ArticlesDissolved and particulate Fe in a hydrothermal plume at 9°45′N, East Pacific Rise:: Slow Fe (II) oxidation kinetics in Pacific plumes
Introduction
Most of the heterogeneous reactions occurring in hydrothermal plumes are driven directly or indirectly by oxidation-reduction reactions and phase transformations of iron. Initially, in the buoyant plume (BP), an estimated 40–90% of the vent fluid iron precipitates as sulfides within the first few meters above a black smoker vent orifice Campbell et al 1988, Edmond et al 1979b, German et al 1991, Kadko et al 1990, Mottl and McConachy 1990, Rudnicki 1995, Rudnicki and Elderfield 1993. Rapid dilution and cooling, accompanied by increases in pH and oxygen concentration, lead to precipitation of the remaining Fe as oxyhydroxides (Mottl and McConachy, 1990). In addition, a poorly constrained portion of the sulfides may settle out of the BP as relatively large particles, while another fraction undergoes oxidative dissolution, adding dissolved Fe (DFe) back to the BP. The net input of many less abundant metals from primary vent fluid to the deep ocean reservoir depends on their tendency to form sulfide precipitates (or other solids), the stability of those solids, and finally on adsorption/coprecipitation with Fe oxyhydroxide phases as they are formed.
A fundamental control on the formation of particulate Fe(III) phases in plumes is the oxidation rate of Fe(II). Calculated and experimentally estimated Fe(II) oxidation half-lives (t1/2) range from 2.1 min at TAG (Rudnicki and Elderfield, 1993) to 10 and 32–42 h at Gorda Ridge and Juan de Fuca Ridge, respectively Chin et al 1994, Massoth et al 1994, Massoth et al 1998. In addition, it has been suggested that variations in vent fluid chemistry may affect Fe oxidation rates in the BP or neutrally buoyant plume (NBP) (Lilley et al., 1995), but these effects have not received quantitative treatment. A better understanding of the controls on Fe transformations in plumes is needed to reconcile these few disperate findings.
Studies of particulate Fe in plumes, measured either directly or by optical proxy, have helped to identify plume locations, been used in comparisons of the magnitude of plumes, and allowed determination and interpretation of the minor element components of Fe particles in NBPs. Nevertheless, understanding of the formation of these particles has been limited because only a few studies have determined DFe in plumes, using methods that may or may not distinguish soluble Fe(II) from colloidal Fe(III), and only a subset of these has determined comparable DFe and particulate Fe (PFe) concentrations in the same plume James and Elderfield 1996, Massoth et al 1998. A general paucity of data for DFe relative to PFe has resulted in an incomplete understanding of Fe chemistry in plumes.
In this paper, we present data addressing the dissolved/particulate partitioning of Fe (and Mn) in a NBP at on the East Pacific Rise (EPR). Plume particle composition is well studied in this region and is notable for low PFe in the NBP, even relative to vent sites <200 km further north at 11°N Haymon et al 1993, Lupton et al 1993, Baker et al 1994, Feely et al 1994a. We present new estimates of Fe oxidation rates supporting our observations of high concentrations of dissolved Fe(II) at 9°45’N, which is oxidized slowly (time scale of a few hours) relative to plume dilution rate, with concurrent aggregation of colloidal Fe(III) products to filterable (>0.4 μm) sizes. Our findings at this site are used as a case study within a general consideration of the factors controlling PFe and DFe partitioning in NBPs associated with vents on all the major ocean ridges. We argue that the major control on the development of PFe in NBPs is the ambient deepwater chemistry (pH, O2) of the ocean basin into which the plume is mixed; site-to-site compositional variations of vent fluid endmembers appear to exert only a secondary influence.
Section snippets
Methods
In 1996 and 1997 a NBP at 9°45’ N on the EPR was sampled using two different techniques (Fig. 1). In April 1996, large volume in situ filtration systems Sherrell 1991, Sherrell et al 1999 were deployed on a hydrowire from the R/V Atlantis II (Voyage 132, Leg 25) to collect particulate samples on Millipore HA filters (0.45 μm pore size, 142 mm diameter). In December 1997, particulate and dissolved phases were sampled at similar depths from the same region (Fig. 1). During the 1997 cruise,
Dissolved and particulate Fe and Mn distributions in the neutrally buoyant plume
Dissolved Mn (DMn) in the NBP samples varied from 194 nM in the most proximal sample to 16 nM in the most distal sample (Fig. 1). Particulate Mn (PMn) is relatively constant throughout the NBP between 1 and 2 nM and is thought to result largely from older resuspended particles (Sherrell et al., 1999). Because of the slow oxidation rate of Mn, dissolved Mn behaves as a near-conservative tracer of vent fluid dilution on the hours-days time scale of plume ages represented by our samples Cowen et
Conclusions
- 1.
High concentrations of dissolved Fe (up to 320 nM) in proximal samples of a neutrally buoyant plume at 9°45′N on the EPR suggests that Fe(II) is not completely oxidized in the buoyant plume.
- 2.
Calculation of Fe(II) oxidation rates at ambient Pacific deepwater pH and [O2] gives Fe(II) half-life of 3.3 h at the 9°45′N vent site. This helps to explain low particulate Fe (∼10–20 nM) in the plume, representing <15% of total Fe in proximal samples (<1 km) but increasing to about 50% at several km
Acknowledgements
The authors acknowledge Greg Ravizza and Linda Godfrey for their superb efforts in obtaining the 1997 samples and for intellectual contributions. We received able shipboard assistance of Laura Magde, Kathy Sullivan, and the Captain and crew of the R/V Atlantis II and the Alvin Group. We are grateful to Dan Fornari and Sandy Williams for the loan of the CTD and transmissometer. Karen Von Damm, Gary Massoth, Frank Millero, Chris Sabine, Clare Reimers, and Peter Rona were generous in providing
References (53)
Patterns of event and chronic hydrothermal venting following a magmatic intrusionnew perspectives from the 1996 Gorda Ridge eruption
Deep-Sea Res. II
(1998)- et al.
Hydrothermal plumes along the East Pacific Rise, 8°40′ to 11°50′NPlume distribution and relationship to the apparent magmatic budget
Earth Planet. Sci. Lett.
(1994) - et al.
Scavenging rates of dissolved manganese in a hydrothermal vent plume
Deep-Sea Res
(1990) - et al.
Techniques for determination of trace metals in small samples of size-fractionated particulate matterPhytoplankton metals off central California
Mar. Chem.
(1999) - et al.
A comparison of the equilibrium constants for the dissociation of carbonic acid in seawater media
Deep-Sea Res
(1987) - et al.
On the formation of metal-rich depostis at ridge crests
Earth Planet. Sci. Lett.
(1979) - et al.
Ridge crest hydrothermal activity and the balances of the major and minor elements in the ocean
Earth Planet. Sci. Lett.
(1979) - et al.
Hydrothermal plume particles and dissolved phosphate over the superfast-spreading southern East Pacific Rise
Geochim. Cosmochim. Acta
(1996) - et al.
Hydrothermal plumes along the East Pacific Rise, 8°40′ to 11°50′NParticle distribution and composition
Earth Planet. Sci. Lett.
(1994) - et al.
The effect of hydrothermal processes on midwater phosphorus distributions in the northeast Pacific
Earth Planet. Sci. Lett.
(1990)