The Mankayan mineral district of northern Luzon, Philippines, hosts several significant ore deposits and prospects of various types within an area of ~25 km2, including the Far Southeast porphyry Cu-Au deposit, the Lepanto high sulfidation epithermal Cu-Au deposit, the Victoria intermediate sulfidation epithermal Au-Ag vein deposit, the Teresa epithermal Au-Ag vein deposit, the Guinaoang porphyry Cu-Au deposit, and the Buaki and Palidan porphyry Cu-Au prospects, all having formed in a period of about 2 m.y., from ~3 Ma. The geologic units include (1) a basement composed of Late Cretaceous to middle Miocene metavolcanic rocks and volcaniclastic rocks; (2) the Miocene 12 to 13 Ma tonalitic Bagon intrusive complex; (3) the Pliocene, ~2.2 to 1.8 Ma, Imbanguila dacite porphyry and pyroclastic rocks; and (4) postmineralization cover rocks, including the ~1.2 to 1.0 Ma Bato dacite porphyry and pyroclastic rocks and the ~0.02 Ma Lapangan tuff.

Extensive advanced argillic alteration crops out for ~7 km along the unconformity between the basement rocks and the Imbanguila dacite formation and consists of quartz-alunite ± pyrophyllite or diaspore, with local zones of silicic alteration and a halo of dickite ± kaolinite. The alteration and its subhorizontal geometry indicate that it is a lithocap or coalesced lithocaps. The northwest-striking portion is ~4 km long and hosts the Lepanto enargite Au ore deposit, also controlled by the Lepanto fault. The Lepanto epithermal deposit is related to the underlying Far Southeast porphyry; the quartz-alunite alteration halo of Lepanto is contemporaneous with the ~1.4 Ma potassic alteration of the porphyry. There are also silicic-advanced argillic alteration patches ~600 m above the Far Southeast orebody at the present surface; these are interpreted to be perched alteration. There is no systematic mineralogical or textural zoning in the Lepanto lithocap that indicates direction to the intrusive source. Most surface samples of the lithocap contain less than 50 ppb Au, despite many being less than a few hundred meters from underground Cu-Au ore.

This study found that several characteristics of the Lepanto lithocap change systematically with distance from the causative intrusion: The alunite absorption peak at ~1,480 nm in the short wavelength infrared (SWIR) spectrum shifts to higher wavelengths where the sample is closer to the intrusive center, due to higher Na and lower K content in the alunite; published experimental studies indicate that high Na/(Na + K) is related to higher formation temperature. High Ca alunite, including huangite, also occurs at locations proximal to the intrusive center. Alunite mineral composition analyzed by laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) indicates that the Pb content decreases toward the intrusive center, whereas Sr, La, Sr/Pb, and La/Pb increase markedly. Whole-rock compositions, using only nonmineralized (taken as Cu <0.1wt % and Au <0.1 ppm) and alunite-bearing samples, show that Pb and Ag/Au, plus Hg and Ag, decrease toward the intrusive center, and Sr/Pb and La/Pb ratios increase. Normalizing whole-rock Pb to the (Na + K) molal content produces a proxy for the alunite mineral composition, and this ratio provides the same indications as the LA-ICP-MS analyses of alunite. The concealed Victoria epithermal veins consist of intermediate sulfidation mineralization on the southwest flank of the porphyry. The veins are not exposed, but their presence at depth is indicated by subtle alteration (illite or interstratified illite and/or smectite or smectite + pyrite) and geochemical (As, Se) anomalies at the surface. The anomalies are strongly dependent on erosion level; no anomalies were found where the surface is >~350 m above the upper extent of the veins. An airborne geophysics survey indicates that the Far Southeast orebody is associated with a wide zone of demagnetization due to extensive magnetite-destructive phyllic alteration. Such low magnetic anomalies on the margin of a large lithocap elsewhere may deserve attention. The directional indicators and mineralization signatures found in this study have the potential to indicate direction to the intrusive center during exploration of similar porphyry-epithermal districts.


Lithocaps are horizontal to subhorizontal blankets of residual quartz and advanced argillic alteration of hypogene origin, occurring over intrusions (Sillitoe, 1995a). They can host high sulfidation epithermal mineralization, particularly within their fracture-controlled roots. Lithocaps are temporally and genetically related to intrusions that may be associated with deeper porphyry-style mineralization (Sillitoe, 1995a, 1999, Sillitoe, 2011; Hedenquist et al., 1998). Lithocaps can have large areal extent (>20 km2; Sillitoe, 1995a) and, because they resist erosion, are typically prominent at the surface, which generally makes them easy to find. The presence of a lithocap of large areal extent (A. Arribas, pers. commun., 1999) is encouraging for exploration at an early stage, as it indicates extensive hydrothermal activity and potential for high sulfidation ore; in addition, there is potential for deeper porphyry and marginal epithermal vein mineralization. Despite the relative ease of finding lithocaps, it may be difficult to further define the location of mineralization within, under, or adjacent to a large lithocap due to the lack of directional indicators. In particular, determining the position of the underlying intrusive center is commonly difficult. In addition, veins on the margin may terminate several hundred meters below the paleosurface, below the level of the lithocap. In a district without outcropping veins, exploration for such veins is typically difficult.

The Mankayan mineral district, northern Luzon, Philippines, was studied in order to develop tools for exploration in such districts, as it is the site of several large intrusion-related ore deposits and prospects. The district lies within a well-defined, 150-km-long belt of porphyry Cu deposits in the Central Cordillera of northern Luzon (e.g., Cooke et al., 2011; Deyell and Hedenquist, 2011; Hollings et al., 2011a, b; Waters et al., 2011; Wolfe and Cooke, 2011). It is one of the country’s richest mining districts, both in terms of proven and potential economic value as well as abundance and diversity of hydrothermal mineralization (Sillitoe and Gappe, 1984). Within an area of ~25 km2 there are several porphyry Cu-Au and epithermal deposits and prospects (Fig. 1). The Lepanto high sulfidation epithermal deposit and associated lithocap host is located above and adjacent to the Far Southeast porphyry Cu-Au deposit (Fig. 1). In addition to the close spatial relationship (Sillitoe, 1983) and an overlap in alteration and ore mineralogy (Garcia, 1991; Hedenquist et al., 1998), the two orebodies have been shown to be contemporaneous and genetically linked (Arribas et al., 1995; Hedenquist et al., 1998). The more recently discovered Victoria intermediate sulfidation vein deposit (Cuison et al., 1998; Claveria, 2001), located within 1 km southwest of the Far Southeast porphyry (Fig. 1), and the adjacent Teresa vein deposit to the south, completes the assemblage of porphyry and related high- and intermediate sulfidation ore deposits in this intrusion-centered setting.

This study examined the surface expressions of the three principal deposits, Lepanto (mined out), Far Southeast (drilled out), and Victoria (now being mined). The ores of these deposits are largely blind, and this study was conducted in order to identify signatures and directional indicators that would have potential application during exploration of similar porphyry-epithermal districts. Such findings should be widely applicable, given the common relationships noted among such deposits (e.g., Sillitoe and Hedenquist, 2003; Sillitoe, 2010) and evidence for transitions between them (e.g., Einaudi et al., 2003).

Discovery histories of ore deposits in the district

The Lepanto enargite Au orebody was worked for Cu and Au at the start of the Ming dynasty (14th century), and by the local people prior to the 1500s, after which the Spanish became involved. Outcrops consisting of vuggy to massive quartz along the Lepanto fault were first mined near the 1,150-m-elevation level, with luzonite-enargite still visible in the old workings in the cliff face (Lepanto is the type locality for luzonite). The Cantabro-Filipino company was the first to conduct large-scale mining in 1865, with at least 1,100 metric tons (t) of Cu produced during a 10-year period. The present underground mining activity dates from 1936, when the Lepanto Consolidated Mining Co. commenced mining until the Japanese took over production; Mitsui produced 11,000 t of Cu during the early 1940s. Lepanto Consolidated Mining Co. resumed mining in 1948, and up to 1996 a total of 36.3 Mt of ore was produced from the Lepanto mine at an average grade of 2.9 percent Cu, 3.4 g/t Au, and 14 g/t Ag, with a total of 0.74 Mt Cu, 92 t Au, and 393 t Ag recovered. The Lepanto mine closed in 1996 with a remaining minable reserve of 4.4 Mt at 1.76 wt percent Cu and 2.4 g/t Au (Claveria et al., 1999a).

The Far Southeast porphyry was discovered in 1980, based in part on the prediction that Lepanto was sitting over a porphyry environment (Sillitoe, 1983). Porphyry fragments at the surface were recognized by Lepanto geologists in outcrops of Imbanguila dacite volcanic products in the 1970s. In 1978, the deeper parts of two drill holes, drilled in 1974 about 4 km southeast of Lepanto, were recognized to contain porphyry-type mineralization; subsequent reassay in mid-1980 identified “appreciable Cu values” (Sillitoe, 1995b, p. 33). In 1979, the Cu and Au mineralized porphyry fragments in outcrop over the eastern end of the Lepanto orebody were recognized as being hosted by a diatreme breccia, i.e., implying derivation from a nearby source at depth (Sillitoe, 1995b). An induced polarization (IP) survey in 1979 east of Lepanto identified a 1.5-km-long chargeability anomaly, and this was tested by a deep drill hole in April 1980. The drilling was unsuccessful, intersecting only pyrite thought to be related to Lepanto. A second hole closer to Lepanto was drilled from ~1,400-m elevation to 1,100-m depth in October 1980, and low-grade (0.16 wt % Cu, 0.31 g/t Au) mineralization of a leucocratic quartz diorite porphyry stock was intersected in the bottom 200 m of the hole. This successful drill hole confirmed a 1980 model developed from original surface observations that the diatreme-hosted porphyry-altered lithic fragments were transported from the central parts of a porphyry deposit to the surface by eruptive activity (Sillitoe, 1983). Lepanto Consolidated Mining Co. then rehabilitated the 700-m-elevation level 1.5 km from the Lepanto workings into an area that had earlier drill hole indications of good Cu and Au mineralization. In late 1980, the first drill hole from this underground position, 1,000 m in depth, intersected a 475-m interval of 0.46 wt percent Cu and 0.41 g/t Au, with grades increasing downward and continuing to the bottom of the hole; this is considered to be the discovery drill hole. Up to 1986, 75 drill holes totaling 38,000 m were completed from underground on a 100- to 75-m grid. An additional 11,700 m of drilling was conducted in 1992 to 1995 by a joint venture between Lepanto and CRA Ltd. of Australia, to a depth of 250 m below sea level. The top of the resource is ~650 m below the surface, and at a 1.0 wt percent Cu equiv cutoff, the deposit contains 356 Mt at 0.73 wt percent Cu and 1.24 g/t Au, among the highest grades known in porphyry deposits of the Philippines (Concepción and Cinco, 1989). The core of the porphyry deposit contains 105 Mt at a grade of 0.86 percent Cu and 2.02 g/t Au at 1.8 percent Cu equiv cutoff, considered to be economic for an underground operation in the 1980s (Concepción and Cinco, 1989).

Epithermal veins at Nayak (Fig. 1) and Suyoc (6 km southeast of Nayak) were exploited by artisanal miners for several years. At Nayak, about 2 km south-southwest of the surface projection of the Far Southeast deposit, outcropping quartz-adularia veins have illite halos and contain sphalerite, galena, and Au. These veins were exposed at ~1,200-m elevation by river-cut erosion. By contrast, at Suyoc the principal ore was quartz-carbonate-chalcopyrite veins (Gonzalez, 1967). In early 1991, a surface drilling program of 12 shallow holes in the Nayak area defined about 0.3 Mt of resources at an average grade of 3 g/t Au. In 1995, exploration shifted to the Tabbac area, northeast of Nayak and closer to the surface projection of the Far Southeast deposit. In May the sixth surface hole of the year was drilled from 1,320-m elevation with a south-southeast inclination. At the 1,200-m-elevation level, clay alteration with comb-texture milky quartz and sphalerite-galena was intersected with up to 0.7 g/t Au. At the 1,000-m-elevation level, strong clay alteration and quartz veins with sphalerite-chalcopyrite were intersected in the same hole, with grades of 1.1 to 25.0 g/t Au over 21.6 m, averaging 3.7 g/t. Subsequently, in September 1995, lateral drilling from the 1,000-m level of Lepanto to the west-southwest tested an area 100 m east of the surface hole. The drilling intersected eight mineralized zones with grades from 1.3 to 193 g/t Au, with the best interval of 3.8 m averaging 112.7 g/t. A crosscut was established from Lepanto at an elevation 50 m below the intersections, and encouraging results led to the development of the Victoria mine, with a 2,500 t/d carbon-in-pulp plant becoming operational in March 1997. The quartz-carbonate epithermal veins had an initial resource estimate of 11 Mt at 7.3 g/t Au; grades of 3 to 9 g/t are continuous over a 400-m-vertical interval, and high-grade ore, >30 g/t Au, is more restricted, with up to 250-m-vertical intervals (Cuison et al., 1998; Disini et al., 1998; Claveria et al., 1999a, b). Subsequent development included a twin-ramp decline from near Nayak completed in late 1999. These declines penetrated a fault zone of broken vein material ~1 km south of the Victoria veins that assayed up to 3 to 4 g/t Au. Subsequent underground exploration drilling in a southwest direction from Victoria found the northern extension of this fault zone with better grades, not as high grade as the Victoria veins but with wider brecciated intervals, and this area was denoted Teresa. In October 2004, after seven years of mining, the remaining Victoria resource was 4.6 Mt averaging 5.68 g/t Au at a 2.8 g/t cutoff, and Teresa was 0.8 Mt averaging 5.73 g/t Au.

The Guinaoang porphyry Cu-Au deposit occurs 3 km southeast of Far Southeast (Fig. 1). This prospect is largely concealed by postmineralization rock and shallow-level advanced argillic alteration (quartz-alunite). The advanced argillic alteration was drilled in the 1970s by a Filipino company exploring the area for Lepanto-like mineralization along the southeast extrapolation of the Lepanto fault, principal host to about 70 percent of the Lepanto orebody. Twelve holes, some more than 500 m deep, were drilled without apparent success. However, during subsequent relogging it was recognized that high sulfidation sulfides overprinted sericitic alteration, and that chalcopyrite is present at greater depths (Sillitoe, 1995b). The area was subsequently mapped by Gold Fields Asia Ltd. in early 1980, who took the initial decision to drill test the area for a porphyry target based on a small outcrop of intermediate argillic alteration beneath hypogene quartz-alunite alteration, as well as the relogging of the earlier drill holes (Sillitoe and Angeles, 1985). Drilling intersected porphyry stockwork from a 200-m depth and outlined an estimated resource of 500 Mt at a grade of 0.4 percent Cu and 0.4 g/t Au. The mineralization is largely hosted by an altered quartz diorite intrusion 200 to 1,000 m below surface.

Geology of the District

The geology of the Mankayan district has been summarized by Gonzalez (1956, 1959), Sillitoe and Angeles (1985), Concepción and Cinco (1989), and Garcia (1991). There are four main units in the district: (1) a Late Cretaceous to middle Miocene basement consisting of Lepanto metavolcanic rocks, Apaoan volcaniclastic rocks, and Balili volcaniclastic rocks; (2) the Miocene tonalitic Bagon intrusive complex; (3) the Pliocene Imbanguila dacitic to andesitic porphyry and pyroclastic rocks, which predates the Far Southeast porphyry Cu-Au mineralization, and hosts much of the Lepanto enargite Au deposit and the Victoria veins; and (4) postmineralization cover rocks, including the Pleistocene Bato dacitic to andesitic porphyritic lava and pyroclastic flow units, and the Recent Lapangan tuff (Fig. 1).

The Lepanto metavolcanic rocks (or “green metavolcanic rocks” of Ringenbach et al., 1990) are the lowermost stratigraphic unit of the Mankayan district and crop out in the western part of the district. This unit consists of indurated, tightly packed andesitic to basaltic lavas with minor turbiditic sedimentary rocks. It was cut by mafic dikes and has subsequently suffered greenschist facies metamorphism. From reconnaissance regional mapping they are inferred to be of Cretaceous-Paleogene age (Fernandez and Pulanco, 1967). This unit has been intruded by the Bagon tonalitic intrusive complex, which has a reported radiogenic age of 12 to 13 Ma (Sillitoe and Angeles, 1985). The Apaoan volcaniclastic rocks are present in the northeastern part of Mankayan and consist of green and red thin-bedded siltstone-sandstone. The Balili group overlies the Lepanto metavolcanic and Apaoan volcaniclastic rocks unconformably and consists of mostly matrix-supported polymictic volcanic conglomerates; fossils indicative of late Oligocene to middle Miocene ages were found within calcareous horizons (Sillitoe and Angeles, 1985).

Both Imbanguila and Bato dacitic units are characterized by complex sequences of volcanic breccias, pyroclastic horizons, and massive porphyritic rocks, as well as dike emplacement during deposit formation (summarized by Hedenquist et al., 1998). These units typically contain easily identified bipyramidal quartz eyes, in addition to feldspar and hornblende phenocrysts. The lithologic and chemical similarities (Hedenquist et al., 1998) make it difficult to determine the exact sequence of activity without conducting further extensive radiometric dating and detailed study of the volcanic facies. The Imbanguila host rock, where dated, was found to be 2.19 ± 0.62 to 1.82 ± 0.36 Ma (n = 4; Arribas et al., 1995). The Bato rocks are younger, 1.18 ± 0.08 to 0.96 ± 0.29 Ma (n = 2). Nearby at Guinaoang, igneous biotite from a lapilli tuff associated with a dacite dome has a K-Ar age of 2.9 ± 0.4 Ma (Sillitoe and Angeles, 1985).

The Imbanguila unit lies over the basement rocks and the elevation of the unconformity, i.e., the base of the Imbanguila unit, is contoured in Figure 2. The contour shows that the Imbanguila unit is associated with two large vents that are close to each other and coalesce at elevations between 900 and 1,000 m. The unconformity is steep in the lower parts with a ridge between the two depressions, becomes gentler upward, and flattens to the northwest (Figs. 2, 3). These two vents are the likely sources of the Imbanguila dacite. The vents are located above the subsequent Far Southeast porphyry alteration and mineralization, which are centered on quartz diorite porphyry dikes that intruded to about 300- to 400-m elevation, below the vents. The distribution of the Imbanguila unit (or units) is spatially related to the Lepanto fault, a splay of the Abra River fault that is part of the Philippine fault system. The Imbanguila unit is also offset by the Lepanto fault, consistent with the multiple-movement history of the fault system.

There are also two breccia bodies at the surface. About 1 km northwest of the surface projection of the Far Southeast deposit, a well-defined diatreme breccia crops out above Lepanto at ~1,300-m elevation (Fig. 1; near the Upper Tram), cutting Imbanguila dacite. The diatreme breccia contains lithic fragments with biotite and magnetite alteration and bornite-chalcopyrite mineralization (D. G. Malicdem, pers. commun. to Sillitoe, 1983, and observed in this study), indicating derivation from an underlying porphyry deposit. Fresh hornblende from the matrix of this breccia yielded a K-Ar age of 1.43 ± 0.21 Ma (Arribas et al., 1995). This age, coupled with the presence of altered lithic fragments, indicates that there was eruptive activity at the same time as the Far Southeast intrusive magmatism and associated hydrothermal activity (secondary biotite age 1.41 ± 0.05 Ma, n = 6; Arribas et al., 1995). The feeder of this unit has not been identified, but it is indistinguishable chemically from Imbanguila and Bato rocks (Hedenquist et al., 1998). A hydrothermal breccia with altered and mineralized porphyry fragments, cemented by sulfide minerals, crops out directly above the Far Southeast deposit (Fig. 1).

Part of the Lepanto orebody (the fault-hosted main ore body; Figs. 3, 4), a minor portion of the Victoria-Teresa veins, and most of the Far Southeast orebody are hosted by basement metavolcanic or volcaniclastic rocks. Despite these basement host rocks, ore deposits in the Mankayan district are spatially and temporally related to the late Pliocene to Pleistocene events of intermediate-composition volcanism (Fig. 1).

The Lapangan tuff forms a thin and discontinuous cover of poorly consolidated dacitic air-fall tuff, which is mainly present in the center of the Mankayan district (Sillitoe and Angeles, 1985). Garcia (1991) reported an age of 18,820 ± 679 years from 14C of humic soil in the eastern part of the district.

Within the Mankayan district, the prominent fault system is a set of faults trending N 50° W and N 40° E (Fig. 1). These faults are considered to be part of the northernmost splay of the Philippine fault, the major tectonic lineament of the Philippine island arc (Sillitoe and Angeles, 1985). The northwest-trending part of the system comprises the steep, northeast-dipping Lepanto fault that hosts much of the Lepanto deposit, with open-space growth into fault-related veins (Gonzalez, 1956). A kinematic analysis of the faulting indicates that the N 50° W-trending faults (e.g., the Lepanto fault) may have formed due to strike-slip movement along the major fault system (Maleterre et al., 1988).

Mineralization in the District

The Lepanto high sulfidation deposit

The Lepanto high sulfidation orebodies are hosted in residual quartz and advanced argillic alteration zones, the latter consisting of hypogene alunite, dickite, kaolinite, pyrite and, locally at depth, diaspore and pyrophyllite (Hedenquist et al., 1998). Gonzalez (1956, 1959) provided much of the early framework for understanding the geology, alteration, and mineralization. Ores are closely associated with vuggy residual to massive residual quartz, collectively referred to as silicic alteration. Approximately 70 percent of the ore is hosted by the Lepanto fault, with strong brecciation of the silicic ore zone in part possibly resulting from synmineral movement that offset alteration zones; the balance of ore is contained in the subhorizontal blanket of the lithocap (Garcia, 1991). Early paragenetic studies (Gonzalez, 1959) noted that enargite and luzonite was followed by chalcopyrite, tennantite, and gold plus electrum with telluride minerals. Tejada (1989) and Claveria (1997, 1998, 2001) further described the paragenesis of Lepanto mineralization, with evidence for early coarse pyrite generated during the largely Cu- and Au-barren leaching event, which formed a core of residual vuggy quartz and halo of advanced argillic alteration. This was followed by the high sulfidation state minerals enargite and luzonite with fine pyrite, largely hosted by the silicic core, and finally by tennantite, chalcopyrite, sphalerite, galena, and tellurides plus selenides. The postenargite stage of sulfides is associated with the introduction of gold and was accompanied by the deposition of anhydrite plus barite gangue minerals. The tennantite of the gold event at Lepanto may be of intermediate- or high sulfidation state, depending on its composition (Jannas et al., 1999), although the presence of chalcopyrite indicates the former. Claveria (2000) noted a geochemical zonation in most ore-related elements in Cu-Au zones, without discriminating ores and wall rocks, from the southeast to the northwest along the trend of the Lepanto deposit. For example, Au, Cu, Sb, Se, and Te generally decrease toward the northwest and Ca decreases distinctly, whereas Ag and Zn slightly increase.

The Far Southeast porphyry deposit

The Far Southeast porphyry is a concealed deposit. The top of the porphyry-type mineralization is at an elevation of ~900 m, ~550 m below the surface, roughly coincident with the base of the enargite Au mineralization of the Lepanto ore-body. At an elevation of 100 m below sea level, the porphyry orebody is elongate in the direction of the regional northwest-trending structure. The Cu and Au grades are concentric around dikes and irregular intrusive bodies of melanocratic quartz diorite porphyry (Concepción and Cinco, 1989) emplaced in the basement. Potassic alteration consists of a biotite-magnetite ± K-feldspar assemblage and is associated with veins of vitreous, anhedral quartz. This alteration is partially to pervasively overprinted by alteration assemblages of chlorite plus hematite and/or white mica sericite-clay-chlorite (SCC; Sillitoe and Gappe, 1984). There is no definitive paragenetic evidence linking Cu sulfide minerals to the early veins of vitreous, anhedral quartz veins (Hedenquist et al., 1998). However, petrographic evidence shows that Cu sulfides are associated mainly with a later event characterized by the formation of euhedral quartz crystals with anhydrite (Hedenquist et al., 1998; Imai, 2000). Cathodoluminescence images show that the early anhedral quartz is overgrown by euhedral quartz (P. Redmond and J. Reynolds, pers. commun., 2000), the latter associated with sulfide deposition and mineral inclusions of illite (Hedenquist et al., 1998). Bleached halos of illite, centimeter to meter wide, accompany these euhedral quartz-anhydrite-white mica-hematite-pyrite-chalcopyrite-bornite veins, both of which cut sericite-clay-chlorite alteration. Gold in the Far Southeast deposit is present as free grains of electrum associated with chalcopyrite and bornite (Concepción and Cinco, 1989) and locally is accompanied by Bi-Te–bearing tennantite (Imai, 2000). Upward and outward from the core of economic porphyry mineralization the pervasive sericite-clay-chlorite assemblage grades from white mica-dominated with minor pyrophyllite locally to an assemblage in which pyrophyllite is abundant, variably accompanied by quartz, anhydrite, and kandite minerals (dickite, nacrite, and kaolinite). This pervasively altered rock is overlain and locally cut by a silicic zone with local alunite that hosts the southeast extent of the Lepanto ore deposit; the alunite halo includes a variable assemblage of anhydrite, diaspore, dickite, and/or pyrophyllite (Hedenquist et al., 2001).

The relationship between the Lepanto high sulfidation deposit and the Far Southeast porphyry deposit

Early ideas suggested that Lepanto was younger than high-temperature alteration of the underlying porphyry system (Sillitoe, 1983). Dating by Arribas et al. (1995), however, found that the biotite of the porphyry-related potassic alteration and the alunite in the halo to the silicic host of the high sulfidation orebody were essentially the same age, 1.41 ± 0.05 (n = 6) and 1.42 ± 0.08 Ma (n = 5), respectively. This study was the first to demonstrate a coeval age of potassic alteration of a porphyry deposit and its overlying advanced argillic alteration, the latter a lithocap that hosts a high sulfidation ore deposit. Shinohara and Hedenquist (1997) and Hedenquist et al. (1998) argued that these synchronous alteration events were related to a coupled hypersaline liquid and a low-salinity vapor, formed by separation as the solvus was intersected by critical fluid at depth. The hypersaline liquid remained at depth and caused the potassic alteration, as evidenced by fluid inclusions. The buoyant vapor ascended to shallower depths to form an acidic condensate that leached the rock and created the residual quartz (silicic) and quartz-alunite alteration.

A phyllic alteration overprint of the porphyry ore body followed, with illite ages of 1.37 to 1.22 ± 0.04 to 0.10 Ma (n = 10; Arribas et al., 1995). Stable isotopic studies of the biotite, alunite, and illite indicated that all formed from aqueous fluids with a dominantly magmatic origin, although the alunite-stable fluid was an acidic condensate of magmatic vapor with a variable meteoric water component, the latter progressively and regularly increasing as distance from the porphyry increased, to a distance 4 km to the northwest (Hedenquist et al., 1998). This study was the first to document in detail that the relatively low-salinity fluid responsible for the phyllic alteration was also magmatic in origin, rather than meteoric as previously thought. This detail followed the original suggestion of a magmatic origin of phyllic alteration by Kusakabe et al. (1990) based on their study of Chilean porphyry deposits.

Mancano and Campbell (1995) conducted the first fluid inclusion study of enargite and determined that there was a regular decrease within the Lepanto orebody of both homogenization temperature (305°–195°C) and salinity (~4–2 wt % NaCl equiv) with increasing distance from the porphyry, over a distance of >2 km. This supported the idea that the enargite (and gold) mineralization of Lepanto formed from fluids that originated from the vicinity of the Far Southeast deposit and flowed laterally to the northwest. Additional fluid inclusion and stable isotope results indicated that it was the phyllic stage of illite-stable fluid, initially an end-member magmatic fluid at ~350°C and 5 wt percent NaCl equiv that was likely responsible for the Lepanto mineralization (Hedenquist et al., 1998).

The Victoria and Teresa veins

The Victoria veins are located southwest of the Far Southeast porphyry, and at their closest point are within a few hundred meters of the porphyry (Figs. 1, 3). Imbanguila dacite porphyry and pyroclastic units host much of the veins, but the veins extend downward into the Balili volcaniclastic and Lepanto metavolcanic units (Fig. 3). Some veins have an arcuate shape, with the strike direction changing from northeast to southeast (Fig. 1). The Victoria veins pinch out upward at ~1,150- (0 zone) to ~1,100-m (4 and 8 zones) elevation, and they extend to below the 700-m-elevation level, locally to the 550-m level. Individual veins are mined over a 300-m vertical interval, have widths up to 8 m, and in general strike north-northeast with a dip to the southeast, with strike extents of up to 600 m. By contrast, the Teresa veins, subsequently recognized southwest of Victoria in the vicinity of the Nayak decline at an elevation of ~1,170 m, tend to occur in wider breccia zones and strike north-south, parallel to the northerly projection of the surface veins at Nayak. The veins crop out at ~1,200-m elevation at Nayak. To the north they are mined at an elevation of ~900 to 1,150 m in the Teresa sector, although narrow veins continue to 850- to 800-m elevation (Claveria et al., 1999a, b).

The alteration halos of the veins have been reported to consist of illite, or locally chlorite, rarely in direct contact with propylitic-altered wall rock (Claveria, 2001; Sajona et al., 2001, 2002). The paragenetic sequence of gangue and ore minerals in the Victoria veins (Claveria, 2001) consists of (1) an early quartz vein stage, associated with an intermediate sulfidation-state assemblage including chalcopyrite, tetrahedrite, and low Fe sphalerite, as well as pyrite and galena; (2) a carbonate stage of rhodochrosite, with similar sulfides; and (3) a late sulfate stage of anhydrite that is sulfide poor. Bornite and hematite are restricted to the early quartz stage, and gold was introduced during the later part of the quartz stage and through the carbonate stage, ending during the sulfate stage.

Sajona et al. (2001) reported homogenization temperatures of ~200° to 260°C and salinities of <1 to about 4 wt percent NaCl equiv for fluid inclusions in sphalerite, quartz, and rhodochrosite from Victoria. On the 1,000-m level in the north, fluid inclusions in rhodochrosite have higher homogenization temperatures than sphalerite. The homogenization temperatures of fluid inclusions in sphalerite, as well as the Fe content in sphalerite, indicate cooling during flow from the south to the north, with oxidation state increasing slightly on cooling, whereas the salinity trend is more complicated.

Claveria (2001) reported that the northwesternmost Victoria veins cut advanced argillic alteration and enargite related to the Lepanto deposit. Similar epithermal quartz veins were recognized earlier to cut enargite mineralization in the main Lepanto deposit, near its base (Garcia, 1991). Such phenomena were also observed in this study, for example, anhydrite + quartz + pyrite ± illite veins cut quartz ± alunite ± pyrophyllite ± diaspore ± dickite assemblages (Fig. 5). Where enargite and advanced argillic alteration is present in the Victoria area, it is clearly overprinted by quartz veins and related mineralization. Samples of illite from the Victoria veins were dated by the Ar-Ar method at 1.31 ± 0.02 Ma (Sakakibara et al., 2001) and 1.14 ± 0.02 and 1.16 ± 0.02 Ma (Hedenquist et al., 2001). For illite in the wall rock adjacent to veins, Sakakibara et al. (2001) reported illite Ar-Ar ages of 1.55 ± 0.03 and 1.31 ± 0.02 Ma and a K-Ar age of 1.50 ± 0.07 Ma. The dating results are generally in agreement with the paragenetic observation, with the ~1.5 Ma wall-rock alteration ages slightly overlapping with the Far Southeast biotite and alunite ages (1.41 ± 0.05, n = 6, and 1.42 ± 0.08 Ma, n = 5, respectively; Arribas et al., 1995). There is a possibility that the wall rocks of the Victoria veins were altered during the nearby Far Southeast-Lepanto event (~1.5 Ma) and were further altered during the subsequent Victoria veining event (1.31–1.14 Ma).

The Teresa veins were dated during this study. Although alteration around the Victoria and Teresa veins appears similar, a sample of illite from the 900-m level of the Teresa vein was dated at 2.22 ± 0.05 Ma by Ar-Ar analysis (App. 1), indicating that this portion of the vein system is older.

Other deposits and prospects

Porphyry alteration assemblages at Guinaoang include sericite-clay-chlorite that overprinted an initial potassic assemblage, superseded by sericitic alteration that has a K-Ar age of 3.5 ± 0.5 Ma (Sillitoe and Angeles, 1985). A relatively small (1.5 km2) lithocap of advanced argillic alteration overlies the porphyry mineralization. The Guinaoang lithocap locally hosts enargite mineralization similar to Lepanto, albeit much smaller in size (Sillitoe and Angeles, 1985; Trudu, 1992). At Buaki, 2 km west of Far Southeast, porphyry-style quartz veins and stockworks crop out at surface. Based on assays of surface samples and underground samples from the 1,150 m-level tunnel, a resource of 30 Mt grading 0.4 percent Cu and 0.5 g/t Au was estimated (Lepanto Consolidated Mining Co., unpub. internal report). The Palidan-Mohong Hill porphyry lithocap prospect 3 km to the south was tested by drilling and tunneling, and the resource is estimated at 100 Mt grading 0.4 percent Cu and 0.4 g/t Au (Gold Fields Asia Ltd., unpub. internal report). Alunite in the Mohong Hill lithocap was dated by K-Ar (A. Arribas, pers. commun., 1996) at 1.66 ± 0.32 Ma.

Surface Alteration Patterns

Lithocap occurrence

The most notable surficial alteration feature in the Mankayan district is the extensive advanced argillic alteration. Nearly continuous outcrops of quartz-alunite occur for over ~7 km from the north end of the airstrip south-southeast to the Palidan porphyry locality (Figs. 1, 2, 6a, b). The elevation of the outcrops ranges from ~1,050 to 1,600 m, mostly at or close to locations where the unconformity between the base of the Imbanguila dacite unit and the basement rocks is intersected by the current surface (Fig. 2). Such hypogene advanced argillic alteration with subhorizontal geometry due to lithologic control constitutes a lithocap (Sillitoe, 1995a).

The northern lithocap extends to the southeast below the surface to an area overlying the Far Southeast porphyry, according to underground information. The rocks on the surface above this section of underground lithocap, however, are fresh, except for some directly above the Far Southeast porphyry (Figs. 1, 2, 4). This ~4-km-long northwest-trending lithocap hosts the Lepanto orebody and is referred to here as the Lepanto lithocap, for clarity (Fig. 3). The dating of five alunite samples from the Lepanto lithocap, collected over a 4-km distance, from the porphyry to airstrip, indicates that the advanced argillic alteration formed at essentially the same time (1.42 ± 0.08 Ma; Arribas et al., 1995). There is a strong structural control on the lithocap, associated with the intersection of the Lepanto fault with the unconformity, a relationship first noted by Gonzalez (1956). The Lepanto fault is the major feeder and many branch faults functioned as subsidiary feeders (Garcia, 1991). Alteration along the faults below the subhorizontal part constitutes the root zones of the lithocap.

There are also patches of advanced argillic alteration located on the surface over the Far Southeast porphyry, 500 to 700 m above the unconformity (Fig. 2) at 1,370- to 1,450-m elevation. These zones constitute stacked lenses of advanced argillic alteration within the lithocap. Narrow faults containing alunite, including huangite (the Ca end-member alunite), are present in fresh Imbanguila dacite porphyry in deep gorges below the level of the surficial patches; these faults are believed to be the fluid conduits for the stacked lenses.

The Lepanto lithocap, as well as the high sulfidation mineralization hosted by portions of it, has been shown to be continuous and genetically related to Far Southeast porphyry by dating, fluid inclusion studies, and O-H isotope studies, summarized above (Arribas et al., 1995; Mancano and Campbell, 1995; Hedenquist et al., 1998). The Buaki porphyry is present on the western margin of the lithocap, but erosion has exposed the core of this small porphyry system, indicating that it is not related to the Lepanto lithocap; surface samples at Buaki contain porphyry-style quartz stockworks overprinted by dickite ± kaolinite that is interpreted to be part of the Lepanto lithocap. If there were a lithocap generated by the Buaki porphyry, it has been eroded away. In summary, there is no geologic, geochemical, or geophysical evidence for any other porphyry systems to be related to the Lepanto lithocap except for Far Southeast.

The southern portion of the lithocap that crops out in the district trends north-northwest. It partially covers the Victoria-Teresa veins, extends south to Mohong Hill (Fig. 6b) where it overlies the Palidan porphyry and continues ~0.5 km farther south. The potentially older age of Mohong Hill alunite (1.66 ± 0.32 Ma), compared to that of alunite samples in the lithocap northwest of the Far Southeast porphyry (1.42 ± 0.08 Ma), suggests that there may have been more than one intrusive center of volatiles that generated acidic condensates along the same unconformity, with leaching and alteration coalescing to form an apparently single alteration zone along the same permeable horizon. This is consistent with the Palidan porphyry that crops out below Mohong Hill (Figs. 1, 2). For this reason, this part of the lithocap alteration in the district is excluded from discussion below on zoning in the Lepanto lithocap.

Lithocap mineralogy and zoning

At Lepanto the lithocap is mostly composed of quartz-alunite ± pyrite, locally with pyrophyllite and/or diaspore. The alunite crystals are typically >60 μm in size with a flaky shape, indicating a hypogene origin, which is consistent with the alunite δ34S compositions of 22 to 25 per mil (Hedenquist et al., 1998). The principal cliff-forming rocks are composed of quartz-alunite, with the disseminated pyrite commonly oxidized at surface due to weathering. Within the quartz-alunite horizons the silicic centers locally show a vuggy texture, but massive silicic rocks are more common, largely along and proximal to the Lepanto fault. Most of the Lepanto ore is hosted in such massive silicic rocks, although not all the massive silicic rocks are mineralized. On the surface most of these massive silicic ± alunite rocks are only weakly anomalous in Au (<50 ppb), except at a few locations such as the Spanish workings. Cathodoluminescence studies of a few massive silicic samples show that they were originally residual quartz with a vuggy texture and the vugs were subsequently filled by euhedral quartz (Fig. 7). Quartz-alunite horizons are surrounded by dickite ± kaolinite (Fig. 4). At the end of the airstrip, the dickite ± kaolinite zone is ~20 m thick both above and below the near-horizontal quartz-alunite zone (Fig. 6a).

Zonation in the root zones of the lithocap is similar to the classic pattern noted by Steven and Ratté (1960) at Summitville, United States, with a silicic core that hosts mineralization changing laterally outward to quartz-alunite then dickite ± kaolinite (Fig. 4), over a scale of 10s of meters. In contrast, the zonation is vertical in subhorizontal lithocap horizons, with quartz-alunite in the middle and dickite ± kaolinite above and below (Figs. 4, 6a). The silicic core is well developed close to the major fluid conduits, such as the Lepanto fault, but is largely missing elsewhere. Laterally along the lithocap horizon there is no obvious zonation in texture or mineralogy from proximal to distal locations relative to the causative intrusion at Far Southeast, or in any other directions.

Mineralogy and zoning over intermediate sulfidation veins

The surface alteration above the Victoria veins is shown in Figure 8. There is weak white mica-pyrite alteration, grading outward to unaltered rocks that only show evidence of weathering. Except for a few millimeter-scale carbonate veinlets, no quartz or thick carbonate veins were found on the surface, despite the Victoria veins being up to 8 m wide a few hundred meters below.

Illite crystallinity (IC), a term introduced by Kubler (1967), can be used to indicate the formation temperature of white mica. Illite crystallinity was originally measured by X-ray diffraction (XRD) in which the XRD-IC was expressed as the half-height width of the 10-Å peak. The dominant control on IC is argued to be temperature (Frey, 1987), with higher temperature increasing the crystallinity of illite (i.e., larger crystals), resulting in a smaller XRD-IC value (sharper peak shape). Where illite crystallinity studies are determined from short wavelength infrared (SWIR) spectral features, the SWIR-IC is defined as the AlOH peak depth at ~2,200 nm divided by the H2O peak depth at ~1,900 nm on a hull quotient SWIR spectrum (Pontual et al., unpub. manual, 1997). In contrast to XRD-IC, the more crystalline mineral results in higher SWIR-IC values. Measurements of XRD-IC and SWIR-IC on the same samples show that the values have a good negative correlation, meaning that either can be used to determine IC (Chang, unpub. data) and thus indicate relative formation temperature; SWIR data are easy and fast to acquire.

Where the tops of the veins are less than ~250 m below the present surface, the outcropping alteration is mostly white mica + pyrite; chlorite is rare and epidote is not observed. The rocks typically also contain halloysite, but this mineral is believed to be caused by weathering. The SWIR-IC of white mica is 0.5 to 0.7 and the white mica is illite or interstratified illite and/or smectite according to XRD identification. Where the tops of the veins are deeper, between ~250 to ~350 m, white mica + pyrite alteration is still present at the surface but the SWIR-IC is smaller, 0.3 to 0.5. This indicates that the white mica is probably smectite, which has been confirmed by XRD analysis. At surface locations where the veins are >350 m below the present surface, only halloysite ± kaolinite ± smectite occur; the presence of igneous magnetite and lack of pyrite here suggests that this alteration is due primarily to weathering (Fig. 8a, b).

Geochemical Patterns: New Insights for Exploration

SWIR spectral variations of alunite

A total of 44 surface samples containing alunite were collected along a 7-km traverse of the quartz-alunite portion of the lithocap (Fig. 9a). Another 25 samples were collected from underground on the 1,150-, 900-, and 700-m levels, between the Lepanto deposit and the northern margin of the Victoria veins, to expand the distribution of alunite samples studied. Most alunite samples are of altered Imbanguila dacitic pyroclastic rocks, as indicated by the residual bipyramidal quartz eyes, whereas a few samples may have replaced rocks of the Balili unit. All the alunite samples were measured using the CODES PIMA-II instrument. In addition, the composition of 15 alunite samples was determined by laser-ablation inductively coupled plasma-mass spectrometry (LA-ICP-MS) for 24 elements (82 analyses), and of these, nine samples were analyzed by electron microprobe (83 analyses). Some other alunite samples from this district were also microprobed. The microprobe and LA-ICP-MS results are listed in Appendices 2 and 3, respectively.

Backscattered electron images indicate that the alunite is typically zoned, with zones 1 to 5 μm wide, which is mainly due to the variations in Na, K, and Ca contents. The major elements of alunite were analyzed using an electron microprobe with a 15-μm beam size. In each sample seven to 11 analyses were obtained. The composition variation within one sample is typically small compared with the variation between samples. For example, the mole Na/(Na + K) ratio of all the analyses ranges from 0.04 to 0.96, whereas the variation of this ratio within a sample is typically <0.30, although in low Na samples the range of this ratio is up to 0.51. Figure 9b further illustrates that the intrasample variation is small relative to the variation of the whole dataset. The Al and S content in all the alunite analyses have little variation, with the average and standard deviation being 20.12 ± 0.35 and 15.03 ± 0.31 wt percent, respectively; the Ca content is typically low but locally up to 3.89 wt percent. PIMA analyses with a ~2- by 10-mm window identified huangite in a few samples (spectrum shown in App. 4). Locally there are small APS minerals included in alunite. A few microprobe analyses of an APS mineral in one sample revealed that it is rich in Sr (10.60–12.61 wt %) with minor Ca (2.53–3.04 wt %), i.e., in the svanbergite-woodhouseite series.

In SWIR spectra, K-Na alunites have a strong absorption feature at about 1,480 nm wavelength (App. 4). The exact position of this feature shifts, related to alunite composition (Thompson et al., 1999). At Mankayan, at least three SWIR spectra were measured on each sample. The variation in the wavelength position of this feature in each sample is typically <2 nm, with only a few exceptions that were up to 4 nm. The average composition of alunite shows a correlation between the alunite Na/(Na + K) ratio and its ~1,480-nm feature, with higher wavelength position corresponding to higher Na content in the alunite (Fig. 9b), despite the variations in both the ratio and the wavelength position for each sample. The higher Na content in turn has been demonstrated to correlate positively with formation temperature, supported by alunite experiments (Stoffregen and Cygan, 1990).

In the Mankayan district, the alunite feature at ~1,480 nm in SWIR spectra varies between 1,479 and 1,495 nm. The wavelength peak position of alunite in the Lepanto lithocap generally decreases with increasing distance from the intrusive center (Fig. 9a), particularly to the far northwest extent of the lithocap. Closer to the Far Southeast porphyry deposit, the peak positions generally have higher wavelengths, regardless of the sample elevation, i.e., from the surface (~1,350-m elev) or underground (1150, 900, and 700 mL); in particular, the wavelengths of the most distant alunite are all low. The samples from lithocap outcrops in the southern part of the district may be related to the Palidan porphyry, but this remains to be tested with further study and dating. Microprobe results of alunite, hosted by Imbanguila dacite, show that alunite is more Ca rich closer to the Far Southeast porphyry (Fig. 9c). Huangite, the Ca end member of alunite, has only been found over the Far Southeast porphyry deposit (Fig. 9c).

Mineral chemistry of alunite

Trace elements in alunite analyzed by LA-ICP-MS included Ca, Sr, La, Ce, Nd, Sm, Eu, Gd, Dy, Er, Yb, Lu, Zr, Ba, Au, Ag, Pb, Sb, Bi, Mn, Fe, Cu, As, and Se. The internal standard used for data reduction was aluminum, which showed little variation in the 83 microprobe analyses. Six to 10 LA-ICP-MS analyses were obtained from most of the samples. Similar to the major elements, variation in these trace elements among analyses of the same sample is typically smaller than the variation of the whole dataset. For example, the Pb of the all the analyses ranges from 1 to 3,616 ppm, whereas Pb variation in analyses of individual samples (max-min) are mostly <1,000 ppm, particularly in low Pb samples (Fig. 9d). The intrasample variation is generally larger in samples containing more Pb, with the largest intrasample variation being ~3,400 ppm in the sample with the highest Pb content (App. 3). Despite the variation in each sample, the samples have significantly different compositions (e.g., Fig. 9d). All the elements plus some element ratios were plotted on base map to examine spatial trends; multiple analyses for each sample were plotted individually, without averaging for each sample, and are represented by the multiple circles at each location (e.g., Fig. 10a, and all other diagrams of this type). These results indicate that the Pb content in alunite from the Lepanto lithocap is lower closer to the intrusive center (Fig. 10a). Silver in most of the analyses is below the detection limits, with samples containing detectable Ag (0.2–3.5 ppm) distal to the intrusive center (App. 5a). In contrast, the Sr content and rare earth elements (REE), represented by La (Apps. 6a, 7a) and particularly the La/Pb and Sr/Pb ratios (Apps. 6b, 7b), generally increase closer to the intrusive center.

Lithocap whole-rock geochemical signatures and patterns

We determined the geochemical signatures of different types of mineralization and associated alteration by analyzing whole-rock compositions, including major elements and 57 trace elements (ACME Lab, Vancouver, Canada). The elements, sample digestion methods, analytical methods, and detection limits are described in Appendix 8. Results below the detection limits were given values of half the detection limit for the purpose of data plotting and interpretation. In total 141 samples were analyzed, with 103 from the surface (App. 9). In addition, 105 previous analyses of whole-rock samples from Lepanto underground workings, many of them constituting ore, were also available for this study (Claveria, 1997).

Gold and related elements

The gold anomalies in the lithocap at the surface, associated with quartz-alunite as well as dickite ± kaolinite zones, are instructive for exploration and assessment of lithocaps elsewhere. A total of 98 analyses of lithocap samples were obtained, of which 60 were from the surface, and these included silicic, quartz-alunite, and dickite ± kaolinite alteration types. Three samples of silicic alteration along structures at the surface contain enargite and >1 g/t Au. They are not considered here for this assessment of gold signature in lithocap alteration, as they represent the later mineralization event, distinct from the early alteration event (Hedenquist et al., 1998).

Lithocap samples were collected from above Far Southeast and south of the Buaki 1,150 mL portal to ~4 km northwest at the end of the airstrip. Samples close to Buaki are about 100 m west of the surface projection of the Lepanto orebody, whereas samples at the end of the airstrip are nearly 2 km northwest from underground Lepanto ore (Fig. 11). This suite of 25 samples returned gold values of 1 to 46 ppb, averaging only 12 ppb. A silicic sample in this area from the Lepanto fault structure contained 0.7 wt percent Cu as enargite, and 1.1 g/t Au, and is excluded from this average. Other samples of lithocap at the surface above distal enargite veins, northwest of the surface projection of the Victoria and Teresa veins, contain 1 to 70 ppb Au, with the 17 samples averaging 27 ppb. Two other samples from the surface projection of a distal enargite vein, both associated with silicic alteration along a structure, contain 1.1 and 4.0 g/t Au and 0.4 and >2 wt percent Cu, respectively, and are excluded from this consideration. Farther south, up to 0.5 km south of Mohong Hill, 13 lithocap samples, possibly related to the Palidan porphyry, contain 4 to 500 ppb Au, averaging 133 ppb. Two samples in quartz-alunite lithocap close to the Guinaoang porphyry deposit contain 23 and 34 ppb Au.

The distribution of Au in lithocap samples does not show any clear patterns related to the position of the Lepanto ore-body, nor does Cu, As, Te, and other ore-related elements, except for higher values in samples of silicic alteration associated with structures. The most notable result is the lack of strong Au enrichment (mostly <50 ppb) in the quartz-alunite lithocap as close as a few hundred meters from the largely blind enargite Au orebody.

Trace element signatures in alunite-bearing lithocap samples

Although gold and other metals do not define any obvious patterns when data from all of the lithocap samples are plotted together, a selected sample set does have some elements and ratios that show trends relative to the causative intrusive center. From the whole sample set, mineralized samples, i.e., defined as containing ≥0.1 wt percent Cu and/or ≥100 ppb Au, were first filtered out. In addition, samples not containing alunite were also excluded. The remaining samples from the Lepanto quartz-alunite lithocap show clear spatial trends relative to the Far Southeast porphyry in the following elements and ratios: Hg, Pb, Ag, Sr/Pb, La/Pb (La as a representative of all REE), and Ag/Au.

Mercury is low (less than ~130 ppb) above or close to the Far Southeast porphyry and is higher to the northwest, along the inferred major fluid-flow direction of the Lepanto lithocap (Fig. 12a). About 4 km to the northwest, at the end of the airstrip, Hg concentrations are the highest, about 4,000 ppb. When all nonmineralized samples (Cu < 0.1 wt % and Au <100 ppb) of all advance argillic alteration types are plotted together, including those without alunite, this trend is less clear (Fig. 12b). Ore samples, which are much higher in Hg (up to 38,000 ppb), further obscure this trend.

Lead also shows a trend of higher values with increasing distance from the Far Southeast intrusive center and the enargite orebodies (Fig. 10b). At proximal locations, the Pb content in the alunite-bearing samples is <40 ppm, whereas about 1 to 2 km northwest of the surface projection of the enargite orebodies, Pb increases to about 90 to 150 ppm. In contrast with Hg, Pb does not increase significantly until far from the intrusive center. When all nonmineralized samples are plotted, including those without alunite, the trend in Pb is obscure (Fig. 10c). Mineralized samples contain much higher Pb (up to 14,700 ppm); adding them further obscures the trend. Other elements or ratios that show trends in nonmineralized, alunite-bearing rocks include Ag, Ag/Au (both decrease toward the porphyry center; App. 5b, c), and Sr/Pb and La/Pb ratios (both increase toward the porphyry center; Apps. 6c, 7c). Other rare earth elements show similar behavior as La.

The fact that clear geochemical trends are detected only in alunite-bearing samples suggests that the anomalies are related to alunite. Normalizing the whole-rock Pb to moles of (Na + K), assuming the Na and K in the whole rock mostly resides in alunite of the quartz-alunite lithocap, improves the trend (Fig. 10d), supporting the contention that Pb in non-mineralized samples is largely hosted by alunite. This assertion is confirmed by LA-ICP-MS analyses of alunite (Fig. 10a), which show that alunite contains up to ~3,600 ppm Pb in the lattice; the alunite mineral chemistry has the same trend as the whole-rock compositions.

Halos above intermediate sulfidation veins

At the present surface 150 to 450 m above the Victoria and Teresa veins, there are weak but detectable As (Fig. 13a) and Se (Fig. 13b) anomalies. Above the surface projections of the veins, As is typically >18 ppm, whereas rocks a few hundred meters from these positions contain less than this amount. For Se the anomalous level above veins is >0.5 ppm; there is a poor correlation between Se and S. For reference, the least altered rock contains 0.2 ppm As and 0.1 ppm Se. Other elements out of the 57 analyzed do not show any consistent difference in rocks at the surface projects of the veins and some distance away. For example, the Au values range from below the detection limit (0.2 ppb) to 46 ppb above the veins, compared to <0.2 to 23 ppb away from the veins laterally. Silver content at the surface above the veins ranges from 8 to 214 ppm, compared to 10 to 240 ppm a few 100 m away from the surface projection of the veins. Thus, geochemical anomalies above the upper limit of the veins are subtle but real for a limited group of elements.


In August 1997, following recognition of the economic significance of the Victoria veins to the south of the then-depleted Lepanto orebody, Lepanto Consolidated Mining Co. decided to fly an airborne geophysical survey over the prospective area. The survey was contracted to World Geoscience and was flown in a helicopter with lines 100 and 200 m apart oriented north-south, with nominal sensor terrain clearance of 40 m. Magnetic data were collected using Scintrex equipment with intrinsic resolution of 0.0001 nT, a cycle rate of 0.1 sec and sample interval of 4 m. Radiometric data were collected using a Picodas 256 channel spectrometer with a cycle rate of 1 sec and sampling interval of 40 m, connected to a 16.75 l NaI crystal. Total magnetic intensity, digital terrain model, and standard 4 channel radiometrics (K, U, Th) were provided as located and gridded data. The discussion presented here relates to a 7-km2 area centered on the Victoria veins, flown at 100-m line spacing; this area covers all the known significant deposits in the Mankayan district.

The Mankayan district is particularly challenging for interpretation of aeromagnetic data, as it contains multiple shallow-dipping volcanic units of different ages and compositions with multiple intrusions and extensive hydrothermal alteration. As a result, the magnetic susceptibility can be expected to vary widely over short distances, and remanence effects must also be anticipated. In addition, the area lies close to the magnetic equator, so magnetic features are displaced as a result of the shallow (20°N) magnetic inclination. In processing of magnetic data acquired at low latitudes, spatial frequency domain reduction to the pole transformations of total magnetic intensity data are used to shift anomalous features to overlie their source. The magnetic data acquired by World Geoscience were presented as images in which reduction to the pole transformations had been done (Fig. 14).

The magnetic character of stratigraphic units was assessed both by magnetic susceptibility measurements on samples and by correlating mapped geology with the aeromagnetic survey results. The basement Lepanto metavolcanic and Balili volcaniclastic rocks are both variably magnetic, with many small localized anomalies recorded by the aeromagnetic survey. The Imbanguila dacite porphyry is effectively nonmagnetic close to the porphyry ore deposit but moderately magnetic elsewhere. The Imbanguila pyroclastic rocks are variably magnetic. The postmineralization Bato pyroclastic deposits are thin (100–200 m) with low measured susceptibility: their apparent character in the aeromagnetic survey may be due to underlying Bato dacite intrusions or older units.

The magnetic expression of mineralization was examined by overlying known orebodies on the magnetic image. It is apparent that the Far Southeast orebody lies in a zone of lower reduction to the pole total magnetic intensity, corresponding to the extensive zone of magnetite destructive alteration; this was due to the phyllic overprint of potassic alteration at depth, as well as more shallow advanced argillic alteration of fresh rock. There is no apparent indication of the deep magnetite-rich potassic alteration that occurs below ~300- to 400-m elevation, i.e., ~1 km below the surface (average susceptibility for core from the Far Southeast potassic zone is 0.05 SI). Forward modeling of a 400-m-diam cylinder with a depth extent of 1,000 m and a top 900 m below the surface indicates that the expected magnetic response of the deep potassic zone is broad and low amplitude, and thus would not be detectable in the presence of other anomalies. The Guinaoang porphyry deposit to the southeast occurs in an area of subdued response and also has no apparent magnetic expression. The Lepanto high sulfidation orebody, and its host lithocap, which has a shallow southeast plunge toward the Far Southeast porphyry, has no apparent magnetic expression. There is a general, although not exact, association between the Victoria and Teresa vein systems and zones of demagnetization or reverse magnetization.

Comparison of topographic lineaments interpreted from the digital elevation model derived from the airborne geophysical data, magnetic lineaments interpreted from the reduction to the pole total magnetic intensity, and the geologic map (Fig. 1) shows that many northwest-trending topographic lineaments correspond to mapped faults. Some of the northeast-trending lineaments, as defined by topography and magnetic data, correspond to mapped faults, but there are many additional lineaments that may represent other structures.

The radiometric image clearly maps out the basement units (Lepanto metavolcanic rocks and Bagon intrusion). The Imbanguila and Bato units have a variable response, between and within units. However, there is no obvious radiometric expression of any of the mineralized systems.

Summary and Discussion

Genetic link between porphyry and lithocap: the basic processes

The genetic link between porphyry-style alteration-mineralization and lithocap was implied in Sillitoe’s (1995a) definition of lithocap. This genetic link was demonstrated in the Mankayan district through field observation, radiometric dating, fluid inclusion studies, and stable isotope studies (Arribas et al., 1995; Mancano and Campbell, 1995; Hedenquist et al., 1998). Typically the formation of porphyry-related alteration involves the early-stage separation of hypersaline liquid and vapor as a magmatic critical fluid intersects its solvus (Henley and McNabb, 1978). The hypersaline liquid causes the deep potassic alteration, whereas the vapor, being buoyant, ascends toward the surface along structures. Magmatic gases such as HCl and SO2 fractionate to the vapor and, after condensing into ground water, form an acidic liquid. The condensate becomes more reactive as it cools and the acids increasingly dissociate, and this causes leaching, forming residual quartz cores with halos of advanced argillic alteration (quartz-alunite ± dickite-kaolinite; Ransome, 1907; Steven and Ratté, 1960; White, 1991; Rye, 1993; Hedenquist et al., 1998). If the ascending, structurally controlled acidic fluid encounters a permeable horizon, either an unconformity or one or more rock units, the reactive fluid will flow along and cause intense leaching and alteration in that layer, creating a lithocap above and lateral from the causative intrusion (Sillitoe, 1995a; Hedenquist and Taran, in prep.).

At Mankayan, the intrusions related to the Far Southeast porphyry deposit exsolved the fluid which subsequently separated to hypersaline liquid and vapor (Shinohara and Hedenquist, 1997), the latter subsequently forming an acidic condensate. The unconformity between the basement and Imbanguila units channeled the flow of this condensate, facilitated by the Lepanto fault, and the leaching and alteration along the path produced the Lepanto lithocap (Hedenquist et al., 1998). The lithocap extends prominently to the northwest from the Far Southeast porphyry, rather than in all the directions, due in part to the nature of the unconformity and the local geology. The unconformity has a cone shape with two vent-related depressions (Fig. 2). It is speculated that the vapors and subsequent acid condensates ascended from the northwestern depression, the northwestern side of which is less steep, and therefore the condensates could flow unobstructed to the northwest. In contrast, the southeastern side of the cone has a steeper slope and turns downward toward the southeastern depression at 900- to 1,000-m elevation (Fig. 2). Any condensate ascending along the unconformity to the southeast would lose its permeable channel at this elevation, as hot fluids tend not to move downward. The rock above this section of the unconformity is mostly dacite porphyry and it has low permeability (Fig. 1), which would hinder further fluid ascent and flow to the southeast. In summary, the geometry of the unconformity and the spatial distribution of the rock units may have caused the asymmetry of the lithocap.

All lithocaps have structural roots, or feeders, along which the acidic condensate ascends. But not all structurally controlled zones of hypogene advanced argillic alteration have associated lithocaps (Sillitoe, 1999), either because they did not intersect permeable horizons, or such horizons have subsequently been eroded. If there are multiple permeable layers or perched water tables, it is possible that multiple, stacked horizons of lithocap will form. In the Mankayan district, the patches of advanced argillic alteration above the Far Southeast porphyry lie far above the basement-Imbanguila unconformity (Fig. 2) along which the majority of the advanced argillic alteration occurs. These patches are probably connected by fractures with the main lithocap below and thus constitute perched horizons. Erosion in the Mankayan district has been minimal since the formation of the Far Southeast-Lepanto system, <500 m based on pressure estimates from fluid inclusion studies at Far Southeast (Hedenquist et al., 1998), in part due to its youth, meaning that only a portion of the quartz-alunite lithocap is presently exposed (Fig. 2; compare the elevation of the unconformity with that of outcropping quartz-alunite, formed at and above the unconformity).

Formation of high sulfidation epithermal deposits: a two-stage process

Hypogene advanced argillic alteration forms during the initial stage of porphyry development, during vapor discharge. Mineralization, if present, forms in a subsequent stage (White, 1991), as seen at Lepanto (Hedenquist et al., 1998) and other high sulfidation deposits (Sillitoe and Hedenquist, 2003). Hedenquist et al. (1998) demonstrated that the Lepanto high sulfidation mineralization is caused by the phyllic-stage lower salinity fluid, following early alteration that produced potassic alteration at depth and the lithocap alteration closer to the surface.

In the Mankayan district, only part of the lithocap is mineralized. The orebodies are located at deeper positions and relatively close to the intrusive source (Figs. 1, 3). The majority of ore (~70%) lies in structural roots, principally the Lepanto fault (Fig. 4). Within the lithocap that extends ~4 km northwest from the intrusion, mineralized roots occur up to ~2.5 km from the intrusive source, whereas most of the mineralized horizons are within ~1.5 km of the intrusion (Fig. 1). Lithocaps elsewhere show many similar characteristics that in general indicate the potential for high sulfidation mineralization, but most are not strongly mineralized. One reason for this may be due primarily to hydrologic reasons, i.e., the subsequent mineralizing fluid that comes after the early acidic condensate may not be able to ascend to the elevation of the lithocap, especially to the upper horizons of a perched lithocap. For example, the patches of advanced argillic alteration at the surface directly above Far Southeast are barren; likewise, the upper lithocap horizon over the Quimsacocha deposit, Ecuador, is also largely barren (IAMGold staff, pers. commun., 2007). Alternatively, the later mineralizing fluid may have been minor or nonexistent (Einaudi et al., 2003). Where the deeper porphyry environment is exposed, the mineralized portion of the lithocap may have been eroded (R.H. Sillitoe, pers. commun., 2007).

Zonation in mineralogy and texture

At Mankayan, the alteration halo to structural feeders of the Lepanto lithocap and ore deposit are zoned (Fig. 4), similar to the classic zonation pattern at the structurally controlled Summitville high sulfidation epithermal deposit (Steven and Ratte, 1960); the alteration zonation, from structure to margin, commonly has a scale of 10s of meters. The horizontal part of the lithocap shows a vertical zonation, with quartz-alunite in the middle and dickite ± kaolinite above and below, also at a scale of 10s of meters. Silicic zones at Lepanto, vuggy or massive in texture, largely occur in proximal locations close to the feeder structure(s) and tend to be absent in distal locations, similar to other lithocap horizons (Sillitoe, 1995a). For exploration of lithocaps up to tens of square kilometers in size, such local alteration zonation is of limited use. At the kilometer scale, there is no systematic mineralogical or textural zonation relative to the intrusive source in the Lepanto lithocap.

Signatures in alunite

Alunite is commonly one of the most abundant minerals in lithocaps, as at Mankayan. SWIR spectral features of alunite and some trace elements in alunite show indications of systematic variations relative to the Far Southeast porphyry source. The alunite absorption peak at ~1,480 nm of the SWIR spectrum shifts to higher wavelength positions closer to the Far Southeast porphyry (Fig. 9a). The Pb content in alunite is lower closer to the intrusive center (Fig. 10a), whereas Sr, Sr/Pb, La, and La/Pb increase (Apps. 6a, b, 7a, b).

The variation in the ~1,480 feature of alunite is related to its Na/(Na + K) composition (Fig. 9b), which is directly correlated to the temperature of formation (Stoffregen and Cygan, 1990). This is consistent with the spatial distribution of the samples relative to the intrusive heat source, with Na-rich samples closest to the Far Southeast porphyry.

Lead in alunite is higher in distal positions (Fig. 10a), probably because Pb substitutes for K, as the ionic radius of Pb2+ (1.32) is closer in size to K+ (1.33) than to Na+ (0.98). Since alunite with higher K forms in distal positions at lower temperature compared to alunite with higher Na, Pb also has a higher concentration in distal alunite. The elements Sr2+ (1.27) and La3+ (1.04) are probably more related to Ca2+ (1.06) than to Na+ or K+, due to both ionic radii and charge considerations. Their contents in alunite have a trend that increases near the intrusive center (Apps. 6a, 7a), consistent with the distribution of high Ca alunite (up to 3.9 wt %) and huangite (Ca alunite) closer to Far Southeast (Fig. 9c). The La/Pb and Sr/Pb ratios magnify these signals. In addition to crystal lattice effects, the solubility of these trace elements at different temperatures may also have played a role. For example, Pb may have a higher solubility at higher temperature in proximal locations, resulting in its transport to cooler, distal location, where it is then incorporated into alunite.

Signatures in whole-rock composition of lithocap samples

The Lepanto lithocap does not have significant anomalies of Au. Lithocap samples of quartz-alunite, even only a few hundreds of meters away from the Lepanto orebodies, contain <50 ppb Au. Gold is not appreciably introduced during the early vapor condensate-related leaching and lithocap development, consistent with the extremely low Au content of near-surface, low-pressure magmatic vapors (Hedenquist et al., 1994; Hedenquist, 1995), in contrast to the higher metal contents in high-pressure volcanic vapor (Hedenquist, 1995), and in vapor-rich inclusions from high-pressure porphyry depths (Heinrich, 2005). This lack of Au in the early, near-surface (i.e., low-pressure) vapors explains the <50 ppb Au anomaly in whole rock within the Lepanto lithocap. Gold content is also low in alunite, dominantly below the detection limits (App. 3). The detection limits are mostly 0.1 to 0.2 ppm, with a few up to 0.6 ppm. As the ionic radius of Au+ (1.37) is similar to K+ (1.33), it is argued that Au would have substituted in alunite if Au had been present at the time of alunite formation. When all whole-rock samples were plotted without discrimination, patterns were not found for any other elements of the 57 elements (Apps. 8, 9; App. 9 is in a digital supplementary) analyzed, as well as many element ratios.

However, this study did find geochemical patterns zoned in the Lepanto lithocap relative to the Far Southeast porphyry, the intrusive source, in selected suites of samples from the Lepanto lithocap, namely nonmineralized (Cu <0.1 wt % and Au <100 ppb), alunite-bearing samples. As noted for this lithocap, Pb, Hg, Ag, and Ag/Au decrease and La/Pb and Sr/Pb increase toward the intrusive center (Figs. 10b, 12a; Apps. 5b, c, 6c, 7c). Selecting non-mineralized samples restricts the samples to all having formed in the early leaching and alteration stage, relatively free of overprinting by later mineralizing fluids which may complicate trends. Consequently, these patterns indicate the direction to the intrusive source of volatiles. Filtering out samples without alunite constrains the samples to be formed under similarly acidic conditions, and to have the same mineralogic nature. The sample filtering practice is essential for selecting rocks formed at approximately the same time and under similar conditions, and by fluids dominated by magmatic characteristics. The selected samples can then be more reliably compared to one another, even though variations in geology may also have an effect.

The trends shown in alunite compositions and the composition of non-mineralized, alunite-bearing whole-rock samples from the lithocap indicate the direction to the paleo-thermal source for the initial-stage leaching and alteration event. Such trends do not point to ore but they help to indicate areas with a higher potential for mineralization, because porphyry-style mineralization occurs in and adjacent to the intrusive source, and high sulfidation mineralization also tends to occur proximal to intrusive source(s) and related structures. To use the vectors, the outcrops of lithocap alteration are assumed to be genetically linked. Any evidence of such a relationship, primarily geological relationships but also dating, increases the confidence in the results.

Surface signatures of concealed intermediate sulfidation epithermal veins

At the present erosional surface above the Victoria intermediate sulfidation veins, there are subtle alteration (illite to interstratified illite/smectite to smectite + pyrite) and geochemical (As, Se) expressions. Combining the results of alteration mineralogy, illite crystallinity, and geochemical anomalies may improve the targeting of such a vein prospect, despite the anomalies being weak. In a district such as at Mankayan with known porphyry and/or high sulfidation deposits, or even a district containing a barren lithocap, alteration and geochemical anomalies can nevertheless provide indications of the location of vein targets worthwhile to drill test. The subtlety of signatures is strongly dependent on the erosion level; therefore, even if such indications are not present at some erosional surfaces, e.g., >350 m above the upper extent of veins—similar to those in the Mankayan district—the potential of the district should not be ignored if there are other indications of mineralization potential, e.g., mineralized fragments in an exposed diatreme, or within 2 to 3 km of a porphyry deposit or lithocap.

Geophysical signatures

In the Mankayan district airborne magnetic imagery is complex and difficult to interpret, reflecting the combined effects of low magnetic inclination, magnetic topography and remnant magnetization in an area with complex and variable volcanic stratigraphy and many intrusions. The airborne geophysical data have not provided any signatures that can be confidently related directly to ore bodies. There is no direct surface expression of the magnetic alteration in the ~1,000-m deep core of the Far Southeast porphyry deposit, but the ore body occurs in a large zone with obvious demagnetization, providing an indirect guide. Importantly, this survey indicates that not all porphyry deposits are associated with a positive magnetic anomaly, if magnetite-bearing potassic alteration is overprinted by other alteration types. There are several magnetic anomalies of relative lows present in the Mankayan district (Fig. 14), but the one above Far Southeast is special in that it is at the end of a ~4 km elongated lithocap. This indicates that integrating geophysical findings with geological knowledge may help increase confidence in targeting. In a district that contains many low magnetic anomalies, those closest to the margin of a large lithocap may deserve priority in follow-up investigation.

The Guinaoang porphyry deposit and the Lepanto high sulfidation deposit have no apparent geophysical expression. The Victoria and Theresa vein systems can be tentatively related to a zone of probable demagnetization, consistent with the white mica-pyrite alteration. Radiometric data reflect bedrock units, but show no effects of mineralization. Numerous linear features are apparent according to digital elevation model and reduction to the pole total magnetic intensity images. Most have been confirmed to be faults by mapping, whereas some northeast-trending lineaments have not been recognized in mapping and deserve exploration as possible sites for mineralized veins. Thus, the geophysical survey, while not pointing directly to ore, has provided potential targets for investigation in the Mankayan district.

General Implications for Exploration

The presence of a lithocap indicates an epithermal level of erosion, and the potential for epithermal and/or porphyry mineralization nearby. This is clearly demonstrated in the Mankayan district and at many other locations worldwide. New exploration tools and lessons that can be learnt from this study in the Mankayan district include the following points.

Alteration mapping, aided by SWIR equipment, is essential to assess the advanced argillic lithocap environment, as is mapping of lithology and structures. However, alteration by itself may be insufficient to point to the causative intrusive source. Vectors found in this study include the following: (1) the alunite peak position at ~1480 nm on SWIR spectrum shifts toward higher wavelength in samples that are closer to the intrusive source of acidic condensates; (2) in alunite the Pb content decreases closer to the intrusive center, whereas Sr, La, La/Pb, and Sr/Pb increase; (3) the whole-rock composition of alunite-containing, non-mineralized (<0.1 % Cu and <100 ppb Au) lithocap samples can also point to the intrusive source: Hg, Pb, Ag, and Ag/Au decrease and La/Pb and Sr/Pb increase toward the intrusive center. Normalizing whole-rock Pb to the (Na+K) moles is a proxy for the alunite mineral composition, and as demonstrated in the Mankayan district, the ratio provides the same indication as alunite compositions. This is useful, as whole-rock analyses are much less expensive, with a faster turnaround time, than LA-ICP-MS analyses. These vectors point to the causative intrusive source that is the potential center to mineralization, including porphyry and high sulfidation deposits.

The gold anomaly in the quartz-alunite alteration of a lithocap can be quite low (<50 ppb), even within a few hundred meters of the surface projection of underground ores. A prospect should not be discarded only based on a low level of gold anomaly in quartz-alunite alteration, particularly if the lithocap is large. It is critical to locate the structural feeders of lithocaps, as high sulfidation mineralization—if present—is most likely to be concentrated there. In addition, the presence of silicic alteration is much more indicative of prospect potential than pervasive quartz-alunite or clay alteration, particularly that associated with structures. Stacked lithocap horizons may occur, and in such cases the lower layers have a better chance of being mineralized (e.g., Quimsacocha).

Low magnetic anomalies on the margin of a large lithocap, particularly at the end of an elongate, structurally controlled lithocap, deserve special attention, as the anomaly may be caused by demagnetization due to porphyry-style alteration, e.g., either phyllic or advanced argillic overprint of potassic alteration.

White mica-pyrite alteration, coupled with various elements of the epithermal suite—both As and Se at Mankayan—may indicate the presence of concealed intermediate sulfidation epithermal veins, especially in a district with known porphyry or high-sufidation mineralization or large lithocaps. The signatures are strongly dependent on the erosion level, and are typically quite subtle in positions several hundred meters over veins, and hence an assessment of the degree of erosion is essential.


We are grateful to Mr. Bryan Yap, President of Lepanto Consolidated Mining Company, for permission to conduct this study in the Mankayan district, for providing access to the air magnetic and radiometric survey results, and for permission to publish the outcome of this study. This work is part of AMIRA Project P765, completed in December, 2006, at CODES, University of Tasmania. The project was sponsored by Anglo American, Anglo Gold Ashanti, Gold Fields, Newcrest, Newmont, Placer Dome, and Teck Cominco, as well as Barrick in the final year, with additional funding provided by the Australian Research Council. We acknowledge the discussion and ideas provided by representatives of the sponsoring companies, and assistance by the AMIRA research coordinator Alan Goode. We thank Lyndon Bradish and Froilan Conde, plus many other Lepanto staff, including Bene, Danny, Ed, Louie, Perfecto, Perry, Ric, and Willy, for their assistance, and Dave Braxton for help in the field. Paddy Waters of Anglo American Philippines generously provided logistic support and funding for some of the whole-rock geochemical analyses. We appreciate the help from CODES staff, including Simon Stephens for sample preparation, Sarah Gilbert and Leonid Danyushevsky for LA-ICP-MS analyses, and June Pongratz for report preparation. We thank Roger Stoffregen for comments on an early version of the manuscript, and John Thompson and Raymond Jannas for useful reviews.

A digital supplement to this paper is available at <http://economicgeology.org/>
© 2011 Gold Open Access: this paper is published under the terms of the CC-BY license.

Supplementary data