DSDP Site 594, Chatham Rise, New Zealand - Late Neogene planktonic foraminiferal biostratigraphy revised

A restudy of the planktonic foraminifera at DSDP Leg 90, Site 594, has been made and consequently a revised scheme of late Neogene zones has been necessary. The Globorotalia sphericomiozea Zone (including G. conomiozea) is contracted to occupy a much smaller interval just below the Miocene/Pliocene boundary and the G. puncticulata, G. inflata and G. truncatulinoides Zones have been expanded. In the uppermost Miocene and Lower Pliocene, the standard regional planktonic foraminiferal datums provide a good correlation with DSDP Leg 29 Hole 284, DSDP Leg 90 Hole 593 and with New Zealand on-land palaeomagnetically-zoned sequences. They do not, however, correlate with the magnetochronology adopted in the Leg 90 Initial Reports for this interval in Hole 594 in which Chrons 4–6 are offset downwards against the foraminiferal datums. It is possible that there are unrecognized hiatuses in the Plio-Pleistocene. Higher in the Pliocene, very low species diversity of the cold-water faunas, scarcity of warm-water foraminiferal zonal species, and their much later appearance in Hole 594, contribute to a recognition that the Subtropical Convergence has been an effective barrier against the southwards penetration of most warmer water planktonic species. Neoacarinina blowi Thompson and Globorotalia crozetensis Thompson, are recorded from the New Zealand area for the first time


INTRODUCTION
DSDP Leg 90, Holes 594, 594a and 594b, were sited in the Bounty Trough on the southern edge of the Chatham Rise, due east of Oamaru at Lat. 45"31.41'S, Long. 179"56.88'E, in a water depth of 1204 m (Fig. 1).
The primary scientific objectives were to obtain a high resolution stratigraphic sequence from the Quaternary, to the late Palaeogene using the hydraulic piston corer and extended core barrel. The location, south of the Subtropical Convergence, is well placed to record oscillating glacial and interglacial episodes accompanied by alternating Subantarctic and cool subtropical assemblages of the fossil planktonic microfauna and microflora (Fig. 1). Hole 594, drilled to a subbottom depth of 1716.9m, recovered a late Early Miocene to Holocene sequence and Holes 594a and 594b were drilled to depths of 1851.3m and 1255.1 m respectively to recover intervals of poor recovery from cores 17, 22, 32 and 33.
The initial reports of DSDP Leg 90 recorded that many zonal indicators of the calcareous nannoplankton were very rare or absent and the zonation adopted had to rely on auxiliary species. Reworking of nannofossils from older deposits was common and their preservation was generally poor. The Miocene/Pliocene zonal boundary could only be approximately located and the Pliocene Zones "18, "17, "13 and the NN8/NN7 boundary could not be identified (Lohman, 1985). Siliceous microfossils, radiolarians, diatoms and silicoflagellates occur throughout the cored sequence, but the stratigraphic correlation of the Neogene t Dr Norcott de B. Hornibrook died on April 18th, 1994. radiolaria was difficult because they were very scarce and much diluted by terrigenous material and lacked warmwater species in the Pliocene (Caulet, 1985). Diatom assemblages present throughout the hole enabled recognition of 11 of 14 zones in the Middle Miocene to Quaternary Southern Ocean diatom zonation (Ciesielski, 1985).
Silicoflagellates were present in sufficient numbers to recognize 6 zones, somewhat modified for the cold-water character of the assemblages (Locker & Martini, 1985). Study of the Upper Miocene silicoflagellates from Site 594 by Bukry (1985) showed that large abundances of Mesocena and Distephanus between 180m and 290m indicated very cool conditions compared to their occurrence in more northern Leg 90 Pacific sites.
Recognizing that a revision of the foraminiferal biostratigraphy of this interval was necessary the authors re-examined shipboard splits of Cores 1-23 from Hole 594. The spacing of the available samples is not as close as the authors would have preferred as a basis for a more detailed revision, and the assemblages are sparse in some intervals. However, they have provided a basis sufficient to allow a much improved record of the foraminifera1 ranges. The resulting amended biostratigraphy (Table 1 and Plates 1 & 2) incorporates some revised and some new species identifications. The range of Globorotalia sphericorniozea, which had been combined with G. puncticulata puncticuloides by Jenkins & Srinivasan (1985), is shortened and confined to the interval from Core 20(4) 91-92cm to Core 20(2) 90cm. Globorotalia juanai, which is a new record, becomes extinct in Core 20(2) 90-92 cm. Cloborotalia crassaforrnis is recorded only in Core 19(4) 90-92 cm and in Core 15 cc; the upper part of its range, as initially recorded in Core 9 cc to Core 4(5) cc, in the Leg 90 Initial Reports (p. 828, table 16A), is referred to G. crassacarina Scott et al. (1990). The entry of G. crassula is recorded in Core 9 cc, and G. truncatulinoides is recorded earlier, from Core 7(2)  , 1985), this revision has resulted in a contraction of the G. sphericorniozea Zone, the upper part of which is included in the G. puncticulata Zone, and an expansion of the G. inflata and G. truncatulinoides Zones.

CORRELATION OF PLANKTONIC FORAMINIFERAL ZONES AND DATUMS
With the revised ranges of planktonic foraminifera from Hole 594, it is possible to correlate the Upper Miocene and Lower Pliocene part of the sequence with onshore New Zealand stages and also with DSDP Holes 284, 284A and 593 in the Tasman Sea (Figs 2, 3).
The Globorotalia rnioturnida Zone, present in Cores 23 cc    (Edwards, 1987;Wright & Vella, 1988). The interval of overlapping ranges of G. sphericorniozea, G. juanai and G. rnargaritae in Core 20 does not support Ciesielski's conclusion, (1985, fig. 3), that the basal Gilbert is missing in a hiatus. The base of the Opoitian Stage (basal Gilbert, Chron 4, c.5Ma) is at the base of the G. puncticulata Zone in Core 19 cc.  Edwards 1987.) Compared to the established magnetostratigraphic positions of these planktonic foraminifera1 datums in New Zealand on-land sequences, the following magnetostratigraphic correlatives adopted by Barton & Bloemendal (1985, fig. 26)  Either all of these datums occur much later in Hole 594 or the magnetostratigraphic interpretaion is incorrect as has been claimed previously by Wright & Vella (1988).
The low-diversity cool-water assemblages and the relatively large sampling intervals combine to make correlations with the Upper Pliocene and Pleistocene in the latitude of New Zealand, north of the Subtropical Convergence, rather uncertain above Core 19, particularly as G. crassaforrnis, which has useful coiling reversal zones in the north, is extremely rare. Two sinistral specimens in Core 19(4) 90-92cm are consistent with an Opoitian (Gilbert) age but the two dextral individuals in Core 1. 5 cc, if an adequate sample of the population, are more consistent with a Mangapanian (Late Pliocene, upper Gauss) age.
Individual morphotypes of Gioborotaiiu puncticulata puncticuioides appear in populations of G. puncticulatu puncticulata in Core 18 cc and more abundant typical individuals, with strongly compressed margins and flared chambers, dominate mixed populations from Core 16 cc upwards. A similar sequence is also present in Hole 284 where rare morphotypes of Globorotalia puncticulata puncticuloides appear prior to large populations of this subspecies, abundant in the lower part of Core 7 (Hornibrook, 1982) near the Gilbert/Gauss boundary which was recorded in nearby Hole 593 (Fig.1).
An obvious anomaly is the sudden appearance of typical 98 populations of Globorotalia inflata in Core 15 cc at the Gauss/Matuyama boundary in Hole 594 without any earlier record of the precursor transitional G. puncticulata-G. inflata populations which are usually present in the southwest Pacific. A sedimentary hiatus eliminating most of the Gauss in the interval of non-recovery in Core 17, proposed by Ciesielski (1985, fig. 3), could possibly account for the apparently late appearance of G. inflata although this species has not yet been recorded in the equivalent interval in Hole .594A, Core 10. Evidence that G. inflata appeared in Subantarctic waters around southern New Zealand about 1 Ma later than it did north of the Subtropical Convergence does not seem to support Wei's suggestion (Wei, 1991) that the enhancement of the Antarctic Intermediate water-mass might have created new niches and triggered off the evolution of this species at about 3.5 Ma.
Globorotalia crassacarina recorded by Scott et al. (1990) from the Nukumaruan to Castlecliffian, Upper Pliocene to Pleistocene in New Zealand, is recorded earliest in Hole 594 Core 10 cc, rare and coiling dextrally. This is in contrast to its dominantly sinistral coiling habit in Hole 284 in which it appears at the bottom of Core 5 (Lower Matuyama in Hole 593) about 1 Ma earlier than in Hole 594.
Globorotalia crassula appears first, rare and atypical in Core 9 cc, becoming more abundant and having an atypical morphology from 7 cc upwards as high as Core 3(2) 90-92 cm. Compared with its entry in Holes 284 and 593, at the Gauss/Matuyama boundary, its earliest appearance in the uppermost Matuyama in Hole 594 is about 1 Ma later.
Globorotalia truncatuiinoides appears first, rare and atypical, with a small umbilicus, in Core 7 cc, becoming more common in Core 3 cc which contains a fairly typical, strongly keeled population with a well developed, open umbilicus resembling individuals in the Pleistocene Castlecliffian and in the uppermost 3 cores of Hole 284   (Scott et a/., 1990), equal to Brunhes in adjoining Hole 593.
The last appearance of Globorotalia puncticulata puncticulata is recorded in Core 9(3) 90-92cm in the vicinity of the Gauss/Matuyama boundary, co-occurring with G. puncticulata puncticuloides. Its last appearance in Hole 593 is in Core 8 below the Gauss Chron (Hoskins, 1990) whereas Wei (1991) gives its last occurrence datum, possibly including G. puncticula~a puncticuloides, as c. 2.35 Ma, in the basal Matuyama. Hornibrook et al. (1989) and Scott et al. (1990) show the last appearance of G. puncticulata puncticulata in the Waipipian (Gauss Chron) and that of G. puncticulata puncticuloides high in the Nukumaruan (Upper Matuyama) and possibly in the Castlecliffian (Brunhes). Hornibrook (1982) recorded G. puncticulata puncticuloides from as high as Core 3 and possibly Core 2 (Brunhes) in Hole 284, but Hoskins (1990) recorded it from only as high as Core 6 (Lower Matuyama) in Hole 593.

NEW RECORDS OF SPECIES IN THE NEW ZEALAND AREA
Globorotalia crozetensis and Neoacarinina blowi were originally described by Thompson (1973), from a piston core RC8-39, Lat. 42"53'S, Long. 40°21 'E, from the western Crozet Basin in the southwest Indian Ocean, in Upper Pleistocene sediments. Neither species has been recorded previously from the southwest Pacific.
G. crozetensis (Plate 2, figs. 4-6) ranges from Core 18 cc (Lower Pliocene) to Core 2(2) 90-92 cm (Pleistocene) in Hole 594. An imbricate arrangment of chambers on the spiral side is particularly characteristic of this species which seems to have been restricted to cool seas.
The species of Globigerina, with a wide open umbilicus and tending to produce end kummerform chamberlets (Plate 2, fig. 9) is provisionally referred to G. umbilicata of Orr & Zeitzeff (1971), and was first recorded from the Pliocene of California.

IMPLICATIONS FOR CORRELATIONS BETWEEN THE SUBANTARCTIC AND THE COOL SUBTROPICAL ZONES
Hole 594 has been valuable for providing a record of late Neogene microfaunal and microfloral plankton south of the Subtropical Convergence during the Miocene, Pliocene and Pleistocene. It provides examples of low species diversity compared with the biota of the northern part of New Zealand and clearly demonstrates the intolerance to cold water of some species such as Globorotalia crassaformis which was common in warmer waters in the Pliocene from Lat. 42"s northwards. Unfortunately, doubts regarding the interpretation of the magnetostratigraphy and the presence of possible sedimentary hiatuses introduce uncertainty in evaluating the reliability of the foraminiferal biostratigraphy as a basis for north-south correlation. The recognition of a hiatus between 1.45 to 0.73Ma by Dersch & Stein (1991), based on their interpretation of the sedimentology, is an example of this problem.
Although planktonic foraminiferal faunal diversity south of the Subtropical Convergence was very low in the uppermost Miocene and Lower Pliocene in Cores 23-16 in Hole 594, it is possible to recognise the following New Zealand stages (Figs 2, 3): Upper Tongaporutuan, (Globorotalia rniotumida Zone, in part); Kapitean, (G. sphericomiozea Zone); and Opoitian, (G. puncticulata Zone, in part).
However, the correlation with the standard geomagnetic polarity sequence adopted by Barton & Bloemendal (1985), if correct, implies that all the standard regional foraminiferal datums, well integrated with magnetostratigraphy in on-land sections (Edwards, 1987), occur c . 1 Ma later at Site 594 than they do north of the Subtropical Convergence. While it could be argued that such an upward offset of regional datums was due to the effect of the Convergence, acting as a barrier to the southwards spread of warm-water species, their close spacing and their same order of appearance and disappearance as at DSDP Leg 29 Site 284 and Leg 90 Site 593 and in New Zealand on-land sequences, makes that reasoning seem unlikely. Furthermore, the late extinction level of the warm-water Miocene species, Globorotalia miotumida near the base of the Thvera in Hole 594, but observed to occur within the lower part of Chron 5 or earlier in the above-mentioned Sites (Fig. 3), implies the unlikely conclusion that this species was able to exist c. 1 Ma longer in cold southern waters, becoming extinct in the Early Pliocene when temperatures were rising again.
What is clear is that the following important foraminiferal datums occur in the same upwards order both south and north of the Subtropical Convergence: LOD Globorotalia miotumida, FOD G. conomiozea, FOD G. sphericomiozea, FOD G. margaritae, LOD G. juanai, FOD G. puncticulata, FOD G. crassaformis, FOD G. puncticulata puncticuloides (earlier in the south). The following species all appear later south of the Subtropical Convergence: FOD G. inJEata, FOD G. crassacarina, FOD G. crassula, FOD G. truncatulinoides s.1. (Fig. 3).
Globorotalia puncticulata puncticuloides populations are especially abundant and typical in Hole 594. Watkins et al. (1973) identified these forms as falling within the normal variation of Early-Middle Pleistocene Subantarctic G. puncticulata populations. Typical populations are especially common in the Upper Pliocene Lower Nukumaruan assemblages containing the cold water mollusc Chlamys delicatula (Hornibrook, 1981) in the North Island East Coast Basin of New Zealand, where they may have more potential value as indicators of incursions of cold Subantarctic water rather than as reliable index fossils for regional correlation.

The uppermost Miocene and Lower Pliocene in DSDP
Leg 90 Hole 594 in the Bounty Trough can be correlated, using planktonic foraminifera, with the same interval of that age in DSDP Leg 29 Hole 284 and DSDP Leg 90 Hole 593 in the Tasman Sea and in New Zealand with on-land sequences, but not with the interpretation of the magnetic polarity reversals by Barton & Bloemendal (1985: 1312).
Assuming that the interpretation of the magnetic polarity reversals through most of the Gauss and the overlying sequence by Barton & Bloemendal (1985) is dependable, and has not been affected by the presence of unrecognized hiatuses in sedimentation, their adopted magnetochronology would appear to be a more reliable basis for correlation with that of Hole 593 (p. 1306, fig.  22) than the foraminiferal biostratigraphy based on our rather widely spaced sampling intervals. Planktonic foraminifera have not been shown to be reliable for correlation across the Subtropical Convergence in the Late Pliocene at Leg 90 DSDP Site 594.