RESULTS

The ¹³C and ¹⁸O data from Holes 1220B and 1221C obtained from bulk carbonate and benthic foraminifers are shown in Table T1 and in Figures F1 and F2.

The carbon isotope excursion is present at both sites and appears on all ¹³C records. The magnitude of the excursion in Hole 1220B is 2.01 in the N. truempyi record, 1.35 in the Cibicidoides spp. record, and 1.44 in the bulk carbonate record. In Hole 1221C, the magnitude of the carbon isotope excursion is 0.96, 0.83, and 0.82 for N. truempyi, Cibicidoides spp., and bulk carbonate, respectively. The ¹³C records from Hole 1220B are larger in magnitude than those produced from Hole 1221C.

The ¹³C values from Leg 199 sites are lower than the magnitude observed at higher latitudes, such as the 2.6 excursion seen in N. truempyi at Site 690 in the Weddell Sea (Kennett and Stott, 1991) but is of similar magnitude to those observed in Deep Sea Drilling Project and ODP Sites 527, 738, and 1051 and the Alamedilla Section in Spain (–2.1, –1.6, –1.9, and –1.7, respectively) (Katz et al., 1999; Lu et al., 1998; Lu and Keller, 1993; Thomas and Shackleton, 1996). All of these sites (including ODP 690) contain small gaps in the ¹³C record like the Leg 199 sites because of some combination of carbonate dissolution or absence of appropriate foraminifers for isotopic analysis.

There appears to be a systematic offset between the ¹³C records of bulk carbonate, N. truempyi, and Cibicidoides spp. in both Holes 1220B and 1221C. The Cibicidoides spp. record consistently has the lightest values of ¹³C, whereas the bulk carbonate record consistently has the heaviest ¹³C values of both sites. The average offset between Cibicidoides spp. and N. truempyi is 0.21 ± 0.29 (N = 86); between Cibicidoides spp. and the bulk carbonate record, the offset is 1.10 ± 0.38 (N = 56); and the offset between the N. truempyi record and the bulk carbonate record is 0.86 ± 0.31 (N = 52). Our results differ from those of Katz et al. (2003), who report that ¹³C of Cibicidoides spp. is typically more positive than that of N. truempyi. We can only note that the interspecies offset in our results is internally consistent and is based upon well-preserved foraminifers in Hole 1220B that are selected with a narrowly defined species concept even if we have not identified the species of Cibicidoides used in our analysis. Unfortunately, Cibicidoides spp. is known to contain a large number of species that are likely to have distinctly different isotope fractionation effects, and these may account for the contrast between our results and those of Katz et al. (2003).

The PETM is recognized only in the benthic foraminiferal ¹⁸O records from Hole 1220B, where an excursion of –0.65 is observed in the N. truempyi record and –0.86 is observed in the Cibicidoides spp. record. In contrast, the bulk carbonate ¹⁸O values at this site actually increase across the P/E boundary to unusually heavy values of ¹⁸O, reaching up to +4.0. It is unlikely that these ratios represent actual deep-sea ¹⁸O values in a time when the world's oceans experienced a ubiquitous increase in temperature. These anomalous values are coincident with the carbonate dissolution horizon, where dolomite crystals are abundant. It is likely that the ¹⁸O values in this part of the bulk carbonate record are of diagenetic origin, rather than a primary signal. Bulk carbonate ¹³C in Hole 1220B also yields unusually positive ratios, suggesting that carbon isotopes have also been affected by diagenesis.

In Hole 1221C, neither the bulk carbonate record of ¹⁸O nor the benthic foraminiferal ¹⁸O records have a negative excursion related to the PETM. The data from the bulk carbonate record have much scatter and do not show a clear pattern, whereas the benthic foraminifer records show an inverse pattern from that expected: ¹⁸O values increase across the P/E boundary. We also attribute this unusual result to diagenesis.