Benthic foraminifera and environmental turnover across the Cretaceous/Paleogene boundary at Blake Nose (ODP Hole 1049C, Northwestern Atlantic)

Abstract

Sediments recovered at lower bathyal ODP Site 1049 on Blake Nose (Northwestern Atlantic) offer an opportunity to study environmental changes at the Cretaceous/Paleogene (K/P) boundary relatively close to the Chicxulub impact structure on the Yucatan peninsula, Mexico. In Hole 1049C, the boundary is located at the base of a 9-cm-thick layer with abundant spherules, considered to be impact ejecta. Uppermost Maastrichtian oozes below, and lowermost Danian pelagic oozes above the spherule-bed contain well-preserved bathyal benthic foraminifera. The spherule-bed itself, in contrast, contains a mixture of shallow (neritic) and deeper (bathyal) species, and specimens vary strongly in preservation. This assemblage was probably formed by reworking and down-slope transport triggered by the K/P impact. Across the spherule-bed (i.e., the K/P boundary) only ∼7% of benthic foraminiferal species became extinct, similar to the low extinction rates of benthic foraminifera worldwide. Quantitative analysis of benthic foraminiferal assemblages and morphogroups in the >63-μm size fraction indicates a relatively eutrophic, stable environment during the latest Maastrichtian, interrupted by a sudden decrease in the food supply to the benthos at the K/P boundary and a decrease in diversity of the faunas, followed by a stepped recovery during the earliest Danian. The recovery was probably linked to the gradual recovery of surface-dwelling primary producers.

Introduction

During the last decades, the Cretaceous/Paleogene (K/P) boundary (commonly known as the Cretaceous/Tertiary (K/T) boundary) has been studied extensively, but there is still controversy regarding the origin of the characteristic K/P deposits and the cause(s) of the global mass extinctions. Most researchers accept the hypothesis that the deposition of a spherule-rich clastic sediment unit at the K/P boundary in sections around the Gulf of Mexico and North Atlantic was caused by destabilization on the continental margin caused by an asteroid impact at Chicxulub, on the northern Yucatan peninsula, followed by mass wasting (e.g., Bohor, 1996, Smit et al., 1996, Bralower et al., 1998, Klaus et al., 2000, Soria et al., 2001, Alegret et al., 2002a, Alegret et al., 2002b, Norris and Firth, 2002). Some, however, argue that these deposits resulted from changing sea-level and regional tectonics Keller and Stinnesbeck, 1996, Stinnesbeck et al., 1996, Keller et al., 1997. Most researchers accept that calcareous micro- and nannoplankton suffered a catastrophic, sudden mass extinction at the K/P boundary (e.g., Luterbacher and Premoli-Silva, 1964, Romein, 1977, Romein and Smit, 1981, Smit, 1990, Molina et al., 1998), although others maintain that the extinction was more gradual or stepped (e.g., Keller, 1989a, Keller, 1989b). At many locations, specimens of typically Cretaceous species of planktic foraminifera occur somewhat above the level of the most severe extinction marked by the signature of an asteroid impact such as high iridium-levels and spherules, but there is considerable evidence that these represent reworking (e.g., Huber et al., 2002, Soria et al., 2001).

Benthic foraminifera, in contrast, do not show significant extinction above background levels at the end of the Cretaceous (Culver, 2003), but show temporary changes in community structure, which have been interpreted as resulting from the collapse of the pelagic food web and the subsequent drop in food supply to the benthos (e.g., Thomas, 1990a, Thomas, 1990b, Kuhnt and Kaminski, 1993, Alegret et al., 2001a, Alegret et al., 2002a, Alegret et al., 2002b, Alegret et al., 2003, Culver, 2003).

During Ocean Drilling Program Leg 171B, the K/P boundary interval was cored at Blake Nose, Northwestern Atlantic (Fig. 1), ~1600 km from the Chicxulub impact site (Norris et al., 1998). A complete K/P interval was recovered in three holes (1049A, 1049B and 1049C) on the eastern margin of Blake Nose, at a present water depth of 2671 m. ODP Site 1049 was a re-drill of Deep Sea Drilling Project (DSDP) Site 390 (Benson et al., 1978). A paleodepth of about 2500 m during the late Maastrichtian has been assigned to this site Frank and Arthur, 1999, D'Hondt and Arthur, 2002. This paleodepth was constructed by backtracking, which in our opinion is not valid because the site is not located on oceanic basement Dillon and Popenoe, 1988, Norris et al., 2001b.

Maastrichtian benthic foraminifera from Site 390 were described superficially in the site chapter (Benson et al., 1978), where the paleodepth was estimated to be between 600 and 1600 m on benthic foraminiferal evidence. A few samples from Site 390 were included in the study by Widmark and Speijer, 1997a, Widmark and Speijer, 1997b, who argued for a paleodepth of about 1200 m, in agreement with Benson et al. (1978). Widmark and Speijer (1997b) tentatively placed Site 390 in a region of enhanced northern summer upwelling (their Fig. 2), and thus presumably enhanced seasonal productivity, but at considerably lower intensity than upwelling at the North African margin.

Various aspects of the sediment record from Blake Nose Site 1049 have been studied in detail. The geochemistry was documented by Martı́nez-Ruiz et al., 2001, Martı́nez-Ruiz et al., 2002, inorganic chemistry and mineralogy by Speed and Kroon (2000) and the planktic foraminiferal turnover across the K/P boundary by Norris et al., 1998, Norris et al., 1999 and Huber et al. (2002). No detailed benthic foraminiferal studies have been performed after the original shipboard description (Norris et al., 1998). We present the first detailed analysis of uppermost Maastrichtian and lowermost Danian benthic foraminiferal assemblages at Blake Nose, infer the paleobathymetry of the deposits, and illustrate the environmental changes across the K/P boundary.

We studied upper Maastrichtian and lower Danian samples from Hole 1049C (30°08.5370′N, 76°06.7271′W), where the K/P boundary is located at the base of a spherule-bed intercalated between the uppermost Maastrichtian and the lowermost Danian pelagic foraminiferal and nannofossil oozes (e.g., Huber et al., 2002). There is evidence for uppermost Maastrichtian soft-sediment deformation, but the series is reported to be biostratigraphically complete and without stratigraphic repetition Norris et al., 1999, Huber et al., 2002. Deformation features in the Maastrichtian part of the section have been related to seismicity caused by the impact at the K/P boundary Klaus et al., 2000, Norris and Firth, 2002.

We used samples from Core 1049C-8X. There are non-recovered intervals of several meters between Core 1049C-8X and the overlying and underlying cores. We limited our study to Core 1049C-8X, limiting our study to a fully recovered interval of sediment.

There is a sharp contact between the upper Maastrichtian ooze and the ‘spherule-bed’ at 1049C-8X-99 cm (113.09 mbsf). In Hole 1049C, this bed is a 9-cm-thick layer with green spherules (Fig. 3) that have been interpreted to be asteroid impact ejecta and consist mainly of smectite and carbonate Norris et al., 1999, Klaus et al., 2000, Martı́nez-Ruiz et al., 2001, Martı́nez-Ruiz et al., 2002. The spherule-bed contains reworked Cretaceous planktic foraminifera, shocked quartz, and clasts of dolomite, limestone and chert Norris et al., 1998, Norris et al., 1999, Klaus et al., 2000. The spherule-bed is topped by a 3-mm-thick limonitic layer with goethite concretions, which is enriched in iridium (Smit et al., 1997). The limonitic layer is overlain by a 7-cm-thick layer of dark grey, burrow-mottled clays, enriched in Ir, with Cretaceous and Danian planktic foraminifera (e.g., Norris et al., 1998, Klaus et al., 2000). Above this burrow-mottled clay is a 15-cm-thick white mud layer containing Danian planktic foraminifera and calcareous nannofossils. The white layer is thicker in Hole 1049C than in Holes 1049A and 1049B, and is overlain by a grey–green ooze with abundant, well-preserved, Danian planktic foraminifera and calcareous nannofossils.

We use the biostratigraphic zonation in the works of Norris et al. (1999) and Huber et al. (2002), who identified the upper Cretaceous planktic foraminiferal Abathomphalus mayaroensis Biozone in the lower four studied samples (Fig. 3). They did not find specimens of Plummerita hantkeninoides, which marks the uppermost part of the Maastrichtian, but they identified the calcareous nannofossil Micula prinsii Biozone, typical for the latest Maastrichtian. In the lower Danian, they recognized the Pα (Parvularugoglobigerina eugubina) and P1a Biozones Fig. 2, Fig. 3.

According to Norris et al. (1999), there may be an unconformity between the spherule-bed and the overlying, grey, mottled clays. In the K/P stratotype (El Kef, Tunisia), the boundary is defined at the base of a dark clay layer containing an Ir anomaly and microspherules (e.g., Cowie et al., 1989, Arenillas et al., 2002). Therefore, we consider that the K/P boundary at Blake Nose is located just below the spherule-bed (Smit et al., 1996), and we include this layer into the P0 Biozone (Fig. 3) following Arenillas et al. (2002), although Huber et al. (2002) left the spherule-bed unzoned.

We analyzed 30 samples from sections 1049C-8X-1 to 8X-5 (106.21–113.59 mbsf), comprising the upper 52 cm of the Maastrichtian and the lower 7 m of the Danian (Fig. 2). Samples are spaced at decimeter-intervals in the Maastrichtian, at centimeters in the lowermost Danian, and between 20 and 80 cm in upper section 8X-5 and sections 8X-4 to 8X-1.