This paper describes and reviews the use of calcrete (caliche) for Au exploration in the southern Yilgarn Craton of Western Australia. There are very strong correlations between Au and the alkaline earth metals, particularly in residual soils. However, where transported overburden exceeds about 10 m, the use of calcrete to indicate underlying Au mineralization is severely limited. Even where transported overburden is less than 10 m, other factors, such as thick clay, may restrict geochemical signals from reaching the calcrete. In some cases, gold in calcrete occurring over mineralization may have been sourced from upslope.
Several lessons are to be learnt from calcrete sampling in the Yilgarn WA. Some of the more important being:
Calcrete is an excellent sample medium for Au exploration because its physico-chemical properties appear to mimic those of the Au species found in soil, so that Au and Ca commonly accumulate in the same part of the soil profile.
Calcrete consists of a variety of materials and morphologies derived from the parent material on and within which it develops, although dominated by calcite and dolomite.
Gold found in calcrete may have been transported laterally up to several hundred metres – that is, its presence does not necessarily indicate Au mineralization directly below.
Regolith classification (including mapping and profiles) is essential to correctly interpret Au anomalies developed in calcrete.
If the thickness of transported overburden exceeds 10 m and/or if lacustrine sediments are present, experience to date suggests that calcrete will probably be ineffective as a sample medium.
For the last decade, many Australian companies have been using calcrete as a preferred geochemical sample medium for Au exploration. Their experience, coupled with detailed scientific research, is both considerable and significant. Indeed, several lessons are to be learnt from the use of calcrete for Au exploration, particularly from the Yilgarn Craton where the technique was first used. Calcretes (or caliche) are secondary carbonates, principally consisting of calcite and dolomite, that may precipitate in regolith where the average annual rainfall is less than about 600 mm. They commonly also contain parent material (e.g. colluvium, laterite, rock and saprolite) on and within which they form. Pedogenic calcretes are those that form in unsaturated (vadose) soil horizons and are widely distributed in the Yilgarn Craton, south of about 30°S (the ‘Menzies Line’, ⇓Fig. 1; ⇓Butt et al. 1997), seemingly more abundant over more basic rocks (⇓Anand et al. 1997) and towards the southeast (⇑Fig. 1). In contrast, groundwater calcretes form in saturated (phreatic) environments, typically in the axes of major drainages in the northern Yilgarn Craton. This paper describes the use of pedogenic calcretes for Au exploration in the Yilgarn Craton. The term ‘calcrete’ here is used in its broadest sense and includes powdery accumulations and clay coatings of carbonate through to massive indurated forms (⇓Netterburg 1980).
Historically, the use of calcrete as a geochemical guide to precious metal deposits began in Russia in the late 1950s (⇓McGillis 1967). One of the first exploration studies of calcrete in the Yilgarn Craton was undertaken for Ni deposits. Initially, however, it was considered as a geochemical diluent (⇓Mazzuchelli 1972) and other media, such as the residual soils themselves, were sampled instead. ⇓Cox (1975) gave further consideration to its use when looking for Ni in the Kambalda area (Western Australia). However, many saw the calcrete as a problem rather than a tool and efforts were made to upgrade the metal content of samples by removing the carbonate using dilute acid (e.g. ⇓Garnett et al. 1982). Other, mainly base metal, studies in South Africa and Australia followed during the next decade (e.g. ⇓Mazzucchelli et al. 1980; ⇓Frick 1985; ⇓Leduc 1986; ⇓Guedria et al. 1989; ⇓Tordiffe et al. 1989; ⇓Harrison 1990). A general association between Au and calcrete has been recorded in several anecdotal accounts (e.g. ⇓Kriewaldt 1969; K. Schulz, pers. comm., 1985) and in the scientific literature (e.g. ⇓Mann 1984a, ⇓b; ⇓Smith & Keele 1984; ⇓Lawrance 1988; ⇓Smith 1987; ⇓Glasson et al. 1988). Calcrete was, however, never systematically selected as an exploration medium in this period.
Calcrete has become a specific sample medium for Au exploration in Australia since the late 1980s as a result of research by CSIRO and, later, by CRC LEME (Cooperative Research Centre for Landscape Evolution and Mineral Exploration). In 1987, CSIRO commenced a research project with a consortium of exploration companies through AMIRA (Australian Mineral Industries Research Association Ltd) to improve geological, geochemical and geophysical methods for mineral exploration that would facilitate the location of blind, concealed or deeply weathered Au deposits. In late 1987, the first detailed research on calcrete above a Au deposit commenced at Bounty (120 km S of Southern Cross, Western Australia; ⇑Fig. 1) with spectacular results (⇓Lintern 1989; ⇓Lintern et al. 1992; ⇓Lintern & Butt 1993). For the first time a precious metal was demonstrated to be highly correlated (not merely associated) with Ca (calcrete); this was strong evidence for the highly soluble and mobile nature of Au in the soil. The results also showed, not unexpectedly, that Fe and many other elements had been diluted by the calcrete, consistent with the earlier base metal studies, and so were not correlated with Ca. The correlation was widespread at the Bounty Deposit in soils typical of those found throughout the auriferous greenstones of the Eastern Goldfields of WA. Thus, the potential for the widespread use of calcareous soil (calcrete) for exploration purposes was enormous. Further research confirmed the relationship to be robust throughout the southern Yilgarn Craton and indicated that calcrete was a geochemical sample medium that could be used with confidence, subject to certain important provisos.
Calcrete sampling reached a significant milestone as an exploration technique when its use was publicly acknowledged to have located the 0.5 million ounce Challenger Gold Deposit in 1995, the first significant discovery in the Gawler Craton, South Australia, for many decades. Subsequently, in a series of conference proceedings, company reports, magazine reviews and newspaper articles describing Au exploration in the Gawler Craton, calcrete was highlighted and recommended by many as the principal sampling technique to be used.
Interestingly, by the end of the 1990s, calcrete was being acclaimed as the sampling medium of choice back in the Yilgarn Craton by more exploration companies than the early 1990s, many of whom were not familiar with the earlier WA research, but had been convinced by the successes in SA (e.g. East Kundana discoveries, ⇓Boyer & Grivas 1999; Golden Cities, ⇓Kehal et al. 1999; Ghost Crab, ⇓Miller & McLeod 1999; Carosue Dam, ⇓Langworthy & Joyce 1999). Clearly, it took several years to filter through to the whole exploration community, even in the area where it was first recognized! As with many other geochemical sampling media, such as soil, rock chip or stream sediments, the popularity of calcrete has benefited from analytical laboratories providing low-cost, rapid chemical analyses with low detection limits. Calcrete has gained a foothold in many other parts of the world for Au exploration, including North and South America (⇓Smee 1999; S. Gatehouse, pers. comm., 2000; L. Bettenay, pers. comm., 1999), and no doubt will increase in popularity as new discoveries are made with it.
One of the common problems associated with calcrete sampling is the failure to recognize calcrete, particularly the powdery accumulations and clay coatings. These latter forms commonly do not impart a paler colour compared to equivalent non-calcareous soil, particularly if the carbonate content is low. In addition, other pale materials such as gypsum, silcrete and kaolinite may be mistaken for calcrete. The use of dilute hydrochloric acid is a simple test, except where dolomite is dominant. Scratching the dolomite into a powder or the use of stronger acid may be warranted in these cases. ⇓Edgecombe (1997) reported mis-identification of fossiliferous marine limestone or a thin coating of carbonate on saprolite as calcrete. In both cases, the sample will effervesce but a closer inspection of the material will reveal its true character. Where occurrence only as thin coatings is possible, so that carbonate forms only a small proportion of the sample, analysis for Ca is recommended; the presence of gypsum may complicate interpretation. In sandy areas, sieving is beneficial in ‘upgrading’ the calcrete sample that would otherwise be diluted by Si. Care should be taken when ‘soil sampling’ in terrains where calcrete is known to occur, because it is commonly found just below the surface. The depth of sampling may be crucial, with higher Au concentrations and anomalies merely due to the intermittent presence of calcrete unrecognized by the sampler. Calcrete is used as a road base, particularly in dune areas, where it is only too readily and inappropriately sampled by inexperienced field technicians. In testing the potential for calcrete sampling, orientation surveys should not be conducted by using 0–1 m or 1–2 m drill cuttings from deep drilling programs in lieu of specifically collected, due to potential cross-hole contamination.
Most standard digests and techniques for Au analysis of calcrete are suitable including aqua regia with AAS graphite furnace, fire assay, cyanide leach techniques with ICP-MS and neutron activation. Samples should be pulverized prior to analysis to release Au occluded in carbonate fragments, lithorelics or other material. The size of the sample to be analysed can be relatively small e.g. 10 g, because Au associated with the carbonate is usually sub-micron in size. For acid digestions, there must be sufficient acid present to both neutralize the carbonate and then dissolve the Au. Routine analyses by aqua regia–graphite furnace AAS will give detection limits of about 1 ppb which is acceptable for most applications; the acid digestion has the advantage of providing a solution that can be analysed for other elements, such as Ca and base metals. Sub-ppb detection limits are recommended in some terrains (e.g. aeolian dunes) using techniques such as cyanide leach with ICP-MS finish, or pre-adsorption onto activated carbon followed by neutron activation. The application of a number of partial and selective extractions has been tested at several sites (⇓Gray et al. 1999). Reagents and techniques tested included iodide, MMI (Mobile Metal Ion), Enzyme Leach, HCl and acetate-hydroxylamine, but none appeared to offer any significant advantage over total digestion methods in detecting the signal of buried mineralization in calcareous soils.
THE IMPORTANCE OF REGOLITH CLASSIFICATION
As with any geochemical sampling technique, knowledge of the weathering profile, style of mineralization, geology, climate, geomorphology and landform evolution is crucial. With it, the explorer is guided towards the most appropriate sample media to use, the grid size for sampling, the elements to be analysed and the choice of analytical technique. This information has been integrated into dispersion models that enable exploration case histories to be classified, to facilitate understanding of the process and to make valid comparisons between prospects. One of the more useful schemes for sub-tropical terrains has been proposed by ⇓Butt & Zeegers (1992) who broadly divided the regolith into three Types (A, B and C) based on the degree of preservation of the weathering profile. This classification has been particularly useful for exploring in the Yilgarn Craton. The three Types were further classified according to modifications of the pre-existing profile e.g., alteration, the recent accumulation of secondary materials (such as calcrete), and the presence or absence of transported material. In their scheme, a fully preserved or complete weathering profile (Type A) is described as one that has saprock, saprolite, mottled zone, and, importantly, a lateritic duricrust. Type B are those in which profiles are partially eroded to saprolite and/or laterite is absent. Type C are those in which the earlier regolith has been entirely eroded (or was never present). The uppermost horizons (lateritic residuum, saprolite or bedrock) are parent materials to residual soils, or are immediately overlain by sediments in depositional areas. As noted by ⇓Butt et al. (2000), classification is not possible in depositional areas without drilling or other information. For surface sampling procedures, such as calcrete sampling, the model classification is appropriate for relict (laterite-dominated) or erosional landscape regimes in which soils are developed largely from residual materials. However, in depositional regimes, the characteristics of the transported overburden, especially its thickness, are possibly of more immediate consequence. This change of emphasis is reflected by an additional category, Type T. Thus, the simplified and modified profiles of Butt & Zeegers (incorporating the lower pedolith and mottled zone in Type B, and ignoring the genetic theme) and incorporating Type T can be summarized as follows: Type A, weathered bedrock with lateritic residuum; Type B, weathered bedrock with no lateritic residuum; Type C, saprock-bedrock with thin residual soil; and Type T, transported overburden overlying Type A, B or C.⇓
Type T includes sites concealed by palaeochannel sediments. Palaeochannels are a common feature of the Australian continent and present considerable difficulties to exploration. In the Yilgarn Craton, most palaeochannels appear to be of Eocene age and are filled with clayey, sandy and lignitic fluvial, lacustrine and/or marine sediments. Some channels are older and contain Permian sediments. The channels are not easy to detect and the courses they follow are poorly defined. In the south, the palaeochannels commonly host, or are related to, significant Au mineralization, generally within the basal sands or in underlying saprolite. In many cases, the origin of the mineralization in the sediments has not been determined. It is possible that many channels are incised along intermittently mineralized shears. Gold deposits within and beneath palaeochannels present a special problem for exploration because they are usually concealed by barren sedimentary units, that may be 15 to over 50 m thick. Nevertheless, surficial anomalies have been reported by exploration companies directly above buried mineralization in some locations, commonly associated with calcrete e.g., Baseline (near Panglo, R. Howard, Pancontinental Ltd, pers. comm., 1992) and Ghost Crab (near Steinway, M. Miller, Newcrest Ltd, pers. comm. 1996).
Thus, as demonstrated by the case histories that follow, it remains necessary to establish the presence and nature of the transported overburden, and, if possible, the underlying residual regolith, by scout drilling prior to sampling. There are no documented case studies for Type C regolith.
CALCRETE CASE STUDIES
Type A: weathered bedrock with lateritic residuum
Mulline – soil developed on lateritic residuum
At Mulline Prospect, 140 km NW of Kalgoorlie (⇑Fig. 1), the regolith is mostly preserved and lateral dispersion of Au from primary mineralization has created a halo characteristic of lateritic Au deposits (⇓Fig. 3a). The saprolite is overlain by pseudo-bedded calcrete and Fe-rich pisoliths which are in turn overlain by a calcareous, nodule-rich loam and then a sandy loam with Fe-rich nodules (⇑Fig. 3b). The lateritic residuum has undergone extensive alteration and disintegration by precipitation of calcrete as coatings, veins, nodules, cements and indurated sheets. Gold is present in both the Fe-rich and the Ca-rich components of the profile and its distribution reflects their relative proportions (⇑Fig. 3c–f, ⇓Lintern & Butt 1991). The carbonate has diluted the abundances of Fe and trace elements (including Au) associated with the lateritic residuum. For Au, however, the accumulation in the carbonate offsets the dilution, so there is a net increase in Au abundance. In the upper portion of the trench profiles (shaded area), Au appears to be correlated with the carbonate but, in the lower portion, Au is associated with the older Fe-rich materials (⇑Fig. 3e). Accordingly, there is no simple correlation between Au and either Fe or Ca through the whole of the sampled profile.
Type B: weathered bedrock with no lateritic residuum
Bounty – soil developed on saprolite and thin colluvium
At the Bounty Gold Deposit, 240 km SW of Kalgoorlie (⇑Fig. 1), pre-existing laterites have been partly eroded and soils are mostly developed on thin colluvium and saprolite (⇓Fig. 4a). The site consists of undulating valleys and rises with local relief of 25–50 m. Powdery dolomitic calcrete has developed extensively within the clay-rich surficial material to depths of 1–2 m, beneath which is a non-calcareous, heavy red clay merging to saprolite within 6 m of the surface. The precise boundary between transported and in situ regolith is difficult to determine but the colluvium probably has not moved any appreciable distance (<50 m). Detailed profile sampling shows that the concentrations of Ca, Mg and Au are closely correlated (⇑Fig. 4b). Their abundances generally increase steadily from the surface, reaching maxima by one metre and then decline, with little present below two metres. Whereas there is good agreement between Au and the alkaline earth metals vertically (0–2 m) as demonstrated by the soil profile data, the correlation is not apparent in a lateral sense when the geochemistry of the bulk soil (c. 0–2 m) as collected by the auger is compared along a traverse (⇑Fig. 4d); obviously not all the calcrete is rich in Au. However, auger sampling is clearly successful in collecting the Au-rich calcrete horizon. Profile 4 was collected on the edge of calcrete-poor lateritic residuum, which explains the low Ca and Mg contents reported for the corresponding augered samples. Gold enrichment in calcareous soil downslope from the main mineralized zone (⇑Fig. 4a, c, d) to the east of Profile 1 is probably displaced. The results demonstrate that it is essential to sample the carbonate-rich horizon for Au consistently during soil surveys preferably using a power-auger and collecting the top one or two metres (⇓Lintern 1989; ⇓Lintern et al. 1990, ⇓1992).
Panglo – soil developed on saprolite
At the Panglo Gold Deposit, 30 km NNW of Kalgoorlie (⇑Fig. 1), the regolith has been truncated to the mottled zone (see also Scott & Howard 2001). The geology of the area consists of shales with mafic and ultramafic volcanic rocks adjacent to a granitic pluton; a palaeochannel and associated sediments occur to the west and south of the deposit (⇓Fig. 5a). The deposit is located beneath a salt-scalded area with groundwater located within 10 m of the surface. There is significant supergene enrichment at 40 m depth, but overlying saprolite is leached, with very low Au contents (<50 ppb). For the pedolith, a strong relationship between Au and alkaline earths was found in calcareous soils from the top 10 cm (⇑Fig. 5b, ⇓Lintern & Scott 1990; ⇓Lintern 1996). However, geochemical results from deeper, less calcareous units within the top few metres indicate that the association between Au and the alkaline earths is more tenuous; here, higher concentrations of Au do not necessarily correspond with high concentrations of Ca. This is probably due to residual particulate Au present within mottles or weathered rock (⇓Fig. 6). Thus Au-rich (>100 ppb) regolith material, when sub-sampled, did not show any particular affinity between Au and Ca or Fe. In the top 4 m, particulate Au in the upper saprolite is a relict that has survived intense weathering by being partly armoured in vein quartz and by secondarily precipitated Fe and Si. Some Au has probably been chemically re-mobilized into the top 10 cm from the Au grains by capillarity and evapo-transpiration.
Type T: transported overburden overlying Type A, B or C
Argo and Apollo – soil developed on transported overburden (colluvium)
The Argo Gold Deposit and Apollo Prospect are located c. 500 m apart about 100 km SE of Kalgoorlie (⇑Fig. 1) in a broad colluvial plain of low relief, adjacent to a salt lake. Mineralization is in saprolite at about 15 m depth at Apollo and 20–30 m at Argo; at the latter, it extends into a palaeochannel following the old land surface. At both sites, the residual regolith profile consists only of saprolite, generally 20–30 m thick (⇓Fig. 7). At least three sedimentary units cover the saprolite: (1) variably mottled, red and grey lacustrine clays, 2–7 m thick at Apollo and thicker at Argo; (2) a calcareous clay-rich to sandy-clay unit from about 0.2–2.0 m depth, with locally abundant carbonate-coated nodules (1–2 cm), carbonate coatings on clay, lithorelics (derived from saprolite outcropping 1–2 km away), and a dark Mn-rich horizon at c. 1.5 m depth; (3) a surficial, non-calcareous, clay-rich sand, mostly 10 to 20 cm thick (but up to 2 m in places), covering the entire area. The palaeochannel has incised into the saprolite 250 m to the south of the Apollo Prospect and cuts E–W across the southern boundary of the nearby Argo mine pit. The palaeochannel has a maximum depth of 60 m and average width of 400 m, and has been infilled by sediments of presumed Tertiary and Quaternary age.
Samples were collected by shallow augering into the carbonate horizon across both deposits with additional samples taken from shallow soils and deeper soil pits. These samples had Au concentrations of 5–20 ppb but with no discrimination between mineralized and background areas (⇓Fig. 8a, ⇓Lintern & Gray 1995b; ⇓Lintern et al. 1997). At Apollo, Au concentrations over mineralization in some samples were marginally higher than those over background: (i) near-surface soils (>10 ppb, ⇑Fig. 8b), (ii) sub-surface soil grab samples (up to 37 ppb, ⇑Fig. 8b) and (iii) deeper soil profile samples (profile L, up to 23 ppb, ⇑Fig. 8c). However, the anomalous samples were commonly not calcareous. At Argo, there are no apparent differences in the distribution characteristics, or total Au content, between mineralized and background areas. In the seven soil profiles (only B, E and J illustrated), Au concentrations were generally less than 10 ppb, with maxima occurring between 0.3 and 0.5 m (23 ppb) in the Ca-rich horizon (⇑Fig. 8c), and in the narrow Mn-rich horizon at 1.2–1.8 m depth (24 ppb).
Augering the top 1–2 m is the best practical sampling technique for collection of the calcrete, although this, in itself, is not effective in detecting mineralization at either site. Some non-calcareous soils may be as rich in Au as the calcrete. While the higher Au concentrations at Apollo over mineralization are anomalous, on a broader local scale (including Argo) they do not appear to be significant. The variable thicknesses and type of overburden, and generally low Au concentrations, have made the interpretation of the surface geochemistry at Apollo unclear and the source of the Au equivocal.
Safari – soil developed on transported overburden (colluvium)
The Safari Gold Prospect is 200 km NNE of Kalgoorlie (⇑Fig. 1) and 9 km NE of the margin of Lake Raeside. It is situated on a gentle, sandy, colluvial valley that slopes to the southwest. Gold is primarily associated with quartz veins within an anastomosing shear. The mean Au content of the top 1 m of the saprolite gives a very strong anomaly peaking over mineralization (1000 ppb), compared to an elevated and noisy background of 10–50 ppb Au.
The palaeosurface is much steeper and more variable than the present land surface at Safari (⇓Bristow et al. 1996; ⇓Fig. 9a). The residual regolith is truncated to a clay-rich saprolite and is almost completely blanketed by c. 4–10 m of transported overburden. This comprises up to 1 m of aeolian and colluvial sand, overlying a polymictic assemblage containing 2–10% angular, weakly weathered rock fragments in a matrix of sand, silt and, in places, clay. The coarse material commonly increases towards the base. Post-depositional modification of the sediments is widespread; most significant is the widespread calcification from about 0.5–5.0 m below the surface. Mottles of carbonate, a few tens of millimetres in diameter, occur only 20 cm below the surface, but carbonate morphology below this is uncertain (probably as coatings) because the only samples are drill cuttings. Beneath the zone of intense calcification, the sediments are variably indurated by Si and Fe oxides. Fresh rock is generally encountered 10–20 m below the unconformity. The bedrock is reasonably fresh where it subcrops, but has been brecciated by calcification in the upper few metres.
Samples were collected at 0.5 m intervals to 10 m depth in a specially drilled traverse across the mineralization. A pilot hole was drilled at each site to minimize cross hole contamination. Anomalous Au (22–60 ppb) is present in the carbonate horizon, with enrichment strongest from 0.5–2.5 m depth directly over mineralization (⇓Fig. 9) (⇓Bristow et al. 1996). Gold contents above background (i.e. >7 ppb) occur in the calcareous horizon for 800 m across strike of the mineralization. An anomaly in the top half metre, with excellent contrast, peaks directly over the primary mineralization with concentrations exceeding 5 ppb for over 600 m across strike. Despite higher absolute Au contents within the calcrete horizon, preferentially sampled highly calcareous fragments do not increase anomaly contrast, and the Au/Ca ratio in these is consistently lower than a bulk sample from the same interval. The Au anomaly associated with carbonate at Safari accurately reflects mineralization, with optimum sampling between 0.5 and 2.5 m depth.
The Safari case study, showing continuous Au enrichment from mineralization through 4–10 m of sediments, is a convincing example of dispersion into calcrete within transported overburden not recorded elsewhere.
Steinway – soil developed on colluvium above palaeochannel sediments
The Steinway Gold Prospect is a sub-economic Au deposit, located 25 km S of Kalgoorlie in a palaeochannel (⇑Fig. 1 and ⇓Fig. 10a). The central and northern parts of the area have a variable thickness (>20 m) of transported overburden of partly consolidated clays, sands and silts (⇑Fig. 10b). Beneath the transported overburden, the regolith is truncated to saprolite that overlies a bedrock of mafic andesites, trachytes, porphyritic tuffs and black shales. There are two types of mineralization at Steinway: (i) saprolite-hosted supergene mineralization located at c. 30 to 40 m just beneath the palaeochannel unconformity and (ii) primary mineralization associated with quartz stockwork veining within mafic andesites. The Ca- (as calcrete) and Fe-rich soil (0–1 m) overlying Steinway is anomalous in Au, with a maximum concentration of 150 ppb within a local threshold of 24 ppb (⇑Fig. 10c). The anomaly is one of the strongest in the area that has been drill-tested. Not all the anomalies have proved to have mineralization beneath them (⇑Fig. 10a, c), nor all mineralization have a surface expression.
The distribution of Au in a soil profile overlying the mineralization was investigated in detail (⇓Lintern & Gray 1995a; ⇓Lintern & Craig 1996; ⇑Fig. 10d, e). The soil profile consisted principally of a heavy red clay loam with calcrete occurring as coatings on the clay and as pendant-shaped powdery accumulations or pods that were more prominent in the top metre of the profile. Nearly black, vitreous, sub-rounded ferruginous granules, a few millimetres in size, were found to be abundant throughout the soil profile, commonly in lenses, and were separated from the clay and carbonate for detailed study. The principal results are summarized as follows:
Gold is associated with both ferruginous granules and calcrete in the soil as reflected in the size fraction geochemistry (⇑Fig. 10d).
Some Au in the calcrete is highly soluble in weak extractants (including deionized water), and is probably present as colloidal particles or in a ‘chemical’ form.
Some ferruginous granules contain microscopically-visible particulate Au with Au concentrations of individual, ferruginous granules being extremely variable (<40–15 000 ppb).
Relict primary fabrics and visible Au were observed in ferruginous granules.
The colluvium and palaeochannel sediments beneath the soil are essentially barren of Au.
These results suggest that the ferruginous granules are the immediate source of Au in the soil, and that both are derived by mechanical dispersion from upslope, rather than as Au migrating chemically from the underlying mineralization. The highly-soluble Au in the calcrete is probably derived from (i) the weathering of particulate Au in the ferruginous granules, and/or (ii) direct chemical dispersion from a similar upslope source as the ferruginous granules themselves.
The results from Steinway indicate that there may not be a causal link between anomalous Au in calcrete and underlying mineralization and that, in areas with thick transported overburden, sampling of calcrete, at best, may indicate the potential of the (sub-)catchment. Significantly, 400 m SW of Steinway, calcareous soils above the Greenback Gold Deposit have background (<24 ppb Au) concentrations yet mineralization is higher in the profile and has been mined. It is suggested, therefore, that for depositional environments where palaeochannels are anticipated, wider sampling intervals could be used, with a follow-up requirement that deep samples be collected, including basal sediments and/or ferruginous saprolite.
Higginsville – soil developed on colluvium above palaeochannel sediments
The Higginsville Gold Deposits are 120 km S of Kalgoorlie (⇑Fig. 1) within undulating terrain that has valleys filled with colluvium, and sub-crop confined to the higher areas. Draining the old Higginsville area (including the Vine workings) are the Mitchell and Challenge–Swordsman palaeochannels partly filled with sediments related to marine incursions in the Eocene (⇓Fig. 11a, b). Gold mineralization is present in basal sands, grits and conglomerates (⇑Fig. 11a). The palaeochannels overlie major shears and altered Archaean rocks containing patchy primary mineralization, which suggests that the channel-hosted Au deposits are probably derived from proximal Archaean sources. Ferruginous and calcareous soils, aeolian sands and large areas of partly weathered, remnant, Tertiary lacustrine, marine and fluvial sediments overlie deeply weathered Archaean basement that is truncated to saprolite (⇑Fig. 11a). The resulting regolith is complex, with fresh rock occurring below 80 m.
Soil samples were collected from an auger traverse (0–1 m) (i) across the Pluto deposit (Challenge–Swordsman palaeochannel), (ii) across the Mitchell-4 deposit (Mitchell palaeochannel), and (iii) downslope from the outcropping Vine deposit to the buried North Graveyard deposit (upper reaches of the Mitchell palaeochannel) (⇑Fig. 11b, ⇓Lintern et al. 1996). In addition, near surface soils (0–0.1 m) and samples from eleven soil profiles were collected.
There is a strong anomaly and an association between Au, calcrete and mineralization in residual soils at the Vine deposit (Type B regolith profile, ⇑Figs 11b and ⇓12a). Dispersion may extend southwards to the shallowly-buried (<10 m) North Graveyard deposit (Type T), although higher Au contents here may be related to higher carbonate contents. Importantly, the Au content of the soils at Mitchell and Pluto over the palaeochannels were not related to underlying mineralization. The highest concentration of Au in soil (425 ppb) was found in profile G, immediately down drainage of the Aphrodites pit (⇓Fig. 12a). Near-surface samples (0–0.1 m) from the Mitchell traverse had lower Au concentrations (maximum 11 ppb) than the 0–1 m soils (maximum 20 ppb) with neither sample type reflecting the occurrence of buried mineralization. There was a generally positive correlation between Au and Ca in the soils (⇑Fig. 12a, b).
Fifty individual ferruginous granules were randomly selected from soil profile G. As at Steinway, the ferruginous granules had highly variable Au contents ranging from <100–51 000 ppb. The Au-rich (over 250 ppb) granules are pervasively ferruginized, with some having partly-preserved lithic fabrics. Some granules contain Au particles up to 10 μm in size, although no Ag was detected within them. As with Steinway, it is probable that Au released from the ferruginous granules is the immediate source of much of the Au associated with the carbonates in overlying soils. The granules are probably derived from auriferous ferruginous saprolite (Type B regolith) or lateritic duricrust (Type A) upslope.
Weathering has complex effects on chemical and physical dispersion. Calcrete, however, is unusual since it can be both readily dissolved and be mechanically moved as nodules or pisoliths. For Au, it appears that calcrete provides a host for (i) hydromorphically-mobilized ‘chemical Au’ occurring as organic ligands or other complexes and/or (ii) mechanically-mobilized Au particles of various size ranges (μm to cm) that become re-located within the re-mobilized Ca. Thus Au anomalies in calcrete may be found in a variety of landscape positions relative to their source.
Calcrete has both advantages and disadvantages for mineral exploration both for Au and base metals (⇓Butt 1992). The presence of calcrete may be advantageous to exploration in the following circumstances:
The mobility of some Au to the calcareous horizons of soils appears to be primarily governed by the processes of meteoric water infiltration and evapo-transpiration. This may give rise to or enhance a near-surface expression to concealed primary or secondary mineralization
Calcrete represents a consistent, easily-identified, sampling medium (generally corresponding to a B or illuvial horizon) for exploration purposes.
The principal disadvantages are:
Many pedogenic calcretes represent absolute additions to soils developed on pre-existing deep weathering profiles which, in many instances, have been partly or fully truncated prior to soil and calcrete formation. The concentrations of many mobile elements associated with economic mineralization may have been already reduced by leaching during the initial deep weathering and later pedogenesis, so that the addition of calcrete causes dilution and depresses anomaly contrasts still further. This is particular true for base metals.
The high pH prevailing with calcrete reduces chemical mobility of many elements (significantly, not Au) and hence restricts the development of Au pathfinder anomalies, e.g. As.
The addition of calcrete may act as a diluent to other sample media, e.g. laterite or skeletal soils developed on bedrock.
Calcrete may occur in different forms and/or positions in the soil profile representing different ages, stages of development or climatic episodes that may have different levels of metal concentration or dilution. For example, calcrete many metres thick may not be particularly useful as a sample medium, nor would calcrete developed high in the profile of an aeolian dune.
In deeply weathered terrains, the degree of regolith truncation (if any) and the presence of transported overburden are both significant. In depositional terrain, such overburden may influence the type, interpretation and cost of a geochemical sampling programme, including those involving calcrete. Ideally, the nature and thickness of transported material should be determined, since these may have a bearing on the ability of Au to migrate to the surface. This is not always possible, particularly for regional soil surveys where drilling is limited. The presence of even a thin layer of transported material can have enormous effects on the use of calcrete as a sampling medium. Two basic strategies have been adopted to explore areas dominated by transported overburden. The first approach has been to drill through the overburden and into the underlying bedrock (itself usually weathered and potentially leached of elements of interest). This approach is expensive, but relatively low-risk, and is commonly used on prospects where one or more areas of mineralization have already been identified. The second approach involves the collection of materials (including calcrete) at or near the surface and relies on the premises that (i) if the sediments are old, diagenesis or post-depositional weathering will have caused migration of pathfinder or target elements to the surface, (ii) there is active dispersion of such elements or (iii) where the transported overburden is thin, bioturbation can physically mix sediments and re-locate particles to the near surface. This approach is high risk, for it depends upon the ability of such processes to give surface expression to mineralization through many tens of metres of transported overburden, which has been doubted by many exploration geochemists. There are apparently convincing case studies that demonstrate that the use of surficial sampling media is effective in certain circumstances. However, it is clear that local geology and environment are of major importance and studies are commonly insufficiently detailed to make an accurate judgement. The evidence presented here (e.g. Steinway, Argo-Apollo, Higginsvile) and elsewhere (⇓Lintern & Gray 1995c, ⇓d; ⇓Gray et al. 1999) suggests that the presence of either thick sediments (>10 m depth) and/or impervious lacustrine sediments will provide a severe hindrance to any upwards aqueous mobilization of Au.
LESSONS FOR EXPLORATION PRACTICE – SUMMARY
Several lessons are to be learnt from calcrete sampling in the Yilgarn Craton. These include:
Calcrete is an excellent sample medium for Au exploration because it appears that its physico-chemical properties mimic that of the Au species found in soil, so that Au and Ca often accumulate in the same part of the soil profile.
Calcrete (including calcareous soil) has a variable composition that partly depends on the nature of the host material and amount of calcite and dolomite precipitation.
Gold in calcrete may have been transported laterally up to several hundred metres, so that its source may not be presence does not be directly below.
Mapping and classification of the regolith settings (e.g. as Type A, B, C or T) are essential for the correct interpretation of Au anomalies developed in calcrete.
Calcrete anomalies are strong and coherent particularly in areas where lateritic residuum, saprolite or thin colluvium dominate (Types A, B and presumably C).
If the thickness of transported overburden exceeds 10 m and/or if lacustrine sediments are present, experience to date suggests that calcrete will probably be ineffective as a sample medium.
In transported (Type T) regolith settings, the Au–Ca correlation may be weaker than in residual settings (Type A, B and C); this is probably due to factors including the presence of detrital Au and different physico-chemical regimes e.g. caused by periodic flooding.
The extra expense involved with the use of partial or selective extraction procedures for calcrete samples may not be warranted over the simpler total digest method.
The most cost-effective sampling procedure is by power auger drilling and combining the cuttings through the calcrete horizon (top metre or so). Other techniques, such as surficial soil sampling or drilling and routinely sampling at a specified depth, may be inappropriate because calcrete, and associated Au anomalies, may be overlooked.
The author wishes to acknowledge the support provided by the CSIRO, AMIRA Ltd. and sponsors of projects P241, P241A and P409 for funding, allowing access to sites, supplying drilling rigs and operators, geological plans, geochemical data and samples. The author would also like to acknowledge his co-workers for assisting in many of the case histories. A. D. Vartesi is thanked for the drafting of the diagrams. Comments by C. R. M. Butt, K. M. Scott, N. Rutherford and others greatly improved earlier versions of the manuscript. CRC LEME is supported by the Australian Cooperative Research Centres Program. Many of the referenced unpublished reports are or will soon become ‘open file’; see CRC LEME web site http://leme.anu.edu.au/ for details.
- © 2001 AAG/The Geological Society of London