4.2 Benthic Cr supply: the flux of Cr to bottom waters from pore waters
Pore waters were sampled from calcareous sediments (~3350 m water depth) collected from a shallower feature punctuating the abyssal plain (~4000 to 5000 m) in the Tasman Basin (Figure 1). Regional sedimentation rates are on the order of 1-2 cm kyr-1 (Cochran & Osmond, 1976). Organic matter is around 1% at the sediment surface, though upper sediments remain oxic and organic carbon content decreases with depth (Table 1, Table S7). Surface sediment Mn/Al (~0.016) and Fe/Al (~0.53) are higher than upper continental crust values (0.009 and 0.48 respectively, Rudnick & Gao, 2003; Table S7), suggesting slight authigenic oxide enrichment, though Fe-Mn oxides are not a dominant phase. Sediment Mn/Al begins to decrease below 11 cm, suggesting oxic conditions in the upper 10 cm and more reducing conditions below this depth.
Pore water dissolved [Cr] shows a shallow sub-surface maximum (0-1 cm below the sea floor) of 47.4 nmol kg-1, approximately an order of magnitude greater than local bottom waters (4.81 nmol kg-1, Figure 3), and pore water [Cr] generally decreases exponentially with depth. The lowest pore water concentrations are observed in the deepest samples (6.7 nmol kg-1 at 9-10 cm depth), which remain elevated relative to bottom water. This upper pore water Cr maximum is consistent with observations by Shaw et al. (1990). In contrast to our observations of maximum pore water [Cr] at the sediment surface, Shaw et al. (1990) observed Cr removal in the uppermost oxic sediments, above the near-surface pore water [Cr] maximum. The absence of Cr removal in our uppermost oxic sediments may reflect the strong differences in bulk sediment composition between the lithogenic-dominated sites in Shaw et al. (1990) and our carbonate-dominated site.
A near-surface pore water maximum in [Cr] suggests pore waters may act as a diffusive source of dissolved Cr to the ocean. Bottom waters composed of CDW are enriched in dissolved Cr and are isotopically distinct ([Cr] = 4.81 nmol kg-1, δ53Cr = 0.76 ± 0.03 ‰) from CDW at nearby stations not in contact with sediments ([Cr] = 3.95 nmol kg-1, δ53Cr = 0.84 ± 0.03 ‰, 2SD, n = 6 samples ≥ 3000 m at stations TS8 and PS2, Table S9). Given the gradient of [Cr] from the pore water to the overlying bottom water, the diffusive flux of dissolved Cr can be estimated using Equation 1
\(\text{Flux}_{\text{Cr}}\ =\ D_{s}\frac{\text{ΔC}}{\text{Δz}}\)Equation 1
where Ds is the Cr diffusion coefficient corrected for temperature and tortuosity (supplemental material, see also Abbott et al., 2015), giving an estimated benthic flux of ~3.2 nmol Cr cm-2yr-1.
To contextualize this, if global oxic sediments were characterized by a benthic source of similar magnitude to that observed at our site, the global benthic Cr flux would be comparable to or larger than riverine inputs, currently believed to be the dominate Cr source (Bonnand et al., 2013; supplemental material), consistent with box model estimates (Jeandel & Minster, 1987). While quantitative estimates of global benthic Cr fluxes await greater data availability across diverse sediment types, our data identify that benthic sources are at least locally important and support previous studies that: (1) have argued for benthic Cr fluxes based on globally distributed elevated deep water [Cr] (Cranston, 1983; Murray et al., 1983; Jeandel & Minster, 1987), and (2) have invoked elevated pore water [Cr] to explain Cr enrichments in planktonic foraminifera in sediments compared to water column samples (Remmelzwaal et al., 2019).
We use the bulk sediment composition to examine potential sources of Cr to pore waters by applying mass balance calculations based on the average composition of the upper 4 cm, reported Cr concentrations of these phases, and regional sedimentation rates (Table 1). These calculations suggest that neither lithogenic nor carbonate-hosted Cr can account for the Cr flux out of these carbonate-rich and detrital-poor sediments. Nor is the Cr content of surface phytoplankton (~1-21 ppm, Martin & Knauer, 1973) high enough to explain observations (Table 1). Therefore, we suggest a combination of mechanisms causing Cr enrichments in particles delivered to sediments can explain the observed pore water data. First, respiration may result in a relative Cr enrichment in particles reaching the seafloor because the release of Cr from particles appears decoupled from organic matter respiration (see sections 4.1, 4.3). At the same time, scavenging of particle-reactive Cr(III) may increase the Cr content on biogenic and non-biogenic particles. Regardless of which process(es) are involved, the Cr released from these sediments must largely be derived from the water column based on mass balance and is incorporated into the sediments as particle-adsorbed Cr that has been scavenged onto particle surfaces in the upper ocean (Connelly et al., 2006; Semeniuk et al., 2016; Janssen et al., 2020), and at intermediate depths (see section 4.3), while detrital Cr may be more inert. Mn oxides in surface sediments may also facilitate oxidative release of Cr(III) coupled to Mn reduction.
To better constrain the origin of dissolved Cr in pore water we calculate its δ53Cr from mass balance, using observed bottom and deep water [Cr] and δ53Cr (bottom water: [Cr] = 4.81 nmol kg-1, δ53Cr = 0.76 ± 0.03 ‰; deep water: ([Cr] = 3.95 nmol kg-1, δ53Cr = 0.84 ± 0.03 ‰, 2SD; supplemental material). Pore water δ53Cr is estimated at 0.34 ± 0.25 ‰ (2 SEM propagated error), which is heavier than silicate earth (-0.124 ± 0.101 ‰, Schoenberg et al., 2008). While the closeness of these ranges indicates that some lithogenic component may contribute to the pore water flux, mass balance calculations, combined with the likely refractory nature of lithogenic Cr (e.g. Bauer et al., 2019), indicate that it is unlikely that lithogenic Cr contributes significantly to the pore water flux. In contrast, pore water δ53Cr is lighter than dissolved δ53Cr in the upper 1000 m (δ53Cr ≈ 1.1-0.9 ‰) by a similar amount as the enrichment factor predicted by the global δ53Cr–[Cr] array (ε ≈ -0.7 ‰). In other words, removal of dissolved Cr to particles in the upper 1000 m with an enrichment factor following the global array would result in δ53Crparticulate similar to the calculated δ53Crpore water.
Although the exact origin of pore water Cr is hard to discern and may include biogenic, scavenged, and lithogenic sediment components, our mass balance calculations highlight the importance of internal processes acting to redistribute Cr within the water column, transferring dissolved Cr from upper waters to deep waters. Therefore, the benthic flux may act more as an attenuation of the Cr sink term associated with particle export than an entirely new Cr source. Pore water data from diverse sediment types are needed to determine whether a truly new contribution from lithogenic material may be regionally important in more lithogenic-rich sediments.