University of Southern California Dissertations and Theses

The distributions and geochemistry of iodine and copper in the Pacific Ocean

(USC Thesis Other)

The distributions and geochemistry of iodine and copper in the Pacific Ocean

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

The Distributions and Geochemistry of Iodine and Copper in the Pacific Ocean

By

Rintaro Moriyasu

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(OCEAN SCIENCES)

May 2022

Copyright 2022 Rintaro Moriyasu

ii

Did you find it in the sandy ground?
Does it emulate the ocean sound?

-Aaron Freeman & Michael Melchiondo

This thesis is dedicated to my wife, Callie.

iii

Acknowledgements
This thesis is the accumulation of all that I have done over the past five years. Without
the support and kindness from my family, I would not have been able to finish this research nor
my doctorate, and for that I would like to thank them. Specifically, I would like to thank my
wife, Callie, without whom I would not have finished this doctorate. Truthfully, this PhD was the
longest grind of my life, and half of it happened during the COVID-19 pandemic. Callie was
there to support me, during this difficult time for humanity, and so I am dedicating this doctoral
thesis to her. I would also like to thank the rest of my family, mainly my parents Eiichi and
Mikako, who had the patience to watch me spend five years doing this. They would often send
words of encouragement and offer me financial support, without which I could not have spent
years on this research while being paid minimum wage.
I would also like to thank my advisor, James Moffett, for the support and mentorship he
has offered over the years. His meticulous attention to scientific details has improved both my
research and manuscript writing abilities. My dissertation committee members were also helpful
in providing feedback.
Finally, I would like to thank all of the friends that I have made during my time at USC.
Specifically, I want to thank the friends I made in the Moffett and John labs, Kenny Bolster,
Alexis Floback, Xiaopeng Bian, and Shun-Chung Yang; they helped me measure total dissolved
metals for many of my samples. Without them, this doctorate would have likely taken twice the
time that it actually took to complete it.
iv

TABLE OF CONTENTS

Dedication ii
Acknowledgements iii
List of Tables vi
List of Figures vii-viii
Abstract ix-x
Introduction 1

Chapter 1: The Distribution and Redox Speciation of Iodine in the Eastern Tropical North
Pacific Ocean
1.0.0 Abstract 3
1.1.0 Introduction 4
1.2.0 Sampling 7
1.3.0 Methods 8
1.4.0 Results 9
1.5.0 Discussion 33
1.6.0 Conclusions 39
1.7.0 Acknowledgements 40

Chapter 2: Determination of Inert and Labile Copper on GEOTRACES Samples Using a
Novel Solvent Extraction Method
2.0. Abstract 41
2.1. Introduction 42
2.2. Sampling 45
2.3. Materials and Reagents 47
2.4. Methods 48
2.5. Competitive Ligand Exchange 51
2.6. Results 55
2.7. Application of Method to GEOTRACES Sample 60
2.8. Discussion 70
2.9. Acknowledgements 72

Chapter 3: The Distribution of Iodine Redox Speciation along the GEOTRACES GP15
Transect
3.0. Abstract 73
3.1. Introduction 74
3.2. Sampling 77
3.3. Materials and Instrumentation 81
3.4. Methods 81
3.5. Results 84
3.6. Discussion 91
3.7. Conclusions 101
3.8. Acknowledgements 102

v

Chapter 4: The Impact of Copper Lability on Distribution and Cycling along the
GEOTRACES GP15 Transect
4.0. Abstract 103
4.1. Introduction 104
4.2. Sampling 107
4.3. Materials and Instrumentation 109
4.4. Results 111
4.5. Discussion 119
4.6. Conclusions 126
4.7. Acknowledgements 126

References 127

Appendices 138
Appendix A: Supplemental Information for Chapter 1 138
Appendix B: Supplemental Information for Chapter 4 184

vi

List of Tables
Table 2-1 List of Chelators, Concentrations, and Detection Windows 53
Table 2-2 Frozen Samples from GA03 59
Table 2-3 Conditional Stability Constant of Strong Copper Ligands 67

Table 3-1 Intercalibration of Iodide measurements 87
Table 3-2 Rate of Iodide Oxidation Calculation 97

Table 4-1 Columbia River Outflow of Inert Copper 118
Table 4-2 Destruction of Inert Copper by natural UV-irradiation 125
vii

List of Figures
Figure 1-1 Map of R/V Roger Revelle and Falkor cruises 7
Figure 1-2 Scatterplot of Salinity and Nitrite 10
Figure 1-3 Scatterplot of Iodide 11
Figure 1-4 Transect Profiles of Revelle Cruise Pt. 1 13
Figure 1-5 Transect Profiles of Revelle Cruise Pt. 2 15
Figure 1-6 Transect Profiles of Falkor Cruise 17
Figure 1-7 Profiles at Station 1 of Revelle 21
Figure 1-8 Plot of Temperature versus Salinity at Station 3 of Revelle 22
Figure 1-9 Profiles at Station 3 of Revelle 23
Figure 1-10 Profiles at Station 34 of Revelle 25
Figure 1-11 Profiles at Station 17 of Revelle 27
Figure 1-12 Profiles at Station 9 of Falkor 28
Figure 1-13 Iodide at 110
o
W, 14
o
N from Revelle and Falkor 29
Figure 1-14 Casts 1 and 2 of Falkor Station 2 30
Figure 1-15 Profiles of Cast 1 at Station 2 of Falkor 31
Figure 1-16 Profiles of Cast 2 at Station 2 of Falkor 32

Figure 2-1 Map of Samples Taken from GA03 46
Figure 2-2 Map of Stations 4 and 9 of GP15 and Catalina Island 47
Figure 2-3 Sample Extraction over Time 56
Figure 2-4 Copper Extracted versus Oxine Concentration 57
Figure 2-5 Results from DTPA Titration Experiment 58
Figure 2-6 Depth Profiles of Labile, Total, and Inert Cu from Station 9 62
Figure 2-7 Competitive Ligand Exchange of Samples from Station 9 68

Figure 3-1 Full Transect of GP15 78
Figure 3-2 Shelf Stations of GP15 79
Figure 3-3 Main Transect Stations of GP15 80
Figure 3-4 Iodide Measurements Using Argon and Nitrogen at Station 13 85
Figure 3-5 Difference in Measurements at Station 13 86
Figure 3-6 Intercalibration of Iodide Measurements at Station 35 of GP15 87
Figure 3-7 Transect Profiles of GP15 88
Figure 3-8 Nonlinear Regression of samples from station 23 and 3 91
Figure 3-9 Surface Iodide Concentrations along a Latitudinal Gradient 93
Figure 3-10 Iodide and Fluorescence Depth Profiles 94
Figure 3-11 Scatterplots of Iodide Concentrations to Fluorescence 95
Figure 3-12 Surface Iodide plotted against Net Primary Productivity 96
Figure 3-13 Equatorial Iodide and Nitrite Depth Profiles 98
Figure 3-14 Equatorial Iodide Transect Profile with Nitrite Overlay 99

Figure 4-1 The map of the GP15 Transect with markers for Cu speciation 108
Figure 4-2 Map of Columbia River with markers 109
Figure 4-3 Distributions from Station 7 to 35 112
Figure 4-4 Distributions along the Shelf of Alaska 115
viii

Figure 4-5 Scatter of Atlantic vs. Pacific for Percent Inert Cu 117
Figure 4-6 Percent Labile Cu and Iodide at Stations 7 and 14 124
ix

Abstract
The distributions of iodine and copper speciation were measured along three different
cruises, within the (mainly North) Pacific Ocean, one of which was a GEOTRACES cruise.
GEOTRACES is an international collaborative effort to elucidate the biogeochemical cycling of
various trace elements and isotopes (TEIs). Many of these elements are considered nutrients, and
their distributions and cycling have biological impacts, especially on phytoplankton.
Marine iodine is mostly found in its two inorganic forms: iodate and iodide. The former
is found throughout most of the oxygenated ocean while the latter was thought to be enriched in
the surface ocean and in anoxic regions, referred to as Oxygen Deficient Zones (ODZs). In this
thesis, high iodide concentrations were found in these ODZs, which cannot be explained by the
in situ reduction of available iodate, and likely to originate as a “plume” from the reducing shelf
margins off of the Mexican coast. These high iodide values were referred to as “excess iodine”
and found to persist even at the edges of these ODZs, likely to do with the slow abiotic oxidation
rate of iodide by oxygen and other terminal electron acceptors (chapter 1). These measurements
were repeated in the GEOTRACES GP15 which left from the shelf of Alaska and transected to
Tahiti. Although high rates of sulfate reduction have been predicted to occur off of Alaska, no
large plume of highly concentrated iodide was observed along this cruise. Instead, this cruise
offered the opportunity to find that: previous iodide measurements made within the South Pacific
are likely erroneous, and the rate of surface iodide oxidation using a non-linear fit with
complementary beryllium measurements (chapter 3).
Copper, a unique element, is considered a nutrient at low concentrations and a toxin
higher concentration. It is also considered a “hybrid” element for its depth profiles which exhibit
behaviors of both “nutrient-like” and “scavenged” distributions; this behavior is exhibited in the
x

form of a linear depth profile which increase from surface to benthic waters. Most (≥99%) of the
dissolved copper pool is bound by organic molecules and considered biologically inert, and
many workers study the compounds that bind the metal through electrochemical methods. One
hypothesis, for the linear depth distribution, is that some of these compounds bind to copper so
tightly that it prevents scavenging by sinking particulate matter. This tightly bound fraction of
dissolved copper is referred to as “inert copper”. This work explored this hypothesis and required
the creation of a novel method to physically separate (via a process of liquid-liquid extraction)
dissolved copper into two pools: labile and inert (chapter 2). The method was then applied to, the
previously mentioned transect, GEOTRACES GP15 (chapter 4). The data reveal that 60-90% of
marine copper is rendered chemically inert throughout the ocean. Labile copper, on the other
hand, exhibits a profile similar to other scavenged elements at most locations: higher
concentrations at surface waters with a sharp decline with depth due to scavenging by particles.
There appears to also be a benthic source of labile copper, as well, which is likely due to the
destruction of the binding capacity of the particulate with diagenesis. Our findings conclude that
copper complexation by, ligands and organic materials, are not all reversible, as suggested by our
current paradigm.
1

Introduction
This PhD thesis focuses on the speciation of iodine and copper in the Pacific Ocean. For
marine iodine, it mainly takes the form of iodate, the oxidized form, and remains fairly
conservative throughout the world’s oceans (Luther et al., 1995). However, in certain regions
where the waters are permanently anoxic, the reduced form iodide becomes the dominant form
(Luther & Farrenkopf., 2002). One such classification of anoxic seawater is referred to as
Oxygen Deficient Zones (ODZ), characterized by its low oxygen and nitrite accumulation. As
such, iodide has been shown to be tracers for dissolved oxygen and in situ redox conditions
(Cutter et al., 2018). Iodide is also prevalent in surface waters where biology and photochemistry
help the formation of the reduced species (Chance et al., 2014, 2019). The two chapters on iodine
focus on the redox speciation of iodine in the ODZ, referred to as the Eastern Tropical North
Pacific, and the GEOTRACES GP15 transect from Alaska to Tahiti.
Copper speciation, on the other hand, is based on the complexation of the metal ion,
rather than a product of environmental redox conditions. Copper is commonly referred to as the
“Goldilocks” element since it is both a nutrient and a toxin depending on its concentration. It is
also known for having a depth distribution which increases linearly with depth. This type of
distribution is referred to as a hybrid of the nutrient and scavenged elements. There are two
hypotheses for this behavior. The first of these two is referred to as reversible scavenging (Little
et al., 2014) and was previously applied to Thorium-230 because
230
Th also has a linear
distribution (Bacon & Anderson., 1982). However, unlike
230
Th, which has a ubiquitous source
through the α-decay of Uranium-234, the main input of copper is fluvial (Little et al., 2014;
Richon & Tagliabue., 2019). Copper does not have a ubiquitous source throughout the water
column, so a direct comparison between the two is unreasonable.
2

The second hypothesis for the linear depth distribution of copper is the existence of an
inert fraction. The concept of inert copper was previously measured in coastal waters with high
concentrations of organic compounds (Kogut & Voelker., 2003). Additionally, the rate of copper
scavenging has been found to be affected by complexation by ligands since these ligands prevent
the adsorption of copper onto the surface of sinking particles (van den Berg et al., 1987). The
only way to accurately model copper, using the AWESOME Ocean Circulation Inverse Model
(John et al., 2020), was to invoke this inert fraction (Seth John – personal communication). The
two chapters, on the topic of copper in this thesis, involve the development of a method to
quantify inert copper and the application of this method to aforementioned GEOTRACES GP15
transect.
This thesis also involved several firsts in the field. For iodine, this includes: the first
experimental estimate of iodide turnover in the ocean, the first cross basin surface to seafloor
section of iodine, and the first ever intercomparison of iodine speciation methods between
voltammetry and mass spectrometry. For copper, this was the first demonstration that ≥ 90% of
the total dissolved fraction of the metal being inert. Finally, by quantifying the inert fraction, it
demonstrated that assumptions underlying previous speciation measurements have been wrong,
in regard to the equilibration of copper species.
3

Chapter 1
The Distribution and Redox Speciation of Iodine in the Eastern Tropical
North Pacific Ocean
Authors: Rintaro Moriyasu, Natalya Evans, Kenneth M. Bolster, Dalton S. Hardisty, and James
W. Moffett
Originally published in: Global Biogeochemical Cycles, 34
Published on: January 28
th
, 2020
1.0 Abstract
The distributions of iodate (IO3
-
), iodide (I
-
), nitrite (NO2
-
) and oxygen (O2) were
determined on two cruises along the Eastern Tropical North Pacific (ETNP) in 2018. Analyses
were performed on two zonal and one meridional transect through the large Oxygen Deficient
Zone (ODZ). This ODZ is characterized by an extensive region of low O2 with high rates of
denitrification and NO2
-
accumulation. Iodine distribution and chemistry resembled that of the
two other large ODZs, the Arabian Sea and Eastern Tropical South Pacific (ETSP). Within the
zone of NO2
-
accumulation, iodine was present as I
-
while IO3
-
concentrations were low to
undetectable. The most reducing conditions were associated with a specific water mass, referred
to as the 13 °C Water, the same water mass that contains the reducing zone within the ETSP.
Iodide maxima centered around a potential density of 26.5 kg/m
3
, same as in the ETSP. One
difference between the ETSP and ETNP is that residual IO3 was always present in the ETSP,
whereas in the ETNP, it was frequently undetectable. Although estimates for IO3
-
reduction are
near instantaneous, according to Arabian Sea data, the observations from the ETSP and ETNP
require slower reduction rates on the order of decades. Throughout much of the ODZ, iodine
concentrations were higher than the mean oceanic value. This “excess iodine” is attributed to
lateral inputs from sedimentary margins. These results suggest that, like the ETSP, margin inputs
are significant throughout the basin and influence the nitrogen and iron cycles as well.
4

1.1.0 Introduction
Iodine, a biologically important trace element, is found in most of the world’s ocean at
concentrations averaging 470 nM (Luther et al., 1995). It exists primarily in the form of iodate
(IO3
-
), which is thermodynamically favored, but it is also found as iodide (I
-
) at low nanomolar
concentrations in most oxygenated water; this is with the exception of surface waters where IO3
-

is biologically reduced to I
-
(Luther et al., 1995; Tsunogai & Sase., 1969; Wong & Brewer.,
1976; Chance et al., 2014). The sum of the concentrations of IO3
-
and I
-
have been found to equal
the concentration of total iodine (Rue et al., 1997). At the surface, I
-
accumulates to 50-250 nM
from the reduction of IO3
-
by phytoplankton (Farrenkopf et al., 2002). Comparatively, the abiotic
oxidation of I
-
to IO3
-
does not occur readily in nature because the oxidation by oxygen (O2) is
kinetically unfavorable (Luther et al., 1995). Luther et al., (1995) argued that the oxidation, by
both triplet and singlet oxygen species, is unfeasible without light (<473 nm based on the
energetic differences between the Highest Occupied Molecular Orbital of I
-
and the Lowest
Occupied Molecular Orbital of triplet oxygen) which does not penetrate below the euphotic zone.
Oxidation by reactive oxygen species (ROS), peroxides, hydroxyl anion, and hydroxyl radicals
are possible, however, and produce hypoiodous acid (HOI) which reforms to I
-
or reacts readily
with organic matter to form organo-iodine compounds (Luther et al., 1995). Methyl-iodide and
other volatile, halogenated methyl species have been found to be produced in the North Atlantic
Ocean; these organic-iodine species have been found to be the main transport of iodine from
seawater into the atmosphere (Lovelock et al., 1973; Moore & Tokarcsyk., 1992). Butler &
Smith., (1980) also demonstrated that I
-
can be converted to IO3
-
with UV-irradiation and
peroxides.
5

In contrast to the slow oxidation kinetics of I
-
in oxygenated water, the reduction of IO3
-

in the absence of O2 is likely to be the result of biological activity and may occur more rapidly;
Farrenkopf et al., (1997) observed seemingly instantaneous reduction of IO3
-
during shipboard
incubation in the Arabian Sea, and two separate studies have observed dissimilatory IO3
-

reduction in seawater mediums inoculated with bacteria (Farrenkopf et al., 1997; Amachi et al.,
2007). Babbin et al., (2017) hypothesized that IO3
-
can act as an oxidant for nitrite (NO2
-
) at the
top of the Oxygen Deficient Zone (ODZ); this reaction forms nitrate (NO3
-
) and I
-
. Cutter et al.,
(2018) showed that IO3
-
persisted at low concentrations throughout the ETSP ODZ, an important
observation discussed throughout this paper. ODZs are defined in this paper as regions where O2
is absent or negligible and NO3
-
is the primary terminal electron acceptor.
Iodide concentrations are relatively low in comparison to IO3
-
in the oxygenated water
column, but I
-
dominates iodine speciation in the ODZ of the Arabian Sea and the Eastern
Tropical South Pacific (ETSP) (Farrenkopf et al., 2002; Cutter et al., 2018). Compared to the
other two ODZs, where detailed transects have been evaluated, iodine speciation in the Eastern
Tropical North Pacific (ETNP) is less well studied. Rue et al., (1997) reported a single profile
showing quantitative conversion of IO3
-
to I
-
, within the ODZ, suggesting that the iodine redox
couple is a sensitive indicator of its redox environment; in this case, the total iodine was
unchanged throughout the ODZ. In striking contrast, the Arabian Sea (Farrenkopf et al., 2002)
and the ETSP (Cutter et al., 2018) exhibited strong local maxima in total iodine within their
ODZs. This “excess” in total iodine is defined as the difference between the observed total iodine
and the expected total iodine concentration. Several approaches have been used to determine the
“expected” iodine concentration. Farrenkopf et al., (2002) used an empirical relationship
between I
-
and salinity to obtain their expected value within the ODZ. Cutter et al., (2018) used a
6

similar approach to calculate an expected IO3
-
, within the ODZ, based on an empirical
relationship between phosphate (PO4
3-
) and IO3
-
derived below the hypoxic water.
Farrenkopf et al., (2002) argued that excess iodine in the Arabian Sea must be derived
from reducing shelf sediments rather than remineralization of sinking organic particles because
the I:C ratio has a much higher sedimentary organic matter than particulate organic matter in the
water column. They reported an excess in iodine concentrations throughout the Arabian Sea.
Similarly, Cutter et al., (2018) found, in the ETSP, iodine concentrations exceeding total
inorganic iodine found in oxic marine waters. These excesses in I
-
concentrations cannot be
explained by in situ IO3
-
reduction (Cutter et al., 2018).
The single iodine profile from the ETNP, reported by Rue et al., (1997), showed that IO3
-

was undetectable throughout the ODZ where O2 was absent. This implied that in the absence of
O2, IO3
-
is quickly consumed, presumably as a terminal electron acceptor in respiration, which is
consistent with experimental observations of IO3
-
reduction in seawater (Farrenkopf et al., 1997;
Amachi et al., 2007). However, in the ETSP, residual IO3
-
was detected throughout the ODZ in
all stations which suggests that IO3
-
utilization may be under kinetic control. The data from the
ETNP and ETSP suggest that iodine distribution in the ODZ cannot be expressed as a simple
function of O2. In some cases, as in Rue et al., (1997), the redox speciation appears to be
controlled by oxygen distribution. However, in the ETSP, the approach to the equilibrium
predicted distribution is incomplete, and the sharp local maxima seem similar to those observed
in the profiles for the nitrogen cycle (i.e. NO2
-
) rather than for O2. The Arabian Sea has features
common to both of these end members (Farrenkopf et al., 2002).
In order to learn more about these factors, which lead to the distribution of iodine we see
in the water column within and between ODZs, we conducted a survey of iodine chemistry
7

throughout the ODZ of the ETNP. Since there is currently only one reported profile from
VERTEX III, 15 °N, 107 °W (shown as “X” in Figure 1) (Rue et al., 1997), it was the least
studied of all three ODZs with respect to iodine. We carried out the most extensive
measurements of iodine in a single study to date, in the context of a broad array of hydrographic
and nitrogen cycle parameters, and evaluated relationships relative to O2, NO2
-
, as well as excess
iodine. We also used the data to verify a water mass analysis, reported in a companion paper,
Evans et al. (2020), which provide insight into the details of our results.
1.2.0 Sampling
Samples were collected from two cruises: one aboard the R/V Roger Revelle from March
26
th
to May 2
nd
, 2018, and the other aboard the R/V Falkor from June 24
th
to July 17
th
2018. The
cruise tracks are shown in Figure 1.
Figure 1. Map of combined Revelle and Falkor cruises. Stations prefixed by “RR” connotes stations taken aboard the Revelle;
stations prefixed by “F” connotes stations taken aboard the Falkor. X on map denotes location of VERTEX III (Vertical
Transport and Exchange 15

°N, 107 °W, 19-26 November, 1982).

Samples were collected, using a Seabird CTD, with calibrated sensors for temperature,
conductivity, pressure, fluorometer, transmissometer, and O2 concentrations. Each one of these
CTD carousels held 24 12 L GO-FLO bottles which were closed at selected depths.
8

Samples were filtered using a 0.2 µm (AcroPak
TM
200) filter within 2 hours of sample
collection using a Masterflex
®
pump. Filtered samples were measured as quickly as possible,
which was typically within 6 hours of filtration, and were frozen in cases where this was not
feasible. Previous studies have shown that iodine is stable on the timescales of months to years in
filtered seawater (Campos et al., 1997). One study of IO3
-
reduction, in the Arabian Sea, found
that samples collected from the ODZ (quantitatively I
-
) were found to show no oxidation of I
-

(Farrenkopf et al., 1997). This was even after prolonged exposure of the samples to atmospheric
O2 while sitting on the lab bench.
1.3.0 Methods
The method of I
-
determination was adapted from Rue et al., (1997) which was originally
adapted from Luther et al., (1988). This method uses a cathodic square wave stripping
voltammetry with a Hanging Mercury Drop Electrode (HMDE) and a calomel reference
electrode. The procedure requires a 10 mL seawater sample to be treated with 150 μL of 0.2%
Triton X 100 (Sigma Aldrich – BioX grade) and purged under argon gas for five minutes. Once
purged, 50 μL of 2 M anhydrous sodium sulfite (0.1 M in sample) (Millipore GR ACS) is added
to the sample to avoid O2 interference (Tian & Nicolas., 1995). The sample is then purged for an
additional minute before the measurement is made. Duplicate measurements were taken for each
sample with a drop size of 6, deposition time of 30 seconds, and 5 seconds of quiet time. Scan
increments were set to 2 mV with scan range set between -140 to -700 mV; the square wave
amplitude and frequency were set respectively to 25 mV and 125 Hz. Iodide concentrations were
determined through the method of standard additions using potassium iodide (KI) (Sigma
Aldrich – ACS grade). Precision was determined to be ±4 nM (standard deviation with 95%
confidence for a sample found to be 243 nM in concentration after 5 replicates) in the laboratory,
9

but this precision was also found to fluctuate to an order of magnitude higher depending on the
wear of the capillary and swell of the ocean when measurements were done shipboard.
Iodate determination was also adapted from Rue et al., (1997) who adapted their method
from Wong & Brewer., (1977). Measurements were made on a spectrophotometer (Perkin Elmer
Lambda 35) using a 10 cm quartz cuvette. 1 mL of 0.12 M sulfanilamide (Sigma Aldrich – ACS
grade) in 1% sulfuric acid (Macron Fine Chemicals) was added to each 25 mL sample to prevent
NO2
-
interference. The sample is left sitting for 5 minutes after which 1 mL I
-
solution (0.12 M
KI in deionized water) is added to the sample to form triiodide (I3
-
). The sample is then measured
for I3
-
at 353 nm within one minute of iodide addition. IO3
-
concentrations were determined by
the method of standard additions using KIO3 (Baker Analyzed – ACS Reagent). Precision, based
on 5 replicates, was ±12 nM for a seawater sample found to be 286 nM.
1.4.0 Results
Figure 2 shows NO2
-
and salinity versus potential density for the entire data set. Nitrite
within the secondary nitrite maximum (SNM) is clustered within a remarkably narrow density
range centered at 26.2 kg/m
3
. This SNM coincides with a maximum in salinity and represents a
distinct water mass referred to in the literature as 13 °C Water (13CW) or Equatorial Sub Surface
Water (ESSW) (Evans et al., 2020). In the ETNP, it is more widely referred to as 13CW (Fiedler
& Talley., 2006; Nürnberg et al., 2015), so we will adopt that convention here. The ODZ ([O2] <
1 µM) spans a much wider range of potential density of 25.5-27.25 kg/m
3
. Therefore, the salinity
feature must play an important role in constraining nitrogen cycle processes to a limited area.
Regardless of mechanism, this feature is also associated with the highest iodine accumulation
(Figure 3), particularly the excess iodine (in this case, the amount exceeding the mean
concentration of 475 nM). Remarkably, the relationship between density and iodine is very
10

similar off of the ETSP (Figure 3) for the ETNP and reflects the importance of the 13CW at both
locations. The importance of this water mass in both systems may reflect the salinity gradient
enhancing the pycnocline at the base of the oxycline (Evans et al., 2020).
Figure 2. A scatterplot of salinity, and nitrite at various isopycnals at all stations measured from both cruises

11

Figure 3. Scatterplot on the left is from of all iodide concentrations (nM) at all isopycnals from both cruises from the ETNP.
Scatterplot on the right consists of data taken from the ETSP.

1.4.1 Sectional data for all parameters on each transect
Figures 4 through 6 show sectional data for O2, NO2
-
, I
-
, IO3
-
and excess iodine (defined
below) for each transect. Each section reveals a large plume of I
-
in the upper ODZ that is
roughly coincident with NO2
-
and associated with a large deficit in IO3
-
. For the two zonal
transects (Figures 4 and 5), there is clearly a much higher concentration near the coast, with I
-

concentrations decreasing offshore and the plume of I
-
becoming much thinner. These transects
also reveal that the most significant depletion of IO3
-
occurs in the upper ODZ closest to the
coast. However, both transects also reveal isolated pockets of high NO2
-
, I
-
and low IO3
-

extending as far as 125 °W. These features are largely centered between the 26-26.5 kg/m
3

potential density anomaly contours shown on the map which is common for the dataset as a
whole shown in Figures 2 and 3.
12

The zonal transect (Figure 6) shows that NO2
-
and I
-
, as well as IO3
-
depletion, do not
extend north of 19 °N. This is evidently the northern boundary of the denitrifying ODZ. Stations
north of this all had detectable O2 with the exception of Station 3 which had <1 µM O2 at one
depth, 200 m (26.4 kg/m
3
isopycnal). There is a strong maximum in I
-
and NO2
-
in the southern
section of this meridional transect as it crosses the two zonal transects in the core of the ODZ.
Excess iodine was calculated by taking the sum of measured I
-
and IO3
-
and subtracting a
mean value of total iodine (475 nM) below hypoxic depths from both cruises. This is slightly
higher than the mean global average of 470 nM determined by Luther et al., (1995). This
calculation is different than the approaches used by Cutter et al., (2018) who used empirical
ratios between IO3
-
to PO4
3-
to determine an expected value. However, we found these
correlations were poor in our data and previously published work resulting in large error
propagation in the excess iodine calculation (see Supporting Information). The differences
between the Cutter approach and our approach yield differences in phosphate replete surface
waters but only minor differences in the ODZ and do not influence our conclusions. Farrenkopf
et al., (2002) used salinity to calculate a conservative IO3
-
value to derive excess iodine. The
salinity range in our study was not sufficient for this correlation to have an impact and resulted in
additional error propagation as with PO4
3-
.
Excess iodine was highest at the western end of the zonal transects and decreased moving
offshore since it is derived from a benthic source. The Revelle zonal transect (Figure 5),
however, revealed a striking maximum in excess iodine well offshore. Presumably, this excess
was derived from the margin and separated from the larger maxima to the east by a water mass
centered at around 120 °W which is probably NEPIW (Evans et al., 2020) and not influenced as
directly by the margin. The Falkor transect (Figure 5), which did not go as far west, failed to
13

sample this feature. This transect, however, also shows a general decline in excess iodine going
offshore but interspersed with local minima at 108 °W and 113 °W. This fine structure was also
observed in the NO2
-
data. Excess iodine was detected throughout the basin which suggests that
the impact of the margin extends throughout the ODZ to at least 110 °W and potentially further
west in isolated pockets. Clearly, physical processes influence the patchiness in iodine and NO2
-

distribution which is addressed in more detail in the discussion section.
Figure 4. Transect profiles of oxygen, iodide, iodate, and excess iodine along the western transect from 105.6-128 °W.

14

15

Figure 5. Transect profiles of oxygen, iodide, iodate, and excess iodine along the western transect from 102
°
W 14 °N to 119 °W
18 °N.

16

17

Figure 6. Transect profiles of oxygen, iodide, iodate, and excess iodine along 110

°W from 22.6 to 14 °N.

18

1.4.2 Excess Iodine
Luther et al., (1995) estimated a mean oceanic value of 470 nM iodine in most of the
world’s oceans; this is generally consistent with observations from individual reports from the
Arabian Sea, ETNP, and ETSP (Cutter et al., 2018; Farrenkopf et al., 2002; Rue et al., 1997). We
determined excess iodine by subtracting the average total iodine concentration (475 nM) from
depths below the ODZ from the total iodine concentration (i.e. the sum of IO3
-
and I
-
). We
flagged several samples where the total iodine (the sum of I
-
and IO3
-
) was below 380 nM. Since
there are no reports of iodine removal in these waters, we assume that there was a problem with
19

the instrumentation, particularly the hanging drop electrode for these samples. The data are
reported in the tables but not used for the figures.
When analyzing excess iodine of the transect from 105.6-128 °W, this investigation
found that excess iodine spanned a broader depth range closer to the shore (Figure 4). This
excess diminishes to a narrower depth range between the 112-120

°W latitudes. Beyond 120

°W,
there is an increase in I
-
and excess iodine between 200-400 m which reflect an external source.
Excess iodine is greatest within a water mass encompassed between potential density anomalies
of 26.0-26.5 kg/m
3
(σθ). This water mass is likely to be the 13CW which has been previously
identified within the ETSP as an accumulation zone for high concentrations of reduced species
(Peters et al., 2018) and high rates of denitrification and anammox. In Figure 5 (transect from
102

°W, 14

°N – 119

°W, 18 °N), this study observed a similar trend; a broader I
-
feature was
observed closer to the coast between 200-600 m which narrowed offshore where it was largely
confined to the 13CW. This transect did not go as far west and no increase in I
-
concentration at
the western end was observed. For the final transect (Figure 6), the highest concentration of
excess iodine is furthest north with pockets of high I
-
scattered throughout this transect.
1.4.3 Representative Stations Within and Outside the Transect
Further details are revealed by examining depth profiles for representative stations within
and outside of the transect. Revelle Station 1 is outside of the ODZ but is close enough to
reducing margin sediments that it has high concentrations of I
-
and excess iodine (Figure 7).
Revelle Station 3 has the highest concentration of I
-
observed in this study in a pronounced
surface maximum (Figure 8, 9). Revelle Station 34 was not on any of the transects but was the
only shelf station with high I
-
throughout the water column (Figure 10). Falkor Station 9 and
Revelle Station 17 were crossover stations with similar distributions of iodine species despite a
20

lengthy interval between sampling. Falkor Station 2 was chosen because two separate CTD casts
were deployed spanning a two-day period with surprisingly different features. Data, for all
stations, are included in the Supporting Information.
1.4.3.1 Revelle Station 1 (110 °W 22.7 °N)
Revelle Station 1 (Figure 7) was north of the ODZ boundary. In the absence of
denitrification, we might expect there to be no I
-
in the subsurface waters. Instead, there was a
well-developed I
-
maximum and IO3
-
minimum (Figure 7), where O2 concentration drops to its
lowest point, which is spatially below the IO3
-
minimum. There appear to be consistently 150-
200 nM concentrations of I
-
and ~400 nM IO3
-
at this station. The exception is at 400 m where
O2 concentration is 1.15 µM. Iodide and IO3
-
concentrations are found to be equal at this depth.
Since a subsurface I
-
maximum remains detectable, even in the presence of O2, oxidation rates
must be slow enough for it to persist during advection from coastal sediments and denitrifying
zones within the OMZ to the south and east.
21

Figure 7. Profiles of iodine species and oxygen at Revelle Station 1 (110

°W 22.7 °N)

22

1.4.3.2 Revelle Station 3 (110
°
W 21.6
°
N)
The highest I
-
/total iodine/excess iodine was observed at Revelle Station 3 (Figure 9). At
the surface, we observed 1188 nM I
-
and 211 nM IO3
-
. Total iodine was 1400 nM which caused
excess iodine to reach 925 nM. Most likely, this high total iodine concentration reflects lateral
inputs from reducing shallow sediments since this station is close to several broad shelf areas.
The relatively low IO3
-
concentration at the surface is unusual for oxygenated water and suggests
that biological reduction by phytoplankton may be occurring. Iodide increases to 1.0 µM at 200
m where the nominal O2 concentration was below or near the nominal detection limit of the CTD
O2 sensor. The most difficult feature to explain is the minimum in I
-
between 100 m and 150 m.
This station managed to capture Pacific Subarctic Upper Water advected via the California
Current from approximately 80
m to 120 m as seen by the
distinctly lower salinity of these
samples (Figure 8). This water
mass originates from the Gulf of
Alaska (Thomson &
Krassovski., 2010) and therefore
does not match the local water
masses.
Figure 8. Plot of Temperature (
o
C) versus
Salinity (PSU) of Revelle Station 3. The
numerical values above each point refers
to depth of sampling in meters. Values between 80-120 m show a distinct minima in salinity and does not match local water
masses.
23

Figure 9. Profiles of chemical species at Revelle Station 3 (110 °W 21.6 °N).

24

1.4.3.3 Revelle Station 34 (105.7

°W, 21.3

°N)
Station 34 was the depth profile taken above the reducing sediment at the coastal shelf
station (Figure 10). From 65-450 m, all O2 concentrations were <1 µM. Iodide and IO3
-
were
measured in the water column from the surface to bottom. At the surface, we measured 147 nM
IO3
-
and 429 nM I
-
. The rest of the water column from 65-450 m (bottom depth = 455 m) had
>700 nM total iodine which was predominantly I
-
; the average IO3
-
concentration below the
surface was only 48 nM. The water overlying the sediment, sampled using a multi-corer, had an
I
-
concentration of >900 nM. There was no detectable IO3
-
found in this bottom water sample.
25

Figure 10. Profiles of chemical species at Revelle Station 34 (105.7 °W, 21.3 °N).

26

1.4.3.4 Falkor Station 9/Revelle Station 17 (110

°W, 14

°N)
Revelle Station 17 was measured as well on the Falkor cruise as Station 9 and was a
crossover station. Profiles from Figure 11 were measured aboard the Revelle on April 2
nd
, 2018.
Profiles from Figure 12 were measured aboard the Falkor on July 6
th
, 2018. Similar profiles for
IO3
-
and I
-
are detected on both cruises. The Iodide maximum occurred at 120 m depth
(corresponding to the 26.0 kg/m
3
isopycnal) aboard the Revelle and at 145 m depth
(corresponding to the 26.2 kg/m
3
isopycnal) aboard the Falkor. A smaller, secondary I
-
maximum
occurred aboard the Revelle at 300 m (corresponding to the 26.6 kg/m
3
isopycnal) and at 250 m
(corresponding to the 26.6 kg/m
3
isopycnal) aboard the Falkor (Figure 13).
27

Figure 11. Profiles of chemical species at Revelle Station 17 (110

°W, 14

°N).

28

Figure 12. Profiles of chemical species at Falkor Station 9 (110

°W, 14

°N).

29

Figure 13. Iodide measurements from 110

°W, 14

°N measured aboard both the Falkor and Revelle plotted against potential
density.

1.4.3.5 Falkor Station 2 (103 °W, 14

°N)
At Falkor Station 2, CTD casts were deployed over multiple days. Cast 1 was taken on
June 30
th
and cast 2 was taken on July 2
nd
, 2018 (Figure 15 for cast 1 and Figure 16 for cast 2).
The depth profiles shown in this paper for this station were deployed and collected two days
apart. While the first cast showed a modest excess iodine within the ODZ, the second cast
showed a pronounced excess with a sharp maximum of 813 nM I
-
at a depth of 365 m (see
Supporting Information); this is an excess of 338 nM I
-
above our average concentration. This
sharp maximum is attributed to excess iodine transported via mesoscale eddy from the shelf,
which propagated westward while we were at this station. Using satellite altimetry, as described
in Evans et al., (2020), the first cast had a sea surface height anomaly of 0.807 m whereas the
second cast had a sea surface height anomaly of 0.815 m which indicates that the core of the
30

anticyclonic eddy was approaching from cast 1 to cast 2. This trend can also be seen in Figures
6a and 6b in Evans et al., (2020). Although Evans et al., (2020) focused primarily on mesoscale
eddies within the 26.2-26.5 kg/m
3
isopycnals, and this excess iodine feature is centered at the
26.8 kg/m
3
, the core of the NEPIW water mass, subsurface eddies could also influence that
isopycnal as well.
Figure 14. Casts 1 and 2 of Falkor station 2, respectively. Excess iodine profiles plotted against potential density.

31

Figure 15. Profiles of chemical species at Falkor Station 2, Cast 1 (103

°W, 14

°N). A second oxygen profile is provided that is on
a smaller scale.

32

Figure 16. Profiles of chemical species at Falkor Station 2, Cast 2 (103

°W, 14 °N). A second oxygen profile is provided that is
on a smaller scale.

33

1.5.0 Discussion
1.5.1 Iodine redox cycling in the absence of Oxygen
The results from this study represent the largest survey of iodine speciation to date in an
ODZ. Previous works have evaluated at details of iodine speciation in other low O2 settings in
marginal marine basins, such as the Chesapeake Bay (Wong & Zhang., 2003), Cariaco Basin
(Wong & Brewer., 1977), Mediterranean Sea (Ullman et al., 1990), Orca Basin (Wong et al.,
1985), Baltic Sea (Truesdale et al., 2013), Black Sea (Luther & Campbell., 1991), and Saanich
Inlet (Emerson et al., 1979). In addition to this study of the ETNP, one previous work has looked
at a single profile from a station nearby our transect (Rue et al., 1997), and studies have looked at
vertical transects through the ETSP (Cutter et al., 2018), Arabian Sea (Farrenkopf et al., 1997,
2002), and Benguela (Chapman., 1983) ODZs. Despite the large spatial, chemical, and physical
differences between each of these settings, they each reveal similar patterns of IO3
-
minima
overlapping with minima in O2, and NO2
-
maxima. These patterns are generally interpreted to
reflect dissimilatory IO3
-
reduction and denitrification under low O2 conditions which have
similar Gibbs Free Energy Yields (Farrenkopf et al., 1997). However, because of important
physical differences between estuarine anoxic basins and marginal upwelling systems, we restrict
our discussion below to comparisons of the ETNP, ETSP, and when applicable to the Arabian
Sea ODZs. Thermodynamically, I
-
is favored in the absence of O2. Previously, Rue et al., (1997)
reported a single profile from a station located within our sampling region (Figure 1) that showed
I
-
and IO3
-
profiles with no excess iodine which was analogous to many stations we measured
offshore. Similar to this previous study, we observe in many, but not all, stations that IO3
-
is
completely absent when O2 is undetectable. This consistent absence of or low concentration of
IO3
-
within the ODZ of the ETNP stand in sharp contrast to observations from the ETSP where
34

IO3
-
concentrations were often near 200 nM throughout the core of the ODZ where O2
concentrations were below detection (Cutter et al., 2018). We also observed many stations that
seemed to display profiles with sharp maxima in I
-
concentrations coinciding with the SNM
(between 26-26.8 kg/m
3
) which was noted previously by Farrenkopf et al., (1997, 2002). Often,
such maxima were associated with a high concentration of excess iodine. In general, excess
iodine is highest within the ODZ and increases towards the margin. This was not observed
previously at the ETNP and is consistent with a margin source for excess iodine as previously
proposed to explain the ubiquity of similar features in both the ETSP and Arabian Sea
(Farrenkopf et al., 2002; Cutter et al., 2018).
The similarities between I
-
and NO2
-
indicate that I
-
accumulation may be linked to
nitrogen cycling in the short term since IO3
-
is a useful terminal electron acceptor in bacterial
respiration. Laboratory studies have shown that IO3
-
is directly reduced by bacteria during the
oxidation of organic matter in low O2 waters (Farrenkopf et al., 1997; Amachi et al., 2007), and
it has been recently proposed that denitrifiers can also utilize IO3
-
during the oxidation of NO2
-
(Babbin et al., 2017). Ultimately, these processes lead to the complete removal of IO3
-
in the
absence of O2.
The most unexpected and potentially important finding is that I
-
and NO2
-
maxima are
associated with the same water mass, the 13CW, in the ETNP and ETSP. This water mass arises
in the western Pacific and spreads eastward (Frenger et al., 2018). It is associated with a local
salinity maximum. This enhances the pycnocline at the base of the oxycline and may contribute
to the persistence of the upper boundary of the ODZ. Below the oxycline, the core of the 13CW
has a very modest density gradient, and this may enhance lateral advection which is driven by
eddies. “Puddies” are active in these density horizons and are probably important in the lateral
35

advection of excess iodine into the basin (Frenger et al., 2018). Currently, there are not enough
data to firmly establish the importance of the 13CW Water Mass in controlling the boundaries of
the ODZ relative to other processes such as primary production in the overlying water. However,
it is important to learn more about how this water mass is formed, and how its properties may
change in the future.
In the Arabian Sea, NO2
-
and I
-
are also centered on the 26.5 kg/m
3
density surface. There
is also a local maximum in salinity immediately above this feature that is associated with the
Persian Gulf Water (PGW) (Acharya & Panigrahi., 2016). It is surprising that the properties of
this water mass are very similar to 13CW despite its very different source. PGW appears to be
enriched in both Fe and iodine that are derived from the continental margin, but it is unclear if
the source is margin sediments off the Indian coast or the Persian Gulf itself.
Iodate is usually reduced to I
-
within the OMZ (where [O2] < 1 µM), however we found
at specific stations (Revelle Stations 5, 21, 24, 25, and 26; Falkor Stations 13, and 18) that even
when O2 is below one micro-molar (the detection limit of the CTD), IO3
-
concentrations ranged
from 50-200 nM. In recent years, I:Ca ratio in carbonates have been a popular paleo-proxy for O2
concentrations (Lu et al., 2010; Hardisty et al., 2014). Since IO3
-
can exist at relatively high
concentrations even in “anoxic” conditions, these results may be useful in calibrating the I:Ca
proxy which takes advantage of surface water planktonic foraminifera I:Ca ratio (Hoogakker et
al., 2018). This proxy assumes that IO3
-
exists only in oxygenated water while the dominant
iodine species is I
-
in oxygen deficient waters (Hoogakker et al., 2018). In anoxic waters, where
all of the iodine should be in the form of I
-
, iodine is not incorporated into the carbonate structure
of planktonic foraminifera (Hardisty et al., 2017). While most of the IO3
-
is reduced to I
-
, at most
stations within the OMZ, many stations had ~15-30 nM of IO3
-
remaining and sometimes higher.
36

Comparable results in IO3
-
concentrations were seen in the O2 deficient waters of the ETSP
(Cutter et al., 2018). These findings suggest that using the I/Ca ratio as a proxy for oxygenation
may need to consider that there is a lag between deoxygenation and IO3
-
disappearance that is
probably on the order of several decades to a century, based on differences in residual IO3
-

between the ETSP and the older waters in the ETNP. Our data provide less insight on the rate of
appearance of IO3
-
associated with the onset of oxygenation because we did no high-resolution
sampling around the boundaries of the ODZ where the 13CW becomes oxygenated. There are no
estimates of the rate of oxidation of I
-
to IO3
-
in seawater in the literature.
1.5.2 Excess iodine
The data reveal excess iodine throughout the ETNP. All three ODZs (ETNP, ETSP, and
Arabian Sea) have a substantial excess iodine signal which indicate that margin to basin transport
is important for all of them. The large region containing excess iodine implications for the
transport of other species as well. Cutter et al., (2018) reported that I
-
and Fe (II) coincided
within the ODZ and were both transported from the margin. Estimates of lateral transport were
obtained using Ra-228 and constrained by upwelling rates calculated from Be-7 to support this
conclusion. It seems likely that Fe transport from the margin to the interior of the ODZ is
important in the Arabian Sea and ETSP as well since Fe (II) has also been measured there.
Excess iodine is a specific tracer of lateral transport from shelf sediments because the
iodine: carbon ratio, associated with remineralization at the sediment water interface, is at least
100x higher than in sinking particles offshore (Kennedy & Elderfield., 1987b; Farrenkopf et al.,
2002). Flux experiments by Kennedy & Elderfield., (1987b) also demonstrate I
-
as the dominant
redox state of iodine transferred from sediments to seawater. Because the small I:C ratio in
sinking particles underlie this important paradigm of excess iodine, we looked at the I:C ratio in
37

suspended particles more closely since the estimate of 10
-4
used by Farrenkopf et al., (2002)
came from a single noisy correlation of IO3
-
versus PO4
3-
in the South Atlantic (Wong et al.,
1976). Unfortunately, there are no measurements of both POC and iodine in marine particles.
Wong et al., (1976) measured particulate iodine in the South Atlantic and reported values
ranging from 2-127 ng/kg of seawater. Using estimates of POC from a compilation by Martiny et
al., (2014) in this region, we found that the highest possible value of I:C is about 10
-5
. Recently,
I
-
uptake has been studied in phytoplankton (de la Cuesta & Manley., 2009; Bergeijk et al.,
2016). From these data, estimates of I:C range from 10
-8
to 10
-6
. Thus, the estimate of 10
-4
used
by Farrenkopf et al., (2002) is quite conservative, in this context, and their conclusions are
strongly supported by these other studies.
A pore water I
-
source most prominent in reducing waters is also supported by a number
of studies evaluating the speciation of pore water iodine, diffusive fluxes of iodine across the
sediment-water interface, and comparisons between I:C in oxic versus reducing sediments.
Specifically, studies evaluating pore water iodine speciation reveal that it is dominated by I
-
due
both to reducing conditions and the input of I
-
from organic matter during sediment diagenesis
(Kennedy & Elderfield., 1987a, b; Anschutz et al., 2000). Similar to ammonia, I
-
increases in
pore waters during diagenesis to concentrations in excess of that of seawater by several fold
which often reach several µM or higher. This gradient results in net transport of I
-
to overlying
bottom waters which has been documented in a range of studies (Kennedy & Elderfield., 1987a,
b; Anschutz et al., 2000; Ullman & Aller., 1985). To date, no study has directly compared these
fluxes between pore waters to oxic versus anoxic bottom waters, but evidence from sediments is
consist with the likelihood of increased iodine retention at the sediment water interface in oxic
versus anoxic settings. Specifically, I:C ratios at the sediment water interface of anoxic
38

sediments mimic those discussed above for phytoplankton and POC while sediments directly
underlying oxic bottom waters have much higher I:C ratios on the order of 10
-3
(Price & Calvert.,
1973; Price & Calvert., 1977; Kennedy & Elderfield., 1987a,b; Lu et al., 2008). The higher I:C
ratios under oxidizing conditions appears to reflect a combination of sorption to Mn and Fe-
oxides (Fox et al., 2009; Ullman & Aller., 1985) as well as the oxidation of I
-
to HOI and
subsequent reaction with sedimentary organic matter (Francois et al., 1987). The diminished
contribution of scavenging of iodine at the sediment water interface is anticipated to directly
increase the diffusive flux of I
-
to seawater in anoxic relative to oxic bottom waters.
1.5.3 Effects of local physics – a water mass analysis
Evans et al., (2020) carried out a water mass analysis of the ETNP based on the sections
from these cruises. This work shows that NO2
-
, I
-
and excess iodine are primarily associated with
a single water mass, the 13CW. Peters et al., (2018) showed that the upper ODZ in the ETSP
with active denitrification was mostly contained within this water mass. It is not surprising that
this water mass would be associated with NO2
-
and I
-
in our study. Evans et al., (2020) showed
that most of the small-scale structure in the sections reported in Figures 5 through 7 can be
accounted for by the contribution of the 13CW which show spatial and temporal variabilities.
For example, in Figure 5, the “hot spot” of I
-
is associated with a high percentage of the 13CW
water mass relative to Northeast Pacific Intermediate Water (NEPIW), the other main water mass
along this density horizon. NEPIW brings water from outside the ODZ into this density horizon.
It is low in I
-
and NO2
-
but may still have O2 levels below the detection limit of the CTD sensor.
The low I
-
and NO2
-
at 120

°W on this transect is associated with a higher percentage of NEPIW
relative to 13CW (Evans et al., 2020). The temporal variability in iodine distribution, revealed in
Figures 13 and 14, can also be explained largely by variability in the relative importance of these
39

two water masses. Indeed, eddies generated by poleward undercurrents flowing along the
continental margin transport ESSW westward (Evans et al., 2020). Evans et al., (2020)
demonstrated that sea surface height anomalies, during these cruises, indicate the presence of
mesoscale eddies. The water mass analysis can also explain why such eddies are enriched in I
-
. A
cruise in 2012 included a station on the shelf at 18.8 °N, -104.6 °W, east of our transects. The
13CW shoaled over the shelf centered at 150 m which enabled I
-
to be mobilized from the shelf
sediments and enriched in this water mass (Evans et al., 2020). 13CW may act as a conduit in the
shelf-to-basin shuttle for reduced species from shelf sediments to the deep interior of the ETNP.
1.6.0 Conclusion
The distribution and speciation of iodine is controlled by a combination of physical,
chemical and biological processes. Iodate reduction within the ODZ leads to I
-
becoming the
predominant species with maxima at the same potential density anomaly (26.5 kg/m
3
) in both the
ETSP and ETNP. There are also substantial inputs of excess iodine from the continental margin
which extend throughout the basin, and it is significant that the 26.5 kg/m
3
potential density
anomaly extends over the shelf benthic boundary layer in both hemispheres. These basic features
are modified by physical processes that influence the relative contributions of specific water
masses within the ODZ. Eddies generated off the continental margin are probably driving these
processes and contribute to the fine structure and variability that we observe. Iodine speciation is
clearly linked to the distribution of O2, but the similar distributions of iodine and NO2
-
indicate
that it is linked to the nitrogen cycle as well, perhaps because of the use of IO3
-
by denitrifiers as
a terminal electron acceptor.
40

1.7.0 Acknowledgments
Nitrite measurements were made possible by both Dr. Rick Keil, Dr. Allen Devol, Dr.
Andrew Babbin and their respective labs. I:C ratio calculations were made possible by the data
provided to us by Dr. Dan Ohnemus. We would also like to acknowledge the cooperation from
the captains, crew members, and marine technicians aboard the R/V Roger Revelle and the R/V
Falkor for operating the equipment to make this study possible. Funding for this project came
from the National Science Foundation OCE1636332. A Schmidt Ocean Institute grant to Dr.
Karen Casciotti and Dr. Andrew Babbin. supported the FK180624 cruise. Additional support
provided by the MIT Ally of Nature and Heflinger Funds to Dr. Babbin. Dalton Hardisty’s
participation in and contribution to the project is funded through NSF OCE1829406. The data
from this paper is being submitted to the Biological and Chemical Oceanography – Data
Management Office (BCO-DMO).
41

Chapter 2
Determination of Inert and Labile Copper on GEOTRACES Samples Using a
Novel Solvent Extraction Method
Authors: Rintaro Moriyasu, and James W. Moffett
Originally published in: Journal of Marine Chemistry, 239
Published on: February 20
th
, 2022
2.0.0 Abstract
Copper, in seawater, is predominantly bound by organic ligands of unknown
composition. Complexation has been thermodynamically characterized using synthetic ligand
competition experiments which assumes equilibrium amongst all chelators within the system.
However, equilibration times are often constrained by wall loss issues with the synthetic ligands.
Here, a solvent extraction methodology, was utilized to avoid the wall-loss problems. Using an
exceptionally high concentration of a strong copper chelator, oxine (8-hydroxyquinoline), at least
six hours of equilibration are required to reach steady state between the competing ligand and the
labile copper in seawater. This is much longer than equilibration times used in previously
published works. This method was optimized by using samples from GEOTRACES expeditions
in the North Pacific and North Atlantic Oceans. Surprisingly, 60-90% of the copper was not
exchangeable with oxine under these conditions. We define this fraction as “inert”, and these
data, which include profiles as deep as 1000 m in the North Pacific, suggest that this is a
widespread feature. Our results suggest that there are two distinct pools of labile and inert
copper, rather than an assemblage of similar complexes with incremental differences in stability
constants. The results have important implications for the marine geochemistry of copper and its
bioavailability. Complexation has been shown to limit bioavailability and influences scavenging
and copper residence time. Moreover, a basic paradigm of copper speciation methodologies, that
even strong Cu complexes are relatively labile, is likely incorrect.
42

2.1.0 Introduction
Within aquatic regimes, dissolved copper (dCu) is predominantly bound by organic
ligands; indeed, among all of the first-row transition metals, copper (Cu) is known to form the
strongest, most stable complexes due to its small ionic radius and its propensity to undergo Jahn-
Teller distortion (Irving & Williams., 1953). It has been estimated that ≥99% of Cu is bound by
ligands (van den Berg et al., 1987; Coale & Bruland., 1988; Moffett & Dupont., 2007; Bundy et
al., 2013). This is significant because both Cu bioavailability and toxicity have been found to be
functions of free cupric ion concentration ([Cu
2+
]) rather than of total dCu concentration (Coale
& Bruland., 1988; Sunda & Lewis., 1978; Sanders et al., 2015), and Cu toxicity can occur at
[Cu
2+
] as low as 10
-12
M for cyanobacteria (Brand et al., 1986).
Marine water column Cu profiles are generally depleted at the surface and increase linearly
with depth; this is thought to be a result of particulate reversible scavenging as its main sink, mid-
depth remineralization, and remobilization via benthic flux (Boyle et al., 1977). The linearity of
the depth profile is primarily attributed to reversible scavenging (Little et al., 2013) where the
slope of the depth profile is a function of scavenging rate (Richon & Tagliabue., 2019). Copper
ligands are also thought to play a major role in controlling the rate of Cu scavenging by particulate
matter (Davis & Leckie., 1978; Laglera & van den Berg., 2003). Free Cu
2+
, rather than Cu species
that are strongly bound by organic ligands, will be scavenged more easily; organic-bound Cu, on
the other hand, will remain dissolved throughout the water column (van den Berg et al., 1987).
Even among organic ligands that bind Cu, there is a range in binding strength. In many studies, at
least two classes of ligands are identified; L1 ligands have stronger binding strengths and have
concentrations comparable to ambient Cu (Moffett et al., 1990; Whitby et al., 2018), while L2
ligands have weaker binding strengths and are at significantly higher concentrations than L1
43

(Bundy et al., 2013). The strength of these ligands have been found to be on a spectrum where
ligand types are divided by their conditional stability constants (𝐾 ′
𝑐𝑜𝑛𝑑 ) of approximately 10
13
-
10
16
for L1-ligands, and 10
10
-10
13
for L2-ligands (Whitby et al., 2018), but these values can vary
depending on the window of detection (α) which is a sum of the conditional stability constants
(𝐾 ′
1
𝑐𝑜𝑛𝑑 and 𝛽 ′
2
𝑐𝑜𝑛𝑑 for the mono and bis-complex, respectively) of the artificial ligand multiplied
by the concentrations of artificially added competing ligand (Equation 1), used for one particular
electrochemical method known as the Competitive Ligand Exchange – Adsorptive Cathodic
Stripping Voltammetry (CLE-AdCSV). Marine Synechoccocus sp. has been found to produce
chelators in the presence of high Cu concentrations; these ligands were observed to have similar
𝐾 ′𝑐𝑜𝑛𝑑 to L1-type ligands in both seawater samples (Moffett et al., 1990) and cultures (Moffett &
Brand., 1996). Organic ligands that are not produced by microbial processes, often referred to as
humic substances, can also chelate Cu in the open ocean (Whitby & van den Berg., 2015; Whitby
et al., 2018).
It has been proposed that some strongly binding Cu compounds (log K’ ≥ 13) are
kinetically non-labile (Kogut & Voelker., 2003). Kogut & Voelker., (2003) define “inert Cu” as
Cu species that are non-exchangeable between the various Cu pools on a timescale of up to 48
hours. The analytical window, based off of the strength and concentration of the competing ligand
used for the electrochemical method, constrains the kinetic binding constant of the strongly
binding ligand (KCuL) as being greater than 10
17.3
; these workers constrain this value based on the
α of the added ligand, salicylaldoxime (SA), at 1 mM concentration (log(αSA) = 8.5). Kogut &
Voelker., (2003) attribute the formation of this fraction to the complexation of dCu by colloids.
They were able to eliminate this fraction and other organic ligands by UV-irradiating their samples
prior to measurement. Ruacho et al., (2020) also assume the existence of inert Cu in their work.
44

In Waquoit Bay, MA, Moffett et al., (1997) reported Cu so tightly bound that even 500 µM
of benzoylacetone was unable to detect any signal on the CLE-AdCSV; this further illustrates the
issues many workers have with this method’s ability to resolve organic-bound Cu. Leeuwen &
Town., (2005) also demonstrate the kinetic limitations of the CLE-AdCSV method. These workers
show that there is a disequilibrium between the added ligands and strongly bound complexes
originating from the slow rates of dissociation which leads to an overestimation in the conditional
stability constants of metal-ligand complexes. Leeuwen & Town., (2005) claim that no definitive
conclusions can be made about this fraction that is undetectable by electrochemical means and
raise questions about the validity of conditional stability constants derived by CLE-AdCSV.
Another evidence of nonexchangeable Cu was reported by Ma et al., (2006); these workers used
Donnan dialysis and a radioisotope,
64
Cu (t1/2 = 12 hours), to isotopically label the exchangeable
pool of Cu between solid and solution phases in water extracted from soil samples. Ma et al.,
(2006) also observed an overestimation of labile Cu in their isotope dilution calculation because
not all of the organically bound Cu was exchangeable with their radiotracer, and this non-
nonexchangeable fraction is thought to be colloidal. This fraction likely affects local cycling of Cu
by preventing particulate scavenging which could have an effect on the global Cu cycling.
In this work, we developed a novel method for measuring the concentration of kinetically
inert Cu on the Inductively Coupled Plasma – Mass Spectrometer. This method requires the use
of solvent extraction, which was utilized previously as an alternative to CLE-AdCSV (Moffett &
Zika., 1987a), while adding a much stronger Cu ligand, 8-hydroxyquinoline, at higher
concentrations. This will result in a higher detection window, allowing for maximum extraction of
the labile Cu fraction and creating a physical separation from the inert fraction. Moreover, it
enables much longer equilibration times (at least two weeks) by eliminating the problem of wall
45

loss that constrains voltammetric methods. The method has been applied to samples from a
geographically diverse range of open ocean environments.
2.2.0 Sampling
Samples used for the initial development of the method were obtained from
GEOTRACES GA03 aboard the R/V Knorr (KN204-01), shown on the map in Figure 1. The
ship was equipped with the GEOTRACES CTD/rosette (Model 32G, Sea-Bird Electronics). This
standard GEOTRACES rosette contains 24 Teflon-coated 12 L GO-FLO bottles (Model 10812
T, General Oceanics) with hydrographic sensors for dissolved oxygen, algal pigments,
conductivity, temperature, pressure, and beam transmittance. These samples were preserved by
filtering through 0.2 µm Acropak
TM
200 Supor capsule filter (Pall Corporation) and frozen at -20
°C. Locations, where seawater was sampled during KN204-01, are shown in Figure 1. Samples
used for assessing the method through the DTPA (Diethylenetriamine pentaacetate) titration
experiment were obtained from Wrigley Institute’s Flow-through seawater system on Santa
Catalina Island (33.26 °N, 118.29 °W), shown on map in Figure 2. These samples were filtered,
using 0.2 µm Acropak
TM
200 Supor capsule filters, sent back to the lab, and measured within 1
week of collecting.
46

Figure 1. Map of samples taken from GEOTRACES GA03. Samples taken aboard the R/V Knorr on cruise (KN-204-01) between
November 6
th
to December 11
th
, 2011.

Samples used for application of the method were taken during the first half of
GEOTRACES GP15 aboard the R/V Roger Revelle (RR1814) and were sampled at stations 4
(54.7 °N, 155.1 °W) and 9 (44.5 °N, 152.0 °W), shown on Figure 2. Station 9 samples ranged
from GEOTRACES Sample # 13048-13070 (ranging between 25-1000 m depths), taken on
October 7
th
, 2018, and consisted of a 12-point depth profile. These samples were also collected
using the standard GEOTRACES CTD/rosette, and total and labile Cu were measured. Inert Cu
was calculated from all samples taken aboard RR1814. Samples were also sent back filtered
through a 0.2 µm Acropak
TM
200 Supor capsule filters and frozen, to prevent wall-loss and
47

changes in speciation, from Hilo, HI. Freezing samples at -20

°C have been previously shown to
preserve the speciation of Cu in samples (Jacquot & Moffett., 2015).
Figure 2. Map of GP15 Stations 4 (54.7 °N, 155.1 °W) and 9 (44.5 °N, 152.0 °W), and Santa Catalina Island (33.2 °N, 118.0
°W).

2.3.0 Materials and Reagents
2.3.1 Materials
All seawater samples were frozen and stored within 1 L Fluorinated Low-Density
Polyethylene (Nalgene; FLPE) bottles. Sample bottles were acid-washed prior to RR1814 cruise
following standard trace element and metal acid washing procedure.
The reaction vessels used for solvent extraction were 15 mL Standard Teflon Vials with
rounded interiors (Savillex: Catalog #200-015-20) with 33 mm PFA Closures (Savillex: Catalog
#600-033-01) and 7 mL Standard Teflon Vials with flat interiors (Savillex: Catalog #200-007-
10) with 24 mm PFA Closures (Savillex: Catalog #600-024-01).
48

2.3.2 Reagents
Toluene (Sigma Aldrich; CAS #: 108-88-3), used for solvent extraction, was distilled
prior to usage using a Teflon still. A 0.1 molar (M) solution of 8-hydroxyquinoline (Sigma
Aldrich; oxine; CAS# 148-24-3; ACS grade) was prepared in Optima
TM
grade methanol (Sigma
Aldrich; CAS# 67-56-1). A 5% solution of Optima
TM
grade nitric acid (HNO3; Thermo Fisher;
CAS# 7697-37-2) in Milli-Q
®
water (Millipore; 18.2 MΩ/cm) was used for the back-extraction.
Concentrated Optima
TM
grade hydrogen peroxide (Fisher; H2O2; CAS#: 7722-84-1) was used to
treat for organic interference when measuring the concentration of total Cu on GEOTRACES
GP15 samples (Baconnais et al., 2019). For the titration experiment with diethylenetriamine
pentaacetate (DTPA) (Sigma Aldrich; CAS# 67-43-6), a 0.1 M solution of the added ligand was
made in Milli-Q
®
water with pH adjusted to 10 using Optima
TM
grade Ammonium Hydroxide
(Fisher; CAS# 1336-21-6).
2.4 Methods
2.4.1 Solvent Extraction Methodology
Frozen samples were defrosted overnight in a water bath. Once defrosted, 15 mL of
sample was used to pre-treat the Teflon vials overnight. A 5 mL aliquot of seawater sample is
measured into each vial, and 500 µM of oxine was pipetted into each aliquot. A 4 mL aliquot of
toluene was added to each vial and allowed to equilibrate on the shaker for 6 hours. Once
equilibrated, 3 mL of added toluene was pipetted into three separate 7 mL Teflon flasks (1 mL
per vial for triplicates) and taken to dryness on a hotplate at 160 °C. 1 mL of 5% Optima grade
HNO3 was added to each Teflon vial; the vials were left on a shaker overnight to allow for back-
extraction of the labile Cu fraction into aqueous layer. The acid used for back-extraction is then
measured using isotope dilution on the Thermo Element 2
TM
Inductively Coupled Plasma – Mass
49

Spectrometer (ID-ICP-MS) on medium resolution. Due to low concentrations of the labile
fraction, ESI Apex Q desolvating nebulizer (Thermo Fisher, Parts #: 1299560) was used during
the measurements. Samples were measured within 1 week of defrosting to retain speciation.
Thorough cleaning of Teflon vials, both 7 and 15 mL was essential to avoid significant
memory effects. Prior to use, all vials were washed in a 10% HCl solution bath for at least 1-
week. This was followed by filling the vials with 3 M Optima grade HNO3 and leaving them
closed on a hotplate at 100 °C overnight. Once removed from the hotplate, the vials were rinsed
with high-purity Milli-Q water and filled with 1% Optima grade HNO3 (Fisher) for storage.
Teflon vials and closures were used because they are impermeable to toluene.
Pre-treatment of the Teflon vials used in the equilibration and solvent extraction steps (15
mL vial) is crucial since the acid washing process can leave a pH memory effect on the vials.
Un-neutralized acid can cause uneven extraction of labile Cu among replicates, but by leaving 15
mL of the seawater sample in the vials overnight, this can be avoided. The sample used to pre-
treat is disposed of and replaced with fresh sample for analysis.
The 1 L FLPE bottles used to store, and freeze samples were thoroughly cleaned using an
acid washing protocol. This consisted of a 10% Citranox (Alconox) solution bath for 1 day,
followed by a 10% hydrochloric acid (HCl; Fisher; CAS# 7647-01-0) solution bath for at least 1-
week. The bottles were then filled with 1% Optima grade HCl solution and left to soak for 2
days. Between each step of the process, the bottles were thoroughly rinsed out using Milli-Q
water. These were then stored and shipped empty inside of 2 re-usable, re-sealable zipper bags.
The toluene used for the solvent extraction method was distilled prior to usage which
lowered the Cu blanks to below detection. A simple 375 mL Teflon Sub-boiling Still Assembly
(Savillex; Catalog# 520-1-1-2) was used for this process along with standard heating tape as the
50

heat source on one end. A Styrofoam container filled with ice was placed on the other end of the
distillation apparatus to cool down the vaporized solvent.
2.4.2 Total Copper Measurements
Total Cu measurements for the GEOTRACES GA03 samples were done previously by
Jacquot & Moffett., (2015) using a single-batch nitrilotriacetatic acid (NTA) resin extraction
which was followed up by ID-ICP-MS; the procedure is described there. These values were
compared with [Cu]Total measurements using the solvent extraction method developed in this
article which used a mercury lamp to UV-irradiate the seawater sample for 2 hours (Jelight
Company, Inc.; UVO-Cleaner Model No. 342) in a Teflon Vial with a quartz window. Once
oxidized, samples were left overnight prior to extraction. This treatment was assumed to destroy
any organic matter binding both the inert and labile Cu fractions, enabling for the yield of our
protocol in a system where 100% of the Cu should be complexed by oxine.
Total Cu measurements for station 9 of GEOTRACES GP15 utilized a method adapted
from Rapp et al., (2017). Samples were acidified to a pH of 1.7 and pre-concentrated using the
SC-DX seaFAST (Elemental Scientific; M-SFS2-MG-52) automated pre-concentration solid
phase extraction columns and measured on the Element 2 ID-ICP-MS on medium resolution.
Measurements were made in triplicates.
2.4.3 Equilibration Time Experiment
The equilibration time for the solvent exchange method was varied to see if the amount
of time given for ligand-solvent equilibration would change the concentration of extractable,
labile Cu. This experiment was also done to ensure that all Cu that is reactive to the added ligand
would be extracted. The range of equilibration times used ranged from 1-336 hours
(approximately two weeks). When leaving samples on the shaker during equilibration, all
51

samples were wrapped within a black, opaque bag to prevent photo-oxidation/degradation of Cu-
binding ligands.
2.5.0 Competitive Ligand Exchange
2.5.1 Theory
The theory behind the competitive ligand exchange/solvent extraction methodology for
Cu was developed by Moffett & Zika., (1987a) who coupled extraction with analysis on the
Graphite Furnace Atomic Absorbance Spectroscopy (GFAAS). That method used acetylacetone,
while the current method uses the much stronger chelator, 8-hydroxyquinoline, commonly
referred to as oxine (ox). The theory is analogous to competitive ligand approaches using other
analytical methodologies, such as voltammetry. Modifying the equation developed by Campos &
van den Berg., (1994), we have equation (1):
𝛼 𝐶𝑢 (𝑜𝑥 )𝑥 =
[𝐶𝑢 (𝑜𝑥 )
𝑥 ]
[𝐶𝑢
2+
]
= 𝛽 2
𝑐𝑜𝑛𝑑 [𝑜𝑥 ]
2
+ 𝐾 1
𝑐𝑜𝑛𝑑 [𝑜𝑥 ] (1)
where αCu(ox)x is the partitioning coefficient between the ligand-reactive “labile Cu”
([Cu(ox)x]), and cupric ions ([Cu
2+
]); K1
cond
and β2
cond
are the conditional stability constants for
the mono- and bis- oxine complexes with Cu. The partitioning coefficient is also referred to as
the analytical window of the ligand, often used to quantify the binding strength of the added
ligand system. Labile Cu is defined as:
[𝐶𝑢 ]
𝑙𝑎𝑏𝑖𝑙𝑒 = [𝐶𝑢 (𝑜𝑥 )
𝑥 ] = [𝐶𝑢𝐿 ]
𝑤𝑒𝑎𝑘 + [𝐶𝑢
2+
] + [𝐶 𝑢 ′
] (2)
where:
[𝐶𝑢 (𝑜𝑥 )
𝑥 ] = [𝐶𝑢 (𝑜𝑥 )
2
] + [𝐶𝑢 (𝑜𝑥 )
+
] (3)
The labile fraction is the sum of weakly-bound organic Cu complexes ([CuL]weak), free cupric
ions, and inorganic Cu species ([Cu’]). This equation can be simplified because, at higher added
52

ligand concentrations, the mono-complex is insignificant (Buck & Bruland., 2005; Bundy et al.,
2013). Since we are using 500 µM oxine, equation (1) can be simplified down to:
𝛼 𝐶𝑢 (𝑜𝑥 )2
=
[𝐶𝑢 (𝑜𝑥 )
2
]
[𝐶𝑢
2+
]
= 𝛽 2
𝑐𝑜𝑛𝑑 [𝑜𝑥 ]
2
(4)
and [𝐶𝑢 (𝑜𝑥 )
𝑥 ] = [𝐶𝑢 (𝑜𝑥 )
2
] (5)
since we are forming mostly the bis-complex with negligible concentrations of the non-
extractable, charged mono-complex. Total dissolved Cu ([Cu]T) in our system is then:
[𝐶𝑢 ]
𝑇 = [𝐶𝑢𝐿 ]
𝑠𝑡𝑟𝑜𝑛𝑔 + [𝐶𝑢 ]
𝑙𝑎𝑏𝑖𝑙𝑒 (6)
where [CuL]strong is the fraction of the Cu pool bound by strongly binding organic ligands. This
fraction is so tightly bound that it will not exchange with the added ligands, even with a detection
window of 9.52 (log(β2
cond
) of 16.12 at 500µM of oxine) which is a much higher value than those
used by previous studies by sometimes orders of magnitudes (Table 1), resulting in:
[𝐶𝑢 ]
𝑖𝑛𝑒𝑟𝑡 = [𝐶𝑢𝐿 ]
𝑠𝑡 𝑟𝑜𝑛𝑔 (7)
[𝐶𝑢 ]
𝑙𝑎𝑏𝑖𝑙𝑒 = [𝐶𝑢 (𝑜𝑥 )
2
] (8)
and [𝐶𝑢 ]
𝑖𝑛𝑒𝑟𝑡 = [𝐶𝑢 ]
𝑇 − [𝐶𝑢 ]
𝑙𝑎𝑏𝑖𝑙𝑒 (9).
The method developed in this study directly measures [Cu]labile while [Cu]T is measured
using standard seaFAST automated pre-concentration columns followed by measurement on the
ICP-MS. From the difference, an operationally defined “inert” Cu fraction (bound by organic
ligands with α ≥ 9.52) is be calculated.
53

Table 1. List of Chelators, concentrations, and detection windows (Log(αAL)) used here and previous studies.

Chelator used

[Chelator] (µM) Log(αAL)

Campos & van den Berg,
(1994)

8-hydroxyquinoline (oxine)

1

5.02

Whitby et al., (2018)

Salicylaldoxime (SA)

10

5.1

Bundy et al., (2013)

SA

25

5.8

Kogut & Voelker, (2003)

SA

500; 1000

7.9; 8.5

This Study

oxine

500

9.52

2.5.2 DTPA Titration Experiment
The oxine-toluene system was calibrated against a known Cu chelator, diethylenetriamine
pentaacetate (DTPA). This enabled us to calculate the empirical side reaction coefficient
(αCu(ox)2) (Equation 1) for the method to compare with other methods and to set lower limits on
how strongly bound inert Cu might be. Reliable stability constant data are available for DTPA
Cu complexes, compiled in Twiss & Moffett., (2002). A seawater sample from Santa Catalina
Island was UV-irradiated for 2 hours to remove all pre-existing organic chelators of Cu; these
samples were then equilibrated for 3 hours with varying concentrations of DTPA (0-2.46mM)
before equilibration with oxine. Solvent Extraction was performed as above. These results are
shown in Figure 5.
2.5.3 Choice of Ligand
Oxine was chosen as the competing ligand primarily for its high affinity with Cu. It was
previously used by Campos & van den Berg., (1994) as the competing ligand to achieve the
54

highest range of detection windows (6.4-9.0). Comparatively, previous studies that have used
solvent extraction have used salicylaldoxime (Dahl., 1968), and acetylacetone (Moffett & Zika.,
1987a; Moffett et al., 1990). These competing ligands, however, are more useful for establishing
an equilibrium between artificial and natural ligands which enable estimates of natural ligand
binding strengths to be calculated; such ligands work best if they have a moderate binding
strength that are comparable to natural ligands (Donat & van den Berg., 1992; Campos & van
den Berg., 1994). In comparison, our objective is to overwhelm the naturally occurring ligands
so that all forms of Cu that are kinetically labile will be transferred to the oxine-complexed pool.
This was achieved by using high concentrations of a strongly binding Cu ligand; this method
reaches the highest detection window, αAL (Table 2), allowing for the extraction of most, if not
all, chemically labile Cu species, [Cu]labile.
2.5.4 Choice of Solvent
Toluene (methylbenzene) and carbon tetrachloride were found to be good solvents for the
solvent extraction method by Moffett & Zika., (1987a). Toluene was chosen over carbon
tetrachloride in previous studies, such as Moffett et al., (1990) since it is much less toxic to
handle. Both solvents, however, were found to not affect the electrostatic properties of seawater
(Moffett & Zika., 1987a).
2.6.0 Results
2.6.1 Equilibration Time Experiment
The time taken between the addition of ligand-solvent system and extraction was varied
for GEOTRACES sample # 12576 ([Cu]total=1.85 nM; 50 m depth at station 4 of GP15) and can
be found in Figure 3. An equilibration time of ≤ 3 hours resulted in <10% of [Cu]total extracted.
When equilibration time becomes ≥6 hours, ≥25% of [Cu]total extracted which we expect to be
55

the maximum extractable/labile Cu. This value does not change for over two weeks which shows
that all extractable copper species with 500 µM oxine were extracted by 6 hours. In addition, this
experiment shows that there is no significant wall-loss from leaving samples for 2 weeks in the
Teflon vials with the solvent.
56

Figure 3. Sample 12576 (50 m, Station 4 of GP15) had the equilibration time between addition of ligand-solvent system and
solvent removal varied. This equilibration time ranged between 1-336 hours (corresponding to 2 weeks).

2.6.2 Oxine Concentration Dependence
The oxine concentration for solvent extraction was determined using samples from
GEOTRACES GA03 (GEOTRACES # 7564 at 235.8 m and #7709 at 187.1 m) and varying
oxine concentration to find the optimal concentration for oxine addition to maximize extraction
(Figure 4). Results of all GA03 samples can be found in Table 2. A concentration of 500 µM of
oxine was found to maximize the extraction of [Cu]labile while less oxine resulted in ≤ [Cu]labile.
There is a general decrease in extractable Cu for additions of oxine >500 µM; this is thought to
be caused by the formation of a charged, copper tris-oxine complex which does not partition into
the organic solvent layer.
57

Figure 4. Copper extracted versus Oxine concentration for GEOTRACES Samples 7564 and 7709.

2.6.3 DTPA Titration Experiment
The detection window used in this method was determined by titrating the oxine with
diethylenetriamine pentaacetate (DTPA) added to UV-irradiated seawater from Santa Catalina
Island ([Cu]Total = 3.15±0.06 nM). This conditional stability constant of this Cu ligand in
seawater was previously calculated by Twiss & Moffett., (2002); these workers found that
log (𝐾 𝐷𝑇𝑃𝐴 𝑐𝑜𝑛𝑑 ) is 12.42. By UV-oxidizing the seawater samples, the naturally existing Cu ligands
are removed (Donat & van den Berg., 1992; Campos & van den Berg., 1994; Bundy et al., 2013;
Whitby et al., 2018). Prior to this experiment, samples from GEOTRACES GA03 were used as
standards to show that UV-irradiation of seawater can indeed destroy all pre-existing ligands and
can achieve quantitative extraction of total copper (Table 2). [Cu]T measured using this method
(UV-ox Cu) was compared with those measured by Jacquot & Moffett., (2015) who used a
nitrilotriacetic acid (NTA) resin extraction with an isotope spike on the ICP-MS. Varying
concentrations of DTPA (from 0-2.5 mM) were added to seawater solutions free of organic-
bound Cu. Copper bound by DTPA forms a charged complex which does not partition into the
solvent phase.
58

The results (Figure 5) showed that in the absence of any strong, organic Cu ligand, all Cu
is extractable, and in the presence of increasing concentrations of these ligands, the concentration
of labile Cu (extractable Cu) decreases until below detection. Using the results from the DTPA
experiment (where [DTPA] = 0.98 mM), we calculated the log(𝛽 ′
2
𝐶𝑜𝑛𝑑 .
) of oxine with the known
log(𝐾 ′
1
𝐶𝑜𝑛𝑑 .
) of DTPA (Twiss & Moffett, 2002) and concentrations of extractable and total Cu.
The log(αCu(ox)2) was then calculated to be 9.52 (Table 2).
Figure 5. Results from the Diethylenetriamine Pentaacetate (DTPA) titration experiment. Samples were titrated with 0-2.5 mM
DTPA against 500 µM 8-hydroxyquinoline (oxine). Top figure is actual concentration of Cu extracted (nM) from Santa Catalina
Island seawater sample against the amount of DTPA added (M). Bottom figure is percent Cu extracted (of total 3.15±0.06 nM)
against added DTPA.

59

Table 2. Frozen samples taken from GEOTRACES GA03 (NAZT) measured using solvent extraction
methodology with 500 µM 8-hydroxyquinoline. Samples were analyzed with and without a 2-hour UV-
irradiation/oxidation step to destroy all naturally existing chelators.

1
Refers to GEOTRACES Sample ID number
2
Samples were UV-irradiated under a mercury lamp for 2 hours and extracted with 500µM oxine.
3
CuT obtained from BCO-DMO – measured by Jacquot & Moffett., (2014) using nitrilotriacetic acid (NTA) extraction.
4
A % Recovered when comparing totals measured via UV-oxidation solvent-extraction and on the totals previously via NTA
extraction.

2.6.4 Extraction of Copper without the Added Ligand
A control experiment, where solvent extraction of Cu was measured without using any
added copper ligands, was done previously by Moffett & Zika., (1987a). These workers found
that no measurable Cu (less than 1%), measurable on the GFAAS, partitioned into the solvent
phase from the aqueous when extraction was done without acetylacetone. This experiment was
repeated under the solvent extraction protocol proposed by this study. The sample used was
GEOTRACES # 7709 (at 187.1 m depth) from the GA03 cruise. All replicates (n=9) yielded
extractable Cu below detection and were unquantifiable (data not included). Without an added
ligand, neither labile nor inert Cu will extract into the solvent phase. This experiment was
replicated (n=3) for GEOTRACES #12576 (depth of 50 m and used for the equilibration time
variance experiment) for 48 hours; similar to equilibrating without oxine for 6 hours on #7709,
the measured [Cu]labile was below detection.
G#
1
[Cu]labile
(nM)

SD Depth
(m)
UV-ox.
Cu
(nM)
2

SD [Cu]Total
(nM) –
Jacquot
3

SD %
Labil
e
%
Inert
%
Recovered
4

7564

0.56

0.06

235.8

1.593

0.002

1.65

0.03

34.9

65.1

97

7709

0.33

0.03

187.1

0.76

0.01

0.84

0.05

45.3

54.7

91

7907

0.36

0.02

95.1

0.85

0.04

0.90

0.06

42.6

57.4

94
60

2.7.0 Application of Method to GEOTRACES Samples
2.7.1 Water Column Profile
The solvent extraction method reported in this article was applied to samples from station
9 of GP15 (Figure 6). Error bars, for inert Cu, were propagated from measurements made for
total and labile Cu. Labile Cu was found to be highest at the surface and decreases rapidly below
the fluorescence maxima. Below 100 m, ≥90% of dCu was found to be consistently inert. This
appears to be in stark contrast with samples measured along GA03 (n=3) where depths ≥ 100 m
all had between 30-50% labile Cu. A detailed explanation of these trends requires analysis of
other samples from this expedition (GP15) and others. However, one explanation is that waters
recently exposed to the surface may have more labile Cu through the release of inert Cu by
photochemical reactions. This is in line with the findings of Moffett & Zika., (1987a, b), who
saw an increase in cupric ion activity which they attributed to the photolysis of Cu-binding
ligands. These workers saw a 20-fold increase in free cupric ions which are a part of the labile
Cu pool. Increase in free cupric ions at the surface was also observed in the oligotrophic
Sargasso Sea (Moffett et al., 1990). Moffett & Zika., (1987a) also saw a decrease in the organic
ligand concentrations below the euphotic zone which they attribute to decomposition. At station
9, however, we observed that most of the dissolved Cu below the euphotic zone to be inert, and
our current understanding of Cu-binding ligands would conclude that this inert pool would be
strongly bound by organic ligands. Another explanation is that surface waters contain more
recent inputs of Cu, and the formation of inert complexes are slow. The differences between the
Atlantic and Pacific samples, while preliminary at this stage, are also consistent with the notion
that older waters have more time for the formation of inert complexes or for the transfer of labile
Cu into the inert pool. Moffett et al., (1990) tested whether or not photo-oxidation of seawater
61

removed the strong Cu-binding ligand, the L1-class. They left a sample from the Sargasso Sea
(140 m depth) in sunlight for 4 hours a day for 5 days straight in a quartz flask; the sample was
then measured on the CLE-AdCSV where they found that the L1-class ligand, which was
previously detected, had become undetectable. A control, kept in the dark for the same period of
time, resulted in no loss of ligands. The local labile Cu minima appears to coincide with the
chlorophyll maxima. Similar results were reported in Moffett et al., (1990) where the depth of
maximum Cu-binding occurred at the chlorophyll maxima as well.
62

Figure 6. Depth profiles of labile, total, and inert copper species at station 9 (12-point depth profile), accompanied by profiles of
temperature, percent of total copper that is inert at each depth, temperature, and surface labile copper overlaid with surface
fluorescence. Concentrations of labile copper is measured on the ICP-MS and indicated by a red profile while concentration of
inert copper and percent inert are derived values and indicated in orange.

63

64

2.7.2 Implications of Inert Cu for the Electrochemical Determination of free [Cu
2+
] and
Conditional Stability Constants and Concentrations of Copper-Binding Ligands
There are several important implications of the data generated by this method. It is
necessary to reconsider how previous Cu speciation data were interpreted using CLE-AdCSV.
With electrochemical methods, unknown parameters are solved for using titration data and the
following equation:
𝐶𝑢 [𝐴𝐿 ]
2
[𝐶𝑢 ]
𝑇𝑜𝑡𝑎𝑙 =
𝛽 2
[𝐴𝐿 ]
2
1+ 𝛴 𝐾 𝑖 𝐿 𝑖 +𝛽 2
[𝐴𝐿 ]
2
(10)
where [AL] is the added ligand, [Cu]Total is the total dissolved Cu detected by CLE-AdCSV, β2 is
the conditional stability constant of the added ligand, Ki and Li is the conditional stability
constant and concentration of the naturally occurring ligand, respectively. Modifying equation
(9) and applying it to (10), we obtain:
𝐶𝑢 [𝐴𝐿 ]
2
[𝐶𝑢 ]
𝑇𝑜𝑡𝑎𝑙 − [𝐶𝑢 ]
𝐼𝑛𝑒𝑟𝑡 =
𝛽 2
[𝐴𝐿 ]
2
1+ 𝛴 𝐾 𝑖 𝐿 𝑖 +𝛽 2
[𝐴𝐿 ]
2
(11)
and therefore,
𝐶𝑢 [𝐴𝐿 ]
2
[𝐶𝑢 ]
𝐿𝑎𝑏𝑖𝑙𝑒 =
𝛽 2
[𝐴𝐿 ]
2
1+ 𝛴 𝐾 𝑖 𝐿 𝑖 +𝛽 2
[𝐴𝐿 ]
2
(12)
This allows for a better approximation of Ki and Li because the electrochemical titrations are not
accurately accounting for inert Cu, causing an overestimation of ligand concentrations and their
conditional stability constants.
At station 9, we measured one shallow and one deep sample (25 and 750 m depths) from
the profile using CLE-AdCSV, following the protocols outlined in Jacquot & Moffett., (2015).
Using 2 µM of the Cu chelator salicylaldoxime (SA), a conditional stability constant (K) and
ligand concentration ([L]) were calculated with a single ligand model on a nonlinear fit,
previously described by Gerringa et al., (1995) and Moffett & Dupont., (2008) where:
𝐾 =
[𝐶𝑢𝐿 ]
[𝐶𝑢
2+
][𝐿 𝑓 ]
(13)
Labile Cu
65

and
[𝐿 ] = [𝐶𝑢𝐿 ] + [𝐿 𝑓 ] (14).
The α of the natural ligand along with equation (13) can be rearranged into the reciprocal of the
Langmuir isotherm (15).

[𝐶𝑢𝐿 ]
[𝐶𝑢
2+
]
=
𝐾 [𝐿 ]
1+𝐾 [𝐶𝑢
2+
]
(15)
When weak and inorganic ligands are included, the equation becomes:

𝛴 [𝐶𝑢𝐿 ]
[𝐶𝑢
2+
]
= 𝛴 𝐾 𝑖 𝐿 𝑖 (𝑖 >1)
+
𝐾 [𝐿 ]
1+𝐾 [𝐶𝑢
2+
]
(16)
which was solved using Sigmaplot version 8.0 for K and [L], assuming ΣKiLi (i>1) = 200
(Moffett., 1995). Using the conditional stability constant (K’1) of the strongest detectable ligand
(L1), [L1], and total Cu ([Cu]T), the following quadratic can be formed to solve for [Cu
2+
]
(Moffett & Dupont, 2008):
0 = −[𝐶𝑢 ]
𝑇 + (1 + 𝐾 ′
1
[𝐿 1
] − 𝐾 ′
1
[𝐶𝑢
𝑇 ])𝑥 + 𝐾 ′
1
𝑥 2
(17)
where
[𝐶𝑢
2+
] =
𝑥 𝛴 𝑖 =2
𝐾 𝑖 𝐿 𝑖 (18).
As mentioned previously, ΣKiLi (i>1) is estimated as 200. Using both [Cu]T and the labile Cu
measured used in this method, the values for K’1, [L1], and [Cu
2+
] were calculated.
Results for conditional stability constants (K) and ligand concentration ([L]) are reported
in Table 3 and Figures 7A-D. Additional calculations were performed for titration data in the
North Atlantic reported in Jacquot & Moffett., (2015). These samples were collected at the same
stations and similar depths to the data reported in Table 2 (none of the sample depths actually
titrated in that study were available for comparison). Based on the data in Table 2, we assumed a
representative value of 50% inert Cu for samples on GA03 and applied that assumption to the
66

titration data in Jacquot & Moffett., (2015). In each case, subtracting the inert fraction from the
total Cu results in a substantially lower conditional stability constant that was obtained by
applying a single ligand model using total Cu. This is a more realistic estimate for the binding
constant of the complex comprising the labile fraction than the single ligand model, since the
existence of an inert fraction implies that there must be at least two ligand pools. The total ligand
concentration (Table 3) is defined as the sum of the “labile” ligand calculated from the non-
linear regression plus the inert Cu (assuming the inert Cu is a ligand-bound complex). It is larger
than the ligand concentration determined without considering inert Cu because the “labile
ligand” is similar to the value derived from equation (11). The excess ligand, defined as the total
ligand concentration minus the total Cu, is much larger when equation (11) is used. A higher
ligand concentration, combined with a weaker stability constant, means that the free [Cu
2+
]
estimates are basically unchanged with the exception of the 25 m sample at GP15 Sta. 9. These
results suggest that future analysis of electrochemical titration data should include an estimate of
the inert fraction.
67

Table 3. Conditional stability constant (K’1) of strong Copper ligand, strong ligand concentration [L1], and
free cupric ion concentration [Cu
2+
] measured on two depths from GP15 station 9 and 3 depths from GA03
using Competitive Ligand Exchange – Adsorptive Cathodic Stripping Voltammetry. Data obtained from
voltammeter were fitted to a nonlinear regression using Sigmaplot version 8. Total L ([L]Total) is defined as
the sum of the [L1] and the concentration of inert Cu while Excess L ([L]Excess) is defined as [L]Total subtracted
by [Cu]Total.
Sample and Cu used Depth
(m)
[Cu]Total
(nM)
Log(K’1)

[L1]
(nM)
[L]Total
(nM)
[L]Excess
(nM)
Log([Cu
2+
])

13050 with labile Cu

750

2.24

14.22

1.43

3.51

1.27

-14.86

13050 with total Cu

750

2.24

14.98

2.54

4.62

2.38

-14.86

13070 with labile Cu

25

1.20

14.99

1.46

3.42

2.22

-14.9

13070 with total Cu

25

1.20

14.79

1.47

3.43

2.23

-14.53

GA03 St. 18 with Labile Cu

110

0.53

14.28

1.26

1.79

1.26

-14.39

GA03 St. 18 with Total Cu

110

1.05

14.63

1.26

1.26

0.21

-14.20

GA03 St. 20 with Labile Cu

135

0.41

13.69

1.29

1.70

1.29

-13.88

GA03 St. 20 with Total Cu

135

0.82

13.94

1.30

1.30

0.48

-13.87

GA03 St. 22 with Labile Cu

51

0.49

13.83

1.22

1.71

1.22

-13.94

GA03 St. 22 with Total Cu

51

0.97

14.24

1.25

1.25

0.28

-13.93
68

Figure 7. (A-D). Samples from station 9 (GEOTRACES# 13050 and 13070) of GP15 were analyzed using the traditional
Competitive Ligand Exchange – Adsorptive Cathodic Stripping Voltammeter (CLE-AdCSV). Titration data were then run
through Sigmaplot vers. 8 to calculate values of K’1, [L1], and [Cu
2+
]. Results of this non-linear regression were made using the
measured labile Cu concentrations (6A – 13050 and 6C – 13070 using labile Cu) and total Cu concentrations (6B – 13050 and
6D – 13070 using total Cu). These results can be found in Table 3.

69

70

2.8.0 Discussion
The current paradigm underlying all speciation methods for Cu is that it is at equilibrium
amongst all of the competing ligands (natural and artificial) in a titration experiment. Many
previous studies using voltammetry, including those that use oxine, limit the equilibration time to
less than an hour. Beyond this time period, wall loss within the electrode assembly creates an
artifact. Our data suggest that at least 6 hours are required to reach a steady state (Figure 3).
Beyond that, there is little change between 6 hours and two weeks. These findings reveal several
problems when using competitive ligand exchange techniques with voltammetry. Firstly,
estimates of speciation parameters can change depending on whether or not the inert fraction is
excluded from the denominator in the speciation calculations. Secondly, since we showed that 6
hours was required for equilibration with the naturally occurring ligands in the labile fraction,
even with an exceptionally high competing ligand concentration (Figure 3), the earlier
voltametric methods have likely not reached equilibrium, given equilibration times of an hour or
less. This is certainly the case for studies that use oxine, such as Campos & van den Berg.,
(1994). Oxine is likely representative of other ligands, used in voltammetry, which have similar
oxygen donor ligands.
Our findings suggest that, when inert Cu is treated separately in calculations, estimates of
the stability constants are lower, but the excess ligand concentration increases. Therefore, the
estimates of free Cu
2+
are similar to those determined without considering an inert fraction. This
is not surprising because free Cu
2+
, in a sample, is determined by the equilibrium between Cu
and labile ligands in a sample even if this is only a small fraction of the total Cu pool. Recently,
there have been important advances in the modeling of titration data derived from competitive
71

ligand methods (Pižeta et al., 2015). However, these models are only useful if the assumptions
used to derive the original data are valid.
While “inert” remains operationally defined, the data in Figure 3 show that the two pools
we have designated as inert and labile are completely different in their exchangeabilities.
Alternatively, had there been a continuum of ligands with incrementally increasing lability, one
might have expected a linear relationship between oxine complex formation and time over the
two weeklong period which we did not see. This sharp break between the two pools creates
opportunities to characterize each fraction using other tools in the future such as an examination
of stable isotope composition. The sharp break also suggests that there may be fundamental
differences in the speciation, perhaps with inert Cu consisting of colloids (Jensen et al., 2020),
sulfidic nanoparticles (Luther & Rickard., 2005), or proteins; labile Cu is comprised of strong
but exchangeable, low molecular weight chelators.
If the valence of inert copper is Cu (I), it is possible that a Cu (I)-selective chelator may
identify a larger labile fraction. Moffett et al., (1985) and Moffett & Zika, (1988) showed that Cu
(II) chelators, even those that interact weakly with Cu (I), can greatly facilitate its oxidation and
incorporation into the Cu (II) pool. However, these workers were focusing on simple inorganic
complexes of Cu (I). Copper (I) associated with a protein or reduced sulfur may react slowly
with an oxygen-donor ligand, such as oxine.
The data presented here are not sufficient to present definitive conclusions about
distributional trends in the world’s oceans. However, it is likely that the inert fraction forms
slowly below the surface since there is a higher fraction in the older waters of the North Pacific
and lower in the euphotic zone and in the North Atlantic. The presence of an inert fraction of Cu
72

may have important implications for its geochemistry in the water column since it is anticipated
that its biological availability and exchange with sinking particles will be strongly inhibited.
2.9.0 Acknowledgements
Seawater sampling for this project was made possible by the scientists participating in the
GEOTRACES GP15 cruise. We would like to acknowledge the cooperation of the captain and
crew of the R/V Roger Revelle. We would also like to acknowledge the assistance of Seth John’s
lab – Shun-Chung Yang, and Xiaopeng Bian for measuring and sharing total copper
concentrations. Funding for this project was made possible by the National Science Foundation,
OCE-1636332, and awarded to James W. Moffett.
73

Chapter 3
The Distribution of Iodine Redox Speciation along the GEOTRACES GP15
Transect
Authors: Rintaro Moriyasu, Kenneth M. Bolster, Dalton S. Hardisty, David C. Kadko, and James
W. Moffett
Originally submitted to: Global Biogeochemical Cycles

3.0.0 Abstract
The distribution of iodate and iodide were measured along the GEOTRACES GP15
meridional transect at 152 °W from the shelf of Alaska to Papeete, Tahiti. The transect is divided
into two sections for the purposes of this manuscript: the oxygenated waters near the shelf of
Alaska, as a point of comparison for the reducing shelf of the various Oxygen Deficient Zones
that have previously been studied, and the main meridional transect. Our results reveal a dynamic
redox cycle of iodine in the upper water column. Iodide concentrations are highest in the
permanently stratified tropical mixed layers which reflects biological production on greater than
seasonal timescales. Using vertical mixing coefficients (Kz), derived from complementary
7
Be
data, the rate of in situ iodide oxidation (λ) from the top 150 m and iodide half-life (t1/2), were
calculated for two stations. The resulting turnover times, spanning 1-2 decades, contribute to the
surface maxima of iodide and strong gradients in iodide below the euphotic zone which agree
with estimates from a global iodine model. No evidence, for significant inputs of iodine from the
atmosphere, Alaskan continental margin, or seafloor, was shown in the data. These results are
consistent with previous work suggesting that the absence of oxygen in waters overlying
reducing margin sediments is a necessary precondition for large subsurface plumes of iodide
observed in oxygen deficient zones but not off the Alaskan continental margin.
74

3.1.0 Introduction
Iodine is predominantly found in its inorganic forms of iodate (IO3
-
) and iodide (I
-
) in
marine environments. It is of biological importance to humans (Hetzel., 1983) and marine
microorganisms (Farrenkopf et al., 1997; Wong., 2001; Amachi et al., 2007). On average, total
iodine, which is defined as the sum of the inorganic forms (Rue et al., 1997), is found to be semi-
conservative at concentrations ranging between 400-500 nM (Wong et al., 1985; Luther et al.,
1995; Chance et al., 2014) where the oxidized form of IO3
-
is dominant in most of the
oxygenated seawater (Wong & Brewer., 1974). However, there are exceptions where the
concentrations of I
-
are enriched; these exceptions include anoxic regions, referred to as Oxygen
Deficient Zones (ODZ), and surface ocean waters. For the former, iodine concentrations and
speciation measurements have been made for all three of the major ODZs: Arabian Sea
(Farrenkopf et al., 1997; Farrenkopf et al., 2002), Eastern Tropic South Pacific (ETSP; Cutter et
al., 2018), and the Eastern Tropic North Pacific (ETNP; Moriyasu et al., 2020). These areas
show high I
-
concentrations as a result of the dissimilatory reduction of IO3
-
into I
-
that occurs in
low oxygen (O2) environments (Farrenkopf et al., 1997; Amachi et al., 2007). Several studies
have measured the rate of reduction of IO3
-
into I
-
in these ODZs; studies, on samples obtained
from the Arabian Sea, found near instant reduction of IO3
-
during shipboard incubations as well
as the making of standards for electrochemical methods (Farrenkopf et al., 1997; Farrenkopf and
Luther., 2002). In contrast, a study in the ETNP (Hardisty et al., 2021) found reduction to occur
much more slowly at only ≤15 nanomolar (nM)/day and in incubation samples taken from depths
above the oxycline of the ODZ. These results, along with tracer experiments and model outputs
(Babbin et al., 2017), have concluded that IO3
-
is likely to be a terminal electron acceptor for
nitrite (NO2
-
), forming nitrate (NO3
-
) and I
-
. Additionally, IO3
-
has also been found to be a
75

terminal electron acceptor for acetate reduction, releasing I
-
as a side product (Reyes-Umana et
al., 2021).
The excess in total iodine concentration has been previously proposed to be a tracer for
O2 concentration and the redox condition of seawater (Cutter et al., 2018). In the ETSP and
ETNP ODZs, plumes of concentrated I
-
were previously observed; neither dissimilatory nor
abiotic reduction of IO3
-
can explain the magnitude of this plume. The excess iodine diffuses
from the reducing shelf sediment into the anoxic waters (Cutter et al., 2018; Moriyasu et al.,
2020) and has also been suggested as a tracer for shelf and margin inputs (Cutter et al., 2018).
Excess iodine was previously calculated by subtracting the mean total iodine concentration from
the measured total iodine, the sum of IO3
-
and I
-
concentrations (Moriyasu et al., 2020). These
previous works linked both the secondary nitrite maxima and I
-
maxima with the 13C water
mass, centering around a potential density (σθ) of 26.5 kg/m
3
, flowing through both the ETSP
and ETNP (Evans et al., 2020; Moriyasu et al., 2020). In contrast to these studies, which have
focused on hypoxic and anoxic regions, the GP15 meridional transect sampled in the oxygenated
sections of the North and South Pacific.
The other exception, where high concentrations of I
-
are found, is near the surface (upper
150 m). The origin of surface I
-
is thought to be associated with processes that occur during algal
growth (Hepach et al., 2020); surface I
-
has been suggested as a tracer for biological activity
(Tsunogai & Sase., 1969; Wong & Brewer., 1976; Luther et al., 1995; Farrenkopf et al., 2002;
Chance et al., 2014). The concentrations of sea surface I
-
have been shown to form a gradient
with latitude; while higher latitudes have between 50-100 nM concentrations of surface I
-
, lower
latitudes can have up to ≥200 nM (Nakayama et al., 1989; Farrenkopf et al., 2002; Carpenter et
al., 2021; Chance et al., 2020).
76

Surface I
-
is important, for its direct impact to marine iodine emissions, and have been
modeled and known to impact tropospheric ozone by reducing and forming iodine oxide radicals
(Sherwen et al., 2016). These radicals act as nucleation sites for cloud formation which can also
impact the climate of a region (Saiz-Lopez et al., 2012). Iodine radicals can also produce
hypoiodous acid (HOI) which reacts with organic molecules to form organo-iodine compounds
(Luther et al., 1995). These organo-iodine species were previously thought to be the main
transport of iodine from the ocean surface into the atmosphere (Lovelock et al., 1973; Moore &
Tokarczyk., 1992), but more recent studies of iodine mass balances have shown that organo-
iodine accounts for only 20% of marine iodine emissions (Jones et al., 2010; Prados-Roman et
al., 2015) while the rest originates from sea surface inorganic species (Carpenter et al., 2013)
which further highlights the importance of sea surface I
-
.
This manuscript explores both the input of excess iodine in the oxygenated margin of the
Alaskan shelf and the sea surface I
-
concentrations along the GEOTRACES GP15 meridional
transect. Additionally, using I
-
depth profiles from two stations straddling the equator and their
corresponding
7
Be data, the rate of I
-
oxidation from the mixed layer will be explored. Finally,
this manuscript will discuss methodological differences with other workers, primarily with
Cutter et al., (2018), which has caused overestimations in I
-
concentrations in the oxygenated
water columns. This difference becomes apparent in station 35 (-10.5 °N, 152.0 °W), a cross-
over station between GEOTRACES GP15 and GP16 transects.

77

3.2.0 Sampling
Samples were taken aboard the R/V Roger Revelle (RR1814 and RR1815) along the
GEOTRACES GP15 transect along the Pacific Ocean from the shelf of Alaska to Papeete, Tahiti
from station 1 (56.1 °N, 157.0 °W) to station 35 (-10.5 °N, 152.0 °W) between September 18
th
to
November 24
th
, 2018. The stations sampled along this transect can be found in Figure 1. For the
purpose of analyzing the results of this dataset, the transect is further divided into shelf stations
(stations 1-4 in Figure 2) and the main Pacific meridional transect (stations 7-35* in Figure 3).
(*Note: not all stations between 8 and 35 were collected for this manuscript).

78

Figure 1. Full transect of GP15. Marked on this map are the locations sampled for iodine speciation.

79

Fig 2. Shelf Stations of GP15 – points have been labeled with station numbers which correspond with other GP15 measurements.
Only stations sampled for iodine are marked.

80

Fig 3. Stations along the main transect of GP15 – points have been labeled with station numbers which correspond with other
GP15 measurements. Only stations sampled for iodine are marked.

The samples were collected, using the GEOTRACES trace metal clean carousel attached
with a Seabird CTD, with calibrated sensors for temperature, conductivity, pressure, fluorometer,
81

transmissometer, and oxygen. Each CTD carousel held 24 GO-FLO bottles which were closed at
selected depths. Samples were filtered using a 0.2 µm Acropak
TM
200 Supor capsule filter (Pall
Corporation) and frozen at -20 °C to preserve speciation. Previous studies on iodine speciation
have shown that iodine remains stable over longer periods of time (from months to years) when
filtered and frozen (Campos., 1997).
3.3.0 Materials
Each seawater sample was frozen and stored inside a 1 L Fluorinated Low-Density
Polyethylene (Nalgene; FLPE) bottle. Sample bottles were acid-washed prior to RR1814 and
RR1815 cruise following standard trace element and metal acid washing procedure according to
the GEOTRACES sampling manual (Cutter et al., 2017). While trace element and metal clean
procedures were not necessary for the preservation and measurement of iodine speciation, the
samples were originally collected for trace metal measurements.
3.4 Methods
The electrochemical analyses, for the quantification of I
-
, were made using the BASi
Controlled Growth Mercury Electrode Stand. The Calomel Reference Electrode (BASi; Part #:
EF-1352), and the Platinum Wire Auxiliary Electrode (BASi; Part #: MW-1032) were used for
analyses. The measurements, for IO3
-
concentrations, used a UV/Vis spectrophotometric method
(Perkin Elmer Lambda 35). Samples were held in a quartz cuvette with a cell length of 10 cm.
Measurements were made between wavelengths of 300-500 nm.
3.4.1 Iodide
Iodide was measured using the Hanging Mercury Drop Electrode (HMDE) with the
Cathodic Stripping Square Wave Voltammetry (CSSVW) setting. The method was adapted from
Tian & Nicolas., (1995) and Cutter et al., (2018); these workers originally adapted their methods
82

from Luther et al., (1988). The method, used in this work, was further adapted in Moriyasu et al.,
(2020). The major difference, between the Moriyasu et al., (2020) and the Cutter et al., (2018)
methods, is the inert gas used to purge O2 from the sample. While Cutter et al., (2018) used
nitrogen (N2) gas, Moriyasu et al., (2020), and this work, additionally, opted to use the denser
Argon (Ar). We noticed differences in measured I
-
concentrations between purging with N2 and
Ar. This difference was seen when measuring a crossover station between the GEOTRACES
GP15 and GP16 transects at station 35 of GP15 (-10.5 °N, 152.0 °W). Density likely plays a role
in the efficacy of keeping ambient oxygen out of the sample container.
The CSSVW method adds a 150 µL aliquot of 0.2% solution of Triton-X 100 (Sigma
Aldrich – BioX grade; CAS#: 9036-19-5) to a 10 mL sample prior to being purged under inert
Ar gas. The samples then require a minimum of 5 minutes to purge off any remaining O2
interference. After said purge, a 50 µL aliquot of 2 M sodium sulfite (Na2SO3; Millipore GR
ACS; CAS#: 7757-83-7) is added to prevent any remaining O2 from causing interference,
according to the protocol of Tian and Nicolas., (1995). The sample is then purged for an
additional minute and analyzed on the square wave setting. Each sample was measured with a
mercury drop size of 6, deposition time of 30 seconds, and a quiescence time of 5 seconds. Scan
increments were set to 2 mV while the scan range was set between -140 to -700 mV. The square
wave amplitude and frequency for the square wave stripping voltammetry were set respectively
to 25 mV and 125 Hz. Standard addition, with potassium iodide (KI; Sigma Aldrich – ACS
grade; CAS#: 7681-11-0), was used to determine I
-
concentrations in samples. Precision of
measurement was determined to be ±4 nM (standard deviation with 95% confidence for a sample
found to be 243 nM in concentration after 5 replicates) in the laboratory (Moriyasu et al., 2020).
83

In addition, for select samples from the crossover station 35, I
-
was also quantified via Ion
Chromatography Inductively Coupled Plasma - Mass Spectrometer (Thermofisher iCAP-TQ) at
Michigan State University. Methods were according to Hardisty et al., (2020, 2021) eluting I-
from 10 mL of seawater with 18% TMAH 2 M nitric on a AG1-X8 Bio-Rad anion exchange
resin. The additions of known I
-
concentrations and iodine-free 18.2 mΩ −cm water through the
same column procedure were used to assess yields and blanks. Each sample was independently
processed in duplicate or triplicate throughout the entire procedure.
3.4.2 Iodate
Iodate determination was adapted from Rue et al., (1991) who adapted their method from
Wong & Brewer., (1974). A 1 mL aliquot of 0.12 M sulfanilamide (Sigma Aldrich – ACS grade;
CAS#: 63-74-1) in 1% sulfuric acid (Macron Fine Chemicals) was added to a 25 mL sample.
The sample was equilibrated with the sulfanilamide solution for 5 minutes, after which a 1 mL
aliquot of a 0.12 M KI (Sigma Aldrich – ACS grade) solution in deionized water was added. The
sample was then measured within a minute from the addition of the KI aliquot; the absorbance
peak was then measured at 350 nm wavelength. The IO3
-
concentration was determined through
standard addition using potassium iodate (KIO3; Baker Analyzed
TM
– ACS Reagent; CAS#:
7758-05-6). Precision, based on 5 replicates, was found to be ±12 nM for a seawater sample with
286 nM IO3
-
(Moriyasu et al., 2020).
3.5 Results
3.5.1 Excess Iodine
Excess iodine was calculated by subtracting an estimated average total iodine
concentration from the sum of measured IO3
-
and I
-
, according to the work by Moriyasu et al.,
(2020). Since the region of study was outside of any ODZs, all depths were included into our
84

calculation for the mean value since there were no inputs of margin I
-
in any of our samples.
Previously, Moriyasu et al., (2020) found the average concentration to be 475 nM by averaging
inorganic iodine measured outside of the ODZ (depths ≥1000 m). This was likely to be a slight
overestimate as a result of residual inputs of excess iodine from within the ODZ and adjacent
shelf sediments (Evans et al., 2020). Here, we found that the average total iodine concentration
was 430 nM and that total iodine concentrations rarely go above 450 nM in oxygenated waters,
except in shelf and surface samples where total iodine exceeded 470 nM on average. Our data,
within the central North Pacific, agree with Nakayama et al., (1989) who reported that total
iodine concentrations were generally conservative at ≤450 nM.
3.5.2 Inert Gas for Oxygen Interference in the Electrochemical Determination of Iodide
The difference in I
-
concentrations, when using Ar instead of N2 gas, is significant
(Figure 4, 5). Samples from station 13 (34.0 °N, 152.0 °W) were measured on two separate
occasions with one of the two inert gases. Iodide concentrations at all depths were of greater
magnitude, when using N2, suggesting that a five-minute purge with N2 did not remove all O2
from the sample.
85

Fig 4. Iodide concentration measured on the Cathodic Stripping Voltammeter using either Argon (Ar – black line) or Nitrogen
(N2 – red line) at Station 13 of GP15 (34.0 °N, 152.0 °W)

86

Fig 5. Measured difference between purging with argon versus nitrogen gases at station 13 (34.0 °N, 152.0 °W). Difference
between measurements ranged between ~40-60 nM below 100 m while differences increased to ≥ 100 nM above.

Station 35 of this transect (-10.5 °N, 152.0 °W) was a crossover station with station 36 of
the GEOTRACES GP16 transect. This provided us a good opportunity for inter-calibration with
Cutter et al., (2018) (Figure 6). With the exception of the surface (≤75 m), our measured I
-

concentrations were on average 40-60 nM lower than those measured by Cutter et al., (2018).
The CTD data from both cruises also show that the seawater was oxygenated. A few samples
from station 35 (450 and 1200 m) were also measured for I
-
on the Inductively Coupled Plasma –
Mass Spectrometer (Table 1) which found ≤15 nM I
-
at those depths and corroborated the
precision of our measurements. The difference in inert gas will likely not cause problems in
anoxic waters where problems with O2 interference are less likely to occur; however, in
oxygenated waters, we recommend purging with argon.
87

Fig 6. Iodide depth profile from stations 35 of GP15 and 36 of GP16 (-10.5 °N, 152 °W). Samples, at this station, for GP16 were
originally collected in December of 2013, and data reported in Cutter et al., (2016) (red line) while samples for GP15 were
collected in November of 2018 (black line).

Table 1. Intercalibration of Iodide measurements made between Inductively Coupled Plasma – Mass Spectrometer and
Hanging Drop Mercury Electrode. Measurements were made at three stations along the GP15 transect in either the North
or South Pacific.

3.5.3 Iodine along the Main Transect
Iodide concentrations of the top 300 m of the GP15 transect can be found in Figure 7A
with corresponding IO3
-
(Figure 7B), total iodine (Figure 7C), and excess iodine (Figure 7D)
concentrations, respectively. Individual depth profiles for each demi and super stations can be
found in the supplemental figures. At almost all stations, along the main transect, we observed a
Sample
#
Station Depth
(m)
Lat.
(°N)
Long.
(°W)
Mass Spec Iodide
(nM)

Electrode Iodide
(nM)
13050 9 750 44.5 152.0 14.5±1.4 (n=3) 15±5
13358 13 350 34.5 152.0 9.3±1.1 (n=2) 12±2
15444 35 1200 -10.5 152.0 8.1±0.6 (n=3) 9.3±0.2
15456 35 450 -10.5 152.0 4.4±3.8 (n=2) 11.1±0.8
88

sub-surface I
-
maxima. In almost all cases, while the I
-
concentrations were highest at the
shallowest samples (between 20-30 m), there were noticeable increases in I
-
concentrations
between 50-100 m depths. Below 100 m, the I
-
concentrations dropped to values ≤20 nM where
concentrations fluctuate in a range between 2-15 nM. Deep water (≥5000m) samples were
measured for iodine speciation at the super stations (8, 14, 23, 29, and 35). These deep samples
showed a slight elevation of I
-
at roughly 50 nM. The elevation is likely to be a result of the
benthic release of I
-
from the decomposition of organic material at the sediment-water interface
(Kennedy & Elderfield., 1987 a, b). Individual station data is available on BCO-DMO.
Fig 7 (A-D). The distributions of iodide, iodate, total iodine, and excess iodine along the main GP15 transect (off of the shelf)
from 49.5 °N to -10.5 °N along the longitude of 152.0 °W.

B
A
89

3.5.4 Iodine along the Shelf of Alaska
Unlike in previous studies, where the focus was on the reduced waters on the shelf
(Cutter et al., 2018; Evans et al., 2020; Moriyasu et al., 2020), this study was able to observe I
-

on the oxygenated shelf of Alaska at stations 1-4 (56.1 °N, 157.0 °W to 54.7 °N, 155.0 °W).
While these stations had observably elevated I
-
concentrations, ranging from 50-120 nM even at
depths as deep as 1500 m, we did not observe a noticeable plume of excess iodine. By station 4
(54.7 °N, 155.0 °W), I
-
concentrations below the euphotic zone were ≤50 nM, indistinguishable
from the rest of the open ocean water along the main transect (Figure 7D).
C
D
90

3.5.5 1-Dimensional Model for the Estimation of Iodide Oxidation
The vertical turbulent diffusion rate (Kz) at two stations, 23 (7.5 °N, 152.0 °W) and 35 (-
10.5 °N, 152.0 °W), were calculated based on
7
Be data, resulting in 2.0 and 13.0 m
2
/d,
respectively. Using these Kz values and their corresponding I
-
concentrations, we calculated the
rate of I
-
oxidation at these stations using equation 1, originally derived in Kadko, (2017):
𝐶 (𝑧 )
= 𝐶 0
𝑒 𝛼 (𝑧 −𝐻 )
(1)
where C0 is the constant concentration in the mixed layer, C(z) is the concentration at depth z (in
m), and H is the mixed layer depth at the station. The value of α is defined as:
𝛼 = −√
𝜆 𝐾 𝑧 (2).
where Kz is in m
2
/day and 𝜆 is the decay constant in 1/day. Results were shown in Table 2. Using
a non-linear regression on MatLab, rate of oxidation (λ), and half-life (t1/2) were calculated for
these stations (Figure 8).
91

Fig 8. Non-linear regressions of Station 23 (7.5 °N, 152.0 °W) and Station 35 (-10.5 °N, 152.0 °W). The fit was used to calculate
the λ, the rate of in situ I
-
oxidation above the mixed layer.

3.6.0 Discussion
3.6.1 Upper Water Column Iodide and Biological Activity
Throughout the transect, the highest I
-
concentrations were in the mixed layer, reflecting
the highest rates of IO3
-
reduction, the lowest rates of I
-
oxidation, or both. In general, mixed
layer I
-
concentrations increased going from high to low latitudes (Figure 9). This may reflect
more time for I
-
to accumulate in the permanently stratified tropical stations compared to the
subtropical and temperate stations where seasonal mixing would decrease I
-
concentrations. Our
92

findings are consistent with previous work (Nakayama et al., 1989; Chance et al., 2014; Chance
et al., 2019) and are discussed in the following sections in the context of model-derived estimates
of I
-
turnover.
At most stations along the GP15 transect, the I
-
depth profiles resembled the fluorescence
depth profiles (Figure 10), indicating that the surface I
-
seen here are related to biological
activity. Iodide concentrations plotted against fluorescence at individual stations show a good
correlation (Figure 11). The data are plotted to include and exclude surface samples (above the
fluorescence maxima) where data points in red are samples from below the fluorescence
maxima, and points in black are above the maxima. Above the fluorescence maxima, the
correlation between I
-
and fluorescence falls off at most stations. The I
-
concentrations, at these
depths, typically were above the iodide-fluorescence curve; this may be a result of cell
senescence, speculated by Bluhm et al., (2010).
93

Fig 9. Iodide concentrations from the top 150 m of the water column across latitudes on the GP15 transect.

94

Fig 10. Iodide (black) and fluorescence (red) plotted against depth at demi and super stations in the North Pacific along the
GP15 transect (49.5 °N, 152.0

°W to 0.0 °N, 152.0 °W). The data plotted correspond to the top 400 m of each station.

95

Fig 11. Scatter plots of iodide concentration to chlorophyll fluorescence at individual stations. The black points and line are data
which include samples above the fluorescence maxima while the red points and line exclude these data points. Effectively, the
visible black points, which are not overlapped by the red dots, are surface samples.

The comparison of I
-
with fluorescence suggests that IO3
-
reduction by photoautotrophs is
a probable source. Phytoplankton have been shown to reduce IO3
-
to I
-
, and this may be
associated with oxidative stress (Hepach et al., 2020) which is consistent with elevated I
-

accumulated in sunlit shallow tropical mixed layers. At stations 13, 14, and 17, local maxima in
I
-
concentrations coincide maxima in chlorophyll fluorescence which may reflect enhanced
production by phytoplankton at those depths. Hardisty et al., (2021) reported high rates of IO3
-

96

reduction at ~95 m in the ETNP. Previous compilations have shown that I
-
concentrations are not
a simple function of primary production as other factors, such as temperature and mixed layer
depth, are more important (Chance et al., 2014). Iodide within the mean mixed layer depth was
also plotted against net primary productivity for all (Figure 12). Surprisingly, I
-
decreases with
increasing productivity. This suggests that seasonal mixing of the more northerly stations is a
key determinant in how much iodide accumulates in the mixed layer. The permanently stratified,
oligotrophic southern stations enable I
-
accumulation at higher levels.
Figure 12. Iodide concentrations in the mean mixed layer depth plotted against Net Primary Productivity along GP15 transect
from stations 7-35.

R² = 0.3369
0 0.3 0.6 0.9 1.2
0
0.3
0.6
0.9
1.2
0
60
120
180
240
0 225 450 675 900
Iodide (nM)
Net Primary Productivity (mg/(m^2*d))
97

3.6.2 Rate of Upper Water Column Iodide Oxidation
Our explanation for the trends in Figure 12 suggests that the turnover of I
-
must be longer
than a seasonal time scale in order to account for greater I
-
accumulation in permanently
stratified regimes. Indeed, results from the 1-D model show that I
-
is oxidized slowly below the
euphotic zone, with a half-life between 4-22 years (Table 2). In general, the lifetimes for I- in
marine waters have been estimated to be as low as <6 months to up to 40 years (reviewed in
Chance et al., 2014), which further corroborates slow oxidation kinetics. Our estimated range is
in good agreement with the recent estimate of 16 years from a global ocean model for surface I
-

(Wadley et al., 2020). As mentioned by Luther et al., (1995), the abiotic oxidation of I
-
by
oxygen is highly unlikely since the reaction by either singlet or triplet O2 is kinetically
unfavorable; the likelihood of oxidation by peroxides and ozone are also low as they are not
widely available in seawater.
Table 2. Rate of iodide oxidation calculated by forming a 1-dimensional model utilizing
7
Be as a tracer for vertical
upwelling. The equation C(z)=C(0)e
α(z-H)
was fitted to iodide depth profiles at various stations along the GP15 transect to
produce the decay constant of surface iodide, λ (/day). The half-life (t1/2) was also produced by the non-linear regression
performed on Matlab.
Cruise Station Lat. (
o
N) Long. (
o
W) λ (/day) t1/2 (days)
GP15 23 7.5 152.0 4.47E-04 1549.5
GP15 35 -10.5 152.0 1.01E-04 6885.9
Geometric
Mean
N/A N/A N/A 1.56E-04 4452.5

Previous studies have related I
-
oxidation to the first step of nitrification, the oxidation of
ammonium into NO2
-
(Truesdale et al., 2001; Long et al., 2015; Hughes et al., 2021). This
reaction, conducted by ammonium oxidizing bacteria and archaea, has been shown to be a light
inhibited process (Merbt et al., 2012). Photoinhibition of I
-
oxidation would explain why I
-
is
persistent in the mixed layer, particularly the shallow, permanently stratified mixed layers in the
tropics, which is a global phenomenon. However, it does not necessarily mean a link with
98

nitrification. Other reactions mediated by multi-copper oxidases, the same family of enzymes
that include ammonium mono-oxygenase, also are light inhibited. Moreover, iodide-selective
pathways have been reported in microbes that are mediated by multi-copper oxidases (Yeager et
al., 2017).
The data, in Figure 13 show that I
-
drops off sharply at the same depth as the primary
nitrite maxima in nearly every station. The primary nitrite maxima are thought to be a zone
where ammonium oxidation to NO2
-
is occurring at a high rate, slightly exceeding NO2
-

oxidation (the second step in nitrification). Unlike in the ODZ in the ETNP, the water that
contained the highest concentrations of the reduced species, I
-
and NO2
-
, was not the 13C water
which is found at the 26.5 kg/m
3
isopycnal (Figure 14). Especially around the equator, where I
-

concentrations were highest, the maxima were found at 23.0 and 24.0 kg/m
3
for I
-
and NO2
-
,
respectively.
Fig. 13. Depth profiles of iodide and nitrite at the four stations near and at the equator: 23 (7.5 °N, 152.0 °W), 29 (0.0 °N, 152.0
°W), 34 (-7.5 °N, 152.0 °W), and 35 (-10.5 °N, 152.0 °W).

99

Fig 14. Transect profile of iodide at the four stations nearest the equator, overlayed with nitrite concentrations, and plotted
against potential density instead of depths.

3.6.3 Gulf of Alaska as a Source of Iodine to the North Pacific
Previous studies have shown elevated concentrations of excess I
-
associated with inputs
from the continental shelf. Within ODZs, excess iodine appears to be associated with significant
benthic sources associated with highly reducing sediments off of the west coast of India
(Farrenkopf and Luther., 2002), Peru (Cutter et al., 2018), and southern Mexico (Moriyasu et al.,
2020). High iodine fluxes under these conditions are consistent with the hypothesis of François
(1987) that I
-
is displaced by sulfide under reducing conditions and that I
-
is released in
association with organic matter oxidation, similar to ammonia (Kennedy & Elderfield, 1987a, b;
Anschutz et al., 2000). A global model of sulfate reduction rates in sediments suggests that the
Alaskan continental margin near the terminus of the GP15 transect is a “hot spot” for sulfate
reduction (Bowles et al., 2014). Thus, we anticipated a high concentration of excess iodine in
shelf waters, possibly extending as an offshore plume. Results showed that excess iodine values
were much lower in the shelf waters than off Peru, India, or Mexico. While the Alaskan shelf is
oxidizing, excess iodine is the sum of I
-
and its oxidation product, IO3
-
, and so it should be
insensitive to oxidation. It seems more likely that benthic fluxes of iodine from the Alaskan shelf
100

sediments are simply lower than in those other regimes. At present, we do not have an
explanation for this finding. Importantly, we note that other shelf studies of iodine speciation
have also shown total iodine to be conservative (Luther et al., 1991) despite the likelihood of
high pore-water iodine in locally reducing sediments.
Our results indicate the likelihood that benthic fluxes of iodine from the Alaskan shelf
sediments are simply lower than in those other regimes characterized by bottom water hypoxia.
This could be from lower pore water iodine generally; however, given substantial sulfate
reduction and associated organic matter remineralization in the region, we suspect that there is
excess iodine in the pore waters and that dissolved iodine is being sequestered at the benthic
boundary layer at the transition between the reducing sediments and overlying oxic water
column. Specifically, it has been previously demonstrated that the organic iodine to total organic
carbon (I/TOC) ratios are higher at the sediment-water interface of localities underlying
oxygenated water columns relative to those of underlying low oxygen water columns (Zhou et
al., 2017; Price & Calvert., 1973, 1977; Kennedy & Elderfield., 1987 a, b; Lu et al., 2008). While
both types of localities have active iodine pore water inputs, as I
-
, in association with organic
remineralization, the explanation for the difference in I/TOC ratios seems related to active redox
cycling of both iodine and metals at the benthic-boundary layer of oxic localities. Specifically,
Anschutz et al., (2000) demonstrated sharp excess I
-
gradients in the upper 3 cm of bioturbated
sediments, which they hypothesized were related to I
-
oxidation to both I2 and IO3
-
. François.,
(1987) demonstrated that I
-
is not removed from the dissolved phase, but I2 and IO3
-
are. I2
quickly reacts with organic matter to form iodinated organic compounds (François., 1987) and
IO3
-
can be absorbed onto the surfaces of oxides (Fox et al., 2009). We note that, while seldom
measured, IO3
-
production from I
-
has been demonstrated in pore fluids (Kennedy & Elderfield.,
101

1987 a, b). Together, decreases in I
-
from oxidation and sequestration in sediments have the
impact of decreasing benthic I
-
fluxes. However, no study has directly compared benthic iodine
fluxes from sediments underlying gradients in bottom water oxygen content and thus further
work is needed to determine if and if so, why – benthic iodine fluxes, and in turn excess water
column iodine – differ between localities.
3.6.4 Iodate, Total Iodine, and Excess Iodine on the Transect
Most of the water column of the transect was dominated by IO3
-
, which was fairly
conservative, consistent with earlier works (Nakayama et al., 1989). Cutter et al., (2018) reported
low but persistent I
-
concentrations throughout the deep-water column, but we argue, based off
an intercalibration of the mercury electrode-Ar, mercury electrode-N2, and IC-ICP-MS
techniques for I
-
determination, that this was a result of an artifact in the protocol used in that
work. Our results are more consistent with those of Nakayama et al., (1989) who used a
completely different technique. Total iodine showed little variation on the transect except for the
slight increase near the margin, as noted above. Increases in I
-
in the mixed layer were
compensated for by decreases in IO3
-
.
3.7.0 Conclusions
This work synthesizes the distributions of the two main inorganic iodine species along
the GEOTRACES GP15 transect and determine the rate of I
-
oxidation in the surface oceans. Our
results reveal a dynamic redox cycle of iodine in the upper water column. Turnover times
spanning 1-2 decades contribute to the surface maxima and strong gradients in I
-
below the
euphotic zone and agree with estimates from a global iodine model. Our estimates were made
possible by complimentary
7
Be analyses performed on the GEOTRACES GP15 cruise.
Permanently stratified, shallow mixed layers in the tropics are associated with high I
-

102

accumulation. We found no evidence for significant inputs of iodine from the atmosphere,
Alaskan continental margin, or seafloor. The three major ODZs are more important for iodine
redox cycling, although other sources, such as major rivers, are understudied.
3.8.0 Acknowledgements
We would like to thank Karen Casciotti’s lab for the nitrite measurements used for the
analyses of this manuscript. We would also like to acknowledge the cooperation from the
captain, crew members, and marine technicians aboard the R/V Roger Revelle for operating the
equipment to make data collection possible. Funding for this project came from the National
Science Foundation OCE-1636332. Funding support for Dalton Hardisty’s IC-ICP-MS
measurements came from NSF OCE-1829406. Measurements for iodine on the ICP-MS, used in
this manuscript, were done by Alexi Schnur. The dataset used in this manuscript is being
submitted to the Biological & Chemical Oceanography Data Management Office (BCO-DMO).

103

Chapter 4
The Impact of Copper Lability on Distribution and Cycling along the
GEOTRACES GP15 Transect
4.0.0 Abstract
Copper (Cu) has traditionally been considered as an element with both “nutrient-type”
and “scavenged-type” distribution characteristics, often leading to the characterization of
“hybrid-type”. Dissolved Cu is typically depleted in surface waters, which is interpreted as
reflecting biological uptake, and generally exhibits a linear increase in concentration with depth.
Such linear profiles have been typically attributed to the reversible scavenging of Cu onto
sinking particles, the known main removal of Cu. The particulate Cu sinks to the benthic layer
and is re-released which is believed to be the cause of accumulation of dissolved Cu in the deep
ocean. Most of the dissolved Cu pool is bound by some form of organic molecules which are
commonly known as ligands. These organic ligands have been found previously to limit
bioavailability and control the rate of scavenging by sinking particles. Some of the dCu has been
found to bind so strongly to these organic ligands that the Cu atoms are not released on a
timescale of ocean circulation or of common biogeochemical processes, thus rendering it
kinetically inert. Here, we apply a solvent extraction methodology for speciation between
kinetically labile and inert Cu on samples taken from the GEOTRACES GP15 transect (shelf of
Alaska to Pape’ete, Tahiti) and from the Columbia River. Additionally, frozen samples, from
several stations along the GEOTRACES GA03 (Jacquot & Moffett., 2015) transect, were
measured using the method as a point of reference between the beginning and ending of the
oceanic thermohaline circulation. Our findings are, that chemically inert species of Cu play a
major role in global copper cycling, and that the labile species are a small minority of the total
dissolved pool and likely scavenged rapidly throughout the water column.
104

4.1.0 Introduction
Within oceanic regimes, Copper (Cu) is considered a micronutrient, at low ambient
conditions, as it is serves as a cofactor for proteins in many biological processes. Most notably,
Cu is used for electron transport by certain marine diatoms (Peers & Price., 2006), denitrification
(Granger & Ward., 2003; Ward et al., 2008), ammonia oxidation by archaea (Walker et al., 2010;
Amin et al., 2013; Jacquot et al., 2014), and oxidases for high-affinity iron transport in high-
nutrient low-chlorophyll regions (Maldonado et al., 2006). However, at elevated concentrations,
often from anthropogenic input, Cu can be a toxin (Moffett et al., 1997); it has been found to
disrupt Manganese and Zinc uptake in diatoms (Sunda & Huntsman., 1983), green algae (Sunda
& Huntsman., 1998), and cyanobacteria (Mann et al., 2002)
Copper bioavailability and toxicity are dependent on the concentration of free cupric ion
(Cu
2+
), rather than total dissolved Cu (Sunda & Lewis., 1978; Coale & Bruland., 1988). Nearly
all of the dissolved Cu pool is bound by organic molecules (van den Berg et al., 1987; Coale &
Bruland., 1988; Moffett & Dupont., 2007), commonly referred to as ligands, and a large fraction
of these ligands are of biogenic origin. Since even a cupric ion concentration as low as 10
-11
M
can be toxic to organisms, many microorganisms are suspected to produce organic ligands as a
means to buffer the concentration of [Cu
2+
] to safe levels while having enough to maintain
biological activity (Bruland & Lohan., 2003). Cyanobacteria have an even lower resistance to
Cu-toxicity, resulting in a decrease in reproduction due to ambient cupric ion concentrations as
low as 10
-12
M (Brand et al., 1986). As a result of this buffering behavior, Cu is often referred to
as the “Goldilocks” element.
Copper species and their organic ligands are typically determined through
electrochemistry. Using an added synthetic ligand to compete with the naturally occurring
105

ligands, it is possible to thermodynamically characterize the Cu pool and associated chelators.
These ligands are typically categorized by the range of their conditional stability constants
(𝐾 ′
1
𝑐𝑜𝑛𝑑 =
[𝐶𝑢 𝐿 1
]
[𝐶𝑢
2+
][𝐿 1
]
), a unitless measure of binding strength; stronger ligands, commonly
referred to as L1, have a K1’
cond
of 10
13
-10
16
(Bundy et al., 2013), which are thought to be
microbially produced to detoxify cupric ions (Moffett & Brand., 1996). Weaker L2 ligands, on
the other hand, are organics that are thought to be terrestrial or marine degradation products
(Whitby et al., 2018). The strength of these ligands is operationally defined based on the
conditional stability constant of the added chelator. Typically, the strength of the added synthetic
ligand is quantified by the partition coefficient (α) which is defined as:
𝛼 𝐶𝑢 (𝐴𝐿 )
𝑥 =
[𝐶𝑢 (𝐴𝐿 )
𝑥 ]
[𝐶𝑢
2+
]
= 𝛽 2
𝑐𝑜𝑛𝑑 [𝐴𝐿 ]
2
+ 𝐾 ′
1
𝑐𝑜𝑛𝑑 [𝐴𝐿 ] (1)
where [Cu(AL)x] is the concentration of the complex formed between Cu and the added ligand
([AL]), and [Cu]
2+
is the concentration of the cupric ions. This partition coefficient is equal to the
sum of the conditional stability constants of the mono- and bis- complexes. This α is referred to
as the analytical window of the ligand. Strong L1-type ligands are thought to be the product
many marine organisms use to buffer cupric ion concentrations. Indeed, Synechoccosus sp. were
found to produce ligands with K1’
cond
similar to L1-type ligands in culture experiments (Moffett
& Brand., 1996). These workers found that the concentration of L1 ligands fluctuated when the
cultures were exposed to excess cupric ions.
The major source of oceanic Cu is thought to be fluvial input (Bruland & Lohan., 2014)
while the major sink of Cu is scavenging (Boyle et al., 1977). However, Aeolian inputs of Cu
may also be important, seeing as there are large surface Cu maxima observed in the Indian
Ocean, where there are dust inputs from deserts (Saager et al., 1992). Copper is also thought to
be introduced into surface waters via aerosol inputs; for example, in the Pacific Ocean, there are
106

Cu inputs from cities on the Asian continent (Wang et al., 2016). The near linear depth profile is
explained by sedimentary diffusion of Cu species into deep waters and mid-depth scavenging of
Cu (Boyle et al., 1977). These workers hypothesized that the scavenged, particulate Cu, which
settle in the benthic layer, is re-released, and recycled into the water column through biological
and chemical activity.
Another hypothesis for the linear distribution of dissolved Cu is reversible scavenging by
particles (Little et al., 2013; Roshan & Wu., 2015; Richon & Tagliabue., 2019). This model was
originally applied to
230
Th by Bacon & Anderson., (1982). Similar to the radioisotope Thorium-
230 (
230
Th), also known to have a linear depth profile, Cu is thought to be reversibly scavenged
onto the surfaces of sinking particles. However, unlike
230
Th which has a source throughout the
water column from the α-decay of Uranium-234, Cu is circulated and cycled throughout the
oceans, so that a model, such as the one applied to
230
Th, is likely not appropriate.
Copper bound by inorganic ligands, referred to as inorganic Cu (Cu’), and cupric ions are
thought to be the species removed from the surface water by scavenging (Sunda & Huntsman,
1995; Little et al., 2013; Richon & Tagliabue., 2019); the prominent inorganic ligand that binds
Cu is thought to be bicarbonate (Byrne & Miller., 1985). This inorganic pool of Cu has been
hypothesized to sorb on and off of particles while organic Cu remains in the dissolved phase.
Davis & Leckie., (1978) have shown that strong-binding metal ligands control the rate of metal
scavenging on sinking particles. Since then, the importance of L1-type Cu ligands on the rate of
scavenging have been elucidated in estuaries (van den Berg et al., 1987; Laglera & van den
Berg., 2003). These organic ligands affect the partitioning between the dissolved and particulate
phases and act as a buffer for the metal ion in the dissolved phase.
107

An alternative to the former two hypotheses, for the linearly increasing distribution of Cu
is the existence of an inert Cu pool. This fraction of the Cu pool was previously proposed by
Kogut & Voelker., (2003); they operationally define this fraction as chemically inert and non-
exchangeable with added ligands (1 mM of the Cu chelator, salicylaldoxime) in their titration
experiments, resulting in a log(αsa) of 8.5 and an equilibration period of 24 and 48 hours,
although they do not report a significant change in inert Cu between the two equilibration times.
Their results show that the log(K’) of the inert fraction is likely to be ≥ 13. The voltammetry,
utilized by these workers, is limited in two ways for the purpose of measuring inert Cu. The first
issue is wall-loss. Voltammetric techniques require an equilibration period where the in-situ
metal can form new complexes with the added ligand. This assumption, of an inert fraction, was
also recently used by Ruacho et al., (2020) to explain the distribution of Cu along the
GEOTRACES GP16 transect.
Using a solvent extraction technique, developed by Moriyasu & Moffett., (2022), this
work elucidates the distribution of labile and inert Cu along the GEOTRACES GP15 “Pacific
Meridional Transect” from the shelf of Alaska to Papeete, Tahiti. This dataset, along with
samples preserved from the GEOTRACES GA03 transect (Jacquot & Moffett., 2015), allow us
to compare the Cu lability between the Pacific and Atlantic. Additionally, we explore the sources
of both labile and inert pools.
4.2.0 Sampling
Samples were obtained from the GEOTRACES GP15 cruise aboard the R/V Roger
Revelle (RR1814 from September 18
th
- October 21
st
of 2018, and RR1815 from October 24
th
-
November 23
rd
of 2018), shown on the map in Figure 1. The Revelle was equipped with the
standard GEOTRACES CTD/rosette (Model 32G, Sea-Bird Electronics). This rosette contained
108

24 Teflon-coated 12 L GO-FLO bottles for trace metal clean sampling (Model 10812 T, General
Oceanics) with sensors for dissolved oxygen, conductivity, algal pigments, pressure, beam
transmittance, and temperature. These samples were preserved by filtering them through 0.2 µm
Acropak-200 Supor capsule filter (Pall Corporation) and freezing them in 1 L Fluorinated Low-
Density Polyethylene (Nalgene; FLPE) bottles. Sample bottles were acid-washed prior to
RR1814 and RR1815 cruises according to the GEOTRACES cookbook (Cutter et al., 2017).
Fig 1. Map of the GP15 transect with markers for stations measured for labile, inert, and total copper.

109

Samples were also taken from the Columbia River off of the coast of Oregon using a
carboy as a sampler (Figure 2). Samples were filtered with a GFD + 0.2 µm Supor and frozen to
preserve speciation. Total Cu measurements for these samples were also made by mass
spectrometry. These samples served as end members for fluvial input.
Fig 2. Map of the Columbia River where samples were taken.

4.3.0 Methods and Reagents
4.3.1 Measurement of Labile Copper
Labile Cu measurements were made using a solvent extraction methodology, described in
Moriyasu & Moffett., (2022), and measured using the Thermo Element 2
TM
Isotope Dilution
Inductively Coupled Plasma – Mass Spectrometer (ICP-MS). In this method, seawater samples
are equilibrated with the organic solvent, methylbenzene (toluene; Sigma Aldrich; CAS #: 108-
88-3) at a ratio of 4:5 solvent to seawater. After which, 500 µM of the strong Cu-binding ligand,
8-hydroxyquinoline (oxine; Sigma Aldrich; CAS# 148-24-3; ACS grade), is added. The labile
110

portion of the Cu pool forms a nonpolar bis-complex with the oxine which fractionates into the
organic solvent layer over the course of 6 hours. The toluene is then physically separated from
the seawater and evaporated on a hotplate. The labile Cu is then re-dissolved into 1 mL of 5%
Optima grade nitric acid solution per replicate (HNO3; Thermo Fisher; CAS# 7697-37-2) in 15
mL polypropylene Falcon tubes (VWR; Catalog #89049-172). These samples were then
measured on an Element 2 (Thermo Fisher) ICP-MS using isotope dilution. All chemistry
involving toluene and heating were done in 15 mL Standard Teflon Vials with rounded interiors
(Savillex; Catalog #200-015-20) with 33 mm PFA Closures (Savillex: Catalog #600-033-01) and
7 mL Standard Teflon Vials with flat interiors (Savillex: Catalog #200-007-10) with 24 mm PFA
Closures (Savillex: Catalog #600-024-01) as typical trace metal clean polypropylene tubes are
permeable to toluene and not resistant to heat.
4.3.2 Measurement of Total Copper
Total Cu measurements for GEOTRACES GP15 were adapted from a method from Rapp
et al., (2017). Samples were acidified to a pH of 1.7, and concentrated Optima
TM
grade
hydrogen peroxide (Fisher; H2O2; CAS#: 7722-84-1) was added to oxidize organic interferences,
according to the methods of Baconnais et al., (2019). The samples were then pre-concentrated
using the SC-DX seaFAST (Elemental Scientific; M-SFS2-MG-52) automated pre-concentration
solid phase extraction columns and measured on the ICP-MS at medium resolution.
4.3.3 Photodegradation Experiment
Two GEOTRACES samples from station 8 of GP15 (47.0 °N, 152.0 °W) were placed
within quartz flasks with stoppers and exposed to natural sunlight in Los Angeles (34.0 °N,
118.3 °W) from February 7-14
th
, 2022. Samples from 2100 and 5000 m were chosen as
representatives of mid- and deep-water samples that were unlikely to have been exposed to
sunlight in many years prior to collection and were expected to contain a high fraction of inert
111

Cu. The weather was projected to be good, and it was thought to be an ideal week to conduct an
experiment on the destruction of inert Cu. After a week, samples went through the solvent-
extraction protocol described in Moriyasu & Moffett., (2022).
4.4.0 Results
Most of the water column throughout the transect was dominated by inert Cu. At almost
all stations, we observed maxima in labile Cu concentrations at surface depths (Supplemental
Figures). At the surface, the labile Cu as a percent of total dissolved Cu is ranges between 30-
40% while at most other depths on the transect, this value is typically between 10% or less of the
total. These maxima can also be observed in the transect profiles (Figures 3 and 4). While we
observed a slight increase in labile Cu with increasing depth, the total Cu concentrations also
increase, resulting in an overall lower fraction of labile to inert copper. The GP15 transect can
largely be divided into groupings of shelf stations and main transect stations
112

Figure 3 (A-E). The distributions of labile, total, inert, and percent inert, and percent labile Cu along the GP15 transect from
station 7 to station 35 (49.5 °N, 152.0 °W to -10.5 °N, 152.0 °W).

A
B
C
D
113

4.4.1 Stations along the Main Transect
Depth profiles for ten stations were taken along the cruise transect from 49.5 °N, 152.0
°W to -10.5 °N, 152.0 °W (see supplemental figures for specific station numbers and
coordinates). Total Cu increased approximately linearly with depth (figure 3B), as previously
observed, and the labile fraction of Cu was high throughout the surface of the entire main
transect (figures 3A, E). At most stations, labile Cu was highest in the shallowest samples and
decreased rapidly with depth and ranged between 20-50% of total Cu. Aside from the stations
around the equator, which had mid-depth maxima (at ~3000 m), the Cu pool between 1000-4000
m was made up of mostly of the inert fraction. At depths ≥ 4000 m, both the total, labile, and
114

inert Cu concentrations sharply increased. Additionally, the percent labile Cu is shown to
increase to 15-30% of total Cu, compared to less than 10% labile Cu found in mid-depth
samples, suggesting that the sediment layer is a source of labile Cu.
4.4.2 Stations along the Shelf of Alaska
Depth profiles for four stations were taken along the Alaskan shelf (stations 1-4). The
deepest sample obtained from the shelf was at 1595 m from station 3 (55.1 °N, 155.7 °W). At
these stations, labile Cu (figures 4A, E) concentrations are elevated. Total Cu (figure 4B) was
highly concentrated at the station closest to shore (station 1), reaching concentrations ≥ 4 nM at
the surface. At this station, however, we found that ≤ 10% of the total dCu was labile (figures
4C, D), likely corresponding to both the high concentrations of total dCu and high particulate
concentrations on the shelf of Alaska (Turner et al., 2017). Comparatively, labile Cu (and as a
result % labile Cu) was greater at station 2 than 1 (supplemental figures S1 and S2). It is likely
that a similar release process, observed in waters overlying the deep sea sediment, is occurring in
the waters above the sediment on the shelf. The rest of the shelf stations had % labile Cu at the
surface comparable to the rest of the transect.
115

Fig 4 (A-E). Labile, total, inert, percent inert, and percent labile copper transect profiles for stations 1-4 that were sampled
along the shelf of Alaska.

A
B
116

C
D
E
117

4.4.3 Stations from the North Atlantic GEOTRACES GA03 Transect
Frozen samples (see Jacquot & Moffett., 2015 for sampling and preserving) from five
stations (3, 14, 16, 18 and 22) along the GEOTRACES GA03 transect were measured for Cu
lability to provide depth profiles. Specifically, station 22 (19.4 °N, 29.4 °W) offered an
opportunity to compare deep samples between the North Atlantic and Pacific oceans (Figure 5).
Labile Cu, at this station, was higher at comparable depths in the North Pacific; the labile
fraction, at station 22 of GA03, ranged from 15-40% throughout the water column and never fell
below 10% even at intermediate depths (2000-4000 m). Full station data are shown in the
Supplemental Figure (S15-19).
Fig 5. Percent labile Cu (% Labile Cu) by depth of the Pacific (GP15) versus the Atlantic (GA03).

0 0.3 0.6 0.9 1.2
0
0.3
0.6
0.9
1.2
0
1300
2600
3900
5200
0 25 50 75 100
Depth (m)
% Labile Cu
Pacific
Atlantic
118

4.4.4 River End Member
Four samples were taken from the Columbia River outflow off of the coast of Oregon on
the 28
th
of September 2021; these samples were taken to act as riverine end members (46.15 °N,
123.59 °W to 46.13 °N, 123.44 °W). The Columbia River was chosen since it is one of the
largest riverine sources in the Northeast Pacific Ocean where many of the stations are located.
The salinities at all four locations were all less than 1 PSU. Copper species for these samples are
reported in Table 1. The fraction of reactive Cu in all of these samples was ≤ 8% of the total
with the highest labile Cu being found at the location closest to the Pacific (46.25 °N, 123.99
°W).
Table 1. Samples taken along the Columbia River and respective Cu species concentrations. All samples were taken from
the surface using a carboy.

*Sample # is independent of GEOTRACES sample numbering. Please refer to Figure 2 for locations.
1
% Labile refers to the concentration of labile Cu as a percent of total Cu concentration.
2
% Inert refers to the concentration of inert Cu as a percent of total Cu concentration.

Sample
#*
Lat. (°N) Long. (°W) Labile Cu
(nM)
Total Cu
(nM)
Inert Cu
(nM)
% Labile
1
% Inert
2

1

46.25

123.99

0.49

6.40

5.92

7.59

92.41

2

46.20

123.93

0.27

6.90

6.62

3.97

96.03

3

46.19

123.88

0.39

6.57

6.18

5.97

94.03

4 46.23 123.74 0.20 8.02 7.83 2.47 97.53

119

4.5.0 Discussions
4.5.1 Labile Copper – Inputs from Benthic and Surface Boundaries
Labile Cu has the highest fraction in two distinct places: the surface and benthic layers.
The release of benthic Cu has been previously observed by workers, such as Boyle et al.,
(1977) and Klinkhammer., (1980), who have explained the release to be a destruction of the
Cu-scavenging properties of particulates via diagenesis, and Cu is likely to increase in
lability from this process.
Another source of labile Cu appears to be the surface boundary which is likely the result
of aeolian inputs or internal conversion of inert to labile Cu via photochemical processes.
While we cannot rule out dust inputs within the scope of this work, we can explore the
possibility of light exposure as a source of labile Cu. This would corroborate previous works
that showed a decrease in strong Cu-binding ligands as a result of UV-irradiation by sunlight
(Moffett & Zika., 1987; Moffett et al., 1990; Laglera & van den Berg., 2006; Sato et al.,
2021).
Previously, Moriyasu & Moffett., (2022) showed that inert Cu becomes solvent
extractable after 2 hours of exposure under a UV-lamp. From these results and the observed
surface maxima of labile Cu, it is likely that the compounds causing Cu to be chemically
nonreactive undergo photo-degradation. This hypothesis was tested by leaving two
GEOTRACES samples from station 8 (47.0 °N, 152.0 °W) in quartz flasks and exposing it to
natural sunlight. Results of this experiment is in Table 2.
120

Table 2. Destruction of Inert Copper by natural UV radiation. Samples were left at 34.0 °N, 118.3 °W for the
span of one week in February of 2022. Both samples were taken from Station 8 of GP15 (47.0 °N, 152.0 °W).
1
Refers to GEOTRACES Sample ID number.
2
Labile Cu fraction measured previously using method described in Moriyasu & Moffett., (2022).
3
Samples were UV-irradiated naturally on the roof of Ahmanson Center for Biology, University of Southern California (34.0 °N,
118.3 °W) from February 7-14
th
, 2022.
4
[Cu]Total obtained following protocol by Rapp et al., (2017) and Baconnais et al., (2019)
5
A % Recovered when comparing Cu measured post-UV oxidation of ligands and total measured.

As shown in the table above, after one week of exposure to UV radiation, ≥95%, of the
previously unavailable, inert Cu becomes extractable in organic solvent. This supports the
hypothesis that photo-degradation of Cu binding compounds at the surface is one potential
source of labile Cu. Both sources of labile Cu, benthic and surface, do not appear to be related to
any specific water mass since they are ubiquitous throughout the transect.
4.5.2 River Input and the Terrestrial Source of Inert Copper
Labile and inert Cu speciation measurements were made at four separate points along the
Columbia River. The Columbia River is responsible for the input of approximately 2×10
11
m
3
of
water, and with it 4.9×10
10
moles of organic carbon into the Pacific, annually (Dahm et al.,
1981). As such, the Columbia has a major impact on the waters in the Northeast Pacific, a large
section of the GP15 transect. Previous studies have attributed riverine input to be the main source
of Cu into the ocean (Bruland). Two models of the biogeochemical cycling of Cu (Little et al.,
2014; Richon & Taliabue., 2019) also attribute rivers as the source. The model by Little et al.,
G
#
1

[Cu]l
abile

(nM)
2

Dep
th
(m)
[Cu]labile
after
exposure
(nM)
3

[Cu]Total
(nM)
4

Initial
%
Labile
Initi
al %
Iner
t
%
Recovere
d
5

12
91
9

1.24
±0.0
4

505
0

6.03±0.02

6.23±0.0
1

10.3

80.1

96.8

12
97
1

0.32
±0.0
2

210
0

3.12±0.02

3.280±0.
01

9.0

91.0

95.1
121

(2014) attribute a Cu flux of approximately 0.6-0.8 Gmol/year while the Richon & Tagliabue.,
(2019) attribute 6.7 Gmol/year. Along with this input, Cu-binding compounds, such as humic
substances and ligands, are also likely sourced from rivers, however, there have yet to be any
calculations as to their fluxes. This difference in fluvial fluxes of Cu has caused an order of
magnitude difference in residence time between the two models. Historically, the paradigm has
been that Cu has a long residence time of around 5000 years (Boyle et al., 1977) since dCu
concentrations increased linearly with salinity which would agree with the model Little et al.,
(2014). This new fluvial flux by Richon & Tagliabue., (2019) results in a residence time of only
400-500 years.
Our data indicate that rivers, or at least the Columbia, contain low concentrations of
labile Cu. This is in line with the findings of previous workers, who found that much of the Cu in
coastal waters was inert (Kogut & Voelker., 2003). From the data, it appears that heavily
complexed, chemically inert Cu have fluvial sources. This could suggest that inert Cu does not
necessarily all form during the in situ complexation of labile Cu in the water column, but that
one major source of inert Cu may be terrestrial. Studies done on estuarine sites found that the
majority of the Cu-binding compounds were terrestrially derived humic substances (Kogut &
Voelker., 2001; Muller & Batchelli., 2013). While humic acids and substances have been found
to have lower binding constants than L1-type ligands (Whitby et al., 2018), a similar terrestrial
compound may be responsible for inert Cu.
4.5.3 GA03 Atlantic Ocean and the In Situ Formation of Inert Copper
Compared to the Pacific, the general trend seen in the North Atlantic was higher labile Cu
fractions at all depths. The increase in the inert fraction from the Atlantic to the Pacific Ocean
can be found in figure 5, and specific profiles of North Atlantic stations can be found in the
122

Supplemental Figures. The data would suggest that starting in the North Atlantic, inert Cu
concentration increases over the course of the global ocean conveyor belt, with the oldest waters
in the North Pacific accumulating the highest concentrations, indicating that at least part of the
inert Cu pool is formed in situ through the complexation of labile Cu with time.
4.5.4 Copper Lability and the Distribution of Dissolved Copper
Looking at the overall transect profile and individual depth profiles, labile Cu appears to
act more similarly to a “scavenged-type” element. An input of Cu at the surface, as a result of
aeolian, or photochemical degradation, with mid-depth removal by sinking particulates, and
benthic recycling; the labile Cu depth profile is reflective of this cycle.
Both the distributions of labile and inert Cu play a role in the biogeochemical cycling of
the total dissolved metal. The two pools of dissolved Cu are also likely to have different
residence times. Whether the labile species is rapidly complexed into the inert species, or
rapidly scavenged into a third pool of Cu (i.e., particulate Cu), the equilibrium between labile
and inert species is currently unknown. Labile Cu should have a fairly short residence time in the
range of hundreds of years, similar to other scavenged elements, such as aluminum which has a
residence time of around 200 years (Orians & Bruland., 1985). The inert fraction, which
increases linearly with depth along with total Cu, would likely have residence times similar to
more “traditional” total Cu residence times.
Neither the labile nor inert fraction appear to be associated with any specific water mass
(Lawrence et al., 2022). This makes sense when considering that the main sources of labile
Cu/sinks of inert Cu are photo-degradation at the surface and diagenesis at the benthic depths;
these processes appeared to be ubiquitous throughout the main transect. There, however, was one
measured mid-depth maximum in labile Cu concentration at station 14 (32°N, 152.0 °W),
123

centered around a depth of 3000 m. This may be associated with Pacific Deep Water, but we
have no explanation as to why there was no mid-depth maximum at station 8 (47.0 °N, 152.0
°W). The stations between these two stations only sampled to a depth of 1200 m. Finally, surface
labile Cu was compared with GP15 surface iodide data collected by Moriyasu et al., (under
review) since both chemical species have surface maxima, likely due to photochemistry and
phytoplankton activity (Farrenkopf & Luther., 2002; Amachi et al., 2007) (figure 6 A-D).
Figures 6A and C are depth profiles at station 7 (49.5 °N, 152.0 °W) and 14 (32.0 °N, 152.0 °W)
while B and D are scatterplots of the two species from the top 300 m. At station 14, percent
labile Cu and iodide appear to follow a near linear trend, however this trend is not apparent at
other stations. This is likely a result of permanent stratification occurring at station 14 while most
other stations exhibit seasonal stratification. Both species likely form, as a result of
photochemistry, and then they are quickly removed by mixing. Due to the stratification going on
at station 14, these processes were preserved.
124

Fig. 6 (A-D). Depth profiles of percent labile Cu and iodide concentrations (nM), and scatter plots of percent labile Cu
versus iodide concentrations for stations 7 and 14.
A
B
125

C
D
126

4.6.0 Conclusion
Our data show that the overall linear depth profile of Cu can likely be attributed to the
inert Cu pool which make up the majority of the total Cu pool. However, when observing the
labile Cu distribution, it acts much more like a typical “scavenged-type” element which has a
surface maximum throughout the GP15 transect. This surface source is likely a result of photo-
degradation of inert Cu by solar radiation. Additionally, we observe a source of benthic labile
Cu, likely to be re-released labile Cu from the sediment layer, but this does not contribute much
to the overall total Cu profile since inert Cu concentrations are so high. Fluvial end-member
samples also show that the Cu pool is also comprised almost entirely of inert Cu, suggesting a
potential terrestrial source of the inert fraction. Finally, inert Cu has higher concentrations in the
Pacific compared to the North Atlantic which likely has to do with a mix of two processes: the
flux of inert Cu from rivers, and the accumulation of inert Cu over the span of the thermohaline
circulation of the global ocean.
4.7.0 Acknowledgements
Sampling for this dataset was made possible by the scientists and crew participating in the
GEOTRACES GP15 cruise. We would also like to thank the crew of the R/V Roger Revelle who
participated in the RR1814 and RR1815 cruises (the first and second legs of the GP15 transect).
Finally, we would also like to acknowledge the assistance of Seth John’s lab at the University of
Southern California – Shun-Chung Yang, and Xiaopeng Bian for measuring and sharing total
copper concentrations. Funding for this project was made possible by the National Science
Foundation, OCE-1636332, and awarded to James W. Moffett. Finally, the International
GEOTRACES Programme is possible, in part, thanks to the support from the U.S. National
Science Foundation (Grant OCE-1840868) to the Scientific Committee on Oceanic Research
(SCOR).
127

References
1. Acharya, S. S., & Panigrahi, M. K. (2016). Eastward shift and maintenance of Arabian Sea
oxygen minimum zone: Understanding the paradox. Deep Sea Research Part I:
Oceanographic Research Papers, 115, 240–252. https://doi.org/10.1016/j.dsr.2016.07.004
2. Amachi, S., Kawaguchi, N., Muramatsu, Y., Tsuchiya, S., Watanabe, Y., Shinoyama, H., &
Fujii, T. (2007). Dissimilatory iodate reduction by marine Pseudomonas sp. strain
SCT. Appl. Environ. Microbiol., 73(18), 5725-5730. https://doi.org/10.1128/AEM.00241-07
3. Amin, S. A., Moffett, J. W., Martens-Habbena, W., Jacquot, J. E., Han, Y., Devol, A., et al.
(2013). Copper requirements of the ammonia-oxidizing archaeon Nitrosopumilus maritimus
SCM1 and implications for nitrification in the marine environment. Limnology and
Oceanography, 58(6), 2037–2045. https://doi.org/10.4319/lo.2013.58.6.2037
4. Anschutz, P., Sundby, B., Lefrançois, L., Luther, G. W., & Mucci, A. (2000). Interactions
between metal oxides and species of nitrogen and iodine in bioturbated marine sediments.
Geochimica et Cosmochimica Acta, 64(16), 2751–2763. https://doi.org/10.1016/S0016-
7037(00)00400-2
5. Babbin, A. R., Peters, B. D., Mordy, C. W., Widner, B., Casciotti, K. L., & Ward, B. B.
(2017). Multiple metabolisms constrain the anaerobic nitrite budget in the Eastern Tropical
South Pacific: Nitrogen Dynamics in the Eastern Tropical South Pacific. Global
Biogeochemical Cycles. https://doi.org/10.1002/2016GB005407
6. Baconnais, I., Rouxel, O., Dulaquais, G., & Boye, M. (2019). Determination of the copper
isotope composition of seawater revisited: A case study from the Mediterranean Sea.
Chemical Geology, 511, 465–480. https://doi.org/10.1016/j.chemgeo.2018.09.009
7. Bluhm, K., Croot, P., Wuttig, K., & Lochte, K. (2010). Transformation of iodate to iodide in
marine phytoplankton driven by cell senescence. Aquatic Biology, 11(1), 1–15.
https://doi.org/10.3354/ab00284
8. Boyle, E. A., Sclater, F. R., & Edmond, J. M. (1977). The distribution of dissolved copper in
the Pacific. Earth and Planetary Science Letters, 37(1), 38–54.
https://doi.org/10.1016/0012-821X(77)90144-3
9. Bowles, M. W., Mogollón, J. M., Kasten, S., Zabel, M., & Hinrichs, K.-U. (2014). Global
rates of marine sulfate reduction and implications for sub–sea-floor metabolic activities.
Science, 344(6186), 889–891. https://doi.org/10.1126/science.1249213
10. Brand, L. E., Sunda, W. G., & Guillard, R. R. L. (1986). Reduction of marine phytoplankton
reproduction rates by copper and cadmium. Journal of Experimental Marine Biology and
Ecology, 96(3), 225–250. https://doi.org/10.1016/0022-0981(86)90205-4
11. Bruland, K. W., & Lohan, M. C. (2003). Controls of Trace Metals in Seawater, 23-47,
Pergamon, Oxford.
12. Buck, K. N., & Bruland, K. W. (2005). Copper speciation in San Francisco Bay: A novel
approach using multiple analytical windows. Marine Chemistry, 96(1), 185–198.
https://doi.org/10.1016/j.marchem.2005.01.001
128

13. Bundy, R. M., Barbeau, K. A., & Buck, K. N. (2013). Sources of strong copper-binding
ligands in Antarctic Peninsula surface waters. Deep Sea Research Part II: Topical Studies in
Oceanography, 90, 134–146. https://doi.org/10.1016/j.dsr2.2012.07.023
14. Butler, E. C. V., & Smith, J. D. (1980). Iodine speciation in seawaters—the analytical use of
ultra-violet photo-oxidation and differential pulse polarography. Deep Sea Research Part A.
Oceanographic Research Papers, 27(6), 489–493. https://doi.org/10.1016/0198-
0149(80)90058-8
15. Byrne, R. H., Miller, W. L. (1985). Copper (II) carbonate complexation in seawater.
Geochemica et Cosmochemica Acta, 49(8), 1837-1844. https://doi.org/10.1016/0016-
7037(85)90153-X
16. Campos, M. L., A., & van den Berg, C. M. (1994). Determination of copper complexation
in sea water by cathodic stripping voltammetry and ligand competition with
salicylaldoxime. Analytica Chimica Acta, 284(3), 481–496. https://doi.org/10.1016/0003-
2670(94)85055-0
17. Campos, M. L. A. M. (1997). New approach to evaluating dissolved iodine speciation in
natural waters using cathodic stripping voltammetry and a storage study for preserving
iodine species. Marine Chemistry, 57(1), 107–117. https://doi.org/10.1016/S0304-
4203(96)00093-X
18. Carpenter, L. J., MacDonald, S. M., Shaw, M. D., Kumar, R., Saunders, R. W., Parthipan,
R., et al. (2013). Atmospheric iodine levels influenced by sea surface emissions of inorganic
iodine. Nature Geoscience, 6(2), 108–111. https://doi.org/10.1038/ngeo1687
19. Carpenter, L. J., Chance, R. J., Sherwen, T., Adams, T. J., Ball, S. M., Evans, M. J., et al.
(2021). Marine iodine emissions in a changing world. Proceedings of the Royal Society A:
Mathematical, Physical and Engineering Sciences, 477(2247), 20200824.
https://doi.org/10.1098/rspa.2020.0824
20. Chance, R., Baker, A. R., Carpenter, L., & Jickells, T. D. (2014). The distribution of iodide
at the sea surface. Environ. Sci.: Processes Impacts, 16(8), 1841–1859.
https://doi.org/10.1039/C4EM00139G
21. Chance, R. J., Tinel, L., Sherwen, T., Baker, A. R., Bell, T., Brindle, J., et al. (2019). Global
sea-surface iodide observations, 1967–2018. Scientific Data, 6(1), 286.
https://doi.org/10.1038/s41597-019-0288-y
22. Chance, R., Tinel, L., Sarkar, A., Sinha, A. K., Mahajan, A. S., Chacko, R., et al. (2020).
Surface Inorganic Iodine Speciation in the Indian and Southern Oceans From 12°N to 70°S.
Frontiers in Marine Science, 7, 621. https://doi.org/10.3389/fmars.2020.00621
23. Chapman, P. (1983). Changes in iodine speciation in the Benguela current upwelling
system. Deep Sea Research Part A. Oceanographic Research Papers, 30(12), 1247-1259.
https://doi.org/10.1016/0198-0149(83)90083-3
24. Coale, K. H., & Bruland, K. W. (1988). Copper complexation in the Northeast Pacific: NE
Pacific copper complexation. Limnology and Oceanography, 33(5), 1084–1101.
https://doi.org/10.4319/lo.1988.33.5.1084
129

25. Cutter, G., Casciotti, K., Croot, P., Geibert, W., Heimbürger, L.-E., Lohan, M., et al. (2017).
Sampling and Sample-handling Protocols for GEOTRACES Cruises. Version 3, August
2017. 139pp. & Appendices. http://dx.doi.org/10.25607/OBP-2
26. Cutter, G. A., Moffett, J. W., Nielsdóttir, M. C., & Sanial, V. (2018). Multiple oxidation
state trace elements in suboxic waters off Peru: In situ redox processes and
advective/diffusive horizontal transport. Marine Chemistry, 201, 77–89.
https://doi.org/10.1016/j.marchem.2018.01.003
27. Coale, K. H., & Bruland, K. W. (1988). Copper complexation in the Northeast Pacific: NE
Pacific copper complexation. Limnology and Oceanography, 33(5), 1084–1101.
https://doi.org/10.4319/lo.1988.33.5.1084
28. Dahl, I. (1968). The application of salicylaldoxime in solvent extraction. Analytica Chimica
Acta, 41, 9–14. https://doi.org/10.1016/S0003-2670(01)80354-8
29. Dahm, C. N., Gregory, S. V., & Kilho Park, P. (1981). Organic carbon transport in the
Columbia River. Estuarine, Coastal and Shelf Science, 13(6), 645–658.
https://doi.org/10.1016/S0302-3524(81)80046-1
30. Davis, J. A., & Leckie, J. O. (1978). Effect of adsorbed complexing ligands on trace metal
uptake by hydrous oxides. Environmental Science & Technology, 12(12), 1309–1315.
https://doi.org/10.1021/es60147a006
31. de laCuesta, J. L., & Manley, S. L. (2009). Iodine assimilation by marine diatoms and other
phytoplankton in nitrate-replete conditions. Limnology and Oceanography, 54(5), 1653–
1664. https://doi.org/10.4319/lo.2009.54.5.1653
32. Emerson, S., Cranston, R. E., & Liss, P. S. (1979). Redox species in a reducing fjord:
equilibrium and kinetic considerations. Deep Sea Research Part A. Oceanographic
Research Papers, 26(8), 859-878. https://doi.org/10.1016/0198-0149(79)90101-8
33. Evans, N., Boles, E., Kwiecinski, J. V., Mullen, S., Wolf, M., Devol, A. H., et al. (2020).
The role of water masses in shaping the distribution of redox active compounds in the
Eastern Tropical North Pacific oxygen deficient zone and influencing low oxygen
concentrations in the eastern Pacific Ocean. Limnology and Oceanography, 65(8), 1688–
1705. https://doi.org/10.1002/lno.11412
34. Farrenkopf, A. M., Dollhopf, M. E., Chadhain, S. N., Luther III, G. W., & Nealson, K. H.
(1997). Reduction of iodate in seawater during Arabian Sea shipboard incubations and in
laboratory cultures of the marine bacterium Shewanella putrefaciens strain MR-4. Marine
Chemistry, 57(3–4), 347–354. https://doi.org/10.1016/S0304-4203(97)00039-X
35. Farrenkopf, A. M., & Luther III, G. W. (2002). Iodine chemistry reflects productivity and
denitrification in the Arabian Sea: evidence for flux of dissolved species from sediments of
western India into the OMZ. Deep Sea Research Part II: Topical Studies in Oceanography,
49(12), 2303–2318. https://doi.org/10.1016/S0967-0645(02)00038-3
36. Fiedler, P. C. & Talley, L. D. Hydrography of the eastern tropical Pacific: A review.
Progress in Oceanography 69, 143–180 (2006).
https://doi.org/10.1016/j.pocean.2006.03.008
130

37. Fox, P. M., Davis, J. A., & Luther, G. W. (2009). The kinetics of iodide oxidation by the
manganese oxide mineral birnessite. Geochimica et Cosmochimica Acta, 73(10), 2850–
2861. https://doi.org/10.1016/j.gca.2009.02.016
38. François, R. (1987). The influence of humic substances on the geochemistry of iodine in
nearshore and hemipelagic marine sediments. Geochimica et Cosmochimica Acta, 51(9),
2417–2427. https://doi.org/10.1016/0016-7037(87)90294-8
39. Frenger, I., Bianchi, D., Stührenberg, C., Oschlies, A., Dunne, J., Deutsch, C., et al. (2018).
Biogeochemical Role of Subsurface Coherent Eddies in the Ocean: Tracer Cannonballs,
Hypoxic Storms, and Microbial Stewpots? Global Biogeochemical Cycles, 32(2), 226–249.
https://doi.org/10.1002/2017GB005743
40. Gerringa, L. J. A., Herman, P. M. J., & Poortvliet, T. C. W. (1995). Comparison of the
linear Van den Berg/ Ružić transformation and a non-linear fit of the Langmuir isotherm
applied to Cu speciation data in the estuarine environment. Marine Chemistry, 48(2), 131-
142. https://doi.org/10.1016/0304-4203(94)00041-B
41. Granger, J., & Ward, B. B. (2003). Accumulation of nitrogen oxides in copper-limited
cultures of denitrifying bacteria. Limnology and Oceanography, 48(1), 313–318.
https://doi.org/10.4319/lo.2003.48.1.0313
42. Hardisty, D. S., Lu, Z., Planavsky, N. J., Bekker, A., Philippot, P., Zhou, X., & Lyons, T.
W. (2014). An iodine record of Paleoproterozoic surface ocean oxygenation. Geology,
42(7), 619–622. https://doi.org/10.1130/G35439.1
43. Hardisty, D. S., Lu, Z., Bekker, A., Diamond, C. W., Gill, B. C., Jiang, G., et al. (2017).
Perspectives on Proterozoic surface ocean redox from iodine contents in ancient and recent
carbonate. Earth and Planetary Science Letters, 463, 159–170.
https://doi.org/10.1016/j.epsl.2017.01.032
44. Hepach, H., Hughes, C., Hogg, K., Collings, S., & Chance, R. (2020). Senescence as the
main driver of iodide release from a diverse range of marine phytoplankton. Biogeosciences,
17(9), 2453–2471. https://doi.org/10.5194/bg-17-2453-2020
45. Hetzel, B. (1983). IODINE DEFICIENCY DISORDERS (IDD) AND THEIR
ERADICATION. The Lancet, 322(8359), 1126–1129. https://doi.org/10.1016/S0140-
6736(83)90636-0
46. Hoogakker, B. A. A., Lu, Z., Umling, N., Jones, L., Zhou, X., Rickaby, R. E. M., et al.
(2018). Glacial expansion of oxygen-depleted seawater in the eastern tropical Pacific.
Nature, 562(7727), 410–413. https://doi.org/10.1038/s41586-018-0589-x
47. Hughes, C., Barton, E., Hepach, H., Chance, R., Pickering, M. D., Hogg, K., et al. (2021).
Oxidation of iodide to iodate by cultures of marine ammonia-oxidising bacteria. Marine
Chemistry, 234, 104000. https://doi.org/10.1016/j.marchem.2021.104000
48. Irving, H., & Williams, R. J. P. (1953). 637. The stability of transition-metal complexes.
Journal of the Chemical Society (Resumed), (0), 3192–3210.
https://doi.org/10.1039/JR9530003192
131

49. Jacquot, J. E., Horak, R. E., Amin, S. A., Devol, A. H., Ingalls, A. E., Armbrust, E. V., et al.
(2014). Assessment of the potential for copper limitation of ammonia oxidation by Archaea
in a dynamic estuary. Marine Chemistry, 162, 37–49.
https://doi.org/10.1016/j.marchem.2014.02.002
50. Jacquot, J. E., & Moffett, J. W. (2015). Copper distribution and speciation across the
International GEOTRACES Section GA03. Deep Sea Research Part II: Topical Studies in
Oceanography, 116, 187–207. https://doi.org/10.1016/j.dsr2.2014.11.013
51. Jensen, L. T., Wyatt, N. J., Landing, W. M., & Fitzsimmons, J. N. (2020). Assessment of the
stability, sorption, and exchangeability of marine dissolved and colloidal metals. Marine
Chemistry, 220, 103754. https://doi.org/10.1016/j.marchem.2020.103754
52. John, S. G., Liang, H., Weber, T., DeVries, T., Primeau, F., Moore, K., et al. (2020).
AWESOME OCIM: A simple, flexible, and powerful tool for modeling elemental cycling in
the oceans. Chemical Geology, 533, 119403.
https://doi.org/10.1016/j.chemgeo.2019.119403
53. Jones, C. E., Hornsby, K. E., Sommariva, R., Dunk, R. M., Glasow, R. von, McFiggans, G.,
& Carpenter, L. J. (2010). Quantifying the contribution of marine organic gases to
atmospheric iodine. Geophysical Research Letters, 37(18).
https://doi.org/10.1029/2010GL043990
54. Kadko, D. (2017). Upwelling and primary production during the U.S. GEOTRACES East
Pacific Zonal Transect. Global Biogeochemical Cycles, 31(2), 218–232.
https://doi.org/10.1002/2016GB005554
55. Kennedy, H. A., & Elderfield, H. (1987a). Iodine diagenesis in non-pelagic deep-sea
sediments. Geochimica et Cosmochimica Acta, 51(9), 2505–2514.
https://doi.org/10.1016/0016-7037(87)90301-2
56. Kennedy, H. A., & Elderfield, H. (1987b). Iodine diagenesis in pelagic deep-sea sediments.
Geochimica et Cosmochimica Acta, 51(9), 2489–2504. https://doi.org/10.1016/0016-
7037(87)90300-0
57. Klinkhammer, G. P. (1980). Early diagenesis in sediments from the eastern equatorial
Pacific, II. Pore water metal results. Earth and Planetary Science Letters, 49(1), 81–101.
https://doi.org/10.1016/0012-821X(80)90151-X
58. Kogut, M. B., & Voelker, B. M. (2001). Strong Copper-Binding Behavior of Terrestrial
Humic Substances in Seawater. Environmental Science & Technology, 35(6), 1149–1156.
https://doi.org/10.1021/es0014584
59. Kogut, M. B., & Voelker, B. M. (2003). Kinetically Inert Cu in Coastal Waters.
Environmental Science & Technology, 37(3), 509–518. https://doi.org/10.1021/es020723d
60. Laglera, L. M., & van den Berg, C. M. G. (2003). Copper complexation by thiol compounds
in estuarine waters. Marine Chemistry, 82(1), 71–89. https://doi.org/10.1016/S0304-
4203(03)00053-7
132

61. Laglera, L. M., & van den Berg, C. M. G. (2006). Photochemical oxidation of thiols and
copper complexing ligands in estuarine waters. Marine Chemistry, 101(1–2), 130–140.
https://doi.org/10.1016/j.marchem.2006.01.006
62. Lawrence, R. M., Shrikumar, A., Le Roy, E., Swift, J. H., Lam, P. J., Cutter, G., &
Casciotti, K. L. (2022). Water mass analysis of the 2018 US GEOTRACES Pacific
Meridional Transect (GP15) (preprint). Oceanography.
https://doi.org/10.1002/essoar.10510438.1
63. Leeuwen, H. P. V., & Town, R. M. (2005). Kinetic Limitations in Measuring Stabilities of
Metal Complexes by Competitive Ligand Exchange-Adsorptive Stripping Voltammetry
(CLE-AdSV). Environmental Science & Technology. 39(18), 7217-7225.
https://doi.org/10.1021/es050367%2B
64. Little, S. H., Vance, D., Siddall, M., & Gasson, E. (2013). A modeling assessment of the
role of reversible scavenging in controlling oceanic dissolved Cu and Zn distributions:
MODELING OCEANIC CU AND ZN. Global Biogeochemical Cycles, 27(3), 780–791.
https://doi.org/10.1002/gbc.20073
65. Little, S. H., Vance, D., Walker-Brown, C., & Landing, W. M. (2014). The oceanic mass
balance of copper and zinc isotopes, investigated by analysis of their inputs, and outputs to
ferromanganese oxide sediments. Geochimica et Cosmochimica Acta, 125, 673–693.
https://doi.org/10.1016/j.gca.2013.07.046
66. Long, A., Dang, A., Xiao, H., & Yu, X. (2015).The Summer Distribution of Dissolved
Inorganic Iodine along 18 N in the South China Sea. Journal of Marine Science: Research
& Development, 05(03). https://doi.org/10.4172/2155-9910.1000169
67. Lovelock, J. E., Maggs, R. J., & Wade, R. J. (1973). Halogenated Hydrocarbons in and over
the Atlantic. Nature, 241(5386), 194–196. https://doi.org/10.1038/241194a0
68. Lu, Z., Hensen, C., Fehn, U., & Wallmann, K. (2008). Halogen and 129I systematics in gas
hydrate fields at the northern Cascadia margin (IODP Expedition 311): Insights from
numerical modeling. Geochemistry, Geophysics, Geosystems, 9(10).
https://doi.org/10.1029/2008GC002156
69. Lu, Z., Jenkyns, H. C., & Rickaby, R. E. (2010). Iodine to calcium ratios in marine
carbonate as a paleo-redox proxy during oceanic anoxic events. Geology, 38(12), 1107–
1110. https://doi.org/10.1130/G31145.1
70. Lu, Z., Hoogakker, B. A. A., Hillenbrand, C.-D., Zhou, X., Thomas, E., Gutchess, K. M., et
al. (2016). Oxygen depletion recorded in upper waters of the glacial Southern Ocean. Nature
Communications, 7, 11146. https://doi.org/10.1038/ncomms11146
71. Lu, W., Wörndle, S., Halverson, G. P., Zhou, X., Bekker, A., Rainbird, R. H., et al. (2017).
Iodine proxy evidence for increased ocean oxygenation during the Bitter Springs Anomaly.
Geochemical Perspectives Letters, 53–57. https://doi.org/10.7185/geochemlet.1746
72. Luther, G. W., Swartz, C. B., & Ullman, W. J. (1988). Direct determination of iodide in
seawater by cathodic stripping square wave voltammetry. Analytical Chemistry, 60(17),
1721–1724. https://doi.org/10.1021/ac00168a017
133

73. Luther, G. W., & Campbell, T. (1991). Iodine speciation in the water column of the Black
Sea. Deep Sea Research Part A. Oceanographic Research Papers, 38, S875–S882.
https://doi.org/10.1016/S0198-0149(10)80014-7
74. Luther, G. W., Wu, J., & Cullen, J. B. (1995). Redox Chemistry of Iodine in Seawater:
Frontier Molecular Orbital Theory Considerations. In C. P. Huang, C. R. O’Melia, & J. J.
Morgan (Eds.), Aquatic Chemistry (Vol. 244, pp. 135–155). Washington, DC: American
Chemical Society. https://doi.org/10.1021/ba-1995-0244.ch006
75. Luther, G. W., & Rickard, D. T. (2005). Metal Sulfide Cluster Complexes and their
Biogeochemical Importance in the Environment. Journal of Nanoparticle Research, 7(4–5),
389–407. https://doi.org/10.1007/s11051-005-4272-4
76. Ma, Y. B., Lombi, E., Nolan, A. L., & McLaughlin, M. J. (2006). Determination of labile
Cu in soils and isotopic exchangeability of colloidal Cu complexes. European Journal of
Soil Science, 57(2), 147–153. https://doi.org/10.1111/j.1365-2389.2005.00723.x
77. Maldonado, M. T., Allen, A. E., Chong, J. S., Lin, K., Leus, D., Karpenko, N., & Harris, S.
L. (2006). Copper-dependent iron transport in coastal and oceanic diatoms. Limnology and
Oceanography, 51(4), 1729–1743. https://doi.org/10.4319/lo.2006.51.4.1729
78. Mann, E. L., Ahlgren, N., Moffett, J. W., & Chisholm, S. W. (2002). Copper toxicity and
cyanobacteria ecology in the Sargasso Sea. Limnology and Oceanography, 47(4), 976–988.
https://doi.org/10.4319/lo.2002.47.4.0976
79. Martiny, A. C., Vrugt, J. A., & Lomas, M. W. (2014). Concentrations and ratios of
particulate organic carbon, nitrogen, and phosphorus in the global ocean. Scientific Data, 1,
140048. https://doi.org/10.1038/sdata.2014.48
80. Merbt, S. N., Stahl, D. A., Casamayor, E. O., Martí, E., Nicol, G. W., & Prosser, J. I.
(2012). Differential photoinhibition of bacterial and archaeal ammonia oxidation. FEMS
Microbiology Letters, 327(1), 41–46. https://doi.org/10.1111/j.1574-6968.2011.02457.x
81. Moffett, J.W., Zika, R.G., Petasne, R.G. (1985). Evaluation of bathocuproine for the
spectro-photometric determination of copper(I) in copper redox studies with applications in
studies of natural waters. Analytica Chimica Acta, 175, 171–179.
https://doi.org/10.1016/S0003-2670(00)82729-4
82. Moffett, J. W., & Zika, R. G. (1987a). Solvent extraction of copper acetylacetonate in
studies of copper (II) speciation in seawater. Marine Chemistry, 21(4), 301–313.
https://doi.org/10.1016/0304-4203(87)90053-3
83. Moffett, J. W., & Zika, R. G. (1987). Photochemistry of Copper Complexes in Sea Water. In
Photochemistry of Environmental Aquatic Systems (Vol. 327, pp. 116–130). American
Chemical Society. https://doi.org/10.1021/bk-1987-0327.ch009
84. Moffett, J.W., Zika, R.G. (1988). Measurement of copper(I) in surface waters of the
subtropical Atlantic and Gulf of Mexico. Geochimica et Cosmochimica Acta, 52, 1849–
1857. https://doi.org/10.1016/0016-7037(88)90008-7
134

85. Moffett, J. W., Zika, R. G., & Brand, L. E. (1990). Distribution and potential sources and
sinks of copper chelators in the Sargasso Sea. Deep Sea Research Part A. Oceanographic
Research Papers, 37(1), 27–36. https://doi.org/10.1016/0198-0149(90)90027-S
86. Moffett, J. W. (1995). Temporal and spatial variability of copper complexation by strong
chelators in the Sargasso Sea. Deep Sea Research Part I: Oceanographic Research Papers,
42(8), 1273–1295. https://doi.org/10.1016/0967-0637(95)00060-J
87. Moffett, J. W., & Brand, L. E. (1996). Production of strong, extracellular Cu chelators by
marine cyanobacteria in response to Cu stress. Limnology and Oceanography, 41(3), 388-
395. https://doi.org/10.4319/lo.1996.41.3.0388
88. Moffett, J. W., Brand, L. E., Croot, P. L., & Barbeau, K. A. (1997). Cu speciation and
cyanobacterial distribution in harbors subject to anthropogenic Cu inputs. Limnology and
Oceanography, 42(5), 789–799. https://doi.org/10.4319/lo.1997.42.5.0789
89. Moffett, J. W., & Dupont, C. (2007). Cu complexation by organic ligands in the sub-arctic
NW Pacific and Bering Sea. Deep Sea Research Part I: Oceanographic Research Papers,
54(4), 586–595. https://doi.org/10.1016/j.dsr.2006.12.013
90. Moore, R. M., & Tokarczyk, R. (1992). Chloro-iodomethane in N. Atlantic waters: A
potentially significant source of atmospheric iodine. Geophysical Research Letters, 19(17),
1779–1782. https://doi.org/10.1029/92GL01796
91. Moriyasu, R., Evans, N., Bolster, K. M., Hardisty, D. S., & Moffett, J. W. (2020). The
Distribution and Redox Speciation of Iodine in the Eastern Tropical North Pacific Ocean.
Global Biogeochemical Cycles, 34(2). https://doi.org/10.1029/2019GB006302
92. Moriyasu, R., & Moffett, J. W. (2022). Determination of inert and labile copper on
GEOTRACES samples using a novel solvent extraction method. Marine Chemistry,
104073. https://doi.org/10.1016/j.marchem.2021.104073
93. Muller, F. L. L., & Batchelli, S. (2013). Copper binding by terrestrial versus marine organic
ligands in the coastal plume of River Thurso, North Scotland. Estuarine, Coastal and Shelf
Science, 133, 137–146. https://doi.org/10.1016/j.ecss.2013.08.024
94. Nakayama, E., Kimoto, T., Isshiki, K., Sohrin, Y., & Okazaki, S. (1989). Determination and
distribution of iodide- and total-iodine in the North Pacific Ocean - by using a new
automated electrochemical method. Marine Chemistry, 27(1–2), 105–116.
https://doi.org/10.1016/0304-4203(89)90030-3
95. Nürnberg, D., Böschen, T., Doering, K., Mollier‐Vogel, E., Raddatz, J., & Schneider, R.
(2015). Sea surface and subsurface circulation dynamics off equatorial Peru during the last
17 kyr. Paleoceanography, 30(7), 984–999. https://doi.org/10.1002/2014PA002706
96. Orians, K. J., & Bruland, K. W. (1985). Dissolved aluminium in the central North Pacific.
Nature, 316(6027), 427–429. https://doi.org/10.1038/316427a0
97. Peers, G., & Price, N. M. (2006). Copper-containing plastocyanin used for electron transport
by an oceanic diatom. Nature, 441(7091), 341–344. https://doi.org/10.1038/nature04630
135

98. Peters, B. D., Jenkins, W. J., Swift, J. H., German, C. R., Moffett, J. W., Cutter, G. A., et al.
(2018). Water mass analysis of the 2013 US GEOTRACES eastern Pacific zonal transect
(GP16). Marine Chemistry, 201, 6–19. https://doi.org/10.1016/j.marchem.2017.09.007
99. Pižeta, I., Sander, S. G., Hudson, R. J. M., Omanović, D., Baars, O., Barbeau, K. A., et al.
(2015). Interpretation of complexometric titration data: An intercomparison of methods for
estimating models of trace metal complexation by natural organic ligands. Marine
Chemistry, 173, 3–24. https://doi.org/10.1016/j.marchem.2015.03.006
100. Prados-Roman, C., Cuevas, C. A., Hay, T., Fernandez, R. P., Mahajan, A. S., Royer, S.-J., et
al. (2015). Iodine oxide in the global marine boundary layer. Atmospheric Chemistry and
Physics, 15(2), 583–593. https://doi.org/10.5194/acp-15-583-2015
101. Price, N. B., & Calvert, S. E. (1973). The geochemistry of iodine in oxidised and reduced
recent marine sediments. Geochimica et Cosmochimica Acta, 37(9), 2149–2158.
https://doi.org/10.1016/0016-7037(73)90013-6
102. Price, N. B., & Calvert, S. E. (1977). The contrasting geochemical behaviours of iodine and
bromine in recent sediments from the Namibian shelf. Geochimica et Cosmochimica Acta,
41(12), 1769–1775. https://doi.org/10.1016/0016-7037(77)90209-5
103. Rapp, I., Schlosser, C., Rusiecka, D., Gledhill, M., & Achterberg, E. P. (2017). Automated
preconcentration of Fe, Zn, Cu, Ni, Cd, Pb, Co, and Mn in seawater with analysis using
high-resolution sector field inductively-coupled plasma mass spectrometry. Analytica
Chimica Acta, 976, 1–13. https://doi.org/10.1016/j.aca.2017.05.008
104. Reyes-Umana, V., Henning, Z., Lee, K., Barnum, T. P., & Coates, J. D. (2021). Genetic and
phylogenetic analysis of dissimilatory iodate-reducing bacteria identifies potential niches
across the world’s oceans. The ISME Journal. https://doi.org/10.1038/s41396-021-01034-5
105. Richon, C., & Tagliabue, A. (2019). Insights Into the Major Processes Driving the Global
Distribution of Copper in the Ocean From a Global Model. Global Biogeochemical Cycles,
33(12), 1594–1610. https://doi.org/10.1029/2019GB006280
106. Roshan, S., & Wu, J. (2015). The distribution of dissolved copper in the tropical-subtropical
north Atlantic across the GEOTRACES GA03 transect. Marine Chemistry, 176, 189–198.
https://doi.org/10.1016/j.marchem.2015.09.006
107. Ruacho, A., Bundy, R. M., Till, C. P., Roshan, S., Wu, J., & Barbeau, K. A. (2020). Organic
dissolved copper speciation across the U.S. GEOTRACES equatorial Pacific zonal transect
GP16. Marine Chemistry, 225, 103841. https://doi.org/10.1016/j.marchem.2020.103841
108. Rue, E. L., Smith, G. J., Cutter, G. A., & Bruland, K. W. (1997). The response of trace
element redox couples to suboxic conditions in the water column. Deep Sea Research Part
I: Oceanographic Research Papers, 44(1), 113–134. https://doi.org/10.1016/S0967-
0637(96)00088-X
109. Saiz-Lopez, A., Plane, J. M. C., Baker, A. R., Carpenter, L. J., von Glasow, R., Gómez
Martín, J. C., et al. (2012). Atmospheric Chemistry of Iodine. Chemical Reviews, 112(3),
1773–1804. https://doi.org/10.1021/cr200029u
136

110. Sato, M., Ogata, N., Wong, K. H., Obata, H., & Takeda, S. (2021). Photodecomposition of
natural organic metal-binding ligands from deep seawater. Marine Chemistry, 230, 103939.
https://doi.org/10.1016/j.marchem.2021.103939
111. Saiz-Lopez, A., Plane, J. M. C., Baker, A. R., Carpenter, L. J., von Glasow, R., Gómez
Martín, J. C., et al. (2012). Atmospheric Chemistry of Iodine. Chemical Reviews, 112(3),
1773–1804. https://doi.org/10.1021/cr200029u
112. Sherwen, T., Schmidt, J. A., Evans, M. J., Carpenter, L. J., Großmann, K., Eastham, S. D.,
et al. (2016). Global impacts of tropospheric halogens (Cl, Br, I) on oxidants and
composition in GEOS-Chem. Atmospheric Chemistry and Physics, 16(18), 12239–12271.
https://doi.org/10.5194/acp-16-12239-2016
113. Sunda, W. G., & Lewis, J. A. M. (1978). Effect of complexation by natural organic ligands
on the toxicity of copper to a unicellular alga, Monochrysis lutheri1. Limnology and
Oceanography, 23(5), 870–876. https://doi.org/10.4319/lo.1978.23.5.0870
114. Sunda, W. G., & Huntsman, S. A. (1983). Effect of competitive interactions between
manganese and copper on cellular manganese and growth in estuarine and oceanic species
of the diatom Thalassiosira1,2. Limnology and Oceanography, 28(5), 924–934.
https://doi.org/10.4319/lo.1983.28.5.0924
115. Sunda, W. G., & Huntsman, S. A. (1995). Regulation of copper concentration in the oceanic
nutricline by phytoplankton uptake and regeneration cycles. Limnology and Oceanography,
40(1), 132–137. https://doi.org/10.4319/lo.1995.40.1.0132
116. Sunda, W. G., & Huntsman, S. A. (1998). Interactions among Cu
2+
, Zn
2+
, and Mn
2+
in
controlling cellular Mn, Zn, and growth rate in the coastal alga Chlamydomonas. Limnology
and Oceanography, 43(6), 1055–1064.
117. Thomson, R. E., & Krassovski, M. V. (2010). Poleward reach of the California
Undercurrent extension. Journal of Geophysical Research, 115(C9).
https://doi.org/10.1029/2010JC006280
118. Tian, R. C., & Nicolas, E. (1995). Iodine speciation in the northwestern Mediterranean Sea,
method and vertical profile. Marine Chemistry, 48(2), 151–156.
https://doi.org/10.1016/0304-4203(94)00048-I
119. Truesdale, V. W., Watts, S. F., & Rendell, A. R. (2001). On the possibility of iodide
oxidation in the near-surface of the Black Sea and its implications to iodine in the general
ocean. Deep Sea Research Part I: Oceanographic Research Papers, 48(11), 2397–2412.
https://doi.org/10.1016/S0967-0637(01)00021-8
120. Tsunogai, S., & Sase, T. (1969). Formation of iodide-iodine in the ocean. Deep Sea
Research and Oceanographic Abstracts, 16(5), 489–496. https://doi.org/10.1016/0011-
7471(69)90037-0
121. Turner, J. S., Pretty, J. L., & McDonnell, A. M. P. (2017). Marine particles in the Gulf of
Alaska shelf system: Spatial patterns and size distributions from in situ optics. Continental
Shelf Research, 145, 13–20. https://doi.org/10.1016/j.csr.2017.07.002
137

122. Twiss, M. R., & Moffett, J. W. (2002). Comparison of Copper Speciation in Coastal Marine
Waters Measured Using Analytical Voltammetry and Diffusion Gradient in Thin-Film
Techniques. Environmental Science & Technology, 36(5), 1061–1068.
https://doi.org/10.1021/es0016553
123. Ullman, W. J., & Aller, R. C. (1985). The geochemistry of iodine in near-shore carbonate
sediments. Geochimica et Cosmochimica Acta, 49(4), 967–978.
https://doi.org/10.1016/0016-7037(85)90311-4
124. Ullman, W. J., Luther III, G. W., De Lange, G. J., & Woittiez, J. R. W. (1990). Iodine
chemistry in deep anoxic basins and overlying waters of the Mediterranean Sea. Marine
Chemistry, 31(1-3), 153-170. https://doi.org/10.1016/0304-4203(90)90036-C
125. van Bergeijk, S. A., Hernández, L., Zubía, E., & Cañavate, J. P. (2016). Iodine balance,
growth and biochemical composition of three marine microalgae cultured under various
inorganic iodine concentrations. Marine Biology, 163(5). https://doi.org/10.1007/s00227-
016-2884-0
126. van den Berg, C. M. G., Merks, A. G. A., & Duursma, E. K. (1987). Organic complexation
and its control of the dissolved concentrations of copper and zinc in the Scheldt estuary.
Estuarine, Coastal and Shelf Science, 24(6), 785–797. https://doi.org/10.1016/0272-
7714(87)90152-1
127. van den Berg, C. M. G., & Donat, J. R. (1992). Determination and data evaluation of copper
complexation by organic ligands in sea water using cathodic stripping voltammetry at
varying detection windows. Analytica Chimica Acta, 257(2), 281–291.
https://doi.org/10.1016/0003-2670(92)85181-5
128. Wadley, M. R., Stevens, D. P., Jickells, T. D., Hughes, C., Chance, R., Hepach, H., et al.
(2020). A Global Model for Iodine Speciation in the Upper Ocean. Global Biogeochemical
Cycles, 34(9), e2019GB006467. https://doi.org/10.1029/2019GB006467
129. Walker, C. B., Torre, J. R. de la, Klotz, M. G., Urakawa, H., Pinel, N., Arp, D. J., et al.
(2010). Nitrosopumilus maritimus genome reveals unique mechanisms for nitrification and
autotrophy in globally distributed marine crenarchaea. Proceedings of the National
Academy of Sciences, 107(19), 8818–8823. https://doi.org/10.1073/pnas.0913533107
130. Wang, F., Chen, Y., Meng, X., Fu, J., & Wang, B. (2016). The contribution of
anthropogenic sources to the aerosols over East China Sea. Atmospheric Environment, 127,
22–33. https://doi.org/10.1016/j.atmosenv.2015.12.002
131. Ward, B. B., Tuit, C. B., Jayakumar, A., Rich, J. J., Moffett, J., & Naqvi, S. W. A. (2008).
Organic carbon, and not copper, controls denitrification in oxygen minimum zones of the
ocean. Deep Sea Research Part I: Oceanographic Research Papers, 55(12), 1672–1683.
https://doi.org/10.1016/j.dsr.2008.07.005
132. Whitby, H., & van den Berg, C. M. G. (2015). Evidence for copper-binding humic
substances in seawater. Marine Chemistry, 173, 282–290.
https://doi.org/10.1016/j.marchem.2014.09.011
138

133. Whitby, H., Posacka, A. M., Maldonado, M. T., & van den Berg, C. M. G. (2018). Copper-
binding ligands in the NE Pacific. Marine Chemistry, 204, 36–48.
https://doi.org/10.1016/j.marchem.2018.05.008
134. Wong, G. T. F., & Brewer, P. G. (1974). The determination and distribution of iodate in
South Atlantic waters. Journal of Marine Research, 32, 25-36.
135. Wong, G. T. F., Brewer, P. G., & Spencer, D. W. (1976). The distribution of particulate
iodine in the Atlantic Ocean. Earth and Planetary Science Letters, 32(2), 441–450.
https://doi.org/10.1016/0012-821X(76)90084-4
136. Wong, G. T. F., & Brewer, P. G. (1977). The marine chemistry of iodine in anoxic basins.
Geochimica et Cosmochimica Acta, 41(1), 151–159. https://doi.org/10.1016/0016-
7037(77)90195-8
137. Wong, G. T., Takayanagi, K., & Todd, J. F. (1985). Dissolved iodine in waters overlying
and in the Orca Basin, Gulf of Mexico. Marine Chemistry, 17(2), 177-183.
https://doi.org/10.1016/0304-4203(85)90072-6
138. Wong, G. T. F. (2001). Coupling iodine speciation to primary, regenerated or new
production: a re-evaluation, 18. https://doi.org/10.1016/S0967-0637(00)00097-2
139. Wong, G. T., & Zhang, L. S. (2003). Seasonal variations in the speciation of dissolved
iodine in the Chesapeake Bay. Estuarine, Coastal and Shelf Science, 56(5-6), 1093-1106.
https://doi.org/10.1016/S0272-7714(02)00310-4
140. Yeager, C. M., Amachi, S., Grandbois, R., Kaplan, D. I., Xu, C., Schwehr, K. A., &
Santschi, P. H. (2017). Microbial Transformation of Iodine: From Radioisotopes to Iodine
Deficiency. In Advances in Applied Microbiology (Vol. 101, pp. 83–136). Elsevier.
https://doi.org/10.1016/bs.aambs.2017.07.002
141. Zhou, X., Jenkyns, H. C., Lu, W., Hardisty, D. S., Owens, J. D., Lyons, T. W., & Lu, Z.
(2017). Organically bound iodine as a bottom-water redox proxy: Preliminary validation and
application. Chemical Geology, 457, 95–106.
https://doi.org/10.1016/j.chemgeo.2017.03.016
139

Appendices
Appendix A: Supplemental Information for Chapter 1
Comparison of methods for calculating excess iodine
Iodide (I
-
) is frequently observed in surface waters due to iodate (IO3
-
) reduction by
phytoplankton (Farrenkopf et al. 2002, de la Cuesta & Manley 2009; Chance et al. 2014). It
would be beneficial for modeling studies to be able to determine what fraction of the iodide (I
-
)
observed in the surface ocean, where total iodine tends to be slightly depleted, is the result of in
situ phytoplankton reduction versus excess iodine transferred from the shelf (e.g., Wong and
Zhang, 2003). In calculating excess iodine in the ETSP ODZ, Cutter et al. (2018) used a linear
regression of the relationship between IO3
-
and phosphate in the deep ocean in order to calculate
a predicted IO3
-
depletion. The calculated IO3
-
deficit was assumed to reflect in situ reduction to
I
-
; IO3
-
went through stoichiometric conversion to I
-
and subtracted from measured I
-
, thus giving
“excess iodine”. This approach is based largely on observations of Elderfield & Truesdale (1980)
of highly consistent and correlative relationships between specific IO3
-
and specific phosphorous
for deep sea waters. For Cutter et al. 2018, they used a subset of the stations (GEOTRACES
GP16 stations 9 and 11 samples taken at depths greater than 501 m) to define their correlation.
The calculated slope, however, is significantly steeper than the iodine to phosphorous ratio in
biological material (de la Cuesta & Manley 2009; Bergeijk et al. 2016), and the R
2
is 0.28,
indicating a high degree of uncertainty in the slope calculation (Supporting Figure A). Such poor
correlation would lead to significant compounded uncertainty in predicted IO3
-
depletions at the
surface, especially since PO4
3-
concentrations at the surface are much lower than the
concentrations from the deeper waters (>501m) used to calculate the relationship. We also note
140

that similarly poor or inconsistent relationships between IO3
-
and PO4
3-
have been found for the
shelf and other water masses since the original observations of Elderfield and Truesdale (1980)
(e.g., Chapman 1983).

Supporting Figure A. Dissolved phosphate and iodate concentrations measured below 500 m (below the hypoxic zone) at all stations
on the GEOTRACES GP16 cruise. The dashed line shows the results of a linear regression of the data. The correlation is
negative but non-statistically significant (r = -0.065, p = 0.17). Data downloaded from BCO-DMO.

The GEOTRACES GP16 cruise did not find an overall correlation between IO3
-
and
phosphate in the deep ocean, as shown in Figure 17. Variations in the iodine to phosphorous
ratios in the deep ocean probably reflect water mass mixing processes, rather than biological
141

utilization. With a larger database of iodine and phosphorous concentrations in surface waters, or
more information about iodine to phosphorous ratios in marine phytoplankton, this approach
might be able to calculate excess iodine values in the surface more accurately. In the absence of
such datasets, we chose to calculate excess iodine by treating total iodine as essentially
conservative and then subtracting the mean total iodine concentration found in deep waters from
I
-
. This approach still accounts for in situ reduction of IO3
-
to I
-
and makes our values more
precise, at the risk of slightly underestimating excess iodine observed near the ocean surface.
142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

Appendix B: Supplemental Information for Chapter 4

186

187

188

189

190

191

192

193

194

195

196

197

198

199

Abstract (if available)

Abstract The distributions of iodine and copper speciation were measured along three different cruises, within the (mainly North) Pacific Ocean, one of which was a GEOTRACES cruise. GEOTRACES is an international collaborative effort to elucidate the biogeochemical cycling of various trace elements and isotopes (TEIs). Many of these elements are considered nutrients, and their distributions and cycling have biological impacts, especially on phytoplankton.
Marine iodine is mostly found in its two inorganic forms: iodate and iodide. The former is found throughout most of the oxygenated ocean while the latter was thought to be enriched in the surface ocean and in anoxic regions, referred to as Oxygen Deficient Zones (ODZs). In this thesis, high iodide concentrations were found in these ODZs, which cannot be explained by the in situ reduction of available iodate, and likely to originate as a “plume” from the reducing shelf margins off of the Mexican coast. These high iodide values were referred to as “excess iodine” and found to persist even at the edges of these ODZs, likely to do with the slow abiotic oxidation rate of iodide by oxygen and other terminal electron acceptors (chapter 1). These measurements were repeated in the GEOTRACES GP15 which left from the shelf of Alaska and transected to Tahiti. Although high rates of sulfate reduction have been predicted to occur off of Alaska, no large plume of highly concentrated iodide was observed along this cruise. Instead, this cruise offered the opportunity to find that: previous iodide measurements made within the South Pacific are likely erroneous, and the rate of surface iodide oxidation using a non-linear fit with complementary beryllium measurements (chapter 3).
Copper, a unique element, is considered a nutrient at low concentrations and a toxin higher concentration. It is also considered a “hybrid” element for its depth profiles which exhibit behaviors of both “nutrient-like” and “scavenged” distributions; this behavior is exhibited in the form of a linear depth profile which increase from surface to benthic waters. Most (≥99%) of the dissolved copper pool is bound by organic molecules and considered biologically inert, and many workers study the compounds that bind the metal through electrochemical methods. One hypothesis, for the linear depth distribution, is that some of these compounds bind to copper so tightly that it prevents scavenging by sinking particulate matter. This tightly bound fraction of dissolved copper is referred to as “inert copper”. This work explored this hypothesis and required the creation of a novel method to physically separate (via a process of liquid-liquid extraction) dissolved copper into two pools: labile and inert (chapter 2). The method was then applied to, the previously mentioned transect, GEOTRACES GP15 (chapter 4). The data reveal that 60-90% of marine copper is rendered chemically inert throughout the ocean. Labile copper, on the other hand, exhibits a profile similar to other scavenged elements at most locations: higher concentrations at surface waters with a sharp decline with depth due to scavenging by particles. There appears to also be a benthic source of labile copper, as well, which is likely due to the destruction of the binding capacity of the particulate with diagenesis. Our findings conclude that copper complexation by, ligands and organic materials, are not all reversible, as suggested by our current paradigm.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Investigating the global ocean biogeochemical cycling of alkalinity, barium, and copper using data-constrained inverse models

PDF

Anaerobic iron cycling in an oxygen deficient zone

PDF

Comparative behavior and distribution of biologically relevant trace metals - iron, manganese, and copper in four representative oxygen deficient regimes of the world's oceans

PDF

The distribution and speciation of copper across different biogeochemical regimes

PDF

Discerning local and long-range causes of deoxygenation and their impact on the accumulation of trace, reduced compounds

PDF

The marine biogeochemistry of nickel isotopes

PDF

Concentration and size partitioning of trace metals in surface waters of the global ocean and storm runoff

PDF

Modeling deep ocean water and sediment dynamics in the eastern Pacific Ocean using actinium-227 and other naturally occurring radioisotopes

PDF

Investigations on marine metal cycling through a global expedition, a wildfire survey, and a viral infection

PDF

Transcriptional and morphological impacts of copper on Mytilus californianus larval development in current and future ocean conditions

PDF

Going with the flow: constraining the lateral advection of redox-active metals from continental margins under differing oxygen regimes

PDF

The distribution of B-vitamins in two contrasting aquatic systems, and implications for their ecological and biogeochemical roles

PDF

B-vitamins and trace metals in the Pacific Ocean: ambient distribution and biological impacts

PDF

Iron-dependent response mechanisms of the nitrogen-fixing cyanobacterium Crocosphaera to climate change

PDF

The dynamic regulation of DMSP production in marine phytoplankton

PDF

A trace metal investigation of the radiocarbon anomaly in the eastern equatorial Pacific during the last glacial termination

PDF

Future impacts of warming and other global change variables on phytoplankton communities of coastal Antarctica and California

PDF

Particulate organic carbon flux calculated from ²³⁴Th measurements and sediment traps in the eastern tropical south Pacific

$Germanium-silicon fractionation in a continental shelf environment: insights from the northern Gulf of Mexico$

PDF

Germanium-silicon fractionation in a continental shelf environment: insights from the northern Gulf of Mexico

PDF

Actinium-227 as a tracer for mixing in the Deep Northeast Pacific

Asset Metadata

Creator Moriyasu, Rintaro (author)

Core Title The distributions and geochemistry of iodine and copper in the Pacific Ocean

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Ocean Sciences

Degree Conferral Date 2022-05

Publication Date 04/12/2022

Defense Date 03/09/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag biogeochemical cycling,Copper,GEOTRACES,Iodine,method development,OAI-PMH Harvest,Pacific Meridional Transect,Pacific Ocean,redox,speciation,trace element and nutrients,trace metal chemistry

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Moffett, James (committee chair), Hammond, Douglas (committee member), John, Seth (committee member), Smith, Adam (committee member)

Creator Email moriyasu@usc.edu,rintaromoriyasu@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC110937260

Unique identifier UC110937260

Document Type Dissertation

Format application/pdf (imt)

Rights Moriyasu, Rintaro

Type texts

Source 20220412-usctheses-batch-922 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu