1. Introduction
The Porcupine Seabight (Fig. 1) initially gained fame at the end of the nineteenth century for its special deep-water habitats (Thomson, Reference Thomson1873; Le Danois, Reference Le Danois1948). Hovland, Croker & Martin (Reference Hovland, Croker and Martin1994) were the first to publish seismic profiles picturing mound structures up to 200 m high and more than 1500 m long. Very-high-resolution seismic profiling has confirmed the presence of large seabed (carbonate) mounds in the Porcupine Basin (Henriet et al. Reference Henriet, De Mol, Pillen, Vanneste, Van Rooij, Versteeg, Croker, Shannon, Unnithan, Bouriak and Chachkine1998; De Mol et al. Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002; Huvenne, Croker & Henriet, Reference Huvenne, Croker and Henriet2002). The mounds occur in three well-delineated provinces, featuring distinct morphologies: the Belgica mounds on the eastern flank (De Mol et al. Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002; Van Rooij et al. Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003), the Hovland mounds in the north (Hovland, Croker & Martin, Reference Hovland, Croker and Martin1994; De Mol et al. Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002) and a large number of buried Magellan mounds further to the northwest (Huvenne, Croker & Henriet, Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002; V. Huvenne, unpub. Ph.D. thesis, Ghent Univ. Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003). The mounds have been defined as deep-water coral banks where colonies of dead and living corals (namely Lophelia pertusa and Madrepora oculata) interact with sediment drifts in a dynamic environment (De Mol et al. Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002; Foubert et al. Reference Foubert, Beck, Wheeler, Opderbecke, Grehan, Klages, Thiede, Henriet, Freiwald and Roberts2005). According to De Mol et al. (Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002) and Van Rooij et al. (Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003), mound growth probably started during Pliocene times after periods of erosion or non-deposition. The mounds reach heights of up to 250 m and widths of up to 5 km and are enclosed by sediment drifts (De Mol et al. Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002; Huvenne, Croker & Henriet, Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002; Van Rooij et al. Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003). Despite the diverse and large seismic datasets gathered over more than 20 cruises during the last decade, the knowledge of the internal structures, the initiation and growth of these impressive sea-floor features within the Porcupine Seabight and their interaction with sedimentary processes occurring along the continental margins remained elusive. During Integrated Ocean Drilling Program Expedition 307 in May 2005 aboard the JOIDES Resolution, Challenger mound, a prominent mound structure covered with dead cold-water coral rubble in the Belgica mound province, was drilled to the moundbase to unveil the internal structure of the mounds in Porcupine Seabight (Site U1317, Fig. 1). To constrain the stratigraphical framework of the slope–mound system and identify and correlate erosional surfaces observed in seismic sections with global palaeoceanographic events, two further sites, located down- and upslope from Challenger mound (sites U1316 and U1318, Fig. 1) were drilled. This study focuses on IODP Site U1318 (51°26.16′N, 11°33.0′W) located in the eastern part of the Porcupine Basin at the upper slope edge of the continental margin of Ireland.
Location of drill sites IODP 1318A, 1318B and 1318C. Inset: location of the Porcupine seabight.
Dinoflagellate cysts and other organic-walled phytoplankton are frequently used as a tool for unravelling biostratigraphical relationships in the Neogene of the North Atlantic realm (Costa & Downie, Reference Costa, Downie, Montadert and Robert1979; Brown & Downie, Reference Brown, Downie, De Graciansky and Poag1984; Head, Norris & Mudie, Reference Head, Norris, Mudie, Srivastava, Arthur and Clement1989a,c; de Verteuil & Norris, Reference De Verteuil and Norris1996) and for the reconstruction of the palaeoenvironment (Louwye, Head & De Schepper, Reference Louwye, Head and De Schepper2004; Piasecki, Reference Piasecki2003; Versteegh et al. Reference Versteegh, Brinkhuis, Visscher and Zonneveld1996; Warny & Wrenn, Reference Warny and Wrenn2002). This paper details the depositional history and palaeoenvironment in the eastern Porcupine Seabight through a high-resolution dinoflagellate cyst analysis of the Neogene sediment drift sequence at IODP Site 307 in holes U1318B and 1318C. An assessment of the age of the unconformity on which the Belgica mounds are rooted is given, together with a reconstruction of the palaeoenvironment during early Neogene times. Furthermore, highest and lowest occurrences of dinoflagellate cysts are linked to the palaeomagnetic record and provide new absolute ages for dinoflagellate cyst events in the North Atlantic realm.
2. Palaeoceanographic and geological background
The Porcupine Seabight forms a deep embayment in the Atlantic shelf off the southwestern coast of Ireland and is enclosed by shallow platforms, consisting of Precambrian and Palaeozoic rocks: the Slyne Ridge to the north, the Irish continental shelf to the east and Goban Spur to the south (Fig. 1). Only a relatively small opening towards the deeper North Atlantic Ocean is present to the southeast. The underlying structure of the Porcupine Basin is a Middle to Late Jurassic failed rift of the proto-North Atlantic Ocean (Naylor & Shannon, Reference Naylor and Shannon1982; Moore & Shannon, Reference Moore and Shannon1992). The centre of the basin is filled with 10 km of sediments deposited during the Cenozoic post-rift period (Shannon, Reference Shannon1991). Recent sedimentation is mainly pelagic to hemipelagic, although probably reworked foraminiferal sands can be found on the upper slope of the eastern continental margin. The main sediment supply zone is located on the Irish and Celtic shelves, whereas input from Porcupine Bank seems to be rather limited (Rice et al. Reference Rice, Billet, Thurston and Lampitt1991).
High-resolution seismic profiles through the Belgica mound province perpendicular to the slope reveal that the coral banks are rooted on a continuous but irregular erosional surface (De Mol et al. Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002). In the mound area, this erosional surface forms the boundary between seismic units P1 and P3 (Fig. 2) and reflects a deeply incised substratum. The seismic facies of the lowermost unit P1 is characterized by gentle basinward-dipping parallel strata (Van Rooij et al. Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003). Within a certain depth interval, however, sigmoidal deposits are observed (De Mol et al. Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002), interpreted as upslope-migrating sediment waves within a sediment drift unit (Van Rooij et al. Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003) (Fig. 2). Unit P3 consists of drift deposits of unspecified late Neogene age (Van Rooij et al. Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003). Seismic stratigraphical analyses suggest that the Belgica mounds were already in place before deposition of this unit (Van Rooij et al. Reference Van Rooij, Blamart, Kozachenko, Henriet, Viana and Rebesco2007). The latter authors furthermore state that the mounds were large enough to influence the intensity of the currents and consequently the sedimentary pattern.
Seismic profiles of the Belgica mound province (for location of the profiles, see Fig. 1), with location of the drill site 1318B, seismic units P1, P2 and P3, and unconformity reflectors RD3, RD2 and RD1 (dashed lines – unconformity reflectors; dotted line – upper limit of drift deposits not observed in U1318B; dashed line with question mark indicates location of a yet unmapped unconformity). Seismograms: Jean-Pierre Henriet, Renard Center for Marine Geology, Ghent University, Belgium.
The IODP Hole U1318B investigated penetrated seismic units P1, P2 and P3. Due to the upslope position of the site, only the upper part of unit P3 was cored, leaving out the base of the drift unit (Fig. 2). Unit P2 is intercalated between units P1 and P3 and is an acoustically transparent layer characterized by low-amplitude reflectors (De Mol et al. Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002; Van Rooij et al. Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003). This unit is absent downslope from Hole 1318B. The upper boundary of unit P2 appears erosive and is, according to Van Rooij et al. (Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003), related to the intra-Neogene margin-wide erosive event RD1. This event eroded units P2 and P1 downslope from Hole 1318B and modelled the palaeotopography before deposition of unit P3 and the onset of mound growth.
Dating of erosional events by De Mol et al. (Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002), Van Rooij et al. (Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003) and Van Rooij et al. (Reference Van Rooij, Blamart, Kozachenko, Henriet, Viana and Rebesco2007) mostly relied on geometrical arguments, interbasin correlations and correlations with DSDP site 548 on Goban Spur (Fig. 1) (de Graciansky et al. Reference De Graciansky, Poag, Cunningham, Loubere, Masson, Mazzullo, Montadert, Müller, Otsuka, Reynolds, Sigal, Snyder, Vaos, Waples, De Graciansky, de Poag, Cunningham, Loubere, Masson, Mazzullo, Montadert, Müller, Otsuka, Reynolds, Sigal, Snyder, Vaos and Waples1985; McDonnell & Shannon, Reference McDonnell, Shannon, Shannon, Haughton and Corcoran2001; Pearson & Jenkins, Reference Pearson, Jenkins, Summerhayes and Shackleton1986; Stoker, van Weering & Svaerdborg, Reference Stoker, van Weering, Svaerdborg, Shannon, Haughton and Corcoran2001). According to De Mol et al. (Reference De Mol, Van Rensbergen, Pillen, Van Herreweghe, Van Rooij, McDonnell, Huvenne, Ivanov, Swennen and Henriet2002), the erosional surface on which the Belgica mounds are rooted is of probable Miocene age, while the Unit P3 is of supposed Late Pliocene to Pleistocene age. Van Rooij et al. (Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003) refine the age model considerably and propose an Early to Middle Miocene age for unit P1, an inferred Middle Miocene–Middle Pliocene age for unit P2 and a Quaternary age for unit P3. The RD2 unconformity between units P1 and P2 at drill site U1318B is caused by an early Middle Miocene erosional event, while the RD1 unconformity between P1 and P3 in the Belgica mound area, and P2 and P3 in the drill site area is related to a Late Pliocene erosional event. Van Rooij et al. (Reference Van Rooij, Blamart, Kozachenko, Henriet, Viana and Rebesco2007) propose a Middle Pleistocene age for the onset of the drift sedimentation while the initial settling of coral banks started during Early Pleistocene times.
Van Rooij et al. (Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003) also presented a reconstruction of the local depositional environment associated with the coral bank settling (Figs 2, 3). Unit P1 was deposited under a relatively calm environment except for a limited zone of the slope where upslope-migrating sediment waves, indicative of strong bottom currents, were deposited. The subsequent erosional event RD2 truncating unit P1 was caused by the introduction of Norwegian Sea Water (NSW) into the North Atlantic Ocean and a wide-spread bottom-current flow under the modern-day oceanographic regime. Following the deposition of unit P2, a margin-wide erosional event during the Late Pliocene significantly eroded unit P2 and P1, creating the irregular and terraced palaeotopography on which coral bank initiation took place. According to the latter authors, this period heralds the influence of the glacial/interglacial variability on the oceanographic regime at the dawn of the Quaternary.
Seismic, lithostratigraphical and biostratigraphical interpretation with calcareous nannofossils of drill Site U1318B. Core recovery (black) and position of studied samples at holes U1318B and U1318C. Wavy lines indicate unconformities, dashed lines indicate uncertain position of boundary.
The IODP Hole U1318B investigated was drilled in a water depth of 409 m upslope from the Belgica mound area and recovered 213 m of sediment. A visual description of the sedimentological features is given in Expedition 307 Scientists (Reference Ferlman, Kano, Williams and Henriet2006), who distinguished three lithostratigraphical units (Fig. 3). Unit 1 comprises the interval 0–82.0 m below the sea-floor (mbsf) and unit 2 the interval 82.0–86.2 mbsf. Seismic analysis suggests the presence of a yet unmapped unconformity (J.-P. Henriet, pers. comm.) separating lithological units 1 and 2 (Figs 2, 3). A distinct unconformity is observed at 86.2 mbsf and is represented by a 5 to 10 cm thick conglomerate of black pebbles (possibly bivalves) and granules. Unit 3 comprises the sequence downhole from the RD1 moundbase unconformity at 86.2 mbsf to 241.0 mbsf. The sequence is lithologically rather uniform and consists of silty clay, fine-grained sand and clayey silt. The upper 10 cm of unit 3 consists of a distinct bivalve bed unconformably overlain by the conglomerate at the base of unit 2. In general, the carbonate content of unit 3 varies between 27 and 50 wt% (Expedition 307 Scientists, Reference Ferlman, Kano, Williams and Henriet2006). Within this unit, three subunits can be distinguished. Subunit 3A (86.2–127.4 mbsf) consists of silty clays interbedded with well-sorted fine-grained sand and silt with erosive boundaries at their base. Robust bivalves occur sporadically in the upper part of subunit A. Subunit 3B (127.4–190.3 mbsf) consists of silty clay with a carbonate content varying between 10 and 20 wt%. Bioturbation is abundant. Subunit 3C (190.3–241.0 mbsf) is composed of silty clay to fine-grained sand with a varying carbonate content of approximately 25 to 35 wt%. Bioturbation is abundant and bivalves are rare. Expedition 307 Scientists (Reference Ferlman, Kano, Williams and Henriet2006) correlate the boundary between lithostratigraphical units 3C and 3B with the boundary between seismic units P1 and P2 (Fig. 3).
A biostratigraphical analysis with calcareous nannofossils of core catcher samples from holes 1318A and 1318B was carried out during IODP Expedition 307 and provided a first relative dating of the sediments from Site 1318 (Expedition 307 Scientists, Reference Ferlman, Kano, Williams and Henriet2006) (Fig. 3). The late Middle Pleistocene to Late Pleistocene Emiliana huxleyi Zone (Recent–0.26 Ma) was recognized at a depth of 8.9 mbsf (core 1318A-1H) and 73.6 mbsf (core 1318A-8H). A sample at a depth of 84.6 mbsf (core 1318A-9H) belongs to the small Gephyrocapsa Zone and has an Early Pleistocene age. The unconformity at 86.2 mbsf is clearly reflected in the nannofossil record. A sample at 104.3 mbsf (core 1318A-11H), below the unconformity, yields diagnostic species with a highest occurrence at 3.6 Ma, which implies an Early Pliocene or older age. Samples 127.3 mbsf (core 1318B-14H) and 186.0 mbsf (core 1318B-20X) yield characteristic early Middle Miocene species. Nannofossils below 195.5 mbsf are badly preserved and a relative dating could not be proposed.
3. Material and methods
3.a. Palynological analysis
In Hole 1318B, lithostratigraphical unit 3 below the RD1 moundbase unconformity was analysed for dinoflagellate cysts and other marine organic-walled phytoplankton. A total of 60 samples were taken at a regular interval from core 11H (upper depth limit: 90.5 mbsf) to the deepest core 27X (lower depth limit: 241.5 mbsf) (Figs 3, 4). As no sediment was recovered from cores 15X and 16X in Hole 1318B, five samples from the corresponding depth interval in Hole 1318C (cores 8X, 9X and 10X) were analysed (Fig. 3). The first step in the processing consisted in the drying of about 30 cc of sample. About 15 to 25 g of dry sediment were prepared using standard palynological maceration techniques at the Research Unit Palaeontology at Ghent University (Louwye, Head & De Schepper, Reference Louwye, Head and De Schepper2004). The chemical processing involves demineralization with hydrochloric acid (HCl) and hydrofluoric acid (HF). Before demineralization, two Lycodpodium clavatum tablets (batch no. 483216, X=18583, s=±764 per tablet) were added for counting the concentration of dinoflagellate cyst and other marine palynomorphs. The residue was sieved on a 16 μm nylon screen and stained with safranin-O. Glycerine jelly was used as a mounting medium, and the slides were sealed with transparent nail varnish. No oxidation or ultrasound treatments were applied since these techniques could have damaged the dinoflagellate cysts or caused selective loss of species.
(a, b, c) Distribution of dinoflagellate cysts and other organic-walled palynomorphs in holes 1318B and 1318C. For location of samples see Figure 3.
A minimum of 300 marine palynomorphs, mainly dinoflagellate cysts, green algae and acritarchs, were counted systematically in each slide, and additionally the terrestrial palynomorphs, mainly bisaccate pollen (Fig. 4). Also, the presence of organic linings of foraminifers was noted. The rest of the slide was then scanned for rare species and well-preserved specimens for microphotography. The latter were not included in the count. The preservation of the dinoflagellate cysts varies between moderate to exceptional, and the assemblage can be considered as diverse and rich. Protoperidiniacean cysts are present throughout the section studied, indicating an unlikely preservational bias through oxidation. Photomicrographs were taken with a Zeiss MRc5 camera mounted on a Zeiss Axioplan2 microscope. The taxonomy of the species recorded (Fig. 4) follows Fensome & Williams (Reference Fensome and Williams2004). The timescale of Lourens et al. (Reference Lourens, Hilgens, Shackleton, Laskar, Wilson, Gradstein, Ogg and Smith2005) is used throughout this paper. The slides are housed in the collection of the Research Unit Palaeontology of Ghent University, Belgium.
3.b. Palaeomagnetic measurements
Shipboard palaeomagnetic measurements were conducted on cores from holes U1318A, U1318B and U1318C. Remanence measurements and alternating field (AF) demagnetizations were performed using a long-core cryogenic magnetometer (2G Enterprises model 760-R), permanently installed on the R/V JOIDES Resolution. This instrument is equipped with a direct-current superconducting quantum interference device (DC-SQUID) and has an inline AF demagnetizer capable of reaching peak fields up to 80 mT. The spatial resolution measured by the width at half-height of the pickup coils response is <10 cm for all three axes, although they sense a magnetization over a core length up to 30 cm. The magnetic moment noise level of the cryogenic magnetometer is ∼10−9 emu or 10−6 A/m for 10 cm3 rock volume. The practical noise level, however, is affected by the magnetization of the core liner (∼8 × 10−6 A/m) and the background magnetization of the measurement tray (∼1 × 10−5 A/m). Measurements were undertaken using the standard IODP magnetic coordinate system (+x = vertical upward from the split surface of archive halves, +y = left split surface when looking upcore, and +z = downcore). Alternating-field demagnetization of natural remanent magnetization was conducted up to 20 mT in 5 mT steps on core 307-U1318A-1H. Based on this demagnetization experiment (Fig.5a, b), the other sections were demagnetized at 10 and 15 mT with a spatial resolution of 5 cm. NRM and magnetization after two-step demagnetization were measured on archive halves. Discrete samples were taken on the working halves of cores in Hole U1318B.
(a) Stepwise AF demagnetization (5, 10, 15, and 20 mT) of Section 307-U1318A-1H1 at 0.25 mbsf. (b) Stepwise AF demagnetization (5, 10, 15, and 20 mT) of Section 307-U1318A-1H3 at 3.15 mbsf. (c) Stepwise AF demagnetization (0, 2.5, 5, 7.5, 10, 12.5, 15, 17.5, 20, 22.5, 25, 30, 35, 40, 45, and 50 mT) of Section 307-U1316A-2H5 at 164.5 mbsf. Left panels represent orthogonal projections of endpoints of the magnetization vector. Open circles – projections on vertical plane; solid circles – projections on horizontal plane. Right panels represent equal area projection of the magnetization vector during demagnetization.
Shore-based measurements were carried out with the cryogenic magnetometer (2G Enterprises model 760-R) installed at the geophysical centre of Dourbes (KMI, Belgium) and equipped with direct-current superconducting quantum interference device (DC-SQUID) magnetic sensors. An inline AF demagnetizer has the capability to reach peak fields up to 100 mT. The SQUID-sensors can detect magnetizations up to 10−12 A/m. The more precise detection limit, compared to the same instrument installed aboard the JOIDES Resolution, can be explained by lower interference of magnetic and electrical noise. The measurements on the discrete samples confirmed the isolation of the characteristic remanent magnetization for the uppermost sections at 10–15 mT. Obtaining stable remanent magnetizations in the lowermost sections was difficult (Fig. 5c). However, characteristic remanent magnetizations could be calculated by the standard three-dimensional least-squares fit to palaeomagnetic vector data via principal component analysis (PCA) on AF-demagnetization steps 7.5 mT, 10 mT and 12.5 mT, respectively (Kirschvink, Reference Kirschvink1980). Visualization and analysis of the palaeomagnetic data were performed in the windows-based ‘palaeomagnetic analysis program’, developed by Zhang & Ogg (Reference Zhang and Ogg2003).
4. Discussion of the organic-walled phytoplankton assemblage
A total of 89 in situ dinoflagellate cyst species, 13 acritarch species and one green alga were recorded (Fig. 4). Every sample is dominated by the Spiniferites/Achomosphaera spp. group. Neogene species of the genus Spiniferites are usually thin-walled and easily recognized. However, they have a limited biostratigraphical potential, which is the reason for the incorporation in this large amalgam of species. The notable exception, Achomosphaera andalousiensis andalousiensis, has a lowest occurrence (LO) in the upper part of the Serravallian (Gradstein, Ogg & Smith, Reference Gradstein, Ogg and Smith2005), and was therefore not included in the group. Thick-walled representatives of Spiniferites spp. are recorded in every sample, and it is probable that a small number of them are to be considered as reworked from the Cretaceous. Since no distinction was made between Neogene and pre-Neogene Spiniferites and Achomosphaera, the size of this allochthonous component of the assemblage remains unknown. Reworking of other pre-Neogene species is relatively unimportant and never exceeds more than 3% of the assemblage (core 1318B-20X). Although not detailed in Figure 4, reworking of Cretaceous species is more frequent than of Palaeogene species.
Batiacasphaera micropapillata, Batiacasphaerta minuta, Dapsilidinium pseudocolligerum, Lingulodinium machaerophorum machaerophorum, Nematosphaeropsis labyrinthus and Operculodinium centrocarpum centrocarpum are abundant species in almost every sample. Cousteaudinium aubryae and Cleistosphaeridium placacanthum are present abundantly in the lower part of the section studied, respectively, in cores 1318B-17X to 27X, and in cores 1318B-18X to 27X. In the latter cores, Polysphaeridium zoharyi is present continuously. Unipontedinium aquaeductum was recorded in the assemblage from core 1318B-17X through 11H, that is, the upper part of the core. Habibacysta tectata is present in large numbers in every sample of cores 1318C-9X through 1318B-11H. The remainder of the assemblage is present in considerably lower numbers. The heterogeneous group ‘Round Brown Cysts’ mostly comprises thin-walled dark brown cysts with probable Brigantedinium affinities. The preservation of these heterotrophic cysts is moderate to poor and hinders a specific determination. A selection of dinoflagellate cysts species in open nomenclature is given in Figure 6.
Photomicrographs of selected dinoflagellate cysts from Hole U1318B. All photomicrographs taken in bright field. Various magnifications. (a, b) Batiacasphaera sp. 1 Edwards Reference Edwards, Robert and Schnitker1984. High focus on ornamentation (a) and low focus on archaeopyle. Maximum width central body 44 μm; wall thickness 2 μm. (c, d) Echinidinium sp. A. High focus on cyst wall (c) and slightly lower focus (d). Orientation uncertain. Maximum diameter without processes 46 μm; maximum length processes 10 μm. (e, f) Echinidinium sp. B. High focus on cyst wall (e) and slightly lower focus (f). Orientation uncertain. Maximum diameter without processes 45 μm; maximum length processes 8 μm. (g, h) Echinidinium sp. C. High focus (g) and low focus (h). Orientation uncertain. Maximum diameter without processes 39 μm; maximum length of processes 8 μm. (i–k) Lejeunecysta sp. A. High focus on cingulum (i), slightly lower focus on epicyst (j) and low focus on hypocyst (k). Maximum diameter at cingulum 71 μm. (l–n). Selenopemphix sp. A. High focus on epicyst and cingulum (l), slightly lower foci on archeaopyle (m, n). Maximum diameter 90 μm. (o, p) Trinovantedinium sp. A. High focus on cingulum (o), slightly lower focus on hypocyst (p) and low focus on ventral surface (q). Maximum diameter at cingulum 87 μm, length of processes 8 μm.
5. Review of selected Lower and Middle Miocene dinoflagellate cyst studies in the North Atlantic Realm
A detailed and comprehensive account of global Miocene dinoflagellate cyst studies is given in Head, Norris & Mudie (Reference Head, Norris, Mudie, Srivastava, Arthur and Clement1989c). These authors discuss in great detail the biostratigraphical studies with dinoflagellate cysts from mostly DSDP and ODP sites in the North Atlantic realm and contiguous areas: Norwegian–Greenland Sea, DSDP Leg 38 (Manum, Reference Manum, Talwani and Udintsev1976), Rockall Plateau, DSDP Leg 48 (Costa & Downie, Reference Costa, Downie, Montadert and Robert1979), Denmark (Piasecki, Reference Piasecki1980), Goban Spur, DSDP Leg 80 (Brown & Downie, Reference Brown, Downie, De Graciansky and Poag1984), Rockall Plateau, DSDP Leg 81 (Edwards, Reference Edwards, Robert and Schnitker1984) and Vøring Plateau ODP Leg 104 (Manum et al. Reference Manum, Boulter, Gunnarsdottir, Rangnes, Scholze, Eldholm, Thiede and Taylor1989). According to Head, Norris & Mudie (Reference Head, Norris, Mudie, Srivastava, Arthur and Clement1989c), sample resolution and dinoflagellate cyst diversity and richness in the latter studies is mostly high, while the availability and quality of the independent biostratigraphical control is highly variable. Traditional biostratigraphical control is usually provided by calcareous nannoplankton and foraminifers, and only a few sites have stratigraphical control through magnetostratigraphy or radiometric dating. The dinoflagellate cyst biozonations from these sites have not only been refined or redefined during the last decades, the chronostratigraphical position of the dinoflagellate cyst events has altered many times since their publication. In light of the recent advancements in Lower and Middle Miocene biostratigraphy and the publication of a new Neogene timescale (Lourens et al. Reference Lourens, Hilgens, Shackleton, Laskar, Wilson, Gradstein, Ogg and Smith2005), it is necessary to review certain aspects (ranges of key species, ages of boundaries) of these biozonations. Furthermore, recent studies made clear that many of the discrepancies in the stratigraphical ranges of some species between sites are related to differing palaeoenvironmental conditions, such as sea-surface temperature, upwelling and palaeodepth of the depositional area. A brief re-evaluation of some Lower and Middle Miocene key species with stratigraphical or palaeoecological significance from the above mentioned studies, and present in this study, is therefore given (Fig. 7).
Stratigraphical distribution of selected dinoflagellate cysts in Hole 1318B according to literature: 1 – Zevenboom (Reference Zevenboom1995); 2 – Munsterman & Brinkhuis (Reference Munsterman and Brinkhuis2004); 3 – Williams et al. (Reference Williams, Brinkhuis, Pearce, Fensome, Weegink, Exon, Kennett and Malone2004); 4 – de Verteuil & Norris (Reference De Verteuil and Norris1996). Biozonation according to de Verteuil & Norris (Reference De Verteuil and Norris1996).
5.a. Baffin Bay–Labrador Sea
Head, Norris & Mudie (Reference Head, Norris, Mudie, Srivastava, Arthur and Clement1989a,b,c) reported on Miocene and Lower Pliocene high-resolution dinoflagellate cyst studies from Baffin Bay (Upper Miocene to lowermost Pliocene: Hole 645E, Leg 105) and from the Labrador Sea (Lower Miocene to lower Upper Miocene: Site 646, Leg 105). Approximately 100 dinoflagellate cyst taxa were recovered from the Lower to Middle or lower Upper Miocene in Hole 645E in Baffin Bay. Five dinoflagellate cyst biozones are defined (BB1 to BB5), and ages mainly rely on dinoflagellate cyst datums, and where available, on biostratigraphy with other microfossils and magnetostratigraphy. Biozones BB3 and BB4 are of Middle Miocene age. The LO of Labyrinthodinium truncatum defines the lower boundary of zone BB3, and is considered as an important datum. Other species with a LO within this zone are Invertocysta tabulata and Habibacysta tectata. The base of zone BB4 is defined by the LO of Unipontedinium aquaeductum. The upper boundary of zone BB4 is defined by the LO of the acritarch Leiosphaeridia sp., and tentatively placed in the late Middle Miocene or early Late Miocene. Unfortunately, the age of this zone is poorly constrained by other microfossil groups.
5.b. Salisbury Embayment, USA
De Verteuil (Reference De Verteuil, Mountain, Miller, Blum, Poag and Twitchell1996, Reference De Verteuil, Miller and Snyder1997) and de Verteuil & Norris (Reference De Verteuil and Norris1996) published their results of the biostratigraphical analysis with dinoflagellate cysts of the continental slope and rise off New Jersey (ODP Leg 150), the Cape May and Atlantic City boreholes (New Jersey Coastal Plain, USA), and the Miocene in the Salisbury Embayment (Atlantic margin, USA). Their zonation covers the uppermost Oligocene to the uppermost Miocene and holds ten biozones with an average zonal duration of 1.8 Ma (Fig. 7). The biostratigraphical study is mostly based on classic onshore Miocene sections but the data was integrated with planktonic foraminiferal and calcareous nannofossil biostratigraphical data from ODP Leg 150. The zonation proved readily applicable to Miocene successions from other mid-latitude sites such as the North Sea Basin (Louwye, Reference Louwye2002, Reference Louwye2005; Louwye, De Coninck & Verniers, Reference Louwye, De Coninck and Verniers2000).
The Cousteaudinium aubryae Interval Zone (DN3: middle Lower Miocene to upper Lower Miocene) is a ‘gap’ zone defined as the interval from the highest occurrence (HO) of Exochosphaeridium insigne to the LO of Labyrinthodinium truncatum. The assemblage is characterized by high numbers of the eponymous species Apteodinium spiridoides and Cleistosphaeridium placacanthum. The Distatodinium paradoxum Interval Zone (DN4: uppermost Lower Miocene to lower Middle Miocene) is defined as the interval from the LO of Labyrinthodinium truncatum to the HO of Distatodinium paradoxum. Apteodinium spiridoides and Cousteaudinium aubryae disappear with this zone. The lower boundary, however, is easily defined by the large numbers of Labyrinthodinium truncatum, an important dinoflagellate cyst horizon according to de Verteuil & Norris (Reference De Verteuil and Norris1996). The Batiacasphaera sphaerica Interval Zone (DN5: middle Middle Miocene) is defined as the interval from the HO of Distatodinium paradoxum to the HO of Cleistoaphaeridium placacanthum. The range of Unipontedinium aquaeductum is restricted to this zone. De Verteuil & Norris (Reference De Verteuil and Norris1996) consider the LO of Unipontedinium aquaeductum as an important dinoflagellate cyst horizon across the North Atlantic and Mediterranean. In the Salisbury Embayment, Habibacysta tectata and Apteodinium tectatum appear and disappear in the base of this zone, respectively. The Selenopemphix dionaeacysta Interval Zone (DN6: upper Middle Miocene) is defined as the interval from the HO of Cleistophaeridium placacanthum to the LO of Cannosphaeropsis passio. The Cannosphaeropsis passio Range Zone (DN7: upper Middle Miocene to uppermost Middle Miocene) holds the entire range of the eponymous species. No other relevant bioevents within this zone are noted.
5.c. The Netherlands
Munsterman & Brinkhuis (Reference Munsterman and Brinkhuis2004) presented a biozonation for the Miocene of the Netherlands (southern North Sea Basin). They defined 14 formal dinoflagellate cyst zones and three subzones. No direct calibration of their data through other microfossil biostratigraphy or magnetostratigraphy was provided. However, a number of calibrated dinoflagellate cyst events from the Mediterranean (mainly Zevenboom, Reference Zevenboom1995) and the Atlantic realm were integrated into their zonal scheme (Fig. 7). Eight zones were defined in the Middle Miocene, based on following dinoflagellate cyst events: LO of Labyrinthodinium truncatum, LO and HO of Unipontedinium aquaeductum, HO of Cousteaudinium aubryae, HO of Palaeocystodinium ventricosum (sensu Zevenboom, Reference Zevenboom1995), LO of Achomosphaera andalousiensis, HO of Cerebrocysta poulsenii, HO of Cannosphaeropsis poulsenii. Munsterman & Brinkuis (Reference Munsterman and Brinkhuis2004) place the LO of Labyrinthodinium truncatum in the North Sea Basin at 15.8 Ma, which is, compared to other biozonations, relatively late. The authors, however, suspect a delayed entry of this species (D. Munsterman, pers. comm.). The LOs of the latter species and Cerebrocysta poulsenii coincide in the Netherlands, an event also recorded in our study.
5.d. Other studies
An overview of some key dinoflagellate cyst datums with an estimated correlation to magnetostratigraphy and planktonic foraminifer zones is given by Gradstein, Ogg & Smith (Reference Gradstein, Ogg and Smith2005). Labyrinthodinium truncatum has its LO at the Langhian–Burdigalian boundary, which is placed at 15.97 Ma. The LO of Unipontedinium aquaeductum coincides with the HO of Cousteaudinium aubryae at 14.9 Ma.
Williams et al. (Reference Williams, Brinkhuis, Pearce, Fensome, Weegink, Exon, Kennett and Malone2004) gave an overview of global dinoflagellate cyst events for the Late Cretaceous through the Neogene based on a high-resolution analysis of two sites offshore of Tasmania (ODP Leg 189) and literature data (Fig. 7). The proposed bioevents are calibrated against detailed magnetostratigraphical results from ODP Leg 189. Dinoflagellate cysts events are grouped into low-, mid- and high-latitude associations for both hemispheres.
6. Palaeomagnetic and biostratigraphical dating of Hole U1318B and Hole U1318C
Only the inclination data from the characteristic remanent magnetizations were retained for the interpretation of a magnetostratigraphical framework (Fig. 8). Magnetostratographical dating was possible by correlating the inclination records with the geomagnetic polarity timescale from Lourens et al. (Reference Lourens, Hilgens, Shackleton, Laskar, Wilson, Gradstein, Ogg and Smith2005) (Table 1).
Inclination records at peak fields of 15 mT and interpreted magnetostratigraphical framework in holes U1318B and U1318C. Large dots represent magnetization directions obtained by the standard least-squares method on discrete samples (Kirschvink, Reference Kirschvink1980). Black horizontal lines represent the interpreted chrons/subchrons.
Overview of the chrons and subchrons encountered in holes U1318B and U1318C
The boundary chron/subchron ages are according to Lourens et al. (Reference Lourens, Hilgens, Shackleton, Laskar, Wilson, Gradstein, Ogg and Smith2005).
The inclination data cluster around approximately 66° above 82.0 mbsf and 79.9 mbsf in holes U1318B and U1318C, respectively. Below these depths, inclination data show a small shift, which can be explained by the presence of more reworked and coarser material in the intervals between 82.0 and 86.2 mbsf (Hole U1318B) and 79.9 and 84.2 mbsf (Hole U1318C). The inclinations of the uppermost sections correspond with the reference inclination of the present geomagnetic field at the drilling site, which is 66.36°, according to the International Geomagnetic Reference Field 2005. These positive inclinations can be interpreted as a normal polarity zone corresponding with the Brunhes Chron, which has an age <0.78 Ma.
Below these depths, inclination data become more scattered. Also, magnetic intensities drop below these depths to low values (Expedition 307 Scientists, Reference Ferlman, Kano, Williams and Henriet2006). The extremely depleted values of the magnetic intensities in the lowermost sections are greater than can be accounted for by dilution with magnetite-poor sediments and suggest that post-depositional destruction of magnetite has occurred, explaining the scattered directional data at these depths. Florindo, Roberts & Palmer (Reference Florindo, Roberts and Palmer2003) showed that dissolution of magnetite is a common feature in siliceous sedimentary environments, whereby thermodynamic calculations indicate that magnetite is unstable under conditions of elevated dissolved silica concentrations (and appropriate Eh-pH conditions), and predict that magnetite will break down to produce iron-bearing smectite. Pore water analyses in the off-mound sediments indeed show an increase in silica concentrations in the lowermost sections (Expedition 307 Scientists, Reference Ferlman, Kano, Williams and Henriet2006).
However, a tentative magnetostratigraphical framework can be proposed below 86 mbsf and 84 mbsf for holes U1318B and U1318C, respectively, based on the general inclination trends and taking into account the presence of an important unconformity and some smaller-scaled erosive boundaries (Expedition 307 Scientists, Reference Ferlman, Kano, Williams and Henriet2006). The unconformity is localized at 86.2 and 84.2 mbsf in holes U1318B and U1318C, respectively, and corresponds with the RD1 moundbase unconformity (Fig. 2). The palynological analysis showed that a latest Middle Miocene age can be proposed for the sediments immediately below the unconformity. This palynological data was used as a palaeomagnetic tiepoint.
Another tiepoint is provided by a Sr-isotope analysis (Kano et al. Reference Kano, Ferdelman, Williams, Henriet, Ishikawa, Kawagoe, Takashima, Kakizaki, Abe, Sakai, Browning and Li2007). Although this isotope study focuses on the origin and growth history of Mound Challenger (Hole U1317), some Sr-isotope analysis on skeletal aragonite from bivalves at Site 1318 were performed. The Sr-isotope analysis of a bivalve at a depth of 140.83 mbsf in Hole U1318C (Core 9X, Fig. 3) gives a mean age of 13.38 Ma (12.74 Ma to 14.62 Ma, range of ages from the isotopic value ± 2σ on the upper/lower limit age curve), and confirms the palynological data (Fig. 9).
Distribution of selected dinoflagellate cyst in Hole 1318B and Hole 1318C, together with the palaeomagnetic interpretation of the sequence, and biozonation according to de Verteuil & Norris (Reference De Verteuil and Norris1996). Solid line – continuous occurrence, X – scattered occurrence; R? – possibly reworked; R – reworked. Dashed horizontal line – uncertain boundary.
Figures 8, 9 and Table 1 represent the magnetostratigraphical framework, and the boundary Chron ages according to Lourens et al. (Reference Lourens, Hilgens, Shackleton, Laskar, Wilson, Gradstein, Ogg and Smith2005). The palaeomagnetic analysis indicates an age between late Burdigalian (Chron C5Cn.3n: 16.72 Ma) and late Serravallian at 86 mbsf (C5An.1n: 12.01 Ma) for the sections studied (Table 1, Figs 8, 9).
Unipontedinium aquaeductum has a continuous, but restricted, range in the upper part of the section studied from sections 1318B-17X-3 (lower part of Chron C5Acn, 14.10 Ma) to 1318B-11H-4 (base of Chron C5An.2n, approximately 12.4 Ma) (Figs 4, 9). The higher occurrence of a sole, fragmented specimen is considered not to be in situ. Zevenboom (Reference Zevenboom1995) placed the LO of Unipontedinium aquaeductum at the Cessole, Italy in the upper part of polarity subchron C5Bn.2n (15.1 Ma following Lourens et al. Reference Lourens, Hilgens, Shackleton, Laskar, Wilson, Gradstein, Ogg and Smith2005), while the highest common occurrence of Unipontedinium aquaeductum at Cassinasco, Italy is located in the basal part of polarity subchron C5An (12.4 Ma following the timescale of Lourens et al. Reference Lourens, Hilgens, Shackleton, Laskar, Wilson, Gradstein, Ogg and Smith2005). These dates were also used by Munsterman & Brinkhuis (Reference Munsterman and Brinkhuis2004) in their southern North Sea dinoflagellate cyst zonation. Williams et al. (Reference Williams, Brinkhuis, Pearce, Fensome, Weegink, Exon, Kennett and Malone2004) placed the LO of Unipontedinium aquaeductum at 15 Ma, based on data of de Verteuil & Norris (Reference De Verteuil and Norris1996). The latter authors gave an overview of the global, non-calibrated LOs and HOs of Unipontedinium aquaeductum, in the lower Middle Miocene and the middle Middle Miocene, respectively. Higher stratigraphical occurrences are considered to be reworked.
Cannosphaeropsis passio is continuously recorded in the uppermost part from sections 1318B-12H-1 (base of Chron C5Ar.1r) to 1318B-11H-1 (C5An.1r), and has thus in this study a range from 12.73 Ma to 12.12 Ma (Figs 4, 9). Organic-walled phytoplankton from the uppermost core 1318B-10H, immediately below the unconformity, were not examined in this study, and a slightly younger occurrence might be possible (Figs 4, 9). The HO would then lie in Chron C5An.1n at 12.01 Ma. Many authors consider this species as an index species for the late Middle Miocene (e.g. de Verteuil & Norris, Reference De Verteuil and Norris1996; Munsterman & Brinkhuis, Reference Munsterman and Brinkhuis2004; Strauss, Lund & Lund-Christensen, Reference Strauss, Lund and Lund-Christensen2001; Williams et al. Reference Williams, Brinkhuis, Pearce, Fensome, Weegink, Exon, Kennett and Malone2004). According to the latter, Cannopshaeropsis passio has a restricted occurrence from 12.8 Ma to the Serravallian–Tortonian boundary, based on records from de Verteuil & Norris (Reference De Verteuil and Norris1996). This implies that the considered section has a middle to late Serravallian age. Zevenboom (Reference Zevenboom1995) recorded a single specimen, as Cannosphaeropsis utinensis, in the base of the Tortonian in the Mazzapiedi section (Italy), more specifically in the top of subchron C5r, corresponding with an age of approximately 11 Ma (Lourens et al. Reference Lourens, Hilgens, Shackleton, Laskar, Wilson, Gradstein, Ogg and Smith2005). Munsterman & Brinkhuis (Reference Munsterman and Brinkhuis2004) found a similar occurrence above the Serravallian–Tortonian, but did not specify the exact position. Most probably, both findings above the Middle–Upper Miocene boundary can be attributed to reworking.
Both subspecies Labyrinthodinium truncatum truncatum and Labyrinthodinium truncatum modicum are present through the greater part of the section studied; they are only absent in lowermost core 1318B-27X. The LO lies in section 1318B-26X-2 in the base of Chron C5Cn.2r at 16.54 Ma (Figs 4, 9). Although many authors place the LO of Labyrinthodinium truncatum just below the Burdigalian–Langhian boundary (e.g. de Verteuil & Norris, Reference De Verteuil and Norris1996; Munsterman & Brinkhuis, Reference Munsterman and Brinkhuis2004; Strauss et al. Reference Strauss, Lund and Lund-Christensen2001), no precise age is specified. Williams et al. (Reference Williams, Brinkhuis, Pearce, Fensome, Weegink, Exon, Kennett and Malone2004) located the LO of this species at 16.5 Ma and the HO at 7.85 Ma (late Tortonian), based on the records of de Verteuil & Norris (Reference De Verteuil and Norris1996).
The biozonation of de Verteuil & Norris (Reference De Verteuil and Norris1996) can readily be applied to the studied sequences (Fig. 9). The Cousteaudinium aubryae Interval Zone (DN3: middle Lower Miocene to upper Lower Miocene) is recognized in the two lowermost samples, immediately below the LO of L. truncatum. Apteodinium spiridoides, a major component of the assemblage in the type area, was not recorded in the Porcupine Basin, while Cleistosphaeridium placacanthum is ubiquitous. The Distatodinium paradoxum Interval Zone (DN4: uppermost Lower Miocene to lower Middle Miocene) is recognized from sections 1318B-26X-2 to 1318B-21X-7 where the eponymous species has its HO. The recognition of the Batiacasphaera sphaerica Interval Zone (DN5: middle Middle Miocene), defined as the interval from the HO of Distatodinium paradoxum to the HO of Cleistosphaeridium placacanthum, proves difficult. The latter species has its highest continuous occurrence in section 1318B-14H-1, where the upper boundary of this zone could be placed. However, Unipontedinium aquaeductum, a species restricted to this zone in the Salisbury Embayment type area, occurs continuously, higher in the studied sequence. Consequently, the location of the upper boundary of the DN5 Zone is unsure and tentatively positioned above section 1318B-14H-1. The Selenopemphix dionaeacysta Interval Zone (DN6: upper Middle Miocene) is a ‘gap’ zone (see Section 5.d). The Cannosphaeropsis passio Range Zone (DN7: upper Middle Miocene to uppermost Middle Miocene) is recognized from sections 1318B-12H-1 (123–125) to the top section 1318B-11H-1 (123–126).
In conclusion, palaeomagnetic and biostratigraphical data indicate a late Burdigalian to late Serravallian age for the studied section. An overview of the magnetostratigraphical and biostratigraphical tiepoints versus the depth (mbsf) for holes U1318B and U1318C is presented in Figure 10. Based on the magnetostratigraphical tiepoints, an average sedimentation rate could be calculated of 3.3 cm ka−1. However, this value should be interpreted with care as compaction might have played an important role. The age estimates from magnetostratigraphy and biostratigraphy are in broad agreement; both suggest ages from late Burdigalian to late Serravallian.
Age (Ma) versus depth (mbsf) for palaeomagnetic, biostratigraphical and isotopic tiepoints at Site U1318.
An early to middle Langhian to late Serravallian age can be attributed to seismic unit P2. The seismic discontinuity RD2, separating seismic units P1 and P2, and lithostratigraphical units 3C and 3B, most probably represents a hiatus of minor magnitude since it is not readily reflected as a major shift in the dinoflagellate cyst assemblage. The location of the boundary between biozones DN4 and DN5 a few metres below the discontinuity might be a reflection of this hiatus. The detailed lithological analysis of the cores does not provide more information on this discontinuity, and further study is thus needed. Only a broad late Burdigalian to early Langhian age can be proposed for the upper part of seismic unit P1.
The youngest sediments in the sequence studied are thus of late Serravallian age, and it remains unknown whether the unconformity RD1 located at the top of the section studied is precisely located at the Serravallian–Tortonian boundary or in the uppermost Serravallian. Nevertheless, a correlation of the upper boundary of the section studied, that is, the unconformity in core 10H, can be proposed with the major sequence boundary Ser4/Tor1 at 10.5 Ma (Hardenbol et al. Reference Hardenbol, Thierry, Farley, Jacquin, de Graciansky, Vail, de Gradiansky, Hardenbol, Jacquin and Vail1998).
7. Palaeoenvironmental indices
The composition of a fossil dinoflagellate cyst assemblage is largely determined by the palaeoenvironment. Parameters such as temperature, salinity and nutrient availability influence the contemporanous assemblage and thus also the fossil assemblage. However, numerous factors interfere with an interpretation of a fossil assemblage in terms of past environments. A thorough review of the history of palaeoecological studies with dinoflagellate cysts and its restrictions was given by Dale & Dale (Reference Dale, Dale and Haslett2002). One of the main limiting factors in the Porcupine Basin for a reconstruction of the Miocene palaeoenvironment based on dinoflagellate cysts is the location of the drill site. Hole U1318B was drilled at a depth of 420 m and the site can be considered as located on the upper slope. Considering this depth, it is clear that site U1318B has known an oceanic influence since early Neogene times, since no elements in the sedimentological or palaeontological record point to a neritic, shelfal environment. A truly neritic fossil assemblage does not contain any oceanic species such as Impagidinium spp., but on the other hand, an assemblage from the oceanic realm or continental slope contains not only true oceanic species, but also many neritic species transported from the shelf to the upper slope depositional area (Dale, Reference Dale, Jansonius and McGregor1996; Dale & Dale, Reference Dale and Dale1992, Reference Dale, Dale and Haslett2002). In this respect, a local palaeoenvironmental reconstruction is limited at drill site U1318B, but a regional range for the palaeoenvironmental indices covering not only the Porcupine Basin but the southwestern Irish shelf is assumed.
7.a. Diversity
The diversity index represents the number of dinoflagellate cysts species recorded in one sample; other marine palynomorphs were not included in the diversity index (Fig. 11).
Diversity, reworking and ratios in holes 1318B and 1318C. The position of the samples is given, together with the stratigraphical position. Dashed line – smooth curve.
7.b. Temperature signal
Several authors have proposed proxies for reconstructions of the variation in sea-surface temperature, the so-called W/C ratios. In his detailed palynological study of the Pliocene MIS M2 cooling event in DSDP Hole 610 (Rockall Trough, eastern North Atlantic), S. De Schepper (S. De Schepper, unpub. Ph.D. thesis, Cambridge Univ. 2006) gave an extensive overview of the species used in several warm/cold indices proposed by Boessenkool et al. (Reference Boessenkool, Brinkhuis, Schonfeld and Targarona2001), Marret & Zonneveld (Reference Marret and Zonneveld2003), Versteegh & Zonneveld (Reference Versteegh and Zonneveld1994) and Warny (S. Warny, unpub. Ph.D. thesis, Univ. Catholique de Louvain, 1999). The W/C ratio is usually calculated as nW/(nW+nC) with n the number of specimens counted, W representing the warm-water indicating species and C the cold-water indicating species (Fig. 11). Different species were used by these authors for the establishment of the temperature signal. Versteegh (Reference Versteegh1994) stated that the choice of certain species depends on environmental factors, and it is conceivable that the temperature signal is an expression of ecological factors. De Schepper (S. De Schepper, unpub. Ph.D. thesis, Cambridge Univ. 2006) constructed W/C curves according to the above-mentioned authors and calculated the correlation of the W/C curves with a δ18O isotopic curve of the MIS M2 interval from DSDP Hole 610A. The R2 values vary between 0.3427 (Boessenkool et al. Reference Boessenkool, Brinkhuis, Schonfeld and Targarona2001) and 0.5673 (S. Warny, unpub. Ph.D. thesis, Univ. Catholique de Louvain, 1999). Based on published data and the data from DSDP Hole 610A, De Schepper (S. De Schepper, unpub. Ph.D. thesis, Cambridge Univ. 2006) considers Impagidinium aculeatum, Impagidinium patulum, Impagidinium strialatum, Melitasphaeridium choanophorum, Selenopemphix nephroides, Spiniferites hyperacanthus and Spiniferites mirabilis as warm-water species, and Bitectatodinium tepikiense, Filisphaera filifera and Impagidinium pallidum as cold-water species. The curve obtained with these species gives a R2 value of 0.6686 with the δ18O isotopic curve.
Lingulodinium machaerophorum, Polysphaeridium zoharyi, Tuberculodinium vancampoae and Operculodinium israelianum are species with an affinity for equatorial–temperate waters (Dale, Reference Dale, Jansonius and McGregor1996). Head (Reference Head2003) suggests a warm-water association for Nannobarbophora gedlii. Impagidinium paradoxum, Impagidinium patulum and Impagidinium strialatum are also considered to be temperate to tropical species (Marret & Zonneveld, Reference Marret and Zonneveld2003), while Tectatodinium pellitum may be regarded as a subtropical to tropical coastal species. Head (Reference Head1997) regards Melitasphaeridium choanophorum as a good indicator for relatively warm surface waters. According to Head (Reference Head1993), the distribution of the genus Sumatrodinium suggests affinities with tropical and subtropical to warm temperate waters. Dale (Reference Dale, Jansonius and McGregor1996) considers Bitectatodinium tepikiense and Impagidinium pallidum as species characteristic for the subpolar zone. On the basis of its fossil occurrences, Head, Norris & Mudie (Reference Head, Norris, Mudie, Srivastava, Arthur and Clement1989c) and Head (Reference Head1994) assume that high numbers of Habibacysta tectata indicate cool conditions. Versteegh (Reference Versteegh1994) also considers Habibacysta tectata as a cool-water species.
Given the close location to DSDP Hole 610A, the selection by De Schepper (S. De Schepper, unpub. Ph.D. thesis, Cambridge Univ. 2006) of warm-water and cold-water indicating species was followed, supplemented with Impagidinium paradoxum, Lingulodinium machaerophorum, Nannobarbophora gedlii, Operculodinium israelianum, Polysphaeridium zoharyi, Sumatradinium druggii, Sumatradinium soucouyantiae, Sumatradinium hamulatum, Tectatodinium pellitum and Tuberculodinium vancampoae as warm-water indicating species and Bitectatodinium tepikiense, Filisphaera filifera, Impagidinium pallidum and Habibacysta tectata as cold-water species.
7.c. Neritic/Oceanic signal and terrestrial influence
An index relying on the ratio between neritic and oceanic species does not represent the distance of the depositional area from the coastline, but is a proxy for the degree of transport of neritic species from the shelf to the ocean, or as a measure for the changes of the neritic water mass influence in the depositional area (Versteegh, Reference Versteegh1994). The ratio is calculated as N/O=nN/(nN+nO), with n the number of specimens counted, N representing the neritic species and O the truly oceanic species (Fig. 11).
According to Dale (Reference Dale, Jansonius and McGregor1996), species of the genus Impagidinium are considered to be oceanic, while representatives of the genus Nematosphaeropsis have neritic–oceanic affinties. In their ecological overview of the geographical distribution of dinoflagellate cysts, Marret & Zonneveld (Reference Marret and Zonneveld2003) regard Polysphaeridinium zoharyi, Selenopemphix nephroides, Tectatodinium pellitum and Tuberculodinium vancampoae as neritic cysts. Habibacysta tectata and species of Sumatradinium spp. are considered to be neritic species by Head (Reference Head1993, Reference Head1994). De Schepper (S. De Schepper, unpub. Ph.D. thesis, Cambridge Univ. 2006) considers Edwardsiella sexispinosum as an oceanic species. Apteodinium spp., Dapsilidinium spp., Lingulodinium machaerophorum and Systematophora are also thought to reflect marginal marine conditions (Zevenboom, Brinkhuis & Visscher, Reference Zevenboom, Brinkhuis and Visscher1994). The boundary between both realms is inhabited by Operculodinium centrocarpum (Dale, Reference Dale, Jansonius and McGregor1996). In addition to the above mentioned species, De Schepper (S. De Schepper, unpub. Ph.D. thesis, Cambridge Univ. 2006) considers Achomosphaera andalousiensis andalousiensis, Bitectatodinium tepikiense, Bitectatodinium? serratum, Invertocysta lacrymosa, Melitasphaeridium choanophorum, Operculodinium? eirikianum, Selenopemphix dionaeacysta, the Spiniferites/Achomosphaera spp. Group and Trinovantedinium glorianum as neritic species.
7.d. Productivity signal
Eutrophic or nutrient-rich marine environments are often characterized by a high primary productivity of phytoplankton. The nutrients are brought to deep shelfal areas by upwelling watermasses and to coastal areas by river discharge. Phytoplankton associations in such areas are often dominated by diatoms, although dinoflagellates are usually present, though in much lower numbers (Dale, Reference Dale, Jansonius and McGregor1996). Protoperidinioid dinoflagellates favour these areas because of the presence of abundant prey. Dale (Reference Dale, Jansonius and McGregor1996) reviews dinoflagellate cyst studies in upwelling areas and underlines the importance of protoperidinioid dinoflagellates as a primary signal for upwelling. Their abundance versus gonyaulacoid cysts is considered as a proxy for palaeoproductivy. Reichart & Brinkhuis (Reference Reichart and Brinkhuis2003) demonstrated the positive correlation between absolute abundance of protoperinioid cysts and upwelling. Although protoperidiniod dinoflagellate cysts are sensitive to aerobic decay (Zonneveld, Versteegh & de Lange Reference Zonneveld, Versteegh and de Lange2001), Reichart & Brinkhuis (Reference Reichart and Brinkhuis2003) further showed that the principal trends in the number of protoperidinioid cysts per gram of sediment do not alter significantly after intense oxidation. Their presence, or sometimes dominance, in every sample of this study, suggests that post-depositional oxidation did not alter the assemblages. Protoperidinioid dinocysts with brown, smooth walls, rounded outlines and indications of an archeopyle are grouped as ‘Round brown cysts’ (Fig. 4). These include mostly species of the genus Brigantedinium. Representatives of the genera Lejeunecysta, Selenopemphix and Trinovantedinium are cysts produced by heterotrophic dinoflagellates living on the inner shelf, and may represent upwelling (Reichart & Brinkhuis, Reference Reichart and Brinkhuis2003). However, the oceanic species Impagidinium paradoxum and the outer neritic to oceanic Nematosphaeropsis labyrinthus, clearly indicate the influence of oligotrophic (oceanic) water masses (Dale, Reference Dale, Jansonius and McGregor1996). The P/G ratio is calculated as nP/(nP+nG) with n the number of specimens counted, P reprensenting peridinioid cysts (round brown cysts, Echinidinium spp., Lejeunecysta spp., Sumatradinium spp. and Trinovantedinium spp.) and G representing all other dinoflagellate cysts excluding the goniodomacean Tuberculodinium vancampoae and Geonettia? sp. ind., the polykrikacean Polykrikos and the organic membranes of calcareous cysts (Fig. 11).
7.e. Terrestrial or continental influence
According to Versteegh (Reference Versteegh1994), the ratio between the sporomorphs (spores and pollen) and the marine palynomorphs is a measure for the terrestrial influence versus the marine influence and in this manner is a measure for the distance between the depositional area and the coast. Site U1318B is located approximatively 100 km from the present-day coastline and, as already mentioned, is located on the upper slope. It is unsurprising that the greater part of the terrestrial palynomorphs consist of bisaccate pollen brought into the depositional area by wind. Non-bisaccate pollen and spores are rare. The ratio is calculated as S/D=nS/(nS+nD) with n the number of specimens counted, S representing pollen and spores, and D dinoflagellate cysts and acritarchs (Fig. 11). We assume that the majority of the pollen and spores are primarily transported by wind rather than by rivers, since most pollen are bisaccate and of the Pinus type. The high percentage of bisaccate pollen in our dataset may be overrepresented in marine sediments because of their high buoyancy (Traverse, Reference Traverse and Traverse1988).
7.f. Reworking
The reworking is given as the relative proportion to the in situ marine palynomorphs (Fig. 11). The amount of reworked dinoflagellate cysts and other palynomorphs is indicative of sea-level variations (Sturrock, Reference Sturrock, Emery and Myers1996).
8. Palaeoenvironmental considerations
The dinoflagellate cyst assemblages are diverse and comprise species with different ecological preferences. Species with oceanic preferences such as Impagidinium spp. and Nematosphaeropsis sp. are present in every sample, although they never dominate the assemblage. Unipontedinium aquaeductum, here considered as an outer neritic species because of its absence in shallow marine environments (de Verteuil & Norris, Reference De Verteuil and Norris1996), is distinctly present. The majority of the species can be considered as neritic, while truly shallow marine species such as Lingulodinium machaerophorum and Polysphaeridium zoharyi are present to rare. Considering the location of Hole U1318B on the upper slope at a depth of 420 m, the dominance of these neritic species confirms the transport of truly neritic species from the shelfal area to the deeper depositional area.
Between the base of the studied sequence at 241 mbsf (core 1318B-27X) to approximately 144 mbsf (section 1318B-9X-7), the warm/cold ratio (Fig. 11) ranges between 0.9 and 1, and indicates a dominance of thermophilic dinoflagellate cyst species. Around 139 mbsf (section 1318B-9X-3), a distinct drop to values lower than 0.5 is observed, reflecting a distinct increase of cold-water indicating dinoflagellate cysts over approximately 35 m, followed by a short return of warm-water indicating species. This sudden shift at 139 mbsf is also indicated by the other ratios, although to a lesser degree. The S/D ratio (the expression of the continental influence) plunges towards values below 0.01. The productivity signal (the P/G ratio) and the neritic signal (the N/O ratio) both decrease slightly; this is caused by a decrease in protoperidinioid cysts and an increase in oceanic cysts, respectively.
Throughout the studied section, the dinoflagellate cyst diversity is considered relatively high, between 30 and 40 species per sample. The reworking remains low at values below 2%, although an abrupt influx of allochtonous elements, coupled with a rise in dinoflagellate cyst diversity, in core 20X at 180 mbsf is observed. Reworking tends to be higher during periods of low sea-level and lower during periods of high sea-level. Low sea-levels cause the river base level of erosion to lower, which results in higher amounts of terrestrial derived organisms and reworked fossil assemblages from the continent transported to the sea.
9. Discussion and conclusion
During the latest Oligocene, a warm phase was established during which the extent of the Antarctic ice sheets reduced and the temperature of bottom water increased. This warming period culminated in the Middle Miocene Climatic Optimum, which is well documented in the marine and continental realm from several proxies and lasted approximately from 17 to 14.5 Ma (Böhme, Reference Böhme2003; Zachos et al. Reference Zachos, Pagani, Sloan, Thomas and Billups2001). A gradual global cooling phase set in after 14.5 Ma and was characterized by several short-lived glaciations, the so-called Mi-isotope zones Mi3, Mi4, Mi5 and Mi6 of Miller et al. (Reference Miller, Wright and Fairbanks1991), with a maximum age of 13.7, 12.9, 11.9 and 10.3 Ma, respectively (Miller et al. Reference Miller, Mountain, Browning, Kominz, Sugarman, Christie-Blick, Katz and Wright1998). Distinct excursions in a δ18O curve, based on a benthic stable isotope record from the southeast Atlantic (ODP Site 1085), are dated at 13.8, 13.2, 11.7 and 10.4 Ma and correlated by Westerhold, Bickert & Rohl (Reference Westerhold, Bickert and Rohl2005) to the Mi3, Mi4, Mi5 and Mi 6 events of Miller et al. (Reference Miller, Wright and Fairbanks1991) and Miller et al. (Reference Miller, Mountain, Browning, Kominz, Sugarman, Christie-Blick, Katz and Wright1998). The stepwise cooling during the Middle Miocene climatic transition is further documented by Shevenell, Kennett & Lea (Reference Shevenell, Kennett and Lea2004) and Abels et al. (Reference Abels, Hilgen, Krijgsman, Kruk, Raffi, Turco and Zachariasse2005), who consider Mi3a at 14.2 Ma a minor step in the cooling phase and Mi3b at about 13.8 Ma a major step. Based on Mg/Ca data from planktonic foraminifers, Shevenell, Kennett & Lea (Reference Shevenell, Kennett and Lea2004) assume sea-surface temperature cooled 6 to 7°C during the Middle Miocene climate transition in the southwest Pacific. The astronomical forcing by long-period orbital cycles is considered to be the driving factor behind the Neogene climate changes (Zachos et al. Reference Zachos, Pagani, Sloan, Thomas and Billups2001). Based on integrated stratigraphy and astronomical tuning of a middle Miocene sequence in Malta, Abels et al. (Reference Abels, Hilgen, Krijgsman, Kruk, Raffi, Turco and Zachariasse2005) demonstrated that the major cooling step at 13.82±0.03 Ma coincides with minimum eccentricity values associated with the 400 ka cycle and minimum obliquity amplitudes of the 1.2 Ma cycle. The observed temperature drop (Fig. 11) can be paralleled with the onset of the global, climatic cooling phase which occurred after the Middle Miocene Climatic Optimum. The relative dating (see Section 6) places the distinct cooling event at 139 mbsf in Chron C5Bn at approximately 13.60 Ma, and suggests a correlation with the Mi3 event of Miller et al. (Reference Miller, Mountain, Browning, Kominz, Sugarman, Christie-Blick, Katz and Wright1998).
The sea-level fall associated with the Middle Miocene climatic transition shows a considerable geographical variability. δ18O data from the Marion Plateau (off NE Australia) indicate a sea-level fall of 50.0±5.0 m (John, Karner & Mutti, Reference John, Karner and Mutti2004), while a multidisciplinary study on the New Jersey continental slope indicates a sea-level drop of 25 m±5 m (Miller et al. Reference Miller, Mountain, Browning, Kominz, Sugarman, Christie-Blick, Katz and Wright1998). According to Westerhold, Bickert & Rohl (Reference Westerhold, Bickert and Rohl2005), the sea-level drop in the southeast Atlantic associated with the cooling during the late Middle Miocene is estimated at 43 m. Based on a multidisciplinary study of the New Jersey continental platform, Miller et al. (Reference Miller, Mountain, Browning, Kominz, Sugarman, Christie-Blick, Katz and Wright1998) demonstrate a good correlation between sequence boundaries and increases of δ18O, and postulate the expansion of ice sheets since about 42 Ma to be the main controlling factor on formation of sequence boundaries. Given the location of Site U1318B and the depth of the present-day sea-floor at 420 m, it is unlikely that the sea-level lowering will significantly influence the dinoflagellate cyst associations. The neritic/oceanic signal (Fig. 11) points to an increase of oceanic, oligothropic species and a reduction of the neritic species. The latter would then implicate a reduction of the downslope transport of neritic dinoflagellate cysts to the depositional area, which is unusual. Westerhold, Bickert & Rohl (Reference Westerhold, Bickert and Rohl2005) note a distinct enhancement of the transport of shelf-derived terrigenous matter during periods of cooling linked to increased ice-sheet growth on Antarctica. However, the lowering of the productivity together with the increase in oceanic, and thus oligothrophic, species in the Porcupine Basin rather points to a reduction or perhaps even to a shutdown of the upwelling after 14 Ma. At the same time, a collapse of the opal deposition in the North Atlantic realm is observed (Cortese et al. Reference Cortese, Gersonde, Hillenbrand and Kuhn2004). According to these authors, the waxing of the Antarctic glaciers and the development of the North Atlantic Deep Water influenced positively the fractionation of nutrients coupled to the establishment of an anti-estuarine circulation (exchange of deep, nutrient-rich waters for surface, nutrient-poor waters) in the North Atlantic. This circulation impeded the upwelling of silica-rich Antarctic Bottom Water in the North Atlantic.
A reconstruction of the Miocene depositional history based on the dinoflagellate cyst analysis partly confirms the interpretation of Van Rooij et al. (Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003). Seismic unit P1 was deposited in a relatively calm environment during the later part of the Early Miocene and the earlier part of the Middle Miocene, although upslope-migrating sediment waves suggest the presence of bottom current flow (Fig. 12). Unconformity RD2 is genetically related to the introduction of Norwegian sea water in the North Atlantic (van Rooij et al. Reference Van Rooij, De Mol, Huvenne, Ivanov and Henriet2003). Dinoflagellates indicate that this event took place during early Middle Miocene times. The hiatus associated with the unconformity is probably of minor magnitude and as such is not reflected by a clear break in the palynological record. The subsequent deposition of the acoustically transparent unit P2 then took place during the remainder of Middle Miocene times. Reflector RD1 terminates the Middle Miocene sequence, and reflects a basin-wide erosional event deeply affecting units P2 and P1. The unconformity can be dated as terminal Middle Miocene or earliest late Miocene, and correlated with sequence boundary Ser4/Tor1 at 10.5 Ma of Hardenbol et al. (Reference Hardenbol, Thierry, Farley, Jacquin, de Graciansky, Vail, de Gradiansky, Hardenbol, Jacquin and Vail1998). Seismic unit P3 is younger than 0.78 Ma and is thus of Middle Pleistocene or younger age.
Depositional history in the Belgica mound province. (a, b) Deposition of seismic Unit P1 and formation of sediment waves during Early Miocene and early Middle Miocene times. Subsequent formation of unconformity RD2 during early Middle Miocene times. (c) Deposition of the acoustically transparent Unit P2 during Middle Miocene times and subsequent erosion (unconformity RD1) (d). (e) Deposition of seismic Unit P1 during post-Middle Miocene times. The direction and size of the grey-shaded arrows illustrate the direction and vigour of the bottom currents, allowing drift deposition when small and erosive when large.
The Neogene megasequences of the NW European Atlantic margin are bounded by regional unconformities, preserved in the shelfal deposits and continuing into the deep water basins (Stoker et al. Reference Stoker, Hoult, Nielsen, Hjelstuen, Laberg, Shannon, Praeg, Mathiesen, van Weering and McDonnell2005). It is clear that eustatic sea-level changes cannot account solely for the observed stratal geometry, and unconformities, along the northwestern Atlantic margin, and these authors point out that middle to late Cenozoic tectonic activity plays a major role in the regional sedimentary and oceanographic dynamics.
Acknowledgements
Prof. Dr Joseph Hus is thanked for the use of the palaeomagnetic measure infrastructure at the Geophysical Centre of the Royal Meteorological Institute in Dourbes and for his excellent support during the measurements. Yuji Fuwa, Klayton Curtis, Margaret Hastedt and Trevor Williams are thanked for their palaeomagnetic help aboard the JOIDES Resolution during IODP expedition Leg 307. The authors would like to thank the IODP and Transocean crews and the scientific staff of the JOIDES Resolution on IODP Expedition Leg 307. The IODP staffs at the Bremen Core Repository and at Texas A&M University in College Station are thanked for their support during the post-cruise sampling activities. A part of this study has been carried out under the framework of the FWO Genesis and ESF Moundforce project. A. Foubert is a Ph.D. student funded through a FWO-fellowship. D. Van Rooij is a post-doctoral fellow of the FWO Flanders. The discussions with Jan De Coninck on dinoflagellate cyst taxonomy and morphology are appreciated. Jean-Pierre Henriet (Renard Centre of Marine Geology, Ghent University) introduced the first author to the Porcupine Basin and its intriguing geology. His interest and co-operation are kindly acknowledged. The careful reviews by J. Riding (British Geological Survey) and an anonymous reviewer are appreciated. Sabine Vancauwenberghe (Research Unit Palaeontology, Ghent University) is kindly thanked for her technical assistance during palynological maceration.
