units and basalt (II) (Figure 6C), shows that the absolute amounts of the elements increase from the assumed protolith, basalt (II), through hornblende- plagioclase gneiss and iron formation to calc-silicate gneiss.

Figure 6 - Chemical variation diagrams for the footwall gneiss units: footwall rocks plot as a cluster in the TiO₂*50-Zr-Y diagram (A), suggesting a common origin, and overlap the same plot for the underlying mafic volcanics (B); and C) is the linear plot of Zr vs. Y for basalt (II) and footwall rocks. Symbols: hornblende-plagioclase gneiss, filled diamonds; iron formation, cross; calc-silicate gneiss, open circles; and basalt (II), filled circles.

The alteration of the volcanic protolith of the hornblende- plagioclase gneiss may have taken place following transportation, sorting and deposition of volcaniclastic detritus. However, alteration gradients were not detected in the unit, and it is considered more likely that alteration, or submarine weathering of volcaniclastic material in a hydrothermally active basin, preceded the final sedimentary processes.

Ore Zone Rocks

The units of the ore zone are quartz-hornblende-plagioclase gneiss and quartz-plagioclase gneiss. Quartz-hornblende- plagioclase gneiss, which generally overlies iron formation, is the host to the sulphide ore; and quartz-plagioclase gneiss forms the hanging wall to the ore.

The gradational contact relationships are reflected in the compositions of the units in that they are difficult to define on the basis of geochemistry. When considered as a single unit, the major element content is quite variable. Silica varies from 44.89 to 86.99 wt % with a mean value of 59.90 wt %. The TiO₂ content varies from 0.24 to 1.66 wt %. The Al₂O₃ content is higher than in the footwall units ranging from 5.34 to 19.57 wt %, with a mean of 15.04 wt %, but 27 of the 31 samples have Al₂O₃ contents between 12 and 20 wt %. The CaO content of the unit is similarly variable (2.02 to 21.70 wt %, with a mean value of 9.41 wt %), with 11 of the 31 samples having CaO contents between 10 and 13 wt %. The mean values for MgO and Fe203 are low compared to other units in the stratigraphic section (2.40 and 5.79 wt % respectively). The alkali contents are variable, with up to 6.75 wt % Na20 and up to 3.28 wt % K20. The two units of the ore zone may be distinguished in terms of their major element signatures, with relatively high CaO concentrations in quartz-plagioclase gneiss, and relatively high Fe₂O₃ contents in quartz-hornblende-plagioclase gneiss (Figure 7A).

Figure 7 - Chemical variation diagrams for ore zone units: A) the units of the ore zone are differentiated on the basis of their relative proportions of calcium and total iron (expressed as Fe₂O₃); B) the TiO₂*50-Zr-Y plot of the ore zone units shows a wider dispersion than in the footwall rocks (see Figure 6); and C) the negative Al₂O₃-SiO₂ correlation in the ore zone rocks is due to the dilution effect of hydrothermal, exhalative SiO₂. Symbols: quartz-plagioclase gneiss, open triangles; and quartz-hornblende- plagioclase gneiss, filled circles.

The quartz content of the ore zone units distinguishes them in the field from the hornblende-plagioclase gneiss and magnetite iron formation of the footwall. This is reflected in the geochemistry of the ore zone units (Figure 7C), in that the majority of the major elements in both ore zone units, have a strong negative correlation with SiO₂ due to the dilution effect of hydrothermal, exhalative silica (SiO₂-CaO, r=-0.68; MgO, r=-0.54; F₂O3, r=-0.68; Al₂O₃, r=-0.64; and TiO₂, r=-0.63).

In the plot of the "immobile" elements TiO₂*50-Zr-Y (Figure 7B) the relative values are dispersed over an area on the diagram that includes, but extends beyond, the plot of the samples from the footwall rocks (Figure 6A).

Assuming isochemical metamorphism, a strong correlation between K₂O, Ba, Rb, Na₂O and Zr (r varies from 0.70 to 0.42) suggests that the abundances of these elements in the original basin were controlled by K-bearing clays such as illite. However, the correlation of these elements with Al₂O₃ is weak for K₂O (r=0.17) and moderate for the others (0.64 to 0.30). A possible explanation for this relationship is that kaolinite was a constituent of the more aluminous protolith (up to 19.57 wt % Al₂O₃).

The precursor to the ore zone units was probably sediment of mixed exhalative and clastic composition, in which sulphides, magnetite, chert, and calcite represent the exhalative component, and the aluminosilicates are indicative of clastic protolith. The local development of fine laminations in the units is also compatible with such an interpretation. Calc- silicate rocks with similar mineralogy and composition in massive sulphide deposits associated with sedimentary rocks, notably Broken Hill-type deposits, have been identified as exhalative units (Parr and Plimer, 1993). In Table 1, a whole rock analysis of a calc-silicate exhalative unit from the Bergslagen Ore Province, Sweden (Parr, 1988), is compared to the composition of the quartz- plagioclase gneiss.

The stratigraphic succession in the mine area records the history of a volcanic-sedimentary basin in a volcanic arc environment (probably a back-arc basin), as it evolved from a period of mafic volcanism and greywacke-type sedimentation of proximal derivation, into a time in which clastic sediments competed with exhalative, chemical sediments.

The clastic material in the three units that comprise the footwall was formed by the sedimentary deposition of material derived from the alteration of mafic volcanics in a submarine, hydrothermally active environment. The exhalites were deposited in the following sequence: carbonate-silica of the calc-silicate gneiss unit overlying the mafic flows and tuffs, iron oxide- carbonate of the magnetite iron formation, Cu-Fe sulphide and iron oxide-silica of the ore, and silica- carbonate of the quartz-plagioclase gneiss forming the hanging wall.

In the map area, indications of mineralization, in the form of minor amounts of iron and copper sulphides, occur in the ore zone units for their entire strike length. Coombe (1991) states that mineralization similar to that at Anglo-Rouyn "is known at this favourable horizon over a strike length of more than 20 km, from Williams Peninsula in the southwest, northeastwards to Rob Lake." It should be noted, however, that the "favourable horizon" referred to by Coombe is defined by its proximity (within 20 to 25 m) to the contact with the Jepson Lake pluton.

The Cu-Fe sulphides occur as discrete sheets of sulphides, immediately overlying and closely associated with the magnetite iron formation. The magnetite iron formation is thickest in the northeast part of the mine area (Figure 2) and decreases southwestwards to pinch out at the main pit where the sulphide ore units plunge to the southwest. As a consequence, the ore is not exposed at surface southwest of the Main Pit. It is assumed from the stratigraphic and structural relationships, that footwall magnetite iron formation plunges to the southwest with the sulphides of the ore units. These relationships suggest that the magnetite iron formation occupied a more restricted aerial extent in the volcanic- sedimentary basin than the exhalites of the underlying calc-silicate gneiss unit and the quartz-plagioclase gneiss of the hanging wall. Furthermore, the structural relationships at the Main Pit suggest that the base metal sulphides were deposited in the basin at the edge of the iron oxide-carbonate exhalite (magnetite iron formation). These empirical stratigraphic relationships, among the exhalite facies, could be applied to exploration within the "favourable horizon" defined by Coombe (1991).

The Anglo-Rouyn deposit has many of the definitive characteristics of a Besshi-type massive sulphide deposit (Slack, 1993): (i) the metal content of the sulphide ore is Cu, ±Zn, Au, low Pb, (ii) the association of the ore body with mafic volcanic rocks and greywacke sediments in a probable back-arc setting, (iii) the sheet-like form of the sulphide units, and the absence of any evidence of focused discharge sites, and (iv) the close association of the ore with a succession of exhalative-clastic sediments, in which silica- carbonate exhalite is widely dispersed, iron oxide (magnetite iron formation) is more restricted, and the sulphide ore is associated with the iron oxide exhalite.

The project was funded under the Geoscience program of the Canada-Saskatchewan Partnership Agreement on Mineral Development (1990-95). Our understanding of the geology Zr and structural geology of the area benefited greatly from discussions in the field with T.I.I. Sibbald of the Saskatchewan Geological Survey, and K.H. Poulsen and F. Robert of the Geological Survey of Canada. Data on the petrography of the rock units and ore were provided by Karen Krey.

Barrett, T.J. and MacLean, W.H. (1994): Chemostratigraphy and hydrothermal alteration in exploration for VHMS deposits in greenstones and younger volcanic rocks; in Lentz, D.R. (ed.), Alteration and alteration processes associated with ore-forming systems, Geol. Assoc. Can., Short Course Notes, v11, p433-467.

Bhatia, M.R. and Crook, K.A.W. (1986): Trace element characteristics of greywackes and tectonic setting discrimination of sedimentary basins; Contrib. Mineral. Petrol., v86, p181-193.

Camire, G.E., Lafleche, M.R., and Ludden, J.N. (1993): Archean metasedimentary rocks from the northwestern Pontiac subprovince of the Canadian Shield: Chemical characterization, weathering and modelling of the source area; Precamb. Resear., v62, p285-305.

Coombe Geoconsultants Ltd. (1991): Base metals in Saskatchewan; Sask. Energy Mines, Open File Rep. 91-1, 218p.

Feng, R. and Kerrich, R. (1990): Geochemistry of fine-grained clastic sediments in the Archean Abitibi greenstone belt, Canada: Implications for provenance and tectonic setting; Geochim. Cosmochim. Acta, v54, p1061-1081.

Forsythe, L.H. (1971): The geology of the Nemeiben Lake area (east half) and the geology of the mineral deposits in the Nemeiben Lake-Stanley Lake areas, Saskatchewan; Sask. Dep. Miner. Resour., Rep. 115, 178p.

Hoffman, P.F. (1990): Subdivision of the Churchill Province and extent of the Trans-Hudson Orogen; in Lewry, J.F. and Stauffer, M.R. (eds.), The Early Proterozoic Trans-Hudson Orogen of North America, Geol. Assoc. Can., Spec. Pap. 37, p15-39.

La Flèche, M.R. and Cam ire, G. (1996): Geochemistry and provenance of metasedimentary rocks from the Archean Golden Pond sequence (Casa Berardi mining district, Abitibi subprovince); Can. J. Earth Sci., v33, p676-690.

Lewry, J.F. and Slimmon, W.L. (1985): Compilation bedrock geology, Lac la Ronge, NTS Area 73P/73I; Sask. Energy Mines, Rep. 225, 8p.

Lewry, J.F., Thomas, D.J., Macdonald, R., and Chiarenzelli, J. (1990): Structural relations in accreted terranes of the Trans-Hudson Orogen, Saskatchewan: Telescoping in a collisional regime?; in Lewry, J.F. and Stauffer, M.R. (eds.), The Early Proterozoic Trans-Hudson Orogen of North America, Geol. Assoc. Can, Spec. Pap. 37, p75-94.

Maxeiner, R.O. (1994): Geology of the MacKay Lake and Anglo-Rouyn Mine areas, southern La Ronge Domain (part of NTS 73P-7); in Summary of Investigations 1994, Saskatchewan Geological Survey, Sask. Energy Mines, Misc. Rep. 94-4, p44-52.

Parr, J.M. (1988): The metasediments associated with stratabound base metal mineralization, Ljunsnarsberg District, central Sweden; Geol. en Mijnb., v67, p189-202.

Parr, J.M. and Plimer, I.R. (1993): Models for Broken Hill-type lead-zinc-silver deposits; in Kirkham, R.V., Sinclair, W.D., Thorpe, R.I., and Duke, J.M. (eds.), Mineral Deposit Modelling, Geol. Assoc. Can., Spec. Pap. 40, p253-288.

Pearce, J.A. (1975): Basalt geochemistry used to investigate past tectonic environments on Cyprus; Tectonophysics, v25, p41-67.

Roberts, R.G. and Krey, K. (1992): The Anglo-Rouyn mine, La Ronge area: A preliminary report; in Summary of Investigations 1992, Saskatchewan Geological Survey, Sask. Energy Mines, Misc. Rep. 92-4, p184-188.

___ (1995): The structural geology and lithogeochemistry of the Anglo-Rouyn deposit, Lac la Ronge, Saskatchewan; in Richardson, D.G. (ed.), Investigations completed by the Saskatchewan Geological Survey and the Geological Survey of Canada under the Geoscience Program of the Canada-Saskatchewan Partnership Agreement on Mineral Development (1990-1995), Geol. Surv. Can., Open File 3119, Sask. Energy Mines, Open File Rep. 95-3, p147-176.

Sibbald, T.I.I. (1986): Bedrock geological mapping, Sulphide Lake area (part of NTS 73P-7); in Summary of Investigations 1986, Saskatchewan Geological Survey, Sask. Energy and Mines, Misc. Rep. 86-4, p63-64.

Slack, J.F. (1993): Descriptive and grade-tonnage models for Besshi-type deposit massive sulphide deposits; in Kirkham, R.V., Sinclair, W.D., Thorpe, R.I., and Duke, J.M. (eds.), Mineral Deposit Modelling, Geol. Assoc. Can., Spec. Pap. 40, p343-371.

Stauffer, M.R. (1984): Manikewan: An Early Proterozoic ocean in central Canada, its igneous history and orogenic closure; Precamb. Resear., v25, p257-281.

Taylor, S.R. and McLennan, S.M. (1985): The Continental Crust: Its Composition and Evolution. An Examination of the Geological Record Preserved in Sedimentary Rocks; Blackwell, Oxford, 307p.

Winchester, J.A. and Floyd, P.A. (1977): Geochemical discrimination of different magma series and their differentiation products using immobile elements; Chem. Geol., v20, p325-343.

R. Gwilym Roberts and R.O. Maxeiner