Origin of Iron Isotope Signatures in Boreal Estuaries

LICENTIATE T H E S I S

Department of Civil, Environmental and Natural Resources Engineering Division of Geosciences and Environmental Engineering

Origin of Iron Isotope Signatures in Boreal Estuaries

Sarah Conrad

ISSN 1402-1757 ISBN 978-91-7583-137-4 (print)

ISBN 978-91-7583-138-1 (pdf) Luleå University of Technology 2014

Sarah Conrad Or ig in of Ir on Isotope Signatur es in Bor eal Estuar ies

ISSN: 1402-1757 ISBN 978-91-7583-XXX-X Se i listan och fyll i siffror där kryssen är

O ^{RIGIN OF} I ^RON I ^SOTOPE S ^{IGNATURES IN}

B ^OREAL E ^STUARIES

S ^ARAH C ^ONRAD

Department of Civil, Environmental and Natural Resources Engineering Luleå University of Technology

Luleå, Sweden

Supervisors

Johan Ingri, Lena Alakangas, Emma Engström

Printed by Luleå University of Technology, Graphic Production 2014 ISSN 1402-1757

ISBN 978-91-7583-137-4 (print) ISBN 978-91-7583-138-1 (pdf) Luleå 2014

www.ltu.se

The geochemistry of iron (Fe) during freshwater transport and estuarine mixing has been investigated. Espescially the changes in the Fe-isotope signature have been studied. The fate of Fe-isotopes during estuarine mixing has been poorly studied. Sampling was performed in Kalix River, Kalix and Råne estuary, and in the open Bothnian Bay, Northern Baltic Sea. Water samples were filtered with 0.22 μm membrane filters. Both particulate (> 0.22 μm) and colloidal fractions (< 0.22 μm) were analyzed.

Iron particles and colloids, with a negative Fe-isotope signature, are formed during spring flood in forested catchments. These Fe complexes are associated with organic carbon (OC), and probably have a mixed oxidation state (Fe(II,III)-OC). Negative colloids are labile and flocculate and/or oxidize during riverine transport. Therefore, no negative colloids are detectable in the HVWXDULHVRUWKHRSHQ%RWKQLDQ%D\:LWKLQWKHHVWXDULHVWZRW\SHVRI˜

Fe signatures were measured: negative particles and positive colloids. The open Bothnian Bay shows a third distinct group of positive particles. This group mirrors the rapid removal of Fe colloids and particles at low salinities. Most of the Fe has been removed from surface water at salinities below 1.0 psu.

Data in this study show that the Fe-isotopes can be used to trace the origin and cycling of iron particles and colloids in the boreal landscape.

A ^BSTRACT

The thesis is based on the following papers

Conrad S., Ingri J., Wuttig K., and Rodushkin I. Iron isotope signatures in boreal estuaries and the Bothnian Bay. Manuscript.

Ingri J., Nordblad F., Engström E., Conrad S., and Rodushkin I. Iron isotope signatures in brown rivers. Manuscript.

Abstracts in conference proceedings

Conrad S. and Ingri J. 2014. Iron isotope signatures in low salinity estuaries, Northern Baltic Sea. The International Carbon Conference, Reykjavik, Iceland.

This project has received funding from the European Union´s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 290336 and the Swedish Research Council.

L ^{IST OF PAPERS}

First of all, I would like to acknowledge my supervisors Johan Ingri, Lena Alakangas and Emma Engström to give me the opportunity to take part in such an interesting project. Johan Ingri provided great support during the last two years, especially the last few months.

I´m grateful for my colleagues that became friends. Thanks for welcoming me at LTU. Special thanks to Fredrik for helping me with lab and field work, to Susi and Tobi, without their help everything would have been so much harder and to Dmytro who takes care about me.

I can´t thank my family enough, without them I wouldn´t be me. The same applies to all my friends I gained during the years. Thank you, Steffen, for letting me do this, supporting me in every single minute and believing in us.

A CKNOWLEDGEMENTS

C ^ONTENTS

I NTRODUCTION

S ^{COPE OF THE THESIS}

T ^HE B ^ALTIC S ^{EA AND THE} B ^OTHNIAN B ^AY I ^{RON IN NATURAL WATERS}

I ^{RON ISOTOPES} M ^{ATERIAL AND} M ^ETHODS

S ^{TUDY AREA}

S ^{AMPLING AND MATERIAL} M ^{EMBRANE FILTRATION} A ^NALYSIS

E ^LEMENT CONCENTRATION BY ICP-SFMS

I ^{RON ISOTOPE RATIO} MEASUREMENTS BY MC-ICP-MS S ^{UMMARY OF} R ^ESULTS

F ^UTURE W ^ORK R ^EFERENCES

P ^APER I - I ^{RON ISOTOPE SIGNATURES IN BOREAL ESTUARIES AND THE} B ^OTHNIAN B ^AY

P ^APER II - I ^{RON ISOTOPE SIGNATURES IN BROWN RIVERS}

1

1

1

3

3

5

5

7

7

8

8

9

11

15

17

S ^{COPE OF THE THESIS}

This thesis is part of an European research programme, which focuses on the advances of isotope measurement to quantify and understand fundamental processes that control metal transport in the environment.

The main research goal is to understand the distribution, speciation, isotopic composition, transport and biogeochemical behaviour of Fe in the Bothnian Bay and its catchment. In this thesis the pathways of Fe from freshwater sources, across freshwater-brackish interfaces, pelagic waters to sediment have been studied.

T ^HE B ^ALTIC S ^{EA AND THE} B ^OTHNIAN B ^AY

The Baltic Sea is the largest brackish water basin in the world and is located between 53.0° N to 66.0°N. Its halocline lies in 60 to 80 m water depth. It is divided into several basins, and its only water exchange is through the Kattegat to the North Sea and the Atlantic Ocean (Figure 1). The main part Baltic Proper is followed by the Gulf of Bothnia in the north, the Gulf of Finland in the northeastern direction and the Gulf of Riga in the east (Leppäranta and Myrberg, 2009). The Gulf of Bothnia is located between Sweden´s east coast and Finland´s west coast and is subdivided into the Bothnian Sea and the Bothnian Bay. It covers an area of 117,000 km

. Due to the high freshwater input through rivers from Sweden and Finland a salinity gradient from north to south exists. The Bothnian Sea can be classified as brackish water while the Bothnian Bay can be classified as freshwater. The Bothnian Bay is unaffected by tides and is ice covered for about five to seven months per year. It lies between 63.5° and 65.5° N and has an area of 36,800 km

with an average depth of 41 m and the deepest place at the Luleå deep with 146 m. The catchment area extends 260,675 km

, 44 % of this is located in Sweden while the rest is

I NTRODUCTION

finish property. The major freshwater rivers draining to the Bothnian Bay are Kemijoki (Finland, 556 m

s

), Lule River (Sweden, 506 m

s

), Torne River (Finland and Sweden, 388 m

s

) and Kalix River (Sweden, 295 m

s

).

K e mijok i Lule Ka lix T o

rne

Figure 1 - The Baltic Sea and their drainage basins. The major rivers Kemijoki, Lule

River, Torne River and Kalix River are highlighted.

I ^{RON IN NATURAL WATERS}

Iron is the 4

most abundant element in the earth´s crust and the most abundant trace element in marine and terrestrial organisms. Nonetheless, the bioavailability of Fe in waters is low, because of its low solubility in oxic waters (Liu and Millero, 2002) and the formation of colloids (Sanudo-Wilhelmy et al., 1996) and complexes with organic matter. It plays a key role in essential biogeochemical processes, through it changes in redox states. Iron has two oxidation states in water, Fe(II), and Fe(III), Fe(II) predominates in anoxic waters and Fe(III) in oxic waters (Liu and Millero, 2002). Iron is delivered to the open ocean by rivers, aeolian dust, icebergs (and continental runoff), hydrothermal activity and by recycling from shelf sediments (e.g. Wells et al., 1995). It has been shown, that coastal sediments can be major suppliers of iron to the deep sea (Elrod et al., 2004; Lam and Bishop, 2008). Most of riverine Fe is removed in estuaries (Sholkovitz, 1976). Total Fe is therefore low in the open ocean, which can be a limiting factor for photosynthesis.

Colloidal Fe is mainly transported in complexes with organic matter and as Fe- rich oxyhydroxides (Hassellöv and von der Kammer, 2008). At the freshwater- saltwater interface, a salt-induced flocculation occurs, and most of the particulate and colloidal Fe will be removed from the water column (Sholkovitz, 1976).

Our data show almost complete removal of Fe below 1.0 psu in the estuaries.

I ^{RON ISOTOPES}

Iron has four natural isotopes,

Fe (5.8%),

Fe (91.72 %),

Fe (2.2 %), and

58

Fe (0.28 %), (Taylor et al., 1992), whereas

Fe is the stable isotope (half-life time ca. 3.1*1022 years) and

Fe is the dominant isotope. A common way of defining the Fe-isotope signature is the proportion of these two natural isotopes in the sample.

∂ ⁵⁶ Fe= ( ⁵⁶ Fe/ ⁵⁴ Fe) _Sample

( ⁵⁶ Fe/ ⁵⁴ Fe) _IRMM -1 *1000‰

The Fe-isotope signature is quantified in permil (‰) deviation relative to the isotopic composition of a reference material (e.g. IRMM-014). In low- WHPSHUDWXUHHQYLURQPHQWVDYDULDWLRQRIÆRIWKH˜

Fe signature in natural samples has been observed (Beard et al. 2003). Iron isotopes show systematic changes in seawater over the last 1.7 Ma (Zhu et al., 2000). Rivers with high Fe and organic carbon show temporal variations. Malinovsky et al. (2005) showed DYDULDWLRQLQ˜

Fe of 1.0 - 2.0 ‰ due to seasonal redox variations in water and sediment samples in freshwater lakes.

Fractionation of Fe isotopes is mainly related to the oxidation of Fe. Iron is

LQYROYHGLQUHGR[WUDQVLWLRQV)H,,WHQGWRKDYHD˜

Fe signal, which is lighter

than the Fe(III) signature. Bullen et al. (2001) showed that abiotic fracionation

of Fe(II) into Fe(III) causes isotopic fractionation. Precipitated ferrihydrites,

had a heavier signal than the source material.

S ^{TUDY AREA}

The Bothnian Bay is the northern most part of the Baltic Sea and was sampled during the KBV005 cruise in May/June 2013. Its catchment is about 260,700 km

and fed by several rivers, e.g. Kalix River and Råne River, whose estuaries were sampled as well. The brackish system is affected by the freshwater compositions, which are formed by the different river catchments. The high freshwater input and the low tide-influence lead to low salinity (e.g. 0.2 psu).

The rivers Kalix and Råne show, despite their size, similar geochemical behaviours. The catchments differ by the source area while the Kalix River has its source in the Caldonides; the Råne River originate in the lake Råne träsk, which is woodland dominated (Figure 2). Therefore, the south-east flowing Kalix River is affected by mica shist, quartzite and amphibolite, which are the main minerals of the Caldonides (pre-Cambrian age). The catchment area covers 18,130 km

and is ice-covered for about five month per year. The total drainage area is 23,846 km

with an annual discharge of 295 m

s

. The drainage area is mostly woodland (55-65 %) with coniferous, 17 to 20 % is covered by wetland, lakes cover about 4 %, and farmland and population cover less than 1 % (SMHI, 2014). Quaternary deposits contain mainly till and show well developed podzol profiles (Fromm, 1965). The catchment area of the Råne River covers 4207 km

, with 72 % covered by woodland, 24 % wetland, 3.7 % lakes, and less than 1 % covered by farmland and population (SMHI, 2014).

The yearly precipitation ranges from 1000-1500 mm in the mountains to 400-700 mm at the coast. About 45 % of the precipitation is snow. The main hydrological event for the sampling site is the spring flood when the discharge increases up to ten times compared to base flow.

M ^{ATERIAL AND} M ^ETHODS

50 100 km 0

18°

20° 22°

68° 68°

Torne River

Kalix RiverKalix River

Ängeså River

Narkån

Äihämä Tvärån River

Lina River Kaitum River

Vassara River Rautas River

Lake

Torne träsk Lake

Torne träsk

KEBNEKAISE

KALIX KALIX

Luleå

 Luleå



67° 67°

66°

Bön River Abisko River

V itta ngi R

iv er

Kalix River

20° 22°

66°

Tär en ö d R ive r Tär en ö d

R ive r

10°E 20°E

20°E 10°E

70°N

65°N

60°N

NORWAY SWEDEN

FINLAND

Arctic circle

Killingi

Ö.Lansjärv Ö.Lansjärv

Karlsborg Karlsborg Råneå

Råneå Jukkasjärvi

Jukkasjärvi

GÄLLIVARE GÄLLIVARE

Kalix River

KIRUNA KIRUNA

Råne River Råne River

Figure 2 - Kalix and Råne River with tributaries and the borders of their catchments.

S ^{AMPLING AND MATERIAL}

Two pristine river estuaries of Northern Sweden were sampled during spring flood 2013. Water samples were taken directly from the vessel KBV005 in Kalix and Råne Estuary, as well as in the central Bothnian Bay (Map of the sampling stations can be found in Paper I). Four stations in each estuary (KA1 to -4 and RA1 to -3, -5) and two stations in the central Bothnian Bay (A5 and A13) were chosen. The samples were taken with Limnos Water Samplers in 0.5 m, 5.0 m, and 10.0 m depths. The Limnos Samplers consist of a duroplastic plastic, titanium screws and two 1L polyethylene Nalgene bottles. Briefly, they were cleaned successively with mucasol, 10 % HCl (s.p.), and 10 % HNO

(u.p.).

Between the different steps, they were rinsed with high purity water and dried under a clean bench (Cutter, 2010). Between the sampling stations, the bottles were rinsed with 10% HCl (s.p.) and high purity water. The Limnos sampler was lowered with a winch to the desired depth and opened by a drop weight.

On board, the tubing of the bottles was locked, and the water was directly pumped, with a peristaltic pump, from the bottles to prevent contamination.

Samples were taken in Kamlunge, in the Kalix River. This sampling location is approximately 20 km upstream from the river mouth. Samples from Kamlunge were taken during two different sampling campaigns. One campaign was in 1991 and 1992 (sampling details, treatment, and analytical information see Ingri et al., 2006), while the second campaign was between March and October 2006. Unfiltered samples were collected about 4 m from land with acid-washed, metal-free sampling equipment (Engström et al., 2010).

Additionally samples from two small first order catchments, Västrabacken and Kallkällebäcken, in the Krycklan area were taken. The Krycklan catchment consists of forests, mires, streams and lakes that make up 70 % of the area in Sweden. The site provides the most advanced long-term field research facility in operation in the boreal biome. Västrabäcken has 100 % forest within the catchment, whereas Kallkällebäcken is dominated by peatland. More detailed information about these catchments can be found in Laudon et al. (2013, and references therein).

M ^{EMBRANE FILTRATION}

The samples were filtered in situ through Millipore membrane filters (diameter

142 mm; 0.22 μm), which were locked in polycarbonate filter holders. The

filter holders were cleaned with 5 % HNO

(s.p.) and rinsed and dried under a

clean bench. The Millipore filters were cleaned for one week in 5 % acetic acid and afterwards rinsed with high purity water (Ödman et al., 1999). They were stored in high purity water until usage. The filtered samples (colloidal phase) were used for onboard experiments and aliquots were stored in acid-leached polyethylene bottles and conserved with HNO

(s.p.). The Millipore filters (particulate phase) were stored in 5 % HNO

cleaned petri dishes, which were bagged in zip logs and kept at – 18°C until chemical investigation.

Filtration resulted in different sample fraction, throughout the thesis the samples are classified as:

A ^NALYSIS

Marine environmental analysis like pH, temperature, alkalinity, dissolved organic carbon (DOC) and additional parameter, were carried out by Umeå Marina Forskningcentrum on unfiltered samples with a CTD-sond SBE 911 All laboratory sample preparations and analysis were performed in cooperation with ALS Scandinavia AB (Luleå, Sweden).

E ^LEMENT C ONCENTRATION BY I ^NDUCTIVELY C ^OUPLED P ^LASMA S ^ECTOR F ^IELD M ^ASS S PECTROPHOTOMETER (ICP-SFMS)

For the element analysis the water samples were diluted (2-200 fold) in 10 % HNO

; the degree of dilution is dependent on the salinity of the sample.

At least two dilutions of each sample were done, one high dilution for major elements and one low dilution for minor and trace elements. The filters were treated with a 10 mL of 1000:1 mixture of HNO

/ HF (s.p.) overnight following by closed-vessel digested in a microwave oven. The digests were

Sample Acronym Size Paper unfiltered

iron uFe Unfiltered water sample Paper II Paper I Paper II dissolved

iron dFe Water passing a 0.22 μm

membrane filter Paper II Paper I Paper II Paper I Paper II total iron tFe Sum of pFe and dFe,

respectively cFe, results particulate

iron pFe Particles remaining on a 0.22 μm membrane filter

colloidal

iron cFe Water passing a 0.22 μm

filter

further diluted in 10 % HNO

.

Wide range of major, minor and trace elements were measured in diluted water samples and filter digests by ICP-SFMS (ELEMENT XR, Thermo Scientific, Bremen, Germany) using a combination of internal standardization (Indium added at 2 μgL

to all measurement solutions) and external calibration. Details of the analytical procedure, as well as instrument parameters and measurement conditions, can be found elsewhere (Rodushkin and Ruth, 1997 and Rodushkin et al., 2005).

F ^{E ISOTOPE RATIO} MEASUREMENTS

For the isotope measurements, the water samples and the digested filters were evaporated till dryness, and the residuals were re-dissolved in 1 mL 8M HCl (s.p.), see above. Iron was separated from the matrix elements by ion exchange (Ingri et al., 2006). Fraction containing Fe was evaporated to dryness; 50 μl of concentrated HNO

was pipetted directly to residue following by addition of 5 mL MQ-water. For samples with relatively high Fe content, concentration in measurement solutions was adjusted to 2 mgL

by dilution with 1% HNO

. Evaporated separates of water samples with low Fe concentrations were diluted to approximately 40-50 μgL

and measured using high-efficiency dessolvation nebulizer (Aridus) in a separate analytical sequence.

Isotope ratio measurements were performed by MC-ICP-MS (NEPTUNE and NEPTUNE PLUS). Instrumental mass-bias was corrected using combination of internal standardization (Nickel added at 5 mgL

to all measurement solutions) and bracketing isotope standards matching sample solutions in Fe concentration and acid strength. Delta values were calculated against IRMM-14 CRM (Ingri et al., 2006).

In-house quality control samples (prepared by sequential dilutions of

SPECTROSCAN 10000 mgL

Fe element standard for atomic spectroscopy

from TEKNOLAB, Drøbak, Norway) were analyzed at the beginning and

the end of each analytical session in order to ensure internal consistency of

analytical results. Fe isotope ratios in this material were measured on a regular

EDVLVDW$/6ODERUDWRU\IURPDQGPHDQ˜

Fe of -0.236±0.031 ‰ (n>120,

one sigma) makes it a suitable reproducibility control.

The dissolved Fe (dFe) concentrations (Figure 3) along the river show that the main sources of Fe during the spring flood are forested areas. Paper II describes the detailed distribution of dFe from the source areas along the main channel.

The data show high dFe values during spring flood in the woodland catchment and low dFe from the mountain area. During summer, the tributaries show high dFe values, while the main channel has a low dFe concentration. In the estuaries, the removal of dFe is obvious. In the Råne and Kalix Estuary a removal of about 80 % of colloidal Fe (cFe) is taking place below 1.0 psu salinity.

The tFe concentration is about 10 μmol/L at the outermost estuarine station and about 1 μmol/L in the open Bothnian Bay, hence significant removal of particulate Fe (pFe) and cFe also occurs in the open Bothnian Bay.

The cFe/pFe ratio decreases from 80 at the headwaters to 2 in the estuaries, respectively 0.5 in the Bothnian Bay. The change in the ratio is indicating significant formation of particles from the colloid fraction during transport.

One of the main findings of this study is the removal of approximately 80 % of cFe and pFe below 1.0 psu salinity. Earlier studies suggested major removal of Fe between 5 to 15 psu (Sholkovitz et al., 1978; Escoube et al., 2009).

S ^{UMMARY OF} R ^ESULTS

0 10 40

20 30

15 45

25 35

5

dissolved Fe μM

Peatland Mountain Woodland

Main channel Råne Estuar y

Kalix Estuar y

Bothnian Ba y

Figure 3- Dissolved Fe concentration in Kalix River, Kalix Estuary, Råne Estaury, and the Bothnian Bay. Samples were taken during spring flood in 1991, 2006, and 2013.

Negative particles and colloids form in the forested area during spring flood.

7KH ˜

Fe signature of pFe values are about 0.5 ‰ more negative than the G)H)LJXUH2QWKHFRQWUDU\˜

dFe from the peatland area has a positive signature. Baseflow signatures for both catchments are negative (forest: -0.9 ‰;

SHDWODQGÆ8QILOWHUHGVDPSOHVVKRZDGURSRI˜

Fe from +0.5 ‰ in the source area down to 0.0 ‰ at Kamlunge during spring flood. Throughout WKH \HDU WKH ˜

uFe values show strong temporal variation, but the signal is DOZD\VSRVLWLYH(DUOLHUVWXGLHVVKRZHGDQHJDWLYH˜

pFe signature during spring IORRGDQGDSRVLWLYH˜

pFe signature during the rest of the year (Ingri et al., 2006). Three groups of Fe-isotope signatures can be observed during spring flood in the river, negative particles, negative colloids and positive colloids.

It has been shown that DOC has a large influence on the oxidation of Fe(II) to Fe(III) and the coherent formation of particles and colloids in the source waters (Perdue et al., 1976). At the oxic-anoxic interface in peatlands positive cFe is formed, while in the forested areas negative cFe occurs. It appears that site specific Fe signatures are formed, which are regulated by DOC and pH.

The formation of negative cFe is related to high DOC, which partially prevent the oxidation of Fe(II) to Fe(III). Therefore Fe(II) can co-precipitate with Fe(III) and forms Fe(II,III)-OC nano colloids. During transport these colloids can be transformed to positive Fe(III)-OH colloids by photoreduction and/or oxidation. Negative pFe is formed by flocculation of negative cFe.

In the estuaries, two groups of Fe-isotope signatures have been observed.

&ROORLGDO)HKDVDSRVLWLYH˜

Fe signature, while pFe is negative (Figure 4).

Data from paper I suggest no significant changes during flocculation processes in the estuaries. The same has been shown in previous studies (Escoube et al., 2009). In the open Bothnian Bay a third group of Fe-isotopes has been observed, positive particles.

Pédrot et al. (2011) concluded that two types of particulate-colloidal Fe are formed in the presence of terrestrial humic substances. Iron occurring within mixed Fe nanoparticles-organic matter colloids-particles, and in more crystalline Fe colloids (Pédrot et al., 2011). Data in this study suggest that particles with

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.7 0.6 0.5

0.3 0.4

0.2 0.1

-1.0 -0.9 -0.8 -0.7 -0.6

∂

Fe ‰

Peatland W oodland

Main channel Estuar y

Bothnian Ba y

par ticulate

dissolv ed dissolv ed unfilter ed par ticulate dissolv ed par ticulate

Figure 4 - Fe-isotope signature of the source areas, peatland and woodland, the main channel, the estuaries and the Bothnian Bay. The hatched area shows the range of isotopic signature in the samples, of the different locations, during the sampling events.

The information next to the hatched area defines the sort of sample (dissolved, particulate

or unfiltered).

a negative Fe-isotope signature are removed below 1.0 psu. The Fe-isotope

signature of the woodland catchment is formed in the riparian zone. The

riparian zone is fed by water from podzol profiles, which is depleted in heavy

Fe-isotopes. These podzol soils support the fractionation of Fe-isotopes, due to

their low pH (Fekiacova at al., 2013). Dissolved Fe(II) is oxidized, which leads

to an enrichment of light isotopes in the water. Teutch et al. (2005) showed that

the adsorption of Fe(II) onto newly formed Fe(III)-oxyhydroxides yields a very

light soil water component with values down to -3.0 ‰ indicating that heavier

Fe is preferentially transferred to the newly formed Fe(III)-OH surfaces.

The second half of the thesis programme is focused on the cycling of P in the Bothnian Bay. The Gulf of Bothnia is an area with net deposition of phosphorus, in contrast to southern parts of the Baltic Sea. Mass balance calculations suggest that more than 60 % of the yearly sediment sequestering capacity of phosphorus in the Baltic Sea is located in the Gulf of Bothnia. The project focuses on the mechanisms, causing the high sequestering of phosphorus in sediments and, as we suggest, is of importance for future management of the whole Baltic Sea and related catchments. Hence, the distribution, speciation, isotopic composition, transport and biogeochemical behaviour of Fe in the Baltic Sea have recently become important issues. This program has studied the pathways of Fe from the freshwater sources, across the freshwater-brackish interface, pelagic waters to sediment, and will be focussed on phosphorus cycling.

F ^UTURE W ^ORK

Beard BL, Johnson CM, Von Damm KL, and Poulson RL (2003) Iron isotope constraints on Fe cycling and mass balance in oxygenated Earth oceans. Geology V.31, no.7, 629-632.

Cutter G (2010) Sampling and sample-handling protocols for GEOTRACES cruises.

Elrod VA, Berelson WM, Koale KH, and Johnson KS (2004) The flux of iron from continental shelf sediments: A missing source for global budgets. Geophysical Research Letters 31.

Engström E, Rodushkin I, Ingri J, Baxter DC, Ecke F, Österlund H, and Öhlander B (2010) Temporal isotopic variations of dissolved silicon in a pristine boreal river.

Chemical Geology 271, 142-152.

Escoube R, Rouxel OJ, Sholkovitz E, Donard OFX (2009) Iron isotope systematics in estuaries: The case of North River, Massachusetts (USA). Geochimica et Cosmochimica Acta 73, 4045-4059.

Fantle MS and DePaolo DJ (2004) Iron isotopic fractionation during continental weathering. Earth and Planetary Science Letters 228, 547–556.

Fromm E (1965) Beskrivning till jordartskartan over norrbottens län, nedanför lappmarksgränsen, SGU ser. Ca. 41, 1-151, (In Swedish with english summary).

Ingri J, Malinovsky D, Rodushkin I, Baxter DC, Widerlund A, Andersson P, Gustafsson Ö, Forsling W, Öhlander B (2006) Iron isotope fractionation in river colloidal matter, Earth and Planetary Science Letters 245, 792-798.

Lam PJ and Bishop JKB (2008) The continental margin is a key source of iron to the HNLC North Pacific Ocean. Geophysical Research Letters 35.

Laudon H, Tabermann I, Ågren A, Futter M, Ottosson-Löfvenius M, and Bishop K (2013) The Krycklan Catchment Study—A flagship infrastructure for hydrology, biogeochemistry, and climate research in the boreal landscape. Water Resources Research 49, 7154-715.

Leppäranta M and Myrberg K (2009) Physical Oceanography of the Baltic Sea, Springer Berlin Heidelberg New York, ISBN 978-3-540-79702-9.

R ^EFERENCES

Liu X and Millero FJ (2002) The solubility of iron in seawater. Marine Chemistry 77, 43-54.

Ödman F, Ruth T, Pontér C (1999) Validation of a field filtration technique for characterization of suspended particulate matter from freshwater. Part 1. Major elements. Applied Geochemistry 14, 301–317.

Rodushkin I and Ruth T (1997) Determination of Trace Metals in Estuarine and Sea- water Reference Materials by High Resolution Inductively Coupled Plasma Mass Spectrometry. Journal of Analytical Atomic Spectrometry 12, 1181–1185.

Rodushkin I, Nordlund P, Engström E, and Baxter DC (2005) Improved multi- elemental analyses by inductively coupled plasma-sector field mass spectrometry through methane addition to the plasma. Journal of Analytical Atomic Spectrometry 20, 1250-1255.

Pédrot M, Le Boudec A, Davranche M, Dia A, and Henin O (2011) How does organic matter constrain the nature, size and availability of Fe nanoparticles for biological reduction? Journal of Colloid and Interface Science 359, 75-85.

Perdue ME, Beck KC, and Reuter JH (1976) Organic complexes of iron and aluminium in natural waters. Nature 260, 418-420.

Sanudo-Wilhelmy SA, Rivera-Duarte I, and Flegal AR (1996) Distribution of colloidal trace metals in the San Francisco bay estuary. Geochimica and Cosmochimica Acta 60, 4933-4944.

Sholkovitz ER (1976) Flocculation of dissolved organic and inorganic matter during the mixing of river water and seawater. Geochimica et Cosmochimica Acta 40, 831.

Sholkovitz ER (1978) The flocculation of dissolved Fe, Mn, Al, Cu, Ni, Co and Cd during estuarine mixing. Earth and Planetary Science Letters 41, 77-86.

Swedish Meteorological and Hydrological Institute (SMHI) (http://www.smhi.se/, October 2014).

Wells ML, Price NM, and Bruland KW (1995) Iron chemistry in seawater and its relationship o phytoplankton: a workshop report. Marine Chemistry 48, 157-182.

Zhu XK, O’Nions RK, Guo Y, and Reynolds BC (2000) Secular variation of iron

isotopes in North Atlantic deep water. Science 287, 2000–2002.

P ^APER I

I ^{RON ISOTOPE SIGNATURES IN}

BOREAL ESTUARIES AND THE B ^OTHNIAN B ^AY

Conrad S., Wuttig K., Rodushkin I., and Ingri J., in preparation. Iron

isotope signatures in boreal estuaries and the Bothnian Bay.

A ^BSTRACT

Water samples were taken in Kalix River and Råne River estuaries, in northern Sweden, and in the open Bothnian Bay. The water was filtered with membrane filters (0.22 μm), and both particlulate and colloidal phase were analyzed.

Element concentrations were determined by ICP-SFMS on diluted water samples and filter digests. Iron isotopes were separated from matrix elements by ion exchange and measured with an MC-ICP-MS.

Both estuaries show low salinity (0.01 to 2.56 psu), high DOC (408 to 727 μmol/L), and high Fe (up to 20 μmol/L) concentrations. The colloids of WKHVH´EURZQHVWXDULHVµVKRZHGSRVLWLYH˜

Fe values, whereas particles showed PDLQO\QHJDWLYHLVRWRSHVLJQDWXUHV,QWKHRSHQ%RWKQLDQ%D\SRVLWLYH˜

Fe values were measured in the particulate phase. Flocculation of particles and colloids did not change the Fe-isotope signature. At a salinity of 1.0 psu, almost all Fe had been removed from the water.

I NTRODUCTION

It has long been recognized that Fe is closely related with organic carbon in brown freshwater (Perdue et al., 1976). Although the importance of iron oxides for the cycling of DOC has long been known, just recently has the role of sedimentary reactive Fe phases for OC storage been estimated to be quantitatively important on a global scale, the ”rusty carbon sink” (Lalonde et al., 2012). The coupled cycling of Fe and terrigenous DOC at oxic-anoxic transitions may explain some of the “missing” terrestrial derived DOC in the oceans (Bianchi, 2011). In spite of the importance, little is known about these complexes during freshwater and estuarine transport. Lalonde et al. (2012) proposed that Fe-OC complexes are formed during early diagenesis in the sediment. Data in this study indicate that these complexes might be formed already on land. Ingri et al. (in prep.) indicate that Fe isotopes can be used to trace the origin of Fe-OC complexes.

It is well established that most of the Fe-rich particulate and colloidal material in rivers flocculate and settle in the mixing zone between freshwater and sea water (Sholkovitz et al., 1978). Hence, the river input of Fe into the ocean is greatly modified by processes, that occur during the mixing of freshwater and coastal water (Eckert and Sholkovitz, 1976; Sholkovitz, 1976, 1978; Boyle et al., 1977;

Mayer, 1982; Hunter, 1990).The fate of Fe isotopes during estuarine mixing

has been little studied. To the author’s knowledge only one comprehensive

estuarine study has been published (Escoube et al., 2009), and Fe isotope data from estuarine water with salinity below 1.0 psu has not previously been presented. We have used the same sampling and analytical methods as used by Escoube et al. (2009) in this study to make it possible to compare between the estuaries. Samples along Kalix and Råne Estuary were taken during spring IORRG  DQG DQDO\]HG IRU WKHLU WUDFH PHWDO FRPSRVLWLRQ DQG WKHLU ˜

Fe VLJQDWXUHV &ROORLGDO )H LQ WKHVH´EURZQ ULYHUVµ VKRZV SRVLWLYH ˜

Fe values, whereas the particulate phase has mainly a negative isotope signature (Ingri et al., 2006; Ingri et al., in prep.). The fate of these two groups of Fe-isotope signatures during estuarine mixing is discussed in this study

S ^{AMPLING SITE AND ANALYTICAL METHODS}

S ^{AMPLING SITE}

The Bothnian Bay (63.5° and 66°N) is covered up to six month per year with

ice. Its catchment is about 260,700 km

and fed by several rivers. The high

freshwater input and the low tide-influence lead to low salinity (maximum

3.0 psu in surface water). The large Kalix River and the smaller Råne River

show, despite their different size, similar geochemical behaviors during spring

flood. The Kalix River has its source in the Caledonian Mountains, whereas

Råne River originates in the lake Råne träsk in the woodland. Kalix river

catchment area covers 18,130 km

and is ice- and snow-covered for about

five month per year. The total drainage area is 23,846 km

with an annual

discharge if 295 m

s

. The drainage area is mostly woodland (55-65 %) with

coniferous, 17 to 20 % is covered by wetland, lakes cover 4 %, and farmland and

population cover less than 1 % (SMHI). Quaternary deposits contain mainly till

and show well developed podzol profiles (Fromm, 1965). The catchment area

of the Råne River covers 4207 km

, with 72 % covered by woodland, 24 %

wetland, 3.7 % lakes, and less than 1 % covered by farmland and population

(SMHI). The yearly precipitation ranges from 1000-1500 mm in the mountains

to 400-700 mm at the coast. About 45 % of the precipitation is snow. The main

hydrological event for the sampling site is the spring flood when the discharge

increases up to ten times compared to base flow.

S ^{AMPLING AND MATERIAL}

Water samples were taken directly from the vessel KBV005 in Kalix and Råne Estaury, as well as in the central Bothnian Bay (Figure 1). Four stations in each estuary (K1 to -4 and R1 to -3, -51) and two stations in the central Bothnian Bay (A5 and A13) were sampled.

The samples were taken with Limnos Water Samplers at 0.5 m, 5.0 m, and 10.0 m depths. The Limnos Samplers consist of a duroplastic plastic, titanium screws and two 1 L polyethylene Nalgene bottles. The polyethylene bottles were cleaned successively with mucasol, 10 % HCl (s.p.), and 10 % HNO

(u.p.).

Between the different steps they were rinsed with high purity water and dried under a clean bench. Between the sampling stations the bottles were rinsed with 10 % HCl (s.p.) and high purity water. The Limnos sampler was lowered with a winch to the desired depth and opened by a drop weight. On board the tubing of the bottles was locked and the water was directly pumped from the bottles to prevent contamination. The samples were filtered in situ through Millipore membrane filters (diameter 142 mm; 0.22 μm), which were locked in polycarbonate filter holders. The filter holders were cleaned with 5 % HNO

(s.p.) and rinsed and dried under a clean bench. The Millipore filters were

18 km

R1/11

K1/11 K2/21

K3/31 K4/41 R51

R3/31 R2/21

N 65°

N 66°

N 64°

Bothnian Bay

Bothnian Bay A5 A13

E 21° E 25°

Figure 1 - Map of Northern Sweden showing the sampling locations in Råne and Kalix

Estuary (Råne stations: R1 to R3 in May and R11 to R51 in June; Kalix stations: K1 to

K4 in May and K11 to K41 in June) and the Bothnian Bay (A5 and A13 in May; A51 and

A131 in June).

cleaned for one week in 5 % acetic acid and afterwards rinsed with high purity water (Ödman et al., 1999). They were stored in high purity water until usage.

The filtered samples (colloidal phase) were used for onboard experiments and aliquots were stored in acid-leached polyethylene bottles and conserved with HNO

(s.p.). The Millipore filters (particulate phase) were stored in 5 % HNO

cleaned petri dishes, which were bagged in zip logs and kept at – 18°C until chemical investigation.

A ^{NALYTICAL METHODS}

Basic parameter like oxygen saturation, pH, DOC, and additional parameters were carried out by Umeå Marina Forskningcentrum on unfiltered samples.

All laboratory sample preparations and analyzes were done in collaboration with ALS Scandinavia AB (Luleå, Sweden).

E ^LEMENT CONCENTRATION BY INDUCTIVELY COUPLED PLASMA SECTOR

FIELD MASS SPECTROPHOTOMETER (ICP-SFMS)

For the element analysis the colloidal samples were diluted (2-200 fold) in 10 % HNO

; the degree of dilution is dependent on the salinity of the sample.

At least two dilutions of each sample were done one high dilution for major elements and one low dilution for minor and trace elements. The filters were treated with a 10 ml of 1000:1 mixture of HNO

/HF (s.p.) overnight following by closed-vessel digested in a microwave oven. The digests were further diluted in 10 % HNO

(s.p.).

Major, minor, and trace elements were measured in diluted water samples and filter digests by ICP-SFMS (ELEMENT XR, Thermo Scientific, Bremen, Germany) using a combination of internal standardization (Indium added at 2 μgL

to all measurement solutions) and external calibration. Details of the analytical procedure, as well as instrument parameters and measurement conditions, can be found elsewhere (Rodushkin and Ruth, 1997 and Rodushkin et al., 2005).

I ^{RON ISOTOPE RATION} MEASUREMENTS

For the Fe-isotope measurements, the colloidal samples and the digested

filters were evaporated till dryness, and the residuals were re-dissolved in 1

mL 8 M HCl (s.p.), see above. Iron was separated from the matrix elements

by ion exchange (Ingri et al., 2006). Fraction containing Fe was evaporated to dryness; 50 μL of concentrated HNO

(s.p.) was pipetted directly to residue following by addition of 5 mL MQ-water. For samples with relatively high Fe content, concentration in measurement solutions was adjusted to 2 mgL

by dilution with 1% HNO

(s.p.). Evaporated separates of water samples with low Fe concentrations were diluted to approximately 40-50 μgL

and measured using high-efficiency dessolvation nebulizer (Aridus) in a separate analytical sequence.

Isotope ratio measurements were performed by MC-ICP-MS (NEPTUNE and NEPTUNE PLUS). Instrumental mass-bias was corrected using combination of internal standardization (Nickel added at 5 mgL

to all measurement solutions) and bracketing isotope standards matching sample solutions in Fe concentration and acid strength. Delta values were calculated against IRMM- 14 CRM (Ingri et al., 2006).

In-house quality control samples (prepared by sequential dilutions of SPECTROSCAN 10000 mgL

Fe element standard for atomic spectroscopy from TEKNOLAB, Drøbak, Norway) were analyzed at the beginning and the end of each analytical session in order to ensure internal consistency of analytical results. Iron isotope ratios in this material were measured on a regular EDVLVDW$/6ODERUDWRU\IURPDQGPHDQ˜

Fe of -0.236±0.031 ‰ (n>120, one sigma) makes it a suitable reproducibility control for Fe isotope ratio measurements in low fractionated samples (Baxter et al., 2006). The Fe isotopic VLJQDWXUH LV GHFODUHG DV ˜

Fe in permil (‰). It is calculated by the formula below and displays the ratio of

Fe to

Fe in the sample and a standard.

R ^ESULTS

Discharge values from Kalix and Råne River from 2010 until 2013 are shown in figure 2, the main graph shows detailed discharge information between April and August 2013. Spring flood maximum for Råne River was in mid-May 2013 while the maximum discharge in Kalix River occurred four days later.

Kalix River discharge is about four times higher, than Råne River discharge.

Sampling was carried out in May (21.-23.) and June (05.-07.) 2013 in both estuaries.

∂ ⁵⁶ Fe= ( ⁵⁶ Fe/ ⁵⁴ Fe) _Sample

( ⁵⁶ Fe/ ⁵⁴ Fe) _IRMM -1 *1000‰

Hence, the collected data show the behavior of elements and isotopes during and after maximum spring flood.

G ^EOCHEMICAL CLASSIFICATION

The main geochemical parameters (temperature, pH, salinity, etc.) measured during the sampling events in late May and early June 2013 are summarized in table 1.

Salinity in Kalix Estuary increases from 0.01 to 1.04 psu in May, and from 0.08 to 2.64 during June (Figure 3). DOC ranges between 411 and 684 μmol/L during both samplings and decreases along the estuary. The oxygen saturation ranges from 89.2 to 107.7 %, at most stations the values decrease with depth (Figure 4). pH ranges between 6.8 and 7.5 during the sampling events. During May pH increases from the river mouth to the outermost station.

Salinity in Råne Estuary ranges from 0.01 to 2.08 psu in May and from 0.22 to 0.83 psu during June. In May, the salinity increases with distance to the coast. DOC ranges from 408 to 718 μmol/L and decreases from the innermost

Spring flood 2013

Discharge m

/s

Date

04/01/2013 06/24/201 07/08/2013 07/22/2013

3

06/10/2013 05/27/2013 05/13/2013 04/29/2013 04/15/2013

Kalix River Råne River

1000

400 1800

1600

1400

1200

0 200 600 800

05/19/201 3

440 m³/s 05/23/2013

1550 m

3/s

2010 2011 2012 2013

2000

0 500 1000 1500

Discharge m

/s

Year

Figure 2 – Discharge in m

s

from 2010 to 2013 in Råne and Kalix River. The main

graph shows the discharge during spring flood 2013 with maximum values at certain

dates for both Kalix and Råne River (Datasource: SMHI).

Depth (m) 0.00.0

2.0 4.0 6.0

12.0 10.0 8.0

0.5 1.0 1.5 2.0 2.5 3.0

R1

R2 R3 A5

A13

Kalix Estuary Råne Estuary Bothnian Bay

May June M

A Y

J U N E

M A Y

J U N E

RÅNE ESTUARY KALIX ESTUARY

R21 R31 R51 R11

A131 A51 0.00.0

2.0 4.0 6.0

12.0 10.0 8.0

0.5 1.0 1.5 2.0 2.5 3.0

Depth (m)

Salinity psu

K1

K4 K3 K2 0.00.0

2.0 4.0 6.0

12.0 10.0 8.0

0.5 1.0 1.5 2.0 2.5 3.0

K11

K41 K31 K21

0.00.0

2.0 4.0 6.0

12.0 10.0 8.0

0.5 1.0 1.5 2.0 2.5 3.0

300

DOC (μmol/L)

600 550 500 450 400

350 650 700 750

R1

R3 R2 A13

A5 0.0 2.0 4.0 6.0

12.0 10.0 8.0

Depth (m)

R11

R21 R31 R51 A131 A51

0.0 2.0 4.0 6.0

12.0 10.0 8.0

Depth (m)

300 350 400 450 500 550 600 650 700 750

K11 K21

K31 K41 0.0

2.0 4.0 6.0

12.0 10.0 8.0

300 350 400 450 500 550 600 650 700 750 K1

K3 K2

K4 0.0

2.0 4.0 6.0

12.0 10.0 8.0

300 350 400 450 500 550 600 650 700 750

Figure 3 - Depth distribution of salinity (psu)

and DOC (μmol/L) for Råne and Kalix Estuary

during May and June 2013. Bothnian Bay stations

are included in the Råne Estuary graphs.

Oxygen (%)

2.0 4.0 6.0

12.0 10.0 Depth (m) 8.0

93 98 103 108 113

R1

R3 R2 A5 A13

0.088

R11

R31 R41 R21

A51

A131 2.0

4.0 6.0

12.0 10.0 Depth (m) 8.0

0.088 93 98 103 108 113

K1

K4 K3 K2

2.0 4.0 6.0

12.0 10.0 8.0

0.088 93 98 103 108 113

K11

K21

K31 K41

2.0 4.0 6.0

12.0 10.0 8.0

0.088 93 98 103 108 113

pH

6.4 6.6 6.8 7.0 7.2 7.4 7.6 7.8 8.0

R1

R3

R2 A5A13

2.0 4.0 6.0

12.0 10.0 Depth (m) 8.0

0.0

R11 R41 R31

R21

A51 A131 2.0

4.0 6.0

12.0 10.0 Depth (m) 8.0

0.06.4 6.6 6.8 7.0 7.2 7.4 7.6 7.8 8.0

K1

K4 K3 K2 2.0

4.0 6.0

12.0 10.0 8.0

0.06.4 6.6 6.8 7.0 7.2 7.4 7.6 7.8 8.0

K11

K21

K31 K41 2.0

4.0 6.0

12.0 10.0 8.0

0.06.4 6.6 6.8 7.0 7.2 7.4 7.6 7.8 8.0

May June M A Y

J U N E

M A Y

J U N E

RÅNE ESTUARY KALIX ESTUARY

Kalix Estuary Råne Estuary Bothnian Bay

Figure 4 - Depth distribution of oxygen

(saturation %) and pH for Råne and Kalix Estuary

during May and June 2013. Bothnian Bay stations

are included in the Råne Estuary graphs.

to the outermost station. Oxygen saturation ranges from 89.9 to 103.8 % in May and from 97.4 to 107.4 % in June. The overall trend is increasing oxygen saturation from the river mouth to open Bothnian Bay. Råne Estuary has pH values between 6.4 and 7.4, with lower values close to the river mouth; similar pH values have been observed in the brown estuary of Öre River by Forsgren and Jansson (1992).

Salinity in the Bothnian Bay varies slightly around 2.9 psu for both stations and weeks, station A51 (June) shows lower values around 2.48 psu. The DOC ranges between 348 and 396 μmol/L, with higher values during the second sampling. Oxygen saturation in the open Bothnian Bay ranges from 101.7 to 110.3 % while station A13 has higher values than station A5 during both sampling events. pH ranges from 7.7 to 7.9 and show no large differences between May and June.

Chlorophyll-a values (Figure 5) were measured at 5 m depth for each station and ranges from 0.8 to 3.6 mg/m

. Kalix Estuary values of May and June range from 1.4 to 2.6 mg/m

, and Bothnian Bay values for both samplings range between 2.4 and 2.9 mg/m

. The only difference between May and June was observed in Råne Estuary (0.8 to 1.8 mg/m

in May and 2.0 to 3.6 mg/m

in June).

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

0.0 3.0 2.5 2.0 1.5 1.0 0.5 3.5 4.0

Salinity psu

Chl-a mg/m

Kalix Estuary Råne Estuary Bothnian Bay

May June Sampling date

Figure 5 - Scatter plot of chlorophyll-a (mg/m

) and salinity (psu) in 5 m depth at each sampling

location.

I ^{RON CHEMISTRY AND ISOTOPIC} COMPOSITION

Table 2 shows all measured iron concentrations and isotopic signatures and their standard deviation for more detailed information.

We have defined the filtered fraction (< 0.22 μm) as dissolved fraction, dFe and the fraction on the filters (> 0.22 μm) as particulate fraction, pFe. In the rivers and the estuaries dFe is almost the same as cFe (colloidal fraction), (Ingri et al., 2000). We, therefore, use the label cFe for the dissolved fraction in the following text. Total Fe concentrations are calculated from cFe and pFe and are labeled tFe.

I ^RON CONCENTRATION

In the Kalix Estuary, a slight decrease of cFe from the surface water samples to the deepest samples at each station was measured (Figure 6). While the stations K1 to K3 show concentrations of about 17.0 μmol/L, station K4 shows lower concentrations of 5.7 (surface) to 5.0 μmol/L (bottom) in May. In June, the depth profiles in Kalix Estuary are less defined. No clear patterns can be observed. In the Råne Estuary, the cFe concentrations decrease with depth and with distance to the river mouth. The cFe concentration in Råne Estuary is similar to Kalix Estuary, the values range from 16 to 6 μmol/L in the upper 5 m during May. The difference between the surface water samples and the 5 m samples is minor, while the difference to the 10 m samples ranges between 50 and 75 %. In June, the range is 14 to 6 μmol/L. Station R11 to R51 show decreasing cFe concentration with depth. The concentrations also decrease with distance to the river mouth. cFe data from Escoube et al. (2009) range from 0.0 to 8.0 μmol/L, which is approximately 50 % less cFe than in this study.

The pFe concentrations in Kalix estuary are lower than cFe concentrations (Figure 7). At station K1 to K3 the concentration is decreasing with depth, the ratio cFe/pFe is increasing and shows values from 3.0 to 6.0. Besides station K1, there is a decreasing trend of pFe from coast to the open ocean. In June, concentrations are generally lower. The overall trend is decreasing concentration with depth, whereas the decrease from surface water to the 5 m samples is not as distinct as the decrease from 5 m to 10 m depth. All stations show values between 8.5 at surface and 1.2 μmol/L at 10 m depth. The pFe concentrations are decreasing with depth and with distance to the river mouth. The cFe/pFe ratios are lower than in May (1.1 – 4.2) and is decreasing with depth (Figure 8). The pFe concentrations in are lower than the cFe and are not following a clear pattern.

Råne and Kalix estuary show similar pFe concentrations ranging from 6 to

0.0 2.0 4.0 6.0 8.0 10.0 12.014.0 16.0 18.0 20.0 0.0

2.0

4.0

6.0

12.0 10.0 8.0

Depth (m)

R51

R31

R11 R21

A131 A51

cFe (μmol/L)

0.0 2.0 4.0 6.0 8.0 10.0 12.014.0 16.0 18.0 20.0 0.0

2.0

4.0

6.0

12.0 10.0 8.0

Depth (m)

R1

R3 R2 A13A5

0.0 2.0 4.0 6.0 8.0 10.0 12.014.0 16.0 18.0 20.0 0.0

2.0

4.0

6.0

12.0 10.0 8.0

K1

K4 K3 K2

0.0 2.0 4.0 6.0 8.0 10.0 12.014.0 16.0 18.0 20.0 0.0

2.0

4.0

6.0

12.0 10.0 8.0

K31 K21 K41 K11

Kalix Estuary Råne Estuary Bothnian Bay

May June

RÅNE ESTUARY KALIX ESTUARY

M A Y

J U N E

Kalix Estuary Råne Estuary Bothnian Bay

May June

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

0.0

2.0

4.0

6.0

12.0 10.0 8.0

Depth (m)

R31 R21 R51 A131A51

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

0.0

2.0

4.0

6.0

12.0 10.0 8.0

K21 K31 K41

K11

pFe (μmol/L)

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

0.0

2.0

4.0

6.0

12.0 10.0 8.0

Depth (m)

R3 R1

R2 A13

A5

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

0.0

2.0

4.0

6.0

12.0 10.0 8.0

K1

K4 K3 K2

MA Y

UJ NE

RÅNE ESTUARY KALIX ESTUARY

Figure 6 - Depth distribution of cFe (μmol/L) for Råne and Kalix Estuary during May and June 2013. Bothnian Bay stations are included in the Råne Estuary graphs.

Figure 7 - Depth distribution of pFe (μmol/L)

for Råne and Kalix Estuary during May and

June 2013. Bothnian Bay stations are included

in the Råne Estuary graphs.

0.5 μmol/L along the estuary. At station R1 and R2 the 5 m samples show the highest concentrations, while station R3 shows decreasing concentration with depth. The ratio cFe/pFe is decreasing with depth at all stations, varying between 2.6 and 4.8.

The pFe concentrations of the Råne Estuary show decreasing values with depth during June. Compared to May the values at R11 are three times higher (15.5 μmol/L). Station R2 shows almost no variation in May, while station R31 shows slightly higher values compared to May. Station R51 shows slightly decreasing values with depth. The cFe/pFe ratio is varying from 0.91 to 3.13.

In the Bothnian Bay (A5 and A13) the cFe and pFe results show no variation with depth. The values at A5 are slightly higher than at station A13, the ratios cFe/pFe are 0.75 to 1.52 during May and 0.47 to 1.30 during the June. The pFe Data in Escoube et al. (2009) range from 2.0 to 8.0 μmol/L during the sampling.

tFe (μmol/L)

Depth (m)

Kalix Estuary Råne Estuary Bothnian Bay

May June

RÅNE ESTUARY KALIX ESTUARY

M A Y

J U N E 0.0 5.0

R1

R3 R2 A13A5

R31R51

R11 R21

A131 A51

K1

K4 K3 K2

K21 K41 K31 K11

30.0 25.0 20.0 15.0

10.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0

0.0 5.0 10.0 15.0 20.0 25.0 30.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0

0.0 2.0 4.0 6.0 8.0 10.0 12.0

0.0 2.0 4.0 6.0 8.0 10.0 12.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 0.0

2.0 4.0 6.0 8.0 10.0 12.0

Depth (m)

Figure 8 - Depth distribution of tFe (μmol/L) for Råne and

Kalix Estuary during May and June 2013. Bothnian Bay stations

are included in the Råne Estuary graphs.

The tFe in Råne Estuary shows a decrease of approximately 40 to 60 % from station R1/R11 to the other stations within this estuary (Figure 8). The difference between May and June is mainly seen at the innermost station (R1/

R11). In June the values of the stations R21-R51 are very similar to each other, around 10 μmol/L. Total Fe decreases from 29.5 to 0.5 μmol/L in the open Bothnian Bay. In Kalix Estuary, the tFe values decrease from 24.5 to 6.1 μmol/L in May and from 14.5 to 2.5 μmol/L in June.

DOC, cFe and pFe show exponential decrease with increasing salinity (Figure 9). DOC decreases from 650 to 420 μmol/L, cFe decreases from 18 to 4 μmol/L, and pFe decreases from 6 to 2 μmol/L. The relation between cFe and DOC in the estuaries and the open Bothnian Bay is shown in Figure 10.

While cFe decreases from 18 to almost 0 μmol/L in the open Bothnian Bay, DOC concentration decreases about 50 % (Figure 10).

20

0 2 4 6 8 10 12 14 16 18

0.0 0.5 1.0 1.5 2.0 2.5 3.0

750

300 400 450 500 550 600 650 700

DOC μmol/L

Salinity psu

F e μmol/L

350

cFe pFe

May

DOC cFe pFe DOC

20

0 2 4 6 8 10 12 14 16 18

0.0 0.5 1.0 1.5 2.0 2.5 3.0

750

300 400 450 500 550 600 650 700

DOC μmol/L

Salinity psu

F e μmol/L

350

cFe pFe

June

DOC cFe pFe DOC

Figure 9 - Exponential trend lines for cFe, pFe and DOC along

the estuaries during May and June. The black line displays DOC,

the grey line displays cFe and the dashed grey line displays pFe.

I RON ISOTOPES

All determined isotope signatures for cFe are positive in the Kalix Estuary (Figure 11), ranging from 0.37 to 0.45 ‰ during May and from 0.11 to 0.48 ‰ in June. The overall trend during May is a slight decrease with depth.

In June the isotope signature in Kalix Estuary is broadening, at K11 and K21 the isotopic signature is increasing while we determined decreasing profiles at stations K31 and K41. The most distinct change in these profiles is from 5.0 m to 10.0 m depth, with a shift from 0.40 to 0.16 ‰, respectively 0.48 to 0.11 ‰.

In the Råne Estuary cFe shows only positive values, and signatures very similar to Kalix estuary during spring flood in May (0.32 to 0.48 ‰, and 0.37 to 0.49 ‰ during June). In June, the samples show decreasing values at the stations R11 and R21, and slightly increasing signals with depth at the stations R31 DQG5)RUWKHVDPSOHVRIWKH%RWKQLDQ%D\WKHGHWHUPLQDWLRQRI˜

cFe signal was below detection.

0RVW˜

pFe values in the Kalix Estuary are negative (Figure 12). There is no clear trend with depth, the changes between the stations and with depth are small. The isotopic signature ranges from -0.02 to -0.13 ‰ during May and from -0.06 to 0.13 ‰ in June. The pFe isotope signature within the Råne Estuary is negative with the exception of station R21 (June). During May, the

300 350 400 450 500 550 600 650 700 750

0 18 16 14 12 10 8 6 4 2 20

DOC μmol/L

cF e μmol/L

Kalix Estuary Råne Estuary Bothnian Bay

May June

Figure 10 - Scatter plot of cFe and DOC for all sampling locations in May and

June.

overall trend is increasing with depth at all stations. In June, results are more complex, at station R11 the values decrease, while the other stations show an increasing signature.

,Q FRQWUDVW DOO WKH ˜

pFe values in the open Bothnian Bay show a positive signature. The values range from 0.08 to 0.12 ‰ during May and from 0.05 to 0.10 ‰ in June. The overall trend at both stations is slightly decreasing isotopic signature with depth.

Kalix Estuary Råne Estuary

May June

∂

cFe ‰

0.0 0.1 0.2 0.3 0.4 0.5 0.6

0.0

2.0

4.0

6.0

12.0 10.0 8.0

Depth (m)

R31 R21R51 R11

0.0 0.1 0.2 0.3 0.4 0.5 0.6

0.0

2.0

4.0

6.0

12.0 10.0 8.0

K41 K31 K21

K11 12.0

0.0 0.1 0.2 0.3 0.4 0.5 0.6

0.0

2.0

4.0

6.0

10.0 8.0

Depth (m)

R3 R2

R1

0.0 0.1 0.2 0.3 0.4 0.5 0.6

0.0

2.0

4.0

6.0

12.0 10.0 8.0

K1 K2

K3

K4

RÅNE ESTUARY KALIX ESTUARY

M A Y

J U N E

)LJXUH'HSWKGLVWULEXWLRQRI˜

cFe (‰) for Råne

and Kalix Estuary during May and June 2013.

Kalix Estuar y Råne Estuar y Bothnian Ba y

Ma y June Sampling date

0.0-0.14 0.0 2.0 4.0 6.0 12.010.08.0

Depth (m)

0.02-0.02-0.06-0.100.060.100.14 R51R31

R11 A131A51

∂

F e ‰

Depth (m)

0.02-0.02-0.06-0.100.060.100.14 R1 R2R3A5A13

0.0-0.14 0.0 2.0 4.0 6.0 12.010.08.0

Depth (m)

0.02-0.02-0.06-0.10 K2K3

K4

K1 0.0-0.14 0.0 2.0 4.0 6.0 12.010.08.0

Depth (m)

0.02-0.02-0.06-0.10 K11 K31K41K21

RÅNE ESTU AR Y KALIX ESTU AR Y M A Y J U N E

)LJXUH'HSWKGLVWULEXWLRQRI˜

pFe (‰) for Råne and Kalix Estu-

ary during May and June 2013. Bothnian Bay stations are included in

the Råne Estuary graphs.

D ^ISCUSSION

Råne Estuary is a semi-closed water-body, similar to Öre Estuary (Forsgren and Jansson, 1992), partly isolated behind several islands. Kalix Estuary is open to the Bothnian Bay and can exchange water masses easily. In Råne Estuary only station R51 is located in the deep-water zone while in Kalix Estuary station K3 and K4 are located in the deep-water zone (Figure 1). This difference is seen in the salinity, DOC, oxygen and pH depth profiles (Figure 3 and 4). All profiles show a well-mixed layer in the upper 5 m in both estuaries (for pH the upper 10 m layer is well-mixed). In the open Bothnian Bay, the uppermost 10 m are well-mixed. Råne Estuary samples show almost no variation with depth, which suggest a well-mixed water column. In contrast the deepest samples in Kalix Estuary during June, mirror influence of bottom water from the open Bothnian Bay with higher salinity values and lower DOC and oxygen concentrations.

This is known as salt wedge behavior (Holliday and Liss, 1976). In this kind

1.0 2.0 3.0 4.0 5.0 6.0 7.0

0.0

-0.2 -0.1 0.1 0.5

0.2 0.3 0.4 0.6

cFe/pFe δ

Fe ‰

Colloids Particles

Particles Bothnian Bay

May June

Flocculation

Oxidation

)LJXUH6FDWWHUSORWIRU˜

Fe of the colloidal and the particulate phase for all sampling

locations during May and June. Plotted versus cFe/pFe ratio.