Sediments recovered from Site 1084 represent a relatively continuous hemipelagic section spanning the last 4.7 m.y. Micropaleontological studies were carried out on core-catcher samples from Hole 1084A. Additional samples from within the cores were examined to improve the biostratigraphic resolution. An integrated biostrati-graphic framework composed of both calcareous and siliceous microfossils was established (Fig. 11), resulting in a well-constrained age model for Site 1084. Sedimentation rates range from 10 to 27 cm/k.y. with highest values located within the last 1 m.y. A second episode of high sedimentation rate (17 cm/k.y.) is associated with an upper Pliocene diatom-rich interval. These sediments (at an approximate age interval of 2–3 Ma.) are rich in the pennate diatom Thalassiothrix antarctica and may represent mat deposits similar to the ones discovered in the eastern equatorial Pacific during Leg 138.
Calcareous nannofossils were studied in core-catcher samples from Hole 1084A. Additional samples from within the cores were examined close to datum events to improve the stratigraphic resolution. Overall abundance ranges from abundant to very abundant within the top 280 mbsf (Cores 175-1084A-1H through 32X) and the bottom 190 mbsf (Cores 46X through base) of Site 1084. Nannofossil abundance shows minima within Cores 175-1084A-33X through 45X, a sedimentary section composed of diatom-rich clays (lithostratigraphic Subunit IC; see "Lithostratigraphy" section, this chapter) grading downcore to diatom oozes (upper part of Unit II). Preservation within this diatom-rich interval is moderate to poor, but improves in the over- and underlying carbonate-rich sequences.
Calcareous nannofossils provided 13 biohorizons (Table 3). These datum events were constrained within an average depth interval of 3 mbsf. Within the sampling resolution, the sedimentation appears continuous throughout the entire section. Based on the youngest identified datum (last occurrence [LO] of Gephyrocapsa caribbeanica acme, 0.26 Ma), no sediments recovered from Site 1084 were younger than 0.09 Ma (i.e., Zone NN21b is missing from the top cores). Paleomagnetic evidence (see "Paleomagnetism" section, this chapter) suggests that Site 1084 terminated within the lower Pliocene sediment at 4.8 ± 0.2 Ma. The oldest identified biostratigraphic datum event (LO of Amaurolithus tricorniculatus; 4.5 Ma, mean depth = 570.7 mbsf) is offset from the C3n.1n paleomagnetic Chron by ~30 mbsf, an indication of reworking of the nannofossil assemblages within the bottom part of the section. Despite the reworking of bottom assemblages, the sedimentation rate pattern suggests that Site 1084 terminated within Zone NN14.
Subzone NN21b, which describes the last 0.09 m.y., was not recognized in the uppermost core of Site 1084. The 0.26 Ma datum event (Zone NN21/NN20 boundary) was identified at the mean depth of 48.6 mbsf between Samples 175-1084A-6H-5, 110 cm, and 6H-6, 110 cm.
The LO of Pseudoemiliania lacunosa, which defined the Zone NN20/NN19 boundary, was constrained between Samples 175-1084A-9H-5, 110 cm, and 9H-CC (mean depth = 77.9 mbsf).
In addition to the zonal boundary events, six biohorizons were identified within this interval. These include the top and bottom ranges of Reticulofenestra asanoi, a species whose LO and first occurrence (FO) are close to the Brunhes/Matuyama boundary and the onset of the Jaramillo Chron, respectively. Sedimentation rates within Zone NN19 are the highest recorded throughout Site 1084, with a maximum of 27 cm/k.y. from 0.83 to 1.06 Ma. The LO of the Small Gephyro-capsa acme (Weaver, 1993), dated at 0.6 Ma (middle oxygen-isotopic Stage 15), was identified at the mean depth of 97.1 mbsf and can therefore be used to approximate the age of a major lithologic change within Hole 1084A (boundary between lithostratigraphic Subunits IA and IB; see "Lithostratigraphy" section, this chapter).
The Zone NN19/NN18 boundary (1.95 Ma) was identified at the mean depth of 313.5 mbsf with the LO of Discoaster brouweri between Samples 175-1084A-34X-CC and 36X-CC. Because of poor preservation of the nannofossil assemblages as well as barren samples, the depth location of this zonal boundary event could not be constrained further. Zone NN17 was lumped with Zone NN18 because of the sparse occurrence of D. pentaradiatus, whose LO datum event defines the Zone NN18/NN17 boundary.
The LO event of Reticulofenestra pseudoumbilica, one of the few non-Discoaster index species within the Neogene, was constrained within the interval bounded by Samples 175-1084A-60X-4, 108 cm, and 60X-4, 114 cm. This datum event defines the Zone NN16/NN15 boundary. The Zone NN17/NN16 boundary was identified at the mean depth of 425.5 mbsf (LO of D. surculus between Samples 46X-CC and 47X-CC).
This short interval (0.68 m.y. in duration) was identified between 552.9 and 570.8 mbsf. The base of Zone NN15 is probably offset by ~30 mbsf because of possible reworking of the index species Amaurolithus tricorniculatus, whose LO event defines the Zone NN15/NN14 boundary.
The uppermost Sample (175-1084A-1H-CC) is dominated by Globigerina bulloides and contains abundant Globorotalia inflata and Neogloboquadrina pachyderma. Other components of the assemblage include Globigerina umbilicata, Globorotalia hirsuta, G. scitula, G. truncatulinoides, and N. dutertrei (Table 4). The assemblages are generally dominated by G. bulloides and G. inflata (Table 4). Sporadic occurrences of Globigerinoides ruber indicate an input of a warmer fauna (Cores 175-1084A-7H-CC, 8H-CC, 14H-CC, 15H-CC, 25X-CC, 26X-CC, 38X-CC, 40X-CC, 41X-CC, 44X-CC, 48X-CC, and 50X-CC). The presence of G. ruber correlates with intervals of higher color reflectance values that are interpreted as probable periods of interglacial sedimentation (see "Lithostratigraphy" section, this chapter).
It is difficult to determine whether the absence of marker species is caused by ecological conditions or by selective dissolution. For example, the dominance of G. inflata may be ecological and indicate oligotrophic waters (e.g., the assemblages identified in the Benguela Current system by Little et al., 1997). It may also be a preservation signal; dissolution-susceptible species such as G. bulloides and G. margaritae, although expected, are commonly absent from the assemblages. Both cool (e.g., N. pachyderma (sinistral) and warm (H. siphonifera, G. ruber, and N. dutertrei) water faunas are present in the same assemblages downcore and may indicate an increased contribution from cooler Southern Ocean waters. N. pachyderma (sinistral) is generally present during the interval where subantarctic diatoms are found (Cores 175-1084A-6H-CC and 11H-CC through 50X-CC).
G. truncatulinoides is present consistently in Samples 175-1084A-1H-CC through 17H-CC and in Sample 175-1084A-30X-CC. Samples 175-1084A-19X-CC through 29X-CC and 31X-CC through 37X-CC are severely affected by dissolution, and so the exact position of the Pliocene/Pleistocene boundary, which is defined on the FO of G. truncatulinoides, could not be determined. It falls between Samples 175-1084A-30X-CC and 38X-CC. The boundary between Zones Pt1b and Pt1a occurs at the LO of G. tosaensis (~0.65 Ma; Sample 175-1084A-12H-CC) and is placed at 100 mbsf. Pliocene Zones PL3–PL6 are undifferentiated because of the absence of index species.
The species G. crassaformis viola is restricted to the late Pliocene and earliest Pleistocene. It is present in the interval between Samples 175-1084A-28X-CC through 55X-CC. A single specimen of G. margaritae, which is restricted to the early Pliocene, is present in Sample 55X-CC. It is not present elsewhere in the core and may be reworked because an early Pliocene age is in disagreement with calcareous nannofossil age estimates. Calcareous nannofossils also show evidence for reworking. Although G. margaritae is a temperate species, it is quite susceptible to dissolution.
Benthic foraminifers were studied in core-catcher samples from Hole 1084A. The overall abundance of benthic foraminifers at Hole 1084 is high with some exceptions: Samples 175-1084A-37X-CC (335.53 mbsf) and 49X-CC (450.97 mbsf) were barren of benthic foraminifers, and Samples 175-1084A-22X-CC (182.77 mbsf), 46X-CC (422.12 mbsf), and 52X-CC (477.83 mbsf) contained only traces or rare benthic foraminifers (Table 5). The preservation was good throughout Hole 1084A.
As at Site 1082 the benthic foraminiferal fauna studied at Hole 1084A correlates well with the different lithostratigraphic units recognized (see "Lithostratigraphy" section, this chapter). The benthic foraminiferal species dominating lithostratigraphic Subunit IA (nannofossil clay and nannofossil ooze) are Bulimina aculeata, B. exilis, and Nonionella turgida (Table 5; Fig. 12). The species B. exilis is present in high relative abundance throughout Hole 1084A, except for lithostratigraphic Unit IV, in which it is absent. The other two species are more or less restricted to lithostratigraphic Subunit IA.
Lithostratigraphic Subunit IB (diatom-bearing clay to nannofossil-rich diatomaceous clay with clayey nannofossil ooze) spans the uppermost upper Pliocene and lower Pleistocene sediments; it is dominated by B. exilis and Uvigerina hispidocostata (Table 5; Fig. 12). Several species are present in high abundances in a pulse-like manner in this interval; Bolivina seminuda (Sample 175-1084-13H-CC [117.72 mbsf], ~91%), Bulimina aculeata (24X-CC [206.98 mbsf], ~51%), Stilostomella spp. (25X-CC [219.65 mbsf], ~52%), Epistominella exigua (27X-CC [235.79 mbsf], ~20%), and U. auberiana (28X-CC [248.62 mbsf], ~21%; see Table 5; Fig. 12).
Lithostratigraphic Subunit IC (diatomaceous clay) is dominated by Cibicidoides wuellerstorfi and Pullenia bulloides; several other species are present, but in low relative abundances (Table 5; Fig. 12). Lithostratigraphic Unit II (clay-rich nannofossil diatom ooze to clay-rich diatomaceous nannofossil ooze) and Subunit IC are characterized by the common occurrence of diatom mats. The benthic foraminifer fauna is much the same as in lithostratigraphic Subunit IC, with the addition of overall high abundance of Bulimina exilis and Oridorsalis umbonatus and high abundance of Astrononion novo-zealandicum and Melonis barleeanum in the lower part (Table 5; Fig. 12).
Lithostratigraphic Unit III (nannofossil clay) spans the lowermost late Pliocene and is dominated by B. exilis together with Epistomin-ella exigua and Oridorsalis umbonatus (Table 5; Fig. 12).
The lowermost section of Hole 1084A reaches the lower Pliocene sediment and belongs to lithostratigraphic Unit IV (nannofossil ooze). The benthic foraminiferal assemblage in this interval is dominated by Bolivina subaenarensis, M. barleeanum, O. umbonatus, P. bulloides, and Uvigerina auberiana (Table 5; Fig. 12).
Radiolarians are abundant and well preserved in almost all samples examined down to 175-1084A-53X-CC (Table 6). From Sample 175-1084A-54X-CC through 65X-CC, radiolarians are abundant and generally moderately preserved. Radiolarian fauna indicates a Quaternary to early Pliocene age for Hole 1084A. No apparent reworking was identified.
The radiolarian zones used for this hole are those of Caulet (1991) and Motoyama (1996). Moore's (1995) tropical zonation, which has been applied to previous holes from Sites 1075 to 1083, was not used for this hole because of the complete absence of diagnostic taxa Buccinosphaera invaginata, Collosphaera tuberosa, Anthocyrtidium angulare, and Pterocanium prismatium, and because of the scarcity of Lamprocyrtis neoheteroporos. There are some difficulties in applying the established Antarctic zonations (Caulet, 1991; Lazarus, 1992) to the whole sedimentary sequence at Hole 1084A because of the absence of index species, such as Prunopyle titan, Helotholus vema, and Amphymenium challengerae, and because of apparent differences in stratigraphic ranges of Eucyrtidium calvertense and Stichocorys peregrina. The presence of Axoprunum angelinum and Cycladophora pliocenica, however, allows us to use a part of the Antarctic zones of Caulet (1991).
The absence of A. angelinum indicates that the uppermost cores (175-1084A-1H-CC through 8H-CC) are assigned to Zone NR1 of Caulet (1991).
Samples 175-1084A-9H-CC through 32X-CC belong to the Pleistocene Zone NR2 or Zone NR3/NR4 of Caulet (1991). Since Phormostichoartus pitomorphus was not identified, the zonal boundary between Zone NR2 and Zone NR3/4 was not determined. Caulet (1991) defined the bottom of the Zone NR3/NR4 by the last common occurrence of C. pliocenica. Considering the inconsistent occurrence of this species at Hole 1084A, the LO of the species in Sample 175-1084A-33X-CC is correlated with the bottom of Zone NR3/NR4, thereby approximating the Pliocene/Pleistocene boundary.
The FO of C. davisiana indicates an age of 2.7 Ma for Sample 175-1084A-48X-CC. Samples 175-1084A-33X-CC through 48X-CC between the LO of C. pliocenica and the FO of C. davisiana can be correlated with the P. prismatium zone of Moore (1995).
The presence of S. pylomaticus indicates that the age of the lowest sample (175-1084A-65X-CC) is no older than 5.2 Ma; thus, Samples 175-1084A-49X-CC through 65X-CC can be correlated to a zonal sequence from the lower part of the Cycladophora sakaii Zone to the S. pylomaticus Zone in the North Pacific (Motoyama, 1996).
Diatom counts and identifications were carried out using smear slides and acid-treated, sieved (20 µm) material from core-catcher samples from Hole 1084A (Table 7). Diatom preservation is moderate throughout Hole 1084A. As has been the case at Sites 1081 and 1082, overall diatom abundances are low or zero in the lower Plio-cene sediments. Diatom abundances reach a maximum in the late Pliocene (Table 7; Fig. 13). In contrast to the other sites, diatom abundances, although highly variable, remain moderately high (common to abundant) throughout the Pleistocene (Table 7).
In addition to the "background" diatom assemblage composed of a mixture of upwelling-indicator (Chaetoceros resting spores and Thalassionema nitzschioides var. nitzschioides) and oceanic species (e.g., Azpeitia nodulifer, A. tabularis, Hemidiscus cuneiformis, and Thalassiothrix spp.), we recorded many more cold-water markers characteristic of the Southern Oceans (e.g. Hemidiscus karstenii, Fragilariopsis kerguelensis, Proboscia barboi, Thalassiosira lentiginosa, and T. kolbei) than at Sites 1081, 1082, and 1083. The Antarctic/subantarctic flora is present consistently between Samples 175-1084A-50X-CC (~460 mbsf) and 30X-CC (~266 mbsf; Fig. 13) and sporadically in Samples 175-1084A-6H-3, 80 cm, and 6H-CC. Rhizosolenia hebetata, a cold-water indicator, is observed in Samples 175-1084A-11H-CC and 12H-CC, 16H-CC through 19X-CC, and 25X-CC (Table 7). The intervals of cold-water flora may indicate periods of increased advection of subantarctic waters into the Benguela Oceanic Current (BOC) and concomitant intensified northward Benguela Current transport, an assumption that is also supported by the presence of the Antarctic radiolarian C. pliocenica, and by the sparse occurrence of the low–middle latitude nannofossil genus Discoaster (Fig. 13).
Chaetoceros resting spores and setae of C. cinctus, C. debilis, C. diadema, C. lorenzianus, C. radicans, and C. constrictus/vanheurckii are particularly abundant in Pleistocene sediments, and they may be regarded as indicators of high coastal upwelling productivity from the Lüderitz upwelling cell (see Shannon and Nelson, 1996). They are also abundant during highest overall abundance times in the late Pliocene within Subunit IC and Unit II (see "Lithostratigraphy" section, this chapter; Fig. 13). Within the late Pliocene (at an approximate age interval of 2–3 Ma), core-catcher Samples 175-1084A-37X-CC (~335 mbsf) through 51X-CC (470 mbsf) are rich in the pennate diatom Thalassiothrix antarctica (and probably other Thal-assiothrix species as well), accompanied by lanceolate forms of Thal-assionema nitzschioides. Most of the Thalassiothrix group specimens are fragmented. Species of the genus Thalassiothrix are characterized by narrow (1.5–6 µm), long (400 µm to 4 mm), straight to twisted cells. They form an interlocking meshwork (this could readily be seen under a binocular scope) within which abundant Chaetoceros spores, Thalassionema nitzschioides, and calcareous nannofossils are found. In addition, scattered specimens of Actinocyclus curvatulus, A. octonarius, Actinoptychus senarius, Azpeitia nodulifera, Coscinodiscus spp., Hemidiscus cuneiformis, H. karstenii, Nitzschia reinholdii, N. fossilis, Proboscia barboi, and Stephanopyxis spp. are observed.
Site 1084 is located within the area of extension of the upwelling filaments in the frontal zone between the BCC and the BOC. The intervals of greatest diatom abundances in the late Pliocene are recorded by a mixed/Thalassiothrix-rich assemblage representative of two different oceanographic regimes that meet over Site 1084: (1) Chae-toceros spores and setae as a proxy of colder upwelling waters transported by the BCC and (2) mixed warm oceanic (e.g., A. nodulifera and H. cuneiformis) and Southern Ocean species (e.g., T. antarctica and P. barboi) as proxies of BOC waters. These intervals apparently represent mat deposits, which may be similar to the ones discovered in the eastern equatorial Pacific during Leg 138 (Kemp and Baldauf, 1993). The fact that these Thalassiothrix-rich sediments occur during persistent subantarctic water-mass influence at Site 1084 (Fig. 13) may relate to more vigorous surface circulation leading to the development of stronger frontal systems (among the BCC, the upwelling filaments, and the BOC), facilitating the concentration of diatom cells in the plankton and, consequently, greater downward flux through the water column (see Kemp et al., 1995, and references therein). Alternatively, silicate-rich subsurface waters were advected from the Southern Ocean, or by poleward undercurrents, or both.
Because of the mixture of warm, temperate, and Southern Ocean species and the occasional lack of biostratigraphic markers from either oceanographic regime, diatom biostratigraphic zonations are difficult to apply to Hole 1084A; consequently, they are not given here. Instead, biostratigraphic events, such as the FO or LO of a species, have been assigned to the sample containing the first or last observed specimens (Table 7; Fig. 11).