| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology 45(12) © Oxford University Press 2004; all rights reserved
Intrusion and Crystallization of a Spinifex-Textured Komatiite Sill in Dundonald Township, Ontario
1 LGCA, UMR5025 CNRS, UNIVERSITÉ JOSEPH FOURIER, GRENOBLE, FRANCE
2 MINERAL EXPLORATION RESEARCH CENTRE, DEPARTMENT OF EARTH SCIENCES, LAURENTIAN UNIVERSITY, SUDBURY, ONTARIO, CANADA
3 OTTAWA–CARLETON GEOSCIENCES CENTRE, UNIVERSITY OF OTTAWA, OTTAWA, ONTARIO, CANADA
4 PRECAMBRIAN GEOSCIENCE SECTION, ONTARIO GEOLOGICAL SURVEY, SUDBURY, ONTARIO, CANADA
RECEIVED NOVEMBER 14, 2003; ACCEPTED SEPTEMBER 1, 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSGEOLOGICAL SETTINGFIELD RELATIONSTEXTURAL VARIATIONSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Although komatiite has been defined as an ultramafic volcanic rock characterized by spinifex texture, there is a growing recognition that similar textures can also form in high-level dykes and sills. Here, we report the results of a petrological and geochemical investigation of a
5 m thick komatiite sill in Dundonald Township, Ontario, Canada. This unit forms part of a series of komatiites and komatiitic basalts, some of which clearly intruded unconsolidated sediments. The komatiite sill is differentiated into a spinifex-textured upper part and an olivine cumulate lower part. Features characteristic of the upper sections of lava flows, such as volcanic breccia and a thick glassy chilled margin, are absent and, instead, the upper margin of the sill is marked by a layer of relatively large (1–5 mm) solid, polyhedral olivine grains that grades downwards over a distance of only 2 cm into unusually large, centimetre-sized, skeletal hopper olivine grains. This is underlain by a 1 m thick zone of platy spinifex-textured olivine and coarse, complex, dendritic, spinifex-textured olivine. The texture of the olivine cumulate zone in the overlying unit is uniform right down to the contact and a lower chilled margin, present at the base of all lava flows, is absent. The textures in the sill and the overlying unit are interpreted to indicate that the sill intruded the olivine cumulate zone of the overlying unit. Thermal modelling suggests that soon after intrusion, a narrow interval of the overlying cumulate partially melted and that the liquid in the upper part of the sill became undercooled. The range of olivine morphologies in the spinifex-textured part of the sill was controlled by nucleation and crystallization of olivine in these variably undercooled liquids. KEY WORDS: komatiite; intrusion; spinifex texture; olivine
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSGEOLOGICAL SETTINGFIELD RELATIONSTEXTURAL VARIATIONSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
A few years after Mike O'Hara finished his work on the Scourie picritic intrusions (O'Hara, 1961
), Naldrett & Mason (1968) published one of the first complete petrological descriptions of a komatiite. At that time, about a year before Viljoen & Viljoen's (1969a, 1969b) papers defining komatiite as a new type of volcanic rock, most of the petrological community shared Bowen's doubts about the existence of ultramafic magmas (Bowen, 1956). Naldrett & Mason (1968) presented detailed descriptions of skeletal olivine morphologies in a series of ultramafic units in Dundonald Township, Ontario. Although they recognized that these textures were attributable to very rapid cooling, they hesitated to call the units lavas and instead interpreted them as a series of high-level intrusions.
The komatiites in Dundonald Township were subsequently largely neglected as attention turned to better exposed units 50 km to the east in Munro Township (Pyke et al., 1973) and 5 km to the northeast around the Alexo mine (Barnes, 1983; Arndt, 1986). Muir & Comba (1979) studied the Ni–Cu–PGE mineralization in central Dundonald Township, and recognized numerous ‘interflow’ breccias containing mixtures of graphitic material, volcanic clasts and sulfides; consequently, they interpreted the units that hosted the mineralization as extrusive. In 1989, a large area of outcrop in Dundonald Township, immediately north of the komatiite occurrences described by Muir & Comba (1979), was stripped of vegetation during evaluation of the sulfide mineralization by Falconbridge Ltd. Mapping of the stripped outcrops by Davis (1997, 1999) indicated that the ‘interflow’ breccias were peperites and that the thin komatiitic basalts intercalated with them were sills. More detailed mapping has shown that many, perhaps most, of the komatiitic units in the sequence are intrusive (Houlé et al., 2002a, 2002b, in preparation; Cas et al., 2003).
The question of whether many other komatiites might also be intrusive was brought into the limelight by Grove, de Wit, Parmen and coworkers (Grove et al. 1994, 1997; Parman et al., 1996), who suggested that the komatiites in the Barberton greenstone belt in South Africa were emplaced as a series of mid-crustal sills, rather than as lava flows. This question is very important because Barberton is the type locality of komatiite, where this type of volcanic ultramafic rock was first described by Viljoen & Viljoen (1969b). Grove, de Wit, Parman and coworkers advanced two principal arguments to support their interpretation. First, on the basis of field and petrographic observations, they concluded that the komatiites of the Barberton Belt are discordant and intrusive. Second, they proposed that komatiite magma was hydrous and formed in Archean subduction zones—a possibility previously suggested by Allègre (1982). Because the solubility of water in silicate liquids depends strongly on pressure, a hydrous magma will not reach the surface without degassing and is more likely to intrude into the upper crust than to erupt on the surface. Grove et al. (1997) suggested that the Barberton komatiites formed in this way. Although it has long been clear that some komatiitic rocks are intrusive (e.g. Williams, 1979; Pyke, 1982; Davis, 1999; Stone & Stone, 2000; Beresford & Cas, 2001), and that some komatiitic magmas may have contained as much as1–2% H2O (Stone et al., 1997), detailed mapping and petrographic studies by numerous workers in Australia, Brazil, Canada, Finland, South Africa and Zimbabwe have shown, through the identification of hyaloclastite flow-top breccias and other features indicative of surface eruption, that most komatiites are actually extrusive.
An evaluation of komatiite units in which intrusive relationships appear unequivocal indicates that these units invaded sequences of unconsolidated sediments. This is to be expected because unconsolidated sediments are much less dense than komatiitic magma and, for rheological reasons, more difficult for rising magma to traverse than brittle volcanic rock. However, the question of whether komatiites are capable of intruding other komatiites or mafic volcanic rocks, as suggested by Grove and coworkers for the komatiites in Barberton, remains unanswered.
The purpose of this study is to provide a detailed field and petrologic description of a 5 m thick komatiite unit at Dundonald Beach (Figs 1–4) and to present arguments for an intrusive origin. We compare the features of the Dundonald komatiite sill with those of well-documented komatiite flows in Canada, Australia and southern Africa, and show that apart from the upper contact, which provides convincing evidence of an intrusive origin, there is little to distinguish the sill from a typical spinifex-textured komatiite flow. We then demonstrate, using mineralogical and textural evidence, that the magma that formed the Dundonald Beach sill contained very little water, and that an intrusive setting by no means indicates that the parent komatiite was hydrous.
|
|
|
|
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSGEOLOGICAL SETTINGFIELD RELATIONSTEXTURAL VARIATIONSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The entire stripped outcrop area of Dundonald Beach (Fig. 2) was mapped at 1:250 scale between 2000 and 2003 as part of M. G. Houlé's PhD dissertation at Laurentian University and the University of Ottawa (Canada), and as part of collaborative projects with Université de Grenoble (France) and Monash University (Australia). In the present study, we mapped the two westernmost exposures of the komatiite sill at a scale of 1:50 (Fig. 3). Samples from the outcrop shown in Fig. 3 were slabbed and the polished surfaces were scanned to produce the images reproduced in Fig. 5. Thin sections of the samples were imaged (Figs 6 and 7) and studied in detail. With the exception of irregular patches in the lower olivine cumulate zones, secondary minerals have pseudomorphically replaced all of the olivine and much of the pyroxene. Although this alteration precludes detailed study of primary mineral compositions within the unit, it has accentuated the contrast between dark serpentinized olivine and paler matrix, highlighting the relict igneous textures. Variations in the sizes and orientations of olivine crystals in the upper part of the unit were documented by tracing their long dimensions by hand on images of outcrops, scanned slabs or thin sections, then analyzing the results using MATLAB®. The results are shown in Fig. 8.
|
|
|
|
|
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSGEOLOGICAL SETTINGFIELD RELATIONSTEXTURAL VARIATIONSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Dundonald Township is located
40 km northeast of the city of Timmins in northern Ontario (Fig. 1). The rocks are part of the 2719–2710 Ma Kidd–Munro assemblage of the Abitibi greenstone belt (Ayer et al., 2002a, 2002b), which in this area comprises mainly mafic and ultramafic metavolcanic rocks, minor felsic metavolcanic rocks and lesser pelitic metasedimentary rocks (Davis, 1997, 1999; Houlé et al., 2002a, 2002b). The rocks are macroscopically folded and locally faulted, but the strain was strongly partitioned, so the degree of penetrative deformation within most units is quite low. The metamorphic grade is lowermost greenschist facies (Jolly, 1982).
Figure 2 is a simplified geological map of the ‘Dundonald Beach’ area. The volcanic–sedimentary sequence in this area dips and faces towards the south, and comprises a lower unit of felsic volcaniclastic rocks overlain by a series of komatiitic basalts and komatiites. The units in the western part of the outcrop are thicker (up to 10 m) and better differentiated with olivine and pyroxenespinifex-textured upper zones and pyroxene and olivine cumulate lower zones. Eastward along strike these units become much thinner (1–3 m) and grade into fine-grained, massive komatiitic basalt sills, peperites and graphitic metasedimentary rocks as described by Davis (1997, 1999), Houlé et al. (2002a, 2002b; in preparation), Cas & Beresford (2001) and Cas et al. (2003). The komatiitic basalts are overlain at both ends of the outcrop by thicker, more massive olivine cumulate komatiite units, which contain the high-grade, low tonnage Ni–Cu–PGE mineralization described by Muir & Comba (1979), and, in the western part of the outcrop, the spinifex-textured unit that is the subject of this study. The upper part of the sequence is composed of another series of thick differentiated komatiitic basalts.
|
FIELD RELATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSGEOLOGICAL SETTINGFIELD RELATIONSTEXTURAL VARIATIONSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The spinifex-textured komatiite sill described in this paper is exposed at the southern margin of the outcrop along a strike length of
150 m (Fig. 2). In the south-western part of the Dundonald Beach outcrop (Fig. 2), it is 6 m thick and comprises a 1·2–1·5 m thick upper olivine spinifex-textured zone underlain by a 3·5 m thick lower olivine cumulate zone (Fig. 3). In the part of the outcrop immediately east of this area (see Fig. 2), the outcrop is less massive, but the sill is also 6 m thick and contains only a very thin (0·5 m) upper olivine spinifex-textured zone. It is overlain by another unit (only partly exposed along the southern margin of the outcrop shown in Fig. 3) of similar thickness, but with a fine-grained upper random olivine spinifex-textured zone containing abundant amygdales. In the southeastern part of the Dundonald Beach outcrop (see Fig. 2), the sill is composed almost entirely of olivine cumulate, clearly transgresses underlying komatiitic basalts, and hosts small amounts of very high tenor Ni–Cu–PGE mineralization. The upper contact in this area is sheared, but appears to be capped by a thin zone of peperite. The lower contacts of the sill are not as well exposed as the upper contacts, partly because they have been trenched for Ni–Cu–PGE mineralization. In the area shown in Fig. 3, an interval of fine-grained komatiite separates the olivine cumulate from the underlying komatiitic basaltic unit. A thin layer of shale occurs sporadically along the contact. The fine-grained komatiite is poorly exposed and highly weathered and was not sampled.
The contact between the olivine cumulate of the overlying unit and the top of the komatiite sill is slightly curved and has a gentle convex-upward form (see upper left part of Fig. 3). The transition from the olivine cumulate in the lower part of the overlying unit to the spinifex-textured rock in the upper part of the sill is sharp, and there is no indication of chilling or thermal metamorphism along the margins of either unit. The texture of the olivine cumulate at the base of the overlying unit is uniform right down to the contact (Figs 5 and 6).
The uppermost part of sill is a very fine-grained dark grey–green rock containing, within 1 cm of the contact, millimetre-sized polyhedral olivine grains (Fig. 5a and Layer 1 in Fig. 6). The dimensions of the olivine grains increase rapidly downwards into the unit, and at a distance of only 5 cm from the contact, the rock consists of stubby, randomly oriented, centimetre-sized skeletal (hopper) olivines in a fine-grained groundmass of pyroxene and altered glass (Fig. 5 and Layer 2 in Fig. 6).
The textures, particularly the morphologies of the olivine crystals, vary considerably within the spinifex zone. From top to base, there is a transition from a thin upper zone of randomly oriented bladed olivines (Fig. 4), through a zone composed of ‘books’ of parallel platy olivines (Fig. 5b), to a zone of coarse dendritic olivines (Fig. 5c). This texture, which appears to be peculiar (in this area) to the sill, consists of large (up to 20 cm long and 1 cm wide) olivine dendrites that are sparsely distributed and randomly oriented within a matrix of finer platy olivines, pyroxene needles and altered glass. The texture of this and other units is described in detail below.
The B1 zone, a layer of horizontally oriented olivine tablets between the spinifex and olivine cumulate zones, is about 10 cm thick and planar. It is oriented parallel to the overall strike of the magmatic and sedimentary units (Figs 2–4), which indicates that it was horizontal when deposited. The divergence between the orientation of the B1 zone and that of the upper contact contrasts with the undulating nature of the upper surface of the komatiite unit, suggesting that the latter is a magmatic, rather than structural, feature.
The olivine cumulates in the sill and the overlying unit have an identical, relatively uniform, fine-grained mesocumulate texture comprising >60% olivine in a matrix of fine-grained acicular pyroxene and altered glass.
|
TEXTURAL VARIATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSGEOLOGICAL SETTINGFIELD RELATIONSTEXTURAL VARIATIONSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Upper contact zone (sample DUN1)
Textures at the upper contact are illustrated in a photo of the contact in outcrop (Fig. 4b), a scanned image of a large thin section (Fig. 5a), a mosaic of photomicrographs (Fig. 6a) and a drawing of the olivine grains (Fig. 6b). At all of these scales, the texture in the olivine cumulate zone of the overlying unit remains uniform all the way down to the contact. In a band only 5–10 mm wide and immediately above the contact, the matrix between the olivine grains has a slightly darker colour, but it is not clear whether this is as a result of baking from the komatiite sill or enhanced alteration related to focused fluid circulation along the contact. The olivine cumulate above the contact contains 60–70% solid, sub-equant, sub-millimetre olivine crystals, and rare larger grains, in a matrix of fine-grained pyroxene and very fine-grained altered brown glass. Most of the olivine grains are altered to serpentine or chlorite, but in irregularly distributed patches, many of the grains remain unaltered.
The uppermost centimetre of the sill containssub-equant, moderately to highly skeletal grains of olivine in a matrix of fine olivine wafers, acicular pyroxene grains and altered brown glass (Figs 4 –6). The contact is inconspicuous, even at the thin-section scale. It is not marked by a fracture or suture, or by a zone of chilling. Instead, there is a marked increase in the size of the olivine grains, from 0·5 mm in the overlying olivine cumulate, to1–5 mm in the uppermost margin of the sill. The larger olivine grains have distinctive shapes and orientations: at the contact, they are solid, subhedral and non-skeletal, whereas downward over a distance of a few millimetres, the habit changes from solid to skeletal. The transition is well illustrated in the grain just right of centre in Fig. 6a; as shown in Fig. 7b, it has a sub-vertical orientation, a euhedral, slightly skeletal morphology at its base, and appears to be rooted in equant cumulus olivine grains at the top. This grain apparently nucleated on the olivine grains at the base of the overlying unit and grew downward into the liquid of the sill. Deeper in the sill, but still within 2–3 cm of the upper contact, the olivine grains have a well-developed ‘hopper’ morphology. They are relatively stubby (aspect ratio 2–4) with highly skeletal, ribbed interiors and, in some cases, distinctive parabolic ‘snouts’ (Figs 5b and 7a). With the exception of chromite, all magmatic minerals in the upper contact zone have been totally replaced by secondary minerals.
In contrast to the textures in the olivine cumulate of the overlying unit, which are uniform, the olivine grains in the upper border zone of the sill have characteristic sizes and morphologies that change with distance from the contact, as described above. The border zone itself maintains a relatively constant thickness and follows the gentle undulations of the contact.
Random (DUN2) and platy (DUN3) olivine spinifex-textured zone
The textures in these zones are similar to those in many komatiite lava flows. Random olivine spinifex texture (Fig. 4b) consists of 50% randomly oriented, fine to coarse (0·5–7 cm), moderately skeletal olivine blades in a matrix of fine-grained pyroxene grains and altered glass. The pyroxene grains have an acicular habit and are zoned from pigeonite cores (now altered to chlorite) to augite margins. Platy olivine spinifex texture (Fig. 5b) consists of 40% coarse (1–2 cm wide, 5–10 cm long) parallel books of moderately skeletal olivine wafers in a similar matrix. The books are oriented in a near-random manner, but more commonly closer to parallel than perpendicular to the top of the unit.
Coarse dendritic spinifex-textured zone (DUN4)
This name is given to a texture that has not previously been described in the literature. The dominant element of the texture are very large olivine crystals, several centimetres long and up to 1 cm wide, with a highly complex internal structure (Figs 4b and 5b). These crystals commonly have a fine parabolic snout at one end (Figs 5b and 7a)—a characteristic feature of dendrites. In the centre of each crystal, complexly ribbed cells have grown outward from a central spine. The crystals are relatively sparsely distributed and randomly oriented in a matrix of finer platy olivines, acicular pyroxene grains and altered glass. All of the olivine is altered to serpentine or chlorite, but magmatic pyroxene is preserved in patches in the groundmass. Similar textures occur in thick komatiite flows, komatiitic sills and komatiitic dykes at several other localities (e.g. Kambalda, Western Australia; Eldorado Township, Ontario; McArthur Township, Ontario).
B1 zone (DUN5)
As in some (but not all) spinifex-textured lava flows, the spinifex zone is separated from the olivine cumulate zone by a 5–10 cm thick layer containing fine, slightly skeletal, tabular olivine grains oriented parallel to the upper contact of the sill (Fig. 5c). Many of the large dendritic olivine crystals in the lowermost part of the spinifex zone are oriented at a high angle to the B1 layer. Some appear to have nucleated on B1 olivine grains and to have grown upwards into the overlying sill, indicating that the B1 zone formed prior to crystallization of the lowermost part of the spinifex zone (e.g. Pyke et al., 1973; Arndt, 1986; Lesher & Keays, 2002).
Olivine cumulate zone (DUN6, 8, 9)
The olivine cumulate of the komatiite sill is very similar to that of the overlying unit (see description above) and to the cumulate zones of many komatiite lava flows. One difference between the sill and the overlying unit in this locality is the presence in the cumulate zone of the sill of irregular, centimetre-sized patches within which the olivine grains have a skeletal morphology. In some of the patches, the grains have an elongate tabular form and the texture is very similar to random spinifex texture. As in the overlying unit, olivine grains in irregular regions are unaltered.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSGEOLOGICAL SETTINGFIELD RELATIONSTEXTURAL VARIATIONSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Environment of emplacement
The graphitic, framboidal sulfidic sediments in theDundonald area (Muir & Comba, 1979
; Houlé et al., 2002a, 2002b, in preparation) indicate a deep-water environment of deposition. The fluidal and blocky peperites in the lower part of the stratigraphic sequence (Fig. 2) indicate that the sediments were unconsolidated when the komatiitic magmas were emplaced. The very fine-grained skeletal pyroxene and altered glass in the mesostases of all of the rocks, including the cumulates, indicates rapid cooling. Together, these features provide evidence that the komatiitic units were emplaced in or on a series of unconsolidated sediments close to or at the ocean floor.
Discrimination between komatiite flows and sills
Dann (2000) has described a number of features that can be used to distinguish between komatiite lava flows and sills. Many of these apply specifically to the volcanic sequence in the Barberton greenstone belt and are based on relations between komatiite and pillow lava, so they are not applicable in this area. Of the others, the most useful are the presence of flow-top breccias and hyaloclastites, vesicular flow tops and pipe vesicles in lava flows. Dann also mentioned the specific form of an inflated lava flow, in which an upward domed roof overlies a thickened portion of the flow. As the only feature diagnostic of a sill, he cited locally cross-cutting relationships with the host rock. In the following section, we discuss these features and present others that we believe are diagnostic in evaluating the volcanic setting of the komatiites at Dundonald Beach.
Flow-top breccias
The upper surfaces of lava flows are in direct contact with air or water and cool much more rapidly than the upper contacts of sills. As a consequence, many komatiite flows are capped by characteristic flow-top hyaloclastite breccias containing shard- or bubble-wall fragments and a wide variety of quench textures (e.g. Munro Township: Pyke et al., 1973; Arndt, 1976; Alexo: Arndt, 1986; Zwishevane: Renner et al., 1993). Even where hyaloclastites have not formed, lava flows are characterized by relatively thick upper chilled margins containing sparse, small and highly skeletal crystals and abundant volcanic glass. The upper margins of sills, in contrast, are in contact with poorly conductive solid rocks, cool much less rapidly and should form less glass. Where present, flow-top breccia and aphanitic upper chilled margins provide unequivocal evidence of an extrusive origin.
Polyhedral jointing
Many komatiite flows, including some at the classicBarberton (Viljoen & Viljoen, 1969b; Dann, 2001) and Munro Township (Pyke et al., 1973) localities, do not have extensive flow-top breccias. Upper chilled margins may be present in both lava flows and sills, but they are thicker in lava flows and are cut by distinctive polyhedral joints (Figs 4 and 8). How deep the polyhedral joints penetrate into the flow depends on several factors, including the thickness of the flow and whether the flow is immediately covered by another flow. Thinner flows tend to have relatively thick zones of polyhedral jointing and very thin flows may be polyhedrally jointed throughout. Rapidly emplaced units have thinner zones of polyhedral jointing. Where present, polyhedral jointing provides unequivocal evidence of an extrusive origin.
Grain-size variations
Because of differences in cooling rates, at given distances from the tops of sills, the crystals should be larger, less elongate and less skeletal than those at a similar positions in lava flows (Fig. 8).
Lower chilled margins
Although the upper chilled margin of a flow may be removed by thermomechanical erosion from an overlying flow or sill, a lower chilled margin can only be removed by intrusion from below. If the textures in the lower part of the unit remain uniform right down to the contact between the two units, with no evidence of an increase in the proportion of glass or the appearance of quench crystals that marks a chilled lower contact, the contact is most likely intrusive. Importantly, however, sills commonly intrude along the contacts between units, so the presence of a lower chilled margin cannot be used to discount intrusion.
Transgressive upper contacts
Although the lower contacts of both lava flows and sills may be transgressive, the upper contacts of flows cannot be transgressive. If the komatiite unit or apophyses cut across the structure of the overlying unit, there is no doubt that the komatiite unit was emplaced later and is intrusive.
Baked contacts
Similarly, both lava flows and sills may bake underlying rocks, but a flow cannot bake the rocks that overlie it unless the overlying flow was emplaced before the underlying flow cooled completely. If these rocks are significantly baked (thermally metamorphosed), there is no doubt that the komatiite unit was intrusive.
Peperitic textures
Peperites form by disintegration of magma intruding and mingling with unconsolidated, typically wet sediments (White et al., 2000; Skilling et al., 2002). Peperitic textures may form along the upper and lower margins of sills, but not along the upper contacts of lava flows. The presence of peperite along the upper margin of a unit is unequivocal evidence for an intrusive origin.
Overall form and relationship with surrounding rocks
The upper surfaces of lava flows are free and easily deformed, and their lower surfaces conform to the underlying topography, which may also be irregular. When new magma flows into the interior, the roof of the flow may rise, producing a characteristic domed form (e.g. Dann, 2000), or lava may break out from the flow to form a separate unit. In either case, the upper and lower surfaces are irregular on an intermediate to large scale and the overlying rocks are undeformed. If a sill intrudes along a plane of weakness into solid, brittle rocks, it will be planar and its upper and lower margins should match. However, if magma intrudes at shallow depth into unconsolidated sediments, the intruded material may deform and the upper surface of the intrusion may be domed and irregular. Lava flows and sills may both be planar and concordant, and flows and sills may both thermomechanically erode underlying rocks, but only flows can have irregular or domed upper surfaces that have not deformed overlying rocks.
Features indicative of an intrusive origin for the komatiite unit at Dundonald Beach
Based on the above discussion, we can now summarize the evidence for an intrusive origin for the spinifex-textured sill at Dundonald Beach.
There is no lower chilled margin at the base of the overlying unit. Textures in the olivine cumulate zone of the overlying unit are uniform all the way down to the lower contact, with no sign of chilling at the base. The olivine cumulate zone of the overlying unit is interpreted to have been truncated during intrusion of the underlying sill.
There is no sign of chilling or other flow-top structures along the upper contact of the sill. In contrast, the grains immediately below the upper contact are solid and polyhedral to rounded, and they are distinctly larger than the olivine grains in the cumulate zone of the overlying unit. These grains give way downwards to large (5–10 mm), equant, highly skeletal hopper olivine grains with unusual morphologies, as illustrated in Figs 4–6. Breccias, polyhedral joints and other structures characteristic of lava flows are absent.
The variations in grain sizes are quite different from those observed in typical komatiite lava flows (Fig. 8: right side), which have a relatively thick (7 cm) aphanitic zone in which no crystals are visible in hand samples, underlain by a thick (15 cm) random spinifex textured zone in which the average length of the olivine blades gradually increases from about 0·6 to 4·5 mm, underlain by a platy spinifex zone of variable thickness, which contains books with an average length exceeding 10 mm. The grain-size variations in the upper part of the komatiite sill are quite different (Fig. 8: left side). Even in the uppermost centimetre of the unit, individual olivine grains are visible and have an average length of 0·7 mm. At a depth of 0·5–6 cm, the average length of random spinifex-textured olivine crystals varies from 2·8 to 8·5 mm. In komatiite flows, this grain size is only reached at a distance of about 25 cm from the top of the unit. Deeper in the sill, at a depth of about 30 cm, the grain sizes are about the same as those in a flow. The differences between a flow and a sill are less pronounced in the interior, because once a solid crust has formed at the top of the unit, the cooling conditions are very similar. In both cases, if olivine phenocrysts have had time to accumulate (via settling or in situ crystallization), a lower olivine cumulate layer develops and spinifex textures form through downward crystallization of olivine (or pyroxene) in the thermal and chemical gradient in the liquid just beneath the solid crust (Faure et al., 2002).
Vesicles are most abundant in a zone within 1 cm of the upper contact of the sill. The largest vesicles occur at a distance of 2–3 mm from the contact and, immediately adjacent to the contact, there are only a few, smaller, sparsely distributed vesicles. Some of the vesicles are filled with quenched silicate liquid, which apparently drained into the vesicles during cooling (Fig. 7d); others are filled with secondary minerals (Fig. 6). A chilled zone containing abundant small vesicles, as might form during the quenching of lava, is absent. Although it is possible that the upper part of the unit may have been melted, destroying any previous vesicles, the present distribution is consistent with the komatiite sill being intrusive into the overlying unit. Beresford et al. (in preparation) have shown that the komatiitic units in the eastern side of the outcrop (Fig. 2) commonly have vesicular tops; they demonstrated using field relations and textures that the volatiles were derived through interaction between the komatiite magma and unconsolidated, water-rich sediments.
The komatiite sill directly overlies a series of thinner komatiitic basalt units that clearly intruded into a sequence of unconsolidated sediments (Davis, 1997, 1999; Houlé et al., 2002a, 2002b, in preparation; Cas et al., 2003).
The possibility that the two units illustrated in Fig. 3 erupted rapidly enough to impede the formation of chilled margins between them was considered, but the grain-size variations (Fig. 8) and absence of polyhedral jointing militate against such an interpretation. In places where the structures and textures of lavas indicate that they have been emplaced as compound lava lobes in rapid succession (e.g. Pyke Hill), the boundaries between flows are defined by the 10–20 cm thick, aphanitic, polyhedrally jointed upper zones in the underlying flows and 2–5 cm thick, aphanitic, lower chilled margins in the overlying flows (Pyke et al., 1973). In the Dundonald units, in contrast, the thickness of the upper fine-grained margin is less than 1 cm and there is no chill whatsoever at the base of the overlying unit.
We have also considered the possibility that the komatiites at Dundonald Beach, which are thicker and contain a greater proportion of olivine cumulate rock than those on Pyke Hill, may represent magma conduits and remained hotter for a longer period of time than thinner flows, thereby impeding the formation of chilled margins. However, even very thick (up to 100 m) olivine cumulate lava channels and channelized sheet flows in the Kambalda area exhibit well-defined upper and lower aphanitic margins (Lesher et al., 1984; Lesher, 1989).
Finally, the possibility that the komatiites at Dundonald Beach represent beheading and changing crystallization dynamics in a reactivated lava channel complex, analogous to the interpretation for internal zones of branching crescumulate olivine and vesicular zones within the thick olivine mesocumulate lava channel complexes at Kambalda (Lesher et al., 1984; Cowden, 1988; Lesher, 1989) was also considered. However, the compound units at Kambalda do not contain internal chilled margins or spinifex-textured rocks like those at Dundonald Beach and, as argued below, the textures within the lower unit in Fig. 3 suggests emplacement from below—not from above. For these reasons, we prefer to interpret the contact between these units as intrusive and attribute its irregular nature to thermomechanical erosion of the hot lower part of the overlying unit during intrusion of the lower unit. Along strike, the lower komatiitic unit clearly cuts down into the underlying komatiitic basalt units (Fig. 2).
On the basis of all these arguments, we conclude that the komatiite unit that is the subject of this study, and most, if not all, of the other komatiitic rocks in the Dundonald Beach area are best interpreted as a series of interfingering, mainly concordant intrusions.
Interpretation of textures in the upper part of the sill
The textures and olivine habits in the upper part of the komatiite sill are highly unusual and require further explanation. More specifically, we need to account for the lack of chilling at the upper margin of the sill and in the lower margin of the overlying unit, the lack of evidence of baking and other thermal effects, the large size and polyhedral habit of the olivine grains immediately below the contact, and the change in habit of olivine and chromite grains within the spinifex zone. We must also explain the changes in olivine morphologies from the upper contact zone of the komatiite sill; from relatively large (compared with cumulus olivines of the overlying unit) subhedral to rounded, solid polyhedral olivine (Layer 1 in Fig. 6) 
highly skeletal, subequant hopper olivine (Layer 2) skeletal bladed olivine in the random spinifex layer large parallel olivine plates and coarse dendrites in the lower part of the spinifex layer (Figs 4 and 5). All of these features can be explained if we take into account the mechanisms of nucleation and crystal growth in silicate liquids, and the likely temperature distribution in the sill immediately after its emplacement.
From the work of experimental petrologists, particularly Donaldson (1978, 1982), we know that the morphology of olivine crystallizing in rapidly cooled mafic–ultramafic magmas depends primarily on (1) the cooling rate and/or degree of undercooling, (2) the nucleation mechanism, and (3) the composition of the liquid. Slow cooling or crystallization from modestly undercooled liquids results in polyhedral olivine morphologies; faster cooling or crystallization from highly undercooled liquids leads to more elongate and more skeletal grains. Because of differences in viscosity and melt structure, at a constant cooling rate, olivine that crystallizes in basaltic magma has a more skeletal and elongate habit than olivine crystallizing in an ultramafic magma. Faure et al. (2002) have demonstrated that when olivine crystallizes within a thermal gradient, such as along the margin of a flow or sill, the olivine grains acquire a preferred orientation and have a more skeletal habit than olivine that crystallizes in the absence of a thermal gradient. Finally, pre-existing crystals act as nuclei and lead to the growth of crystals with less skeletal morphologies than those that nucleate homogeneously (Lofgren, 1983).
Figure 9 shows the results of modelling the evolution of temperatures in a sill and the overlying unit. The parameters used in the model are given in the figure caption. Although a chilled margin forms immediately after emplacement, it is immediately remelted as heat is transferred from the sill to the overlying rock. Twenty seconds after emplacement, the temperatures in the silicate liquid within a few millimetres of the contact are significantly below the liquidus, and a millimetre-wide zone of supersolidus temperatures, a zone of partial melt, has formed at the base of the overlying unit. Sixty minutes after emplacement, the zone of sub-liquidus temperatures extends several centimetres downward into the sill and the partially molten zone in the overlying unit is about 1 cm thick. The textures in the rocks at both sides of the contact, and particularly the morphologies of the olivine grains, can be explained by the temperature variations shown in the figure. The irregular nature of the contact and especially the embayments filled with liquid from the sill (labelled ‘e’ in Fig. 5a) and the presence of isolated packets of olivine grains at the base of the olivine cumulate (labelled ‘p’ in the same figure) result from partial melting at the base of the overlying unit. The randomly oriented hopper olivine crystals in the upper few centimetres of the sill nucleated homogeneously and grew from moderately to highly undercooled liquid in the upper part of the sill. The large, solid, polyhedral crystals in Layer 1, immediately below the contact (Fig. 6a and b), nucleated heterogeneously and grew on the olivine grains in the overlying cumulate rock. The platy and coarse dendritic spinifex crystals deeper in the sill grew later, at lower cooling rates in a thermal gradient—conditions that Faure et al. (2002, in preparation) have shown to lead to constrained growth, elongate morphologies and preferred orientations.
|
L = 2·8 x 103 kg/m3, heat conductivity 
L = 1 W/K/m, in the solid, Intrusive komatiites and the water contents of komatiite magmas
Grove, de Wit, Parman and coworkers have argued in a series of papers (Grove et al., 1994
, 1996, 1999; Parman et al., 1997) that komatiites in the Barberton belt, South Africa, were emplaced as sills at a depth of about 6 km. They proposed that the parental komatiite magma contained up to 6% H2O, and that this magma crystallized within the sills before all the H2O had escaped. They attributed the growth of spinifex textures and the compositions of calcic pyroxene interstitial to olivine in spinifex-textured rock to crystallization in a hydrous magma at a pressure of about 0·2 GPa. Although the possibility of an intrusive origin for some of these units has since been eliminated by the detailed mapping of Dann (2000), the theory remains correct. If a hydrous magma becomes trapped within the middle to upper crust, the magma may crystallize before the H2O escapes. If, on the other hand, the komatiite erupts at the surface as a lava flow, the low solubility of water at low pressure and the very low viscosities of komatiite magmas should provoke rapid degassing of the lava (Green, 1974; Arndt et al., 1998; Cashman & Blundy, 2000). Parman et al. (2001) and Dann (2001) have suggested that sluggish kinetics may inhibit degassing, citing as evidence certain boninites that appear to have erupted as non-vesiculated magmas and that contain glass with up to 6% H2O. Those workers suggested that hydrous komatiite may also erupt in a non-equilibrium, non-degassed state. How does our study of the Dundonald komatiites, and of komatiites in other parts of the Abitibi belt, contribute to the problem?
A high-Mg unit in Boston Township, 100 km south-east of Dundonald (Fig. 1), has a very peculiar chemical composition, being unusually rich in Fe, Ti and highly incompatible trace elements. Stone et al. (1987, 1997) identified magmatic amphibole in several samples from this unit and argued that the magma originally contained 1–2% H2O. This magma appears to have formed through low-degree melting of a slightly hydrous, chemically anomalous part of the mantle. Detailed mapping of the upper contact of this unit by Houlé et al. (in preparation) indicates that it is intrusive in two separate locations: apophyses from the upper chilled margin transgress overlying rocks and the upper margin contains fragments of the overlying rocks. There is little doubt that this komatiite contained a small amount of H2O and that it was emplaced as sill. Although it remains possible that the H2O was acquired locally, by assimilation of sediments, it is likely that a picritic magma enriched in Fe, Ti and incompatible trace elements was derived from an enriched, hydrous source.
In contrast, the intrusive komatiites at Dundonald, like most komatiites in the Abitibi belt (Sproule et al., 2002) and other late Archean belts, are strongly depleted in highly incompatible lithophile elements (e.g. U, Th, Nb, Ta, LREE). As noted by Arndt et al. (1998), unless the magma was systematically enriched in H2O relative to these other elements, it must have been essentially anhydrous. The komatiite sill does contain vesicles, but they are very sparse (always less than 1% of the rock) and are confined to the uppermost part of the unit, consistent with only minor volatile contents.
Other constraints on the initial water-content of the magma come from more involved arguments. Although these komatiites are intrusive in the strictest sense of the term, they erupted very close to the ocean floor. Houlé et al. (2002a, 2002b, in preparation) and Cas et al. (2003) demonstrated clearly that some of the komatiitic basaltic units below the komatiite sill invaded unconsolidated and water-saturated sediments, suggesting that dense komatiite magma rose through the crust until it encountered a layer of weak, low-density sediment, which it intruded instead of continuing to the surface. It is difficult to judge the thickness of the sediment pile, but based on the minor amount of sediment in the sequence, it is unlikely to be more than a few hundred metres. Even if we add the pressure imposed by several kilometres of ocean water, the total pressure is unlikely to have reached 0·1 GPa. Komatiite magma can dissolve no more than a few hundred ppm H2O at equilibrium under these conditions. Parman et al. (2001) and Dann (2001) have suggested that a water-rich magma might have been emplaced in a metastable state, unable to nucleate and degas, but as noted above, sills crystallize more slowly than flows, so water would be more likely to have nucleated in sills than in flows, facilitating exsolution and accumulation beneath the cap of overlying olivine cumulate rock, forming highly vesicular zones or large gas cavities, as observed in the upper parts of continental flood basalts (Self et al., 1997). This picture is in marked contrast to the very sparsely vesicular nature of the upper part of the Dundonald komatiite sill. As mentioned earlier, the vesicular upper parts of the komatiitic unit in the eastern part of the outcrop have been attributed by Beresford et al. (in preparation) to interaction between the komatiite magma and the enclosing unconsolidated sediments.
In the case of komatiite lava flows, it is conceivable that the magma could have lost volatiles during lava fountaining and that the lava flows subsequently crystallized from degassed lava. Such an interpretation is at odds, however, with the rarity of vesicular, scoriaceous, or fragmental komatiite in the Abitibi belt and elsewhere. The fragmental komatiites at Scotia (Page & Schmulian, 1981), in Finland (Saverikko, 1985) and on Gorgona Island (Echeverria & Aitken, 1986), which had been interpreted as pyroclastic deposits, are now interpreted as epiclastic deposits (Scotia) or hyaloclastites (Finland, Gorgona). In any case, because the Dundonald komatiites invaded unconsolidated sediments before reaching the surface, they would have had no opportunity to degas through lava fountaining.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSGEOLOGICAL SETTINGFIELD RELATIONSTEXTURAL VARIATIONSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
On the basis of detailed field and petrographic study, we conclude that the sill and underlying units are intrusive, formed from essentially anhydrous komatiite liquid emplaced in a volcano-sedimentary succession very close to the ocean floor. These rocks may, therefore, be considered as the hypabyssal component of the komatiitic volcanic series. Magmatically, they are identical to komatiite flows and they owe their intrusive origin only to the nature of the rocks that they encountered on their way to the surface.
|
ACKNOWLEDGEMENTS |
|---|
The fieldwork for this project was supported by grants from the Jacques Cartier Centre in Lyon (to N.T.A.) and the Natural Sciences and Engineering Council of Canada (to C.M.L.). We have benefited greatly from discussions in the field with Paul Davis, Harold Gibson, Phil Thurston, Ray Cas and Steve Beresford. We thank Mike Higgins, François Faure, Marjorie Wilson and an anonymous reviewer for comments that helped us improve themanuscript.
* Corresponding author. E-mail: arndt{at}ujf-grenoble.fr
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSGEOLOGICAL SETTINGFIELD RELATIONSTEXTURAL VARIATIONSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Allègre, C. J. (1982). Genesis of Archaean Komatiites in a Wet Ultramafic Subducted Plate. In: Arndt, N. T. & Nisbet, E. G. (eds) Komatiites. London: George Allen and Unwin, pp. 495–500.
Arndt, N. T. (1986). Differentiation of komatiite flows. Journal of Petrology 27, 279–301.
Arndt, N. T. (1994). Komatiiites. In: K. C. Condie (ed.) Archean Crustal Evolution. Amsterdam: Elsevier, pp. 11–44.
Arndt, N. T., Albarède, F., Cheadle, M. J., Ginibre, C., Herzberg, C., Jenner, G., et al. (1998). Were komatiites wet? Geology 26, 739–742.
Ayer, J. A., Amelin, Y., Corfu, F., Kamo, S., Ketchum, J., Kwok, K., et al. (2002a). Evolution of the southern Abitibi greenstone belt based on U–Pb geochronology: autochthonous volcanic construction followed by plutonism, regional deformation and sedimentation. Precambrian Research 115, 63–95.[CrossRef][Web of Science]
Ayer, J. A., Ketchum, J. W. F. & Trowell, N. F. (2002b). New Geochronological and Neodymium Isotopic Results from the Abitibi Greenstone Belt, with Emphasis on the Timing and the Tectonic Implications of Neoarchean Sedimentation and Volcanism. In: Summary of Field Work and Other Activities 2002. Ontario Geological Survey, Open File Report 6100, pp. 5-1–5-16.
Barnes, S.-J. (1983). A comparative study of olivine and clinopyroxene spinifex flows from Alexo, Abitibi greenstone belt, Canada. Contributions to Mineralogy and Petrology 83, 293–308.[CrossRef][Web of Science]
Beresford, S. W. & Cas, R. A. F. (2001). Komatiitic invasive lava flows. Canadian Mineralogist 39, 525–535.[CrossRef][Web of Science]
Bowen, N. L. (1956). The Evolution of the Igneous Rocks. New York: Princeton University Press, 334 pp.
Cas, R. A. F. & Beresford, S. W. (2001). Field characteristics and erosional processes associated with komatiitic lavas: implications for flow behavior. Canadian Mineralogist 39, 505–524.[CrossRef][Web of Science]
Cas, R. A. F., Houlé, M., Beresford, S., Lesher, M., Fowler, A., Arndt, N., et al. (2003). Are spinifex-cumulate textures komatiitic rocks units necessarily lavas? Archean examples from Dundonald Beach, Abitibi greenstone belt, Canada. 23rd IUGG General Assembly, Sapporo, Japan.
Cashman, K. & Blundy, J. (2000). Degassing and crystallization of ascending andesite and dacite. In: Francis, P., Neuberg, J. & Sparks, R. S. J. (eds) Causes and Consequences of Eruptions of Andesite Volcanoes. London: Royal Society, pp. 1487–1513.
Cowden, A. (1988). Emplacement of komatiite lava flows and associated nickel sulfides at Kambalda, Western Australia. Economic Geology 83, 436–442.
Dann, J. C. (2000). The Komati Formation, Barberton Greenstone Belt, South Africa, part I: new map and magmatic architecture. South African Journal of Earth Sciences 6, 681–730.
Dann, J. C. (2001). Vesicular komatiites, 3·5-Ga Komati Formation, Barberton Greenstone Belt, South Africa: inflation of submarine lavas and origin of spinifex zones. Bulletin of Volcanology 63, 462–481.[CrossRef][Web of Science]
Davis, P. C. (1997). Volcanic stratigraphy of the late Archean Kidd–Munro assemblage in Dundonald and Munro Townships and genesis of associated nickel and copper–zinc volcanogenic massive sulfide deposits, Abitibi greenstone belt, Ontario, Canada. M.Sc. thesis, University of Alabama, Tuscaloosa, 201 pp.
Davis, P. C. (1999). Physical Volcanology of Komatiites Flows at Dundonald Beach and Alexo Mine, Dundonald and Clergue Townships, Kidd–Munro Assemblage, Ontario Portion and the Abitibi Greenstone Belt. Day 2: Guidebook Series 1. Sudbury, Ont.: Mineral Exploration Research Centre, Laurentian University, 27–44.
Donaldson, C. H. (1978). An experimental investigation of olivine morphology. Contributions to Mineralogy and Petrology 57, 297–300.[CrossRef]
Donaldson, C. H. (1982). Spinifex-textured komatiites: a review of textures, mineral compositions and layering. In: Arndt, N. T. & Nisbet, E. G. (eds) Komatiites. London: George Allen and Unwin, pp. 211–244.
Echeverria, L. M. & Aitken, B. (1986). Pyroclastic rocks: another manifestation of ultramafic volcanism of Gorgona Island, Colombia. Contributions to Mineralogy and Petrology 92, 428–436.[CrossRef][Web of Science]
Faure, F., Arndt, N. T. & Libourel, G. (2002). Crystallisation of plate spinifex texture at 1 atm. pressure in a thermal gradient. Goldschmidt Conference Abstract. Geochimica et Cosmochimica Acta Special Supplement, A225.
Green, D. H. (1974). Genesis of Archaean peridotitic magmas and constraints on Archaean geothermal gradients and tectonics. Geology 3, 15–18.
Grove, T. L., Gaetani, G. A. & de Wit, M. J. (1994). Spinifex textures in 3·49 Ga Barberton Mountain Belt komatiites: evidence for crystallization of water-bearing, cool magmas in the Archean. EOS Transactions, American Geophysical Union 75(Spring), 354.
Grove, T. L., Gaetani, G. A., Parman, S., Dann, J. & de Wit, M. J. (1996). Origin of spinifex textures in 3·49 Ga komatiite magmas from the Barberton Mountainland, South Africa. EOS Transactions, American Geophysical Union 77, 281.
Grove, T. L., de Wit, M. J. & Dann, J. (1997). Komatiites from the Komati Type Section, Barberton, South Africa. In: de Wit, M. J. & Ashwal, L. D. (eds) Greenstone Belts. Oxford: Oxford Science Publications, pp. 422–437.
Grove, T. L., Parman, S. W. & Dann, J. C. (1999). Conditions of magma generation for Archean komatiites from the Barberton Mountainland, South Africa. In: Fei, Y., Bertka, C. M. & Mysen, B. O. (eds) Mantle Petrology: Field Observations and High-Pressure Experimentation. Houston: Geochemical Society, pp. 155–167.
Houlé, M. G., Gibson, H. L., Lesher, C. M., Davis, P. C. & Arndt, N. T. (2002a). The role of subvolcanic sills and peperites in facilitating the eruption of Archean komatiitic flows at the Dundonald Beach Ni–Cu–(PGE) Deposit, northern Ontario. GSA Annual Meeting, Denver, CO, p. 513.
Houlé, M. G., Davis, P. C., Lesher, C. M. & Arndt, N. T. (2002b). Extrusive and intrusive komatiites, komatiitic basalt and peperite and ore genesis at the Dundonald Ni–Cu–(PGE) Deposit, Abitibi Greenstone Belt, Canada. 9th International Platinum Symposium Abstract Volume, Billings, MT, pp. 181–184.
Jolly, W. T. (1982). Progressive metamorphism of komatiites and related Archaean lavas of the Abitibi area, Canada. In: Arndt, N. T. & Nisbet, E. G. (eds) Komatiites. London: George Allen and Unwin, pp. 247–266.
Lesher, C. M. (1989). Komatiite-associated nickel sulfide deposits. In: Whitney, J. A. & Naldrett, A. J. (eds) Ore Deposition Associated with Magmas. Reviews in Economic Geology 4, 45–101.
Lesher, C. M. & Keays, R. R. (2002). Komatiite-associated Ni–Cu–(PGE) deposits: mineralogy, geochemistry, and genesis. In: Cabri, L. J. (ed.) The Geology, Geochemistry, Mineralogy, and Mineral Beneficiation of the Platinum-Group Elements. Canadian Institute of Mining, Metallurgy and Petroleum Special Volume 54, 579–617.
Lesher, C. M. & Stone, W. E. (1996). Exploration geochemistry of komatiites. In: Wyman, D. (ed.) Igneous Trace Element Geochemistry: Applications for Massive Sulphide Exploration. Geological Association of Canada, Short Course Notes 12, 153–204.
Lesher, C. M., Arndt, N. T. & Groves, D. I. (1984). Genesis of komatiite-associated nickel sulphide deposits at Kambalda, Western Australia: a distal volcanic model. In: Buchanan, D. L. & Jones, M. J. (eds) Sulphide Deposits in Mafic and Ultramafic Rocks. London: Institution of Mining and Metallurgy, pp. 70–80.
Lofgren, G. E. (1983). Effect of heterogeneous nucleation on basaltic textures: a dynamic crystallization study. Journal of Petrology 24, 229–255.
Muir, J. E. & Comba, C. D. A. (1979). The Dundonald deposit: an example of volcanic-type nickel–sulfide mineralization. Canadian Mineralogist 17, 351–360.
Naldrett, A. J. & Mason, G. D. (1968). Contrasting Archaean ultramafic igneous bodies in Dundonald and Clergue Townships, Ontario. Canadian Journal of Earth Sciences 5, 111–143.
O'Hara, M. J. (1961). Petrology of the Scourie Dyke. Mineralogical Magazine 32, 848–865.
Page, M. L. & Schmulian, M. L. (1981). The proximal volcanic environment of the Scotia nickel deposits. Economic Geology 76,1469–1479.
Parman, S., Grove, T. L., Dann, J. & de Wit, M. J. (1996). Pyroxene compositions in 3·49 Ga komatiite: evidence of variable H2O content. EOS Transactions, American Geophysical Union 77, 280.
Parman, S., Dann, J., Grove, T. L. & de Wit, M. J. (1997). Emplacement conditions of komatiite magmas from the 3·49 Ga Komati Formation, Barberton Greenstone Belt, South Africa. Earth and Planetary Science Letters 150, 303–323.[CrossRef][Web of Science]
Parman, S., Grove, T. L. & Dann, J. (2001). The production of Barberton komatiites in an Archean subduction zone. Geophysical Research Letters 28, 2513–2516.[CrossRef][Web of Science]
Pyke, D. R. (1982). Geology of the Timmins area, District of Cochrane. Ontario Division of Mines, Geological Report 219, 141 pp.
Pyke, D. R., Naldrett, A. J. & Eckstrand, O. R. (1973). Archean ultramafic flows in Munro Township, Ontario. Geological Society of America Bulletin 84, 955–978.
Renner, R., Nisbet, E. G., Cheadle, M. J., Arndt, N. T., Bickle, M. J. & Cameron, W. E. (1993). Komatiite flows from the Reliance Formation, Belingwe Belt, Zimbabwe: I—petrography and mineralogy. Journal of Petrology 35, 361–400.[Web of Science]
Saverikko, M. (1985). The pyroclastic komatiite complex at Sattasvaara in northern Finland. Bulletin of the Geological Society of Finland 57, 55–87.
Self, S., Thorvaldur, T. & Keszthelyi, L. (1997). Emplacement of continental flood basalt lava flows. In: Mahoney, J. J. & Coffin, M. F. (eds) Large Igneous Provinces. Washington, DC: American Geophysical Union, pp. 381–410.
Skilling, I., White, J. D. L. & McPhie, J. (2002). Peperite: review of magma–sediment mingling. Journal of Volcanology and Geothermal Research 114, 1–17.[CrossRef][Web of Science]
Sproule, R. A., Lesher, C. M., Ayer, J. A., Thurston, P. C. & Herzberg, C. T. (2002). Spatial and temporal variations in the geochemistry of komatiitic rocks in the Abitibi greenstone belt. Precambrian Research 115, 153–186.[CrossRef][Web of Science]
Stone, M. S. & Stone, W. E. (2000). A crustally contaminated komatiitic dyke–sill–lava complex, Abitibi greenstone belt, Ontario. Precambrian Research 102, 21–46.[CrossRef][Web of Science]
Stone, W. E., Jensen, L. S. & Church, W. R. (1987). Petrography and geochemistry of an unusual Fe-rich basaltic komatiite from Boston Township, northeastern Ontario. Canadian Journal of Earth Sciences 24, 2537–2550.[Web of Science]
Stone, W. E., Déloule, E., Larson, M. S. & Lesher, C. M. (1997). Evidence for hydrous high-MgO melts in the Precambrian. Geology 25, 143–146.
Viljoen, M. J. & Viljoen, R. P. (1969a). Archaean vulcanicity and continental evolution in the Barberton region, Transvaal. In: Clifford, T. N. & Gass, I. (eds) African Magmatism and Tectonics. Edinburgh: Oliver and Boyd, pp. 27–39.
Viljoen, M. J. & Viljoen, R. P. (1969b). The geology and geochemistry of the lower ultramafic unit of the Onverwacht Group and a proposed new class of igneous rocks. Geological Society of South Africa, Special Publication 2, 55–85.
White, J. D. L., McPhie, J. & Skilling, I. (2000). Peperite: a useful genetic term. Bulletin of Volcanology 62, 65–66.[CrossRef][Web of Science]
Williams, D. A. C. (1979). The association of some nickel sulfide deposits with komatiitic volcanism in Rhodesia. Canadian Mineralogist 17, 337–349.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. T. Stiegler, D. R. Lowe, and G. R. Byerly Abundant pyroclastic komatiitic volcanism in the 3.5-3.2 Ga Barberton greenstone belt, South Africa Geology, October 1, 2008; 36(10): 779 - 782. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. G. Houle, H. L. Gibson, C. M. Lesher, P. C. Davis, R. A. F. Cas, S. W. Beresford, and N. T. Arndt Komatiitic Sills and Multigenerational Peperite at Dundonald Beach, Abitibi Greenstone Belt, Ontario: Volcanic Architecture and Nickel Sulfide Distribution Economic Geology, September 1, 2008; 103(6): 1269 - 1284. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Hanley THE ROLE OF ARSENIC-RICH MELTS AND MINERAL PHASES IN THE DEVELOPMENT OF HIGH-GRADE Pt-Pd MINERALIZATION WITHIN KOMATIITE-ASSOCIATED MAGMATIC Ni-Cu SULFIDE HORIZONS AT DUNDONALD BEACH SOUTH, ABITIBI SUBPROVINCE, ONTARIO, CANADA Economic Geology, March 1, 2007; 102(2): 305 - 317. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||