Palaeoenvironment of Maastrichtian ostracods from ODP Holes 1049B, 1050C and 1052E in the Western North Atlantic

The Maastrichtian ostracods recovered from ODP Holes 1049B, 1050C and 1052E on the Blake Nose, Western North Atlantic, are investigated. The three sites are located on a depth transect encompassing middle to lower bathyal, Late Cretaceous palaeodepths. Fourteen samples ranging in age from early to late Maastrictian are investigated from Hole 1052E, which is the shallowest site. The early Maastrichtian G. falsostuarti–G. gansseri Zone of Hole 1052E yields rare ostracods. The species richness, abundance and faunal density are on average considerably higher in the late Maastrichtian R. fructicosa and A. mayaroensis Zones of Hole 1052E, possibly, at least partly, as a result of palaeoceanographical changes that were also responsible for the disappearance of the inoceramid bivalves at this location. A palaeobathymetrical comparsion among the late Maastrichtian ostracod assemblages recorded from Holes 1049B, 1050C and 1052E shows that the faunal density and mean number of taxa are inversely correlated with palaeodepth; however, the dominance of the platycopid genus Cytherella increases with palaeodepth. A dominance of platycopids may signify environmental stress related to low oxygen content. The dominance of the benthic foraminifer Nuttalides trumpeyi in the Late Cretaceous of Holes 1049B and 1050C provides additional evidence of oxygen deficiency. From a total of 28 genera recorded from Holes 1049B, 1050C and 1052E, 14 were previously recorded from Hole 689B, a high latitude hole in the Southern Ocean, and show that many ostracod genera display a wide latitudinal distribution in the Late Cretaceous deep sea, although more geographically restricted genera are also present, analogous with modern and Tertiary oceans.


INTRODUCTION
During the Maastrichtian, world-wide biotic changes occurred which included the global extinction of the inoceramid bivalves (MacLeod, 1994; and the tropical rudist reef faunas (Johnson & Kauffman, 1990). There were also major latitudinal shifts in the distribution of planktonic foraminifers and calcareous nannoplankton (Huber & Watkins, 1992).
According to Saltzman & Barron (1982), the formation of Late Cretaceous deep water occurred both in low latitude marginal seas by evaporation forming warm, saline bottom waters and at high latitudes by the subsidence of cool surface water. For the Maastrichtian, Barrera & Huber (1990) and Barrera et al. (1997) showed that planktonic and benthic foraminifera1 6"O isotopes provide evidence of a trend of cooling of surface and intermediate waters between 74 and 68 Ma at high southern latitudes. Superimposed on this trend, there is isotopic evidence of a temporarily short event in the early Maastrichtian (between 71 and 70 Ma) during which the oceanic circulation may have become more typical of the Recent thermohaline pattern with an intensified production of cool, well-oxygenated bottom and intermediate waters at high latitudes Barrera et al., 1997). Before and after this event, the oceanic circulation was probably more typical of the Late Cretaceous pattern with a stronger influence of warm saline plumes formed at low latitudes (Barrera et al., 1997). According to MacLeod (1994), the last inoceramids were adapted to warm, oxygen-deficient conditions. Their extinction was possibly, at least partly, a consequence of such changes in oceanic circulation as occurred between 71 and 70 Ma (Macleod, 1994). In addition, an increase in the s7Sr/x6Sr ratio of sea water at 71 Ma indicates intensified continental weathering, and may be an expected result of a global regression (Barrera et al., 1997). Such a regression may have drained the low-and mid-latitude epicontinental seas and temporarily reduced the formation of warm, saline bottom waters. The extinction of rudist reefs has been related to the disappearance of hypersaline surface water produced in shallow epicontinental seas (Johnson & Kauffman, 1990).
Little is known about how the Maastrichtian palaeoceanographic perturbations that occurred between 74 and 68 Ma affected nannofossils, benthic foraminifers and ostracods at lower latitudes. Low latitude sites with good recovery and preservation through this time interval are rare. One of the objectives of Leg 171B, which included five sites drilled along a depth transect on the Blake Nose in the Western North Atlantic, was to study patterns of turnover in middle Maastrichtian microbiota. Benson et al. (1984Benson et al. ( , 1985 have previously shown the usefulness of ostracod studies in detecting major palaeoceanographic changes in the Cenozoic world ocean. The objective of the present preliminary study is partly to describe faunal changes in middle bathyal ostracod assemblages throughout the Maastrichtian at Site 1052. The intention is also to describe late Maastrichtian palaeobathymetrical changes in the composition of bathyal ostracod assemblages along a depth transect on the Blake Nose encompassing ODP Sites 1049, 1050 and 1052 ( Fig. 1). Swain (1978) and Guernet (1982) have previously described Cretaceous and Palaeogene ostracods, respectively, from Site 390 (DSDP Leg 44) on the Blake Nose. Previous studies, partly or completely focused on Late Cretaceous deep-sea ostracods (i.e. from a palaeodepth > 500 m) of the Atlantic and Southern Oceans include Benson (1975Benson ( , 1977, Benson et al. (1984Benson et al. ( , 1985, Damotte (1979Damotte ( , 1988, Dingle (1981), Swain (1973Swain ( , 1978Swain ( , 1983, Majoran et al. (1997),   .

METHODS
A total of 24 Maastrichtian samples (20 cm3 each) were studied from ODP Holes 1049B, 1050C and 1052E (see Table 1). All samples were dried and subsequently immersed in deionized water and placed on a rotary table for 24 h. They were then washed over a 63pm sieve and dried. The dried samples were sieved through a 125 pm sieve. All ostracods retained from this sieve fraction were picked under a binocular microscope and arranged on faunal slides. Almost all the specimens picked consisted of single valves or valve fragments. One valve was counted as one specimen, as were identifiable broken specimens that were greater than half the size of a complete valve. Broken specimens less than half the size of a complete valve were counted as fragments. The specimens were mainly identified to the level of the species, although some featureless genera, difficult to separate at specific level, were lumped together into multispecific categories, e.g. Cytherella spp., Krithe spp. and Argilloeciu spp. The ostracod valve accumulation rate (OVAR) (Majoran et al., 1997) was taken as a measure of the faunal density of ostracods, or more precisely the number of ostracods produced per unit area and unit time. The OVAR (the number of valves and unidentified fragments per cm2 and kyr) was calculated as NDS, where N is the number of valves and unidentified fragments per gram of sediment, D is the dry density of the sediment in g/cm3 and S is the sedimentation rate in cm/kyr. The OVAR corresponds to the benthic foraminifer accumulation rate (BFAR) of Herguera & Berger (1991), which has been suggested as a proxy for palaeoproductivity. The BFAR DMGUG.N.Atl.1-42. hypothesis is based on the assumption that the number of benthic foraminifera produced per cm2 and kyr is related to the supply of organic material to the seafloor which in turn is positively correlated with the productivity in the photic zone (Herguera & Berger, 1991). An exception is in areas with a pronounced oxygen minimim zone such as the Oman Margin where Naidu & Malmgren (1995) showed that the BFAR values were negatively correlated with the productivity of surface waters during the Holocene. They suggested that this may be due to the influence of low dissolved oxygen concentrations on benthic foraminiferal assemblages. According to Naidu & Malmgren (1995), BFAR may not be generally taken as a measure of productivity.

OSTRACOD DISTRIBUTION
Most specimens recorded are juveniles, which makes precise taxonomic determinations difficult. Most species are represented by only a few individuals. All ostracod species recorded were blind and therefore indicative of palaeodepths exceeding 700-800m (Benson, 1975).

Site 1052
Hole 1052E (29'57.08'N, 76'37.61'W) was drilled on the upper part of the Blake Nose at a water depth of 1343.5m, being one of a suite of holes associated with the shallowest site of Leg 171B. It penetrated 175.8 m of Maastrichtian sediments, which, on the basis of lithology, can be divided into two subunits: ( I ) a late Maastrichtian interval of 87.0 m (1052E-18R-3, 0 cm, to 1052E-27R-2, 38 cm), consisting of greenish grey to light greenish grey nannofossil chalk with clay to clayey nannofossil chalk with metre-scale alternations between lighter and darker intervals; and (2) an early Maastrichtian interval of 88.8m (1052E-27R-2, 38 cm, to 1052E-36R-3, 122 cm), consisting dominantly of light greenish grey to very light greenish grey nannofossil chalk to nannofossil chalk. The older subunit differs from the younger by generally being lighter and more carbonaterich, by a higher frequency of bioturbated intervals and apparent slump deposits, and by the presence of inoceramid shell remains (Norris et al., 1998).
The benthic foraminifers are represented by oligotaxic faunas in the Upper Cretaceous of Site 1052. Eouvigerina subsculptura constitutes the dominant element in the benthic foraminiferal community. This species is common in high trophic levels in the Maastrichtian of the Tethyan realm and is thus indicative of high productivity over the Blake Nose (Norris et al., 1998).
A total of 14 samples was investigated for ostracods from the Maastrichtian of Hole 1052E. Five of the samples were taken from the G. falsostuarti-G. gansseri Zone, two from the R. fructicosa Zone and seven from the A . mayaroensis Zone ( Table  2). The five samples of the G. falsostuarti-G. gansseri Zone (all within the lower Maastrichtian lithological subunit) are impoverished with respect to ostracods and contain mainly rare specimens of Cytherella spp. The species richness and abundance of ostracods are on average considerably higher in the nine samples of the R. fructicosa and A . mayaroensis Zones (all within the upper Maastrichtian lithological subunit) ( Table 2). From these nine samples a total of 594 specimens (not fragments) were recorded and a total of 47 species identified. Many species are represented by only a few or single specimens. The most dominant taxa are Cytherella spp. (total relative abundance of 30.3%), Krithe spp. (total relative abundance of 20.2%) and Argilloecia sp. 1 and spp. (total relative abundance of 9.6%). Other relatively common species are 'Brachycythere' sp., Bairdia spp., Cytheropteron spp. and Platyleberis sp. Rarer species with a total relative abundance > 1 9'0 include Aversovaha spp., Bythoceratina sp., Eucythere cf. circumcostata, Eucytherura sp. 2, Paraphysocythere sp., Imhotepia sp. and Projindobythere? sp. Of additional interest is the relatively high diversity of cytherurid species (six species of Eucytherura and four species of Hemiparacytheridea), although most of these species are represented by single or very few specimens. Figure 2A shows the variation in OVAR among the samples studied from Hole 1052E. In each calculation of the OVAR, the sedimentation rate was set to 2.2cm/kyr, which is the mean sedimentation rate for the Maastrichtian and Danian (Norris et al., 1998). In all the calculations, the dry density was set to 1.566 g/cm3, which is the mean density in the interval from core section 1052E-20R-1 to 1052E-35R-1 (Norris et al., 1998). In the G. falsostuarti-G. gansseri Zone, the OVAR is very low and ranges between 0.15 and 0.43. The OVAR increases across the zonal boundary between the G. .falsostuarti-G. gansseri and R. fructicosa zones and reaches a value of 7.97 in the oldest sample of the R. fructicosa Zone (1052E-27-1: 98-102) before it decreases to 0.81 in the subsequent sample of this zone (1 052E-26R-1 : 1 38-1 43 cm). The OVAR increases across the zonal boundary between the R. ,fructicosa and A . mayaroensis   zones and attains a value of 3.85 in the oldest sample of the A . mayaroensis Zone (1052E-25R-1: 113-117cm). The OVAR drops in the second oldest sample of this zone (1052E-24R-1: 109-1 13 cm) to a value of 0.39, before it gradually increases to 24.62 in the subsequent four samples of this zone. The OVAR of the youngest sample of the A . muyaroensis Zone is 17.63 (see Table 2; Fig. 2A). The variation in the number of species and specimens among the samples studied in Hole 1052E follows an almost identical pattern to the variation in OVAR ( Fig. 2A-C). It can possibly be inferred that the variation in OVAR is related to the variation in productivity, with relatively higher productivity in the late Maastricthian than the early Maastrichtian of Hole 1052E. Alternatively, the differences in OVAR between the early and late Maastrichtian are due to significantly lower oxygen concentrations at the sediment-water interface in the early Maastrichtian. It cannot be ruled out, however, that part of the differences observed between the early and late Maastrichtian ostracods are due to the effects of dilution from variations in the sedimentation rate. It is also possible that the more impoverished ostracod fauna of the early Maastrichtian (compared with the late Maastrichtian fauna) is a result of selective removal of taxa by winnowing or dissolution.

Site 1050
Hole 1050C (30°6.00'N, 76°14.10W) was drilled at a water depth of 2296.5 m and represents an addendum to Holes 1050A and 1050B. It was obtained to recover an equivalent sequence of the Cretaceous that was sampled at the shallower Site 1052. Five late Maastrichtian samples from the A . muyaroensis Zone were studied for ostracods from a lithological subunit consisting of nannofossil claystone, calcareous nannofossil claystone and nannofossil foraminifera1 chalk. A total of 117 specimens (not fragments) were recorded and 14 different species identified ( Table 3). The most abundant species are Cytherella sp(p). (total relative abundance 47.8%), Krithe spp. (total relative abundance 23.1 Yo) and Argilloecia sp. 1 and spp. (total relative abundance 12.0%). Other species are represented by only a few or single specimens. Cardobairdia sp. and Pterygocythereis sp. are two species recorded in Hole 1050C that were not recorded in Hole 1052E. Cardobuirdia is a deep-water genus common in the Caribbean Tertiary (Van den Bold, 1974) and also elsewhere in the deep sea.

Site 1049
Site 1049 represents a reoccupation of DSDP Site 390, which was drilled on the eastern margin of the Blake Nose lOkm downslope from Site 1050. It represents the deepest site of Leg 171B. Hole 1049B (30°08.54", 76'06.73'W) was drilled at a water depth of 2670.8m. Five samples were studied for ostracods from the R . fructicosa and A. mayaroensis zones of the late Maastrichtian of this core. They were obtained from a lithological subunit characterized by greenish grey light grey, and pale green clayey nannofossil ooze and clayey nannofossil chalk which is laminated to slightly bioturbated. Ostracods were rare in these samples, with a total of 25 specimens recorded and eight species identified (Table 4). Nine unidentified fragments of ostracods were also recorded. Cytherella sp(p). is the dominant species with a total abundance of 52.0%. Cardobairdia, with a

Polarity interval
C 3 0 N ? Site-Hole 1 0 4 9 6 may be responsible for the lower abundance of ostracods at the deeper sites. There is, however, no evidence of differential assemblage sorting among the various sites as the ontogenetic distribution of the various ostracod species are similar at all The mean OVAR, the mean number of taxa and the mean number of specimens were calculated for the late Maastrichtian of each of Holes 1049B, 1050C and 1052E. Each of these parameters is highest in Hole 1052E and lowest in Hole 1049B, and are thus negatively correlated with palaeodepth (Figs 3A-C). The palaeobathymetrical differences in OVAR and abundance (number of specimens) are probably not explained in terms of the effect of dilution of the ostracods by a higher sedimentation rate at greater depths. Norris et al. (1998) estimated the mean sedimentation rate for the Maastrichtian of Sites 1052, 1050 and 1049 to be 2.2, 1.7 and 0.36cm/kyr, respectively. Thus there is also a negative correlation between the sedimentation rate and the palaeodepth. Alternatively, it Table 4. Maastrichtian ostracods from Hole 1049B.  total of four specimens and a total relative abundance of l6%, is the second most abundant species. Other species are represented by one or two specimens only.

Core-section
Hole 1052E and 1050C are both dominated by the genera Cytherella, Krithe and Argilloecia (Tables 2 and 3). The most dominant species of Hole 1049B are Cytherella spp. (Table 4). Figure 3D shows the variation in the relative abundance of the most dominant species among the three holes. There is a greater dominance from the sum of the three dominant genera Cytherella, Krithe and Argilloecia in the late Maastrichtian of Hole 1050C than in Hole 1052E. The observed increase in dominance with an increase in palaeodepth is further noted for the genus Cytherella. The relative abundance of Cytherella is approximately 30% in Hole 1052E and approximately 50% in Holes 1050C and 1049B (Fig. 3D). An increase in dominance and a decrease in diversity generally imply increased stress levels. The higher dominance, particularly by Cytherella at the deeper sites, may possibly be explained in terms of environmental stress related to low oxygen content. According to Whatley (1991), Boomer & Whatley (1992 and Whatley et al. (1994), cytherellids may flourish under low oxygen conditions (contrary to many other ostracod taxa) due to their filter-feeding habit in that they manage to obtain a greater volume of water which they circulate across their respiratory surface. That the oxygen content was relatively low during the Maastrichtian of Sites 1049 and 1050 is consistent with the observations from benthic foraminifers, where the dominance of Nuttalides truempyi and the low numbers of Gavelinella beccariyormis may indicate that the benthic community was indeed influenced by low oxygen, warm, saline, deep water circulation (Norris et al., 1998). Thus it is possible that the differences in OVAR among the three holes are largely related to oxygen concentrations at the sediment-water interface.
There may be other explanations that possibly tie together the observed palaeobathymetric differences in species composition, OVAR, number of species and number of specimens among the various sites investigated, related to, for example, palaeoproductivity and palaeobathymetric differences in food supply and sediment characteristics. Table 5 lists previous records of the various ostracod genera identified from the Maastrichtian of the present sites studied. The previous records relate to studies of Campanian-Maastrichtian deep-sea ostracods from the North and South Atlantic and from the Southern Ocean and off southeast Africa (see Benson, 1975Benson, , 1977Damotte, 1979Damotte, , 1988Dingle, 1981;Swain, 1973Swain, , 1978Swain, , 1983Majoran et a/., 1997. The comparison with Dingle (1981) relates only to those species from his assemblages recovered from samples with an estimated palaeodepth > 500 m. The following taxonomic interpretations are presently made: Neocythere sp. 19 in Damotte, 1988belongs to Paraphysocythere Dingle, (1969; Bythocypris of Sites 689 and 698 in Majoran et a/. (1997) and

DISCUSSION
The replacement of the depleted ostracod assemblage of the G.
.falsostuarti-G. gansseri zones by the generally richer and more diversified ostracod assemblage of the R. fructicosa and A . mayaroensis zones in Hole 1052E occurs near the top of magnetostratigraphic subchron C31R (see Table 2). The replacement seems to post-date the short-lived global palaeoceanographic episode observed in southern high latitudes in the early Maastrichtian between about 71 and 70 Ma (see MacLeod, 1994). The boundary between subchrons C31R and C31N is dated to 68.657 Ma (Gradstein et al., 1994). It is important to note that the replacement of the ostracod assemblages seems to coincide with the disappearance of inoceramids. Inoceramid prisms occur in samples 10S2E-35l R , 110-113cm to 1052E-29R-1, 121-125cm, but were not found in samples 1052E-27R-I, 98-102cm to 1052E-20R-1, 100-103 cm (Norris et al., 1998). According to MacLeod (1994), inoceramids were adapted to life in warm, oxygen-deficient conditions on a substrate which had a low population of burrowing organisms and few or no shell-crushing predators. Their extinction was possibly a consequence of the global change in oceanic circulation during which warm, oxygen-poor bottom water was replaced by more vigorously circulating, cool, oxygenated Antarctic bottom water (MacLeod, 1994). That extinction was also accompanied by an increase in the populations of burrowing organisms. MacLeod (1994) Haq et al. (1987) is estimated to be S0m in sections in Alabama (see Barrera, 1994 and references cited therein). It is possible that the observed changes in the composition of ostracod assemblages across subchrons C31R to C31N are, in part, due to the global palaeoceanographic changes responsible for the extinction of the inoceramids. A lowering of the sea level across the early and late Maastrichtian boundary may, for example, have resulted in an increase in ventilation and the disappearance of the oxygen minimum zone from the location of Site 1052 and may also explain the increase in OVAR.