Available online at www.sciencedirect.com
Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
www.elsevier.com/locate/gca
The continental shelf benthic iron flux and its isotope composition
Silke Severmann a,*, James McManus b, William M. Berelson c, Douglas E. Hammond c
b
a
Department of Earth Sciences, University of California Riverside, Riverside, CA 92521, USA
College of Oceanic and Atmospheric Sciences, Oregon State University, Corvallis, OR 97330, USA
c
Department of Earth Sciences, University of Southern California, Los Angeles, CA, 90089, USA
Received 23 September 2009; accepted in revised form 19 April 2010; available online 27 April 2010
Abstract
Benthic iron fluxes from sites along the Oregon–California continental shelf determined using in situ benthic chambers,
range from less than 10 lmol m2 d1 to values in excess of 300 lmol m2 d1. These fluxes are generally greater than previously published iron fluxes for continental shelves contiguous with the open ocean (as opposed to marginal seas, bays, or
estuaries) with the highest fluxes measured in the regions around the high-sediment discharge Eel River and the Umpqua River.
These benthic iron fluxes do not covary with organic carbon oxidation rates in any systematic fashion, but rather seem to
respond to variations in bottom water oxygen and benthic oxygen demand. We hypothesize that the highest rates of benthic
iron efflux are driven, in part, by the greater availability of reactive iron deposited along these river systems as compared to
other more typical continental margin settings. Bioirrigation likely plays an important role in the benthic Fe flux in these systems as well. However, the influence of bottom water oxygen concentrations on the iron flux is significant, and there appears
to be a threshold in dissolved oxygen (60–80 lM), below which sediment–ocean iron exchange is enhanced. The isotope
composition of this shelf-derived benthic iron is enriched in the lighter isotopes, and appears to change by 3& (d56Fe) during
the course of a benthic chamber experiment with a mean isotope composition of 2.7 ± 1.1& (2 SD, n = 9) by the end of the
experiment. This average value is slightly heavier than those from two high benthic Fe flux restricted basins from the California Borderland region where d56Fe is 3.4 ± 0.4& (2 SD, n = 3). These light iron isotope compositions support previous
ideas, based on sediment porewater analyses, suggesting that sedimentary iron reduction fractionates iron isotopes and produces an isotopically light iron pool that is transferred to the ocean water column. In sum, our data suggest that continental
shelves may export a higher efflux of iron than previously hypothesized, with the likelihood that along river-dominated margins, the benthic iron flux could well be orders of magnitude larger than non-river dominated shelves. The close proximity of
the continental shelf benthos to the productive surface ocean means that this flux is likely to be essential for maintaining ecosystem micronutrient supply.
Ó 2010 Elsevier Ltd. All rights reserved.
1. INTRODUCTION
Within bioturbated or physically mixed continental shelf
and slope marine sediments, the cycling of iron can play an
important role in carbon degradation (Froelich et al., 1979;
Aller et al., 1986; Canfield and Des Marais, 1993; Thamdrup et al., 1994; Thamdrup, 2000). In such locations, the
*
Corresponding author. Present address: Department of Earth &
Planetary Sciences and Institute of Marine & Coastal Sciences,
Rutgers University, New Brunswick, NJ 08901, USA. Tel.: +1 732
932 6555x236.
E-mail address: [email protected] (S. Severmann).
0016-7037/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.gca.2010.04.022
high rates of electron transport, which are typically associated with carbon oxidation, rely on multiple electron transport pathways, with much of this electron transport taking
place within the upper few centimeters of the sediment
package. Carbon oxidation may thus generate a dissolved
pool of reduced iron within the sediment package which
can be transported to the overlying water column via fluid
advection and diffusion (Aller, 1998, 2004). This process is
likely to be an important route for trace metal supply to the
biologically productive waters typical of the continental
shelf (Johnson et al., 1999; Elrod et al., 2004, 2008; Lam
et al., 2006; Nishioka et al., 2007). This benthic–pelagic exchange can be quantified indirectly from porewater profiles
Continental shelf iron flux and its isotope composition
or directly from benthic incubation chamber experiments
(e.g., Klinkhammer et al., 1982; Heggie et al., 1986; Archer
and Devol, 1992; Johnson et al., 1992; Reimers et al., 1992;
Berelson et al., 2003). However, because these environments are physically and chemically dynamic, and because
sands rather than the more easy-to-sample muds often
dominate shelf environments (de Haas et al., 2002), it is difficult to quantify a global iron flux from these settings.
In this manuscript we present benthic incubation chamber data from muddy sites along the Oregon and California
continental margin (Fig. 1) to estimate the magnitude of the
benthic iron flux in these regions and the isotope composition of that benthic flux. Our work places a primary emphasis on continental shelf sediments neighboring two river
systems of the Pacific Northwest—the Umpqua River of
southern Oregon and the Eel River of northern California.
For comparison, we also discuss previously published and
new results from the central-to-southern California shelf
and Borderland Basin systems. Collectively these data suggest that sediments on continental shelves, particularly systems neighboring high sediment discharge rivers (e.g., the
Eel River) have a higher rate of iron sedimentary efflux than
more typical continental margin settings (e.g., Elrod et al.,
2004). The key characteristics that differentiate the margin
systems discussed here are the reactive iron delivery from
the continent, organic carbon rain rate, and the activity
of macrobenthos, and consequently the mode of benthic–
pelagic solute exchange (i.e., advective versus diffusive).
Our thesis regarding the high benthic efflux of iron is con-
Fig. 1. Map of our three study areas along the California and
Oregon continental margin (basemap created using the USGS map
generator Map-It at http://woodshole.er.usgs.gov/mapit/).
3985
sistent with previous observations of higher dissolved iron
concentrations in surface waters on the broader shelves
off northern California compared to central and southern
California (Chase et al., 2005a). Of particular importance
with respect to global processes, the high continental iron
efflux at the northern shelf sites occurs in the face of relatively high dissolved bottom water oxygen concentrations
(60–140 lM at our northern shelf sites compared to typically 610 lM in the San Pedro and Santa Monica Basins;
Elrod et al., 2004).
Furthermore, our data show that the benthic iron flux at
all of our sites is enriched in the lighter isotopes of iron,
probably reflective of isotope fractionation during iron
reduction and iron recycling in the sediments as well as in
the bottom waters (Severmann et al., 2006, 2008; Staubwasser et al., 2006). We have observed distinct differences in the
isotope compositions of the benthic iron efflux over the
course of the benthic chamber incubation experiments that
can be related to the depositional regime of these different
shelf systems. A key finding presented here implicates the
importance of reactive solid phase delivery via rivers for
iron release within sediments, which may lead to a large
advective or diffusive metal transport to the ocean.
1.1. Iron geochemistry
Despite its high crustal abundance, iron has a low solubility under the oxidizing conditions typical of the contemporary ocean, often resulting in low iron concentrations in
the upper ocean (Martin and Fitzwater, 1988; Martin and
Gordon, 1988; Martin et al., 1989, 1991; Cullen, 1991;
Kuma et al., 1996; Johnson et al., 1997; Coale et al.,
1998). Such low surface ocean iron concentrations can limit
ocean fertility with this “iron limitation” being most notably expressed within the high nutrient low chlorophyll
(HNLC) regions of the open ocean (references as above).
Here, iron supply is restricted because of low iron inputs
by atmospheric deposition and upwelling (Duce and Tindale, 1991; Johnson et al., 1997; Archer and Johnson,
2000; Fung et al., 2000; Jickells and Spokes, 2001; Cassar
et al., 2007).
In contrast to the open ocean, however, coastal systems
are proximal to a rich source of iron, as well as other microand macronutrients, i.e., the continent. Notwithstanding
this proximity, coastal systems can also be limited by iron
availability (Pakulski et al., 1996; Hutchins and Bruland,
1998), and even where production may not be limited by
available iron, ecosystem structure can be influenced by
iron availability (Brand et al., 1983; Brand, 1991; Sunda
and Huntsman, 1995; Hutchins et al., 1999; Crawford
et al., 2003; Leblanc et al., 2005).
One principal difference between the iron cycle of coastal
systems as compared to the ocean interior is that for coastal
systems it is believed that sediments are an important iron
source to the overlying water column (Croot and Hunter,
1998; Hutchins and Bruland, 1998; Hutchins et al., 1999;
Johnson et al., 1999, 2001; Anderson and Raiswell, 2004;
Elrod et al., 2004; Raiswell and Anderson, 2005; Raiswell,
2006; Elrod et al., 2008; Lohan and Bruland, 2008),
whereas dust and glacial erosion might play a more prom-
3986
S. Severmann et al. / Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
1980b, 2001; Canfield and Des Marais, 1993; Thamdrup
et al., 1994). During upwelling events the combined dissolved and fine particulate iron fraction can be delivered
to the overlying water column, as the benthic boundary
layer communicates with the surface ocean (Fig. 2). This
process can thus lead to elevated dissolved and fine particulate iron concentrations within the water column overlying
the continental shelf (Chase et al., 2005a; Elrod et al., 2008;
Lohan and Bruland, 2008).
inent role in the remote oceanic regions of the central
oceans and the Southern Ocean (Jickells et al., 2005; Mahowald et al., 2005, 2009; Cassar et al., 2007). The paradigm
that dust is the principal source of reactive iron to the surface ocean in remote regions is increasingly being challenged, however, and several studies have highlighted the
important role of coastal sediments as an iron source far beyond the immediate vicinity of the shallow shelf environments (Elrod et al., 2004; Lam et al., 2006; Nishioka
et al., 2007; Lam and Bishop, 2008; Raiswell et al., 2008).
Continental weathering supplies coastal regions with particulate iron, largely via river transport (for a recent review see
Poulton and Raiswell, 2002). Once it is delivered to the
coastal zone, terrestrial material makes its way to the seabed where it can be acted upon by physical, chemical,
and biological processes (Fig. 2, and Colbert, 2004; Wetz
et al., 2006). Under one scenario, physical resuspension of
fine particulate sediment provides a source of iron that
can be accessed by surface biota following its vertical transport to the euphotic zone (Hutchins et al., 1998; Chase
et al., 2005b). An extension of that idea asserts that terrestrial iron undergoes repetitive oxidation-reduction steps
within the benthos before a fraction of it is re-injected into
the bottom water, thereby creating a biologically-accessible
dissolved and fine particulate iron reservoir (Fig. 2, and
Johnson et al., 2001; Poulton and Raiswell, 2002). This
repetitive recycling is imposed upon iron because of the
constant interplay between iron oxide reduction coupled
to organic matter mineralization and iron oxidation coupled to advective supply of oxygenated bottom water.
Bioirrigation or other physical mixing processes likely dominate exchange at the sediment–ocean boundary (Aller,
2. STUDY SITES
A considerable amount of iron diagenesis research has
focused on deltaic systems dominated by large rivers (e.g.,
Aller, 1998; McKee et al., 2004); part of the emphasis for
this manuscript is on sites proximal to small rivers. We
present results from two cruises in 2007 along the southern
Oregon and northern California shelf (herein called BIF-I
and BIF-II cruises, where BIF stands for Benthic Iron
Flux). The BIF-I cruise took place in April/May and the
BIF-II cruise in September. Our BIF study sites are the
northern continental shelf (70–200 m water depth) adjacent
to the outflow of the Eel and Umpqua Rivers (Fig. 1). This
general region of the Pacific Northwest is one known for
having relatively small river basins, but elevated sediment
transport to the ocean (Griggs and Hein, 1980). The Eel
River in particular is a system with an exceptionally high
terrigenous sediment delivery, having an annual load of
1.5–2.5 1010 kg yr1, making this system the largest terrigenous sediment source on the US West Coast (Milliman
and Syvitski, 1992; Wheatcroft and Sommerfield, 2005).
The outer shelf to upper slope of this margin accumulates
Winter
Rivers
FeOOH
FeOOH
FeOOH
FeOOH
Spring/Summer
Corg
Corg
Feaq
Feaq
dissim
ilatory
Feaq
iron re
Feaq
ductio
n
Nitrate
& Pho
sphate
Fig. 2. This figure is a conceptual depiction of the wind-driven, cross-shelf circulation that occurs in eastern boundary upwelling regions. This
particular representation focuses on the onshore transport within the bottom boundary layer and the offshore transport (bottom figure) in the
surface mixed layer, driven by the along-shelf wind stress (Lentz, 1995). Note that these two layers merge over the inner shelf, leading to rapid
and efficient communication between the benthic boundary layer and the euphotic zone (also see Allen et al., 1995).
Continental shelf iron flux and its isotope composition
sediment at rates of 0.2–1.4 cm yr1, with an area-normalized rate of 0.4 cm yr1 (Sommerfield and Nittrouer,
1999; Wheatcroft and Sommerfield, 2005). This region of
sediment accumulation accounts for approximately one
third of all the sediment discharged by this system (Wheatcroft et al., 1997; Wheatcroft and Borgeld, 2000).
Although the Eel River represents the largest sediment
source to the region (Wheatcroft and Sommerfield, 2005),
much of that sediment (two thirds) appears to be transported elsewhere along the margin, presumably down the
continental slope and canyon (Wheatcroft, 2000). The shelf
near the Umpqua River also retains a significant fraction of
its sediment (Wheatcroft and Sommerfield, 2005), but this
system is smaller than the Eel. Both rivers deliver fresh,
highly reactive material to the shelf bed as the climate of
the Pacific Northwest is wet, and major flood events are
common to this system. Sediments in this shelf region are
typically bioturbated down to several tens of centimeters
depth (Wheatcroft and Sommerfield, 2005; Wheatcroft,
2006) with bioirrigation as a major pathway for solute exchange, which was confirmed by the frequent encounter
of macrofauna during processing of sediment cores. The final two aspects of our study site choices have to do with (1)
the timing of upwelling, and (2) the possibility for late-season hypoxia. In terms of upwelling, the Pacific Northwest
margin is characterized by the upwelling season typically
being well under way by June (Bakun, 1973). Although
much less is known about the occurrence of hypoxia in
the region, it is possible that low oxygen conditions on
the shelf may be a regular event—at least along the Oregon
margin (Service, 2004; Chan et al., 2008). In some regards
our selected sites are unusual in that they represent shelf
areas where sediment is accumulating; however, these sites
offer an opportunity for assessing the benthic iron efflux
to this region’s coastal zone. In comparison to our Pacific
Northwest continental shelf sites, we also discuss previously
published data from Monterey Bay (Berelson et al., 2003;
Elrod et al., 2004), and present new data for samples collected from Morro Bay in July 2001 (McManus et al.,
2003; Hammond et al., 2004) (Fig. 1). The terrigenous sediment delivery flux at these central California shelf sites is
likely to be smaller compared to the Pacific Northwest
BIF sites, but as with the BIF sites, both of the central California shelf sites are quite likely to be occupied by macrofauna, and biorirrigation is a dominant pore water
transport process. We also present both new and previously
published data (McManus et al., 1997; Elrod et al., 2004)
from two California Borderland Basin sites (San Pedro
and Santa Monica Basins) for comparison of their iron
fluxes and isotope compositions to those measured along
the two shelf areas (see Table 1 for station overview).
3. METHODS
The benthic incubation chambers used in this work have
been described previously (Berelson and Hammond, 1986;
Berelson et al., 2003). Briefly, each “benthic lander” contains three PVC chambers with an approximate volume of
7 ± 1 l and is stirred by a rotating paddle (7 RPM). Each
chamber draws six samples (200 ml) during incubation at
3987
preprogrammed time intervals. To assess chamber volume
we introduce a spike of CsCl into each chamber. Upon
recovery, samples are stored in various types
P of collection
tubes before processing for analysis of pH, CO2, oxygen,
and trace metals. Trace metal samples are filtered through
0.45 lm filters with Polyethersulfone membrane (WhatmanÒ GD/X Syringe filters), which is operationally defined
here as the total dissolved fraction, and trace metal samples
are acidified to pH 1.7 with purified HCl. Samples for
oxygen analysis (8 ml) were collected and stored in valved,
glass ampoules. Dissolved oxygen was analyzed on board
ship within an hour of lander recovery using an optode
(PreSens GmbH, Regensburg, Germany), which typically
has an uncertainty of ±1 lM.
Sediment cores were retrieved using a multi-corer (Barnett et al., 1984) and immediately placed into a cold room
for further processing at in situ temperature (4 °C). Sediment pore fluids were obtained by centrifugation. For this
technique sediments were placed in centrifuge tubes under
a nitrogen atmosphere. After centrifugation, porewater
samples were filtered using 0.45 lm filters with Polyethersulfone membrane. Wet sediments for two cores recovered
during BIF-I (Eel River 70 m site and Umpqua River
105 m sites) were analyzed for their operationally defined
highly reactive iron contents using 1 M hydroxylamine–
HCl solution in 25% v/v acetic acid (Chester and Hughes,
1967; Poulton and Canfield, 2005). Extractions were performed at room temperature over 48 h, and this procedure
has been shown to extract easily reducible iron oxides (ferrihydrite and lepidocrocite).
In this communication we report the dissolved oxygen,
total dissolved iron and organic carbon respiration rates
from the benthic incubation chambers. In addition, the iron
isotope composition of a number of benthic incubation
samples is presented. Total dissolved iron concentrations
from the BIF-I and BIF-II lander samples were measured
by isotope dilution (ID) using enriched 57Fe spike. Spiked
samples were purified using Nitriloacetic Acid (NTA)
Superflow resin, following the protocol described in Lohan
et al. (2005). For this procedure an aliquot of H2O2 was
added to the acidified sample (pH 1.7) before passing it
through a small in-line column containing 0.2 ml of the
NTA resin. After removing the salt-matrix with ultra-pure
water, samples were eluted with 2 M HNO3. Analyses of
purified ID samples were conducted on an Axiom high-resolution inductively coupled plasma mass spectrometer
(HR-ICP-MS) at Oregon State University, average reproducibility was ±1% based on replicate analyses. In addition
to the ID method, samples from BIF-II, Morro Bay and the
Borderland basins were also analyzed by developing a standard curve using Sc and Ge as internal monitor. Analysis of
10-fold diluted samples (no pre-concentration) was performed on an Agilent 7500ce quadrupole ICP-MS equipped
with a collision cell at the University of California Riverside, average reproducibility for this method was ±7%.
Comparison of ID versus standard calibration on select
samples from BIF-II show good agreement between the
two methods (average reproducibility ±11%) Obvious outliers that do not reproduce within our expected range, most
likely due to contamination during or after sampling, were
3988
Table 1
Station overview and flux data summary.
Dissolved
Fe flux
(lmol m 2 d1)f
Carbon oxidation
rate
(mmol m 2 d1)f
45.0
n/d
37.6
52.3
n/d
n/d
n/d
34.8
n/d
5.1 ± 0.4
8.0 ± 2.0
6.5 ± 0.7
5.2 ± 0.2
5.9 ± 0.7
6.1 ± 0.6
6.5 ± 1.0
3.3 ± 1.3
4.7 ± 0.4
No lander
deployment
13 ± 4
12 ± 4
29
332 ± 30
17 ± 10
97 ± 41
55 ± 8
52 ± 19
21 ± 8
10 ± 4
17 ± 4
11 ± 5
12 ± 7
10 ± 9
23 ± 12
18 ± 10
17 ± 8
7±6
112
64
133
153
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
26 ± 43
85
6±2
11 ± 4
13 ± 3
7±2
7±1
15 ± 2
3
4
n/d
n/d
n/d
n/d
6.0 ± 0.8
12.1 ± 1.5
568 ± 215
412 ± 81
No lander
deployment
3±1
2±1
Longs
Date
Northern Shelf. BIFI+II
Eel River shelf 70 ma
40° 48.0
124° 7.7
Apr-07
Eel River shelf 90 ma
Eel River shelf 90 ma
Eel River shelf 120 ma
Eel River shelf 125 ma
Eel River shelf 125 ma
Umpqua River shelf 105 ma
Umpqua River shelf 105 ma
Umpqua River shelf 190 ma
Umpqua River shelf 190 m a
40°
40°
40°
41°
41°
43°
43°
43°
43°
57.0
57.0
43.7
0.0
0.0
56.0
56.0
55.2
55.2
124°
124°
124°
124°
124°
124°
124°
124°
124°
18.1
18.1
28.6
20.3
20.3
19.2
19.2
41.4
41.4
Apr-07
Sep-07
Apr-07
May-07
Sep-07
Apr-07
Sep-07
Apr-07
Sep-07
2
3
1
2
2
2
2
2
3
93
142
83
94
125
60
78
64
60
Central California shelf
Morro Bay 100 m b
Morro Bay 200 mb
Monterey Bay 100 mc
Monterey Bay 100 m c
35°
35°
36°
36°
17.9
17.9
44.8
44.8
120°
121°
121°
121°
58.4
1.6
55.6
55.6
Jul-01
Jul-01
Mar-94
Nov-95
2
1
6
3
California Borderland Basins
San Rsdro Basin 900 m d
Santa Monia Basin 900 me
Santa Monia Basin 900 me
33° 35.7
33° 44.3
33° 44.3
Jul-01
Jul-01
Jul-03
3
3
a
b
c
d
e
f
118° 32.0
118° 47.1
118° 47.1
Number
of
chambers
Dissolved O2
flux
(mmol m 2 d1)f
Lats
Bottom
water
diss. O2
(lM)
Bottom
water
diss. Fe
(nM)
81
This study.
Hammond et al. (2004) (lander data except Fe); McManus et al. (2003) (porew ater Fe data); this study (lander Fe data).
Brod et al. (2004) (lander and porew ater data).
This study (lander data); McManus et al. (1997) (porewater Fe data, cores collected in March 1994).
Brod et al. (2004) (lander data); this study (porewater Fe data, cores collected in July 2003).
Errors for all calculated fluxes are 1-SD based on multiple chambers.
S. Severmann et al. / Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
Site name and
sampling depth
(approx.)
Continental shelf iron flux and its isotope composition
excluded from the data presented here. The means of the
initial sample draws for the two BIF cruises are 28 ± 16
and 38 ± 13 nM for BIF-I and II, respectively. These values
typically represent the sample draw approximately 1 h into
the experiment and are assumed to represent a reasonable
estimate of the iron concentration within the benthic
boundary layer. In addition, bottom water samples were
collected 10 m above the seafloor with Niskin bottles
(BIF-I cruise only), and concentrations were determined
by a flow injection technique following Lohan et al.
(2006) (analysis courtesy of Chris Holm, OSU). This method, which is modified after Measures et al. (1995), uses
NTA resin for on-line sample pre-concentration and catalytic spectrophotometric detection with N,N-dimethyl-pphenylenediamine dihydrochloride (FI-NTA-DPD). Analysis of NASS-5 (seawater certified reference material from
the National Research Council of Canada) by the same
technique gave a measured value of 3.4 ± 0.4 nM compared
to the certified reference value of 3.7 ± 0.6 nM. Mean bottom water concentration during our BIF-I cruise was
58 ± 8 nM, i.e. slightly higher than the initial lander draws.
Iron isotope measurements were made on a Thermo
Finnigan High Resolution Multi-Collector ICP-MS (Neptune) at the University of California, Santa Cruz, and at
the Woods Hole Oceanographic Institution Plasma Laboratory. Porewater and lander samples were purified using
two different protocols. Porewater samples were purified
following a standard anion exchange protocol (e.g., Beard
et al., 2003), where an appropriate volume of porewater
(previously acidified to pH <2 for preservation) was mixed
with trace metal grade concentrated HCl to give a final
molarity of 6 M HCl for the sample-acid mix. Following
the initial processing through a large (0.6 ml) ion exchange
column, purified samples were dried down, re-dissolved in
0.2 ml of 6 M HCl and processed through a second, smaller
column with a 0.2 ml resin volume. Lander samples were
purified and pre-concentrated using NTA resin and following the same procedure described above for iron concentration measurements. Following this initial purification step,
samples were processed through a small ion exchange column (same as above) to remove any naturally occurring
copper in the samples. The limited quantity of isotope data
for the benthic chamber experiments is in part governed by
the quantity of iron necessary for isotope analyses, the low
iron concentrations at the start of the experiments, and the
small sample volumes available from each sample draw.
To monitor matrix interference effects during isotope
analysis we prepared an in-house seawater sample by spiking 1–10 ml of trace metal free seawater with a small
amount of iron with known isotope composition. Trace metal free seawater was obtained by passing open ocean surface water through Chelex-100 resin. The ratio of iron to
seawater matrix in our synthetic sample matched the lowest
iron to seawater ratio in our porewater and lander samples.
The measured isotope composition of our synthetic seawater samples was analytically indistinguishable from the pure
iron standard reference material (Table S1, Electronic
Annex).
Following the procedure of Arnold et al. (2004), purified
samples were introduced into the mass spectrometer as 0.2–
3989
2 ppm iron solutions, mixed with equal amounts of copper
standard of known isotope composition (NIST-976 copper
isotope standard), which was measured simultaneously for
mass bias correction. Sample isotope ratios were normalized
to the average of two bracketing standards. Isotope ratios of
56
Fe/54Fe and 57Fe/54Fe are reported using standard delta
notation. Measured ratios are normalized relative to igneous
rocks, which have an average isotope composition of
d56Fe = 0 ± 0.10& (2 SD; Beard et al., 2003). On this scale
the isotope composition of the international iron isotope reference material IRMM-014 is 0.09& for d56Fe. The average analytical precision and the external precision for d56Fe
(2 SD) are very similar at 0.13& and 0.10&, respectively.
Several standard reference materials of known isotope composition, including IRMM-014 (an international iron isotope
standard), SDO-1 (a black shale standard reference material)
BIR-1 (a basalt standard reference material) and an in-house
porewater reference standard (made from natural porewater
samples, iron concentration 55 lM) were measured routinely for each sample batch.
Organic carbon respiration rates were estimated from
the efflux of respiratory CO2. Measurements were made
on 1–3 ml of chamber water using a UIC coulometer and
standards provided by Scripps Institute of Oceanography
(CO2CRMS). In general, this estimate is not as precise as
our other flux estimates.
Some of this lack in precision deP
rives from the
CO2 measurement, which had a large
uncertainty during our BIF-I and BIF-II cruises
in part beP
cause we are measuring a small change in CO2 against a
large background signal. However, the estimated respiratory CO2 flux is also limited by our lack of an accurate estimate for carbonate dissolution (or precipitation) rates in
these settings. Thus our estimates of carbon oxidation rates
are generally presented with a large uncertainty, and we
look to the dissolved oxygen uptake fluxes to further develop the electron transport background for our discussion.
For all experiments presented, analytical precision is not
likely the limiting factor governing our confidence in the
benthic flux estimates. Contamination of samples, either
with ambient bottom water or some other unknown contaminant introduced during handling, and our estimation
of chamber volume and leakage all influence our confidence
in any particular value or flux estimate. Benthic lander
experiments are inherently fraught with a host of potential
sampling artifacts that have been extensively discussed in
the literature (see e.g., Rutgers van der Loeff et al., 1984;
Sundby et al., 1986; Tengberg et al., 2005). For example,
oxidation effects during chamber deployment and recovery
are unavoidable and may cause an underestimation of the
actual flux. However, we note that these oxidation effects
essentially mimic processes that would otherwise occur in
the water column as the benthic efflux is exposed to oxygenated bottom water. In addition, the heterogeneity of flux
values at any given site produce an average flux with an
uncertainty that tends to be as large or larger than the
uncertainty deriving from any analytical uncertainty (Hammond et al., 1996). Despite these limitations, benthic chamber experiments remain one of our most dependable tools
to shed light on the solute exchange between the sediments
and the water column (Tengberg et al., 2005).
3990
S. Severmann et al. / Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
4. RESULTS
4.1. Incubation chamber results
Iron concentrations in the benthic incubation chambers
from all sites generally show an increase with time (Fig. 3)
with a range in iron flux from essentially zero iron flux (one
chamber from Morro Bay) to a high flux exceeding
500 lmol m2 d1 in the low oxygen environments of
the Santa Monica and San Pedro Basins (Table 1). In total,
Lander diss. Fe (nM)
BIF−I − Eel River Shelf
these observations imply a net transport of iron from the
sediments to the overlying water column along the continental margin (Elrod et al., 2004), with a particularly high
benthic iron flux from the shelf region around the Eel River
and the Borderland Basins. In addition, both nitrate (not
shown) and dissolved oxygen within the incubation chambers exhibit a decrease in concentration with time (Fig. 4)
and RCO2 increases, which is consistent with the high rates
of organic carbon oxidation in all of these settings. Evidence for oxygen consumption along with nitrate reduction
BIF−II − Eel River Shelf
8000
8000
6000
6000
4000
4000
2000
2000
0
0
Eel R. Shelf, 90m
Eel R. Shelf, 120m
Eel R. Shelf, 125m
0
10
20
30
40
0
Lander diss. Fe (nM)
BIF−I − Umpqua River Shelf
30
40
BIF−II − Umpqua River Shelf
8000
600
1500
400
1000
6000
20
800
2000
8000
10
6000
500
200
0
0
0
10
20
30
0
4000
4000
2000
2000
0
0
10
20
30
Umpqua R. Shelf, 105m
Umpqua R. Shelf, 190m
0
10
20
30
40
0
Central CA Shelf
20
30
40
CA Borderland Basins
600
Lander diss. Fe (nM)
10
8000
8000
400
200
6000
Morro Bay, 100m
6000
Morro Bay, 200m
0
0
10
20
Monterey Bay, 100m
30
4000
4000
San Pedro B., 900m
Santa Monica B., 900m
(July 2001)
Santa Monica B., 900m
(March 1995)
2000
2000
0
0
0
10
20
Time (hrs)
30
40
0
10
20
30
40
Time (hrs)
Fig. 3. Dissolved iron concentrations in lander chamber during seafloor incubation experiments are distinctly non-linear at several of our
study sites (inserts show the data on an extended scale). Iron concentrations are corrected for dilution over the course of the experiment by
accounting for the inflow of ambient bottom water during a sample draw. Monterey Bay data and Santa Monica Basin data for March 1995
are from Elrod et al. (2004), all other data are from this study.
Continental shelf iron flux and its isotope composition
Northern Shelf, BIF−I
3991
Northern Shelf, BIF−II
Lander diss. oxygen (μM)
160
150
120
100
80
Eel R. Shelf, 90m
Eel R. Shelf, 120m
50
Eel R. Shelf, 125m
40
Umpqua R. Shelf, 105m
Umpqua R. Shelf, 190m
0
0
0
10
20
30
40
0
Time (hrs)
10
20
30
40
Time (hrs)
Fig. 4. Dissolved oxygen concentrations in lander chamber decrease as oxygen is consumed during carbon remineralization in the sediments
and bottom water.
is prevalent not only throughout this study but it is a general feature of this continental margin (Reimers et al., 1992;
Berelson and Stott, 2003), as is the observation of a significant iron benthic flux (Elrod et al., 2004).
When examined as a whole, the BIF-I benthic incubations suggest a non-linear increase in total dissolved iron
as the experiments progress (Fig. 3). We will discuss this
feature later; however, it is important to note here that to
estimate the initial benthic flux—i.e., the change in concentration over time during the first 2–5 h of the experiment—
we chose those data that best approximate a linear relationship between concentration and time as representative of
the benthic flux. Samples that were excluded from the fit
for flux calculations are in parentheses in the data appendix
(Table S1). This approach effectively makes our estimates
of the benthic flux minimum estimates. In contrast to the
BIF-I experiments, during BIF-II the two Umpqua River
sites suggest an approximately linear change in iron concentration with time. However, both Eel River sites show a
curved profile with an increasing benthic iron flux with
time, much like that observed during BIF-I (Fig. 3).
The northern continental shelf environments display
quite high benthic fluxes compared to the central California
shelf and previously reported iron fluxes in this region (Berelson et al., 2003; Elrod et al., 2004), but they are similar to
fluxes reported by Lohan and Bruland (2008) for hypoxic
shelf waters near large rivers. The highest benthic iron
fluxes, however, were measured in the low oxygen settings
of the San Pedro and Santa Monica basins (Fig. 3 and Table 1). Our estimates for these two basins are much higher
than estimates from Elrod et al. (2004) for the same locations. We note, however, that the bottom water oxygen
concentrations (3–4 lM) during sample collection for this
study (Hammond et al., 2004) were lower than the values
reported in Elrod et al. (2004), who measured oxygen concentrations of 9–10 lM during their sampling campaign. It
is plausible that this difference in dissolved oxygen could
drive large differences in the iron flux if the presence of
higher oxygen concentrations in this basin effectively limited the flux of iron. The contrast between our data and that
of Elrod et al. (2004) suggests that our high benthic flux
does not represent a steady state feature of these basins
and that fluxes in these settings can be highly variable.
The iron concentrations at the two central California shelf
sites off Morro Bay are in the same range as the lowest estimates from our northern shelf data and previously published data from Monterey Bay (Fig. 3). One chamber
from Morro Bay 100 m exhibits an iron flux that is essentially zero, and the other chamber from this station exhibits
a non-linear concentration increase with time. The Morro
Bay 200 m site has a higher benthic iron flux; however,
we do not have an initial point for this chamber and this
flux could therefore be non-linear as well. In sum, these values are consistent with the Oregon and northern California
margin.
In addition to the iron concentration data, we also present iron isotope compositions of the dissolved benthic iron
flux. Both the BIF-I chambers we measured from the Eel
River 125 m site suggest that as the total dissolved iron concentration increases, the d56Fe increases to progressively
heavier isotope compositions relative to the initial point
(Fig. 5). The first measured d56Fe value for the red chamber
from this site (Table S1) has an iron isotope composition of
5.0 ± 0.1& (2 SD), and linear extrapolation of this value
to the initial stage of the experiment suggests that the initial
iron flux might have had an isotope composition as low as
6&. These values are among the lightest reported for
natural samples, yet they are comparable to those measured
for sediment porewaters from the mixing zone of a subterranean estuary, where precipitation of Fe-oxyhydroxides
causes the residual dissolved Fe(II) to shift to values as
low as 5& (Rouxel et al., 2008).
For the remaining deployments we measured only the
experiment end points for their isotope composition (Table
S1); therefore, it is not possible to make inferences whether
similar trends are apparent for the other sites. However, all
experiment end points converge on a mean iron isotope
composition of 2.7 ± 1.1& (2 SD).
Benthic chamber isotope compositions from experiments conducted in the San Pedro and Santa Monica Basins have a slightly more negative isotope composition
compared to the experiment end points from the northern
3992
S. Severmann et al. / Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
Northern Shelf, BIF−I
CA Borderland Basins
Eel R. Shelf, 90m
Eel R. Shelf, 120m
δ56Fe (o/oo)
Eel R. Shelf, 125m
Umpqua R. Shelf, 105m
Umpqua R. Shelf, 190m
San Pedro B. 900m
Santa Monica B. 900m
Time (hrs)
Time (hrs)
Fig. 5. Chamber dissolved iron isotope compositions for the northern shelf and southern basin sites show distinctly different trends, which
may be related to the differences in porewater transport mechanism (i.e., bioirrigation or advection on the shelf versus diffusion in the basins).
Error bars are 2-STD; isotope ratios are reported relative to average igneous rocks.
shelf sites, and the d56Fe in the chambers remains relatively
constant over the course of the experiment, despite substantial increases in iron concentration by several lM (compare
Figs. 3 and 5). The average d56Fe value for all the basin
sites is 3.6 ± 0.7& (2 SD).
4.2. Porewater results
Of the continental shelf sites, the highest porewater iron
concentrations are found at the stations closest to the Eel
River with concentration maxima between 100 and
250 lM (Fig. 6). Porewater concentrations in shelf sediments near the Umpqua River are distinctly lower and generally between 50 and 100 lM. Late summer porewater iron
concentrations for cores recovered during the BIF-II cruise
were similar or, as in the case of the Eel 125 m site, even
slightly higher compared to BIF-I porewater concentrations
measured in the spring.
At the central California margin sites, Morro Bay sediments have the lowest porewater iron concentrations with
values and profile shape very similar to the Umpqua River
sediments (Fig. 6c). Prior pore water iron data from the
Santa Monica and San Pedro Basins (McManus et al.,
1997) as well as new data from the Santa Monica Basin suggest maximum values in the 100–250 lM range (Fig. 6).
The iron isotope composition of the BIF region porewaters is light relative to average continental crust and weathering products, with d56Fe values generally ranging between
0.7& and 4.0&, and the lightest values for each individual core occurring near the sediment water boundary
(Fig. 7 and Homoky et al., 2009). The general feature of
dissolved iron in porewaters from bioturbated sediments
having light iron isotope compositions has been noted before for this continental margin (Severmann et al., 2006)
as well as in other localities (Bergquist and Boyle, 2006).
The generally lighter isotope values near the sediment surface suggest that the initial benthic flux will also be lighter;
we note, however, that the sediment–water solute exchange
in this region is likely driven by bioirrigation in addition to
diffusion across the sediment–water interface, thus under
these circumstances the source of the iron-rich fluid could
be well below the sediment–water interface (Aller, 2001).
The porewater iron isotope profile from the Santa Monica
Basin shows a trend similar to that observed at northern
shelf sites, with light values of 3& near the sediment
surface and values between 1& and 2& at depth
(Fig. 7).
5. DISCUSSION
5.1. The continental shelf iron flux
Our estimate of the benthic iron flux off the Eel River for
the April/May 2007 period (330 lmol m2 d1) is the
highest of the Pacific Northwest shelf sites, and is also the
highest flux for any of the shelf sites we have measured
(i.e., including Monterey Bay and Morro Bay). Furthermore, for the two stations near the Umpqua River and
the 125 m station on the Eel River shelf we measured higher
benthic iron fluxes during the April/May sampling period
compared to those measured in September (Table 1). In
the case of the 125 m Eel River station the flux decreased
by more than one order of magnitude between May and
September. The shallower (90 m) Eel River station exhibits
little difference in the flux between the two sampling periods. Also, we do not have data from our third Eel River
station (120 m) for both cruises.
One possible explanation for the apparent spatial and
temporal difference in iron flux is that it may be related
to the higher availability of reactive iron near the Eel River
shelf compared to the Umpqua River shelf (i.e., spatial difference), and during spring as compared to September (i.e.,
temporal difference, see also Elrod et al., 2008). In terms of
the spatial difference this rationale follows from the idea
that the Eel River likely has higher sediment discharge as
compared to the Umpqua. In terms of the temporal differences, the principle is that the winter and spring river discharge can deliver a significant reactive iron input to the
shelf bed and that the remobilization of this iron is greatest
during the high production period that follows the high
Continental shelf iron flux and its isotope composition
Northern Shelf, BIF−I
Northern Shelf, BIF−II
Porewater Fe (μM)
Depth (cm)
0
100
200
3993
Porewater Fe (μM)
300
0
0
0
5
5
10
10
15
15
100
200
300
Eel R. Shelf, 70m
Eel R. Shelf, 90m
Eel R. Shelf, 120m
Eel R. Shelf, 125m
20
Umpqua R. Shelf, 105m
20
Umpqua R. Shelf, 190m
Central CA Shelf
CA Borderland Basins
Porewater Fe (μM)
Porewater Fe (μM)
Depth (cm)
0
100
200
300
0
0
0
5
10
10
20
100
200
300
Morro Bay, 100m
Morro Bay, 200m
Monterey Bay, 100m
15
30
20
40
San Pedro B., 900m
Santa Monica B., 900m
(July 2003)
Santa Monica B., 900m
(March 1995)
Fig. 6. Low porewater iron concentrations at the central California shelf station coincide with low benthic iron flux, whereas seasonal
differences, e.g. at the 125 m Eel River station, are not reflected in the porewater concentrations. Morro Bay data are from McManus et al.
(2003), Monterey Bay data are from Elrod et al. (2004), Santa Monica Basin data for March 1995 are from McManus et al. (1997), all other
data are from this study.
river discharge period (Fig. 2, and Colbert, 2004; Wetz
et al., 2006; Elrod et al., 2008). Central to these ideas is
the point made by Wetz et al. (2006) that the coastline-normalized riverine discharge decreases by >65% from Oregon
to Northern and Central California. In short, there is ample
evidence to suggest that there is a significant input of riverine particulate material in the winter and spring in these regions. Consistent with the hypothesis that reactive iron
availability affects the magnitude of the benthic iron efflux,
we have measured significantly higher concentrations of
highly reactive iron in the sediments below 5 cm from the
Eel River shelf compared to sediments from the Umpqua
Rivers shelf (Fig. 8). We only measured highly reactive iron
contents during the spring and cannot assess directly
whether a depletion of reactive iron has occurred over the
summer season. However, porewater iron concentrations
during late summer are similar to or even slightly higher
than spring porewater concentrations, suggesting that
highly reactive iron is not depleted. In fact, mass balance
calculations based on an average highly reactive iron contents of 0.5 wt.% (Fig. 8), and a mass accumulation rate
that assumes a constant sedimentation rate of 0.4 cm yr1,
an average downcore porosity of 0.7 and a sediment density
of 2.6 g cm3, suggests that the sedimentary reactive iron
delivered to the self can support an annual average benthic
iron flux of P600 lmol Fe m2 d1, i.e. an order of magnitude higher than most of our measured spring and summer
iron fluxes (Table 1). The apparent excess of sedimentary
3994
S. Severmann et al. / Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
BIF−I
porewater
δ56
Borderland Basins
o
Fe ( /oo)
porewater δ56Fe (o/oo)
Depth (cm)
Eel R. Shelf, 70m
Eel R. Shelf, 90m
Eel R. Shelf, 120m
Eel R. Shelf, 125m
Umpqua R. Shelf, 105m
Umpqua R. Shelf, 190m
Santa Monica B. 900m
Fig. 7. Porewater iron isotope compositions show a similar trend at the northern shelf sites and the southern basin sites despite different solute
transport regimes (advective on the shelf versus diffusive in the basins). Error bars are 2-STD; isotope ratios are reported relative to average
igneous rocks.
Highly reactive
sedimentary Fe (wt.%)
0.0
0
0.3
0.6
0.9
Depth (cm)
5
10
15
20
Eel R. Shelf, 70m
Umpqua R. Shelf, 105m
Fig. 8. Sediment on the Eel River shelf (70 m) have higher contents
of reactive iron (determined by standard hydroxylamine–HCl
extraction, Chester and Hughes, 1967) compared to the Umpqua
River shelf site (105 m). Although concentrations are very similar
in the upper 5 cm, we note that bioirrigation likely extends several
tens of centimeters.
reactive iron argues against its diminished availability over
the summer exerting a major control on the seasonal variation in benthic iron flux.
With respect to geographic variations in sedimentary
reactive contents, we observe that pore fluid iron concentrations are generally higher for the Eel River shelf compared
to the Umpqua River shelf, consistent with the higher ben-
thic iron fluxes we measured near the Eel as well as the generally higher dissolved iron concentrations seen in the
incubation chambers (Table 1 and Figs. 3 and 9). In contrast to these sites, which both have relatively high porewater iron concentrations, the porewater iron
concentrations from Monterey Bay are quite low by comparison (Fig. 6 and Elrod et al., 2004), as is the benthic iron
flux. Thus, to first order we observe high porewater iron
concentrations (implying higher reactive iron contents) at
shelf sites where iron fluxes are larger than the general
oceanographic iron flux versus carbon oxidation relationship (Fig. 9a and Elrod et al., 2004). However, this broad
generalization is unlikely to be true in detail, as other factors will certainly influence the transport of iron across
the sediment–seawater boundary. For example, one of the
influential factors that could affect exchange is the sediment
irrigation rate because this process rather than diffusion,
likely dominates transport in shelf environments (Aller,
1980a, 2001; Berelson et al., 2003). Another important
example, as will be examined below, is the bottom water
oxygen concentration. In short, we would not expect a simple relationship between the porewater concentrations and
the large seasonal changes we observe in the benthic flux
simply because variations in oxygen availability would tend
to influence the iron flux as well (Sundby et al., 1986; Pakhomova et al., 2007).
Although the reported fluxes for our study are higher
than previously reported for the Oregon–California margin,
and are indeed comparatively high for open-ocean sites
(e.g., Elrod et al., 2004) they are not anomalous for restricted basins with limited bottom water renewal (Sundby
et al., 1986; Skoog et al., 1996; Friedl et al., 1998; Warnken
et al., 2001; Friedrich et al., 2002; Pakhomova et al., 2007).
Results from our study, specifically the high fluxes measured for the Eel River shelf, emphasizes the point made
Continental shelf iron flux and its isotope composition
Central CA shelf
CA Borderland Basins
Northern shelf, BIF−I
Northern shelf, BIF−II
103
b
Fe flux (µmol m−2 d−1)
Fe flux (µmol m−2 d−1)
a
102
101
100
3995
103
102
101
100
0
5
10
15
20
25
Cox (mmol m−2 d−1)
0
50
100
150
200
Bottom water oxygen (µmol)
Fig. 9. (a) For sites with high sediment reactive iron loading, benthic iron fluxes exceed the general oceanographic relationship for iron flux
versus carbon oxidation (indicated by solid line with dashed lines showing error envelope), which was proposed by Elrod et al. (2004) based on
data from the central California shelf. (b) The apparent correlation between iron flux and bottom water oxygen points to the role of redox
recycling and the activity of benthic biota in regulating sediment–ocean solute exchange.
in Elrod et al. (2004) that their estimate for the shelf contribution to the global ocean iron budget (2.2 109 mol y1)
needs to be considered a minimum. Placing a more specific
value on that flux is difficult; however, it is worth noting
that in some locations at least a portion of this flux can
be detected hundreds to thousands of kilometers from the
ocean–continent boundary (Elrod et al., 2004; Lam et al.,
2006; Nishioka et al., 2007; Lam and Bishop, 2008; Raiswell et al., 2008), making benthic iron flux an important
delivery vector to the global ocean.
5.2. Relationship of chamber iron to dissolved oxygen
One of the more striking relationships that arise from
our data is the non-linear iron concentration versus time
relationship as the incubation experiments evolve. This
relationship appears to be related to the depletion of oxygen during the experiments (Fig. 4). Essentially it appears
that as oxygen decreases below a particular range, which
seems to be centered around 60–80 lM (Fig. 10), the sediment package becomes sufficiently reducing that the flux of
iron can increase. This observation implies a coeval relationship between the availability or remobilization of reactive iron in any one location and the bottom water oxygen
concentration. The expression of this relationship is at least
implied for the 125 m Eel River site where the spring flux is
much greater than all of the other sites, yet the September
flux is puzzlingly low. Here the bottom water oxygen contents for September are greater (125 lM) than for the
spring period (94 lM). At the Umpqua River sites both
time periods have oxygen concentrations of 60–80 lM (Table 1) but the spring Fe fluxes are roughly twice the fall flux.
The importance of oxygen on the benthic iron flux certainly
has been discussed prior to this work (Balzer, 1982; Sundby
et al., 1986; Pakhomova et al., 2007; and references therein), and may also partially explain the low fluxes at the
Monterey Bay site, as this site was reported to have much
higher bottom water oxygen concentrations that ranged between 100 and 185 lM (Elrod et al., 2004). Thus although
we believe that the presence of reactive iron and, more specifically, high pore water iron concentrations, is an important component to the high benthic iron flux, the
influence of dissolved oxygen certainly cannot be ignored
(Sundby et al., 1986; Pakhomova et al., 2007).
This point of the importance of dissolved oxygen is also
apparent in the Morro Bay sites. Here the lower flux occurs
at a Shallower site (100 m) having a bottom water oxygen
concentration of 112 lM (Hammond et al., 2004), whereas
the high iron flux is at a deeper shelf site (200 m), with a
lower organic carbon oxidation rate, but a bottom water
oxygen concentration that is roughly half that of the shallower site (64 lM, Hammond et al., 2004). Another important difference between the two Morro Bay stations is the
higher porewater iron contents at the 200 m site compared
to the shallower 100 m site (Fig. 6).
The San Pedro and Santa Monica Basin sites provide
additional data regarding the importance of oxygen on iron
benthic fluxes. These sites differ significantly in a number of
respects from the shelf sites. An important distinction between the Borderland Basin sites and the other shelf sites
is that in these low oxygen basins the transport across the
sediment–seawater boundary is likely dominated by diffusive rather than advective transport (i.e., bioirrigation).
This difference means that critical oxygen levels could well
3996
S. Severmann et al. / Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
Northern Shelf, BIF−I and BIF−II
Lander diss. Fe (nM)
6000
Eel R. Shelf, 90m
4000
Eel R. Shelf, 120m
Eel R. Shelf, 125m
Umpqua R. Shelf, 105m
Umpqua R. Shelf, 190m
2000
0
100
80
60
40
20
0
Lander diss. oxygen (µM)
Fig. 10. As bottom water oxygen concentrations decrease below a threshold value of 60–80 (indicated by the shaded area), benthic iron
fluxes at sites with high sedimentary reactive iron loading increase from relatively small (linear) flux to large (exponential) flux. Black symbols
are for BIF-I, gray symbols are for BIF-II data.
play a more prominent role in restricting or enhancing iron
transport across that boundary. During the period when
these basins were sampled for this work, oxygen contents
were quite low (3–4 lM, Hammond et al., 2004). These basins are intermittently flushed (Berelson, 1991) and exhibit
sedimentary characteristics of temporal oxygen variability
(Hagadorn et al., 1995; Stott et al., 2000), thus oxygen
can be consumed over time following a flushing period or
event, which would allow the oxygen penetration depth to
shoal and reducing conditions expand over time. This
would have the net effect of remobilizing oxidized iron close
to the sediment–water boundary, and likely lead to high
variability in the benthic flux measured in this basin. Prior
data from this region also showed higher fluxes compared
to the other continental margin sites (Fig. 9a, and Elrod
et al., 2004); however, the iron fluxes from that study were
lower than those reported here while the reported bottom
water oxygen content was higher (9–10 lM, Elrod et al.,
2004).
Overall the benthic iron flux values presented in this study
show a reasonably good correlation with bottom water oxygen (Fig. 9b). The improved correlation compared to the iron
flux versus carbon oxidation relationship (Fig. 9a) suggests
that oxygen is a more dominant variable with respect to benthic iron flux, at least in coastal sediments that have an excess
of reactive iron. It is unlikely, however, that a single parameter can be considered as a master variable that operates
across a broad range of marine settings.
5.3. Iron isotope composition of the benthic flux
Porewaters from bioturbated continental margin sediments near Monterey Bay have previously shown iron isotope compositions to be light relative to average weathering
flux (Severmann et al., 2006). The porewater values gener-
ally range between 1& and 3&, with the lightest compositions near the sediment–water boundary. These
porewater iron isotope compositions imply that if there is
a substantial benthic iron flux along continental margins,
and that iron originates from near the sediment water interface, then the water column iron would be isotopically light.
Our new porewater data reaffirm that indeed the near-surface porewater isotope compositions are light and our benthic incubation chamber data clearly suggest that the
benthic flux is also light (Figs. 5 and 7).
The positive correlation between d56Fe and iron concentrations in the BIF porewaters (Fig. 11) suggests that the
isotope compositions are driven not just by the effect of iron
reduction, but also, at least in part, by the fractional removal of iron into isotopically heavier oxyhydroxide phases
(Bullen et al., 2001; Rouxel et al., 2008). Teutsch et al.
(2005) offer an alternative explanation and suggest that
sorption of Fe(II) onto iron oxide/hydroxide particles is
the primary control on the dissolved iron isotope compositions. Irrespective of the removal mechanism, the pertinent
observation with respect to our BIF porewater and chamber isotope compositions is that partial removal of a Fe(II)
by oxidation or sorption onto freshly precipitated oxyhydroxides would drive the residual dissolved iron pool to
lower d56Fe values.
The effect of partial re-oxidation of dissolved iron on the
isotope compositions is evident in the BIF-I chamber data
from the Eel 125 m site, where d56Fe values increase steadily with decreasing oxygen concentrations and increasing
iron concentration in the chamber (compare Figs. 3–5).
One interpretation of the higher d56Fe values as the incubation experiment progresses is that an increasingly smaller
proportion of Fe(II) is re-oxidized at this site. However,
the absence of such a correlation in the chamber data from
the California Borderland Basin sites, where the isotope
Continental shelf iron flux and its isotope composition
Eel R. Shelf, 70m
Eel R. Shelf, 90m
Eel R. Shelf, 120m
Eel R. Shelf, 125m
Umpqua R. Shelf, 105m
Umpqua R. Shelf, 190m
Porewater δ56Fe (o/oo)
Santa Monica B., 900m
Porewater Fe (μmol/l)
Fig. 11. The positive correlation between porewater isotope
compositions and log of iron concentrations is consistent with
removal of Fe(II) into heavier Fe-oxyhydroxides, leaving behind a
light residual dissolved porewater iron pool.
compositions remain relatively constant despite increasing
iron concentrations (compare Figs. 3 and 5), suggests that
the mechanism controlling the isotope compositions of
the benthic iron source are more complex. Both Borderland
Basins have very low bottom water oxygen and are primarily controlled by diffusive exchange between the sediment
package and the overlying water. Our Santa Monica and
San Pedro isotope results suggest that the dissolved iron efflux is not undergoing a significant transient chemical or
environmental change, but as is the case with the BIF data,
it is clear that the benthic iron flux is indeed light, entirely
consistent with the porewater data in this report and in Severmann et al. (2006).
Another observation that all sites have in common is a
P1& offset between the initial isotope composition during
the flux experiments and the lowest values measured in the
porewaters. A possible explanation is that none of our, or
previous, porewater measurements have fully resolved the
isotope composition in the uppermost portion of the sediment column where the solute exchange is taking place.
Consistent with the d56Fe versus iron concentration relationship outlined above (Fig. 11) it suggests that in both
systems (oxygen-rich and oxygen-poor), a significant proportion of iron is re-precipitated within the benthic boundary layer and re-deposited back into the sediments.
5.4. Effects of decreasing bottom water oxygen on iron flux
and its isotope composition
There is an assumption when assessing benthic chamber
data that the flux of iron across the sediment–seawater
3997
boundary is constant, and that the proportion of iron that
is re-oxidized in the benthic boundary layer also remains
constant. The data collected at the BIF study sites forces
a reanalysis of these assumptions, and there are (at least)
two plausible explanations for how progressively decreasing
bottom water oxygen concentrations may affect the iron
flux from the sediments, and both of them are consistent
with the benthic chamber iron trends.
5.4.1. Effects within the sediments
The first explanation concerns the effect that decreasing
oxygen concentrations would have on benthic infauna
activity, the sediment oxygen penetration depth, and consequently the rate of solute exchange. The curvature observed
in some of the chamber iron concentrations may be considered as a transition from small to large fluxes as porewater
iron diffuses across the oxidized surface layer in the sediments. The oxygen penetration depth in the sediments
was calculated from the initial bottom water oxygen concentrations and the flux of oxygen into the sediments (Table
1, following the approached used in Cai and Reimers, 1995)
and was found to vary between 0.2 and 0.4 cm for our BIF
stations (Table 2). The time it would take for an iron molecule to diffuse a distance of 0.2 to 0.4 cm through surface
sediments with an average porosity of 0.9 at 10 °C (based
on the Stokes–Einstein equation) is 5–20 h, which overlaps
with the time frame during which we have the observed the
transition from small to large flux at these stations. Sites
with shorter oxygen penetration depth show this increase
in iron flux earlier, and it is intriguing that the diffusion
time scale corresponds with the timing of change in the
magnitude of the flux.
Decreasing bottom water oxygen will also affect the
depth of oxygen penetration into the sediments and the
activity of benthic organisms. The conventional wisdom is
that asphyxiation of benthic infauna will shut down the
advective transport and significantly decrease the rate at
which dissolved species are released from the sediments (Aller, 1980b, 2001; Rutgers van der Loeff et al., 1984). In the
case of iron (and other redox-sensitive species), however, a
progressive shoaling of the oxygen penetration depth could
also dramatically increase the diffusive iron flux into the
bottom water (Fig. 12). Because chamber oxygen concentrations in our northern shelf station experiments (Eel
and Umpqua River shelves) did not drop to zero, a thin
but significant oxidized layer would have remained intact
at the sediment–seawater boundary, and benthic pumping
would have persisted. Within the burrows, however, the
oxidized layer likely disappeared, effectively eliminating
any diffusive barrier to the iron-rich pore fluids that are
apparent in the ‘bulk’ porewater profiles at depths >5 cm
(Fig. 6). Under low bottom water oxygen conditions the
iron-rich water that would fill the anoxic burrows is transferred very effectively to the bottom water by infaunal
pumping (Fig. 12b).
In the porewater isotope profile, the oxidized surface
layer coincides with the lowest porewater d56Fe values
(Fig. 7a), as has previously been observed in other continental margin sediments (Severmann et al., 2006). Although
not apparent in the bulk porewater profile, infaunal
3998
S. Severmann et al. / Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
100
0
60
Bottom water O2 (µmol/l)
Fe2+
Fe2+
Fe2+
Fe2+
A
-3
B
Porewater δ56Fe (‰)
-2
-1
Fig. 12. High initial bottom water oxygen concentrations establish an oxidized surface layer with isotopically light porewater iron; the oxic
layer poses a barrier to rapid diffusion. Decreasing bottom water oxygen concentrations causes the worm holes to become anoxic, and
porewater with high iron concentrations and heavier isotope composition is rapidly transferred into the bottom water through pumping by the
benthic biota.
burrows are probably lined by a similar low d56Fe porewater boundary layer. Under the initial conditions of relatively high bottom water oxygen, all porewater iron must
overcome this oxidized boundary before it exits the sediments. As the burrows become increasingly anoxic, iron diffuses more freely into the water contained within the
burrows and the isotope composition progressively shifts
to heavier d56Fe values, approaching that of the deeper
porewater pool (Fig. 7). The deep burrows provide a pathway for rapid transfer of the deeper sourced high [Fe]-high
d56Fe porewater pool to the overlying water, consistent
with a change in benthic chamber iron isotope compositions from very light to progressively heavier values. In
the California Borderland Basin sites, in contrast, bioturbation is likely limited to the upper few millimeters and deep
bioirrigation by macofauna is absent or at least highly reduced compared to the shelf environment as evidenced by
laminations in some locations within the basin (Stott
et al., 2000). Consequently, the porewater iron flux is
sourced from shallow sediment depth throughout the experiment, and the isotope composition does not change over
time.
Under this scenario of decreasing bottom water oxygen
concentrations, this situation may lead to an initial increase
in solute exchange of redox-sensitive species, and yet may
Table 2
Oxygen penetration depth and time for transition from small to
large Fe flux for selected stations.
BIF-I
BIF-II
BIF-I
BIF-II
BIF-I
BIF-II
BIF-I
BIF-II
Station
Oxygen
penetration
depth
Time to large
Fe flux (h)
Eel River shelf 90 m
Eel River shelf 90 m
Eel River shelf 125 m
Eel River shelf 125 m
Umpqua River shelf
105 m
Umpqua River shelf
105 m
Umpqua River shelf
190 m
Umpqua River shelf
190 m
0.38
0.35
0.39
0.43
0.26
14.5
17
7
20
5
0.25
–
0.40
15
0.27
17
ultimately involve a reduction in advective transport when
bioirrigating fauna become asphyxiated. With prolonged
low bottom water oxygen concentrations, bioirrigation will
eventually cease, and porewater iron would be increasingly
titrated by sulfide as bacterial iron reduction is replaced by
sulfate reduction. Previous flux studies from a shallow bay
Continental shelf iron flux and its isotope composition
C t ¼ C 0 eðktÞ
ð1Þ
The iron concentration (Ct at time = t) was calculated
cumulatively for each time interval with a loss term defined
by the time for iron accumulation and a decay term (k), where
k ¼ 0:69=T 1=2
ð2Þ
and
T 1=2 ¼ 34:36 eð0:242½O2 Þ
ð3Þ
as defined by Lohan and Bruland (2008). A further correction in chamber data takes into account the time it took for
sample recovery and processing, including a 5-h time delay
between the last sample draw and shipboard filtration. The
oxygen concentrations applied to each time step were based
on the measured oxygen profile for the chamber in
question.
We applied this model to the results from the Eel River
site at 125 m (Fig. 13). Our observations from two seafloor
incubation experiments, are consistent with a model that as-
Fe conc. − modeled
Fe T1/2 − modeled
O2 conc. − observed
12
BIF−I
8
4000
4
2000
120
100
80
60
0
10
20
40
0
30
12
6000
140
BIF−II
8
4000
4
2000
120
100
80
60
0
0
10
20
0
30
Chamber O2 (μM)
0
Chamber O2 (μM)
140
T1/2 Fe (hrs)
Chamber Fe (nM)
6000
T1/2 Fe (hrs)
5.4.2. Effects within the bottom water
A second explanation for the coupling between iron flux
and bottom water oxygen relates to the affect of iron oxidation kinetics; it essentially concerns the fate of the dissolved
iron flux after it leaves the sediments and mixes into the
bottom or chamber water. The half-life (T1/2) of dissolved
iron within a chamber describes the rate at which Fe(II)
is lost through oxidation; it can be estimated as a function
of oxygen concentration, temperature and pH, and should
be on the order of a few hours (Lohan and Bruland,
2008) under typical sea water, upper ocean conditions. During the seafloor incubation experiments, Fe(III) may be precipitated on the chamber walls, the sediment surface, or
within the sample vials prior to filtration. As the experiments progress, the decrease in chamber oxygen concentration with time causes the iron T1/2 in the benthic chamber
to progressively increase. The assumption is that the
changes in chamber pH during incubation (as much as
0.2 pH units) are secondary to the oxidation rate as a function of oxygen concentration.
To demonstrate the influence of a change in the iron oxidation kinetics we generated a model where a constant initial iron flux was delivered to the chamber (height = 10 cm).
The term ‘initial’ in this case refers to the primary benthic
flux without taking into account any oxic removal that
may take place in the chamber or chamber sample storage
vessel. At 2 h intervals the accumulated iron is subjected to
a loss term such that
Fe conc. − observed
Chamber Fe (nM)
in the Black Sea have shown that sediments may even become a sink for iron as it is scavenged by porewater sulfide
(Pakhomova et al., 2007); therefore, and because of the
dynamics between the benthic biota and chemistry, it is unclear precisely how the magnitude of benthic–ocean exchange will be altered once lower oxygen concentrations
become a persistent feature. However, based on our results,
relatively small changes in bottom water oxygen concentration may promote large changes in the benthic Fe flux, but
as oxygen concentration continues to decline, benthic Fe
fluxes may also lessen.
3999
40
Time (hrs)
Fig. 13. Decreasing bottom water oxygen concentrations change
the kinetics of iron oxidation and increase the half life of dissolved
Fe(II) in the chamber with the effect that less Fe(II) is re-oxidized
and returned to the sediments, potentially explaining some of the
observed curvature in our measured chamber iron concentrations
(see Fig. 3). Measured values for individual chambers from
Eel River station at 125 m depth are well matched by model
outputs for constant iron fluxes of 1 mmol m2 d1 (BIF-I) and
3 mmol m2 d1 (BIF-II), respectively.
sumes an initial iron flux of 1 mmol m2 d1 for BIF-I and
3 mmol m2 d1 for BIF-II, respectively. The two example
chambers had different initial and final chamber oxygen
concentrations, yet both were well fitted by the Lohan
and Bruland (2008) formulation. This outcome suggests
that a remarkably large rate of Fe(III) remobilization occurs within BIF sediments such that it generates an initial
Fe(II) flux of 1–3 mmol m2 d1. The model output also
shows that T1/2 likely increased from 1–3 to 8–11 h over
the duration of the experiments. A longer T1/2 implies that
relatively less dissolved Fe(II) is oxidized to Fe(III) and removed as iron-oxyhydroxide. The effect on the iron isotope
composition is that the iron in the chamber will become
increasingly heavier as the iron oxidation kinetics shift in
favor of a relatively larger dissolved Fe(II) pool (Bullen
et al., 2001; Rouxel et al., 2008). In the California Borderland Basin sites, the initial bottom water oxygen concentrations were very low, suggesting that T1/2 at the start of the
experiments was already comparable to the duration of the
incubation, and oxidative iron loss was minimal. Consequently, the isotope compositions of dissolved iron remained relatively constant during the experiment, even as
the chambers turned anoxic.
Both processes (shoaling of the oxygen penetration
depth and change in iron oxidation kinetics) would lead
4000
S. Severmann et al. / Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
to a progressive increase in the d56Fe values of the chamber
dissolved iron, which makes it difficult to evaluate quantitatively which process is more important. The likelihood is
that both processes occur simultaneously, which explains
why there is not a single threshold value for the transition
from linear to exponential increase in chamber iron concentrations (i.e., low to high flux regime). Rather, we observe
that ‘iron breakthrough’ occurs at bottom water oxygen
concentrations ranging from 60 to 80 lM, presumably
depending on a range of factors, including depth of bioirrigation and porewater iron concentrations.
6. CONCLUSIONS AND IMPLICATIONS FOR PAST
AND FUTURE COASTAL IRON CYCLING
The primary motivation for this study was to assess the
magnitude of the benthic iron flux along the continental
shelf, and to assess if it has a recognizable iron isotope signature that is distinct from other oceanic iron sources. As
predicted from previous studies of porewater isotope compositions in shallow coastal sediments (Bergquist and Boyle, 2006; Severmann et al., 2006), we found that the benthic
iron flux has a light isotope composition that makes it distinct from other continental weathering sources. The mean
d56Fe value of this flux generally ranges between 2.5&
and 3.5&, although it is somewhat sensitive to bottom
water oxygen concentrations, especially in sediments dominated by solute transport associated with bioirrigation. If a
significant portion of this shelf iron flux makes it to the
open ocean it is likely that it is identifiable by its light isotope composition. Further modification of the isotope composition may occur as the iron is transported off shore,
although quantitative utilization of bioavailable iron by
the biota in HNLC regions will likely preserve the isotope
composition of the shelf source in the cells.
Our general thesis regarding the first order processes
that regulate the benthic iron flux is summarized in the
schematic diagram of Fig. 2. The general concept is that
during the lower productivity period of winter, rivers discharge most of their particulate material, and some of that
material will maintain residence along the continental shelf.
During the more productive spring period, primary production is likely fueled in part by the upwelling of nutrient-rich
near-shore waters (Fig. 2, bottom). This production then
produces a pool of reactive organic material that settles
to the seafloor and creates a sedimentary package that
has sufficient reducing potential to mobilize the particulate
metals. The combination of these ideas is partially borne
out in the high “dissolvable” water column iron concentrations, which have been observed previously along the Oregon margin (Chase et al., 2005a), as well as along the
central California margin (e.g., Elrod et al., 2008). Our specific interest here was focused on comparing eastern-boundary sites having high particulate river discharge (Eel and
Umpqua Rivers) with regions of presumably lower particulate discharge (e.g., Monterey Bay). Our general hypothesis
was that the higher supply of continental “reactive” iron in
the region near the rivers would result in a higher benthic
flux as compared to “average” continental margin regions.
Although this idea maintains merit, it is also clear that
other shelf regions (e.g., Morro Bay) may also produce high
benthic iron fluxes and that there is considerable (over an
order of magnitude) variation in continental shelf benthic
fluxes. Part of this variability appears to be linked to variations in dissolved oxygen. For example, the Umpqua River region dissolved oxygen is generally lower than the areas
around the Eel River, and this difference may help maintain
the high benthic iron fluxes during both study periods.
Our experiments highlight the effect of variable bottom
water oxygen on the magnitude of the benthic iron flux in
the context of past and current environmental changes. It
is quite likely that changes in coastal productivity and sea
level have had, or will have, an impact on the coastal and
likely also the oceanic iron cycle (Fig. 14). During the current interglacial or high sea level stand, the flux of iron is
assumed to be relatively high from shelf and low-oxygen regions. During the low sea level stands typical of glacial periods, much of the continental shelf may have been better
ventilated or partially exposed, thus benthic iron delivery
from the shelves would have been minimized during these
periods. In contrast, as sea level rises, more of the shelf
could become depositional, and if these environments also
Fe2+
O2 min
Interglacials
Glacials
Fe2+
Fe2+
O2 min
Future?
Fig. 14. Changing environmental conditions, such as sea-level
stand, ventilation, or surface productivity are likely to affect iron
cycling on continental shelves, with higher coastal benthic iron flux
during high sea-level stands and expanded oxygen minimum zones.
This change in coastal biogeochemical iron cycling may have
implications for productivity further offshore if iron can be
effectively transported to the open ocean.
Continental shelf iron flux and its isotope composition
become increasingly hypoxic the iron supply to the water
column could also increase (Fig. 14, future scenario). If a
feedback is established that further enhances production
and carbon export, the oxygen minimum zone could also
expand, as has been shown to be the case for low-latitude
coastal upwelling zones (Paulmier et al., 2008; Stramma
et al., 2008). This admittedly oversimplified model highlights the need for a firm biogeochemical framework to improve the accuracy of past interpretations and future
predictions.
ACKNOWLEDGMENTS
This work was generously funded by NSF through several
grants: OCE-0624704 to S.S., OCE-9911550, OCE-0241372 and
OCE-0624777 to J.M., OCE-0624443 to W.M.B., and OCE9911608 to D.E.H. We thank the captains and crews of the R.V.
Wecoma and R.V. Pt. Sur for top-notch support, tireless efforts
and good times at sea – we could not have done it without you.
We are also grateful to the many helping hands that assisted with
cruise preparation, sample collection and ship-board analysis,
namely the hands of Tim Riedel, Lisa Collins, Will Beaumont, William “James” Homoky, Jeremy Owens, and Jesse Muratli – we
could not have done it without you either. Kenneth Johnson and
Ginger Elrod are thanked for providing porewater iron concentrations for the Monterrey Bay. Last but not least, we very much
appreciate the constructive and insightful comments from an anonymous reviewer and from Seth John, as well as from the Associate
Editor Dave Burdige, and we thank them for their time and effort.
APPENDIX A. SUPPLEMENTARY DATA
Supplementary data associated with this article can
be found, in the online version, at doi:10.1016/
j.gca.2010.04.022.
REFERENCES
Allen J. S., Newberger P. A. and Federik J. (1995) Upwelling
circulation on the oregon continental shelf. Part I: response to
idealized forcing. J. Phys. Oceanogr. 25, 1843–1866.
Aller R. C. (1980a) Diagenetic processes near the sediment–water
interface of Long Island Sound: II. Fe and Mn. Adv. Geophys.
22, 351–415.
Aller R. C. (1980b) Quantifying solute distributions in the
bioturbated zone of marine sediments by defining an average
microenvironment. Geochim. Cosmochim. Acta 44, 1955–1965.
Aller R. C. (1998) Mobile deltaic and continental shelf muds as
suboxic, fluidized bed reactors. Mar. Chem. 61, 143–155.
Aller R. C. (2001) Transport and reaction in the bioirrigated zone.
In The Benthic Boundary Layer. Transport Processes and
Biogeochemistry (eds. B. P. Boudreau and B. B. Jørgensen).
Oxford University Press, Oxford.
Aller R. C. (2004) Conceptual models of early diagenetic processes:
the muddy seafloor as an unsteady, batch reactor. J. Mar. Res.
62, 815–835.
Aller R. C., Mackin J. E. and Cox R. T. (1986) Diagenesis of Fe
and S in Amazon inner shelf muds: apparent dominance of Fe
reduction and implications for the genesis of ironstones. Cont.
Shelf Res. 6, 263–289.
Anderson T. F. and Raiswell R. (2004) Sources and mechanisms
for the enrichment of highly reactive iron in euxinic Black Sea
sediments. Am. J. Sci. 304, 203–233.
4001
Archer D. and Devol A. (1992) Benthic oxygen fluxes on the
Washington shelf and slope: a comparison of in situ microelectrode and chamber flux measurements. Limnol. Oceanogr. 37,
614–629.
Archer D. E. and Johnson K. S. (2000) A model of the iron cycle in
the ocean. Global Biogeochem. Cycles 14, 269–280.
Arnold G. L., Weyer S. and Anbar A. D. (2004) Fe isotope
variations in natural materials measured using high mass
resolution multiple collector ICPMS. Anal. Chem. 76, 322–327.
Bakun A. (1973) Coastal upwelling indices, west coast of North
America, 1946–71. U.S. Dept. Commerce.
Balzer W. (1982) On the distribution of iron and manganese at the
sediment/water interface: thermodynamic versus kinetic control. Geochim. Cosmochim. Acta 46, 1153–1161.
Barnett P. R. O., Watson J. and Connelly D. (1984) A multiple
corer for taking virtually undisturbed samples from shelf,
bathyal, and abyssal sediments. Oceanol. Acta 7, 399–408.
Beard B. L., Johnson C. M., Skulan J. L., Nealson K. H., Sun H.
and Cox L. (2003) Application of Fe isotopes to tracing the
geochemical and biochemical cycling of Fe. Chem. Geol. 195,
87–117.
Berelson W. M. (1991) The flushing of two deep-sea basins,
Southern California Borderland. Limnol. Oceanogr. 36, 1150–
1166.
Berelson W. M. and Hammond D. E. (1986) The calibration of a
new free-vehicle benthic flux chamber for use in the deep sea.
Deep-Sea Res. A: Oceanogr. Res. Papers 33, 1439–1454.
Berelson W. M., McManus J., Coale K. H., Johnson K. S., Burdige
D. J., Kilgore T., Colodner D., Chavez F. P., Kudela R. and
Boucher J. (2003) A time series of benthic flux measurements
from Monterey Bay, CA. Cont. Shelf Res. 23, 457–481.
Berelson W. M. and Stott L. D. (2003) Productivity and organic
carbon rain to the California margin sea floor: modern and
paleoceanographic perspectives. Paleoceanography 18, 1002.
doi:10.1029/2001PA000672.
Bergquist B. A. and Boyle E. A. (2006) Iron isotopes in the
Amazon River system: weathering and transport signatures.
Earth Planet. Sci. Let. 248, 54–68.
Brand L. E. (1991) Minimum iron requirements of marine
phytoplankton and the implications for the biogeochemical control of new production. Limnol. Oceanogr. 36, 1756–
1771.
Brand L. E., Sunda W. G. and Guillard R. R. L. (1983) Limitation
of marine phytoplankton reproductive rates by zinc, manganese, and iron. Limnol. Oceanogr. 28, 1182–1198.
Bullen T. D., White A. F., Childs C. W., Vivit D. V. and Schulz M.
S. (2001) Demonstration of significant abiotic iron isotope
fractionation in nature. Geology 29, 699–702.
Cai W.-J. and Reimers C. E. (1995) Benthic oxygen flux, bottom
water oxygen concentration and core top organic carbon
content in the deep northeast Pacific Ocean. Deep-Sea Res. I:
Oceanogr. Res. Pap. 42, 1681–1699.
Canfield D. E. and Des Marais D. J. (1993) Biogeochemical cycles
of carbon, sulfur, and free oxygen in a microbial mat. Geochim.
Cosmochim. Acta 57, 3971–3984.
Cassar N., Bender M. L., Barnett B. A., Fan S., Moxim W. J.,
Levy, II, H. and Tilbrook B. (2007) The southern ocean
biological response to aeolian iron deposition. Science 317,
1067–1070.
Chan F., Barth J. A., Lubchenco J., Kirincich A., Weeks H.,
Peterson W. T. and Menge B. A. (2008) Emergence of anoxia in
the California current large marine ecosystem. Science 319, 920.
Chase Z., Hales B., Cowles T. J., Schwartz R. and van Green A.
(2005a) Distribution and variability of iron input to Oregon
coastal waters during the upwelling season. J. Geophys. Res.
110. doi:10.1029/2004JC002590.
4002
S. Severmann et al. / Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
Chase Z., Johnson K. S., Elrod V. A., Plant J. N., Fitzwater S. E.,
Pickell L. and Sakamoto C. M. (2005b) Manganese and iron
distributions off central California influenced by upwelling and
shelf width. Mar. Chem. 95, 235–254.
Chester R. and Hughes M. J. (1967) A chemical technique for the
separation of ferro-manganese minerals, carbonate minerals
and adsorbed trace elements from pelagic sediments. Chem.
Geol. 2, 249–262.
Coale K. H., Johnson K. S., Fitzwater S. E., Blain S. P. G., Stanton
T. P. and Coley T. L. (1998) IronEx-I, an in situ ironenrichment experiment: experimental design, implementation
and results. Deep-Sea Res. II: Top. Stud. Oceanogr. 45, 919–
945.
Colbert D. (2004) Geochemical cycling in a Pacific Northwest
Estuary, Oregon State University, Tillamook Bay, Oregon,
USA.
Crawford D. W., Lipsen M. S., Purdie D. A., Lohan M. C.,
Statham P. J., Whitney F. A., Putland J. N., Johnson W. K.,
Sutherland N., Peterson T. D., Harrison P. J. and Wong C. S.
(2003) Influence of zinc and iron enrichments on phytoplankton
growth in the northeastern subarctic Pacific. Limnol. Oceanogr.
48, 1583–1600.
Croot P. L. and Hunter K. A. (1998) Trace metal distributions
across the continental shelf near Otago Peninsula, New
Zealand. Mar. Chem. 62, 185–201.
Cullen J. J. (1991) Hypotheses to explain high-nutrient conditions
in the open sea. Limnol. Oceanogr. 36, 1578–1599.
de Haas H., van Weering T. C. E. and de Stigter H. (2002) Organic
carbon in shelf seas: sinks or sources, processes and products.
Cont. Shelf Res. 22, 691–717.
Duce R. A. and Tindale N. W. (1991) Atmospheric transport of
iron and its deposition in the oceans. Limnol. Oceanogr. 36,
1715–1726.
Elrod V. A., Berelson W. M., Coale K. H., and Johnson K. S.
(2004) The flux of iron from continental shelf sediments: a
missing source for global budgets. Geophys. Res. Lett. 31,
L12307, doi:10.1029/2004GL020216.
Elrod V. A., Johnson K. S., Fitzwater S. E. and Plant J. N. (2008)
A long-term, high-resolution record of surface water iron
concentrations in the upwelling-driven central California
region. J. Geophys. Res. 113, C11021.
Friedl G., Dinkel C. and Wehrli B. (1998) Benthic fluxes of
nutrients in the northwestern Black Sea. Mar. Chem. 62, 77–88.
Friedrich J., Dinkel C., Friedl G., Pimenov N., Wijsman J.,
Gomoiu M.-T., Cociasu A., Popa L. and Wehrli B. (2002)
Benthic nutrient cycling and diagenetic pathways in the northwestern Black Sea. Estu. Coast. Shelf Sci. 54, 369–383.
Froelich P. N., Klinkhammer G. P., Bender M. L., Luedtke N. A.,
Health G. R., Cullen D., Dauphin P., Hammond D. E.,
Hartman B. and Maynard V. (1979) Early organic matter in
pelagic sediments of the eastern equatorial Atlantic: suboxic
diagenesis. Geochim. Cosmochim. Acta 43, 1075–1090.
Fung I. Y., Meyn S. K., Tegen I., Doney S. C., John J. G. and
Bishop J. K. B. (2000) Iron supply and demand in the upper
ocean. Global Biogeochem. Cycles 14, 281–296.
Griggs G. B. and Hein J. R. (1980) Sources, dispersal and clay
mineral composition of fine-grained sediments off central and
northern California. J. Geol. 88, 541–566.
Hagadorn J. W., Stott L. D., Sinha A. and Rincon M. (1995)
Geochemical and sedimentologic variations in inter-annually
laminated sediments from Santa Monica Basin. Mar. Geol. 125,
111–131.
Hammond D. E., McManus J., Berelson W. M., Kilgore T. E. and
Pope R. H. (1996) Early diagenesis of organic material in
equatorial Pacific sediments: stoichiometry and kinetics. DeepSea Res. II: Top. Stud. Oceanogr. 43, 1365–1412.
Hammond D. E., Cummins K. M., McManus J., Berelson W. M.,
Smith G. and Spagnoli F. (2004) A comparison of methods for
benthic flux measurement along the California margin: shipboard core incubations vs. in situ benthic landers. Limnol.
Oceanogr. Methods 2, 146–159.
Heggie D., Kahn D. and Fischer K. (1986) Trace metals in
metalliferous sediments, MANOP Site M: interfacial pore water
profiles. Earth Planet. Sci. Let. 80, 106–116.
Homoky W. B., Severmann S., Mills R. A., Statham P. J. and
Fones G. R. (2009) Fe isotopes in pore-fluids of deep sea and
continental shelf sediments. Geology 37, 751–754.
Hutchins D. A. and Bruland K. W. (1998) Iron-limited diatom
growth and Si:N uptake ratios in a coastal upwelling regime.
Nature 393, 561–564.
Hutchins D. A., DiTullio G. R., Zhang Y. and Bruland K. W.
(1998) An iron limitation mosaic in the California upwelling
regime. Limnol. Oceanogr. 43, 1037–1054.
Hutchins D. A., Witter A. E., Butler A., Luther G. W. and II I.
(1999) Competition among marine phytoplankton for different
chelated iron species. Nature 400, 858–861.
Jickells T. D., An Z. S., Andersen K. K., Baker A. R., Bergametti
G., Brooks N., Cao J. J., Boyd P. W., Duce R. A., Hunter K.
A., Kawahata H., Kubilay N., laRoche J., Liss P. S., Mahowald N., Prospero J. M., Ridgwell A. J., Tegen I. and Torres R.
(2005) Global iron connections between desert dust, ocean
biogeochemistry, and climate. Science 308, 67–71.
Jickells T. D. and Spokes L. J. (2001) Atmospheric iron inputs to
the Ocean. In The Biogeochemistry of Iron in Seawater (eds. D.
R. Turner and K. A. Hunter). John Wiley, Chichester.
Johnson K. S., Berelson W. M., Coale K. H., Coley T. L., Elrod V.
A., Fairey W. R., Iams H. D., Kilgore T. E. and Nowicki J. L.
(1992) Manganese flux from continental margin sediments in a
transect through the oxygen minimum. Science 257, 1242–1245.
Johnson K. S., Chavez F. P., Elrod V. A., Fitzwater S. E.,
Pennington J. T., Buck K. R. and Walz P. M. (2001) The
annual cycle of iron and the biological response in central
California coastal waters. Geophys. Res. Lett. 28, 1247–1250.
Johnson K. S., Chavez F. P. and Friederich G. E. (1999)
Continental-shelf sediment as a primary source of iron for
coastal phytoplankton. Nature 398, 697–700.
Johnson K. S., Gordon R. M. and Coale K. H. (1997) What
controls dissolved iron concentrations in the world oceans?
Mar. Chem. 57, 137–161.
Klinkhammer G., Heggie D. T. and Graham D. W. (1982) Metal
diagenesis in oxic marine sediments. Earth Planet. Sci. Let. 61,
211–219.
Kuma K., Nishioka J. and Matsunaga K. (1996) Controls on
iron(III) hydroxide solubility in seawater: the influence of pH
and natural organic chelators. Limnol. Oceanogr. 41, 396–407.
Lam P. J. and Bishop J. K. B. (2008) The continental margin is a
key source of iron to the HNLC North Pacific Ocean. Geophys.
Res. Lett., 35.
Lam P. J., Bishop J. K. B., Henning C. C., Marcus M. A.,
Waychunas G. A. and Fung I. Y. (2006) Wintertime phytoplankton bloom in the subarctic Pacific supported by continental margin iron. Global Biogeochem. Cycles 20. doi:10.1029/
2005GB002557.
Leblanc K., Hare C. E., Boyd P. W., Bruland K. W., Sohst B.,
Pickmere S., Lohan M. C., Buck K., Ellwood M. and Hutchins
D. A. (2005) Fe and Zn effects on the Si cycle and diatom
community structure in two contrasting high and low-silicate
HNLC areas. Deep-Sea Res. I: Oceanogr. Res. Pap. 52, 1842–
1864.
Lentz S. J. (1995) U.S. contributions to the physical oceanography
of continental shelves in the early 1990’s. Rev. Geophys., 1225–
1236.
Continental shelf iron flux and its isotope composition
Lohan M. C., Aguilar-Islas A. M. and Bruland K. W. (2006) Direct
determination of iron in acidified (pH 1.7) seawater samples by
flow injection analysis with catalytic spectrophotometric detection: application and intercomparison. Limnol. Oceanogr.
Methods 4, 164–171.
Lohan M. C., Aguilar-Islas A. M., Franks R. P. and Bruland K.
W. (2005) Determination of iron and copper in seawater at pH
1.7 with a new commercially available chelating resin, NTA
Superflow. Analyt. Chim. Acta 530, 121–129.
Lohan M. C. and Bruland K. W. (2008) Elevated Fe(II) and
dissolved Fe in hypoxic shelf waters off Oregon and Washington: an enhanced source of iron to coastal upwelling regimes.
Env. Sci. Technol. 42, 6462–6468.
Mahowald N. M., Baker A. R., Bergametti G., Brooks N., Duce R.
A., Jickells T. D., Kubilay N., Prospero J. M. and Tegen I.
(2005) Atmospheric global dust cycle and iron inputs to the
ocean. Global Biogeochem. Cycles, 19.
Mahowald N. M., Engelstaedter S., Luo C., Sealy A., Artaxo P.,
Benitez-Nelson C., Bonnet S., Chen Y., Chuang P. Y., Cohen
D. D., Dulac F., Herut B., Johansen A. M., Kubilay N., Losno
R., Maenhaut W., Paytan A., Prospero J. M., Shank L. M. and
Siefert R. L. (2009) Atmospheric iron deposition: global
distribution, variability, and human perturbations. Annu. Rev.
Mar. Sci. 1, 245–278.
Martin J. M. and Fitzwater S. E. (1988) Iron deficiency limits
phytoplankton growth in the north-east Pacific subarctic.
Nature 331, 341–343.
Martin J. M. and Gordon R. M. (1988) Northeast Pacific iron
distribution in relation to phytoplankton productivity. DeepSea Res. 35, 177–196.
Martin J. M., Gordon R. M. and Fitzwater S. E. (1991) The case
for iron. Limnol. Oceanogr. 36, 1793–1802.
Martin J. M., Gordon R. M., Fitzwater S. E. and Broenkow W. W.
(1989) VERTEX: phytoplankton/iron studies in the Gulf of
Alaska. Deep-Sea Res. 36, 649–680.
McKee B. A., Aller R. C., Allison M. A., Bianchi T. S. and Kineke
G. C. (2004) Transport and transformation of dissolved and
particulate materials on continental margins influenced by
major rivers: benthic boundary layer and seabed processes.
Cont. Shelf Res. 24, 899–926.
McManus J., Berelson W. M., Coale K. H., Johnson K. S. and
Kilgore T. (1997) Phosphorous regeneration in continental
margin sediments. Geochim. Cosmochim. Acta 61, 2891–2907.
McManus J., Hammond D. E., Cummins K. M., Klinkhammer G.
P. and Berelson W. M. (2003) Diagenetic Ge–Si fractionation in
continental margin environments: further evidence for a nonopal Ge sink. Geochim. Cosmochim. Acta 67, 4545–4557.
Measures C. I., Yuan J. and Resing J. (1995) Determination of iron
in seawater by flow injection analysis using in-line preconcentration and spectrophotometric detection. Mar. Chem. 50, 3–12.
Milliman J. D. and Syvitski J. P. M. (1992) Geomorphic/tectonic
control of sediment discharge to the ocean: the importance of
small mountainous rivers. J. Geol. 100, 525–544.
Nishioka J., Ono T., Saito H., Nakatsuka T., Takeda S.,
Yoshimura T., Suzuki K., Kuma K., Nakabayashi S., Tsumune
D., Mitsudera H., Johnson W. K. and Tsuda A. (2007) Iron
supply to the western subarctic Pacific: importance of iron
export from the Sea of Okhotsk. J. Geophys. Res. 112.
doi:10.1029/2006JC004055.
Pakhomova S. V., Hall P. O. J., Kononets M. Y., Rozanov A. G.,
Tengberg A. and Vershinin A. V. (2007) Fluxes of iron and
manganese across the sediment–water interface under various
redox conditions. Mar. Chem. 107, 319–331.
Pakulski J. D., Coffin R. B., Kelley C. A., Holder S. L., Downer R.,
Aas P., Lyons M. M. and Jeffrey W. H. (1996) Iron stimulation
of Antarctic bacteria. Nature 383, 133–134.
4003
Paulmier A., Ruiz-Pino D. and Garcon V. (2008) The oxygen
minimum zone (OMZ) off Chile as intense source of CO2 and
N2O. Cont. Shelf Res. 28, 2746–2756.
Poulton S. W. and Canfield D. E. (2005) Development of a
sequential extraction procedure for iron: implications for iron
partitioning in continentally derived particulates. Chem. Geol.
214, 209–221.
Poulton S. W. and Raiswell R. (2002) The low-temperature
geochemical cycle of iron: from continental fluxes to marine
sediment deposition. Am. J. Sci. 302, 774–805.
Raiswell R. (2006) An evaluation of diagenetic recycling as a source
of iron for banded iron formations. In Evolution of Early
Earth’s Atmosphere, Hydrosphere and Biosphere-constraints
from Ore Deposits (eds. S. E. Kesler and H. Ohmoto).
Geological Society of America.
Raiswell R. and Anderson T. F. (2005) Reactive iron enrichment in
sediments deposited beneath euxinic bottom waters: constraints
on supply by shelf recycling. In Mineral Deposits and Earth
Evolution (eds. I. McDonald, A. J. Boyce, I. B. Butler, R. J.
Herrington and D. A. Polya). Geological Society Special
Publication 248, London.
Raiswell R., Benning L. G., Tranter M. and Tulaczyk S. (2008)
Bioavailable iron in the Southern Ocean: the significance of the
iceberg conveyor belt. Geochem. Trans. 9. doi:10.1186/14674866-9-7.
Reimers C. E., Jahnke R. A. and McCorkle D. J. (1992) Carbon
fluxes and burial rates over the continental slope and rise off
central California with implications for the global carbon cycle.
Global Biogeochem. Cycles 6, 199–224.
Rouxel O., Sholkovitz E., Charette M. and Edwards K. J. (2008)
Iron isotope fractionation in subterranean estuaries. Geochim.
Cosmochim. Acta 72, 3413–3430.
Rutgers van der Loeff M. M., Anderson L. G., Hall P. O. J.,
Iverfeldt A., Josefson A. B., Sundby B. and Westerlund S. F. G.
(1984) The asphyxiation technique: an approach to distinguishing between molecular diffusion and biologically mediated
transport at the sediment–water interface. Limnol. Oceanogr.
29, 675–686.
Service R. F. (2004) New dead zone off Oregon coast hints at sea
change in currents. Science 305, 1099.
Severmann S., Johnson C. M., Beard B. L. and McManus J. (2006)
The effect of early diagenesis on the Fe isotope compositions of
porewaters and authigenic minerals in continental margin
sediments. Geochim. Cosmochim. Acta 70, 2006–2022.
Severmann S., Lyons T. W., Anbar A., McManus J. and Gordon
G. (2008) Modern iron isotope perspective on Fe shuttling in
the Archean and the redox evolution of ancient oceans. Geology
36, 487–490.
Skoog A., Hall P. O. J., Hulth S., Paxéus N., Van Der Loeff M. R.
and Westerlund S. (1996) Early diagenetic production and
sediment–water exchange of fluorescent dissolved organic
matter in the coastal environment. Geochim. Cosmochim. Acta
60, 3619–3629.
Sommerfield C. K. and Nittrouer C. A. (1999) Modern accumulation rates and a sediment budget for the Eel shelf: a flooddominated depositional environment. Mar. Geol. 154, 227–241.
Staubwasser M., von Blanckenburg F. and Schoenberg R. (2006)
Iron isotopes in the early marine diagenetic iron cycle. Geology
34, 629–632.
Stott L. D., Berelson W., Douglas R. and Gorsline D. (2000)
Increased dissolved oxygen in Pacific intermediate waters due to
lower rates of carbon oxidation in sediments. Nature 407, 367–
370.
Stramma L., Johnson G. C., Sprintall J. and Mohrholz V. (2008)
Expanding oxygen-minimum zones in the tropical oceans.
Science 320, 655–658.
4004
S. Severmann et al. / Geochimica et Cosmochimica Acta 74 (2010) 3984–4004
Sunda W. G. and Huntsman S. A. (1995) Iron uptake and growth
limitation in oceanic and coastal phytoplankton. Mar. Chem.
50, 189–206.
Sundby B. r., Anderson L. G., Hall P. O. J., Iverfeldt k., van der
Loeff M. M. R. and Westerlund S. F. G. (1986) The effect of
oxygen on release and uptake of cobalt, manganese, iron and
phosphate at the sediment–water interface. Geochim. Cosmochim. Acta 50, 1281–1288.
Tengberg A., Hall P. O. J., Andersson U., Linden B., Styrenius O.,
Boland G., de Bovee F., Carlsson B., Ceradini S., Devol A.,
Duineveld G., Friemann J.-U., Glud R. N., Khripounoff A.,
Leather J., Linke P., Lund-Hansen L., Rowe G., Santschi P., de
Wilde P. and Witte U. (2005) Intercalibration of benthic flux
chambers: II. Hydrodynamic characterization and flux comparisons of 14 different designs. Mar. Chem. 94, 147–173.
Teutsch N., von Gunten U., Porcelli D., Cirpka O. A. and Halliday
A. N. (2005) Adsorption as a cause for iron isotope fractionation in reduced groundwater. Geochim. Cosmochim. Acta 69,
4175–4185.
Thamdrup B. (2000) Bacterial manganese and iron reduction in
aquatic sediments. In Advances in Microbial Ecology (ed. B.
Schink). Kluwer Academic/Plenum Publisher, New York.
Thamdrup B., Fossing H. and Jørgensen B. B. (1994) Manganese,
iron, and sulfur cycling in a coastal marine sediment, Aarhus
Bay, Denmark. Geochim. Cosmochim. Acta 58, 5115–5129.
Warnken K. W., Gill G. A., Griffin L. L. and Santschi P. H. (2001)
Sediment–water exchange of Mn, Fe, Ni and Zn in Galveston
Bay, Texas. Mar. Chem. 73, 215–231.
Wetz M. S., Hales B., Wheeler P. A., Chase Z. and Whitney M.
(2006) Riverine input of macronutrients, iron and organic
matter to the coastal ocean off Oregon, USA, during winter.
Limnol. Oceanogr. 5, 2221–2231.
Wheatcroft R. A. (2000) Oceanic flood sedimentation: a new
perspective. Cont. Shelf Res. 20, 2059–2066.
Wheatcroft R. A. (2006) Time-series measurements of macrobenthos abundance and sediment bioturbation intensity on a flooddominated shelf. Prog. Oceanogr. 71, 88–122.
Wheatcroft R. A. and Borgeld J. C. (2000) Oceanic flood deposits
on the northern California shelf: large-scale distribution and
small-scale physical properties. Cont. Shelf Res. 20, 2163–2190.
Wheatcroft R. A. and Sommerfield C. K. (2005) River sediment
flux and shelf sediment accumulation rates on the Pacific
Northwest margin. Cont. Shelf Res. 25, 311–332.
Wheatcroft R. A., Sommerfield C. K., Drake D. E., Borgeld J. C.
and Nittrouer C. A. (1997) Rapid and widespread dispersal of
flood sediment on the northern California margin. Geology 25,
163–166.
Associate editor: David J. Burdige