The Upper Ordovician-Lower Silurian rocks in the Os area, Major Are, Western

SVEIN ERIK INGDAHL

Ingdahl, S. E.: The Upper Ordovician-Lower Silurian rocks in the Os area, Major Bergen Are, . Norsk Geologisk Tidsskrift, Vol. 69, pp. 163-175. 1989. ISSN 0029-196X.

The two younger Ashgillian-Llandoverian sedimentary sequences in the Os area, the Holdhus and the Ulven Groups, have been correlated and are now collectively named the Os Group. The group is divided into fiveformations: the Skarfjell, Vaktdal, Hegglandsdal, V alla and Moberg Formations. No structural/ metamorphic break is found between any of the formations, and the Os Group is claimed to represent a continuous sedimentary sequence related to transgression and regression in a shallow marine to beach environment.

Svein Erik lngdahl, Statoil, Postboks 1212, 5001Bergen, Norway.

The investigated area (Fig. l) is located in the lngdahl 1985). The stratigraphy of the Holdhus southwestern part of the Scandinavian Cale­ and Ulven Groups in various parts of the Os area donides and is a part of the Bergen Arcs, which has been correlated and is believed to belong to have previously been discussed several times in the Os Group (Figs. l, 2). the geological literature (Reusch 1882; Kolderup A structural/metamorphic unconformity has & Kolderup 1940; Kvale 1960; Sturt & Thon recently been postulated between the Skarfjell 1978 and refs. therein). The Bergen Are System and Vaktdal Formations (Saltnes 1984; Sturt (Kolderup & Moncton 1911) consists of five 1983). A discussion of the unconformity is incor­ arcuate beits containing rocks with different porated in this paper. A syn- to post-kinematic lithologies and tectonometamorphic histories event associated with Dl reached upper green­ (Kolderup & Kolderup 1940; Kvale 1960; Sturt & schist facies conditions indicated by the occur­ Thon 1978 and refs. therein). It is now established rence of almandine garnet and pargasitic that the allochthonous Caledonian sequence hornblende (lngdahl 1985; Fossen 1986, 1988). within the Major Bergen Are (MaBA) consists of two major complexes, the Gullfjellet Ophiolite Complex (GOC) (Furnes et al. 1982; Thon 1985; lngdahl l985) and the Complex (Thon The Holdhus Group 1985; Ingdahl 1985 and refs. therein) separated Sediments of Upper Ordovician age have been fromthe overlying Holdhus and Ulven Groups by described both in the Os and the Samnanger­ a major stratigraphic unconformity (Kvale 1960; Osterøy area within the Major Bergen Are Sturt & Thon 1976; Naterstad 1976; Færseth et (Reusch 1882; Kolderup & Kolderup 1940; Fær­ al. 1977; Thon 1982). In the Os area the younger seth et al. 1977). Within the Samnanger area these sequence is preserved in three large, modified sediments were divided into two formations: the synclinal structures (the Os, Hegglandsdal and Grasdalen Formation and the Moberg Con­ Ulven Synclines) (Figs. l and 2). glomerate, making up the Holdhus Group (Fær­ The rocks within the Hegglandsdal and Os Syn­ seth et al. 1977). Within the Os area these rocks clines are termed the Holdhus Group (Færseth et have traditionally been divided into chlorite spa­ al. 1977; Thon 1985; Ingdahl 1985) while the rocks ragmite with conglomerate, mica schist with lime­ within the Ulven Syncline are termed the Ulven stone and polymictic coarse conglomerate Group (Ryan & Skevington 1976). In addition, (Reusch 1882; Kolderup & Kolderup 1940). the Moberg Formation appears in an Fl anticline Kvale (1960) and Thon (1985) suggested that both at Hauglandsosen with the 'quartz-augen gneiss' the chlorite sparagmite and the Moberg Con­ forming a narrow zone in the core (Øvereng 1969; glomerate were older than the Ashgillian lime- SIMPLIFIED GEOLOGICAL MAP

OF THE OS AREA . INGDAHL 1989 GEOLOGICAL SYMBOLS z UJ . _ lithological bomdary � Thrust � � �THE SKARFJELL FORMA Tl ON contact :::>0 ·---F!Ut �� !IIIIITHE VAKTDAL FORMATION >- S 1 (main Scandian f()jation) 1- C]THE GJERTRUDSBERG FORMATION

THE HEGGLANDSDAL FORMATION � METAGREYWACKE I!!!D POLYMICT CONGLOMERATE THE V ALLA FORMATION !:5SlLIMESTONE 0 MICA SCHIST/PHYLLITE ml!]THE MOBERG FORMATION m:;] � L FJELLET OPHIOLITE t�� �E } � OUARTZ·AUGEN GNEISS §THE SAMNANGER COMPLEX

A NORSK GEOLOGISK TIDSSKRIFT 69 (1989) Ordovician-Silurian rocks, Os area, W. Norway 165 THE OS AREA

THE ULVEN SYNCLINE

THE HEGGLANDSDAL SYNCLINE

Fault contact towards the Gullfjellet Ophiolite Complex

Metagreywacke locally current bedded THE with interbedded os cgl. and phyllites SYNCLINE

Gullfjellet Ophiolite Complex

Fig. 2. Stratigraphy of the Upper Ordovician/Lower Silurian cover sequence to the GOC in the Os area, Major Bergen Are. stone. The present author suggests a subdivision glomerate to pebble sandstone, sandstone and of the Holdhus Group into the Hegglandsdal, phyllite. The coarse conglomerate is immature Valla and Moberg Formations, supporting the both compositionally and texturally and contains stratigraphy of Kolderup & Kolderup (1940). In clasts of both igneous and sedimentary origin in the Hegglandsdal Syncline, thrusting of older a matrix of medium-grained lithic and feldspathic rocks (the GOC) abo ve younger rocks (the greywacke. Primary structures other than strati­ Holdhus Group) along the inverted western limb fication and grading are scarce, but channelling has cut out several units of the Holdhus Group, and truncation of sandstone beds by conglomerate making the Hegglandsdal Syncline into a half units are observed locally'. The cobbles in the syncline where mainly the eastern part remains conglomerate unit reftect the underlying base­ (Ingdahl 1985). ment rocks with cobbles solely from the GOC. Greenstone, amphibolite and phyllite/mica schist The Moberg Formation cobbles/pebbles are generally strongly stretched and occur either as lenses or bands. They are The Moberg Formation (the Moberg Conglom­ generally difficultto differentiatefrom the matrix erate of Reusch 1882) within the Major Bergen due to the similar colours and the strain state. Are can be traced from the Os area in the south Previous authors on this subject have reported through the Samnanger area to Osterøy in the quartzite clasts in the conglomerate (Reusch 1882; north (Reusch 1882; Kolderup & Kolderup 1940; Kolderup & Kolderup 1940; Kvale 1960; Færseth Færseth et al. 1977; Henriksen 1981). The Mo berg et al. 1977). This is supported by the present Formation is made up of coarse, polymict con­ author, but all quartzite clasts are interpreted as glomerate with minor amounts of granular con- metachert, which indicates erosion of a formerly

Fig. l. Simplified geological map of the Os area. 166 S. E. Ingdahl NORSK GEOLOGISK TIDSSKRIFT 69 (1989) marine environment closely related to the gillian age. The limestone has been traced from Gullfjellet Ophiolite. Kolderup & Kolderup the Os through the Samnanger to the Osterøy (1940) reported both limestone and marble boul­ area (Færseth et al. 1977). In the Samnanger area ders. They further point to the possibility of an all the rocks above the Moberg Formation were older marble horizon now eroded. Kvale (1960) collectively named the Grasdalen Formation identified two types of limestones - one grey (Færseth et al. 1977). The Grasdalen Formation type occurring within the Middle Phyllite (Valla has also been used synonymously with the Valla Formation) and an older white one, supporting Formation (Ingdahl 1985). As the name was the observations of Kolderup & Kolderup (1940). already in use in Finnmark (Siedlecka & Siedlecki Kvale (1960) claimed that only the white one 1971), it was suggested by the Norsk Stratigrafisk occurs as pebbles in the conglomerate. This Komite that the name should be replaced; sub­ author supports the observation that limestone/ sequently Valla was chosen as the new formation marble boulders/pebbles occur within the name. This formation conformably overlies the Moberg Formation. It is not possible, however, Moberg Formation. At several localities, to relate these to any certain horizon. They may however, the contact is tectonized. In Samnanger well represent an older horizon now eroded, but the limestone shows interfingering relationships they may also represent erosion of more or less with the conglomerate of the Moberg Formation. synsedimentary limestones eroded from nearby The phyllite consists of quartz, white mica and areas as intraformational limestone pebbles/ chlorite as essential minerals; biotite, albite, cli­ cobbles occur locally towards the top. nozoisite, epidote, opaques, apatite, sphene, The coarse, polymictic conglomerate is gen­ zircon, tourmaline, graphite and garnet as access­ erally interbedded with 5 to 100 cm thick massive, ories. A gradual change from conglomerate grey to green sandstone beds. The compositions through sandstone into limestone or calcareous of these sandstones are similar to the con­ phyllite, which reflectsa sedimentary transition, is glomerate matrix and consist of quartz, plagio­ locally observed. In most places, however, either clase, biotite, hornblende, epidote and chlorite limestone or phyllite has sharp and sheared con­ as the essential constituents. Quartz and albite tacts against the conglomerate. The limestone/ also occur as secondary minerals (along with horn­ marble generally forms discontinuous beds occur­ blende, biotite, white mica, epidote and chlorite. ring as limes tone lenses of variable size arranged White mica, calcite, opaques, sphene, tourma­ parallel to the regional strike. A more continu­ line, clinozoisite, apatite and zircon occur as ous outcrop of the limestone occurs in central accessory minerals. Dark grey to black phyllite Hegglandsdal. The limestone, which consists of beds, 0.5-5 cm in thickness, consist of quartz, calcite with quartz, albite, biotite, white mica, white mica, biotite and chlorite as the essential clinozoisite, opaque, sphene and graphite as minerals. Accessories are albite, sphene and cal­ accessories, generally shows a red-brown colour cite. on weathered surfaces and a white to bluish colour A monomictic, granular conglomerate to peb­ on fresh surfaces. The purest limestone is gen­ ble sandstone generally occurs below or is inter­ erally converted to marble. The more impure bedded with the coarse, polymictic conglomerate. varieties with thicker beds of pelitic material are It consists of quartz clasts in a matrix containing less recrystallized and deformed, probably due to white mica, epidote and chlorite. The quartz clasts the pelite taking up most of the strains. It is are generally subangular to subrounded, and vary generally within these less-deformed limestone in diameter from l to 15 mm. The rock is greyish beds that fossils are found. Within the phyllite, green in colour due to the presence of chlorite sandstone horizons occur at several places and and epidote. The sandstone has a higher content some are large enough to be mapped. Thinner of white mica and chlorite, and fewer quartz clasts sandstone horizons are generally boudinaged and than the conglomerate. In addition, it contains are in addition partly transposed. Others, occur­ altered feldspar grains. ring in major fold closures, are tight to isoclinally folded. When Reusch described the fossils from the Ulven and Os areas in 1882, it was quite The Valta Formation a sensation, as they were among the first ever The Valla Formation constitutes phyllite/rnica described from metamorphic rocks. At that time schist, marble and fossiliferous limestone of Ash- it was also generally believed that all mica schists NORSK GEOLOGISK TIDSSKRIFT 69 (1989) Ordovician-Silurian rocks, Os area, W. Norway 167 were of Precambrian age. Fossils are found in mica, clinozoisite, garnet, apatite, sphene, zircon, several places within both the Hegglandsdal and tourmaline and rock fragments. Garnet, horn­ the Ulven synclines. The fauna is not very rich blende, biotite and epidote frequently occur as and individual localities show limited assem­ porphyroblasts. Rock fragments have been blages. All previous authors on the subject identified as diorite, trondhjemite/tonalite, (Reusch 1882; Kiær 1929; Kolderup & Kolderup greenstone, epidosite and 'quartz-augen gneiss'. 1940) agree in correlating the fauna within the The carbonate content decreases upwards. Cal­ Hegglandsdal Syncline with the gastropod lime­ careous greywacke beds occur in the lower part stone in the Oslo area (stage Sa, Ashgillian). of the sequence either directly overlying the con­ glomerate or partly interfingering with or over­ lying the limestone/phyllite. The calcareous The Hegglandsdal Formation greywacke shows both incomplete and complete Within the Hegglandsdal Syncline this formation graded beds. The thicker graded beds fine represents the uppermost part of the Holdhus upwards from coarse to very finesand. Each unit Group. Because of faulting and increased excision is often capped by a fine sandstone bed. Small­ along the inverted western limb the formation scale trough cross-bedding and climbing ripples thins both northward and southward. The Hegg­ are frequently seen in this facies. Rip-up clasts landsdal Formation comprises a sandstone are also common and some of the sandstone beds sequence with minor amounts of conglomerate also contain larger fragments of epidote. The and mica schist. Locally it contains a basal, poly­ plane parallel bedded greywacke unit forms 0. 5- mictic conglomerate which has a sharp contact 2 m thick units of fine to medium-grained lithic with the underlying limestone or greywacke. The and calcareous greywacke. Each unit is made up contact with the overlying greywacke is not of several minor beds which vary from 5 to 30 cm exposed. This conglomerate reaches its maximum in thickness, usually separated by well-sorted thickness of 30 m in the central part of Hegg­ finer-grainedsandstone or pelite laminae. Locally landsdalen. The conglomerate is generally clast­ the interbeds are missing, and bedding planes are supported with minor greywacke matrix. The cob­ marked by abrupt changes in grain size. The bles are well rounded and consist of epidosite, contacts between the different beds, however, are greenstone, 'quartz-augen gneiss', trondhjemite/ always sharp. tonalite, diorite and gabbro. In addition, pebbles of vein quartz are frequent. During the Dl deformation they have been extensively stretched with the long axes in the 2-30 cm range and short axes about 1/2 to 1/ 5 of the long axes. This The Ulven Group conglomerate resembles that of the Moberg For­ Sediments of Lower Silurian age appear in a mation, but is based on structural observations syncline between ophiolitic rocks (Reusch 1882; thought to represent a higher stratigraphical hori­ Kolderup & Kolderup 1940; Ryan & Skevington zon. Along the contact between the GOC and the 1976). The Ulven Syncline (Kolderup & Kolderup sandstones, conglomerates correlated with this 1940) is a tight to isoclinal fold which appears as conglomerate crop out. The western zone of con­ an open fold within the quartzite in the central glomerate is discontinuous and is only locally part around Ulvenvatn. This syncline becomes found between the GOC and the overlying sand­ tighter both northwards and southwards and here stones. This is probably due to a combination of the quartzite in addition approaches an isoclinal palaeotopography and tectonic effects.At Borgen configuration. The rocks within the Ulven Syn­ the conglomerate has its maximum apparent eline are termed the Ulven Group (Inderhaug thickness of about 40 m. 1975; Ryan & Skevington 1976). The sandstones constitute a fining-upward sequence of greywacke comprising plane parallel bedding and graded beds. lts maximum appar­ The Gertrudsberg Formation ent thickness is 200 m in central Hegglandsdal. The greywacke consists of quartz, feldspar and The Gertrudsberg Formation (Ryan & Ske­ rock fragments as essential constituents. Access­ vington 1976) consists of greywacke with poly­ ories are biotite, chlorite, epidote, calcite, white mictic conglomerate and phyllite. Sometimes 168 S. E. Ingdahl NORSK GEOLOGISK TIDSSKRIFf 69 (1989) quartz siltstones and irregular bands of con­ The Skarfjell Formation glomerate occur at or near the lower contact. The conglomerate, denoted the Ulven Conglomerate The Skarfjell Formation (Ryan & Skevington by Kolderup & Kolderup (1940), is composed 1976) represents the uppermost part of the Os of flattened fragments lithologically identical to Group and occupies the hinge of the Ulven Syn­ underlying rocks such as greenstones, diorites, eline. The thickness of the formation varies from epidosites, trondhjernites,gabbros and quartzites O to 230 m. It was deposited on the Vaktdal For­ in a matrix of quartz, feldspars, amphiboles and mation and locally the two formations are inter­ occasionally garnets overgrown by amphibole bedded. The Skarfjell Formation is divided into (homblende) and chlorite. The greywacke shows two members, the conglomerate member and the rare cross-bedding. The clast size is between l sandstone member. The formation generally and 4cm in length and they are normally flattened starts with a basal conglomerate followed by a and elongated subparallel to the contact. Most much thicker sandstone sequence. The con­ fragments are fairly angular. The conglomerate glomerate is generally found directly above the only covers a narrow and short zone with a thick­ phyllite, but locally it is either missing due to ness of less than 5 m on average. The con­ lateral thinning or to a thin layer of sandstone glomerate grades into a protoquartzite sanctstone occurring between the two. Pebbles of quartzite with white mica, chlorite and opaques defining locally occur within the Vaktdal Phyllite dose to the main (Sl) cleavage, which is commonly sub­ the contact with the Skarfjell Formation. The parallel to the layering (SO). The protoquartzitic conglomerate is clast supported with both reverse sandstones are interlayered with phyllitic horizons and normal grading. The conglomerate consists (lnderhaug 1975). of well-rounded and oblate cobbles of white to grey quartzite in a quartzitic matrix. Cobbles of serpentinite, sandstone, phyllite, trondhjemite/ The Vaktdal Formation tonalite and calcareous material make up about This formation consists primarily of phyllites, but 15% of the total amount of cobbles in places. The metasandstones, carbonate and siltstone beds as cobbles show a fairly well definedpreferred planar well as graphite-rich schists are found. The for­ and linear orientation. Both normal and shear mation has a maximum thickness of 125 m on the fractures are seen, and locally the individual parts northwest limb and 10 5m on the southeast limb of the cobble are distorted. The conglomerate at of the Ulven Syncline. At its base, the contact the base contains 50 % or more by volume of between the phyllites of this formation and the rounded clasts of pure white quartz (1-15 cm conglomerates or metasandstones of the Ger­ along the longest axis, 5--10 cm most common) trudsberg Formation (Inderhaug 19 75; Ryan & set in a quartzite matrix. The conglomerates vary Skevington 1976) is sharp. In the northwest, 50 m in thickness up to a maximum of 55m and often above the base, a graptolite fauna has been contain thin intercalations of quartzite (0.3-- obtained (Ryan & Skevington 1976). The top 0. 5 m). Where the conglomerate is absent at the 10 m of this formation contains rare beds of quartz base of the formation, the basal 2m of quartzite siltstone (O.l m) and flattened corals (Ryan & is often calcareous, containing up to 15% by Skevington 1976). The fauna in the Ulven Syn­ volume of calcite and occasional shelly fossils eline has recently been re-investigated and found (Ryan & Skevington 1976). The conglomerate to belong to the M. gregarius Zone (Ryan & grades into the overlying sandstone with several Skevington 1976), which can be correlated with cobble beds in the transition zone. The contact the lower part of stage 6b, Middle Llandoverian between the conglomerate and the sandstone beds in the Oslo area. According to the new subdivision is sharp. The matrix of the conglomerate is com­ of the Silurian sequence in the Oslo Region posed of fine-grained quartz and minor amounts (Worsley et al. 1983), the Vaktdal Formation of sericite/white mica and biotite and locally, should correlate with the Sælabonn and/or the carbonate. The sandstone overlying the con­ Solvik Formations. In the southeast, limestone glomerate consists of about 90% medium-grained bands (1-2 cm), separated by phyllites (2-4cm), quartz, 5% biotite and white mica and about occur in the top 2m. Reusch's etage 6c fauna 5% feldspar, opaques and calcite. Small quartzite probably came from the top few metres of the clasts (2-7 cm) occur irregularly in the lower Vaktdal Formation (Ryan & Skevington 1976). 10 0 m of this formation. Primary structures, such NORSK GEOLOGISK TIDSSKRIFT 69 (1989) Ordovician-Silurian rocks, Os area, W. Norway 169 as cross-bedding and slumping, are frequently glomerate sequence of the Skarfjell Formation seen. In the upper half of this formation rare may be studied in the fold closure of the Ulven bands of quartz siltstone (1-3 cm) and shale (1- Syncline (within the industrial area) (Fig. 3). Here 30 cm) occur. one can see two strong cleavages and one weak cleavage in the phyllite, while only a single strong cleavage is developed in the conglomerates and quartzites. Within the conglomerate, cobbles are Intrusives orientated in the firstfoliation and some are defor­ Granitoid intrusive dykes are observed in the med with the production of a penetrative cleav­ Moberg Formation. These have intruded pre- or age. Pressure shadows are also developed at clast syn-Dl and they have not been rotated completely edges. This cleavage is the first one observed into parallelism with the Sl foliation during this in the Skarfjell Formation. Within the Vaktdal event. This is supported by the strong deform­ Formation bedding (SO) is revealed by thin silty ation of the granitoids with recrystallization of plagioclase and the development of a strong foli­ S1 ation. The sediments of the Ulven Group are cut by lamprophyres. At the quarry at Hjorthaug the intrusion is seen to cut the layering of the sandstone on the western limb of the syncline, while it is subparallel to the layering in the sand­ stone on the eastern limb. The analysis carried out by lnderhaug (1975) gives values within the alkaline field, both in the alumina-alkali and the alkali-silica variation diagram. The analysis is con­ sistent with Kersantite, a plagioclase-bearing lam­ prophyre with biotite phenocrysts, which is genetically related to diorites. Wl) SKARfiiELL FM. E3 VAKTDAL FM. § GERTRUOSBEAG FM.

A 8 GREENSTONE discussion of the postulated EJ GABBRO unconformity between the Skarfjell and Vaktdal formations A structural and metamorphic discordance has recently been claimed to be present between the Skarfjell Formation and the underlying Vaktdal Formatio_n, based on observations of a recently excavated locality at the inverted western limb of the Ulven Syncline (Saltnes 1984; Sturt 1983). They claim that two cleavages in the phyllite of the Vaktdal Formation are truncated by the conglomerate of the Skarfjell Formation. Quartz veins developed parallel to the first foliation in the phyllite are folded by the second generation folds, which are truncated by the conglomerate. The present author rejects this interpretation, however, following instead that of Inderhaug (1975) and Ryan & Skevington (1976) of this being a continuous sequence. This is based on the following observations:

(l) Structures developed in the phyllite of the Fig. 3. Structural relationships observed at the excavated Vaktdal Formation and the sandstone con- locality on the knoll just east of the Ulven camp (LM 023791). 170 S. E. /ngdahl NORSK GEOLOGISK TIDSSKRIFT 69 (1989) laminae and is isodinally folded. A slaty deavage tion seems to be the only result of deformation. is axial planar to these folds and corresponds in However, in a finer-grained, more calcareous orientation to the first deavage in the Skarfjell sandstone bed in a knoll just to the north of the Formation. The deavage is further axial planar excavated locality there is a penetrative foliation. to the Ulven Syndine. A second deavage of This one is crenulated and locally a ·crenulation crenulation type affects the phyllite and dips cleavage is developed. The oldest of these cleav­ towards the east throughout the dosure of the ages is axial planar to the Ulven Syncline. The first Ulven Syndine, indicating that it must be later cleavage, Sl, is subvertical, while the crenulation than the syndine. A corresponding crenulation deavage, S2, dips 60° to the southeast and has deavage is also weakly developed in the rocks of virtually the same orientation as the S2 cleavage the Skarfjell Formation in the binge of the Ulven in the nearby phyllites. In the northern extension Syndine. The third deavage is a crenulation of the Ulven Syncline the same relationship is deavage occasionally developed in the phyllite. observed better and more frequently. Here the first deformation (Dl) is stronger and sub­ (2) At the excavated locality at Ulven on the sequently the boulders within the conglomerate steep to overturned northwest limb of the Ulven are more strongly deformed. The conglomerate Syndine, two cleavages are well developed in the here contains phyllitic horizons at the base. These phyllite (Fig. 3). A slaty deavage is folded around phyllites show two deavages very well developed asymmetric, angular folds with an axial-planar (Fig. 4). crenulation deavage. A further cleavage is devel­ oped in places, but generally only small, angular (5) The two major phases of folding observed folds affect the earlier structures. A distinct crenu­ in the Skarfjell and Vaktdal formations can be lation cleavage is exclusively developed dose to correlated with the two events, Fl and F2, the base of the Skarfjell Formation in a zone observed in the rest of the Os area. about 0.5 m thick along the contact, which seems to represent a zone of intense deformation. None (6) A traverse through the Skarfjell Formation of the crenulation deavages have orientations within the Ulven Syncline shows that because of commensurate with an origin in connection with competence contrasts between different litho­ the development of the Ulven Syndine. logies the effect of Dl varies considerably. The Sl cleavage is only occasionally developed in less (3) The postulated unconformity is between phyl­ competent lithologies, usually rocks with a higher lites and a thick sequence of conglomerates and proportion of mica, while more competent litho­ orthoquartzitic sandstone. Across such contacts logies are uncleaved. one would expect a marked competence contrast (7) At the excavated locality the phyllite is jux­ which would result in refraction of cleavages and taposed with conglomerates with a sandy matrix variation in the degree of deformation. The Sl containing little white mica which would make the deavage in the Vaktdal Formation has the same development of a penetrative cleavage extremely orientation and is coeval with the first deavage difficult. observed in the Skarfjell Formation throughout the syndine. (8) Fossiliferous horizons within the Vaktdal For­ mation are symmetrically disposed around the (4) The occurrence of two strong and one weak Skarfjell Forma ti on in the core of the Ulven Syn­ deavage in the phyllite and only one weak cleav­ eline (C. Magnus, pers. comm.). age in the conglomerate is part of the evidence put forward for the existence of an unconformity (Saltnes 1984; Sturt 1983). The presence of two Sedimentological interpretation and strong cleavages and one weak deavage in the phyllite is considered incontrovertible. However, discussion of the Os Group the present author would argue that locally there The Os Group starts with the Moberg Formation, are also two deavages in the Skarfjell Formation. which is interpreted as an alluvial fan deposit At the excavated contact at Ulven (Saltnes 1984; related to fault scarps dose to an ancient shoreline Sturt 1983) the conglomerates and metasand­ and in part reworked by shoreline processes (Fig. stones seem almost undeformed. Pressure solu- 5). The dasts are generally well rounded and, NORSK GEOLOGISK TIDSSKRIFT 69 (1989) Ordovician-Silurian rocks, Os area, W. Norway 171

Fig. 4. A strong S2 crenulation cleavage developed in the phyllite bed near the base of the quartzite conglomerate of the Skarfjell Formation (LM 063815).

MARINE

r,;ølCoarse polymicl � Æconglonierale �\ Granular conglomerate/ ·'l lJ pebble sandstone O sand \\ 0 shale rocks

Fig. 5. Schematic, interpretative model for the Moberg Formation in Ashgill to pre-Ashgill times, representing large fans composed of cobbles, gravels and sand reworked by marine processes. 172 S. E. Ingdahl NORSK GEOLOGISK TIDSSKRIFT 69 (1989) despite later tectonic strains, in less deformed ophiolitic rocks has been claimed from various areas this is believed to reflecttheir original shape, parts of western Norway. The increased rates indicating either that the conglomerate was of subsidence in Late Ordovician/Early Silurian resedimented or was reworked by shoreline pro­ times could be the effect of isostatic loading by cesses. The cobbles reflectthe immediately under­ nappes thrust towards the southeast during an lying substrate, which suggests that transport was early Caledonian orogeny, but the Jack of evi­ relatively short. The conglomerate was probably dence for such a model makes this highly specu­ deposited on an uneven topography and in fault­ lative. bounded basins, as it varies considerably in thick­ The Valla Formation is stratigraphically situ­ ness laterally. ated above the Moberg Formation. Locally it Some of the sandstone beds are laterally per­ sits on the Moberg Formation with a primary sistent and may represent wave-worked shoreline depositional contact, but generally has a highly­ deposits (Clifton 1973). As the contact between sheared contact with the underlying rocks. The the Moberg Formation and the Valla Formation lithologies of the Valla Forma tion are interpreted is transitional, shallow-marine fossils in the as marginal-marine deposits. This is based on overlying Valla Formation also support a the occurrence of marine fossils and the overall shallow-marine environment. The sandstones fre­ stratigraphic context with a fining-upwards quently form discontinuous lenses within the con­ sequence from conglomerates in the Moberg For­ glomerate, indicating alluvial reworking {Miall mation through sandstones and phyllites to inter­ 1981, 1983) and/or fragmentation during de­ layered impure limestones and phyllites with the formation. Towards th(j top of the sequence, local occurrence of purer limestones. The fine­ limestones frequently occur locally as clast grained, 'phyllitic' and silty sediments indicate a material, suggesting reworking of limestones low-energy environment, while the interbedded and a synsedimentary evolution of conglom­ sandstones probably represent storm deposits. erates and limestones. This is further supported The fining-upward trend could be due to dim­ by observations in the Samnanger area, where inishing sediment supply and fan retreat related to fossiliferous limestones are interbedded with denudation of the source area and/or an increased conglomerates. distance to the fan apex related to backward A granular conglomerate to pebble sandstone faulting of the source area. Transgression and occurs locally between the GOC and the cobble deposition of fossiliferous limestones in Ashgill conglomerate. The granular conglomerate to peb­ time, indicate a tectonically quiet period and ble sandstone is interpreted as more distal alluvial gentle topography in hinterland. Elevated sediments, which during subsequent tectonic domains underlain by resistant lithologies con­ activity or transgression have been overlain by stituted shallow-water areas and islands favour­ coarser conglomerates. able for carbonate sedimentation. This Upper The precise age of the Mo berg Forma ti on can­ Ordovician/Lower Silurian limestone horizon is not be determined, but as limestone overlying it found throughout western Norway and can be contains shallow-marine fossils of Ashgillian age, traced from Karmøy in the south to Lyngen in the Moberg Forma tion must be Ashgillian or pre­ the north, indicating a major break in tectonic Ashgillian. A maximum age for the formation is activity on a regional scale. In support of eustatic indicated by the recently published datings from sea-leve! rises is the worldwide distribution of a plagiogranite differentiate of the GOC giving Llandoverian transgressive formations (Berry & an age of 489 ± 3 Ma and an arc-related Boucot 1973). trondhjemite pluton (the large 'quartz-augen' The Hegglandsdal and the Gertrudsberg for­ gneiss body) gi ving an age of 482 :t:� Ma (Dunning mations have been correlated based on stra­ & Pedersen 1988). A further indication of the age tigraphical position and lithological composition. is given by the Moberg Formation equivalent on The Gertrudsberg Formation may represent a Bømlo (the Roaldsfjord Conglomerate), where more distal facies of the Hegglandsdal Formation. underlying andesites and rhyolites from the Siggjo The rocks of the Hegglandsdal Formation are Complex give Rb/Sr whole rock ages of interpreted as shallow-marine to shelf deposits, 464 ± 16 Ma and 468 ± 23 Ma (Furnes et al. probably related to delta progradation in an active 1983). continental margin setting. Phases of delta pro­ Thrusting/emplacement of both foreland and gradation and subsequent transgression/aban- NORSK GEOLOGISK TIDSSKRIFT 69 (1989) Ordovician-Silurian rocks, Os area, W. Norway 173 donment may explain the observed alternations of Llandoverian age. The transgression may be of conglomerate and carbonate beds. This model related to a remelting of the Upper Ordovician is based on the stratigraphical position above ice age. shallow-marine calcareous sediments and on sedi­ A period of renewed faulting is indicated by mentary structures such as parallel lamination, the deposition of the coarse, reworked marine sharp, planar to erosive bases, normal grading conglomerates and sandstones of the Skarfjell with a turbidite-like appearance and cross-lami­ Formation. The Skarfjell Formation has an eros­ nated greywackes. Each sandstone bed is most ive base and overlies shallow-marine sediments probably the result of a single depositional event. with a slightly angular, stratigraphic uncon­ The cross-laminated greywackes probably refiect formity. The Skarfjell Formation is tentatively local wave reworking. Planar-laminated sand­ correlated with the Utslettefjell Conglomerate on stones are attributed to wave currents and storm­ , which shows the same relationship (Thon surge effects.The horizontal lamination may also et al. 1980). The Skarfjell Formation is inter­ reftect deposition of sand fromsuspension during preted as beach to shallow-marine deposits fol­ storms (Dott & Bourgeois 1981; Hamblin & Wal­ lowing Inderhaug (1975) and Ryan & Skevington ker 1979), i.e. as high-velocity, highly-turbulent (1976). The idea that the Skarfjell Formation ftows. High-energy events are recorded in the should be deposited by braided rivers (Saltnes thicker, laterally-persistent sheet sandstones. 1984, and refs. therein) cannot be supported. A Trough cross-stratifiedsandstones reftectonshore beach to shallow-marine environment for the waves and longshore currents. In central Hegg­ Skarfjell Formation is based on the following landsdal, the carbonate sedimentation (Valla observations in the Ulven area. Formation) was followed by a conglomerate (a) The Skarfjell Formation contains mature which prograded into a marine environment. This sediments which are more likely in shallow­ conglomerate indicates that the transgression was marine to beach environments than in braided discontinuous, with minor regressions resulting in rivers. The sediments are mature because the inftux of coarse material. Whether this was due matrix in the conglomerate is very quartz-rich to eustatic changes in sea leve! or active tectonism, and well sorted. The conglomerate cobbles are however, is impossible to tell. The presence of dominated by quartzite, over 90% locally; they conglomerates in a shallow marine succession may are of about the same size and are well rounded. indicate a nearby high-relief source of terrestrial (b) Individual conglomerate beds show the detritus. The conglomerates were probably trans­ lateral continuity typical for shallow-marine to ported by rivers and reworked into beach gravels. beach environments rather than in braided rivers. Fining-upward sequences of this type are gen­ (c) Low-angle cross-bedding in quartzite. This erally attributed to transgressive cycles in which indicates beach to shallow-marine environments. the rate of subsidence exceeds that of sedi­ In braided rivers one would expect high-angle mentation, related to eustatic sea leve! changes trough cross-bedding. and/or to backward faulting. The upward ( d) Sharp contacts between conglomerate and decrease in calcareous material is related to an sandstone beds. This can also be observed in increase in clastic input. The phyllite horizons braided-river deposits, but is normally associated occurring within the greywacke may represent with channelling. In the Skarfjell Formation ero­ variations in rate of sedimentation with time and sional contacts related to channelling beds cannot of ftow regime capacities. be inferred. The Vaktdal Formation is interpreted as shal­ ( e) The conglomerate beds are texturally uni­ low marine deposits as it shows a sedimentary form. transition upwards into the Skarfjell Formation ( f) Normal grading occurs, but individual con­ which represents beach to shallow marine glomerate beds are generally either ungraded or deposits. The thin siltstones and sandstones are coarsening upwards. This is more typical for considered to be distal representatives of pro­ shallow-marine to beach deposits than for grading fan deltas. These phyllites are probably braided rivers. at !east partiy equivalent to the Lower Llandovery (g) The quartzites are well sorted. graptolitic shales on Stord (Ryan & Skevington (h) The framework component is commonly 1976). The Vaktdal Formation reftects trans­ well sorted in individual conglomerate beds, gression and deposition of clay-rich sediments though successive beds may vary greatly in coarse- 174 S. E. Ingdahl NORSK GEOLOGISK TIDSSKRIFT 69 (1989) ness. Sand infillor small pebble infill locally cause beach deposits where alluvial sediments inter­ bimodal textures. fingerwith ti dal deposits because of several trans­ (i) Local beds of finer material are always cal­ gressive and regressive cycles. The coarse careous. conglomerates at the base of the Holdhus Group (j) The Skarfjell Formation overlies shallow­ were a response to faulting. Transgression and marine pelitic sediments of Llandoverian age deposition of fossiliferous limestones in Ash­ (Ryan & Skevington 1976) without any major gillian time indicate a tectonically quiet period unconformity. and gentle topography. Elevated domains under­ (k) Pebby sandstones probably result from lain by resistant lithologies constituted shallow­ shape filtering on grave! beaches causing pebble water areas and islands favourable for carbonate entrapment of sand beds (Bluck 1967). sedimentation. The shoreline to shelf deposits of (l) Absence of angular fragments in the con­ the Hegglandsdal Formation indicate progra­ glomerate units. At !east some angular fragments dation of a delta. The calcareous sediments were would be expected in most braided river environ­ drowned and overlain by boulders and sandstone. ments. Transgression and deposition of clays of Llan­ (m) Extremely well-rounded clast material in dovery age are only observed within the Ulven the conglomerate beds. Syncline. A period of renewed faulting is indi­ In addition to the characteristics which favour cated by the coarse, reworked marine con­ a shallow-marine to beach environment, the glomerate and sandstone of the Skarfjell Skarfjell Formation lacks some of the main Forma.tion. No structural/metamorphic break characteristics of braided rivers which may sep­ occurs between any of the formations and the arate them from shallow-marine to beach recently postulated structural/metamorphic deposits, such as planar cross-bedding, upwards break between the Vaktdal and Skarfjell for­ fining within beds, ripples, large-scale channels mations is not supported. and fine-grained clastic sediments. The Skarfjell Formation overlies the Vaktdal Formation, which Acknowledgements.- The author thanks Dr. B. Robins and E. Prestholm (Geology lnstitute/Dep. A, University of Bergen), locally contains fossils of Llandoverian age (Ryan and H. Fossen and K. G. Jacobsen (Statoil/Bergen) for their & Skevington 1976). This indicates that the helpful advice and critical comments, which substantially deformation of the Os Group must be post-Lower improved the manuscript. The author also thanks an anonymous Silurian in age. referee for critically reading an early draft of the manuscript and for the many suggestions towards its improvement. Statoil, Bergen is thanked for financial support for my attending the NGF meeting in Trondheim in 1986. Statoil's drawing offices in Conclusion Bergen and Stavanger are acknowledged for help with the figures. The two younger Ashgillian-Llandoverian sedi­ mentary sequences in the Os area are preserved Manuscript received July 1987 in Fl synclines. Both sequences have been cor­ related and collectively named the Os Group. References The group is divided into five formations. In stratigraphic order these are: (1) The Skarfjell Berry, W. B. & Boucot, A. J. 1973: Glacioeustatic control of Formation (uppermost), comprising conglom­ late Ordovician-early Silurian platform sedimentation and fauna! changes. American Bulletin of Geological Society 84, erate and metasandstone. (2) The Vaktdal For­ 275--284. mation, comprising phyllite, limestone and Bluck, B. J. 1967: Deposition of some Old Red Sandstone metasandstone of Llandoverian age. (3) The conglomerates in the Clyde area: a study in the significance Hegglandsdal Formation, comprising sandstone of bedding. Scottish Journal of Geology 3, 139-167. and min or polymictic conglomerate. ( 4) The Valla , H. 1983: The Caledonian Geological patterns of and southern Bømlo. Evidence for Lower Palaeozoic Mag­ Formation, comprising marble, phyllite and fos­ matic Are Development. Unpublished Cand. Real. Thesis. siliferous shallow marine limestone of Ashgillian University of Bergen, Norway, 473 pp. age. (5) The Moberg Formation (lowerrnost), Clifton, H. E. 1973: Pebble segregation and bed lenticularity in comprising monomict granule conglomerate to wave-worked versus alluvial grave!. Sedimentology 20, 173-- 187. pebble sandstone, polymictic coarse conglomer­ Dott, R. H., Jr. & Bourgeois, J. 1981: Sedimentary rocks. In: ate, metasandstone and min or phyllitic interbeds. Encyc/opedia of Science and Technology Yearbook. McGraw­ The Os Group represents shallow marine to Hill. NORSK GEOLOGISK TIDSSKR!Ff 69 (1989) Ordovician-Silurian rocks, Os area, W. Norway 17 5

Dunning,G. R. & Pedersen,R. B. 1988: U /Pb ages of ophiolites Naterstad, J. 1976: Comment on the Lower Palaeozoic Uncon­ and arc-related plutons of the Norwegian Caledonides: impli­ formity in West Norway. American Journal of Science 276, cations for the development of Iapetus. Contributions to 394-397. Mineralogy and Petrology 98, 13-23. Reusch, H. 1882: Silurfossiler & Pressede Konglomerater i Fossen, H. 1986: Metamorphism in the Major and Minor Bergensskifrene. Kristiania (Oslo), 152 pp. Bergen Arcs. Geolognytt 22, Norsk Geologisk Forenings 10. Ryan, P. D. & Skevington, D. 1976: A Reinterpretation of the landsmøte (abstract). Late Ordovician-Early Silurian stratigraphy of the Dyvikvå­ Fossen, H. 1988: The Ulriken Gneiss Complex and the Run­ gen and Ulven-Vaktdal areas, , Western Norway. demanen Formation: a basement-cover relationship in the Norges geologiske undersøkelse 324, 1-19. Bergen Arcs, West Norway. Norges geologiske undersøkelse Saltnes, M. 1984: Deformation of quartzit(} conglomerates, 412, 67-86. Ulven, Os, Western Norway. Unpublished Cand. Real. Furnes, H.,, H. Amaliksen,K. G. &Nordås,J. 1983: Thesis, University of Bergen, Norway. 237 pp. Evidence for an incipient Caledonian (Cambrian) orogenic Siedlecka, A. & Siedlecki, S. 1971: Late Precambrian sedi­ phase in southwestern Norway. Geological Magazine 120, mentary rocks of the Tanafjord-Varangerfjord region of 607-612. Varanger Peninsula, . Norges geologiske Furnes, H., Thon, A., Nordås, J. & Garmann, L. B. 1982: undersøkelse 269, 246-294. Geochemistry of Caledonian metabasalts from some Sturt, B. A. 1983: Late Caledonian and possible Variscan stages Norwegian ophiolite fragments. Contributions to Mineralogy in the Orogenic evolution of the Scandinavian Caledonides and Petrology 79, 295-307. (Abstract)-Symposium Morocco and Pa/eozoic Orogenesis Færseth, R. B., Thon, A., Larsen, S. G., Sivertsen, A. & Rabat, 30-31. Elvestad, L. 1977: Geology of the Lower Palaeozoic Rocks Sturt, B. A., Andersen, T. B. & Furnes, H. 1985: The Skei in the Samnanger-Osterøy Area,Major Bergen Are, Western Group, Leka: An unconformable clastic sequence overlying Norway. Norges geologiske undersøkelse 334, 19-58. the Leka Ophiolite.Jn Gee, D. G. & Sturt, B. A. (eds.): The Hamblin, A. P. & Walker, R. G. 1979: Storm-dominated shal­ Caledonide Orogen-Scandinavia and Related Areas. John low marine deposits: the Fernie-Kootenay (Jurassic) tran­ Wiley, New York. sition, southern Rocky Mountains. CanadianJournalof Earth Sturt, B. A. & Thon, A. 1976: The age of orogenic deformation Science 16, 1673-1690. in the Swedish Caledonides. AmericanJournal of Science 276, Henriksen, H. 1981: A major unconformity within the meta­ 385-390. morphic Lower Palaeozoic Sequence of the Major Bergen. St urt, B. A. & Thon, A. 1978: Caledonides of . Are, Osterøy, Western Norway. Norges geologiske under­ Geological Survey of Canada. IGCP Project 27, 39-47. søkelse 367, 67-75. Thon, A. 1980: Steep shear zones in the basement and associ­ Inderhaug, J. E. 1975: Liafjellsområdets Geologi. Unpublished ated deformation of the cover sequence on Karmøy, S.W. Cand. Real. Thesis, University of Bergen, Norway, 165 pp. Norwegian Caledonides. Journal of Structural Geology 2, 75- Ingdahl,S. E. 1985: Stratigraphy, structural geology and meta­ 80. morphism in the Os area, Major Bergen Are. Unpublished Thon, A. 1985: Late Ordovician and Early Silurian cover Cand. Real. Thesis, University of Bergen, Norway, 667 pp. sequences to the west Norwegian ophiolite fragments: stra­ Kiær, J. 1929: Den fossilførende Ordoviciske-Silurske lagrekke tigraphy and structural evolution. In Gee, D. G. & Sturt, B. på Stord. Bergen Museums Årbok. A. (eds.): The Caledonide Orogen-Scandinavia and Related Kolderup, C. F. & Kolderup, N. H. 1940: Geology of the Areas. John Wiley, New York. Bergen Are System. Bergens Museum Skrifter 20. 137 pp. Thon, A. 1985: The Gullfjellet Ophiolite Complex and the Kolderup, C. F. & Moncton, H. W. 1911: The Geology of structural evolution of the Bergen Are, west Norwegian Cale­ the Bergen District, Norway. Project Geological Association donides. /n:Gee, D. G. &Sturt,B. A. (eds.): TheCaledonide XXIII, l-16. Orogen-Scandinavia and Related Areas. John Wiley, New Kvale, A. 1960: The Nappe Area of the Caledonides in Western York. Norway. Excursion Guide. International Geological Thon, A. Magnus, C. & Brevik, H. 1980: The stratigraphy of Congress, 21st, Norden 1960. Norges geologiske undersøkelse the Dyvikvågen Group, Stord. A revision. Norges geologiske 212e, 43 pp. undersøkelse 359, 31-42. Miall, A. D. 1981: Alluvial sedimentary basins: tectonic setting Worsley, D., Aarhus, N., Bassett, M. G., Howe, M. P. A., and basin architecture. bi Miall, A. D. (ed.): Sedimentation Mørk, A. & Olaussen, S. 1983: The Silurian succession of and tectonics in alluvial basins. Special Paper Geological the Oslo Region. Norges geologiske undersøkelse384, 1-57. Association Canada 23, 1-33. Øvereng, O. 1969: Geologiske undersøkelser i området Miall, A. D. 1983: Basin analysis of fluvial sediments. Special Liafjeli-Bjørnatrynet i Os. Unpublished Cand. Real. Thesis, Publications of the InternationalAssociation of Sedimentology University of Bergen, Norway. 248 pp. 6, 279-286.