The Paleocene–Eocene thermal maximum (PETM, ~55 Ma) was a time of abrupt global warming, when both deep-ocean and equatorial sea-surface temperatures increased significantly (Kennett and Stott, 1991; Thomas and Shackleton, 1996; Zachos et al., 1993, 2003). This event was accompanied by a foraminiferal benthic extinction event (BEE) (e.g., Miller et al., 1987; Thomas, 1998; Thomas and Shackleton, 1996) and a decrease in the carbon isotopic composition of marine carbonate of up to 3% (e.g., Bralower et al., 1995; Kennett and Stott, 1991; Thomas and Shackleton, 1996) during a period of ~30,000 yr and a gradual return to near-initial values during a period of ~150,000 yr (Norris and Röhl, 1999). These changes have been linked to massive release of biogenic methane from gas hydrate dissociation (e.g., Dickens et al., 1995, 1997) and imply that there were major perturbations in the global carbon cycle.
However, the nature of the response of marine export productivity to this event is controversial. Based on an increase in biogenic Ba (total Ba normalized for Ba crustal abundance) at Atlantic (Ocean Drilling Program [ODP] Site 1051) and Southern Ocean (ODP Site 690) sites, Bains et al. (2000) proposed that oceanic productivity increased during the PETM and may have been responsible for drawing down atmospheric CO2, thereby cooling climate. Other evidence supporting increased marine productivity during the PETM includes a peak in biogenic Ba in the Middle Eastern bathyal sediments (Schmitz et al., 1997), Sr/Ca in coccolith carbonate at Site 690 (Stoll and Bains, 2003), and dinoflagellate assemblages that are indicative of increased productivity or increased temperature in coastal oceans (Crouch et al., 2001). Several other studies (e.g., Kelly et al., 1996; Bralower, 2002) have found evidence for decreased productivity for this interval at open-ocean sites in the equatorial Pacific (Site 865) and the Southern Ocean (Site 690) based on nannofossil assemblages. Dickens et al. (2003) proposed an alternative explanation for the increase in barite enrichment in PETM sediments—namely that this may have resulted from enhanced barite preservation because excess dissolved barium was released from seafloor gas hydrates. However, Ba/Ca data in foraminifers during this time interval are not consistent with this observation (Hall et al., 2004).
Understanding export production and organic C burial in the equatorial Pacific could have a major impact on interpretations of the causes and effects of this event. This area is an important upwelling zone and contributes to oceanic export productivity today as likely was also the case during the Paleocene and Eocene (Huber, 2002). We present here some of the first high-resolution records of reactive P and barite, indicators of nutrient burial and export production, respectively, across the Paleocene/Eocene (P/E) boundary in the equatorial Pacific Ocean.
We measured bulk sediment concentrations of reactive phosphorus (the sum of oxide-associated, authigenic, and organic P; sequentially extracted from bulk sediment), and barite (BaSO4; wt%) at high resolution (2 cm) in P/E boundary sediments from ODP Leg 199 Hole 1221C (153.40–154.80 meters below seafloor [mbsf]). Reactive P was measured to separate the signal of P involved in biogeochemical cycling from detrital P. The measurement of barite (separated from bulk sediment) rather than biogenic Ba enabled us to identify the origin of the barite crystals (biogenic vs. diagenetic) by determining whether our samples resembled modern unaltered biogenic barite (Paytan et al., 2002).
Double peaks in total P and total Ba separated by a Mn peak and coincident with a low total Ca zone were measured in shipboard bulk sediment. Bulk Ca was measured on the ship to approximate calcium carbonate concentration in the P/E boundary sections from Sites 1220 and 1221 (Lyle, Wilson, Janecek, et al., 2002). These total concentration data were measured on bulk sediment samples using a lithium metaborate fusion procedure and are subject to shipboard analytical errors (e.g., error in weighing samples during high seas or instrumental noise levels varying with sea state; see discussion in Quintin et al., 2002). We compare our reactive P and total P values determined via sequential extraction in the laboratory to the shipboard total P values, which allows us to evaluate (1) the agreement between the measurements and (2) whether the biologically available P concentration (reactive P) is significantly different from total P. We compare barium concentrations associated with barite in each sample to the shipboard total Ba concentrations to determine the extent of detrital Ba influence on the total Ba concentrations.
We compare the reactive P and barite measurements to a suite of other biogenic productivity proxies (CaCO3, organic C, and biogenic silica) generated by Murphy et al. (this volume). The multiproxy approach to paleoproductivity reconstructions is useful because strengths of some proxies compensate weaknesses of others (see Faul et al. [2003] for detailed evaluation of proxies and references). For example, sulfate reduction can lead to the loss of barite; otherwise, barite is a good export productivity tracer with a high burial efficiency (up to 30%) relative to other biogenic components (e.g., organic C and silica) (Dymond et al., 1992; Paytan et al., 1993; Paytan and Kastner, 1996).
Reactive forms of P delivered to the sediment/water interface, such as P associated with organic matter, are transformed to authigenic P with increasing age and depth in sediments (Ruttenberg and Berner, 1993; Delaney and Anderson, 1997, 2000; Filippelli and Delaney, 1995, 1996; Faul and Delaney, 2000; Faul et al., 2003). Because of this process, referred to as "sink-switching" by Ruttenberg and Berner (1993), the authigenic P phase provides an extractable record of nutrient (P) burial and thus a good indication of P cycling in the ocean. Since the vast majority of P flux to the sediment is associated with organic matter (Delaney, 1998), reactive P in the sediment may also represent organic C burial. The P burial record is preserved in sediments even when the record of organic C is erased as a result of postdepositional oxidation (Anderson et al., 2001; Anderson and Delaney, in press). Here, we use reactive P as an indicator of nutrient P burial and also as a proxy related to organic C burial. We use the ratio of barite to reactive P in sediments as a measure of the ratio of organic matter exported from the surface ocean to depth to nutrients and organic C buried in the sediment. This comparison holds as long as barium and phosphorus are not remobilized diagenetically (Nilsen et al., 2003). We assess what these records may indicate about equatorial Pacific productivity and the potential role of export production on C sequestration across the P/E boundary.