Calcareous Nannofossil Biostratigraphy and Biochronology at ODP Site 1123 (Offshore New Zealand): A Reference Section for the Last 20 Myr in the Southern Ocean

Abstract

The quantitative analysis of the calcareous nannofossil content yield in the 600 m thick succession drilled at ODP Site 1123 (offshore New Zealand), considered as a reference section for the Southern Ocean region, allowed the recognition of 43 bioevents distributed along the last 20 Myr. The correlation with the excellent magnetostratigraphic record resulted in the attribution of numerical ages for the position of the detected horizons. Many of the marker species used in previous zonation were detected also at ODP Site 1123, but others revealed to be absent or of scarce applicability. On the other hand, the good applicability of additional events was verified and proved to be useful for the biostratigraphic subdivision and correlation. The obtained average bio- and chronostratigraphic resolution is about 0.6 Myr along the whole section, which increases to about 0.3 in the Pliocene–Holocene time interval. The final result is a detailed southern mid-to-high latitude nannofossil biochronology for the last 20 Myr, which confirms that the ODP Site 1123 succession represents a reference section for the Southern Ocean.

1. Introduction

In 1998, ODP Site 1123 was drilled 410 km NE of the Chatham Islands (offshore New Zealand) at a depth of 3290 m [1] (Figure 1). It was drilled in order to document the stratigraphy of the northern slopes of the Chatam Rise and to establish the effect of the Deep Western Boundary Current on the Neogene sediment deposition. The succession is about 600 m thick and consists of almost exclusively nano ooze with numerous tephra levels in the upper part. The Marshall Paraconformity, e.g., [2,3], represents a major discontinuity that stops the continuity of the succession and connects Lower Miocene to basal Oligocene deposits, with about 12.5 Myr missing.

Several studies were carried out in recent decades on the ODP Site 1123 succession, which made it possible to reconstruct the paleoceanographic and paleoclimatic history of the Southern Ocean region, especially in the Plio-Pleistocene time interval, e.g., [4,5,6,7,8,9,10]. Particular attention was paid to the evolution of the Deep Western Boundary Current within the complex framework of deep ocean circulation in specific time intervals, e.g., [11,12,13,14,15].

Other studies focused on the bio- and chronostratigraphic features of the selected time interval, according to different groups of fossils, e.g., [16,17,18], but at present a detailed biostratigraphic reconstruction of the entire ODP Site 1123 succession is not available.

Calcareous nannofossils are remnants of calcareous unicellular brown algae widely used for the stratigraphic correlation of marine successions. The on-board preliminary analysis revealed that the calcareous nannofossil assemblages were abundant and generally well preserved in almost all the samples. It was soon clear that the ODP Site 1123 succession represented the most complete sedimentary record currently available for the Lower Miocene–Holocene interval in the study area, well subdivided with a relevant number of biostratigraphic events and the succession also accompanied by an excellent magnetostratigraphic record.

In this paper, we present the results of a quantitative study of the nannofossil assemblages of the ODP Site 1123 succession, which allowed us to identify 43 bioevents, significantly improving the results obtained from the on-board analysis, where a total of 22 events were detected [1]. The excellent magnetostratigraphic data [1] were recalibrated according to the ATS2020 [18] and integrated with the updated nannofossil biostratigraphy; as such, the age calibration of the recognized bioevents comes out in terms of numerical ages (Ma).

The final result was a detailed southern mid-to-high latitude nannofossil biochronology for the last 20 Myr, which confirmed that the ODP Site 1123 succession represents a reference section for the Southern Ocean.

2. Materials and Methods

Cores from ODP Site 1123 recovered a succession of nannofossil ooze, chalk and limestone [1]. The first 250 m were assigned to litostratigraphic Unit I and consisted of light gray clayey nannofossil ooze alternating with white nannofossil ooze, with abundant tephra layers, especially in the upper part. The subsequent 200 m were assigned to the lithostratigraphic Unit II, which was made of homogeneous light greenish gray nannofossil chalk. The lowermost part above the Marshall Paraconformity was about 130 m thick and was assigned to Unit III, and this part consisted of nannofossil mudstone, nannofossil chalk and clay-bearing nannofossil chalk; an interval of about 8 m of thickness was characterized by plastically deformed clayey nannofossil ooze and reported as “debris flow”, and it was present within this unit (Figure 2).

Samples from the nannofossil analysis were collected from cores 1123B-H-1 (H =HPC=Hydraulic Piston Core) to 1123B-H-52 and from 1123C-X-18 (X=XCB=Extended Core Barrel) to 1123C-X-29. Depths of samples were reported in revised meters of the composite depth (mcd) [1]. A total of 368 samples were examined in selected intervals, in addition to those examined onboard by one of the authors of the present paper (ADS). Samples were thus collected at an average distance of 75 cm, which guaranteed the maximum error of 35 cm for the detected bioevents. Standard smear slides [19] were made for all samples using an optical adhesive as a mounting medium and dried under an UV lamp. Calcareous nannofossils were examined by means of a light-polarized microscope at 1000–1600× magnification.

The taxonomy of most of the taxa considered in the present paper were referenced in [19,20,21,22,23,24]. Useful information is also available from the Nannotax3 website [25].

Quantitative methods were adopted to make one bioevent easily correlatable among different successions and to better highlight significant variations in the calcareous nannofossil assemblages [26,27,28].

In detail, bioevents based on species belonging to the Calcidiscus, Discoaster, Gephyrocapsa, Helicosphaera and Sphenolithus genera were estimated within a prefixed number of specimens of the same genus, while the other bioevents were evaluated within a minimum of 500 nannofossils > 4 µm in size.

The definitions of the bioevents detected in the present study (Figure 3) basically follow those proposed by Backman et al., 2012 [29], which are: lowest and highest occurrences of marker species, referred to as base (B) and top (T), respectively; levels in which marker species begin to be common or to decline, referred to as Base common (Bc) or Top common (Tc), respectively. Other biohorizons regarded as relevant are: the base and the top of the paracme (interval of the temporary absence of a taxon), referred to as PB and PT, respectively; the base and the top of the acme (the interval of sharp increase in the abundance of a taxon) referred to as AB and AT, respectively; the level where the replacement of a taxon with another occurs it was indicated as the abundance crossover (X).

The chronostratigraphic framework referred to Raffi et al., 2020 [18], for the Neogene Period and to Pillans and Gibbard, 2012 [30], for the Quaternary, and compared with the New Zealand Geological Timescale [31] (Figure 4).

Figure 3. List of nannofossil bioevents detected at ODP Site 1123 (X) and bioevents adopted as zonal boundaries in standard [15,18,19,32,33] and most recent zonal schemes for oceanic successions [29]; ps stands for present study, nd stands for not detected. Most of the nannofossils index markers are illustrated in Figure A1, Figure A2 and Figure A3 in Appendix A.

Figure 3. List of nannofossil bioevents detected at ODP Site 1123 (X) and bioevents adopted as zonal boundaries in standard [15,18,19,32,33] and most recent zonal schemes for oceanic successions [29]; ps stands for present study, nd stands for not detected. Most of the nannofossils index markers are illustrated in Figure A1, Figure A2 and Figure A3 in Appendix A.

Figure 4. Chronostratigraphic framework adopted in the present study. For the International units stages [18,30]; for the New Zealand regional stages [31].

Figure 4. Chronostratigraphic framework adopted in the present study. For the International units stages [18,30]; for the New Zealand regional stages [31].

3. Neogene–Quaternary Calcareous Nannofossils Biostratigraphy: State of the Art

The biostratigraphic usefulness of calcareous nannofossils, in the Neogene–Quaternary interval, was highlighted in the late 1950s. Pioneering studies are those of Bramlette and Riedel (1954) [34] and subsequently of Bramlette and Wilcoxon, 1967 [35], that represent the basis that gave rise to the standard zonal scheme by Martini, 1971 [32]. Later, the advent of the deep-sea drilling and the recovery of deep-sea sediments led to an ever-increasing improvement in nannofossil biostratigraphic schemes, from Okada and Bukry, 1980 [33] and to Backman et al., 2012 [29] (Figure 3), and this topic field also benefited from advances in other fields of Earth Sciences (e.g., magnetostratigraphy, radiometric dating, cyclostratigraphy, etc.), highlighting the high degree of bio- and chronostratigraphic resolution that this group of organisms can provide in dating sediments of marine origin.

In the following sections, we summarize the main nannofossil biostratigraphic events characterizing the last 20 million years.

3.1. Holocene–Pleistocene (Holocene–Gelasian = Haweran–Nukumaruan)

During the Holocene–Pleistocene time interval, gephyrocapsids are important components of the nannofossil assemblages, indispensable for the biostratigraphic subdivision and correlation of marine successions. In the present study, we adopted the following biometrically based definitions of gephyrocapsids [27,36,37], splitting the group into four categories: (i) Gephyrocapsa specimens with long axis < 4μm, reported as “small Gephyrocapsa”; (ii) Gephyrocapsa specimens with long axis ≥ 4μm and open central area (e.g., G. oceanica s.l. [27]; not including G. carribeanica), reported as “medium Gephyrocapsa”; (iii) Gephyrocapsa specimens with long axis > 5.5 μm (including G. carribeanica), reported as “large Gephyrocapsa” and (iv) Gephyrocapsa specimens with long axis ≥ 4μm, open central area and a bridge nearly aligned with the short axis, referred to as Gephyrocapsa omega (corresponding to Gephyrocapsa sp. 3 of [36]). In the Late Pleistocene, relevant bioevents are the first occurrence of E. huxleyi and the disappearance of P. lacunosa.

3.2. Pliocene–Late Miocene (Piacenzian–Tortonian = Mangapanian–Tongaporutuan)

In this time interval, the Discoaster genus provides the highest number of bioevents, even if the distribution of Ceratholitidae is also noteworthy in the Zanclean and Messinian Stages. They are represented by the Amaurolithus, Nicklithus and Ceratolithus genera, which, however, can be characterized by a very discontinuous presence. Other relevant bioevents are the disappearance of R. pseudoumbilicus and Sphenolithus spp.

3.3. Middle–Early Miocene (Serravallian–Burdigalian = Waiauan–Otaian)

In the Middle–Early Miocene, many useful biostratigraphic events are based on species belonging to the Sphenolithus genus (namely S. belemnos; S. disbelemnos; S. heteromorphus) that are generally well represented in the nannofossil assemblages. Remarkable bioevents in this time interval are represented by the last occurrence (T) of T. carinatus and by the first occurrence (B) of H. ampliaperta, even if the two species are often discontinuously distributed in oceanic sediments.

The Discoaster genus also provides useful bioevents, but the low degree of preservation can in some cases prevent the correct identification of some species, such as Discoaster exilis, which due to recrystallization problems can easily be confused with D. variabilis.

4. Calcareous Nannofossils Biostratigraphy and Biochronology at ODP Site 1123

Nannofossils are generally abundant throughout the sequence; the preservation is good, especially in the top part of the succession, down to about 200 mcd, where dissolution or recrystallization phenomena may be present. Reworking may occur throughout the sequence, in low frequencies.

The study of nannofossil associations at ODP Site 1123 preliminarily focused on recognizing the biostratigraphic events identified in the schemes of [32,33]. Subsequently, the possibility of comparing the data with the more recent zonal scheme of [29] allowed for a more detailed biostratigraphic subdivision of the succession.

Age estimates for the nannofossil bioevents were calculated according to the age model reconstructed for ODP Site 1123. The age model was based primarily on the magnetostratigraphic record, virtually complete from Chron C1n to Chron C6r (Figure 5 and Figure 6), with the correlation of the individual polarity chrons to the geomagnetic polarity time scale being based on shipboard biostratigraphical data from calcareous nannofossils, planktic foraminifera, radiolaria and the identification of the magnetochron reversal pattern [1]. All magnetostratigraphic reversal ages were recalibrated according to the ATS2020 [18].

The average sedimentation rate for lithostratigraphic Units I and II (Figure 2) is respectively 33.4 and 34.9 m/Myr and remarkably uniform (Figure 4 and Figure 5). Possible small hiatuses or condensed intervals of less than 100 kyr are reported at about 270 and 340 mcd (age model in [1]). A major inflexion point at ~465 mcd denoted a change in sedimentation, separating the relatively fast deposition of lithostratigraphic Units I and II from the much slower Unit III where the sedimentation rate decreases to ~11.3 m/Myr. The lowermost portion of Unit III is characterized by an increased sedimentation rate (~32 m/Myr), probably due to the presence of a debris flow.

In the present study, the magnetostratigraphy and the calcareous nannofossils biostratigraphy at ODP Site 1123 were integrated and the numerical ages for each recognized bioevent were obtained through the linear interpolation between magnetic reversals.

Forty-three datum levels have been recognized. Some of them appeared as marker boundaries within the available nannofossil zonal schemes (Figure 3). Other levels have shown their potentiality in stratigraphic correlation and have been easily detected at Site ODP 1123, thus confirming their biostratigraphic usefulness.

The depth datum levels were calculated as mid-points within two subsequent samples and the corresponding age estimates are shown in Figure 7. In

Table 1. Comparison between the age estimates (expressed in Ma) obtained in the present study for the nannofossil bioevents detected at ODP Site 1123 and previous calibration for the same or similar horizon. The 1 stands for present study, 2 stands for Oceanic succession calibration and 3 stands for Mediterranean succession calibration.

Epoch	Bioevent		Marker Species	1	2		3	Reliability Degree	Chronostratigraphic Remarks
Pleistocene-Holocene	1	B	E. huxleyi	0.26	0.29 [27]	0.24 [38]	0.26 [39]	good	Base Late Pleistoc. 0.13 [30]
	2	T	P. lacunosa	0.44	0.43 [29]		0.46 [40]	good	Chibanian GSSP-0.77 [41]
	3	PT	m Gephyrocapsa	1.00	1.06 [42]			good
	4	B	G. omega	1.00			0.96 [40]	good
	5	T	H. sellii	1.13	1.24 [37]		1.26 [40]	moderate
	6	T	C. macintyrei	1.13	1.60 [29]		1.67 [43]	unreliable
	7	PB	m Gephyrocapsa	1.31				good
	8	T	l Gephyrocapsa	1.31	1.25 [37]		1.24 [43]	good
	9	B	l Gephyrocapsa	1.72	1.59 [42]		1.61 [43]	good
	10	B	m Gephyrocapsa	1.85	1.71 [42]	1.69 [44]	1.71 [43]	good	Calabrian GSSP-1.80 [45]
	11	T	D. brouweri	1.96	1.93 [46]	2.06 [44] ~1.93 [47]	1.95 [43]	good
	12	T	D. pentaradiatus	2.61	2.39 [46]		2.51 [48]	good	Gelasian GSSP-2.58 [49]
Late Miocene-Pliocene	13	T	D. surculus	2.64	2.53 [46]	~2.37 [47]	2.55 [48]	good
	14	T	D. tamalis	2.84	2.76 [46]	2.87 [44]	2.82 [40]	good
	15	T	R. pseudoumbilicus	3.74	3.82 [46]	~3.75 [47]	3.85 [50]	good	Piacenzian GSSP-3.60 [51]
	16	Bc	D. asymmetricus	4.14	4.04 [29]		4.11 [50]	good
	17	B	P. lacunosa	4.14	4.00 [52]			good
	18	Bc	H. sellii	4.41			4.62 [53]	moderate
	19	T	D. quinqueramus	5.61	5.53 [29]			good	Zanclean GSSP-5.33 [54]
	20	T	N. amplificus	6.07	5.98 [29]		5.85 [55]	unreliable
	21	B	N. amplificus	6.27	6.82 [29]		6.69 [55] 6.21 [56]	unreliable
	22	B	A. delicatus	7.15			7.13 [55]	good	Messinian GSSP-7.24 [57]
	23	B	A. primus	7.47	7.45 [58]	7.39 [29]	7.42 [55]	good
	24	T	M. convallis	7.68	7.78 [59]		7.78 [55]	moderate
	25	B	D. quinqueramus	8.12	8.10 [44]			good
	26	Bc	D. pentaradiatus	9.45			9.37 [55]	moderate
	27	T	C. coalitus	9.84	9.62 [58]	9.65 [29]		unreliable
	28	B	M. convallis	9.71	9.75 [29]		9.23 [55]	moderate
	29	Bc	D. bellus	10.47	10.64 [58]		10.38 [59]	moderate
	30	B	C. coalitus	10.77	10.89 [58]	10.79 [29]		unreliable
	31	Tc	C. miopelagicus	11.15	11.04 [58]	10.61 [29]	10.90 [59]	moderate
	32	Tc	D. kugleri	11.25	11.61 [58]	11.60 [29]	11.60 [59]	unreliable	Tortonian GSSP-11.61 [60]
Early-Middle Miocene	33	B	D. kugleri	11.90	11.89 [58]	11.88 [29]	11.90 [59]	unreliable
	34	Tc	C. premacintyrei	12.51	12.57 [29]		12.51 [61]	good
	35	T	S. heteromorphus	13.53	13.60 [58]		13.59 [62]	good	Serravallian GSSP-13.82 [63]
	36	T	H. ampliaperta	13.89	14.86 [29]			unreliable
	37	X	D.deflandrei/D.variab.	15.58	15.80 [29]			moderate	Langhian GSSP-15.99 [64]
	38	T	S. disbelemnos	17.76			17.69 [65]	to be tested
	39	B	S. heteromorphus	17.92	17.65 Bc [29]		17.99 [65]	good
	40	Tc	S. belemnos	18.19	17.94 T [29]		18.02 T [65]	moderate
	41	Bc	S. belemnos	18.73	19.01 B [29]		19.12 B [65]	moderate	Base Burdigalian Stage? [66]
	42	T	T. carinatus	19.29	19.18 [29]			moderate
	43	AT	C. abisectus	19.78				to be tested

, we provided a comparison of the ages obtained in the present study with the ages reported for the same horizon in the previous literature. An estimate of the degree of reliability is also given. The identified bioevents and the inferred ages are discussed in the following sections.

4.1. Holocene–Pleistocene (Holocene–Gelasian = Haweran–Nukumaruan)

The most recent nannofossil events recognized at ODP Site 1123 are the first occurrence of E. huxleyi, between samples 1123B-2H-7-10 and 1123B-2H-6-110, and the disappearance of P. lacunosa, between samples 1123B-2H-7-10 and 1123B-2H-6-110.

The Pleistocene Epoch is finely subdivided by the distribution of gephyrocapsids, with the first occurrence of the medium Gephyrocapsa (which approximates the Calabrian GSSP [45]) detected between samples 1123B-8-H4-30 and 1123B-8-H3-105. The biostratigraphically useful distribution interval of large Gephyrocapsa [27,37] was detected between samples 1123B-8-H1-105 and 1123B-8-H1-30 (Base) and between samples 1123B-6-H4-105 and 1123B-6-H4-30 (top). The last was detected as coincident to a paracme interval of the medium Gephyrocapsa; above this horizon, the nannofossil assemblages are dominated by small Gephyrocapsa specimens, up to the first occurrence of Gephyrocapsa omega, here detected between samples 1123B-5-H3-30 and 1123B-5-H2-105. This bioevent coincides with the end of the paracme interval of medium Gephyrocapsa, reported as a “reemG event” by [42].

Age estimates for bioevents from 1 to 10 (Table 1) compare well with the existing literature. Slight differences with the ages previously reported may be related to different methods of calibration (e.g., astrochronology) or to the different geomagnetic polarity time scales adopted [67].

Two further bioevents were reported as biostratigraphically useful in the present time interval, namely the last occurrences (T) of H. sellii and C. macintyrei, both recognized at the same level, between samples 1123B-6-H1-105 and 1123B-5-H5-30. The age estimation obtained for H. sellii T is slightly younger with respect to the previous calibration, thus revealing as moderately reliable, but the high degree of diachroneity characterizing the C. macintyrei T makes it an unreliable event.

From a chronostratigraphic point of view, we could observe that: (i) the base of the Late Pleistocene [30] falls between the base of E. huxleyi and the top of P. lacunosa; (ii) the base of the Middle Pleistocene (Chibanian GSSP at 0.77 Ma, [41]) falls between the top of P. lacunosa and the base of G. omega (coincident with the top paracme of the medium Gephyrocapsa); (iii) the Calabrian GSSP (1.80 Ma at Vrica section, [45]) is approximated by the base of the medium Gephyrocapsa.

Two more useful bioevents based on discoasterids were detected: (i) the disappearance (T) of D. brouweri, which represent the last stock of Discoaster before they became extinct (between samples 1123B-8-H5-105 and 1123B-8-H5-30). The estimated age of this bioevent at ODP Site 1123 fits well with the previous calibrations; (ii) the disappearance of D. pentaradiatus (between samples 1123B-10-H2-105 and 1123B-10-H2-30) that approximates the Gelasian GSSP (base of the Pleistocene epoch (2.58 Ma at Mt. St. Nicola Section [48,68]).

4.2. Pliocene–Late Miocene (Piacenzian–Tortonian = Mangapanian–Tongaporutuan)

The Pliocene epoch was adequately subdivided by the distribution of several species belonging to the Discoaster genus, which were well represented and preserved in this time interval at ODP Site 1123. We easily detected the top of D. surculus (between samples 1123B-10-H3-105 and 1123B-10-H3-30) and D. tamalis (between samples 1123B-11-H3-105 and 1123B-11-H3-30). Discoaster asymmetricus was sporadically present in low frequencies from core 33, but the common presence of the species occurred between samples 1123B-15-H7-31 and 1123B-15-H6-105, together with the first occurrence (B) of P. lacunosa.

The top of R. pseudoumbilicus between samples 1123B-14H-5-30 and 1123B-14H-4-105 and the common presence (Bc) of H. sellii between samples 1123B-16H-6-105 and 1123B-16H-6-30 provided useful tools for subdividing the Early Pliocene; indeed, taxa such as ceratholitids (e.g., C. rugosus and C. acutus) traditionally used as markers of this time interval are very rare or absent at ODP Site 1123. The disappearance of Sphenolithus spp. was not detected as low frequencies of this taxa are also present in the late Pliocene and Pleistocene, probably due to reworking.

The top of Discoaster quinqueramus between samples 1123B-21X-2-30 and 1123B-21X-1-105 indicated that we are very close to the Pliocene/Miocene boundary [18].

Pliocene bioevents are mainly based on Discoaster species, and the estimated ages at ODP Site 1123 well fit with previous calibrations (Table 1). We also provide ages for (i) the top of R. pseudoumbilicus, which approximates the Piacenzian GSSP [51]; (ii) the base of P. lacunose, that fits with the calibration obtained by [52] in central Indian Ocean and (iii) the Bc of H. sellii, slightly younger than the calibration obtained in the Mediterranean [53,69].

The Late Miocene can be subdivided based upon a series of fast-evolving taxa such as Amaurolithus and Catinaster. Although specimens of the Amaurolithus genus are discontinuously distributed within the ODP Site 1123 succession, we were able to detect the B of A. primus between samples 1123B-29X-1-102 and 1123B-29X-1-28, and the B of A. delicatus between samples 1123B-28X-2-105 and 1123B-28X-2-30. The ages obtained for these two bioevents were comparable with those available from the existing literature, and they fall between the Tortonian/Messinian boundary [57].

On the other hand, N. amplificus, considered a useful marker in the late Miocene, was discontinuously recorded in very few specimens. The base and the top of this species were detected between samples 1123B-23X-5-105 and 1123B-23X-5-30 and between samples 1123B-22X-6-105 and 1123B-22X-6-30, respectively. Nevertheless, none of the two bioevents can be considered reliable, as testified by the diachronous ages obtained with respect to previous calibrations already observed in Mediterranean sections [56,70].

Catinaster coalitus is discontinuously distributed in low frequencies from samples 1123B-36X-2-105 to 1123B-39X-2-30 and very rare specimens of C. calyculus occurred within the same interval, thus events based on Catinaster cannot be considered as reliable at ODP Site 1123.

Bad preservation of discoasterids was often aggravated by overgrowth and hampered the recognition of species such as D. hamatus, D. neohamatus, D. loeblichii and D. berggrenii. Nevertheless, we were able to detect some additional reliable bioevents based on the estimate ages and used them for stratigraphic correlations, namely: (i) Bc of D. pentaradiatus between samples 1123B-34X-5-123 and 1123B-34X-5-42; (ii) Bc of D. bellus between samples 1123B-38X-3-105 and 1123B-38X-2-30 and B of D. quinqueramus between samples 1123B-31X-1-105 and 1123B-31X-1-30. Discoaster surculus is very discontinuous in the lower part of its distribution range, thus the Bc of the species was impossible to be established.

Further bioevents characterizing this time interval are the restricted distribution range of M. convallis, with the top found between samples 1123B-30X-1-30 and 1123B-29X-CC; the base was detected between samples 1123B-35X-5-30 and 1123B-35X-4-107. These bioevents ranged from 9.71 to 7.68 Ma, according to our results. The last common occurrence (Tc) of C. miopelagicus was recorded between samples 1123B-40X-4-30 and 1123B-40X-3-105, and the estimated age of this bioevent is in good agreement with the one reported by [58].

The short distribution range of D. kugleri (top between samples 1123B-40X-6-105 and 1123B-40x-6-30; base between samples 1123B-42X-3-30 and 1123B-42x-2-105) characterizes the Tortonian–Serravallian transition [59], but this species is very rare at ODP Site 1123. Thus, in spite of the good agreement of the inferred age for the base of this species, with respect to previous calibrations, neither the base nor the top can be considered reliable bioevents, even if the presence of D. kugleri may help identify the Tortonian GSSP [60].

4.3. Middle–Early Miocene (Serravallian–Burdigalian = Waiauan–Otaian)

Tc of C. premacintyrei, which proved to be a useful tool for stratigraphic correlation, occurred between samples 1123B-44X-cc and 1123B-44x-4-30. The age obtained for this bioevent is comparable with other calibrations, thus confirming its reliability.

T of S. heteromorphus was found between samples 1123B-48-cc and 1123b-48X-3-100, with an estimated age of 13.53 Ma. This event fits well with previous determinations and is the best nannofossil event for approximating the Serravallian GSSP [63].

In this interval, most specimens of Discoaster spp. are overgrown due to re-crystallization, and the recognition of species such as D. signus was impossible. In this condition, we only could observe the crossover (X) between Discoaster deflandrei and D. variabilis, which was detected between samples 1123B-49X-3-105 and 1123B-49X-3-29. The resulting age of 15.58 Ma is slightly younger than that calibrated by [29], but the bioevent can be considered moderately reliable and useful for approximating the Langhian GSSP [64].

T of H. ampliaperta is historically considered a biostratigraphically useful taxon (Table 1), but it was difficult to detect because the species was very rare and discontinuous at ODP Site 1123. Last specimens occurred between samples 1123B-49X-2-30 and 1123B-49X-2-105, but the resulting age of 13.89 Ma is about 1 Myr younger than that reported by previous calibrations.

In Early Miocene, a few significant bioevents were recognized based upon the Sphenolithus genus, namely the B of S. heteromorphus (between sample 1123C-22X-2-105 and 1123C-22X-2-35) and the Bc and Tc of S. belemnos, respectively, between sample 1123C-22X-6-23 and 1123C-22X-5-105 and between 1123C-24X-1-30 and 1123C-23-cc.

The estimated age of 17.92 Ma for the B of S. heteromorphus at ODP Site 1123 fit well with the one obtained by [65] in the Mediterranean, and is older than the one obtained by [29], which did however consider the Bc of the species.

The ages obtained for Tc and Bc of S. belemnos are partially in agreement with previous calibrations which nonetheless consider the T and the B of the species. The quality of these events could also be distorted by the presence of a debris flow that affects this part of the ODP Site 1123 succession.

It is worth mentioning that the base of S. belemnos represents one of the criteria used to recognize the base of the Burdigalian Stage, which still remains inconclusive [18,66].

Sphenolithus disbelemnos is widely known as a useful tool in biostratigraphy [28,65,71,72] and has been recorded since the base of the succession above the Marshall Paraconformity. The Tc of the species was found between 1123C-21X-6-102 and 1123C-21X-6-25 and had an age estimate of 17.76 Ma, comparable with that obtained in some Mediterranean sections [65,69].

The Top of T. carinatus (between 1123C-26X-2-30 and 1123C-26X-2-30) had an age of 19.29 Ma, comparable with the calibration by [29], but it can be considered only moderately reliable due to the scarce and discontinuous presence of the species.

The oldest nannofossil event recognized above the Marshall Paraconformity was a sharp drop in the abundance of large (>10 µm according to Rio et al., 1990 [27]) C. abisectus (between samples 1123C-28X-1-105 and 1123C-28X-1-30), previously observed in other Pacific successions [33,73] and considered a useful tool for detecting the basal Miocene, but no previous age estimates are available. The age obtained for this bioevent at ODP Site 1123 is 19.78 Ma and can be considered moderately reliable pending testing in other successions.

Below the Marshall Paraconformity, at 599.20 mcd, Early Oligocene nannofossil assemblages occurred. Based upon the calculated sedimentation rate, the age of the oldest sediments immediately above the paraconformity was estimated to be 19.9 Ma.

5. Conclusive Remarks

The quantitative analysis of the calcareous nannofossil content of ODP Site 1123 allowed a refined biostratigraphic subdivision via the recognition of 43 bioevents, whose numerical ages were estimated through a comparison with the excellent magnetostratigraphic record available for the same succession. An evaluation of the reliability of the detected bioevents was also given, through a comparison with previous age calibrations.

We were able to detect many of the bioevents used in previous nannofossil zonal schemes but were also to verify the absence or scarce usefulness of well-established stratigraphic markers such as the Ceratholithus and Catinaster genera and some species of the Discoaster genus (e.g., D. berggrenii, D. hamatus, D. loeblichii, D. neorectus and D. signus). In addition, the distribution of taxa traditionally considered as biostratigraphically useful, such as with N. amplificus and H. ampliaperta, was revealed to be poorly applicable to ODP Site 1123 succession. On the other hand, we were able to highlight additional bioevents which confirmed their potentiality for stratigraphic correlations and highly improved the bio- and chronostratigraphic resolution in the last 20 Myr in the Southern Ocean region.

The obtained average bio- and chronostratigraphic resolution, only considering the bioevents characterized by an acceptable degree of reliability, is of about 0.6 Myr along the whole section, which increases to about 0.3 in the Pliocene–Holocene time interval.

However, some uncertainties still remain, such as the low bio- and chronostratigraphic resolution that characterizes the upper part of the Burdigalian and the lower Langhian stages, as well as the need to test the validity of some additional bioevents (e.g., S. disbelemnos; C. abisectus) that should deserve further investigation.