MEDDELANDEN från

STOCKHOLM UNIVERSITETS INSTITUTION för

GEOLOGISKA VETENSKAPER No.351

Isotope-based reconstruction of the biogeochemical Si cycle

Implications for climate change and human perturbation

Xiaole Sun

Stockholm 2012

Department of Geological Sciences Stockholm University SE-106 91 Stockholm

Sweden

A dissertation for the Degree of Doctor of Philosophy in Natural Science

Xiaole Sun Department of Geological Sciences Stockholm University SE-10691 Stockholm Sweden

Abstract

The global silicon (Si) cycle is of fundamental importance for the global carbon cycle. Diatom growth in the oceans is a major sequestration pathway for carbon on a global scale (often referred to as the biological pump). Patterns of diatoms preserved in marine sediment records can reveal both natural and anthropogenic driven environmental change, which can be used to understand silicon dynamics and climate change. Si isotopes have been shown to have great potential in order to understand the Si cycle by revealing both past and present patterns of dissolved Si (DSi) utilization, primarily when diatoms form their siliceous frustules (noted as biogenic silica, BSi). However, studies using Si isotopes are still scarce and only a few studies exist where stable Si isotopes are used to investigate the biogeochemical Si cycle in aquatic systems. Therefore, this thesis focuses on developing analytical methods for studying BSi and DSi and also provides tools to understand the observed Si isotope distribution, which may help to understand impacts of climate change and human perturbations on marine ecosystems. The Baltic Sea, one of the biggest estuarine systems in the world, was chosen as the study site. BSi samples from a sediment core in Bothnian Bay, the most northern tip of the Baltic Sea, and diatom samples from the Oder River, draining into the southern Baltic Sea were measured and reported in Paper II and III, after establishing a method for Si isotope measurements (Paper I). Si isotope fractionation during diatom production and dissolution was also investigated in a laboratory-controlled experiment (Paper IV) to validate the observations from the field. The major result is that Si isotope signatures in BSi can be used as an historical archive for diatom growth and also related to changes in climate variables. There is isotopic evidence that the Si cycle has been significantly altered in the Baltic Sea catchment by human activities.

Keywords: diatoms, biogenic silica (BSi), dissolved Si (DSi), Si isotope fractionation, the Baltic Sea

© Sun, Xiaole. Stockholm 2012 ISBN: 978-91-7447-559-3 Cover: top left: picture of Paper I; top right: clean diatoms (by Xiaole Sun); bottom: the Baltic Sea (by Xiaole Sun); middle right: bathymetry of the Baltic Sea (by Xiaole Sun, data source: HELCOM datasets) Printed by US-AB SU, Stockholm 2012

Isotope-based reconstruction of the biogeochemical Si cycle Implications for climate change and human perturbation

Xiaole Sun

Department of Geological Sciences, Stockholm University, SE-10691, Stockholm, Sweden

This doctoral thesis consists of a summary and four papers, referred to by the Roman numerals. Below is the list of the four papers.

Paper I:

Sun, X., Andersson, P., Land, M., Humborg, C., Mörth, C.-M., 2010. Stable silicon isotope analysis on nanomole quantities using MC-ICP-MS with a hexapole gas-collision cell. Journal of Analytical Atomic Spectrometry 25, 156-162.

Paper II:

Sun, X., Andersson, P., Humborg, C., Gustafsson, B., Conley, D.J., Crill, P., Mörth, C.-M., 2011. Climate dependent diatom production is preserved in biogenic Si isotope signatures. Biogeosciences 8, 3491-3499.

Paper III:

Sun, X., Andersson, P., Humborg, C., Pastuszak, M., Mörth, C.-M., 2012. Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system. Submitted to Journal of Chemical exploration.

Paper IV:

Sun, X., Olofsson, M., Andersson, P., Legrand, C., Humborg, C., Mörth, C.-M., 2012. Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system. Submitted to Geochimica et Cosmochimica Acta.

Paper I is Reproduced by permission of The Royal Society of Chemistry, with the link to the paper via http://pubs.rsc.org/en/content/articlelanding/2010/ja/b911113a All the work in this thesis has been carried out by the author with the exception for Si isotope measurements before 2005 (Land, M.) in Paper I, sampling of the sediment core and measurement of BSi content (Conley, D.) and calculation of mixed layer depth (Gustafsson, B.) in Paper II, sampling of diatoms and concentration measurements (Pastuszak, M.) in Paper III, diatom cultivation (Olofsson, M.) in Paper IV. All of the co-authors contributed to the four papers by important scientific discussion and text revision. Xiaole Sun Stockholm, August 2012

To my grandpa, Yao, I know you would have been proud

of what I have achieved today……

Isotope-based reconstruction of the biogeochemical Si cycle

1

1. Introduction

The aquatic ecosystems on Earth have been tremendously influenced by human activities (Carpenter et al., 1992). Ecosystem response depends mainly on changes in the hydrological cycle and variations in environmental cycles of carbon and other nutrients like N, P and Si. Perturbations of these cycles, natural or anthropogenic, have been given considerable interests in the past decades (Chahine, 1992; Falkowski et al., 2000; Gruber and Galloway, 2008; Laruelle et al., 2009). However, until now, most of these studies have focused on carbon budgets due to carbon ubiquity in the biosphere and nitrogen and phosphorus because of eutrophication as a result of human activities. Silicon (Si) is to a lesser extend investigated in the study of nutrient cycling in ecosystems.

The importance of studying the biogeochemical Si cycle arises from its close relation to the global carbon cycle (Berner, 2004). About three quarters of the primary production in coastal and nutrient rich areas of the world oceans is carried out by diatoms, a phytoplankton group that essentially dissolved Si (DSi) to form their unique siliceous frustule, noted as biogenic silica (BSi). In low nutrient areas diatoms still contribute to about 30% of the marine primary production (Nelson et al., 1995). Diatoms transport carbon from atmosphere to deep oceans by sedimentation, a process also known as the “biological pump” (Buesseler, 1998; Goldman, 1988; Nelson et al., 1995). Moreover, diatoms also form the foundation of the shortest desirable food chains, (Cushing, 1989; Ryther, 1969). Diatoms are sensitive to environmental change because of their direct response to many physical, chemical, and biological changes. When DSi concentrations becomes low, it could severely limit diatom biomass build up, leading to ecosystem shifts from production of diatom to non-siliceous algae (Officer and Ryther, 1980), possibly leading to harmful algae blooms and decreased Si export to the open ocean (Dugdale et al., 1995). After diatoms die, preservation of BSi in sediments allows for investigating diatom production over time (Conley and Schelske, 2002).

Silicon is the second most abundant element in the continental crust after oxygen, 28.8 wt% (Wedepohl, 1995), and weathering of silicate minerals are the main source for DSi and silicate containing particles in aquatic systems. From weathering sources to the ocean, DSi transported by rivers serves diatoms as an indispensable nutrient in many biogeochemical processes (Ragueneau et al., 2006; Ragueneau et al., 2000). This makes BSi and DSi potential proxies for investigating environmental change.

Biogenic silica in sediments have been successfully used for reconstructing past environmental changes including climatic variations, for example relating diatoms to temperature (Rosén et al. 2000, Blass et al. 2007, McKay et al. 2008) and to pH and changes in nutrient concentrations (Andren et al., 1999, Braak and Dame 1989, Bennion et al. 1996, Weckström 2006, Olli et al. 2008). However, these studies are associated with large uncertainty because of the difficulty to discriminate between biological and physical processes as driving forces for the observed variations in BSi records. Here this thesis will show that Si isotope signatures in DSi and BSi can be of good complementary use for constraining those uncertainties.

Si isotope fractionation by diatoms

Silicon has three stable isotopes 28Si, 29Si, 30Si with natural abundances of 92.22%, 4.69% and 3.09%, respectively (De Laeter et al., 2003) and one naturally radioactive isotope, 32Si with a half-life of 178±10 years (Nijampurkar et al., 1998).

It has been shown that diatoms preferentially take up 28Si to produce BSi, resulting in an isotopic fractionation (ɛ30/28-value) of approximately -1.1‰ (De La Rocha et al., 1997); Later, the ɛ-value was shown to be -1.08‰ in freshwater (Alleman et al., 2005; Hughes et al., 2011) whereas as a range of -0.6‰ to -2.2‰ for in situ measurements in oceanic water was indicated (Cardinal et al., 2005; Cardinal et al., 2007; De La Rocha et al., 1997; Varela et al., 2004). The documented δ30Si values in BSi of diatoms ranges from -0.3‰ to +2.6‰ (Alleman et al., 2005; Cardinal et al., 2007; De La Rocha et al., 1998; Varela et al., 2004), which is slightly higher values compared to BSi observed in phytoliths produced in terrestrial systems by vascular plants, ranging from -1.7‰ to +2.5‰ (Douthitt, 1982; Ziegler et al., 2005). During the formation of BSi, DSi (residual) is enriched in heavier Si isotopes. In river waters, the measured δ30Si values has been shown to range from +0.4‰ to +3.4‰ (De La Rocha et al., 2000; Ding et al., 2004; Ding et al., 2011; Engström et al., 2010; Georg et al., 2007; Hughes et al., 2011). DSi in a soil solution shows δ30Si values ranging from -0.8‰ to +1.7‰ (Ziegler et al., 2005) and a similar range of -0.2‰ to +1.3‰ is found in groundwater (Georg et al., 2009; Opfergelt et al., 2011). In seawater, the δ30Si values in DSi range from +0.5‰ to +3.2‰ (Beucher et al., 2008; Cardinal et al., 2005; De La Rocha et al., 2000; Reynolds et al., 2006; Varela et al., 2004). Demarest et al. (2009) showed a Si isotope fractionation of -0.55‰ in enriching residual BSi with heavier Si isotopes observed during BSi dissolution of diatoms sampled

Xiaole Sun

2

from the Southern Ocean if > 20% of BSi was dissolved.

The range of observed Si isotope values imply that tracing enriched or depleted BSi caused by diatom-inferred environmental changes and evolvement with time in geological records is possible.

Quantitative relationship between the Si isotopic compositions of BSi and DSi

Because of the discrimination against heavier Si isotopes during diatom production, δ30Si values of DSi and BSi during diatom production follows a Rayleigh distillation behaviour (Fig. 1), which describes the partitioning of Si isotopes in a closed system between two Si reservoirs: DSi and BSi. Comparing a closed and open system, as the fraction of DSi remaining in water (f) decreases i.e. diatom production, the δ³ᵒSi values of DSi and the produced BSi increases. At the same time, δ³ᵒSi values of accumulated BSi in both systems also increase as DSi concentrations decreases, but only up to the initial δ³ᵒSi value of DSi. This means that the accumulated BSi can never reach a δ30Si value higher compared to the initial DSi. δ30Si values of instantly produced BSi in a closed system differs always by -1.1‰ (ɛ-value) compared to remaining DSi, if assuming a fractionation factor (α) of 0.9989 (De La Rocha et al., 1997). In contrast for an open system, the difference between δ30Si values of the whole pools of DSi and BSi is always 1.1‰.

2. Thesis objectives

So far Si isotope data in the literature are still scarce, and only a few studies have attempted to use stable Si isotopes as tracers for understanding the biogeochemical Si cycle and its variations in aquatic systems. Therefore, the aim of this thesis is to provide

Fig. 1 Expected Si isotope variations in a closed and an open system.

data for Si isotope distribution in important Si reservoirs and to show how these data can be used to reconstruct diatom DSi utilization as a function of climate and environmental change, sometimes also superimposed by anthropogenic disturbance.

Four papers each with their specific objectives are formulated as shown in Fig. 2. The work started with a method development for Si isotope measurements (Paper I), and then the methods were applied to two case studies investigating diatom-regulated Si dynamics (Paper II and III). Afterwards the data derived from the case studies were used to validate results from a lab-controlled experiment for investigating Si isotope fractionation patterns during diatom production and dissolution (Paper IV).

Fig. 2 Flow chart of the four papers in the thesis and their specific objectives

0.00.20.40.60.81.0

δ ³ᵒSi (‰

)

f, fraction of remaining DSi

remaining DSi (closed)

accumulated BSi (closed)

instant BSi (closed)

accmulated BSi (open)

Remaining DSi (open)

Method development

•Paper Ι to develop an analytical technique for precisely and accurately determine the stable Si isotopes using MC-ICP-MS.

Methods Method application

•Paper II to investigate the variations in time of diatom production vs. climate change.

•Paper III to examine the enrichment of heavier Si isotopes following seasonal diatom production patterns.

Data Lab-controlled experiment

•Paper IV to investigate the Si isotope fractionation factor during diatom production and dissolution.

Isotope-based reconstruction of the biogeochemical Si cycle

3

3. Site description

The Baltic Sea is located in Northern Europe, from 53°N to 66°N latitude and from 20°E to 26°E longitude and is usually separated into seven basins: Bothnian Bay, Bothnian Sea, Gulf of Finland, Baltic Proper, Gulf of Riga, Arkona and Kattegat (Fig. 3).

The Baltic Sea is a semi-closed brackish water body, but also one of the many aquatic ecosystems in the world that show long-term decreasing DSi concentrations in the water column, especially from 1970 to 1990 (Conley et al., 1993; Sandén et al., 1991). Due to the decrease in DSi concentration the biogeochemical cycle of Si in the Baltic Sea has changed within the 20th century. One of the reasons is reduction of DSi in riverine transport due to retention of Si in dams (Conley, 2002; Humborg et al., 2000). Changes in water flow paths due to river regulations leading to less weathering, is also a possible explanation for reduction in Si concentrations (Humborg et al., 2002; Humborg et al., 2004). Another reason is eutrophication (Conley et al., 1993; Schelske et al., 1983), which stimulate diatom production and simultaneously increases the accumulation of BSi in sediments of the Baltic Sea (Conley et al., 2008).

The diatom spring bloom in the Baltic Sea is sometimes terminated by the depletion of DSi (Smetacek, 1985), especially in the Gulf of Riga DSi limitation may occur during certain periods of the year (Olli et al., 2008). Studies of different aspects of the biogeochemical Si cycle in the Baltic Sea including riverine Si fluxes, estuarine Si fluxes, and concluded that the decreasing DSi concentrations is an on going problem. This decrease might ultimately lead to Si limitation in the entire Baltic Sea. However, recent reports show the decreasing DSi concentrations are levelling out instead of declining further (Papush and Danielsson, 2006).

In all, the Baltic Sea is a good site for studying the interactions between diatoms and Si under human perturbation. In this thesis, two specific sites have been chosen for more detailed studies: one is the northern part of the Baltic Sea, Bothnian Bay and the other is the Oder River draining into the southern Baltic Sea (Fig. 3).

Bothnian Bay

Bothnian Bay (235 499 km2) is the northern portion of the Baltic Sea (Fig. 3). The area has been subjected to recurring glaciations. The last permanent ice melted from the area about 9300 years ago. The average depth of the bay is 43 m and maximum depth is 147 m. The catchment area is sparsely populated and consists mainly of forest, wetland and only a

small amount of agriculture land (~1%). The total water volume of the bay is about 1500 km3, which is 7% of the total water volume of the Baltic Sea. The total water discharge into the Bothnian Bay is 97 km3 yr-1 (Humborg et al., 2008) and 60% of the water is derived from rivers regulated by dams. River regulation with the purpose of generating electricity started in the beginning of the 20th century and continued until ca. 1970, with most of the dam constructions between 1940 and 1970, e.g. Kemijoki and Luleå River (Humborg et al., 2006). A consequence is decreased fluxes of weathering related elements such as base cations and Si (Humborg et al., 2002; Humborg et al., 2000), due to changes in water pathways through the catchments and particle trapping of BSi behind dams (Humborg et al., 2006). This has led to a decreased DSi load in the northern Baltic Sea (Humborg et al., 2007).

Temperature variations in the water column are > 15 °C annually. Formation of a winter ice cover in Bothnian Bay starts from October to the end of November. The entire bay is generally covered by ice, even during the mildest winters, and the ice cover often remains for more than 120 days per year. During the years 1961-1990, the summer (ice-free period) average maximum surface temperature in the open sea is around 16 ᵒC (Dannenberger et al., 1997). A surface thermocline in Bothnian Bay develops about a month earlier than in the other sub-basins of the Baltic Sea and lasts for about three months. Mean DSi concentrations in Bothnian Bay observed between 1970-2001 are 31 µM in winters and decrease to an average concentration of 23 µM in summers (Danielsson et al., 2008). Residence time for DSi is approximate 3.3 years (Papush et al., 2009).

The Oder River

The Oder River is one of the largest rivers in the catchment basin of the Baltic Sea. For the first 112 km from its source, it passes through the Czech Republic, but its middle reach (a distance of 186 km) constitutes the boundary between Poland and Germany before reaching the Baltic Sea via the Szczecin lagoon (Fig. 3). The total length of the Oder River is 854 km, of which 738 km lies in Poland. The total watershed area has been calculated to 119 000 km2, of which about 90% is in Polish territory. The annual discharge of the Oder River is 17 km3 (Radziejewska and Schernewski, 2008). 62% of the river basin is in agricultural use, of which 54% is arable land, 8% grassland and 32% of the area is covered by forest.

Xiaole Sun

4

Fig. 3 Map of the Baltic Sea area (left) and two maps of study sites, Bothnian Bay (upper right) and the Oder River (lower right, data from source of

USGS and HELCOM). BB=Bothnian Bay, BS=Bothnian Sea, GF=Gulf of Finland, GR=Gulf of Riga, BP=Baltic Proper, AR=Arkona, KA=Kattegat.

Isotope-based reconstruction of the biogeochemical Si cycle

5

Increased anthropogenic influence has accelerated eutrophication in the Oder River over the past 50 years. Intensive algal blooms, low water transparency, oxygen depletion in some parts, and fish mortality have become common. The present state of the Oder Lagoon is described as eutrophic, hypertrophic or polytrophic, depending on the trophic system (Radziejewska and Schernewski, 2008).

4. Materials and methods

Silicon isotope analysis using MC-ICP-MS (Multi Collectors Inductively Coupled Plasma Mass Spectrometry) requires a relatively free-of-matrix solution for obtaining precise and accurate results, therefore, a series of sample preparation steps are adopted in this thesis, and these methods can be possibly applied to various types of Si samples (Fig. 4).

Diatom extraction

The method for sediment diatom extraction in Paper II is a procedure modified from Morley et al. (2004). Organic and inorganic carbon was removed by mixing samples with 15% H2O2 and 10% HCl. After washing the samples (centrifuging between each wash) each sample was sieved through a 10 µm sieve cloth and an 80 µm steel mesh to recover the 10-80 µm fraction (which contains most of the diatoms) and to remove silt and clay. The sieved samples were collected in centrifuge tubes containing sodium polytungstate (SPT). The density of the SPT was adjusted to be between 1.8 g cm-3 and 2.3 g cm-3 to separate diatoms from the remaining residuals. This was followed by drying diatom samples at 40 ˚C.

Diatom alkaline digestion

Clean diatom samples derived from sediments in Paper II and from filtrated river water in Paper III were fused by using solid sodium hydroxide (NaOH) in Ag crucibles at 730°C in a muffle furnace and thereafter washed with MQ-e water and collected in a 0.12 M HCl solution (Georg et al., 2006b).

Another wet digestion method was used in Paper IV, which was adopted from Ragueneau et al. (2005). A 0.25 mol L-1 NaOH solution was added into tubes which contained freeze-dried diatoms collected on polyethersulfone filters. HCl was used to stop the digestion by neutralising the pH. After centrifugation, the supernatant was diluted for further treatment.

DSi extraction from seawater

Dissolved Si in the Baltic Sea water samples in Paper IV was extracted by a magnesium two-step co-

precipitation technique adopted from Reynolds et al. (2006). A 1 mol L-1 NaOH solution was added into seawater samples and this was repeated twice. After each addition, samples were shaken and left for 1 hour for precipitating Mg(OH)2 with Si co-precipitated, and thereafter samples were centrifuged to collect the precipitates. The precipitates were dissolved in 4 mol L-1 HCl. The Si concentration of the removed supernatant was checked to make sure that the Si removal was complete.

Chromatographic purification

The final solutions derived from the methods contain a cationic matrix. Therefore a chromatographic separation method is required to purify the Si solution prior to Si isotope analysis. The used purification process was modified from Georg et al. (2006b). Cation exchange resin AG 50W-X12 (100-200mesh) in H+ form was filled in chromatography columns. The samples were loaded on the column, in form of H4SiO4, which passed through the resin whereas the resin retains the cations.

Silicon isotope analyses

Silicon isotope analyses are described in detail in Paper I using an IsoProbe MC-ICP-MS equipped with a hexapole collision cell at the Laboratory for Isotope Geology, Swedish Museum of Natural History. Due to instrument limitation of only having low-resolution mode (m/Δm ~ 450), measurement of δ³o Si values is not possible with the current set up. The instrumental mass discrimination caused by a number of processes in the plasma and in the transmission through the ion optical system (Becker, 2007) was corrected by using the standard-sample-standard bracketing technique. The sensitivity for 28Si is ~12 V/ 36 µmol L-1 Si (1 mg L-1), with an uptake rate of ~70 µl min-1 via a Cetac® Aridus desolvating nebulizer. The Si isotope ratio of each sample is calculated and is given in δ29Si values (‰) relative to the accepted reference material for silicon isotopes, NBS 28 according to Eq.1. All of the δ30Si values in this thesis are calculated from the relationship δ30Si = δ29Si×1.96, assuming mass-dependent fractionation (Reynolds et al., 2007).

Eq.1

29

28

sample29

29

28

standard

1 1000

Si

SiSi

Si

Si

Xiaole Sun

6

Fig. 4 Sample preparation steps prior to Si isotope analysis using MC-ICP-MS.

MC-ICP-MS for Si isotope analysis

15% H2O

2

10% HCl Seive 10-80 µm

SPT

1.8 – 2.3 g/cm3 Clean diatoms after drying

Diatom extraction from sediments

Solid NaOH

Clean diatoms

0.25 mol L-1

NaOH

Store in

0.12 M HCl

Diatom alkaline digestion

1 mol L-1

NaOH (twice)

DSi extraction from seawater

Store in

0.12 M HCl

Precipitates dissolved in HCl

Chromatographic purification

Isotope-based reconstruction of the biogeochemical Si cycle

7

5. Summary of results

The results of the four papers are briefly described below.

Paper I: Stable silicon isotope analysis on nanomole quantities using MC-ICP-MS with a hexapole gas-collision cell

Paper I reports a method developed for precisely and accurately measuring the δ29Si value (2S.D., 0.2‰) using MC-ICP-MS with a total Si consumption of less than 14 nmol (390 ng Si). Data was collected during a four-year period (2005-2009) and the average δ29Si value of IRMM-018 relative to NBS-28 was measured to -0.95‰ (n=23, 2S.D. 0.16‰), with a 95% confidence interval of ±0.028‰. The mean δ29Si value of the Big-Batch standard was found to be -5.50‰ (n=6, 2S.D. 0.26‰). This demonstrates that the used methods give long-term reproducibility with respect to both precision and accuracy. However, it was also shown that the presence of cations, such as Mg, Na, Al etc., can cause severe matrix effects on the Si isotope measurements.

Paper II: Climate dependent diatom production is preserved in biogenic silica isotope signatures

Paper II is the first case study using the method developed in Paper I. The obtained Si isotope values in diatoms are used to reconstruct diatom production in Bothnian Bay. The data are based on down-core

analysis of Si isotope distribution in BSi covering the period 1820 to 2000. The sediment core is dated using 210Pb. Results show a linear 210Pb decay relation with depth in the sediment core (r2 = 0.97). This linearity suggests minimal bioturbation, indicating an average sedimentation rate in this Bothnian Bay core of approximate 2.1mm yr−1. The δ30Si values of BSi ranges between -0.18‰ and +0.58‰. By assuming Rayleigh distillation, diatom production can be calculated as fraction of remaining DSi (f values) inferred from measured Si isotope values. This reveals that the sediment core record can be divided into two periods, an unperturbed period from 1820 to 1950 and another period affected by human activities from 1950 to 2000 (Fig. 5). The shift in Si isotope values toward higher values after 1950 is most likely caused by large scale damming of rivers.

The production is shown to be correlated with air and water temperature, which in turn were correlated with the mixed layer depth (ML). The deeper ML depth observed in colder years resulted in lower diatom production. These finding corroborated by pelagic investigations in the 1990’s, have clearly shown that the ML depth controls diatom production in the Baltic Sea (Wasmund et al., 1998). Especially after cold winters and deep water mixing, diatom production was low and DSi concentrations were not depleted in the water column after the spring bloom.

Fig. 5 Fraction of the remaining DSi (f values) in the Bothnian Bay water column and f values calculated from observations during the summers 1980 to 2000, plotted together with average summer air temperature through years. The inferior and superior error bars on f values are the first quartile and third quartile for each f value derived from the Monte Carlo simulations (Paper II).

10

12

14

16

18

0

0.2

0.4

0.6

0.8

1

1800 1850 1900 1950 2000

Av

erag

e su

mm

er T

emp

erat

ure

(°C

)

Fra

ctio

n o

f re

mai

nin

g D

Si in

wat

er (

f)

Year

Observed remaining DSiCalculated f (δ³ᵒSi = 1.1‰) Calculated f (δ³ᵒSi = 1.4‰) Annual average summer temp.5 per. Mov. Avg. (Annual average summer temp.)

Xiaole Sun

8

Paper III: Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system

The Oder River is a eutrophied river entering the southern Baltic Sea. The river concentrations in DSi, DIN and DIP display similar variations, i.e. decreasing throughout the spring and summer and increasing in the autumn. In contrast to the dissolved nutrients, BSi shows increasing concentration, which indicates a period of rapid diatom growth in spring and summer, followed by decreasing concentration in late summer and autumn. The analysis of Si:N:P ratios indicates that the diatom production in the Oder River is limited by different nutrients throughout the studied period.

The δ³ᵒSi values increased from a minimum value of +0.75‰ in the spring to a maximum of +3.05‰ at the beginning of August, which are some of the highest values ever recorded in freshwater diatoms. The δ³ᵒSi values started to decrease in late August, and in September the δ³ᵒSi values reached +1.09‰, similar to observed values in April. We suggest that dissolution of BSi becomes more important later in the summer when phosphate limits the diatom production. The Rayleigh model used here to predict δ30Si values of DSi suggests that the initial value is close to +2‰ (before the start of the diatom bloom). This means that there is also a biologic control of the Si entering the river and this is probably caused by Si isotope fractionation during uptake of Si in for example phytoliths.

Paper IV: Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system

Previous studies imply that knowledge about Si isotope fractionation is required for a better understanding of the biogeochemical cycle of Si, in estuarine and coastal areas where primary production stands for some 30-50% of the global marine production. Paper IV reports systematical measurements of variations in Si isotope values of DSi and BSi following diatom growth and dissolution of BSi to investigate Si isotope fractionation patterns during these processes. Two species of diatoms from the Baltic Sea, were selected, Thalassiosira baltica (TBTV1) and Skeletonema marinoi (SMTV1). The Chl a (chlorophyll a) concentrations of the two species increased exponentially, while DSi concentrations decreased during the initial growth period, reaching a level of 200 µg L-1 for TBTV1 and 530 µg L-1 for SMTV1. The DSi concentrations during growth days decreased progressively and the experiment was terminated when DSi concentrations were below 0.4 µmol L-1. The Si isotope values measured in the two species varied by more than 0.7‰ for δ 29Si and 1.3‰ for δ 30Si, ranging from +0.6‰ up to +1.3‰ for δ 29Si, and from +1.2‰ up to +2.6‰ for δ 30Si. The measured δ29Si values in seawater samples increased from +1.35‰ to +2.58‰, δ30Si values from +2.64‰ to +5.07‰, calculated by multiplying δ29Si values with 1.96 (Reynolds et al., 2007). The two species fractionate the Si isotopes during their growth identically (Fig. 6); 29α fractionation factors were 0.9993 ± 0.0004 (2σ, TBTV1) and 0.9992 ± 0.0004 (2σ, SMTV1). The average value for 29α for all diatom sample measurements was 0.99925 ± 0.0003 (2σ). The calculated average value for 30α was 0.9984 ± 0.0004 (2σ).

Fig. 6 Plot of Si isotope compositions of diatoms and DSi following diatom growth as a function of f values (fraction of remaining DSi in water). Lines are expected change for Rayleigh fractionation behaviour in a closed system. Error bars of Si isotope values are 2σ (Paper IV).

0

2

4

6

8

00.51

δ³ᵒSi, ‰

f, fraction of remaining DSi in water

δ³°Si of TBTV1 δ³°Si of DSi DSi (Rayleigh)BSi (Rayleigh)

0

2

4

6

8

00.20.40.60.81

δ³ᵒSi, ‰

f, fraction of remaining DSi in water

δ³°Si of SMTV1 δ³°Si of DSi DSi (Rayleigh)BSi (Rayleigh)

Isotope-based reconstruction of the biogeochemical Si cycle

9

During the dissolution experiment, the Chl a concentrations of SMTV1 gradually decreased to 267 µg L-1, while the DSi concentration increased steadily for 20 days to 50 µmol L-1, and was constant (50-60 µmol L-1) for another 20 days, indicating dissolution of diatom frustules. During the first 25 days in the dark, the siliceous frustules were linearly dissolved (r2=0.9) and 75% of the Si was released back to the medium. Thereafter, during the last 10 days, 90% of the initial BSi was dissolved and released back to the medium. The Si isotope values of SMTV1 and DSi seemed constant during dissolution. Calculation gives the Si isotope fractionation factor during dissolution for 29Si and 30Si as 0.9999 ± 0.0002 and 0.99974 ± 0.00022 (2σ). This is in contrast to findings from an open ocean species, where Si isotope fractionation during dissolution was observed (Demarest et al. 2009).

6. Discussion

Reconstruction of DSi utilization by δ³ᵒSi values

To understand the biogeochemical Si cycle and its link with climate, Si isotope signatures of BSi have been used to trace diatom production by reconstructing DSi utilization in oceans e.g. Reynolds et al. (2006), Beucher et al. (2008), De la Rocha et al. (1998), van den Boorn et al. (2010) and in lakes, e.g. Street-Perrott et al. (2008), Opfergelt et al. (2011). However, to date, no studies have reconstructed DSi utilization by diatoms in estuaries or in coastal areas. Estuaries, with gradients in water turnover times, salinity, nutrients and primary production, are also considered to be of significance for the global biogeochemical nutrient and carbon cycles (Fry, 2002; Voss et al., 2000). Paper II attempts to reconstruct DSi utilization during the past 200 years using a sedimentary record of δ³ᵒSi values in preserved BSi in Bothnian Bay.

Important to all of these applications is the uncertainty in isotope fractionation during Si uptake by diatoms used in the Rayleigh model (Fig. 1), no matter which aquatic ecosystems diatoms live in. A few studies have investigated the variability in the Si isotope fractionation factor during diatom growth among different species (De La Rocha et al., 1997), in the oceans (Street-Perrott et al., 2008; Varela et al., 2004) and in lakes (Alleman et al., 2005). A study by Demarest et al. (2009) also looked into the Si isotope fractionation during diatom dissolution and showed that a Si isotope fractionation factor of -0.55‰ during dissolution may complicate this Si isotope-based reconstruction of DSi utilization. Again, there are no such studies in estuaries. Paper IV examines Si

isotope fractionation patterns during BSi production and dissolution using diatoms isolated from the Baltic Sea. The two species of diatoms examined exhibit an identical Si isotope fractionation factor during growth. This is in agreement with the Si isotope fractionation factors in the oceans and in freshwater. However, no Si isotope fractionation during dissolution was observed, even after 90% of the diatoms were dissolved. This may be attributed to the smaller size of the diatoms living in estuarine systems with lower salinity compared to the open ocean. This suggests that information about the original environmental conditions in estuarine and even coastal systems is directly retained. However, in terms of the complexity of the mechanism behind those biological/physical/ chemical processes, these kinds of studies are important but are still scarce.

Our knowledge of the current diatom production in water and the historical diatom production derived from sedimentary records is largely a result of applying the Rayleigh model. The Si isotope fractionation factor is the most important parameter in the model. Higher Si isotope fractionation factor can cause overestimates of DSi utilization; likewise lower Si isotope fractionation factor can result in underestimates of DSi utilization. The problem to estimate DSi utilization can lead to erroneous conclusions about effects from for example climate change. The variability in Si isotope fractionation should be studied in detail in order to constrain the uncertainty in Si isotope data.

Si isotope-inferred climate variations

One important reason to trace diatom dynamics is the sensitivity of diatoms to changes in environmental conditions, for example temperature. Historical records of air and water temperature are required to be able to understand the causes and mechanisms behind current climate change and impact from human activities. Especially high-latitude systems are predicted to experience the greatest impact from climate fluctuations (Carpenter et al., 1992; Douglas et al., 1994; Smol et al., 2005). Annual average surface water temperature in the Baltic Sea has increased by 1.4 ᵒC during the last 100 years, which is higher compared to 63 other studied large marine ecosystems (Mackenzie and Schiedek, 2007). The widely used method for reconstructing temperature is by analyzing oxygen isotope compositions of diatoms, because of a dependence of oxygen isotope fractionation between BSi and water (Swann and Leng, 2009). Variation in BSi content in sediments is also used to reconstruct air temperature (Blass et al., 2007; McKay et al., 2008).

Xiaole Sun

10

A major result from this thesis is to provide a possible way of connecting δ³ᵒSi values and temperature. This is done by sedimentary δ³ᵒSi BSi values that can be used as an archive for estimating the fraction of DSi utilization (Paper II). This fraction is a function of water turbulence and light condition, which in turn is a function of temperature. The correlation implies that higher temperature causes higher diatom production due to a shallower mixed layer depth, i.e. more diatoms can float to water surface to obtain enough light. Lower temperature leads to lower diatom production due to a deeper mixed layer depth without enough light penetration.

Temperature is predicted to increase in the future (IPCC, 2007), which in general stimulates diatom production. A model developed by Omstedt et al. (2009) indicates that the Baltic Sea acts as a source for CO2 before 1950, but after 1950, the increased primary production enhances the uptake of CO2 due to eutrophication, which makes the Baltic Sea work as both a sink and source of CO2. This suggests that temperature-induced increased diatom production in the future may continue the enhancement of CO2 uptake and possibly shift the Baltic Sea to be a sink for CO2.

Although the sediment records could reveal the past climate and models could help the understanding of the ecosystem’s future, the effect of climate change on ecosystems are difficult to predict. Climate-induced increased water flux also increases nutrient transport to the Baltic Sea, which in turn stimulates diatom growth. It is clear that increased biological processes will have a major influence on the conditions for biota in aquatic systems. This will also affect species composition, distributions, and interactions in ways that are only partly understood at the present time.

Si isotope-inferred human perturbation

Recently, the biogeochemical Si cycle has also been perturbed by human activities. Eutrophication and hydrological alterations have been shown to cause DSi depletion and diatom growth in the Baltic Sea (Conley et al., 2008; Conley et al., 1993; Humborg et al., 2006; Humborg et al., 2008). This thesis shows that both eutrophication and hydrological alterations have most likely resulted in increasing δ³ᵒSi values in the Baltic Sea.

The heavily eutrophied Oder River studied in Paper III is subject to nutrient-limited diatom growth. Depletion of DSi during the excessive diatom bloom leads to very high δ³ᵒSi values in produced BSi, indicating that the Oder River is a strong source for Si with high δ³ᵒSi values, which can probably be traced in the Baltic Sea.

Paper II suggests that there is a shift in δ³ᵒSi values from +1.1‰ measured in a pristine Kalix river (Engström et al., 2010) to +1.4‰ (assumed in Paper II by fitting data into field measurements). This is most likely caused by river damming. Increased BSi behind dams decreases DSi concentrations, leading to increasing δ³ᵒSi values in both BSi and DSi in river water.

By examining Si isotope fractionation during diatom production and dissolution (Paper IV), different Si isotope patterns in the Baltic Sea basins can be suggested. Si-limited diatom production is observed in Gulf of Riga (Olli et al., 2008), which predicts δ³ᵒSi values for BSi of +0.85‰ resulting in the δ³ᵒSi value of residual DSi of +3.64‰. Paper IV also shows that the entire Baltic Sea may face higher δ³ᵒSi values in DSi although a Si-limited diatom production has not been observed in other basins yet. This forecast is supported by Conley et al (2008), where it is suggested that DSi will be a limiting nutrient in the Baltic Sea in the near future.

Enhanced diatom production can possibly induce oxygen depletion in the bottom water by exporting organic matter into bottom water, accelerating development of hypoxia in the Baltic Sea. This can lead to changes in biogeochemical cycles and habitat loss for fish and other biota (Conley et al., 2011; Conley and Johnstone, 1995) since sufficient oxygen is essential to support healthy biological communities (Vaquer-Sunyer and Duarte, 2009). Kabel et al. (2012) concluded that increasing temperature can substantially deteriorate bottom water conditions in the Baltic Sea by stimulating cyanobacteria blooms. Si isotope values of BSi can be used to estimate the order of magnitude of diatom blooms from a given year by analysing sediment samples. By comparing the results with cyanobacteria blooms, the findings may help us to understand what is most important for the hypoxia in the Baltic Sea: the diatom spring bloom or the summer cyanobacteria bloom. However, to what degree the hypoxic areas influences nutrient biogeochemical cycles are not yet fully understood. More studies with multiple approaches and isotope-based techniques are therefore required.

Implications for control of δ³ᵒSi values of Si inputs to the oceans

About 85% of the DSi delivered to the oceans is carried by rivers, which constitute the major Si source for diatom production in the ocean. Rivers may exhibit large variability in concentration and fluxes of DSi and Si isotope signatures. This variability can drive a shift in Si isotopic compositions of BSi in surface waters of the ocean by changing the relative degree of DSi utilization (and therefore the Rayleigh

Isotope-based reconstruction of the biogeochemical Si cycle

11

model). This will finally change the Si isotope values of BSi preserved in sediments and thus influence the sedimentary records. Therefore, river inputs with variable Si flux and δ³ᵒSi values carry important information for the interpretations of the oceanic Si cycles.

Many processes can cause the variability of δ³ᵒSi values of DSi and BSi in rivers over time, such as river damming (Paper II) and uptake of Si by diatoms (Paper III), phytoliths (Ding et al., 2011) and terrestrial plants (Ding et al., 2008). Another process linked to climate is weathering, which also generates varying δ³ᵒSi values, which are also linked climate (Georg et al., 2006a). An important process is retention of DSi and BSi in estuaries (Conley and Malone, 1992; Treguer et al., 1995). Estuaries can be effective traps for DSi and BSi and prevent Si transport to the ocean. However, these processes will be difficult to distinguish from each other since Si isotope fractionation is always small, only up to a few permil. Analysis of δ³ᵒSi values combined with analysis of other indicators, such as Ge:Si and Al:Si ratios may provide additional information about processes controlling isotope composition of Si inputs to the oceans. Investigation of Si isotope signature distribution and behavior in rivers and estuaries will help to reveal internal and external Si source control in estuarine systems over time and also have implications for understanding δ³ᵒSi records in the ocean.

7. Future perspectives

More studies on historical records of climate and environmental changes are needed to understand the past to improve our possibility to predict the future variations. Factors controlling uptake, storage and recycling of Si in and estuarine and coastal systems must be studied in detail. This is important in order to quantitatively assess the response of ecosystems to changes in Si inputs and climate and what impacts these changes will affect the Si cycle in the ocean. These factors will also help to constrain Si budgets for systems with different geological settings in both perturbed and pristine systems. It is also essential that future-monitoring programs includes DSi and BSi measurements and that the monitoring should be extended to samples taken before and after building of dams, as well as river mouths. This will ensure datasets and time series for evaluation and modeling.

Modeling the Si cycle both on global and regional scale is also required for constraining the uncertainties associated with technical, practical and economic limitations, and for revealing how the

resilience of diatom-based ecosystems responding to possible ecosystem changes. The future priorities should be pointed out as well.

The Si isotope fractionation factor should be carefully studied before interpreting Si isotope data in different systems and Si isotope compositions in rivers also need more studies to better understand land-sea fluxes of Si.

Acknowledgements

Time flies! Now it is time to finish my doctoral thesis and have the chance to thank the people who have helped and supported me in all kinds of ways in the past four years.

First of all, I would like to express my sincere gratitude to my supervisors, Carl-Magnus Mörth, Christoph Humborg and Per Andersson. Thank you, Magnus, for offering me the opportunity to walk on this great journey in Stockholm, and for your heartfelt encouragements and supports along the way. I will never forget how many good ideas and inspirations you offered me. Thank you, Christoph, for your resourceful ideas and constructive feedbacks along my work. Thank you, Per, not only for your very helpful guidance and invaluable discussions, but also for great Sushi parties we had together every year. It is fantastic to work with you all!

I am very grateful for my co-authors, Bo Gustafsson, Patrick Crill, Daniel Conley, Marianna Pastuszak, Martin Olofsson, Catherine Legrand and Magnus Land for scientific contributions and insightful comments to greatly improve the manuscripts.

Many thanks to Klara Hajnal and Heike Siegmund for so kindly helping me with lab work and sample measurements; Many thanks also to Hans Schöberg, Torsten Persson and Karin Wallner for assisting me with sample preparation and providing technical knowledge for MC-ICP-MS.

I want to express my gratitude to the Geochemistry group for insightful discussions during lunch seminars. I would like to particularly thank Volker Brüchert for offering me opportunities to be a teaching assistant several times from which I have learnt a lot.

Special thanks to all of my cute former and present fellow PhD students. We were having a lot of fun together.

I am also grateful for the help from Eve Arnold, Monica Rosenblom and Elisabeth Däcker with all of administrative work.

Xiaole Sun

12

Many thanks to all the former and present colleagues at Department of Geological Sciences, Stockholm University and Laboratory for Isotope Geology, Swedish Museum of Natural History for creating a warm working environment and making me so pleasant during my stay in Stockholm.

Funding from the Swedish Research Council (VR. 2007-4731) is greatly acknowledged.

With love and gratitude, I would like to thank all my family and friends in Sweden and in China. The greatest thank you is deserved by my Mummy, my Daddy and my Deguo, for your endless love and always believing in me. Without your love and encouragements, this journey would have been much more difficult.

References

Alleman, L. Y., Cardinal, D., Cocquyt, C., Plisnier, P.-D., Descy, J.-P., Kimirei, I., Sinyinza, D., and André, L., 2005, Silicon Isotopic Fractionation in Lake Tanganyika and Its Main Tributaries: Journal of Great Lakes Research, v. 31, no. 4, p. 509-519.

Becker, J. S., 2007, Inorganic Mass Spectrometry: Principles and Applications, Wiley, 514 p.:

Berner, R. A., 2004, The Phanerozoic Carbon Cycle: CO2 and O2, New York, Oxford University Press.

Beucher, C. P., Brzezinski, M. A., and Jones, J. L., 2008, Sources and biological fractionation of Silicon isotopes in the Eastern Equatorial Pacific: Geochimica et Cosmochimica Acta, v. 72, no. 13, p. 3063-3073.

Blass, A., Bigler, C., Grosjean, M., and Sturm, M., 2007, Decadal-scale autumn temperature reconstruction back to AD 1580 inferred from the varved sediments of Lake Silvaplana (southeastern Swiss Alps): Quaternary Research, v. 68, no. 2, p. 184-195.

Buesseler, K. O., 1998, The decoupling of production and particulate export in the surface ocean: Global Biogeochemical Cycle, v. 12, no. 2, p. 297-310.

Cardinal, D., Alleman, L. Y., Dehairs, F., Savoye, N., Trull, T. W., and André, L., 2005, Relevance of silicon isotopes to Si-nutrient utilization and Si-source assessment in Antarctic waters: Global Biogeochemical Cycle, v. 19, no. GB2007.

Cardinal, D., Savoye, N., Trull, T. W., Dehairs, F., Kopczynska, E. E., Fripiat, F., Tison, J.-L., and André, L., 2007, Silicon isotopes in spring Southern Ocean diatoms: Large zonal changes despite homogeneity among size fractions: Marine Chemistry, v. 106, no. 1-2, p. 46-62.

Carpenter, S. R., Fisher, S. G., Grimm, N. B., and Kitchell, J. F., 1992, Global Change and Freshwater

Ecosystems: Annual Review of Ecology and Systematics, v. 23, p. 119-139.

Chahine, M. T., 1992, The hydrological cycle and its influence on climate: Nature, v. 359, no. 6394, p. 373-380.

Conley, D. J., 2002, Terrestrial ecosystems and the global biogeochemical silica cycle: Global Biogeochem. Cycles, v. 16, p. 1121-1128.

Conley, D. J., Carstensen, J., Aigars, J., Axe, P., Bonsdorff, ., remina, T., aahti, B.- ., umborg, C., onsson, P., otta, ., a nnegren, C., arsson, ., aximov, ., edina, . ., ysia -Pastus a , ., emei aite - i iene , ., alve, ., ilhelms, S., and ille n, ., 2011, ypoxia s ncreasing in the Coastal Zone of the Baltic Sea: Environmental Science & Technology, v. 45, no. 16, p. 6777-6783.

Conley, D. J., Humborg, C., Smedberg, E., Rahm, L., Papush, L., Danielsson, Å., Clarke, A., Pastuszak, M., Aigars, J., Ciuffa, D., and Mörth, C.-M., 2008, Past, present and future state of the biogeochemical Si cycle in the Baltic Sea: Journal of Marine Systems, v. 73, no. 3-4, p. 338-346.

Conley, D. J., and Johnstone, R. W., 1995, Biogeochemistry of N, P and Si in Baltic Sea sediments: response to a simulated deposition of a spring diatom bloom: Marine Ecology Progress Series, v. 122, p. 265-276.

Conley, D. J., and Malone, T. C., 1992, Annual cycle of dissolved silicate in Chesapeake bay: implications for the production and fate of phytoplankton biomass Marine Ecology Progress Series, v. 81, no. 2, p. 121-128.

Conley, D. J., Schelske, C. L., and Stoermer, E. F., 1993, Modification of the biogeochemical cycle of silica with eutrophication: Marine Ecology Progress Series, v. 101, p. 179-192.

Cushing, D. H., 1989, A difference in structure between ecosystems in strongly stratified waters and in those that are only weakly stratified: J. Plankton Res., v. 11, no. 1, p. 1-13.

Danielsson, Papush, L., and Rahm, L., 2008, Alterations in nutrient limitations -- Scenarios of a changing Baltic Sea: Journal of Marine Systems, v. 73, no. 3-4, p. 263-283.

Dannenberger, D., Andersson, R., and Rappe, C., 1997, Levels and patterns of polychlorinated dibenzo-p-dioxins, dibenzofurans and biphenyls in surface sediments from the western Baltic Sea (Arkona Basin) and the Oder River estuarine system: Marine Pollution Bulletin, v. 34, no. 12, p. 1016-1024.

De La Rocha, C. L., Brzezinski, M. A., and DeNiro, M. J., 1997, Fractionation of silicon isotopes by marine diatoms during biogenic silica formation:

Isotope-based reconstruction of the biogeochemical Si cycle

13

Geochimica et Cosmochimica Acta, v. 61, no. 23, p. 5051-5056.

-, 2000, A first look at the distribution of the stable isotopes of silicon in natural waters: Geochimica et Cosmochimica Acta, v. 64, no. 14, p. 2467-2477.

De La Rocha, C. L., Brzezinski, M. A., DeNiro, M. J., and Shemesh, A., 1998, Silicon-isotope composition of diatoms as an indicator of past oceanic change: Nature, v. 395, no. 6703, p. 680-683.

De Laeter, J. R., Böhlke, J. K., De Bièvre, P., Hidaka, H., Peiser, H. S., R., R. K. J., and Taylor, P. D. P., 2003, Atomic weights of the elements. Review 2000 (IUPAC Technical Report): Pure and Applied Chemistry v. 75, no. 6, p. 683-799.

Demarest, M. S., Brzezinski, M. A., and Beucher, C. P., 2009, Fractionation of silicon isotopes during biogenic silica dissolution: Geochimica et Cosmochimica Acta, v. 73, no. 19, p. 5572-5583.

Ding, T., Wan, D., Wang, C., and Zhang, F., 2004, Silicon isotope compositions of dissolved silicon and suspended matter in the Yangtze River, China: Geochimica et Cosmochimica Acta, v. 68, no. 2, p. 205-216.

Ding, T. P., Gao, J. F., Tian, S. H., Wang, H. B., and Li, M., 2011, Silicon isotopic composition of dissolved silicon and suspended particulate matter in the Yellow River, China, with implications for the global silicon cycle: Geochimica et Cosmochimica Acta, v. 75, no. 21, p. 6672-6689.

Ding, T. P., Zhou, J. X., Wan, D. F., Chen, Z. Y., Wang, C. Y., and Zhang, F., 2008, Silicon isotope fractionation in bamboo and its significance to the biogeochemical cycle of silicon: Geochimica et Cosmochimica Acta, v. 72, no. 5, p. 1381-1395.

Douglas, M. S. V., Smol, J. P., and Blake Jr, W., 1994, Marked post-18th century environmental change in high-arctic ecosystems: Science, v. 266, no. 5184, p. 416-419.

Douthitt, C. B., 1982, The geochemistry of the stable isotopes of silicon: Geochimica et Cosmochimica Acta, v. 46, no. 8, p. 1449-1458.

Dugdale, R. C., Wilkerson, F. P., and Minas, H. J., 1995, The role of a silicate pump in driving new production: Deep Sea Research Part I: Oceanographic Research Papers, v. 42, no. 5, p. 697-719.

Engström, E., Rodushkin, I., Ingri, J., Baxter, D. C., Ecke, F., Österlund, H., and Öhlander, B., 2010, Temporal isotopic variations of dissolved silicon in a pristine boreal river: Chemical Geology, v. 271, no. 3-4, p. 142-152.

Falkowski, P., Scholes, R. J., Boyle, E., Canadell, J., Canfield, D., Elser, J., Gruber, N., Hibbard, K., Högberg, P., Linder, S., Mackenzie, F. T., Moore III, B., Pedersen, T., Rosenthal, Y., Seitzinger, S.,

Smetacek, V., and Steffen, W., 2000, The Global Carbon Cycle: A Test of Our Knowledge of Earth as a System: Science, v. 290, no. 5490, p. 291-296.

Fry, B., 2002, Conservative mixing of stable isotopes across estuarine salinity gradients: A conceptual framework for monitoring watershed influences on downstream fisheries production: Estuaries and Coasts, v. 25, no. 2, p. 264-271.

Georg, R. B., Reynolds, B. C., Frank, M., and Halliday, A. N., 2006a, Mechanisms controlling the silicon isotopic compositions of river waters: Earth and Planetary Science Letters, v. 249, no. 3-4, p. 290-306.

-, 2006b, New sample preparation techniques for the determination of Si isotopic compositions using MC-ICP-MS: Chemical Geology, v. 235, no. 1-2, p. 95-104.

Georg, R. B., Reynolds, B. C., West, A. J., Burton, K. W., and Halliday, A. N., 2007, Silicon isotope variations accompanying basalt weathering in Iceland: Earth and Planetary Science Letters, v. 261, no. 3-4, p. 476-490.

Georg, R. B., West, A. J., Basu, A. R., and Halliday, A. N., 2009, Silicon fluxes and isotope composition of direct groundwater discharge into the Bay of Bengal and the effect on the global ocean silicon isotope budget: Earth and Planetary Science Letters, v. 283, no. 1-4, p. 67-74.

Goldman, J. G., 1988, Spatial and temporal discontinuities of biological processes in pelagic surface waters in Rothschild, B. J., ed., Toward a theory on biological-physical interactions in the world ocean: Dordrecht, Kluwer Academic Publishing, p. 273-296.

Gruber, N., and Galloway, J. N., 2008, An Earth-system perspective of the global nitrogen cycle: Nature, v. 451, no. 7176, p. 293-296.

Hughes, H. J., Sondag, F., Cocquyt, C., Laraque, A., Pandi, A., André, L., and Cardinal, D., 2011, Effect of seasonal biogenic silica variations on dissolved silicon fluxes and isotopic signatures in the Congo River: Limnology and Oceanography, v. 56, no. 2, p. 551-561.

Humborg, C., Blomqvist, S., Avsan, E., Bergensund, Y., Smedberg, E., Brink, J., and Mörth, C.-M., 2002, Hydrological alterations with river damming in northern Sweden: Implications for weathering and river biogeochemistry: Global Biogeochem. Cycles, v. 16, p. 1039-1051.

Humborg, C., Conley, D. J., Rahm, L., Wulff, F., Cociasu, A., and Ittekkot, V., 2000, Silicon Retention in River Basins: Far-reaching Effects on Biogeochemistry and Aquatic Food Webs in Coastal Marine Environments: AMBIO, v. 29, no. 1, p. 45-50.

Xiaole Sun

14

Humborg, C., Mörth, C. M., Sundbom, M., and Wulff, F., 2007, Riverine transport of biogenic elements to the Baltic Sea - past and possible future perspectives: Hydrology and Earth System Sciences, v. 11, p. 1593-1607.

Humborg, C., Pastuszak, M., Aigars, J., Siegmund, H., Mörth, C. M., and Ittekkot, V., 2006, Decreased Silica Land-sea Fluxes through Damming in the Baltic Sea Catchment-Significance of Particle Trapping and Hydrological Alterations: Biogeochemistry, v. 77, no. 2, p. 265-281.

Humborg, C., Smedberg, E., Blomqvist, S., Mörth, C.-M., Brink, J., Rahm, L., Danielsson, Å., and Sahlberg, J., 2004, Nutrient variations in boreal and subarctic Swedish Rivers: Landscape control of land-sea fluxes Limnology and Oceanography, v. 49, no. 5, p. 1871-1883.

Humborg, C., Smedberg, E., Medina, M. R., and Mörth, C.-M., 2008, Changes in dissolved silicate loads to the Baltic Sea -- The effects of lakes and reservoirs: Journal of Marine Systems, v. 73, no. 3-4, p. 223-235.

IPCC, 2007, Intergovernmental panel on climate change, WMO, UNEP. Climate change 2007.

Kabel, K., Moros, M., Porsche, C., Neumann, T., Adolphi, F., Andersen, T. J., Siegel, H., Gerth, M., Leipe, T., Jansen, E., and Sinninghe Damste, J. S., 2012, Impact of climate change on the Baltic Sea ecosystem over the past 1,000 years: Nature Clim. Change, v. advance online publication.

Laruelle, G. G., Roubeix, V., Sferratore, A., Brodherr, B., Ciuffa, D., Conley, D. J., Durr, H. H., Garnier, J., Lancelot, C., Phuong, Q. L. T., Meunier, J. D., Meybeck, M., Michalopoulos, P., Moriceau, B., Longphuirt, S. N., Loucaides, S., Papush, L., Presti, M., Ragueneau, O., Regnier, P., Saccone, L., Slomp, C. P., Spiteri, C., and Van Cappellen, P., 2009, Anthropogenic perturbations of the silicon cycle at the global scale: Key role of the land-ocean transition: Global Biogeochemical Cycles, v. 23.

Mackenzie, B. R., and Schiedek, D., 2007, Daily ocean monitoring since the 1860s shows record warming of northern European seas: Global Change Biology, v. 13, no. 7, p. 1335-1347.

McKay, N. P., Kaufman, D. S., and Michelutti, N., 2008, Biogenic silica concentration as a high-resolution, quantitative temperature proxy at Hallet Lake, south-central Alaska: Geophys. Res. Lett., v. 35, no. 5, p. L05709.

Morley, D. W., Leng, M. J., Mackay, A. W., Sloane, H. J., Rioual, P., and Battarbee, R. W., 2004, Cleaning of lake sediment samples for diatom oxygen isotope analysis: Journal of Paleolimnology, v. 31, no. 3, p. 391-401.

Nelson, D. M., Treguer, P., Brzezinski, M. A., Leynaert, A., and Queguiner, B., 1995, Production and dissolution of biogenic silica in the ocean: Revised global estimates, comparison with regional data and relationship to biogenic sedimentation: Global Biogeochemical Cycle, v. 9, no. 3, p. 359-372.

Nijampurkar, V. N., Rao, D. K., Oldfield, F., and Renberg, I., 1998, The half-life of 32Si: a new estimate based on varved lake sediments: Earth and Planetary Science Letters, v. 163, no. 1-4, p. 191-196.

Officer, C. B., and Ryther, J. H., 1980, The possible importance of silicon in marine eutrophication: Marine Ecology Progress Series, v. 3, p. 83-91.

Olli, K., Clarke, A., Danielsson, Å., Aigars, J., Conley, D. J., and Tamminen, T., 2008, Diatom stratigraphy and long-term dissolved silica concentrations in the Baltic Sea: Journal of Marine Systems, v. 73, no. 3-4, p. 284-299.

Omstedt, A., Gustafsson, E., and Wesslander, K., 2009, Modelling the uptake and release of carbon dioxide in the Baltic Sea surface water: Continental Shelf Research, v. 29, no. 7, p. 870-885.

Opfergelt, S., Eiriksdottir, E. S., Burton, K. W., Einarsson, A., Siebert, C., Gislason, S. R., and Halliday, A. N., 2011, Quantifying the impact of freshwater diatom productivity on silicon isotopes and silicon fluxes: Lake Myvatn, Iceland: Earth and Planetary Science Letters, v. 305, no. 1-2, p. 73-82.

Papush, L., and Danielsson, Å., 2006, Silicon in the marine environment: Dissolved silica trends in the Baltic Sea: Estuarine, Coastal and Shelf Science, v. 67, no. 1-2, p. 53-66.

Papush, L., Danielsson, Å., and Rahm, L., 2009, Dissolved silica budget for the Baltic Sea: Journal of Sea Research, v. 62, no. 1, p. 31-41.

Radziejewska, T., and Schernewski, G., 2008, The Szczecin (Oder-) Lagoon, Ecology of Baltic Coastal Waters, in Schiewer, U., ed., Volume 197, Springer Berlin Heidelberg, p. 115-129.

Ragueneau, O., Conley, D. J., Leynaert, A., Ni Longphuirt, S., and Slomp, C. P., 2006, Role of diatoms in silicon cycling and coastal marine food webs, Washington DC, SCOPE, The silicon cycle: Human perturbations and impacts on aquatic systems.

Ragueneau, O., Savoye, N., Del Amo, Y., Cotten, J., Tardiveau, B., and Leynaert, A., 2005, A new method for the measurement of biogenic silica in suspended matter of coastal waters: using Si:Al ratios to correct for the mineral interference: Continental Shelf Research, v. 25, no. 5–6, p. 697-710.

Ragueneau, O., Treuer, P., Leynaert, A., Anderson, R. F., Brzezinski, M. A., DeMaster, D. J., Dugdale, R. C., Dymond, J., Fischer, G., Francis, R., Heinze, C.,

Isotope-based reconstruction of the biogeochemical Si cycle

15

Maier-Reimer, E., Martin-Jezequel, V., Nelson, D. M., and Queuiner, B., 2000, A review of the Si cycle in the modern ocean: recent progress and missing gaps in the application of biogenic opal as a paleoproductivity proxy: Global and Planetary Change, v. 26, no. 4, p. 317-365.

Reynolds, B. C., Aggarwal, J., André, L., Baxter, D., Beucher, C., Brzezinski, M. A., Engström, E., Georg, R. B., Land, M., Leng, M. J., Opfergelt, S., Rodushkin, I., Sloane, H. J., van den Boorn, S. H. J. M., Z., V. P., and D., C., 2007, An inter-laboratory comparison of Si isotope reference materials: Journal of Analytical Atomic Spectrometry, v. 22, p. 561-568.

Reynolds, B. C., Frank, M., and Halliday, A. N., 2006, Silicon isotope fractionation during nutrient utilization in the North Pacific: Earth and Planetary Science Letters, v. 244, no. 1-2, p. 431-443.

Ryther, J. H., 1969, Photosynthesis and Fish Production in the Sea: Science, v. 166, no. 3901, p. 72-76.

Sandén, P., Rahm, L., and Wulff, F., 1991, Non-parametric trend test of baltic sea data: Environmetrics, v. 2, no. 3, p. 263-278.

Schelske, C. L., Stoermer, E. F., Conley, D. J., Robbins, J. A., and Glover, R. M., 1983, Early Eutrophication in the Lower Great Lakes: Science, v. 222, no. 4621, p. 320-322.

Smetacek, V. S., 1985, Role of sinking in diatom life-history cycles: ecological, evolutionary and geological significance: Marine Biology, v. 84, no. 3, p. 239-251.

Smol, J. P., Wolfe, A. P., Birks, H. J. B., Douglas, M. S. V., Jones, V. J., Korhola, A., Pienitz, R., Rühland, K., Sorvari, S., Antoniades, D., Brooks, S. J., Fallu, M. A., Hughes, M., Keatley, B. E., Laing, T. E., Michelutti, N., Nazarova, L., Nyman, M., Paterson, A. M., Perren, B., Quinlan, R., Rautio, M., Saulnier-Talbot, E., Siitonen, S., Solovieva, N., and Weckström, J., 2005, Climate-driven regime shifts in the biological communities of arctic lakes: Proceedings of the National Academy of Sciences of the United States of America, v. 102, no. 12, p. 4397-4402.

Street-Perrott, F. A., Barker, P. A., Leng, M. J., Sloane, H. J., Wooller, M. J., Ficken, K. J., and Swain, D. L., 2008, Towards an understanding of late Quaternary variations in the continental biogeochemical cycle of silicon: multi-isotope and sediment-flux data for a e utundu, t enya, ast frica, since 38  a BP: Journal of Quaternary Science, v. 23, no. 4, p. 375-387.

Swann, G. E. A., and Leng, M. J., 2009, A review of diatom δ18O in palaeoceanography: Quaternary Science Reviews, v. 28, no. 5–6, p. 384-398.

Treguer, P., Nelson, D. M., Van Bennekom, A. J., DeMaster, D. J., Leynaert, A., and Queguiner, B.,

1995, The Silica Balance in the World Ocean: A Reestimate: Science, v. 268, no. 5209, p. 375-379.

van den Boorn, S. H. J. M., van Bergen, M. J., Vroon, P. Z., de Vries, S. T., and Nijman, W., 2010, Silicon isotope and trace element constraints on the origin of ~3.5 Ga cherts: Implications for Early Archaean marine environments: Geochimica et Cosmochimica Acta, v. 74, no. 3, p. 1077-1103.

Vaquer-Sunyer, R., and Duarte, C. M., 2009, Thresholds of hypoxia for marine biodiversity: Proceedings of the National Academy of Sciences, v. 105, p. 15452–15457.

Varela, D. E., Pride, C. J., and Brzezinski, M. A., 2004, Biological fractionation of silicon isotopes in Southern Ocean surface waters: Global Biogeochem. Cycles, v. 18, no. 1, p. GB1047.

Voss, M., Larsen, B., Leivuori, M., and Vallius, H., 2000, Stable isotope signals of eutrophication in Baltic Sea sediments: Journal of Marine Systems, v. 25, no. 3-4, p. 287-298.

Wasmund, N., Nausch, G., and Matthäus, W., 1998, Phytoplankton spring blooms in the southern Baltic Sea ---- Spatio-temporal development and long-term trends: Journal of Plankton Research, v. 20, no. 6, p. 1099-1117.

Wedepohl, H. K., 1995, The composition of the continental crust: Geochimica et Cosmochimica Acta, v. 59, no. 7, p. 1217-1232.

Ziegler, K., Chadwick, O. A., Brzezinski, M. A., and Kelly, . F., 2005, atural variations of δ30Si ratios during progressive basalt weathering, Hawaiian Islands: Geochimica et Cosmochimica Acta, v. 69, no. 19, p. 4597-4610.

Xiaole Sun

16

Paper I

ISSN 0267-9477

Journal of Analytical Atomic Spectrometry

PAPERSun et al.Stable silicon isotope analysis on nanomole quantities using MC-ICP-MS with a hexapole gas-collision cell

COMMUNICATIONGünther et al.Development of direct atmospheric sampling for laser ablation-inductively coupled plasma-mass spectrometry

ASU REVIEWAtomic spectrometry update. Environmental analysis

www.rsc.org/jaas Volume 25 | Number 2 | February 2010 | Pages 93–216

PAPER www.rsc.org/jaas | Journal of Analytical Atomic Spectrometry

Stable silicon isotope analysis on nanomole quantities using MC-ICP-MSwith a hexapole gas-collision cell

Xiaole Sun,*a Per Andersson,b Magnus Land,c Christoph Humborgd and Carl-Magnus M€orthe

Received 8th June 2009, Accepted 13th November 2009

First published as an Advance Article on the web 25th November 2009

DOI: 10.1039/b911113a

We demonstrate in this study that a single focusing multiple collector inductively coupled plasma mass

spectrometer (MC-ICP-MS) equipped with a hexapole gas-collision cell (GV-instrument� Isoprobe) can

precisely determine the d29Si (2S.D., 0.2&) using a total Si consumption of less than 14 nmole (390 ng

Si). Testing and evaluation of background, rinse time, and major matrix effects have been performed in

a systematic way to establish a procedure to measure d29Si in small quantities. Chemical purification

prior to analysis is required to remove potential interferences. For data collected during a four-year

period, the average d29Si value of IRMM-018 relative to NBS-28 was found to be �0.95& (n ¼ 23,

2S.D. 0.16&) with a 95% confidence interval (�0.95 � 0.028&). The mean d29Si value of the Big-Batch

standard was found to be �5.50& (n ¼ 6, 2S.D. 0.26&). Although determination of the d30Si

measurements is not possible, with our current instrument we demonstrate that this system provides

a fast and long-term reliable method for the analysis of d29Si in purified samples with low Si

concentration (18 mM Si).

Introduction

Silicon is one of the most abundant elements in the continental

crust, 28.8 wt%,1 and constitutes the backbone of silicate

minerals, which are continuously altered by weathering

processes. Mass-balance studies of the oceanic Si cycle2 have

demonstrated that the gross input is primarily derived from the

continents (84%) as dissolved silicate (DSi). The largest part by

far is supplied by river transport, but eolian transport (7.5%) is

also important. Submarine weathering of basalt is an additional

source primarily at sea floor hydrothermal areas. As an essential

element, silicon fertilizes the seas by stimulating the production

of diatoms, which are a key phytoplankton group responsible for

fuelling of the pelagic and benthic food webs and dominating the

biological carbon sequestration in the oceans.3–5 Weathering of

continental silicates is also suggested to be a long-term regulator

of atmospheric CO2 by sequestration of CO2 carbonates during

weathering and consecutive storage in the ocean and the deep sea

sediments.6 The silicon cycle and its variation through time are

crucial for the understanding of marine productivity and carbon

sequestration and stable silicon isotopes constitute a tool that

can be used to study these processes.

Variations in the stable silicon isotope composition (28Si, 29Si,30Si with abundances of 92.22%, 4.69% and 3.09%, respectively)7

aDepartment of Geology and Geochemistry, Stockholm University, andLaboratory for Isotope Geology, Swedish Museum of Natural History,Stockholm, Sweden. E-mail: xiaole.sun@geo.su.sebLaboratory for Isotope Geology, Swedish Museum of Natural History,SE-10405 Stockholm, Sweden. E-mail: per.andersson@nrm.secWSP Environmental, SE-12188 Stockholm, Sweden. E-mail: magnus.land@WSPGroup.sedDepartment of Applied Environmental Science, Stockholm University, SE-10691 Stockholm, Sweden. E-mail: christoph.humborg@itm.su.seeDepartment of Geology and Geochemistry, Stockholm University, SE-10691 Stockholm, Sweden. E-mail: magnus.morth@geo.su.se

156 | J. Anal. At. Spectrom., 2010, 25, 156–162

have been documented in geological materials since the early

1950s. It has been shown that minerals formed at different

temperatures were isotopically lighter (i.e. enriched in 28Si) as

a function of decreasing crystallization temperature.8 Based on

theoretical calculations, Grant9 predicted a positive correlation

between 30Si enrichment and increasing silicon content in rocks.

Another study showed that the enrichment of 30Si in coexisting

minerals in igneous rocks increases in the mineral order of bio-

tite, quartz, feldspar, and that granitic rocks were enriched the

most in 30Si compared to meteorites and mafic rocks.10 However,

like most of the other elements above 20 amu, only small isotopic

variations of Si are found in nature (<10&). Therefore, the

analytical methods for determining Si isotope ratios need to have

sufficient accuracy and reproducibility to determine silicon

isotope ratio variations.

Currently, studies on variations in stable Si isotope ratios are

focused on low-temperature processes such as biogenic opal

formation,11–14 clay formation15 and chemical weathering

processes in groundwater aquifers.16,17 Opal formation in the

ocean is dominated by diatom assimilation of dissolved silicate

from the surrounding aqueous phase to build up their silicified

cell wall, a process which occurs in micromole Si concentration

environments. The opal formed has an isotope value of about

1& lower than the source isotope value of DSi,18 i.e., the heavier

Si isotopes are enriched in surface waters as diatoms preferen-

tially sequester the 28Si to form opal. When diatom growth

become Si-limited (DSi < 5 mmol/l) as is often the case in

coastal19,20 and open ocean marine systems,5 the isotope frac-

tionation becomes even higher. Thus, heavier Si isotopes in

surface waters might indicate a shift in the ecosystem, which has

occurred in many coastal waters from siliceous (diatoms) to non-

siliceous phytoplankton species.21–23 Therefore, methods

addressing stable isotope fractionation during opal formation

should be operational on micromole concentrations.

This journal is ª The Royal Society of Chemistry 2010

Until recently, Si isotope data in the literature have been

limited, largely due to analytical difficulties. The traditional

method of analysing Si isotope ratios is by fluorination of silicon

and introducing the resulting SiF4 gas into a gas-source isotope

ratio mass spectrometer (IRMS).10,11,24 A new preparation

method using the acid decomposition of Cs2SiF6 followed by the

analysis of IRMS25 improved the precision compared with the

previous IRMS method. However, this technique is fairly

complicated and requires a potentially hazardous preparation

process.

The development of multiple-collector inductively coupled

plasma mass spectrometry (MC-ICP-MS) has opened up new

possibilities for the determination of stable silicon isotopes with

a precision comparable with that of IRMS.26–30 The major

advantages with ICP source mass spectrometry is that the sample

preparation is simplified and less hazardous, and it requires

shorter analytical time and smaller sample size compared to gas-

source IRMS.

Stable Si isotope measurements using MC-ICP-MS has mainly

utilised double-focusing mass spectrometers with high mass

resolution, i.e. m/Dm > 1000, to overcome the problem with

polyatomic interferences derived from the ICP source. Instru-

mentation used for high-resolution analysis includes the

NuPlasma HR, the NuPlasma 1700 from NU instruments� and

the Neptune from ThermoFisher�.26,30,31

In this study, we have established a technique to analyse stable

silicon isotopes using a single-focusing low-resolution (m/Dm �450) MC-ICP-MS, the IsoProbe from GV instruments�. This

instrument is equipped with a hexapole collision cell, which

works as an energy filter. The technique permits higher sensitivity

and less consumption of Si compared to the double-focusing

high-resolution MC-ICP-MS.26,31,32 Below, we present detailed

investigations of matrix effects, polyatomic interferences as well

as instrument memory and background. Moreover, we docu-

ment the long-term performance over four years using interna-

tional standard materials.

Experimental

Si isotope standards

It is important that the standard materials used are well docu-

mented so that results from different sample measurements can

be normalized and properly compared. Therefore, several stan-

dard materials with different isotope compositions have been

introduced to compile data on silicon isotopes. Previously, the

standard to which samples were compared with was the Caltech

Rose Quartz33 followed by the RM8564-NBS28-Silica Sand

(National Institute of Standards and Technology); the latter is

labelled in the following as NBS-28. NBS-28 is nowadays widely

used as an universally accepted zero reference material for silicon

isotopes, and has a similar Si isotope composition to that of

Caltech Rose Quartz.34 Other standard materials such as the Big-

Batch and diatomite with an isotopic composition different than

NBS-28 have been produced and used as secondary reference

materials.16,17,30–32 An additional reference material, IRMM-018,

has shown a larger than expected variation in isotopic compo-

sitions, which might be due to heterogeneity in this mate-

rial.31,32,35

This journal is ª The Royal Society of Chemistry 2010

Sample preparation

Solid materials obtained as part of the inter-laboratory

comparison of Si isotope reference materials published by Rey-

nolds et al.30 have been prepared and used throughout this study.

This included the standard material NBS-28 and IRMM-018,

and the isotopically highly enriched (both in 29Si and 30Si)

material called Big-Batch (prepared at UCSB, University of

California Santa Barbara, and initially originated from

a commercial Na-metasilicate).

The solid samples of standards were prepared by fusion with

LiBO2 and digested in HCl. About 42.5 mg SiO2 was mixed with

150 mg LiBO2 in a graphite crucible and fused at 1000 �C for 30

minutes. After cooling, the formed glass pearl was digested in

20 ml 2.5 M HCl on a shaking table, resulting in a solution

containing 35.4 mM Si. This solution was immediately diluted 50

times with 0.075 M HCl, yielding a stock solution with

a concentration of 0.7 mM Si in 0.12 M HCl, which was stored in

acid-cleaned high-density polyethylene bottles. It is important to

keep the Si concentration in samples below about 2 mM Si to

avoid polymerization during storage.36 After four-year storage,

the Si concentration in the stock solution was still 0.7 mM Si,

indicating that polymerization did not influence the concentra-

tion during this period. Water used for dilution was obtained

from a Millipore� Milli-Q water purification system. The stock

solution was diluted to 18 mM Si in 0.12 M HCl for Si isotope

measurements.

The Big-Batch standard material contains about 1.5 mg/g

Mo30,37 as a result from the preparation process using a molyb-

date co-precipitation technique, which may cause potential

matrix effects. Screening of our stock solution by ICP-OES

showed that there was about 4 mM Mo. Furthermore about

2.6 mM Li and 2.6 mM B was present in the Si solution.

Instrumentation

All the isotope measurements were carried out in the Laboratory

of Isotope Geology at the Swedish Museum of Natural History

using an IsoProbe MC-ICP-MS. The operating conditions of the

instrument are summarized in Table 1. The ions generated in the

plasma are introduced into the hexapole collision cell, where they

collide/react with a collision gas. This reduces the energy spread

from about 15 V to less than 1 V, and makes the ions suitable,

after acceleration, for direct entry into the mass spectrometer.

The collision cell also reduces the presence of ions interfering

with the measured Si isotopes by collisions between the ions and

the collision gas as described below. The low energy spread of the

ions allows the refocused beam width to be less than the width of

the collector slits, leading to the flat-topped peaks required for

accurate isotope ratio measurements. A large variety of gases and

gas mixtures can be used in the collision cell and for isotope ratio

measurement noble gases such as Ar are frequently used. For

optimal transmission of light isotopes, Helium is selected as

a collision gas.38

A typical torch material is quartz, but for the Si isotope

measurements we used a torch in which the outer tube of the

torch is made of syalon, an aluminium–silicon–oxynitride alloy,

which has a high wear resistance, good durability and low

contamination potential. The inner part of the torch is made of

J. Anal. At. Spectrom., 2010, 25, 156–162 | 157

Table 1 Instrument settings

Multi-collector settings

Cup Mass

Axial 28SiHigh 3 29SiHigh 6 30Si

a Denoted values of these parameters are kept constant during one entiremeasurement, but it may vary between different measurements in thisgiven range.

ICP parameters

Values Units

High tensiona 6000�6020 VCooling Ar flowa 14 l/minIntermediate Ar flowa 0.95�1.05 l/minCollision He flowa 8�9.9 l/minNeb 1 gas flowa 50�60 ml/minHexapole RF amplitudea 30�50 %Outer tube of torch Syalon —Inner tube of torch Alumina —Cone Nickel —

Cetac� Aridus settings

Spray Chamber 95 �CDesolvator 160 �CSweep gas flowa 3.60�3.80 ml/minN2 gas flow 0 ml/min

Fig. 1 Mass scans with the signals in volts for the 28Si, 29Si and 30Si peaks

and possible isobaric interferences from polyatomic ions using 18 mM Si

solutions. The isobaric interferences are given and marked by the shade

area as well as their contribution to the signal in volts. The dashed vertical

lines on low and high mass side show the positions for test measurements

of Si isotope ratios, and the area between them is the measurement

position.

pure aluminium, which not only has extremely high-temperature

stability and wear resistance, but also helps to provide protective

atmospheres at high temperature to eliminate contamination or

impurity.

The Si concentrations both in the standard and unknown

solution were kept at about 18 mM. The solution was introduced

into the plasma using a Cetac� Aridus desolvating nebulizer with

a sample uptake rate of approximate 50–60 mL/min. A 28Si signal

of 8�10 V was usually obtained, but the 28Si signal could occa-

sionally decrease to �7 V during measurements. The formation

of gaseous SiCl4 can be neglected in water-rich solution.

Interferences

The Si isotope ratio measurement using MC-ICP-MS is usually

affected by mass bias, isobaric interferences, and matrix effects.

Potential isobaric interferences include 14N2 and 12C16O on the28Si peak, 14N15N and 12C16O1H on the 29Si peak and 14N16O on the30Si peak. The mass resolution on our Isoprobe is too low (m/Dm

� 450) to fully resolve the polyatomic interferences, but they

could be reduced by using the hexapole collision cell. The three Si

isotopes are influenced by the interfering ions, but compared to

the atomic ions the polyatomic ions are slightly heavier and they

mainly affect the peaks on the high-mass side (Fig. 1). By

adjusting the measurement position to the low-mass side, a well

defined flat-top peak can found for 28Si and 29Si (Fig. 1).

However, due to a relatively large isobaric interference from14N16O on the smallest peak, 30Si, no flat area can be defined and

we are thus not able to measure this isotope (Fig. 1).

158 | J. Anal. At. Spectrom., 2010, 25, 156–162

Keeping the measurement position stable is important and this

was tested by determination of the 29Si/28Si ratio at two positions,

one on the low mass side and the other on the high mass side, of

the flat area for 28Si and 29Si. The 29Si/28Si ratio was measured and

it was found that the ratios determined on both sides of the

measurement position (marked in Fig. 1) differ from the ratio

determined at the measurement position. This demonstrates that

it is necessary to reduce the isobars to a minimum and carefully

adjust the position of the peak before data collection.

By using the hexapole as a reaction cell we can reduce the

isobaric interferences to obtain sufficient peak flatness to

measure the 29Si/28Si ratio. This is similar to what Moureau

et al.39 have demonstrated when injecting N2O in a reaction cell

to remove the isobaric interference 92Zr for 92Mo efficiently and

thus make it possible to accurately measure Mo isotope ratios

also by using an Isoprobe.

Measurement procedure

An individual measurement of a standard/sample consisted of 5

blocks and 12 cycles per block, with the integration time of 10

seconds for each cycle. Before each measurement the inlet system

was rinsed for 180 seconds with 0.12 M HCl. After the rinsing,

there was an uptake period of 180 seconds before data acquisi-

tion commenced. In total it took about 13 minutes to measure

one standard/sample. With an uptake flow of 50�60 mL/min, the

This journal is ª The Royal Society of Chemistry 2010

Fig. 3 (a) examples of upward and (b) downward drift in two

measurements of 29Si/28Si for IRMM-018 vs. NBS-28 and Big-Batch vs.

NBS-28. (a) Nine points with the NBS-28 measured 5 times and (b),

IRMM-018/Big-Batch as unknown solutions was measured 4 times. Both

graphs show the linear trend.

amount of Si consumed during the measurement was approxi-

mately 14 nmol (390 ng).

In order to obtain the relative Si isotope variation, we used the

standard-sample-standard bracketing technique. This method is

frequently applied for determination of isotope ratio using MC-

ICP-MS but requires identical matrix conditions in the sample

and standard solutions.39 Each sample was measured using the

standard-sample-standard bracketing technique including nine

brackets, i.e., nine individual measurements. Before and after

each measurement, a blank solution was introduced for 210

seconds and measured with an integration time of 60 seconds. It

took approximately 160 minutes to acquire each reported isotope

ratio, which was the average value calculated from nine brackets

with two times the standard deviation (2S.D.) or at 95% confi-

dence interval.

Instrument drift

When using the Isoprobe for isotope ratio determination, a drift

in the ratio during the measurement period is commonly

observed. This drift can be both in the upward or downward

direction, i.e., increasing or decreasing ratios with time. Typically

a systematic variation towards a single direction was observed

(Fig. 2). Fig. 2a shows the machine upward drift in an individual

measurement, which was the case for most of the analyses, and

Fig. 2b illustrates the downward drift. The drift in the 29Si/28Si

ratio appeared not only in individual measurements, but was also

observed during the entire analysis using the standard-sample-

standard bracketing technique (Fig. 3). The systematic drift in

the ion beam signal is independent of the integration time. This

type of drift in the isotope ratio is frequently observed also for

other heavier elements such as S and Pb, but can mostly be

Fig. 2 (a) example of upward and (b) downward drift in the measured29Si/28Si ratios from two individual measurements including 60 cycles. The

dotted lines show the 95% confidence interval and the solid line shows the

linear trend.

This journal is ª The Royal Society of Chemistry 2010

effectively corrected using the standard-sample-standard brack-

eting technique.

Mass discrimination or mass bias is observed in all mass

spectrometric techniques and can generally be described as the

time dependent deviation between observed and ‘‘true value’’. In

ICP-MS the mass bias is the result of a number of processes in the

plasma and in the transmission through the ion optical system.40

Generally the mass bias decreases with increasing mass and is

most pronounced for the light isotopes.41 The applied standard-

sample-standard bracketing technique can be used to compen-

sate for mass bias, matrix interferences and backgrounds.28

However, this requires that the signals obtained from both the

standard and the samples are equally high and that the blank

solutions have the same matrix as the analytical solution.

The reported delta values in per mil units were calculated

according to Equation 1,

d 29Si ¼

�29Si28Si

�sample�

29Si28Si

�standard

� 1

0BBB@

1CCCA,1000 (Equation 1)

Matrix effects

Matrix effects are caused by the presence of additional ions, such

as Na+, Mg2+, Al3+, Ca2+, K+, in the Si solution. These ions can

affect the stability of the ion beam signal and lead to inaccurate

results. Presence of Na+ in the analytical solution is known to

suppress the ion beam of the analyte, i.e. Si. In this study, we

made a detailed study of the effect of increasing concentration of

Mg2+ in the analytical solution on the d29Si value. Aluminium,

J. Anal. At. Spectrom., 2010, 25, 156–162 | 159

the most abundant metal in the earth crust, might also be

a potential problem causing matrix effects. To examine these

possible interferences on the d29Si value, a series of measurements

with varying concentrations of Mg and Al were performed by

adding these elements to the IRMM-018 standard solution. The

experiments were performed in 2005 and repeated using different

machine settings during 2008/09.

The presence of Mg shown as an increased Mg/Si ratio (Fig. 4)

seems to increase the spread in the d29Si value even though the

average value of each measurement was similar. At high Mg/Si

ratio, >1, the d29Si seems to decrease towards lower values and

the precision also seems to decrease. This might be due to

changing conditions in the plasma when the ion strength is

increasing.

In contrast to Mg, the presence of Al causes a substantial effect

on the d29Si value and larger errors (Fig. 5). At Al/Si weight ratios

as low as 0.1, the d29Si is lowered by >0.7&. In addition to these

detailed investigations of the effects of Mg and Al, we have

observed that Si intensity can be greatly reduced in solutions with

high concentrations of Na+ and Ca2+, which might also cause

poor measurement accuracy. However, this effect on the d29Si has

not been quantitatively determined.

The formation of 26MgH2+, 27AlH+ and 27AlH2 are potential

problems in that they cause isobaric interferences. However, the

formation of metal hydrides is likely to be small and can most

Fig. 4 Testing the influence of varying ions in the matrix on the d29Si by

using increasing Mg/Si weight ratios. Data were collected both in 2005

and 2008/09, respectively. The used Si concentration was 71 mM in 2005

and 18 mM in 2008/09.

Fig. 5 Testing the influence of varying ions in the matrix on the d29Si by

using increasing Al/Si weight ratios. Data were collected both in 2005 and

2008/09, respectively. The used Si concentration was 71 mM in 2005, and

18 mM in 2008/09.

160 | J. Anal. At. Spectrom., 2010, 25, 156–162

likely be omitted. In addition, van den Boorn et al.42 reported

that anionic species, such as SO42�, could also cause a substantial

matrix effect on the Si isotope measurement.

The Li and B (67.5 mM Li and 65 mM B) in the 18 mM Si

solution of the fused standards materials were not removed.

However, we cannot exclude the possibility for a matrix effect

from those elements on the Si isotope ratio. This was avoided by

addition of Li and B to all solutions to keep the Li and B

concentrations on the same level and thus affect the plasma in

a similar way. We adjusted the amount of Li and B in the samples

and measured the elemental concentrations by ICP-OES before

Si isotope analysis to assure that the matrix of Li and B were

properly adjusted.

Clearly, the presence of alkali metal and alkaline earth metals

as well as Al ions in the analytical solution could severely affect

the precision and accuracy of the Si isotope ratio measurements.

Also other species including both cations and anions could be

a potential problem for high precision Si isotope measurements.

Therefore it is crucial to remove interferences and isobars by

chemical purification prior to the measurements. Alternatively,

matrix effects from elements like Li and B, which are hard to

quantitatively remove by column chemistry, could be eliminated

by determining and adjusting the concentrations in samples.

Background and rinse time test

Since the variations in d29Si values between samples are likely to

be small, even a tiny amount of Si from a previous standard/

sample remaining in the nebuliser and the analytical system

could potentially affect the measured ratio. The system was

rinsed using 0.12 M HCl between each individual measurement

but the effectiveness of this cleaning needed to be tested thor-

oughly. The rinse time was investigated by measuring IRMM-

018 relative to NBS-28 and the results show that a rinse time of at

least 150 seconds between samples is necessary to remove the

previous samples. A rinse time of 180 seconds was applied to

ensure that the entire system was clean. We efficiently eliminated

the instrument memory effects and could reach an error of d29Si

less than 0.2& (2S.D.).

Instrumental background is mainly derived from the analysed

solutions, but the build-up of impurities on the cones as well as

on the transfer optics over time also contributes to the back-

ground. We investigated the background contribution by using

a 180 s rinse time and conducting blank determinations at the

beginning and end of each measurement. The observed back-

ground contributed up to approximately 2.5% on the 28Si and 29Si

intensities. The background was found to be similar before and

after the analysis. Part of this background comes from build-up

of impurities in the inlet area of the instrument. By careful

cleaning it could be reduced to about 0.6%. However, during

a session of measurements the background was generally found

to be stable and could be corrected using the standard-sample-

standard bracketing technique.

Results and discussion

Long-term reproducibility

To determine our ability to precisely measure and reproduce the

d29Si value using a MC-ICP-MS equipped with a collision cell, we

This journal is ª The Royal Society of Chemistry 2010

Fig. 6 The d29Si measured in IRMM-018 relative to NBS-28. Section 1

was performed in 2005 and Section 2 was measured in 2008/09. The same

solutions and the same machine were used in both sections although the

settings of the machine varied between these two periods. Each point

represents the average value of nine brackets in one measurement.

Fig. 7 The d29Si measured in the Big-Batch relative to NBS-28. The two

sections are the results measured in 2005 and 2008/09. Each point

represents the average value of nine brackets in one measurement.

have analysed the IRMM-018 and the Big Batch standard. The

data have been obtained in periods during 2005 and from

October 2008 to August 2009 (Fig. 6 and 7). Data from

measurements of IRMM-018 relative to NBS-28 are reported in

Fig. 6, comprising the first measurement period in 2005 with the

integration time of 5 and 7 seconds and the second in 2008/2009

with the integration time of 10 seconds. The data for IRMM-018

comprising the entire period has been examined statistically

using a two sample t-test assuming equal variance and normally

distributed errors to test whether the measured results in both

periods were similar. No statistically distinct difference between

the two periods could be found (p # 0.0014, n ¼ 23).

Table 2 Summary of the results in this study and comparison with otherpublished data

Standard d29Si (&) 2S.D. n

IRMM-18 �0.95 0.16 21IRMM-18a �0.85 0.14 740Big-Batch �5.51 0.25 6Big-Batcha �5.35 0.3 198Big-Batchb �5.39 0.3 28Big-Batchc �5.39 0.17 15

a Inter-laboratory calibrated Si isotope values.30 b Measurements fromvan den Boorn et al.37 c Measurements from Chmeleff et al.31

This journal is ª The Royal Society of Chemistry 2010

The results for both the IRMM-018 and the Big-Batch are

summarised in Table 2 along with literature data. The average

d29Si value for IRMM-018 during the entire four-year period was

found to be �0.95& (2S.D. 0.16&) with a 95% confidence

interval of �0.028&. The results are in agreement with the value

of�0.85& (2S.D., 0.14&) obtained in the inter-calibration study

by Reynolds et al.30 The precision (2S.D.) for a reported d29Si

value is typically between 0.1 and 0.2&, although smaller and

larger errors (0.061 to 0.27&) occurred occasionally, most

probably caused by machine random errors during measure-

ments.

Similar to IRMM-018, we have also analysed the Big-Batch

standard and our mean d29Si value was found to be �5.50&

(2S.D., 0.26&, Fig. 7). This can be compared to previously

published values of �5.39& (2S.D., 0.17&),31 �5.39& (2S.D.,

0.3&)37 and �5.35& (2S.D., 0.3&) reported by Reynolds et al.30

in the inter-calibration study (Table 2).

The amount of Si consumed during one measurement was

approximate 400 ng per sample. This is comparable with other

methods using double-focusing high-resolution MC-ICP-MS

consuming between 200 and 500 ng Si,26,30,31 but lower than the 3

mg Si typically used by a double-focusing low-resolution MC-

ICP-MS.28 The sensitivity for 28Si was found to be �18 V/36 mM

Si (1 mg/L), which is >3 times higher than 5 � 0.5 V/36 mM

observed by Engstr€om et al.27 and slightly higher than �13 V/

36 mM reported by Georg et al.26 using a high-resolution MC-

ICP-MS. Clearly, our method yields high signals using small

amounts of Si which can produce d29Si values with a precision

similar to other methods.

Conclusion

This study demonstrates that by using a single focusing low-

resolution MC-ICP-MS equipped with a hexapole gas-collision

cell, we can precisely and accurately measure the d29Si (2S.D.,

0.2&) with a total Si consumption of less than 14 nmol (390 ng

Si). Analysis of the same standard material performed during

2005 and 2008/09 gave a similar precision and accuracy. Given

that the instrument’s hardware and settings have been partly

changed during the last four years, the reproducibility in d29Si is

found to be stable. For precise measurement of d29Si values it is

necessary to use a Si solution with low concentrations of inter-

ference ions to avoid matrix effects. Thus, chemical purification

to remove these interferences of the Si solution prior to analysis is

required. Alternatively, matrix effects from elements like Li and

B, which are not easy to remove by using ion-exchange, can be

eliminated by determining and adjusting the concentrations in

samples and standards to the same level.

Acknowledgements

We would like to thank Hans Sch€oberg and Torsten Persson in

the Laboratory for Isotope Geology at the Swedish Museum of

Natural History for assisting with sample preparation and

providing technical support. We are also grateful for the funding

from the Swedish Research Council (VR-2007-4763). We

appreciate insightful comments and constructive suggestions

from two anonymous reviewers, who greatly improved this

manuscript.

J. Anal. At. Spectrom., 2010, 25, 156–162 | 161

References

1 K. Hans Wedepohl, Geochim. Cosmochim. Acta, 1995, 59, 1217–1232.2 P. Treguer, D. M. Nelson, A. J. Van Bennekom, D. J. DeMaster,

A. Leynaert and B. Queguiner, Science, 1995, 268, 375–379.3 P. G. Falkowski, R. T. Barber and V. Smetacek, Science, 1998, 281, 200.4 O. Ragueneau, P. Treuer, A. Leynaert, R. F. Anderson,

M. A. Brzezinski, D. J. DeMaster, R. C. Dugdale, J. Dymond,G. Fischer, R. Francis, C. Heinze, E. Maier-Reimer, V. Martin-Jezequel, D. M. Nelson and B. Queuiner, Global Planet. Change,2000, 26, 317–365.

5 R. C. Dugdale, F. P. Wilkerson and H. J. Minas, Deep-Sea Res., PartI, 1995, 42, 697–719.

6 R. A. Berner, Am. J. Sci., 1991, 291, 339–376.7 J. R. De Laeter, J. K. B€ohlke, P. De Bi�evre, H. Hidaka, H. S. Peiser,

K. J. R. Rosman and P. D. P. Taylor, Pure Appl. Chem., 2003, 75,683–799.

8 J. H. Reynolds and J. Verhoogen, Geochim. Cosmochim. Acta, 1953,3, 224–234.

9 F. S. Grant, Geochim. Cosmochim. Acta, 1954, 5, 225–242.10 D. Tilles, J. Geophys. Res., 1961, 66, 3003–3013.11 C. L. De La Rocha, M. A. Brzezinski, M. J. DeNiro and A. Shemesh,

Nature, 1998, 395, 680–683.12 C. L. De La Rocha, Geology, 2003, 31, 423–426.13 I. Basile-Doelsch, J. D. Meunier and C. Parron, in Nature, Nature

Publishing Group, 2005, vol. 433, pp. 399–402.14 F. Fripiat, R. Corvaisier, J. Navez, M. Elskens, V. Schoemann,

K. Leblanc, L. Andr�e and D. Cardinal, Limnol. Oceanogr.:Methods, 2009, 7, 470–478.

15 F. Robert and M. Chaussidon, in Nature, Nature Publishing Group,2006, vol. 443, pp. 969–972.

16 R. B. Georg, C. Zhu, B. C. Reynolds and A. N. Halliday, Geochim.Cosmochim. Acta, 2009, 73, 2229–2241.

17 R. B. Georg, A. J. West, A. R. Basu and A. N. Halliday, Earth Planet.Sci. Lett., 2009, 283, 67–74.

18 C. L. De La Rocha, M. A. Brzezinski and M. J. DeNiro, Geochim.Cosmochim. Acta, 1997, 61, 5051–5056.

19 C. Humborg, E. Smedberg, M. R. Medina and C.-M. M€orth, J. Mar.Syst., 2008, 73, 223–235.

20 K. Olli, A. Clarke, A. Danielsson, J. Aigars, D. J. Conley andT. Tamminen, J. Mar. Syst., 2008, 73, 284–299.

21 A. Danielsson, L. Papush and L. Rahm, J. Mar. Syst., 2008, 73, 263–283.22 D. J. Conley, C. L. Schelske and E. F. Stoermer, Mar. Ecol.: Prog.

Ser., 1993, 101, 179–192.

162 | J. Anal. At. Spectrom., 2010, 25, 156–162

23 O. Ragueneau, D. J. Conley, A. Leynaert, S. N. Longphuirt andC. P. Slomp, in The Silicon Cycle: Human Perturbations andImpactson Aquatic Systems, eds. V. Ittekkot, D. Unger, C.Humborg and T. A. Nguyen, Island Press, Washington, D.C., 2006,pp. 197–213.

24 C. B. Douthitt, Geochim. Cosmochim. Acta, 1982, 46, 1449–1458.25 M. A. Brzezinski, J. L. Jones, C. P. Beucher, M. S. Demarest and

H. L. Berg, Anal. Chem., 2006, 78, 6109–6114.26 R. B. Georg, B. C. Reynolds, M. Frank and A. N. Halliday, Chem.

Geol., 2006, 235, 95–104.27 E. Engstrom, I. Rodushkin, D. C. Baxter and B. Ohlander, Anal.

Chem., 2006, 78, 250–257.28 D. Cardinal, L. Y. Alleman, J. De Jong, K. Ziegler and L. Andr�e,

J. Anal. At. Spectrom., 2003, 18, 213–218.29 C. L. De La Rocha, Geochem., Geophys., Geosyst., 2002, 3(8), 1–6.30 B. C. Reynolds, J. Aggarwal, L. Andr�e, D. Baxter, C. Beucher,

M. A. Brzezinski, E. Engstr€om, R. B. Georg, M. Land, M. J. Leng,S. Opfergelt, I. Rodushkin, H. J. Sloane, S. H. J. M. van denBoorn, P. Z. Vroon and D. Cardina, J. Anal. At. Spectrom., 2007,22, 561–568.

31 J. Chmeleff, I. Horn, G. Steinhoefel and F. von Blanckenburg, Chem.Geol., 2008, 249, 155–166.

32 B. C. Reynolds, B. R. Georg, F. Oberli, U. Wiechert andA. N. Halliday, J. Anal. At. Spectrom., 2006, 21, 266–269.

33 S. Epstein and H. P. Taylor, Geochim. Cosmochim. Acta, 1970, 2,1085–1096.

34 I. Friedman and J. D. Gleason, Earth Planet. Sci. Lett., 1973, 8, 124.35 T. Ding, D. Wan, R. Bai, Z. Zhang, Y. Shen and R. Meng, Geochim.

Cosmochim. Acta, 2005, 69, 5487–5494.36 R. K. Iler, The chemistry of silica, solubility, polymerization, colloid

and surface properties and biochemistry, John Wiley & Sons, NewYork, 1979.

37 S. H. J. M. van den Boorn, P. Z. Vroon, C. C. van Belle, B. van derWagt, J. Schwieters and M. J. van Bergen, J. Anal. At. Spectrom.,2006, 21, 734–742.

38 E. McCurdy and G. Woods, J. Anal. At. Spectrom., 2004, 19, 607–615.

39 J. Moureau, M. Granet, F. Chartier, G. Favre, H. Isnard andA. Nonell, J. Anal. At. Spectrom., 2008, 23, 1538–1544.

40 J. S. Becker, Inorganic Mass Spectrometry: Principles andApplications, Wiley, 2007.

41 J. S. Becker, J. Anal. At. Spectrom., 2002, 17, 1172–1185.42 S. H. J. M. van den Boorn, P. Z. Vroon and M. J. van Bergen, J. Anal.

At. Spectrom., 2009, 24, 1111–1114.

This journal is ª The Royal Society of Chemistry 2010

Paper II

Biogeosciences, 8, 3491–3499, 2011www.biogeosciences.net/8/3491/2011/doi:10.5194/bg-8-3491-2011© Author(s) 2011. CC Attribution 3.0 License.

Biogeosciences

Climate dependent diatom production is preserved in biogenic Siisotope signatures

X. Sun1,2, P. Andersson2, C. Humborg3,4, B. Gustafsson4, D. J. Conley5, P. Crill 1, and C.-M. Morth1,4

1Department of Geological Sciences, Stockholm University, 10691, Stockholm, Sweden2Laboratory for Isotope Geology, Swedish Museum of Natural History, 10405, Stockholm, Sweden3Department of Applied Environmental Science, Stockholm University, 10691, Stockholm, Sweden4Baltic Nest Institute, Stockholm Resilience Centre, Stockholm University, 10691 Stockholm, Sweden5Department of Earth and Ecosystem Sciences, Lund University, 22362, Lund, Sweden

Received: 28 January 2011 – Published in Biogeosciences Discuss.: 12 April 2011Revised: 16 September 2011 – Accepted: 9 November 2011 – Published: 29 November 2011

Abstract. The aim of this study was to reconstruct diatomproduction in the subarctic northern tip of the Baltic Sea,Bothnian Bay, based on down-core analysis of Si isotopesin biogenic silica (BSi). Dating of the sediment showed thatthe samples covered the period 1820 to 2000. The sedimentcore record can be divided into two periods, an unperturbedperiod from 1820 to 1950 and a second period affected by hu-man activities (from 1950 to 2000). This has been observedelsewhere in the Baltic Sea. The shift in the sediment corerecord after 1950 is likely caused by large scale dammingof rivers. Diatom production was inferred from the Si iso-tope composition which ranged betweenδ30 Si −0.18 ‰ and+0.58 ‰ in BSi, and assuming fractionation patterns due tothe Raleigh distillation, the production was shown to be cor-related with air and water temperature, which in turn werecorrelated with the mixed layer(ML) depth. The sedimentaryrecord showed that the deeper ML depth observed in colderyears resulted in less production of diatoms. Pelagic investi-gations in the 1990’s have clearly shown that diatom produc-tion in the Baltic Sea is controlled by the ML depth. Espe-cially after cold winters and deep water mixing, diatom pro-duction was limited and dissolved silicate (DSi) concentra-tions were not depleted in the water column after the springbloom. Our method corroborates these findings and offers anew method to estimate diatom production over much longerperiods of time in diatom dominated aquatic systems, i.e. alarge part of the world’s ocean and coastal seas.

Correspondence to:X. Sun(xiaole.sun@geo.su.se)

1 Introduction

Dissolved silicate (DSi) is an essential nutrient for diatoms,which play an important role in regulating the uptake and fateof C and N in the world oceans (Smetacek, 1998). Diatomdynamics is an effective measure of environmental changedue to its sensitivity to a variety of physical and ecologicalconditions. In high-latitude ecosystems of the subarctic andarctic region, climate fluctuations on short growing seasonsfor diatoms may lead to major biological influences (Ding etal., 2011). After diatom death and cell lysis, the siliceous cellwalls of diatoms can be well preserved in sediments. This al-lows for the reconstruction of diatom production by analysisof the preservation of biogenic silica (BSi) over time (Con-ley and Schelske, 2002). BSi has been used previously totrack climate-related changes in aquatic production over mil-lennial time scales (Blass et al., 2007; Rosen et al., 2000).Recently, the lacustrine BSi flux was used to reconstruct airtemperature with decadal resolution back to 1580 CE in theSwiss Alps (Blass et al., 2007) and also to quantitatively in-fer summer temperature for the past 2 kyr in South-CentralAlaska (McKay et al., 2002).

The discovery of Si isotopic fractionation of about−1.1 ‰during formation of diatom cell walls (De La Rocha et al.,1997) has provided a proxy for investigation of diatom-related processes. Many other studies have shown thatisotopic fractionation is independent of temperature, inter-species effect and the partial pressure of CO2 (De La Rochaet al., 1997; Milligan et al., 2004; Varela et al., 2004). Hence,variations in theδ30Si values of diatoms are therefore re-lated to DSi utilization efficiency in water during diatom

Published by Copernicus Publications on behalf of the European Geosciences Union.

3492 X. Sun et al.: Climate dependent diatom production is preserved in biogenic Si isotope signatures

production and can provide information on the role of the bi-ological pump (the transport of C into deep water body andthe regulation of atmospheric concentrations of CO2) in pa-leoclimatic and paleoceanographic studies (Brzezinski et al.,2002; De La Rocha et al., 1998; De la Rocha, 2006; Beucheret al., 2007; van den Boorn et al., 2010), but also ongoingenvironmental changes (Reynolds et al., 2006; Brzezinski etal., 2001; De La Rocha et al., 2000; Cardinal et al., 2005;Wille et al., 2010). For example, DSi utilization efficiencyin the Baltic Sea is also controlled by environmental con-ditions such as the supply of macronutrients (N, P, Si) andlight which are functions of the physical mixing regime ofthe upper water column (Wasmund et al., 1998). The phys-ical mixing regime is a function of local wind and tempera-ture conditions and overall climate. The Si isotope signatureof sedimentary BSi could therefore be a potential tracer forclimate variation.

In this study we report an application of Si isotope anal-yses of diatoms to reconstruct diatom production variabilityfrom 1820 to 2000 in the subarctic Bothnian Bay. The stud-ied period can be divided into an unperturbed period from1820 to 1950 and a perturbed period from 1950 to 2000.

2 Study site

Bothnian Bay (235 499 km2), located between the latitude63.5◦ N and 66◦ N, is the northern portion of the Baltic Sea(Fig. 1). The area has been subjected to recurring ice ages.The last permanent ice melted from the area about 9300 yrago. The average depth of the bay is 43 m and its maximumis 147 m. The catchment area is sparsely populated and con-sists mainly of forest, wetland, other natural areas and onlya small amount (∼1 %) of agriculture land. The total wa-ter volume of the Bay is about 1500 km3 which is 7 % ofthe total water volume of the Baltic Sea. Total water dis-charge into Bothnian Bay is 97 km3 yr−1 (Humborg et al.,2008) and, 60 % of the water is derived from rivers regulatedby dams. River regulation started at the beginning of the 20thcentury and continued until ca. 1960, with most of the damconstruction between 1940 and 1960, e.g. on the Kemijokiand Lulea River (Humborg et al., 2006). A consequence ofwhich is a decrease in fluxes of weathering related elementssuch as base cations and Si (Humborg et al., 2002, 2000) dueto changes in water pathways through the catchments andparticle trapping of BSi behind dams (Humborg et al., 2006).This has ultimately led to a decreased DSi load in the North-ern Baltic Sea (Humborg et al., 2007).

Temperature variations in the water column are>15◦Cannually. Formation of a winter ice cover in Bothnian Baystarts between the beginning of October and the end ofNovember. The entire bay is generally covered by ice evenduring the mildest winters and the ice cover is often main-tained for more than 120 days per year. During the years1961–1990, the summer (ice-free period) average maximum

surface temperature in the open sea is above 16◦C (Haa-pala and Alenius, 1994). A surface thermocline in Both-nian Bay develops about a month earlier than in the othersub-basins of the Baltic Sea and lasts for about three months.Mean DSi concentrations in Bothnian Bay observed between1970–2001 are 31 µM in winters and decrease to an averageconcentration of 23 µM in summers (Danielsson et al., 2008).Residence time for DSi is approximate 3.3 yr (Papush et al.,2009).

3 Materials and methods

3.1 Sediment core

A short sediment core (∼38 cm) taken at a depth of112 m in Bothnian Bay, Station 263870 (64◦33.581 N and21◦54.769 E, Fig. 1) was sliced into 1 cm sections andfreeze-dried. This core was dated and sediment accumula-tion rates were determined by analysing210Pb (46.51 keV),214Pb (351.99 keV) and137Cs (661.63 keV) on an EG&GORTEC® co-axial low energy photo spectrometer (LEPS)with a high-purity germanium crystal. Each sediment sectionwas measured for 2 to 3 days after having been left stand-ing for 2 weeks to achieve radionuclide re-equilibrium of de-cay products. Thereafter, the externally calibrated standard(pitchblende, Stackebo, Sweden) was added to 5 of the al-ready measured samples at different depths to determine therelatively efficiency of the gamma detector system. The sed-imentation rate was calculated by Eq. (1):

210Pbx =210Pb0 ·e−λt (1)

where210Pbx is the activity of210Pb per mass weight of sam-ple at depthx, 210Pb0 is the activity of210Pb at the surface(x = 0), λ is the decay constant of210Pb (0.0311 yr−1), and tis the age of the sediment sample. The BSi content in sedi-ments was determined using the method described by Conleyand Schelske (2002).

3.2 Diatom extraction

The method used for extraction of diatoms from the sedi-ment was modified from Morley et al. (2004). Organic car-bon was removed from each sediment slice by repeated mix-ing with 20 ml 15 % H2O2 followed by incubation for 12 hand then heating to 90◦C until no bubbles were evident af-ter subsequent addition of H2O2. Addition of 10 ml 10 %HCl and allowing the reaction to proceed for several hoursremoved inorganic carbon. There was relatively little inor-ganic carbon,<0.1 %, in most of the samples. The sampleswere washed 3 times at each step using MilliQ-e water andcentrifuged between each wash. Thereafter each sample wassieved through a 10 µm sieve cloth and an 80 µm steel meshto recover the 10–80 µm fraction and to remove silt and clay.The 10–80 µm fraction contained most of the diatoms. Purity

Biogeosciences, 8, 3491–3499, 2011 www.biogeosciences.net/8/3491/2011/

X. Sun et al.: Climate dependent diatom production is preserved in biogenic Si isotope signatures 3493

Fig. 1. Map of the study area. The regulated rivers are marked with solid lines. Unregulated rivers are indicated with dashed lines. Thesampling site is marked with a star.

was evaluated with an optical microscope and diatoms werefound to constitute about 95 % of the phytoplankton. Thesieved samples were collected in centrifuge tubes contain-ing sodium polytungstate (SPT, 3Na2WO49WO3·H2O). Thedensity of the SPT was adjusted to be between 1.8 g cm−3

and 2.3 g cm−3 in order to separate diatoms from the re-maining clay and mineral materials. Centrifugation was at4200 rpm for 5 min and repeated until clean diatoms wereretrieved. Subsequently, all diatom samples from each sed-iment slice were carefully rinsed with MilliQ-e water andsieved at 5 µm to remove all traces of SPT. Finally, the di-atom samples were dried at 40◦C for 24 h.

3.3 Diatom fusion and chromatographic purification

About 5 mg of each extracted diatom sample were mixedwith ca. 180 mg solid NaOH and fused in Ag crucibles at730◦C for 10 min in a muffle furnace. This was followed bywashing with MilliQ-e water into a 0.12 M HCl solution inorder to neutralize the NaOH (Georg et al., 2006). The finalstock solution was diluted to 200 ml to yield a Si concentra-tion of ∼10 mg l−1 and stored in pre-cleaned HDPE bottlesto minimize Si polymerization.

Interfering elements remaining after diatom dissolutionwere removed prior to Si isotope analyses by using cation ex-change columns. A protonated 1 ml BioRad cation exchangeresin AG 50W-X12 (100–200 mesh) was placed in 2 ml Bio-Rad columns. The resin rinsing and diatom sample load pro-cesses are reported in Table 1, and the recovery shown inFig. 2.

3.4 Si isotope analysis

The δ29Si value of each sediment slice was measured usinga single focus multiple-collector inductively coupled plasmamass spectrometer (MC-ICP-MS) equipped with a hexapolegas-collision cell. Analytical precision for replicate sampleswas better than 0.2 ‰ (2σ) with a total Si consumption ofless than 14 nmol (390 ng) and the internal reproducibilitywas tested by measuring IRMM-18 and Big Batch (Sun etal., 2010). The30Si could not be measured directly dueto the instrumental limitation of using low mass resolution(m/1m−1

∼450) which does not fully resolve the isobars.However, theδ30Si values were calculated from the mass ra-tio relationshipδ30Si = δ29Si× 1.96 which assumes mass-dependent fractionation (Reynolds et al., 2007).δxSi ‰ wasexpressed as relative to the standard material NBS 28 (Eq.2):

δxSi=

(

xSi28Si

)sample(

xSi28Si

)NBS 28

−1

·1000 (2)

wherex = 29,30.

4 Results and discussion

4.1 Silicon recovery of cation-exchange columns

The quality of the chemical purification is of vital impor-tance to obtain good precision and accuracy of the Si iso-tope analysis. Figure 2 shows the recovery of Si when usingthe cation-exchange column. Predominant Si species afterNaOH fusion and HCl dissolution exhibit no affinity for theresin and thus pass directly through the resin. More than

www.biogeosciences.net/8/3491/2011/ Biogeosciences, 8, 3491–3499, 2011

3494 X. Sun et al.: Climate dependent diatom production is preserved in biogenic Si isotope signatures

Table 1. The purification process used in this study for dissolveddiatom samples from Georg et al. (2006).

BioRad cation-exchange resin AG 50W-X12(100–200 mesh) in H+ form, 1 ml resin bed

Step Solution Volume (ml)Pre-cleaning 4 M HCl 3Pre-cleaning 8 M HCl 3Pre-cleaning 4 M HCl 3Conditioning MQ-e water 6Sample load Acidified diatom samples 2Elution MQ-e water 4

Fig. 2. Recovery of Si using a cation-exchange resin (AG 50W-X12,100–200 mesh) in a 1 ml resin bed.

60 % of the loaded Si is recovered from the initial elution ofthe 2 ml sample aliquot through the resin. The remaining Siin the resin is eluted using 4 ml MilliQ-e water. No break-through of ambient cations is observed if only MilliQ-e wa-ter is added. The cations start to elute when adding 4 M HCl(Fig. 2). Si is quantitatively recovered using this chromato-graphic purification with recovery efficiencies greater than99 %.

4.2 210Pb dating

The sediment core is dated using210Pb as shown in Fig. 3.Figure 3a exhibits a gradually decreasing trend with increas-ing depth. Fig. 3b shows the linear210Pb decay relation withdepths in the sediment core (r2

= 0.97). The linearity sug-gests minimal bioturbation and the slope indicates an averagesedimentation rate in this Bothnian Bay core of approximate2.1 mm yr−1 which is consistent with previous measurementsfrom this area (Grasshoff et al., 1983).

4.3 BSi vs. Si isotopes in BSi

The depth profiles of BSi contents andδ30Si values are plot-ted in Fig. 4 with the corresponding time scale obtainedfrom the 210Pb dating. The BSi content of the sediment

Fig. 3. Downcore variations in210Pb. (A) The depth profile oftotal 210Pb activity in the sediment core. The error bars repre-sent one standard deviation.(B) Correlation between depth andln (210Pb0−

210Pbx), i.e. sample210Pb activity normalized to thesurface value.

core (Fig. 4a) is approximately 3.5 % below 10 cm depthwhich is material deposited before the 1950s. It increasesto a maximum of 7.8 % in the surface sediments. This isdue to increased diatom production, probably caused by an-thropogenic nutrient emissions, especially N and P, to theBaltic Sea (Conley et al., 2008). Four periods of decreasesin BSi contents were observed around 1825, 1850, 1880 and1912. These dates are extrapolated from the sedimentationrate from the210Pb dating. The range ofδ30Si values aredisplayed in Fig. 4b and vary between−0.18 ‰ and 0.58 ‰.Before the 1950s the averageδ30Si is about 0.16 ‰ with asmall negative peak of−0.18 ‰ at 17 cm, i.e. 1910s, fol-lowed by a gradual increase to∼0.5 ‰ toward the surface.In general, increased BSi contents result in higherδ30Si val-ues.

It should be noted that the sedimentary BSi is not sim-ply a measure of the entire Bothnian Bay primary produc-tion, but is a balance between diatom production and disso-lution. The nearly linear sedimentation rate (Fig. 3) suggeststhat the variation in the sedimentation rate does not drive thevariability of the BSi contents. During diatom production,there is Si isotopic fractionation of−1.1 ‰ (De La Rocha etal., 1997). However, dissolution of diatoms also exhibits Siisotopic fractionation of−0.55 ‰ if the dissolution is morethan 20 % of the total amount of BSi (Demarest et al., 2009).This means that theδ30Si values of the preserved BSi canpotentially increase, i.e.28Si is preferentially released dur-ing diatom dissolution. Although no diatom dissolution rateestimates exist for Bothnian Bay, the excellent preservationof diatoms and the low mineralization rate in the bottom wa-ter combined with a shallow water depth of 112 m and rapidsedimentation rates imply that dissolution of BSi is probablylow. This means that the potential shift of Si isotope valuesin BSi caused by dissolution may be ignored in this study.

Biogeosciences, 8, 3491–3499, 2011 www.biogeosciences.net/8/3491/2011/

X. Sun et al.: Climate dependent diatom production is preserved in biogenic Si isotope signatures 3495

Fig. 4. Sediment profiles.(A) BSi content (%);(B) δ30Si values inBSi. The bars in 4 B represent the 95 % confidence interval.

4.4 Quantitatively reconstructing diatom production

The amount of DSi remaining in the water column is con-trolled by diatom production which occurs between themiddle of May and September in Bothnian Bay. Due to therelatively long residence time of DSi compared to a short di-atom growing season, Bothnian Bay can be assumed to bea closed system, i.e. a steady state with respect to the phys-ical input of nutrients and its biological removal, to whichRayleigh distillation equations (Eq. 3) can be applied (Hoefs,2009). The fraction of remaining DSi (f ), is a factor between0 and 1. A number close to 1 indicates that very little DSi istaken up by BSi in the water. Thef -values can be calculatedusingδ30Si values according to Eq. (3):

1−f α

1−f=

δ30SiBSi+1000

δ30Siriver input+1000(3)

whereα = 0.9989 and denotes the Si isotope fractionationfactor during diatom production (De La Rocha et al., 1997);δ30SiBSi andδ30Siriver input representδ30Si of BSi and of theriver input to Bothnian Bay, respectively.δ30SiBSi is the val-ues shown in Fig. 4 and theδ30Siriver input value is taken tobe +1.1 ‰ which is derived from measurements of an unreg-ulated boreal river, the Kalix River (Opfergelt et al., 2011).This represents unperturbed river inflow.

Each point in Fig. 5 represents the calculated 5 yr averagef value using Eq. (3) (each 1 cm section of sediments cor-responds to approximately five years). The error of the cal-culation due to the uncertainty of the measuredδ30SiBSi val-ues was examined by Monte Carlo analysis. 10 000 random

Fig. 5. Fraction of the remaining DSi (f ) in the water column ofBothnian Bay using Eq.( 3) andf -values calculated from observa-tions during the summers 1980 to 2000, plotted together with aver-age summer air temperature through years. The inferior and supe-rior error bars onf -values are the first quartile and third quartile foreachf -value derived from Monte Carlo simulations, respectively.

δ30SiBSi values were generated assuming a normal distribu-tion around a mean value and the standard deviation of thatmean. Here the average of measuredδ30SiBSi values and itsstandard deviation is used to produce 10 000f -values. Thefirst quartile and third quartile for eachf value were calcu-lated and are shown in Fig. 5 as error bars, which means that50 % of thef -values fall in this range. It should also be notedthat the calculated variations inf -values are independent oftheα-value, i.e. the absolute values off will change with theα-value, but the variation will be unchanged.

The air temperature data displayed in Fig. 5 are the ob-served average temperatures between May and September ofeach year from 1824 to 2000 with a 5 yr moving average.By plotting air temperature and the calculated fraction of theremaining DSi (f ) from Si isotope measurements a clear pat-tern emerges where cold periods are associated with highf -values, i.e. low diatom production, and warm periods showrelatively lowerf -values, i.e. high diatom production.

The temperature history and correlation with the isotopi-cally derived f -values stand out particularly for severalpeaks. For example, the last period of cooling associatedwith the Little Ice Age corresponds to the late 19th cen-tury (Bradley and Jonest, 1993). Another cold period in theearly 19th century is possibly caused by the Dalton Mini-mum, a period with low solar activity (Buntgen et al., 2006).The largest amount of remaining DSi,f = 0.98, is observedaround 1910 which corresponds to a period with very coldsummers. More recently, another period with highf -valuesis observed around 1983, which also was a period with coldsummers.

Wasmund et al. (1998) showed that diatom blooms inthe southern Baltic Sea were triggered by the reduction indepth of the mixed layer (ML) improving light conditions

www.biogeosciences.net/8/3491/2011/ Biogeosciences, 8, 3491–3499, 2011

3496 X. Sun et al.: Climate dependent diatom production is preserved in biogenic Si isotope signatures

Fig. 6. Exponential regression fit of the fraction of remaining DSi(f ) vs. the 5 yr moving average of air temperature for each depthinterval.

Fig. 7. Comparison of air temperature and water temperature be-tween 1948 and 1963.(A) Plot of daily air temperature vs. dailywater temperature.(B) Plot of average summer air temperature vs.average summer water temperature.

for phytoplankton growth. A possible explanation for theSi isotope-based diatom-temperature relationship observedin Bothnian Bay is therefore that temperature-induced varia-tions of the ML depth control diatom production. This wouldimply that air temperature regulates the water temperaturewhich is correlated with the ML depth. In fact, Fig. 7ashows daily air temperature plotted vs. daily water temper-ature between 1948 and 1963 giving a correlation coefficientof R2

= 0.78. The correlation is improved if average summertemperatures are used,R2

= 0.93 (Fig. 7b). To further sup-port our hypothesis that diatom growth and air temperatureare linked via the depth of the ML, an estimate of the correla-tion is calculated between stratification strength and surfacewater temperature for each month from May to August (Ap-pendix A, Fig. 8). Generally, the depth of the ML representedby the pycnocline depth is negatively correlated with summersurface temperature. The steepest slope appears in June andJuly (Fig. 8b and 8c), indicating a strong correlation between

the pycnocline depth and the surface temperature. Figure 8ashows a gentle slope possibly due to the ice cover in Maycausing low diatom production. In August, the slope is lesssteep than those in June and July. This indicates that sum-mer stratification isolates the growing diatoms from the deepnutrient reservoir, leading to phosphorus-limited diatom pro-duction (Humborg et al., 2003). Therefore, diatom produc-tion in June and July represents classical spring growth con-ditions and contributes most to the diatoms found preservedin sediments. In summary, diatom production in BothnianBay is controlled by depth of the ML, which is negativelycorrelated with the air temperature.

Figure 5 shows that the fraction of the remaining DSi (f -values) can be divided into two periods, before and after1950. Before 1950, Bothnian Bay is likely to have beenless disturbed due to anthropogenic eutrophication and riverregulations (Humborg et al., 2006) and may be used hereto infer the general relationship between diatom productionand air temperature in Bothnian Bay. Earlier than 1950, theδ30Siriver input value used for the calculations of thef -valueis +1.1 ‰ derived from measurements in the unperturbedKalix River (Engstrom et al., 2010). Although we cannotstate with absolute certainty that the Kalix River value canbe extended to represent all the unperturbed rivers draininginto Bothnian Bay during the studied period it is unlikelythat there is a significant difference inδ30Si values amongNordic unperturbed rivers. This assumption can be made be-cause the catchments have similar soil covers derived fromthe last glacial retreat. Soils in the area are mainly tills of lo-cal origin reflecting the Fennoscandian bedrock. The mineralcomposition of the bedrock in the entire region is dominatedby granite and gneiss and is homogenous. The soils show lit-tle disturbance due to human activities. Homogenous Si iso-topic compositions have been observed in eight of the mostcommon species of vegetation in the boreal forest (Engstromet al., 2008).

However, if we assume that the data before the 1950s rep-resent a relatively undisturbed situation both on land and inthe sea, a positive exponential relationship (Eq. (3),r2

=

0.61) can be derived by the comparison of the fraction ofremaining DSi (f ) with the 5 yr moving average of air tem-perature (T ) from 1824 to 1955 (Fig. 6).

f = 5.1e−0.15T (4)

Equation (4) is valid as long as temperature controls the MLdepth and no nutrient limitation occurs. We also have tokeep in mind that our calculatedf -values are dependent onthe isotopic fractionation factor (α in Eq. 3), which mightvary between different areas. Although the exactα-valuefor Bothnian Bay is lacking, the validation using the databetween 1980 and 2000 indicates that anα-value of 0.9989is representative of the Si isotopic fractionation factor. Forother areas, calculations with Eq. (4) might change. In addi-tion, this method is only valid as long as Rayleigh distillationcan be applied, i.e. the assumption of a closed system, in our

Biogeosciences, 8, 3491–3499, 2011 www.biogeosciences.net/8/3491/2011/

X. Sun et al.: Climate dependent diatom production is preserved in biogenic Si isotope signatures 3497

Fig. 8. Plot of the estimated pycnocline depth vs. water surfacetemperature for May(A), June(B), July (C), August(D). The cor-relation squared is also indicated in each panel.

case, much longer DSi residence time than diatom growthperiods in Bothnian Bay.

After ca. 1950, thef -values exhibit a decrease (Fig. 5).This is not consistent with calculatedf -values using fieldobservations from 1980 and onwards. To get a good fit be-tween observed and calculated data theδ30Siriver input valueused for calculations in Eq.(3) has to be shifted to +1.4 ‰.The shift coincides in time with a period of large scale hydro-electric and flood control projects in the major rivers drain-ing into Bothnian Bay (Humborg et al., 2002, 2006), leavingonly a few rivers unregulated today. There is no change inbedrock and vegetation in the catchment which suggests thatthe isotope composition of source Si is constant. The shiftin the fractionation factor is thus most likely to occur dur-ing the transport of DSi into the Bothnian Bay. Dammingof rivers increases diatom production in the reservoirs be-hind the dams (Humborg et al., 2006) and enriches the re-maining DSi in the river water with the heavier isotopes, ul-timately leading to increasedδ30Siriver input values. To testthis hypothesis, additional analysis of samples from majorrivers draining into Bothnian Bay and behind dams as wellas diatoms should be made. However, assuming that theδ30Siriver input value of +1.4 ‰ is correct, the remaining frac-tion of DSi (f ) can be calculated for the summers between1980 and 2000 and can be compared withf -values calcu-lated with Eq.(4) using the observed air temperatures forthe same period. A paired t-test shows that there is no sig-nificant difference between the observed and calculatedf -values (p ≤0.003).

5 Conclusions

Three conclusions can be drawn from this study in the sub-arctic area of the Baltic Sea, Bothnian Bay as follows:

1. stable Si isotope signature in sedimentary BSi can beused to study variations in time of diatom production;

2. air and water temperatures can be used as a proxy forthe mixed layer depth (which controls diatom produc-tion) and as shown also correlates well with diatom pro-duction;

3. large scale anthropogenic activities such as changing thehydrological regimes in rivers by damming are likely tobe imprinted on the sedimentary Si isotopic record.

Appendix A

Estimations of the relation between stratificationstrength and surface temperature

A1 Estimation of stratification strength fromobservations

A convenient way of analysing continuous stratification is touse the integrated properties profile potential energy and in-tegrated buoyancy, i.e. steric height. This is analogous to theapplications in various systems of the world (e.g. Gustafsson,1999; Bjork et al., 2001; Andersson and Stigebrandt, 2005).The profile potential energy is defined as:

P =g

ρ0

∫ 0

D

(ρ0−ρ)zdz (A1)

B =g

ρ0

∫ 0

D

(ρ0−ρ)dz (A2)

whereg is the acceleration of gravity,ρ0 is the density atdepthD andρ is the density profile.

The two layer equivalents for these are

P =g(ρ0−ρ1)

2ρ0h2 (A3)

B =g(ρ0−ρ1)

ρ0h (A4)

whereρ1 is a surface layer density and h the depth of thesurface layer. From Eqs. (A3–A4) we can derive:

h =2P

B(A5)

g(ρ0−ρ1)

ρ0=

B2

2P(A6)

www.biogeosciences.net/8/3491/2011/ Biogeosciences, 8, 3491–3499, 2011

3498 X. Sun et al.: Climate dependent diatom production is preserved in biogenic Si isotope signatures

Thus, by integrating Eqs. (A1–2) from measured profiles,we can estimate an equivalent homogeneous surface layer us-ing Eqs (A5–6). In the present application, the equivalentdepth h gives us a rough estimate on the depth of the surfacelayer.

A2 Application to Bothnian Bay

606 profiles were extracted from the BED data base. A crite-rion for selection was that the vertical resolution was at least5 m in the upper 20 m and at least 10 m down to the refer-ence depthD, which was chosen to 50 m. Each profile wasinterpolated to 1 m resolution using a cubic spline, with nor-mal conditions at the surface and at a maximal depth of theanalysis.h was calculated for each profile using eq.s (A1–2and 5).

Acknowledgements.We thank Hans Schoberg and Karin Wallnerfrom Laboratory for Isotope Geology, Swedish Museum of NaturalHistory for assisting with sample preparation. Funding from theSwedish Research Council (VR. 2007-4763) is acknowledged. Weare also grateful to the constructive comments from anonymousreviewers and George Swann.

Edited by: I. Semiletov

References

Andersson, H. C. and Stigebrandt, A.: Regulation of the Indone-sian throughflow by baroclinic draining of the North AustralianBasin, Deep-Sea Res. Part I, 52, 2214–2233, 2005.

Beucher, C. P., Brzezinski, M. A., and Crosta, X.: Silicic acid dy-namics in the glacial sub-Antarctic: Implications for the silicicacid leakage hypothesis, Global Biogeochem. Cy., 21, GB3015,doi:10.1029/2006gb002746, 2007.

Blass, A., Bigler, C., Grosjean, M., and Sturm, M.: Decadal-scale autumn temperature reconstruction back to AD 1580inferred from the varved sediments of Lake Silvaplana(southeastern Swiss Alps), Quaternary Res., 68, 184–195,doi:10.1016/j.yqres.2007.05.004, 2007.

Bjork, G., Gustafsson, B. G., and Stigebrandt, A.: Upper layer cir-culation of the Nordic seas as inferred from the spatial distribu-tion of heat and freshwater content and potential energy, PolarRes., 20, 161–168, 2001.

Bradley, R. S. and Jonest, P. D.: “Little Ice Age” sum-mer temperature variations: their nature and relevance torecent global warming trends, The Holocene, 367–376, 3,doi:10.1177/095968369300300409, 1993.

Brzezinski, M. A., Nelson, D. M., Franck, V. M., and Sigmon, D.E.: Silicon dynamics within an intense open-ocean diatom bloomin the Pacific sector of the Southern Ocean, Deep Sea ResearchPart II, Topical Studies in Oceanography, 48, 3997–4018, 2001.

Brzezinski, M. A., Pride, C. J., Franck, V. M., Sigman, D.M., Sarmiento, J. L., Matsumoto, K., Gruber, N., andRau, G. H.: A switch from Si(OH)4 to NO−

3 depletion inthe glacial Southern Ocean, Geophys. Res. Lett., 29, 12,doi:10.1029/2001GL014349, 2002.

Buntgen, U., Frank, D. C., Nievergelt, D., and Esper, J.: SummerTemperature Variations in the European Alps, 755–2004, J. Cli-mate, 19, 5606–5623,doi:10.1175/JCLI3917.1, 2006.

Cardinal, D., Alleman, L. Y., Dehairs, F., Savoye, N., Trull, T. W.,and Andre, L.: Relevance of silicon isotopes to Si-nutrient uti-lization and Si-source assessment in Antarctic waters, GlobalBiogeochem. Cy., 19, GB2007,doi:10.1029/2004GB002364,2005.

Conley, D. J. and Schelske, C. L.: Biogenic Silica, in: Tracking En-vironmental Change Using Lake Sediments: Biological Methodsand Indicators, edited by: Smol, J. P., Birks, H. J. B., and Last,W. M., Kluwer Academic Press, 3, 281–293, 2002.

Conley, D. J., Humborg, C., Smedberg, E., Rahm, L., Papush, L.,Danielsson,A., Clarke, A., Pastuszak, M., Aigars, J., Ciuffa, D.,and Morth, C.-M.: Past, present and future state of the biogeo-chemical Si cycle in the Baltic Sea, J. Marine Syst., 73, 338–346,2008.

Danielsson, P. L. and Rahm, L.: Alterations in nutrient limitations –Scenarios of a changing Baltic Sea, J. Marine Syst., 73, 263–283,2008.

Dannenberger, D., Andersson, R., and Rappe, C.: Levels and pat-terns of polychlorinated dibenzo-p-dioxins, dibenzofurans andbiphenyls in surface sediments from the western Baltic Sea(Arkona Basin) and the Oder River estuarine system, Mar.Poll. Bull., 34, 1016–1024,doi:10.1016/s0025-326x(97)00099-4, 1997.

De La Rocha, C. L.: Opal-based isotopic proxies of paleoenvi-ronmental conditions, Global Biogeochem. Cy., 20, GB2007,doi:10.1029/2005GB002664, 2006.

De La Rocha, C. L., Brzezinski, M. A., and DeNiro, M. J.: Fraction-ation of silicon isotopes by marine diatoms during biogenic silicaformation, Geochim. Cosmochim. Ac., 61, 5051–5056, 1997.

De La Rocha, C. L., Brzezinski, M. A., DeNiro, M. J., andShemesh, A.: Silicon-isotope composition of diatoms as an in-dicator of past oceanic change, Nature, 395, 680–683, 1998.

De La Rocha, C. L., Brzezinski, M. A., and DeNiro, M. J.: A firstlook at the distribution of the stable isotopes of silicon in naturalwaters, Geochim. Cosmochim. Ac., 64, 2467–2477, 2000.

Demarest, M. S., Brzezinski, M. A., and Beucher, C. P.: Frac-tionation of silicon isotopes during biogenic silica dissolution,Geochim. Cosmochim. Ac., 73, 5572–5583, 2009.

Ding, T. P., Gao, J. F., Tian, S. H., Wang, H. B., and Li, M.: Siliconisotopic composition of dissolved silicon and suspended particu-late matter in the Yellow River, China, with implications for theglobal silicon cycle, Geochim. Cosmochim. Ac., 75, 6672–6689,doi:10.1016/j.gca.2011.07.040, 2011.

Engstrom, E., Rodushkin, I.,Ohlander, B., Ingri, J., and Baxter, D.C.: Silicon isotopic composition of boreal forest vegetation inNorthern Sweden, Chem. Geol., 257, 247–256, 2008.

Engstrom, E., Rodushkin, I., Ingri, J., Baxter, D. C., Ecke, F.,Osterlund, H., andOhlander, B.: Temporal isotopic variationsof dissolved silicon in a pristine boreal river, Chemical Geology,271, 142–152,doi:10.1016/j.chemgeo.2010.01.005, 2010.

Georg, R. B., Reynolds, B. C., Frank, M., and Halliday, A. N.:New sample preparation techniques for the determination of Siisotopic compositions using MC-ICPMS, Chem. Geol., 235, 95–104, 2006.

Grasshoff, K., Erhardt, M., and Kremling, K.: Methods of sea wateranalysis, 2nd ed., Weinheim, Germany, Verlag Chemie, 1983.

Biogeosciences, 8, 3491–3499, 2011 www.biogeosciences.net/8/3491/2011/

X. Sun et al.: Climate dependent diatom production is preserved in biogenic Si isotope signatures 3499

Gustafsson, B. G.: High frequency variability of the surface layersin the Skagerrak during SKAGEX, Cont. Shelf Res., 19, 1021–1047, 1999.

Haapala, J. and Alenius, P.: Temperature and salinity, statistics forthe northern Baltic Sea, Finn. Mar. Res., 262, 51–123, 1994.

Hoefs, J.: Stable isotope geochemistry, 6th ed., Springer Berlin Hei-delberg, 9–10 pp., 2009.

Humborg, C., Conley, D. J., Rahm, L., Wulff, F., Cociasu, A., andIttekkot, V.: Silicon Retention in River Basins: Far-reaching Ef-fects on Biogeochemistry and Aquatic Food Webs in Coastal Ma-rine Environments, AMBIO: A Journal of the Human Environ-ment, 29, 45–50, 2000.

Humborg, C., Blomqvist, S., Avsan, E., Bergensund, Y., Smedberg,E., Brink, J., and Morth, C.-M.: Hydrological alterations withriver damming in northern Sweden: Implications for weatheringand river biogeochemistry, Global Biogeochem. Cy., 16, 1039–1051,doi:10.1029/2000gb001369, 2002.

Humborg, C., Danielsson,A., Sjoberg, B., and Green, M.: Nutri-ent land-sea fluxes in oligothrophic and pristine estuaries of theGulf of Bothnia, Baltic Sea, Estuarine, Coastal and Shelf Sci.,56, 781–793,doi:10.1016/s0272-7714(02)00290-1, 2003.

Humborg, C., Pastuszak, M., Aigars, J., Siegmund, H., Morth, C.M., and Ittekkot, V.: Decreased Silica Land-sea Fluxes throughDamming in the Baltic Sea Catchment-Significance of ParticleTrapping and Hydrological Alterations, Biogeochem., 77, 265–281, 2006.

Humborg, C., Morth, C.-M., Sundbom, M., and Wulff, F.: Riverinetransport of biogenic elements to the Baltic Sea – past and possi-ble future perspectives, Hydrol. Earth Syst. Sci., 11, 1593–1607,doi:10.5194/hess-11-1593-2007, 2007.

Humborg, C., Smedberg, E., Medina, M. R., and Morth, C.-M.: Changes in dissolved silicate loads to the Baltic Sea –The effects of lakes and reservoirs, J. Mar. Syst., 73, 223–235,doi:10.1016/j.jmarsys.2007.10.014, 2008.

McKay, N. P., Kaufman, D. S., and Michelutti, N.: Biogenic sil-ica concentration as a high-resolution, quantitative temperatureproxy at Hallet Lake, South-Central Alaska, Geophys. Res. Lett.,35, L05709, 10.1029/2007gl032876, 2008.

Milligan, A. J., Varela, D. E., Brzezinski, M. A., and Morel, F. M.M.: Dynamics of Silicon Metabolism and Silicon Isotopic Dis-crimination in a Marine Diatom as a Function ofpCO2, Limnol.Oceanogr., 49, 322–329, 2004.

Morley, D. W., Leng, M. J., Mackay, A. W., Sloane, H. J., Rioual,P., and Battarbee, R. W.: Cleaning of lake sediment samples fordiatom oxygen isotope analysis, J. Paleolimnol., 31, 391-401,2004.

Opfergelt, S., Eiriksdottir, E. S., Burton, K. W., Einarsson, A.,Siebert, C., Gislason, S. R., and Halliday, A. N.: Quantifyingthe impact of freshwater diatom productivity on silicon isotopesand silicon fluxes: Lake Myvatn, Iceland, Earth Planet. Sci. Lett.,305, 73–82,doi:10.1016/j.epsl.2011.02.043, 2011.

Papush, L., Danielsson,A., and Rahm, L.: Dissolved sil-ica budget for the Baltic Sea, J. Sea Res., 62, 31–41,doi:10.1016/j.seares.2009.03.001, 2009.

Reynolds, B. C., Frank, M., and Halliday, A. N.: Silicon iso-tope fractionation during nutrient utilization in the North Pacific,Earth Planet. Sci. Lett., 244, 431–443, 2006.

Reynolds, B. C., Aggarwal, J., Andre, L., Baxter, D., Beucher, C.,Brzezinski, M. A., Engstrom, E., Georg, R. B., Land, M., Leng,M. J., Opfergelt, S., Rodushkin, I., Sloane, H. J., van den Boorn,S. H. J. M., Z., V. P., and D., C.: An inter-laboratory comparisonof Si isotope reference materials, J. Anal. Atom. Spectrom., 22,561–568, 2007.

Rosen, P., Hall, R., Korsman, T., and Renberg, I.: Diatom transfer-functions for quantifying past air temperature, pH and total or-ganic carbon concentration from lakes in northern Sweden, J. Pa-leolimnol., 24, 109–123,doi:10.1023/a:1008128014721, 2000.

Smetacek, V.: Biological oceanography: Diatoms and the silicatefactor, Nature, 391, 224–225, 1998.

Sun, X., Andersson, P., Land, M., Humborg, C., and Morth, C.-M.: Stable silicon isotope analysis on nanomole quantities usingMC-ICP-MS with a hexapole gas-collision cell, J. Anal. Atom.Spectrom., 25, 156–162,doi:10.1039/B911113A, 2010.

van den Boorn, S. H. J. M., van Bergen, M. J., Vroon, P. Z., deVries, S. T., and Nijman, W.: Silicon isotope and trace ele-ment constraints on the origin of∼3.5 Ga cherts: Implicationsfor Early Archaean marine environments, Geochim. Cosmochim.Ac., 74, 1077–1103,doi:10.1016/j.gca.2009.09.009, 2010.

Varela, D. E., Pride, C. J., and Brzezinski, M. A.: Bio-logical fractionation of silicon isotopes in Southern Oceansurface waters, Global Biogeochem. Cy., 18, GB1047,doi:10.1029/2003gb002140, 2004.

Wasmund, N., Nausch, G., and Matthaus, W.: Phytoplankton springblooms in the southern Baltic Sea–spatio-temporal developmentand long-term trends, J. Plankton Res., 20, 1099–1117, 1998.

Wille, M., Sutton, J., Ellwood, M. J., Sambridge, M., Maher, W.,Eggins, S., and Kelly, M.: Silicon isotopic fractionation in ma-rine sponges: A new model for understanding silicon isotopicvariations in sponges, Earth Planet. Sci. Lett., 292, 281–289,doi:10.1016/j.epsl.2010.01.036, 2010.

www.biogeosciences.net/8/3491/2011/ Biogeosciences, 8, 3491–3499, 2011

Paper III

Silicon isotope enrichment in diatoms during nutrient-limited blooms

in a eutrophied river system

Xiaole Suna1

, Per Anderssonb, Christoph Humborg

c, Marianna Pastuszak

d, Carl-Magnus Mörth

a

a Department of Geological Sciences, Stockholm University, 10691, Stockholm, Sweden

b Laboratory for Isotope Geology, Swedish Museum of Natural History, 10405, Stockholm, Sweden

c Department of Applied Environmental Science, Stockholm University, 10691, Stockholm, Sweden

d Department of Fisheries Oceanography and Marine Ecology, Sea Fisheries Institute, 81332, Gdynia, Poland

Abstract

We examined the Si isotope fractionation in diatoms by following a massive nutrient

limited diatom bloom from a eutrophied natural system. We hypothesized that the Si

isotope fractionation should be larger in comparison to observations in less nutrient rich

environments. The Oder River, which is a eutrophied river draining the western half of

Poland and entering the southern Baltic Sea, shows that a diatom bloom may cause extreme

Si isotope fractionation. The rapid nutrient depletion and fast biogenic silica (BSi) increase

observed during the spring bloom suggests a Rayleigh behavior for a closed system for

dissolved Si (DSi) and BSi in the river at certain time scales. An enrichment factor (ε) of up

to -1.6‰ is found based on observations between April and June, 2004. A very high δ30

Si

value of up to +3.05‰ is measured in diatoms. This is about 2 times higher than previously

recorded δ30

Si in freshwater diatoms. The Rayleigh model used to predict the δ30

Si values

of DSi suggests that the initial value before the start of the diatom bloom is close to +2‰.

This indicates that there is a biological control of the Si isotope compositions entering the

river, probably caused by Si isotope fractionation during uptake of Si in phytoliths. Clearly,

eutrophied rivers with enhanced diatom blooms deliver 30

Si-enriched DSi and BSi to the

coastal ocean, which can be used to trace the biogeochemistry of DSi/BSi in estuaries.

Keywords: biogenic silica, Si isotopes, a eutrophied river system, nutrient limitation

1 Corresponding author: xiaole.sun@geo.su.se

Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system

- 1 -

1. Introduction

Dissolved Si (DSi) is an essential nutrient in aquatic

systems for the growth of siliceous organisms such as

diatoms, radiolarians, and sponges, which play an

important role in carbon sequestration (Goldman,

1988; Nelson et al. 1995). These organisms use Si to

produce structural components made of opal, i.e.

biogenic silica (BSi). The interactions between Si

and aquatic ecosystems have been of wide interest

since Schelske and Stoermer (1971) introduced the

hypothesis that DSi depletion caused by excessive

growth of diatoms was due to increased fluxes of

phosphorus in Lake Michigan, USA. The same

phenomenon has also been described for the other

Great Lakes in North America (Schelske et al. 1983;

Conley et al. 1993). Exhaustion of DSi has also been

reported in Chesapeake Bay (Fisher et al. 1988;

Conley and Malone 1992) and in the Mississippi

River plume, where the DSi concentration decreased

to < 0.93 µmol L-1

(Nelson and Dortch 1996).

Around 90% depletion of DSi has been reported in

High Nutrient Low Chlorophyll (HNLC) coastal

ocean water around Peru (Dugdale et al. 1995), while

approximately 77% utilization of DSi was recorded

in the Southern Ocean (Brzezinski et al. 2001;

Leynaert et al. 2001). This indicates that some

coastal areas are approaching Si limitation and

therefore understanding the interaction between

siliceous organisms and DSi is important.

The total algal growth in an aquatic ecosystem is

primarily regulated by the availability of N and P.

Availability of Si relative to N and P, i.e. Si:N and

Si:P ratios, can influence the composition of the

coastal phytoplankton community. A decreasing Si:N

ratio may exacerbate eutrophication by reducing the

potential for diatom growth, in favor of noxious

flagellates (Officer and Ryther 1980; Conley et al.

1993; Humborg et al. 1997). Non-diatom species are

known to be less available to higher trophic levels

and some non-diatom based food webs are

economically undesirable. Therefore, the proportion

of diatoms in the phytoplankton community is of

primary importance globally for many fisheries

(Struyf et al. 2009).

Silicon isotopes have been recognized as an

important proxy, not only for paleoclimatic and

paleoceanographic studies (e.g. De La Rocha et al.

1998; Brzezinski et al. 2002; Van Den Boorn et al.

2010), but also for tracing present environmental and

ecological changes (e.g. Cardinal et al. 2005;

Engström et al. 2010; Wille et al. 2010). The

principle behind these studies is that diatoms

Fig. 1 Si isotope variations under closed- and open-

system Rayleigh conditions for diatom production.

During diatom production (fraction of remaining DSi

drops), the δ30

Si of DSi and instant BSi increases

correspondingly. At the same time, the δ30

Si of the

accumulated BSi in both systems also increases as DSi

concentration decreases, but only up to the initial δ30

Si of

DSi, i.e. dashed line.

preferentially take up 28

Si during cell wall formation,

resulting in a Si isotope fractionation factor of

approximate -1.1‰ (De La Rocha et al. 1997). This

selective utilization of DSi leads to lower Si isotope

values in diatom opal shells, leaving behind non-

utilized DSi enriched in heavier Si isotopes with time

(higher Si isotope values). Thus, measuring Si

isotopes in DSi in aquatic systems and knowing the

range of isotope fractionation that diatom blooms

produce will allow calculating the magnitude of

diatom blooms. The magnitude of Si isotope

fractionation is apparently neither affected by CO2

levels (Milligan et al. 2004), nor temperature in three

species of diatoms during their growth (De La Rocha

et al., 1997), and it is similar for diatoms both in

freshwater and in marine systems (Alleman et al.

2005). Therefore, the change in Si isotopic

composition generally depends on the availability of

DSi, i.e. input and output of Si to the system, and

whether the system is closed or open. The Si isotope

composition in a closed and open system can be

calculated by assuming a Rayleigh distillation

behavior (Fig. 1). In this way, Si isotopes determined

in DSi may be used to estimate both the magnitude of

diatom blooms and other related fluxes, e.g. the

downward flux of carbon.

0.00.20.40.60.81.0

δ30S

i (‰

)

f, fraction of remaining DSi

remaining DSi (closed)

accumulated BSi (closed)

instant BSi (closed)

accmulated BSi (open)

Remaining DSi (open)

Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system

- 2 -

The Si isotope fractionation caused by diatom

production is also considered to be responsible for an

enrichment of the heavier Si isotopes in river waters

in comparison to primary minerals in igneous rocks

in river catchments (De La Rocha et al. 2000; Ding et

al. 2004; Georg et al. 2006a). Silicon taken up by

vegetation (vascular plants) in the watershed is

another possible explanation for the positive shift to

higher δ29

Si and δ30

Si values in river water (Opfergelt

et al. 2006). Furthermore, the production rate of

diatoms is rather high (Conley and Malone 1992),

which suggests that effects on the DSi concentration

and Si isotope values may occur on very short time

scales. A recent study with monthly data on Si

isotopes in DSi and BSi in the Congo River has

shown seasonal variations explained by diatom

production, but the seasonality effect from diatom

growth in a tropical river is limited (Hughes et al.

2011).

In the Baltic Sea, especially in the Gulf of Riga and

the Gulf of Finland, DSi-limited diatom production

has been reported (Conley et al. 2008; Danielsson et

al. 2008). In rivers such as the Oder River, draining

into the southwestern Baltic Sea, there is evidence

for DSi-limited diatom production. The BSi fraction

has in some cases been reported to comprise 99.8%

of the total Si pool during spring and summer, but in

winter, the fraction of BSi diminishes to only 0.8%

(Humborg et al. 2006; Pastuszak et al. 2008). Thus,

BSi from the Oder River can be considered as an

important source for Si in the Baltic Sea (Humborg et

al. 2006). Since this river system is eutrophic and has

high diatom production with frequently occurring Si

limitation, it is a potentially good site to evaluate

seasonal changes in Si isotopes in river water, and

may even show the maximum possible range of Si

fractionation in a natural aquatic system.

2. Materials and methods

2.1 Nutrient and water flow data sources

The biweekly water sampling for nutrients (nitrogen,

phosphorus, and silicon) was performed at the

lowermost Oder River monitoring station (Krajnik

Dolny) (Fig. 2) by the Sea Fisheries Institute (SFI)

Branch in Świnoujście (Poland) during the period

January 2003 - June 2005 within the EU SIBER

project. Nutrient analyses were done by the SFI

according to standard methods as recommended by

Grasshoff (1983) and UNESCO (1983). The nutrient

data have been published in Pastuszak et al. (2008)

and Pastuszak and Witek (2009a,b). The biweekly

measurements of BSi at the same sampling site were

performed from April 2003 to June 2005. The BSi

samples were prepared in duplicates; 200 ml to 400

ml of water was filtered using polycarbonate 0.6

micron 47 mm filters and Sartorius polycarbonate

holders for vacuum sterile filtration. Filters with the

samples were dried at room temperature and stored in

microbiologically sterilized 47 mm Millipore®

petri dishes. BSi analyses from a portion of the filter

samples were performed according to the method by

DeMaster (1981) (Pastuszak et al. 2008), while the

remaining fraction was subjected to diatom analyses

performed at Stockholm University. The monthly

discharge in the Oder River during 2004 was

obtained from IMWM (2004) (Fig. 3).

Fig.2 Map of the lower Oder River indicating the sampling

site in Krajnik Dolny.

Fig.3 Water discharge in the Oder river during 2004

(source: IMWM, 2004)

2.2 Sample preparation

0

200

400

600

800

wate

r d

isch

arg

e (m

3/s

)

Sampling date

Krajnik Dolny

Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system

- 3 -

Diatom samples from April to September 2004 were

selected in this study due to availability of samples

and BSi contents although DSi samples were not

available since the DSi concentrations were very low

and therefore could not be measured. Each diatom

sample deposited on the filter was mixed with 4 mL

dichloromethane in a polypropylene centrifugation

tube. The filter dissolved rapidly and left diatoms and

insoluble organic and mineral materials residue.

After centrifugation, the top clear solution was

removed and new dichloromethane was added. This

was repeated several times to ensure that the entire

polycarbonate filter was dissolved and removed.

After this procedure the samples were dried at 40 °C

to evaporate all of the dichloromethane.

In order to remove the organic carbon the dried

diatom samples were treated with 15% H2O2 and

heated until all visible bubbles disappeared. This was

followed by treatment with SPT heavy liquid

(3Na2WO49WO3·H2O) to separate the diatom

fraction from the remaining inorganic material. The

SPT density was adjusted to a range between 1.9 g

cm-3

and 2.3 g cm-3

(Morley et al. 2004). Thereafter,

all diatom samples were carefully rinsed with

MilliQ-e water to remove all the SPT, and dried at 40

˚C for 24 h. The samples were examined under an

optical microscope to ensure a pure diatom fraction.

The dried diatoms were fused with NaOH and

dissolved in 0.12 mol L-1

HCl. Finally, a cation-

exchange column was used to remove the matrix of

the diatom solutions. The methods used in this step

was adopted and modified from the preparation

technique by Georg et al. (2006b).

2.3 Si isotope analyses

The Si isotope values of the diatom samples were

determined by MC-ICP-MS (Multi Collector

Inductively Coupled Plasma Mass Spectrometer)

equipped with a hexapole gas-collision cell, capable

of precisely and accurately measuring the δ29

Si value

with a total Si consumption of less than 14 nmol (390

ng). The standard-sample bracketing technique was

employed and each sample was measured at least 5

times. The internal reproducibility was examined by

measuring IRMM-18 and Big Batch, was better than

0.2‰ (2). The detailed instrument settings and

measurement procedure is documented in Sun et al.

2010. The isotope composition is expressed as δ29

Si

per mil (‰) relative to NBS 28 (Eq.1).

29

28

sample29

29

28

28

1 1000

NBS

Si

SiSi

Si

Si

Eq.1

Because of the instrumental limitation of measuring

in low mass resolution (m/Δm ~ 450), the δ30

Si was

not measured in this study. Therefore, the δ30

Si was

calculated using the formula δ30

Si = δ29

Si×1.96 i.e.

assuming a mass-dependent fractionation of Si

isotopes (Reynolds et al., 2007). This relationship has

been demonstrated by others and it has been shown

that long-term reproducibility are within an error of

less than 0.15‰ for δ29

Si and 0.24‰ for δ30

Si

(Engström et al. 2006; Reynolds et al. 2007;

Chmeleff et al. 2008).

3. Results and discussion

3.1 Possible range of Si isotope values in diatoms

The average concentrations of DSi, BSi, DIN

(dissolved inorganic nitrogen) and DIP (dissolved

inorganic phosphate), calculated for the period April

to September 2004, are 37.2 µmol L-1

, 58.5 µmol L-1

,

47.9 µmol L-1

, and 0.5 µmol L-1

, respectively in the

river water (Table 1). The concentrations of DSi,

DIN and DIP display similar variations, i.e.

decreasing throughout the spring and summer and

increasing in the autumn (Fig. 4). In contrast to the

nutrients, BSi shows increasing concentration, which

indicates a period of rapid diatom growth in spring

and summer, and decreasing concentration in late

summer and autumn (Fig. 4). The Oder River shows

peak discharge during snowmelt and early spring

(Fig. 3) and during peak flow the fast flowing water

carries more suspended matter (soil particles)

compared to the slowly flowing water during the

other seasons, leading to the increased water turbidity

and reduced light condition, which is a crucial

parameter in diatom development. Despite abundant

nutrients at this time of the year, diatoms show low

production. After the fast transition from the

snowmelt to late spring, soil particles settle out of the

water column.

There are a number of shortcomings in using N:P:Si

ratios to evaluate the limiting role of nutrients in

phytoplankton growth. These ratios might be

inaccurate when reaching very low concentrations

close to the detection limit, and calculated ratios may

be affected by large errors. Nonetheless, nutrient

ratios have still been widely used (Howarth 1988;

Fisher et al. 1992). It is shown that diatoms in

freshwater having sufficient nutrients produce

biomass with a C:N:Si:P ratio of 106:16:84:1

(Conley et al. 1989; Goldman 1988) and that

variations in the stoichiometry of dissolved inorganic

nutrients can provide evidence for which nutrient is

most likely to limit the biomass production.

Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system

- 4 -

Table 1 Temporal nutrient and BSi concentrations, as well as Si isotopic compositions of BSi expressed by δxSi in the

Oder River. The uncertainty for δxSi is reported by 95% confidence interval (C.I.).

Sampling

date 2004

DSi

(µM)a

BSi

(µM)a

DIN

(µM)a

DIP

(µM)a

DIN:DIP DSi:DIN DSi:DIP δ

29Si

(‰)

δ30

Si

(‰)b

95%

C.I.

5 April 158.3 21.3 186.4 0.9 200.4 0.8 170.2 0.38 0.75 0.24

21 April 78.8 33.5 168.0 0.7 258.4 0.5 121.2 0.58 1.13 0.19

10 May 12.7 68.5 111.4 0.9 131.1 0.1 15.0 0.76 1.49 0.24

24 May 7.2 78.7 74.2 0.3 239.4 0.1 23.3 0.61 1.20 0.22

07 June 0.2 81.3 14.4 0.4 36.1 0.0 0.4 1.12 2.12 0.18

21 June 1.7 94.6 8.4 0.4 22.6 0.2 4.7 1.25 2.35 0.14

08 July 2.0 100.2 1.0 0.1 6.9 4.0 27.5 - - -

23 July 5.8 89.7 0.2 0.1 2.9 34.2 98.7 - - -

6 August 23.0 76.7 0.5 0.5 1.0 48.5 49.9 1.56 3.05 0.21

24 August 32.9 36.2 1.5 0.4 3.7 21.6 80.0 1.29 2.53 0.21

07 Sept. 46.3 14.4 0.6 0.2 2.8 73.4 202.2 0.60 1.18 0.24

21Sept. 73.9 10.4 7.8 0.6 12.1 9.5 114.9 0.56 1.09 0.13

Average 37.2 58.5 47.9 0.5 76.4 16.1 75.7 0.87 1.71 0.092 a Measured by the SFI, Gdynia, Poland and published by Pastuszak et al. (2008) and Pastuszak and Witek (2009)

b Calculated by δ

30Si = δ

29Si×1.96 (Reynolds et al., 2007)

- sample not available

Fig. 4 DIN, DIP, DSi and BSi concentrations in the Oder

River during April to September 2004. The inset shows a

magnified scale for the June and July DIN, DIP and DSi

data (source: Pastuszak et al., 2008; Pastuszak and Witek,

2009).

The DIP concentration decreased to about 0.5 µmol

L-1

during spring blooms (Fig. 4) and thus seems to

be the primary limiting nutrient for the diatom

growth. The DSi:DIP ratio reached the highest value,

with a maximum of about 170 on 5 April 2004

(Table 1). In June there is a sharp decrease in the DSi

concentrations from 1.7 to 0.2 µmol L-1

(Pastuszak et

al. 2008), which is below the level of approximately

2 M reported as a threshold concentration for

diatom growth (Egge and Aksnes 1992). The

DSi:DIP ratio decreased to very low values, 0.4 to

4.7, and thus the limiting nutrient shifted from DIP to

DSi. The δ30

Si values in diatoms increased from

+0.75‰ in April to values more than +2.46‰ in June.

The period from July to September was characterized

by a sharp decrease in the DIN concentrations

(DIN=NO3-N+NO2-N+NH4-N), from 7.8 to 0.2 µmol

L-1

(Pastuszak and Witek 2009a; Pastuszak and

Witek 2009b) simultaneously with the diatom

biomass peak of 100.2 µmol L-1

in July, which is

followed by a decrease due to the exhausted nutrient

supply. The very low DIN concentration during

summer was a result of the low water flux, which

contributes to lower N emissions to the river basin.

Nitrogen mainly originates from diffuse sources,

with agricultural activity playing a key role

(Pastuszak et al. 2011). The DSi:DIN ratio reaches

48.5 whereas the DIN:DIP decreased to as low as 1.0

in August 2004 and the limiting nutrient was

thereafter shifted to nitrogen (Table 1).

The BSi shows positive δ29

Si values throughout the

sampling period with an average of +0.87‰, and a

calculated δ30

Si value of +1.7‰ (Fig. 5). This is

consistent with previously reported δ30

Si values in

natural diatoms ranging from -0.3 to +2.6‰ (De La

Rocha et al., 1998; Varela et al., 2004). However, the

Si isotope compositions in the diatoms show large

variations during spring and summer. The δ30

Si

values increase from a minimum value of +0.75‰ in

the spring to a maximum in excess of +3‰ at the

beginning of August, which is higher than previously

recorded values in freshwater diatoms, although there

is a small drop in May (Fig. 5). The δ30

Si values start

to decrease in late August, and in September the δ30

Si

values reach around +1.09‰ similar to that observed

in April, which is the starting value of the spring

bloom. These changing ratios during the growth

Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system

- 5 -

Fig. 5 The Si isotopic composition of BSi reported as

δ30Si, and DSi and BSi concentrations in the Oder River

during April to September 2004. The δ30

Si error bars

represent the 95% confidence interval. DSi and BSi values

from Pastuszak et al., 2008.

season indicate that the Oder River is facing an

alternating nutrient limitation of the diatom

production, which is simultaneously leading to the

very high δ30

Si values in the diatoms.

3.2 Si isotope fractionation during spring blooms

The relationship between the variations of the δ30

Si

values and DSi and BSi concentrations between

April and June, 2004 (Fig. 5) suggests that the

diatom production dominates, as demonstrated by the

rapid and sharp decrease in the DSi concentration. It

is checked under a light microscope that BSi mainly

originates from the diatom bloom and not from

phytoliths or weathering of silicate minerals

delivered from the terrestrial environment. Fig. 6

shows that δ30

Si values between April and June are in

agreement with the curve of accumulated BSi in a

closed system. Therefore, we hypothesize that the

river system is a closed Si reservoir at the timescale

of diatom blooms, i.e. the production rate of diatoms

is considerably faster than replenishment of new DSi

to the system, especially during the DIP and DSi

dramatically decreasing period. In contrast to the

observed concentrations, a completely open system

would show very small changes in the DSi

concentrations.

The Rayleigh model for a closed system is

introduced here to constrain the isotope variations.

Fig. 6 shows that the measured data from April to

June, 2004 are plotted together with the Rayleigh

model for the accumulated BSi (Eq. 2) (Hoefs 2009).

The reason why this is not extended to September

will be discussed later. Apparently, the measured

data do not fit for the Rayleigh model for an open

system. The fractionation factor (α30/28) of 0.9984 as

well as initial δ30

Si of riverine input is derived from

the model for a closed system expressed by Eq.3

(R2=0.66), which is a transformation of Eq.2 (Fig. 6).

In addition, a curve of the δ30

Si in the remaining DSi

fraction is constructed using the Rayleigh equation

for the depletion of remaining DSi (Eq.4):

30

30

10001

1 1000

BSi

river input

Sif

f Si

Eq.2

0.998430 1

(2.0 1000) 10001

BSi

fSi

f

Eq.3

30 30 ( 1)( 1000) 1000DSi river inputSi Si f Eq.4

where α is the Si isotope fractionation factor during

diatom production; δ30

SiBSi, δ30

SiDSi and δ30

Siriver input

represent δ30

Si in BSi, remaining DSi and the river

input, respectively.

The Si isotope fractionation factor (α30/28) during

diatom production in the Oder River reached 0.9984

corresponding to an enrichment factor εDSi-BSi of -

1.6‰. The error of the calculation due to the

uncertainty of the measured δ30

SiBSi values was

examined by Monte Carlo analysis. 5000 random

δ30

SiBSi values were generated assuming a normal

distribution around a mean value and the standard

deviation of that mean. Each of measured δ30

SiBSi

values and its standard deviation is used to produce

5000 εDSi-BSi values, giving an average εDSi-BSi of -1.6

± 0.04‰ (95% confidence interval). However, the

fractionation factor is still lower than -1.1‰ reported

for diatom culturing laboratory experiments (De La

Rocha et al. 1997) and -1.1‰ ± 0.4‰ found in Lake

Tanganyika (Alleman et al. 2005).

Fig. 6 The δ

30Si plotted vs. the fraction of remaining of

DSi. Si isotope change under closed- and open-system

Rayleigh model for diatom production with α= 0.9984.

Open symbols represent measured data from April to

September 2004.

0

50

100

150

200

0

1

2

3

4

Apr.

05

Apr.

21

May

10

May

24

June

07

June

21

July

08

July

23

Aug.

06

Aug.

24

Sep.

07

Sep.

21

DS

i, B

Si

con

cen

trat

ion

(µ

M)

δ 3

0S

i (‰

)

Sampling date

δ³°Si in BSi

BSi concentration (µM)

DSi concentration (µM)

0

2

4

6

8

0.00.20.40.60.81.0

δ 3

0S

i (‰

)

f, fraction of remaining DSi

measuredDSi (closed)DSi (open)BSi (open)accumulated BSi (closed)instant BSi (closed)

April May

June

September

August

April May

June

August

September

Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system

- 6 -

Compared to marine diatoms, the εDSi-BSi of -1.6‰ is

also lower than the values observed in the Southern

Ocean of -0.6‰ (Demarest et al. 2009), but within

the range of -1.1‰ to -1.9‰ found in the Antarctic

circumpolar current (Varela et al. 2004). The rapid

depletion of DSi leads to a variation in the δ30

Si

values in diatoms in excess of +2.30‰ (Fig.5), which

is around 1.5-2 times higher compared to previously

reported values in diatoms from freshwater systems.

These findings indicate that diatoms are able to

fractionate Si isotopes to very high values in a

freshwater system limited by nutrient supply. As

shown in Fig.6, if the DSi concentration decreases to

a very low level and with a simultaneous increase in

the BSi concentration, the Si isotope values in the

remaining DSi could in the Oder River be extremely

higher than any current recorded values. The

regression curve in Fig. 6 indicates that the BSi

accumulation in the Oder River follows a Rayleigh

behavior for a closed system, which means that Eq.2

could be used to reconstruct diatom production

represented by f values if α30/28 and the Si isotope

values in BSi and DSi are known. It should be noted

that clay formation can also lead to Si isotope

fractionation, and it remains unclear as to what

degree this process is controlling the δ30

Si values in

the Oder River. This question calls for future studies,

with a high-resolution time series of water and

sediment samples.

The regression curve (Eq.3) also suggests that the

δ30

Si value in DSi of river water (δ30

Siriver input) prior

to the spring bloom is around +2‰, i.e. the isotope

starting value for diatom-driven Si isotope

fractionation. This value is in agreement with the

range of BSi in phytoliths documented to be in the

range between -1.4 and +2.8‰ (Douthitt 1982) and

also in the range of DSi in soil solution, -0.8 to

+1.7‰. One possible reason for this relatively high

value of +2‰ could be that the DSi contribution to

the Oder River preferentially comes from the upper

soil layers, where biogeochemical cycling of Si

results in more enriched, positive δ30

Si values, than

in the lower soil layers indicating a biological control

of δ30

Si values export to this aquatic system (Derry et

al. 2005; Ding et al. 2008). This might be a

fundamental process in a variety of geological

settings, supporting Si uptake by vegetation as

important for Si cycling.

In August 2004, the Si isotope composition in BSi

increased to a peak value of +3.05‰, while the DSi

concentration increased with decreasing BSi

concentration. This does not fit Rayleigh distillation

behavior in a closed system (Fig. 6). The high δ30

Si

value is most likely affected by other processes. One

reason could be that the diatom dissolution rate is

enhanced, whereas the diatom production rate is low

due to the nutrient limitations, especially DIN. This

dissolution can lead to preferentially releasing 28

Si,

which can exhibit an enrichment factor of -0.55‰,

enriching the heavier Si isotope in the remaining BSi,

if dissolution is higher than 20% of the total BSi

(Demarest et al., 2009). This can potentially shift the

δ30

Si values in the remaining BSi towards even

higher values. The true mechanism of dissolution-

induced influence on high δ30

Si value is not clear. It

is also possible that increased input of a source with a

high δ30

Si value is coming from upstream eutrophied

areas in the watershed area, mostly consisting of

agricultural lands and forest (Humborg et al. 2006;

Humborg et al. 2008). This would lead to not only

enhanced primary production in the river, but also

increased uptake of DSi by terrestrial and aquatic

vegetation (Ding et al. 2005). This does not seem to

be a likely reason because of rather low river flow

and very limited diatom production. It is also

possible that Si is incorporated into secondary

minerals, e.g. kaolinites, and since 28

Si is

preferentially bonded to secondary mineral phases

this can result in an increasing δ30

Si value of the

dissolved phase (De La Rocha et al., 2000; Ziegler et

al., 2005). However, this process is kinetically very

slow and therefore not very likely. Another possible

reason is that the samples from August is sampled

near the surface of water column, so it is actually a

mixture of instant BSi and accumulated BSi, as

shown in Fig. 6 those data points stand in the middle

of instant and accumulated curves. DSi distribution

in river is not homogeneous, the sampled diatoms is

from a part of water with little DSi, and water

samples from another part of river with more DSi.

This means in Fig. 6, the data points in August are

not associated with real f values, i.e. not real Si

utilization during production of this sampled diatom.

Since eutrophied rivers, such as the Oder River, with

enhanced diatom production and dissolution are

important sources for DSi delivered to the coastal

ocean, the enrichment of 30

Si in DSi makes it

possible to trace the fate and transport of the DSi and

BSi in the Baltic Sea and other similar environments.

4. Conclusion

This rapid nutrient depletion and fast diatom

production during the spring bloom allows a closed

system approximation for the DSi in the river. A

Rayleigh distillation model is used to calculate a Si

isotope fractionation factor of -1.6‰ which is higher

compared to other freshwater systems, -1.1‰ and is

Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system

- 7 -

possibly the upper range of Si fractionation in natural

aquatic systems. The δ30

Si values for the river

diatoms are also constructed and similar to the

observed data showing a seasonal pattern. The

calculated input of the δ30

Si value before the onset of

the diatom bloom is about +2‰. This value coincides

with reported data for phytoliths and therefore

suggests that there is a biologic control of the Si

isotopic composition of the input to the river. A high

δ30

Si value up to 3.05‰ in diatoms is observed in

August with simultaneous rapid and strong depletion

of the major nutrients (N, P and Si). Since eutrophied

rivers, such as the Oder River, involving many

complex processes, the signatures of Si isotopes are

helpful to unravel diatom-related processes in the

Baltic Sea and other similar environments.

Acknowledgements

We would like to thank Hans Schöberg from the

Laboratory for Isotope Geology at the Swedish

Museum of Natural History for assisting with sample

preparation. Thanks also to Eve Arnold for English

checking. We are grateful for the funding from the

Swedish Research Council (VR-2007-4763).

References

Alleman, L.Y., Cardinal, D., Cocquyt, C., Plisnier,

P.-D., Descy, J.-P., Kimirei, I., Sinyinza, D.,

André, L., 2005. Silicon Isotopic Fractionation in

Lake Tanganyika and Its Main Tributaries.

Journal of Great Lakes Research 31, 509-519.

Brzezinski, M.A., Nelson, D.M., Franck, V.M.,

Sigmon, D.E., 2001. Silicon dynamics within an

intense open-ocean diatom bloom in the Pacific

sector of the Southern Ocean. Deep Sea Research

Part II: Topical Studies in Oceanography 48,

3997-4018.

Brzezinski, M.A., Pride, C.J., Franck, V.M., Sigman,

D.M., Sarmiento, J.L., Matsumoto, K., Gruber, N.,

Rau, G.H., 2002. A switch from Si (OH) 4 to NO

3 depletion in the glacial Southern Ocean.

Geophysical research letters 29, No. 12.

Cardinal, D., Alleman, L.Y., Dehairs, F., Savoye, N.,

Trull, T.W., André, L., 2005. Relevance of silicon

isotopes to Si-nutrient utilization and Si-source

assessment in Antarctic waters. Global

Biogeochem. Cycles 19, GB2007.

Chmeleff, J., Horn, I., Steinhoefel, G., von

Blanckenburg, F., 2008. In situ determination of

precise stable Si isotope ratios by UV-

femtosecond laser ablation high-resolution multi-

collector ICP-MS. Chemical Geology 249, 155-

166.

Conley, D.J., Humborg, C., Smedberg, E., Rahm, L.,

Papush, L., Danielsson, Å., Clarke, A., Pastuszak,

M., Aigars, J., Ciuffa, D., Mörth, C.-M., 2008.

Past, present and future state of the

biogeochemical Si cycle in the Baltic Sea. Journal

of Marine Systems 73, 338-346.

Conley, D.J., Kilham, S.S., Theriot, E., 1989.

Differences in Silica Content Between Marine and

Freshwater Diatoms. Limnology and

Oceanography 34, 205-213.

Conley, D.J., Malone, T.C., 1992. Annual cycle of

dissolved silicate in Chesapeake bay: implications

for the production and fate of phytoplankton

biomass Marine Ecology Progress Series 81, 121-

128.

Conley, D.J., Schelske, C.L., Stoermer, E.F., 1993.

Modification of the biogeochemical cycle of silica

with eutrophication. Marine Ecology Progress

Series 101, 179-192.

Danielsson, Papush, L., Rahm, L., 2008. Alterations

in nutrient limitations -- Scenarios of a changing

Baltic Sea. Journal of Marine Systems 73, 263-

283.

De La Rocha, C.L., Brzezinski, M.A., DeNiro, M.J.,

1997. Fractionation of silicon isotopes by marine

diatoms during biogenic silica formation.

Geochimica et Cosmochimica Acta 61, 5051-

5056.

De La Rocha, C.L., Brzezinski, M.A., DeNiro, M.J.,

2000. A first look at the distribution of the stable

isotopes of silicon in natural waters. Geochimica

et Cosmochimica Acta 64, 2467-2477.

De La Rocha, C.L., Brzezinski, M.A., DeNiro, M.J.,

Shemesh, A., 1998. Silicon-isotope composition

of diatoms as an indicator of past oceanic change.

Nature 395, 680-683.

Demarest, M.S., Brzezinski, M.A., Beucher, C.P.,

2009. Fractionation of silicon isotopes during

biogenic silica dissolution. Geochimica et

Cosmochimica Acta 73, 5572-5583.

DeMaster, D.J., 1981. The supply and accumulation

of silica in the marine environment. Geochimica

et Cosmochimica Acta 45, 1715-1732.

Derry, L.A., Kurtz, A.C., Ziegler, K., Chadwick,

O.A., 2005. Biological control of terrestrial silica

cycling and export fluxes to watersheds. Nature

433, 728-731.

Ding, T., Wan, D., Bai, R., Zhang, Z., Shen, Y.,

Meng, R., 2005. Silicon isotope abundance ratios

and atomic weights of NBS-28 and other

reference materials. Geochimica et Cosmochimica

Acta 69, 5487-5494.

Ding, T., Wan, D., Wang, C., Zhang, F., 2004.

Silicon isotope compositions of dissolved silicon

Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system

- 8 -

and suspended matter in the Yangtze River, China.

Geochimica et Cosmochimica Acta 68, 205-216.

Ding, T.P., Tian, S.H., Sun, L., Wu, L.H., Zhou, J.X.,

Chen, Z.Y., 2008. Silicon isotope fractionation

between rice plants and nutrient solution and its

significance to the study of the silicon cycle.

Geochimica et Cosmochimica Acta 72, 5600-

5615.

Douthitt, C.B., 1982. The geochemistry of the stable

isotopes of silicon. Geochimica et Cosmochimica

Acta 46, 1449-1458.

Dugdale, R.C., Wilkerson, F.P., Minas, H.J., 1995.

The role of a silicate pump in driving new

production. Deep Sea Research Part I:

Oceanographic Research Papers 42, 697-719.

Egge, J.K., Aksnes, D.L., 1992. Silicate as regulating

nutrient in phytoplankton competition. Marine

Ecology Progress Series 83, 281-289.

Engström, E., Rodushkin, I., Baxter, D.C., Ohlander,

B., 2006. Chromatographic Purification for the

Determination of Dissolved Silicon Isotopic

Compositions in Natural Waters by High-

Resolution Multicollector Inductively Coupled

Plasma Mass Spectrometry. Analytical Chemistry

78, 250-257.

Engström, E., Rodushkin, I., Ingri, J., Baxter, D.C.,

Ecke, F., Österlund, H., Öhlander, B., 2010.

Temporal isotopic variations of dissolved silicon

in a pristine boreal river. Chemical Geology 271,

142-152.

Fisher, T.R., Harding, L.W., Stanley, D.W., Ward,

L.G., 1988. Phytoplankton, nutrients, and

turbidity in the Chesapeake, Delaware, and

Hudson estuaries. Estuarine, Coastal and Shelf

Science 27, 61-93.

Fisher, T.R., Peele, E.R., Ammerman, J.W., Harding,

L.W.J., 1992. Nutrient limitation of

phytoplankton in Chesapeake Bay. Marine

Ecology Progress Series 82, 51-63.

Georg, R.B., Reynolds, B.C., Frank, M., Halliday,

A.N., 2006a. Mechanisms controlling the silicon

isotopic compositions of river waters. Earth and

Planetary Science Letters 249, 290-306.

Georg, R.B., Reynolds, B.C., Frank, M., Halliday,

A.N., 2006b. New sample preparation techniques

for the determination of Si isotopic compositions

using MC-ICPMS. Chemical Geology 235, 95-

104.

Goldman, J.G., 1988. Spatial and temporal

discontinuities of biological processes in pelagic

surface waters in: Rothschild, B.J. (Ed.), Toward

a theory on biological-physical interactions in the

world ocean. Kluwer Academic Publishing,

Dordrecht, pp. 273-296.

Grasshoff, K., Erhardt, M., Kremling, K., 1983.

Methods of sea water analysis. Verlag Chemie,

Weinheim, Germany.

Hoefs, J. 2009. Stable isotope geochemistry, 6th ed.

Springer Berlin Heidelberg.

Howarth, R. W. 1988. Nutrient Limitation of Net

Primary Production in Marine Ecosystems.

Annual Review of Ecology and Systematics. 19,

89-110.

Hughes, H.J., Sondag, F., Cocquyt, C., Laraque, A.,

Pandi, A., André, L., Cardinal, D., 2011. Effect of

seasonal biogenic silica variations on dissolved

silicon fluxes and isotopic signatures in the Congo

River. Limnology and Oceanography 56, 551-561.

Humborg, C., Ittekkot, V., Cociasu, A., Bodungen,

B.v., 1997. Effect of Danube River dam on Black

Sea biogeochemistry and ecosystem structure.

Nature 386, 385-388.

Humborg, C., Pastuszak, M., Aigars, J., Siegmund,

H., Mörth, C.M., Ittekkot, V., 2006. Decreased

Silica Land-sea Fluxes through Damming in the

Baltic Sea Catchment-Significance of Particle

Trapping and Hydrological Alterations.

Biogeochemistry 77, 265-281.

Humborg, C., E. Smedberg, M. R. Medina, and C.-M.

Mörth. 2008. Changes in dissolved silicate loads

to the Baltic Sea -- The effects of lakes and

reservoirs. Journal of Marine Systems. 73, 223-

235.

IMWM [ed.]. 2004. Miesięczne Biuletyny

Państwowej Służby Hydrologiczno-

Meteorologicznej, (Monthly Bulletins of the

National Hydrological and Meteorological

Service, in Polish). Institute of Meteorology and

Water Management.

Leynaert, A., Tréguer, P., Lancelot, C., Rodier, M.,

2001. Silicon limitation of biogenic silica

production in the Equatorial Pacific. Deep Sea

Research Part I: Oceanographic Research Papers

48, 639-660.

Milligan, A.J., Varela, D.E., Brzezinski, M.A., Morel,

F.M.M., 2004. Dynamics of Silicon Metabolism

and Silicon Isotopic Discrimination in a Marine

Diatom as a Function of pCO2. Limnology and

Oceanography 49, 322-329.

Morley, D.W., Leng, M.J., Mackay, A.W., Sloane,

H.J., Rioual, P., Battarbee, R.W., 2004. Cleaning

of lake sediment samples for diatom oxygen

isotope analysis. Journal of Paleolimnology 31,

391-401.

Nelson, D.M., Treguer, P., Brzezinski, M.A.,

Leynaert, A., Queguiner, B., 1995. Production

and dissolution of biogenic silica in the ocean:

Revised global estimates, comparison with

regional data and relationship to biogenic

Silicon isotope enrichment in diatoms during nutrient-limited blooms in a eutrophied river system

- 9 -

sedimentation. Global Biogeochemical Cycle 9,

359-372.

Nelson, D.M., Dortch, Q., 1996. Silicic acid

depletion and silicon limitation in the plume of

the Mississippi River: evidence from kinetic

studies in spring and summer. Marine Ecology

Progress Series 136, 163-178.

Officer, C.B., Ryther, J.H., 1980. The possible

importance of silicon in marine eutrophication.

Marine Ecology Progress Series 3, 83-91.

Opfergelt, S., Cardinal, D., Henriet, C., Draye, X.,

André, L., Delvaux, B., 2006. Silicon Isotopic

Fractionation by Banana ( Musa spp.) Grown in a

Continuous Nutrient Flow Device. Plant & Soil

285, 333-345.

Pastuszak, M., Conley, D.J., Humborg, C., Witek, Z.,

Sitek, S., 2008. Silicon dynamics in the Oder

estuary, Baltic Sea. Journal of Marine Systems 73,

250-262.

Pastuszak, M., Stålnacke, P., Pawlikowski, K., Witek,

Z., 2011. Response of Polish rivers (Vistula, Oder)

to reduced pressure from point sources and

agriculture during the transition period (1988–

2008). Journal of Marine Systems 94, 157-173.

Pastuszak, M., Witek, Z., 2009a. Odpływ wód i

substancji biogenicznych Wisłą i Odrą z obszaru

Polski, in: Igras, J., Pastuszak, M. (Eds.), Udział

polskiego rolnictwa w emisji związków azotu i

fosforu do Bałtyku-Contribution of Polish

agriculture to emission of nitrogen and

phosphorus compounds to the Baltic Sea: . IUNG-

PIB, Puławy, Poland, pp. 271-306.

Pastuszak, M., Witek, Z., 2009b. Rola estuariów

Odry i Wisły w retencji ładunków azotu i fosforu

wprowadzanych do Bałtyku w odpływie

rzecznym, in: Igras, J., Pastuszak, M. (Eds.),

Udział polskiego rolnictwa w emisji związków

azotu i fosforu do Bałtyku-- Contribution of

Polish agriculture to emission of nitrogen and

phosphorus compounds to the Baltic Sea. IUNG-

PIB, Puławy, Poland, pp. 304-325.

Reynolds, B.C., Aggarwal, J., André, L., Baxter, D.,

Beucher, C., Brzezinski, M.A., Engström, E.,

Georg, R.B., Land, M., Leng, M.J., Opfergelt, S.,

Rodushkin, I., Sloane, H.J., van den Boorn,

S.H.J.M., Z., V.P., D., C., 2007. An inter-

laboratory comparison of Si isotope reference

materials. Journal of Analytical Atomic

Spectrometry 22, 561-568.

Schelske, C.L., Stoermer, E.F., 1971. Eutrophication,

Silica Depletion, and Predicted Changes in Algal

Quality in Lake Michigan. Science 173, 423-424.

Schelske, C.L., Stoermer, E.F., Conley, D.J., Robbins,

J.A., Glover, R.M., 1983. Early Eutrophication in

the Lower Great Lakes. Science 222, 320-322.

Struyf, E., Smis, A., Van Damme, S., Meire, P.,

Conley, D., 2009. The Global Biogeochemical

Silicon Cycle. SILICON 1, 207-213.

Sun, X., Andersson, P., Land, M., Humborg, C.,

Mörth, C.-M., 2010. Stable silicon isotope

analysis on nanomole quantities using MC-ICP-

MS with a hexapole gas-collision cell. Journal of

Analytical Atomic Spectrometry 25, 156-162.

Unesco. 1983. Chemical methods for use in marine

environmental monitoring. Manual and guides:

Intergovernmental Oceanographic Commission.

12, 1-53.

van den Boorn, S.H.J.M., van Bergen, M.J., Vroon,

P.Z., de Vries, S.T., Nijman, W., 2010. Silicon

isotope and trace element constraints on the origin

of ~3.5 Ga cherts: Implications for Early

Archaean marine environments. Geochimica et

Cosmochimica Acta 74, 1077-1103.

Varela, D.E., Pride, C.J., Brzezinski, M.A., 2004.

Biological fractionation of silicon isotopes in

Southern Ocean surface waters. Global

Biogeochem. Cycles 18, GB1047.

Wille, M., Sutton, J., Ellwood, M.J., Sambridge, M.,

Maher, W., Eggins, S., Kelly, M., 2010. Silicon

isotopic fractionation in marine sponges: A new

model for understanding silicon isotopic

variations in sponges. Earth and Planetary Science

Letters 292, 281-289.

Ziegler, K., Chadwick, O.A., Brzezinski, M.A., Kelly,

E.F., 2005. Natural variations of δ30

Si ratios

during progressive basalt weathering, Hawaiian

Islands. Geochimica et Cosmochimica Acta 69,

4597-4610.

Paper IV

1

Effect of diatom growth and dissolution on silicon isotope fractionation

in an estuarine system

Xiaole Suna1

, Martin Olofssonb, Per Andersson

c,

Catherine Legrandb, Christoph Humborg

d,e, Carl-Magnus Mörth

a,e

a Department of Geological Sciences, Stockholm University, SE-10691, Stockholm, Sweden

b Centre for Ecology and Evolution in Microbial model Systems (EEMiS), School of Natural Sciences, Linnæus

University, SE-39182 Kalmar, Sweden. c Laboratory for Isotope Geology, Swedish Museum of Natural History, SE-10405, Stockholm, Sweden

d Department of Applied Environmental Science, Stockholm University, SE-10691, Stockholm, Sweden

e Baltic Nest Institute, Stockholm Resilience Centre, Stockholm University, SE-10691 Stockholm, Sweden

Abstract

Si isotopes provide a powerful tool to reveal past and present patterns in diatom production. Most studies

have focused on Si fractionation factors during diatom growth in open ocean systems and have found

lower Si isotope values in diatom shells (biogenic silica). Recent findings indicate that even the

fractionation of Si isotopes during the physicochemical dissolution of diatom shells in the opposite

direction produces higher δ30

Si values in the remaining biogenic silica (BSi), allowing for the

interpretation of diatom production patterns over geological time scales. However, estuarine and coastal

primary production represents approximately 30-50% of global marine production, and there are hardly

any studies on Si isotope fractionation during either diatom growth or dissolution. In this study, Si isotope

fractionation during diatom growth and the dissolution of the frustule were measured. Two species of

diatoms from the Baltic Sea, one of the largest estuarine systems in the world, were selected for this study.

The results show that both species of diatoms during growth yields an identical Si isotope fractionation

factor of 0.99925 for 29

Si and 0.9984 for 30

Si. In contrast to findings from open ocean species, no Si

isotope fractionation during dissolution was observed even after 90% of the diatoms dissolved. Whether

there is isotope fractionation during dissolution or not will have profound implications for studies using Si

isotopes to interpret the Si cycle in marine and estuarine systems. We propose that the small size of the

diatoms living in estuarine systems with low salinity may explain the non-existence of Si isotope

fractionation during dissolution. Therefore, we suggest that Si isotopes are an instrumental variable

holding information about original environmental conditions of estuarine and even coastal systems.

Finally, we tested the Si isotope fractionation patterns gained from the lab experiments on a sediment core,

corroborating the observed dissolved Si (DSi) uptake rates in the above water column during diatom

growth.

1 Corresponding author xiaole.sun@geo.su.se

Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system

- 1 -

1 Introduction

Land-ocean fluxes from rivers contribute 80% of the

dissolved Si (DSi) load to marine systems (Treguer et al.,

1995); therefore, estuaries are a crucial link between land

and marine systems. One of the most important

characteristics of estuaries is that those systems are

highly dynamic, resulting in gradients in water turnover

times, salinity, nutrients and primary production.

Meanwhile, eutrophication, a major worldwide problem

caused by human activities, is concentrated around

estuaries and shallow coastal waters (Conley et al., 1993;

Fisher et al., 1988; Schelske and Stoermer, 1971). The

high input of nutrients derived from land can result in

high biological activity in estuaries, leading to the

removal of DSi from water and a decrease in the DSi flux

to open oceans. Hence, estuarine systems in general can

be regarded as a sink for soluble silica during low water

turnover and high primary productivity (Treguer et al.,

1995). Due to the complexity of these systems, their

biogeochemical Si cycle is difficult to study and to model

conceptually. To better assess DSi land-ocean fluxes, it is

essential to identify and quantify various biological

processes in estuaries controlling DSi; indeed, the

production, retention and dissolution of diatoms are the

most important regulators for DSi in these processes. The

uncertainty in production, retention and dissolution is

mainly attributed to our poor understanding of the

distribution and biogeochemical fate of BSi and DSi and

also heavily influenced by eutrophication (Conley et al.,

2008; Conley et al., 1993) and river damming (Humborg

et al., 2000; Humborg et al., 1997; Humborg et al., 2006).

Thus, considerable work is required to assess the

estuarine cycling of Si, in particular the role of siliceous

algae, such as diatoms, which are a dominating algae

species and play an essential role in the biogeochemical

cycles of Si (Buesseler, 1998; Nelson et al., 1995).

The recognition that diatoms preferentially take up 28

Si,

resulting in an isotope fractionation (δ-value) of

approximately -1.1‰ (De La Rocha et al., 1997), has

afforded the possibility to estimate past and current

diatom production in various systems because diatom

production quantitatively corresponds to δ30

Si values of

BSi. Currently, an increasing number of studies on

diatoms and Si isotopes have appeared regarding open

ocean and terrestrial systems. In oceans, δ30

Si values of

DSi range from +0.5‰ to +3.2‰, corresponding to δ30

Si

values of BSi in diatoms ranging from -0.3‰ to +2.6‰

(Beucher et al., 2008; Cardinal et al., 2005; De La Rocha

et al., 2000; Reynolds et al., 2006; Varela et al., 2004);

these values are slightly higher than the BSi of phytoliths

produced in terrestrial systems, which range from -1.7‰

to +2.5‰ (Douthitt, 1982; Ziegler et al., 2005). In river

waters, the measured δ30

Si values range from +0.4‰ to

+3.4‰ (De La Rocha et al., 2000; Ding et al., 2004; Ding

et al., 2011; Engström et al., 2010; Georg et al., 2007;

Hughes et al., 2011). DSi in a soil solution also shows

δ30

Si values ranging from -0.8‰ to +1.7‰ (Ziegler et al.,

2005), and a similar range of values is found in

groundwater, -0.2‰ to +1.3‰ (Georg et al., 2009;

Opfergelt et al., 2011). Based on all these measured δ30

Si

values, the Si isotope fractionation factor is indicated to

be -1.08‰ for freshwater (Alleman et al., 2005; Hughes

et al., 2011) and to be in the range of -0.6‰ to -2.2‰ for

in situ measurements of oceanic waters (Cardinal et al.,

2005; Cardinal et al., 2007; De La Rocha et al., 1997;

Varela et al., 2004). However, none of these studies has

been performed for estuarine and coastal areas, which are

responsible for 30-50% of global marine production.

Using Si isotopes to estimate the role of diatoms in

biogeochemical fluxes along the aquatic continuum from

land to the open ocean still requires further study,

especially in estuarine and coastal systems, and there

might be typical ranges of Si isotopes and fractionation

factors in the various realms. When interpreting Si

isotope signatures in diatoms and DSi, much of our

knowledge still comes from assumptions about Si isotope

fractionation factors (Reynolds et al., 2006; Sun et al.,

2011). Moreover, a recent study by Demarest et al. (2009)

shows a Si isotope fractionation value during dissolution

of -0.55‰, enriching residual BSi with heavier Si

isotopes if > 20% BSi is lost during dissolution. This

implies that if Si isotope fractionation is not taken into

account, diatom production may be significantly

overestimated, i.e. higher dissolution-induced δ30

Si

values of BSi are not equivalent to actual production in

water.

Therefore, to study Si isotope fractionation factors and to

evaluate the potential of using Si isotopes as

environmental traces from land to oceans, two estuarine

species of diatoms from the Baltic Sea, one of the largest

estuarine systems in the world, were selected for this

study. We systematically analysed the dynamics of Si

isotope values during diatom growth and dissolution in a

laboratory-controlled system. The resulting fractionation

factors were tested by a sedimentary record of BSi of the

most northern basin of the Baltic Sea, Bothnian Bay.

2 Materials and methods

2.1 Diatom cultures and experimental setup

Two species of diatoms, the solitary Thalassiosira

baltica (TBTV1) and the chain-forming Skeletonema

marinoi (SMTV1), were isolated from the Gulf of

Finland by Dr. Anke Kremp (Finnish Environment

Institute-SYKE, Helsinki, Finland), and cultured at the

School of Natural Sciences, Linnæus University, Sweden.

The stock cultures were grown in f/2 medium (Guillard,

1975) prepared from 1 µm filtered and autoclaved Baltic

seawater (salinity 7 psu). Diatoms were grown at 8-10 °C,

under a photon flux of 200-250 µmol photons m-2

s-1

and

a 16/8h light/dark cycle. To measure Si isotope values

Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system

- 2 -

during diatom growth, cells were collected in the

exponential phase by filtering the stock culture through

5-10 µm nylon nets. Cells were rinsed four times and

resuspended with silicate-free f/2 medium (f/2-Si) to

minimise the input of DSi from the stock cultures. Cells

were inoculated in duplicate into 5 L PET bottles in f/2-

Si Baltic seawater medium at chlorophyll a (Chl a) at

concentrations of 10 µg Chl a L-1

for TBTV1 and 7 µg

Chl a L-1

for SMTV1. The initial Baltic seawater used for

the inoculation of cultures contained approximately 20

µmol L-1

DSi. To ensure a reliable diatom biomass and

remaining DSi for Si isotopic analyses, additional DSi

was added to the culture medium to reach final DSi

concentrations above 60 µmol L-1

. The diatom cultures

were gently bubbled with air to mix the cells and to keep

them in suspension until DSi depletion, and incubated

under the same conditions as the stock culture. TBTV1

and SMTV1 cultures were monitored daily for 5 and 7

days for cell counts, Chl a, DSi concentrations and Si

isotopic analyses.

After DSi was depleted in the two cultures, the growth

experiment was terminated. To measure Si isotope values

during the dissolution of the diatom frustule, a fraction of

the SMTV1 culture (4 L) was distributed into two 2 L

polycarbonate bottles. The bottles were kept in the dark

at 22-24 °C for 44 days. The same set of samples used

during diatom growth was initially taken daily, but less

frequently after the fifth day until 90% of the diatom

frustules were dissolved. In addition, the O2

concentration was measured regularly during dissolution

to ensure that no anoxia occurred.

Duplicate aliquots (2 ml) from each replicate were fixed

with acidic Lugol’s solution, and cells were counted in an

inverted light microscope (Olympus BX50) using a

Palmer-Maloney counting chamber. Samples (2 x 5-15

ml) were filtered (Pall AE/E filters) and extracted with

96 % ethanol, overnight at room temperature for

chlorophyll a fluorometric analysis (Jespersen and

Christoffersen, 1987). To collect diatom cells, 150-200

ml of water samples was filtered daily (alternatively

twice a day if DSi decreased rapidly) through 0.2 µm

polyethersulphone membrane filters. One hundred

millilitres of the filtrate was immediately analysed for

DSi concentration according to Valderrama (1995). The

remaining filtrates (50 ml) were kept at 4 °C, and diatom

cells collected onto the filters were freeze-dried prior to

preparations for Si isotope analysis.

2.2 Sample preparation prior to Si isotope analysis

The method used for BSi digestion was modified from

Ragueneau et al. (2005). Polyethersulphone filters

containing 2-7 mg of freeze-dried diatoms were placed

into 15 ml polypropylene (PP) tubes. A 5 ml solution of

0.25 mol L-1

NaOH was added and the samples were

heated in a water bath for 1 hour at 90 °C. After cooling,

HCl was used to stop the digestion by neutralising pH.

Centrifugation was performed at 4200 rpm for 5 minutes.

The supernatant was transferred to another PP tube and

diluted to 10 ml solution for further treatment.

Dissolved Si in water samples was extracted by a

magnesium 2-step co-precipitation technique adopted

from Reynolds et al. (2006). Two hundred microlitres of

1 mol L-1

NaOH was added to 10 ml of the water samples.

Samples were shaken and left for 1 hour, and thereafter

centrifuged at 4200 rpm for 5 minutes. Another 100 µl of

1 mol L-1

NaOH was dropped into the separated

supernatant to precipitate more Mg(OH)2. After samples

were shaken and left for another hour, they were

centrifuged and the supernatant was removed. The

precipitates were then dissolved in 4 mol L-1

HCl. The

removed supernatant was analysed to make sure that the

Si removal was complete.

The purification of Si from the dissolved diatom solution

and water samples was performed using a BioRad

AG50W-X12 cation-exchange resin. Five hundred

microliters of the samples in 0.2 mol L-1

HCl was loaded

on the column with a 1 ml resin bed and eluted with 4 ml

Mille-Q water. Si having undergone no charges flowed

directly through the resin with Mille-Q water. Detailed

information regarding this step can be found in Sun et al.

(2011).

2.3 Si isotope analysis

The Si isotope compositions of both DSi and BSi were

analysed in the Laboratory for Isotope Geology at the

Swedish Museum of Natural History, Stockholm,

Sweden, using an IsoProbe MC-ICP-MS with a hexapole

collision cell. Generally, this instrument is capable of

precisely and accurately measuring δ29

Si (2 S.D., better

than 0.2‰), although measuring δ30

Si is not possible due

to the instrument’s low resolution (m/Δm~450). The

instrumental mass discrimination is corrected by

applying the standard-sample bracketing technique. The

sensitivity for 28

Si is ~12 V/ 36 µmol L-1

Si (1 mg L-1

),

with an uptake rate of ~70 µl min-1

via a Cetac® Aridus

desolvating nebuliser. The detailed measurement

procedure and the long-term reproducibility shown by

IRMM-18 and NBS 28 have been documented in Sun et

al. (2010).

The Si isotope ratio of each sample is calculated and

expressed in δ29

Si values (‰) relative to the standard,

NBS 28, and according to Eq.1, δ30

Si values are

calculated from the relationship δ30

Si = δ29

Si×1.96, which

assumes mass-dependent fractionation (Reynolds et al.,

2007). 29

28

sample29

29

28

standard

1 1000

Si

SiSi

Si

Si

(Eq.1)

2.4 Data reduction

(1)

Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system

- 3 -

A closed system usually means that a substrate is

continuously removed to form a product, but neither

undergoes any exchange with the outside system. During

the diatom experiment, we set up a closed system in 5 L

containers without replenishing DSi and BSi. To interpret

the Si isotope values of such a system, the Rayleigh

distillation equations are used. The relationship between

Si isotope values in products and substrates can be

expressed by the following equations:

29( 1)

29

0

1000

1000

substrateSif

Si

(Eq.2)

29

29

0

1000 1

1000 1

productSi f

Si f

(Eq.3)

where δ29

Sisubstrate, δ29

Siproduce and δ29

Si0 are Si isotope

values of the substrates, products and initial substrate,

respectively; α is the Si isotopic fractionation factor, and

f is the fraction of remaining substrates in water. By

rearranging Eq. 2, α can be solved as shown in Eq. 4:

29

29

0

1000ln( )

10001

ln

substrateSi

Si

f

(Eq.4)

Alternatively by rearranging Eq. 3, α can be solved as

shown in Eq. 5: 29

29

0

1000ln[1 *(1 )]

1000

ln

productSif

Si

f

(Eq.5)

3 Results

Table 1 and Fig. 1 and 2 summarise the BSi and DSi

concentrations and the Si isotope values of BSi and DSi

during the culture experiment. For TBTV1 and SMTV1

the Chl a concentrations increased exponentially, while

the DSi concentrations decreased during the initial the

DSi concentrations decreased during the initial growth

period (Fig.1). During exponential growth, the maximum

growth rate was 0.42 day-1

for TBTV1 and 1.08 day-1

for

SMTV1. Chl a levels reached 200 µg L-1

for TBTV1 and

530 µg L-1

for SMTV1 (Figure 1A-1 and 1B-1). The DSi

concentrations decreased progressively during growth

days and the experiment was terminated when the DSi

concentrations fell below 0.4 µmol L-1

(Fig. 1A-2 and

1B-2).

Fig.1 also shows that the two species of diatoms have

similar Si isotopic compositions and that the δ30

Si values

increased during diatom growth both in cells and in the

remaining DSi. During diatom growth, the Si isotope

values in the two species varied by more than 0.7‰ for

δ29

Si and 1.3‰ for δ 30

Si, ranging from +0.6‰ up to +1.3‰

for δ29

Si and from +1.2‰ up to +2.6‰ for δ30

Si,

respectively. The two species fractionate Si isotopes

during their growth, producing almost identical 29

α

fractionation factors of 0.9992 ± 0.0003 and 0.9993 ±

0.00038 (2σ), respectively (calculated by Eq. 5). A two-

sample t-test, assuming equal variance, shows that there

is no significant difference between TBTV1 and SMTV1

(p ≤ 0.061). The average value for 29

α over all diatom

sample measurements was 0.99925 ± 0.00034 (2σ),

corresponding to a difference in Si isotopic compositions

between DSi and diatoms up to +0.75‰ for δ29

Si. The

measured δ29

Si values in seawater samples increased

from 1.35‰ to 2.58‰; δ30

Si values are calculated by

multiplying δ29

Si by 1.96 (Reynolds et al., 2007), giving

values ranging from +2.64‰ to +5.07‰. The calculated

average value of 30

α is 0.9984 ± 0.0004 (2σ),

corresponding to a difference of +1.6‰ for δ30

Si, i.e. the

seawater is continually enriched with the heavier Si

isotopes.

Fig.1 Plot of Si isotope values following diatom growth curves as a function of time. A-1 and A-2 are plots of TBTV1 growth; B-

1 and B-2 are plots of SMTV1 growth. Error bars of Si isotope values are 2σ.

0

1

2

3

4

0

200

400

600

δ30S

i, ‰

Ch

l a

(µg L

-1)

Chl a (TBTV1)

δ³°Si of TBTV1

A-1

0

1

2

3

4

0

200

400

600δ

30S

i, ‰

Ch

l a

(µg L

-1)

Chl a (SMTV1)

δ³°Si of SMTV1

B-1

0

2

4

6

0

20

40

60

80

100

0 2 4 6 8 10

δ30S

i, ‰

DS

i co

nce

ntr

atio

n

(µm

ol

L-1

)

Days

DSi conc.

δ³°Si of DSi

(TBTV1)

A-2

0

2

4

6

0

20

40

60

80

100

0 2 4 6 8

δ3

0Si

, ‰

DSi

co

nce

ntr

atio

n

(µm

ol L

-1)

Days

DSi conc.

δ³°Si of DSi (SMTV1)

B-2

Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system

- 4 -

Table 1 Summary of Si concentrations and Si isotope values of DSi and BSi during SMTV1 and TBTV1 production and dissolution, as well as the fractionation factor (α)

calculated by Eq. 4 and 5. The 2σ of the calculated α is shown below the average values.

Days Chl a

(µg/L)

DSi conc.

(µM) f δ

29SiBSi 2σ δ

30SiBSi δ

29Si DSi 2SD δ

30SiDSi

29α by Eq.

5

30α by Eq.

5

SMTV1 production

0 6.8 64.8 1.00 +0.63 0.25 +1.23 +1.35 0.15 +2.64 N/A N/A

2 25.7 57.8 0.89 +0.69 0.15 +1.36 +1.52 0.13 +2.98 0.99930 0.99864

3 73.3 42.1 0.65 +0.80 0.14 +1.57 +1.56 0.13 +3.05 0.99932 0.99866

3.5 145.3 34.3 0.53 +0.86 0.24 +1.70 +1.81 0.10 +3.55 0.99932 0.99868

4 305.0 0.4 0.01 +1.32 0.09 +2.58 N/A N/A N/A 0.99942 0.99832

5 527.7 0.4 0.01 +1.29 0.10 +2.59 N/A N/A N/A 0.99900 0.99801

Average 0.9993 0.99847

2σ 0.00034 0.0004

SMTV1 dissolution

7 550.1 0.8 1.00 +1.29 0.18 +2.54 N/A N/A N/A N/A N/A

9 527.6 0.5 0.96 +1.30 0.22 +2.55 N/A N/A N/A N/A N/A

12 542.5 1.3 0.99 +1.30 0.10 +2.54 N/A N/A N/A N/A N/A

17 499.3 8.8 0.91 +1.29 0.09 +2.53 N/A N/A N/A N/A N/A

22 501.9 20.9 0.91 +1.34 0.19 +2.63 +1.10 0.13 +2.36 0.99977 0.99949

28 467.4 36.0 0.85 +1.27 0.12 +2.48 +1.14 0.25 +2.32 0.99977 0.99965

32 448.1 47.6 0.81 +1.32 0.07 +2.58 +1.23 0.15 +2.40 0.99986 0.99973

43 298.3 48.5 0.54 +1.37 0.14 +2.69 +1.35 0.14 +2.56 1.00000 1.00000

49 270.6 58.6 0.49 +1.36 0.18 +2.66 +1.32 0.10 +2.46 0.99996 0.99992

Average 0.99990 0.99974

2σ 0.00020 0.0004

TBTV1 production

0 11.9 81.8 1.00 +0.61 0.07 +1.20 +1.35 0.15 +2.64 N/A N/A

1 15.3 75.2 0.92 +0.64 0.19 +1.25 +1.43 0.11 +2.81 0.99926 0.99855

2 26.2 67.2 0.82 +0.67 0.11 +1.31 +1.45 0.16 +2.84 0.99925 0.99853

3 42.2 58.7 0.72 +0.65 0.14 +1.27 +1.55 0.14 +3.03 0.99918 0.99838

3.5 43.1 54.3 0.66 +0.70 0.13 +1.36 +1.74 0.09 +3.40 0.99920 0.99842

4 65.0 46.8 0.57 +0.84 0.13 +1.65 +1.66 0.13 +3.25 0.99932 0.99867

4.5 69.2 42.3 0.52 +0.74 0.17 +1.46 +1.80 0.07 +3.52 0.99914 0.99833

5 109.1 26.4 0.32 +0.79 0.19 +1.54 +2.58 0.13 +5.07 0.99896 0.99797

5.5 118.4 19.6 0.24 +1.01 0.12 +1.99 +2.46 0.13 +4.83 0.99925 0.99855

6 133.1 2.9 0.04 +1.24 0.24 +2.43 N/A N/A N/A 0.99920 0.99826

Average 0.9992 0.99842 2σ 0.0002 0.0004

Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system

- 5 -

Fig. 2 Plot of Si isotope values following SMTV1 dissolution

curve as a function of time. A and B are the variation of Chl a

and the DSi concentration, respectively. Error bars of Si

isotope values are 2σ.

During the dissolution experiment, the Chl a

concentrations of SMTV1 gradually decreased to 267 µg

L-1

after the growth period (Fig. 2A), while the DSi

concentration increased steadily for 20 days to 50 µmol

L-1

, and remained constant (50-60 µmol L-1

) for another

20 days (Fig. 2B), indicating the dissolution of diatom

frustules. During the first 25 days in the dark, the

siliceous frustules were linearly dissolved (r2=0.9) and 75%

of the silicate was released back to the medium.

Thereafter, during the last 10 days, the dissolution

released 90% BSi back in to the medium. The Si isotope

values of SMTV1 and DSi appeared to remain constant

during dissolution. Calculation using Eq. 5 gives the Si

isotope fractionation factor during dissolution for 29

Si

and 30

Si as 0.9999 ± 0.0002 and 0.99974 ± 0.0004 (2σ).

4 Discussion

Our findings suggest that i) diatoms from estuarine

systems show identical Si isotope fractionation factors of

0.99925 for 29

Si and 0.9984 for 30

Si, which are very

similar to those of oceanic species; ii) the fractionation

factors during diatom production closely follow the

theoretical values calculated by the Rayleigh distillation

approach; iii) when applying these fractionation factors

to a sediment record, we can reproduce the observed

water column data on the fraction of remaining DSi

during the growth period; iv) in contrast to previous

studies on diatom dissolution, the decrease in BSi in the

lab experiments is not associated with Si isotope

fractionation (at least up to 90% dissolution), making Si

isotopes a powerful tracer for past diatom blooms in

estuarine systems.

4.1 Mechanism of Si isotope fractionation

The calculated Si isotope fractionation factors during the

Baltic Sea diatom’s growth, 29

α and 30

α, were calculated

to be 0.99925 ± 0.00034 (2σ) and 0.9984 ± 0.00035 (2σ),

respectively, which are in the same range as values

previously reported for other cultured marine diatoms:

Thalassiosira weissflogii and Skeletonema costatum (De

La Rocha et al., 1997). Our values are also consistent

with the value of 0.99944 for 29

α found in Lake

Tanganyika (Alleman et al., 2005) and 0.9981-0.9989 for 30

α reported in Antarctic surface water (Varela et al.,

2004).

The exact mechanism of Si isotope fractionation during

diatom growth is difficult to explain. A plot of δ30

Si

values as a function of f values (Fig. 3) shows a very

good agreement between the experimental results and the

Rayleigh model expressed by Eq. 4 and 5. This behaviour

is typical for a kinetic isotope effect in a closed system,

i.e., the Si isotope values vary with f values such that the

remaining DSi is enriched in the heavier Si isotopes. The

fact that diatoms are an effective ‘trap’ for keeping Si in

the water column without exchange with the remaining

DSi does not indicate an isotope equilibrium effect in this

study.

BSi dissolution commences when dissociated hydroxyls

from water weaken and eventually break the adjacent Si-

O bond (Dove and Crerar, 1990). The open linkage binds

with hydroxyl to form Si-OH. This process is repeated

until all Si-O bonds are broken, and the surface then

assumes the form Si(OH)4. This silica-water interaction

tends to be slow and is influenced by many different

factors, but it is commonly believed that temperature, pH,

the presence of solution species, particle size, etc., can

have a great impact on the dissolution of BSi (Fournier et

al., 1982; Iler, 1979; Loucaides et al., 2008). The

experiments conducted in our study and those in the

study by Demarest et al. (2009) were laboratory

controlled, and the variations in temperature are similar;

thus, they are unlikely to be the reason for the difference

in the observed Si isotope fractionation between these

two studies. A possible reason could be that cations

present in the solution are attracted to the negatively

charged surface sites of BSi by electrostatic forces. In

Baltic Sea water with much lower salinity (7 PSU

measured in this experiment) than that in the Southern

Ocean, the surface of BSi is less ‘saturated’ than BSi in

ocean water. This allows hydroxyls to easily approach

the xSi-O bonds on the surface and break them without

biasing Si isotopes. Another possible explanation is that

the effective surface area increases with a decrease in

particle size (e.g., Knauss and Wolery, 1988). Because

0

1

2

3

4

0

200

400

600

δ30S

i, ‰

Ch

l a

(µg L

-1)

Chl a (SMTV1)

δ³°Si of SMTV1

A

0

2

4

6

0

50

100

8 18 28 38 48

δ30S

i, ‰

DS

i co

nce

ntr

atio

n

(µm

ol

L-1

)

Days

DSi conc.

δ³°Si of DSi (SMTV1)

B

Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system

- 6 -

Fig. 3 Plot of Si isotope values of diatoms and DSi following diatom growth as a function of f values (fraction of remaining DSi in

water). Lines are expectations for Rayleigh fractionation in a closed system. Error bars of Si isotope values are 2σ.

the diatom size in low-salinity water with a shallower

mixed layer depth is normally smaller than that of

diatoms growing in ocean water (Litchman et al., 2009),

BSi dissolution in the Baltic Sea could be faster and more

effective, which would make the different Si isotopes

subject to simultaneous dissolution.

Although we try to explain the lack of isotope

fractionation during BSi dissolution, we cannot fully rule

out possible Si isotope fractionation during BSi

dissolution. Perhaps this fractionation is too small to be

detected by the MC-ICP-MS employed in this study. One

limitation is that only δ29

Si can be measured; the mass

difference between 28

Si and 29

Si is harder to detect than

the difference between 28

Si and 30

Si considering the

current analytical error. However, this is unlikely in this

case because the deviation between the calculated δ30

Si

values of remaining BSi after 90% dissolution is still

smaller than the documented analytical errors for δ30

Si, indicating an undetectable Si isotope fractionation

between the δ30

Si values of remaining BSi.

4.2 Applications of the Si isotope fractionation factor

It is well known and has been shown that Si isotopes

have the potential to reveal both past and present patterns

of DSi utilisation mainly by diatoms, which can

contribute to both modern and paleoceanographic studies.

However, Deramest et al. (2009) suggest that Si isotope

fractionation during diatom dissolution may complicate

this Si potential of revealing DSi utilisations and cause

discrepancies between the measured δ30

Si of BSi and the

real oceanic conditions inhabited by diatoms. In contrast,

our observation of unfractionated diatom dissolution

implies that diatoms can still possibly record Si isotope

information generated under original environmental

conditions. We therefore suggest that δ30

Si isotopes of

diatoms smaller than oceanic species and alternatively

growing in low-salinity can be reliably used as a proxy

for estuarine and even coastal studies. In our study, we

tested this notion by applying our measured Si isotope

fractionation factor to our measured δ30

Si values

preserved in BSi derived from Bothnian Bay, the most

northern part of the Baltic Sea.

The Baltic Sea, located in Northern Europe, has a surface

area of 377 000 km2 and a water volume of 21 000 km

3.

As one of the largest estuarine system in the world, it

consists of several basins, from north to south: Bothnian

Bay, Bothnian Sea, Gulf of Finland, Baltic Proper, Gulf

of Riga, Danish Sounds and Kattegat, and finally

connects to the North Sea. The shallow sills between the

North Sea and the Baltic Sea restrict the water exchange

leading to a long water residence time of approximately

20 years (Papush et al., 2009). The residence time of DSi

is 11.2 years for the whole Baltic Sea (Wulff and

Stigebrandt, 1989), which makes the Baltic Sea resemble

a closed system comparable to the experimental set up of

our lab experiments.

There is a biological control on Si isotope signatures in

the Baltic Sea water due to the large proportion of

diatoms in phytoplankton (Klais et al., 2011). Given an 29

α of 0.99925 and the calculated 30

α of 0.9984 for

diatom production, it is possible to address the potential

range of variations in the Si isotope values in the Baltic

Sea. Fig. 3 shows the behaviour of diatom production

following the Rayleigh model and illustrates the possible

Si isotope values in remaining DSi and diatoms

determined by the experiments performed in this study.

Using a δ30

Si value of +1.4‰ in rivers draining into

Bothnian Bay (Sun et al., 2011) as δ30

Si0, an 30

α of 0.9984

and approximately 30% depletion of DSi (Danielsson et

al., 2008), i.e., f = 0.7, +2.0‰ of δ30

Si of remaining DSi

is calculated by Eq. 2. Likewise, +0.15‰ of δ30

Siproduct in

accumulated BSi is calculated by Eq.3. Monte Carlo

analysis is adopted to estimate the uncertainty of this

0

2

4

6

8

00.20.40.60.81

δ 3

0S

i, ‰

f, fraction of remaining DSi in water

δ³°Si of TBTV1

δ³°Si of DSi

DSi (Rayleigh)

BSi (Rayleigh)

A

0

2

4

6

8

00.20.40.60.81

δ30S

i, ‰

f, fraction of remaining DSi in water

δ³°Si of SMTV1

δ³°Si of DSi

DSi (Rayleigh)

BSi (Rayleigh)

B

Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system

- 7 -

Table 2 Potential Si isotope values in the major Baltic Sea basins calculated by measured α of 0.9984 ± 0.00035 (2σ)

Basins Average depletion

of DSi (%)1

f

values δ

30Si of input DSi

δ30

Si of accumulated

diatom (‰) ± 2σ

δ30

Si of remaining

DSi (‰) ± 2σ

Bothnian Bay 30% 0.70 +1.4‰2 +0.15 ± 0.30 +1.95 ± 0.13

Bothnian Sea 40% 0.60 +1.4‰ (assumed) +0.21 ± 0.28 +2.20 ± 0.18

Gulf of Finland 62% 0.38 +1.6‰ (assumed) +0.70 ± 0.21 +3.10 ± 0.35

Gulf of Riga 73% 0.27 +1.6‰ (assumed) +0.85 ± 0.17 +3.64 ± 0.40

Baltic Proper 50% 0.50 +1.6‰ (measured) +0.53 ± 0.25 +2.67 ± 0.25

Kattegat 67% 0.33 +1.6‰ (assumed) +0.75 ± 0.20 +3.32 ± 0.38 1 Calculated by interannual variations in mean DSi concentrations between 1970 and 2000 (Danielsson et al. 2008)

2 Sun et al. (2011)

calculation derived from the measured Si isotope

fractionation factor. Five thousand 30

α are randomly

generated assuming a normal distribution based on the

average 30

α of 0.9984 and its 2σ of 0.00035, giving an

average δ30

Si value of accumulated BSi, +0.15‰ ± 0.30‰

(2σ). This is comparable with our previously reported

δ30Si value of +0.40‰ ± 0.20‰ (2σ) in sedimentary BSi

(Sun et al. 2011) without statistically significant

differences between these values. It also implies that

there is no measurable dissolution-induced Si isotope

fractionation, i.e., BSi preserved in sediments still

preserves information concerning the initial water

conditions.

We further suggest possible Si isotope patterns in the

various basins of the Baltic Sea using the measured Si

isotope fractionation factor to be tested in future studies.

The Baltic Sea shows a range of DSi uptake rates

expressed as the relative amount of DSi depleted (f

values) after the growth period that vary between years

(Wasmund et al. 1997) but also shows a typical range

between the various marine basins. Table 2 shows an

estimate of Si isotope values of diatoms and DSi in

basins of the Baltic Sea based on measurements and other

estimates, not taking into account dissolution effects. The

f values of different basins are derived from average DSi

concentrations measured before and after diatom blooms

between 1970 and 2000 (Danielsson et al., 2008). The

δ30

Si0 value of Baltic Proper before diatom blooms was

measured to be +1.6‰ in this study, resulting in δ30

Si

values of +0.53‰ ± 0.25‰ (2σ) and +2.67‰ ± 0.25‰

(2σ) for δ30

Si in accumulated diatoms and remaining DSi,

respectively. The values 2σ is derived from 5000 values

of δ30Si calculated by randomly generating 5000 30α as

mentioned above. In particular, the δ30

Si values of DSi in

the Gulf of Finland and Gulf of Riga possibly increase to

approximately +3.1‰ ± 0.35‰ (2σ) and +3.64‰ ± 0.40‰

(2σ), respectively. Generally, the Baltic Sea seems to

exhibit increasing δ30

Si values, both in diatoms and in

DSi from north to south caused by increased DSi

consumption during diatom production over the past

several decades. The highest expected Si isotope values

of +0.85‰ ± 0.17‰ of BSi are obtained for the Gulf of

Riga, indicating ongoing DSi depletion (Olli et al., 2008)

mainly caused by eutrophication in its catchments. This

is also a known problem occurring throughout the entire

Baltic Sea, especially in the southern part over the past

several decades. However, our knowledge about DSi and

BSi in the Baltic Sea and in its catchments prior to the

1960s is very limited due to the lack of historical

observations. Therefore, Si isotope values in sedimentary

BSi could be a complementary tool for reconstructing

past Si dynamics by transforming the observed δ30

Si

values of sedimentary BSi to DSi utilisation in water.

5 Conclusion

This study shows that diatom production yields an

identical non-species-dependent Si isotope fractionation

factor of 0.99925 for 29

Si and 0.9984 for 30

Si in an

estuarine system. In contrast to the results of another

study, the Si isotope fractionation during dissolution of

diatom frustules is not observed, possibly due to the

lower salinity of the surrounding water and the smaller

size of the diatoms. This indicates that δ30

Si signatures

preserved in BSi retain direct information about original

environmental conditions. These findings may contribute

to a better understanding of the biogeochemical cycle of

Si in estuarine and coastal environments and hold

implications for marine studies using Si isotopes as a

proxy to estimate the historical development and

magnitude of recent and past diatom blooms.

Acknowledgements

We are grateful to Mireia Bertos Fortis for assistance in

the laboratory. Funding for diatom cultivation was

provided by the Swedish Research Council Formas

through the programme ECOCHANGE “Ecosystem

dynamics in the Baltic Sea in a changing climate

perspective”. This is a contribution to the LnU Centre for

Ecology and Evolution using Microbial Models Systems

(EEMiS). We are also grateful for the funding from the

Swedish Research Council for sample preparation and Si

isotope analysis (VR-2007-4763).

Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system

- 8 -

References

Alleman, L.Y., Cardinal, D., Cocquyt, C., Plisnier, P.-D.,

Descy, J.-P., Kimirei, I., Sinyinza, D., André, L., 2005.

Silicon Isotopic Fractionation in Lake Tanganyika and

Its Main Tributaries. J. Gt. Lakes Res. 31, 509-519.

Beucher, C.P., Brzezinski, M.A., Jones, J.L., 2008.

Sources and biological fractionation of Silicon

isotopes in the Eastern Equatorial Pacific. Geochim.

Cosmochim. Acta 72, 3063-3073.

Buesseler, K.O., 1998. The decoupling of production and

particulate export in the surface ocean. Global

Biogeochem. Cycles 12, 297-310.

Cardinal, D., Alleman, L.Y., Dehairs, F., Savoye, N.,

Trull, T.W., André, L., 2005. Relevance of silicon

isotopes to Si-nutrient utilization and Si-source

assessment in Antarctic waters. Global Biogeochem.

Cycles 19, GB2007.

Cardinal, D., Savoye, N., Trull, T.W., Dehairs, F.,

Kopczynska, E.E., Fripiat, F., Tison, J.-L., André, L.,

2007. Silicon isotopes in spring Southern Ocean

diatoms: Large zonal changes despite homogeneity

among size fractions. Mar. Chem. 106, 46-62.

Conley, D.J., Humborg, C., Smedberg, E., Rahm, L.,

Papush, L., Danielsson, Å., Clarke, A., Pastuszak, M.,

Aigars, J., Ciuffa, D., Mörth, C.-M., 2008. Past,

present and future state of the biogeochemical Si

cycle in the Baltic Sea. J. Mar. Sys. 73, 338-346.

Conley, D.J., Schelske, C.L., Stoermer, E.F., 1993.

Modification of the biogeochemical cycle of silica

with eutrophication. Mar. Ecol. Prog. Ser. 101, 179-

192.

Danielsson, Papush, L., Rahm, L., 2008. Alterations in

nutrient limitations -- Scenarios of a changing Baltic

Sea. J. Mar. Sys. 73, 263-283.

De La Rocha, C.L., Brzezinski, M.A., DeNiro, M.J.,

1997. Fractionation of silicon isotopes by marine

diatoms during biogenic silica formation. Geochim.

Cosmochim. Acta 61, 5051-5056.

De La Rocha, C.L., Brzezinski, M.A., DeNiro, M.J.,

2000. A first look at the distribution of the stable

isotopes of silicon in natural waters. Geochim.

Cosmochim. Acta 64, 2467-2477.

Demarest, M.S., Brzezinski, M.A., Beucher, C.P., 2009.

Fractionation of silicon isotopes during biogenic silica

dissolution. Geochim. Cosmochim. Acta 73, 5572-

5583.

Ding, T., Wan, D., Wang, C., Zhang, F., 2004. Silicon

isotope compositions of dissolved silicon and

suspended matter in the Yangtze River, China.

Geochim. Cosmochim. Acta 68, 205-216.

Ding, T.P., Gao, J.F., Tian, S.H., Wang, H.B., Li, M.,

2011. Silicon isotopic composition of dissolved

silicon and suspended particulate matter in the Yellow

River, China, with implications for the global silicon

cycle. Geochim. Cosmochim. Acta 75, 6672-6689.

Douthitt, C.B., 1982. The geochemistry of the stable

isotopes of silicon. Geochim. Cosmochim. Acta 46,

1449-1458.

Dove, P.M., Crerar, D.A., 1990. Kinetics of quartz

dissolution in electrolyte solutions using a

hydrothermal mixed flow reactor. Geochim.

Cosmochim. Acta 54, 955-969.

Engström, E., Rodushkin, I., Ingri, J., Baxter, D.C., Ecke,

F., Österlund, H., Öhlander, B., 2010. Temporal

isotopic variations of dissolved silicon in a pristine

boreal river. Chem. Geol. 271, 142-152.

Fisher, T.R., Harding, L.W., Stanley, D.W., Ward, L.G.,

1988. Phytoplankton, nutrients, and turbidity in the

Chesapeake, Delaware, and Hudson estuaries. Estuar.

Coast. Shelf Sci. 27, 61-93.

Fournier, R.O., Rosenbauer, R.J., Bischoff, J.L., 1982.

The solubility of quartz in aqueous sodium chloride

solution at 350°C and 180 to 500 bars. Geochim.

Cosmochim. Acta 46, 1975-1978.

Georg, R.B., Reynolds, B.C., West, A.J., Burton, K.W.,

Halliday, A.N., 2007. Silicon isotope variations

accompanying basalt weathering in Iceland. Earth

Planet. Sci. Lett. 261, 476-490.

Georg, R.B., West, A.J., Basu, A.R., Halliday, A.N.,

2009. Silicon fluxes and isotope composition of direct

groundwater discharge into the Bay of Bengal and the

effect on the global ocean silicon isotope budget.

Earth Planet. Sci. Lett. 283, 67-74.

Guillard, R.R.L., 1975. Culture of phytoplankton for

feeding marine invertebrates, in: M.H, S.W.L.a.C.

(Ed.), Culture of Marine Invertebrate Animals.

Plenum Press, New York, pp. 26-60.

Hughes, H.J., Sondag, F., Cocquyt, C., Laraque, A.,

Pandi, A., André, L., Cardinal, D., 2011. Effect of

seasonal biogenic silica variations on dissolved silicon

fluxes and isotopic signatures in the Congo River.

Limnol. Oceanogr. 56, 551-561.

Humborg, C., Conley, D.J., Rahm, L., Wulff, F., Cociasu,

A., Ittekkot, V., 2000. Silicon Retention in River

Basins: Far-reaching Effects on Biogeochemistry and

Aquatic Food Webs in Coastal Marine Environments.

Ambio 29, 45-50.

Humborg, C., Ittekkot, V., Cociasu, A., Bodungen, B.v.,

1997. Effect of Danube River dam on Black Sea

biogeochemistry and ecosystem structure. Nature 386,

385-388.

Humborg, C., Pastuszak, M., Aigars, J., Siegmund, H.,

Mörth, C.M., Ittekkot, V., 2006. Decreased Silica

Land-sea Fluxes through Damming in the Baltic Sea

Catchment-Significance of Particle Trapping and

Hydrological Alterations. Biogeochemistry 77, 265-

281.

Iler, R.K., 1979. The chemistry of silica, solubility,

polymerization, colloid and surface properties and

biochemistry. John Wiley & Sons, New York.

Klais, R., Tamminen, T., Kremp, A., Spilling, K., Olli, K.,

2011. Decadal-scale changes of dinoflagellates and

Effect of diatom growth and dissolution on silicon isotope fractionation in an estuarine system

- 9 -

diatoms in the anomalous Baltic Sea spring bloom.

PLoS ONE 6, e21567. DOI: 10.1371/ journal.pone.

0021567

Litchman, E., Klausmeier, C.A., Yoshiyama, K., 2009.

Contrasting size evolution in marine and freshwater

diatoms. P. Natl. Acad. Sci. USA 106, 2665-2670.

Loucaides, S., Van Cappellen, P., Behrends, T., 2008.

Dissolution of biogenic silica from land to ocean:

Role of salinity and pH. Limnol. Oceanogr. 53, 1614-

1621.

Nelson, D.M., Treguer, P., Brzezinski, M.A., Leynaert,

A., Queguiner, B., 1995. Production and dissolution

of biogenic silica in the ocean: Revised global

estimates, comparison with regional data and

relationship to biogenic sedimentation. Global

Biogeochem. Cycles 9, 359-372.

Olli, K., Clarke, A., Danielsson, Å., Aigars, J., Conley,

D.J., Tamminen, T., 2008. Diatom stratigraphy and

long-term dissolved silica concentrations in the Baltic

Sea. J. Mar. Sys 73, 284-299.

Opfergelt, S., Eiriksdottir, E.S., Burton, K.W., Einarsson,

A., Siebert, C., Gislason, S.R., Halliday, A.N., 2011.

Quantifying the impact of freshwater diatom

productivity on silicon isotopes and silicon fluxes:

Lake Myvatn, Iceland. Earth Planet. Sci. Lett. 305,

73-82.

Papush, L., Danielsson, Å., Rahm, L., 2009. Dissolved

silica budget for the Baltic Sea. J. Sea Res. 62, 31-41.

Ragueneau, O., Savoye, N., Del Amo, Y., Cotten, J.,

Tardiveau, B., Leynaert, A., 2005. A new method for

the measurement of biogenic silica in suspended

matter of coastal waters: using Si:Al ratios to correct

for the mineral interference. Cont. Shelf Res. 25, 697-

710.

Reynolds, B.C., Aggarwal, J., André, L., Baxter, D.,

Beucher, C., Brzezinski, M.A., Engström, E., Georg,

R.B., Land, M., Leng, M.J., Opfergelt, S., Rodushkin,

I., Sloane, H.J., van den Boorn, S.H.J.M., Z., V.P., D.,

C., 2007. An inter-laboratory comparison of Si

isotope reference materials. J. Anal. At. Spectrom 22,

561-568.

Reynolds, B.C., Frank, M., Halliday, A.N., 2006. Silicon

isotope fractionation during nutrient utilization in the

North Pacific. Earth Planet. Sci. Lett. 244, 431-443.

Schelske, C.L., Stoermer, E.F., 1971. Eutrophication,

Silica Depletion, and Predicted Changes in Algal

Quality in Lake Michigan. Science 173, 423-424.

Sun, X., Andersson, P., Humborg, C., Gustafsson, B.,

Conley, D.J., Crill, P., Mörth, C.M., 2011. Climate

dependent diatom production is preserved in biogenic

Si isotope signatures. Biogeosciences 8, 3491-3499.

Sun, X., Andersson, P., Land, M., Humborg, C., Mörth,

C.-M., 2010. Stable silicon isotope analysis on

nanomole quantities using MC-ICP-MS with a

hexapole gas-collision cell. J. Anal. At. Spectrom. 25,

156-162.

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 Reestimate.

Science 268, 375-379.

Varela, D.E., Pride, C.J., Brzezinski, M.A., 2004.

Biological fractionation of silicon isotopes in

Southern Ocean surface waters. Global Biogeochem.

Cycles 18, GB1047.

Wulff, F., Stigebrandt, A., 1989. A time-dependent

budget model for nutrients in the Baltic Sea. Global

Biogeochem. Cycles 3, 63-78.

Ziegler, K., Chadwick, O.A., Brzezinski, M.A., Kelly,

E.F., 2005. Natural variations of δ³oSi ratios during

progressive basalt weathering, Hawaiian Islands.

Geochim. Cosmochim. Acta 69, 4597-4610.

MEDDELANDEN från

STOCKHOLMS UNIVERSITETS INSTITUTION för

GEOLOGISKA VETENSKAPER No. 351*

No. 338 SUNDVALL, R., Water as a trace component in mantle pyroxene: quantifying diffusion, storage capacity and variation with geological environment. Stockholm, 2010.

No. 339 PETTERSSON, C-H., The tectonic evolution of northwest Svalbard. Stockholm, 2010.

No. 340 SVAHNBERG, H., Deformation behaviour and chemical signatures of anorthosites:: Examples from southern West Greenland and south-central Sweden. Stockholm, 2010.

No. 341 ANDERSON, J., JAKOBSSON, M. and OSO 0910 SCIENTIFIC PARTY. Oden Southern Ocean 0910 OSO 0910: Cruise report.

No. 342 BORTHWICK, V., Fundamentals of substructure dynamics: In-situ experiments and numerical simulation. Stockholm, 2011.

No. 343 SILJESTRÖM, S., Single fluid inclusion analysis using ToF-SIMS. Implications for ancient Earth biodiversity and paleoenvironmental studies. Stockholm, 2011.

No. 344 HELL, B., Mapping bathymetry: From measurement to applications. Stockholm, 2011.

No. 345 HANSLIK, D., Late Quaternary Biostratigraphy and Paleoceanography of the central Arctic Ocean. Stockholm, 2011.

No. 346 NEUBECK, A., Low temperature olivine alteration; an experimental approach. Stockholm, 2012.

No. 347 ANDERSSON, R. Lipid biomarkers and other geochemical indicators in paleoenvironmental studies of two Arctic systems: a Russian permafrost peatland and marine sediments from the Lomonosov Ridge. Stockholm, 2012.

No. 348 Duc, N., Abiotic and biotic methane dynamics in relation to the origin of life. Stockholm, 2012.

No. 349 Dalsätt, J., Fossil birds: Contributions to the understanding of avian evolution. Stockholm, 2012.

No. 350 Hammarlund, E. U., Ocean chemistry and the evolution of multicellularity. Stockholm, 2012.

*Formerly/ 1990-2009: MEDDELANDEN från STOCKHOLMS UNIVERSITETS

INSTITUTION för GEOLOGI och GEOKEMI (ISSN 1101-1599)