THE BJERKREIM - SOKNDAL LAYERED
INTRUSION, S.W. NORWAY
by Brian Robins1 and J. Richard Wilson2
2Department of Earth Sciences,
University of Aarhus, 8000 Aarhus C, Denmark
The Bjerkreim-Sokndal Intrusion
(BKSK) (Michot 1960, 1965, Duchesne 1987, Wilson et al. 1996) is a large
(40 km long and up to 15 km wide), Late Proterozoic layered intrusion that
occupies an area of about 230 km2
(Fig. 1). Lithologically the intrusion consists of virtually all of the rock
types belonging to the anorthosite kindred, i.e. andesine anorthosite,
troctolite, leuconorite, norite, gabbronorite, jotunite (hypersthene
monzodiorite), mangerite (hypersthene monzonite), quartz mangerite and igneous
charnockite (hypersthene granite). Anorthosite, leuconorite and norite are
accompanied by ilmenite-rich rocks.
The BKSK is emplaced in
granulite-facies quartzo-feldspathic and mafic gneisses as well as anorthosite
and leuconorite belonging to the Egersund-Ogna (Michot & Michot 1970,
Duchesne & Maquil 1987), Håland-Helleren (Michot 1961) and Åna-Sira (Krause
et al. 1985, Duchesne & Michot
1987) massifs, and xenoliths of all these host rocks are common within the
intrusion itself (Duchesne 1970).
The BKSK and the various anorthosite massifs are cut by members of a suite of small plutons and wide, laterally-persistent dykes of jotunite, some of which are differentiated (Duchesne et al. 1989, Wilmart et al. 1989). The most voluminous of the jotunites that cut the northern part of the BKSK is the Lomland dyke/sill complex (Duchesne et al. 1989). The BKSK is also cut by members of the Egersund swarm of basaltic dykes.
The Bjerkreim-Sokndal Intrusion has
generally been described as a lopolith, but recent detailed mapping shows it to
be a trough-like, discordant intrusion. Modelling of the associated +10-30 mgal
gravity anomaly (Smithson & Ramberg 1979) and a seismic reflection profile
(Deemer & Hurich 1997) shows that the base of the intrusion lies at a depth
of 4-5 km.
Layering within the intrusion is deformed into a deep, doubly-plunging syncline that branches in the south around a dome cored by the Åna-Sira anorthosite massif (Fig. 1). The core of the syncline is occupied by quartz mangerite and charnockite, which do not exhibit modal or textural layering, and these are separated in places from the underlying mangerite by a zone with abundant wall-rock xenoliths. The magnitude of the gravity lows over the granitoids suggest a maximum thickness of about 2 km (Smithson & Ramberg 1979). There is no evidence that the roof of the intrusion is preserved anywhere within the confines of the present outcrop.
Figure 1. Location (A), sketch map (B),
and stratigraphic subdivision of the Bjerkreim lobe (C) of the Bjerkreim-Sokndal Intrusion
The BKSK consists of three lobes;
the Bjerkreim lobe in the north-west, and the smaller Sokndal and Mydland lobes
to the south and south-east respectively (Fig.1B). Modal layering and phase
contacts in the Bjerkreim lobe are disposed in a syncline that plunges
south-east at 20-40°. In the steep limbs of the syncline the cumulates are
foliated, generally in the plane of modal layering. In places, cumulus minerals
form augen in a foliated matrix, small shear zones are developed, and there is
a strong mineral lineation. Linear mineral and magnetic fabrics dominate in the
core of the syncline (Paludan et al.
1994, Bolle et al. 2000). Cumulus plagioclases are strained or
recrystallised to shape-oriented polygonal aggregates, whereas prismatic
Ca-poor pyroxenes are commonly kinked or bent. Uniform paleomagnetic vectors in
different parts of the intrusion (Poorter 1972) suggest that the deformation
and development of the synformal disposition of the layering took place at
temperatures in excess of the Curie point (550-650°C). The synformal
disposition of the layering is inferred to be due to gravitational foundering
(Paludan et al. 1994, Bolle et al.
2000).
The Layered Series in the Bjerkreim lobe has a thickness of >7000m in the axial region of the syncline and can be divided into 6 megacyclic units (MCU 0 to IV) which exhibit characteristic sequences of cumulates (Figs. 1 and 2). The megacyclic units can be further subdivided into zones a-f, based on assemblages of cumulus minerals (Figs. 2 and 3).
The
megacyclic units vary in stratigraphic thickness, lateral persistence and in the
nature of the layer sequences they exhibit. The lower three megacyclic units,
exposed only in the northernmost part of the intrusion, are individually as
much as 1300m thick but show a pronounced southward thinning in the western
limb of the syncline and are not developed in the southern parts of the
Bjerkreim lobe.
Figure 2. Stratigraphy of the Layered Series of the Bjerkreim-Sokndal
Intrusion as developed in the axial region of the Bjerkreim lobe and its
subdivision into macrocyclic units and cumulate zones.
The lowermost cumulates are exposed in the north-western part of the Bjerkreim lobe and consist of plagioclase-hypersthene-ilmenite cumulates (phiC). They are regarded as the top of MCU 0, the rest of which, together with an unknown thickness of cumulates, is hidden. These cumulates are overlain successively by pC, piC and phC belonging to MCU IA (~1300m thick in a profile along the axial trace of the syncline).
Figure 3. Generalised stratigraphy,
cumulus minerals and cryptic layering of the Bjerkreim-Sokndal Layered Intrusion.
Revised after Wilson et al. (1996)
This sequence is repeated in MCU IB (~875m thick) which locally also displays more evolved lithologies with the entry of cumulus Ca-rich pyroxene, followed by apatite and magnetite. MCUs 0-IB are characterised by the presence of plagioclase megacrysts (up to 10 cm long) in all rocks with the exception of the most evolved cumulates at the top of MCU IB.
MCU II (reaching a thickness of
1600m) consists of a thin layer of magnetite-bearing piC overlain exclusively
by phiC. The appearance of cumulus magnetite in the leuconorites at the base of
MCU II and its absence in the overlying cumulates suggests affinities with the
olivine-bearing zones near the bases of succeeding MCUs. The base of MCU II is characterised by a marked regression in the
composition of cumulus plagioclase (Fig. 3).
MCU
III (maximum ~1100m thick) generally has a lower zone (Zone IIIa in Fig. 3) up
to 140m thick that consists mainly of pC, but with interlayered iC, phiC and
hiC. The base of the zone is marked by a thin (<10m) sulphide-bearing
subzone, unique in the Layered Series, that consists of ilmenite norite, mafic
ilmenite norite or massive orthopyroxenite. Zone a is characterised by a
stratigraphic regression to higher-temperature mineral compositions. In the
axial region of the intrusion and on the southern flank, zone a is overlain by
leucotroctolite (zone b) that contains cumulus magnetite in addition to
plagioclase, olivine and ilmenite. Together with the similar rocks near the
base of MCU IV, the zone IIIb cumulates are the most primitive cumulates in the
intrusion. The leucotroctolite is in turn overlain by phiC of zone c, followed
by magnetite norite (Zone d) and gabbronorite (Zone e) with the successive
(re-)entry of cumulus magnetite and then apatite together with Ca-rich
pyroxene. In the eastern flank of the intrusion MCU III is relatively condensed
and zone b is absent. Zone e is only developed in the flanks of the lobe. In
the axial region the base of MCU IV rests on zone d cumulates.
MCU IV (maximum thickness ~1800m) displays a sequence similar to MCU III. MCU IV contains, however, additional, more-evolved cumulates. Michot (1960) recognised the prominent olivine-bearing zone b near the base of MCU IV and referred to it as the "Svalestad horizon". It has a thickness of about 100m and is laterally persistent along strike for about 24km. Olivines in the olivine-bearing zones near the bases of MCU III and IV are partially or completely replaced by orthopyroxene-Fe-Ti oxide symplectites, but the zones are texturally distinctive even where no olivine remains. Small amounts of biotite and hornblende also occur in the olivine-bearing zones. Ca-poor pyroxene is inverted pigeonite in the upper part of MCU IV (Zone f) which grades into overlying mangerite through a jotunitic Transition Zone (TZ) whose base is defined by the re-entry of olivine (~Fo50), which more or less coincides with the appearance of interstitial alkali feldspar (Duchesne et al. 1987). With the appearance of cumulus mesoperthite the rocks grade upwards from jotunite to mangerite, which in turn passes into massive quartz mangerite and, locally, charnockite. Even in these highly-evolved rocks, hydrous phases are not abundant: Calcic amphibole is generally a minor mineral (except in the uppermost part of the granitoids where it may occur as large oikocrysts) and biotite is generally an accessory mineral. The combined thickness of the mangerite, quartz mangerite and charnockite is >350m (Rietmeijer 1979) and may be as much as 2km (Smithson & Ramberg 1979).
Viewed on a broad scale (Figs. 2 and 3), the lower part of the Layered Series is dominated by plagioclase cumulates, the middle part by plagioclase-hypersthene-ilmenite cumulates and the upper part by plagioclase-hypersthene/pigeonite-augite-ilmenite-magnetite-apatite cumulates. Combined with the reversals to relatively primitive mineral assemblages at the bases of the MCUs, this is strong, first-order evidence that the Layered Series crystallised in a continuously fractionating, periodically replenished magma chamber. Replenishment events were few in number and widely spaced in time. The bulk composition of the magma occupying the chamber after each replenishment event can be judged by the relative volumes of the repective types of cumulate that constitute the successive macrocyclic units. In the lowermost units the cumulates represented are predominantly high-temperature varieties and the proportion of more-evolved cumulates generally increases in the upper units. This pattern suggests that the bulk composition of the magma occupying the chamber became progressively more evolved with time.The base of MCU IV seems to reflect the last major influx of magma into the BKSK chamber. This replenishment event appears to have involved a large volume of magma and was associated with very significant expansion of the chamber. The regressive layered sequences beneath the most-primitive leucotroctolitic cumulates in MCUs III and IV are up to 120m thick, showing that replenishment of the magma chamber must have taken place over a prolonged period of time. After the influx of magma reflected by the MCU III/IV transition, fractional crystallisation was apparently uninterrupted and we interpret the stratigraphic transition to mangerite, quartz mangerite and charnockite as reflecting progressive differentiation of the residual magma, probably containing a significant crustal component due to assimilation of country rocks and hybridisation with roof melts.
Ilmenite
is an early-crystallising, almost ubiquitous mineral in the Bjerkreim-Sokndal
intrusion. Substantial concentrations of ilmenite are restricted, however, to
the regressive sequences that occur at the bases of the upper MCUs, especially
units III and IV. Ilmenite and plagioclase-ilmenite cumulates are prominent in
zone IVa in the axial region of the intrusion, and ilmenite-rich melanorites
are found at the base of MCU III to the east of Teksevatnet. This relationship
indicates that the generation of ilmenite-rich cumulates was related to
replenishment of the Bjerkreim-Sokndal magma-chamber and was probably a
consequence of turbulent mixing of the resident magma with inflowing
more-primitive jotunitic magma.
Ilmenite exhibits a general decrease in hematite content through the Layered Series, from 16-20% in the lower part of the series where it is the only cumulus Fe-Ti oxide, to about 2% in the mangerite and quartz mangerite (Duchesne 1972). This pattern is repeated on a smaller scale in the individual MCUs. Ilmenite contains about 0.3% V2O3 in the lower part of the Layered Series and shows an identical variation to hematite content. Manganese in ilmenite, however, increases almost continuously through the Layered Series, from about 0.3% MnO in leuconorites near the base to 1.0% MnO in mangerite; breaks in the trend at contacts between MCUs are slight. Nickel and chromium are enriched in ilmenite at the bases of MCUs where concentrations can be as high as 1000ppm Ni and 1.4% Cr2O3.
Magnetite is a cumulus phase in the upper parts of MCUs IB, III and IV, in the TZ where it occurs in oxide-rich layers, and in the leuconorites and/or leucotroctolites near the bases of MCUs II, III and IV. Its TiO2 content increases systematically from <2% in zone d of MCU III to as much as 19% (corresponding to Usp58Mt42) in the TZ. As with ilmenite, the Mn-concentrations increase (to ~0.25% MnO at the top of MCU IV) and vanadium decreases upwards through the Layered Series (from ~1.3% to 0.02% V2O3 at the top of MCU IV) (Duchesne 1972). Magnetite in leucotroctolite at the base of MCU IV contains lower V concentrations (1.0-0.75% V2O3) than the magnetite in the uppermost part of MCU III and on its reappearance in the upper part of MCU IV, possibly due to the presence in the leucotroctolite of small amounts of amphibole that has a high partition coefficient for V (Jensen et al. 1993). Ni concentrations are generally low (< 40ppm), except in the leucotroctolites at the bases of MCUs III and IV where concentrations are 900-600ppm. Nickel decreases with stratigraphic height in the leucotroctolite at the base of MCU IV. Chromium exhibits the same behaviour as Ni, concentrations in the leucotroctolite at the base of MCU IV varying from 1.4-0.4% Cr2O3. Chromium contents in magnetite elsewhere in the BKSK are very low.
Fine to medium-grained, granular jotunites are present at several places along the steep, discordant northern margin of the BKSK. They occur along the outer margin of an up to 100m thick Marginal Series that separates cumulates belonging to MCU IA and IB and the high-grade gneisses of the metamorphic envelope (Fig. 1). The marginal jotunites are generally sparsely to markedly porphyritic and are considered to be chilled representatives of the magma that was parental to the oldest part of the Layered Series.
The jotunites are evolved basic rocks characterised by high FeOt (11.1-12.9 wt%) and TiO2, MgO in a narrow range between 3.8 and 5.0 and low CaO (5.4-6.7 wt%) (Fig. 4). They exhibit light REE-enriched chondrite-normalised rare-earth patterns with either a small positive Eu-anomaly or none at all, suggesting that previous fractionation or accumulation of plagioclase phenocrysts was very limited. Their compositions are similar to the jotunite trapped between anorthosite blocks enclosed within the plagioclase cumulates of MCU IB at Tjørn that likewise has been claimed to be representative of a parental magma (Duchesne & Hertogen 1988), and also to marginal jotunites of the Hidra Leuconorite (Duchesne et al. 1974).
Figure 4. Major-element composition
and CIPW norm of a typical chilled jotunite from the northern margin of the
Bjerkreim-Sokndal Intrusion. From Robins et al. (1997)
The marginal jotunites have features that are consistent
with a status as the parental magma for the cumulates of MCU IA and IB of the
Layered Series of the BKSK. Textures show clearly that plagioclase was the
first phase to crystallise from the jotunite magma and was followed by Ca-poor
pyroxene and Fe-Ti oxides. The jotunites are rich in TiO2 and poor
in diopside components, compatible with the early crystallisation of cumulus
ilmenite and the delayed appearance of cumulus Ca-rich pyroxene. The chemistry of
the minerals in the jotunites is also comparable with the BKSK Layered Series:
Plagioclases in the jotunites are slightly more sodic and the pyroxenes
decidedly more iron-rich than the equivalent highest-temperature minerals in
the BKSK cumulates. The presence of interstitial K-feldspar and quartz in the
marginal rocks also demonstrates that the jotunite magma had the potential to
produce a significant amount of an acid residual magma, as required by the
presence of the granitoids at the top of the BKSK Layered Series.
Dry melting experiments on a jotunite
collected from Tjørn show that such melts have plagioclase as the sole liquidus
phase to ~7kb at temperatures of 1150-1165°C
and oxygen fugacities of between FMQ-2 and FMQ-4 (Vander Auwera & Longhi
1994). Olivine, ilmenite and Ca-poor pyroxene (which crystallises together with
olivine) appear successively at lower temperature within ~55°C of the
liquidus at pressures up to ~5 kb. Allowing for the low oxygen fugacity in the
melting experiments, which stabilises olivine relative to Ca-poor pyroxene and
suppresses magnetite saturation, and the lower TiO2 of the
experimental starting material (3.5 wt. %) compared with the marginal
jotunites, that reduces ilmenite saturation, the experimental crystallisation sequence
of the jotunite at moderate pressure (5-7 kb)
(plagioclase-hypersthene/olivine-ilmenite) is reasonably similar to that in the
lower part of the BKSK Layered Series, that crystallised at 4-6kb based on the
contact-metamorphic mineral assemblages (Jansen et al. 1985).
Phase equilibria show that jotunitic magmas have compositions that reside on a thermal divide at pressures where they coexist with plagioclase and two pyroxenes and cannot be decendants of mantle-derived basalts (Longhi et al. 1999). They appear to have originated by melting of lower crustal gabbronorites. Recent investigations of the Re-Os systematics of the Rogaland Igneous Province (Schiellerup et al. 2000) also support a crustal origin.
The distribution and contact relationships of the oldest cumulates in the Bjerkreim lobe of the BKSK indicate that the magma chamber during the earliest phases of its evolution was approximately wedge shaped and relatively limited in horizontal and vertical extent. It is probable that the initial development of the magma chamber was controlled by displacements along a normal fault, space being created by more pronounced subsidence of the hanging wall beneath the floor of the embryonic intrusion than the roof rocks. Thus the cumulates forming MCU 0-IB crystallised at the base of a chamber with the form of a half graben, the deepest part of the chamber being along its steep, fault-controlled, north-eastern margin.
The high frequency of anorthositic xenoliths in the early cumulates suggests that the margins of the early magma chamber were mainly composed of rocks belonging to the Egersund-Ogna anorthosite massif while part consisted of granulite-facies quartzofeldspathic and mafic gneisses. The incorporation of large numbers of blocks of anorthosite and leuconorite suggests either that extensive stoping took place along the roof and walls of the chamber or that large numbers of xenoliths were transported into the chamber as it was filled.
During the early stages of evolution of the BKSK chamber there were at least two major episodes of magma recharge, represented by the bases of MCUs IA and IB. Each of these were followed by the crystallisation of exceptionally thick sequences of plagioclase-rich, zone a cumulates on the floor of the magma chamber. The cumulus plagioclase in MCUs IA and IB is distinctly more sodic (An46-39) and the orthopyroxene has a lower mg# than in the succeeding units. Plagioclase is commonly antiperthitic and interstitial quartz and apatite occur in some of the plagioclase-rich cumulates, features that are uncommon in the higher-temperature cumulates elsewhere in the Bjerkreim lobe of the BKSK. These observations suggest that the early, plagioclase-rich cumulates may have crystallised from lower-temperature, more-differentiated magmas than both those emplaced later in the evolution of the magma chamber and those represented by the jotunite marginal chills. Systematic stratigraphic variations in the composition of plagioclase in these rocks are not conspiquous, despite their considerable thicknesses. These relatively evolved magmas emplaced early in the development of the BKSK must have had densities sufficiently low to permit the settling of leuconorite and anorthosite xenoliths and, by inference, also plagioclase primocrysts. The thicknesses of pC and piC, the inconspiquous cryptic variation and the presence of plagioclase and rarer orthopyroxene megacrysts in MCUs IA and IB suggest that the magmas may have been emplaced with significant amounts of crystals, particularly plagioclase, in suspension.
Crystallisation of the multiphase cumulates at the top of MCU IB was interrupted by emplacement of a voluminous batch of jotunite magma from which the cumulates of MCU II were formed. This influx of magma led to elevation of the roof and substantial lateral enlargement of the chamber. Stratigraphic relations show that the edge of the chamber was displaced by >6 km to the southeast and in the southern part of the Bjerkreim lobe cumulates belonging to MCU II are separated from the floor of the intrusion by only a thin marginal zone of plagioclase-rich rocks. The new stretch of floor produced during lateral enlargement of the chamber was not planar. An elevated ridge existed in the Teksvatnet area, and the MCU II sequence and later cumulates thin markedly over this topographic feature. The ~2m thick sequence of phiC that intervenes between the low-temperature cumulates (phci±m±aC) forming the uppermost part of MCU IB and the higher-temperature zone a cumulates in the lower part of MCU II indicates that the replenishment event was not instantaneous but persisted for a period of time. The regressive stratigraphic sequence that crystallised during the prolonged influx of magma was the result of mixing of some of the resident with the inflowing magma. Subsequent to replenishment, the magma chamber was occupied by a lens of liquid residing on a floor of earlier cumulates that dipped inwards at low angles to an axial depression. Continuous cooling and fractional crystallisation of the magma led to formation of a thick sequence of pimC and phiC (MCU II). The magma did not differentiate sufficiently to re-attain saturation in magnetite, Ca-rich pyroxene or apatite before the emplacement of a further batch of magma.
The replenishment event that terminated the crystallisation of MCU II resulted in a further expansion of the magma chamber similar to that accompanying the preceeding magma influx. Judging by the degree of cryptic layering in MCU III, the increase in the depth of the magma occupying the chamber was, however, much less than during the earlier influx and the extent of expansion of the chamber appears to have been more limited. Expansion of the chamber took place by lateral wedging of 1600-3000m towards the south. The edge of the magma chamber in the southernmost part of the Bjerkreim lobe advanced further into the Helleren anorthosite massif, and during this processes numerous blocks of anorthosite and leuconorite were incorporated into the magma chamber.
The final major replenishment event took place after the resident magma had undergone a degree of fractional crystallisation sufficient to stabilise first magnetite, then Ca-rich pyroxene and apatite as liquidus minerals along the more distal (and higher) parts of the floor of the magma chamber while in the axial region of the magma chamber, magnetite-bearing noritic cumulates were still crystallising. This phase of magma influx appears to have been very voluminous and it resulted in a major lateral enlargement of the magma chamber to the south and the development of the Sokndal and Mydland lobes of the BKSK, where only the equivalents of cumulates belonging to MCU IV in the Bjerkreim lobe are represented. During the lateral migration of the edge of the magma chamber from its previous location near the present southern margin of the Bjerkreim lobe, myriads of large blocks and slabs of anorthositic rocks and quartzofeldspathic gneisses were stoped from the enlarging roof of the chamber. Whilst new, high-temperature magma flowed into the lowest part of the magma chamber, the less-dense residual magma was probably decanted southwards into the new Sokndal and Mydland lobes. This major replenishment was followed by almost continuous differentation that eventually led to the formation of mangeritic cumulates.
Duchesne
and Wilmart (1997) have proposed that the crystallisation of mangerite was
terminated by the emplacement of magmas that varied in composition from
jotunite to charnockite. They envisage that the thick uppermost unit of quartz
mangerite and charnockite in the BKSK crystallised from a viscous,
inhomogeneous mixture of residual acid magmas and externally-derived,
differentiated magma, possibly with additional admixture of anatectic melts
derived from the gneisses that formed part of the roof of the magma chamber. In
the Bjerkreim lobe of the intrusion there is, however, an uninterrupted cryptic variation from the TZ
into the overlying granitoids, suggesting continued cooling and fractional
crystallisation of residual magma. The latter was highly contaminated by
assimilation of country rock xenoliths and hybridisation with roof melts.
The BKSK magma chamber was emplaced into quartzo-feldspathic and mafic gneisses as well as massif-type anorthosites. Xenoliths of these country rocks are enclosed in the BKSK cumulates and are exceptionally abundant in places. It is unlikely that the plagioclase-saturated BKSK magmas could have assimilated anorthositic rocks or dry mafic gneisses, but extensive interaction between the magmas occupying the chamber and xenoliths of quartzo-feldspathic gneiss seems very probable. In addition, the granitoids that form the uppermost part of the BKSK may have resulted in part from partial melting of gneisses that formed part of the roof of the magma chamber, and anatectic acid magmas may have mixed with the underlying more basic magmas occupying the bulk of the chamber.
Figure 5. Cryptic variation in
plagioclase compositions and initial Sr87/Sr86 for
bulk-rock samples collected
through the Layered Series in the southern flank of the Bjerkreim
lobe (after Nielsen et al. 1996). The location of sample profile 1
is given in Fig. 1. The stratigraphic column to the left is for the
Layered Series as developed in the axial region of the lobe. Note the
pronounced cryptic regression associated with the MCU III/IV contact.
The initial 87Sr/86Sr ratios (Sr0) of cumulates in the Bjerkreim lobe vary substantially and provide robust evidence of extensive assimilation within the magma chamber. Sr0 shows a general evolution with stratigraphic height in the Bjerkreim Layered Series from 0.705 in MCU II to 0.7086 in the upper part of MCU IV (zone e) (Fig. 5). The trend of increasing Sr0 is interrupted by regressions to values as low as 0.7048 associated with the lower boundaries of MCU III and IV. With the exception of the uppermost part of the Layered Series there is a remarkable antipathetic relation between the cryptic variation as defined by the composition of cumulus minerals (An%, mg# in pyroxenes) and Sr0.
There is relatively little isotopic data for the granitoids that form the uppermost part of the BKSK, and certain aspects of it indicates disturbance of the Rb-Sr system. Wielens et al. (1980) calculated an isochron that gave an initial Sr isotope ratio of 0.7075±0.0028 on the basis of some of the data for these rocks reported by Versteeve (1975). Demaiffe et al. (1986) estimated Sr0 for the quartz mangerite to be ~0.7085. Both of these values coincide with that in the upper part of MCU IV, eliminating earlier isotopic arguments for a separate origin of the granitoids, and in accord with an origin as residual, highly-differentiated magmas. Currently, there is no isotopic evidence that supports an origin for all or part of the quartz mangerites and charnockites exclusively through anatexis of country-rock gneiss. The 87Sr/86Sr ratios (at 930-920Ma) for the gneisses in the vicinty of the BKSK are extremely variable, but generally higher than that of the quartz mangerite (e.g. 0.7196 for gneisses in Gydalen, a short distance to the east of the margin of the BKSK (Versteeve 1975)).
The variation in Sr0 in the Bjerkreim Layered Series is open to interpretation in several different ways. Assimilation of country rocks (either in situ or as xenoliths) or incorporation of anatectic melts may have taken place continually during the cooling and fractional crystallisation of magma that was well-mixed on a chamber scale. Alternatively, the degree of contamination may have increased towards the chamber roof in a stratified magma, due to assimilation of increasing numbers of buoyant xenoliths in the upper parts of the magma chamber or variable amounts of physical mixing with a separate, low-density, anatectic roof melt. Additionally, a contaminated isotopic signature may have diffused from the roof downward through a stratified magma column (see below). In view of the evidence in the BKSK for magma stratification during at least part of the evolution of the chamber, it would seem likely that some or all of these processes were operating during the crystallisation of the cumulates at the base of the magma chamber. Assuming simultaneous assimilation and fractional crystallisation (AFC), the available Sr isotope data are consistent with a ratio between the rates of assimilation and fractional crystallisation of ~0.2 (Tegner et al. 2000).
Several features of the stratigraphic organisation of the cumulates in the Bjerkreim lobe of the BKSK indicate that the magma from which they crystallised was compositionally stratified and that the density stratification was stable over substantial periods of time.
Detailed mapping of phase contacts has demonstrated that the zone a cumulates at the base of MCUs IB and II thin dramatically as they are traced from the north-eastern margin of the intrusion to the south-west, and both zones eventually pinch out within ilmenite norites. These plagioclase-rich cumulates must be contemporaneous with the lower-temperature cumulates into which they pass along the strike of the modal layering towards the margins of the intrusion. The zone a cumulates at the base of MCUs III and IV exhibit a similar geometrical relationship: They are thickest in the axial region of the intrusion but thin towards both margins, and wedge out completely to the east towards the Teksvatnet ridge. The converse relationship is apparent in the uppermost part of MCU III, where the lowest-temperature cumulates (phcimaC) occur in the more marginal parts of the unit but are absent in the axial region of the Bjerkreim lobe where the basal cumulates of MCU IV rest on higher-temperature phimC. These stratigraphic relationships suggest gradients in magma composition and temperature across the floor of the chamber during fractional crystallisation. We suggest that the magma occupying the chamber was stably stratified with density and temperature decreasing and the degree of magma differentiation increasing upwards through the column of magma, while the cumulate-melt interface was sloping, either uniformly towards the north-east margin of the chamber (during the crystallisation of MCU IB) or generally inwards towards the axis of the chamber (during the crystallisation of MCU III). An exception to this generalisation is the ridge in the chamber floor represented near Teksevatnet that existed during the crystallisation of MCU II and later units. The cumulates of MCU II-IV thin over this feature and certain zones wedge out eastwards towards it (e.g. zone IIIb). Evidently, at the stage in the accumulation of the Layered Series represented by MCUs II and III, higher-temperature cumulates were forming in the adjacent basins than on the ridge itself. During the crystallisation of MCU IV, the Teksevatnet topographic high appears to have been largely eliminated.
The angles of 2-15° that exist between certain of the phase contacts and the boundaries between the megacyclic units in the Bjerkreim lobe provide a reasonable minimum estimate of the original slopes of the temporary floor of the magma chamber beneath a stratified magma. The fact that discordances between cryptic and modal layering exist both near the base and the top of MCU III suggest that magma stratification was extremely stable. It persisted for at least as long a period of time as the up to 1050m-thick megacyclic unit took to crystallise.
Whether the column of stratified magma in the BKSK chamber was divided into horizontal liquid layers separated by diffusive interfaces is not clear from field relations. A continuously-stratified magma excludes convective mixing. There is, however, clear evidence within the cumulates of the BKSK for convection during their crystallisation, including cross-lamination, erosional unconformities and troughs. We therefore suggest that the magma was discontinuously stratified, and consisted of horizontal, independantly-convecting and internally homogeneous liquid layers separated by relatively-sharp, diffusive interfaces. Several processes that appear to have operated in the BKSK magma chamber could have led to the development of stratification: Entrainment of resident magma and hybridisation in turbulent fountains during the emplacement of new, less-differentiated and denser magma (Campbell 1996); Repeated emplacement of hot, dense magma along the floor of the chamber, with little mixing with the overlying resident magma (Huppert & Sparks 1980); Varying degrees of mixing between the resident magma and anatectic melts generated along the inclined walls and roof of the magma chamber or assimilation of varying amounts of buoyant wall-rock xenoliths (Campbell & Turner 1987); Compositional convection driven by density differences arising from crystallisation along inclined surfaces (McBirney et al. 1985).
The cryptic variation in mineral chemistry and particularly in initial Sr, Nd and Pb isotope ratios exhibited by a section through the sequence of zone a cumulates above the base of MCU IV (see Figs. 7 & 8 in the description of excursion localities) suggests that the final influx of magma into the Bjerkreim-Sokndal chamber was prolonged and associated with the elevation of the differentiated, contaminated and compositionally-zoned column of resident magma (see Fig. 9 in the description of excursion localities) as well as hybridisation of the inflowing and resident magmas (Jensen et al. 1993, Barling et al. 2000). This resulted in a modal regression from phcmiaC to phiC/piC and culminated in the crystallisation of high-temperature, plagioclase (An53)-olivine (Fo74) cumulates. The modal regression is accompanied by a reverse cryptic variation in mineral compositions and a systematic variation in initial isotopic ratios (e.g. a steady upward decrease in 87Sr/86Sr from 0.7061 to 0.7048), demonstrating that the cumulates crystallised from hybrid magmas with an increasing proportion of the inflowing, more primitive jotunite. Recharge of the magma chamber took place after prolonged fractional crystallisation of magnetite and consequent decrease in the density of the resident magma. Hybridisation is envisaged as taking place in a turbulent fountain with the efficiency of hybridisation of the inflowing magma with the less-dense resident magma decreasing with time.
In contrast, differentiation of the resident magma was arrested at a relatively early stage by the influx of magma marked by the MCU II/III contact. MCU II consists exclusively of a thin basal sequence of plagioclase cumulates and a thick series of phiC. Cumulus magnetite does not make an appearance in MCU II, and it is likely that the resident magma was differentiating with increasing density during its crystallisation. The MCU II/III boundary is characterised by a sulphide-enriched subzone associated with a discontinuous layer of orthopyroxenite or mafic ilmenite norite. This is succeeded by a general regression in mineral compositions that culminates in the zone IIIb troctolitic cumulates in the central and western part of the Bjerkreim lobe. Zone b cumulates are, however, absent in the eastern flank. The stratigraphic relations appear to be consistent with prolonged magma-chamber recharge associated with progressive mixing of the inflowing jotunitic magma and the resident, stratified magma whose basal portion was more dense than the replenishing magma. The sulphide-enriched orthopyroxenite and related melanocratic ilmenite norite that represent the initial response to the replenishment event are explained by crystallisation of hybrid magmas residing in the pyroxene phase volume. Their chamber-wide distribution is inferred to result from mixing taking place some distance above the floor at a level where the plume formed by the inflowing magma reached a level of neutral buoyancy in the compositionally-stratified magma column and then spread laterally throughout the chamber (Fig. 6). As the influx proceeded the resident magma was stripped from the base of the chamber and mixed into the ascending plume as the hybrid layer increased in thickness and became compositionally stratified. Eventually the lower boundary of the hybrid layer reached the floor of the magma chamber. The highest-temperature cumulates (poC, zone IIIb) crystallised from the lowest part of this hybrid layer and were restricted to the central trough on the chamber floor (corresponding to the axial region of the intrusion), while lower-temperature cumulates crystallised simultaneously on the eastern «shelf» from magma higher up in the hybrid layer.
Figure 6. The sequence of events
during the influx of magma reflected by zones IIIa and IIIb depicted in
schematic W-E sections: A) Magma flowed in as a turbulent plume and the
hybrid spread laterally some distance above the floor of the chamber; B)
The hybrid layer thickened and stratified and its lower boundary sank as
resident, denser magma beneath was mixed into the plume. Crystallisation
in the hybrid layer resulted in the sulphide-bearing orthopyroxenite and
equivalents. C) The stratified hybrid layer reached the floor of the
chamber. Troctolite (zone b) crystallised in the axial trough while more
evolved cumulates formed on the eastern shelf.
Barling, J., Weiss, D. & Demaiffe, D. 2000: A Sr-, Nd- and Pb-isotopic investigation of the transition between two megacyclic units of the Bjerkreim-Sokndal layered intrusion, south Norway. Chemical Geology 165, 47-65.
Bolle, O., Diot, H. & Duchesne, J. C. 2000: Magnetic fabric and deformation
in charnockitic igneous rocks of the Bjerkreim-Sokndal layered intrusion
(Rogaland, Southwest Norway). Journal of Structural Geology 22, 647-667.
Campbell, I. H. 1996: Fluid dynamic processes in basaltic magma chambers. In Cawthorn, R. G. (ed.): Layering in Igneous Rocks. Elsevier. pp. 45-76.
Campbell, I. H. & Turner, J. S. 1987: A laboratory investigation of assimilation at the top of a basaltic magma chamber. Jour. Geology 95, 155-172.
Demaiffe, D. Weis, D. Michot, J. & Duchesne, J.-C. 1986: Isotopic constraints on the genesis of the Rogaland Anorthositic suite (soutwest Norway). Chemical Geology 57, 167-179.
Deemer, S. & Hurich, C. 1997: Seismic image of the basal portion of the Bjerkreim-Sokndal intrusion. Geology 25, 1107-1110.
Duchesne, J.-C. 1970: Sur la provenance de xénolithes anorthositiques dans le massif de Bjerkem-Sogndal (Norvège). Ann. Soc. Géol. Belg. 93, 643-656.
Duchesne, J.-C. 1972: Iron-Titanium oxide minerals in the Bjerkreim-Sogndal massif, South-western Norway. Journal of Petrology 13, 57-81.
Duchesne, J.-C. 1987: The Bjerkreim-Sokndal massif. In Maijer, C. & Padget, P. (eds.): The geology of Southernmost Norway. Norges geol. unders. Spec. Pub. 1, 56-59.
Duchesne, J.-C. & Hertogen, J. 1988: Le magma parental du lopolithe de Bkjerkeim-Sokndal (Norvège méridionale). C.R. Acad. Sci. Paris 90,45-48.
Duchesne, J.-C. & Michot , J. 1987: The Rogaland intrusive masses: Introduction. In Maijer C & Padget P (eds.): The geology of Southernmost Norway. Norges geol. unders. Spec. Pub. 1, 48-50.
Duchesne, J.-C. & Wilmart, E. 1997: Igneous Charnockites and related rocks from the Bjerkreim-Sokndal Layered Intrusion (southwest Norway): a jotunite (hypersthene monzodiorite)-derived A-type granitoid suite. Journal of Petrology 38, 337-369.
Duchesne, J.-C., Denoiseux, B. & Hertogen, J. 1987: The norite-mangerite relationships in the Bjerkreim-Sokndal layered lopolith (S.W. Norway). Lithos 20, 1-17.
Duchesne, J.-C., Schärer, U. & Wilmart, E. 1993: A 10Ma period of emplacement for the Rogaland anorthosites, Norway, evidence from U.Pb ages. Terra Nova (Terra Abstracts) 5, 64.
Duchesne, J.-C., Wilmart, E., Demaiffe, D. & Hertogen, J. 1989: Monzonorites from Rogaland (Southwest Norway): a series of rocks coeval but not comagmatic with massif-type anorthosites. Precambrian Research 7, 111-128.
Duchesne, J.-C., Roelandts, I., Demaiffe, D., Hertogen, J., Gijbels, R. & De Winter, J. 1974: Rare-earth data on monzonoritic rocks related to anorthosites and their bearing on the nature of the parental magma of the anorthositic series. Earth Planet Sci. Lett. 24, 325-335.
Huppert, H. E. & Sparks, R. S. J. 1980: The fluid dynamics of a basaltic magma chamber replenished by influx of hot, dense, ultrabasic magma. Contrib. Mineral. Petrol. 75, 279-289.
Jansen, J. B. H., Blok, R. J. P., Bos, A. & Scheelings, M. 1985: Geothermometry and geobarometry in Rogaland and preliminary results from the Bamble area, S. Norway. In Tobi, A. C. & Touret, J. L. R. (eds.) The Deep Proterozoic Crust in the North Atlantic Provinces. NATO ASI Ser. C 158, D. Reidel Publishing Co., Dordrecht, 499-516.
Jensen, J. C., Nielsen, F. M., Duchesne, J.-C., Demaiffe, D. & Wilson, J. R. 1993: Magma influx and mixing in the Bjerkreim‑Sokndal layered intrusion, South Norway: evidence from the boundary between two macrocyclic units at Storeknuten. Lithos 29, 311-325.
Krause, H., Gierth, E. & Schott, W. 1985: Ti-Fe deposits in the South Rogaland igneous complex with special reference to the Åna-Sira anorthosite massif. Norges geol. unders. Bull. 402, 25-37.
Longhi, J.,
Vander Auwera, J., Fram, M. S. & Duchesne, J.-C. 1999: Some phase equilibrium
constraints on the origin of Proterozoic (massif) anorthosites and related
rocks. Journal of Petrology 40,
339-362.
McBirney, A. R., Baker, B. H. & Nilson, R. H. 1985: Liquid fractionation. Part I: basic principles and experimental simulations. Jour. Volc. and Geoth. Res. 24, 1-24.
Michot, J. 1961: The
anorthositic complex of Haaland-Helleren. Norsk geol. tidsskrift 41, 157-172.
Michot, J. & Michot, P. 1970: The problem of the anorthosites. The South Rogaland igneous
complex (South Western Norway). In:Isachsen YW (ed.) Origin of anorthosites and related rocks. New York State Mus. Sci.
Ser. Mem. 18, 399-410.
Michot, P. 1960: La géologie de la catazone: le problème des anorthosites, la palingenèse basique et la tectonique catazonale dans le Rogaland meridionale (Norvège méridionale). Norges geol. unders. 212, 1-54.
Michot, P. 1965: Le magma plagioclasique. Geol. Rundschau 54, 956-976.
Nielsen, F.
M., Campbell, I. H., McCulloch, M. & Wilson, J. M. 1996: A Strontium isotopic investigation of the Bjerkreim‑Sokndal
Layered Intrusion, southwest Norway. Jour.
Petrology 37, 171-193.
Nielsen, F. M. & Wilson, J. R. 1991: Crystallization processes in the Bjerkreim‑Sokndal layered intrusion, south Norway: evidence from the boundary between two macrcyclic units. Contrib. Mineral. Petrol. 107, 403‑414.
Paludan, J.,
Hansen, U. B. & Olesen, N. Ø. 1994: Structural
evolution of the Precambrian Bjerkreim-Sokndal Intrusion, South Norway. Norsk
geol. tidsskrift. 74, 185-198.
Poorter, R. P. E. 1972: Palaeomagnetism of the Rogaland Precambrian (Southwestern Norway). Phys. Earth Planet. Interiors 5, 167-176.
Rietmeijer, F. J. M. 1979: Pyroxenes from iron-rich igneous rocks in Rogaland, S.W. Norway. Geol. Ultraiectina 21, 341pp.
Robins, B., Tumyr, O., Tysseland, M. & Garmann, L.B. 1997: The Bjerkreim-Sokndal Layered Intrusion, Rogaland, SW Norway: Evidence from marginal rocks for a jotunite parent magma. Lithos 39, 121-133.
Schärer, U., Wilmart, E. & Duchesne, J.-C. 1996: The short duration and anorogenic character of anorthosite magmatism: U-Pb dating of the Rogaland complex, Norway. Earth and Planetary Science Letters 139, 335-350.
Schiellerup, H., Lambert, D. D., Prestvik, T., Robins,
B., McBride, J. S. & Larsen, R. B. 2000: Re-Os isotopic evidence for a lower crustal
origin of massif-type anorthosites. Nature 405, 781-784.
Smithson, S. B. & Ramberg, I. B. 1979: Gravity interpretation of the Egersund anorthosite complex, Noway: Its petrological and geothermal significance. Geological Society of America Bulletin 90, 199-204.
Tegner, C., Meyer, G. B., Schiellerup, H., Robins, B., & Wilson, J. R. 2000: Crustal assimilation and fractional crystallisation (AFC) of basalt: r-values from Norwegian layered intrusions, . AGU Fall Meeting, San
Fransisco dec. 15-19, American Geophysical Union Eos Tranactions 81, p. 1357
Vander Auwera, J. & Longhi, J. 1994: Experimental study of a jotunite (hypersthene monzodiorite):
constraints on the parent magma composition and crystallisation conditions (P,
T, fO2) of the Bjerkreim-Sokndal layered intrusion (Norway). Contrib. Mineral. Petrol. 118, 60-78.
Vander Auwera, J. & Longhi, J. 1994: Experimental study of a jotunite (hypersthene monzodiorite):
constraints on the parent magma composition and crystallisation conditions (P,
T, fO2) of the Bjerkreim-Sokndal layered intrusion (Norway). Contrib. Mineral. Petrol. 118, 60-78.
Versteeve, A. J.
1975: Isotope geochronology in the
high-grade metamorphic Precambrian of southwestern Norway. Norges
geol. unders. 318, 1-50.
Weilens, J. B. W., Andriessen, P. A. M., Boelrijk, N. A. I. M., Hebeda, E. H., Priem, H. N. A., Verdurmen, E. A. Th. & Verschure, R. H. 1980: Isotope geochronology in the high-grade metamorphic Precambrian of southwestern Norway: New data and interpretations. Norges geol. unders. 359, 1-30.
Wilmart, E., Demaiffe, D. & Duchesne, J. C. 1989: Geochemical constaints on the genesis of the Tellnes ilmenite deposit, Southwest Norway. Econ. Geol. 84, 1047-1056.
Wilson, J. R. & Sørensen, H. S. 1996: The Fongen-Hyllingen Layered Intrusive Complex, Norway. In Cawthorn, R. G. (ed.): Layering in Igneous Rocks. Elsevier. pp.
303-329
Wilson, J. R., Robins, B., Nielsen, F. M., Duchesne, J.-C. & Vander Auwera, J. 1996: The Berkreim-Sokndal Layered Intrusion, southwest Norway. In Cawthorn, R. G. (ed.): Layering in Igneous Rocks. Elsevier. pp. 231-255.