Site 1137 was drilled primarily to establish the age and rock types forming Elan Bank, a shallow western salient of the central and southern Kerguelen Plateau. The hole penetrated 219.4 m of sediment with 59% recovery. In the first core, a disconformity separates a middle Pleistocene diatom ooze (lithologic Unit I) from an expanded 190-m-thick upper Miocene to uppermost Eocene nannofossil ooze section, which is remarkably devoid of chert (lithologic Unit II). This is underlain by a 20-m-thick glauconitic volcanic sand (lithologic Unit III) in which best recovery was in the last sediment core above basaltic basement. This sand contains exceptionally well-preserved foraminifers, calcareous nannofossils, and a few dinoflagellates and spores dated as late Campanian in age, between ~72 and 76 Ma by a conservative estimate, but between ~74 and 75 Ma using more narrowly defined biostratigraphic criteria based on Leg 183 preliminary drilling results (see discussion below on "Basal Sediment Age"). This provides a minimum age for the basaltic basement beneath the sediments. The basement may be somewhat older here, as seismic stratigraphic studies suggest that we have sampled the top of a basal sedimentary unit that thickens east of the site (see "Background and Objectives" and "Seismic Stratigraphy").
Sedimentation rates in Unit II pelagic sediments are exceptionally high (20 m/m.y.) (Fig. F11). Because the Tertiary sequence of Elan Bank has not been diluted by fine clastic material as at Sites 1138-1140 to the north and northeast, these high rates indicate high siliceous and calcareous planktonic productivity in this locality, which is ~6°S of the present-day Polar Front (Antarctic Convergence).
We examined smear slides prepared for calcareous nannofossil studies of core-catcher samples as well as for diatoms in the Neogene. Diatom preservation was good in the Neogene, and biostratigraphic control by coccoliths was poor because of high-latitude assemblages of low diversity. We made no attempt to process samples for diatom study, and our preliminary results could be greatly improved by shore-based study by diatom specialists. Sample 183-1137A-1R-CC from foraminifer-bearing diatom ooze of lithologic Unit I contains no calcareous nannofossils but does contain abundant diatoms from the Actinocyclus ingens Zone. Although the nominate taxon is abundant, we noted no Fragilariopsis barronii, indicating a mid-Pleistocene age for this sample.
Nannofossils are abundant in Sample 183-1137A-2R-CC, which we assigned to the combined Zone CN11-CN5b, but diatoms are few; thus, we could not determine a more precise age. Nannofloras in Samples 183-1137A-3R-CC to 4R-CC are dominated by reticulofenestrids, particularly R. perplexa, and also belong to Zone CN11-CN5b. They also contain common Denticulopsis dimorpha, indicating a diatom age between 10.6 and 11.7 Ma (assuming no Nitzschia denticuloides are present).
Sample 183-1137A-5R-CC contains the diatom Denticulopsis dimorpha, but the nannoflora is strongly dominated (95% of the assemblage) by Cyclicargolithus floridanus/abisectus, indicating the top of the combined Zone CN5a-CN3. This zone continues downhole through the next two core-catcher samples. Sample 183-1137A-6R-CC contains rare Pontosphaera multipora and is dominated by approximately equal numbers of reticulofenestrids and Coccolithus pelagicus, whereas the subjacent core catcher (7R-CC) contains the first downhole occurrences of Sphenolithus moriformis (rare) and Discoaster sp. cf. D. deflandrei (heavily overgrown) in the hole.
Calcidiscus leptopora/macintyrei is absent in Sample 183-1137A-8R-CC, which we therefore assigned to the combined lower Miocene Zone CN2-CN1. Approximately 65% of the assemblages is composed of Cyclicargolithus and 30% of Coccolithus pelagicus; Coccolithus miopelagicus, Discoaster sp. cf. D. deflandrei, Sphenolithus moriformis, and Helicosphaera granulata are few. We also assigned Samples 183-1137A-9R-CC to 12R-CC to Zone CN2-CN1, with the exception of 10R-CC, for which there was no core. Minor amounts of Oligocene taxa are reworked in these samples. Coccolithus miopelagicus is very abundant, and C. pelagicus and small reticulofenestrids strongly dominate Sample 183-1137A-11R-CC. On the other hand, Cyclicargolithus floridanus/abisectus constitute about 10% of the assemblage in Sample 183-1137A-12R-CC.
Large (up to 18 µm) Reticulofenestra bisecta are abundant in Sample 183-1137A-13R-CC, where they mark the top of the Oligocene and the zone of the same name. The nominate taxon for the next zone downhole, Chiasmolithus altus, is abundant (~5% of the assemblage) in Sample 183-1137A-14R-CC, which is still dominated by Cyclicargolithus (about 95%). Reticulofenestra bisecta is small (up to 12 µm) and only few to common. We also assigned Samples 183-1137A-14R-CC to 18R-CC to the Chiasmolithus altus Zone. Zygrhablithus bijugatus are rare and quite possibly reworked in Sample 183-1137A-15R-CC, but common in Sample 16R-CC. The succession of these three last occurrence datums are apparently quite consistent at all Leg 183 sites at which they have been encountered (Sites 1137 to 1140). Thus, the dissolution-susceptible Z. bijugatus can serve as a useful guide fossil on the central and northern Kerguelen Plateau, where the drill sites are all well above the Oligocene calcite compensation depth.
Isthmolithus recurvus, Reticulofenestra umbilica/hillae (up to 20 µm), Chiasmolithus altus, and Clausicoccus fenestratus are abundant to very abundant in Sample 183-1137A-19R-CC; Chiasmolithus oamaruensis is common to few. The co-occurrence of the first two taxa may indicate a hiatus within the core. Although the range of I. recurvus crosses the Oligocene/Eocene boundary, abundant C. fenestratus plus the strong dominance of C. altus over C. oamaruensis places this sample in the Oligocene (Blackites spinosus Zone or ~CP16; compare with the Eocene/Oligocene sequence at Deep Sea Drilling Project Hole 511 [Leg 71] on the Falkland Plateau and Ocean Drilling Program (ODP) Hole 737B [Leg 119] on the southern Kerguelen Plateau [Wise, 1983, table 1A; Wei and Thierstein, 1991, table 3]).
We did not observe Isthmolithus recurvus in Sample 183-1137A-20R-CC, but it is abundant in 21R-CC. A few Reticulofenestra oamaruensis and Discoaster tani are present in both samples. We did not see, however, Clausicoccus fenestratus; in addition, Chiasmolithus oamaruensis was abundant and considerably more prevalent than C. altus. Although both I. recurvus and R. oamaruensis span the Oligocene/Eocene boundary, the criteria cited in the above paragraph regarding the distribution of C. fenestratus and the chiasmoliths place these two samples on the Eocene side of the boundary, within Subzone CP15b. One notable difference in these samples and the Oligocene Sample 183-1137A-19R-CC is the larger size of Coccolithus, which reaches 19 µm (= C. eopelagicus) in Sample 183-1137A-21R-CC. Other differences may be more apparent than real, in that Sample 183-1137A-21R-CC contains some taxa reworked from the middle to lower Eocene, including Chiasmolithus solitus, C. expansus, Reticulofenestra onusta, Lapideacasis cornuta, and perhaps a single but well-preserved specimen of Discoaster saipanensis and several of Markalius inversus (which is common in Sample 183-1137A-21R-CC and rare in 20R-CC).
Nannofossils were very rare in well-cemented, moldic packstone at the top of Sample 183-1137A-22R-1, 10-14 cm, consisting mostly of a few of the Mesozoic Watznaueria barnesae. Sample 183-1137A-23R-CC, however, contains abundant, well-preserved coccoliths, presumably of the same assemblage as in the superjacent core. These include Biscutum coronum, B. magnum, B. dissimilis, B. constans, Arkhangelskiella cymbiformis, Eiffellithus touriseiffelii, Ahmuellerella octoradiata, Teichorhabdus ethmos, Chiastozygus garrisonii, Misceomarginatus pleniporus, Cribrosphaerella ehrenbergii, Repagulum parvidentatum, and several species of Cretarhabdus, but no Eiffellithus eximius, Aspidolithus parcus, or Nephrolithus. This assemblage is characteristic of the upper Campanian Biscutum coronum Zone, probably of the Repagulum parvidentatum Subzone (although the upper boundary of that subzone, the first occurrence of Nephrolithus corystus, was inconsistent in the Leg 183 sites, possibly because of preservation problems).
All core-catcher samples from the 200-m-thick diatom and nannofossil ooze succession yielded abundant and well-preserved planktonic foraminifers. The single core catcher from the Upper Cretaceous glauconite-bearing sandy packstone (Unit III) that lies unconformably below the Eocene ooze contains surprisingly well-preserved microfossils. Planktonic foraminifers are extremely scarce in this sample, however, because of dilution by common inoceramid prisms, volcanic glass, and mineral grains.
The Pleistocene is represented by ~10 m of diatom ooze in the first core (lithologic Unit I). Sample 183-1137A-1R-CC, contains a low-diversity planktonic fauna dominated by sinistrally coiled Neogloboquadrina pachyderma with occasional Globigerina bulloides and Turborotalia quinquelobula. This low-diversity assemblage belongs to the long-ranging biozone NK7, contains no age-diagnostic species, and is typical of the subantarctic upper Neogene (Pliocene, Pleistocene, and uppermost Miocene). Diatoms provide greater biostratigraphic control in this interval.
The biostratigraphic utility of planktonic foraminifers increases in the Neogene. Globorotalids are common in Sample 183-1137A-2R-CC. Owing to the presence of abundant Neogloboquadrina continusoa, Neogloboquadrina nympha, and Globorotalia scitula, we assign this sample to the uppermost Miocene zonal range NK5-NK6. The fauna in Sample 183-1137A-3R-CC is characterized by notably large forms of both Globorotalia scitula and globigerinids, including Globigerina bulloides, Globigerina falconensis and Globigerina woodi, and rather small Neogloboquadrina spp. This sample does not contain Globorotalia miozea, the marker for the zone below; therefore, we also assigned it to the NK5-NK6 zonal range. The size of faunal elements fluctuates and species dominance (globorotalids over globigerinids) varies throughout the Miocene. This may be an ecological effect, reflecting local or global variations in surface-water temperature. Alternatively, physical processes at the seafloor or during deposition may have winnowed assemblages. Because assemblages dominated by large forms also contain a range of small species, and vice versa, the former explanation is more probable. Sample 183-1137A-4R-CC, which we also place in Zones NK5-NK6, is dominated by rather small globigerinids, with less common, large (~250-300 µm) Globorotalia scitula.
Abundant, large middle Miocene globigerinids occur with Globorotalia miozea in Sample 183-1137A-4R-CC. Because of the presence of a small Neogloboquadrina-type form, we placed this sample in Zone NK5. We did not find this form in the subjacent two samples and, therefore, placed them in the zone below, Zone NK4. The downhole first occurrence datum of Globorotalia miozea occurs in Sample 183-1137A-7R-CC. The planktonic fauna in this sample is composed of large globigerinids and Globorotalia praescitula but no catapsydracids, indicting a late Zone NK3 age.
Globorotalia pseudoscitula does not occur in the subjacent two samples. Because of the presence of Globorotalia zealandica and Paragloborotalia incognita in assemblages dominated by rather small globigerinids, paragloborotalids, Globorotaloides suteri, and catapsydracids, we assigned these samples to the lower Miocene Zone NK2. No core-catcher sample was obtained for Core 183-1137-10R.
Paragloborotalia incognita is absent from Samples 183-1137A-11R-CC to 13R-CC. On this basis we assigned these samples to lowermost Miocene Zone NK1, above the top of the Subbotina euapertura range. We note the presence of extremely large Globigerina brazieri and G. cf. G. labiacrassata in Sample 183-1137A-12R-CC.
Globigerina euapertura, the nominate taxon of the upper Oligocene Zone AP16, first appears in Sample 183-1137A-14R-CC, accompanied by catapsydracids, tenuitellids, Globigerina brazieri, Globorotaloides suteri, and a variety of indeterminate globigerinids of upper Oligocene affinities. We assigned this and the subjacent sample to the AP15-AP16 zonal range. We could not delineate the boundary between AP15 and AP16 (defined by the last appearance datum [LAD] of Globigerina labiacrassata s.s). This was a result of the overlapping range of similar-looking Globigerina cf. Globigerina labiacrassata with the zonal marker.
Chiloguembelina cubenis is present in Sample 183-1137A-16R-CC, indicating that the top of Zone AP14 had been reached. In addition to C. cubensis, Samples 183-1137A-16R-CC and 17R-CC contain Globigerina praebulloides, Globigerina brazieri, and common tenutitellids. In the absence of Subbotina angiporoides, (the LAD of which marks the top of the zone below), we assigned these samples to Zone AP14. S. angiporoides is present in the next interval of samples, Samples 183-1137A-18R-CC to 20R-CC. On this basis, and in the absence of Globigerinatheka index, we assign these samples to Zone AP13. In addition to the distinctive lower Oligocene marker species Subbotina angiporoides, these samples contain catapsydracids, tenutitellids, Chiloguembelina cubensis, and Subbotina utilusindex that are difficult to distinguish from Subbotina linaperta.
The uppermost Eocene is indicated by Globigerinatheka index in Sample 183-1137A-21R-1, 75-79 cm. This sample, which we assigned to the AP11-AP12 zonal range, represents the stratigraphically lowest Cenozoic horizon above the disconformity. We cannot delineate the base of AP12 (the lower boundary of which is denoted by the LAD of Subbotina linaperta [Stott and Kennett, 1990]) because of difficulties in distinguishing S. utilusindex from the zonal marker S. linaperta. Huber (1991) noted that the range of S. linaperta extended into the lower Oligocene on the southern Kerguelen Plateau.
Core 183-1137A-22R contained no material in the core catcher, but glauconite-bearing sandy packstone in Sample 183-1137A-23R-CC yielded rare and moderately well-preserved planktonic foraminifers. This sample represents the stratigraphically lowest sediments above igneous basement. The planktonic assemblage included Globigerinelloides multispinatus, Archeoglobigerina australis, Globotruncana spp. and a single specimen of Globigerinelloides impensus, the nominate taxon for the upper Campanian biozone. Inoceramid prisms are also common in this sample.
Constraints on the age of the oldest sediment above basement in Cores 183-1137A-22R and 23R were derived by shipboard analyses of calcareous nannofossils, planktonic foraminifers, and paleomagnetic stratigraphy. Further shore-based study of the palynomorph assemblages may yield further age-diagnostic taxa.
The richest assemblages were in Core 183-1137A-23R, where we assigned the calcareous nannofossil assemblage to the middle to upper Biscutum coronum Zone and the planktonic foraminifers to the Globigerinelloides impensus Zone. Apart from the stratigraphic charts compiled by Cita et al. (1997), previous Southern Hemisphere high-latitude correlations do not show an overlap between these two zones (Huber, 1992). At this site as well as at Site 1135, however, there was a clear overlap between these two zones, with the extinction (LAD) of G. impensus recorded stratigraphically above rather than well below that of Aspidolithus parcus. We consider it unlikely that G. impensus is reworked at this locality because the site is located reasonably high on Elan Bank, above and presumably upcurrent from most likely sources of contaminants during Campanian times.
For the reasons outlined above, we have adjusted the LAD of G. impensus upward on the biostratigraphic correlation chart (see Fig. F6D in the "Explanatory Notes" chapter), to at least 74 Ma. This level is still within Chron C33n, as determined by Huber (1992) for this LAD. This correlation also adheres to our shipboard magnetic polarity data for this site (see "Paleomagnetism"). We noted a similar set of relationships in Core 183-1135A-39R on the southern Kerguelen Plateau. Assuming, therefore, that this correlation is correct, the basal sediment age is ~74 to 75 Ma. A more conservative estimate would be ~72 to 76 Ma.