Foraminiferal, Palynomorph, and Diatom Biostratigraphy And

Foraminiferal, palynomorph, and diatom biostratigraphy and paleoenvironments of the Torsk Formation: A reference section for the Paleocene-Eocene transition in the western Barents Sea

JENONAGY1, MICHAEL A. KAMINSKIZ, KARI JOHNSEN3and ALEXANDER G. MITLEHNER4 1. Department of Geology, University of Oslo, P.O. Box 1047 Blindern, N-0316 Oslo, Norway 2. Research School of Geological and Geophysical Sciences, Birkbeck College and University College London, Gower Street, London, WCIE 6BT, U.K. 3. Saga Petroleum as., Kjorboveien 16, P.O. Box 496, N-1301, Norway 4. Department of Geography, Edinburgh University, Drumrnond Street, Edinburgh, EH8 9XP, U.K.

ABSTRACT A microfossil-based biostratigraphic reference section is selected for the Paleogene Torsk Formation in the 7119/9-1 well, yocited in the Tromser Basin of the southwestern ~arentsSea. Benthic foraminifera, dinoflagellate cysts, and diatom infillings indicate an early late Paleocene to early Eocene age for the Torsk Formation on the eastern flank of the basin. A comparison of benthic foraminiferal last occurrences in the reference section with that established in neigh- bouring wells illustrates the potential use of these events for high-resolution stratigraphic correlation. Though differing in detail, the record of microfossil last occurrences in the western Barents Sea compares well with the biostratigraphy of the Central North Sea area. A morphogroup analysis of the Paleogene foraminiferal succession of the Tromser Basin reveals a batl~y- metric deepening of the depositional area culminating near the Paleocene-Eocene transition, followed by a shallowing.

INTRODUCTION region is of early Eocene age, the early Paleogene The Barents Sea covers a continental shelf area that biotic record can only be obtained from continental is of a comparable size to the entire North Sea, but margin wells. Fortunately, Paleocene to Eocene the subsurface geology of the area is constrained sediments have been recovered from a number of by fewer than 53 wells, mostly concentrated in the exploratory wells drilled in the Tromser and Ham- Hamrnerfest and Tromser basins (Fig. I). This area merfest basins, off the northern coast of Norway. has been subjected to intense Mesozoic and Ceno- These basins are situated just east of the oceanic zoic tectonic movements that have affected the areas of the Norwegian-Greenland Sea, and consti- configuration of seaways linking the North Atlan- tute the northernmost offshore basins of the North tic with the Arctic Ocean. Therefore, studies of the Sea / Norwegian Sea rift system explored by con- sedimentary and biotic record preserved in the ventional drilling. The main objective of this study Barents Sea basins are important for reconstructing is to assess the stratigraphic and paleoenviron- the history of faunal connections between the mental significance of early Paleogene benthic for- North Atlantic and Arctic oceans, and constraining aminifera, dinoflagellate cysts, and diatoms from a paleobiogeographic patterns in the Boreal Realm. key section in the western Barents Sea. This study Reconstructing the development and history is intended to complement the assessment of of marine connections through the Norwegian- microfossil assemblages from younger Cenozoic Greenland Sea is the prime objective of the ,,Arctic material carried out under the auspices of the ODP Gateways" legs of the Ocean Drilling Program. As ,,Arctic Gateways" initiative. the oldest oceanic crust in the Norwegian Sea Among 12 wells that we have examined from

In: Hass, H.C. & Kaminski, M.A. (eds.) 1997. Contributions to the Micropaleontology and Paleoceanogra- phy of the Northern North Atlantic. Grzybowski Foundation Specinl Publication, no. 5, pp. 15-38 Nagy, Kaminski, Johnsen and Mitlehner

Figure 1. Location map of the western Barents Sea, with main struct~~ralfeatures and locations of 6 of the studied wells. the western Barents Sea area, well 7119/9-1 was Greenland Sea, but the structural basins on the chosen as a reference section because of the pre- Barents Sea platform were formed as a result of sence of a thick Paleogene succession (1030 m earlier (Cretaceous to Paleocene) rifting phases. according to electric logs), rich microfossil assemb- The geology and structural development of lages, and the availability of continuous sampling. the Barents Sea has been the subject of numerous In addition, Knutsen et 01. (1992) published a seis- studies published mainly after drilling in the area mic line that crosses the well site and defined seis- commenced in the early 1980's. The subsurface mic units in this well, so it is now possible to more geology of the area based on seismic evidence has accurately assess the age and significance of regio- been discussed by R~nnevik(1981), RPrmevik & nal seismic reflectors. Altogether, 129 ditch cutting Jacobsen (1984), Roufosse (1987), and Faleide et 01. samples were analysed from this well, with few (1993a,b). Recently, Riis (1992), Cloetingh et nl. exceptions spaced at 10 m intervals. The biostrati- (1992), Walderhaug (1992), and Richardsen ct nl. graphy from well 7119/9-1 is compared with data (1993) have provided estimates of the amount of from nearby wells in the Hammerfest and Tromser Cenozoic upiift and erosion in the western Barents basins. Sea based bn geophysical and petrophysical evidence. The early Cenozoic sedimentary history of STUDY AREA the area has been discussed by Knutsen & Vorren Geological Setting (1991). The Barents Sea overlies an intracratonic area of The area of this study is located in the south- structural basins, platforms, and highs. In the west, western part of the ~arentsSea close to the conti- the area borders the oceanic Norwegian-Greenland nental margin. The region is divided into four main Sea with a passive shear margin marked by the basins: Bjarnerya, Troms~r,Hammerfest, and Nord- Senja Fracture Zone in the south and the Hornsund kapp, with intervening structural highs such as the Fault Zone in the north. The Barents Sea margin Senja Ridge, Loppa High, and Veslemery High (Fig. started to develop in the early Paleogene in 1). The Tromser Basin is connected with the Ham- connection with the opening of the Norwegian- merfest Basin over the Ringvassoy-Loppa Fault Foraminiferal, palynomorph, and diatom biostratigraphy of the Paleogene Torsk Formation 17

Complex. In an easterly direction, the Hammerfest 1993a,b). The top of the Torsk Formation is cut by basin is linked to the Nordkapp Basin over a shal- an erosional unconformity, and it is overlain by low threshold on the southernmost extension of Plio-Pleistocene glacial and glacial-marine sedi- the Bjarmeland Platform. The reference well cho- ments. sen was drilled on the eastern flank of the Tromser Large areas of the Barents Shelf east of the Basin. Senja Ridge were uplifted and subjected to mul- tiple phases of erosion, lasting to the early Plio- Stratigraphical Framework cene. As a result, the upper parts of the Torsk For- The Mesozoic and Cenozoic lithostratigraphy of mation (as much as 1500 m) have been removed, the western Barents Sea was formally outlined by particularly in marginal parts of the basins where Worsley ct 01. (1988). The Paleogene Torsk Forma- the formation thickness is strongly reduced. tion has a rather lnonotonous lithology consisting of light grey to greenish grey, essentially non- MATERIAL & METHODS calcareous claystones. Thin beds of siltstone or Well 7119/9-1 was a wildcat well drilled in the limestone occur rarely throughout the formation, summer of 1984 at 71°24'53.19"N, and and tuffaceous horizons are commonly reported in 19'49'43.26"E in 201 m of water, with Elf Aquitaine the lower part. The thickness of the formation Norge a.s. acting as the operator. Samples for this varies from 200-300 m in wells drilled in marginal micropaleontological study were provided by the areas of the Hamrnerfest Basin, to over 1 krn in the Norwegian Petroleum Directorate in Stavanger. central parts of the basin. The formation thickens to The material analysed consists wholly of washed over 2 km in the undrilled deepest parts of the ditch cutting samples ranging in dry weight from Tromser Basin. 25 to 45 grams. The micropaleontological samples were disintegrated by boiling in sodium carbonate, and washed over a 63pm sieve. Foraminifera and diatoms were picked from the >125pm fraction and mounted on cardboard slides. Palynological samples were analysed in accordance with standard techniques. All depths reported in this paper are quoted below rotary kelly bushing (therefore, a value of 226 m must be subtracted from the reported depth to convert to depth below sea floor).

RESULTS The succession of palynomorphs, diatoms, and benthic foraminifera in the 7119/9-1 well can each be subdivided into distinct assemblages based on the distribution of donlinant taxa, the occurrences of stratigraphically important species, and fluctuations in diversity and abundance (Fig. 3). The chronostratigraphy of the well is based primarily 011 palynomorph assemblages, which can be correlated with the zonations of the northern North Sea published by Schroder (1992) and Mudge & B~~jak (1994, 1996).

a. Palynomorph assemblages Figure 2 Lithostratigraphic scheme of Upper Cretaceous Altogether, 70 samples have been analysed for and Cenozoic deposits of the southwestern Barents Sea palynomorpl~swith a spacing of 10 m except in the (modified after Worsley tt a/. (1988) and Faleide et al. upper 400 m of the well, where the sample spacing (1 993)). is mostly 30 m. The studied interval contains 90 palynomorph taxa, and their stratigraphical distri- The basal contact of the Torsk Formation is bution is presented in Table 1. The p alynomoph unconformable, and represents an important succession is subdivided into six stratigraphically depositional hiatus encompassing the latest Creta- significant assemblages (Fig. 3). ceous and early Paleocene (Fig. 2). This hiatus can From the uppermost part of the formation (at be traced on seismic lines throughout the south- 330 m), a single sample was analysed which con- western Barents Sea (Rernnevik, 1981; Faleide et nl., tains a poor palynomorph assemblage. No age-dia- 18 Nagy, Kaminski, Johnsen and Mitlehner

gnostic species were recorded, and most of the rence of Alisocystn sp. 2 and D. oebisf~ldeizsis and palynomophs present are reworked from Creta- the high content of terrestrial material (pollen, ceous deposits, a feature that is typical of the Plio- spores, plant tissue fragments) indicates that the cene - Pleistocene deposits of the area. The palyno- interval belongs to the lower part of the lower morph succession from the 7119/9-1 well is dis- Eocene, and corresponds to the top of the lower cussed below in order from youngest to oldest. Eocene Subzone E2A of Mudge & Bujak (1994) in the northern North Sea. 1. Alisoc?ystn sp. 2 assemblage Drncodil~izr~izvnrielor~gitudzrril was recorded at 405 m and below. This species is cl~aracteristicof Interval: 405-720 m. the early Eocene. However, D. vnriclongitr~lliirri is thought not to coexist with Alisoc!ystn sp. 2, but in Age: early Eocene slightly younger sediments (Heilmann-Clawen, 1985). Younger early Eocene sediments are there- Samples from this interval contain rich and diverse fore thought to be present between 405 and 330 rn. palynomorph assemblages. The persistent occur- A certain number of species occurring in the rence of Alisoc!jstn sp. 2 (Heilmam-Clausen, 1985) assemblage are probably caved. The presence of and Deflnrzdren oebisfeldensis below 405 m suggest RIzornbodi~ziunzdrnco at 435 and 470 m suggests that an age no younger than early Eocene. This age sediments younger than the middle part of the assignment is supported by the presence of Middle Eocene (Costa & Mai~um,in Vinken, 1988) Chnrlesdowniefl edwnrdsii at 435 m and Cl~nrlesdozu- are present above 405 m. Svnlbnrdelln cooksonine was nien crnssim~iosaat 570 m. The persistent occur- recorded at 405 m. This species I~asa known strati-

Depth Age Benthic foraminifera Palynomorph Diatom Abundance/g ~~~~~~l~~~~ and Assemblages Assemblages ; dominantspecies -5 10 ---&-.-- .. . Rebculophragmium Stellanma spp. 1 amplectens assemblage Deflandrea oeb~sfeldensis 600 - H. excavatus. Alisocysta sp. 2 B muillcamerala. Dracod~niurnvarielongiludum ~f~$~I,"~~~~ K conversa. Charlesdo wniea edwardsii Craspedodiscus a,f, rnoe,leri EARLY Recuwoides

700 -

800 -

900 -

I000 - Rhizarnmina spp, Psammosphaera trinrlatensis

Alisocysta margarita Palaeoperidinium pyrophorum Cerodinium srwtum

1300 - Recllwoldes spp.. H. wailen. K. horrida. Arnrnosphaeroidma

1400 - Psammosphaera fusca acme .Palaeocystodinium H. rugosa. K. conversa. H. jawisi

Figure 3. Chronostratigraphy, foraminifera1distribution, palynomorph assemblages, and diatom assemblages in well 7119/9-1. Foraminiferal, palynomorph, and diatom biostratigraphy of the Paleogene Torsk Formation

graphic range spanning from late Eocene to early m. This species usually occurs in high numbers in Oligocene (Costa & Manum, in Vinken, 1988). In the middle part of the upper Paleocene, and this this assemblage there is also evidence of reworking event probably corresponds to dinoflagellate cyst from the Cretaceous. zone 4 of Heilmann-Clausen (1985). The common occurrence of Palncoperidiizzllri~pyrop110r11111 at 1370 2. Dcflaildrea oebisfelder~sisassemblage 111 probably corresponds to the upper part of dinoflagellate cyst zone 3 of Heilmann-Clause11 (1985), Interval: 720-890 m. which is equivalent to the middle part of the ,,P4" palynomorph zone of Mudge & Bujak (1996). Age: earliest Eocene 5. Palaeocystodirzi~~rnb~~lliforrire assemblage The consistent occurrence of Apcctodiniriri~ qlriri- q~~clntl~n~below 720 m and the presence of the Apec- Interval: 1420-1450 m. todir~i~lrr~11onron~orplz11ii1/paru~1ii~ complex at 720 m suggests an early Eocene age. The relatively com- Age: early late Paleocene 111on occurrence of Dcf7arzdrefl ocbisfeldrnsis at 720 m supports this age assignment. The top of the D. The LO of the species Palneocystodirrillril br~llifi~i.rrri~ oebisfeldrr~sisacme defines the top of the lowermost (=P. a~rstrnlirz~~rrrof Mudge & Bujak, 1996) at 1420 111 Eocene Subzoi~eElc of Mudge & Bujak (1994), and indicates an early late Paleocene age. The conunon correlates with the top of the Balder Formation in occurrence of P. b~llliforn~e,P. pyropl~orllnr and the northern North Sea. Isabelidir~illnl cf. ziiborgense below 1440 m supports The LO of Cerodirzi~~ri~zi~nrdenense is observed this age assignment. This interval corresponds to at 845 m and it occurs consistently below 850 m. dinoflagellate cyst zone 3 and the upper zone 2 of The LO of C. zunrderzcnse defines the top of the Heilmann-Clausen (1985). lowermost Eocene Elb Subzone of Mudge & Bujak The presence of Spongodiriiunr delitierlse at 1450 (1994). m may be due to reworking. This species is known to have its last occurrence in the earliest early 3. Apectodiriiurrr angustr~iiz assemblage Paleocene (Hansen, 1977). According to Schroder (1992) it ranges only to the top of the Ekofisk For- Interval: 890-970 m. mation in the southern Viking Graben.

Age: latest Paleocene 6. Spongodiriir~iizdelitieiise - Odor~tocliitiriacf. costntcz - Cleistospllaeridit~n~sp. 1 assemblage The LO of Apectodinillrir nllglrstuiiz at 890 m indicates the latest Paleocene ,,P6" palynomorph zone of Interval: 1460-1570 m. Mudge & Bujak (1996). The individuals reveal con- siderable morphological variability. Some speci- Age: late Campanian mens have less well-developed horns, and especially the apical horn displays variation in size. The An abrupt change in the organic content of the characteristic acme of Apectodiizi~rii~spp. was not samples takes place at 1460 m, with an increase in recorded in this well. the amount of coal particles and the occurrence of' abundant Cleistosphaeridium sp. 1 and common 4. Alisoc!ystn nlarprita assemblage Cl~ataizgiellaspy. Spoizgodiriiun~delitier~se is relatively coininon at Interval: 970-1420 m. 1460 m. This species is known to occur in the upper Campanian (Lucas-Clark, 1987), but is more cha- Age: middle late Paleocene racteristic of sediments of Maastrichtian and early Paleocene age (Hansen, 1977). The LO of O~ic~rrfo- Thc LO of Alisoc?ysta nrnrgnrita at 970 m together chitiria cf. costatn at 1460 m suggests ail age not with the occurrence of Pnlneoperidirii~~ii~pyropl~or~rir~ younger than the late Cainpanian or early at 1020 m suggests a middle late Paleocene age, Maastrichtian (Tocher, 1987). corresponding to the ,,P5" palynomorph zone of Mudge & B~jak(1996). Scluoder (1992) noted that b. Benthic Foraminiferal Assemblages the range of A. n~nrgnritnis limited to the lower- Samples examined from the uppermost part of the most part of the Lista Formation in the southern Torsk Formation (410-500 m) were barren of fora- Viking Graben. In the studied well, the species minifera. Ninety-three samples were analysed in Areo1iger.n cf. senorierisis displays an acme at 1000 the interval from 510 to 1470 m, mostly at n sample Nagy, Kaminski, Johnsen and Mitlehner

spacing of 10 m. A total of 84 species and open diachronous in offshore Norway, as it characterises 11omenclature taxa were recognised, all of which the late middle Eocene Zone NSR6 of Gradstein & are agglutinated (Table 2). The succession of fora- Backstrom (1996) in Haltenbanken and in the nor- miniferal events can be compared with the integra- t!~ern North Sea. It is rare in our samples. The ted microfossil zonation of the northern North Sea asseinblage is numerically doininated by Rlr~znr~r- and Haltenbanken areas recently established by niirzn spp., with Rec~ir.r~oidesspp., Knrrerlrlirzn spy., Gradstein et nl. (1994) and by Gradstein & Back- and Spiroylectnmrilir~a spectnbilis present in low strom (1996). The benthic foraminifera1 succession numbers. The whole interval is characterised by a in the well can be subdivided into six main taxonomic turnover, and a plateau in the species assemblages: range chart occurs near 810 m, coincident with a peak in the abundance of tub~~larforms. At this 1. Retic~rlophrngnli~mnrilplecterls assemblage level the LOs of several typically Paleocene to lower Eocene species are observed, including Snc- Interval: 510-710 m. cnniniir~n plncentn, S~~li~~rrarrirriirinsp., Knlnn~opsis ,yrzyboii~skii,and Glorrrospirn spy. Both diversity and Age: early Eocene abundance decrease to minimum values near the base of this assemblage. At the top of the Torsk Formation from 510 m to We did not observe any planktonic forainini- 700 m, both abundance and diversity are low. The fera or calcareous benthic foraminifera in the lower assemblage consists mainly of Hnplopl~rn~rnoid~s Eocene interval of the well, nor did we find any cscwntrrs and Biidnshevnrlln ~lllilticniiiernta.At 710 in calcareous specimens in adjacent wells in the there is a sharp increase in foraminifera1 abun- western Barents Sea. Calcareous assemblages dance, and a peak in the frequency of Reticlr- belonging to the Sribbotirln pntngoriicn Zone of Grad- loplzrng~rlilrnrninplccter~s (= the early Eocene mor- stein et nl. (1994) have been observed as far north as photype called Retic~ilophrngnlirrnl irzterrnedilrn~ Haltenbanken (Gradstein & Backstrom, 1996) and (Mjatliuk)) is observed. This assemblage also con- the outer Vmring Plateau (Hulsbos et nl., 1989). Eit- tains common Rliiznninliizn spp., Reclrruoides spp., her the northern boundary of the early Eocene Knrrerl~linn spp., Hnylophrngmoides excnvntlis, and planktonic foraminifera must have existed south of B~rdnshevarllnnrlrlticnnrerntn. The species R. iizternie- the Barents Sea area, or the deep waters of the dilri~l is the nominate taxon of the late early Eocene basin were too corrosive to allow the preservatioi~ to early nuddle Eocene Zone NSR5A of Gradstein of calcareous microfossils. & Backstrom (1996). 3. Hnplophmglrloides aff. c8ger.i assemblage 2. Spiroplectnninririn rinvnrronrln assemblage Interval: 890-1070 m. Inteival: 710-890 in. Age: late Paleocene Age: latest Paleocene to earliest Eocene From 890 m to ca. 1070 111 the diversity of benthic The LOs of many taxa are observed near 710 m, foraminifera is low, and the assemblage is doini- incl~tding that of Spiroylt~ctnl~iniir~nrznvnrrclnr~n, nated by tubular forms and a species of Hnplo- Arrrrrrospllncroidirln pserrdopnucilocirlntn, Hnplo- plirngr~~oidestentatively designated H. aff. ~>~q:~rri. ~)llrngi~ioideskirki, and Anrrrrodisclrs nlncileiitlrs. In the The LO of this species was observed at 910 m, but Central North Sea, the LO of S. linvarronnn is an foraminifera1 abundance near the top of this inter- important event within the lower Eocene Sllbbotinn val is especially low. We therefore arbitrarily pntngonicn Zone of Gradstein et 01. (1994). In the placed the top of this asseinblage at the abundance Haltenbanken area of offshore Norway, the LO of minimum at 890 m. The foraminifera in this inter- S. nnvnr~ronr~nis observed within the middle part of val are not age-diagnostic. Zone NSR5A of Gradstein & Backstrom (1996). Mudge & Bujak (1996) correlate the LO of S. nnvnr- 4. lieticulophrn~nrilrl~lpn~rprrn assemblage ronrrn with the top of the Ypresian. Interval: 1080-1160 m. The interval from 790 to 890 m is characterised by the occurrence of a diminuative form of Airlnro- Age: middle late Paleocene ~rror~gir~rrlirlnalrbrrtnc, a species first described from the North Sea and Labrador Shelf by Gradstein & Between 1080 and 1160 m, both the abundance and Kaminski (1989). This species is apparently diversity of the benthic foraminifera steadily Foraminiferal, palynomorph, and diatom biostratigraphy of the Paleogene Torsk Formation 21

increase downhole. This interval is characterised in the well. There is an abrupt downhole expansion by the consistent common occurrence of Reticlllo- in abundance at 1410 m. The assemblage is distin- phragmi~~nipaupera, an index taxon for the upper guished from overlying assemblages because it is Paleocene in the North Sea and Haltenbanken. It is dominated by species that agglutinate coarse mate- the nominate taxon of the late late Paleocene Zone rial for the construction of their test, such as Hyper- NSR2B of Gradstein & Backstrom (1996). Other a~nl~iinanigosa, Psalnrnosphaera filsca, and an un- important species include Haplophragnloidcs zualteri named species of Recllrvoides. Psamnlosphaem filscn and tubular forms belonging mainly to Rhiznm- increases in relative abundance at 1430 m, near the iilina. Spiroplcctainmina spectabilis is occasionally base of the Torsk Formation. Other common forms present in this interval in low numbers. The LO of in this interval include Reclrrvoidcs spp., Rhab- Hyperamnlinn rl~gosnis observed together with that damnzina spp., Karrerl~lina sp., and Aiilr?ro- of R. paupern at 1080 m. sphaeroidina pselldopallcilocnlata. Three distinctive species are restricted to this assemblage: Prae- 5. Spiro~7lectai1liilinaspcctnbilis assemblage cystarllminn sp., Cn~ldnniillinaOVL~~IL~~L, and Rzchakiizn epigonn. Interval: 1180-1400 m. Samples from the underlying Cretaceous Kviting Formation display further increased abun- Age: middle late Paleocene dance and diversity compared with the Torsk For- mation. The species Anllnodisc~isglabratl~s, Syirosix- A distinct maximum in the abundance of Spiroplec- moilii~ellnsp. 1, Hormosina z~elascocnsis,and Trocllairl- tanlriril~aspectabilis is observed in the interval from nliizn cf. s1ibves~c1r1nrisdisplay LOs at 1460 m, coin- 1180 to 1400 m. The last common occurrence (LCO) cident with the top of the Kviting Formation. of S. spectabilis occurs within the upper Paleocene However, these are taxa that are known to range in the North Sea, and is placed near the base of into younger stratigraphic levels in other areas of Zone NSR2B of Gradstein & Backstrom (1996). In the North Atlantic. The Upper Cretaceous index the studied well, this event occurs between the LO form Cnlidanlnzina glgantea was observed at 1470 in, of Alisocysta mnrpritn and the LC0 of Palneoperi- confirming a Campanian to Maastrichtian age for diiuiinl pyrophorzi~n(at 1370 m), which agrees with this level. its relative position in the RASC zonal model of Gradstein & Backstrom. One unusual feature of c. Diatom assemblages this assemblage is the large proportion of of juve- A total of 35 diatom taxa were identified from 70 nile forms of S. spectabilis. Other common taxa samples. Although poorly preserved (being infilled include Rhiznlizmilza spp., Alnnlosphaeroidilza pseu- and/or replaced by pyrite) and in relatively small dopaziciloc~ilata,Haplophraglnoides zualteri, and Reti- numbers (due to the effects of differential preserva- c~ilophragini~itn/7nlrpera. In the lower part of the tion bias towards more robust forms), enough st1.a- interval, single occurrences of the species Rzelmkiizn tigraphically important taxa were present to enable nlitzimn and Reticlilopllrngmoides jarvisi were obser- the Paleogene succession to be subdivided into five ved. stratigraphically significant assemblages (Fig. 3). In the studied western Barents Sea wells, the Diatoms occur sporadically in the sainples studied, species Aini~ioanitnrll tlzven~nt~rrnyi(the nominate with a major abundance peak developed between taxon of the mid Paleocene Zone NSR 2A of Grad- 850 and 870 m, at around the Paleocene/Eocene stein & Backstrom) is rare or absent. We therefore boundary. This correlates with a similar abundance selected an event that occurs regionally in the peak in the North Sea Basin. Diatoms are also rela- western Barents Sea to characterise this assem- tively abundant between 1110 and 1310 m. The dia- blage. Our S. spcctnbilis assemblage is roughly equi- tom succession is discussed below in order froin valent to the T. nitllz~cnmlrrmyiZone in the North youngest to oldest. Sea and Haltenbanken. 1. Stellnrinzn spp. assemblage 6. Psananos/7haera f~lscaassemblage Interval: 510-550 m. Interval: 1410-1450 m. Various morphologies belonging to the genus Stel- Age: early middle Paleocene larima occur in the uppermost diatomaceous samples studied. Although some of these are difficult The benthic foraminifera1 assemblage at the base of to identify to species level due to poor preserva- the Torsk Formation (1430-1480 m) displays stron- tion, their biconvex shape, small size (

3 - P6hocny Atlantyk Nagy, Kaminski, Johnsen and Mitlehner

characteristic of Sfcllari~~ln.Usually these are nor- the latter distinguished by their lenticular shape, mal (i.e. vegetative) cells of the species S. nlicrotrias, smooth surface and lack of a girdle. Auxospores of but sometimes resting spores are encountered, F. antiqua also occur, wluch are barrel-shaped with distinguished by having a depression at the centre a very wide girdle. Other species present in the of the valve face. The latter are a morphological assemblage include Stullnrilrzn sp. and Coscirzodisc~ls adaptation found in many diatom species, and are niorsinil~ls. formed in response to seasonal depletions of nutrients (see Round ct 01. 1990, pp. 37-40 for a fuller dis- 5. Tlmlassiosiropsis iuittialza assemblage cussion). Smaller morphologies bearing a strong resemblance to S. ~~licrotriasalso occur in some Interval: 1120-1350 m. samples in the lower Eocene interval. The late Paleocene interval is dominated by rela- 2. Stellarir~lai~licrotrias, Ferzestrclla arztiqlla, Craspedo- tively great numbers of the large, discoidal species disclrs aff. rrlocllrri assemblage Tllalnssiosiropsis zuittiann. This taxon is found in other areas during this interval, including the Rus- Interval: 560-710 m. sian Platform and the North Sea, and indicates fair- ly unrestricted connections between these areas. This assemblage is characterised by the continued The interval from 1350 to 1470 m is barren of dia- presence of Stellariirla nricrotrins, though in low toms. abundance; in addition to the first occurrence of the large, biconvex species Ferzcstrella arltiylla, also DISCUSSION in small numbers. Sporadic appearances of the a. Stratigraphic Correlation ,,doughnut-shaped" genus Crnspedodisclrs (which For the purposes of assessiilg the utility of the bear a resemblance to the species C. r~loelleri.)are assemblages observed in the 7119/9-1 well for also present in this assemblage. Poor preservation regional stratigraphic correlation, we compiled a precl~tdesspecies determination. In the interval graphic overview of foraminifera1 stratigraphic 720 to 810 in, diatoms are completely absent. events that were observed in six wells comprising an east-west transect from the Senja Ridge to the 3. Ferzestrcllo alrtiqua, Coscirzodisc~ls nlorsianus transition between the Ham~nerfestand Nordkapp assemblage basin (Fig. 1). For these purposes, only the last occurrences (LO) and last coinmon occurrences Interval: 850-960 m. (LCO) of selected common, cosmopolitan, deey- water benthic foraminifera were included (Fig. 4). Diatoms reappear at 820 111, sligl~tlyabove the Parts of the well sections analysed display low Paleoce~~e/Eoceneboundary as delimited by paly- abundance and diversity of benthic foraminifera. nomorphs. From 850 m, a rich diatom assemblage In spite of this, the strata reveal several important is present, dominated by Ferzestrclla antiqun. Other biostratigraphical events, such as the LOs of lietic- species include Stcllarir~lanricrotrins, the biconcave 1lloplzra~gtr1inn1arilplectclls, Spiroplecfnnlnlirln rza7mr.ro- species Coscinodiscns nrorsiarl~~s,the triangular arm, and of the dinoflagellate cyst Apcctodinilrnl taxon Trirlacria resirla, and sporadic occurrences of alzg~lst~lni.In the Paleocene part of the trailsect the Tllalnssiosiropsis wittinno. The latter is similar in LOs of Rcticlllophragnlinrzl pnlrpera, Hypcrarnrrrirln shape to Stellarii7ra rrlicrotrias but much larger, with rllgosn, and the LC0 of Spiroplectnrr~ir~irznspectnbilis an extreinely smooth surface and with a pro- are observed in all wells, whereas the LOs of Her.- nounced dimple at the centre of the valve. Other lilosilzn excclsa, Rzellakinn rrzirlinrn, Praec!/stanl~nir.lrr species include the benthic taxa Actiriopt!~cllusselza- sp., and Ca~ldanlr~zinagignntea are restricted to rills and A11lncodisclls hirtlls, indicative of shal- fewer wells. The events included in the transect lowing conditions. Diatoms appear only sparsely have also been recorded from in the Viking Graben between 970 and 1030 m. of the Central North Sea (i.e. Gradstein ct al., 15)94), and may serve as the basis for a wider foramini- 4. Fcrlestrella arztiqlla assemblage feral zonation of the lower Paleogene of offshort? Norway. Interval: 1040-1110 m. In the three westernmost wells, the LO of R. ar~zplecterzsis observed near the eroded upper con- Huge (~250pm) specimens of E arztiqlla dominate tact of the Torsk Formation, therefore its last strati- this assemblage, often characterised by the pre- graphic occurrence in this area is probably not cor- sence of a veneer of original silica on the surface. relative to its true extinction. The occurrence of S. Both vegetative cells and resting spores are found; rznvarroa~znnear the top of the studied interval in Foraminiferal, palynomorph, and diatom biostratigraphy of the Paleogene Torsk Formation

B Barren of microfossils v v Tuff horizon

Figure 4. Foraminifera1 last occurrences in an W-E transect of wells from the Tromsn and Hammerfest basins, plotted at a common depth scale (meters below rig floor) to give an impression of the subsurface geometry of the Torsk For- mation. The position of a distinctive tuff horizon and the LO of a dinocyst approximating the Paleocene/Eocene boundary is also indicated. the three western wells suggests correlation with lage downhole. The LO of a distinct, very coarsely the lower Eocene Si~bbotirza patagoizica Zone of agglutinated tubular species, here identified as H. Gradstein ct nl. (1988; 1994), although we did not rtlgosa, is closely associated with that of R. pallprra. observe any specimens of the nominate species of These events, together with the last cormnon occur- this zone, or any other calcareous benthic forami- rence of S. spectabilis are remarkably continuous nifera that are known to occur in the lower Eocene over the whole study area despite obvious varia ti- of the North Sea. ons in sedimentation and paleobathymetry. Howe- East of Block 7120, the Eocene sediments have ver, the earliest late Paleocene Psanzniosplzncr~~ been removed by erosion, and the Torsk Formation assemblage (indicated in Fig. 4 by the LO of IJlncl- penetrated by the two eastern wells is dated as c!/stn7?imi7za sp.) is not present in the two eastern- wholly within the late Paleocene according to paly- most wells, suggesting that the pre-Cenozoic topo- nomorphs. The top of the R. paupera asseinblage in graphy of the basin did influence the coinposition the studied wells is easily recognised by an abrupt of the benthic foraminifera1 assemblages 111 the change to more a diversified foraminifera1 assemb- North Sea region, Praccysta17~17rir?asp. is regarded to Nagy, Kaminski, Johnsen and Mitlehner

be a deep-water species (middle bathyal) in the among Eocene last occurrences than among events Paleocene, therefore, its absence from the eastern recorded from the Paleocene. However, in this case wells may be a result of ecological exclusion. the outliers can be partially explained by sampling A total of 74 stratigraphic last occurrences of bias. This is attributed to the small size of the stu- foraminifera were recorded in the nearby 7120/7-3 died samples resulting in low specimen numbers well in the western Hammerfest Basin, with 50 in many Eocene samples. events in common with those in the 7119/9-1 well. A second regression curve was calculated from These two wells give the best correlation of last oc- a subset of data points, indicated as shaded currence data observed between any two records squares in Fig. 5. These points represent species investigated so far from the Torsk Formation. found in the Troms~Basin, which are listed as stra- When the depths of all shared LOs are plotted in a tigraphically important in the Paleocene to ~luddle standard graphic correlation plot, a regression Eocene probabilistic stratigraphy of the Central equation can be calculated (Fig. 5). The regression North Sea by Gradstein et nl. (1994).The regression coefficient can be interpreted as a measure of the line constructed from the important North Sea spe- ,,reliability1' of the last occurrence events. The cor- cies alone does not differ greatly from the previous relation between the LOs in the two holes can best curve, but the regression coefficient is much higher be described by a linear regression. Polynomial (~~=0.91).This suggests that the North Sea benthic and exponential solutions of the regression equa- foraminiferal zonation of Gradstein ct nl. (1994) can tion did not yield markedly higher regression coef- give excellent stratigraphic results in the western ficients, suggesting that there were no observable Barents Sea. However, other points also fall close significant hiatuses or changes in sedimentation the correlation line as well, indicating that addi- rate between the two sites. In this solution, the tional index forms are also present in the Tromso- regression coefficient calculated using all the Harnmerfest Basin area. A complete discussion of observed LOs gave a value of ~~=0.71.Outliers the stratigraphic utility of faunal events is beyond from the regression line in all cases consist of rare the scope of this study, and must be based on detai- species. In a later stage of biostratigraphic analysis, led analysis of a greater number of well sections. these outliers are normally deleted from the data set. The ,,scattero in the data points is greater

1600

All species: y = 392,71 + 0,72892~ RA2 = 0,710 1400 d, \ Q, r 1200 - Q, ,000 .-c 5 Q 800 nQ,

600

400 200 400 600 800 1000 1200 1400 Depth in well 712017-3

Figure 5. Graphic correlatiol~of foraminiferal LOs and LCOs in Wells 7119/9-1 and 7120/7-3 in the Troms~and Ham- merfest basins. Axes values are the depths in meters BKB of events observed in both wells. Foraminiferal, palynomorph, and diatom biostratigraphy of the Paleogene Torsk Formation

b. Unconformities pension-feeding tubular forms, and the opportuni- The benthic foraminiferal and dinoflagellate cyst stic species I? filscn are found beneath the Western biostratigraphy investigated here provides signifi- Boundary Undercurrent (Kaminski, 1985). The late cant new constraints for the age and nature of Paleocene deep water conditions in the south- unconformities, which are expressed as important western Barents Sea appear to have developed seismic horizons in the western Barents Sea. Two during a longer time period, because during the seismic horizons were examined here in detail: Late Cretaceous the Tromsa and Sarvestsnaget basins continued to subside (Faleide cJtal., 199%). It Tlzc Bnsnl Tertiary: This is the ,,Sequence boundary is likely that fine-grained sediments bypassed the B" horizon of Rannevik (1981), which marks the area, and that the coarse agglutinated asseinl2lage top of the rotated fault blocks east of the contact be- is testimony to the presence of periodic bottom cur- tween oceanic and continental crust. Rnmevik rents that affected the sea floor. correlated this reflector with the oldest Tertiary sediments in Spitsbergen, and correctly regarded it Top Torsk Forrm7tion: The Torsk Formation is uncon- as representing the base of the upper Paleocene. He formably overlain by Plio-Pleistocene glacial ma- added that older Paleocene sediments may be pre- rine and glacial sediments of the Nordland Group. sent in the eastern part of the Barents Sea area. Seismic lines interpreted by Faleide ct al. (1993a,b) Faleide ct 01. (1993b) recogiused the regional nature show truncation of seismic reflectors within the of the ,,Base Tertiary" reflector, and noted that it Torsk Formation in an easterly direction, indicating constitutes a hiatus spanning the Cretaceous-Terti- that the top of the Paleogene succession has been ary transition. On the southwestern Loppa High clearly eroded, while tlucker and younger Paleo- this reflector defines an angular unconformity gene sediments are present in the west. Our results truncating Cretaceous strata (Knutsen & Vorren, confirm this interpretation. In the easternmost 1991). locality tl~etop of the Torsk formation is witl~inthe Our observations suggest that the ,,Base Ter- middle late Paleocene R. pnilpera Zone (se~zsziGrad- tiary" reflector in the basin areas indeed represents stein et nl., 1988), whereas in the west, sediments a submarine hiatus. Earlier studies of the south- belonging to the late early Eocene R. arnplect~izs western Barents Sea (e.g. Worsley et al., 1988) have Zone underlie the Cenozoic unconformity. suggested an early Campanian age for the under- The timing and causes of the Cenozoic erosion lying Kviting Formation, but based on the occur- is a matter of discussion. In the western part of the rence of Cnudnirziniizn gignizten and Spongodiizi~li~z Barents Sea, in the region of the Sarvestsnaget deltiense in our samples, the formation may include Basin and the Senja Ridge, uplift of rift flailks is strata as young as late Cainpanian to Maastrichtian associated with Paleogene continental breakup. in places. In the wells where we sampled both the The main phase of deformation took place during lowermost Torsk Formation and the uppermost initial breakup in the early Eocene, but some faults Kviting Formation, we observed little change in were active into early Oligocene times. Faleide c2t171. the paleobathymetric significance of the benthic (1993b) estimated that as much as 1 km of Paleo- foraminiferal assemblage on either side of the dis- gene sediments were removed by erosion in this conformity. Assemblages from both formations in area. This erosion is related to uplift along the Senja the Tromsa Basin and central part of the Hammer- Fracture Zone during Eocene time (Vignes, 1995). fest basin contain deep-water benthic foraminiferal This event created a western source area for solnc taxa such as Prnecysta~n~nina,Cnudnmminn, Pseudo- sediments in the Tromsa Basin. Erosion in the bolivinn, Rzclzakinn, and abundant tubular forms Hammerfest Basin was of greater magnitude, such as Rlziznnznziiza and Rhnbdanzmina, which indi- according to estimates from vitrinite reflectance cate a middle bathyal (or deeper) depositional and fission track analysis in the range of 1200-1500 environment. m. A study of the porosity, cementation, and oxy- In the Tromsa Basin and western part of the gen isotopes in the Jurassic St0 Formation in the Hammerfest Basin, the basal assemblage of the 7120/9-1 well (present depth = 1500 m) indicatecl Torsk Formation contains numerous tubular forms that the at least 1500 in of overburden has been and several other taxa that utilize coarse agglutina- removed at this locality (Walderhaug, 1992).Iiegio- ted material in the construction of their test. At nal uplift in the soutl~westernBarents Sea inay some localities, a maximum in Psanzi~zosplzacrafzisca have been associated with or caused by Oligocene is observed. The distinctive composition of the volcanism (Faleide ct al., 1991), a Miocene sei-level basal assemblage in the Torsk Formation invites fall (Reemst & Cloetingl~,1994), increased levels of comparison with areas affected by deep-sea bot- intraplate compression in the Neogene (Cloetingh tom currents. In the modern North Atlantic, high et al., 1990, 1992), and neotectonic activity resulting abundances of coarsely agglutinated forms, sus- from Pleistocene deglaciations (Karpuz et al., 1991). 26 Nagy, Kaminski, Johnsen and Mitlehner

c. Benthic foraminiferal paleoecology and paleo- vals with high benthic foraminifera1 abundance bathymetry and diversity reflect condensed l~orizons.The A striking feature of the foraminiferal assemblages nature of the benthic foraminifera1 abundance pa t- in the studied wells is the absence of calcareous tern in the 7119/9-1 well suggests cycles of higher taxa (both bentl~icand planktonic). The assem- sediment input, although changes in sediment tex- blages include significant proportions of deep- ture are not apparent trom gamma ray logs. water agglutinated foraminifera belonging to the Knutsen & Vorren (1991) and Knutsen ct a/. single-cl~ambered (tubular and globular forms) (1992) noted that the Loppa High served as a sedi- and multicl~amberedgroups. The low availability ment source area during the late Paleocene, and of carbonate in the southwestern Barents sea that in the northwestern part of the Tromsa Basin during the early Paleogene has probably produced prograding clinoforms are visible in seismic reflec- unfavorable conditions for the development of cal- tion profiles. The decrease in foranuniferal abun- careous faunas, as well as for the preservation of dance and diversity in this part of the section any calcareous microfossil tests originally present. (above the Spiroplectclr?~tiri~laspectabilis assemblage) T11e wholly agglutinated nature of the assemblages may reflect hig11 terrigenous influx related to and the presence of taxa that are known from bas- increased sediment supply from the north, which nal facies 111 the Central North Sea points to deep- caused conditions unfavorable for the fauna. water (middle bathyal) conditions in the Troins~ The functional morphology of benthic forami- Basin. Howevel; it is a well known fact that deep- nifera can provide important insights into the eco- sea organisms can be found living in shallower logy of the benthic environment. Foranuniferal test environments at high latitudes (Murray, 1895), so shape is thought to reflect differences in life posi- tlus estimate must be regarded as a lower limit. tion and feeding strategy ('Jones & Charnock, 1985). The accum~~lationrate of benthic foran~inifera Cl~angesin the relative proportions of the various at a given locality can change in response to morphogroups through time can reflect changes in fluctuations in water depth, the flux of marine environmental parameters such as water depth, organic matter to the sea floor, as well as the sedi- current strength, and organic flux to the sea floor. mentation rate (Herguera & Berger, 1991). The For this study, we have devised a n~orpho- foraminifera1 abundance has been measured quan- group scheme for classifying the benthic forainini- titatively in well 7119/9-1 (Fig. 3). The record dis- fera according to test morphology, and by analogy plays fluctuations of two orders of magnitude with modern faunas, for interpreting their pleoe- within the Torsk Formation. Fluctuations of this cology. Paleogene assemblages froin the Barents magnitude cannot be explained by changes in sur- Sea can be subdivided into four groups and nine face productivity and water depth alone, and more subgroups (Fig. 6). Forms in Group 1 (tubular taxa) likely reflect variations in the flux of terrigenous are normally found in higher abundance in bathyal sediments. In sequence stratigraphic terms, inter- to abyssal habitats, reaching an acine in the lower bathyal zone in the lnodern North Atlantic (Schroeder, 1986). Because they are epifaunal suspension feeders, they are adapted to life in areas of low organic flux. In the modern Norwegian-Greenland Sea, an abrupt shift from infaunal to predominantly epifaunal inor- phogroups is observed at a water depth of 1,500 m (Corliss & Chen, 1988). The infaunal Group 3, by contrast, is better adapted to life in areas of increased organic flux and hig-

Figure 6. Morphogrouy scheme and inferred paleoecology of Paleo- gene agglutinated forai~~inifera from the western Rarents Sea. Foraminiferal, palynomorph, and diatom biostratigraphy of the Paleogene Torsk Formation 27

M 2a M3a M3b M 1 M4 M2c globular, rounded elongate tubular flat planispiral keeled semi-infaunal planispiral infaunal 0 20 40 60 0 20 40 60 0 20 40 60 0 20 40 60 0 20 40 60 0 20 40 60 80

Figure 7. Relative proportions (in percent) of agglutinated foraminifera] morphogroups in the Paleogene of well 7119/9-1.

her nutrient supply. This group is often observed in the tubular morphogroup. We believe that the large proportions near the shelf/slope break in the maximum proportions of the tubular morpho- Gulf of Mexico, or in areas where the surface layer group at this level reflect the maximum extent of of the sea floor is affected by bottom currents the Paleocene transgression in the Barents Sea area, (Kaminski, 1985). and therefore represent a maximum flooding sur- We have undertaken a morphogroup analysis face in sequence stratigraphic terms. of the assemblages from the 7119/9-1 well, and Above this interval, a shallowing of the site is results are sl~ownin Figure 7. Taxa belonging to the indicated by the decreasing frequency of the tubu- tubular morphogroup la are common to dominant lar morphogroup 1 and its partial disappearance in tluoughout the studied interval, except in the the later part of the early Eocene. The keeled upper 170 in where they are rare or absent. High morphogroup 2c displays a maximum in the lower abundances of this group recalls the so-called and upper parts of the interval, expressing the first flysch-type agglutinated assemblages of the Paleo- phases of the transgression and the subsequent cene of the North Sea and Alpine-Carpathian regi- regression, respectively. It is of interest to note that ons. The frequency variations of the tubular mor- the infaunal rounded planispiral morphogroup 3a phogroup are interpreted as reflecting changes in attains maximum frequency in the upper part of the paleobathymetry of the site. In the lower part of the interval, corresponding to the early Eocene the section, the relative proportion of tubes increa- portion of the Torsk Formation. As a general rule, ses upsection, suggesting a gradual subsidence of infaunal organisms can only exist in large numbers the basin. Maximum water depth is represented by in areas where organic matter is buried in sufficient the interval between 900 and 1100 m, where the amounts. This observation agrees with the general proportion of this morphogroup reaches 70%. It is trend of increasing surface-water paleoproduc- natural to assume that the water depth during the tivity in the North Atlantic region during the latter latest Paleocene was middle bathyal or deeper. In part of the early Eocene, reflected by the onset of the seismic section published by Knutsen et 01. biosiliceous sedimentation (Berggren & Olsson, (1992, Fig. 11) a double high-intensity reflector is 1985). present between 900 and 1000 m which correlates The ,,deep infaunal" morphogroup 3b dis- precisely to our double peak in the proportions of plays variable but low abundance throughout the 28 Nagy, Karninski, Johnsen and Mitlehner

studied interval. Since this group normally reaches Gradstein, 1981), and the MacKenzie Basin of off- maximum proportions in eutrophic areas or in shore northern Canada (McNeil, 1990) have certain dysaerobic environments such as those found similarities with those from the Barents Sea, in par- beneath the oxygen minimum zone (Kaminski et a/. ticular the occurrence of Stellaritna species. 1995), we conclude that the bottom water in the The sporadic occurrence of diatoms in the sam- Tromsm Basin was not oxygen-deficient during the ples studied is related to periods of restricted water deposition of the Torsk Formation. Howevel; circulation as a result of lowered sea-level, in con- increased numbers of the genus Karrer~lliiln are junctioil with increased nutrient influxes leading to observed in the Psatimzos~~lzaeraassemblage at the diatoin blooms (Calvert, 1966; Tl~unellet ill., 1994) base of the Torsk Formation. This is consistent with in a relatively small and insolated basin. These the finding of elevated proportions of deep infauna periods often coincide with a decline in both spe- in areas disturbed by bottom currents (Kaminski, cies numbers and diversity of benthic foraminifera, 1985). The above patterns are complex, and ana- a pattern also found in the North Sea (Mitlehi~er, lysis of additional sections is needed to provide a 1994; 1996). This phenomenon can be attributed to more con~pletepicture of the subsidei~cehistory of periods of high surface water productivity, with the western Barents Sea. increased nutrient levels leading to diatom blooins and eutrophication in the upper water column, and d. Diatom paleoecology oxygen depletion of the lower water column The most commonly occurring diatom taxon in the (Codispoti, 1989). Conversely, increased numbers samples studied is Stellarirr~nnlicuotrins, a species of benthic foraminifera coincide wit11 a general found intermittently through the entire Paleogene. absence of diatoms, a situation wlucll signifies It occurs in all of the samples from 510 to 710 m, improved water circulation and increased oxyge- and between 845 and 1240 m. This species also has nation in the lower water column. an intermittent stratigraphic distribution in the A particularly significant period of increased Paleogene of the North Sea (Mudge & Bujak 1994; diatom productivity occurred at the Paleocene- Mitlehner, 1994). Some of the diatom taxa recog- Eocene transition, and is represented in well nised in well 7119/9-1 are more stratigraphically 7119/9-1 by an abundant and diverse diatoin restricted. Tlu7lassiorir(1psis wittinna occurs more or assemblage. The presence of benthic species less continuously through the upper Paleocene sec- (Actinopt!yclzlcs serzilritls, Auk7codisctrs sp.) indicates tion, but is absent thereafter; whilst appearing spo- that there were shallow neritic conditions in the radically between 560 and 1100 m is the large (up vicinity. A similar diatom abundance peak is recor- to 300 prn diameter) species Ferzestrella antiqlln (= ded in the North Sea at the same interval, and it is Coscinodiscrrs sp. 1 of Bettenstaedt et nl., 1962), a likely that similar mechanisms, involving a taxon which marks the base of the Eocene in the lowered sea-level with increased influxes of water- North Sea but which has a longer range in the borne and windblown terrigenously and volcani- Barents Sea, extending from the upper Paleocene cally derived nutrients to a restricted basin led to through the lower Eocene. It has a similar range in the proliferation of diatoms at that time. the Paleogene of the Urals and Volga Basin (Glezer et ill. 1974), suggesting a connection of the Barents CONCLUSIONS Sea with the Turgai Sea which extended across the Among the six wells located along a transect from Russian Platform during the Paleogene. Other spe- the Senja Ridge through the Troms~lBasin to the cies co-occurring with E nntiqua include Coscin- Harnrnerfest Basin, well 7119/9-1 is chosen as a odisczis rnorsznnlcs and Trinncrin reginn. This assem- stratigraphic reference section. Analysis of foraini- blage as a whole is very similar to coeval diatoin niferal, palynomorph, and diatom assemblages in assemblages from the North Sea area (Hughes, this well in the Troinsm Basin enable a reliable 1981; Malm et nl., 1984; Homam, 1991; Mitlehner, determination of the chronostratigraphy of the 1994) as well as the Arctic Ocean (Dell'Agnese & Torsk Formation. A total of 86 palynomorpl~,84 Clark, 1994), although there are fewer sinularities foraminiferal, and 35 diatom taxa have been found. to the Paleogene assemblages sporadically reco- The succession of late Paleocene to early Eocene vered by DSDP Leg 38 from the less restricted Nor- assemblages compares well with the North Sea wegian Basin (Schrader & Fenner 1976; Dzinoridze zonations of Mudge & Bujak (1994, 1996), Grad- rt nl., 1978). This assemblage similarity documents stein et nl. (1994), Gradstein & Backstrom (1996), the beginning of the early Eocene transgressive and Mitlehner (1994). The stratigraphic occurren- phase, related to a worldwide warming event ces of several foraminiferal and dinocyst last occur- which occurred in all areas (Rea et nl., 1990; Robert rence events are consistent in wells across the Senja & Chamley, 1991). Other pyritised diatom assemb- Ridge - Hammerfest Basin transect, suggesting that lages, recovered from the Labrador Sea (Thomas & the biostratigraphical record of the 7119/9-1 well Foraminiferal, palynomorph, and diatom biostratigraphy of the Paleogene Torsk Formation

may serve as a standard for stratigraphic correla- gisk Tidskrift, 72,229-235. tion in the southwestern Barents Sea. Micropaleon- Codispoti, L.A. 1989. Phosphorus vs. nitrogen limitation of new and export production. In: Berger, W.H., tological and palynological evidence suggest that Smetacek, V.S. & Wefer, G. (eds.), Productivity of the the top of the Kviting Formation may be as young Ocean: Present and Past, 377-394, Wiley & Sons, Chi- as late Campanian to Maastrichtian. Chester. The foraminiferal assemblages of the studied Corliss, B.H., & Chen, C. 1988. Morphotype patterns of Norwegian Sea deep-sea benthic foraminifera and wells consist exclusively of agglutinated taxa, ecological implications. Geology, 16, 716-719. suggesting reduced availability of dissolved cal- Dell'A ese, D.J. & Clark, D.L. 1994. Siliceous microfos- cium carbonate in the depositional basins. The sils Emthe warm late Cretaceous and early Cenozoic composition of the benthic foraminifera1 assembla- Arctic Ocean. Ioi~riznlof Paleontology, 68,31-47. ges at the base of the Torsk Formation and in the Dzinoridze, R.N., Jouse, A.P., Koroleva-Golikova, G.S., Koslova, G.E., Nagaeva, G.S., Petrushevskaya, M.C. & uppermost part of the underlying Kviting Forma- Strelnikova, N.I. 1978. Diatom and radiolarian Ceno- tion reflects a deep-water (middle or upper zoic stratigraphy, Norwe ian Basin. DSDP Leg 38 bathyal) depositional environment. There is no Initial Reports, Deep Sen Drikiinx Project, Srpplei~~e~zt,38- paleontological evidence of a shallowing of the 41,289-427. environment in connection with the ,,basal Tertiary Faleide, J.I., Gudlaugsson, S.T., Eldholm, O., Myhre, A.M., & Jackson, H.R. 1991. Deep seismic transects unconformity" in the Tromsa Basin. Analysis of the across the sheared western Barents Sea-Svalbard con- benthic foraminiferal morphogroups in the re- tinental margin. Tecto7?opl~ysics,189, 73-89. ference section indicates a deepening of the basin Faleide, J.I., VZr,E.,, & Gudlaugsson, S.T. 1993a. Late until the latest Paleocene, followed by a shallowing Mesozoic - enozoic evolution of the southwestern Barents Sea. In: Parker, J.R. (ed.) Petrolen~r~Geology qf of the basin in the early Eocene. As with the benthic northzuest Ellrope: Proceediizgs of the 4th Confereilce, 933- foraminifera, the Paleogene diatom assemblages 950, Geological Society, London. recovered show marked similarities to those pre- Faleide, J.I., VdpE.,, & Gudlaugsson, S.T. 1994. Late sent in the North Sea, although there are a number Mesozoic - enozoic evolution of the southwestern Barents Sea in a regional rift-shear tectonic setting. of differences in the stratigraphic ranges of some Marine and Petrolel~nzGeology, 10, 186-214. taxa. Glezer, Z.I., Jouse, A.P., Makarova, I.P., Proshkina- Lavrenko, A.I. & Sheshukova-Poretskaya, V.S. 1974. ACKNOWLEDGEMENTS The Dintoilzs o the USSR Fossil and Recent. Vul. 1. Aka- demii Nauk fSSR, Botanical Institute, Leningrad, 403 We thank the Norwegian Petroleum Directorate pp., 16 figs., 20 tbs., 93 pls. (In Russian). and STATOIL for kindly providing samples. We are Gradstein, EM., & Backstrom, S. 1996. Cainozoic biostra- grateful to the STATOIL-VISTA Programme for tigraphy and palaeobathymetry, northern North Sea funding the project. Mufak Naoroz kindly assisted and Haltenbanken. Norsk Geologisk Tidskrift, 76, 3-32. with the sample processing. The manuscript bene- Gradstein, F.M., & Kaminski, M.A. 1989. Taxonomy and fited from comments by Geoff Eaton, Jan-Inge biostratigraphy of new and emended species of Ceno- zoic deep-water a glutinated foraminifera from the Faleide, Felix Gradstein, and Ann Holbourn. This Labrador and ~ort%Seas. Miciopaleo~zlology,35, 72-92. is contribution no. 49 of the Deep-water Agglutina- Gradstein, EM., Kaminski, M.A., & Berggren, W.A. 1988. ted Foraminifera Project. Cenozoic foraminiferal biostratigraphy of the central North Sea. Abl~andl~~ngender Geologischer~ Bl~i?desn~l- stalt, 41, 97-108. REFERENCES Gradstein, F.M., Kaminski, M.A., Berggren, W.A., Kristi- Barron, J.A. 1993. Diatoms. In: Lipps, J.H. (ed.) Fossil Pro- ansen, I.L., & D'Ioro, M.A. 1994. Cenozoic biostrati- knryotes and Protists, 155-167, Blackwell, Boston. raph of the North Sea and Labrador Shelf. Micrupn- Berggren, W.A., & Olsson, R.K. 1985. North Atlantic Eontojdyy, 40 (Supplement), 1-152. Mesozoic and Cenozoic paleoceanography. In: Vogt, Hansen, J.M. 1977. Dinoflagellate stratigraphy and echi- P.R. & Tucholke, B.E. (eds.). The Geology of North Ame- noid distribution in upper Maastrichtian and Danian rica, voll~llzeM, The North Atlantic Region, 565-587, deposits from Denmark. Blllletin of the Geological Geological Society of America. Society of Deiznznrk, 26, 1-26. Bettenstaedt, F., Fahrion, H., Hiltermann, H., & Wick, W. Heilmann-Clausen, C. 1985. Dinoflagellate stratigraphy 1962. Tertiar Norddeutschlands. In: Simon, W. & Bar- of the upper Danian to Ypresian in the Vibor 1 Bore- tenstein, H. (eds.) Arbeitskreis del~tscllerMikropnlnonto- hole, central Jylland. Dee,imrks Gmlogiske Un&rsokelsi,, logen: Leitfossilien der MikropnlZiontologie, 432 pp. & pls., Series A7, 1-69. 2 vols. Gebriider Borntrager, Berlin. Herguera, J.C., & Berger, W.H. 1991. Paleoproductivity Calvert, S.E. 1966. Accumulation of diatomaceous silica from benthic foraminifera abundance - glacial to post- in sediments of the Gulf of California. Bl~lletinof the glacial change in the West-Equatorial Pacific. Geology, Geological Society of Anzerica, 77, 569-596. 19, 1173-1176. Cloetingh, S., Gradstein, F.M., Kooi, H., Grant, A.C., & Homann, M. 1991. Die Diatomeen der Fur-Formation Kaminski, M.A. 1990. Did plate reorganisation cause (Alt-tertiar) aus dem Limfjord-Cebiet, Nordjutland/ rapid late Neogene subsidence around the Atlantic? Danemark. Geologiscl~esInlzrbuch, Hannover, Reihe A, Iolfrnnl oftlze Geologic01 Society of London, 144,43-58. 123,285 pp., 57 pl. Cloetingl~,S., Reemst, P., Kooi, H., & Fanavoll, S. 1992. Hurhes, M.J. 1981. Contribution on Oligocene and Intraplate stresses and the Post-Cretaceous uplift and Ikcene microfossils from the southern North Sea. 111: subsidence in northern Atlantic basins. Norsk Grolo- Neale, J.W. & Brasier, M.D. (eds.) Microfossils fro111Fos- 30 Nagy, Kaminski, Johnsen and Mitlehner

sil nizd Recrnt Shelf Srns, 186-294, Ellis Honvood, Chi- M~tdge,D.C., & Bujak, J.P., 1994. Eocene stratigraphy of chester. the North Sea Basin. Mnrinr nlzd IJctroler~iirGco/ogy, 11, lacqut, M. & Thouvenin, J. 1975. Lower Tertiary tuffs 166-181. and volcanic activity in the North Sea. Irz: Woodland, Mudge, D.C., & Bujak, J.P., 1996. An integrated stratigr'l- A.W. (ed.) Petrokrrn~nizd thc' Cor~tiiir~ntnlSlirlf c$ Nor- hy for the Paleocene and Eocene of the North Sea. I~I: tlriaest Elrropc, VoI. 1: Geolosy, 455-465, Elsevier, Bar- Rnnx, R.w.o,B., Corfield, R.M., ~r Dunay, R.E. (eds.) king. Correlntioi~o thcv Enrly Pnleoge~zein Nurtlizuest EIIIZI;J('. Jones, R.W., & Charnock, M.A. 1985. ,,Morphogroups" of Geological 4ociety Special Publication, 101, 91-113. agglutinatin 7 foraminifera, their life positions and Murray, I., 1895. A summary of scientific results. li~y7or.ts feeding habits and potential applicability in o tlie Scientific Reslrlts oftlre HMS Cllnlle~zgrr,1608 pp. (pa1eo)ecological studies. Revue de Poliobiologic, 4,311 - Jyre & Spattisw~nde,London. 320. Rea, D.K., Zachos, J.C., Owen, R.M. & Gingerlich, P.D. Kaminski, M.A. 1985. Evidence for control of abyssal 1990. Global change at the Paleocene - Eocene bo~uid- ~lutinatedforaminifera1 community structure by ary: climatic and evolutionary consequences of tecto- %strate disturbance - results from the HEBBLE area. nic events. P~~lneogeo~yroplry,Pnl~ieoclii~zntolo~~~, Pnlnc!ocJ- Marine Geology, 66, 113-131. colosy, 79, 117-128. Karninski, M.A., Boersma, A,, Tyszka, J., & Holbo~un, Reemst, P., '., Cloetingh, S. 1994. Tectonostratigraphic A.E.L. 1995. Response of deep-water ag lutinated for- modelline of Cenozoic uulift and erosion in the south- aminifera to dysoxic conditions in the &lifornia Bor- western krents Sea. ~n;inc*arid Petroleir~nGcolog!/, 11, derland basins. Grz!/boicski Foli~idntiorl Sprcinl Piiblicn- 478-490. tiolz, 3, 131-140. Richardsen, G., Vorren, T.O., & Tmrudbakken, B.O. 1993. Kaminski, M.A., Gradstein, EM., & Berggren, W.A. 1989. Post-Early Cretaceous uplift and erosion in the Paleogene benthic foraminifer biostrati raphy and southern Barents Sea: a disc~~ssio~~based on analvsis paleoecology at Site 647, southern ~abrszorSea Pro- of seismic interval velocities. Norsk Geologisk Tidskrift, crc.llin

Nordmela Formations in the Hammerfest Basin - a 117: Dalland, A., Worsley, D., & Ofstad, K. (eds.) A diagenetic approach. Norsk Gcologisk Tidskrift, 72, 32 1 - litl~ostratigraphicscheme for the Mesozoic and Ceno- 323. zolc succession offshore mid- and northern Nor~voy Worsley, D., Johansen, R., & Kristensen, S.E. 1988. The NPD B1rl'ehl',41 42-61. Mesozoic and Celiozoic succession of Troms~flaket. 32 Nagy, Kaminski, Johnsen and Mitlehner

Table 1. Palynornorph distribution in samples from Well 7119/9-1. Depth in metres.

330xxxrxxr 405Xx r I: X~~XX~I:XI)IXXXX 435 X X XI X * X IXXXXXI 470 r )I I~XXXX~~XXII~IXXI~XX~ 510 X X x x x x XXX x )L x XIXX XXX 570 X r XXXXXXI x XIX x x ~~~x~xlt~x 610 X x I XXXX Xx X XX 670 X X r x XI X XX XI* XXXXX 710 r x x X X I )I 720 r x xxxxx X x XXX XXX XI 730 r x X IXX XXX x X x K 740 X x X Xxxxx X IX X 1 X x r 750 X x I. XXX X XI 760 x x x l X I X Xxx X XI X 770 X 1 XI XXXX*XX xrr x x X r XX< xii x 780 r x x r XXXXX X XI X X X )I 790 1 i x X XXX XlXXX X x X XXLX 800 I x X X 810 x x x x x I * X 820 X x x I X X X I X X 830 X x x * X XX * X 8.15 X )I X K XXXIX XII X X I X 850 x x x x xxx XX XI )I X 860 X x x x * X X X XI: X I X X 870 X r X X x x 880 x x X X XX XX X XXX 890 X x X I X* XI I X XX X X 900 X x X X X X X X xx xxx * X 910 X x x I 920 x x x x x x X XXX I X x x R30 X x K x * X x x xx x x 940 X x x I x X X I 950 X x x x x X X x x 960x1~~x x XX xx XXX x x XX Xxx 970 X x X XX* r xx XI?. 980 x x I x XI xxx 990 X x x x 1000 x x X X xx r X XX XXX 1010 X X x x x x x 1020Xx X X * 1030 x r x x X x x x K xxr 1040 x r x x XIEX 1OGOx x x x x x x XI x X 1070X x x X X x x XXXIX 1080~x r x X x XXX 1090 r x x x x x 11OOX" X X xxrx *X I 1120 X x X X 1150 x r r 1180 X x x x XEX Xxx X X 1240 x x x XI X X X X * X X 1270 X x x 1330 x x x 1370 X x 1380~x x xxx x XI 1390 x x r r xxx x x XXX x 14ooXxx~ x x x 1410 X X x 1420 X x x x 1430 r x rXX 1440 X x x x 1450 X x r 146" X x r x X X 1470 X x 1480 x r X X x x X x 1490 x x 1500 X X Foraminiferal, palynomorph, and diatom biostratigraphy of the Paleogene Torsk Formation 33

Table 1 (cont.). Palynomorpl~distribution in samples from Well 7llY/Y-1. Depth in metres.

x x x X X X X X X X XX x X X xxx XXXX x X X X x x x x r XXI XX X X X x x X iI XXXX x x

X x X XXXXX XX liI XXI X XXXX X x X X K X X IX I XX

X X X X XXXX XX XX t X X X x X XXX X X X rx X iI XX X XXXX

x x x XrxX li X X X x X XXXXX XXXXX xr xx X x X11X X XX XXXX 1418C XX X XX X I491 i X XXX xxx x x 150( Nagy, Kaminski, Johnsen and Mitlehner

Table 2. Benthic foraminifera1 distribution in samples from Well 7119/9-1. Depth in metres. Foraminiferal, palynomorph, and diatom biostratigraphy of the Paleogene Torsk Formation

Table 2 (continued). Benthic foraminifera1 distribution in sainples from Well 7119/9-1. Depth in metres.

38 Nagy, Kaminski, Johnsen and MitEehncr

Plate 3. Strati~r~igl~icnI1jnltlcwtant llmthic ftjrarninifera irr.rm t2ttnll 711 911. 1. Nofhir~rol}lrst~? (Cr~yL~t~~~~ki). x.77. 2. If!~f-'ri?ir~trlil~,~rlly~lw Verdeni~rs k \511 Yinte, 13?0m, x28. 3. H!/~r,.mtrrt?rirtnrr~~yosir, s15, 3. Ctrirdr~r~rr~~;/!~r~~s~.t~lsn Dyia- Z~II~1, 13~0111, ~33. 5. Knh~tiropjkcyr:!jlx?~ivkii (r)ylazanka), 134'). ~140.6. S/~ir~>/>/~ct~t~~~~)ii;in.y~t~ct,r/~il~s(Grzyht>~vski) ~i~egalorpliae~.ic,1370111, as[). 7. S~1iroplt~c/nrrr;rr31rniprcmljilix niicfi~spaeric. 1370111, YXD. 8. Sflrrorir~~lnjtl:lcr,rt tn (Cirsybtl- 1vr; kil, I32th11.~78. 9. II:fqlrokrr~fl c-!ligo~ln (IZ7ctld k), 1380m. x70. 10. lizr~ltnkiuorirjr~i!irtr Cushrna~~ A ltenz, 1370n1, v 145. 11. Hi~~~lo~~l~~-i~y:~~~iit