University of Southern California Dissertations and Theses

The distribution and speciation of copper across different biogeochemical regimes

(USC Thesis Other)

The distribution and speciation of copper across different biogeochemical regimes

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
1

The distribution and speciation of copper across
different biogeochemical regimes
by
Jeremy E. Jacquot
Submitted to the Faculty of the USC Graduate School on December 18, 2013, in
partial fulfillment of the requirements of the degree of Doctor of Philosophy

Thesis Supervisor:
Dr. James W. Moffett
Title: Professor, Dept. of Biology, University of Southern California

2
Abstract
Copper (Cu) is an essential micronutrient that functions as a cofactor in many important
enzymatically-mediated pathways including denitrification, methane oxidation and ammonia
oxidation. Yet it can also be a potent toxicant, inhibiting phytoplankton reproduction and growth
rates, at picomolar-level concentrations. In natural waters, over 99.9% of dissolved Cu is
complexed by strong organic ligands of biological origin. As a result, concentrations of the
bioavailable fraction, Cu
2+
, are often over a thousand-fold lower (~10
-14
– 10
-13
mol L
-1
) than
dissolved Cu concentrations. The two main controls on the distribution of dissolved Cu, and
Cu
2+
by extension, are organic complexation and scavenging by particles and biological
processes. This thesis examines the distribution and speciation of Cu in the North Atlantic and
eastern tropical South Pacific (ETSP) Oceans and in Hood Canal, an estuary in Puget Sound,
WA, to better understand how regimes with very distinct biogeochemistries influence Cu
cycling. It also explores the relationship between Cu bioavailability and nitrogen cycle processes
in the ETSP and Hood Canal, two regimes with high nitrification activity.
Cu speciation was characterized using competitive ligand exchange-adsorptive cathodic
stripping voltammetry (CLE-ACSV), and total dissolved Cu concentrations were measured using
isotope dilution inductively coupled plasma mass spectrometry (ID-ICP-MS). Overall, results
indicate that Cu
2+
concentrations are maintained at uniformly low levels throughout the water
column by strong organic ligands and scavenging, regardless of location, but particularly in areas
with high nitrification activity. These processes buffer Cu
2+
concentrations to a narrow, “steady-
state” range that rarely varies by more than two orders of magnitude. Dissolved Cu
concentrations were significantly more variable but generally exhibited the profile characteristics
identified by other researchers.

3
In the ETSP, Cu
2+
levels typically reached their lowest values near the chlorophyll and
primary nitrite maxima layers, indicating heavy drawdown by ammonia-oxidizing archaea
(AOA) and nitrate-reducing diatoms. Dissolved Cu concentrations in offshore waters, away
from the Peruvian coastline, were some of the lowest measured values to date. In the Hood
Canal, dissolved Cu levels, though much higher than in open ocean waters, were almost
unchanged from over 30 years ago, suggesting that anthropogenic inputs have not greatly
affected this system. Cu
2+
concentrations in the upper water column often approached an
experimentally determined threshold below which ammonia oxidation by AOA may become Cu-
limited. Although photoinhibition seems like the more plausible control on nitrification activity,
it seems likely that Cu bioavailability also plays a role. Finally, in the North Atlantic Cu
2+

concentrations remained low through the water column and across the transect, rarely exceeding
10
-13
mol L
-1
. The lowest concentrations were measured in the nutrient-rich upwelling regime
off the Mauritanian coastline while the highest concentrations were measured at depth within the
subtropical central gyre region. Dissolved Cu profiles exhibited many unique features and
showed clear geographic trends while still mostly conforming to the general properties identified
by other researchers in the Atlantic and Pacific Oceans.

4
Table of Contents

Chapter 1. Introduction 8
1.1 Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2 Copper speciation and the emergence of the “free ion model” (FIM). . . . . . . . 10
1.3 Analytical methods to measure dissolved Cu species. . . . . . . . . . . . . . . . 11
1.4 Copper, ammonia oxidizing archaea (AOA) and the nitrogen cycle. . . . . . . . . 14
1.5 Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.6 Figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Chapter 2. The speciation of copper across active gradients in nitrogen-cycle processes in
the eastern tropical South Pacific 26
2.1 Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.2 Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.3 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.4 Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.4.1 Sample collection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.4.2 Nitrite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.4.3 Total dissolved copper. . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.4.4 Copper speciation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.5 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.5.1 Nitrite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.5.2 Total dissolved copper. . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

5
2.5.3 Copper speciation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.6 Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.7 Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.8 Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.9 Figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Chapter 3. Assessment of the potential for copper limitation of ammonia oxidation by
Archaea in a dynamic estuary 53
3.1 Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.2 Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.3 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.4 Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.4.1 Study site and sample collection. . . . . . . . . . . . . . . . . . . . . . . 57
3.4.2 Nutrient and ammonia oxidation rate analyses. . . . . . . . . . . . . . . 59
3.4.3 Dissolved Cu analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.4.4 Cu
2+
and organic ligand analyses. . . . . . . . . . . . . . . . . . . . . . 60
3.5 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.5.1 Study site and hydrography. . . . . . . . . . . . . . . . . . . . . . . . . 62
3.5.2 Nutrient and ammonia oxidation rate analyses. . . . . . . . . . . . . . . 62
3.5.3 Dissolved Cu analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.5.4 Cu
2+
and organic ligand analyses. . . . . . . . . . . . . . . . . . . . . . 63
3.6 Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.6.1 The distributions of dissolved Cu, Cu
2+
and L. . . . . . . . . . . . . . . . 66

6
3.6.2 Strong Cu
2+
drawdown near the chlorophyll and primary nitrite maxima. . 69
3.6.3 Does low Cu bioavailability limit ammonia oxidation by AOA? . . . . . 71
3.7 Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.8 Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.9 Figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

Chapter 4. Copper distribution and speciation across the U.S. North Atlantic
GEOTRACES section 89
4.1 Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.2 Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.3 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.4 Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.4.1 Sample collection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.4.2 Dissolved copper analyses. . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.4.3 Cu
2+
and organic ligand analyses. . . . . . . . . . . . . . . . . . . . . . 94
4.4.4 Nutrients. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
4.5 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
4.5.1 Dissolved copper analyses. . . . . . . . . . . . . . . . . . . . . . . . . . 98
4.5.2 Cu
2+
analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
4.5.3 Organic ligand analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4.6 Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.6.1 Dissolved Cu analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.6.2 Cu
2+
, organic ligand and conditional stability constant analyses. . . . . . 112

7
4.7 Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
4.8 Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
4.9 Figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
4.10 Appendix. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

Chapter 5. Synthesis and Conclusions 164
5.1 Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5.2 Specific contributions made by this thesis. . . . . . . . . . . . . . . . . . . . . 165
5.2.1 Is it time to reconsider the importance of Cu limitation? . . . . . . . . . 165
5.2.2 Dissolved Cu on continental margins and in low O
2
waters. . . . . . . . 167
5.3 Anticipating the future: Where to go from here. . . . . . . . . . . . . . . . . . . 169
5.4 Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

8
Chapter 1
Introduction

Overview
The last 40 years have marked a period of unprecedented growth in our understanding of the
oceanic cycling of trace elements such as iron (Fe), zinc (Zn), cobalt (Co) and copper (Cu).
Research programs like GEOSECS (Geochemical Ocean Section Study) paved the way for the
early cruises of the 1970s and 1980s that established the foundations of what we know today
about these elements’ complex biogeochemistries. Since then, rapid advances in the accuracy
and sensitivity of analytical methodologies and instrumentation and in the use of clean sampling
and handling techniques (Sohrin and Bruland 2011; Cutter and Bruland 2012) have enabled
marine geochemists to organize basin-wide expeditions and develop large datasets, culminating
in the inception of the international GEOTRACES program during the early 2000s. At the same
time, molecular biologists and microbial ecologists have made equally huge strides in
understanding how phytoplankton and other microorganisms access and utilize these
micronutrients to drive a variety of globally important biogeochemical processes including
denitrification and carbon fixation. Working at the intersection of their disciplines, marine
geochemists and biologists have made even greater contributions and continue to break new
ground to this day.
While researchers have devoted a great deal of attention to studying the distribution and
speciation of Fe ever since Martin and Fitzwater (1988) made their landmark discovery that low
Fe availability limited phytoplankton growth in high nutrient low chlorophyll (HNLC) regions
like the subarctic North Pacific, some trace elements, notably Cu, have not received as much

9
scrutiny. Early studies examining the speciation and distribution of dissolved Cu (Boyle et al.
1977, 1981; Bruland 1980; Bruland and Franks 1983; Van den Berg 1984; Coale and Bruland
1988, 1990) spurred efforts to investigate its effects on phytoplankton and primary productivity
(Brand et al. 1986; Coale 1991; Morel et al. 1991; Gerringa et al. 1995; Moffett and Brand 1996;
Moffett et al. 1997; Sunda and Huntsman 1998) and the crucial role played by organic
complexation in controlling the concentration of Cu
2+
, its bioavailable fraction, in a wide range
of marine systems (Donat and Van den Berg 1992; Campos and Van den Berg 1994; Moffett
1995). Yet since that initial burst of activity efforts to further probe Cu’s biogeochemistry,
particularly in oceanic systems, have been fewer and far between (Mann et al. 2002; Buck and
Bruland 2005; Maldonado et al. 2006; Moffett and Dupont 2007; Sander et al. 2007; Annett et al.
2008), and our understanding of the processes that contribute to the metal’s unique hybrid profile
remains limited.
The work described in this thesis seeks to expand on this considerable foundation by
presenting speciation and distribution datasets of dissolved Cu—including the first basin-scale
transect dataset—in several locations with very distinct biogeochemical regimes while also
exploring the heretofore understudied role Cu plays in potentially controlling nitrification
activity by the ubiquitous ammonia-oxidizing archaea (AOA). This chapter will provide some
general background on Cu speciation and the concept of the “free ion model,” describe the major
methodologies used to perform the work described in later chapters, and explain the central role
Cu plays in the nitrogen cycle. Chapter 2 delves into a study of the speciation of Cu in the
eastern tropical South Pacific (ETSP), an oligotrophic regime with a prominent oxygen
minimum zone (OMZ), and how it may be related to the formation of the primary nitrite
maximum (PNM) feature. Chapter 3 next looks at the speciation and distribution of Cu in Hood

10
Canal, a fjord in western Puget Sound, Washington, where nitrification activity is dominated by
AOA and Cu
2+
concentrations often fall to levels below which ammonia oxidation may become
limited. Chapter 4 presents Cu distribution and speciation results from the U.S. North Atlantic
GEOTRACES cruises, the first basin-wide Cu transect to date. Finally, chapter 5 summarizes
the results from the previous chapters and looks ahead to predict how the study of Cu and other
trace metals may evolve in the near future.

Copper speciation and the emergence of the “free ion model” (FIM)
The speciation of dissolved Cu in natural waters is overwhelmingly dominated (> 99%) by
Cu-binding organic ligands of primarily biological origin (Coale and Bruland 1988, 1990;
Moffett 1990). Cu can exist as the free, hydrated cupric ion and form complexes with inorganic
ligands like carbonates (CO
3
2-
), hydroxides (OH
-
) and chlorides (Cl
-
) but these forms are
generally very short-lived and labile and account for only a fraction of dissolved Cu species
(Turner et al. 1981; Donat et al. 1994). Studies carried out through the late 80s to 90s
determined that a strong Cu-binding class of organic ligands, referred to collectively as L
1
, was
primarily responsible for complexing dissolved Cu and maintaining ambient Cu
2+
concentrations
around ~10
-13
mol L
-1
(Coale and Bruland 1988, 1990; Donat and Van den Berg 1992; Moffett
1995). These ligands tended to occur at concentrations (~1 – 4 nmol L
-1
) in excess of dissolved
Cu levels and subsequent studies would show that they were produced by cyanobacteria and
diatoms like Synechococcus spp. and Ditylum brightwellii, respectively (Moffett and Brand
1996; Gerringa et al. 1995; Moffett et al. 1997), in order to mitigate the deleterious effects of
elevated Cu
2+
concentrations. Cu
2+
concentrations as low as ~1 pmol L
-1
were found to
significantly inhibit the growth and reproductive rates of a variety of phytoplankton species

11
(Brand et al. 1986) by, among other things, competitively inhibiting the uptake of other
micronutrients like Mn
2+
and Zn
2+
(Sunda and Huntsman 1983, 1998). Many of these and earlier
pioneering studies would lead to the realization that it was the free, cupric ion form, and not total
dissolved Cu, that constituted the bioavailable fraction and to the development of the “free ion
model” (FIM) (Sunda and Guillard 1976; Morel 1983).
At its most basic level, the model states that there exists an equilibrium between the free
cupric ions in solution and the ions bound by organic ligands or to active sites on microbial cell
membranes and particles (Fig. 1). In practice, this meant that marine organisms would only
respond to the free cupric ion concentration and not to complexed dissolved Cu forms—an
assumption underlying all future experimental and in situ studies of Cu complexation that
became known as the “natural organic matter (NOM) corollary” (Hudson 2005). While the FIM
has sometimes been criticized for making too many assumptions about the speciation of Cu in
natural systems based on the use of synthetic ligand-like chelators (Campbell 1995; Campbell et
al. 2002) or for being too uniformly applied to marine organisms (Ytreberg et al. 2011), it has
retained its importance as a cornerstone in our understanding of not only Cu speciation but trace
metal speciation as a whole.

Analytical methods to measure dissolved Cu species
Growing interest in the speciation and biological effects of trace metals like Cu led to the
development of novel electrochemical methods that remain widely used to this day. Competitive
ligand exchange adsorptive cathodic stripping voltammetry (CLE-ACSV), the method that
enabled a significant proportion of the measurements described in later chapters, was first
developed and refined by Constant M. G. Van den Berg and his associates using a hanging

12
mercury (Hg) drop electrode (HMDE) in a number of studies dating back to the 1980s and 1990s
(Van den Berg 1984, 1986, 1988; Van den Berg and Donat 1992; Donat and Van den Berg 1992;
Campos and Van den Berg 1994). The method was quickly adopted by others to study the
speciation of various trace metals and remains the best, if not only, option to measure and assess
organic ligand concentrations and properties.
In CLE-ACSV, a well-characterized synthetic added ligand (AL) is added to a sample in
order to set up a competitive equilibrium with the natural ligands (L) for Cu
2+
(or any other trace
metal of interest) (Van den Berg 1988). Unlike anodic stripping voltammetry (ASV), another
commonly used electrochemical method that detects labile Cu—which includes Cu
2+
and
inorganically complexed Cu—but is significantly less sensitive, CLE-ACSV is an indirect
approach because it only detects the electrochemically active complexes formed by Cu
2+
and AL
(Cu(AL)
x
) (Bruland et al. 2000; Buck and Bruland 2005) where

[Cu(AL)
x
] = [Cu(AL)
2
0
] + [Cu(AL)
+
] (1)
and [Cu
T
] = [CuL] + [Cu’] + [Cu(AL)
x
] (2)

[Cu
T
] refers to the total dissolved Cu concentration, [CuL] refers to the fraction of dissolved Cu
that is organically complexed by natural ligands, and [Cu’] refers to the labile Cu fraction. The
complexes formed by Cu
2+
and the AL are adsorbed onto the hanging Hg drop (the natural
ligand-Cu complexes are not), and a negative potential is applied after a predetermined period of
time known as the deposition time (Bruland and Lohan 2003). The cathodic stripping current
produced by the reduction of Cu
2+
to Cu
0
is measured and then combined with similar
measurements made in a series of complexiometric titrations in which the amount of added Cu
2+

13
is gradually ramped up while keeping the AL concentration constant. The rationale behind this
approach is that the natural ligands will initially outcompete the AL for the labile Cu fraction,
producing no current or very small readings. As the amount of added Cu
2+
is gradually
increased, however, the natural ligands will eventually become saturated and all of the excess
Cu
2+
will become complexed by the AL, producing significantly larger readings. Combining
these current readings with the sensitivity of the instrument allow for the determination of the
natural ligand concentration and conditional stability constant (K) and by extension the Cu
2+

concentration in the sample.
Whereas the methods and instrumentation used to measure Cu
2+
and organic ligand
concentrations have not greatly changed in the last 30 years, those to measure dissolved Cu
concentrations have become increasingly more precise and accurate. Over the past decade, one
of the methods of choice, and the one used in this thesis, has become isotope dilution inductively
coupled plasma mass spectrometry (ID-ICP-MS). In isotope dilution, the ratio of the natural
abundance of two isotopes of a given element (in this case
63
Cu and
65
Cu) is changed by adding a
known quantity of one isotope, called a spike, to the sample (Wu and Boyle 1998). Assuming
the natural abundance of the isotopes being measured is already known, the dissolved Cu
concentration of the original sample can be determined using the following equation:

[Cu
T
] =
€
f 63 spike
f 63 natural
* [Cu
spike
] *
€
spike volume
sample volume

*

€
Rmeasured – Rspike
Rnatural – Rmeasured
"
#
$
%
&
'

(3)

where f
63
is the fraction of the abundance of
63
Cu over the abundance of all the Cu isotopes and R
is the ratio of
65
Cu:
63
Cu. One of the most attractive aspects of isotope dilution is that recovery
efficiency does not affect the accuracy of the results because only the isotopic ratio is needed

14
(Sohrin and Bruland 2011). Before isotope dilution can be successfully applied, however, Cu (or
another trace element analyte of choice) has to be separated from seawater in order to prevent the
major ion matrix from interfering with the measurements. This is done using a solid-phase
extraction preconcentration method with chelating resins to bind the Cu from the spike and
seawater at the appropriate pH. Following an equilibration period, the sample is decanted and
the resins washed multiple times with clean Milli-Q water to remove any remaining seawater,
after which clean dilute nitric acid is added to the resins to elute Cu.

Copper, ammonia oxidizing archaea (AOA) and the nitrogen cycle
Cu plays an integral role in the nitrogen cycle, acting as a cofactor for the enzymes associated
with nitrous oxide reduction, nitrite reduction and ammonia oxidation, the first step of
nitrification (Philippot 2002; Francis et al. 2005). Nitrification is the microbial oxidation of
ammonia (NH
3
) to nitrate (NO
3
-
), and it makes oxidized forms of nitrogen accessible to
denitrification and anammox, two important pathways for the loss of fixed nitrogen from the
ocean (Lam et al. 2009; Santoro et al. 2010). The oxidation of ammonia to hydroxylamine
(NH
2
OH), the first of two steps in the oxidation to NO
2
-
, is catalyzed by ammonia
monooxygenase (AMO), a multi-subunit enzyme encoded by the amoA, amoC and amoB genes
(Principi et al. 2009). Multiple studies have clearly demonstrated that Cu is a cofactor for AMO
(Arp et al. 2002). For instance, Ensign et al. (1993) showed that the addition of Cu activates
AMO in cell-free extracts while Bedard and Knowles (1989) demonstrated that Cu-selective
chelators inactivated the enzyme. Because it is well conserved and is an integral component of
AMO, researchers have often used amoA, which encodes AMO’s catalytic α-subunit, as a
molecular marker to detect ammonia oxidation and quantify its rate (Francis et al. 2005).

15
It was widely assumed until fairly recently that all aerobic ammonia oxidizers were bacteria.
While several metagenomic surveys had suggested that crenarchaeota possessed distantly related
amoA genes, it wasn’t until the discovery and isolation of Nitrosopumilus maritimus SCM1
(hence referred to as SCM1) that researchers first began to appreciate their considerable role in
nitrification activity worldwide (Könneke et al. 2005; Treusch et al. 2005). Around the same
time, Francis et al. (2005) discovered that the amoA gene was commonly found in areas of the
ocean with active nitrogen cycle processes, including suboxic water columns and the base of the
euphotic zone, suggesting that archaea are ammonia oxidizers. They identified these archaea as
putative ammonia oxidizing archaea (AOA). They also found that these AOA were ubiquitous
throughout the water column with distinct communities living in distinct habitats with little
overlap. A number of recent studies have demonstrated that ammonia-oxidizing archaea (AOAs)
have a high demand for copper (Hallam et al. 2006; Walker et al. 2010; Blainey et al. 2011). An
analysis of SCM1’s genome revealed multiple genes encoding for blue Cu proteins similar to the
plasto- and sulfocyanins; this suggests that it uses primarily Cu-dependent mechanisms of
electron transfer instead of the Fe-based system commonly found in AOBs (Walker et al. 2010).
Taken together, these findings indicate that the contributions made by AOA to oceanic
nitrification rates—and the oceanic nitrogen cycle as a whole—have likely been vastly
underestimated and that the role of Cu bioavailability in potentially controlling their activity
deserves more scrutiny. While reported instances of Cu limitation remain rare (Peers et al. 2005;
Maldonado et al. 2006), the recent experimental determination of a relatively elevated Cu
2+

concentration threshold (~10
-13.4
mol L
-1
) below which ammonia oxidation activity is
significantly reduced in SCM1 (Amin et al. 2013) suggests that AOA may be Cu limited in many
parts of the water column and throughout the ocean. Tackling this problem will require a better

16
understanding of both Cu’s biogeochemistry in relation to the nitrogen cycle and AOA and much
more extensive Cu speciation and distribution datasets, which are still lacking.

Bibliography
Amin, S. A., J. W. Moffett, W. Martens-Habbena, J. E. Jacquot, Y. Han, A. Devol, A. E. Ingalls,
D. A. Stahl, and E. V. Armbrust. 2013. Copper requirements of the ammonia-oxidizing
archaeon Nitrosopumilus maritimus SCM1 and implications for nitrification in the marine
environment. Limnol. Oceanogr. In press.
Annett, A. L., S. Lapi, T. J. Ruth, and M. T. Maldonado. 2008. The effects of Cu and Fe
availability on the growth and Cu : C ratios of marine diatoms. Limnol. Oceanogr. 53(6):
2451–2461, doi:10.4319/lo.2008.53.6.2451
Arp, D. J., L. A. Sayavedra-Soto, and N. G. Hommes. 2002. Molecular biology and biochemistry
of ammonia oxidation by Nitrosomonas europaea. Arch. Microbiol. 178: 250 – 255,
doi:10.1007/s00203-002-0452-0
Bedard, C. and R. Knowles. 1989. Physiology, biochemistry, and specific inhibitors of CH
4
,
NH
4
+
, and CO oxidation by methanotrophs and nitrifiers. Microbiol. Rev. 53: 68 – 84.
Blainey, P. C., A. C. Mosier, A. Potanina, C. A. Francis, and S. R. Quake. 2011. Genome of a
low-salinity ammonia-oxidizing archaeon determined by single-cell and metagenomic
analysis. PLoS ONE 6(2): e16626, doi:10.1371/journal.pone.0016626
Boyle, E. A., F. Sclater, and J. Edmond. 1977. The distribution of dissolved copper in the
Pacific. Earth and Planet. Sci. Lett. 37: 38–54, doi:10.1016/0012-821X(77)90144-3

17
Boyle, E. A., S. S. Huested, and S. P. Jones. 1981. On the distribution of copper, nickel, and
cadmium in the surface waters of the North Atlantic and North Pacific Ocean. J. Geophys.
Res. 86: 8048–8066, doi:10.1029/JC086iC09p08048
Brand, L. E., W. G. Sunda, and R. R. L. Guillard. 1986. Reduction of marine phytoplankton
reproduction rates by copper and cadmium. J. Exp. Mar. Biol. Ecol. 96: 225–250,
doi:10.1016/0022-0981(86)90205-4
Bruland, K. W. 1980. Oceanographic distributions of cadmium, zinc, nickel, and copper in the
North Pacific. Earth and Planet. Sci. Lett. 47: 176–198, doi:10.1016/0012-821X(80)90035-
7
Bruland, K. W., and R. P. Franks. 1983. Mn, Ni, Cu, Zn and Cd in the western North Atlantic, p.
395–414. In C. S. Wong, E. Boyle, K. W. Bruland, J. D. Burton and E. D. Goldberg [eds.],
Trace metals in sea water. Academic.
Bruland, K. W., E. L. Rue, J. R. Donat, S. A. Skrabal, and J. W. Moffett. 2000. Intercomparison
of voltammetric techniques to determine the chemical speciation of dissolved copper in a
coastal seawater sample. Anal. Chim. Acta 405: 99–113, doi:10.1016/S0003-
2670(99)00675-3
Bruland, K. W., and M. C. Lohan. 2003. Controls of trace metals in seawater, p. 23–47. In H.
Elderfield [ed.], The oceans and geochemistry. Academic.
Buck, K. N., and K. W. Bruland. 2005. Copper speciation in San Francisco Bay: A novel
approach using multiple analytical windows. Mar. Chem. 96: 185–198,
doi:10.1016/j.marchem.2005.01.001

18
Campbell, P. G. C. 1995. Interactions between trace metals and aquatic organisms: a critique of
the free-ion activity model, p. 45–102. In A. Tessier and D. R. Turner [eds.], Metal
speciation and bioavailability in aquatic systems. Academic.
Campbell, P. G. C., O. Errécalde, C. Fortin, V. Hiriart-Baer, and B. Vigneault. 2002. Metal
bioavailability to phytoplankton—applicability of the biotic ligand model. Comp. Biochem.
Physiol. 133C: 189–206, doi:10.1016/S1532-0456(02)00104-7
Campos, M. L. A. M., and C. M. G. van Den Berg. 1994. Determination of copper complexation
in sea water by cathodic stripping voltammetry and ligand competition with
salicylaldoxime. Anal. Chim. Acta 284: 481–496, doi:10.1023/A:1009653002399
Coale, K. H., and K. W. Bruland. 1988. Copper complexation in the Northeast Pacific. Limnol.
Oceanogr. 33(5): 1084–1101, doi:10.4319/lo.1988.33.5.1084
Coale, K. H., and K. W. Bruland. 1990. Spatial and temporal variability in copper complexation
in the North Pacific. Deep-Sea Res. 37: 317–336, doi:10.1016/0198-0149(90)90130-N
Coale, K. H. 1991. Effects of iron, manganese, copper, and zinc enrichments on productivity and
biomass in the subarctic Pacific. Limnol. Oceanogr. 36(8): 1851–1864,
doi:10.4319/lo.1991.36.8.1851
Cutter, G. A., and K. W. Bruland. 2012. Rapid and noncontaminating sampling system for trace
elements in global ocean surveys. Limnol. Oceanogr.: Methods 10: 425–436,
doi:10.4319/lom.2012.10.425
Donat, J. R., and C. M. G. Van Den Berg. 1992. A new cathodic stripping voltammetric method
for determining organic copper complexation in seawater. Mar. Chem. 38: 69–90,
doi:10.1016/0304-4203(92)90068-L

19
Donat, J. R., K. A. Lao, and K. W. Bruland. 1994. Speciation of dissolved copper and nickel in
South San Francisco Bay: a multi-method approach. Anal. Chim. Acta 284: 547–571,
doi:10.1016/0003-2670(94)85061-5
Ensign, S. A., M. R. Hyman, and D. J. Arp. 1993. In vitro activation of ammonia
monooxygenase from Nitrosomonas europaea by copper. J. Bacteriol. 175(7): 1971–1980.
Francis, C. A., K. J. Roberts, J. M. Beman, A. E. Santoro, and B. B. Oakley. 2005. Ubiquity and
diversity of ammonia-oxidizing archaea in water columns and sediments of the ocean.
Proc. Nat. Acad. Sci. U.S.A. 102(41): 14683–14688, doi:10.1073/pnas.0506625102
Gerringa, L. J. A., J. Rijstenbil, T. Poortvliet, J. Van Drie, and M. Schot. 1995. Speciation of
copper and responses of the marine diatom Ditylum brightwellii upon increasing copper
concentrations. Aquat. Toxicol. 31: 77–90, doi:10.1016/0166-445X(94)00053-S
Hallam, S. J., K. T. Konstantinidis, N. Putnam, C. Schleper, Y. Watanabe, J. Sugahara, C.
Preston, J. R. de La Torre, P. M. Richardson, and E. F. Delong. 2006. Genomic analysis of
the uncultivated marine crenarchaeote Cenarchaeum symbiosum. Proc. Natl. Acad. Sci.
U.S.A. 103(48): 18296–18301, doi:10.1073/pnas.0608549103
Hudson, R. J. M. 2005. Trace metal uptake, natural organic matter, and the free-ion model. J.
Phycol. 41: 1–4, doi:10.1111/j.1529-8817.2005.41101.x
Könneke, M., A. E. Bernhard, J. R. de La Torre, C. B. Walker, J. B. Waterbury, and D. A. Stahl.
2005. Isolation of an autotrophic ammonia-oxidizing marine archaeon. Nature 437: 543–
546, doi:10.1038/nature03911
Lam, P., G. Lavik, M. M. Jensen, J. van de Vossenberg, M. C. Schmid, D. Woebken, D.
Gutiérrez, R. Amann, M. S. M. Jetten, and M. M. M. Kuypers. 2009. Revising the nitrogen

20
cycle in the Peruvian oxygen minimum zone. Proc. Nat. Acad. Sci. U.S.A. 106(12): 4752–
4757, doi:10.1073/pnas.0812444106
Maldonado, M. T., A. E. Allen, J. S. Chong, K. Lin, D. Leus, N. Karpenko, and S. L. Harris.
2006. Copper-dependent iron transport in coastal and oceanic diatoms. Limnol. Oceanogr.
51(4): 1729–1743, doi:10.4319/lo.2006.51.4.1729
Mann, E. L., N. Ahlgren, J. W. Moffett, and S. W. Chisholm. 2002. Copper toxicity and
cyanobacteria ecology in the Sargasso Sea. Limnol. Oceanogr. 47(4): 976–988,
doi:10.4319/lo.2002.47.4.0976
Martin, J. H., and S. E. Fitzwater. 1988. Iron deficiency limits phytoplankton growth in the
north-east Pacific subarctic. Nature 331: 341–343, doi:10.1038/331341a0
Moffett, J. W., R. G. Zika, and L. E. Brand. 1990. Distribution and potential sources and sinks of
copper chelators in the Sargasso Sea. Deep-Sea Res. 37: 27–36, doi:10.1016/0198-
0149(90)90027-S
Moffett, J. W. 1995. Temporal and spatial variability of copper complexation by strong chelators
in the Sargasso Sea. Deep-Sea Res. 42: 1273–1295, doi:10.1016/0967-0637(95)00060-J
Moffett, J. W., and L. E. Brand. 1996. Production of strong, extracellular Cu chelators by marine
cyanobacteria in response to Cu stress. Limnol. Oceanogr. 41(3): 388–395,
doi:10.4319/lo.1996.41.3.388
Moffett, J. W., L. E. Brand, P. L. Croot, and K. A. Barbeau. 1997. Cu speciation and
cyanobacterial distribution in harbors subject to anthropogenic Cu inputs. Limnol.
Oceanogr. 42(5): 789–799, doi:10.4319/lo.1997.42.5.789
Moffett, J. W., and C. L. Dupont. 2007. Cu complexation by organic ligands in the sub-arctic
NW Pacific and Bering Sea. Deep-Sea Res. 54: 586–595, doi:10.1016/j.dsr.2006.12.013

21
Morel, F. M. M. 1983. Principles of aquatic chemistry, 1
st
ed. Wiley.
Morel, F. M. M., R. J. Hudson, and N. M. Price. 1991. Limitation of productivity by trace metals
in the sea. Limnol. Oceanogr. 36(8): 1742–1755, doi:10.4319/lo.1991.36.8.1742
Peers, G., S. A. Quesnel, and N. M. Price. 2005. Copper requirements for iron acquisition and
growth of coastal and oceanic diatoms. Limnol. Oceanogr. 50(4): 1149–1158, doi:
10.4319/lo.2005.50.4.1149
Philippot, L. 2002. Denitrifying genes in bacterial and archaeal genomes. Biochim. Biophys.
Acta Gene Struct. Expr. 1577: 355–376, doi:10.1016/S0167-4781(02)00420-7
Principi, P., F. Villa, B. Giussani, E. Zanardini, F. Cappitelli, and C. Sorlini. 2009. The effect of
copper on the structure of the ammonia-oxidizing microbial community in an activated
sludge wastewater treatment plant. Microb. Ecol. 57: 215 – 220, doi:10.1007/s00248-008-
9432-5
Sander, S. G., A. Koschinsky, G. Massoth, M. Stott, and K. A. Hunter. 2007. Organic
complexation of copper in deep-sea hydrothermal vent systems. Environ. Chem. 4: 81–89,
doi:10.1071/EN06086
Santoro, A. E., K. L. Casciotti, and C. A. Francis. 2010. Activity, abundance and diversity of
nitrifying archaea and bacteria in the central California Current. Environ. Microbiol. 12:
1989–2006, doi:10.1111/j.1462-2920.2010.02205.x
Sohrin, Y., and K. W. Bruland. 2011. Global status of trace elements in the ocean. Trends Anal.
Chem. 30(8): 1291–1307, doi:10.1016/j.trac.2011.03.006
Sunda, W. G., and R. R. L. Guillard. 1976. The relationship between cupric ion activity and the
toxicity of copper to phytoplankton. J. Mar. Res. 34(4): 511–529.

22
Sunda, W. G., and S. A. Huntsman. 1983. Effect of competitive interactions between manganese
and copper on cellular manganese and growth in estuarine and oceanic species of the
diatom Thalassiosira. Limnol. Oceanogr. 28(5): 924–934, doi:10.4319/lo.1983.28.5.924
Sunda, W. G., and S. A. Huntsman. 1998. Interactions among Cu
2+
, Zn
2+
, and Mn
2+
in
controlling cellular Mn, Zn, and growth rate in the coastal alga Chlamydomonas. Limnol.
Oceanogr. 43(6): 1055–1064, doi:10.4319/lo.1998.43.6.1055
Treusch, A. H., S. Leininger, A. Kletzin, S. C. Schuster, H.-P. Klenk, and C. Schleper. 2005.
Novel genes for nitrite reductase and Amo-related proteins indicate a role of uncultivated
mesophilic crenarchaeota in nitrogen cycling. Environ. Microbiol. 7: 1985–1995,
doi:10.1111/j.1462-2920.2005.00906.x
Turner, D. R., M. Whitfield, and A. G. Dickson. 1981. The equilibrium speciation of dissolved
components in freshwater and sea water at 25°C and 1 atm pressure. Geochim.
Cosmochim. Acta 45: 855–881, doi:10.1016/0016-7037(81)90115-0
Van den Berg, C. M. G. 1984. Determination of the complexing capacity and conditional
stability constants of complexes with copper(II) with natural organic ligands in seawater by
cathodic stripping voltammetry of copper-catechol ions. Mar. Chem. 15: 1–18,
doi:10.1016/0304-4203(84)90035-5
Van den Berg, C. M. G. 1986. The determination of trace metals in sea-water using cathodic
stripping voltammetry. Sci. Total Environ. 49: 89–99, doi:10.1016/0048-9697(86)90233-0
Van den Berg, C. M. G. 1988. Adsorptive cathodic stripping voltammetry and
chronopotentiometry of trace metals in sea water. Anal. Proc. 25: 265–266.
Van den Berg, C. M. G., and J. R. Donat. 1992. Determination and data evaluation of copper
complexation by organic ligands in sea water using cathodic stripping voltammetry at

23
varying detection windows. Anal. Chim. Acta 257: 281–291, doi:10.1016/0003-
2670(92)85181-5
Walker, C. B., J. R. de La Torre, M. Klotz, H. Urakawa, N. Pinel, D. Arp, C. Brochier-Armanet,
P. Chain, P. Chan, A. Gollabgir, J. Hemp, M. Hügler, E. A. Karr, M. Könneke, M. Shin, T.
J. Lawton, T. Lowe, W. Martens-Habbena, L. A. Sayavedra-Soto, D. Lang, S. M. Sievert,
A. C. Rosenzweig, G. Manning, and D. A. Stahl. 2010. Nitrosopumilus maritimus genome
reveals unique mechanisms for nitrification and autotrophy in globally distributed marine
crenarchaea. Proc. Nat. Acad. Sci. U.S.A. 107(19): 8818–8823,
doi:10.1073/pnas.0913533107
Wu, J., and E. A. Boyle. 1998. Determination of iron in seawater by high-resolution isotope
dilution inductively coupled plasma mass spectrometry after Mg(OH)
2
coprecipitation.
Anal. Chim. Acta 367: 183–191, doi:10.1016/S0003-2670(98)00145-7
Ytreberg, E., J. Karlsson, S. Hoppe, B. Eklund, and K. Ndungu. 2011. Effect of organic
complexation on copper accumulation and toxicity to the estuarine red macroalga
Ceramium tenuicorne: a test of the free ion activity model. Environ. Sci. Technol. 45:
3145–3153, doi:10.1021/es1039166

24
Figure Legends
Fig. 1. A conceptual diagram of the free ion model (FIM) illustrating the equilibrium between
the different forms of dissolved Cu. Most dissolved Cu is organically complexed by natural
ligands ([Cu-L]), and the free, hydrated form (Cu
2+
) can adsorb onto particle or microbial cell
membrane active sites.

25

Fig. 1.

26
Chapter 2
The speciation of copper across active gradients in nitrogen-cycle processes in the eastern
tropical South Pacific

Acknowledgments
We thank Douglas G. Capone, Chief Scientist, the Captain and crew of the R/V Atlantis, and
the University of Southern California Plasma Mass Spectrometry Facility. We would also like to
thank the two anonymous reviewers for their helpful comments and suggestions. This research
was supported by the National Science Foundation Chemical Oceanography Program through a
grant to James W. Moffett.

27
Abstract
Copper (Cu) complexation and distribution were characterized using competitive ligand
exchange adsorptive cathodic stripping voltammetry (CLE-ACSV) and isotope dilution
inductively coupled plasma mass spectrometry (ID-ICP-MS) along two transects (20°S and
10°S) in the eastern tropical South Pacific (ETSP). In the southern and westernmost stations, Cu
showed upper water column depletion to values as low as ~0.26 nmol L
-1
, the lowest
concentrations ever reported. However, Cu levels were much higher within the secondary nitrite
maxima (SNM) of the oxygen minimum zone (OMZ) in the northern (10°S) transect. The
enrichment of Cu in the reducing conditions of the OMZ has not been reported before and
probably reflects remineralization and offshore transport from the shelf. Free [Cu
2+
] was
typically low throughout the water column, ranging from about 3.15 × 10
-15
mol L
-1
to 1.34 × 10
-
13
mol L
-1
, and depth profiles exhibited similar features to those for dissolved Cu though they
showed more variability near the surface. Offshore and beyond the influence of the OMZ, the
lowest dissolved and free [Cu
2+
] was within the primary nitrite maxima (PNM), where ammonia
oxidation and nitrate reduction rates are important. This finding is of interest because the two
competing explanations for the PNM—iron (Fe) limitation of diatoms and high rates of ammonia
oxidation relative to nitrite oxidation—have high Cu requirements. The low concentrations of
free Cu
2+
measured here could impose significant constraints on the rates of these processes.

28
Introduction
Copper (Cu) plays an integral role in the nitrogen cycle, acting as a cofactor for the enzymes
associated with nitrous oxide reduction (nitrous oxide reductase; NoSZ), nitrite reduction (nitrite
reductase; NiRK), and ammonia oxidation (ammonium monooxidase; AMO) (Philippot 2002;
Francis et al. 2005). There is an important Cu requirement for microbes catalyzing these
processes and, as a result, Cu limitation may link the N and Cu cycles under certain conditions.
By analogy, previous workers have argued that the high iron (Fe) content of the nitrogenase
enzyme leads to widespread Fe limitation of nitrogen fixation, which has been supported by
previous observations (Kustka et al. 2002). While the distribution and speciation of copper have
been studied extensively, there are no reports of Cu speciation in the context of nitrogen cycling.
This study explored the relationship between Cu distribution and speciation in the eastern
tropical South Pacific along two transects that extended from approximately 80°W to 100°W and
from 100°W to 82°30’W at 20°S and 10°S, respectively.
The eastern tropical South Pacific (ETSP) is the site of one of the world’s three largest
oceanic oxygen minimum zones (< 10 µmol L
-1
O
2
) (Stewart et al. 2011), whose depth varies
between 50 and 400 m, and is a hotspot for denitrification and anaerobic ammonium oxidation
(anammox), the two major pathways for the loss of fixed nitrogen (N
2
) (Lam et al. 2009). The
sampling plan included waters within the oxygen minimum zone (OMZ) dominated by anammox
and denitrification and characterized by a well-defined secondary nitrite maximum (SNM)
between roughly 200 and 400 m. There is also a well-defined primary nitrite maximum (PNM)
throughout the region. The formation of the PNM has been linked to both bacterial nitrification
and nitrite (NO
2
-
) excretion by phytoplankton at the base of the euphotic zone (Lomas and
Lipschultz 2006) so we hypothesized that free Cu
2+
levels would reach a minimum at or near the

29
PNM due to the combined effect of ammonia oxidizing archaeal (AOA)-mediated ammonia
oxidation and diatom-mediated nitrate reduction. In recent work, Lam et al. (2009) determined
that ammonia oxidation and nitrate reduction accounted for 6 – 33% and 67 – 94% of total nitrite
production in the upper OMZ, respectively. Furthermore, because some denitrifiers require Cu
for both nitrite reduction and nitrous oxide reduction, we also expected to see low free Cu
2+

levels at or near the SNM.
Until recently, it was widely assumed that aerobic ammonia oxidizers were primarily bacteria
(AOB) (Francis et al. 2005). While some metagenomic surveys had suggested that
crenarchaeota possessed distantly related amoA genes (Treusch et al. 2005), it was not until the
discovery and isolation of Nitrosopumilus maritimus strain SCM1 (heretofore referred to as
SCM1) by Könneke et al. (2005) that microbiologists began to truly appreciate the contribution
of archaea to the global nitrogen cycle. Soon thereafter, Francis et al. (2005) discovered that a
previously characterized archaeal amoA gene was ubiquitous in areas of the ocean where
nitrification is important, including the Black Sea and the eastern tropical north Pacific (ETNP),
the site of another of the world’s largest pelagic OMZs. In light of their findings, and because
AOB typically constitute only around 0.1% of bacterial communities, the authors reason that
AOA’s contribution to the global nitrogen cycle has likely been significantly underestimated
(Francis et al. 2005). Since then, a number of studies have identified large numbers of diverse
AOA living in both the well-oxygenated and hypoxic waters of the Chilean OMZ (Molina et al.
2010). An analysis of SCM1’s genome revealed the presence of heavily Cu-dependent systems
for electron transport, comprising multiple genes encoding for blue Cu proteins similar to plasto-
and sulfocyanins, and ammonia oxidation (Walker et al. 2010). By contrast, AOB use a Fe-
dependent electron transport system (Walker et al. 2010). Despite the unique circumstances of

30
its discovery, SCM1’s gene content and order are very similar to those of other oceanic
crenarchaeal populations, so it is therefore not unreasonable to presume that their demand for Cu
is similarly elevated (Könneke et al. 2005; Walker et al. 2010). Recent work by S. Amin (pers.
comm.) has even shown that Cu limitation can be induced in SCM1 at levels that closely
resemble those we observed around the PNM. These findings imply that AOA in the upper
OMZ could be subject to similar constraints.
Copper limitation has also been previously demonstrated in Thalassiosira oceanica (Peers et
al. 2005). This oceanic diatom possesses an iron (Fe) transport system that is induced under Fe-
limiting conditions and has a Cu requirement that outstrips that of its coastal relatives
(Maldonado et al. 2006). In Maldonado et al.’s (2006) study, T. oceanica isolates were able to
take up Fe under Fe-limiting conditions as long as sufficient bioavailable Cu was supplied.
Unlike the chlorophyll-rich upwelled waters of the Peruvian coast, the offshore waters that
overlie the OMZ are much less productive and can become Fe-limited (Bruland et al. 2005). In
such a regime, nitrite reduction, which is mediated by the Fe-containing ferredoxin, would likely
become the rate-limiting step in nitrate assimilation by diatoms, resulting in an accumulation of
nitrite in the surrounding waters due to cellular excretion and a more pronounced drawdown of
Cu (Milligan and Harrison 2000). Light limitation in the lower euphotic zone would be expected
accelerate this process (Lomas and Lipschultz 2006). Indeed, work by Sedwick et al. (2005) at
several locations in the Sargasso Sea revealed that the lowest Fe concentration consistently
occurred at or around the depth of PNM. While several studies have shown that some
phytoplankton can substitute flavodoxin, which does not require Fe, for ferredoxin under Fe-
limiting conditions (La Roche et al. 1995), Milligan and Harrison’s (2000) work has indicated
that T. pseudonana, despite making flavodoxin, also released nitrite under similar conditions.

31
Fe-limited diatoms would thus find themselves in direct competition with AOA over already
scarce levels of Cu throughout much of the euphotic zone. One might therefore expect this
competition to culminate in the formation of a local free Cu minimum at the PNM.
Here we present a Cu speciation data set for the ETSP, spanning regimes where nitrogen
fixation, denitrification, anammox and nitrification are presumed to be important, and the
corresponding nitrite measurements. Despite the importance of Cu to the nitrogen cycle, oceanic
speciation data sets are still rare—most still originate from the North Pacific (Coale and Bruland
1988, 1990; Moffett and Dupont 2007), and we believe this study represents the first such
extensive effort in the ETSP.

Methods
Sample collection—Samples were collected aboard the R/V Atlantis during the AT-15-61
cruise between 29 January 2010 and 03 March 2010 (Fig. 1) from 20 m down to 1000 m using 5
L Teflon coated external spring Niskin-type bottles (Ocean Test Equipment) mounted on a trace
metal clean rosette (Sea-Bird Electronics). The rosette was lowered over the side of the ship on a
Kevlar line, and the bottles were preprogrammed to trip on the downcast at pre-specified depths.
After retrieving the bottles and pressurizing them with filtered compressed nitrogen gas, the
seawater samples were forced through Teflon tubing and acid-cleaned 0.2 µm Acropak capsules
(Pall Corporation) into 250 mL low-density polyethylene bottles (LDPE, Nalgene) in a laminar
flow bench inside a positive pressure clean room enclosure. The bottles were rigorously cleaned
in a sequential four-step process: 1) soaked for at least 24 hours in a 5% Citranox acid detergent
bath (Alconox), 2) soaked for at least 24 hours in a 10% hydrochloric acid bath (HCl; Van
Waters and Rogers (VWR) International), 3) filled with 10% HCl and baked at 60°C upright and

32
upside down (to properly leach the threads around the cap) for at least 48 hours and 4) filled with
0.1% trace metal grade HCl (Optima, Fisher) and baked at 60°C again for at least 48 hours. In
between each step, the insides and outsides of the bottles were thoroughly rinsed at least five
times with Milli-Q water (18.2 MΩ; Millipore).
Nitrite—Reactive NO
2
-
was measured spectrophotometrically after Strickland and Parsons
(1968). Briefly, small Acropak-filtered sample aliquots (15 mL or less) were each made to react
with an acidified sulfanilamide solution to form a diazonium compound. The compound was
then mixed with N-(1-Naphthyl)-ethylenediamine dihydrochloride to produce a colored azo dye,
whose extinction was measured on a Shimadzu ultraviolet (UV)-1700 ultraviolet visible (UV-
VIS) spectrophotometer to determine the NO
2
-
concentration of each sample.
Total dissolved copper—All samples were acidified to below pH 2 by the addition of
concentrated trace metal grade HCl (Optima, Fisher) and stored for at least one month before the
analyses were made. The total dissolved Cu concentration ([Cu
T
]) was determined using a single
batch nitrilotriacetatic acid (NTA) resin extraction and isotope dilution inductively coupled
plasma mass spectrometry (ICP-MS) method adapted from Lee et al. (2011) on a Finnegan
Element 2 (Thermo Scientific). Total dissolved Fe was measured simultaneously but those
results will be reported elsewhere (Y. Kondo and J. W. Moffett unpubl.).
The NTA Superflow resin (Qiagen), a chelating resin, was added to the samples during the
preparation stage to isolate and concentrate the dissolved Cu. In order to minimize
contamination to the samples from the beads, the NTA resin was cleaned using the following
procedure (Lee et al. 2011). First, the NTA resin solution (25 mL) was poured into a clean 50
mL polypropylene centrifuge tube (Corning) and then washed five times with Milli-Q water. In
between washes, the tube was spun down in a 5810-R centrifuge (Eppendorf) maintained at 8°C

33
for 10 min at 4000 revolutions per minute (rpm). The supernatant was decanted and Milli-Q
water was added for the next wash. The resin was then washed five times with 1.5 mol L
-1
trace
metal grade HCl (Optima, Fisher) and several more times with Milli-Q water after that to bring
the pH of the solution above 4. Though the exact pH value was not necessarily of consequence,
it served to indicate that all of the HCl had been removed from the solution. The centrifugation
conditions were kept the same. To condition the resin, 0.5 mol L
-1
trace metal grade HCl was
added and the solution was refrigerated for four to five days at ~4°C. For the final cleaning step,
the resin solution was washed five times with 0.5 mol L
-1
trace metal grade HNO
3
(Optima,
Fisher). The resin solution was placed on an analog shaker (Thomas Scientific) for several hours
for the first wash and then left overnight on the shaker for the last wash. After the final wash, the
resin solution was again washed at least five times with Milli-Q water until the pH had risen
above 4 in order to remove all of the HNO
3
. The resin solution was diluted twofold with 25 mL
Milli-Q water and stored in the refrigerator. This primary resin suspension contained sufficient
beads to prepare fifty 50 mL batches of working resin suspension, which were used for the
sample preparation. The working resin suspension is a 1:50 dilution of the primary resin
suspension in Milli-Q water; 25 µL of the working resin suspension contains ~100 – 400 beads.
The 15 mL polypropylene centrifuge tubes (VWR) used for the sample preparation were
cleaned in a four-step process. They were first soaked in 10% HCl at 60°C for 48 hours and
rinsed at least five times with Milli-Q water. They were then soaked in 1% trace metal grade
HCl at 60°C for 48 hours and rinsed at least five more times with Milli-Q water. The caps were
soaked in trace metal grade HCl and rinsed with Milli-Q in the same sequential process.
Following the soaks, the tubes were filled to a positive meniscus with 1% trace metal grade HCl
and capped. They were then baked at 60°C overnight. After emptying the tubes and rinsing them

34
seven times with Milli-Q water, they were filled once more to a positive meniscus with 0.1%
trace metal grade HCl and capped until use.
Upon analysis, the tubes were emptied and rinsed seven times with Milli-Q water and at least
once with the sample. They were filled with ~7.5 mL of sample (with the exact volume
determined gravimetrically) and spiked with enough
65
Cu-enriched spike (British Drug Houses
(BDH) Aristar Plus, VWR) to bring the final concentration to ~2 nmol L
-1
. The pH of the
samples was then increased to at least 4 with 1.5 mL of 0.1 mol L
-1
trace metal grade ammonium
acetate buffer; the buffer was prepared by combining 0.1 mol L
-1
trace metal grade ammonium
hydroxide (NH
4
OH; Optima, Fisher) and 0.1 mol L
-1
trace metal grade acetic acid (CH
3
COOH;
Optima, Fisher). The higher pH was necessary to accurately measure [Cu] with this method (Lee
et al. 2011). In order to measure Fe simultaneously, 0.1 mL of 1.5 mol L
-1
trace metal grade
hydrogen peroxide (H
2
O
2
; Optima, Fisher) was also added to each sample (Y. Kondo and J. W.
Moffett unpubl.). The samples were then left to equilibrate for at least an hour at room
temperature to allow for the complete oxidation of Fe
2+
to Fe
3+
(Lee et al. 2011). Next, 200 µL
(~800 beads) of the working resin suspension was added to each sample, and the tubes were
placed on a shaker for four to five days. The samples were then centrifuged for 10 min at 4000
rpm, and the seawater was carefully siphoned off to leave only the resin beads at the bottom.
The beads were washed with 3 mL Milli-Q water to remove salts and the tubes were once again
centrifuged using the same settings. This step was repeated twice more to ensure that all of the
salts were removed. After the final wash, 1 mL of 5% trace metal grade HNO
3
(Optima, Fisher)
was added to each tube and, after leaving them on the shaker again for one to two days, the
samples were ready for analysis. The isotope ratios (
65
Cu:
63
Cu) were determined on the
instrument in medium resolution mode. Procedural seawater blanks were prepared the same way

35
using ~0.2 mL low trace metal surface seawater from the 2009 Pacific GEOTRACES
intercomparison cruise ([Cu] = 0.59 ± 0.009 nmol L
-1
(n = 3)) in lieu of 7.5 mL of sample. All of
the samples were analyzed in triplicate. The average detection limit and internal blank value for
this method for Cu were both 0.030 nmol L
-1
(n = 3). The Sampling and Analysis of Iron (SAFe)
reference standards S1 and D1 (Johnson et al. 2007) were measured alongside the samples to
assess the accuracy of the method. The resulting concentrations (consensus values: 0.51 ± 0.05
nmol L
-1
(S1), 2.27 ± 0.11 nmol L
-1
(D1)) (http://www.geotraces.org/science/intercalibration)
were 0.54 ± 0.046 nmol L
-1
(n = 3) and 2.35 ± 0.19 nmol L
-1
(n = 3) for S1 and D1, respectively.
These values are within the range of the latest consensus numbers.
Copper speciation—Samples from five stations (5, 7, 9, 11, and 12) were collected for the Cu
speciation analyses in 1 L fluorinated ethylene propylene (FEP, Nalgene) or 1 L fluorinated low-
density polyethylene (FLPE, Nalgene) bottles. Samples were refrigerated at ∼4°C, and most
analyses were performed within four days of collection at room temperature (∼25°C). The rest
were frozen at -20°C and transferred to a shore-based laboratory freezer after the cruise. These
samples were thawed in a refrigerator for two days and then analyzed within seven days. While
onboard measurements are preferable to shore-based ones because of the potential for changes in
speciation during freezing, their time-intensive nature makes it difficult for all of them to be
analyzed during a cruise. It is worth noting that previous studies (Buck et al. 2012) found no
significant difference between the Cu speciation results obtained from fresh samples and those
obtained from frozen (-20°C) samples. Recent results from the US GEOTRACES cruise support
that conclusion (J. E. Jacquot and J. W. Moffett unpubl.). The free Cu
2+
concentrations, ligand
concentrations ([L]) and conditional stability constants (K) were determined using a competitive
ligand exchange adsorptive cathodic stripping voltammetry (CLE-ACSV) method adapted from

36
Buck and Bruland (2005) and Moffett and Dupont (2007). Salicylaldoxime (SA; ≥ 98%, Aldrich)
was used as the added ligand.
The measurements were carried out on a BioAnalytical Systems (BASi) Controlled Growth
Mercury Electrode (CGME) set to the Static Mercury Drop setting (drop size: 14 or 16) and
interfaced with a BASi Epsilon ε2 voltammetric analyzer. It was mounted on a vibration-free
platform to minimize vibration-related interferences. The instrument settings, set in accordance
with previous studies (Buck and Bruland 2005; Moffett and Dupont 2007), were: method:
differential pulse stripping voltammetry; deposition potential: -0.15 V (vs. Ag:AgCl (3 mol L
-1

NaCl) reference electrode and platinum wire counter electrode); stir rate: 600 rpm; deposition
time, t
d
= 300 – 600 s; quiet period, 10 s; scan range, -0.15 to -0.65 V, scan rate, 20 mV s-1; step
potential, 4 mV; pulse width, 35 ms; pulse period, 0.2 s; pulse amplitude, 0.05 V.
The titrations were performed as follows. Each sample was divided into 20 mL aliquots in
125 mL Teflon bottles (FEP, Nalgene). The pH of each sample was then adjusted to closely
match its ambient pH with either 1 mol L
-1
4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
(HEPES; ≥ 99.5%, Sigma) buffer or 1 mol L
-1
4-(2-hydroxyethyl)-1-piperazinepropanesulfonic
acid (EPPS; ≥ 99.5%, Sigma) buffer and varying amounts of 1 mol L
-1
trace metal grade sodium
hydroxide (NaOH; Fluka). This was done because pH has been shown to influence SA’s binding
capacity and we wanted to best approximate the ambient conditions for our titrations (Kogut
2002). Side reaction (α) coefficients for CuSA complexes, which reflect SA’s ability to
outcompete the natural Cu-binding organic ligands, were obtained from Campos and van den
Berg (1994) and adjusted for pH using the following relationship described by Kogut (2002):

[HSA
!
]
[SA
!
]
=
1
10
pKa
1
- pH
+ 1 + 10
-pKa
2
+ pH

37

where [HSA
-
] represents the concentration of the singly deprotonated SA species and [SA]
f
=
[SA] not complexed by Cu. The conditional proton dissociation constants, pKa
1
and pKa
2
, are
8.85 and 11.39, respectively, at an ionic strength of 1.0 mol L
-1
and are 8.84 and 11.46,
respectively, at an ionic strength of 0.5 mol L
-1
(Moffett and Dupont 2007).
A 10
-2
mol L
-1
Cu stock solution in Milli-Q was prepared from anhydrous cupric sulfate
(CuSO
4
; ≥ 99.99% trace metals basis, Aldrich) and 1 µmol L
-1
sub-stocks were prepared daily by
serial dilution. A 10
-2
mol L
-1
SA stock solution in Milli-Q was prepared and 10
-3
mol L
-1
sub-
stocks were prepared daily by serial dilution. The aliquots were spiked with varying Cu
concentrations (0 – 16 nmol L
-1
) and left to equilibrate for at least an hour. Subsequently, SA (1
or 2.5 µmol L
-1
) was added and the samples were allowed to equilibrate for at least 2 hours,
which was sufficient time to obtain steady-state values (Moffett and Dupont 2007). These
detection windows, defined as the optimal SA concentrations to obtain the lowest detection limit
of the analytical signal and minimize error propagation in the calculations to determine Cu
2+
,
have been used successfully by other researchers (Buck and Bruland 2005; Moffett and Dupont
2007) and were thus chosen to make our data comparable with theirs. Ten-milliliter sub-samples
were transferred to a Teflon sample cup and attached to the electrode. Prior to analysis, the sub-
samples were purged for 3 min with high-purity N
2
. We began the analyses using the 1 µmol L
-1

analytical window, but after Sta. 5 and 7 it became apparent that Cu was more strongly
complexed than we had expected so we switched to a higher analytical window (2.5 µmol L
-1
) as
we moved into the heart of the OMZ (Sta. 9 and 11). For Sta. 12, we returned to the lower
analytical window because the conditions there were more oligotrophic.

38
Results
Nitrite—Nitrite concentrations within the PNM were generally higher in the northern transect,
along 10°S (Fig. 2C), than they were in the southern one, along 20°S (Fig. 2D). With the
exception of Sta. 3, at which it slightly exceeded 1 µmol L
-1
, [NO
2
-
] in the southern transect
typically averaged between 0.01 and 0.6 µmol L
-1
and was often undetectable. By contrast,
[NO
2
-
] in the northern transect regularly exceeded 1 µmol L
-1
, reaching values as high as around
4 µmol L
-1

at Sta. 11, the heart of the OMZ. Secondary nitrite maxima (SNM) were also
observed at Sta. 9 through 11, coincident with the depth of the OMZ. At Sta. 10, [NO
2
-
]
was

slightly higher in the SNM than the PNM.
Total dissolved copper—Depth profiles for Cu exhibited the general features reported
elsewhere: depletion of Cu in surface waters and a gradual increase with depth (Boyle et al.
1977). However, a closer examination of individual profiles reveals differences that provide new
information.
Total dissolved Cu concentrations were typically more elevated in the northern transect (Fig.
2E) than they were in the southern transect (Fig. 2F), particularly in the euphotic zone. The
northern transect was closer to the coast and had a pronounced SNM, which the southern transect
lacked; both factors appear related to the north-south differences. At Sta. 12 through 5, [Cu] in
the upper 200 m was generally ~0.4 – 0.7 nmol L
-1
and dipped as low as ~ 0.26 nmol L
-1
. At Sta.
7 through 11, [Cu] in the upper 200 m generally ranged from ~ 0.6 to 0.95 nmol L
-1
. Low [Cu]
in the upper water column suggests extensive scavenging near the surface, likely the result of
biological uptake. Indeed, it is noteworthy that the concentrations reported at Sta. 12 through 5
in the upper water column are some of the lowest oceanic values reported anywhere. Such low
concentrations suggest that biological scavenging is stronger here than in regions like the North

39
Pacific, where surface concentrations are typically higher (Boyle et al. 1977; Coale and Bruland
1990).
Higher [Cu] at Sta. 10 – 12, closer to the coast, could indicate inputs from shelf waters and
shelf and slope sediments. Interestingly, dissolved [Cu] was elevated close to and within the
SNM in the northern transect relative to the same depths in the southern transect, which lacked a
SNM. This suggests that processes contributing to the accumulation of NO
2
-
and dissolved Cu
are somehow related.
Copper speciation—Free Cu
2+
concentrations were low, ranging between
~3.15 × 10
-15
mol L
-1
and 1.34 × 10
-13
mol L
-1
(Figs. 3A–E), indicating strong organic
complexation throughout the water column—consistent with previous findings in the North
Pacific (Coale and Bruland 1988; Moffett and Dupont 2007). These concentrations are also up
to one order of magnitude lower than those reported in the North Pacific by Moffett and Dupont
(2007). The lowest concentrations of free Cu
2+
at each station except Sta. 5 (Fig. 3B–E,
indicated by the arrow) were at the base of the euphotic zone, either within or just below the
PNM. For these stations, the lowest [Cu
2+
] there was significantly lower than the depths
immediately above and below it, suggesting that processes at the base of the euphotic zone are
related to these low values. Sta. 5 (Fig. 3A) was the only station where the minimum value was
not located at the base of the euphotic zone and where the local minimum value was not
significantly different from the values above and below it. This station had a much more weakly
developed PNM with maximum nitrite concentrations at the other stations three- to fifteen-fold
higher.
Conditional stability constants were high at all stations and ranged from 12.5 to 14.0 (Table
1). Several of the highest values (> 13.5) were for samples collected from below 400 m,

40
demonstrating that Cu
2+
is strongly complexed even at depth. The log K values for Sta. 9 and 11
equaled or exceeded 13.0 at all depths; though not as consistently high, most of the values for
Sta. 12 and several of those for Sta. 7 exceeded 13.0 as well. At Sta. 7 and 9, the highest log K
values were associated with the PNM, where we observed local minima in [Cu
2+
]. The highest
ligand concentrations (3.25 – 6.46 nmol L
-1
) were measured at Sta. 7, one of the most
oligotrophic stations. The stations within the OMZ region (Sta. 9 and 11) exhibited a much
larger range (1.02 – 4.24 nmol L
-1
). The highest ligand concentrations were never directly
correlated with the PNM, and the local [Cu
2+
] minima we observed there seem to be driven by
the conditional stability constants. That suggests that low free [Cu
2+
] around the PNM arises
because of strong biological uptake, with the residual Cu bound to only the very strongest
ligands, rather than a local maximum in ligand concentration due to production of a strong ligand
by microbes localized within that zone.

Discussion
The remarkably low [Cu
2+
] and dissolved [Cu] values reported in this study are indicative of
strong organic complexation and extensive biological scavenging in the euphotic zone. Free
Cu
2+

minima were within or slightly below the PNM at all stations except Sta. 5, which was
characterized by a much weaker PNM. Oceanic diatom strains are believed to possess a high Cu
requirement for an inducible Fe acquisition mechanism that can be activated when [Fe] becomes
extremely low (Peers et al. 2005). The low dissolved [Fe] (generally ≤ 0.1 nmol L
-1
) (Y. Kondo
and J. W. Moffett unpubl.) observed at most stations suggests that this mechanism could be an
important control on Cu levels in this region. Indeed, the results from culture work with T.
oceanica (Maldonado et al. 2006) seem to indicate that Cu limitation can become significant

41
when [Cu
2+
] falls below ~10
-14
mol L
-1

—further exacerbating Fe limitation in diatoms—which
was observed frequently in this study, especially within the PNM.
Recent culture studies with an ammonium oxidizing archaeon (S. A. Amin pers. comm.) have
also indicated that Cu limitation on growth can occur once [Cu
2+
] approaches the 10
-14
mol L
-1

threshold. The association between the [Cu
2+
] minima and PNM suggests competing demands
for Cu by AOA and phytoplankton engaged in nitrate reduction, both processes resulting in the
accumulation of nitrite at this depth. Our data indicate that [Cu
2+
] is at or below the putative
threshold for growth limitation of Fe-limited diatoms at 3 of the 5 stations. It is not possible for
us to predict how Cu limitation might affect the competition between AOA and phytoplankton
for ammonia without further work. However, Cu is required for all forms of ammonium
monooxygenase, and its importance in archaea appears widespread (Walker et al. 2010). By
contrast, Cu limitation has been studied in a relatively small number of diatom taxa. Therefore it
seems likely that Cu limitation, if important, imposes a more fundamental constraint on
nitrification than photosynthetic production. The high numbers of AOA that have been reported
to thrive in the ETSP OMZ waters could explain why [Cu
2+
] often remains low even below the
PNM, however (Molina et al. 2010; Stewart et al. 2011).
One notable aspect of our findings is the presence of high dissolved [Cu] in the SNM we
encountered at Sta. 10 and 11. These concentrations were significantly higher than those we
observed at the same depths in the southern transect, perhaps because the northern transect is
closer to benthic sources at the same isopycnals. Another source of copper in the SNM is the
active remineralization of sinking organic matter from the productive surface waters through
dissimilatory nitrate reduction (Stewart et al. 2011). Such inputs are probably higher at Sta. 11
than Sta. 9 and may contribute to the higher free Cu
2+
concentrations at Sta. 11. Elevated [Cu] in

42
the SNM is surprising because it is often assumed that Cu levels in anoxic regimes such as
OMZs should be low due to the formation of kinetically inert sulfide precipitates (Saito et al.
2003). While OMZs are nitrate-replete, and sulfide is therefore thermodynamically unstable,
there is considerable evidence that it is important. Theberge et al. (1997) found that sulfide
levels within the Arabian Sea OMZ were slightly elevated relative to the surface waters, which
they attributed primarily to the release of sulfur from the respiration of organic matter. They
concluded that virtually all of this sulfide was bound to Cu and Zn. A plausible source of this
sulfide has been reported by Canfield et al. (2010) who suggest that sulfate reduction may be
more active in OMZ waters than previously thought because of a recently discovered cryptic
sulfur cycle. The accumulation of excess dissolved Cu under the conditions observed in this
study suggests that such complexes enhance rather than diminish the solubility of Cu within
OMZs, perhaps by decreasing bioavailability. One important gap in our knowledge here is the
valence of Cu in the SNM—or elsewhere for that matter. Leal and Van den Berg (1998)
demonstrated that it is impossible to distinguish strong Cu
+
and Cu
2+
complexes using CSV,
which they demonstrated with model Cu
+
complexes that may actually be present in seawater.
Moffett and Zika (1988) reported Cu
+
in surface waters that was presumed derived from
photochemical processes, but concluded that their methodology would not detect Cu
+
strongly
bound by reduced sulfur.
In summary, our results suggest that Cu is tightly complexed throughout the water column in
the ETSP, and this finding is consistent with the high biological demand for Cu we anticipate
based on physiological studies conducted by other investigators (Peers and Price 2005;
Maldonado et al. 2006). High biological demand is likely to be associated with the various
nitrogen cycle processes that are active in this region, particularly in the PNM, where the lowest

43
free [Cu
2+
] was observed. The distribution of total dissolved Cu is intriguing and requires more
study. The extremely low total dissolved [Cu] in the euphotic zone of the southern transect—the
lowest levels reported worldwide—are surprising given the highly oligotrophic conditions there.
The high dissolved [Cu] in the SNM is also novel and unexpected and must reflect high inputs
coupled with accumulation as strong, soluble complexes.

Bibliography
Boyle, E. A., F. Sclater, and J. Edmond. 1977. The distribution of dissolved copper in the
Pacific. Earth Planet. Sci. Lett. 37: 38–54, doi:10.1016/0012-821X(77)90144-3
Bruland, K. W., E. L. Rue, G. J. Smith, and G. R. DiTullio. 2005. Iron, macronutrients and
diatom blooms in the Peru upwelling regime: Brown and blue waters of Peru. Mar. Chem.
93: 81–103, doi:10.1016/j.marchem.2004.06.011
Buck, K. N., and K. W. Bruland. 2005. Copper speciation in San Francisco Bay: A novel
approach using multiple analytical windows. Mar. Chem. 96: 185–198,
doi:10.1016/j.marchem.2005.01.001
Buck, K. N., J. Moffett, K. A. Barbeau, R. M. Bundy, Y. Kondo, and J. Wu. 2012. The organic
complexation of iron and copper: an intercomparison of competitive ligand exchange-
adsorptive cathodic stripping voltammetry (CLE-ACSV) techniques. Limnol. Oceanogr.:
Methods 10: 496–515, doi:10.4319/lom.2012.10.496
Campos, M. L. A. M., and C. M. G. Van den Berg. 1994. Determination of copper complexation
in sea water by cathodic stripping voltammetry and ligand competition with salicylaldoxime.
Anal. Chim. Acta 284: 481–496, doi:10.1016/0003-2670(94)85055-0
Canfield, D. E., F. J. Stewart, B. Thamdrup, L. De Brabandere, T. Dalsgaard, E. F. Delong, N. P.

44
Revsbech, and O. Ulloa. 2010. A cryptic sulfur cycle in oxygen-minimum-zone waters off
the Chilean coast. Science 330: 1375–1378, doi:10.1126/science.1196889
Coale, K. H., and K. W. Bruland. 1988. Copper complexation in the Northeast Pacific. Limnol.
Oceanogr. 33(5): 1084–1101, doi:10.4319/lo.1988.33.5.1084
Coale, K. H., and K. W. Bruland. 1990. Spatial and temporal variability in copper complexation
in the North Pacific. Deep-Sea Res. 37: 317–336, doi:10.1016/0198-0149(90)90130-N
Francis, C. A., K. J. Roberts, J. M. Beman, A. E. Santoro, and B. B. Oakley. 2005. Ubiquity and
diversity of ammonia-oxidizing archaea in water columns and sediments of the ocean. Proc.
Nat. Acad. Sci. U.S.A. 102(41): 14683–14688, doi:10.1073/pnas.0506625102
Johnson, K. S., V. Elrod, S. Fitzwater, J. Plant, E. Boyle, B. Bergquist, K. Bruland, A. Aguilar-
Islas, K. Buck, M. Lohan, G. J. Smith, B. Sohst, K. Coale, M. Gordon, S. Tanner, C.
Measures, J. Moffett, K. Barbeau, A. King, A. Bowie, Z. Chase, J. Cullen, P. Laan, W.
Landing, J. Mendez, A. Milne, H. Obata, T. Doi, L. Ossiander, G. Sarthou, P. Sedwick, S.
Van den Berg, L. Laglera-Baquer, J. Wu, and Y. Cai. 2007. Developing standards for
dissolved iron in seawater. Eos Trans. AGU 88(11): 131–132, doi:10.1029/2007EO110003
Kogut, M. B. 2002. Cu speciation in estuaries and coastal waters. Ph.D. thesis. Massachusetts
Institute of Technology.
Könneke, M., A. E. Bernhard, J. R. de La Torre, C. B. Walker, J. B. Waterbury, and D. A. Stahl.
2005. Isolation of an autotrophic ammonia-oxidizing marine archaeon. Nature 437: 543–546,
doi:10.1038/nature03911
Kustka, A., E. J. Carpenter, and S. A. Sañudo-Wilhelmy. 2002. Iron and marine nitrogen
fixation: Progress and future directions. Res. Microbiol. 153: 255–262, doi:10.1016/S0923-
2508(02)01325-6

45
La Roche, J., H. Murray, M. Orellana, and J. Newton. 1995. Flavodoxin expression as an
indicator of iron limitation in marine diatoms. J. Phycol. 29: 520–530, doi:10.1111/j.1529-
8817.1995.tb02545.x
Lam, P., G. Lavik, M. M. Jensen, J. van de Vossenberg, M. C. Schmid, D. Woebken, D.
Gutiérrez, R. Amann, M. S. M. Jetten, and M. M. M. Kuypers. 2009. Revising the nitrogen
cycle in the Peruvian oxygen minimum zone. Proc. Nat. Acad. Sci. U.S.A. 106(12): 4752–
4757, doi:10.1073/pnas.0812444106
Leal, M. F. C., and C. M. G. Van den Berg. 1998. Evidence for strong copper(I) complexation by
organic ligands in seawater. Aquat. Geochem. 4: 49–75, doi:10.1023/A:1009653002399
Lee, J.-M., E. A. Boyle, Y. Echegoyen-Sanz, J. N. Fitzsimmons, R. Zhang, and R. A. Kayser.
2011. Analysis of trace metals (Cu, Cd, Pb, and Fe) in seawater using single batch
nitrilotriacetate resin extraction and isotope dilution inductively coupled plasma mass
spectrometry. Anal. Chim. Acta 686: 93–101, doi:10.1016/j.aca.2010.11.052
Lomas, M. W., and F. Lipschultz. 2006. Forming the primary nitrite maximum: Nitrifiers or
phytoplankton? Limnol. Oceanogr. 51(5): 2453–2467, doi:10.4319/lo.2006.51.5.2453
Maldonado, M. T., A. E. Allen, J. S. Chong, K. Lin, D. Leus, N. Karpenko, and S. L. Harris.
2006. Copper-dependent iron transport in coastal and oceanic diatoms. Limnol. Oceanogr.
51(4): 1729–1743, doi:10.4319/lo.2006.51.4.1729
Milligan, A. J., and P. J. Harrison. 2000. Effects of non‐steady‐state iron limitation on nitrogen
assimilatory enzymes in the marine diatom Thalassiosira weissflogii (Bacillariophyceae). J.
Phycol. 36: 78–86, doi:10.1046/j.1529-8817.2000.99013.x
Moffett, J. W., and R. G. Zika. 1988. Measurement of copper(I) in surface waters of the
subtropical Atlantic and Gulf of Mexico. Geochim. Cosmochim. Acta. 52: 1849–1857,

46
doi:10.1016/0016-7037(88)90008-7
Moffett, J. W., and C. L. Dupont. 2007. Cu complexation by organic ligands in the sub-arctic
NW Pacific and Bering Sea. Deep-Sea Res. 54: 586–595, doi:10.1016/j.dsr.2006.12.013
Molina, V., L. Belmar, and O. Ulloa. 2010. High diversity of ammonia-oxidizing archaea in
permanent and seasonal oxygen-deficient waters of the eastern South Pacific. Environ.
Microbiol. 12: 2450–2465, doi:10.1111/j.1462-2920.2010.02218.x
Peers, G., S. A. Quesnel, and N. M. Price. 2005. Copper requirements for iron acquisition and
growth of coastal and oceanic diatoms. Limnol. Oceanogr. 50(4): 1149–1158, doi:
10.4319/lo.2005.50.4.1149
Philippot, L. 2002. Denitrifying genes in bacterial and archaeal genomes. Biochim. Biophys.
Acta Gene Struct. Expr. 1577: 355–376, doi:10.1016/S0167-4781(02)00420-7
Saito, M. A., D. M. Sigman, and F. M. M. Morel. 2003. The bioinorganic chemistry of the
ancient ocean: The co-evolution of cyanobacterial metal requirements and biogeochemical
cycles at the Archean–Proterozoic boundary? Inorg. Chim. Acta 356: 308–318,
doi:10.1016/S0020-1693(03)00442-0
Sedwick, P. N., T. M. Church, A. R. Bowie, C. M. Marsay, S. J. Ussher, K. M. Achilles, P. J.
Lethaby, R. J. Johnson, M. M. Sarin, and D. J. McGillicuddy. 2005. Iron in the Sargasso Sea
(Bermuda Atlantic Time-series Study region) during summer: Eolian imprint, spatiotemporal
variability, and ecological implications. Glob. Biogeochem. Cycles 19: GB4006,
doi:10.1029/2004GB002445
Stewart, F. J., O. Ulloa, and E. F. Delong. 2011. Microbial metatranscriptomics in a permanent
marine oxygen minimum zone. Environ. Microbiol. 14: 23–40, doi:10.1111/j.1462-
2920.2010.02400.x

47
Strickland, J. D. H., and T. R. Parsons. 1968. A practical handbook of seawater analysis, 2nd ed.
Fisheries Research Board of Canada.
Theberge, S. M., G. W. Luther III, and A. M. Farrenkopf. 1997. On the existence of free and
metal complexed sulfide in the Arabian Sea and its oxygen minimum zone. Deep-Sea Res. II
44: 1381–1390, doi:10.1016/S0967-0645(97)00012-X
Treusch, A. H., S. Leininger, A. Kletzin, S. C. Schuster, H.-P. Klenk, and C. Schleper. 2005.
Novel genes for nitrite reductase and Amo-related proteins indicate a role of uncultivated
mesophilic crenarchaeota in nitrogen cycling. Environ. Microbiol. 7: 1985–1995,
doi:10.1111/j.1462-2920.2005.00906.x
Walker, C. B., J. R. de La Torre, M. Klotz, H. Urakawa, N. Pinel, D. Arp, C. Brochier-Armanet,
P. Chain, P. Chan, A. Gollabgir, J. Hemp, M. Hügler, E. A. Karr, M. Könneke, M. Shin, T. J.
Lawton, T. Lowe, W. Martens-Habbena, L. A. Sayavedra-Soto, D. Lang, S. M. Sievert, A.
C. Rosenzweig, G. Manning, and D. A. Stahl. 2010. Nitrosopumilus maritimus genome
reveals unique mechanisms for nitrification and autotrophy in globally distributed marine
crenarchaea. Proc. Nat. Acad. Sci. U.S.A. 107(19): 8818–8823,
doi:10.1073/pnas.0913533107

48
Table 1. Speciation data for Sta. 5 – 12. Note that the log K interval (in parentheses) is
asymmetric.

Station Depth (m) [L] (nmol L
-1
) log K
5 20 2.09 ± 0.44 13.1 (12.9 – 13.2)

140 2.32 ± 0.29 12.9 (12.7 – 13.0)

165 2.78 ± 0.51 12.7 (12.5 – 12.8)

550 2.85 ± 0.33 12.7 (12.5 – 12.9)

1000 3.33 ± 0.32 12.8 (12.6 – 12.9)
7 25 4.71 ± 0.75 12.6 (12.5 – 12.7)

60 5.79 ± 0.41 12.8 (12.8 – 12.9)

80 3.33 ± 0.24 13.8 (13.6 – 13.9)

90 4.53 ± 0.43 12.6 (12.5 – 12.7)

130 3.30 ± 0.32 13.4 (13.1 – 13.5)

160 3.80 ± 0.29 12.7 (12.6 – 12.8)

230 3.25 ± 0.32 13.3 (13.1 – 13.4)

350 3.61 ± 0.47 13.0 (12.7 – 13.1)

1000 6.46 ± 0.60 12.5 (12.4 – 12.6)
9 20 3.08 ± 0.16 13.7 (13.6 – 13.7)

50 1.02 ± 0.04 14.0 (13.9 – 14.0)

90 3.26 ± 0.14 14.0 (13.9 – 14.0)

95 3.06 ± 0.30 14.0 (13.9 – 14.1)

110 3.21 ± 0.21 13.7 (13.5 – 13.8)

135 4.24 ± 0.33 13.1 (13.0 – 13.1)

325 3.84 ± 0.12 13.9 (13.8 – 13.9)

600 2.98 ± 0.26 13.4 (13.3 – 13.5)
11 20 3.28 ± 0.46 13.8 (13.6 – 13.9)

50 1.85 ± 0.24 13.8 (13.7 – 13.9)

60 2.41 ± 0.26 13.4 (13.3 – 13.6)

75 3.21 ± 0.15 13.8 (13.8 – 13.9)

120 3.54 ± 0.30 14.0 (13.8 – 14.1)

250 2.37 ± 0.12 14.0 (13.9 – 14.0)

340 4.11 ± 0.23 13.3 (13.3 – 13.4)

380 3.33 ± 0.45 13.2 (12.7 – 13.4)

450 2.45 ± 0.22 13.2 (13.0 – 13.3)

700 2.61 ± 0.38 13.0 (12.7 – 13.1)

1000 3.93 ± 0.30 13.8 (13.6 – 14.0)
12 20 2.01 ± 0.52 12.6 (12.4 – 12.7)

80 1.69 ± 0.17 13.4 (13.3 – 13.4)

150 1.20 ± 0.13 13.6 (13.5 – 13.7)

250 2.01 ± 0.18 13.2 (13.0 – 13.3)

300 2.02 ± 0.26 13.3 (13.1 – 13.4)

340 3.41 ± 0.29 12.6 (12.4 – 12.7)

400 1.55 ± 0.25 13.9 (13.6 – 14.1)

800 2.19 ± 0.15 13.9 (13.8 – 14.0)

1000 2.24 ± 0.15 13.8 (13.7 – 13.9)

49
Figure Legends
Fig. 1. Map of the cruise track in the eastern tropical South Pacific Ocean that was sampled
from 19 January to 03 March 2010 aboard the R/V Atlantis.

Fig. 2. Upper 1000 m oceanographic sections of dissolved oxygen ((A) north; (B) south), nitrite
((C) north; (D) south), and total dissolved Cu ((E) north; (F) south)) concentrations. The
southern section ranged from 80
o
W to 100
o
W along 20
o
S (Sta. 2, 3, 4, 5, 12). The northern
section ranged from 82
o
30’W to 100
o
W along 10
o
S (Sta. 7, 8, 9, 10, 11).

Fig. 3. Depth profiles of log [Cu
2+
] (closed circles), total dissolved Cu (open triangles) and
nitrite (open squares) for Sta. (A) 5, (B) 7, (C) 9, (D) 11, and (E) 12. Note the arrow indicating
the [Cu
2+
] minimum. Error bars for log [Cu
2+
] and total dissolved [Cu] represent error
propagation from the calculation of [L] and K and standard deviation values (n=3), respectively.

50

Fig. 1.

Ocean Data View
10
o
S
15
o
S
20
o
S
100
o
W 95
o
W 90
o
W 85
o
W 80
o
W 75
o
W
Longitude
Latitude
12 2 3 4 5
7 89 10 11

51

Fig. 2.

Sta. 7 Sta. 8 Sta. 9 Sta. 10 Sta. 11
Sta. 7 Sta. 8 Sta. 9 Sta. 10 Sta. 11
Sta. 7 Sta. 8 Sta. 9 Sta. 10 Sta. 11
Sta. 5 Sta. 4 Sta. 3 Sta. 2 Sta. 12
Sta. 5 Sta. 4 Sta. 3 Sta. 2 Sta. 12
Sta. 5 Sta. 4 Sta. 3 Sta. 2 Sta. 12
0
200
400
600
800
1000
0
200
400
600
800
1000
0
200
400
600
800
1000
0
200
400
600
800
1000
0
200
400
600
800
1000
0
200
400
600
800
1000
2.5
2
1.5
1
0.5
0
3
2.5
2
1.5
1
0.5
0
3
1.5
1
0.75
0.5
0.25
0
1.75
1.25
1.5
1
0.75
0.5
0.25
0
1.75
1.25
100
80
60
40
20
0
100
80
60
40
20
0
Depth (m) Depth (m) Depth (m)
100
o
W 95
o
W 90
o
W 85
o
W
100
o
W 95
o
W 90
o
W 85
o
W
100
o
W 95
o
W 90
o
W 85
o
W
100
o
W 95
o
W 90
o
W 85
o
W 80
o
W
100
o
W 95
o
W 90
o
W 85
o
W 80
o
W
100
o
W 95
o
W 90
o
W 85
o
W 80
o
W
A) DO
(µmol L
-1
)
C) NO
2
-
(µmol L
-1
)
E) Cu
(nmol L
-1
)
B) DO
(µmol L
-1
)
D) NO
2
-
(µmol L
-1
)
F) Cu
(nmol L
-1
)
Longitude Longitude

52

Fig. 3.

S
S S
S
S
S
S
S
S
S S
S
S
S
S
S
S
S
S
S
S
S
S
S
S S
S S
S
SS
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
log Cu
2+
(mol L
-1
) log Cu
2+
(mol L
-1
) log Cu
2+
(mol L
-1
)
-15 -14.5 -14 -13.0 -13.5 -12.5 -15 -14.5 -14 -13.0 -13.5 -12.5 -15 -14.5 -14 -13.0 -13.5 -12.5
0 0.5 1.0 1.5
Dissolved Cu (nmol L
-1
)
2.0 0 0.5 1.0 1.5
Dissolved Cu (nmol L
-1
)
2.0 0 0.5 1.0 1.5
Dissolved Cu (nmol L
-1
)
2.0
0 0.5 1.0 1.5
Dissolved Cu (nmol L
-1
)
2.0 0 0.5 1.0 1.5
Dissolved Cu (nmol L
-1
)
2.0
log Cu
2+
(mol L
-1
)
-15 -14.5 -14 -13.0 -13.5 -12.5
log Cu
2+
(mol L
-1
)
-15 -14.5 -14 -13.0 -13.5 -12.5
Nitrite (μmol L
-1
)
0 1 1.5 2 4 0.5 3.5 2.5
Nitrite (μmol L
-1
)
0 1 1.5 2 4 0.5 3.5 2.5
Nitrite (μmol L
-1
)
0 1 1.5 2 4 0.5 3.5 2.5
Nitrite (μmol L
-1
)
0 1 1.5 2 4 0.5 3.5 2.5
Nitrite (μmol L
-1
)
0 1 1.5 2 4 0.5 3.5 2.5
0
200
400
600
800
1000
Depth (m)
0
200
400
600
800
1000
Depth (m)
0
200
400
600
800
1000
0
200
400
600
800
1000
0
200
400
600
800
1000
B) Sta. 7 C) Sta. 9
D) Sta. 11 E) Sta. 12
A) Sta. 5
Cu
log Cu
2+
NO
2
-

53
Chapter 3
Assessment of the potential for copper limitation of ammonia oxidation by Archaea in a
dynamic estuary

Acknowledgments
We are grateful to Captain Ray McQuin and First Mate Greg Buikema of the R/V Clifford A.
Barnes and to Maija Heller, Katherine Heal, Laura Truxal, Kathy Krogslund, Davey French, and
Yang Han for assistance in sample collection and nutrient measurements. We would also like to
thank Jason Visser at the University of Southern California for assisting us with the ICP-MS.

54
Abstract
The distribution and speciation of copper (Cu) in Hood Canal, a fjord in western Puget Sound,
Washington, was studied over a 2-year period. In Hood Canal, ammonia oxidation is largely
dominated by ammonia-oxidizing archaea (AOA), which have high Cu requirements for other
processes as well. Dissolved Cu was slightly depleted in the upper water column, and
concentrations were almost unchanged from measurements made in the late 1970s, ranging from
4.08 – 6.12 nmol L
-1
. Although this implies that the biological demand is small relative to the
large and relatively constant inventory of dissolved Cu, and that Cu limitation is therefore
unlikely to influence rates of biological processes, speciation measurements indicated that
dissolved Cu is strongly complexed by organic ligands. As a result, bioavailable Cu
2+

concentrations were considerably lower, varying from 6.14 × 10
-15
mol L
-1
to 1.36 × 10
-12
mol L
-
1
. Furthermore, Cu
2+
displayed a clear trend over most sampling periods, with Cu
2+

concentrations one to two orders of magnitude higher below 20 m in the deeper, saline waters,
and exhibiting minima in the upper 15 m. The major freshwater input to Hood Canal is not an
important source of ligands, which suggests that the ligands are likely produced biologically in
the water column and have slow turnover times. In general, ammonia oxidation rates varied
considerably but were lowest in the upper water column where Cu
2+
concentrations were also
lowest.

55
Introduction
There is considerable interest in the speciation of copper (Cu) in estuaries, arising from its
toxicity to marine organisms (Moffett et al. 1997; Buck and Bruland 2005). More recently there
has been interest in Cu limitation of various enzymatic processes that are important in the carbon
and nitrogen cycles (Peers et al. 2005; Maldonado et al. 2006; Jacquot et al. 2013). Cu limitation
can arise when dissolved Cu is tightly complexed by natural organic ligands and Cu
2+

concentrations are sufficiently low (Moffett et al. 2012). Possibly the most Cu-dependent
members of the marine microbial community are the ammonia-oxidizing archaea (AOA)
(Walker et al. 2010). A genomic survey of the recently isolated ammonia-oxidizing archaeon
Nitrosopumilus maritimus SCM1 (hereon called SCM1; Könneke et al. 2005) suggested a high
reliance on Cu for many aspects of its basic physiology (Walker et al. 2010). In addition to
encoding the Cu-dependent metalloenzyme ammonium monooxygenase, which catalyzes the
oxidation of ammonia (NH
3
) to hydroxylamine in the first step of nitrification (Vajrala et al.
2013), the SCM1 genome encodes numerous multi-Cu oxidases and blue Cu proteins resembling
sulfo- and plastocyanin enzymes that are associated with the electron transport system (Walker et
al. 2010). Genes encoding similar blue Cu proteins have been discovered in other AOA
genomes (Hallam et al. 2006; Blainey et al. 2011), suggesting that a high demand for Cu may be
more widespread among AOA than previously thought.
This contrasts sharply with ammonia-oxidizing bacteria (AOB), the other major ammonia
oxidizers (Ensign et al. 1993), which use iron (Fe)-dependent metalloenzymes in their electron
transport system and should therefore have a lower reliance on Cu. Several studies have also
shown that AOA are often ubiquitous in oceanic sites with high nitrification activity and that
their numbers can outstrip those of AOB by often considerable margins (Francis et al. 2005;

56
Beman et al. 2010). These findings underscore the likely magnitude of the contributions of AOA
to nitrification worldwide and the oceanic nitrogen cycle as a whole.
More recently, Jacquot et al. (2013) revealed the existence of a tight correlation between the
depths of the primary nitrite maxima (PNM) and local Cu
2+
minima in the oligotrophic waters of
the eastern tropical South Pacific (ETSP). They postulated that this feature arose because of
heavy Cu
2+
drawdown by AOA and Fe-limited diatoms engaged in ammonia oxidation and
nitrate reduction, respectively. Both processes are mediated by Cu-dependent metalloenzymes
and produce nitrite (Francis et al. 2005; Maldonado et al. 2006). Peers et al. (2005) reported that
Cu limitation in the oceanic diatom Thalassiosira oceanica (equivalent to 65% of its maximal
growth rate, µ
max
) can occur when the Cu
2+
concentration drops below ~1.25 × 10
-14
mol L
-1
.
Amin et al. (2013) demonstrated Cu limitation in SCM1 (76% of µ
max
) when the Cu
2+

concentration falls below ~2 × 10
-13
mol L
-1
. Cu
2+
concentrations in the ETSP, particularly
within the PNM, often fell well below either putative limitation threshold (Jacquot et al. 2013).
Although a lack of supporting data ultimately precluded Jacquot et al. (2013) from determining
whether ammonia oxidation by AOA or nitrate reduction by Fe-limited diatoms was most
responsible for the feature, they argued that the ubiquity of AOA in the region (Molina et al.
2010) and their absolute dependence on Cu (Walker et al. 2010) meant that Cu limitation had the
potential to impose larger constraints on ammonia oxidation.
One objective of our study was to evaluate whether the measured threshold for Cu limitation
reported by Amin et al. (2013) is relevant in coastal environments. We selected a study site in
Hood Canal, a long (110 km), narrow (1 – 2 km) sub-basin of the Puget Sound estuary in
Washington (Newton et al. 2011), where AOA have recently been shown to dominate ammonia
oxidation (Horak et al. 2013). Spring blooms in Hood Canal are long-lived, intense and last

57
from early spring, or even late winter, until late summer. Bloom dynamics typically switch from
light limitation in the winter to nutrient limitation—generally nitrate (Mackas and Harrison
1997)—in the summer (Devol et al. 2011). Given the basin’s high biological productivity (4 g C
mg
-2
d
-1
on average; Newton et al. 2011) and abundant AOA (Horak et al. 2013), we anticipated
a strong biological demand for Cu, particularly in the upper water column. The biological
uptake of Cu is associated with diatoms and AOA and may also be important for other taxa that
have not been studied. A survey of the Cu requirements in the existing literature suggests that
AOA are the most easily Cu-limited microbes studied to date and are therefore the focus of our
study (Amin et al. 2013).
In this paper, we characterize the distribution and speciation of Cu in seawater in Hood Canal
on 4 occasions from July 2011 to August 2012 along with the distribution of ammonia and
ammonia oxidation rates. The objective was to determine if there was a drawdown in dissolved
Cu and Cu
2+
in areas of AOA activity and to see if the Cu
2+
concentration might fall below the
limiting threshold identified in Amin et al. (2013).

Methods
Study site and sample collection—Sampling for this study was performed over a 2-year
period during 4 cruises aboard the R/V Clifford A. Barnes: CB960 (17 July to 22 July 2011);
CB974 (7 May to 13 May 2012); CB980 (16 July to 22 July 2012); and CB985 (24 August to 30
August 2012) near an ocean remote chemical analyzer (ORCA) buoy (Hoodsport; latitude,
47°25'18.48"N; longitude, 123°6'45.36"W) operated by the School of Oceanography at the
University of Washington (http://orcabase.ocean.washington.edu/). Water samples from the
surface (2 m) down to just above the bottom (115 m) were collected over several days using 10 L

58
Teflon-coated Go-Flo bottles (General Oceanics) attached to Kevlar wire. Upon retrieval the
bottles were immediately pressurized with filtered compressed nitrogen gas and the water was
forced through acid-cleaned Teflon tubing and acid-cleaned 0.2 µm Acropak capsules (Pall
Corporation) into 1 L fluorinated low-density polyethylene (FLPE) bottles (Nalgene, Nalge Nunc
International) and 250 mL low-density polyethylene (LDPE) bottles (Nalgene, Nalge Nunc
International) on deck for speciation and dissolved analyses, respectively. The bottles were
washed in a sequential 4-step process: 1) soaked for at least 24 hours in a 5% Citranox acid
detergent bath (Alconox), 2) soaked for at least 24 hours in a 10% hydrochloric acid bath (HCl;
Van Waters and Rogers (VWR) International), 3) filled with 10% HCl and baked at 60°C upright
and upside down (to properly leach the threads around the cap) for at least 48 hours, and 4) filled
with 0.1% trace metal grade HCl (Optima, Fisher) and baked at 60°C again for at least 48 hours.
In between each step the insides and outsides of the bottles were thoroughly rinsed at least 5
times with Milli-Q water (18.2 MΩ; Millipore). As a final measure, the bottles were rinsed at
least 5 times with sample seawater prior to collection to ensure that all the acid had been
removed. The 125 mL polycarbonate bottles (Nalgene, Nalge Nunc International) used to collect
samples for the ammonia oxidation rate measurements were also washed with 10% HCl and
rinsed multiple times with Milli-Q water.
Following the CB974 cruise, two freshwater samples were also collected on 14 May 2012 at
13:00:00 h by hand in the same acid-cleaned 1 L FLPE bottles from the North Fork Skokomish
River upstream of the Hood Canal Basin (latitude, 47°19'9.48"N; longitude, 123° 8'22.20"W).
One sample was collected 5 ft from the bridge (hereon called FW
1
) while the other was collected
directly under the bridge (hereon called FW
2
). These samples were transferred to a refrigerator
at the University of Southern California within 2 days and filtered with a vacuum filtration

59
system using acid-cleaned 0.2 µm pore membrane filters (Whatman). They were then stored in
the refrigerator and analyzed within a week.
Nutrient and ammonia oxidation rate analyses—Samples for nutrient measurements (nitrate
(NO
3
-
), nitrite (NO
2
-
), ammonium (NH
4
+
), phosphate (PO
4
3-
) and silicate (SiO
4
4-
)) and ammonia
oxidation rates were collected using a conductivity, temperature, depth (CTD) rosette (Sea-Bird
Electronics) equipped with 12 10 L Niskin bottles. Dissolved oxygen (O
2
) concentrations were
measured with a CTD sensor package (SBE-43; Sea-Bird Electronics) on the Hoodsport ORCA
buoy mooring and were calibrated against Winkler O
2
determinations for all but the CB974
cruise. Ammonium and nitrite concentrations were measured within 2 hours of sample
collection onboard using the o-phtaldialdehyde (OPA) fluorescence method (Kérouel and
Aminot 1997; Holmes et al. 1999) and spectrophotometrically (Grasshoff et al. 1999),
respectively. Samples for NO
3
-
, PO
4
3-
, and SiO
4
4-
were filtered with a 0.22 µm size pore filter
and frozen at -20°C for shore-based analysis at the University of Washington. Those analyses
were performed at the Marine Chemistry Laboratory of the School of Oceanography using a
Technicon AutoAnalyzer II (UNESCO, 1994).
Samples for ammonia oxidation rate measurements were collected in duplicates or triplicates
as described in Horak et al. (2013) by rinsing the polycarbonate bottles 3 times with seawater
from the appropriate Niskin bottle before filling them with minimal headspace after 3
overflowing volumes. Before they were sealed, 50 nM
15
NH
4
+
(99 atom percent (at %),
Cambridge Isotope Laboratories, Massachusetts) was added to each bottle. Samples were
incubated either in dark surface seawater incubators (CB960, CB974) or in situ in darkened
bottles on a free-floating array in the water column (CB980, CB985). The temperature of
incubation did not affect the ammonia oxidation rates (Horak et al. 2013). The initial δ
15
N of the

60
NO
2
-
+ NO
3
-
pool for each depth was sampled immediately after the addition of
15
NH
4
+
. The
incubations were stopped by flash-freezing 50 mL aliquots in a dry-ice ethanol slurry and the
aliquots were stored at -20°C until analysis at the University of Washington. We used the azide
method to reduce
15
NO
2
-
+
15
NO
3
-
to N
2
O for isotope ratio mass spectrometry analysis, and
ammonia oxidation rates were calculated as described in Horak et al. (2013). We applied a
correction to the rates for ammonia oxidation kinetics for all depths (K
m
= 98 nM; Horak et al.
2013).
Dissolved Cu analyses—The samples were acidified to pH 1.7 by the addition of concentrated
trace metal grade HCl (Optima, Fisher) and stored for at least 2 months before the analyses were
made. Dissolved Cu concentrations were measured using a single-batch nitrilotriacetatic acid
(NTA) resin extraction and isotope dilution inductively coupled plasma mass spectrometry (ID-
ICP-MS) method on a Finnegan Element 2 (Thermo Scientific) adapted from Lee et al. (2011) as
described in Jacquot et al. (2013). The working resin solution used for the measurements was
prepared from a concentrated NTA Superflow chelating resin (Qiagen).
Cu
2+
and organic ligand analyses— The samples were frozen at -20°C onboard and
transferred to a laboratory freezer at the University of Southern California after the cruises for
analysis. They were thawed in a refrigerator at 4°C for 2 – 4 days and were then analyzed within
a week. Recent work has shown no significant difference between Cu speciation results obtained
from fresh samples versus those obtained from frozen ones (Bruland et al. 2000; Buck et al.
2012). The Cu
2+
concentrations, organic ligand (L) concentrations and their corresponding
conditional stability constants (K) were characterized using a competitive ligand exchange
adsorptive cathodic stripping voltammetry (CLE-ACSV) adapted from Jacquot et al. (2013) and
Moffett and Dupont (2007) with salicylaldoxime (SA; ≥ 98%, Aldrich) as the competing ligand.

61
The measurements were carried out on a BioAnalytical Systems (BASi) Controlled Growth
Mercury Electrode set to the Static Mercury Drop setting (drop size: 14) and interfaced with a
BASi Epsilon ε2 voltammetric analyzer. The instrument settings, based on those used in
previous studies with the same electrode model (Buck and Bruland, 2005) were differential pulse
stripping voltammetry; deposition potential = -0.15 V (vs. Ag:AgCl (3 M NaCl) reference
electrode and platinum wire counter electrode); stir rate = 600 revolutions per minute (rpm);
deposition time, t
d
= 60 s; quiet period = 10 s; scan range = -0.15 to -0.65 V, scan rate = 20 mV
s
-1
; step potential = 4 mV; pulse width = 35 ms; pulse period = 0.2 s; pulse amplitude = 0.05 V.
Individual samples were subdivided into 20 mL aliquots in 40 mL or 60 mL Teflon bottles
(Nalgene, Nalge Nunc International). The pH of each sample was then adjusted to closely match
its ambient pH, which ranged from 7.3 to 8.4 and was measured in the laboratory using an
Orion* 3-Star Plus pH bench top meter (Thermo Scientific), with either 1 mol L
-1
3-(N-
morpholino)propanesulfonic acid (MOPS; ≥ 99.5%, Sigma), 1 mol L
-1
4-(2-hydroxyethyl)-1-
piperazineethanesulfonic acid (HEPES; ≥ 99.5%, Sigma) buffer or 1 mol L
-1
4-(2-hydroxyethyl)-
1-piperazinepropanesulfonic acid (EPPS; ≥ 99.5%, Sigma) buffer and varying amounts of 1 M
trace metal grade sodium hydroxide (NaOH; Fluka). Other details about our method, including
an explanation for why we adjusted the pH, can be found in Jacquot et al. (2013). The only
changes we made in this study were to increase the equilibration time after the addition of cupric
sulfate to at least 2 hours and to decrease the equilibration time after the addition of SA to at
least half an hour to reduce the potential for Cu wall loss (K. N. Buck pers. comm.). We also
used 5 µM SA, yielding a stronger detection window than we used in oceanic studies since we
expected to find much stronger organic complexation by natural ligands in this estuary (Jacquot
et al. 2013). Moreover, we analyzed a subset of samples from the CB960 and CB974 cruises

62
with a higher detection window, 10 µM, to probe the presence of even stronger organic ligands.
Recent studies have demonstrated the importance of employing multiple detection windows to
obtain a more complete picture of organic Cu complexation (Buck and Bruland 2005; Bundy et
al. 2012).

Results
Study site and hydrography—Salinity, temperature, and dissolved O
2
concentration depth
profiles showed strong gradients in the upper 10 m and varied slightly between cruises,
especially in the mixed layer, although they remained mostly uniform below the mixed layer (Fig.
1A–D). Average water temperature varied from 8.4°C near the bottom to 14.6°C at the surface
with values ranging from 12°C during CB974 (Fig. 1B) to 18.7°C during CB985 (Fig. 1D).
CB985 featured the sharpest temperature gradient, and thus one of the strongest density gradients,
with the temperature at the surface (2.6 m) decreasing from 18.7°C to 12.1°C at 7.4 m. Salinity
increased from 25 at the surface to 30 near the bottom during CB960, CB980 and CB985 (Fig.
1A, C–D); the surface waters were less saline (23) during CB974 but the deep waters were just
as saline as the rest (Fig. 1B). The depth of the pycnocline remained shallow over the four
cruises, occurring between 5 and 10 m, and was influenced by the warmer, less saline surface
waters. Dissolved O
2
concentrations remained in excess of 200 µmol kg
-1
in the surface waters
and stayed above 60 µmol kg
-1
in the deep (Fig. 1A–D).
Nutrients and ammonia oxidation rates—Concentrations of NO
3
-
, PO
4
3-
, and SiO
4
4-
for all
cruises were mostly characteristic of post-bloom dynamics: depleted at the surface and enriched
at depth (Fig. 2A–D, 3A–D) (Paulson et al. 1993; Devol et al. 2011). Deep-water NO
3
-

concentrations were 30 – 35 µmol L
-1
(Fig. 2A–D) while those for PO
4
3-
and SiO
4
4-
were 2.83 –

63
3.82 µmol L
-1
and 82.5 – 86.3 µmol L
-1
, respectively (Fig. 3A–D). NH
4
+
concentrations ranged
from 0 – 0.74 µmol L
-1
while NO
2
-
concentrations ranged from 0 – 0.42 µmol L
-1
(Fig. 2A–D); in
both cases they reached their highest levels during CB974 and CB985 (Fig. 2B, D). Ammonia
oxidation rates were highly variable over the four cruises and ranged from 0 – 316 nmol L
-1
day
-1

(Fig. 4A–D). The highest oxidation rates, remaining elevated even at depth, were measured
during CB960 and CB985 (Fig. 4A, D).
Dissolved Cu analyses—Dissolved Cu concentrations were relatively uniform, mostly
varying between 4 and 5 nmol L
-1
and exhibiting almost conservative profiles (Fig. 4A–D). The
lowest and highest concentrations overall were 4.08 nmol L
-1
and 6.12 nmol L
-1
. We observed
surface maxima during all cruises except CB985, when this feature was obscured by a large
subsurface maximum (Fig. 4D). For CB960, CB974, and CB980 there was a small but
significant local minimum in the dissolved Cu concentration in the vicinity of the chlorophyll
and primary nitrite maxima that was most pronounced during CB980 (Fig. 4A–C). However,
during CB985 there was a subsurface maximum in Cu (Fig. 4D) that cannot be explained by
anomalies in temperature or salinity (Fig. 1D).
Cu
2+
and organic ligand analyses—Most Cu
2+
concentrations ranged between 10
-14
mol L
-1

and 5 × 10
-13
mol L
-1
; the lowest and highest values were 6.14 × 10
-15
mol L
-1
and 1.36 × 10
-12

mol L
-1
(Fig. 4A–D). The lowest Cu
2+
concentrations were always found within the upper 15 m
of the water column and were located near the depths of the chlorophyll and/or primary nitrite
maxima. Minima in Cu
2+
were especially pronounced relative to the concentrations immediately
above and below them during the CB960 and CB974 cruises (Fig 4A–B); the lowest Cu
2+
concentration, 6.14 × 10
-15
mol L
-1
, was also observed during the CB960 cruise (Fig. 4A). The
highest Cu
2+
concentrations were measured at depths in which either the ligand and dissolved Cu

64
concentrations were very close or the dissolved Cu concentrations exceeded the ligand
concentrations. This demonstrates that even small variations in dissolved Cu concentrations can
cause large changes in Cu
2+
concentrations (Fig. 5).
Generally, Cu
2+
concentrations increased with depth. Salinity also increased gradually with
depth, and the strongest correlation was between salinity and Cu
2+
concentrations in the deeper
waters (Fig. 6). The exception was the bottommost depth (105 m) during the CB974 cruise when
the Cu
2+
concentration decreased relative to the concentration of the next deepest sample (90 m)
(Fig. 4B). Furthermore, the Cu
2+
concentration typically increased immediately below the base
of the mixed layer (20 m). The sharpest increase was observed during the CB974 cruise, which
also featured the highest chlorophyll concentrations, as indicated by fluorescence (Fig. 4B). At
the same time, increases in the Cu
2+
concentration co-occurred with increases in the dissolved Cu
concentration. This trend was obscured during the CB985 cruise by the unexpected increase in
the dissolved Cu concentration near the base of the mixed layer (Fig. 4D). CB985 also featured
a local Cu
2+
minimum around 50 m (Fig. 4D). CB980 was characterized by relatively weak
chlorophyll and PNM (Fig 2C; 4C), which is partly reflected in the near uniform Cu
2+

concentrations. Not only are the deep concentrations very similar, but even the surface values—
which generally display greater variability—are fairly consistent. The other cruises, which
featured more prominent PNM and/or chlorophyll maxima, showed larger gradients in Cu
2+

concentrations.
Ligand concentrations generally ranged between 4 and 6 nmol L
-1
; the lowest and highest
values were 2.98 nmol L
-1
and 7.23 nmol L
-1
(Table 1). The highest concentrations were found
within the upper 20 m though some elevated concentrations were also found at depth.
Freshwater samples (Table 2) were analyzed to estimate the importance of riverine sources of

65
ligands to the excess complexation in the upper water column. We titrated the North Fork
Skokomish River samples (FW
1
and FW
2
) directly but recognized that these results might
overestimate the concentration of Cu-binding ligands from the river in Hood Canal because of
calcium and magnesium ion interactions or flocculation during mixing. Therefore we conducted
a mixing experiment in which we added one part low trace metal filtered surface seawater
collected from the North Pacific Ocean ([Cu] = 0.59 ± 0.009 nmol L
-1
) to two parts freshwater
(from FW
1
) to obtain a sample with a salinity of 11 (assuming salinities of 35 and 0 for the
saltwater and freshwater, respectively) and a final dissolved Cu concentration of 2.03 nmol L
-1
.
Results from this analysis revealed a relatively low ([L] = 2.33 nmol L
-1
) but strong (log K =
14.2) pool of ligands and a low Cu
2+
concentration (4.19 × 10
-14
mol L
-1
). We then calculated the
ligand concentration in the original river water using the following relationship:

L
tot
= (L
GEOTRACES
+ 2L
River
)/3

Where L
tot
refers to the ligand concentration of the mixed sample (L
tot
= 2.33 nmol L
-1
),
L
GEOTRACES
refers to the ligand concentration of the Pacific GEOTRACES sample (L
GEOTRACES
=
1.2 nmol L
-1
; Buck et al. 2012), and L
River
refers to the ligand concentration of the river sample.
Solving for L
River
yields a value of 1.4 nmol L
-1
. This is smaller than the ligand concentrations of
the FW
1
and FW
2
samples, but that is not surprising in view of the aforementioned
considerations. These calculations therefore suggest that the river contribution to the ligand pool
in the surface waters is small. Even at the lowest salinity (25) the river contributes at most one
third of the water so the ligand contribution is about 0.5 nmol L
-1
, or ~10% of the ligand pool
observed.

66
The organic ligand datasets from the CB960 and CB974 samples analyzed with the 10 µmol
L
-1
SA detection window (Table 3) indicate that the concentrations of these stronger ligands were
more similar between cruises than were the concentrations of ligands in the samples analyzed
with the 5 µmol L
-1
SA detection window (Table 1). The concentrations of the stronger ligands
were generally lower than those of the weaker ligands and exhibited less range. The lowest and
highest values were 4.20 nmol L
-1
and 6.14 nmol L
-1
; the highest concentration was measured
near the depth of the chlorophyll maximum during the CB974 cruise.
Conditional stability constants for all samples generally exceeded 10
13
and were highest
around the depths of the chlorophyll and primary nitrite maxima; log K values ranged between
12.9 and 14.7 (Table 1). Log K values for the samples processed with 10 µmol L
-1
SA (Table 3)
were uniformly higher than those for the same samples run with 5 µmol L
-1
SA and ranged
between 13.8 and 15.6. The highest log K values for the samples processed with the lower
detection window were most often associated with the chlorophyll maximum (CB960, CB974,
and CB985). This was also true for the CB960 samples analyzed with the higher detection
window, but the highest log K value for the CB974 cruise was found within the PNM instead of
the chlorophyll maximum. Also, the log K values for the samples collected from the chlorophyll
maxima and PNM at the higher detection window were much closer to each other than they were
for the corresponding samples at the lower detection window.

Discussion
The distributions of dissolved Cu, Cu
2+
and L—The most salient feature of dissolved Cu
chemistry in Hood Canal was its relative uniformity in time and space. Dissolved Cu
concentrations were relatively constant and showed little range over the time interval of the

67
project (Fig. 4A–D). The distribution and concentrations were similar to those of Paulson et al.
(1993) who, like us, saw little to no depletion in the dissolved Cu concentration in the surface
waters. This indicates that there has been little change in dissolved Cu concentrations in Hood
Canal over the last 25 years, in spite of anthropogenic inputs.
Cu
2+
concentrations in the mixed layer displayed some variability but were more uniform at
depth. These concentrations are similar to values measured in coastal waters off Cape Cod,
Massachusetts, where dissolved Cu concentrations were also similar (Moffett et al. 1997). They
are, however, substantially lower than Cu
2+
concentrations measured in polluted harbors (Donat
et al. 1994; Moffett et al. 1997), suggesting that anthropogenic pollution of Cu may not be a
major issue in Hood Canal. In addition, they are similar to estimates of Cu
2+
concentrations
reported in the oligotrophic North Pacific Ocean by other investigators (Coale and Bruland 1988;
Moffett and Dupont 2007) despite the different systems.
Ligand concentrations (Table 1) also spanned a small range and were similar to the dissolved
Cu concentrations, but they kept the Cu
2+
concentrations at levels comparable to oceanic regions
(Moffett and Dupont 2007; Jacquot et al. 2013). The strongest ligands were always observed
within the upper 20 m, where they were also most abundant, and were likely produced
biologically. Elevated concentrations of strong ligands were also found near the bottom on
occasion, perhaps indicating a benthic source or lateral inputs from near shore, but ligands were
generally weaker in the deeper, more saline waters. These slightly weaker, less plentiful ligands
could be made up of old, refractory dissolved organic matter or sinking biological debris from
the surface waters. Like dissolved Cu, these older, weaker ligands could have a longer residence
time in the lower water column.

68
It’s worth noting, however, that some of the highest ligand concentrations (Table 1) were
found during the CB960 cruise despite the low biological activity (Fig. 4A). The results from
Amin et al.’s (2013) study show that SCM1 does not produce Cu-binding organic ligands, even
under Cu limitation, so these ligands may derive from shallow sediment inputs transported along
isopycnal lines. Moreover, ligand concentrations (Table 1) during the CB974 cruise were not
exceptionally high despite coinciding with the strongest chlorophyll maximum of the 4 cruises
(Fig. 4B). Therefore, while our data suggest an upper water column—presumably biological—
source of ligands, the distribution cannot be directly linked to chlorophyll.
The elevated dissolved Cu concentrations measured in all of the surface samples were not
derived from a dissolved river source alone since the dissolved Cu concentration of the major
river source is low (Table 2). More likely, dissolved Cu derived from particulate sources in
rivers and runoff becomes remobilized from near-shore sediments and contributes to the surface
maximum by isopycnal transport. Likewise, while the conditional stability constant of the
ligands in the mixed sample was comparable to that of ambient ligands measured in the upper
water column of Hood Canal (Table 2), mass balance calculations suggest that dissolved riverine
ligands cannot account for the overall elevation of ligands near the surface. Ligands could come
from shallow benthic boundary sources or be produced in situ by phytoplankton and bacteria that
are more abundant in the upper water column than at depth.
It seems likely that complexation buffers dissolved Cu to low concentrations throughout most
of the water column, substantially reducing the amount of bioavailable Cu
2+
and leading to the
long residence time of dissolved Cu in the study site and to potential Cu limitation scenarios. In
effect, organic complexation creates a virtual chemical chemostat by which a large inventory of
dissolved Cu in the water column is buffered against removal by scavenging processes and

69
biological uptake. Yet while the Cu
2+
concentration can be drawn down through higher
biological uptake by phytoplankton and AOA to quite low levels, it never seems to drop below
~10
-15
– 5 × 10
-15
mol L
-1
. This buffering phenomenon has been observed by others in different
systems (Bruland and Lohan 2003; Jacquot et al. 2013) but the reasons why remain unclear.
Strong Cu
2+
drawdown near the chlorophyll and primary nitrite maxima—The lowest Cu
2+

concentrations were consistently measured within the upper 15 m of the water column and
coincided with the depths of the chlorophyll and/or PNM. This probably reflects both the
production of strong ligands in this biologically active zone and the complete utilization of labile
or weakly complexed dissolved Cu by microorganisms. The cruises for which the depths of the
PNM and Cu
2+
minima were closest (CB974 and CB985) had the most pronounced PNM (Fig.
2B, D; 4B, D). This feature is often attributed to high rates of ammonia oxidation, in excess of
nitrite oxidation (Buchwald et al. 2012), so a high Cu demand seems to be associated with its
formation.
Other than CB980, during which NH
4
+
levels peaked at 0.06 µmol L
-1
(Fig. 2C), NH
4
+

concentrations were sufficiently high in the mixed layer to theoretically support maximal
ammonia oxidation rates based on the K
m
(98 ±14 nmol L
-1
) determined by Horak et al. (2013).
However, there was not a consistent correlation between high NH
4
+
concentrations and high
ammonia oxidation rates, suggesting that factors other than substrate availability may have been
limiting ammonia oxidation in the upper water column. For instance, whereas CB985 featured
elevated ammonia oxidation rates throughout the water column CB974 rates were much lower
(Fig. 4B, D). Yet Cu
2+
concentrations were extremely low (~10
-14
mol L
-1
) during both cruises
(Fig. 4B, D). During CB974, AOA, despite having sufficient NH
4
+
in the mixed layer (Fig. 2B),
seemed to be constrained by the low Cu
2+
concentrations drawn down by phytoplankton (Fig.

70
4B). The high NH
4
+
and NO
2
-
concentrations observed during this cruise could therefore be
indicative of rapid remineralization of sinking organic matter and excretion by phytoplankton
due to incomplete assimilatory nitrate reduction, respectively, rather than ammonia oxidation by
AOA (Lomas and Lipschultz 2006).
Our data suggest that the Cu
2+
minima arise primarily from both a drawdown of dissolved Cu
and an increase in the concentration and/or conditional stability constants of the ligands. Since
the ligand and dissolved Cu concentrations are so close, only a slight drawdown of Cu greatly
decreases the Cu
2+
concentration and hence the bioavailability of Cu, and vice-versa (Moffett et
al. 1997). Therefore, even small changes in dissolved Cu concentrations are significant (Fig. 6).
The median Cu
2+
concentration of all our data (1.25 × 10
-13
mol L
-1
) is remarkably similar to the
Cu limitation threshold concentration for SCM1 (~2 × 10
-13
mol L
-1
) proposed by Amin et al.
(2013). This raises the important question of whether Cu
2+
concentrations in this system are
primarily controlled by the high biological demand of AOA or if other processes control it and
consequently limit AOA activity—or some combination of these alternatives. Our findings
suggest that both scenarios may be applicable in this system.
While Jacquot et al. (2013) theorized that high Cu demand by Fe-limited diatoms helped
control Cu
2+
concentrations in the ETSP, Cu
2+
bioavailability in Hood Canal is unlikely to be
strongly influenced by diatoms. Coastal diatoms such as Chaetoceros decipiens and
Thalassiosira weissflogii have a significantly lower Cu limitation threshold (~10
-15
mol L
-1
;
equivalent to 77% and 76% of µ
max
, respectively) (Annett et al. 2008) than SCM1 and,
presumably, AOA in Hood Canal. Furthermore, unlike the ETSP, where dissolved Fe
concentrations were ≤ 0.1 nmol L
-1
(Y. Kondo and J. W. Moffett unpubl.) and thus potentially
limiting for nitrate reduction, dissolved Fe concentrations in Hood Canal are likely replete (≤ 10

71
nmol L
-1
; Paulson et al. 1993). This leaves AOA as the principal known source of Cu demand in
this system.
Does low Cu bioavailability limit ammonia oxidation by AOA?—A comparison of our data
with the Cu limitation threshold for SCM1 experimentally derived by Amin et al. (2013)
suggests that Cu
2+
concentrations in the Hood Canal are poised around the point at which Cu
limitation was important in culture, with values in the upper water column predicted to be
limiting and values at depth predicted to be Cu-replete. There are exceptions to this
generalization, particularly above and below the chlorophyll maxima, where the gradients in
Cu
2+
concentrations can be significant. But, overall, the results suggest that Cu limitation may
be a constraint on ammonia oxidation by AOA in the upper water column, where ammonia
oxidation rates are lowest.
The lowest Cu
2+
concentrations measured during all the cruises were often below the Cu
limitation threshold for SCM1 (Amin et al. 2013), and thus Cu limitation has the potential to
significantly reduce ammonia oxidation rates in this AOA-dominated environment (Horak et al.
2013). Yet more work is required to elucidate the relative influence of Cu limitation on
ammonia oxidation rates. Archaeal ammonia oxidation by laboratory cultures is also known to
be strongly light-limited (Merbt et al. 2011) so the relative importance of Cu limitation and light
inhibition need to be established by i situ experimentation at this site. In this system, light
limitation, not Cu limitation, is likely more important at the shallowest samples where the Cu
2+

concentration is generally quite high. But Cu limitation may be more important at the base of the
mixed layer, where light intensities are low. Cu limitation has also been demonstrated for
diatoms in culture (Maldonado et al. 2006). Diatoms are abundant and represent a significant
competitor for ammonia in the mixed layer. Other unrecognized taxa may also compete with

72
AOA and diatoms for Cu in these waters.

Bibliography
Amin, S. A., J. W. Moffett, W. Martens-Habbena, J. E. Jacquot, Y. Han, A. Devol, A. E. Ingalls,
D. A. Stahl, and E. V. Armbrust. 2013. Copper requirements of the ammonia-oxidizing
archaeon Nitrosopumilus maritimus SCM1 and implications for nitrification in the marine
environment. Limnol. Oceanogr. In press.
Annett, A. L., S. Lapi, T. J. Ruth, and M. T. Maldonado. 2008. The effects of Cu and Fe
availability on the growth and Cu : C ratios of marine diatoms. Limnol. Oceanogr. 53(6):
2451–2461, doi:10.4319/lo.2008.53.6.2451
Beman, J. M., R. Sachdeva, and J. A. Fuhrman. 2010. Population ecology of nitrifying Archaea
and Bacteria in the Southern California Bight. Environ. Microbiol. 12(5): 1282–1292,
doi:10.1111/j.1462-2920.2010.02172.x
Blainey, P. C., A. C. Mosier, A. Potanina, C. A. Francis, and S. R. Quake. 2011. Genome of a
low-salinity ammonia-oxidizing archaeon determined by single-cell and metagenomic
analysis. PLoS ONE 6(2): e16626, doi:10.1371/journal.pone.0016626
Bruland, K. W., E. L. Rue, J. R. Donat, S. A. Skrabal, and J. W. Moffett. 2000. Intercomparison
of voltammetric techniques to determine the chemical speciation of dissolved copper in a
coastal seawater sample. Anal. Chim. Acta 405: 99–113, doi:10.1016/S0003-
2670(99)00675-3
Bruland, K. W., and M. C. Lohan. 2003. Controls of trace metals in seawater, p. 23–47. In H.
Elderfield [ed.], The oceans and geochemistry. Academic.
Buchwald, C., A. E. Santoro, M. R. McIlvin, and K. L. Casciotti. 2012. Oxygen isotopic

73
composition of nitrate and nitrite produced by nitrifying cocultures and natural marine
assemblages. Limnol. Oceanogr. 57(5): 1361–1375, doi:10.4319/lo.2012.57.5.1361
Buck, K. N., and K. W. Bruland. 2005. Copper speciation in San Francisco Bay: A novel
approach using multiple analytical windows. Mar. Chem. 96: 185–198,
doi:10.1016/j.marchem.2005.01.001
Buck, K. N., J. Moffett, K. A. Barbeau, R. M. Bundy, Y. Kondo, and J. Wu. 2012. The organic
complexation of iron and copper: an intercomparison of competitive ligand exchange-
adsorptive cathodic stripping voltammetry (CLE-ACSV) techniques. Limnol. Oceanogr.:
Methods 10: 496–515, doi:10.4319/lom.2012.10.496
Bundy, R. M., K. A. Barbeau, and K. N. Buck. 2012. Sources of strong copper-binding ligands
in Antarctic Peninsula surface waters. Deep-Sea Res. II 90: 134–146,
doi:10.1016/j.dsr2.2012.07.023
Coale, K. H., and K. W. Bruland. 1988. Copper complexation in the Northeast Pacific. Limnol.
Oceanogr. 33(5): 1084–1101, doi:10.4319/lo.1988.33.5.1084
Devol, A., W. Ruef, J. Newton, and C. Smith. 2011. High-frequency marine observation in Hood
Canal. In J. Newton [ed.], Hood Canal dissolved oxygen program integrated assessment &
modeling study report. Academic.
Donat, J. R., K. A. Lao, and K. W. Bruland. 1994. Speciation of dissolved copper and nickel in
South San Francisco Bay: a multi-method approach. Anal. Chim. Acta 284: 547–571,
doi:10.1016/0003-2670(94)85061-5
Ensign, S. A., M. R. Hyman, and D. J. Arp. 1993. In vitro activation of ammonia
monooxygenase from Nitrosomonas europaea by copper. J. Bacteriol. 175(7): 1971–1980.
Francis, C. A., K. J. Roberts, J. M. Beman, A. E. Santoro, and B. B. Oakley. 2005. Ubiquity and

74
diversity of ammonia-oxidizing archaea in water columns and sediments of the ocean.
Proc. Nat. Acad. Sci. U.S.A. 102(41): 14683–14688, doi:10.1073/pnas.0506625102
Grasshoff, K., K. Kremling, and M. Erhard. 1999. Methods of seawater analysis, 3
rd
ed. Wiley-
VCH.
Hallam, S. J., K. T. Konstantinidis, N. Putnam, C. Schleper, Y. Watanabe, J. Sugahara, C.
Preston, J. R. de La Torre, P. M. Richardson, and E. F. Delong. 2006. Genomic analysis of
the uncultivated marine crenarchaeote Cenarchaeum symbiosum. Proc. Natl. Acad. Sci.
U.S.A. 103(48): 18296–18301, doi:10.1073/pnas.0608549103
Holmes, R. M., A. Aminot, R. Kérouel, B. A. Hooker, and B. J. Peterson. 1999. A simple and
precise method for measuring ammonium in marine and freshwater ecosystems. Can. J.
Fish. Aquat. Sci. 56: 1801–1808, doi:10.1139/f99-128
Horak, R. E. A., W. Qin, A. J. Schauer, E. V. Armbrust, A. E. Ingalls, J. W. Moffett, D. A. Stahl,
and A. H. Devol. 2013. Ammonia oxidation kinetics and temperature sensitivity of a
natural marine community dominated by Archaea. ISME J., doi:10.1038/ismej.2013.75
Jacquot, J. E., Y. Kondo, A. N. Knapp, and J. W. Moffett. 2013. The speciation of copper across
active gradients in nitrogen-cycle processes in the eastern tropical South Pacific. Limnol
Oceanogr. 58(4): 1387–1394, doi:10.4319/lo.2013.58.4.1387
Kérouel, R., and A. Aminot. 1997. Fluorometric determination of ammonia in sea and estuarine
waters by direct segmented flow analysis. Mar. Chem. 57: 265–275, doi:10.1016/S0304-
4203(97)00040-6
Könneke, M., A. E. Bernhard, J. R. de La Torre, C. B. Walker, J. B. Waterbury, and D. A. Stahl.
2005. Isolation of an autotrophic ammonia-oxidizing marine archaeon. Nature 437: 543–
546, doi:10.1038/nature03911

75
Lee, J.-M., E. A. Boyle, Y. Echegoyen-Sanz, J. N. Fitzsimmons, R. Zhang, and R. A. Kayser.
2011. Analysis of trace metals (Cu, Cd, Pb, and Fe) in seawater using single batch
nitrilotriacetate resin extraction and isotope dilution inductively coupled plasma mass
spectrometry. Anal. Chim. Acta 686: 93–101, doi:10.1016/j.aca.2010.11.052
Lomas, M. W., and F. Lipschultz. 2006. Forming the primary nitrite maximum: Nitrifiers or
phytoplankton? Limnol. Oceanogr. 51(5): 2453–2467, doi:10.4319/lo.2006.51.5.2453
Mackas, D. L., and P. J. Harrison. 1997. Nitrogenous nutrient sources and sinks in the Juan de
Fuca Strait/Strait of Georgia/Puget Sound estuarine system: assessing the potential for
eutrophication. Estuar. Coast. Shelf Sci. 44: 1–21, doi:10.1006/ecss.1996.0110
Maldonado, M. T., A. E. Allen, J. S. Chong, K. Lin, D. Leus, N. Karpenko, and S. L. Harris.
2006. Copper-dependent iron transport in coastal and oceanic diatoms. Limnol. Oceanogr.
51(4): 1729–1743, doi:10.4319/lo.2006.51.4.1729
Merbt, S. N., D. A. Stahl, E. O. Casamayor, E. Martí, G. W. Nicol, and J. I. Prosser. 2011.
Differential photoinhibition of bacterial and archaeal ammonia oxidation. FEMS Microbiol.
Lett. 327: 41–46, doi:10.1111/j.1574-6968.2011.02457.x
Moffett, J. W., L. E. Brand, P. L. Croot, and K. A. Barbeau. 1997. Cu speciation and
cyanobacterial distribution in harbors subject to anthropogenic Cu inputs. Limnol.
Oceanogr. 42(5): 789–799, doi:10.4319/lo.1997.42.5.789
Moffett, J. W., and C. L. Dupont. 2007. Cu complexation by organic ligands in the sub-arctic
NW Pacific and Bering Sea. Deep-Sea Res. 54: 586–595, doi:10.1016/j.dsr.2006.12.013
Moffett, J. W., C. B. Tuit, and B. B. Ward. 2012. Chelator-induced inhibition of copper
metalloenzymes in denitrifying bacteria. Limnol. Oceanogr. 57(1): 272–280,
doi:10.4319/lo.2012.57.1.272

76
Molina, V., L. Belmar, and O. Ulloa. 2010. High diversity of ammonia-oxidizing archaea in
permanent and seasonal oxygen-deficient waters of the eastern South Pacific. Environ.
Microbiol. 12: 2450–2465, doi:10.1111/j.1462-2920.2010.02218.x
Newton, J., C. Bassin, A. Devol, J. Richey, M. Kawase, and M. Warner. 2011. Overview and
results synthesis. In J. Newton [ed.], Hood Canal dissolved oxygen program integrated
assessment & modeling study report. Academic.
Paulson, A. J., H. C. Curl Jr., and R. A. Feely. 1993. The biogeochemistry of nutrients and trace
metals in Hood Canal, a Puget Sound fjord. Mar. Chem. 43: 157–173, doi:10.1016/0304-
4203(93)90222-A
Peers, G., S. A. Quesnel, and N. M. Price. 2005. Copper requirements for iron acquisition and
growth of coastal and oceanic diatoms. Limnol. Oceanogr. 50(4): 1149–1158, doi:
10.4319/lo.2005.50.4.1149
United Nations Educational, Scientific and Cultural Organization (UNESCO). 1994. Protocols
for the Joint Global Ocean Flux Study (JGOFS) core measurements. IOC Manuals and
Guides 29.
Vajrala, N., W. Martens-Habbena, L. A. Sayavendra-Soto, A. Schauer, P. J. Bottomley, D. A.
Stahl, and D. J. Arp. 2013. Hydroxylamine as an intermediate in ammonia oxidation by
globally abundant marine archaea. Proc. Nat. Acad. Sci. U.S.A. 110(3): 1006–1011,
doi:10.1073/pnas.1214272110
Walker, C. B., J. R. de La Torre, M. Klotz, H. Urakawa, N. Pinel, D. Arp, C. Brochier-Armanet,
P. Chain, P. Chan, A. Gollabgir, J. Hemp, M. Hügler, E. A. Karr, M. Könneke, M. Shin, T.
J. Lawton, T. Lowe, W. Martens-Habbena, L. A. Sayavedra-Soto, D. Lang, S. M. Sievert,
A. C. Rosenzweig, G. Manning, and D. A. Stahl. 2010. Nitrosopumilus maritimus genome

77
reveals unique mechanisms for nitrification and autotrophy in globally distributed marine
crenarchaea. Proc. Nat. Acad. Sci. U.S.A. 107(19): 8818–8823,
doi:10.1073/pnas.0913533107

78
Table 1. Concentrations and conditional stability constants of organic ligands for dissolved Cu
for all cruises using the 5 µmol L
-1
SA detection window. Note that the log K ranges are
asymmetric. The depth marked with the asterisk is the depth of the chlorophyll maximum. The
depth marked with the double asterisk is the depth of the PNM or the depth closest to it.
Cruise Depth (m) L (nmol L
-1
) Log K
CB960
3 6.53 ± 0.077 14.2 (14.2 – 14.3)

8 4.97 ± 0.33 13.6 (13.4 – 13.7)

9* 6.50 ± 0.14 14.5 (14.4 – 14.5)

13 5.77 ± 0.13 13.9 (13.8 – 14.0)

15** 6.63 ± 0.36 13.5 (13.4 – 13.6)
18 7.23 ± 0.22 14.0 (13.9 – 14.1)
25 6.06 ± 0.16 13.4 (13.4 – 13.5)
40 4.90 ± 0.18 13.5 (13.4 – 13.6)
60 5.07 ± 0.17 13.8 (13.7 – 13.9)
90 4.96 ± 0.20 13.2 (13.1 – 13.3)
CB974 2 5.06 ± 0.21 13.4 (13.4 – 13.5)

5 5.02 ± 0.19 13.6 (13.5 – 13.7)

10* 5.58 ± 0.12 14.7 (14.7 – 14.8)

12** 5.82 ± 0.19 14.5 (14.5 – 14.6)

16 5.20 ± 0.20 13.9 (13.8 – 14.0)

20 5.74 ± 0.22 14.1 (14.0 – 14.1)

30 4.28 ± 0.20 13.3 (13.2 – 13.4)

40 4.39 ± 0.19 13.7 (13.7 – 13.8)

50 5.22 ± 0.27 13.4 (13.3 – 13.5)

70 5.94 ± 0.35 13.1 (13.0 – 13.1)

90 3.87 ± 0.12 14.1 (14.0 – 14.2)

105 6.29 ± 0.25 13.5 (13.5 – 13.6)
CB980 2 4.63 ± 0.14 14.4 (14.3 – 14.5)

5** 5.16 ± 0.12 14.1 (14.0 – 14.1)

7 3.23 ± 0.23 14.2 (14.0 – 14.4)

11* 5.63 ± 0.10 14.0 (14.0 – 14.1)

13 5.23 ± 0.18 13.7 (13.6 – 13.7)

16 4.75 ± 0.086 14.0 (14.0 – 14.1)

20 6.21 ± 0.17 13.7 (13.6 – 13.7)

30 2.98 ± 0.36 13.3 (12.9 – 13.5)

40 3.83 ± 0.32 12.9 (12.8 – 13.0)

50 5.57 ± 0.19 13.6 (13.5 – 13.7)

70 6.41 ± 0.21 13.3 (13.2 – 13.3)

90 6.33 ± 0.32 13.3 (13.2 – 13.3)

115 4.26 ± 0.25 13.9 (13.8 – 14.0)
CB985 5 5.83 ± 0.12 14.5 (14.5 – 14.6)

7 6.18 ± 0.22 13.6 (13.5 – 13.6)

12* 6.89 ± 0.10 14.7 (14.7 – 14.8)

15** 6.14 ± 0.18 13.8 (13.8 – 13.9)

20 5.73 ± 0.27 13.0 (12.9 – 13.0)

79

40 5.33 ± 0.13 14.3 (14.2 – 14.3)

50 6.40 ± 0.36 13.8 (13.8 – 13.9)

70 4.28 ± 0.16 13.7 (13.5 – 13.7)

90 4.64 ± 0.12 13.5 (13.5 – 13.6)

115 4.83 ± 0.096 14.0 (13.9 – 14.0)

80
Table 2. Dissolved Cu concentrations, log Cu
2+
, organic ligand concentrations, and conditional
stability constants for freshwater (FW
1
and FW
2
) and mixed water samples from CB974. Note
that the freshwater samples were analyzed with the 10 µmol L
-1
SA detection window while the
mixed sample was analyzed with the 5 µmol L
-1
SA detection window. Note that the log Cu
2+

and log K ranges are asymmetric.

Samples Cu (nmol L
-1
)
Log Cu
2+
(mol L
-1
)
L (nmol L
-1
) Log K
FW
1
2.74 ± 0.12 -12.9 (-13.0 to -12.8) 1.69 ± 0.17 16.0 (15.7 – 16.2)
FW
2
2.03 ± 0.049 -14.8 (-14.9 to -14.8) 2.05 ± 0.081 16.4 (16.3 – 16.5)
Mixed 2.03 -13.4 (-13.5 to -13.1) 2.33 ± 0.17 14.2 (14.1 – 14.3)

81
Table 3. Concentration and conditional stability constants of organic ligands for dissolved Cu
for CB960 and CB974 cruises using the 10 µmol L
-1
SA detection window. Note that the log K
ranges are asymmetric. The depth marked with the asterisk is the depth of the chlorophyll
maximum. The depth marked with the double asterisk is the depth of the PNM or the depth
closest to it.

Cruise Depth (m) L (nmol L
-1
) Log K
CB960 8 5.65 ± 0.23 14.3 (14.2 – 14.4)

9* 4.81 ± 0.097 14.9 (14.8 – 14.9)

13 5.02 ± 0.11 14.1 (14.0 – 14.1)

15** 5.36 ± 0.10 14.5 (14.5 – 14.6)
18 6.10 ± 0.11 14.6 (14.5 – 14.6)
25 4.49 ± 0.098 13.9 (13.8 – 14.0)
40 4.66 ± 0.10 14.1 (14.0 – 14.1)
90 4.68 ± 0.13 13.8 (13.7 – 13.8)
CB974 2 5.35 ± 0.15 14.3 (14.2 – 14.3)

5 4.68 ± 0.15 14.4 (14.3 – 14.5)

10* 5.11 ± 0.16 15.3 (15.2 – 15.4)

12** 6.14 ± 0.14 15.6 (15.6 – 15.7)

16 4.20 ± 0.070 14.6 (14.6 – 14.7)

20 5.75 ± 0.27 14.3 (14.2 – 14.3)

30 4.95 ± 0.20 14.1 (14.0 – 14.1)

105 5.03 ± 0.17 13.9 (13.9 – 14.0)

82
Figure Legends
Fig. 1. Representative depth profiles of temperature (solid line), salinity (dashed line), and
dissolved O
2
concentrations (dotted line) for (A) CB960, (B) CB974, (C) CB980, and (D)
CB985. Data for all but CB974 were retrieved from the ORCA buoy; dissolved O
2

concentrations for that cruise were not Winkler-corrected.

Fig. 2. Representative depth profiles of NO
3
-
(closed circles), NO
2
-
(open triangles), and NH
4
+

(open squares) concentrations for (A) CB960, (B) CB974, (C) CB980, and (D) CB985.

Fig. 3. Representative depth profiles of PO
4
3-
(closed circles), and SiO
4
4-
(open triangles)
concentrations for (A) CB960, (B) CB974, (C) CB980, and (D) CB985.

Fig. 4. Depth profiles of log Cu
2+
(closed circles), dissolved Cu concentrations (open triangles),
fluorescence (solid line), and NH
3
oxidation rates (open squares) for (A) CB960, (B) CB974, (C)
CB980, and (D) CB985. Error bars for log Cu
2+
and dissolved Cu represent error propagation
from the calculation of L and K and standard deviation values (n=3), respectively.

Fig. 5. Representative titration dataset showing how even small changes in the dissolved Cu
concentration can cause large changes in log Cu
2+
.

Fig. 6. Log Cu
2+
plotted against salinity for CB960, CB974, CB980, and CB985.

83

Fig. 1.

84

Fig. 2.

85

Fig. 3.

86

Fig. 4.

87

Fig. 5.

88

Fig. 6.

89
Chapter 4
Copper distribution and speciation across the U.S. North Atlantic GEOTRACE section

Acknowledgments
We are grateful to the chief scientists (Drs. Bill Jenkins, Ed Boyle and Greg Cutter), Captains
Adam Seamans and Kent Sheasley, the crew of the R/V Knorr, and the members of the
GEOTRACES scientific party for their considerable help and guidance. We are also very
grateful to Yoshiko Kondo for help in developing the ICP-MS method and for running some
samples. We want to thank Jessica Fitzsimmons, Randelle Bundy, Rachel Shelley, Ana Aguilar-
Islas and Peter Morton in particular for all their hard work collecting the samples for the entire
science party. This work was funded by an NSF Chemical Oceanography grant to James W.
Moffett.

90
Abstract
Copper (Cu) distribution and speciation were measured along a meridional transect in the
North Atlantic Ocean from Woods Hole, Massachusetts, to Mauritania as part of the U.S.
GEOTRACES program. Dissolved Cu profiles showed surface depletion, sub-surface and deep
scavenging, and a linear increase with depth, but many also exhibited unique features and clear
geographic trends. Concentrations ranged from 0.43 nmol L
-1
to 3.07 nmol L
-1
. The highest
concentrations were measured in the deep waters above the seafloor just west of the Cape Verde
Islands, but concentrations in the upper water column were generally higher at the western end of
the transect than at the eastern end. The westernmost stations on or near the U.S. east coast
continental shelf exhibited surface maxima that decreased in magnitude moving east to Bermuda,
possibly reflecting declining aerosol inputs from North America. Cu
2+
was tightly controlled by
organic complexation and scavenging with concentrations ranging from 1.54 fmol L
-1
to 1.07
pmol L
-1
. Some of the lowest Cu
2+
and dissolved Cu concentrations were measured between
Mauritania and the Cape Verde Islands, an area with high upwelling-fueled biological
productivity. Cu
2+
and dissolved Cu concentrations were low within several benthic nepheloid
layers and a hydrothermal plume above the Trans-Atlantic Geotraverse (TAG) vent field,
suggesting intense scavenging by non-biogenic particles. Our findings show that dissolved Cu
profiles display more variability than previously thought but that Cu
2+
concentrations are almost
uniformly low from surface to seafloor and confined to a narrow biological “steady-state” or
“buffered” range across the transect.

91
Introduction
The GEOTRACES program (www.geotraces.org) provides an unprecedented opportunity to
study the distributions and geochemistry of over 30 trace elements and isotopes across multiple
ocean basins and timescales. Copper (Cu) was selected as a core parameter because it is a vital
micronutrient that facilitates many biogeochemically significant processes including nitrous
oxide reduction (Stiefel 2007), photosynthesis (Peers and Price 2006), aerobic ammonia
oxidation (Ensign et al. 1993; Walker et al. 2010), and iron (Fe) acquisition by diatoms
(Maldonado et al. 2006). It, alongside other essential trace metal micronutrients like iron, zinc
(Zn) and cobalt (Co), therefore has the capacity to shape and profoundly influence phytoplankton
communities and, in turn, global patterns of primary productivity (Sunda 2012). However, it is
also toxic in its bioavailable cupric ion form, Cu
2+
, and is known to inhibit cyanobacterial
reproduction at concentrations as low as 1 pmol L
-1
(Brand et al. 1986) by catalyzing the
formation of reactive oxygen species (Aruoma et al. 1991) or by competitively inhibiting the
uptake and metabolism of manganese (Mn) (Sunda et al. 1981). In natural waters, over 99.8% of
all dissolved Cu is tightly complexed by strong organic ligands, reducing Cu
2+
concentrations by
over a thousand-fold (Bruland and Franks 1983; Donat et al. 1994). The resulting Cu
2+

concentrations, averaging ~10
-14
to 10
-13
mol L
-1
, are therefore neither too low to sustain
biological uptake by cyanobacteria and phytoplankton nor too high to inhibit growth and
reproduction rates (Bruland and Lohan 2003).
Copper depth profiles display a “hybrid” distribution and thus share characteristics of both
nutrient-like and scavenged elements; these include pronounced biological uptake in the euphotic
zone and intense scavenging by sinking particulate matter throughout the water column but
particularly at depth (Bruland and Lohan 2003). In the Atlantic as well as the Pacific Ocean,

92
dissolved Cu concentrations are typically depleted in the surface waters due to biological uptake.
Below the surface, concentrations increase slowly but continuously with depth, never reaching a
mid-water maximum as do nutrient-type elements (Boyle et al. 1977, 1981; Bruland 1980). This
unique linear profile is maintained by the continuous exchange of the Cu
2+
fraction between the
dissolved and particulate phases with the net result being surface and bottom concentrations that
rarely vary by over an order of magnitude—especially in the Atlantic, where they can vary by as
little as four- to five-fold (Bruland 1980; Bruland and Franks 1983). While several earlier
studies had shown that Cu
2+
concentrations increased rapidly below the euphotic zone as the
concentration of strong organic ligands fell off (Coale and Bruland 1988, 1990), a number of
recent ones have demonstrated that Cu
2+
is in fact strongly buffered by natural organic ligands to
concentrations as low as ~10
-14
mol L
-1
even at depth (Moffett and Dupont 2007; Buck et al.
2012). These studies indicate that the two primary controls on the distribution of dissolved Cu,
and by extension Cu
2+
, in the ocean are subsurface scavenging and organic complexation.
What remains unclear, however, is the extent to which each factor influences the distribution
of dissolved Cu at various depths and across areas with very distinct biogeochemical properties.
For example, whereas one might anticipate dissolved Cu to be principally controlled by organic
complexation in a productive, upwelling-fueled regime such as the Peruvian coastline, one might
expect dissolved Cu to be more equally controlled by both complexation and sub-surface
scavenging on the North American continental margin. The major limitation to past studies is
that they were too limited in geographic scope to fully address this question. The U.S.
GEOTRACES North Atlantic cruises marked the first opportunity to investigate the distribution
of dissolved Cu and Cu
2+
from the surface to the seafloor across a basin-wide zonal transect.

93
Herein we present the results of this study and combine our findings with those of other
investigators to achieve a better understanding of the oceanic cycle of dissolved Cu.

Methods
Sample collection—Over 400 samples were collected from 17 stations aboard the R/V Knorr
during the KN199-4 and KN204-1 cruises (Fig. 1) as part of the US GEOTRACES program.
The KN199-4 cruise started from Lisbon, Portugal, on 15 October 2010 and ended at the Cape
Verde Islands on 4 November 2010 while the KN204-1 cruise started from Woods Hole,
Massachusetts, on 6 November 2011 and ended at the Cape Verde Islands on 11 December 2011.
The US GEOTRACES carousel (Model 32G, Sea-Bird Electronics) (Cutter and Bruland 2012)
was equipped with 24 Teflon-coated 12 L GO-FLO bottles (Model 108012T, General Oceanics)
and hydrographic sensors for pressure, conductivity, salinity, temperature, oxygen, beam
transmittance, and fluorescence (Fig. 2–4) housed in titanium pressure cases to minimize the risk
of sample contamination. The carousel system was lowered over the side of the ship with
conducting Vectran cable and flushed with seawater on the way down; the samples were
collected during the upcast. Upon recovery, clear polyethylene shower caps were placed on the
tops and bottoms of the bottles and they were brought into a ISO Class-5 clean laboratory van
(Silhouette Steel) for sub-sampling under HEPA filtered air. The GO-FLOs were pressurized
with filtered (0.2 µm) compressed air and the samples were forced through clean Teflon
perfluoroalkoxy (PFA) tubing and acid-cleaned 0.2 µm Acropak Supor capsule filters (Pall
Corporation) into 1 L fluorinated low-density polyethylene (FLPE, Nalgene) bottles. All
sampling bottles were cleaned in a sequential 4-step process: 1) soaked in a 5% Citranox
(Alconox) acid detergent bath for at least a day to remove organics, 2) soaked in a 10%

94
hydrochloric acid bath (HCl; Van Waters and Rogers (VWR) International) for at least a day, 3)
filled with 10% HCl and baked at 60°C upright and upside down to properly leach the threads
around the cap for at least 2 days, and 4) filled with 0.1% trace metal grade HCl (Optima, Fisher)
and baked at 60°C again for at least 2 days. In between each step, the insides and outsides of the
bottles were thoroughly rinsed at least 5 times with Milli-Q water (18.2 MΩ; Millipore). After
the final step, the 0.1% trace metal grade HCl was decanted and the bottles were rinsed 7 more
times to ensure the removal of all acid. They were then left to dry in a laminar flow bench for
several days before being individually double-bagged and labeled.
Dissolved copper analyses—One hundred twenty-five milliliters of seawater was transferred
from each 1 L FLPE bottle into a corresponding acid-washed and pre-conditioned 125 mL low-
density polyethylene (LDPE, Nalgene) bottle. The samples were then acidified to a pH of 1.7
with 250 µL concentrated trace metal grade HCl (Optima, Fisher) and stored for at least two
months prior to beginning the analyses back onshore. Dissolved Cu concentration were
determined using a
65
Cu spike and a single batch nitrilotriacetatic acid (NTA) resin (Qiagen)
extraction and isotope dilution inductively coupled plasma mass spectrometry (ID-ICP-MS)
method adapted from (Lee et al. 2011) on a Finnegan Element 2 (Thermo Scientific) and
described in (Jacquot et al. 2013).
Cu
2+
and organic ligand analyses—The remaining seawater from the 1 L FLPE bottles was
used to measure Cu
2+
and organic ligand (L) concentrations. Most samples were frozen at -20°C
and moved to USC following the cruise but a subset was refrigerated at 4°C for onboard
analyses; these were performed within a week of collection at room temperature (∼25°C) in a
laminar flow bench inside a clean laboratory van. The frozen samples were thawed in a
refrigerator for at least 2 days and then analyzed within a week inside a positive pressure clean

95
room enclosure. The free Cu
2+
concentrations, organic ligand concentrations and conditional
stability constants (K) were determined using a competitive ligand exchange adsorptive cathodic
stripping voltammetry (CLE-ACSV) method adapted from Buck and Bruland (2005) and
described in Jacquot et al. (2013) with salicylaldoxime (SA; ≥ 98%, Aldrich) as the added ligand.
The competition strength of SA, which determines the type of natural ligands that can be
detected by the method, is represented by the side reaction coefficient, α
Cu(SA)x
, defined as (Buck
and Bruland 2005):

α
Cu(SA)x
=
Cu SA
x
Cu
2+
= β
2
· [SA]
2
+ K
1
· [SA] (1)

where [Cu(SA)
x
] = [Cu(SA)
2
0
] + [Cu(SA)
+
] and β
2
and K
1
are the respective conditional stability
constants of the Cu(SA)
2
0
and Cu(SA)
+
complexes. The conditional stability constants of these
complexes were previously determined by Campos and van den Berg (1994) at different
salinities so that log β
2
= 15.78 – (0.53 · log (salinity)) and log K
1
= 10.12 – (0.37 · log
(salinity)). In practice, the higher the α
Cu(SA)x
used (and therefore the more competitive the
Cu(SA)
x
complexes become), the stronger the class of Cu-binding ligands that is detected,
resulting in smaller measured L concentrations and higher K values (Bruland et al. 2000; Buck et
al. 2012).
The measurements were made on a BioAnalytical Systems (BASi) Controlled Growth
Mercury Electrode (CGME) set to the Static Mercury Drop setting (drop size: 14 or 16) and
interfaced with a BASi Epsilon ε2 voltammetric analyzer. During the cruises it was mounted on
a vibration-free platform made of polyvinyl chloride (PVC) to minimize interferences (Aldrich et
al. 1999). The instrument settings were identical to those used in Jacquot et al. (2013).

96
Each sample was divided into 20 mL aliquots in 10 40 or 60 mL Teflon bottles (FEP,
Nalgene). The pH of each aliquot was then fixed to closely match the sample pH, which was
measured onboard following sample collection using an Orion* 3-Star Plus pH benchtop meter
(Thermo Scientific) calibrated against 3 buffers (Fig. 5), with 200 µL of either 1 mol L
-1
4-(2-
hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES; ≥ 99.5%, Sigma) buffer (pKa = 7.55) or
1 mol L
-1
4-(2-hydroxyethyl)-1-piperazinepropanesulfonic acid (EPPS; ≥ 99.5%, Sigma) buffer
(pKa = 8.0) and 0–50 µL of 1 mol L
-1
trace metal grade sodium hydroxide (NaOH; Fluka).
Sample pH influences the binding capacity of SA (Kogut 2002) and purging can increase pH (J.
E. Jacquot unpubl.) so the use of buffer was necessary. Side reaction coefficients were adjusted
for pH using the following relationship described by Kogut (2002):

[HSA
!
]
[SA
!
]
= (2)

where [HSA
-
] represents the concentration of the singly deprotonated SA species and [SA]
f
=
[SA] not complexed by Cu. The conditional proton dissociation constants, pKa
1
and pKa
2
, are
8.85 and 11.39, respectively, at an ionic strength of 1.0 mol L
-1
(Moffett and Dupont 2007).
A 10
-2
mol L
-1
Cu
2+
stock solution in Milli-Q was prepared from anhydrous cupric sulfate
(CuSO
4
; ≥ 99.99% trace metals basis, Aldrich) and 1 µmol L
-1
sub-stocks were prepared daily by
serial dilution. A 10
-2
mol L
-1
SA stock solution in Milli-Q was prepared weekly and 10
-3

mol L
-1
sub-stocks were prepared daily by serial dilution. The aliquots were spiked with varying
Cu concentrations (0 – 16 nmol L
-1
) and left to equilibrate for at least 2 hours. Subsequently, SA
(1, 2, or 2.5 µmol L
-1
) was added and the samples were allowed to equilibrate for at least 30 min,
1
10
pKa
1
- pH
+ 1 + 10
-pKa
2
+ pH

97
which was sufficient time to obtain steady-state values (Moffett 1995). These detection windows
have been used successfully by other researchers (Buck and Bruland 2005; Moffett and Dupont
2007) and were thus chosen to make our data comparable with theirs. Ten-milliliter sub-aliquots
were transferred to a Teflon sample cup and attached to the electrode. Prior to analysis, the sub-
aliquots were purged for 3 min with high-purity N
2
gas. We used the 2 µmol L
-1
and 2.5 µmol L
-
1
detection windows for the sample analyses because they have yielded the best results for us in
the past, but we also used the 1 µmol L
-1
detection window to further probe the weaker ligand
pool in several deep samples, where we expected to find more.
Nutrients—Micromolar concentrations of nitrate (NO
3
-
), nitrite (NO
2
-
), phosphate (PO
4
3-
), and
silicate (SiO
4
4-
) were measured onboard at all stations by the Ocean Data Facility (ODF) group
of the Scripps Institution of Oceanography (SIO) using a segmented continuous-flow
AutoAnalyzer 3 (SEAL Analytical) (Fig. 6–7). NO
3
-
and NO
2
-
concentrations were measured
using a modified version of the method described in Armstrong et al. (1967). Briefly, seawater
was passed through a cadmium (Cd) column in which NO
3
-
was reduced to NO
2
-
. NO
2
-
was
diazotized with sulfanilamide and combined with N-(1-naphthyl)-ethylenediamine to form a red
dye whose absorbance was subsequently measured with a 10 mm flow cell at 540 nm. The same
procedure was followed to measure NO
2
-
concentrations but without a Cd column. PO
4
3-

concentrations were measured using a modified version of the Bernhardt and Wilhelms (1967)
method. Acidified ammonium molybdate was added to seawater to produce phosphomolybdic
acid. Dihydrazine sulfate was then added to the acid to produce phosphomolybdous acid, a blue
compound, whose absorbance was then measured in a 10 mm flow cell at 820 nm. SiO
4
4-

concentrations were measured as described in Armstrong et al. (1967). Acidified ammonium
molybdate was added to seawater to produce silicomolybdic acid and it was then reduced to form

98
silicomolybdous acid, a blue compound, by the addition of stannous chloride. The absorbance of
the blue compound was then measured at 660 nm using a 10 mm flow cell.

Results
Dissolved copper analyses—Dissolved Cu profiles (Fig. 8A) demonstrated many of the
characteristics (e.g., surface depletion, linear increase with depth, sub-surface and deep
scavenging) that other researchers have identified in both the Atlantic and Pacific Oceans (Boyle
et al. 1977, 1981; Bruland 1980). The concentrations were similar to, but generally lower than,
those measured by other researchers in the North Atlantic (Boyle et al. 1981; Yeats et al. 1995;
Saager et al. 1997). They were strongly correlated with SiO
4
4-
concentrations (R
2
= 0.68) (Fig.
9C), particularly at the open ocean stations, but were very weakly correlated with NO
3
-
or PO
4
3-

concentrations (Fig. 9A–B). Concentrations ranged from 0.43 nmol L
-1
at the surface to 3.07
nmol L
-1
near the seafloor and were lower overall in the upwelling region at the eastern end of
the transect than at the western end. The deep concentrations at the western end were similar to
those at the eastern end, however. The highest concentrations were measured in the dense, cold,
and fresher deep waters of Sta. 20 and 22, immediately west of the Cape Verde Islands. Yet
many profiles also exhibited features that provide new insight into the behavior of Cu through
the water column and across areas with distinct biogeochemical regimes.
The 6 westernmost stations (Sta. 1, 2, 3, 6, 8, and 10) featured surface maxima that gradually
weakened as we moved east past Bermuda, perhaps reflecting declining continental runoff and
anthropogenic aerosols from the North American continent. Cu-enriched coastal waters mixing
with open ocean waters would also tend to reduce the coastal dissolved Cu signature as we
moved off the continental shelf and into the central gyre region. These trends are supported by

99
the strong correlation (R
2
= 0.80) between salinity and dissolved Cu concentrations in surface
samples collected from the upper 50 m (Fig. 10). Dissolved Cu concentrations remained
elevated below the surface and in most of the water column at these stations—especially the 3
located on the U.S. east coast continental margin—suggesting some combination of benthic
inputs and lateral inputs from the shelf in addition to aerosols. The surface maxima at the first 3
stations rapidly decreased to form sub-surface minima within the upper 200 m of the water
column, likely indicating strong biological uptake as evidenced by the high fluorescence
readings in this region (Fig. 3). Some of the subsurface local maxima may have also reflected
inputs from the different water masses that make up the North Atlantic Deep Water (NADW) in
this region. The Deep Western Boundary Current (DWBC) is the primary conduit of NADW
from high-latitude water mass formation areas to the subtropics (Peña-Molino et al. 2011) and is
occupied by Labrador Sea Water (LSW) at intermediate depths. In this study, the LSW layer,
made up of the upper LSW (ULSW) and central LSW (CLSW) sub-layers, extended from about
750 m to 2500 m (W. M. Smethie unpubl.) (Fig. 11). Below the LSW, the Overflow Water
(OW) layer, made up of the Denmark Straight Overflow Water (DSOW) and Iceland-Scotland
Overflow Water (ISOW) sub-layers, extended from 2500 m to about 4500 m. Most of the local
maxima were observed between 750 m and 2500 m, which could indicate a potential dissolved
Cu source from LSW.
Remarkably, the surface concentrations at the 2 westernmost and shallowest stations were
almost identical to or even exceeded the bottom concentrations, with the profiles appearing
almost conservative below 500 m (especially Sta. 2). Their structure could be attributable to
rapid scavenging by sinking particulate organic matter (POM) from the surface waters, where
large-celled diatoms were abundant (D. C. Ohnemus and P. J. Lam unpubl.), and by sediments

100
resuspended from the shelf by bottom-scouring DWBC waters. Sta. 6, 8, and 10 had thick
benthic nepheloid layers (BNLs) that became most pronounced—in terms of suspended
particulate matter load—just above the ocean floor (D. C. Ohnemus and P. J. Lam unpubl.). At
Sta. 6, the layer’s scavenging effect on dissolved Cu became noticeable below 4000 m, with the
concentration staying essentially fixed around ~2.05 nmol L
-1
. At Sta. 10, the dissolved Cu
concentration varied more than it did at station 6 but remained around ~2.24 nmol L
-1
below
4250 m. The scavenging effect seemed much less apparent, if non-existent, at Sta. 8 as the
concentration increased all the way to the bottom.
Depth profiles at the stations between Bermuda and the Cape Verde Islands were generally
similar in structure, featuring slight to pronounced surface minima and concentrations that
gradually increased with depth, but with some notable departures. The Trans-Atlantic
Geotraverse (TAG) vent field station (Sta. 16) profile, for instance, showed intense scavenging
of dissolved Cu within a hydrothermal plume located 200 m above the seafloor. Within the ~170
m wide plume, dissolved Cu concentrations ranged from 1.44 nmol L
-1
to 1.66 nmol L
-1
. By
comparison, the concentrations for the samples collected directly above and below the plume
were 2.07 nmol L
-1
and 2.66 nmol L
-1
, respectively. Sta. 18 had pronounced surface (1.32 nmol
L
-1
at 2 m) and subsurface maxima (1.65 nmol L
-1
at 235 m), but it is unclear why. The maxima
and elevated concentrations cannot be traced to any unusual features in other datasets so it is
possible that the samples were contaminated. At Sta. 20, the dissolved Cu concentration more
than doubled between 40 m (0.61 nmol L
-1
) and 440 m (1.31 nmol L
-1
), the depth of a local
maximum. The concentration also rapidly increased between 1900 m and 3900 m, practically
doubling in value (1.36 nmol L
-1
to 2.51 nmol L
-1
), after which it remained almost constant down
to the seafloor, increasing by only 0.28 nmol L
-1
. While the reason for the accelerated increase

101
in the dissolved Cu concentration is unclear, the lack of change in the concentration below 3900
m is likely due to strong scavenging. Sta. 22 was also characterized by both a surface minimum
and local maximum (at 900 m) though both were much less pronounced, with the dissolved Cu
concentration only increasing from 0.91 nmol L
-1
to 1.54 nmol L
-1
. It also featured a local
minimum at 1500 m. The increase in concentration between 1800 m and 3900 m (1.45 nmol L
-1

to 2.80 nmol L
-1
) was similar to that between 1900 m and 3900 m at Sta. 20, and the
concentration similarly remained almost unaltered down to the seafloor, increasing by 0.27 nmol
L
-1
(though the bottom depth was ~1000 m shallower than that for Sta. 20). Interestingly, Sta. 22
featured the highest dissolved Cu concentration (3.07 nmol L
-1
) of the transect despite not being
the deepest station. Low CFC-11 and CFC-12 concentrations in the deep waters of Sta. 20 and
22 indicates the presence of old, relatively unventilated waters (W. M. Smethie unpubl.), as do
the elevated thorium-230 (
230
Th) concentrations measured there (R. F. Anderson pers. comm.).
These findings suggest that the high dissolved Cu concentrations observed at depth at Sta. 20 and
22 are the result of dissolved Cu accumulating in these older water masses over a long period of
time. The high
230
Th concentrations also suggest minimal scavenging pressure when compared
to other Sta.
The 4 stations between Mauritania and the Cape Verde Islands (Sta. 25, 26, 27, and 28)
featured some of the lowest dissolved Cu concentrations within the upper water column, likely
due to the region’s high nutrient levels (Fig. 6). High aerosol fluxes from the African continent,
as evidenced by the elevated surface dissolved iron (Fe) and aluminum (Al) concentrations (M.
Hatta and C. I. Measures unpubl.), seemingly had little effect upon the surface dissolved Cu
concentrations as most stations featured surface minima, unlike the westernmost stations. Only
Sta. 27 displayed a pronounced surface maximum, but that elevated concentration could have

102
been the result of sample contamination as it is significantly higher than the concentration at the
same depth for the other 3 stations in the area. The lowest concentrations were clustered within
the upper 200 m near the depth of the chlorophyll maximum. Sta. 28, the station closest to
Mauritania, had a large BNL about 100 m above the seafloor in which dissolved Cu was strongly
scavenged with the concentration within the layer (1.54 nmol L
-1
at 2905 m) remaining almost
unchanged from that above it (1.52 nmol L
-1
at 2600 m). Yet only 100 m below, right above the
seafloor, the dissolved Cu concentration increased by over 40% to 2.18 nmol L
-1
. The marked
bottom increase likely reflects inputs from the shelf sediments. It is interesting that dissolved Cu
concentrations at Sta. 28 are not higher than those at Sta. 27, 26, and 25, particularly at depth,
given its proximity to the Mauritanian coast and the presence of nutrient-rich upwelled waters. It
may be that any putative increase in dissolved Cu concentration is being counterbalanced by high
upwelling-fueled biological drawdown, resulting in the concentrations not varying much
between Mauritania and the Cape Verde Islands.
Cu
2+
analyses—One of the most striking aspects of this dataset is how uniformly low Cu
2+

concentrations were across the transect and how little they varied below the euphotic zone (Fig.
8B). The measured values in this study were generally much lower than those from previous
studies done in the Atlantic and Pacific, particularly at depth, though surface measurements were
often within the same order of magnitude (Moffett 1995; Moffett and Dupont 2007). However,
as more recently published datasets from either ocean basin are largely unavailable, it is difficult
to know whether to ascribe these discrepancies simply to varying methodologies or to actual
differences in concentration and complexation. Most concentrations ranged between ~10
-14
and
10
-13
mol L
-1
; the lowest and highest concentrations were 1.54 fmol L
-1
and 1.07 pmol L
-1
. The
median and average concentrations were 22.3 fmol L
-1
and 60.5 fmol L
-1
, respectively. Cu
2+

103
concentrations were very weakly correlated with SiO
4
4-
concentrations (Fig. 12C) but did not
correlate with NO
3
-
and PO
4
3-
concentrations (Fig. 12A–B).
The lowest concentrations were measured in the euphotic zone at the 4 stations between
Mauritania and the Cape Verde Islands, which also had some of the lowest dissolved Cu
concentrations of the transect, and were likely the result of strong biological drawdown in the
area. Concentrations within the upper 200 m of these productive waters ranged mostly from 1.54
fmol L
-1
to 7.63 fmol L
-1
. The highest concentrations were found near the seafloor and in the
surrounding deep waters at the 2 stations immediately west of the Cape Verde Islands (Sta. 20
and 22), where the highest dissolved Cu concentrations were also measured. Depth profiles often
did not display much structure below the euphotic zone—in which there was often significant
small-scale variability—outside of the occasional local maximum or minimum. Another
exception occurred when the concentrations of dissolved Cu and organic ligands were very close,
or when the latter exceeded the former, which would cause the Cu
2+
concentration to
significantly increase; indeed, the highest Cu
2+
concentrations were mostly at depths where the
ligand concentration was close to or less than the dissolved Cu concentration.
In spite of the elevated dissolved Cu concentrations, the Cu
2+
concentrations within the
euphotic zone at the 2 westernmost stations were not exceptionally higher than those in the
Mauritanian area, suggesting that scavenging by biological uptake is also an important control on
the continental margin. Higher concentrations were found further down the water column and
above the seafloor, however, likely because of remineralizing sinking POM and resuspended
benthic inputs. The significant decrease in the Cu
2+
concentration from 1500 m to 2050 m, or
right above the seafloor (despite the concurrent increase in the dissolved Cu concentration), at
Sta. 1 indicates that the resuspended shelf sediments could also represent a source of organic

104
ligands, as could potentially LSW waters. The same could be true for Sta. 3, where a small
increase in the bottom dissolved Cu concentration led into only a slight increase in the Cu
2+

concentration. The scavenging effects of the BNLs on the bottom Cu
2+
concentrations at Sta. 6,
8, and 10 were similar to those on the dissolved Cu concentrations. At both Sta. 6 and 8, the
Cu
2+
concentration either decreased within the layer or remained stable. No scavenging effect
was discernible at Sta. 10, however, as the bottom Cu
2+
concentration increased.
Between Bermuda and the Cape Verde Islands the Cu
2+
profiles were mostly similar, showing
gradients in the euphotic zone around the depth of the chlorophyll maximum, typically in the
form of a local minimum, but otherwise gradually increasing with depth, mirroring the dissolved
Cu profiles. Most stations featured subsurface local maxima between 500 m and 2500 m that
could again be attributable to remineralizing sinking organic material. At Sta. 16, Cu
2+
, like
dissolved Cu, was intensely scavenged out of the water column in the hydrothermal plume, with
the concentrations within it falling to levels observed in the upper water column. Whereas the
dissolved Cu concentration increased by almost 40% below the plume, the Cu
2+
concentration
increased only slightly. Concentrations at Sta. 20 showed the strongest gradients throughout the
water column, ranging from 7.52 fmol L
-1
at 40 m to 1.07 pmol L
-1
at 4300 m (the bottom
concentration was 0.62 pmol L
-1
), because of how close the dissolved Cu and organic ligand
concentrations were below 3000 m.
Cu
2+
concentrations at Sta. 25 – 28 were consistently among the lowest of the transect and
showed little variability from surface to seafloor, especially compared with Sta. 20 and 22. The
absolute lowest concentrations of this section, and of the transect, were measured in the upper
100 m of Sta. 28 around the depths of the chlorophyll maximum and primary nitrite maximum
(PNM) layers (Fig. 3; 7). Yet Cu
2+
concentrations within the upper 500 m at Sta. 28 also showed

105
strong gradients, varying by almost an order of magnitude, suggesting rapid remineralization and
turnover of organic matter. Concentrations within and below the nepheloid layer were almost
unaltered relative to that of the next shallowest depth, again indicating scavenging by the
suspended particulate matter. Like Sta. 28, Sta. 25 – 27 featured pronounced Cu
2+
minima
around the depth of the chlorophyll maximum. While surface Cu
2+
concentrations remained
uniformly low in this area, deep concentrations gradually increased moving west from Sta. 28 to
Sta. 25 as the water column also deepened.
Organic ligand analyses—Ligand concentrations mostly exceeded dissolved Cu
concentrations across the transect and were typically highest in the mid water column or in the
deep waters, perhaps reflecting a source from remineralized sinking organic matter (Fig. 8C; 13).
The lowest and highest concentrations were 0.82 nmol L
-1
and 5.26 nmol L
-1
(Table 1). Organic
ligands with high conditional stability constants (K) were detected throughout the water column,
but the strongest ones were typically clustered in the upper 1000 m, suggesting a biological
origin (Fig. 14). Log K values ranged from 12.5 to 14.8 and the median value was 13.7.
Throughout most of the transect strong organic ligands acted as a buffer for Cu
2+
concentrations,
restricting them within a narrow concentration range spanning one order of magnitude (~10
-14
to
10
-13
mol L
-1
)—effectively a biological “steady-state” range. The few times that Cu
2+

concentrations moved out of that range occurred when biological productivity was high, as was
the case in the upwelled waters between Mauritania and the Cape Verde Islands, or when
dissolved Cu concentrations exceeded ligand concentrations. Even so, Cu
2+
concentrations did
not move too far out of that range, usually decreasing or increasing by an order of magnitude at
most, respectively.

106
At Sta. 1, 3, and 6 at the western end of the transect, elevated concentrations of strong
biologically-produced ligands (log K > 13.3) in the upper 250 m kept Cu
2+
concentrations lower
than in the mid water column (500 m to 2000 m) despite comparably high dissolved Cu
concentrations. The elevated ligand concentrations may have also helped stabilize the surface
maxima, particularly at Sta. 1 and 6 where they are very consistent. The BNLs at Sta. 6, 8, and
10 were not strong ligand sources; instead, it seemed as though the ligands were scavenged as
their concentrations fell within the layers at Sta. 8 and 10 and became almost identical to
dissolved Cu concentrations. While the bottom Cu
2+
concentration at Sta. 10 rose
correspondingly, the two deepest concentrations at station 8 changed very little compared to the
concentrations at the depths above where ligand concentrations outstripped dissolved Cu
concentrations.
Between Sta. 12 and 18, organic ligand concentrations remained high in the upper water
column as dissolved Cu concentrations there decreased, sometimes exceeding them four- to
fivefold. Ligand concentrations below the euphotic zone continued to exceed dissolved Cu
concentrations, but usually by reduced margins. Despite significantly lower productivity levels
strong ligands remained abundant in the upper 1000 m although weaker ligands (log K < 13)
were also found in the mid and deep water columns. At Sta. 16, elevated concentrations of very
strong ligands (log K > 14) were measured in and directly below the hydrothermal plume,
suggesting that the hydrothermal vent represents a significant source of ligands and contributing
to the very low Cu
2+
concentrations. These strong ligands may help stabilize a large fraction of
dissolved Cu and prevent more from being scavenged within the plume. Very strong ligands
were also found further up the water column, between 1000 and 2000 m, and in the adjoining
waters of Sta. 18, perhaps indicating vertical advection from the hydrothermal plume waters. At

107
Sta. 20 and 22 ligand concentrations continued to exceed dissolved Cu concentrations in the
upper water column but below 3000 m and 2000 m, respectively, the opposite began to happen
more frequently, leading Cu
2+
concentrations in the deeper waters to rise by up to an order of
magnitude compared to mid water concentrations.
In the area between Mauritania and the Cape Verde Islands, organic ligand concentrations
remained well in excess of dissolved Cu concentrations throughout the water column. Moreover,
the log K values of these ligands were very high from surface to deep waters, often exceeding 14
even near the seafloor. While a significant fraction of these strong ligands probably derive from
biological processes in the euphotic zone, some may also originate from the deep upwelled
waters and from laterally advected shelf sediments. Within and below the BNL at Sta. 28, ligand
concentrations were elevated but the ligands’ log K values (13.7 – 13.8) were slightly lower than
those of the surface and mid water ligands. These ligands were therefore also weaker than the
ligands found within the hydrothermal vent plume at Sta. 16 but comparable in strength to those
found within the BNLs at Sta. 6, 8, and 10. The difference between the ligand and dissolved Cu
concentrations below 2000 m narrowed moving west between Sta. 28 and Sta. 25, resulting in
the gradual increase in deep Cu
2+
concentrations over the same interval.
The results of the complexiometric titrations performed using a lower detection window, 1
µmol L
-1
SA (Table 2), revealed the presence of organic ligands with sometimes much smaller
log K values but their concentrations were much more similar to those of the stronger ligand
class (measured using either the 2 or 2.5 µmol L
-1
SA detection window) than had been expected.
The log K values ranged from 12.1 to 13.9—the latter value only slightly smaller than some of
the highest log K values—while the ligand concentrations ranged from 2.25 nmol L
-1
to 5.22
nmol L
-1
. There was no clear trend in the concentrations and log K values although it does seem

108
that even the weaker ligands found near the seafloor at the stations between Bermuda and the
Cape Verde Islands remain quite strong, particularly at Sta. 16 and 18 where log K > 13.5, as
opposed to the weaker ligands found at depth at the stations west of Bermuda. Concentrations of
the weak ligands found within the central gyre region were similar to or even lower than those of
the stronger ligands, which was unexpected since weak ligands are typically more abundant.

Discussion
Dissolved Cu analyses—The profile structure and distribution of dissolved Cu across the
North Atlantic were defined by 3 primary features: the surface maxima of the westernmost
stations, the BNLs and hydrothermal plume, and the high upwelling-fueled primary productivity
of the easternmost stations. These features helped delineate the geographic boundaries that
separated the distinct biogeochemical regimes that we encountered during the cruises: the broad
western boundary continental shelf, the central gyre region, and the upwelling system off
western Africa.
Sta. 1, 2, 3, 6, 8, and 10 appeared to be more influenced by Cu-enriched coastal waters and
anthropogenic aerosols, as indicated by the close correlation between lower salinity and higher
dissolved Cu values in the surface layer (Fig. 10), resulting in the pattern of observed maxima
and elevated upper water column concentrations. Continental and fluvial runoff would be
expected to decrease as one moved away from the North American coastline, and that trend
seemed to also be reflected in the gradually declining surface dissolved Cu values from Sta. 1 to
8. Boyle et al. (1977) posited that the surface maximum that they observed during their study
could be the result of a continental shelf source after noting the simultaneous presence of a
surface maximum in
228
Ra (Knauss et al. 1978). Similarly, elevated
228
Ra concentrations were

109
measured within the upper 500 m at Sta. 1 – 10, suggesting that large fluxes of terrigenous
sediments from the broad shelf may have also influenced upper water dissolved Cu
concentrations (M. Charette, P. Henderson, P. Morris and W. S. Moore unpubl.). Because of
western Africa’s much narrower shelf, one would not expect these Cu-releasing processes to
significantly influence deep and upper water column dissolved Cu values at the eastern end of
the transect, and indeed
228
Ra levels there were only elevated near the surface and within the
nepheloid layer at Sta. 28 (M. Charette et al. unpubl.). The absence of elevated concentrations
throughout the water column at Sta. 25 – 28 despite the presence of upwelled waters can likely
be attributed to high biological uptake in the area counteracting any potential increases in
dissolved Cu levels.
The varying scavenging effects of the BNLs and hydrothermal plume on dissolved Cu may be
related to their intensity, or SPM load, and composition. The BNLs at Sta. 6 and 8 were much
more intense than the BNL at Sta. 28 but their particle compositions differed in that the
Mauritanian layer had significantly more Mn oxides (MnO
2
) and Fe oxyhydroxides (Fe(OH)
3
).
Sta. 10 and Sta. 28 had similar SPM loads but the particle composition of Sta. 10’s BNL was
more similar to that of Sta. 6 and 8 (D. C. Ohnemus and P. J. Lam unpubl.). While the
scavenging effects of the nepheloid layers are apparent in the dissolved Cu profiles of Sta. 6, 10,
and 28, they are not at Sta. 8. Sediment inputs at Sta. 8 may be partially masking the effects of
the layer but it is difficult to say without having more measurements between 4500 m and 5000
m. More measurements may have revealed that the dissolved Cu concentration remained mostly
constant between those two depths, reflecting scavenging by the BNL. It is worth noting that the
nepheloid layers’ differing SPM loads and composition did not seem to significantly influence
the extent to which dissolved Cu was scavenged.

110
The hydrothermal plume’s SPM, though much smaller than the nepheloid layers’, is made up
of up to 50% Fe(OH)
3
(D. C. Ohnemus and P. J. Lam unpubl.) and oxyhydroxides are very
effective scavengers of dissolved Cu (Trefry et al. 1985; Trocine and Trefry 1988; German et al.
1991). The significantly higher fraction of Fe(OH)
3
in the hydrothermal plume’s particles, over
an order of magnitude higher than what was found in Sta. 28’s BNL (D. C. Ohnemus and P. J.
Lam unpubl.), was likely the reason for the large discrepancy between the extent to which
dissolved Cu was scavenged at Sta. 16 relative to Sta. 6, 8, 10, and 28. Indeed, Trefry et al.
(1985) found that suspended particles in hydrothermal plumes along the MAR were primarily
composed of Cu- and Zn-enriched Fe oxyhydroxides while German et al. (1991) observed that
Cu and Zn appeared to be preferentially removed (relative to Fe) from neutrally buoyant plume
particles. German et al. (1991) also observed that while plume material can be dispersed
laterally by deep ocean currents and thereby transported great distances away from the vent site,
particulate material becomes trapped within the neutrally buoyant plume and does not travel
far—only about 200 to 400 m above the seafloor. Trocine and Trefry (1988) similarly noted that
unlike Fe particulate Cu, Cd, and Zn values fell precipitously away from the vent source,
reflecting differential settling and dissolution rates for Cu-, Cd-, and Zn-bearing phases. As a
result, while plume material may provide a source of dissolved Cu to the deep water column at
nearby stations, the scavenging effects of the plume remain localized to the vent site. Sander and
Koschinsky (2011) recently estimated that hydrothermal vent fluids could account for up to 14%
of the deep ocean Cu budget so, alongside remineralized POM, the TAG vent site could
represent one of the largest sources of dissolved Cu to the mid and deep water columns of the
central gyre stations of this transect.

111
The stations between Bermuda and the Cape Verde Islands were located in the central gyre
region and were therefore not significantly affected by either anthropogenic inputs or upwelled
waters. They were, however, the deepest stations sampled during the cruises, and one, Sta. 16,
was situated atop the TAG vent field on the MAR. Furthermore, while productivity levels in this
region were low compared to the western- and easternmost stations, dense aggregations of
Trichodesmium spp., as indicated by significant amounts of “tufts” and “puffs,” were observed in
the surface waters, especially around Sta. 16, 18 and 20 (D. C. Ohnemus and P. J. Lam unpubl.).
This was in accordance with the findings of Davis and McGillicuddy Jr. (2006), who also
discovered high abundances of Trichodesmium colonies in the tropical and subtropical North
Atlantic, particularly in the western Sargasso Sea (west of Sta. 14). Trichodesmium require Cu
for their photosynthetic and respiratory electron transport systems (Bernroitner et al. 2008), and
large colonies are known to provide a habitat for various bacteria, including other cyanobacteria
(Capone et al. 1997). These Trichodesmium colonies and associated microbial populations may
have therefore been partially responsible for the low upper water concentrations.
Pronounced surface minima and low upper water column concentrations were characteristic
of the dissolved Cu distribution at the 4 easternmost stations. Influxes of dissolved Cu from the
Mauritanian shelf and coastal waters and upwelled waters did not appear to greatly influence
concentrations anywhere in the water column. This region was also subject to elevated dust
inputs from North Africa but, unlike the North American aerosols, they did not seem to affect
dissolved Cu concentrations at Sta. 25 – 28. In a study investigating the fractional solubility of
Cu in various marine aerosols, Sholkovitz et al. (2010) determined that the solubility of Cu in
Saharan dust was much lower than that in anthropogenic aerosols with non-Saharan sources.
Our results and those of other cruise participants (R. U. Shelley and P. Morton unpubl.) seem to

112
indicate that anthropogenic aerosols originating from North America influence surface dissolved
Cu concentrations while Saharan aerosols from North Africa would not, in agreement with
Sholkovitz et al.’s conclusions (2010). Although the aerosol Cu loadings in the North African
and North American aerosols were very similar, the fractional solubility of Cu in each aerosol
type was very different. The average solubility of Cu in North African aerosols during both
cruises was 2.2 ± 1.2% (and 1.1% specifically for the area between Mauritania and the Cape
Verde Islands) while that in North American aerosols was 25 ± 14%; both means are comparable
to the ranges determined by Sholkovitz et al. (2010; 1 – 7% and 10 – 100%, respectively).
Furthermore, Shelley and Morton (unpubl.) determined that the enrichment factor for Cu in the
North African aerosols was ~1, or crustal in origin based on the average upper continental crust
(UCC) value (25 ppm) calculated by Taylor and McLennan (1995), whereas the enrichment
factor for Cu in the North American aerosols ranged between ~15 to ~40. The discrepancies in
the fractional solubilities and enrichment factors of Cu between these two aerosol types could
account for why surface dissolved Cu concentrations were elevated at the western end of the
transect but not at the eastern end.
Cu
2+
, organic ligand and conditional stability constant analyses—Unlike dissolved Cu, Cu
2+

profiles were defined by their uniformly low concentrations throughout most of the water
column, with values ranging primarily between ~10
-14
and 10
-13
mol L
-1
. Our findings are in
good agreement with those of Moffett and Dupont (2007), who also found Cu
2+
to be strongly
complexed even at depth in the North Pacific and Bering Sea, yielding deep Cu
2+
and ligand
concentrations comparable to surface ones. It is interesting that our surface Cu
2+
and ligand
measurements were often similar to those of past studies, regardless of location or methodology
(Coale and Bruland 1990; Moffett 1995; Moffett and Dupont 2007). Other than the upwelling

113
region between Mauritania and the Cape Verde Islands where we measured the lowest Cu
2+

concentrations, values in the euphotic zone across the transect typically ranged between ~ 10 and
50 fmol L
-1
, similar to the concentration ranges (~10 – 40 fmol L
-1
) measured by other
researchers (Coale and Bruland 1990; Moffett and Dupont 2007). The presence of several sub-
surface maxima, typically between 500 m and 2000 m, were noted, but the poor correlations
between Cu
2+
and nutrient concentrations (Fig. 12A–C) suggest that they are not the result of
remineralization.
Ligand concentrations in this study ranged from ~1.5 to 4 nmol L
-1
while those in other
studies ranged from 0.9 to 4 nmol L
-1
(Coale and Bruland 1990; Moffett 1995; Moffett and
Dupont 2007). These similarities suggest that there is a “steady-state” or “buffered” Cu
2+

concentration range that is maintained by various natural processes—primarily scavenging and
organic complexation—common to both ocean basins so as to simultaneously prevent
concentrations from becoming too low, thus potentially inducing Cu limitation, or too high and
therefore toxic. For instance, whereas growth rates of the oceanic diatom Thalassiosira
weissflogii and the ammonia-oxidizing archaeon (AOA) Nitrosopumilus maritimus can be Cu-
limited by Cu
2+
concentrations of ~12.6 fmol L
-1
and < 0.2 pmol L
-1
, respectively (Peers et al.
2005; Amin et al. 2013), Synechococcus spp. growth rates can be significantly inhibited by
concentrations ≥ 1 pmol L
-1
(Brand et al. 1986).
At the westernmost stations, biological uptake and subsurface scavenging controlled the
distribution of Cu
2+
throughout the water column. Mid to deep water concentrations were often
similar with bottom values sometimes as low as surface values, likely indicating scavenging by
large sinking particles from the surface and sediment inputs or laterally advected inputs from the
coast. Many of these sediment inputs may also be sources of strong organic ligands, resulting in

114
the low values measured across the shelf and within the BNLs. As was the case for dissolved
Cu, Cu
2+
concentrations were reduced or kept near constant by scavenging within the nepheloid
layers but were significantly drawn down within the hydrothermal plume. The Cu
2+
values that
we obtained within the plume (4.48 – 6.5 fmol L
-1
) were significantly lower than those obtained
by Klevenz et al. (2012; ~0.5 pmol L
-1
) within the hydrothermal fluids of two other vent sites
along the MAR. Similarly, our L values (2.6 – 3.03 nmol L
-1
) were much lower than those
obtained by either Klevenz et al. (2012; 51 – 207 nmol L
-1
) or Sander et al. (2007; 60 – 4460
nmol L
-1
) although our log K values were significantly higher (~14.3 versus ~12.19 and 12.48 –
13.46). These large discrepancies stem in large part from the fact that our samples were
collected from the plume 200 – 350 m above the vent field whereas the other researchers’
samples were obtained from hydrothermal fluids diffusing from cracks or black smokers on the
field away from the main vent source. Significant concentrations of Cu, and by extension Cu
2+
,
would therefore still be present in dissolved form and stabilized by equally large or larger
organic ligand concentrations (Sander and Koschinsky 2011). Several hundred meters above the
vent field, most of that dissolved Cu would likely have been scavenged away by sulfides and Fe
oxyhydroxides and what remains would likely be complexed by exceedingly strong organic
ligands, perhaps strong thiols (Sander and Koschinsky 2011), as was the case in this study.
While it is true that the ligand concentrations reported by Klevenz et al. (2012) encapsulate those
of both strong and weak ligands whereas ours only count the former, the differences are simply
too great to be explained away by this or any other methodological deviation, however
important. Our complexiometric analyses using the same detection window (1 µmol L
-1
SA)
employed by Klevenz et al. (2012) detected a ligand class with only slightly reduced log K
values (13.6 – 13.8) but unusually low concentrations (2.25 – 2.93 nmol L
-1
) as well (Table 2).

115
Their findings and our own underline the large gradients in dissolved Cu, organic ligand, and
Cu
2+
concentrations that can exist in hydrothermal fluids separated by only a few hundred
meters.
While not as low as the concentrations at the western- and easternmost stations, Cu
2+

concentrations in the upper water column of the central gyre region between Bermuda and the
Cape Verde Islands were still quite low—perhaps more so than we would have expected given
these waters’ low primary productivity. Yet the ligand concentrations were also elevated,
comparable to those measured in much more productive waters, consistently exceeding dissolved
Cu concentrations in the euphotic zone, and their log K values were typically greater than 13,
suggesting a biological origin. Moreover, the presence of multiple subsurface local Cu
2+

maxima between 500 and 2000 m points to there being significant amounts of sinking POM from
the surface waters. Indeed, Ohnemus and Lam (unpubl.) measured high concentrations of both
sinking and suspended POM in these waters and also observed significant quantities of
Trichodesmium tufts and puffs indicative of large aggregations. Biological uptake by these
Trichodesmium colonies and the cyanobacteria they house (Capone et al. 1997; Bernroitner et al.
2008) could therefore account for the relatively low Cu
2+
concentrations in the euphotic zone.
Jones and Thomas (1988) observed that large Trichodesmium aggregations also produced
significant quantities of ligand-like marine humics that tightly complexed Cu
2+
in the Great
Barrier Reef Lagoon, so the same phenomenon could be at play here and partially explain why
ligand concentrations remained elevated. While Trichodesmium colonies would not be expected
to contribute to export production because of their buoyancy (A. N. Knapp pers. comm.), the
remineralization of sinking POM of cyanobacterial origin could explain the presence of the
multiple subsurface Cu
2+
maxima at these stations.

116
The consistently lowest Cu
2+
concentrations of the transect were measured in the upwelling-
fueled waters between the Cape Verde Islands and Mauritania. More specifically, the lowest
concentration overall, 1.54 fmol L
-1
, was measured near the depths of the chlorophyll and
primary nitrite maxima at Sta. 28, suggesting that the same processes that gave rise to this feature
in the eastern tropical South Pacific (ETSP)—namely ammonia oxidation by ammonia-oxidizing
archaea (AOA) and Fe acquisition by Fe-limited diatoms—may be at play here as well (Jacquot
et al. 2013). Ligand concentrations and log K values were very similar to those measured at the
western end of the transect, and the Cu
2+
depth profiles also showed little variability between the
surface and deep water concentrations. Yet even in these biologically active waters Cu
2+

concentrations never fluctuated by more than an order of magnitude compared to surface
concentrations at the other stations and never dipped below 1 fmol L
-1
. As for the westernmost
stations, bottom and near bottom concentrations remained almost constant, likely reflecting
scavenging and organic complexation by sediment inputs and laterally advected inputs from the
coast. Again much like the other stations the presence of local subsurface maxima within the
upper water column likely indicated remineralization while those below occurred at depths at
which the dissolved Cu and ligand concentrations were very similar. The reasoning behind the
appearance of several subsurface local minima was less clear, however, and could indicate the
combined effects of scavenging and strong organic complexation.

117
Bibliography
Aldrich, A. P., M. Jahme, and C. M. G. Van den Berg. 1999. Antivibration electrode support for
shipboard stripping voltammetry. Electroanal. 11: 1155–1157, doi:10.1002/(SICI)1521-
4109(199911)11:15<1155::AID-ELAN1155>3.3.CO;2-Y
Amin, S. A., J. W. Moffett, W. Martens-Habbena, J. E. Jacquot, Y. Han, A. Devol, A. E. Ingalls,
D. A. Stahl, and E. V. Armbrust. 2013. Copper requirements of the ammonia-oxidizing
archaeon Nitrosopumilus maritimus SCM1 and implications for nitrification in the marine
environment. Limnol. Oceanogr. In press.
Armstrong, F. A. J., C. R. Sterns, and J. D. H. Strickland. 1967. The measurement of upwelling
and subsequent biological processes by means of the Technicon AutoAnalyzer and
associated equipment. Deep-Sea Res. 14: 381–389, doi:10.1016/0011-7471(67)90082-4
Aruoma, O. I., B. Halliwell, E. Gajewski, and M. Dizdaroglu. 1991. Copper-ion-dependent
damage to the bases in DNA in the presence of hydrogen peroxide. Biochem. J. 273: 601–
604.
Bernhardt, H., and A. Wilhelms. 1967. The continuous determination of low-level iron, soluble
phosphate and total phosphate with the AutoAnalyzer. Tech. Symp. I: 385–389.
Bernroitner, M., M. Zamocky, M. Pairer, P. G. Furtmueller, G. A. Peschek, and C. Obinger.
2008. Heme‐copper oxidases and their electron donors in cyanobacterial respiratory
electron transport. Chem. Biodivers. 5: 1927–1961, doi:10.1002/cbdv.200890180
Boyle, E. A., F. Sclater, and J. Edmond. 1977. The distribution of dissolved copper in the
Pacific. Earth and Planet. Sci. Lett. 37: 38–54, doi:10.1016/0012-821X(77)90144-3
Boyle, E. A., S. S. Huested, and S. P. Jones. 1981. On the distribution of copper, nickel, and
cadmium in the surface waters of the North Atlantic and North Pacific Ocean. J. Geophys.

118
Res. 86: 8048–8066, doi:10.1029/JC086iC09p08048
Brand, L. E., W. G. Sunda, and R. R. L. Guillard. 1986. Reduction of marine phytoplankton
reproduction rates by copper and cadmium. J. Exp. Mar. Biol. Ecol. 96: 225–250,
doi:10.1016/0022-0981(86)90205-4
Bruland, K. W. 1980. Oceanographic distributions of cadmium, zinc, nickel, and copper in the
North Pacific. Earth and Planet. Sci. Lett. 47: 176–198, doi:10.1016/0012-821X(80)90035-
7
Bruland, K. W., and R. P. Franks. 1983. Mn, Ni, Cu, Zn and Cd in the western North Atlantic, p.
395–414. In C. S. Wong, E. Boyle, K. W. Bruland, J. D. Burton and E. D. Goldberg [eds.],
Trace metals in sea water. Academic.
Bruland, K. W., E. L. Rue, J. R. Donat, S. A. Skrabal, and J. W. Moffett. 2000. Intercomparison
of voltammetric techniques to determine the chemical speciation of dissolved copper in a
coastal seawater sample. Anal. Chim. Acta 405: 99–113, doi:10.1016/S0003-
2670(99)00675-3
Bruland, K. W., and M. C. Lohan. 2003. Controls of trace metals in seawater, p. 23–47. In H.
Elderfield [ed.], The oceans and geochemistry. Academic.
Buck, K. N., and K. W. Bruland. 2005. Copper speciation in San Francisco Bay: A novel
approach using multiple analytical windows. Mar. Chem. 96: 185–198,
doi:10.1016/j.marchem.2005.01.001
Buck, K. N., J. Moffett, K. A. Barbeau, R. M. Bundy, Y. Kondo, and J. Wu. 2012. The organic
complexation of iron and copper: an intercomparison of competitive ligand exchange-
adsorptive cathodic stripping voltammetry (CLE-ACSV) techniques. Limnol. Oceanogr.:
Methods 10: 496–515, doi:10.4319/lom.2012.10.496

119
Campos, M. L. A. M., and C. M. G. Van den Berg. 1994. Determination of copper complexation
in sea water by cathodic stripping voltammetry and ligand competition with
salicylaldoxime. Anal. Chim. Acta 284: 481–496, doi:10.1016/0003-2670(94)85055-0
Capone, D. G., J. P. Zehr, H. W. Paerl, B. Bergman, and E. J. Carpenter. 1997. Trichodesmium, a
globally significant marine cyanobacterium. Science 276: 1221–1229,
doi:10.1126/science.276.5316.1221
Coale, K. H., and K. W. Bruland. 1988. Copper complexation in the Northeast Pacific. Limnol.
Oceanogr. 1084–1101, doi:10.4319/lo.1988.33.5.1084
Coale, K. H., and K. W. Bruland. 1990. Spatial and temporal variability in copper complexation
in the North Pacific. Deep-Sea Res. 37: 317–336, doi:10.1016/0198-0149(90)90130-N
Cutter, G. A., and K. W. Bruland. 2012. Rapid and noncontaminating sampling system for trace
elements in global ocean surveys. Limnol. Oceanogr.: Methods 10: 425–436,
doi:10.4319/lom.2012.10.425
Davis, C. S., and D. J. McGillicuddy Jr. 2006. Transatlantic abundance of the N
2
-fixing colonial
cyanobacterium Trichodesmium. Science 312: 1517–1520, doi:10.1126/science.1123570
Donat, J. R., K. A. Lao, and K. W. Bruland. 1994. Speciation of dissolved copper and nickel in
South San Francisco Bay: a multi-method approach. Anal. Chim. Acta 284: 547–571,
doi:10.1016/0003-2670(94)85061-5
Ensign, S. A., M. R. Hyman, and D. J. Arp. 1993. In vitro activation of ammonia
monooxygenase from Nitrosomonas europaea by copper. J. Bacteriol. 175: 1971–1980.
German, C. R., A. C. Campbell, and J. M. Edmond. 1991. Hydrothermal scavenging at the Mid-
Atlantic Ridge: Modification of trace element dissolved fluxes. Earth and Planet. Sci. Lett.
107: 101–114, doi:10.1016/0012-821X(91)90047-L

120
Jacquot, J. E., Y. Kondo, A. N. Knapp, and J. W. Moffett. 2013. The speciation of copper across
active gradients in nitrogen-cycle processes in the eastern tropical South Pacific. Limnol
Oceanogr. 58(4): 1387–1394, doi:10.4319/lo.2013.58.4.1387
Jones, G. B., and F. G. Thomas. 1988. Effect of terrestrial and marine humics on copper
speciation in an estuary in the Great Barrier Reef Lagoon. Aust. J. Mar. Fresh. Res. 39(1):
19–31.
Klevenz, V., S. G. Sander, M. Perner, and A. Koschinsky. 2012. Amelioration of free copper by
hydrothermal vent microbes as a response to high copper concentrations. Chem. Ecol. 28(5):
405–420, doi:10.1080/02757540.2012.666531
Knauss, K. G., T.-L. Ku, and W. S. Moore. 1978. Radium and thorium isotopes in the surface
waters of the East Pacific and coastal Southern California. Earth and Planet. Sci. Lett. 39:
235–249, doi:10.1016/0012-821X(78)90199-1
Kogut, M. B. 2002. Cu speciation in estuaries and coastal waters. Ph.D. thesis. Massachusetts
Institute of Technology.
Lee, J.-M., E. A. Boyle, Y. Echegoyen-Sanz, J. N. Fitzsimmons, R. Zhang, and R. A. Kayser.
2011. Analysis of trace metals (Cu, Cd, Pb, and Fe) in seawater using single batch
nitrilotriacetate resin extraction and isotope dilution inductively coupled plasma mass
spectrometry. Anal. Chim. Acta 686: 93–101, doi:10.1016/j.aca.2010.11.052
Maldonado, M. T., A. E. Allen, J. S. Chong, K. Lin, D. Leus, N. Karpenko, and S. L. Harris.
2006. Copper-dependent iron transport in coastal and oceanic diatoms. Limnol. Oceanogr.
51(4): 1729–1743, doi:10.4319/lo.2006.51.4.1729

121
Mann, E. L., N. Ahlgren, J. W. Moffett, and S. W. Chisholm. 2002. Copper toxicity and
cyanobacteria ecology in the Sargasso Sea. Limnol. Oceanogr. 47(4): 976–988,
doi:10.4319/lo.2002.47.4.0976
Moffett, J. W. 1995. Temporal and spatial variability of copper complexation by strong chelators
in the Sargasso Sea. Deep-Sea Res. 42: 1273–1295, doi:10.1016/0967-0637(95)00060-J
Moffett, J. W., and C. L. Dupont. 2007. Cu complexation by organic ligands in the sub-arctic
NW Pacific and Bering Sea. Deep-Sea Res. 54: 586–595, doi:10.1016/j.dsr.2006.12.013
Murrell, J. C., I. R. McDonald, and B. Gilbert. 2000. Regulation of expression of methane
monooxygenases by copper ions. Trends Microbiol. 8: 221–225, doi:10.1016/S0966-
842X(00)01739-X
Peers, G., S. A. Quesnel, and N. M. Price. 2005. Copper requirements for iron acquisition and
growth of coastal and oceanic diatoms. Limnol. Oceanogr. 50(4): 1149–1158, doi:
10.4319/lo.2005.50.4.1149
Peers, G., and N. M. Price. 2006. Copper-containing plastocyanin used for electron transport by
an oceanic diatom. Nature 441: 341–344, doi:10.1038/nature04630
Peña-Molino, B., T. M. Joyce, and J. M. Toole. 2011. Recent changes in the Labrador Sea Water
within the Deep Western Boundary Current southeast of Cape Cod. Deep-Sea Res. 58:
1019–1030, doi:10.1016/j.dsr.2011.07.006
Pohl, C., P. L. Croot, U. Hennings, T. Daberkow, G. Budeus, and M. R. von d Loeff. 2011.
Synoptic transects on the distribution of trace elements (Hg, Pb, Cd, Cu, Ni, Zn, Co, Mn,
Fe, and Al) in surface waters of the Northern- and Southern East Atlantic. J. Mar. Syst. 84:
28–41, doi:10.1016/j.jmarsys.2010.08.003

122
Saager, P. M., H. J. de Baar, J. de Jong, R. F. Nolting, and J. Schijf. 1997. Hydrography and
local sources of dissolved trace metals Mn, Ni, Cu, and Cd in the northeast Atlantic Ocean.
Mar. Chem. 57: 195–216, doi:10.1016/S0304-4203(97)00038-8
Sander, S. G., A. Koschinsky, G. Massoth, M. Stott, and K. A. Hunter. 2007. Organic
complexation of copper in deep-sea hydrothermal vent systems. Environ. Chem. 4: 81–89,
doi:10.1071/EN06086
Sander, S. G., and A. Koschinsky. 2011. Metal flux from hydrothermal vents increased by
organic complexation. Nature Geoscience 4: 145–150, doi:10.1038/NGEO1088
Sarradin, P.-M., M. Waeles, S. Bernagout, C. Le Gall, J. Sarrazin, and R. D. Riso. 2009.
Speciation of dissolved copper within an active hydrothermal edifice on the Lucky Strike
vent field (MAR, 37°N). Sci. Total Environ. 407: 869–878, doi:10.1016/j.ecss.2009.06.011
Sholkovitz, E. R., P. N. Sedwick, and T. M. Church. 2010. On the fractional solubility of copper
in marine aerosols: Toxicity of aeolian copper revisited. Geophys. Res. Lett. 37: L20601,
doi:10.1029/2010GL044817
Stiefel, E. I. 2007. Bioinorganic chemistry and the biogeochemical cycles, p. 7–30. In I. Bertini,
H. Gray, E. Stiefel, and J. Valentine [eds.], Biological inorganic chemistry. University
Science Books. p. 7–30.
Sunda, W. G., R. T. Barber, and S. A. Huntsman. 1981. Phytoplankton growth in nutrient rich
seawater: importance of copper-manganese cellular interactions. J. Mar. Res. 39: 567–586.
Sunda, W. G. 2012. Feedback interactions between trace metal nutrients and phytoplankton in
the ocean. Front. Microbio. 3(204), doi:10.3389/fmicb.2012.00204
Taylor, S. R., and S. M. McLennan. 1995. The geochemical evolution of the continental crust.
Rev. Geophys. 33(2): 241–265, doi:10.1029/95RG00262

123
Trefry, J. H., R. P. Trocine, G. P. Klinkhammer, and P. A. Rona. 1985. Iron and copper
enrichment of suspended particles in dispersed hydrothermal plumes along the mid‐
Atlantic Ridge. Geophys. Res. Lett. 12: 506–509, doi:10.1029/GL012i008p00506
Trocine, R. P., and J. H. Trefry. 1988. Distribution and chemistry of suspended particles from an
active hydrothermal vent site on the Mid-Atlantic Ridge at 26 N. Earth and Planet. Sci.
Lett. 88: 1–15, doi:10.1016/0012-821X(88)90041-6
Walker, C. B., J. R. de La Torre, M. Klotz, H. Urakawa, N. Pinel, D. Arp, C. Brochier-Armanet,
P. Chain, P. Chan, A. Gollabgir, J. Hemp, M. Hügler, E. A. Karr, M. Könneke, M. Shin, T.
J. Lawton, T. Lowe, W. Martens-Habbena, L. A. Sayavedra-Soto, D. Lang, S. M. Sievert,
A. C. Rosenzweig, G. Manning, and D. A. Stahl. 2010. Nitrosopumilus maritimus genome
reveals unique mechanisms for nitrification and autotrophy in globally distributed marine
crenarchaea. Proc. Nat. Acad. Sci. U.S.A. 107(19): 8818–8823,
doi:10.1073/pnas.0913533107
Yeats, P. A., S. Westerlund, and A. R. Flegal. 1995. Cadmium, copper and nickel distributions at
four stations in the eastern central and south Atlantic. Mar. Chem. 49: 283–293,
doi:10.1016/0304-4203(95)00018-M

124
Table 1. Speciation data for Sta. 1 – 28 using 2 and 2.5 µmol L
-1

SA detection windows. Note
that the log K interval is asymmetric.
Station Depth (m) L (nmol L
-1
) Log K
1
2 1.99 ± 0.11 13.9 (13.8 – 14.0)

30 2.52 ± 0.12 14.0 (14.0 – 14.1)

90 2.58 ± 0.056 14.2 (14.1 – 14.2)

110 2.39 ± 0.15 13.7 (13.6 – 13.7)

185 2.43 ± 0.21 13.5 (13.4 – 13.6)
235 2.58 ± 0.17 13.5 (13.4 – 13.6)
525 1.67 ± 0.35 12.9 (12.6 – 13.0)
965 2.11 ± 0.13 13.6 (13.6 – 13.7)
1200 1.63 ± .094 13.7 (13.6 – 13.7)
1500 2.07 ± 0.15 13.5 (13.5 – 13.6)
2050 2.63 ± 0.13 14.4 (14.3 – 14.5)
3 28 2.58 ± 0.22 13.8 (13.7 – 13.9)

60 3.46 ± 0.30 13.3 (13.2 – 13.3)

85 1.65 ± 0.15 14.0 (13.9 – 14.1)

185 0.82 ± 0.073 13.9 (13.8 – 14.0)

417 2.43 ± 0.21 13.5 (13.4 – 13.6)

665 1.62 ± 0.094 13.8 (13.7 – 13.8)

1350 1.66 ± 0.17 13.0 (12.9 – 13.0)

1800 3.55 ± 0.56 12.6 (12.5 – 12.6)

2900 4.93 ± 0.65 12.8 (12.7 – 12.9)
3275 2.36 ± 0.14 13.5 (13.5 – 13.6)
6 65 1.54 ± 0.087 14.8 (14.7 – 14.8)

90 2.24 ± 0.097 14.0 (14.0 – 14.1)

110 2.29 ± 0.20 13.6 (13.6 – 13.7)

166 2.08 ± 0.042 14.3 (14.2 – 14.3)

289 2.36 ± 0.36 13.0 (12.9 – 13.1)

420 1.75 ± 0.089 14.1 (14.0 – 14.2)

600 2.29 ± 0.098 13.7 (13.6 – 13.7)

1085 4.35 ± 0.67 12.5 (12.4 – 12.6)
1300 2.30 ± 0.13 13.4 (13.4 – 13.5)
1600 2.89 ± 0.21 13.2 (13.1 – 13.3)
2400 2.77 ± 0.28 12.9 (12.8 – 13.0)
2700 2.16 ± 0.095 13.7 (13.6 – 13.7)
3600 1.94 ± 0.18 13.3 (13.2 – 13.4)
4535 2.65 ± 0.14 13.6 (13.5 – 13.7)
8 2 1.61 ± 0.08 14.0 (13.9 – 14.0)

30 1.29 ± 0.10 13.9 (13.8 – 14.0)

65 2.57 ± 0.50 13.2 (13.0 – 13.3)

110 2.35 ± 0.22 13.3 (13.2 – 13.4)

285 1.93 ± 0.22 13.4 (13.3 – 13.5)

420 1.26 ± 0.11 14.1 (14.1 – 14.2)

1000 2.23 ± 0.24 12.8 (12.7 – 12.9)

1200 1.65 ± 0.08 13.8 (13.7 – 13.8)

1800 1.23 ± 0.11 13.4 (13.3 – 13.5)

2400 1.89 ± 0.16 13.4 (13.3 – 13.5)

125

3000 3.45 ± 0.48 12.8 (12.7 – 12.9)

3600 3.45 ± 0.68 12.5 (12.4 – 12.6)

4250 3.01 ± 0.79 12.5 (12.3 – 12.7)

4575 2.08 ± 0.12 13.6 (13.5 – 13.7)

4900 2.02 ± 0.16 13.7 (13.5 – 13.8)
10 40 1.67 ± 0.15 13.9 (13.8 – 13.9)

84 1.45 ± 0.084 14.3 (14.3 – 14.4)

96 3.73 ± 0.24 13.3 (13.3 – 13.4)

135 1.76 ± 0.12 14.2 (14.1 – 14.3)

285 1.49 ± 0.15 13.6 (13.5 – 13.7)
575 1.51 ± 0.073 14.4 (14.3 – 14.4)
800 2.77 ± 0.27 13.1 (13.0 – 13.2)
1350 1.99 ± 0.067 13.8 (13.8 – 13.9)
1650 5.20 ± 0.82 12.6 (12.5 – 12.7)
3300 2.16 ± 0.12 14.0 (13.9 – 14.1)
3900 4.26 ± 0.30 13.0 (12.9 – 13.0)
4531 2.29 ± 0.13 13.6 (13.5 – 13.7)
12 2 2.29 ± 0.41 13.7 (13.5 – 13.8)
40 1.60 ± .10 14.1 (14.1 – 14.2)
65 1.76 ± 0.071 14.7 (14.7 – 14.8)
99 1.42 ± 0.087 14.2 (14.2 – 14.3)
136 1.91 ± 0.11 14.1 (14.1 – 14.2)
351 1.75 ± 0.12 13.8 (13.7 – 13.8)
500 2.08 ± 0.14 13.7 (13.6 – 13.7)
750 2.15 ± 0.15 13.4 (13.3 – 13.4)
1000 1.79 ± 0.11 14.1 (14.1 – 14.2)
1350 3.49 ± 0.28 12.9 (12.9 – 13.0)
1800 2.56 ± 0.12 13.5 (13.5 – 13.6)
2550 2.18 ± 0.11 14.1 (14.0 – 14.1)
3600 2.02 ± 0.11 13.4 (13.4 – 13.5)
4700 3.72 ± 0.21 13.4 (13.3 – 13.4)
5100 2.24 ± 0.09 13.8 (13.7 – 13.8)
5615 2.92 ± 0.15 13.7 (13.6 – 13.8)
14 2 2.93 ± 0.26 13.2 (13.1 – 13.3)
100 3.06 ± 0.30 13.3 (13.3 – 13.4)
285 3.05 ± 0.21 13.4 (13.4 – 13.5)
650 3.85 ± 0.28 13.1 (13.0 – 13.2)
965 2.91 ± 0.41 12.8 (12.7 – 12.9)
1449 2.55 ± 0.29 12.5 (12.4 – 12.6)
1800 2.37 ± 0.16 13.4 (13.4 – 13.5)
2100 3.66 ± 0.39 12.8 (12.7 – 12.9)
2700 3.55 ± 0.16 13.2 (13.2 – 13.3)
3400 3.80 ± 0.29 12.9 (12.8 – 12.9)
4225 4.56 ± 0.21 13.0 (13.0 – 13.1)
16 2 1.70 ± 0.31 13.5 (13.3 – 13.6)
40 2.87 ± 0.34 13.0 (12.9 – 13.0)
92 4.45 ± 0.66 12.7 (12.6 – 12.8)
135 1.77 ± 0.13 14.1 (14.1 – 14.2)
235 1.29 ± 0.084 13.9 (13.8 – 14.0)

126
510 2.09 ± 0.072 14.0 (13.9 – 14.0)
800 2.32 ± 0.20 14.3 (14.1 – 14.3)
1200 2.28 ± 0.21 13.3 (13.2 – 13.4)
1500 1.63 ± 0.083 14.0 (13.9 – 14.1)
1800 2.30 ± 0.14 13.9 (13.8 – 14.0)
2100 2.69 ± 0.063 13.9 (13.85 – 13.9)
2400 2.47 ± 0.18 13.3 (13.2 – 13.4)
3000 3.19 ± 0.21 13.2 (13.1 – 13.2)
3224 2.60 ± 0.12 14.4 (14.3 – 14.5)
3255 2.93 ± 0.20 14.4 (14.3 – 14.5)
3333 3.03 ± 0.11 14.1 (14.1 – 14.2)
3393 2.70 ± 0.12 14.4 (14.4 – 14.5)
3589 4.34 ± 0.090 14.2 (14.2 – 14.3)
18 2 2.58 ± 0.21 13.9 (13.8 – 14.0)
40 4.01 ± 0.30 13.4 (13.3 – 13.4)
110 1.91 ± 0.13 14.1 (14.0 – 14.1)
286 2.29 ± 0.061 14.4 (14.3 – 14.5)
350 2.69 ± 0.19 13.7 (13.6 – 13.8)
600 1.49 ± 0.074 14.2 (14.1 – 14.3)
775 1.71 ± 0.095 13.8 (13.7 – 13.8)
1325 2.02 ± 0.083 14.1 (14.0 – 14.1)
1800 1.78 ± 0.085 14.2 (14.1 – 14.3)
2550 2.97 ± 0.28 13.3 (13.2 – 13.4)
3000 5.26 ± 0.39 12.8 (12.7 – 12.9)
3800 3.11 ± 0.10 13.7 (13.7 – 13.8)
4300 3.13 ± 0.14 14.0 (13.9 – 14.0)
20 40 1.82 ± 0.11 13.8 (13.8 – 13.9)
75 3.37 ± 0.40 13.4 (13.3 – 13.5)
135 3.05 ± 0.39 13.1 (13.0 – 13.2)
300 2.00 ± 0.55 13.1 (12.8 – 13.2)
600 2.38 ± 0.38 13.0 (12.9 – 13.1)
1061 1.65 ± 0.18 13.4 (13.3 – 13.5)
1625 2.55 ± 0.23 13.1 (13.0 – 13.1)
2301 2.63 ± 0.21 13.1 (13.0 – 13.1)
3100 1.92 ± 0.074 13.7 (13.6 – 13.7)
3900 2.40 ± 0.26 13.4 (13.2 – 13.6)
4300 2.22 ± 0.15 13.4 (13.3 – 13.5)
4900 2.68 ± 0.13 13.3 (13.2 – 13.4)
5800 2.76 ± 0.14 13.5 (13.4 – 13.6)
22 2 1.80 ± 0.096 14.2 (14.1 – 14.2)
51 1.69 ± 0.19 13.7 (13.6 – 13.8)
125 1.57 ± 0.06 14.3 (14.2 – 14.3)
236 2.34 ± 0.15 13.7 (13.7 – 13.8)
390 3.25 ± 0.18 13.5 (13.5 – 13.6)
665 1.52 ± 0.12 13.6 (13.4 – 13.7)
900 1.99 ± 0.064 14.2 (14.1 – 14.2)
1200 1.81 ± 0.10 13.8 (13.7 – 13.9)
1800 3.22 ± 0.17 13.3 (13.3 – 13.4)
2100 2.20 ± 0.18 13.5 (13.4 – 13.6)

127
2700 3.26 ± 0.25 13.1 (13.0 – 13.2)
3300 2.75 ± 0.13 13.9 (13.8 – 14.0)
3900 3.71 ± 0.38 13.0 (12.9 – 13.0)
4200 3.94 ± 0.23 13.3 (13.2 – 13.4)
4985 2.72 ± 0.16 13.7 (13.6 – 13.9)
24 71 1.84 ± 0.14 13.5 (13.5 – 13.6)
135 3.98 ± 0.45 13.1 (13.0 – 13.2)
235 1.51 ± 0.14 13.8 (13.7 – 13.9)
601 1.55 ± 0.11 13.6 (13.5 – 13.6)
1099 1.42 ± 0.073 13.7 (13.6 – 13.7)
1750 1.59 ± 0.23 13.2 (13.1 – 13.3)
2250 1.71 ± 0.22 13.6 (13.4 – 13.7)
2750 3.23 ± 0.24 13.7 (13.7 – 13.8)
25 37 2.72 ± 0.19 13.7 (13.6 – 13.7)
73 1.95 ± 0.16 13.9 (13.8 – 14.0)
90 2.41 ± 0.098 14.0 (13.9 – 14.0)
135 1.92 ± 0.089 13.9 (13.8 – 13.9)
235 1.99 ± 0.082 13.8 (13.7 – 13.8)
285 1.97 ± 0.10 14.2 (14.1 – 14.3)
450 2.62 ± 0.12 13.6 (13.5 – 13.7)
600 3.06 ± 0.19 13.3 (13.2 – 13.3)
800 1.51 ± 0.088 13.7 (13.6 – 13.8)
1100 1.94 ± 0.18 13.2 (13.1 – 13.3)
1500 1.76 ± 0.083 13.8 (13.7 – 13.8)
1750 1.86 ± 0.11 14.0 (13.9 – 14.0)
2000 1.46 ± 0.093 13.6 (13.5 – 13.7)
2250 2.08 ± 0.097 14.3 (14.2 – 14.3)
2500 2.02 ± 0.10 13.9 (13.8 – 13.9)
2750 2.05 ± 0.082 14.0 (14.0 – 14.1)
3000 2.59 ± 0.14 13.7 (13.6 – 13.8)
3200 1.93 ± 0.11 13.8 (13.7 – 13.9)
3500 3.04 ± 0.16 13.3 (13.2 – 13.4)
26 2 1.93 ± 0.14 13.8 (13.8 – 13.9)
29 2.78 ± 0.15 13.6 (13.6 – 13.6)
63 3.08 ± 0.17 13.8 (13.7 – 13.8)
85 2.35 ± 0.081 14.3 (14.2 – 14.3)
135 1.96 ± 0.061 14.1 (14.1 – 14.2)
180 2.44 ± 0.12 14.0 (13.9 – 14.0)
250 2.33 ± 0.11 13.9 (13.9 – 14.0)
349 1.95 ± 0.073 13.9 (13.9 – 14.0)
450 2.50 ± 0.14 13.3 (13.3 – 13.3)
600 2.13 ± 0.073 13.9 (13.9 – 14.0)
800 1.90 ± 0.12 13.7 (13.6 – 13.8)
964 2.34 ± 0.12 13.6 (13.6 – 13.7)
1100 2.13 ± 0.10 13.9 (13.8 – 14.0)
1500 2.38 ± 0.10 13.7 (13.7 – 13.8)
2000 1.87 ± 0.089 14.0 (13.9 – 14.1)
2500 2.12 ± 0.081 14.0 (13.9 – 14.0)
3000 2.65 ± 0.34 13.4 (13.2 – 13.5)

128

3200 2.31 ± 0.13 14.1 (14.0 – 14.2)
3300 2.42 ± 0.086 14.0 (13.9 – 14.0)
27 31 2.86 ± 0.18 13.9 (13.9 – 14.0)
50 2.82 ± 0.15 13.9 (13.8 – 14.0)
85 3.52 ± 0.14 13.6 (13.6 – 13.7)
100 2.52 ± 0.12 14.0 (13.9 – 14.0)
140 2.20 ± 0.17 13.8 (13.7 – 13.8)
185 1.59 ± 0.088 14.2 (14.1 – 14.2)
235 2.50 ± 0.30 13.5 (13.4 – 13.6)
290 1.41 ± 0.10 13.8 (13.7 – 13.8)
420 2.09 ± 0.10 13.8 (13.8 – 13.9)
500 1.87 ± 0.12 13.5 (13.5 – 13.6)
665 1.91 ± 0.081 13.7 (13.7 – 13.8)
965 2.19 ± 0.066 14.0 (14.0 – 14.1)
1100 1.92 ± 0.095 13.7 (13.6 – 13.7)
1250 1.49 ± 0.083 13.9 (13.8 – 14.0)
1500 1.57 ± 0.13 13.5 (13.3 – 13.5)
1750 2.18 ± 0.11 13.6 (13.5 – 13.6)
2000 1.31 ± 0.072 13.8 (13.7 – 13.9)
2250 1.74 ± 0.11 14.3 (14.2 – 14.4)
2500 1.99 ± 0.11 14.1 (14.0 – 14.2)
2700 1.87 ± 0.096 13.8 (13.8 –13.9)
3000 2.03 ± 0.086 13.7 (13.7 – 13.8)
3281 1.96 ± 0.097 13.8 (13.8 – 13.9)
3323 2.42 ± 0.10 13.8 (13.7 – 13.8)
28 2 3.16 ± 0.73 13.5 (13.2 – 13.7)
28 2.26 ± 0.32 13.8 (13.6 – 14.0)
50 3.40 ± 0.20 14.1 (14.0 – 14.2)
89 2.26 ± 0.15 14.1 (14.0 – 14.2)
135 2.08 ± 0.27 13.2 (13.1 – 13.4)
185 2.23 ± 0.066 14.1 (14.0 – 14.2)
285 2.02 ± 0.19 13.4 (13.3 – 13.6)
360 2.38 ± 0.14 13.8 (13.7 – 13.9)
665 2.23 ± 0.16 13.9 (13.9 – 14.0)
965 2.91 ± 0.36 13.1 (13.0 – 13.3)
1100 2.04 ± 0.095 13.6 (13.5 – 13.6)
1200 2.41 ± 0.20 13.4 (13.3 – 13.6)
1400 1.73 ± 0.049 14.1 (14.1 – 14.2)
1800 1.51 ± 0.26 14.3 (14.0 – 14.6)
1900 1.55 ± 0.13 14.4 (14.3 – 14.5)
2100 1.74 ± 0.09 14.1 (14.1 – 14.2)
2300 1.80 ± 0.055 14.0 (13.8 – 14.0)
2600 2.28 ± 0.14 13.9 (13.8 – 14.0)
2905 2.69 ± 0.31 13.7 (13.5 – 14.0)
3005 3.37 ± 0.27 13.8 (13.6 – 14.0)

129
Table 2. Speciation data for Sta. 6 – 22 using 1 µmol L
-1

SA detection window. Note that the
log K interval is asymmetric.

Station Depth (m) L (nmol L
-1
) Log K
6
3600 5.22 ± 0.49 12.1 (12.1 – 12.2)
8 4900 3.68 ± 0.36 12.4 (12.3 – 12.5)
14 4225 3.01 ± 0.17 12.9 (12.8 – 13.0)
16 3224 2.25 ± 0.12 13.6 (13.5 – 13.7)

3589 2.93 ± 0.15 13.8 (13.7 – 13.9)
18 4300 3.30 ± 0.12 13.9 (13.8 – 14.0)
20 5800 3.06 ± 0.14 13.1 (13.0 – 13.2)
22 4985 2.86 ± 0.21 13.2 (13.0 – 13.4)

130
Figure Legends
Fig. 1. Map of the combined cruise tracks in the North Atlantic Ocean that were sampled
aboard the R/V Knorr in 2010 and 2011.

Fig. 2. Oceanographic sections of (A) temperature (in °C), (B) salinity, and (C) dissolved
oxygen concentrations (in µmol L
-1
).

Fig. 3. Upper 200 m oceanographic section of fluorescence.

Fig. 4. Temperature-salinity diagram for all stations. Each colored T-S line corresponds to a
different station.

Fig. 5. Oceanographic section of sample pH values measured prior to performing
complexiometric titrations.

Fig. 6. Oceanographic sections of (A) NO
3
-
, (B) PO
4
3-
, and (C) SiO
4
4-
concentrations (in µmol
L
-1
).

Fig. 7. Upper 250 m oceanographic section of NO
2
-
concentrations (in µmol L
-1
) at Sta. 25 – 28.

Fig. 8. Oceanographic sections of (A) dissolved Cu (in nmol L
-1
), (B) Cu
2+
(in fmol L
-1
), and
(C) L concentrations (in nmol L
-1
).

131
Fig. 9. Dissolved Cu concentrations (in nmol L
-1
) versus (A) NO
3
-
, (B) PO
4
3-
, and (C) SiO
4
4-

concentrations (in µmol L
-1
). The regression lines are all 1
st
order, linear.

Fig. 10. Upper 50 m dissolved Cu concentrations (in nmol L
-1
) vs. salinity values for Sta. 1 –
10. Regression line is 1
st
order, linear.

Fig. 11. Oceanographic section of neutral density (γ
n
) boundary lines (in kg m
3-
) for values
between 27.8 and 28.2 with water masses indicated. AABW is Antarctic Bottom Water, ISOW
is Iceland-Scotland Overflow Water, DSOW is Denmark Straight Overflow Water, CLSW is
Central Labrador Seawater, and ULSW is Upper Labrador Seawater.

Fig. 12. Cu
2+
concentrations (in fmol L
-1
) versus (A) NO
3
-
, (B) PO
4
3-
, and (C) SiO
4
4-

concentrations (in µmol L
-1
). The regression lines are all 1
st
order, linear.

Fig. 13. Oceanographic section for excess L concentrations (in nmol L
-1
).

Fig. 14. Oceanographic section of log K values.

132

Fig. 1.

133

Fig. 2.

134

Fig. 3.

135

Fig. 4.

136

Fig. 5.

137

Fig. 6.

138

Fig. 7.

139

Fig. 8.

140

Fig. 9.

141

Fig. 10.

142

Fig. 11.

143

Fig. 12.

144

Fig. 13.

145

Fig. 14.

146
Appendix

Depth profiles of log [Cu
2+
] (closed circles), total dissolved Cu (closed inverted triangles) and
organic ligands (L) (open squares) for Sta. 1 – 28. Error bars for log [Cu
2+
] and total dissolved
[Cu] represent error propagation from the calculation of [L] and K and standard deviation values
(n=3), respectively. Transmittance profiles are shown for the stations at which the presence of
BNLs (Sta. 6, 8, 10 and 28) or a hydrothermal plume (Sta. 16) was noted.

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164
Chapter 5
Synthesis and Conclusions

Overview
Following the early, seminal breakthroughs of the 1970s and 1980s, the history of discoveries
in trace metal chemistry has largely been one of slow, gradual progress interspersed with periods
of punctuated leaps. Many of these later findings can be attributed as much to the
unconventional thinking of pioneering researchers like John Martin (Martin and Fitzwater 1988)
as to the drastic improvements made in our analytical toolsets and clean sampling techniques. In
recent years, one could argue that significant discoveries made in this field have often gone hand
in hand with major advances made in molecular biology and, increasingly, genomics. While our
core understanding of the physicochemical processes that shape oceanic trace metal distributions
has remained mostly intact since the 1980s, our knowledge of the biogeochemical processes that
also mediate these cycles has grown by leaps and bounds.
Most of the findings presented in this thesis would not have been possible without the many
contributions made by both marine geochemists and microbiologists to our broader
understanding of Cu uptake and toxicity, metalloenzymes as a whole, and the nitrogen cycle.
The very recent discovery and isolation of the ammonia-oxidizing archaeon Nitrosopumilus
maritimus SCM1 (Könneke et al. 2005) and subsequent physiological studies identifying its
dependence, and that of other closely related AOA, on Cu-based proteins (Hallam et al. 2006;
Walker et al. 2010; Blainey et al. 2011) underline the necessity of combining cutting-edge
biology and chemistry to continue pushing the fields forward. In the future, further advances in
analytical instrumentation and global survey programs like GEOTRACES will help make the

165
routine collection of high-resolution trace element datasets—whose scarcity up to the present has
been a severe limitation—simple and systematic (Cutter and Bruland 2012). Available
GEOTRACES datasets have already deepened our understanding of many key trace elements,
but their real value will emerge from the process and modeling studies that will use them as a
jumping-off point for developing and testing novel hypotheses about these element cycles and
other large-scale biogeochemical phenomena.

Specific contributions made by this thesis
Is it time to reconsider the importance of Cu limitation?—The remarkable uniformity of Cu
2+

concentration ranges across distinct biogeochemical environments and through the water column
is one of the major findings of this thesis. With few exceptions, Cu
2+
levels were low and ranged
from 10
-15
mol L
-1
to 10
-13
mol L
-1
at all of the study sites, reflecting the combined effects of
scavenging by particles and organic complexation—but particularly the latter. Even in Hood
Canal, an estuarine system with elevated dissolved Cu concentrations, Cu
2+
values rarely
exceeded the upper bound of that range. The lowest Cu
2+
values were consistently observed in
the upper water column where biological activity was highest while the most elevated values
were observed in deeper waters or at depths in which the dissolved Cu concentration exceeded
the organic ligand concentration.
In the ETSP and Hood Canal, local Cu
2+
minima often coincided with the depths of the
chlorophyll or primary nitrite maxima although the correlation was much stronger for the former
in the waters overlying the region’s OMZ. Ammonia oxidation rates in Hood Canal were also
lowest in the mixed layer, but it is not yet clear whether that can be attributed to Cu limitation of
ammonia oxidation by AOA, which have been shown to dominate nitrification in this system

166
(Horak et al. 2013), or photoinhibition. Only one PNM was observed off the Mauritanian coast
in the North Atlantic Ocean, but it was also the depth around which the lowest Cu
2+

concentrations of the transect were measured. As at the other study sites, the lowest Cu
2+
values
across the North Atlantic transect were found near the chlorophyll maximum. Cumulatively, the
results from chapters 2 through 4 indicate that [Cu
2+
] is effectively buffered to a narrow range by
a strong class of ligands so as to maintain concentrations that are neither “too low” nor “too
high,” but “just right” for most marine microbes to thrive (Bruland and Lohan 2003). This holds
particularly true for the euphotic zone, where concentrations often span only an order of
magnitude (10
-15
mol L
-1
to 10
-14
mol L
-1
) because of higher biological demand.
Past investigations of Cu speciation have focused almost exclusively on the metal’s toxicity to
a wide array of phytoplankton (Brand et al. 1986; Moffett and Brand 1996; Moffett et al. 1997;
Sunda and Huntsman 1998). The revelations that [Cu
2+
] is tightly complexed by strong organic
ligands from surface to seafloor and that a heretofore understudied but ubiquitous class of
archaea, AOA (Francis et al. 2005; Könneke et al. 2005), is heavily reliant on the metal should
prompt a reconsideration of the relevance of Cu limitation in all aquatic systems. Indeed, the
recent discovery that SCM1 can become Cu-limited at Cu
2+
concentrations below ~ 2 × 10
-13
mol
L
-1
(Amin et al. 2013) implies that AOA may be Cu-limited in many parts of the ocean and that
their collective contributions to the marine nitrogen cycle may be significantly diminished as a
result. Reduced upper water column Cu
2+
concentrations are also known to exacerbate Fe
limitation in oceanic diatoms such as Thalassiosira oceanica by preventing the upregulation of a
Cu-dependent high affinity Fe-uptake system (Peers et al. 2005; Maldonado et al. 2006). Yet,
other than these two examples, documented instances of Cu limitation are still uncommon. The
pairing of increasingly sophisticated genomic tools and large speciation datasets like the ones

167
presented in this thesis should not only help improve our understanding of Cu limitation at a
genetic level but also enable the discovery of new examples.
Dissolved Cu on continental margins and in low O
2
waters—Unlike Cu
2+
, dissolved Cu depth
profiles showed considerable variability from one study site to the next while still mostly
conforming to the general features (e.g., slight surface minimum due to biological uptake,
gradual, linear increase with depth, etc.) identified by previous researchers (Boyle et al. 1977,
1981; Bruland 1980; Bruland and Franks 1983). The notable exception to this trend was Hood
Canal, where dissolved Cu concentrations measured during 4 separate cruises over 2 years rarely
varied by more than 25%, presumably because of the stabilizing effect of organic complexation
and a lack of major sources or sinks.
Open ocean dissolved Cu concentrations in the North Atlantic were similar to previously
published values, but the concentrations measured in the waters on and off the Peruvian
continental shelf were among the lowest ever recorded. Dissolved Cu concentrations off the
Mauritanian shelf, while not as low, were among the lowest values measured in the North
Atlantic. The two areas were also highly productive and characterized by oxygen-depleted
waters, suggesting that similar processes may be controlling the distribution of dissolved Cu in
both systems. Recently, Biller and Bruland (2013) noticed a similar drawdown of dissolved Cu
in the waters overlying California’s continental margin, another narrow eastern boundary shelf
featuring hypoxic waters and high productivity levels. They attributed this to denitrifier activity
and removal of dissolved Cu by non-biogenic particles, an observation previously also made by
Sunda and Huntsman (1995) and Bruland (1980). By contrast, dissolved Cu concentrations in
the productive waters overlying the western boundary continental margin off North America
were typically elevated, particularly at the surface where low-salinity Cu-enriched coastal waters

168
mixed with higher-salinity open ocean waters. These divergent results indicate that dissolved
[Cu] exhibits greater variability in dynamic systems that are buffeted by a range of dynamic
biogeochemical processes.
One of the more interesting findings of this thesis is that dissolved Cu concentrations are
unexpectedly enriched within the SNM of the Peruvian OMZ compared to non-OMZ waters
around the same depth range. Though this enrichment likely reflects inputs from organic matter
remineralization and shelf sediments along isopycnal lines, it goes against expectations that
dissolved Cu concentrations should be low in anoxic regimes due to the formation of sulfide
precipitates (Saito et al. 2003). There is new evidence to show that sulfate reduction may be a
more important process in OMZ waters than once believed (Canfield et al. 2010) so it is possible
that the complexes formed by dissolved Cu and sulfide actually increase, rather than diminish,
the solubility of dissolved Cu under these circumstances. A recent investigation (J. Vedamati
and J. W. Moffett unpubl.) into the distribution of dissolved Cu in the Arabian Sea OMZ yielded
results that were more in line with the predictions made by Saito et al. (2003). They observed
heavy depletion of dissolved [Cu] within the SNM, which is consistent with the view that
dissolved Cu would be scavenged out of the water column through the formation of insoluble
Cu-sulfide precipitates and/or that it would drawn down by denitrifiers. While not an OMZ, the
same scavenging process appeared to be at play within the hydrothermal plume waters of the
TAG vent field, where Fe oxyhydroxides and sulfides are both expected to be abundant (Trefry
et al. 1985; Trocine and Trefry 1988; German et al. 1991). With some predicting an expansion
of OMZs in tropical waters under the influence of climate change (Stramma et al. 2008), it is
now more important than ever that we devote more attention to studying the distribution and
speciation of redox-active metals like Cu in these systems.

169
Anticipating the future: Where to go from here
One of the major challenges to modern oceanography is understanding how, and to what
extent, climate change and ocean acidification specifically will change trace element and
biogeochemical cycles over the coming decades. Rising temperatures and falling seawater pH
will significantly alter the distribution of micronutrients like Fe, Zn and Cu by modifying their
inorganic and organic speciation and by accelerating the rates of chemical reactions (Hoffman et
al. 2012). These changes will have profound influences on the composition of marine microbial
communities throughout the water column and in turn on oceanic biogeochemical cycles as a
whole. Constraining these effects will therefore be crucial for researchers to accurately model
and predict what the ocean of the future will look like.
To that end, Millero et al. (2009) published an article in which they assessed the likely
influence of lower seawater pH on the inorganic and organic speciation of several trace metals,
including Cu. The continuing drop in carbonate (CO
3
2-
) concentrations brought on by decreasing
seawater pH will increase the free Cu
2+
fraction, with potentially mixed results for different
members of the microbial community. For instance, while AOA—which may be potentially Cu-
constrained across vast regions of the ocean—could benefit from an influx of Cu
2+
, other
vulnerable species like Prochlorococcus would suffer. While the potential long-term effects of
ocean acidification on organic speciation are not yet as clear, some studies have already
demonstrated that lower seawater pH decreases organic complexation of Cu in both saltwater
and freshwater systems (Averyt et al. 2004; Louis et al. 2009). The results of these studies are
instructive, but any viable long-term projections will also need to consider how varying aerosol
fluxes, melting sea ice, and changes in land use affect these trace elemental cycles. A warming
ocean is also expected to accelerate the expansion of OMZs in tropical and sub-tropical waters

170
(Stramma et al. 2008), with important consequences for marine life and oceanic nutrient cycles
like the nitrogen cycle. Larger OMZs could mean higher rates of fixed nitrogen loss from
denitrification and anammox (Gruber 2011), and the reduced availability of fixed nitrogen could
have significant repercussions for a variety of microorganisms. The general deoxygenation of
seawater and a changing oceanic nitrogen cycle would directly affect the speciation and
distribution of dissolved Cu in many areas of the ocean in ways that were previously explored in
this chapter.
While CLE-ACSV remains an indispensable tool for measuring Cu
2+
concentrations and for
probing the complexing strength and concentrations of organic ligands, it is clear that it suffers
from several limitations. One is that large speciation datasets are still frustratingly rare even
though the method was initially conceived almost 30 years ago (Van den Berg 1984).
Complexiometric titrations are also slow and time-consuming, and different research groups
often use different instrument settings, reagent concentrations and modeling/statistical methods
to calculate the parameters of interest. A recent intercomparison of CLE-ACSV techniques
(Buck et al. 2012) determined that different groups were able to obtain the same consensus
values but also noted that there were still problems with existing practices that needed to be
addressed for the results of international collaborations such as GEOTRACES to have value.
Fortunately, efforts are now underway to better coordinate these measurements by conducting
more frequent intercomparison exercises and adopting a more unified set of practices.
Automating titrations is the next logical step and the only way to ensure that this technique keeps
up with the analytic demands of programs like GEOTRACES, and there has now been some
progress in that area (S. Sander unpubl.). Novel methods such as the one described in Boiteau et
al. (2013) will help address one of the other major complaints about CLE-ACSV: the fact that it

171
reveals nothing about the structure of the ligands it identifies. The ability to determine ligand
structures could provide valuable insights into the composition of the heretofore ambiguous-
sounding “ligand pool” at any given depth and enable the characterization of its various
constituents. The availability of ever-larger datasets and new analytical tools will allow future
workers to continue disentangling the intricacies of the oceanic Cu cycle and predict how it will
change in a warming, acidic ocean.

Bibliography
Amin, S. A., J. W. Moffett, W. Martens-Habbena, J. E. Jacquot, Y. Han, A. Devol, A. E. Ingalls,
D. A. Stahl, and E. V. Armbrust. 2013. Copper requirements of the ammonia-oxidizing
archaeon Nitrosopumilus maritimus SCM1 and implications for nitrification in the marine
environment. Limnol. Oceanogr. In press.
Averyt, K. B., J. P. Kim, and K. A. Hunter. 2004. Effect of pH on measurement of strong copper
binding ligands in lakes. Limnol. Oceangr. 49(1): 20–27, doi:10.4319/lo.2004.1.20
Biller, D. V., and K. W. Bruland. 2013. Sources and distributions of Mn, Fe, Co, Ni, Cu, Zn, and
Cd relative to macronutrients along the central California coast during the spring and
summer upwelling season. Mar. Chem. 155: 50–70, doi:10.1016/j.marchem.2013.06.003
Blainey, P. C., A. C. Mosier, A. Potanina, C. A. Francis, and S. R. Quake. 2011. Genome of a
low-salinity ammonia-oxidizing archaeon determined by single-cell and metagenomic
analysis. PLoS ONE 6(2): e16626, doi:10.1371/journal.pone.0016626
Boiteau, R. M., J. N. Fitzsimmons, D. J. Repeta, and E. A. Boyle. 2013. Detection of iron ligands
in seawater and marine cyanobacteria cultures by high-performance liquid

172
chromatography–inductively coupled plasma-mass spectrometry. Anal. Chem. 85: 4357–
4362, doi:10.1021/ac3034568
Boyle, E. A., F. Sclater, and J. Edmond. 1977. The distribution of dissolved copper in the
Pacific. Earth Planet. Sci. Lett. 37: 38–54, doi:10.1016/0012-821X(77)90144-3
Boyle, E. A., S. S. Huested, and S. P. Jones. 1981. On the distribution of copper, nickel, and
cadmium in the surface waters of the North Atlantic and North Pacific Ocean. J. Geophys.
Res. 86: 8048–8066, doi:10.1029/JC086iC09p08048
Brand, L. E., W. G. Sunda, and R. R. L. Guillard. 1986. Reduction of marine phytoplankton
reproduction rates by copper and cadmium. J. Exp. Mar. Biol. Ecol. 96: 225–250,
doi:10.1016/0022-0981(86)90205-4
Bruland, K. W. 1980. Oceanographic distributions of cadmium, zinc, nickel, and copper in the
North Pacific. Earth and Planet. Sci. Lett. 47: 176–198, doi:10.1016/0012-821X(80)90035-
7
Bruland, K. W., and R. P. Franks. 1983. Mn, Ni, Cu, Zn and Cd in the western North Atlantic, p.
395–414. In C. S. Wong, E. Boyle, K. W. Bruland, J. D. Burton and E. D. Goldberg [eds.],
Trace metals in sea water. Academic.
Bruland, K. W., and M. C. Lohan. 2003. Controls of trace metals in seawater, p. 23–47. In H.
Elderfield [ed.], The oceans and geochemistry. Academic.
Buck, K. N., J. Moffett, K. A. Barbeau, R. M. Bundy, Y. Kondo, and J. Wu. 2012. The organic
complexation of iron and copper: an intercomparison of competitive ligand exchange-
adsorptive cathodic stripping voltammetry (CLE-ACSV) techniques. Limnol. Oceanogr.:
Methods 10: 496–515, doi:10.4319/lom.2012.10.496

173
Canfield, D. E., F. J. Stewart, B. Thamdrup, L. De Brabandere, T. Dalsgaard, E. F. Delong, N. P.
Revsbech, and O. Ulloa. 2010. A cryptic sulfur cycle in oxygen-minimum-zone waters off
the Chilean coast. Science 330: 1375–1378, doi:10.1126/science.1196889
Cutter, G. A., and K. W. Bruland. 2012. Rapid and noncontaminating sampling system for trace
elements in global ocean surveys. Limnol. Oceanogr.: Methods 10: 425–436,
doi:10.4319/lom.2012.10.425
Francis, C. A., K. J. Roberts, J. M. Beman, A. E. Santoro, and B. B. Oakley. 2005. Ubiquity and
diversity of ammonia-oxidizing archaea in water columns and sediments of the ocean.
Proc. Nat. Acad. Sci. U.S.A. 102: 14683–14688, doi:10.1073/pnas.0506625102
German, C. R., A. C. Campbell, and J. M. Edmond. 1991. Hydrothermal scavenging at the Mid-
Atlantic Ridge: Modification of trace element dissolved fluxes. Earth and Planet. Sci. Lett.
107: 101–114, doi:10.1016/0012-821X(91)90047-L
Gruber, N. 2011. Warming up, turning sour, losing breath: ocean biogeochemistry under global
change. Phil. Trans. R. Soc. A 369: 1980–1996, doi:10.1098/rsta.2011.0003
Hallam, S. J., K. T. Konstantinidis, N. Putnam, C. Schleper, Y. Watanabe, J. Sugahara, C.
Preston, J. R. de La Torre, P. M. Richardson, and E. F. Delong. 2006. Genomic analysis of
the uncultivated marine crenarchaeote Cenarchaeum symbiosum. Proc. Natl. Acad. Sci.
U.S.A. 103(48): 18296–18301, doi:10.1073/pnas.0608549103
Hoffmann, L. J., E. Breitbarth, P. W. Boyd, and K. A. Hunter. 2012. Influence of ocean warming
and acidification on trace metal biogeochemistry. Mar. Ecol. Prog. Ser. 470: 191–205,
doi:10.3354/meps10082

174
Horak, R. E. A., W. Qin, A. J. Schauer, E. V. Armbrust, A. E. Ingalls, J. W. Moffett, D. A. Stahl,
and A. H. Devol. 2013. Ammonia oxidation kinetics and temperature sensitivity of a
natural community dominated by Archaea. ISME J., doi:10.1038/ismej.2013.75
Könneke, M., A. E. Bernhard, J. R. de La Torre, C. B. Walker, J. B. Waterbury, and D. A. Stahl.
2005. Isolation of an autotrophic ammonia-oxidizing marine archaeon. Nature 437: 543–
546, doi:10.1038/nature03911
Louis, Y., C. Garnier, V. Lenoble, D. Omanovic, S. Mounier, and I. Pizeta. 2009.
Characterisation and modeling of marine dissolved organic matter interactions with major
and trace cations. Mar. Environ. Res. 67: 100–107, doi:10.1016/j.marenvres.2008.12.002
Maldonado, M. T., A. E. Allen, J. S. Chong, K. Lin, D. Leus, N. Karpenko, and S. L. Harris.
2006. Copper-dependent iron transport in coastal and oceanic diatoms. Limnol. Oceanogr.
51(4): 1729–1743, doi:10.4319/lo.2006.51.4.1729
Martin, J. H., and S. E. Fitzwater. 1988. Iron deficiency limits phytoplankton growth in the
north-east Pacific subarctic. Nature 331: 341–343, doi:10.1038/331341a0
Millero, F. J., R. Woosley, B. Ditrolio, and J. Waters. 2009. Effect of ocean acidification on the
speciation of metals. Oceanography 22: 72–85, doi:
Moffett, J. W., and L. E. Brand. 1996. Production of strong, extracellular Cu chelators by marine
cyanobacteria in response to Cu stress. Limnol. Oceanogr. 41(3): 388–395,
doi:10.4319/lo.1996.41.3.388
Moffett, J. W., L. E. Brand, P. L. Croot, and K. A. Barbeau. 1997. Cu speciation and
cyanobacterial distribution in harbors subject to anthropogenic Cu inputs. Limnol.
Oceanogr. 42(5): 789–799, doi:10.4319/lo.1997.42.5.789

175
Peers, G., S. A. Quesnel, and N. M. Price. 2005. Copper requirements for iron acquisition and
growth of coastal and oceanic diatoms. Limnol. Oceanogr. 50(4): 1149–1158, doi:
10.4319/lo.2005.50.4.1149
Saito, M. A., D. M. Sigman, and F. M. M. Morel. 2003. The bioinorganic chemistry of the
ancient ocean: The co-evolution of cyanobacterial metal requirements and biogeochemical
cycles at the Archean-Proterozoic boundary? Inorg. Chim. Acta 356: 308–318,
doi:10.1016/S0020-1693(03)00442-0
Stramma, L., G. C. Johnson, J. Sprintall, and V. Mohrholz. 2008. Expanding oxygen-minimum
zones in the tropical oceans. Science 320: 655–658, doi:10.1126/science.1153847
Sunda, W. G., and R. R. L. Guillard. 1976. The relationship between cupric ion activity and the
toxicity of copper to phytoplankton. J. Mar. Res. 34: 511–529.
Sunda, W. G., and S. A. Huntsman. 1995. Regulation of copper concentration in the oceanic
nutricline by phytoplankton uptake and regeneration cycles. Limnol. Oceanogr. 40(1):
132–137, doi:10.4319/lo.1995.40.1.132
Sunda, W. G., and S. A. Huntsman. 1998. Interactions among Cu
2+
, Zn
2+
, and Mn
2+
in
controlling cellular Mn, Zn, and growth rate in the coastal alga Chlamydomonas. Limnol.
Oceanogr. 43(6): 1055–1064, doi:10.4319/lo.1998.43.6.1055
Trefry, J. H., R. P. Trocine, G. P. Klinkhammer, and P. A. Rona. 1985. Iron and copper
enrichment of suspended particles in dispersed hydrothermal plumes along the mid‐
Atlantic Ridge. Geophys. Res. Lett. 12: 506–509, doi:10.1029/GL012i008p00506
Trocine, R. P., and J. H. Trefry. 1988. Distribution and chemistry of suspended particles from an
active hydrothermal vent site on the Mid-Atlantic Ridge at 26 N. Earth and Planet. Sci.
Lett. 88: 1–15, doi:10.1016/0012-821X(88)90041-6

176
Turner, D. R., M. Whitfield, and A. G. Dickson. 1981. The equilibrium speciation of dissolved
components in freshwater and sea water at 25°C and 1 atm pressure. Geochim.
Cosmochim. Acta 45: 855–881, doi:10.1016/0016-7037(81)90115-0
Van den Berg, C. M. G. 1984. Determination of the complexing capacity and conditional
stability constants of complexes with copper(II) with natural organic ligands in seawater by
cathodic stripping voltammetry of copper-catechol ions. Mar. Chem. 15: 1–18,
doi:10.1016/0304-4203(84)90035-5
Walker, C. B., J. R. de La Torre, M. Klotz, H. Urakawa, N. Pinel, D. Arp, C. Brochier-Armanet,
P. Chain, P. Chan, A. Gollabgir, J. Hemp, M. Hügler, E. A. Karr, M. Könneke, M. Shin, T.
J. Lawton, T. Lowe, W. Martens-Habbena, L. A. Sayavedra-Soto, D. Lang, S. M. Sievert,
A. C. Rosenzweig, G. Manning, and D. A. Stahl. 2010. Nitrosopumilus maritimus genome
reveals unique mechanisms for nitrification and autotrophy in globally distributed marine
crenarchaea. Proc. Nat. Acad. Sci. U.S.A. 107(19): 8818–8823,
doi:10.1073/pnas.0913533107

Abstract (if available)

Abstract Copper (Cu) is an essential micronutrient that functions as a cofactor in many important enzymatically-mediated pathways including denitrification, methane oxidation and ammonia oxidation. Yet it can also be a potent toxicant, inhibiting phytoplankton reproduction and growth rates, at picomolar-level concentrations. In natural waters, over 99.9% of dissolved Cu is complexed by strong organic ligands of biological origin. As a result, concentrations of the bioavailable fraction, Cu²⁺, are often over a thousand-fold lower (~10⁻¹⁴ – 10⁻¹³ mol L⁻¹) than dissolved Cu concentrations. The two main controls on the distribution of dissolved Cu, and Cu²⁺ by extension, are organic complexation and scavenging by particles and biological processes. This thesis examines the distribution and speciation of Cu in the North Atlantic and eastern tropical South Pacific (ETSP) Oceans and in Hood Canal, an estuary in Puget Sound, WA, to better understand how regimes with very distinct biogeochemistries influence Cu cycling. It also explores the relationship between Cu bioavailability and nitrogen cycle processes in the ETSP and Hood Canal, two regimes with high nitrification activity. ❧ Cu speciation was characterized using competitive ligand exchange-adsorptive cathodic stripping voltammetry (CLE-ACSV), and total dissolved Cu concentrations were measured using isotope dilution inductively coupled plasma mass spectrometry (ID-ICP-MS). Overall, results indicate that Cu²⁺ concentrations are maintained at uniformly low levels throughout the water column by strong organic ligands and scavenging, regardless of location, but particularly in areas with high nitrification activity. These processes buffer Cu²⁺ concentrations to a narrow, "steady-state" range that rarely varies by more than two orders of magnitude. Dissolved Cu concentrations were significantly more variable but generally exhibited the profile characteristics identified by other researchers. ❧ In the ETSP, Cu²⁺ levels typically reached their lowest values near the chlorophyll and primary nitrite maxima layers, indicating heavy drawdown by ammonia-oxidizing archaea (AOA) and nitrate-reducing diatoms. Dissolved Cu concentrations in offshore waters, away from the Peruvian coastline, were some of the lowest measured values to date. In the Hood Canal, dissolved Cu levels, though much higher than in open ocean waters, were almost unchanged from over 30 years ago, suggesting that anthropogenic inputs have not greatly affected this system. Cu²⁺ concentrations in the upper water column often approached an experimentally-determined threshold below which ammonia oxidation by AOA may become Cu-limited. Although photoinhibition seems like the more plausible control on nitrification activity, it seems likely that Cu bioavailability also plays a role. Finally, in the North Atlantic Cu²⁺ concentrations remained low through the water column and across the transect, rarely exceeding 10⁻¹³ mol L⁻¹. The lowest concentrations were measured in the nutrient-rich upwelling regime off the Mauritanian coastline while the highest concentrations were measured at depth within the subtropical central gyre region. Dissolved Cu profiles exhibited many unique features and showed clear geographic trends while still mostly conforming to the general properties identified by other researchers in the Atlantic and Pacific Oceans.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

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

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

PDF

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

PDF

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

PDF

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

PDF

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

PDF

Distribution and impact of algal blooms leading to domoic acid events in southern California

PDF

Biological nitrogen fixation associated with living and decomposing macroalgae

PDF

Thermal acclimation and adaptation of key phytoplankton groups and interactions with other global change variables

PDF

Temporal variability of marine archaea across the water column at SPOT

PDF

Spatial and temporal dynamics of marine microbial communities and their diazotrophs in the Southern California Bight

PDF

The molecular adaptation of Trichodesmium to long-term CO₂-selection under multiple nutrient limitation regimes

PDF

Effects of global change on the physiology and biogeochemistry of the N₂-fixing cyanobacteria Trichodesmium erythraeum and Crocosphaera watsonii

PDF

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

PDF

The accuracy of core incubations to determine benthic fluxes of 222-radon, 224-radium and 228-radium

PDF

Characterizing protistan diversity and quantifying protistan grazing in the North Pacific Subtropical Gyre

PDF

The marine, neutrophilic, and chemolithoautotrophic iron-oxidizing bacteria: insights into the physiology of Zetaproteobacteria and the discovery of novel iron-oxidizing Gammaproteobacteria

PDF

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

PDF

Oxygen uptake rates in the thermocline of coastal waters: assessing the role of carbon and inorganic nutrient inputs

Asset Metadata

Creator Jacquot, Jeremy E. (author)

Core Title The distribution and speciation of copper across different biogeochemical regimes

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Marine and Environmental Biology

Publication Date 10/03/2013

Defense Date 08/31/2013

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

Tag ammonia oxidation,archaea,chemical oceanography,Copper,Marine biology,nitrification,nitrogen,OAI-PMH Harvest,trace metals

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Moffett, James W. (committee chair), Berelson, William M. (committee member), Hutchins, David A. (committee member), Sanudo-Wilhelmy, Sergio A. (committee member)

Creator Email jacquot@usc.edu,jjacquot@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-334086

Unique identifier UC11296276

Identifier etd-JacquotJer-2076.pdf (filename),usctheses-c3-334086 (legacy record id)

Legacy Identifier etd-JacquotJer-2076.pdf

Dmrecord 334086

Document Type Dissertation

Format application/pdf (imt)

Rights Jacquot, Jeremy E.

Type texts

Source 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 a...

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