X hits on this document

PDF document

California Institute of Technology, GPS Division, Pasadena, California 91125, USA - page 3 / 4





3 / 4

that the source of Fe(II) in the Precambrian ocean is thought to be hydrothermal (Holland, 1973), a relevant question is whether a layer of anoxygenic Fe(II)-oxidizing phototrophs living below a layer of cy- anobacteria could have accounted for the complete oxidation of hydro- thermal Fe(II) entering the ancient ocean at depth before any Fe(II) potentially could have reached the oxic layers of the water column (Fig. 3A).

To calculate the thickness of this layer, we set the flux of Fe(II)

from the deep ocean (J


C/ z


C/z) equal to the

experimentally determined oxidation rate per cell multiplied by the average planktonic cell density and by the layer thickness (z). The value of z is thus a function of the diffusion coefficient (D), the pho- totrophic oxidation rate (time, t) per cell {[1/ ] [dFe(II)/dt]}, the cell density ( ), and the difference ( C) between the concentration of Fe(II) in the lowest part of the ocean, i.e., 0.5 mM (Holland, 1973; Morris, 1993), and the zero concentration of Fe(II) assumed in the case of

complete oxidation by the bacteria: z2



Note that because the cell density is multiplied by the oxidation rate per cell, an explicit term for cell density does not appear in the equation. Figure 3B predicts the thickness of a layer of anoxygenic phototrophs that would be required to completely oxidize Fe(II), as- suming a range of values for the photosynthetic Fe(II) oxidation rate per liter of water [dFe(II)/dt, representing the average oxidation rate for the entire layer] to account for different concentrations of bacteria that may have been present, as well as for different oxidation rates per cell associated with different light intensities and different Fe(II) con- centrations. The diffusion coefficient has also been varied to account for lack of knowledge of the physical processes responsible for mixing in the deep Precambrian ocean. Using the modern global mean eddy- diffusion rate and an Fe(II) oxidation rate relevant for deeper water

[0.014 mM/day: the rate determined at 150 lux (3

mol quanta m



1), 0.07 mM/day, reduced to 20% to account for the narrow wave-

length range in deeper water (Fig. 2B)], a 17.6-m-thick layer would be sufficient to oxidize all of the hydrothermal Fe(II) before it could reach the cyanobacterial-oxic zone. Assuming a higher Fe(II) oxidation rate, a slower Fe(II) diffusion rate or a lower concentration of Fe(II) in the ocean predicts an even thinner layer and vice versa. This is shown for Fe(II) oxidation rates varying one order of magnitude from the exper- imentally determined oxidation rate (Fig. 3B). A 17.6-m-thick layer of anoxygenic phototrophs as calculated from the experimentally deter- mined Fe(II) oxidation rate (as well as thicknesses of 5.6 and 55.6 m calculated for tenfold lower and higher oxidation rates, respectively) is reasonable given that anoxygenic phototrophs have been reported to extend from depths of 60 m to 110 m in the Black Sea (Repeta et al., 1989). This implies that even if cyanobacteria were present in the surface mixed layer, anoxygenic phototrophs could have been respon-

sible for total Fe(II) oxidation.

The amount of Fe(III) minerals that could have been precipitated

in an ancient ocean by anoxygenic phototrophs can be estimated using the thickness predicted by this model (17.6 m) and the corresponding Fe(II) oxidation rate [0.014 mM Fe(II)/day]. This is a conservative estimate with regard to light intensity because light-intensity– dependent oxidation rates determined for strain F4 were much higher ( 4–7 times) than those obtained for strain SW2 (Fig. 2A). Using an area equivalent to that covered by the Hamersley Basin in Western Australia [ 1011 m2 (Konhauser et al., 2002)], 9.0 1012 mol Fe/yr could have been oxidized and then precipitated by anoxygenic phototrophs. This is the same order of magnitude as the maximum rate

necessary to deposit the Hamersley BIF (4.5 hauser et al., 2002).


mol Fe/yr; Kon-

A caveat to consider is whether light attenuation by cells in the water column positioned above the Fe(II) oxidizing phototrophs would

GEOLOGY, November 2005

Figure 3. A: Simple box model of ancient stratified ocean with cya- nobacteria colonizing surface mixed layer, above layer of anoxygen- ic photoautotrophic Fe(II)-oxidizing bacteria, with a hydrothermal source of Fe(II) underneath. Concentration of dissolved Fe(II) is as- sumed to be ~0.5 mM (Holland, 1973; Morris, 1993). Question mark and double arrow point to thickness of layer (calculated in B) of anoxygenic phototrophs supported by Fe(II) input. B: Calculation of thickness of layer of Fe(II)-oxidizing anoxygenic photoautotrophs supported by Fe(II) flux from deep ocean. Shown is dependence of this thickness on Fe(II) oxidation rate for different diffusion coeffi- cients (D): (1) molecular diffusion of dissolved Fe(II) (D 1 10 cm2/s; Sobolev and Roden, 2001); (2) eddy diffusion measured for open oceans (D 0.1 cm2/s; Ledwell et al., 1993); and (3) theoreti- cally predicted eddy diffusion coefficient for modern ocean (D 1 cm2/s; Wunsch and Ferrari, 2004). Dotted lines indicate predicted thickness for this layer (17.6 m), assuming Fe(II) oxidation rate at low wavelength conditions as determined experimentally (1.4 10 M/day). Dashed lines and gray bar indicate thicknesses for this layer (5.6 and 55.6 m) if oxidation rates were one order of magnitude lower 5 5

or higher.

appreciably affect the rate of Fe(II) oxidation that we have assumed. Accordingly, we determined the spectral absorption cross section of phototrophic Fe(II) oxidizers (i.e., the cell-number normalized absorp- tion spectra), and calculated how much a 17.6-m-thick layer of anox- ygenic phototrophs would reduce the light intensity. Our calculations suggest that this effect would be minimal and would still allow enough light to be present for efficient Fe(II) oxidation, even at the base of this layer (Fig. 1). Similarly, the light intensity attenuation coefficients of phytoplankton and of natural waters containing cyanobacterial chlo- rophylls (Kirk, 1994) suggest that Fe(II) oxidation rates would be neg- ligibly affected by an overlaying layer of cyanobacteria (not shown in Fig. 1). Because our calculations are based on conservative assump- tions, even considering limiting factors such as temperature, nutrients,


Document info
Document views13
Page views13
Page last viewedWed Jan 18 02:25:39 UTC 2017