Benthic Foraminiferal Zn/Ca, a New Tracer of Deep Water Palaeocirculation

Abstract

Deep water formation has been implicated as an important driving force of major climatic changes in the geologic record. Our knowledge of the past distribution and circulation of deep water masses has primarily been derived from 813C and Cd/Ca measurements of fossil benthic foraminifera. For example, North Atlantic Deep Water (NADW) today carries signatures of high 813C and low Cd; benthic foraminiferal data suggest that during the last glacial maximum (LGM), NADW shoaled to become Glac ia l North At lant ic In termedia te Water (GNAIW), allowing low-~13C, high-Cd Antarctic Bottom Water (AABW) to penetrate farther into the North Atlantic (e.g. Boyle and Keigwin, 1982; Duplessy et al., 1988). Unfortunately, these two tracers often give conflicting or ambiguous results (Boyle, 1992), indicating the need for additional proxies. Zn holds the promise of being another useful palaeotracer. Its dissolved profile in the modem ocean is nutrient-like, with near-zero concentrations in surface waters and maximum concentrations below 1000 m depth (Bruland et al., 1978). Dissolved Zn covaries strongly with dissolved Si, probably because the two elements have similar rates and sites of uptake and regeneration, in analogy to the relationship between Cd and P. Deep water formation and circulation create large gradients of Zn and Si in today's deep ocean. Dissolved Zn concentrations range from ~1 nmol kg -1 in the deep North Atlantic to ~9 nmol kg -1 in the deep North Pacific. There is a particularly large (five to seven-fold) increase in Zn between the deep North Atlantic and the deep Southern Ocean. This is much larger than the corresponding Cd gradient, and suggests that Zn may be a very sensitive palaeotracer of AABW and its interactions with NADW. Several criteria must be met before Zn can be used as a palaeoceanographic proxy. First, modem benthic foraminifera must accurately record the Zn concentrations of overlying bottom waters. We test this hypothesis by performing a coretop calibration, in which we compare ZrdCa ratios of Holocene foraminifera to modem seawater Zn concentrations predicted from seawater Si measurements. Secondly, glacial-age foraminifera must also be accurate Zn recorders, with no serious diagenetic influences on Zn/Ca ratios. We make a preliminary examination of this issue by measuring Zn/Ca in several LGM samples. Finally, the oceanic inventory of Zn should ideally be constant with time. The oceanic residence time of dissolved Zn is poorly constrained, with estimates ranging from 3 to 30 kyr. We will assess any potential change in the Zn inventory by measuring Zn/Ca in the deep glacial equatorial Pacific, where Zn fluctuations due to circulation changes are expected to be small. If significant changes are evident, a glacial global inventory could eventually be made and a whole-ocean correction could be applied to glacial data, such as that used for 813C (Duplessy et al., 1988).