View(s): 412 (4 ULg)
Download(s): 0 (0 ULg)
JUNIOR MEETING 27.9.2001
1Stratigraphy and chitinozoans of the Tier d'Olne hill (Ombret, Condroz Inlier): part of the Brabant Basin?
3Universiteit Gent, Laboratorium voor Paleontologie,
4Krijgslaan 281/S8, B-9000 Gent
5The Tier d'Olne hill (also known as the "colline 118") is situated along the Meuse river just north of the village of Ombret, 8 km east of Huy. It is situated in the eastern part of the Condroz Inlier, between the Dinant Synclinorium in the SE and the Namur Synclinorium in the NW.
6New fieldwork allowed detailed description of the lithostratigraphy of the rocks of the Tier d'Olne hill and sampling for palynology and sedimentology. These rocks were formerly described by Martin et al. (1970) as the Ombret Formation, consisting of a lower unit of black shale (more than 50 m in thickness) with rare, always thin micaceous sandstone intercalations and an upper sandstone-slate member of turbiditic nature with obliquely stratified quartzitic sandstone up to several dm thick (about 300 m in thickness) (Verniers et al., in press). Based on acritarchs and chitinozoans, Martin et al. (1970) deduced a Late Ordovician age for the formation, probably Caradoc. Paleontological studies by Hance et al. (1991) also evidenced a Late Ordovician age for the Ombret Formation.
7We separated the lower unit as a unit on its own: the Tier d'Olne Formation, consisting of dark grey slate, with few intercalations of sometimes cross-laminated micaceous coarse siltstone to fine sandstone beds, with a few to maximum 6 cm thickness. In section, the slate bears an alternation of millimetric siltstone and mudstone laminae, separated by diffuse boundaries, probably due to bioturbation. Bioturbational siltstone lenses (± 1mm diameter) occur in the mudstone laminae. Some levels contain fine-grained pyrite incorporated in the siltstone laminae. The minimal thickness of the formation is estimated at 208 m. The formation contains mostly flattened, moderately preserved and often broken chitinozoans. At the base, we can define a local biozone (OMB-1, in 2 samples) by the presence of Belonechitinal sp. A, yet undated. The biozone OMB-2 is characterized by the presence of Ancyrochitina spp., found by Martin et al. (1970). Part of this biozone (subbiozone OMB-2a) contains beautifully preserved specimens of Calpichitina lenticularis, together with Desmochitina cocca, Laufeldochitina robusta, Cyathochitina calixl and Cyathochitina campanulaeformis, not observed by Martin et al. (1970). The chitinozoans indicate an Abberreidian, Llandeilian, Aurelucian and possibly early Burrellian age. The biozone OMB-3 contains Spinachitina bulmani?, Conochitina minnesotensis, Rhabdochitinal cf. gallica, together with Desmochitina cocca and Cyathochitina campanulaeformis, dated as Aurelucian to Burrellian (see Figure 1).
8The litho- and biostratigraphy of the Tier d'Olne Formation are dissimilar to those of the Huy and Sart-Bernard Formations of the central area of the Condroz Inlier, but the formation bears good lithostratigraphical resemblances to the Brabant Massif Rigenée Formation. The chitinozoans of the Tier d'Olne Formation are however younger than the assemblages in Brabant Massif biozone 4 in the lower part of the Rigenée Formation (the upper part of the formation is devoid of chitinozoans) and older than the Brabant Massif biozone 5 (Samuelsson & Verniers, 2000).
9About 20 volcano-sedimentary layers occur at the base of the Ombret Formation in a 12 m thick zone. The thickness reaches 11 cm maximum. They are intercalated between turbidites. Higher, the formation is made up of a repetition of light-grey, cross-laminated, micaceous fine sandstone beds, medium grey, parallel laminated siltstone beds and dark-grey, compact laminated mudstone beds. These beds correspond respectively to the Tc-, Td- and Te-intervals of a Bouma turbidite sequence. In the lower part of the Ombret Formation, the Tc-intervals are thin (often less than 10 cm) and the turbidites are interpreted as low-energetic. The sequence contains generally Te-intervals, with many pyrite-like bands and bioturbation features. In the upper part of the Formation, Tc-intervals are thicker (up to around 40 cm) and can show convolute bedding, load casts, current structures and groove casts; in general, Te-intervals are relatively thin and can contain bioturbation features. The minimal thickness is estimated at 328m, but is probably higher, as the displacement of an important inverse fault is not known (Valcke, unpublished Master thesis, RUG, 2001).
10The major part of the Ombret Formation contains the biozone OMB-4; it is characterized by the presence of Belonechitina robusta, Conochitina minnesotensis, Rhabdochitinal cf. gallica, Spinachitina bulmanil, Cyathochitina campanulaeformis and Desmochitina cocca. It corresponds with the Brabant Massif biozone 5 in the Ittre Formation, with a Burrellian age (Samuelsson & Verniers, 2000). The biozone OMB-4a (at the base of biozone 4) yields next to the species of OMB-4 also Desmochitina erinacea. Biozone OMB-5, at the top of the Ombret Formation, contains the same assemblage as
11biozone OMB-4, but Belonechitina robusta is lacking in the OMB-5 biozone (see Figure 1).
12Two crinoid-stems are found in a Te- interval in the upper part of the Ombret Formation. It is the first macrofossil found in the Ombret Formation, and also the first crinoid in an Ordovician turbiditic facies of Belgium.
13Only limited litho- and biostratigraphical similarities are found between the Ombret Formation and the Upper Ordovician formations of the other areas of the Condroz inlier. In contrast, good lithostratigraphical resemblances exist with the Brabant Massif Ittre Formation and in a lesser amount to the Bornival Formation. Further study on the volcano-sedimentary layers needs to investigate the relation of the tuffs at the base of the Ombret Formation with the tuffs at the base of the Ittre Formation (Corin, 1964).
14We interprète that the Tier d'Olne Formation was deposited in a (deep) shelf environment and the Ombret Formation in a deep environment where turbidity currents were active. At the start of the deposition of the Ombret Formation, there was volcanic activity in some vicinity of the depositional basin, evidenced by the volcano-sedimentary layers at the base of the of the formation and by the presence of volcanic particles at the base of some Tc-intervals (Martin et al., 1970). The tectonic setting of this volcanism requires further investigation.
15Taking into account the litho- and biostratigraphical arguments mentioned above, we conclude that the Tier d'Olne and the Ombret Formations are probably deposited in the same sedimentary basin as the Rigenée and the Ittre Formations, named the Brabant Basin in Verniers et ai, 2001, and therefore lateral equivalents.
17CORIN, F, 1963. Sur les roches éruptives de la tranchée d'Hasquempont, canal de Charleroi. Bulletin de la Société belge de Géologie, 72, 94-98.
18HANCE, L., STEEMANS, P., GOEMAERE, E., SOMERS, Y., VANDENVEN, G., VANGUESTAINE & VERNIERS, J., 1991. Nouvelles données sur la Bande de Sambre-et-Meuse à Ombret (Huy, Belgique). Annales de la Société Géologique de la Belgique 114, pp. 253-364. MARTIN, F, MICHOT, P. & VANGUESTAINE, M., 1970. Le flysch Caradocien de Ombret. Annales de la Société géologique de Belgique, 93, 337-362.
19SAMUELSSON, J & VERNIERS, J., 2000. Ordovician chitinozoan biozonation of the Brabant Massif, Belgium. Review of Palaeobotany and Palynology 113, 105-129. VALCKE, S., 2001. Structurele opbouw van de noordrand van de Condrozstrook te Ombret. Unpublished Master Thesis, Ghent University. VERNIERS, J., HERBOSCH, A., VANGUESTAINE, M., GEUKENS, F, DELCAMBRE, B., PINGOT, J.L., BELANGER, I., HENNEBERT, M., DEBACKER, T, SINTUBIN, M. & DE VOS, W., 2001. Cambrian-Ordovician-Silurian lithostratigraphic units (Belgium). In Bultynck. & Dejonghe, eds., Guide to a revised lithostratigraphic scale of Belgium. Geologica Belgica, 4/1-2: 5-38.
20Structural analysis of the northern part of the Condroz Inlier at Ombret
22Laboratorium voor Paleontologie, Universiteit Gent
23The Condroz Inlier is a thin band of Silurian and Ordovician rocks, in the north unconformably overlain by the Middle Devonian of the Namur Synclinorium and in the south by the Lower Devonian of the Dinant Synclinorium. The Condroz Inlier has been subdivided in three subareas: a northern part, at Ombret, with cleavage; a central part, the main part of the Condroz Inlier, without cleavage; and a southern part, at Puagne and the Fond d'Oxhe, with cleavage (Graulich, 1961; Michot, 1979). One of the main problems of the Condroz Inlier is the difference in geological history of these three subareas. The Lower Palaeozoic of the Condroz Inlier is usually described in relation to the Lower Palaeozoic elsewhere in Belgium. This is either the Lower Palaeozoic of the Caledonian deformed Brabant Massif, north of the Condroz Inlier, or the Lower Palaeozoic of the Caledonian and Variscan deformed Ardenne massifs, south of the Condroz Inlier. The relationship can be established based on sedimentology, litho- and biostratigraphy or based on the timing of the deformations, Caledonian or Variscan. For such an investigation detailed structural studies are needed in the different parts of the Condroz Inlier. The present study concentrates on the northern part of the Condroz Inlier at Ombret.
24In a sedimentological, litho- and biostratigraphical study, Vanmeirhaeghe (2001 ) has described both formations of the Ombret area and dated them with Chitinozoa. The oldest of the two, the Tier d'Olne Formation, has an age of lowermost Caradoc or uppermost Llanvirn and consists of clayey to sandy shales. The younger Ombret Formation has a Burellian age and consists of turbidite deposits. Our study focuses on the structural elements of the same area at Ombret.
25Field work and the statistical and geometrical analysis of the structural observations result in two detailed cross-sections, one along a 200 m long outcrop of the Ombret Formation at the foot of the hill 'Tier d'Olne', and another, more south, along an outcrop of the Tier d'Olne Formation. The overall structure is an overturned limb, consisting of moderate to steeply south dipping layers, younging to the north. This large overturned limb encloses an antiform and a synform, separated by a region of bedding with normal polarity and cut by two reverse faults. The siliciclastic rocks of the Ombret and the Tier d'Olne Formations show a uniformly south dipping cleavage. This cleavage is important, as it is unique for the area around Ombret in comparison to the surrounding Condroz Inlier, where it is virtually absent. The cleavage is well defined and ideal to describe its relation with the other structures. In the Ombret formation the cleavage is developed as a pencil structure in the homogeneous, more coarsely grained layers and exhibits a moderate to well developed planar structure in the layers with the finest grain size. In the Tier d'Olne Formation the cleavage planes are well developed and closely spaced.
26A relative chronology of the deformations at Ombret could be established. Firstly, there is evidence for a pre-cleavage folding in the Ombret Formation. The folding can have taken place before the lithification, e.g. by slumping, or after the lithification, in an early tectonic stage. Secondly, a syn-cleavage deformation consists of folds with axial plane cleavage in both formations, and also of growth fibres and bedding-parallel quartz veins in the coarse grained layers of the Ombret Formation. Thirdly, a post-cleavage deformation consists of kink bands, faults and fault related quartz in both the Ombret and the Tier d'Olne Formation.
27In order to determine an age for the deformations in Ombret our data need to be placed in a regional context. In Ombret the overall trend of the syn-cleavage structures as well as of the post-cleavage structures, is NE/ SW. This fits with the trend of the Caledonian and Variscan deformations in the eastern Condroz Inlier. It implies that one cannot distinguish Caledonian of Variscan structures by looking only at the trend. Moreover, the tectonic history at Ombret can include reversing, in which formerly south vergent, Caledonian structures are overturned during the Variscan orogen and become north vergent, like the Variscan structures. As the deformations at Ombret have been related chronologically to the cleavage, the age of the cleavage can be used deduce an age for the other deformations. The geometry of the cleavage on its own cannot give an age. The southwards dip can be either a primary Variscan dip or a Caledonian dip, tilted northwards. However, if one takes into account that cleavage needs a rather high temperature to form (Kisch, 1991 ), the well developed cleavage in the Ombret and the Tier d'Olne Formations can be related to the anchizonal metamorphism observed in both formations (Hance et al., 1991). If we can determine the age of the metamorphism, we also have the age of the cleavage. According to literature, the stratigraphical-regional distribution of the metamorphism near Ombret and elsewhere in Belgium (Fielitz and Mansy, 1999; Hance et al., 1991; Steemans, 1994) can learn us more about the age of metamorphism and cleavage at Ombret. We noticed e.g. a structural continuity between the reverse structure of the Llanvirn-Caradoc at Ombret and the reverse structure of the Middle-Upper Devonian of the Namur Synclinorium just north of the Ombret area, which is in strong contrast with the observed discontinuity in the degree of metamorphism. The Llanvirn-Caradoc of the Ombret area shows an anchizonal degree of metamorphism and a well
28developed cleavage while the Middle-Upper Devonian is diagenetic and shows no cleavage. If we consider the Variscan metamorphism near the Variscan front zone as syn-orogenic, we would expect a gradual change in the degree of metamorphism. No important fault is observed between both, as the structure is a continuous inverted limb, and hence another explanation is proposed for the metamorphic anomaly at Ombret, i.e. a pre-Variscan, deformation age for the metamorphism and consequently for the cleavage, very likely Caledonian.
29The cross-section through the eastern Condroz Inlier, deduced in this study, is based on the foregoing and on the observation that the Ombret and the Tier d'Olne formations show a remarkable resemblance respectively with the Ittre and the Rigenée Formations of the Brabant Massif. It is supposed that all four formations were probably deposited in the same Brabant Basin (Vanmeirhaeghe, 2001 ). Based on the Caledonian age of the metamorphism and of the syn-cleavage deformations, we conclude that the northern part of the Condroz Inlier very probably forms part of the Brabant fold belt. Figure 1A shows a schematic cross-section of the southern Brabant Massif after the Caledonian and before the Variscan deformation. The dark grey colour indicates the Lower Palaeozoic rocks with a Caledonian cleavage (Brabant tectofacies, comp. Michot, 1979). In the southern part of the Brabant Massif, this cleavage is generally dipping northwards and the folds are south vergent (figure 1 A) (Debacker, 2001 ). In the part coloured light grey on figure 1A, there is no development of a Caledonian cleavage (Condroz tectofacies, comp. Michot, 1979). After the Variscan deformation, the inverted limb of the southern Namur Synclinorium was formed. Figure .B shows a detail of figure A after the Variscan orogen. On this hypothetical, schematic cross-section through the eastern Condroz Inlier one can see that, as a consequence of the Variscan orogen, the north dipping cleavage of the southern Brabant Massif becomes a south dipping cleavage, and the south vergent folds turn into north vergent folds, as is observed at Ombret. Furthermore during the Variscan deformation, two reverse faults south of the Tier d'Olne Formation were formed. These are the Faille d'Ombret (1), earlier defined by Michot (1969) and the Faille d'Ombret (2) introduced in this study. The latter is suggested by the stratigraphical gap over a distance less than 50 m between the Lower Ordovician (at Ombret), becoming younger to the north and the Ashgill (south of the Lower Ordovician), becoming younger to the south. An additional argument for the fault is the contrast in cleavage development and metamorphism. Both faults have brought the central Condroz Inlier (diagenetic and without cleavage), the southern Condroz Inlier and the northern Dinant Allochton (lower anchizonal and with a moderately developed cleavage) closer to the northern Condroz Inlier at Ombret (anchizonal and with a well developed cleavage).
30From the foregoing we have set up a hypothesis for the age of the deformations at Ombret. The pre-cleavage fold can be related to the unstable conditions during the sedimentation of the Ombret Formation, that is Caradoc (Vanmeirhaeghe, 2001). These unstable conditions during the Caradoc are also considered as the probable cause of turbidite deposits and slumping in the Ittre Formation, the equivalent of the Ombret formation in the Brabant Massif (Debacker et al, 2001). The syn-cleavage deformation and metamorphism have a Caledonian age, that is late Early Devonian (Mansy et ai, 1999; Debacker, 2001 ), and the post-cleavage deformation has a Variscan age, that is Late Carboniferous (Mansy et ai, 1999).
32DEBACKER, T.N. (2001 ). Palaeozoic deformation of the Brabant Massif within eastern Avalonia: how, when and why? (Ph.D. Thesis, Laboratorium voor Paleontologie, Universiteit Gent.).
33DEBACKER, T.N., SINTUBIN, M. & VERNIERS, J. (2001 ). Large-scale slumping deduced from structural and sedimentary features in the Lower Palaeozoic Anglo-Brabant fold belt, Belgium. Journal of the Geological Society, London 158, 341-352.
34FIELITZ, W. & MANSY, J.-L. (1999). Pre- and synorogenic burial metamorphism in the Ardenne and neighbouring areas (Rhenohercynian zone, central European Variscides). Tectonophysics 309, 227-256. GRAULICH, J.M. (1961). Le sondage de Wépion. Mémoires pour servir à l'Explication des Cartes Géologiques et Minières de la Belgique 2, 1-102. HANCE, L., STEEMANS, P., GOEMAERE, E., SOMERS, Y., VANDENVEN, G., VANGUESTAINE, M. & VERNIERS, J. ( 1991 ). Nouvelles données sur la Bande de Sambre-et-Meuse à Ombret (Huy, Belgique). In: Proceedings of the international meeting on the Caledonides of the Midlands and the Brabant Massif (Brussels, 20-23 September 1989), (L. ANDRÉ, A. HERBOSCH, M. VANGUESTAINE & J. VERNIERS, eds), pp. 253-264. Annales de la Société Géologique de Belgique 114(1). KISCH, H.J. (1991). Development of slaty cleavage and degree of very-low-grade metamorphism. A review. Journal of Metamorphic Geology 9(6), 735-750. MANSY, J.L.,EVERAERTS,M.& DE VOS, W. ( 1999). Structural analysis of the adjacent Acadian and Variscan fold belt in Belgium and northern France from geophysical and geological evidence. Tectonophysics 309,99-116. MICHOT, P. (1969). La faille d'Ombret. Annales de la Société Géologique de Belgique 92, 243-254. MICHOT, P. (1979). La faille mosane et la phase hyporogénique bollandienne, d'âge emsien dans le rameau calédonien condruso-brabançon. Annales de la Société Géologique de Belgique 101, 321-335. STEEMANS, Ph. (1994). Géologie de la région de Puagne, Bande de Sambre-et-Meuse (Belgique). Hypothèses basées sur l'état de maturation des palynomorphes fossiles. Comptes Rendus de F Académie des Sciences Paris 318, Série II, 1551-1556. VANMEIRHAEGHE, J. (2001 ). Litho- en biostratigrafie met Chitinozoa en sedimentologie van het Boven-Ordovicium in de heuvel Tier d'Olne (Ombret, Condrozstrook). (M.Se. thesis, Laboratorium voor Paleontologie, Universiteit Gent.).
35Freshwater molluscs and biostrati-graphy of the rift deposits to the north of Lake Edward
37Baharak BASHAR, Achiel GAUTIER & Dirk VAN
39Research Unit Palaeontology, Dept. of Geology and
40Soil Sciences, Ghent University, Krijgslaan 281 /
41S8, B-9000 Ghent, Belgium
43This study concerns the fossil freshwater molluscs and the biostratigraphy of the rift deposits to the north of Lake Edward, situated in the northern part of the Western Rift Valley, on the Congolese side. The investigated material was collected by de Heinzelin in 1958 from outcrops in the Upper Semliki region and has remained unstudied since. This material derives from lacustrine deposits (Lusso Beds), dating to the Pliocene. The fossils are concentrated in peloidic ironstone beds, characteristic for the Lusso Beds. This study aimed to compare the recently established biostratigraphy for the area with previous work, taking the latest taxonomie revisions of the malacofauna of this region into account. This study resulted in an adjustment of the stratigraphical ranges of some species, possibly caused by regional differences in the composition of the molluscan assemblages. From the biozonal comparisons it has become evident that at the end of the deposition of the Lusso Beds an impoverishment of the endemic molluscan fauna took place. The above has led to the conclusion that the previously established biostratigraphies are still valid, but are less detailed than the biozonation incorporated in this study.
44Holocene depositional history and profile type mapping of the central part of the IJzer palaeovalley (Belgian coastal plain)
46Laboratoires associés de Géologie-Pétrologie-
47Géochimie, B20, Université de Liège
48* Present address: Laboratoire de Géologie des Argiles
49et Sédimentologie des Silicoclastiques, B18, Université
50de Liège, Sart Tilman, 4000 Liège, Belgium
51The aim of this study was to map Holocene deposits and to investigate the evolution of the sedimentary environments and the depositional history during the post-glacial transgression in a part of the Belgian coastal plain. The central part of the IJzer palaeovalley was selected and, during the summer of 2000, 78 new hand-augered gouge cores were made, completing the database of the Belgian Geological Survey.
52The method of profile type mapping worked out by Streif ( 1978) was applied. First of all a general profile type map was made, delineating the major differences in the Holocene sequence, such as: presence or absence of basal peat, areas where the sequence consists of an alternation of peat beds and mud, or entirely of clastic deposits. This map does not show yet the details of the different facies, but give a clear overview of the Holocene geological setting. The detailed or special profile type map shows the different facies of the subsoil, organised in 19 special profile types. Such a map allows a three dimensions view as well as the understanding of the relationship between the different sedimentological units.
53Because the profile type maps do not reveal any information about the thickness of the Holocene sequence, an isohypse map of the pre-Holocene deposits was made, based on numerous boreholes reaching the Holocene/ Pleistocene boundary. It clearly shows the existence of two palaeovalleys former to the Holocene transgression and a network of depressions caused by vertical erosion of the tidal channels during the Late Holocene.
54These three maps permit the understanding of the different sedimentary facies depending on the pre-Holocene morphology. Moreover, these maps, used here in a scientific aim, can have a practical use in geotechnics. Indeed, field stability in a coastal plain is directly in relation with the thickness and nature of soft surface sediments.
55Three selected boreholes have been sampled for minera-logical and grain size analyses. The results prove a constancy in the source of Holocene clay sediments and the impact of peat beds on the nature of the underlying sediments. They confirm the non-existence of different transgressive phases in the Late Holocene, a strongly discussed subject in the Belgian literature. Grain size analyses provide granulometric characterization of the sedimentary environments defined macroscopically.
56The last part deals with the Holocene depositional history of the studied area. It is based on the interpretation of the boreholes correlated in cross-sections as well as on the results from the mapping, together with radiocarbon datings. The depositional history shows that the environmental evolution is in relation with the rate of relative sea-level (RSL) rise (according to the Belgian RSL curve worked out by Denys and Baeteman, 1995), the morphology of the pre-Holocene deposits and the position of tidal channels.
57The rapid RSL rise prior to ca. 7800 cal. BP allowed the deposition of a thick clastic sediment body in the deepest part of the palaeovalleys on the basal peat. A first retardation in the rate of RSL rise at ca. 7800 cal. BP resulted in an emersion of the sedimentary environments, however, temporarily and not always at the same location, leading to the formation of intercalated peat beds in the mud. This typical alternation of peat beds and mud is not caused by regressive phases but is a function of the position of the tidal channels. The second retardation in the rate of RSL rise at ca. 6000 cal. BP caused the further infilling of the area, evolving into a freshwater swamp. The accumulation of peat lasted for about 3000 years resulting in a two meters thick peat bed. Finally as from ca. 2500 cal. BP, tidal conditions were again installed in the study area. During this time, the plain was characterized by the development of tidal channels eroding the underlying deposits and causing the deposition of a more than one meter thick mud layer all over the coastal plain while channels were filled by sand.
58As theoretically discussed by Allen (2000), sediments autocompaction is briefly described. Finally, the impact of reclamation on the uppermost sediments leading to geochemical modifications is analysed.
60Allen, J.R.L., 2000. Holocene coastal lowlands in NW Europe: autocompaction and the uncertain ground. In: Pye, K. & Allen, J.R.L. (Eds). Coastal and Estuarine Environments: sedimentology, geomorphology and archaeology. Geological Society, London, Special Publications, 175: 239-252.
61Denys, L. & Baeteman, C., 1995. Holocene evolution of relative sea level and local mean high water spring tides in Belgium - a first assessment. Marine Geology, 124: 1-19. Streif, H., 1978. A new method for the representation of sedimentary sequences in coastal regions. Proceedings 16"' Coastal Engineering Conference, ASCE, Hamburg: 1245-1256.
62Geochemistry of the cumulates from the Bjerkreim-Sokndal layered intrusion, southwest Norway
64Laboratoires associés de Géologie, Pétrologie et Géochimie, Université de Liège, B-4000 Liège, Belgium
65The Bjerkreim-Sokndal layered intrusion (BKSK) is situated in the Rogaland anorthosite Province, SW Norway. Each term of the AMC series is exposed: anorthosite, leuconorite, troctolite, norite, gabbronorite, jotunite, mangerite, quartz mangerite and charnockite. This intrusion is vertically subdivided into two parts: the lower part, containing cumulate rocks and called the " layered series " and the upper part, formed by acidic rocks. The cumulates with a thickness of ca 7000 m in the Bjerkreim lobe are subdivided in 6 megacyclic units (MCU). Each MCU's base shows more primitive compositions, resulting from a new influx of undifferentiated magma. Each MCU itself can be divided into zones (a to f) depending on the cumulus minerals association. The plagioclase is ubiquitous and is accompanied by orthopyroxene (opx) and ilmenite (Imagnetite ±olivine) in zones a to d and by opx, ilmenite, magnetite, apatite and clinopyroxene (cpx) in zones e and f.
66Recent investigations on BKSK have essentially been devoted to the identification of the parental magma, to the evolution of major elements in minerals and to iso-topic ratios. The results concluded in a detailed description of the sequence and of the conditions (P, T, fD ) of crystallization as well as of magma chamber processes, such as magma influx, hybridation and contamination by the roof.
67The method of this study is built up on whole-rock compositions. It aims at specifying the controlling factors of the compositional variations. New interpretations are based on 111 analyses of the major elements (62 coming from the Paul Michot's collection) and 44 trace elements analyses. Separated minerals analyses also complete the database.
68From the study of the cumulate major elements, we can draw the following conclusions. It is possible to distinguish two groups of rocks: cumulates without neither apatite nor cpx ("leuconorites") and cumulates with apatite and cpx ("gabbronorites"). In Harker's diagrams, each group defines a linear trend. The identification of the two poles demonstrates that the first one is made up of plagioclase and the second one is a mixing of the other minerals (opx and ilmenite for leuconorites; opx, ilmenite, magnetite, apatite and cpx for gabbronorites). We can thus conclude that, inside each group, the variation of the cumulate composition mainly results from the variation of the plagioclase proportion. This also implies that the relative proportion of the different mafic minerals (oxides, pyroxenes and apatite) remains constant during each specific stage of evolution. The main aspects of this study have been presented elsewhere (Charlier & Duchesne, 2001).
69REE patterns perfectly differentiate between the two types of cumulates. Leuconorites are poor in REE and present a large positive anomaly. The REE content of these cumulates is perfectly correlated with the P,0. content, when the rocks have no cumulus apatite. The P,05 content is related to the variable abundance of interstitial liquid. So, the richer a cumulate is in incompatible elements, the higher the proportion of interstitial liquid. In contrast, gabbronorites contain 8 to 10 times more REE and present flat normalized patterns. If we analyze the influence of gabbronoritic minerals on the REE patterns, we can see that apatite plays the dominant role on the whole-rock composition. The presence of interstitial liquid, as well as the other minerals proportions, hardly influences the REE distribution and abundance.
70Mineral analyses, which were performed to understand the REE whole-rock patterns, have made it possible to describe the REE distribution between the minerals of the cumulates. The REE partition coefficients of opx, apatite and cpx could be calculated. Moreover, these data have been completed by more plagioclase analyses to assess the potentialities of a detailed study of the REE evolution in this mineral through the layered series. First interpretations lead us to assess that the liquid became richer in REE until the appearance of the cumulus apatite. At that moment, the REE cumulate composition became almost the same as the liquid composition. As a result, the REE content of the liquid remained the same in the following differentiation. From these data, we can also notice that the positive Eu anomaly of the plagioclase keeps rising in each MCU. The role of oxygen fugacity, of the anorthite content of plagioclase and the influence of the structure of the liquid require a more detailed study to be ascertained.
71The knowledge of the trace elements composition enables us to apply the modelling proposed by Bédard ( 1994). It consists in calculating the equilibrium distribution of trace elements among the minerals of cumulate rocks, and the concentration of trace elements in the coexisting liquids. Calculated liquid compositions are similar to the jotunitic liquids proposed by Vander Auwera et al. ( 1998). Moreover, thanks to the knowledge of the BKSK's parental magma and its evolution, the modelling enables us to constrain the proportion of interstitial liquid. Even if the latter can vary, it does not seem to be larger than 10%. So the cumulates of BKSK can definitely be considered as adcumulates.
72Finally, a model for the formation of cumulate rocks is proposed and permits to clarify our interpretations, previous results and the type of modal layering observed in BKSK. An in situ crystallization of the minerals, as proposed by Vander Auwera & Longhi (1994), combined with an oscillatory nucleation of plagioclase are the main mechanisms of the cumulate formation.
74Bédard, J.H., 1994. A procedure for calculating the equilibrium distribution of trace elements among the minerals of cumulate rocks, and the concentration of trace elements in the coexisting liquids. Chemical Geology: 118: 143-153.
75Charlier, B. & Duchesne, J.C., 2001. Whole-rock geochemistry of the Bjerkreim-Sokndal layered series: bearing on crystallization processes of cumulus rocks (abstract). NGU Report GEODE field workshop 2001 on ilmenite deposits in the Rogaland anorthosite province, S Norway: 29-30.
76Vander Auwera, J. & Longhi, J., 1994. Experimental study of a jotunite (hypersthene monzodiorite): constraints on the parent magma composition and crystallisation conditions (P, T, f02) of the Bjerkreim-Sokndal layered intrusion (Norway). Contributions to Mineralogy and Petrology, 118: 60-78.
77Vander Auwera, J., Longhi, J. & Duchesne, J.C., 1998. A liquid line of descent of the jotunite (hypersthene monzodiorite) suite. Journal of Petrology, 39 : 439-468.
78Quartz veins and double-sided mullions in Lochkovian metasediments in the High-Ardenne Slate Belt (la Flèche Quarry, Bertrix, Belgium)
80a Structural Geology & Tectonics Group, Katholieke Universiteit Leuven, Redingenstraat 16, B-3000 Leuven
81b Fysico-chemische Geologie, Katholieke Universiteit Leuven, Celestijnenlaan 200C, B-3001 Heverlee
82Lower Devonian metasediments are cropping out in the La Flèche Quarry, situated in the High-Ardenne Slate Belt. They consist of a variation of sandstones and pelitic material. A large number of quartz veins are present in the sandstone layers, oriented at high angles to the bedding. The interface between sandstones and pelites is buckled in-between the veins. Such structures have been called boudins in the High-Ardenne Slate Belt, although their origin is a matter of controversy. These structures can, however, also be linked to the mullions in Dedenbom. The quarry La Flèche at Bertrix offers the unique possibility to study these structures because of the good degree of exposure. In this study, a combined structural, microthermometric and geothermometric study allows proposing a kinematic model for the development of the quartz veins and associated structures. The structural study shows that firstly the veins developed in the sandstones, secondly the interface between sandstones and pelites was buckled in-between the veins and finally the layers underwent folding and a slaty cleavage developed in the pelitic layers. The microthermometry of primary C02-N2 and H,O-CO,-N2 fluid inclusions in the quartz veins and the geothermometric study of the chlorites in the quartz veins provide the exact P-T conditions of vein development. Quartz veins developed during deep burial, early in the deformation history of the Lower Devonian metasediments by hydraulic fracturing. This was, however, only possible if there was already a compressive tectonic force, inferring that vein development occurred during the onset of the Variscan orogeny. Subsequent Variscan layer-parallel shortening caused the buckling of the interfaces of the sandstones with the pelitic material. Quartz veins acted as mechanical barriers. This resulted in the development of double-sided mullions, pinned by the pre-existing veins. Subsequently, these polyphase structures behave passively during the protracted Variscan deformation, resulting in folding and slaty cleavage development.
83Keywords : Boudin, hydraulic fracturing, Lower Devonian, mullion, quartz veins, Variscan orogeny.