the distribution and behavior of dissolved and …kbruland/people/brown/mtb_dissertation... ·...
TRANSCRIPT
UNIVERSITY OF CALIFORNIA
SANTA CRUZ
THE DISTRIBUTION AND BEHAVIOR OF DISSOLVED ANDPARTICULATE ALUMINUM IN COASTAL WATERS OF THE
NORTHEAST PACIFIC OFF OREGON AND WASHINGTON AND IN THENORTHERN GULF OF ALASKA
A dissertation submitted in partial satisfaction of the requirements for the degree of
DOCTOR OF PHILOSOPHY
in
OCEAN SCIENCES
by
Matthew T. Brown
December 2009
The Dissertation of Matthew T. Brownis approved:
________________________________Professor Kenneth W. Bruland, Chair
________________________________Professor Margaret L. (Peggy) Delaney
________________________________Senior Scientist Kenneth Johnson, Ph.D.
________________________________Professor Andrew M. Moore
___________________________________Tyrus MillerVice Provost and Dean of Graduate Studies
iii
TABLE OF CONTENTS PAGE
Abstract ………………………………………………………………………...…..xiii
Acknowledgements ………………………………………………………………....xv
INTRODUCTION ……..………………………………………………………...…..1
References ………………………………………………………………..…16
CHAPTER 1: AN IMPROVED FLOW INJECTION ANALYSIS METHOD FORTHE DETERMINATION OF DISSOLVED ALUMINUM IN SEAWATER ….…26
Abstract …………………………………………………………………..…27
Introduction ……………………………………………………………..…..27
Materials and procedures ……………………………………………..…….29
Reagents ..………………………………………………………..…..29
Procedure ………………………………………………………..….29
Methods and assessment …………………………………………………....30
Column loading pH ………………………….…………………..….30
Al mass balance ……………………………………..……….….….30
Eluting acid concentration ……………………………………….…31
Column conditioning ………………………………………….……31
Standardization ……………………………………………….…….32
Blanks and detection limits ……………………………………..…..32
Application ……………………………………………………..…...32
Discussion ………………………………………………………………..…34
Comments and recommendations ……………………………………..……34
iv
References ………………………………………………………………..…34
CHAPTER 2: DISSOLVED AND PARTICULATE ALUMINUM IN THECOLUMBIA RIVER AND COASTAL WATERS OF OREGON ANDWASHINGTON: BEHAVIOR IN NEAR-FIELD AND FAR-FIELD PLUMES…..36
Abstract …………………………………………………………………..…37
1. Introduction …………………………………………………………..…..37
2. Methods ………………………………………………………………..…38
2.1. Sample collection and filtration ……………………………..…38
2.2. Analytical methods …………………………………………….38
3. Results ………………………………………………………………..…..39
3.1. Surface transects ………………………………………….……39
3.1.1. Downwelling/northward plume transects ……….…...39
3.1.2. Upwelling/southwest plume transects ….…..……..….41
3.2. River and estuary ……………………………………….……...42
4. Discussion ……………………………………………………………..…42
4.1. Comparison with previous data ……………………….……….42
4.2. Dissolved aluminum and silicic acid within theColumbia River estuary ………………………………….…44
4.3. Downwelling plume formation, plume dispersion,and chemical characteristics ………………………….…….46
4.4. Upwelling plume formation, plume dispersion,and chemical characteristics ………………………….…….47
4.5. Particulate Al in the estuary, near-field,and far-field plumes ……………………………..………….49
v
4.6. Broad scale implications …………………..………………..….49
5. Conclusions ……………………………………………….………..…….50
Acknowledgements …………………………………….…………..……….50
References …………………………………………………………..………51
CHAPTER 3: DISSOLVED ALUMINUM, PARTICULATE ALUMINUM, ANDSILICIC ACID IN NORTHERN GULF OF ALASKA COASTAL WATERS:GLACIAL/RIVERINE INPUTS AND EXTREME REACTIVITY …………....….52
Abstract …………………………………………………………..…..……..53
1. Introduction ……………………………………………………..………..54
2. Methods …………………………………………………………..………58
2.1. Sample collection and filtration ……………………..…………58
2.2. Analytical methods ……………………………………….……61
3. Results ………………………………………………………………..…..64
3.1. Surface transects ………………………………………….……64
3.1.1. Transect 1 ……………………………...………..……64
3.1.2. Transect 3 ……………………….…………..………..66
3.1.3. Transect 4 ………………………………...…..………68
3.1.4. Transect 5 ………………………………..….………..69
3.1.5. Transect 6 ………………………………...……..……71
3.2. Water column dissolved and particulate Al inthe shelf waters ……………………………………….…….71
vi
4. Discussion ………………………………………………………..………74
4.1. Comparison with previous data ……………………….……….74
4.2. Dissolved Al in the low salinity coastal waters ……………..…77
4.3. A contrast in silicic acid in two low salinity plumes ……....…..82
4.4. Surface transects and relationships between totalsuspended solids, particulate Al, and dissolved Al …………84
4.5. Dissolved Al vs. soluble Al ……………………………....……89
5. Conclusions ……………………………………………………..………..91
Acknowledgements ……………………………….……………..………….92
References …………………………………………………………..………94
CONCLUSIONS ……………………………………………………….…………130
vii
LIST OF FIGURES ……………………..……………………………………….Page
INTRODUCTION
Figure 1. Al speciation and solubility as a function of pH based on the solubility of gibbsite in freshwater from Stumm and Morgan (1996) ……………………………………….…… 20
Figure 2. Dissolved Al vertical profiles from the GEOTRACES North Atlantic intercalibration cruise near Bermuda and from the GEOTRACES Pacific speciation cruise at the SAFe station in the North Pacific subtropical gyre ……………………………………………………….…….21
Figure 3. The net scavenging rate constant of Th-234 in surface waters of the Pacific Ocean ………………………………..……22
Figure 4. Surface water dissolvable aluminum in the north Pacific Ocean along a transect between Hawaii and Monterey ……..….23
Figure 5. A schematic of the processes governing dissolved Al in a model ocean ……………………………………….…….24
Figure 6. MADCOW model calibration comparing surface water dissolved Al-derived dust deposition estimates to independent dust deposition estimates ………………….....…25
CHAPTER 1: AN IMPROVED FLOW INJECTION ANALYSIS METHOD FORTHE DETERMINATION OF DISSOLVED ALUMINUM IN SEAWATER
Figure 1. Al-FIA system manifold using in-line preconcentration of seawater onto Toyopearl AF-Chelate 650M iminodiacetate resin ………….………….…………….…..……28
Figure 2. Effect of sample pH on the preconcentration of aluminum Onto the Toyopearl AF-Chelate 650M resin column …..…..…..30
Figure 3. Standard curves generated from duplicate analyses of Al standards analyzed with and without a column- conditioning step using a 1-min sample loading time (2.5 ml sample loaded) …………………………………..….…..32
viii
Figure 4. Time-dependence study of the effect of sample Acidification on aluminum concentrations in 3 D2 Samples collected during the 2004 SAFe cruise ………..………33
Figure 5. Dissolved aluminum concentrations and salinity as a function of longitude from a surface water transect near 46°N across the far-field Columbia River plume …...…….……34
CHAPTER 2: DISSOLVED AND PARTICULATE ALUMINUM IN THECOLUMBIA RIVER AND COASTAL WATERS OF OREGON ANDWASHINGTON: BEHAVIOR IN NEAR-FIELD AND FAR-FIELD PLUMES
Figure 1. The location of surface transects ……………………...…..……39
Figure 2. Data from downwelling near-field plume surface transect 4 off the Oregon coast ……………………….…..…….40
Figure 3. Data from downwelling far-field plume surface transect 1 off the Washington coast ……………………..…...…40
Figure 4. Data from downwelling far-field plume surface transect 6 off the Washington coast ………………………...…..41
Figure 5. Data from upwelling near-field plume surface transect 11 off the Oregon coast ………………….....…….……42
Figure 6. Data from upwelling intermediate-field plume surface transect 12 off the coast of Oregon ………………….…...……..43
Figure 7. Data from upwelling far-field plume surface transect 14 off the Oregon coast ………………………………...….……43
Figure 8. Dissolved Si vs. salinity (top) and dissolved Al vs. salinity (bottom) for the river, estuary, and near-field plume transect 4 ……………………………………………..….45
Figure 9. Property-salinity plots for downwelling plume sampling (river, estuary, near-field T4, far-field T1 and T6) ………..……46
Figure 10. Silicic acid (top) vs. salinity and dissolved Al (bottom) vs. salinity (T4, T1, and T6) ……………………………..……47
ix
Figure 11. Property-salinity plots for upwelling near-field Transect 11, intermediate-field Transect 12, and
far-field Transect 14 ……………………………………...……48
CHAPTER 3: DISSOLVED ALUMINUM, PARTICULATE ALUMINUM, ANDSILICIC ACID IN NORTHERN GULF OF ALASKA COASTAL WATERS:GLACIAL/RIVERINE INPUTS AND EXTREME REACTIVITY
Figure 1. Schematic of the general circulation in the Gulf of Alaska …...103
Figure 2. Coastal transect map for the 2007 Eddy cruise in the northern and central Gulf of Alaska …………………………..105
Figure 3. Temperature and salinity (top), dissolved aluminum and silicic acid (middle), and leachable particulate (LP) and total particulate (TP) aluminum along Transect 1 near Yakutat in the northern Gulf of Alaska …………….…….107
Figure 4. Temperature and salinity (top), dissolved and soluble aluminum and silicic acid (middle), and leachable particulate (LP) and total particulate (TP) aluminum (bottom) along Transect 3 in the Copper River
outflow region ……..………………………………………..…109
Figure 5. Temperature and salinity (top), dissolved aluminum and silicic acid (middle), and leachable particulate (LP) and total particulate (TP) aluminum (bottom) along Transect 4 (GAK line) in the northern Gulf of Alaska ……………….…111
Figure 6. Temperature and salinity (top), dissolved aluminum and silicic acid (middle), and leachable particulate (LP) and total particulate (TP) aluminum (bottom) along Transect 5 in and near Cook Inlet in the northern Gulf of Alaska …………………………………………………………113
Figure 7. Temperature and salinity (top) and dissolved aluminum, leachable particulate (LP) aluminum, and silicic acid (bottom) along Transect 6 starting in close proximity to the Aialak Glacier near Seward, Alaska ………………………115
x
Figure 8. Temperature (A), salinity (B), silicic acid (C), dissolved Al (D), leachable particulate Al (E), and total particulate Al (F) along Transect 2 shelf stations YAK1, YAK2, and YAK 3 across the continental shelf ……………………….117
Figure 9. Temperature (A), salinity (B), silicic acid (C), dissolved Al (D), leachable particulate Al (E), and total particulate Al (F) at vertical profile stations GAK-A, GAK-B, and GAK-C along Transect 4 (GAK line) …………………….……119
Figure 10, Temperature (A), salinity (B), silicic acid (C), dissolved Al (D), leachable particulate Al (E), and total particulate Al (F) at Kodiak source water (KSW) stations 1 (KSW1) and station 2 (KSW2) ……………………………………….…121
Figure 11. Dissolved aluminum vs. salinity (left) and silicic acid vs. salinity (right) along Transect 1 Plume A (black), Transect 3 (grey), and Transect 6 (open triangles) in the northern Gulf of Alaska ………………………………………123
Figure 12. Dissolved aluminum vs. salinity (left) and silicic acid vs. salinity (right) for T1 Plume A and T1 Plume B ……...…..125
Figure 13. Underway ship fluorescence and biogenic silica in surface waters along Transect 1 ………………………………127
Figure 14. Dissolved Al vs. total suspended solid (TSS) concentrations for Transects 1, 3, and 4 ………..……...…...…129
xi
LIST OF TABLES ………………………...……………………………………..Page
CHAPTER 1: AN IMPROVED FLOW INJECTION ANALYSIS METHOD FORTHE DETERMINATION OF DISSOLVED ALUMINUM IN SEAWATER
Table 1. Times of sample loading onto the Toyopearl AF- Chelate 650M resin column for given sample aluminum concentration ranges …………………..………..……30
Table 2. Percent Al recovery from seawater using the Toyopearl AF-Chelate 650M IDA resin as outlined in the Al mass balance experiment …………….………31
Table 3. Results of acid-eluent study to determine minimum concentration of hydrochloric acid needed to efficiently elute aluminum from the IDA resin ……………………………………………………..….31
Table 4. Dissolved Al data from surface (S) water and deep (1000 m; D1 and D2) samples collected from 30°N, 140°W during the 2004 SAFe cruise ………………..……..……32
CHAPTER 2: DISSOLVED AND PARTICULATE ALUMINUM IN THECOLUMBIA RIVER AND COASTAL WATERS OF OREGON ANDWASHINGTON: BEHAVIOR IN NEAR-FIELD AND FAR-FIELD PLUMES
Table 1. Sample data from the three Columbia River estuary stations (E1,E2, and E3 with E1 being closest to the river) and two Columbia River stations (R1 and R2) ………...……..…44
Table 2. Salinity, silicic acid, leachable particulate (LP), total particulate (TP), and percent-leachable aluminum data from downwelling near-field Transect 4, downwelling far-field Transect 1, and downwelling far-field Transect 6 …......50
Table 3. Salinity, silicic acid, leachable particulate (LP) aluminum, total particulate (TP) aluminum, and percent leachable aluminum data from upwelling near-field Transect 11, upwelling intermediate-field Transect 12, and upwelling far-field Transect 14 ………………………..……50
xii
LIST OF TABLES (cont’d) ………..……...……………………………………..Page
CHAPTER 3: DISSOLVED ALUMINUM, PARTICULATE ALUMINUM, ANDSILICIC ACID IN NORTHERN GULF OF ALASKA COASTAL WATERS:GLACIAL/RIVERINE INPUTS AND EXTREME REACTIVITY
Table 1. Dissolved aluminum, leachable particulate (LP) aluminum, and total particulate (TP) aluminum data with corresponding hydrographic data for Transect 1 near Yakutat in the northern Gulf of Alaska …………...……..………99
Table 2. Dissolved aluminum, soluble aluminum, leachable particulate (LP) aluminum, and total particulate (TP) aluminum data with corresponding hydrographic data for Transect 3 near the Copper River outflow in the northern Gulf of Alaska ………………………………………………….100
Table 3. Dissolved aluminum, leachable particulate (LP) aluminum, and total particulate (TP) aluminum data with corresponding hydrographic data for Transect 4 (GAK line) starting near Seward and heading southwest in the waters of the subarctic Alaskan gyre ………………………..…………100
Table 4. Dissolved aluminum, leachable particulate (LP) aluminum, and total particulate (TP) aluminum data with corresponding hydrographic data for Transect 5 starting in lower Cook Inlet .………………………………………..…..101
Table 5. Dissolved aluminum and leachable particulate aluminum With corresponding hydrographic data for Transect 6 starting Just off the Aialik Glacier in Aialak Bay southeast of Seward, AK …………………………………………………….101
ABSTRACT
Matthew T. Brown
THE DISTRIBUTION AND BEHAVIOR OF DISSOLVED ANDPARTICULATE ALUMINUM IN COASTAL WATERS OF THE
NORTHEAST PACIFIC OFF OREGON AND WASHINGTON AND IN THENORTHERN GULF OF ALASKA
The behavior of Al in high-productivity coastal regions dominated by
freshwater inputs of terrestrial material is not well characterized. Investigating the
sources of both dissolved and particulate Al in coastal waters of Oregon and
Washington as well as in the northern Gulf of Alaska and gaining insight into the
behavior of these Al fractions in both surface waters and at depth were the goals of
this dissertation research. To this end, significant modifications to an existing flow-
injection analysis method were carried out that resulted in a portable, shipboard
analysis method that utilized a commercially-available preconcentration resin, had a
lower detection limit, and was easy to use.
The coastal waters of Oregon and Washington are significantly influenced by
freshwater input from the Columbia River plume. Dissolved and particulate Al
concentrations were significantly greater in the river than in the coastal waters that
mixed to form the plume while riverine dissolved Al concentrations were low relative
to other major world rivers. Dissolved Al within the Columbia River estuary showed
a significant removal (~ 60%) at salinities between 0 and 10 with salt-induced
flocculation colloidal Al complexes and enhanced particle scavenging being probable
explanations for Al removal. Dilution of the plumes advecting from near-field to far-
field with lower dissolved Al surface waters as well as particle scavenging along the
flow path appeared to be controlling dissolved Al distributions.
In northern Gulf of Alaska coastal waters, glacial/riverine inputs were
correlated with extremely high dissolved and particulate Al concentrations in low-
salinity coastal plumes with the dissolved Al existing entirely in the soluble (< 0.03
µm) fraction. The endmember Al characteristics of fresh glacial melt was found to be
significantly different than that of riverine input that had interacted with
soil/vegetation. A consistency in Si:Al ratios from different samplings was attributed
to the chemical weathering of biotite. Finally, an extreme decreasing gradient of
dissolved Al in surface waters was observed moving offshelf where some of the
lowest dissolved Al concentrations in the world oceans was observed. The residence
time of dissolved Al in coastal shelf waters was estimated to be ~ 10 days.
xv
ACKNOWLEDGEMENTS
It is my pleasure to acknowledge and give thanks to all the people who made this
dissertation possible and supported me along the way. First, I would like to thank
Ken Bruland for his guidance and patience over the last four years as my dissertation
advisor. He is an amazing scientist and, more importantly, a wonderful human being
and he brought out the best in me. I am also grateful to my dissertation committee
(Peggy Delaney, Ken Johnson, and Andy Moore) for all their support, advice, and
guidance in this process. I also thank Rob Franks (Marine Analytical Labs) and
Geoffrey Smith for their significant help and guidance in dealing with sampling,
instrumentation, and analytical issues. I am especially thankful and grateful to all the
Bruland labmates that have given me guidance, laughter, and support over the past
few years: Maeve Lohan, Ana Aguilar-Islas, Kristen Buck, Carolyn Berger, Bettina
Sohst, Sherry Lippiatt, and Dondra Biller. Thank you to all the amazing friends and
people who have kept me sane and balanced over this time of my life, my yoga
teacher Mark Stephens, my various meditation teachers and the sangha at Vipassana
Santa Cruz—you don’t know what your unspoken support has meant. Lastly, thank
you to my family—Mom, Dad, Megan, and Molly. Thank you for encouraging me
when I wanted to quit, for believing in me when I was ready to give up. I love you all
eternally.
xvi
The text of this dissertation includes reprints of the following previously published
papers, and one submitted manuscript:
1. Brown, M.T, Bruland, K.W., 2008. An improved flow-injection analysismethod for the determination of dissolved aluminum in seawater. Limnologyand Oceanography: Methods, 6, 87-95.
2. Brown, M.T., Bruland, K.W., 2009. Dissolved and particulate aluminum inthe Columbia River and coastal waters of Oregon and Washington: behaviorin near-field and far-field plumes. Estuarine, Coastal and Shelf Science, 84,171-185.
3. Brown, M.T., Lippiatt, S.M., Bruland, K.W., 2009. Dissolved aluminum,particulate aluminum, and silicic acid in northern Gulf of Alaska coastalwaters: glacial/riverine inputs and extreme reactivity. Submitted andin review, Marine Chemistry.
Kenneth W. Bruland (a co-author listed in these publications) directed and supervised
the research which forms the basis for the dissertation.
xvii
My contribution to Chapters 1, 2, and 3 was as follows:
Chapter 1: a) the assembly of instrumentation used, including research into the
individual parts that would make the working whole. b) Conducted initial testing of
instrument and preconcentration resin. c) Conducted method development and system
optimization studies discussed in the “Assessment” section of the publication. d)
conducted the analysis of all SAFe samples and Columbia River samples presented
herein. e) conducted the statistical analysis of all data presented in the publication.
Chapter 2: a) Analysis of all dissolved aluminum data presented in the publication. b)
analysis of all the leachable and particulate data presented in the publication. c)
analysis of existing hydrographic data and data interpretation
Chapter 3: a) Analysis of all the soluble and dissolved Al data presented in the
publication. b) analysis of existing hydrographic data and data interpretation.
1
INTRODUCTION
Aluminum (Al) is the third most abundant element in the Earth’s crust (8.2%
by weight; Taylor, 1964) yet the oceanic levels of dissolved Al are found at trace
concentrations, less than a few nM in the Pacific Ocean. Early work showed that in
fresh waters the solubility of Al is dependent on pH with a minimum solubility
occurring at pH 6.7 (Reesman et al., 1969). Based on data by Baes and Mesmer
(1976) on the solubility of gibbsite (Al(OH)3 solid) in freshwater , Stumm and
Morgan (1996) show that the minimum Al solubility at 25ºC occurs at pH 6.5-6.7
(Figure 1). While Al can exist as organic complexes in natural waters (Perdue et al.,
1976), Al speciation in the oceans is thought to be dominated largely by the dissolved
inorganic hydrolysis species Al(OH)2+, Al(OH)3
0, and Al(OH)4-. In model speciation
calculations in seawater at pH 8.2, Turner et al. (1981) show that Al exists entirely in
complexes with hydroxide ion, OH-, forming the hydrolysis species. At low pH (< 4)
the speciation is dominated by the free Al3+ ion. At higher pH (≥ 7), including that of
most natural waters, the Al speciation is thought to be dominated by Al (OH)4- or
aqueous Al(OH)30 (Figure 1). Byrne et al. (1988) show that at pH 8.2, similar to
most surface ocean waters, the dominant hydrolysis species is Al(OH)4- at both 5ºC
and 25ºC. However, at a slightly lower pH of 7.6 and at a temperature of 5ºC the
dominant hydrolysis species is modeled to be aqueous Al(OH)30. Sources of Al to the
worlds oceans include atmospheric inputs, riverine inputs, and, to a much lesser
degree, dissolution from sediments.
2
Al can exist in the ocean in either a dissolved form or a particulate form. The
dissolved fraction is operationally defined as that which passes through a given filter
size, usually 0.2-0.45 µm pore size, and includes truly soluble species such as the
dissolved inorganic hydrolysis species mentioned above, as well as a potential
colloidal fraction. For the purposes of this dissertation, the truly soluble species are
those which pass through a 0.03 µm polyethylene hollow fiber flow-through filter
(SteraporeTM, Mitsubishi Rayon) while colloidal Al is associated with the 0.03-0.4 µm
size range. Colloidal Al is dissolved Al complexed with large colloidal-size organic
ligands such as humic and/or fulvic acids and potentially colloidal size inorganic
precipitates. Particulate Al, that fraction of Al retained by a 0.2-0.45 µm pore size
filter, can be either leachable particulate Al or refractory particulate Al. Leachable
particulate Al is Al associated with easily dissolved particulate phases such as
amorphous Al oxyhydroxide coatings on particles, calcium and magnesium
carbonates, adsorbed within biogenic opal, or associated with other particulate
organic phases. Refractory particulate Al is relatively inert Al that is locked within
mineral lattices. Together, the leachable and refractory particulate Al make up the
total particulate Al.
The concentration of dissolved Al in riverine waters is highly variable,
ranging from ~ 50 nM to > 1 µM (Hydes and Liss, 1977; Mackin and Aller, 1984a;
Morris et al., 1986; Upadhyay and Sen Gupta, 1995; Takayanagi and Gobeil, 2000;
Brown and Bruland, 2009). Oceanic levels of dissolved Al are much lower and found
3
at trace concentrations, ranging by more than two orders-of-magnitude from less than
0.2 nM in productive waters of the subarctic North Pacific to > 25 nM in surface
waters of the eastern N. Atlantic where high eolian dust input from the Saharan desert
is observed (Orians and Bruland, 1986; Kramer et al., 2004; Measures et al., 2008;
Brown et al., submitted 2009).
Vertical distributions of Al in the subtropical gyres of the Atlantic and the
Pacific Ocean generally exhibit scavenged-type distributions with higher
concentrations in the surface, a mid-depth minimum, and an increase at depth. In the
North Atlantic, Hydes (1979) measured dissolved Al concentrations of ~ 36 nM in
surface waters, a mid-depth minimum of ~ 21 nM at 1000 m, and an increase to
values > 30 nM at 4000 m near the bottom. More recently on a transect from Iceland
to Brazil across the Saharan dust plume, Measures et al. (2008) observed surface
water dissolved Al concentrations of ~2-37 nM, with the maximum values
corresponding with the location of the large dust plume emanating from the Sahara
desert and minimum values occurring in the high-latitude North Atlantic near 60°N.
Intermediate (1000 m) water Al concentrations ranged from ~ 6-20 nM along the
cruise track. Very recently, as part of the new GEOTRACES intercalibration effort,
vertical profiles of dissolved Al (Brown and Bruland, unpublished data) were
collected at the Bermuda Atlantic Time Series (BATS) station in the summer of 2008
and at the former SAFe station in the oligotrophic subtropical North Pacific gyre in
late spring 2009. The two profiles are shown in Figure 2. The North Atlantic profile
agrees with previous studies showing surface dissolved Al concentrations of ~ 28 nM
4
and a mid-depth minimum of ~ 13 nM. The GEOTRACES North Pacific dissolved
Al data show significantly lower values compared to the North Atlantic with surface
water dissolved Al concentrations of ~ 4 nM and a decrease with depth to values of ~
0.5 nM at 2000m. Similarly, Orians and Bruland (1986) observed significantly lower
dissolved Al concentrations relative to the North Atlantic of 5 nM in the surface
waters of the subtropical N. Pacific Ocean, a decrease to 0.3 nM Al at 2000m, and an
increase to 2 nM at 5300m near the bottom. This scavenged-type of inter-basin
fractionation of Al observed in the deep waters of the ocean basins, the greatest for
any trace metal, is due to the particle-reactive nature of Al and its subsequent
scavenging and removal from seawater along the path of global thermohaline
circulation (Bruland and Lohan, 2003). It should be noted that the oceanic residence
time of Al with respect to scavenging removal has been estimated to be 50-150 years
(Orians and Bruland, 1986), roughly an order of magnitude less than the mixing time
of the deep oceans via thermohaline circulation. Very recently, Middag et al. (2009)
reported very low dissolved Al concentrations (~ 1 nM) in Arctic Ocean surface
waters, nutrient-type profiles with dissolved Al increasing up to ~ 28 nM in deeper
Arctic Ocean waters, and a strong correlation with silica indicating a biological
influence on the distribution of Al.
The fact that oceanic levels of dissolved Al are relatively low compared to
that of riverine systems implies one, if not several, removal mechanisms of Al in the
delivery of riverine and/or lithogenic Al to oceanic waters. Bruland and Lohan
(2003) note that Al has hydrolysis chemistry similar to that of thorium, an extremely
5
particle-reactive element. The scavenging intensity of 234Th within the surface photic
zone has been shown to be dramatically increased in productive regions of the
subarctic North Pacific and California Current system relative to the oligotrophic
gyres of the North and South Pacific (Figure 3). The mean life of 234Th with respect to
removal via particle scavenging was estimated to be 6-10 days in surface waters of
the subarctic Alaskan gyre and California Current system while estimates in the
subtropical North Pacific were 100-300 days. Orians and Bruland (1986) concluded
that dissolved Al removal from surface waters is similar in pattern but at a slower
rate. In addition, Gehlen et al. (2002) showed that Al can also be removed from
seawater as it is incorporated into diatoms during silica frustule biosynthesis. This
biological mechanism of Al removal explains the nutrient-type dissolved Al profiles
and strong correlation with silica observed by Middag et al. (2009) in the Arctic
Ocean.
Han et al. (2008) have developed a global model for surface water Al cycling
which estimates that particle scavenging of Al dominates removal processes in ~ 70%
of the surface ocean while biological uptake of Al in siliceous frustule biosynthesis
accounts for ~30% of Al removal. They suggest that biological uptake will exceed
particle scavenging in productive regions where silica production is high and dust
flux is low. Additional removal mechanisms of Al from seawater during river-ocean
mixing have also been proposed: salt-induced flocculation of dissolved organic-Al
complexes upon river-ocean mixing (Sholkovitz, 1976) and Al removal with respect
6
to an authigenic aluminosilicate mineral formation equilibrium (Mackin and Aller,
1984a,b).
In a transect from subtropical oligotrophic N. Pacific waters south of Hawaii
to the high productivity waters of the California Current system, Orians and Bruland
(1986) observed a decrease in surface dissolved Al values of nearly an order of
magnitude from 2-5 nM near Hawaii in the core of the North Pacific subtropical gyre
to 0.3-0.7 nM in the California Current system. The residence time of dissolved Al in
the surface waters of the California Current system was estimated to be 100 days,
likely shorter during times of intense production. Similarly, in a later transect from
California to Hawaii, Johnson et al. (2003) observed surface dissolvable (20 µm
filtered, pH 3.3) Al values of 4.6-6.6 nM in the oligotrophic waters near Hawaii and
values of ~ 0.5 nM in the California Current (Figure 4). The low concentrations of Al
within the California Current system were attributed to enhanced particle scavenging
by sinking biogenic particles. These particles provide surface adsorption sites for
dissolved Al, transporting it out of the surface water (Orians and Bruland, 1986).
The behavior of Al in moving from open-ocean regimes to the outer edge of
coastal upwelling zones has been characterized as discussed above and is well-
understood. Increased eolian inputs of Al to surface waters of the subtropical North
Pacific gyre and decreased particle scavenging in the low-productivity surface waters
leads to elevated dissolved Al concentrations. Moving into the high productivity
waters of the California Current system, increased scavenging of dissolved Al onto
biogenic particle surfaces and/or pseudo-biological uptake of dissolved Al into
7
diatom frustules during biosynthesis lead to decreased dissolved Al concentrations in
surface waters. Much less is known regarding the behavior of Al on the continental
side of high productivity coastal environments. A schematic of the behavior of
dissolved Al in surface waters of the oligotrophic gyre, high productivity coastal
regions, and the continental sided of these regions is shown in Figure 5. Rivers are a
significant source of Al to the world’s oceans, largely due to the weathering of
aluminosilicate minerals in continental rock. On the continental side of these high
productivity coastal regions such as the California Current system, it is possible that a
number of mechanisms are influencing the behavior of dissolved and particulate
aluminum: inorganic precipitation reactions, organic flocculation of dissolved Al
during river and estuarine mixing, passive and pseudo-active particle Al scavenging,
diffusion of Al into sediments, and sediment resuspension (Figure 5). In terms of
quantifying inputs of Al to the world’s oceans, the fate of terrigenous Al supplied by
rivers and streams in the coastal environment as Al-enriched river plumes and/or
coastal water mixes with surrounding waters needs to be investigated. It is the
behavior and distribution of Al on the continental side of these high productivity
coastal regions that is the focus of this dissertation.
Dissolved Al in seawater can serve as a valuable tracer of both eolian and
fluvial inputs to the oceans (Orians and Bruland, 1986; Hydes, 1989; Chou and
Wollast, 1997; Measures and Vink, 2000; Measures et al., 2005). Measures and Vink
(2000) showed that dissolved Al concentrations in surface waters of the open ocean
can provide reliable estimates of atmospheric dust fluxes to the world ocean. Their
8
MADCOW (measurement of Al for dust calculation in oceanic waters) model
postulates that in surface waters of the open ocean the input of dissolved Al is entirely
due to the dissolution of atmospheric dust and that the removal of dissolved Al is due
to passive particle scavenging through particle flux from biological processes. The
relationship between dust input estimates as derived from surface water aluminum
concentrations using the MADCOW model and independent dust estimates from
remote sampling towers on island or ships as shown in Measures and Vink (2000) is
shown in Figure 6.
The relationship between surface water dissolved Al concentrations and
atmospheric dust deposition is of great interest in a time when global estimates of the
delivery of bio-limiting elements such as iron via dust deposition to open ocean
surface waters are needed to model the global climate system. Utilizing estimates of
Fe solubility from atmospheric dust in surface waters, it is then possible that surface
water dissolved Al concentrations can provide accurate estimates of Fe input to the
open ocean. In a somewhat analogous manner, an initial premise of this dissertation
was that dissolved Al could potentially be used as a tracer of Fe input to coastal
surface waters from low-salinity plumes entering the ocean. Chase et al. (2007)
propose that rivers along the U.S. west coast deliver significant amounts of Fe to
coastal waters during high-flow winter conditions that is then trapped on the shelf and
re-mobilized for phytoplankton use during summer upwelling conditions. While
dissolved Fe will be biologically utilized quite rapidly in surface waters, it was
hypothesized that dissolved Al with a presumed longer residence time than dissolved
9
Fe in coastal surface waters might be used to estimate Fe inputs from the fluvial
sources.
Thus, the ability to accurately measure dissolved Al in seawater, particularly
with portable ship-board instrumentation/methods capable of near real-time rapid and
continuous measurements, is very valuable and there has been significant interest
shown in understanding the marine biogeochemistry of Al in both open-ocean and
coastal environments.
Due to the low concentrations of dissolved Al in seawater and the presence of
a matrix of major ions in seawater such as Ca2+, Mg2+, Na+, and Cl- at concentrations
six or seven orders of magnitude greater than the dissolved Al concentrations, a
method that provides for the removal of the seawater matrix and increases the
concentration of the analyte (e.g,, dissolved Al) is highly desirable. Flow-injection
(FI) methods are ideal for these and other reasons. First, FI methods provide a
closed-system by which contamination is minimized. Second, FI methods are small,
portable, and can be easily adapted to running on-board ship for real-time sample
determinations. Third, and most importantly, FI methods can be adapted to include a
preconcentration step that increases the analyte concentration in the reagent stream
and a seawater matrix removal step, both utilizing an in-line preconcentration
column. Finally, preconcentration columns can be placed in-line on reagent flow lines
for a particular reagent(s) in order to clean the reagent in case there are contamination
issues .
10
A FI method for the determination of dissolved Al utilizing a commercially-
available preconcentration chelating resin followed by fluorometric detection of an
Al-lumogallion complex is the result of my initial dissertation research and is
presented in Chapter 1 (Brown, M.T., Bruland, K.W., 2008, An improved flow
injection analysis method for the determination of dissolved aluminum in seawater,
Limnology and Oceanography Methods, 6, 87-95). This method is based on the
earlier work of Resing and Measures (1994), which utilized the preconcentration of
Al onto resin-immobilized 8-hydroxyquinoline (8-HQ). While 8-HQ is very useful as
a preconcentration resin due to its affinity for several metals, it is not commercially
available and the synthesis described by Dierssen et al. (2001) has proven to be
difficult and time-consuming with significant batch-to-batch variation in resin quality.
The Resing and Measures (1994) method makes use of the fluorescence of an
Al-lumogallion complex at an excitation wavelength of 489 nm and an emission
wavelength of 559 nm. The first major modification to the Resing and Measures
(1994) method was the replacement of 8-HQ with Toyopearl AF-Chelate 650M, a
commercially-available hydroxylated methacrylic polymer base resin bead
derivatized with iminodiacetate (IDA) functional groups. The IDA functional groups
serve as a tridentate ligand (with the two negatively charged oxygen atoms on the
carboxylate groups and the unshared pair of electrons on the nitrogen atom serving as
electron donors) for complexing various metal cations (Cd2+, Cu2+, Mn2+, Ni2+, and
Pb2+; Warnken et al., 2000). The initial focus of my Chapter 1 research was
characterizing the Toyopearl AF-Chelate 650M resin in terms of its capacity to bind
11
Al in solution. The second major modification was the introduction of a column-
conditioning step prior to the sample loading process in order to bring the resin to the
correct pH for binding Al. As will be shown in Chapter 1, this step leads to an
increase in sensitivity. Other modifications include in-line pH adjustment of acidified
samples, a changed acid-eluent strength, and a change in the eluent matrix from
seawater to Milli-Q deionized water. This work provided an easy-to-use, highly
sensitive method for the determination of Al in seawater that can be adapted to
measure a wide range of Al concentrations. The new and improved method presented
in Chapter 1 was used for the subsequent analysis of river, estuary, and seawater
samples in the Columbia River region off Oregon and Washington (Chapter 2) and in
the Gulf of Alaska (Chapter 3).
In Chapter 2 (Brown, M.T., Bruland, K.W., 2009, Dissolved and particulate
Al in the Columbia River and coastal waters of Oregon and Washington: Behavior in
near-field and far-field plumes, Estuarine, Coastal and Shelf Science, 84, 171-185), I
present and interpret high-resolution dissolved Al data and particulate Al data from
the Columbia River plume region off the coasts of Oregon and Washington. The
behavior of dissolved and particulate Al within the Columbia River estuary as well as
in low-salinity plumes during upwelling and downwelling conditions is investigated.
The Columbia River region is of interest as the Columbia River is the major source of
freshwater being delivered to the northeast Pacific Ocean. During the late summer,
the Columbia River accounts for ~ 90% of the freshwater entering the seas between
the Strait of Juan de Fuca and San Francisco Bay (Barnes et al., 1972). The Columbia
12
River plume forms as coastal seawater intrudes and mixes with river water at or near
the mouth of the Columbia River estuary. This plume is found confined tightly along
the Washington coast north of the river mouth during downwelling conditions when
wind stress is to the north and offshore, southwest of the river mouth during
upwelling conditions when wind stress is to the south. Previous work has shown that
chlorophyll concentrations along the Washington and Oregon coasts vary widely,
with higher concentrations observed in the Columbia River plume and along the
Washington coast north of the Columbia River (Landry et al., 1989; Thomas and
Strub, 2001). It is hypothesized that iron (Fe) supply from the Columbia River might
be a cause of the difference in observed chlorophyll concentrations (Hickey and
Banas, 2003).
An initial premise of this research was that dissolved and particulate Al could
be a tracer of the low-salinity Columbia River plume, identifying its presence and the
source of Fe, even if Fe concentrations had been biologically drawn down. My goals
for this research were to elucidate the behavior of dissolved and particulate Al within
the Columbia River estuary as well as investigate the dynamics of dissolved and
particulate Al within low salinity plumes in moving from near-field plumes to far-
field plumes a greater distance from the river mouth. The study provided the unique
opportunity to observe the Columbia River plume both north and south of the river
mouth during downwelling and upwelling periods, respectively.
The Columbia River plume Al study was conducted as part of the final River
Influence on Shelf Ecosystems (RISE) cruise in May/June 2006. High resolution
13
dissolved Al samples and large-volume unfiltered samples for filtrations and
subsequent particulate analysis were collected in low-salinity plumes as well as in the
Columbia River and estuary during both upwelling and downwelling conditions.
Hydrographic (temperature and salinity) data and macronutrient data was used in
conjunction with the dissolved and particulate Al data to elucidate the behavior of Al
within the estuary and low-salinity plumes. Both dissolved Al and particulate Al
concentrations were significantly greater in the river that in the coastal seawater that
mixes to form the plume. Dissolved Al within the estuary showed a ~60% removal at
salinities ranging from 0 to 10 and mechanisms for this removal are discussed. Data
collected from near-field and far-field low-salinity plumes showed that both dissolved
and particulate Al concentrations decrease with distance from the river mouth.
Particle scavenging of Al and dilution of the plume with lower Al oceanic waters are
discussed as well. However, even over 100 km from the mouth of the Columbia
River, both dissolved and particulate Al in a low-salinity far-field plume were over an
order-of-magnitude greater in concentration than in surrounding ocean waters.
While the coastal region off Washington and Oregon is largely influenced by
the freshwater input of the Columbia River, the coastal northern Gulf of Alaska
(GoA) is a region dominated by numerous glaciers and rivers emptying into the
region. The GoA is a semi-enclosed basin of the North Pacific Ocean, bordered to the
west, north, and east by a mountainous coastline containing many glaciers and rivers
and open to the subarctic Alaskan gyre waters to the south. Annual freshwater
discharge into the GoA (~ 2.4 x 104 m3 s-1) is comparable to that of the Mississippi
14
River (~ 2.0 x 104 m3 s-1) and reaches a maximum in the late summer months due to a
melting snowpack and late-summer rains (Brabets, 1997). Approximately 20% of the
watershed of the GoA is covered by glaciers (Royer, 1981). These glaciers have been
referred to as “buzzsaws” of the mountainous regions they cover due to the rapid
erosion they cause (Spotila et al., 2004) and are thus responsible for large inputs of
sediment to the freshwater discharge into the GoA.
In Chapter 3 (Brown, M.T., Lippiatt, S.M., Bruland, K.W., 2009, Dissolved
Al, particulate Al, and silicic acid in northern Gulf of Alaska coastal waters:
glacial/riverine inputs and extreme reactivity, submitted to Marine Chemistry) I
present the first-ever dissolved Al data as well as particulate Al and silicic acid data
in northern Gulf of Alaska (GoA) coastal waters. The focus of this chapter was
largely the coastal region of the northern GoA influenced by the significant sediment-
laden freshwater discharge to the region but also included the offshelf waters of the
Alaskan subarctic gyre and the transition zone in-between. The study aimed to
characterize dissolved and particulate Al distributions in northern GoA coastal waters
and relate freshwater inputs and Al removal mechanisms to observed Al
concentrations.
Determinations of dissolved Al and collection of filters for particulate Al
analysis were done on-board ship during a research cruise aboard the R/V Thomas G.
Thompson in the northern GoA during August/September 2007. Sampling took place
over four near-shore coastal surface water transects and one cross-shelf surface water
transect which encompassed the low-salinity coastal waters of the Alaska Coastal
15
Current as well as the offshelf, high-nutrient, lower-than-expected chlorophyll waters
of the Alaskan subarctic gyre. Vertical profile samples were also collected at several
stations along the shelf and at the shelf break.
Low salinity plumes sampled along the coastal transects due to the
glacial/riverine inputs were dramatically enriched in dissolved Al, particulate Al, and
silicic acid relative to surrounding coastal waters. Dissolved Al in the low-salinity
plumes was observed to exist nearly entirely in the soluble form. The percent-
leachable Al fraction in these low salinity plumes was quite low, indicative of the
largely refractory nature of the lithogenic material being delivered to the coastal
waters. A consistency in Si:Al ratios from different time periods point to a common
source for dissolved Al and silicic acid in the low salinity plume waters emanating
from two rivers in the region and this source is discussed in terms of its chemical
weathering. It appears also that the freshwater endmember characteristics are
different for a fresh glacial meltwater source as opposed to glacial melt that has then
been delivered to the ocean via riverine input that has traveled over the underlying
bedrock, vegetation, and soil. A final note of interest are the dramatically decreased
dissolved and particulate Al concentrations observed in mid- and off-shelf subarctic
Alaskan gyre waters.
16
References
Baes, C.F., Mesmer, R.E. 1976. Hydrolysis of Cations. Wiley Publishers, New York.
Barnes, C.A., Duxbury, A.C., Morse, B.A., 1972. Circulation and selected propertiesof the Columbia River effluent at sea. In: Pruter, A.T., Iverson, D.L. (Eds.), TheColumbia River Estuary and Adjacent Ocean Waters. Univ. of Washington Press.Seattle, WA, pp.41-80.
Brabets, T.P., 1997. Geomorphology of the lower Copper River, Alaska. USGeological Survey Professional Paper, 1581.
Brown, M.T., Bruland, K.W., 2009. Dissolved and particulate aluminum in theColumbia River and coastal waters of Oregon and Washington: Behavior in near-fieldand far-field plumes. Estuarine, Coastal and Shelf Science, 84, 171-185.
Brown, M.T., Lippiatt, S.L., Bruland, K.W. Dissolved aluminum, particulatealuminum, and silicic acid in northern Gulf of Alaska coastal waters: glacial/riverineinputs and extreme reactivity. Submitted to Marine Chemistry, Fall 2009.
Bruland, K.W., Lohan, M.C., 2003. Controls of trace metals in seawater, In: Treatiseon Geochemistry, Vol. 6 (Elderfield, H., Holland, H., and Turekian, K., eds.).Elsevier/Pergamon Press, 23-48.
Byrne, R.H., Kump, L., Cantrell, K.J., 1988. The influence of temperature and pH oftrace metal speciation in seawater. Marine Chemistry, 25, 163-181.
Chase, Z., Strutton, P., Hales, B., 2007. Iron links river runoff and shelf width tophytoplankton biomass along the U.S. West Coast. Geophysical Research Letters, 34,L04607, doi: 10.1029/2006GL028069.
Chou, L., Wollast, R., 1997. Biogeochemical behavior and mass balance of dissolvedaluminum in the western Mediterranean Sea. Deep Sea Research II, 44, 741-768.
Dierssen, H., Balzer, W., Landing, W., 2001. Simplified synthesis of an 8-hydroxyquinoline chelating resin and a study of trace metal profiles from JellyfishLake, Palau. Marine Chemistry, 73, 173-192.
Gehlen, M., Beck, L., Calas, G., Flank, A. –M., Van Bennekom, A.J., VanBeusekom, J.E.E., 2002. Unraveling the atomic structure of biogenic silica: Evidenceof the structural association of Al and Si in diatom frustules. Geochimica etCosmochimica Acta, 66, 9, 1601-1609.
17
Han, Q., Moore, J.K., Zender, C., Measures, C., Hydes, D., 2008. Constrainingoceanic dust deposition using surface ocean dissolved Al. Global BiogeochemicalCycles, 22, GB2003, doi:10.1029/2007GB002975.
Hydes, D.J., Liss, P.S., 1977. The behavior of dissolved Al in estuarine and coastalwaters. Estuarine and Coastal Marine Science, 5, 755-769.
Hydes, D.J., 1979. Aluminum in seawater: control by inorganic processes. Science205, 1260-1262.
Hydes, D.J. 1989. Seasonal variations in dissolved aluminum concentrations incoastal waters and biological limitation of the export of the riverine input ofaluminum to the deep sea. Continental Shelf Research, 9, 919-929.
Johnson, K.S. Elrod, V.A., Fitzwater, S.E., Plant, J.N., Chavez, F.P. et al. 2003.Surface ocean-lower atmosphere interactions in the Northeast Pacific Ocean gyre:Aerosols, iron, and the ecosystem response. Global Biogeochemical Cycles, 17, 2,1063, doi:10.1029/2002GB002004.
Kramer, J., Laan, P., Sarthou, G., Timmermans, K.R., de Baar, H.J.W., 2004.Distribution of dissolved aluminum in the high atmospheric input region of thesubtropical waters of the North Atlantic Ocean. Marine Chemistry, 88, 85-101.
Landry, M.R., Postel, J., Peterson, W., Newman, J., 1989. Broad-scale distributionalpatterns of hydrographic variables on the Washington/Oregon shelf. In: Landry,M.R., Hickey, B.M. (Editors), Coastal Oceanography of Oregon and Washington.Elsevier Oceanography Series, 47. New York, NY. pp. 1-40.
Mackin, J.E., Aller, R.C., 1984a. Dissolved Al in sediments and waters of the EastChina Sea: Implications for authigenic mineral formation. GeochimicaCosmochimica Acta, 48, 281-297.
Mackin, J.E., Aller, R.C., 1984b. Processes affecting the behavior of dissolvedaluminum in estuarine waters. Marine Chemistry, 14, 213-232.
Measures, C.I., Vink, S., 2000. On the use of dissolved aluminum in surface waters toestimate dust deposition to the ocean. Global Biogeochemical Cycles, 14, 317-327.
Measures, C.I., M.T. Brown, Vink, S., 2005. Dust deposition to the surface waters ofthe western and central North Pacific inferred from surface water dissolved aluminumconcentrations. Geochemistry, Geophysics, and Geosystems, 6,doi:10.1029/2005GC000922.
18
Measures, C.I., Landing, W.M., Brown, M.T., Buck, C.S., 2008. High-resolution Aland Fe data from the Atlantic Ocean CLIVAR-CO2 Repeat Hydrography A16NTransect: Extensive linkages between atmospheric dust and upper oceangeochemistry, Global Biogeochemical Cycles, 22, GB1005, doi:10.1029/2007/GB003042.
Middag, R., de Baar, H.J.W., Laan, P., Bakker, K., 2009. Dissolved Al and the siliconcycle in the Arctic Ocean. Marine Chemistry, doi: 10.1016/j.marchem.2009.08.002.
Morris, A.W., Howland, R.J.M., Bale, A.J., 1986. Dissolved aluminum in the Tamarestuary, southwest England. Geochimica Cosmochimca Acta, 50, 189-197.
Orians, K.J., Bruland, K.W., 1986. The biogeochemistry of aluminum in the PacificOcean. Earth and Planetary Science Letters, 78, 397-410.
Perdue, E.M., Beck, K.C., Reuter, J.H., 1976. Organic complexes of iron andaluminum in natural waters. Nature, 260, 418-420.
Reesman, A. L., Pickett, R.E., Keller, W.D., 1969. Aluminum ions in aqueoussolutions. American Journal of Science, 267, 99-113.
Resing, J.A., Measures, C.I., 1994. Fluorometric detection of Al in seawater by flowinjection analysis with in-line preconcentration. Analytical Chemistry, 66, 4105-4111.
Royer, T.C., 1981. Baroclinic transport in the Gulf of Alaska Part II: A fresh waterdriven coastal current. Journal of Marine Research, 39, 2, 251-266.
Sholkovitz, E.R., 1976. Flocculation of dissolved organic and inorganic matter duringthe mixing of river water and seawater. Geochimica et Cosmochimica Acta, 40, 831-845.
Spotila, J.A., Buscher, J.T., Meigs, A.J., Reiners, P.W., 2004. Long-term glacialerosion of active mountain belts: Example of the Chugach-St. Elias Range, Alaska.Geology, 32, 501-504
Stumm, W. Morgan, J.J. 1996. Aquatic Chemistry: Chemical Equilibria and Rates inNatural Waters, 3rd edition. John Wiley and Sons, Inc. Pub., New York.
19
Takayanagi, K., Gobeil, C., 2000. Dissolved aluminum in the Upper St. Lawrenceestuary. Journal of Oceanography, 56, 517-525.
Taylor, S.R., 1964. Abundance of chemical elements in the continental crust: a newtable. Geochimica Cosmochimica Acta, 28, 1273-1285.
Thomas, A., Strub, P.T., 2001. Cross-shelf phytoplankton pigment variability in theCalifornia Current. Continental Shelf Research, 21, 1157-1190.
Turner, D.R., Whitfield, M., Dickson, A.G., 1981. The equilibrium speciation ofdissolved components in freshwater and seawater at 25ºC and 1 atm pressure.Geochimica Cosmochimica Acta, 45, 855-881.
Upadhyay, S., Sen Gupta, R., 1995. The behavior of aluminum in waters of theMandovi estuary, West Coast of India. Marine Chemistry, 51, 261-276.
Warnken, K., Tang, D., Gill, G., Santschi, P., 2000. Performance optimization of acommercially available iminodiacetate resin for the determination of Mn, Ni, Cu, Cd,and Pb by on-line preconcentration inductively coupled plasma-mass spectrometry.Analytica Chimica Acta, 423, 265-276.
20
Figure 1. Al speciation and solubility as a function of pH based on the solubility ofgibbsite in freshwater from Stumm and Morgan (1996). Note that the theoreticalminimum solubility of Al occurs at a pH of ~ 6.5-6.7.
21
Figure 2. Vertical profiles of dissolved Al from the GEOTRACES North AtlanticIntercalibration cruise (open circles) at the Bermuda Atlantic Time Series station inthe subtropical North Atlantic gyre and from the GEOTRACES Pacific speciationcruise (black diamonds) at the SAFe station in the subtropical North Pacificgyre.(Brown and Bruland, unpublished data).
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0 5 10 15 20 25 30 35
Dissolved Al (nM)D
epth
(m
)
GEOTRACESN. AtlanticIntercalibration Profile (BATS)
GEOTRACESN. Pacific Profile(SAFe station)
22
Figure 3. The net scavenging rate constant of Th-234 from the surface mixed layerof the North Pacific Ocean as shown in Bruland and Lohan (2003). Note that thehighest net scavenging rate constants occur in the high productivity waters of theCalifornia Current system and the Gulf of Alaska while the lowest net scavengingrate constants occur in the low productivity waters of the subtropical gyres.
23
Figure 4. Surface water dissolvable aluminum along a transect between Hawaii andMonterey from Johnson et al. (2003). Note the order-of-magnitude decrease indissolvable aluminum in the high productivity waters of the California Currentrelative to the surface waters of the subtropical North Pacific gyre.
California Current
Subtropical N. Pacificgyre
24
Figure 5. A schematic of the processes governing dissolved Al in a model ocean.The ocean is divided into the oligotrophic gyre (left), high productivity coastal regionsuch as the California Current (middle), and the continental side of the highproductivity coastal region (right). It is the behavior and distribution of dissolved Alon the continental side of the high productivity coastal region (outlined by the blackcircle) that is the focus of this dissertation.
Atmospheric Dust InputAtmospheric Dust Input
Oligotrophic gyreOligotrophic gyre(elevated Al)(elevated Al)ττ = 3-5 years = 3-5 years
AlAlAl
AlAl Al
AlAl
???
Much less is known regarding the behavior of aluminumon the continental side of high productivity coastal
environments….
RiverRiverInputInput
or
Diffusion and/orSediment resuspension
-increased eoliandust input
-less export productionand particle scavengingof Al
Al
AlAl
AlAl
-decreased eoliandust input
-more export productionand particle scavengingof Al
High Prod. CoastalHigh Prod. CoastalRegionRegion
(Decreased Al)(Decreased Al)ττ ~ 50-100 days ~ 50-100 days
Inorganicprecipitation
Al Organic flocculation
Al
AlAl
AlAl Particle
scavenging
25
Figure 6. MADCOW model calibration comparing surface water dissolved Al-derived dust estimates with independent deposition estimates from Measures andVink (2000).
26
CHAPTER 1:
AN IMPROVED FLOW INJECTION ANALYSIS METHOD FOR THEDETERMINATION OF DISSOLVED ALUMINUM IN SEAWATER
Matthew T. Brown and Kenneth W. Bruland (2008).Limnology and Oceanography Methods, 6: 87-95.
27
28
29
30
31
32
33
34
35
36
CHAPTER 2:
DISSOLVED AND PARTICULATE ALUMINUM IN THE COLUMBIA RIVERAND COASTAL WATERS OF OREGON AND WASHINGTON: BEHAVIOR IN
NEAR-FIELD AND FAR-FIELD PLUMES
Matthew T. Brown and Kenneth W. Bruland (2009).Estuarine, Coastal, and Shelf Science, 84: 171-185.
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
CHAPTER 3:
DISSOLVED ALUMINUM, PARTICULATE ALUMINUM, AND SILICIC ACIDIN NORTHERN GULF OF ALASKA COASTAL WATERS:
GLACIAL/RIVERINE INPUTS AND EXTREME REACTIVITY
Matthew Brown, Sherry M. Lippiatt, and Kenneth W. Bruland (2009).Submitted to Marine Chemistry
53
Dissolved aluminum, particulate aluminum, and silicic acid in northern Gulf ofAlaska coastal waters: glacial/riverine inputs and extreme reactivity
Brown, Matthew T., Lippiatt, S.M., and Bruland, K.W. (2009)Submitted to Marine Chemistry
Abstract
The coastal northern Gulf of Alaska receives significant fluvial inputs from
the numerous glaciers and rivers that border the region. The distribution of dissolved
(soluble and total) and particulate aluminum (leachable and total) was examined in
coastal surface water transects and vertical profile samples from shelf stations in the
northern Gulf of Alaska in August and September 2007. Both dissolved and
particulate aluminum concentrations were dramatically increased in low-salinity
plumes, indicative of the significant riverine/glacial input to coastal waters in the
region. The percent-leachable particulate Al fraction in these low salinity plumes was
quite low (~ 7%), indicative of the largely refractory nature of the lithogenic material
being delivered to the coastal waters. A consistency in Si:Al ratios from different
time periods in the freshwater endmembers that mix to form the low-salinity plumes
is discussed in terms of the weathering of biotite in this region. The dissolved Al in
these coastal waters appears entirely in the soluble (< 0.03 µm) fraction, likely a
consequence of the freshwater sources to the region being very low in dissolved
organic carbon and organic colloidal complexes. An extreme decreasing gradient of
dissolved Al in surface waters was observed moving offshelf and into the waters of
the Alaskan subarctic gyre where some of the lowest dissolved Al concentrations
54
reported in the world ocean were observed. A high-degree of particle scavenging of
dissolved Al in the coastal waters is discussed with a residence time of dissolved Al
in coastal shelf waters estimated to be ~ 10 days.
1. Introduction
Aluminum (Al) is the third most abundant element in the Earth’s crust (8.2%
by weight; Taylor, 1964) yet the oceanic levels of dissolved Al are found at trace
concentrations, less than a few nM in surface waters of the Pacific ocean. Sources of
Al to the worlds oceans include riverine inputs, atmospheric inputs, and dissolution
from sediments. The concentrations of dissolved Al in riverine waters are highly
variable, ranging from ~ 50 nM to > 1 µM (Hydes and Liss, 1977; Mackin and Aller,
1984a; Morris et al., 1986; Upadhyay and Sen Gupta, 1995; Takayanagi and Gobeil,
2000). Oceanic levels of dissolved Al are found at trace concentrations, ranging from
less than 1 nM in intermediate and deep waters of the N. Pacific to > 25 nM in
surface waters of the eastern N. Atlantic where high eolian dust input from the
Saharan desert is observed (Orians and Bruland, 1986; Kramer et al., 2004; Measures
et al., 2008; Buck et al., 2009). While Al can exist as organic complexes in natural
waters (Perdue et al., 1976), Al speciation in the oceans is thought to be dominated
largely by the dissolved inorganic hydrolysis species Al(OH)2+, Al(OH)3
0, and
Al(OH)4- with the Al(OH)4
- species being dominant at the pH of surface seawater.
The fact that oceanic levels of dissolved Al are relatively low compared to
that of riverine systems implies one, if not several, removal mechanisms of Al in the
delivery of riverine Al to oceanic waters. Walker et al. (1988) showed that
55
aluminosilicate clay mineral surfaces can act as adsorption sites for aqueous Al(III)
hydrolysis species, a scavenging mechanism removing dissolved Al from seawater.
Bruland and Lohan (2003) note that Al has hydrolysis chemistry similar to that of
thorium (Th), an extremely particle-reactive element. They show that the mean life of
234Th with respect to removal via particle scavenging is ~ 6-10 days in productive
surface waters of the subarctic Alaskan gyre and California Current system while
estimates in the oligotrophic subtropical North Pacific were 100-300 days. Orians and
Bruland (1986) argued that dissolved Al removal would be similar in pattern but at a
slower rate. They attributed the low (< 1 nM) surface water Al concentrations of the
California Current to scavenging of Al by biogenic particles in the high-productivity
surface waters with high export production. More recently, Gehlen et al. (2002)
showed that Al can also be removed from seawater as it is incorporated into diatoms
during silica frustule biosynthesis. Han et al. (2008) have developed a global model
for surface water Al cycling which estimates that particle scavenging of Al dominates
removal processes in ~ 70% of the surface ocean while biological uptake of Al in
siliceous frustule biosynthesis accounts for ~30% of Al removal. They suggest that
biological uptake will exceed particle scavenging where silica production is high and
dust flux is low. Additional removal mechanisms of Al from seawater during river-
ocean mixing have also been proposed: salt-induced flocculation of dissolved
organic-Al complexes upon river-ocean mixing (Sholkovitz, 1976) and Al removal
with respect to an authigenic aluminosilicate mineral formation equilibrium (Mackin
and Aller, 1984a,b).
56
The Gulf of Alaska (GoA) is a semi-enclosed basin of the North Pacific
Ocean, bordered to the west, north, and east by a mountainous coastline containing
many glaciers and rivers and open to the subarctic Alaskan gyre waters to the south.
The circulation of the GoA is dominated by two current systems. The upwelling,
cyclonic, subarctic Alaskan gyre is bounded by the Alaska Current in the northeastern
GoA and the Alaskan Stream in the northwestern GoA (Figure 1). As the northward
flowing Alaska Current approaches the head of the GoA, the current turns
southwestward, following the isobaths and forming the Alaskan Stream (Stabeno et
al., 2004). Further inshore over the continental shelf, the circulation is dominated by
the Alaskan Coastal Current (ACC; Figure 1), a coastal feature with a marked
freshwater core (salinity ~ 26-29) observed from Icy Pt. in the northeastern GoA to
Unimak Pass near the end of the Aleutian island chain (Stabeno et al., 2004). The
ACC drives the surface circulation over the continental shelf and controls the
transport of lithogenic and biogenic material. Royer et al. (1981) showed that the
ACC is a baroclinic, coastal feature resulting predominantly from a large coastal
freshwater discharge and, to a lesser degree, local downwelling winds.
Annual freshwater discharge into the GoA (~ 2.4 x 104 m3 s-1) is comparable
to that of the Mississippi River (~ 2.0 x 104 m3 s-1) and reaches a maximum in the late
summer months due to a melting snowpack and summer rains (Brabets, 1997).
Approximately 20% of the watershed of the GoA is covered by glaciers (Royer,
1981). These glaciers have been referred to as “buzzsaws” of the mountainous
regions they cover due to the rapid erosion they cause (Spotila et al., 2004) and are
57
thus responsible for large inputs of sediment to the freshwater discharge into the
GoA. Feely et al. (1979) suggested that the Copper River basin is the major source of
sediment discharge to the northern GoA with peak discharge occurring from June
through September. The wind field of the GoA is dominated by strong cyclonic
winds from fall through spring as the region is the end of the North Pacific storm
track (Stabeno et al., 2004). These along-shore winds result in Ekman pumping in the
subarctic gyre and downwelling along the coast. This downwelling helps confine the
freshwater discharge mentioned above along the coast leading to the formation of the
ACC with a marked freshwater core. Feely et al. (1979) measured suspended
particulate loads of > 6 mg l-1 in the coastal surface waters of the northern GoA
influenced by the Copper River. Assuming 8.2 wt% Al measured in particulate
material collected in the surface mixed layer (Feely et al., 1979), which agrees with
the average wt% Al found in continental crust (Taylor, 1964), this equates to an
estimate greater than 18 µM total particulate Al assuming that the particulate material
is predominantly lithogenic. Anders et al. (2003) report riverine dissolved Al
concentrations of 1.8-3.0 µM at several sites at or near the mouth of the Copper
River. Thus, it follows that dissolved and particulate Al concentrations in northern
GoA coastal waters should be dramatically increased relative to the offshore waters of
the subarctic gyre due to significant glacial/riverine inputs along the coast and
relatively little atmospheric input of Al to the central Alaskan subarctic gyre (Duce
and Tindale, 1991).
58
This paper describes the distributions of soluble, dissolved, and particulate
(leachable and total) Al as well as silicic acid in northern GoA coastal waters during
the relatively high runoff, late summer months of August-September 2007. This is
the first-ever dissolved Al data reported in the coastal waters of the northern GoA.
Such data is lacking in high-latitude regions with glacial inputs. The study aimed to
characterize dissolved and particulate Al distributions in northern GoA coastal waters
and relate freshwater inputs and Al removal mechanisms to observed Al
concentrations. A coastal source of dissolved Al and silicic acid to the region is
discussed in terms of chemical weathering and the distribution of dissolved Al
between colloidal and soluble forms is investigated in low salinity coastal waters
dominated by glacial/riverine input. In addition, dissolved and particulate Al were
examined in waters of the Alaskan subarctic gyre (GAK line) and in coastal upwelled
waters of lower Cook Inlet to compare/contrast with observed Al distributions in
regions dominated by coastal runoff.
2. Methods
2.1 Sample Collection and Filtration
Seawater samples were collected aboard the R/V Thomas G. Thompson
(University of Washington) in the northern GoA from August 15, 2007 to September
20, 2007 in conjunction with an NSF-funded study of mesoscale eddies in the
northern and central GoA. Surface (~ 1m) sampling was conducted using an
underway clean surface pump “fish” system described in detail elsewhere (Bruland et
59
al., 2005; Aguilar-Islas and Bruland, 2006; Lohan and Bruland, 2006). Briefly, this
system utilizes an all PTFE Teflon TM diaphragm pump and PFA Teflon TM tubing
mounted to a PVC depressor vane 1 m above a 20-kg PVC fish, allowing for clean
surface water sampling while underway at speeds of 4-10 knots. Concomitant
underway surface salinity and temperature measurements were obtained using a YSI
CTD Sonde attached to the PVC fish. The YSI CTD Sonde system was calibrated
with the ship’s SeaBird CTD system, which, in turn, was calibrated by the University
of Washington’s CTD group just before the research cruise got underway. Underway
fluorescence data was obtained through the ship’s flow-through seawater system.
Vertical profile samples were collected using 30-L Go-Flo bottles (General Oceanics)
deployed on Kevlar line. Vertical profiles of hydrographic data (temperature, salinity,
etc.) were acquired through the use of the ship’s CTD. Samples for dissolved Al
collected from the “fish” system were filtered in-line through acid-cleaned 0.45 µm
Teflon TM capsule filters (GE Osmonics Capsule filters) unless specified otherwise
and samples for dissolved Al collected from the 30-L Go-Flo’s were filtered through
acid-cleaned 0.4 µm Nuclepore polycarbonate filters.
For particulate Al samples, an unfiltered water sample was collected from the
surface fish system or 30-L Go-Flo bottle directly into a 2-L acid-cleaned low-density
polyethylene (LDPE) bottle. All unfiltered samples were kept cold and dark prior to
filtration, which occurred within a few hours after sample collection. Unfiltered water
samples were filtered under trace metal clean conditions in a class 100 clean bench
using an in-line filtration apparatus as described in Berger et al. (2008). Briefly,
60
unfiltered samples passed through acid-cleaned 47mm, 10 µm filters followed by
47mm, 0.2 µm filters (NucleporeTM polycarbonate track-etched membrane filters)
mounted in polypropylene filter sandwiches (MilliporeTM). This filtration technique
captured 0.2 µm-10 µm and > 10 µm particulate material utilized in both leachable
particulate Al and refractory particulate Al analyses. Filters were folded into eighths
and placed in acid-cleaned 2 mL polypropylene vials and stored frozen. Filter blanks
were processed as described above with the exception that no sample passed through
the filters.
Soluble (< 0.03 µm) Al samples were collected by passing an unfiltered
sample first through a 47mm, 10 µm filter as mentioned above followed by an in-line
acid-cleaned 0.03 µm (200 kDa) polyethylene hollow fiber flow-through filter
(Sterapore TM, Mitsubishi-Rayon, Tokyo, Japan). More detail for this procedure is
given in Hurst and Bruland (2007) while protocols for filter maintenance are outlined
in Nishioka et al. (2001).
Unfiltered samples for total suspended solids (TSS) were collected from the
surface fish system and filtered through pre-weighed 47mm, 0.4 µm Nuclepore
polycarbonate filters. Filters were then rinsed with aliquots of MQ water (> 18 mega-
ohm deionized water from Milli-Q water purification system). These filters were then
dried and weighed and the total suspended material concentrations were determined
by mass difference and filtered volumes. Filtered volumes were approximately 0.3-
0.5 L. These filters were then digested with sodium hydroxide for the determination
of biogenic silica (BSi; Brzezinski and Nelson, 1989).
61
In addition, two samples for dissolved Al were collected from the Copper
River outside of Chitina, AK on September 24, 2007. Grab samples were collected
directly from the river and filtered through acid-cleaned, 0.45 µm Nuclepore
polysulfone membrane filters. The samples were acidified to pH 1.8 back in the
laboratory at UC Santa Cruz and sat for ~ 6 months prior to analysis.
2.2 Analytical Methods
Macronutrients (nitrate+nitrite, [referred to herein as nitrate] and silicic acid)
were determined underway directly from the outflow of the fish system.
Measurements were taken every two minutes on a Lachat QuikChem 8000 Flow
Injection Analysis system using standard colorimetric methods (Parsons et al., 1984).
Filtered samples for dissolved and soluble Al analyses were acidified on-
board ship to pH 1.7-1.8 using sub-boiled quartz distilled 6 N hydrochloric acid (4 ml
6 N HCl per L of sample, ~ 0.024 M HCl added to the sample). Samples were
allowed to sit for at least one hour prior to analysis. Shipboard determinations of
dissolved and soluble Al were made under trace metal clean conditions in a class 100
clean space using the flow injection method of Brown and Bruland (2008) which
utilizes the chelation of dissolved and/or soluble Al onto a commercially-available
iminodiacetate (IDA) resin. For dissolved Al concentrations greater than ~ 100 nM, a
direct injection technique utilizing a small-volume sample loop was used rather than
the IDA preconcentration resin (see Brown and Bruland, 2008 for details). Blank
analysis was performed using acidified MQ water which was loaded onto the IDA
62
column for a similar time as the samples being analyzed. Working Al standards were
made by serial dilutions of a 1000 ppm Al stock standard (SPEX). Al standards made
by standard additions of the working standards to volumes of low-Al seawater were
analyzed daily. This method has a detection limit of 0.1 nM Al when
preconcentrating 10 mL of sample and a precision of 2.5% based on replicate
analyses of a 5 nM Al sample when preconcentrating 2.5 mL of sample.
Procedures for estimating leachable particulate Al (LP Al) were performed on
select 47mm Nuclepore filters according to an optimal method outlined in Berger et
al. (2008). LP Al is Al associated with easily dissolved particulate phases such as Al-
oxyhydroxide coatings on particle surfaces, associated with Fe or Mn oxyhdyroxides,
adsorbed or in acid labile forms with biogenic particles, and Al that might possibly be
associated with carbonates. In short, filters are subjected to a two-hour weak acid
leach (25% acetic acid, pH ~2) with a mild reducing agent (0.02M hydroxylamine
hydrochloride) and an initial short heating step (10 minutes at 90°-95°C). The
leachate was then transferred to a quartz beaker with the filter being rinsed with 4 1-
ml aliquots of sub-boiled quartz distilled water. The leachate solution was then
acidified with 100 µl of concentrated trace metal grade (TMG) nitric acid (HNO3) and
evaporated down to dryness on a hotplate. The dry residue was then brought up in a
second 100 µl of concentrated TMG HNO3 and evaporated to dryness again. Finally,
the ensuing residue was brought up in 8 mL of 1N TMG HNO3 containing a 10ppb
Ga/1ppb Rh internal standard spike.
63
Refractory particulate Al (RP Al) is relatively inert Al that is locked within
mineral lattices. In order to estimate RP Al, all leached filters (with the remaining RP
Al fraction) were bomb-digested in a microwave oven with 3 ml concentrated TMG
HNO3 and 1 ml concentrated TMG hydrofluoric acid in PFA TeflonTM bombs with
pressure relief valves (Savillex, Minnetonka, MN). The resulting solution was
evaporated down to dryness on a hotplate. The dry residue was brought up in 8 ml of
1N quartz-distilled HNO3 containing an internal standard solution (10 ppb Ga and 1
ppb Rh) and heated briefly (~ 10 min) on a hot plate before being transferred to an
acid-cleaned 15 ml LDPE bottle.
The concentrations of LP Al and RP Al were measured at the University of
California Santa Cruz using a Thermo-Electron Element 1 high-resolution inductively
coupled mass spectrometer (ICPMS) (Berger et al, 2008; Hurst and Bruland, 2007).
It should be noted that total particulate Al (TP Al) concentrations presented herein are
the sum of the LP Al and the RP Al concentrations. Both acetic-acid leachate and
bomb-digestion solutions of 1N TMG HNO3 with the Rh/Ga internal standard were
transferred to acid-cleaned 7 ml vials and analyzed in medium resolution on the
ICPMS. Al concentrations were calculated based on calibration of instrument
response against Al standards ranging from 1 ppb to 500 ppb prepared from dilution
of 1000 ppm stock Al solution (SPEX; Edison, NJ; Fisher Scientific).
64
3. Results
3.1 Surface Transects
3.1.1 Transect 1
The locations of the northern GoA transects are shown in Figure 2. Tables 1-
5 show silicic acid, dissolved Al, and particulate Al data as well as corresponding
hydrographic data for each surface transect.
Transect 1 was sampled on August 19-20, 2007 beginning at 58.75°N,
138.16°W, southeast of Yakutat, AK with the intention of sampling the plume of the
Alsek River that has a maximum flow in July and August (USGS Station 15129000;
Alsek River). The transect initially moved in a general northwest direction along and
parallel to the inner shelf to near Dry Bay which is the entry point of the Alsek River
into the ocean. After the plume of the Alsek River was encountered two pronounced
“doglegs” were then undertaken that carried the sampling further offshelf and then
back to the innershelf. The transect ended at 59.95°N, 143.31°W. During the day
prior to the start of the transect winds were quite variable ranging from light (~1-2
knots) westerly winds to ~6-8 knots from the east-southeast. Initial temperature and
salinity of the shelf waters were ~16°C and 31, respectively, with silicic acid and
dissolved Al concentrations of 13-14 µM and ~30 nM. Initial LP Al was ~35 nM and
initial TP Al was ~120 nM.
Both temperature and salinity decreased dramatically upon entering the Alsek
River plume to minimum values of ~9°C and 7, respectively, indicative of a marked
freshwater plume near 59.09°N, 138.85°W, herein referred to as T1 Plume A (Figure
65
3). In these low-salinity plume waters silicic acid concentrations increased by a
factor of 2.5 to ~33 µM while dissolved Al concentrations increased by over a factor
of 40 to a maximum value of ~1300 nM. TSS concentrations increased from
background concentrations of ~3-8 mg l-1 to concentrations of 14-19 mg l-1 (data not
shown). LP Al concentrations increased to ~1750 nM (50-fold increase) while TP Al
concentrations increased to a maximum value greater than 25,000 nM (> 200-fold
increase; Figure 3).
Moving out of the low-salinity T1 Plume A waters, both temperature and
salinity increased to values of ~15.5°C and 30.6, respectively, with concomitant
decreases in silicic acid and dissolved Al concentrations to near-initial values of ~14
µM and 50 nM, respectively. Significant decreases in TSS concentrations, LP Al (<
10 nM), and TP Al (< 50 nM) concentrations were also observed outside of the
plume.
A second low-salinity plume was observed at 59.91°N, 142.00°W on the inner
shelf off Icy Bay, a region with multiple streams and rivers influenced by the Bering,
Guyot, and Malaspina Glaciers. This second low-salinity plume, herein referred to as
T1 Plume B, had similar minimum temperatures as observed in the first low-salinity
plume (~9°C) but salinity only decreased to minimum values of ~19.2. TSS
concentrations increased to ~15 mg l-1. Interestingly, dissolved Al concentrations
increased to nearly 500 nM (> 10-fold increase) in these low-salinity plume waters
yet silicic acid concentrations actually decreased from values of ~12.5-14 µM outside
of the plume to ~5 µM in the low salinity plume core (Figure 3). Marked increases in
66
fluorescence measured from the ship’s underway flowthrough system (data not
shown) were observed from ~ 141.1°W to ~142.4°W which correspond well to the
decreased silicic acid concentrations observed in Figure 2, likely indicating a diatom
bloom and the resultant biological uptake of silicic acid from surface waters. LP Al
increased to ~500 nM and TP Al increased to ~4850 nM in this second low salinity
plume (Figure 3).
Moving out of T1 Plume B, dissolved Al, LP Al, and TP Al decreased
significantly while silicic acid concentrations increased from ~ 2.5 µM to 14 µM. A
corresponding decrease in fluorescence was observed moving into higher salinity
(30.8-31.4) waters.
3.1.2 Transect 3
Transect 3, sampled on August 27, 2007, surveyed the Copper River outflow
region in the northern GoA. The transect started at 59.72°N, 144.89°W and moved
initially in a northeast direction along Kayak Island to 59.97°N, 144.49°W. From
there the transect moved in a west-northwest direction along the inner shelf offshore
of the mouth of the Copper River to 60.31°N, 146.13° W. Finally, the transect
headed in a south-southeast direction to 60.13°N, 145.91°W. Winds were blowing
from the east-southeast at ~15-20 knots during the day prior to the start of the
transect.
Temperature and salinity for the first third of Transect 3 were quite uniform at
13.5-13.8°C and 28.0-28.2, respectively (Figure 4). Silicic acid concentrations were
67
~9-10 µM and dissolved Al concentrations were 160-180 nM over this interval. TSS
concentrations were quite consistent at 3-4 mg l-1 (data not shown). LP Al
concentrations ranged from ~20-75 nM while TP Al concentrations ranged from
~230-340 nM. Salinity decreased significantly in a low salinity Copper River plume
from ~ 28 to a minimum value of ~ 19.2 at 60.19°N, 145.40°W (Julian Day 240.94 in
Figure 4). Dissolved Al increased to concentrations in excess of 600 nM while silicic
acid concentrations increased to ~22 µM. TSS concentrations increased to ~12.5 mg
l-1(data not shown). LP Al and TP Al concentrations increased to ~720 nM and
12,000 nM, respectively.
After exiting the low salinity plume surface waters, salinity increased to ~27.5
with corresponding decreases observed in dissolved Al, LP Al, TP Al, silicic acid,
and TSS concentrations. A second low-salinity plume region with minimum
salinities (~ 19.5) found near 60.29°N, 146.08°W (Julian Day 241.05 in Figure 4) was
observed with increases in dissolved Al and silicic acid to > 600 nM and ~21 µM,
respectively (Figure 4). These values were nearly identical to those observed in the
first low-salinity region sampled. Interestingly, leachable particulate Al and total
particulate Al only increased to 330 nM and 5000 nM, respectively, compared to
values of ~720 nM and 12,000 nM in the first low-salinity region. Similarly, the TSS
concentrations increased to only ~7.5 mg l-1 in this low-salinity plume while
increasing to ~12.5 mg l-1 in the first low-salinity plume. Once out of the second-low
salinity plume surface water, significant decreases were observed in dissolved Al, LP
Al, TP Al, and silicic acid.
68
Soluble Al concentrations along Transect 3 were nearly identical to the
dissolved Al concentrations (Figure 4). In the higher salinity (~ 28.2) waters in the
first third of the transect soluble Al concentrations ranged from ~ 130 nM to 180 nM
while increasing to maximum values of 620 nM and 680 nM in the two lower salinity
(~19.2-19.5) plumes sampled. The average ratio of soluble Al to dissolved Al along
Transect 3 was 0.98. Thus, it appears that dissolved Al existed almost entirely in the
soluble (< 0.03 µm) fraction with virtually no Al in the colloidal (0.03 µm – 0.45 µm)
size fraction.
3.1.3 Transect 4
Starting at 59.82°N, 149.45°W along the inner shelf just offshore of Seward,
Alaska, Transect 4 followed the GAK (Seward) line which has been sampled by the
University of Alaska since 1970 (Stabeno et al., 2004). The transect ran from the
inner shelf in a southeast direction across the continental shelf and shelf break and
into the waters of the subarctic Alaskan gyre out to 58.15°N, 147.86°W. Winds were
variable the day before the transect started, ranging from light and variable winds to
moderate (~6-8 knots) from both the south and the north.
Initial temperature and salinity on the inner shelf were ~13.4°C and 25.4,
respectively, with this salinity being the lowest observed along the entire transect.
Transect maxima in dissolved Al (152 nM), LP Al (37 nM), and TP Al (450 nM)
were observed in these relatively lower salinity inner shelf surface waters (Figure 5).
69
No TSS samples were taken here. Silicic acid concentrations were quite low with
values less than 1 µM.
In mid-shelf surface waters salinity increased to values of ~ 29-31 and
temperature decreased slightly to ~ 12.8°C. A dramatic, rapid decrease in dissolved
Al was observed moving offshore across the shelf with concentrations of 2-4 nM
observed in the mid-shelf region (Figure 5). LP Al and TP Al concentrations also
decreased significantly to ~ 3-7 nM and ~ 40-90 nM, respectively, while silicic acid
concentrations increased to concentrations of 10-15 µM. TSS concentrations were ~
3-5 mg l-1.
In the HNLC subarctic Alaskan gyre surface waters offshore from the shelf
break, salinity became quite constant at ~32.5 (Figure 5). It should be noted that
dissolved Fe concentrations in these waters were very low (~50-80 pM; Lohan, pers.
comm.) while nitrate concentrations were ~9 µM. Silicic acid concentrations in these
waters were ~10-12 µM. Dissolved Al concentrations in these waters were extremely
low (< 0.1 nM) and LP Al and TP Al concentrations had decreased to ~3 nM and 10-
12 nM, respectively. TSS concentrations decreased to 1.7-2.7 mg l-1.
3.1.4 Transect 5
Starting at 59.37°N, 152.52°W in lower Cook Inlet north of Kodiak Island,
Transect 5 ran in a southerly direction followed by a southeasterly direction through
Kennedy-Stevenson entrance, which separates the waters of Cook Inlet from the open
70
continental shelf waters, to 58.95°N, 151.72°W. The transect then ran in a
northeasterly direction to an ending location of 59.19°N, 150.83°W off the Kenai
Peninsula just south of Gore Point. Winds were from the south-southeast at 5-10
knots during the day prior to the start of the transect.
Initial temperatures in lower Cook Inlet and Kennedy-Stevenson entrance
were relatively cool (9-11°C) while salinities were elevated at 30.5-31.7 (Figure 6).
Silicic acid concentrations in the relatively cooler surface waters were also elevated at
~ 16-20 µM. Dissolved Al concentrations in these cooler, nutrient rich surface waters
were ~ 3-6 nM (Figure 6). Both LP Al concentrations and TP Al concentrations in
these higher salinity (31.4-31.8), silicic acid-rich (16-20 µM) waters were variable,
ranging from 5-10 nM and 38-110 nM, respectively.
After leaving lower Cook Inlet and passing through Kennedy-Stevenson
Entrance heading northeast towards Gore Point temperatures began to increase
significantly and salinities began to decrease, indicative of warmer, fresher coastal
surface waters and the ACC. Temperatures increased to maximum values of ~ 13°C
while salinities decreased to minimum values of ~ 28.2. Silicic acid concentrations
decreased to minimum values of ~ 0.7 µM in these warmer, fresher coastal waters
with corresponding increases in dissolved Al concentrations to maximum values of ~
27-30 nM. Narrow bands of increased salinity, elevated silicic acid, and decreased Al
were observed in this region (Julian day 245.98-246.00; Figure 6), possibly indicative
of mixing between the warmer, fresher coastal waters and the deep-mixed, upwelled
waters found in lower Cook Inlet and Kennedy-Stevenson Entrance.
71
3.1.5 Transect 6
Starting at 59.86°N, 149.72°W just off the Aialik Glacier in Aialik Bay southeast
of Seward, AK, Transect 6 was sampled on September 3, 2007 prior to a personnel
exchange in Seward. The transect was relatively short and ended on the GAK line
(Transect 4) at 59.67°N, 149.30°W. The influence of the Aialik Glacier was evident
as the initial temperatures and salinities were somewhat low at ~ 10.4°C and 23.1.
Dissolved Al and LP Al in these lower salinity waters in close proximity to the
glacier were elevated at ~ 210 nM and ~ 380 nM while silicic acid concentrations
were ~ 2.0-2.5 µM. As the ship moved away from the influence of the glacier,
salinity increased significantly to maximum values of ~ 30.6. Silicic acid
concentrations in these surface waters decreased to < 1 µM while dissolved Al values
decreased to 60-80 nM (Figure 7). As the transect approached the GAK line
dissolved Al values decreased to ~3-5 nM and LP Al decreased to ~ 5 nM.
3.2 Water Column Dissolved and Particulate Al in the Shelf Waters
While the focus of this manuscript is on the surface water dissolved and
particulate Al, samples collected at select vertical profile stations over the GoA shelf
are presented to support interpretation of the surface water data. Transect 2 consisted
of a series of vertical profile stations (both CTD casts and Go-Flo sampling) starting
from the inner shelf near Yakutat (YAK1) and extending across the shelf and into the
waters of the Alaskan subarctic gyre (Figure 2). Vertical profiles of temperature,
72
salinity, silicic acid, dissolved Al, LP Al, and TP Al for shelf Stations 1, 2, and 3
(YAK1, YAK2, YAK3) along Transect 2 are shown in Figure 8. YAK1, YAK2, and
YAK3 were located at the inner-shelf, mid-shelf, and outer-shelf just at the shelf
break, respectively.
The salinity profile at YAK1 compared to YAK 2 and YAK3 (which look very
similar) shows the influence of the lower salinity ACC along the inner shelf not just
in the near-surface waters but in the deeper waters as well (Figure 8B). For example,
at a depth of 50 m salinity at YAK1 is 31.9 while it increases to ~ 32.2 at YAK2 and
YAK3. Silicic acid concentrations at YAK1 are somewhat higher in the water column
than at YAK 2 and YAK3, particularly from ~ 50m to 100m (Figure 8C). Significant
decreases in dissolved Al, LP Al, and TP Al are observed in moving from inner shelf
YAK1 to the mid- and outer-shelf stations (Figure 8D, 8E, and 8F). At 50 m depth
dissolved Al concentrations decrease from ~ 13 nM at YAK 1 to ~ 3 nM at YAK 3
while TP Al concentrations decrease from 700 nM to ~ 50 nM.
Three vertical profile stations were occupied along Transect 4 (GAK line)
beginning with nearshore, innershelf Station 1 (GAK-A) and ending with offshelf
Station 3 (GAK-C) which sampled HNLC deeper waters of the subarctic Alaskan
gyre. Vertical profiles of temperature, salinity, silicic acid, dissolved Al, LP Al, and
TP Al for Transect 4 (GAK line) stations GAK-A, GAK-B, and GAK-C are shown in
Figure 9. The temperature profiles show warmer subsurface (~30-110m) waters at
innershelf GAK-A relative to GAK-C, likely a result of the cooler, upwelled waters
of the Alaskan gyre (Figure 9A) while the salinity profiles clearly show the influence
73
of the lower-salinity ACC from the surface to ~ 50 m depth at GAK-A (Figure 9B).
Subsurface dissolved Al concentrations range from ~ 2 nM to 13 nM at GAK-A and
GAK-B with a maximum value found at 10 m depth at nearshore GAK-A. Dissolved
Al, LP Al, and TP Al concentrations at offshelf GAK-C are significantly lower than
concentrations observed at GAK-A and GAK-B (Figure 9D, 9E, and 9F).
Finally, two vertical profile stations designated Kodiak Source Water (KSW)
stations were sampled on the continental shelf northeast of Kodiak Island (KSW1)
and in the waters of the lower Cook Inlet/Shelikof Strait (KSW2; see Figure 2).
Vertical profiles of temperature, salinity, dissolved Al, LP Al, and TP Al are shown in
Figure 10. Water column temperature was slightly lower and water column salinity
was higher at KSW1 relative to KSW 2 (Figures 10A and 10B). However, the
influence of the lower-salinity ACC was not observed at either station. With the
exception of a single data point at 25m at KSW1 (dissolved Al ~9 nM) water column
dissolved Al values were quite consistent at both stations with concentrations of ~1-3
nM. No significant variation with depth was observed (Figure 10D). Both LP Al and
TP Al were relatively low in the upper water column (0-100m) with increases in
concentration near the bottom indicative of a benthic nepheloid layer (Figure 10E and
10F).
74
4. Discussion
4.1 Comparison with Previous Data
The dissolved and particulate Al data reported here are the first reported direct
measurements of Al in the northern and central GoA. Thus, there is no Al data for
direct comparison in the region. The data do show tremendous variation in moving
from coastal surface waters into surface waters of the Alaskan subarctic gyre.
The background dissolved Al in coastal waters along Transect 1 and Transect
3 was elevated at ~20-70 nM (salinity ~ 30.4-31.3) and 160-180 nM (salinity ~28.0),
respectively. In comparison to another shelf region dominated by fluvial input,
Brown and Bruland (2009) report much lower background dissolved Al values of ~4-
10 nM at salinities of ~30 in near-field Columbia River plumes off the Oregon coast.
Ren et al. (2006) reported dissolved Al concentrations of ~30-60 nM in low-salinity
(23.0-27.0) shelf waters of the East China Sea dominated by Yangtze river discharge,
In the most nearshore sample of a survey of Arctic Ocean surface waters, Measures
(1999) observed a dissolved Al concentration of ~22 nM. Maxima in dissolved Al in
the Arctic interior were not correlated with fluvial input which has been shown to be
relatively insignificant to the region (Kenison Falkner et al., 1997). It appears that
dissolved Al concentrations in the GoA coastal waters are quite elevated compared to
other coastal shelf regions. In addition, the dissolved Al maxima in the low salinity
plume waters along Transects 1 and 3 are significantly greater than those reported by
Brown and Bruland (2009) in Columbia River plumes off the coast of Oregon and
Washington. Maximum dissolved Al concentrations of ~1.3 µM and ~500 nM were
75
observed along Transect 1 at salinities of ~7 and 19.2, respectively. Maximum
dissolved Al reported by Brown and Bruland (2009) in the low salinity (~12) near-
field Columbia River plume waters was only ~100 nM.
In a survey of the distribution and transport of suspended material in the
northern Gulf of Alaska, Feely et al. (1979) observed particle concentrations of 1.5 -
6.7 mg l-1 in plumes of turbid water extending offshore from the northern GoA
coastline in the Copper River region and just west of Kayak Island. The source of the
turbid water was attributed to be discharge of sedimentary material from the Copper
River and from coastal streams draining various glaciers east of Kayak Island.
Particle analysis indicated mostly inorganic material of terrestrial origin (Feely and
Cline, 1977). Landing and Feely (1981) measured 8.2 wt % Al in a composite of
three suspended particulate material samples from northern GoA. Assuming 8.2 % Al
in northern GoA particulate material, 1.5-6.7 mg l-1 equates to 4600-20,000 nM total
particulate Al . This estimate is in good agreement with the TP Al concentrations
observed in the three low salinity plumes in the ACC off Yakutat (Transect 1) and the
Copper River (Transect 3) where total maximum particulate Al concentrations ranged
from 4800 nM to 25,000 nM. It is clear that the coastal discharge of the northern GoA
delivers tremendous amounts of particulate Al to coastal waters.
The particulate Al concentrations from the northern GoA surface waters show
more similarity to the Oregon coast shelf waters than do the dissolved Al
concentrations. In low-salinity, near-field Columbia River plume surface waters
sampled in summer 2004, Berger et al. (2008) measured LP Al concentrations of 220-
76
920 nM and TP Al concentrations of ~ 2,000-8,700 nM at salinities of 21.2-26.4.
Brown and Bruland (2008) measured LP Al concentrations of ~ 440-720 nM Al and
TP Al concentrations of ~ 8000-12,000 nM in near-field Columbia River plume
surface waters in summer 2006. In comparison, LP Al concentrations in the low
salinity plume sampled along Transect 3 (salinity ~ 20.4) were 720 nM and TP Al
concentrations were 12,000 nM.
A drastic decrease in water column and surface water Al is observed in
moving across the shelf from the nearshore northern GoA coastal waters offshore into
the Alaskan subarctic gyre. In moving from the inner shelf near Yakutat towards the
shelf break along Transect 2 (Figure 8) subsurface dissolved Al concentrations in the
water column decreased from inner shelf values of 7-16 nM (YAK1) to outer shelf
values of 1-2 nM (YAK3). A similar dramatic subsurface gradient in both LP Al and
TP Al was observed in moving across the shelf off Yakutat. Surface water dissolved
Al concentrations along Transect 4 (GAK line) decreased from values greater than 50
nM along the innershelf to concentrations less than 0.2 nM along the outer GAK line
(Transect 4) in the HNLC surface waters of the GoA. These are some of the lowest
surface water dissolved Al concentrations reported in the world oceans. Similar
minimal dissolved Al values (~ 0.3 nM) were observed between 70°S and 45°S
latitude along 150ºW longitude in the South Pacific during the CLIVAR P16S repeat
hydrography cruise (Measures and Landing, pers. comm.) The extremely low
dissolved Al values in these waters are attributed to a lack of eolian dust input to the
region. Along transects between California and Hawaii, Orians and Bruland (1986)
77
and Johnson et al. (2003) observed surface water dissolved Al and dissolvable Al of ~
0.3-0.7 nM and ~0.5 nM, respectively, in California Current surface waters. In this
case, the dissolvable Al includes the dissolved Al fraction plus a portion of the weak
acid LP Al(samples were filtered only through a 20 µM pre-filter and pH adjusted in-
line to 5.5 prior to analysis; Johnson et al., 2003). The low concentrations of
dissolved Al within the California Current system are attributed to enhanced particle
scavenging by sinking biogenic particles which provide surface adsorption sites for
dissolved Al, transporting it out of the surface water (Orians and Bruland, 1986). It is
likely that the extremely low dissolved Al concentrations in the subarctic gyre waters
of the northern GoA are due to a combination of intense scavenging of fluvially-
derived dissolved Al in the particle-rich coastal waters and a lack of significant
atmospheric dust input to the region.
4.2 Dissolved Al in the Low Salinity Coastal Surface Waters
Property-salinity plots for dissolved Al and silicic acid are shown in Figure 11
for the low-salinity plume water encountered along Transect1 Plume A, Transect 3,
and Transect 6. For both dissolved Al and silicic acid in Transect 1 Plume A,
Transect 3, and Transect 6, a very linear relationship with salinity is observed
indicating conservative behavior and control largely by physical mixing of the high
silicic acid, high dissolved Al freshwater endmember with a relatively low dissolved
Al, low silicic acid coastal endmember (Figure 8). Extrapolation of the mixing lines
to zero salinity for Transect 1 Plume A and Transect 3 yield average dissolved Al and
78
silicic acid freshwater endmember concentrations of ~ 1770 nM and 43 µM,
respectively, for the August 2007 sampling period. The extrapolation of the dissolved
Al data from the low-salinity plume near Yakutat Bay and the Alsek River to a
theoretical zero-salinity endmember (1792 nM; Transect 1 Plume A) is remarkably
similar to that of the Copper River (1740 nM; Transect 3) while the silicic acid
extrapolated zero-salinity endmember is 39 µM and 48 µM, respectively, for Transect
1 Plume A and Transect 3. The average extrapolated dissolved Al zero-salinity
endmember value of 1770 nM observed in late August 2007 is slightly higher than the
1100-1400 nM dissolved Al concentrations observed in the two grab samples from
the Copper River collected September 24, 2007. Average world riverine
concentrations of dissolved Al and silicic acid are 1480±740 nM Al and 150 µM
H4SiO4 (Meybeck, 1988; Treguer et al., 1995). Thus, it appears that the freshwater
sources of the two low salinity plumes measured on Transect 1 Plume A and Transect
3 are comparable to world rivers in terms of dissolved Al yet are relatively silicic acid
poor.
Anders et al. (2003) measured dissolved (0.45 µm-filtered) Al and silicic acid
at a number of sites within the Copper River basin of south central Alaska in July
2000. Close to the mouth of the Copper River, an average dissolved Al concentration
of 2500±400 nM and an average silicic acid concentration of 63±2 µM was observed
from nine sampling sites. Although the dissolved Al and silicic acid concentrations
observed by Anders et al. (2003) in the lower Copper River are greater than either the
79
Copper River dissolved values we observed in September 2007 or the theoretical
zero-salinity endmember dissolved Al or silicic acid concentrations derived from the
mixing lines in Figure 11 from late August 2007, the Si:Al (mol:mol) ratios from the
two different sampling periods for the freshwater endmember are remarkably
consistent at 24.3-25.2. This consistency in the Si:Al ratios is indicative of a common
source of Si and Al to the freshwaters of the region. The linear relationship of
dissolved Al and silicic acid with respect to salinity and the consistency of the Si:Al
ratio of the freshwater end member and the plume waters both point to the lack of
dissolved Al removal during active mixing and formation of the plume. It appears
that for silicic acid and dissolved Al there is simple conservative mixing between the
river and coastal seawater to form the Alsek River and Copper River plumes.
Biotite, a hydrous potassium aluminosilicate
[K(Mg,Fe)3(Al,Fe)Si3O10(F,OH)2], has been shown by many researchers to be the
dominant silicate mineral contributing to the dissolved load of subglacial streams
even when it is only a minor constituent of the associated bedrock (Blum et al., 1994;
Axtmann and Stallard, 1995; Anderson et al., 2000). Nesbitt and Young (1996)
suggest that silicate weathering in subglacial environments is dominated by biotite
weathering because biotite is susceptible to physical damage by glacial erosion and
abrasion leaving fresh mineral surfaces to be weathered. At the Bench Glacier in
south central Alaska Anderson et al. (2000) found biotite to be the dominant silicate
phase to chemically weather.
80
In a laboratory study of biotite dissolution, Malmstrom and Banwart (1997)
show a strong pH dependence on the stoichiometry of biotite dissolution. Under
acidic conditions, a strong preferential release of mineral framework metal ions
(Mg,Fe, and Al) was observed relative to Si. At pH 2, steady-state rates of biotite
dissolution (moles m-2 h-1) of Si, Al, and Fe were 10-6.2, 10-6.2, and 10-6.1, respectively.
However, in moving toward the near-neutral pH region, there was a transition to
preferential release of Si. At a pH of 8, similar to that of the Copper River (Anders et
al., 2003), steady-state rates of biotite dissolution for Si, Al, and Fe were as follows: ~
10-7.0 moles Si m-2 h-1, 10-8.0 moles Al m-2 h-1, and 10-8.75 moles Fe m-2 h-1.
The dissolution of biotite is described as a multi-site reaction where each
framework element is released independently of the others (Guy and Schott, 1989).
Malmstrom and Banwart (1997) concluded that biotite was weathered incongruently
to vermiculite during the weathering process. In moving from pH ~2 towards the
near-neutral pH region, the conditional rate constant (moles biotite h-1 moles-1) for
release of Al in the incongruent dissolution of biotite decreases from 10-3.0 to 10-5.0
while that for Fe release decreases from 10-2.3 to 10-6.0. The conditional rate-constant
(moles biotite h-1 moles-1) for Si dissolution decreases from ~ 10-3.5 to ~ 10-5.0. It
should be noted that in the near-neutral pH region the conditional rate constants of
metal release for incongruent dissolution of biotite are at a minimum for both Al and
Fe (Malmstrom and Banwart, 1997). Based on the given biotite dissolution rates,
Si:Al ratios are expected to be ~ 10:1. However, removal of released ions through re-
81
adsorption of ions into the primary mineral phase or secondary mineral formation will
affect the observed Si:Al ratios.
Malmstrom and Banwart (1997) specifically noted that Fe was normally
retained in the biotite solid phase as compared to all other ions during the laboratory
experiments. While significant increases in both dissolved Al and silicic acid were
observed in low salinity plumes along Transects 1 and 3, it is worth noting that
dissolved Fe concentrations remained remarkably constant at 2-3 nM with no
appreciable increases in concentration in the freshwater plumes (Lohan, pers. comm.;
Lippiatt et al., in prep.). It is likely that the trends observed in dissolved Al, Fe, and
silicic acid concentrations in the zero-salinity river endmember that mixes to form the
low-salinity plumes in the coastal northern GoA are a result of biotite weathering,
dissolution, and secondary mineral formation within the weathering basin.
Another point related to the dissolved Al concentrations in the low salinity
plume surface waters is that there is a distinct difference in the zero-salinity
endmember for dissolved Al and silicic acid concentrations between Transect 6,
which sampled the fresh glacial melt of Aialik Glacier southeast of Seward, and
Transects 1 and 3 which sampled river runoff in the Copper River region and
eastward, likely originating from glacial melt but traveling over vegetation, soil, and
bedrock in the glacial melt/riverine delivery to the ocean. The theoretical zero-salinity
endmember dissolved Al and silicic acid concentrations for the low salinity plumes
along Transects 1 and 3 are 1770 nM and ~43 µM, respectively (Si:Al = 24.3). The
zero-salinity endmember values for Transect 6 were significantly lower at 900 nM
82
dissolved Al and 15 µM silicic acid (Si:Al = 16.7; Figure 11). These differences are
reflective of the chemical weathering and precipitation reactions that takes place as
fresh glacial meltwater travels across soil, vegetation, and bedrock prior to
discharging into the ocean. Similarly, Anderson et al. (2000) noted an increase in
silicate denudation rates with increasing distance from a glacier due to the
establishment of vegetation causing an increase in silicate weathering.
4.3 A Contrast in Silicic Acid in Two Low Salinity Plumes
Property-salinity plots for dissolved Al and silicic acid within the two low-
salinity plumes along Transect 1 (T1 Plume A and T1 Plume B) are shown in Figure
12. Silicic acid along T1 Plume A showed a linear relationship with salinity
indicative of physical control of silicic acid distributions within the plume by dilution
of the relatively silicic acid-rich river endmember with a lower silicic acid coastal
endmember. T1 Plume B showed drastically different distributions with silicic acid
concentrations decreasing in the low salinity plume waters. At a salinity of ~20,
silicic acid concentrations in the low salinity waters of T1 plume B were roughly ~6
µM. The mixing line along T1 Plume A predicts silicic acid concentrations of ~22
µM at a salinity of 20. Thus, there appeared to be a silicic acid drawdown of ~15 µM
in the low salinity waters of T1 Plume B. It should be noted that nitrate values in the
low salinity plume waters of T1 Plume A increased from < 0.05 µM to greater than 2
µM while maximum nitrate values in T1 Plume B showed were only ~0.07 µM.
83
It could be argued that T1 plume B is an “aged” plume coming from the same
source as T1 plume A. However, the temperatures of the low-salinity plume core on
both T1 Plume A and T1 Plume B are both ~ 9°C while background temperatures
along Transect 1 are 15-16.5°C. If T1 plume B were an “aged” plume coming from
the same source as T1 Plume A, it could be argued that the temperature of the plume
would warm after mixing with warmer background surface waters. Alternatively, the
source region of T1 Plume B is fed directly by the Bering, Guyot, and Malaspina
glaciers while the source of T1 Plume A is the Alsek River. It could be that these
glacial sources were responsible for T1 Plume B and therefore delivered lower
concentrations of silicic acid relative to dissolved Al as was observed in Transect 6.
It could also be that the zero-salinity riverine endmember of T1 Plume B was initially
cooler than that of T1 Plume A and the agreement of the temperatures in the two
plumes is coincidental as both T1 Plume A and T1 Plume B warmed upon mixing
with warmer coastal surface waters.
Underway fluorescence and biogenic silica (BSi) data along Transect 1 are
presented in Figure 13. The fluorescence data showed significant increases in the
region from 141°W longitude to 142.5°W longitude, which correspond well with the
location of T1 Plume B. Concomitant increases in BSi from background values of ~
0.2 µM to 6-10 µM were observed. These increases in BSi are somewhat lower in
magnitude than the ~ 15 µM silicic acid removal observed in T1 Plume B. However,
it is evident that a diatom bloom was occurring in the low-salinity waters of T1 Plume
B. It is likely that a combination of both lower silicic acid: dissolved Al ratios due to
84
the glacial source of T1 Plume B and then further silicic acid assimilation by biota
lead to the low silicic acid concentrations in T1 Plume B.
Gehlen et al. (2002) provided evidence for the incorporation of Al during
siliceous frustule biosynthesis. In laboratory diatom cultures, the authors reported
maximum Al:Si ratios of ~ 7.00 x 10-3 while similar ratios were observed in two
natural marine diatom assemblages. Figure 12 also shows dissolved Al-salinity
relationships for T1 Plume A and T1 Plume B. At a salinity of ~ 20, the mixing line
for T1 Plume A where no silicic acid drawdown was observed yields a dissolved Al
concentration of ~ 600 nM. It is interesting to note that in the T1B plume core
waters, where a significant increase in BSi was observed, two samples at a salinity of
~20 show dissolved Al concentrations of ~ 500 nM, 100 nM less than what would be
predicted from the mixing line of T1 Plume A. Utilizing a potential Al removal of ~
100 nM with a silicic acid removal of ~ 10-15 µM yields Al:Si (mol:mol) uptake
ratios of ~ 6.7x10-3 to 10x10-3, very similar to the Al:Si (mol:mol) ratios observed by
Gehlen et al. (2002). While it is realized that two data points do not give substantial
evidence of biological Al uptake, the data is consistent with biological Al uptake
occurring in these surface waters.
4.4 Surface Transects and Relationships Between Total Suspended Solids, Particulate
Al, and Dissolved Al
The surface transects (Transects 1, 3, 4, 5, and 6) can be divided into three
groups based on differing hydrography. Transect 1 off Yakutat Bay, Transect 3 off
85
the Copper River, and Transect 6 off the Aialik glacier in Aialik Bay sampled lower
salinity coastal surface waters dominated by fluvial/glacial input with significantly
increased dissolved and particulate Al concentrations. Transect 4 (along the GAK
line), once further offshore from the influence of coastal runoff, largely sampled
midshelf and offshelf surface waters characterized by relatively higher salinities and
drastically decreased dissolved Al and particulate Al concentrations. Transect 5 near
the southern tip of the Kenai Peninsula sampled the cooler, higher salinity surface
waters of lower Cook Inlet brought to the surface by intense tidal mixing at the
Kennedy Entrance. Similarities and differences in dissolved and total particulate Al
as well as TSS concentrations from these three different groups will be discussed.
Of all the surface waters sampled in this study, dissolved Al, LP Al, and TP
Al were at maximum values in the coastal low salinity plume waters of Transects 1
and 3 in the ACC along the northeast boundary of the Gulf of Alaska (dissolved Al >
1200 nM, LP Al > 1700 nM, and TP Al > 20000 nM along Transect 1 at a salinity <
7). In these low-salinity waters the TSS concentrations (14-19 mg l-1) were also at a
maximum for the study, indicative of the significant input of lithogenic material from
the mountainous coastline to the coastal waters. All of these parameters decrease
significantly along Transects 1 and 3 at higher salinities where the freshwater
endmember has been diluted with a higher salinity, low Al coastal seawater
endmember. It is reiterated that the dissolved Al concentrations as well as LP Al
concentrations in the low salinity glacial melt waters of Transect 6 were significantly
lower than those observed in plume waters of Transects 1 and 3 (discussed in Sec.
86
4.2). TSS concentrations in the coastal seawater strongly influenced by the glacial
melt water were only ~ 4 mg l-1.at a salinity of ~ 11.5.
Percent leachable particulate Al (% LP Al; calculated as (LP Al/TP Al)x
100%) averaged 7.4±1.1% (n =4) in the low salinity plumes along Transects 1 and 3
(Tables 1 and 2). This average value agrees very well with a % LP Al of 7.3±0.6%
reported by Brown and Bruland (2008) in a near-field low salinity Columbia River
plume off the Oregon coast. The relatively low % LP Al values are indicative of the
relatively refractory nature of particulate Al in the weathered continental material
being delivered to the coastal waters. The refractory Al portion consists of Al bound
in mineral lattices in crustal material while the leachable portion is likely Al-
hydroxide coatings on particle surfaces or Al associated with biogenic material
(Berger et al., 2008).
With the exception of a narrow band of relatively lower-salinity (~25) surface
water with elevated dissolved Al, LP Al, and TP Al water just offshore along
Transect 4 (GAK line), the midshelf and outershelf surface waters of Transect 4 were
relatively higher salinity (~ 30.0-32.5) waters with drastically lower concentrations of
dissolved Al, LP Al, and TP Al. This same variation was observed in the water
column values of dissolved Al, LP Al, and TP Al at vertical profile station GAK-A
relative to GAK-C with significantly higher concentrations at GAK-A relative to
GAK-C (Figure 9). Dissolved Al, LP Al, and TP Al concentrations were ~ 2-3 orders
of magnitude lower in the high salinity surface waters of Transect 4 relative to the
low salinity plume coastal waters of Transects 1 and 3, likely a result of the extreme
87
scavenging of the high Al signal in the coastal waters of the northern GoA. TSS
concentrations were only ~ 20% of that observed in the low salinity waters along
Transects 1 and 3. The decreases in TSS, LP Al, and TP Al are likely due to particle
settling as the particle-rich, lower salinity coastal waters move away from their
source. The drastic decrease in dissolved Al is very likely due to adsorption of Al
onto particle surfaces and removal of Al as these particles sink out.
As mentioned previously, in comparison to central Pacific oligotrophic gyre
surface waters, the residence time of Al in waters of the subarctic Alaskan gyre is
suggested to be significantly less (3-4 years vs. ~ 35 days; Orians and Bruland, 1986).
In particle-rich northern GoA coastal shelf waters, it is argued here that dissolved Al
might be removed even more rapidly, reflected in the extreme decreasing gradient of
dissolved Al between the low-salinity coastal waters and the higher salinity midshelf
and outershelf waters where some of the lowest reported dissolved Al concentrations
in the world ocean have been observed. It is worth noting that the % LP Al along
Transect 4 (Table 3) increases to values of ~ 22%-44%, significantly greater than %
LP Al in Transect 1 and 3 low-salinity plume waters. This increase in the % LP Al
fraction is suggestive of removal of dissolved Al onto particle surfaces. Utilizing the
water column Al data from the shelf vertical profile stations (YAK1, YAK2, YAK3,
GAK-A, GAK-B, KSW1, and KSW2) an average water column dissolved Al
concentration of ~ 4 nM can be estimated. Using a freshwater input of 2.4 x 104 m3 s-
1 (Brabets, 1997), a zero-salinity endmember Al concentration of 1700 nM, and a
continental shelf length, width, and depth, respectively, of 1500 km, 50km, and 100
88
m, an average residence time of dissolved Al in northern GoA shelf waters of ~ 10
days is estimated. While this calculation contains a degree of uncertainty, it clearly
shows that the removal of dissolved Al in these particle-rich coastal waters is
relatively rapid compared to oligotrophic gyre surface waters.
Figure 14 shows dissolved Al concentrations as a function of TSS
concentrations for Transects 1, 3, and 4. There is a remarkable degree of variability in
dissolved Al concentrations for a given TSS concentration. For example, at a TSS
concentration of ~ 4 mg l-1, dissolved Al concentrations from the three surface
transects range from > 600 nM along Transect 3 off the Copper River to < 1 nM
along Transect 4 along the GAK line, a difference of over two orders of magnitude.
At TSS concentrations of ~ 19-20 along Transect 1, dissolved Al concentrations
range from ~ 25 nM to 1200 nM. Thus, a clear relationship between dissolved Al and
TSS concentrations does not exist in northern GoA coastal waters.
The cooler, nutrient-rich waters of lower Cook Inlet and Kennedy Stevenson
Entrance sampled along the first half of Transect 5 (Figure 6) are indicative of deep,
persistent tidal mixing that occurs in the region, continually supplying nutrients to the
surface waters (Stabeno et al., 2004). Vertical shear measurements at Kennedy
Stevenson entrance to lower Cook Inlet showed mixing to depths greater than 50m.
This steady supply of nutrients to the surface waters of lower Cook Inlet makes it one
of the most productive high-latitude shelf regions in the world (Sambrotto and
Lorenzen, 1987). It follows then that the surface water dissolved Al concentrations in
lower Cook Inlet might be related to subsurface dissolved Al concentrations along the
89
shelf. Indeed, the relatively low ~ 2-6 nM dissolved Al surface water concentrations
in lower Cook Inlet (Figure 6) agree very well with subsurface dissolved Al values at
shelf vertical profile stations (KSW1 and KSW2, Figure 10) near Kodiak Island and
along Transect 4 (GAK-A and GAK-B, Figure 9). Surface water LP Al
concentrations in lower Cook Inlet (~ 4-15 nM) are relatively low and agree with
water column LP Al concentrations at GAK-A and GAK-B. The water column LP Al
concentrations at innershelf YAK1 are significantly greater, likely due to the close
proximity to the particle-rich freshwater source. The % LP Al in the upwelled surface
waters of lower Cook Inlet (Table 4) was variable at 8.1±2.8% (n=5) yet resembles
the same fraction as that observed in the low-salinity plume surface waters of
Transect 1 off Yakutat Bay and Transect 3 off the Copper River. It is possible that the
relatively low dissolved Al concentrations in the surface waters of lower Cook Inlet
are due to scavenging and removal of dissolved Al onto particle surfaces as both
biogenic and lithogenic particles sink from the surface waters to depth and the
subsequent mixing of this low dissolved Al deeper water back to the surface.
However, a corresponding increase in % LP Al indicative of Al scavenging might be
expected similar to Transect 4, yet this is not observed.
4.5 Dissolved Al vs. Soluble Al
As mentioned in Section 3.2, in the low salinity, high discharge coastal region
near the Copper River along Transect 3, nearly all of the dissolved Al was in the
soluble (< 0.03 um) fraction (Figure 4). This finding is in contrast to previous work
90
by Brown and Bruland (2009) that showed nearly 65% of dissolved Al in a low-
salinity (~ 15) plume off the Oregon coast near the Columbia River was in the
colloidal form, possibly existing as humic acid or fulvic acid complexes.
Interestingly, it was noted that in terms of dissolved organic carbon (DOC) content,
the Columbia River (~ 100 µM DOC) was relatively deplete in its concentrations of
DOC compared to other rivers.
In a study of a glacial estuary in Southeast Alaska, Loder and Hood (1972) observed
very low concentrations of DOC (~ 12 µM) in an outflow stream from a glacier while
concentrations in the glacial estuary ranged from ~ 12 µM to 40 µM. In a study of a
glacial river-floodplain system in Switzerland, Tockner et al. (2002) measured ~ 25
µM in glacial melt water. Thus, it appears that concentrations of DOC in glacial melt
are significantly lower that that of the Columbia River or other major rivers. The
amount of DOC in the Copper River outflow region will be determined not only by
glacial meltwater but by the vegetation over which the glacial melt flows. Anderson
et al. (2000) noted an increase in silicate denudation rates with increasing distance
from a glacier due to the establishment of vegetation which also could release humic
acids to the glacial melt. However, if the vegetation is sparse and not a major source
of DOC to the glacial/riverine runoff, DOC concentrations in the Copper River
outflow region would be quite low. It is argued here that very low DOC
concentrations and a subsequent lack of colloidal organic complexes in the runoff
waters lead to the dissolved Al in surface waters existing entirely in the soluble
91
fraction. It is suggested that low DOC concentrations in the region might also be
linked to the relatively low, invariant dissolved Fe concentrations observed in the
low-salinity plumes (Lippiatt et al., in prep.).
5. Conclusions
The large glacial melt/riverine discharge from the mountainous region of
central coastal Alaska provides a significant source of dissolved and particulate Al to
northern GoA coastal waters. With regards to other major world rivers, freshwater
sources to the northern GoA appear to be rich in dissolved Al yet low in silicic acid
relative to average river concentrations. Low salinity plumes in these coastal waters
were rich in dissolved Al, both leachable and total particulate Al, and total suspended
solids. The percent-leachable particulate Al fraction in these low salinity plumes was
quite low (~ 7%), indicative of the largely refractory nature of the lithogenic material
being delivered to the coastal waters. Si:Al ratios from different time periods point to
a common source for dissolved Al and silicic acid in the low-salinity plume waters
originating from the Copper and Alsek Rivers observed in Transects 1 and 3,
respectively (Si:Al ~24-25). However, the coastal surface waters strongly influenced
by fresh glacial melt along Transect 6 were unique with much lower silicic acid
relative to dissolved Al (Si:Al ~ 16.7). It is suggested that the major source of
dissolved Al and silicic acid in northern GoA coastal waters is the weathering of
biotite, which has been shown to be the predominant mineral to contribute to the
dissolved load in Alaskan subglacial streams. The relative distributions of silicic acid,
92
dissolved Al, and dissolved Fe (data not reported here) seem to agree with laboratory
studies on biotite weathering. Differences in dissolved Al and silicic acid in the
freshwater endmember from fresh glacial melt as opposed to glacial melt/riverine
runoff that has traveled over vegetation, soil, and bedrock. The dissolved Al in these
coastal waters appears entirely in the soluble (< 0.03 µm) fraction. This is likely a
consequence of the freshwater sources to the region being very low in dissolved
organic carbon and a lack of organic colloidal Al-complexes.
Concentrations of dissolved Al in outershelf, HNLC surface waters in the
northern GoA are some of the lowest reported in the worlds oceans. These low
dissolved Al concentrations are likely due to an extreme degree of particle scavenging
of the coastally-derived dissolved Al in the nearshore waters of the northern GoA
(supported by significant increases in the percent-leachable particulate Al fraction), as
well as a low amounts of atmospheric Al input to the region. Residence times of
dissolved Al in the northern GoA shelf waters are estimated to be on the order of 10
days.
Acknowledgements
We thank the captain and crew of the R/V Thomas G. Thompson for their
assistance on this research expedition. We deeply appreciate the efforts of Bettina
Sohst with the nutrient analyses, Geoffrey Smith with the sample collection, and
Michael Lawrence for providing the samples for soluble Al analysis. Finally, we
thank Andrew Schroth for his efforts in providing the grab samples from the Copper
93
River, Alaska. This work was supported by National Science Foundation grant OCE-
0526601 to Kenneth W. Bruland.
94
References
Aguilar-Islas, A.M., Bruland, K.W., 2006. Dissolved manganese and silicic acid inthe Columbia River plume: a major source to the California Current and coastalwaters off Washington and Oregon. Marine Chemistry, 101, 233-247.
Anders, A.M., Sletten, R.S., Derry, L.A., Hallet, B., 2003. Germanium/silicon ratiosin the Copper River Basin, Alaska: Weathering and partitioning in periglacial versusglacial environments. Journal of Geophysical Research, 108, F1, 6005, doi:10.1029/2003JF000026,
Anderson, S.P., Drever, J.I., Frost, C.D., Holden, P., 2000. Chemical weathering inthe foreland of a retreating glacier, Geochimica et Cosmochimica Acta, 64, 1173-1189.
Axtmann, E.V., Stallard, R.F., 1995. Chemical weathering in the South CascadeGlacier basin, comparison of subglacial and extra-glacial weathering, inBiogeochemistry of Seasonally Snow-Covered Catchments, Tonnessen, K.A. et al.,Eds. IAHS-AISH Publishers, 228, 431-439.
Berger, C.J.M., Lippiatt, S.M., Lawrence, M.G., Bruland, K.W., 2008. Theapplication of a chemical leach technique for estimating labile particulate aluminum,iron, and manganese in the Columbia River plume and coastal waters off Oregon andWashington. Journal of Geophysical Research, 113, doi:10.1029/2007JC004703.
Blum, J.D., Erel, Y., Brown, K., 1994. 87Sr/86Sr ratios or Sierra Nevada stream waters:implications for relative mineral weathering rates. Geochimica et CosmochimicaActa, 58, 5019-5025.
Brabets, T.P., 1997. Geomorphology of the lower Copper River, Alaska. USGeological Survey Professional Paper, 1581.
Brzezinski, M.A., Nelson, D.M., 1989. Seasonal changes in the silicon cycle within aGulf Stream warm-core ring. Deep-Sea Research 36, 1009–1030.
Brown, M.T., Bruland, K.W., 2008. An improved flow-injection analysis method forthe determination of dissolved aluminum in seawater. Limnology and OceanographyMethods, 6, 87-95.
Brown, M.T., Bruland, K.W., 2009. Dissolved and particulate aluminum in theColumbia River and coastal waters of Oregon and Washington: behavior in near-fieldand far-field plumes. Estuarine and Coastal Shelf Science,doi: 10.1016/j.ecss.2009.05.031.
95
Bruland, K.W., Lohan, M.C., 2003. Controls of trace metals in seawater, In: Treatiseon Geochemistry, Vol. 6, Elderfield, H., Holland, H., Turekian, K., eds.).Elsevier/Pergamon Press, 23-48.
Bruland, K.W., Rue, E. L., Smith, G.J., Ditullio, G.R, 2005. Iron, macronutrients, anddiatom blooms in the Peru upwelling regime: brown and blue waters of Peru. MarineChemistry, 93, 81-103.
Buck, C.S., Landing, W.M., Resing, J.A., Measures, C.I., 2009. The solubility anddeposition of aerosol Fe and other trace elements in the North Atlantic Ocean:Observations from the A16N/CLIVAR CO2 repeat hydrography section. MarineChemistry, doi:10.1016/j.marchem.2008.08.003, in press.
Duce, R.A., Tindale, N.W., 1991. Atmospheric transport of iron and its deposition inthe ocean. Limnology and Oceanography, 36, 1715-1726, 1991.
Feely, R.A., Cline, J.D., 1977. The distribution, composition, and transport ofsuspended particulate matter in the northeastern Gulf of Alaska, southeastern BeringShelf and Lower Cook Inlet. In: Environmental Assessment of Alaskan ContinentalShelf, Annual Report to Outer Continental Shelf Environmental Assessment Program,Environmental Research Laboratories, National Oceanic and AtmosphericAdministration, 13, 89-171.
Feely, R.A., Baker, E.T., Schumacher, J.D., Massoth, G.J., Landing, W.M., 1979.Processes affecting the distribution and transport of suspended matter in the northeastGulf of Alaska. Deep Sea Research, 26A, 445-464.
Gehlen, M., Beck, L., Calas, G., Flank, A. –M., Van Bennekom, A.J., VanBeusekom, J.E.E., 2002. Unraveling the atomic structure of biogenic silica: Evidenceof the structural association of Al and Si in diatom frustules. Geochimica etCosmochimica Acta, 66, 9, 1601-1609.
Guy, C., Schott, J., 1989. Multisite surface reaction versus transport control duringthe hydrolysis of a complex oxide. Chemical Geology, 78, 181-204.
Han, Q., Moore, J.K., Zender, C., Measures, C., Hydes, D., 2008. Constrainingoceanic dust deposition using surface ocean dissolved Al. Global BiogeochemicalCycles, 22, GB2003, doi:10.1029/2007GB002975.
96
Hurst, M.P., Bruland, K.W., 2007. An investigation into the exchange of iron andzinc between soluble, colloidal, and particulate size-fractions in shelf waters usinglow-abundance isotopes as tracers in shipboard incubation experiments. MarineChemistry, 103, 211-226.
Hydes, D.J., Liss, P.S., 1977. The behavior of dissolved Al in estuarine and coastalwaters. Estuarine and Coastal Marine Science, 5, 755-769.
Johnson, K.S., Elrod, V.A., Fitzwater, S.E., Plant, J.N., Chavez, F.P., 2003. Surfaceocean-lower atmosphere interactions in the Northeast Pacific Ocean gyre: Aerosols,iron, and the ecosystem response. Global Biogeochemical Cycles, 17, 2, 1063,doi:10.1029/2002GB002004.
Kenison Falkner, K., Church, M., Measures, C.I., LeBaron, G., Thouron, D., Jeandel,C., Stordal, M.C., Gill, G.A., Mortlock, R., Froelich, P., Chan, L.H., 1997. Minorand trace element chemistry of Lake Baikal, its tributaries, and surrounding hotsprings. Limnology and Oceanography, 42, 329-345.
Kramer, J., Laan, P., Sarthou, G., Timmermans, K.R., de Baar, H.J.W., 2004.Distribution of dissolved aluminum in the high atmospheric input region of thesubtropical waters of the North Atlantic Ocean. Marine Chemistry, 88, 85-101.
Landing, W.M., Feely, R.A., 1981. The chemistry and vertical flux of particles in theNortheastern Gulf of Alaska. Deep-Sea Research 28, 1, 19-37. 1981.
Loder, T.C., Hood, D.W., 1972. Distribution of organic carbon in a glacial estuary inAlaska. Limnology and Oceanography, 17, 3, 349-355.
Lohan, M.C., Bruland, K.W., 2006. The importance of vertical mixing for the supplyof nitrate and iron to the Columbia River plume: implications for biology. MarineChemistry, 98, 260-273.
Mackin, J.E., Aller, R.C., 1984a. Processes affecting the behavior of dissolvedaluminum in estuarine waters. Marine Chemistry, 14, 213-232.
Mackin, J.E., Aller, R.C., 1984b. Dissolved Al in sediments and waters of the EastChina Sea: Implications for authigenic mineral formation. Geochimica etCosmochimica Acta, 48, 281-297.
Malmstrom, M., Banwart, S., 1997. Biotite dissolution at 25°C: the pH dependence ofdissolution rate and stoichiometry. Geochimica et Cosmochimica Acta, 61, 14, 2779-2799.
97
Measures, C.I., 1999. The role of entrained sediments and sea ice in the distributionof aluminum and iron in the surface waters of the Arctic Ocean. Marine Chemistry,68, 59-70.
Measures, C.I., Landing, W.M., Brown, M.T., Buck, C.S., 2008. High-resolution Aland Fe data from the Atlantic Ocean CLIVAR-CO2 Repeat Hydrography A16NTransect: Extensive linkages between atmospheric dust and upper oceangeochemistry, Global Biogeochemical Cycles, 22, GB1005, doi:10.1029/2007/GB003042.
Meybeck, M., 1988. How to establish and use world budgets of riverine materials. pp.247-272. In Physical and Chemical Weathering in Geochemical Cycles, ed. by A.Lerman and M. Meybeck, Kluwer Academic Publishers, Dordrecht.
Morris, A.W., Howland, R.J.M., Bale, A.J., 1986. Dissolved aluminum in the Tamarestuary, southwest England. Geochimica et Cosmochimica Acta, 50, 189-197.
Nesbitt, H.W., Young, G.M., 1996. Petrogenesis of sediments in the absence ofchemical weathering: effects of abrasion and sorting on bulk composition andmineralogy. Sedimentology, 43, 341-358.
Nishioka, J., Takeda, S., Wong, C.S., Johnson, W.K., 2001. Size-fractionated ironconcentration in the northeast Pacific Ocean: distribution of soluble and smallcolloidal iron. Marine Chemistry, 74, 157-179.
Orians, K.J., Bruland, K.W., 1986. The biogeochemistry of aluminum in the PacificOcean. Earth and Planetary Science Letters, 78, 397-410.
Parsons, T.R., Maita, Y., Lalli, C.M., 1984. A Manual of Chemical and BiologicalMethods for Seawater Analysis. Pergamon Press.
Perdue, E.M., Beck, K.C., Reuter, J.H., 1976. Organic complexes of iron andaluminum in natural waters. Nature, 260, 418-420.
Ren, J.L., Zhang, J., Li, J.B., Yu, X., Liu, S., Zhang, E., 2006. Dissolved aluminum inthe Yellow Sea and East China Sea−Al as a tracer of Changjiang (Yangtze River)discharge and Kuroshio incursion. Estuarine, Coastal, and Shelf Science, 68, 165-174.
Royer, T.C., 1981. Baroclinic transport in the Gulf of Alaska Part II: A fresh waterdriven coastal current. Journal of Marine Research, 39, 2, 251-266.
98
Sambrotto, R.N., Lorenzen, C.J., 1987. Phytoplankton and Primary Production. In:Hood, D.W., Zimmerman, S.T. (Eds.), The Gulf of Alaska, US Department ofCommerce, USA, pp. 249-282.
Sholkovitz, E.R., 1976. Flocculation of dissolved organic and inorganic matter duringthe mixing of river water and seawater. Geochimica et Cosmochimica Acta, 40, 831-845.
Spotila, J.A., Buscher, J.T., Meigs, A.J., Reiners, P.W., 2004. Long-term glacialerosion of active mountain belts: Example of the Chugach-St. Elias Range, Alaska.Geology, 32, 501-504.
Stabeno, P.J., Bond, N.A., Hermann, A.J., Kachel, N.B., Mordy, C.W., Overland,J.E., 2004. Meteorology and oceanography of the Northern Gulf of Alaska.Continental Shelf Research, 24, 859-897.
Takayanagi, K., Gobeil, C., 2000. Dissolved aluminum in the Upper St. Lawrenceestuary. Journal of Oceanography, 56, 517-525.
Taylor, S.R., 1964. Abundance of chemical elements in the continental crust: a newtable. Geochimica et Cosmochimica Acta, 28, 1273-1285.
Tockner, K., Malard, F., Uehlinger, U., Ward, J.V., 2002. Nutrients and organicmatter in a glacial river-floodplain system. Limnology and Oceanography, 47, 1, 266-277.
Treguer, P., Nelson, D.M., Van Bennekom, A.J., DeMaster, D.J., Leynaert, A.,Queguiner, B., 1995. The silica balance in the world ocean: a re-estimate. Science,268, 375-379.
Upadhyay, S., Sen Gupta, R., 1995. The behavior of aluminum in waters of theMandovi estuary, West Coast of India. Marine Chemistry, 51, 261-276.
Walker, W.J., Cronan, C.S., Patterson, H.H., 1988. A kinetic study of aluminumadsorption by aluminosilicate clay minerals. Geochimica et Cosmochimica Acta, 52,55-62.
99
Table 1. Dissolved aluminum, leachable particulate (LP) aluminum, and totalparticulate (TP) aluminum data with corresponding hydrographic data for Transect 1near Yakutat in the northern Gulf of Alaska. For the dissolved Al and silicic aciddata, some data points were interpolated from the preceding and following data pointsif no data existed for a given particulate sample.
Latitude(°N)
Longitude(°W)
Salinity Temp.(° C)
SilicicAcid(µM)
Diss.Al
(nM)
LPAl
(nM)
TPAl
(nM)
%LPAl
58.756 138.179 30.92 16.10 13.6 28.4 35.9 121 29.758.912 138.479 29.85 15.97 13.8 84.2 99.6 1041 9.658.954 138.553 28.35 15.34 15.0 233.1 229.8 2761 8.359.029 138.699 18.35 11.46 22.2 795 1652 8632 7.759.096 138.826 7.44 9.12 32.8 1192 1760 25320 6.959.083 138.994 11.83 10.06 29.1 1151.6 1202 14162 8.559.058 139.262 24.86 15.37 20.5 421.9 159.6 1420 11.259.028 139.569 28.34 16.54 15.6 216.3 23.3 170 13.759.046 139.877 28.16 16.65 15.5 204.5 20.2 180 11.259.209 140.000 28.96 16.42 15.7 135.4 24.4 228 10.759.338 140.104 30.13 15.76 15.6 84.8 21.4 158 13.559.539 140.240 29.78 16.17 14.7 104.2 21.2 201 10.559.584 140.609 30.39 15.19 12.3 62.4 19.3 142 13.659.624 140.886 30.50 15.57 14.4 56.6 7.9 45.7 17.259.662 141.146 27.14 12.39 3.2 177.0 194.6 1486 13.159.753 141.451 30.34 14.53 11.9 87.3 63.6 379 16.859.837 141.740 25.93 11.20 3.3 359.8 396 3221 12.359.902 141.952 23.75 9.94 5.3 452 505 4827 10.559.804 142.278 31.11 15.20 13.0 72.0 10.1 56.8 17.759.666 142.617 31.07 15.55 13.3 43.54 17.2 150 11.459.971 143.163 31.29 15.29 13.6 18.10 7.1 35.7 19.8
100
Table 2. Dissolved aluminum, soluble aluminum, leachable particulate (LP)aluminum, and total particulate (TP) aluminum data with corresponding hydrographicdata for Transect 3 near the Copper River outflow in the northern Gulf of Alaska. Forthe soluble Al data, the symbol “--” means that no soluble aluminum data obtainedfor that sample.
Latitude(°N)
Longitude(°W)
Salinity Temp.(° C)
SilicicAcid(µM)
Diss.Al
(nM)
SolubleAl
(nM)
LPAl
(nM)
TPAl
(nM)
%LPAl
59.835 144.714 27.90 13.84 9.3 179.4 174.2 19.6 226.6 8.760.043 144.623 27.73 13.87 9.6 155.0 -- 74.2 333.2 22.360.164 145.172 25.93 14.16 12.2 313.2 -- 293 4092 7.260.194 145.435 20.83 13.36 19.8 631.5 620.6 725 11955 6.160.237 145.693 23.53 13.90 15.7 355 364.8 187 2568 7.360.280 146.029 20.19 14.04 19.5 600 -- 337 5062 6.760.271 146.085 20.11 14.03 20.0 617 -- 306 3674 8.360.157 145.959 28.65 14.21 10.5 120 116 38.8 217 17.9
Table 3. Dissolved aluminum, leachable particulate (LP) aluminum, and totalparticulate (TP) aluminum data with corresponding hydrographic data for Transect 4(GAK line) starting near Seward and heading southwest in the waters of the subarcticAlaskan gyre.The symbol “--" denotes that no particulate data was obtained for the given sample.
Latitude(°N)
Longitude(°W)
Salinity Temp.(° C)
SilicicAcid(µM)
Diss.Al
(nM)
LPAl
(nM)
TPAl
(nM)
%LPAl
59.804 149.443 26.18 13.25 0.33 118 37.3 451 8.359.744 149.381 28.47 13.08 1.3 25.5 18.3 307 5.959.709 149.341 30.61 13.07 7.0 2.9 11.2 71.4 15.759.623 149.257 30.22 13.15 5.0 3.2 8.9 61.9 14.359.558 149.205 29.90 13.31 1.7 2.9 7.3 42.2 17.359.417 149.066 30.98 12.75 14.0 2.7 -- 87.8 --59.296 148.943 31.92 12.82 14.4 1.2 3.4 15.8 21.659.099 148.748 31.20 12.31 15.1 2.8 12.8 29.7 43.158.973 148.623 31.40 12.85 15.6 3.0 3.3 12.1 27.458.715 148.366 32.31 13.67 15.1 0.52 3.9 8.9 44.358.423 148.078 32.44 13.19 11.1 0.13 -- -- --58.261 147.948 32.45 13.17 11.4 0.09 -- -- --58.180 147.882 32.45 13.13 10.5 0.08 -- -- --
101
Table 4. Dissolved aluminum, leachable particulate (LP) aluminum, and totalparticulate (TP) aluminum data with corresponding hydrographic data for Transect 5starting in lower Cook Inlet. The symbol “--" denotes that no data was obtained forthe given sample.
Latitude(°N)
Longitude(°W)
Salinity Temp.(° C)
SilicicAcid(µM)
Diss.Al
(nM)
LPAl
(nM)
TPAl
(nM)
%LPAl
59.355 152.514 30.94 10.39 12.5 4.2 15.6 305 5.159.242 152.491 30.97 10.11 13.2 4.7 10.1 107 9.559.160 152.496 31.44 9.58 17.9 3.0 4.7 39.2 11.959.055 152.246 31.72 8.83 20.3 4.1 5.9 106 5.658.992 151.939 31.30 9.84 13.6 3.8 9.7 113 8.659.006 151.516 29.95 11.94 7.7 6.9 6.6 87.0 7.659.095 151.207 28.49 13.05 2.0 16.7 12.9 127 10.159.184 150.854 28.48 13.04 1.0 18.9 23.7 -- --
Table 5. Dissolved aluminum and leachable particulate aluminum withcorresponding hydrographic data for Transect 6 starting just off the Aialik Glacier inAialak Bay southeast of Seward, AK. The symbol “--" denotes that no data wasobtained for the given sample.
Latitude(°N)
Longitude(°W)
Salinity Temp.(° C)
SilicicAcid(µM)
Diss.Al
(nM)
LPAl
(nM)
TPAl
(nM)
%LPAl
59.839 149.710 22.95 10.29 2.3 210 379 -- --59.757 149.693 25.81 12.80 0.4 120 104 -- --59.668 149.497 26.78 13.39 0.3 68 25.7 -- --59.630 149.293 29.94 12.85 0.9 3.8 4.6 -- --
102
Figure 1. Schematic of the general circulation in the Gulf of Alaska. The flow of theAlaska Current and the Alaskan Stream are denoted by solid arrows while the flow ofthe Alaska Coastal Current is denoted by dashed arrows. The region shaded in whiteis the 0-100 m depth contour. The contours shown are 100m, 250m, 500m, 1000m,and 4000m
103
104
Figure 2. Coastal transect map for the 2007 Eddy Cruise in the northern and centralGulf of Alaska. Coastal surface water transects are shown with the black lines whilevertical profile stations are given by white dots. The depth contour intervals shownare 100m, 250m, 500m, 1000m, and 4000m.
105
106
Figure 3. Temperature and salinity (top), dissolved aluminum and silicic acid(middle), and leachable particulate (LP) and total particulate (TP) aluminum alongTransect 1 near Yakutat in the northern Gulf of Alaska.
107
108
Figure 4. Temperature and salinity (top), dissolved and soluble aluminum and silicicacid (middle), and leachable particulate (LP) and total particulate (TP) aluminum(bottom) along Transect 3 in the Copper River outflow region of the northern Gulf ofAlaska.
109
110
Figure 5. Temperature and salinity (top), dissolved aluminum and silicic acid(middle), and leachable particulate (LP) and total particulate (TP) aluminum (bottom)along Transect 4 (GAK line or Seward line) in the northern Gulf of Alaska.
111
112
Figure 6. Temperature and salinity (top), dissolved aluminum and silicic acid(middle), and leachable particulate (LP) and total particulate (TP) aluminum (bottom)along Transect 5 in and near Cook Inlet in the northern Gulf of Alaska.
113
114
Figure 7. Temperature and salinity (top) and dissolved aluminum, leachableparticulate (LP) aluminum, and silicic acid (bottom) along Transect 6 starting in closeproximity to the Aialak Glacier near Seward, AK. Note that no total particulatealuminum data was obtained for this transect.
115
116
Figure 8. Temperature (A), Salinity (B), silicic acid (C), dissolved Al (D), leachableparticulate Al (LP Al; (E)), and total particulate Al (TP Al; (F)) along Transect 2 shelfstations YAK1,YAK2, and YAK3 across the continental shelf in the northern GoA.YAK1 was located along the inner shelf (bottom depth = 94m); YAK2 was locatedmid-shelf (bottom depth = 135m); YAK3 was located at the outer shelf just at theshelf break (bottom depth = 630m).
117
118
Figure 9. Temperature (A), salinity (B), silicic acid (C), dissolved Al (D), leachableparticulate Al (LP Al; (E)), and total particulate Al (TP Al; (F)) at vertical profilestations GAK-A, GAK-B, and GAK-C along Transect 4 (GAK line) across thecontinental shelf in the northern GoA. GAK-A was located along the inner shelf(bottom depth = 94m); GAK-B was located mid-shelf (bottom depth = 153m); GAK-C was located offshelf in the HNLC waters of the Alaskan subarctic gyre (bottomdepth = 1450m).
119
120
Figure 10. Temperature (A), salinity (B), silicic acid (C), dissolved Al (D), leachableparticulate Al (LP Al; (E)), and total particulate Al (TP Al; (F)) at Kodiak SourceWater (KSW) Stations 1 (KSW1) and Station 2 (KSW2) on the continental shelf inthe northern GoA. KSW1 was located along the inner shelf (bottom depth = 215m)southeast of the Kenai Peninsula; KSW2 was located along the inner shelf within theShelikof Strait/lower Cook Inlet (bottom depth = 185m).
121
122
Figure 11. Dissolved aluminum vs. salinity (left) and silicic acid vs. salinity (right)along Transect 1 Plume A (black), Transect 3 (grey), and Transect 6 (open triangles)in the northern Gulf of Alaska.
123
124
Figure 12. Dissolved Al vs. salinity (left) and silicic acid vs. salinity (right) for T1Plume A and T1 Plume B. The regression line shown is the regression line for the T1Plume A data where no silicic acid drawdown was observed. The two points withinthe circle are from T1 Plume B where increases in biogenic Si and fluorescence wereobserved.
125
126
Figure 13. Underway ship fluorescence and biogenic silica in surface waters alongTransect 1.
127
128
Figure 14. Dissolved Al vs. total suspended solid (TSS) concentrations for Transects1,3, and 4 in the northern GoA.
129
130
CONCLUSIONS
The research associated with this dissertation resulted initially in a sensitive,
portable, easily adaptable shipboard flow injection analysis method to measure
dissolved Al across a wide range of concentrations in seawater. The dissertation also
investigated the behavior of dissolved and particulate Al in two unique coastal
environments influenced by large fluvial inputs: the Columbia River estuary and
coastal waters of Oregon and Washington influenced by the Columbia River plume
and the coastal waters of the northern Gulf of Alaska, a region highly influenced by
significant glacial and riverine discharge.
The modifications described in Chapter 1 to the pre-existing flow injection
analysis method for dissolved Al in seawater resulted in a method that is significantly
easier to use and more sensitive. This was accomplished through the use of a
commercially-available iminodiacetate preconcentration resin, the addition of a
column-conditioning step to enhance the complexation of Al onto the
preconcentration resin, and the in-line buffering of samples to the proper pH for
binding onto the preconcentration resin. The resulting method has been used to
measure dissolved Al concentrations greater than 1 µM in coastal low-salinity plumes
and concentrations less than 0.5 nM in surface waters of the California Current in the
high-nutrient, lower-than-expected chlorophyll surfaces waters of the subarctic
Alaskan gyre, a concentration range of over three orders of magnitude. The
portability of the method allows for near real-time analysis of dissolved Al in
seawater which is a valuable asset when studying the influence of coastal plumes or
131
eolian deposition to oceanic surface waters for which dissolved Al can serve as a
tracer.
A study of both dissolved and particulate Al in the Columbia River, Columbia
River estuary, and in low-salinity Columbia River plume waters along the coasts of
Washington and Oregon during both upwelling and downwelling conditions was
presented in Chapter 2. It was shown in this study that while the concentrations of
dissolved Al in the Columbia River were low compared to other major world rivers, a
likely result of low concentrations of dissolved organic matter such as humic and
fulvic acids in the Columbia River that serve to hold Al in solution, concentrations of
both dissolved and particulate Al were significantly greater in the river endmember
than in the coastal waters that mix to form the plume. Within the Columbia River
estuary it was found that both dissolved Al and silicic acid concentrations decreased
with increasing salinity. While silicic acid showed conservative mixing behavior
within the estuary, the dissolved Al-salinity relationship within the Columbia River
estuary showed a significant removal (~60%) at salinities between 0 and 10. This
dissolved Al removal was attributed to a combination of salt-induced flocculation of
colloidal Al complexes and enhanced particle scavenging within the estuarine
turbidity maximum.
During downwelling conditions, the Columbia River plume was observed
north of the Columbia River hugging the Washington coast. Silicic acid showed a
relatively conservative behavior in the downwelling plumes. However, the far-
northern transects did show a degree of biological removal of silicic acid as the plume
132
advected northward along the Washington coast. In moving from near-field to far-
field downwelling plumes, dissolved Al appeared to be controlled largely by dilution
of the plume with lower Al, higher-salinity waters. Leachable and total particulate Al
in the downwelling plumes also showed a decrease in concentration with an increase
in salinity as the plume mixes with lower particulate Al, higher salinity waters.
During upwelling conditions, the Columbia River plume advected offshore
and southwest of the Columbia River mouth. While silicic acid showed a fairly
conservative behavior in moving from the near-field to far-field within the upwelling
plume, dissolved Al exhibited a non-conservative behavior with an observed removal
of Al within the far-field plume. Passive particle scavenging of Al onto lithogenic
particle surfaces within the plume and an uptake of dissolved Al into siliceous
frustules during diatom frustule biosynthesis are possible explanations for this
removal. However, even with this nonconservative behavior, both dissolved and
particulate Al concentrations were an order of magnitude greater in the upwelling far-
field plume over 100 km from the Columbia River mouth than the surrounding
surface waters. This illustrated the potential use of dissolved Al as a tracer of the
input of the biologically-required micronutrient Fe to the coastal waters of
Washington and Oregon from the Columbia River plume.
Further work within the Columbia River estuary is necessary with regards to
suspended sediment dynamics to assess the relative roles of salt-induced flocculation
of dissolved Al and/or particle scavenging onto lithogenic particle surfaces to explain
the dissolved Al removal. In addition, elucidation of the relative roles of biological
133
uptake of dissolved Al during diatom frustule biosytnhesis and passive particle
scavenging removal mechanism in the upwelling far-field plumes would be useful in
understanding the biogeochemical cycles of dissolved Al in these low-salinity plume
waters.
In Chapter 3, I presented a study of soluble, dissolved, and particulate Al in
the coastal waters of the northern Gulf of Alaska with the soluble and dissolved Al
data being the first of its kind in the region. A distinct contrast in both dissolved and
particulate Al was observed between the coastal waters of the Alaska Coastal Current
influenced by the significant glacial and riverine discharge to the region and the
offshore surface waters of the central subarctic Alaskan gyre.
Both dissolved Al and particulate Al in low salinity coastal plume waters of
the Copper River and Alsek River showed dramatically elevated concentrations with
maximum dissolved Al and total particulate Al greater than 1.2 µM and 25 µM,
respectively. The percent-leachable Al fraction was quite low in these waters (~7%)
indicative of the highly refractory nature of the lithogenic material being delivered to
the coastal waters. The relatively low-salinity waters of the Alaska Coastal Current
also appeared to be enriched in dissolved and particulate Al relative to offshore
waters. Interestingly, the dissolved Al in the coastal waters was found entirely in the
soluble fraction. Very low concentrations of dissolved organic matter and a
subsequent lack of colloidal Al complexes in the glacial/riverine runoff to the coastal
waters is the probable explanation for a lack of any colloidal Al.
134
Comparison of silicic acid:dissolved Al ratios from different sampling periods
for the Copper River region of the northern Gulf of Alaska were remarkably
consistent. The weathering of biotite, which has been shown to be the dominant
mineral contributing to the dissolved load of subglacial streams in the region,
dissolution, and secondary mineral formation explain the observed distributions of
silicic acid, dissolved Al, and dissolved iron in northern Gulf of Alaska coastal
waters. It was also found that a strictly glacial melt endmember was significantly
different in terms of its silicic acid and dissolved Al concentrations than meltwater
that had traveled across the underlying bedrock, soil, and vegetation through riverine
delivery to the coastal waters.
In contrast, the offshore surface waters of the subarctic Alaskan gyre showed
very low concentrations of both dissolved and particulate Al. The dissolved Al
concentrations (< 0.2 nM) in these surface waters are some of the lowest dissolved Al
values reported in the world oceans. These low concentrations are likely a result of
an extreme degree of particle scavenging occurring in the coastal waters of the region,
supported by significant increases in the percent-leachable Al particulate fraction in
these waters. It was estimated that the residence time of dissolved Al in shelf waters
of the northern Gulf of Alaska was on the order of 10 days as compared to the
estimated 3-4 year residence time of dissolved Al in North Pacific oligotrophic
subarctic gyre surface waters.
Further work is needed in the northern Gulf of Alaska to first investigate the
mechanism of Al removal in these waters, whether through passive particle
135
scavenging or biological incorporation of dissolved Al. Secondly, it has been seen
that dissolved Al is elevated in eddy core waters that transport high-iron coastal
waters offshore into the central gyre. The use of dissolved Al as a tracer of this
process should be investigated.
As mentioned in the introduction section, it was hoped that dissolved Al could
serve as a tracer of the input of biologically required Fe to the coastal water of
Oregon and Washington and in the northern Gulf of Alaska. In coastal waters of
Oregon and Washington, it was found that in low salinity waters of the Columbia
River plume to a first approximation increases in dissolved Al were accompanied by
increases in dissolved Fe. However, it is now understood that an additional source of
Fe to the Columbia River plume can come from Fe(II) in subsurface waters during
upwelling periods. In addition, the residence time of dissolved Al in the Columbia
River plume water will likely vary with particle loads. Thus, the use of dissolved Al
as a tracer of Fe input has limitations in this region. Similarly, in the northern Gulf of
Alaska, extreme reactivity of dissolved Al in shelf waters was observed with a
residence time estimated to be ~10 days. This extreme reactivity of dissolved Al will
limit the use of dissolved Al as a tracer of Fe input.
The research presented in this dissertation resulted in an ideal method for
rapid, near real-time analysis of dissolved Al in seawater. In addition, the research
herein improved our understanding of the biogeochemical behavior of dissolved Al in
two unique coastal environments in the northeast Pacific Ocean strongly influenced
by riverine and /or glacial discharge to the oceans.
136