Formation, uptake and bioaccumulation of methylmercury in coastal seas – a Baltic Sea case study

Aleksandra Skrobonja

Department of Chemistry Umeå 2019 This work is protected by the Swedish Copyright Legislation (Act 1960:729) Dissertation for PhD ISBN: 978-91-7855-177-4 Cover: Sampling cruise, incubation, and mesocosm experiment by Aleksandra Skrobonja Electronic version available at http://umu.diva-portal.org/ Printed by: The service centre in KBC, Umeå University, Sweden 2019

Mojim najdražima. Hvala što verujete u mene.

Table of Contents Abstract...... ii List of publications ...... iii Enkel sammanfattning på svenska ...... v 1. Introduction ...... 1 1.1. Mercury as a pollutant ...... 1 1.2. Formation of MeHg ...... 2 1.3. Uptake of MeHg ...... 3 1.4. Bioaccumulation of MeHg ...... 4 1.5. Aims of the thesis...... 6 2. Materials and Methods...... 7 2.1. Study sites ...... 7 2.2. Experimental approaches ...... 8 2.2.1. Field sampling ...... 8 2.2.2. Determination of total Hg and MeHg in aqueous samples and biota ...... 8 2.2.3. Methylation and demethylation rate constants ...... 9 2.2.4. DOC, humic matter content and ancillary data ...... 10 2.2.5. Microalgal cultivation ...... 10 2.2.6. Mesocosm experiment ...... 11 3. Results and discussion ...... 14 3.1. Organic matter input impacts MeHg formation and cycling (Paper I) ...... 15 3.2. The role of water column redoxclines on MeHg formation and cycling (Paper II) 18 3.3. Uptake kinetics of MeHg in a freshwater alga exposed to MeHg-thiol complexes (Paper III) ...... 21 3.4. Multiple impacts of humic-rich dissolved organic carbon on methylmercury accumulation in heterotrophic pelagic food webs (Paper IV) ...... 27 3.5. Future research remarks ...... 33 4. Conclusions ...... 35 5. Acknowledgements ...... 37 6. Literature ...... 39

i

Abstract

Methylmercury (MeHg) is a potent neurotoxin which can bioaccumulate to harmful levels in aquatic food webs. Methylmercury formation is a predominantly biotic process which involves phylogenically diverse microorganisms (e.g. iron- or sulfate-reducing bacteria). The formation of MeHg is related to the presence of organic matter (OM) which contains substrates essential for methylating microbes and reduced sulfur ligands (thiols, RSH) that form strong bonds with inorganic mercury (HgII) and affect its bioavailability. In aquatic systems, MeHg is bio-concentrated from the water column to the base of the food web and this step is crucial for MeHg levels found at higher trophic levels. Trophic transfer processes of MeHg in the food web are also of great importance. Discharge of OM in coastal areas affects light conditions needed for phytoplankton growth, and promotes heterotrophy, i.e. bacteria production. This may lead to a shift from the phytoplankton- based to the longer bacteria-based (microbial loop) food web and influence the amount of bioaccumulated MeHg in higher trophic levels. Methylmercury levels in predatory biota is thus affected by the bioavailability of HgII for methylation (studied in Paper I & II), MeHg speciation in the water column, crucial for MeHg incorporation at the base of the food web (Paper III), and the structure of the pelagic food web (Paper IV). In this thesis, it is shown that OM can act as a predictor of dissolved MeHg levels in estuarine and coastal systems. It impacts MeHg levels both by affecting HgII bioavailability (through Hg complexation with humic matter) and the activity of methylating microbes (providing metabolic electron donors) (Paper I). Moreover, elevated concentrations of particulate and dissolved HgII and MeHg, are associated with the presence of pelagic redoxclines in coastal seas. The redoxcline affects HgII speciation in the water column and its bioavailability for methylation (Paper II). It is further shown that the molecular structure of ligands in MeHg complexes affects the kinetics of MeHg uptake in phytoplankton. Rate constants for association of MeHg to the cell surface of a green algae were higher in treatments containing smaller thiol ligands of simpler structure than in treatments with larger thiols and more “branched” structure (Paper III). Finally, it is demonstrated that MeHg bioaccumulation in zooplankton can increase in systems with highly heterotrophic food webs and enhanced loadings of terrestrial OM (Paper IV). Such conditions are expected to occur in northern latitude coastal systems following climate changes.

Key words: Mercury, methylmercury, bioaccumulation, mesososm, isotope tracers, methylation, demethylation, stability constant, kinetic model, coastal sea, ICPMS, LC-MS/MS

ii

List of publications Publications included in this thesis I. Organic matter drives high interannual variability in methylmercury concentrations in a subarctic coastal sea Anne L. Soerensen, Amina T. Schartup, Aleksandra Skrobonja and Erik Björn Environmental Pollution, 2017, 229, 531-538

II. Deciphering the Role of Water Column Redoxclines on Methylmercury Cycling Using Speciation Modeling and Observations from the Baltic Sea Anne L. Soerensen, Amina T. Schartup, Aleksandra Skrobonja, Sylvain Bouchet, David Amoroux, Van Liem-Nguyen and Erik Björn Global Biogeochemical Cycles, 2018, 32, 1498-1513

III. Uptake kinetics of methylmercury in a freshwater alga exposed to methylmercury complexes with environmentally relevant thiols Aleksandra Skrobonja, Zivan Gojkovic, Anne L. Soerensen, Per-Olof Westlund, Christiane Funk and Erik Björn Environmental Science & Technology, 2019, 53, 13757-13766

IV. Multiple impacts of humic-rich dissolved organic carbon on methylmercury accumulation in heterotrophic pelagic food webs Aleksandra Skrobonja, Sonia Brugel, Anne L. Soerensen, Evelina Griniene, Agneta Andersson and Erik Björn Manuscript in preparation

Paper I was reprinted with permission from Environmental Pollution. Copyright © 2017, Elsavier Ltd. All rights reserved. Paper II was reprinted with permission from Global Biogeochemical Cycles. Copyright © 2018, American Geophysical Union. All rights reserved. Paper III was reprinted with permission from Environmental Science & Technology. Copyright © 2019, American Chemical Society. All rights reserved.

iii

Author contribution:

Paper I. The author participated in the planning and execution of one of the field campaigns used in the study, performed the experimental analysis from samples collected during several field campaigns and commented on the manuscript.

Paper II. The author participated in the planning and execution of one of the field campaigns used in the study, conducted most of the experimental work, participated in the data evaluation and commented on the manuscript.

Paper III. The author formulated the scientific research objectives and led the planning of the study, performed most of the experiments and data processing and was lead author on the manuscript.

Paper IV. The author was involved in formulating the scientific research objectives and approaches, led the planning of the study, performed the experiments related to the behavior of mercury in the system, processed the data and contributed significantly to the writing of the manuscript.

iv

Enkel sammanfattning på svenska Metylkvicksilver (MeHg) är ett potent neurotoxin som kan bioackumuleras till skadliga nivåer i akvatiska näringsvävar. Bildning av metylkvicksilver är en övervägande biotisk process som involverar olika mikroorganismer (t.ex. järn- eller sulfatreducerande bakterier). Bildningen av MeHg är relaterad till förekomst av organiskt material (OM) som innehåller substrat som är viktiga för metylerande mikrober och som innehåller reducerade svavelligander (tioler, RSH) som bildar starka bindningar med oorganiskt kvicksilver (HgII) och påverkar biotillgängligheten. I akvatiska system biokoncentreras MeHg från vatten till näringsvävens bas och detta steg är avgörande för MeHg koncentrationer på högre trofiska nivåer. Överföring av MeHg mellan trofinivåer i näringsvävar är också av stor betydelse. Tilförsel av OM i kustområden påverkar ljusförhållanden som behövs för växtplanktons tillväxt och främjar heterotrofi, dvs. bakterieproduktion. Detta kan leda till ett skift från den växtplanktonbaserade till den längre bakteriebaserade näringsväven och påverka mängden av MeHg som bioackumuleras i högre trofiska nivåer. Metylkvicksilvernivåer i rovdjursbiota påverkas av biotillgängligheten av HgII för metylering (studerad i artiklarna I & II), speciation av MeHg i vatten, avgörande för MeHg inkorporering vid basen av näringsväven (artikeln III) och strukturen av den pelagiska näringsväven (artikeln IV). I denna avhandling visas det att OM kan fungera som en prediktor för halten av löst MeHg i estuarier och kustsystem. Organiskt material påverkar halterna av MeHg både genom biotillgänglighet av HgII (genom Hg-komplexering med humusämnen) och aktiviteten för metylerande mikrober (genom att tillhandahålla metaboliska elektrondonatorer) (artikeln I). Förhöjda koncentrationer av partikulärt och löst HgII och MeHg är dessutom kopplade till förekomst av pelagiala redoxkliner i kustnära hav. Redoxkliner påverkar speciationen av HgII i vatten och biotillgängligheten för metylering (artikeln II). Det visas vidare att molekylstrukturen för ligander i MeHg-komplex påverkar kinetiken för upptag av MeHg i växtplankton. Konstanter för associering av MeHg till cellytan i en grönalg är högre i behandlingar med små tiolligander med enklare struktur än i behandlingar med större tioler och mer "grenad" struktur (artikeln III). Slutligen visas det att MeHg bioackumulering i djurplankton kan öka i system med mycket heterotrofiska näringsvävar och ökad tillförsel av terrest OM (artikeln IV). Sådana förhållanden förväntas öka i kustnära system på nordliga latituder i och med klimatförändringar.

v

1. Introduction 1.1. Mercury as a pollutant Mercury (Hg) is a global pollutant which causes adverse health effects to humans and wildlife. It is a naturally occurring element in the environment, comprising the Earth’s crust with an average mass abundance of 0.08 parts per million (ppm).1 The Hg ores may contain up to 12-14% of Hg, most commonly found as cinnabar (HgS) ore.2,3 Sources of Hg release into the atmosphere can be of both natural (e.g. volcanic eruptions) and anthropogenic origins (processes involving burning of fossil fuels, mining and other industrial activities).4–6 Mercury is also present in the atmosphere, mainly in the form of gaseous elemental mercury (Hgo) in a relatively low range of concentrations (1-170 ng m- 3).7,8 Both of these types of sources are predominantly emitting inorganic forms of Hg, however, the more toxic organic form of Hg, methylmercury (MeHg), is the dominant Hg species found in rice and fish.9–11

Excessive exposure to Hg can result in damage to the central nervous system (CNS) and in severe cases lead to death. Higher risk for Hg exposure was linked to people working in mines under poor conditions and insufficient knowledge.7 However, the main source of Hg to humans involves ingestion of fish and rice originating from Hg contaminated waters and rice paddies.9,12–14 The first incident involving consumption of MeHg contaminated fish and shellfish was officially documented in 1956 in Minamata city, Japan.15 With thousands of patients diagnosed, and a high mortality rate, the disease including MeHg intoxication was named “The Minamata disease”.15,16

During the 1950s and 1960s, the use of organic Hg compounds as fungicides in agriculture and pulp and paper industry was widespread in Sweden. Moreover, high amounts of Hg were discharged into the water and air from chlorine-alkali industries as well as from incineration of waste containing Hg.17 Today, one of the most serious concerns is the elevated concentration of Hg in piscivorous fish (e.g. pike), as it exceeds the levels considered safe for human consumption (0.1 µg of Hg/kg body weight/day).18 Furthermore, it has been reported that elevated hair concentration of Hg (0.58 µg g-1) was detected in about 10% of children born in Sweden and this could be linked to the mother’s diet affected by Hg levels surpassing health guidelines and regulations.7,19 1

Considering the bioaccumulating, biomagnifying properties of MeHg and its effect on the population and the environment, it is of great importance to improve the understanding of MeHg formation, entrance into the base of the aquatic food web, bioaccumulation and biomagnification in the food web.

1.2. Formation of MeHg Methymercury is formed from inorganic mercury (HgII) predominantly via biological reactions which take place in suboxic and anoxic environments. The methylation process has been coupled with the presence of sulfate reducing bacteria (SRB), iron reducing bacteria (IRB), and more recently, methanogens and other microbes.20–25 In the oxic water column, HgII can be converted to MeHg in the water column during the process of organic matter (OM) remineralization (microbes break down the OM) and within particulate organic matter (POM, fraction of OM that does not pass through a 0.2 µm filter) aggregates with low oxygen micro-environments.26–30 Bioavailability of HgII is an important factor for Hg methylation as it can limit the amount of available HgII for uptake by methylating microogranisms. In nature, HgII availability and speciation are controlled by pH and redox conditions, concentration of inorganic sulfide (S(-II)) due to the strong affinity to Hg, and properties of organic matter (OM). Precursor material, source origin, maturity, composition and other characteristics of OM can affect HgII methylation. Namely, OM can carry substrates necessary for methylation microbes and promote methylation.31 On the other hand, OM can contain ligands which form strong bonds with HgII, such as organic ligands with a sulfhydryl (thiol) group (RS-) in their structure, and repress methylation.32–37 Dissolved organic matter (DOM) is defined in this work as the fraction of OM which passes through the 0.2 µm filter, and HgII and MeHg in marine environments may be present complexed with OM in the dissolved form, depending on the binding affinity to ligands and ligand distribution between the dissolved and particulate phase.35,38 Redox potential is another important factor controlling MeHg production in water columns as elevated concentrations of total mercury (THg) and MeHg

-1 39– have been reported in hypoxic (<2 mL O2 L ) and anoxic (total depletion of O2) zones. 41

2

In this thesis work, research on MeHg formation was conducted based on observations from the Baltic Sea. In the northern Baltic Sea, freshwater input creates a north-south salinity gradient and causes a decrease in the fraction of terrestrially discharged (allochthonous) to in situ produced (autochthonous) DOM along this gradient.42,43 During spring and fall, loading of allochthonous DOM to the estuarine and nearshore sites increases.44,45 In the southern basins of the Baltic Sea, phytoplankton blooms caused by excess nutrient and OM runoffs increase the system’s biological oxygen demand (BOD) and contribute to the formation of permanently hypoxic and anoxic zones (35-40% of the surface area).46–48 Mercury in the surface waters binds to allochthonous and autochthonous particulate organic matter (POM) and sinks into the deeper water where it is released during the microbial OM remineralization process. Furthermore, increases in dissolved THg at the hypoxic-anoxic interface have been reported in other systems with hypoxic and anoxic conditions and these observations were linked to the dissolution of iron (Fe) aggregates which trap OM and Hg.49,50 Redox potential can further induce different distributions between Hg-thiol and Hg-sulfide complexes and Hg partitioning between the dissolved and particulate phases which all affect Hg bioavailability and reactivity.

One of the objectives in this thesis work was to explore how OM composition and concentration as well as hypoxic/anoxic conditions impact MeHg production and fate in the water column of coastal seas (Paper I & II).

1.3. Uptake of MeHg Once MeHg is produced in the water column, it can be bio-concentrated to the base of the food web. This step is crucial for MeHg levels observed higher up the food web as the concentration from the water to the basal producers is increased by a factor of 104 to 107.38,51–56 Cellular uptake of metals by microorganisms from the aqueous phase involve diffusion of the metal (or metal complex) to the cell surface which is then followed either by passive diffusion through the cell wall, or formation of a metal complex with ligands located on the cell wall or biological membranes and subsequent active transport into the cytoplasm.57–59 The association of MeHg to the cell surface would be regulated by the competition between its binding to the ligands in the dissolved phase and at the cell 3

surface.This is hypothesized in the Biotic Ligand Model (BLM) which has been previously established for several trace metals, but not yet for Hg.60,61 In natural waters, MeHg forms strong complexes with organic ligands which contain reduced sulfur groups in the DOM pool, and the chemical speciation of MeHg thus controls its bioavailability and governs the uptake rate by phytoplankton or bacteria.32–34,62 Various studies have investigated the relation between aqueous speciation of MeHg and cellular uptake as well as the mechanistic aspect of the MeHg uptake process.52,62–67 While some studies suggested passive diffusion as the main uptake mechanism for MeHg uptake in unicellular algae63,66,68, evidence for active transport being involved in this process has been found in several other cases.52,54,62

In Paper III, we investigate how the molecular structure and size of six thiol compounds commonly present in the environment affect cellular uptake kinetics of MeHg in the green freshwater alga Selenastrum capricornutum. The uptake of MeHg was determined in short-term exposure experiments on the laboratory scale for whole cells and the cytoplasmic fraction. This data was further applied to two kinetic models which were used for determination of the rate constants involved in the cellular uptake process of MeHg.

1.4. Bioaccumulation of MeHg Following the bioconcentration step from the dissolved phase to the base of the pelagic food web, MeHg is mainly accumulated in zooplankton and higher trophic levels via grazing and less through direct uptake of MeHg from the water column.67 Herbivorous zooplankton feed on phytoplankton and microbes whereas the diet of omnivorous zooplankton in addition includes ingestion of smaller zooplankton. Trophic transfer up the food web is more efficient for MeHg species and this results in higher percentage of MeHg in the THg pool as the trophic level increases.53,69,70 The concentration of MeHg in top consumers is influenced by the total length of the food web and the bioaccumulation factor (BAF) for each step (usually varying between 0.5 to 1.0 log units).71,72 Increase in MeHg from one trophic level to another is defined as biomagnification factor (BMF). It has been proposed that longer food webs such as those found in heterotrophic systems result in elevated concentrations of MeHg in zooplankton.73 Moreover, factors like growth

4

rate (biodilution), trophic level of the prey and lifespan and body size of the consumer may impact and control MeHg levels in these predators.70,72,74,75

The Baltic Sea is a coastal, semi-enclosed sea with spatial gradients in salinity, temperature, proportions of allochthonous to autochthonous OM and, consequently, share of bacteria versus phytoplankton as primary producers. These factors could be important controllers for MeHg formation, bioaccumulation and biomagnification in different areas of the Baltic Sea. Recently, it has been suggested that MeHg production and bioavailability are reduced in marine systems with higher influence of allochthonous over autochthonous OM.27,76 However, aqueous concentration of MeHg is most often higher in the estuarine coastal areas than in the offshore waters due to the transport of MeHg by river discharge. Higher content of allochthonous OM can promote heterotrophy and cause changes in the structural changes at the base of the pelagic food web, potentially enhancing MeHg bioaccumulation.77 To test this hypothesis, we have performed a mesocosm experiment (Paper IV).

5

1.5. Aims of the thesis This thesis work was conducted to improve the current understandings of MeHg formation, incorporation in the base of the food web and bioaccumulation. The individual aims of the thesis work are (1) Deciphering the role and impact of OM and water column redoxclines on MeHg formation and cycling in a subarctic coastal sea (using observations from the Baltic Sea) (Paper I & II). (2) Establishment of comprehensive kinetic models describing MeHg uptake rate constants in phytoplankton, based on MeHg speciation in the medium (Paper III). (3) Exploration of the OM and food web structure influence on bioaccumulated MeHg in higher trophic levels (zooplankton) (Paper IV). The summary figure showing contributions of each Paper to address MeHg formation, uptake and bioaccumulation is shown in Figure 1.

Figure 1. A summary figure illustrating used approaches and studied processes of this thesis work.

6

2. Materials and Methods 2.1. Study sites Water and seston samples were collected from different areas of the Baltic Sea during two individual cruises for three consecutive years (2014, 2015 and 2016), i.e. in total six cruises. The field samplings included cruises with Umeå Marine Science Center crew onboard the Swedish Coast Guard’s ship KBV181 in the northern basin of the Baltic Sea (Bothnian Bay and Bothnian Sea) whereas samples from the southern areas of the Baltic Sea (the Baltic Proper, Belt Seas) were collected onboard R/V Aranda research ship in collaboration with Swedish Meteorological and Hydrological Institute (SMHI). The map, locations and sampling transect are shown in Figure 2.

Figure 2. Map of the Baltic Sea showing the locations of sampling stations as well as the sampling transect. Station names in black color indicate stations where anoxic conditions were not detected, stations marked in red illustrate the locations with anoxic conditions in deep water while the names in purple indicate stations where anoxia was present only some years. (Reproduced from Paper II with permission from American Geophysical Union) 7

2.2. Experimental approaches 2.2.1. Field sampling Water and seston samples were collected for three consecutive years, during six cruises in the Baltic Sea (data published in Paper I & Paper II): September 2014 (SEP14-S); July 2015 (JUL15-S) and July 2016 (JUL16-S) in the Southern Baltic (Baltic Proper and Belt Seas), and September 2014 (SEP14-N), August 2015 (AUG15-N) and August 2016 (AUG16-N) in the Northern Baltic Sea (Bothnian Bay and Sea). All water samples were collected using 2 L Niskin bottles attached to a rosette or 5 L Teflon-lined General Oceanic (GO-FLO) bottles. Total Hg samples were collected in 125 mL amber glass (I-CHEM certified™) or PFA and/or FEP (Nalgene) bottles while the samples for MeHg analysis (monomethylmercury (MMHg) + dimethylmercury (DMHg)) were collected in 125 or 250 mL Teflon® or HDPE bottles pre- cleaned with trace metal grade HCl (Suprapur HCl, Merck). A subset of MeHg samples in 2015 and 2016 were filtered through 0.22 µm hydrophilic PTFE filters using Teflon or Sterivex 0.22 µm filter units. All samples were preserved with hydrochloric acid (1% v/v, Suprapur® HCl 30%, Merck Millipore) and stored in the dark at 4°C until they were further processed.

2.2.2. Determination of total Hg and MeHg in aqueous samples and biota

Total Hg in natural aqueous samples was detected using atomic fluorescence spectrometry following USEPA 1631.78 All Hg species in the samples were oxidized to HgII using bromine monochloride (BrCl) which was followed by HgII reduction to volatile

0 78 elemental Hg (Hg ) using tin(II) chloride (SnCl2) solution. In the mesocosm water samples, elemental Hg was purged and the analysis was done using an on-line cold vapor generation (HGX-200, Cetac) system connected to a PerkinElmer/Sciex ELAN DRCe ICPMS. Total Hg in biota samples was measured using a Direct Mercury Analyzer (Leco DMA 80) following the manufacturer’s method specifications.

Methylmercury content in water samples was analyzed based on previously described methods.79,80 In short, water samples were pH adjusted to 4.8 using a 2 M acetate buffer and sodiumhydroxide (NaOH). Further, MeHg species were ethylated by addition of

8

sodium tetraethylborate (STEB) and purged and trapped onto Tenax adsorbent matrix. Ethylated mercury species were then thermal-desorbed to gas chromatography coupled with inductively coupled plasma mass spectrometry (TDGC-ICPMS, 6890 Agilent GC and 7700 ICPMS).81 Biota samples for MeHg analyses were digested for 2 h at 60 °C using tetramethylammonium hydroxide (TMAH) and sonication.82 The pH value was subsequently adjusted to 4.8 by adding concentrated acetic acid, after which the samples were ethylated, purged and trapped, and analyzed in the same way as the water samples.

2.2.3. Methylation and demethylation rate constants

II -1 The rate constants for inorganic Hg methylation and MeHg demethylation (km, kd (d ), respectively) in Paper II and IV were determined using enriched stable isotope tracers and followed the experimental and data treatment described by Rodriguez-Gonzales et al.83 The rate constants were determined in incubation experiments which involved both light exposed and samples kept in the dark. In Paper II, the light incubations were performed on water samples collected from 5-10 m and in a few cases from 20 m, during the 2015 and 2016 field sampling onboard. The dark incubations were performed in 2014, 2015 and 2016 for water collected from three different depths (<15 m, 30-40 m and >60 m). Each sample was spiked with isotopically enriched Me204Hg and 198HgII and the incubations were stopped by acidification with 0.1 M HCl. For 2014 and 2016 samples, the incubations were stopped at t=0 and 24 h while the incubations for samples collected in 2015 were stopped at t=0, 4, 8 and 24h. In the mesocosm experiment (Paper IV), the incubations were stopped at t=0 and 14 h and only the dark incubations were performed. Following acidification, the samples were spiked with isotopically enriched Hg standards for isotope dilution analysis and the samples were analyzed as previously described.

9

2.2.4. DOC, humic matter content and ancillary data Concentrations of oxygen, sulfide, nitrate, ammonium, phosphorous, silicate, chlorophyll a, DOC and humic matter content are measured on a monthly basis as part of the Swedish National Monitoring Program. The samples are analyzed following HELCOM guidelines84 and it is possible to download the quality controlled data from the SMHI’s website (Paper I & II).85 Humic matter is measured using fluorescence spectroscopy at 350/450 nm excitation/emission wavelength using quinine sulfate as a calibration standard (Papers I, II & IV).84 Concentrations of DOC were determined using a Shimadzu TOC-5000 high temperature catalytic oxidation instrument with NDIR detection after sample acidification and purging.86 Carbon concentration was then calculated using potassium hydrogen phthalate as a standard.

2.2.5. Microalgal cultivation The green unicellular microalga Selenastrum capricornutum (Culture Collection of Algae, Goettingen University, SAG) was used to assess MeHg uptake in assays with six different MeHg-thiol complex treatments and a control (Paper III). All glassware was autoclaved and acid washed following trace metal clean protocols. In the pre-inoculum, 30 ml of algal culture was grown axenically in 100 mL Erlenmeyer flasks. The medium used in this study was sterile Bold’s basal medium, depleted in Cu, Zn and EDTA, with addition of 30 mM sodium nitrate (3N BBM)87. The culture was kept in a closed orbital shaker at 115 rpm,

-2 -1 25°C and 100 µmolph m s of white light for seven days. The pH was maintained at 6.6 ±

0.2 by the balance between CO2 addition (acidifying the medium) and gradual nitrate assimilation by the alga (uptake of protons).88,89 The medium was prepared without Cu and Zn to avoid their competition with MeHg for cellular binding sites, and EDTA was left out to avoid competitive binding of MeHg. The culture was transferred to 100 ml Erlenmeyer flasks with 70 ml of the same BBM medium, maintained for 5 days at 20°C,

-2 -1 under 100 µmolph m s of white light and bubbled with 3% CO2 in air at a constant flow of 1 L min-1. No contamination of the culture was detected under the light microscope (Leica DMi1, 40× magnification). Population density was measured each day using Beckman Coulter Multisizer 3 cell counter (aperture size 70 µm, analytical volume 100 µL, sample dilution 200×) following manufacturer’s method specifications. Optical density

10

was measured at 750 nm wavelength on a UV/Vis Spectrophotometer (Varian Carry 50 Bio) using a 10 mm light path polystyrene cuvette. The culture was inoculated and further cultivated in a 5 L Duran® borosilicate glass bottle with a GL 45 screw neck. The bottle was filled up to 4.8 L with depleted BBM. The culture was illuminated with a white LED panel

-2 -1 -1 (650 µmolph m s ) and bubbled with 3% CO2 in air at a constant flow rate of 1 L min . After ~48 h, the cell density reached ~ 4 000 000 cell ml-1 and the content from the 5 L bottle was evenly divided into seven different trace metal clean and autoclaved 1 L Duran® borosilicate bottles. The control (MeHgOH) and six MeHg-thiol complexes were then added to separate bottles to the final concentration of 1.5 nM MeHg and 90 nM thiol. Each solution was further split into two replicates and the flasks were maintained

-2 -1 for 8 h at 20°C using a white LED panel (650 µmolph m s ) bubbled with a constant flow of 1 L/min of 3% CO2 in air during the exposure.

2.2.6. Mesocosm experiment Brown, humic rich soil was sampled from a location close to the Öre River (coordinates 63° 59.5469′ N, 19° 80.9809′ E). The soil was homogenized manually and sieved through 4-6 mm sieve nets. Following, 9.8 kg of the soil was mixed with 2.3 kg of chelating resin (Amberlite™ IRC748 and Diaion® CR11, sodium form, Sigma Aldrich) and 65 L of MQ water (18.2 MΩ cm-1, at 25°C and 5 ppb total organic carbon (TOC)) for the extraction of DOC. A mechanical pump was immersed into the extraction tank to allow vigorous stirring during 48 h of extraction. The extract was sequentially passed through 100 µm and 15 µm nets, centrifugated (10000 rpm, 10min, rotor JA-14, 6*250ml bottles) and stepwise filtered through a filtration system compiled of 25, 5, 3 and 1 µm filters. In Paper IV we refer to the <1 µm extract fraction as DOC, even though it has not passed through a 0.2 µm filter. The TOC content of the extract was 2.4 g L-1 and total Hg and MeHg concentrations were 8.9 ± 1.4 pmol g-1 (d.w) and 32 ± 4.8 fmol g-1 (d.w), respectively.

Twelve double-mantled high-density polyethylene (HDPE) indoor mesocosms (5  0.74 m in diameter) were used, located at the Umeå Marine Sciences Center facility, Umeå University (Paper IV). Using a pumping system, the mesocosms were simultaneously filled with 1.2 mm filtered brackish water (average salinity 3.79 ± 0.01 PSU, DOC 3.83 ± 0.07 mg L-1) from two inlets located 800 m from land (in the Öre River estuary) at water depths of 11

2 and 8 m which allowed even distribution of planktonic communities. For each DOC treatment (Level 1: DOC 5.5 mg L-1, Level 2: DOC 6.5 mg L-1 and Level 3: DOC 8.0 mg L-1) and the Reference, three replicate mesocosms were prepared and maintained. As light sources, 150 W metal halogen lamps were used (MASTER Colour CDM-T 150W/942 G12 1CT) and the light:dark cycle was set to 12:12 hours. The photosynthetically active radiation level of the lamps was adjusted to be 350 ± 17.5 µmol s-1 m-2 at 1 cm below the water surface. To achieve a high rate of convective stirring of the water column, the bottom sections of the mesocosms were continuously heated with a hot (40°C) 40% propylene glycol solution, while a cold (-4°C) solution was added to the top parts. The middle sections were set to maintain 15°C and thus controlled the overall temperature in the water column (average 15.2 ± 0.3 °C), while upholding convective stirring with a complete mixing for 20 min.90

- 3- + Nitrate (NO3 ), phosphate (PO4 ) and ammonium (NH4 ) solutions were prepared from pure salts of NaNO3, NaH2PO4  H2O and NH4Cl, respectively and added to the mesocosms as nutrients (detailed addition scheme is presented in Table S2 (Paper IV).

II The following Hg enriched isotopes, purchased as HgO or HgCl2 from Oak Ridge National Laboratory (TN, USA) were used in the experiment: 199Hg (91.09%), 200Hg (96.41%) and 201Hg (96.17%). The Me200Hg and Me201Hg enriched stock solutions were prepared following the procedures described elsewhere.91 Isotope standards 199Hg (10 µM) and Me201Hg (1 µM) were pre-equilibrated with 5 mL of the DOC extract in 5 L of MQ water by being left to mix on a magnetic stirrer overnight prior to the mesocosm additions (details specified in Table S3, Paper IV). Water samples were regularly analyzed for total Hg (HgT) and MeHg concentration after sampling and the additions would be accordingly adjusted the next sampling day to maintain a constant level of HgT (2 pM) and MeHg (0.2 pM).

Water samples were taken from the mesocosms at 1 m depth twice a week (Mondays and Thursdays) for determination of total Hg and MeHg concentrations, bacteria and primary production rates, and concentrations of DOC, humic matter, nutrients and chlorophyll (Chl a). Once a week, 50 mL and 25 L of water were collected for taxonomic classification and counting of nano- and microplankton, and zooplankton, respectively. 12

Water for total Hg and MeHg determination was sampled in acid-washed bottles of 125 mL (I-Chem Boston Round Narrow-Mouth Amber Glass Bottles) and 250 mL (Nalgene Fluorinated Narrow-Mouth HDPE Bottles), respectively, using a peristaltic pump with online filtration (0.2 µm PTFE membrane filters) and Teflon tubing. Total Hg samples were oxidized by adding of 1.25 mL of a BrCl solution78 prior to acidification. MeHg samples were acidified using hydrochloric acid (1% v/v, Suprapur® HCl 30%, Merck Millipore). Following, 200HgII or Me200Hg were added as internal standards for isotope dilution analysis and the samples were stored at 4°C until analysis.

Vertical irradiance profiles of photosynthetically active radiation were measured using a LICOR-193SA light sensor. Light measurements were carried out at the water surface and depths of 0.1, 0.2, 0.4, 0.6, 1, 1.5, 2, 2.5, 3, 3.5 and 4 m. The diffuse light attenuation coefficient (Kd) was calculated from the slope of the linear regression of the natural logarithm of down-welling irradiance versus depth. After each water sampling, the mesocosm tanks were refilled with 0.2 µm filtered seawater to compensate for the volume lost during the sampling procedures (~25 and ~10 L for sampling days including and not including plankton taxonomic determination, respectively).

Four seston size fractions were collected at the end of the experiment (day 36) from the entire water column above 0.77 m from the mesocosm bottom, using a series of 200-, 90- , 50- and 20 µm mesh size plankton nets for water filtration. The plankton samples were frozen and stored at -20°C until they were freeze-dried and analyzed for δ13C, δ15N and MeHg concentration.

13

3. Results and discussion It has been established that OM is a driver of HgII methylation and that OM composition as well as concentration have an impact MeHg levels measured in a certain system.27,37,92– 95 However, not many studies have looked into how interannual concentration and composition of OM affect MeHg formation and its biogeochemical cycling within systems. In Paper I, we used the Baltic Sea as our study system and explored how characteristics and concentration of OM affect aqueous MeHg concentrations. We used Hg and OM measurements in a range from 2014 to 2016 and also discussed how interannual variability of OM parameters may impact future MeHg concentrations, especially under predicted climate change scenarios. Furthermore, excess of nutrients and OM in coastal systems such as the Baltic Sea, transported by runoff from e.g. sewage treatment facilities or agricultural lands, cause phytoplankton blooms which, eventually, results in reduced oxygen levels or complete depletion in certain zones. A number of studies has evaluated the importance of chemical speciation of Hg on its cycling and bioavailability to biota, however, transformation pathways of Hg in coastal seas with water column redox gradients are understudied.36,96,97 Hence, in Paper II we used concentrations of THg and Hg species (e.g. Hg0, HgII, MeHg, DMHg) to assess MeHg production and its fate under hypoxic and anoxic conditions. Paper II uses the complete dataset from the field cruises 2014-2016 whereas in Paper I, a subset of this data is used since the focus of the paper is placed on the Bothnian Bay and Sea basins of the Baltic Sea. In Paper III we studied MeHg uptake by a green algae model organism in a microcosm-scale experiment. Contrasting results from previous work on MeHg uptake processes bring to attention the fact that more detailed studies are needed to understand MeHg bioavailability and uptake mechanisms in phytoplankton when MeHg is complexed with low molecular mass (LMM) thiols.52,62–67 In Paper III we evaluated how the molecular structure of six LMM thiols, detected in the Baltic Sea, affects cellular uptake kinetics of MeHg in the freshwater alga Selenastrum capricornutum when exposed to MeHg-thiol complexes. In paper IV, we studied MeHg bioaccumulation under increased loadings of terrestrial organic matter as it has been predicted that terrestrial runoff may increase in large regions, particularly in the northern hemisphere, due to the predicted global warming scenarios. It has been

14

previously demonstrated that increased input of terrestrial OM can enhance the bioaccumulation factor (BAF) of MeHg in zooplankton by shifting the pelagic food web from phytoplankton-based to more bacteria-based.73 We conducted a mesocosm experiment with three treatments containing different levels of increased DOC concentrations, mimicking hypothetical climate change scenarios, and a reference system with no additions of terrestrially derived OM. Using concentrations of MeHg and various biological parameters (e.g. primary and bacteria production rates, taxonomic classification), we calculated the BAF of MeHg in zooplankton, described the food web structure in the four treatments and discussed the outcome with regards to the predicted increase in precipitation followed by enhanced input of terrestrial OM to coastal systems, induced by climate change.

3.1. Organic matter input impacts MeHg formation and cycling (Paper I) Organic matter contains strong binding ligands for HgII such as reduced sulfur groups, and substrate molecules which are essential for metabolism of heterotrophic microbes.31,35 Since methylating organisms require both bioavailable HgII and metabolic supplements to produce MeHg, OM composition and concentration are one of the most important factors that control MeHg ambient concentrations. In the Northern Baltic Sea, most of the freshwater enters the system through the Bothnian Bay and creates north to south gradient in salinity and proportions of allochthonous and autochtonous DOM (fraction of organic matter in solution that passes through a 0.2 µm filter). The amount of terrestrially derived DOM decreases from the coastal regions over the Bothnian Bay and Sea, however the total DOM levels are relatively constant in the offshore waters.42 Furthermore, it has been shown that there is a ~300 µM refractory level of DOC (concentration of organic carbon in the DOM fraction) which is not readily utilized by microbes, thus the remaining portion of labile DOM (concentration >300 µM) from the total DOM pool has potential to be metabolized by microbes and cause OM remineralization in this area.45,98

The trend showing decrease in total Hg following the salinity gradient from the Råne estuary (5.95 pM, station RA1 in Figure 1, inner estuary, and 2.00 pM, station RA2, outer estuary), over the Bothnian Bay (1.2 ± 0.30 pM, stations A5 and A13) and the Bothnian Sea (0.84 ± 0.24 pM, stations C3, C14, GA1) is presented in Figure 3A. Figure 3B illustrates 15

the same decreasing trend for MeHg from the maximum measured concentration in the inner estuary (306 fM) to the average offshore concentrations (80 ± 25 fM in 2014,

16

Figure 3. Water MeHg and DOC concentration. A) solid lines indicate significant linear regressions for the Bothnian Bay and Bothnian Sea; B) The color of squares indicates the concentration of humic matter content. (Reproduced from Paper I with permission from Elsevier Ltd)

From aqueous phase in estuarine systems, Hg can be removed either via evasion or settling with particles. Moreover, significant amounts of DOC may be lost through flocculation and settling in areas where freshwater and water with higher salinities mix.101 In our study, the loss of total Hg due to the conservative mixing of salinities 1 to 2 causing DOM flocculation and settling, is estimated to be 49% (Figure 3A and Figure S1, Paper I). Looking into the effect of POC, DOC, suspended particulate matter (SPM) and particle 17

partitioning coefficient, we calculate that 1 – 11% of total Hg is lost through particle settling in the Råne estuary.102–104 We furthermore estimate 8 – 17% of dissolved Hg to be lost through DOM flocculation and settling, and prescribe the rest of the remaining loss of total Hg from the estuary (43-82%) to be due to Hg0 evasion from the system.101,105,106 Even though some fraction of the MeHg pool undergoes the same removal transformation as HgII (i.e. particle and flocculation settling), HgII methylation and MeHg demethylation processes also come into play, affecting the proportions of HgII and MeHg when comparing estuarine and offshore water columns.

We propose that the DOC and humic matter content can be used to explain the variability in MeHg concentration in the Northern Baltic Sea. We further suggest that this variability is controlled by the OM remineralization rate (with labile DOC used as a proxy) and HgII availability (explained by the humic matter content effect), both contributing to net HgII methylation.

3.2. The role of water column redoxclines on MeHg formation and cycling (Paper II) The importance of redoxclines in water columns has been shown through elevated THg

-1 and MeHg concentrations reported for hypoxic (<2 mL O2 L ) and anoxic (total depletion

39–41 of O2) zones. To be able to predict the Hg fate following increased spread of oxygen depleted zones in marine ecosystems, better understanding of Hg chemical speciation and cycling across the redoxclines is needed.47,107 The Baltic Sea is a shallow, semi-closed system with a large inflow of freshwater from runoff.108,109 The runoff usually contributes to an increase in the amount of nutrients and OM (e.g. from water treatment facilities or agricultural soils), which induces phytoplankton blooms and eventually causes oxygen depletion in certain areas.47,107 Since the majority of methylating microbes thrive in hypoxic or anoxic oxygen conditions, these oxygen depleted zones present potential hotspots for MeHg formation. In the surface water, Hg can bind to the particulate organic matter (POM, fraction of OM that does not pass through a 0.2 µm filter) transported by runoff or formed from decaying biological material, and sink to deeper waters where it can be released during remineralization of OM. The released HgII is potentially available 18

for methylating microbes at or below the redoxcline. Previous studies conducted in the Black sea found elevated concentration of THg at the interface between the hypoxic and anoxic zone, and it was suggested that dissolution of iron (Fe) oxides with the capability to trap Hg and OM in their aggregates is causing these increases in THg levels.49,50 Methylmercury concentrations measured in the Baltic Sea were higher in the hypoxic layer than in the surface water, and highest in the anoxic zone.39,110 Furthermore, it has been suggested that HgII and MeHg in coastal seas will mainly bind to reduced organic sulfur ligands and form complexes which could partition between the dissolved and the particulate phase depending on the redox conditions, and this could impact Hg bioavailability and reactivity.36 Spatial distribution of averaged THg and MeHg for all cruises is presented in Figures 4A and 4B. To summarize, total and methylmercury concentrations were all significantly higher in the hypoxic and anoxic layer compared to the surface one (range of 0.5 to 10.7 pM for THg and

19

Figure 4. Spatial distribution of A) total Hg, B) MeHg C) HgII methylation rate constants and D) dark MeHg demethylation rate constants averaged across all sampling campaigns. Grey lines indicate the normoxic-hypoxic interface and white lines the hypoxic-anoxic interface. (Reproduced from Paper II with permission from American Geophysical Union)

Spatial distribution of HgII methylation and MeHg demethylation rate constants in unfiltered water samples is presented in Figure 4C and 4D. The methylation rate constants in two samples obtained from the normoxic zone were detected (2.4 x 10-4 – 4.8 x 10-4 d- 1) and in four samples from the anoxic water (16 x 10-4 – 48 x 10-4 d-1). Methylation rate constants for HgII were 4-20 times higher in the anoxic water than in the normoxic layer. The range of MeHg dark demethylation rate constants was

20

explained by the presence of the same bacteria responsible for HgII methylation and higher bioavailability of the dissolved MeHg-sulfide complexes.116–118 Furthermore, most of the MeHg variability across the anoxic zone is found close to the hypoxic-anoxic interface and we suggest three possible explanations for this observation: 1) direct mixing of water mass between the two zones, 2) a change in chemical speciation of HgII which affect its bioavailability for methylation and 3) the activity of methylating microbes depending on the concentration of electron donors or acceptors (OM composition,

2- Fe(III)/SO4 ). Results from our speciation model suggest that, for the rest of the anoxic zone, chemical speciation of HgII does not considerably vary and will not drive variability in MeHg concentrations or HgII methylation.

Exposure of MeHg to biota in water columns with redoxclines would primarily be impacted through the organisms feeding in the hypoxic zone. As phytoplankton are less likely to be found in this zone, zooplankton can take up MeHg directly from the water or through ingestion of OM and smaller zooplankton. The 2-6 time increase in MeHg concentration could thus result in significant exposure for zooplankton and the effect of grazing in the hypoxic zone remains to be addressed, especially considering the spread of oxygen depleted areas in coastal zones worldwide.

3.3. Uptake kinetics of MeHg in a freshwater alga exposed to MeHg- thiol complexes (Paper III) Methylmercury is bio-concentrated from water to the base of the food web by a factor of 104 to 107 and this step is thus crucial for MeHg concentrations found at higher trophic levels in aquatic ecosystems.38,51–53 Uptake of MeHg involves diffusion from the medium to the cell surface which is followed either by diffusion through the cell wall or binding to reactive sites at the cell wall or the membrane and transport into the cytoplasm.57–59 The bioavailability of MeHg is controlled by its chemical speciation in the medium and this is a key factor when determining cellular uptake rate and accumulation by bacteria or phytoplankton.32,62 However, previous studies have reported contrasting results with regards to the type of mechanism involved in the cellular uptake process.52,54,62,63,66,119 We evaluated how the chemical structure and stability constant of MeHg complexed with six environmentally relevant low molecular mass thiols affect cellular uptake kinetics of

21

MeHg in a freshwater green microalga Selenastrum capricornutum. We tested the assumption that MeHg bound to hydroxide ligands in the control, or to thiols with simple structure (cysteine (Cys), mercaptoacetic acid (MAC) and 2-mercaptopropionic acid (2MPA)) would have higher uptake rate by our model organism than MeHg complexed with thiol ligands of larger size and structure complexity (glutathione (GSH), N-acetyl-L- cysteine (NACCys) and N-acetyl-Penicillamine (NACPEN)). We further developed two kinetic models and applied them to different algal fractions to calculate the rate constants for the processes involved in cellular uptake of MeHg. Methylmercury mole fractions in whole cells for individual MeHg-thiol treatments and the control during the entire exposure period (8 h) are presented in Figure 5, and the points represent measured values while the solid lines are the modeling results (Model 1). The measured MeHg fractions and the corresponding concentrations are shown in Figures S4 and S5 (Supporting Information, SI, Paper III). Already after 15 minutes, the fraction of MeHg associated with the algal cells increased rapidly reaching 0.33 ± 0.03 – 0.50 ± 0.09 in the MeHg-thiol treatments and 0.77 ± 0.03 in the control. At a time point between 2 and 4 h the cell associated MeHg fraction plateaued in the MeHg-thiol treatments, and after approximately 1 h in the control. We identified three groups of the tested complexes with an increasing rate of MeHg association with the cells (GSH, 2MPA, NACPEN, NACCys < MAC, Cys < Control). This observation is in line with the hypothesis that cellular uptake of MeHg bound to ligands with lower complex stability (such as hydroxide) or to small thiols of simple structure (MAC, Cys) is higher than for MeHg complexes with larger and/or branched-chain thiols (NACCys, NACPEN, 2MPA and GSH). During the first 60 min of the exposure, MeHg concentrations was measured in the cytoplasmic fraction of S. capricornutum and MeHg mole fractions were calculated. Modeling results (Model 2) for these experimental data are presented in Figure 6 while the corresponding concentrations are found in Figure S6 (SI, Paper III). At 1 h of exposure, the cytoplasmic fraction of MeHg ranged from 0.070 ± 0.007 to 0.150 ± 0.003 in the MeHg-thiol treatments, and was 0.107 ± 0.008 for the control. The cytoplasmic accumulation of MeHg was slow when compared to MeHg association to the cell membrane. Rapid association of MeHg to the cells was observed by Gorski et al.65 The proportions of cytoplasmic MeHg in our study are lower than what has previously been 22

reported, however, it cannot be excluded that the cytoplasmic mole fraction of MeHg would continue to increase beyond the fraction measured at 1 h and steady-state conditions predicted by Model 2. Two kinetic models for rate constant determination were used in this study. The two-site Model 1 was used to describe the exchange kinetics of MeHg between the bulk medium and whole cells and two rate constants were calculated: MeHg cellular association (kas) and clearance (kc) rate constants. The kas values for the thiol treatments ranged from [29.3

-3 -3 ± 2.9] x 10 (MeHg-GSH) to [42.5 ± 2.7] x 10 (MeHg-Cys) and kas for the control was determined to be [97.6 ± 4.1] x 10-3 min-1. The three-site model (Model 2) was used to describe the MeHg exchange between the bulk medium, cell membranes and cytoplasmic fraction via rate constants for MeHg adsorption (kad) and desorption (kd) to cell membranes and cellular internalization (ki) and efflux (ke). Using the rate constants determined by Model 1 and 2, we calculated thermodynamic constants between the whole cells and the medium (kBA=kas /kc), cell membranes and the medium (kCA=kad /kd) and cell membranes and the cytoplasmic fraction (kDC=ki /ke). The rate and thermodynamic constants for all treatments and the control are presented in Table 1. The calculated kas indicated that association of MeHg to the cell was slower when MeHg is bound to thiol ligands with larger size and/or branched structure (GSH, NACCys, NACPEN, 2MPA) than when it is bound to hydroxide or smaller thiols with simpler structure (Cys, MAC). Results reported by Lee and Fisher67 and our results suggest that MeHg has lower bioavailability when it is bound to thiol ligands (log K = 16.7 – 17.5)120–122 than to less thermodynamically stable ligand such as hydroxide (log K = -4.5). However, the relative cellular uptake rate differs depending on the phytoplankton species.

23

Figure 5. Methylmercury (MeHg) mole fractions in whole cells and medium. The curves represent the fitted data wilt Model 1 with a 95% confidence interval while the points are the experimental data. (A) GSH, (B) 2MPA, (C) NACPEN, (D) NACCys, (E) MAC, (F) Cys, (G) Control. (Reproduced from Paper III with permission from American Chemical Society)

24

Figure 6. Methylmercury (MeHg) mole fractions associated with cell membranes, cytoplasmic fraction and medium. The curves represent the fitted data with Model 2 with a 95% confidence interval while the points are the experimental data. (A) GSH, (B) 2MPA, (C) NACPEN, (D) NACCys, (E) MAC, (F) Cys, (G) Control. (Reproduced from Paper III with permission from American Chemical Society)

25

Table 1. Calculated MeHg rate constants (association kas, clearance kc, adsorption kad, desorption kd, internalization ki, efflux ke) for MeHg-thiol treatments and the control using Model 1 and Model 2, equilibrium values (KBA, KCA, KDC), model merits of fit and statistical p value for kas and kc. (Reproduced from Paper III with permission from American Chemical Society)

MODEL 1 MODEL 2

kas KBA kc Merit of kad kd KCA ki ke KDC Merit of fit (%) fit (%) (×10-3 min-1) (×10-3 min-1) (×10-3 min-1) (×10-3min-1) (×10-3 min-1) (×10-3 min-1)

GSH 29.3 ± 2.9 A*, a 26.4 ± 0.6 1.1 ± 1.1 0.75 29.4 1.8 16.3 14 100 0.14 1.72

2MPA 30.0 ± 1.9 A, a 15.5 ± 0.2 1.9 ± 0.8 0.27 30.1 2.7 11.2 18.6 170 0.11 0.41

NACPEN 30.8 ± 2.3 A, a 22.2 ± 0.4 1.4 ± 0.9 0.42 32.1 1.7 19.0 101 700 0.14 2.68

NACCys 33.2 ± 2.7 A, a 18.8 ± 0.4 1.8 ± 1.0 0.41 33.0 2.5 13.2 98 700 0.14 0.73

MAC 39.0 ± 4.6 A, b 30.7 ± 0.9 1.3 ± 1.6 0.24 40.1 1.8 21.7 23.8 140 0.17 0.65

Cys 42.5 ± 2.7 A*, b 43.5 ± 0.7 1.0 ± 1.0 0.17 42.6 1.3 33.0 28 200 0.14 0.42

Control 97.6 ± 4.1 B 35.2 ± 0.4 2.8 ± 0.8 0.03 98.7 3.3 29.9 16.5 150 0.11 0.11

Different capital letters indicate statistically significant (p < 0.05, one-way ANOVA, Tukey HSD post-hoc test) differences between individual treatments while the asterisk symbol (*) marks that ka for GSH is close to being statistically significant from Cys (p = 0.057). Different lowercase letters indicate statistically significant (p < 0.05, one-way ANOVA, Tukey HSD post-hoc test) differences between the average ka of GSH, 2MPA, NACPEN and NACCys and the average ka of Cys and MAC.

26

We propose that the observed higher cellular uptake rate for thermodynamically less stable complexes and complexes with small ligands can be explained by uptake mechanisms involving metal-binding to the cell surface ligands and formation of a new complex with the cell surface biotic ligand prior to internalization (FIAM or BLM). However, we do not exclude passive diffusion as a transport mechanism. Although the fraction of MeHg adsorbed to the cell membrane was higher in the control than the thiol treatments, the cytoplasmic fraction of MeHg was not and this suggests that the rapid adsorption to the cell surface ligands may not necessarily lead to internalization. Overall, our results point to the potential importance of thiol ligand composition controlling MeHg incorporation in the food web in aquatic ecosystems.

3.4. Multiple impacts of humic-rich dissolved organic carbon on methylmercury accumulation in heterotrophic pelagic food webs (Paper IV)

Changes in climate is predicted to lead to regional increases in e.g. precipitation and snowmelt which is accompanied by enhanced input of terrestrially derived (allochthonous) OM to coastal systems.123–127

It has been suggested that increased loading of terrestrial OM may increase total Hg input to the systems by promoting MeHg formation in the water column.81,128 As another consequence of enhanced loading of humic rich OM, brownification of water decreases light penetration and suppresses photosynthetic primary production.129,130 However, OM may promote bacterial growth and increase their share in the pelagic food web structure thereby decreasing the energy transfer efficiency from the base of the food web to top predators.125,131–136

While bacteria are unable to be directly consumed by mesozooplankton due to their size as opposed to phytoplankton, they are readily grazed on by protozoans.137,138 Mesozooplankton also feed on protozoans and therefore larger share of bacteria as the base in heterotrophic food webs (microbial loop food webs) results in a higher number trophic levels and lower energy transfer efficiencies to top consumers.135,139 As MeHg 27

biomagnifies with each trophic level, higher MeHg levels can be hypothesized in top predators in systems favoring a bacteria-based food web compared to a phytoplankton- based one. In support of this hypothesis, Jonsson et al73 reported 2-7 times enhanced MeHg bioaccumulation factor (BAF) in zooplankton in a mesocosm experiment following a shift in food web structure from net autotrophic (46% bacteria production of the total basal production) to net heterotrophic (72% bacteria production). The shift in food web structure was induced by loadings of terrestrial OM relevant for regional climate scenarios for northern Scandinavia. In Paper IV we studied the effect of humic-rich DOC loading on the pelagic lower food web structure and bioaccumulation of MeHg in an experimental mesocosm study. We investigated if MeHg bioaccumulation is further increased, or if it levels-off, in food webs with higher proportions of bacteria production than previously studied. An improved understanding of the quantitative relationship between food web structure and MeHg bioaccumulation is important for reliable predictions of MeHg in aquatic biota for different environmental change scenarios. We used four treatments with different levels of DOC: Reference (4.0 mg L-1), Level 1 (5.5 mg L-1), Level 2 (6.5 mg L-1) and Level 3 (8.0 mg L-1) simulating the current average DOC level in the Baltic Sea (Reference) and increases by 38%, 63% and 100%, respectively (DOC and humic matter content results are presented in Figures S15 and S16, Paper IV).

The primary production rate (PP) followed the expected treatment effects whlie the bacteria production rate (BP) was high in all treatments (Table 2, Figure 7). The zooplankton production rate (ZP) was low in all treatments and it was used to calculate the theoretical food web efficiency (FWE) (Eq.S1, SI, Paper IV) which resulted in low values across treatments (Table 2). The δ15N results for the seston size fractions were used for calculation of the trophic level distance between the smallest (20-50 µm) fraction and other three (Eq. S2, Paper IV). The δ15N baseline in the systems is not known, and likely differed among treatments due to the added DOC and the total number of trophic levels up to zooplankton could therefore not be calculated. Therefore, we could not establish if there were any differences among treatments in the number of trophic levels for the part of the food web constituted by organisms <20 µm.

28

Phytoplankton species were classified into four groups based on their trophicity: heterotrophic (HT) organisms larger and smaller than 20 µm, and the sum of autotrophic and mixotrophic organisms (referred to as AU) larger and smaller than 20 µm (Figure 2, Figure S5, Paper IV). The taxonomic compositions of the AU and HT groups were overall very similar among treatments (Figures S6-7, Paper IV). Moreover, the abundance of heterotrophic nanoflagellates (HNF), serving as an indicator of heterotrophy within a system, was similar in all treatments (~1.0 mg C m-3) suggesting similar heterotrophy for all treatments. (Figure S8, Paper IV). However, our data indicate that the autotrophic pathway was more prominent for the Reference. Due to the high BP and the low abundance of edible AU phytoplankton in all treatments, we suggest that the major part of the energy went through the heterotrophic food web (1: bacteria, 2: flagellates, 3: ciliates, 4: mesozooplankton). In addition, we propose that a larger fraction of energy flow was channeled through the more efficient autotrophic food web (1: phytoplankton (<20 µm), 2: mesozooplankton) in the Reference than in the Level 1-3 treatments.

The additions of humic DOC increased concentrations of both dissolved total Hg and MeHg (Table 1; Figures S1 and S2, Paper IV). This result was likely a consequence of decreased partitioning of Hg to particulate matter (and mesocosm walls) due to formation of dissolved Hg complexes with thiol groups in the added DOC. Decreased photodemethylation of MeHg caused by water brownification also could have contributed to higher dissolved MeHg levels. While these observations corroborate with what was previously reported by Jonsson et al.73, it is important to note that high concentrations of humic DOM may also decrease MeHg bioavailability.27,62,64,76,140

The concentration of MeHg measured in seston showed increased from the smallest (20- 50 µm) to the largest fraction (>200 µm). This result is explained by their different trophic position (based on δ15N) in the pelagic food web (Table 1). The corresponding log MeHg BAF values ranged from 4.5 to 4.7 (20-50 µm) and from 5.7 to 6.0 (>200 µm )which is at the upper end of what has been found in previous studies.27,38,53,69,141 This suggests that changes in allochthonous DOC loadings can impact the range of BAFs in estuarine and coastal systems.

29

The addition of humic DOC in our study could contribute to large differences in MeHg bioavailability. We attempted to correct for these differences in a separate uptake incubation experiment with the green alga Selenastrum capricornutum. The highest uptake was observed for the Reference system and the lowest uptake for the Level 3. Our results indicated that across the large quantitative DOC ranges found in coastal to open ocean systems, the bioavailability of MeHg may have a large impact on BAFs, as previously observed by Schartup et al.27 The corrected BAF (BAF’) resulted in a maximum increase of 68% compared to the uncorrected BAF values for the Level 3 treatment (Figure 3, Paper IV). These results suggest that bioavailability of dissolved MeHg is affected by the presence of humic DOC and that this can further impact MeHg BAFs. Furthermore, the correction resulted in new statistical differences between the Level 3 and the Level 1-2 treatments in MeHg bioaccumulation, illustrating the contribution of differences in the food web processes.

The slope of the regression of log[MeHg] versus δ15N in a food web has been used to evaluate the biomagnification potential within a system.142 The biomagnification slopes within the lower food web (20−50 µm to >200 µm seston size fractions) ranged between 0.35-0.50 (Table 1). This is in the higher end of slopes found for higher levels of the food webs (plankton <200 µm and up to fish and birds, usually ranging from 0.2-0.4)143 and suggests efficient transfer of MeHg in the lower food web. The higher slope of the Reference system (Table 1) suggests a more efficient transfer of MeHg per trophic level than in the other treatments and indicates that higher accumulation of MeHg (due to more trophic levels) can be counteracted by a lower MeHg transfer efficiency per trophic level in a heterotrophic compared to autotrophic food web.

The results from our study and what was reported by Jonsson et al. suggest that an increase in BP proportion from 45% to 90% induced by enhanced humic DOC discharges may enhance MeHg bioaccumulation by one order of magnitude in zooplankton (Figure 8). Moreover, these results are relevant for coastal systems such as the Gulf of Bothnia, and other systems in the northern hemisphere, expected to be affected by increased humic DOC loadings following climate scenarios.

30

600

500

400

300

200

100

0 Reference Level 1 Level 2 Level 3 PP (mgC m-2 d-1) BP (mgC m-2 d-1)

Figure 7. Primary and bacterial biomass production rates averaged for the whole duration of the mesocosm experiment for the four treatments (Reference, Level 1, Level 2 and Level 3). Bars present the mean from three mesocosom replicate treatments with one SD error.

5.9

5.7 µm

5.5

5.3

5.1

4.9

4.7

4.5 0 10 20 30 40 50 60 70 80 90 100

log of average MeHg BAF in seston >50 seston in BAF MeHg average of log % Bacterial production of the total basal production

Figure 8. Average of logarithmic values of methylmercury (MeHg) bioaccumulation factor (BAF) in seston > 50 μm as a function of the % bacteria production rate of the total (bacteria + primary production) basal production. Red data points are average for each of the four different treatments in the current studies, and blue data points are reproduced from Jonsson et al. The data points are expressed as the mean of three mesocosom replicate treatments shown with one horizontal and one vertical standard error bar. 31

Table 2. Averaged (± SD) values of measured parameters for the whole duration of the experiment. For log BAF and log BAF’, the statistical tests were run on the data before the logarithm was applied. Reference Level 1 Level 2 Level 3 DOC mg L-1 3.79 ± 0.03A 4.48 ± 0.03 B 5.09 ± 0.07C 5.83 ± 0.06D Humic matter content µg L-1 12.85 ± 0.01 A 15.98 ± 0.15 B 17.84 ± 0.05 C 20.57 ± 0.16 D HgT pM 1.61 ± 0.11A 3.14 ± 0.06BC 3.60 ± 0.15CD 3.69 ± 0.23D MeHg pM 0.17 ± 0.01A 0.21 ± 0.04AB 0.28 ± 0.03B 0.35 ± 0.05c MeHg 20-50 µm pmol g-1 d.w. 8 ± 7 10 ± 2 9 ± 5 12 ± 1 MeHg 50-90 µm pmol g-1 d.w. 154 ± 33 138 ± 12 140 ± 35 147 ± 7 MeHg 90-200 µm pmol g-1 d.w. 118 ± 5 59 ± 34 78 ± 21 129 ± 38 MeHg >200 µm pmol g-1 d.w. 172 ± 32 138 ± 55 185 ± 70 203 ± 117 log BAF MeHg 20-50 µm 4.58 ± 0.34 4.68 ± 0.08 4.49 ± 0.18 4.52 ± 0.10 log BAF MeHg 50-90 µm 5.95 ± 0.14A 5.81 ± 0.05A 5.69 ± 0.15A 5.587 ± 0.003B log BAF MeHg 90-200 µm 5.84 ± 0.02A 5.40 ± 0.15B 5.43 ± 0.17B 5.55 ± 0.09AB log BAF MeHg >200 µm 6.00 ± 0.12 5.79 ± 0.09 5.80 ± 0.22 5.71 ± 0.32 log BAF’ MeHg 20-50 µm 4.58 ± 0.34 4.74 ± 0.08 4.53 ± 0.18 4.74 ± 0.10 log BAF’ MeHg 50-90 µm 5.95 ± 0.14 5.88 ± 0.05 5.73 ± 0.15 5.813 ± 0.003 log BAF’ MeHg 90-200 µm 5.84 ± 0.02A 5.47 ± 0.15B 5.47 ± 0.17B 5.78 ± 0.09A log BAF’ MeHg >200 µm 6.00 ± 0.12 5.85 ± 0.09 5.84 ± 0.22 5.93 ± 0.32 TDN µM 18.4 ± 0.5 A 22.8 ± 0.8 B 26.0 ± 0.3 C 30.7 ± 0.6 D TDP µM 0.45 ± 0.03 A 0.59 ± 0.07 B 0.66 ± 0.01 C 0.75 ± 0.01 D

- A A AB B NO3 µM 4.9 ± 0.2 4.9 ± 0.9 5.2 ± 0.4 5.7 ± 0.3

+ AB A AB B NH4 µM 0.37 ± 0.01 0.26 ± 0.03 0.32 ± 0.03 0.49 ± 0.03

3- PO4 µM 0.31 ± 0.02 0.32 ± 0.06 0.33 ± 0.03 0.35 ± 0.02 Chl a mg m-3 2.5 ± 0.2A 2.3 ± 0.2A 1.8 ± 0.2B 2.0 ± 0.1B pH 7.89 ± 0.06A 7.80 ± 0.07B 7.69 ± 0.05C 7.68 ± 0.03C Kd m-1 1.19 ± 0.02A 1.62 ± 0.02 A 1.92 ± 0.08 B 2.12 ± 0.01 C PP mg C m-2 d-1 81.6 ± 11.5A 59.3 ± 14.9B 47.6 ± 6.9BC 40.1 ± 1.2C BP mg C m-2 d-1 207 ± 12A 301 ± 87B 346 ± 148B 307 ± 32B ZP mg C m-2 d-1 2.8 ± 1.2 2.1 ± 0.3 1.9 ± 1.0 2.2 ± 1.3 BP/(PP+BP) 0.68 ± 0.03A 0.81 ± 0.02B 0.84 ± 0.05BC 0.87 ± 0.01C FWE % 0.9 ± 0.4A 0.5 ± 0.1B 0.4 ± 0.1B 0.5 ± 0.3AB δ15N 20-50 um ‰ 2.9 ± 0.1A 2.6 ± 0.4AB 2.6 ± 0.1B 2.4 ± 0.2AB δ15N 50-90 um ‰ 5.4 ± 0.3A 5.2 ± 0.2AB 4.8 ± 0.2B 5.1 ± 0.1AB δ15N 90-200 um ‰ 5.2 ± 0.2 5.1 ± 0.7 5.1 ± 0.4 5.1 ± 0.5

32

δ15N >200 um ‰ 5.9 ± 0.2 5.7 ± 0.3 5.6 ± 0.2 5.6 ± 0.2 δ13C 20-50 um ‰ -19.4 ± 0.8A -22.2 ± 1.7AB -22.7 ± 1.4AB -24.0 ± 1.5B δ13C 50-90 um ‰ -16.0 ± 0.8 -17.6 ± 2.2 -17.9 ± 2.8 -17.5 ± 1.4 δ13C 90-200 um ‰ -16.7 ± 1.1 -17.8 ± 2.6 -18.4 ± 1.5 -17.8 ± 1.5 δ13C >200 um ‰ -16.2 ± 1.1 -16.9 ± 2.3 -17.5 ± 2.8 -17.3 ± 2.1 k (δ15N) 0.49 ± 0.05 A 0.35 ± 0.05B 0.42 ± 0.06C 0.37 ± 0.04BC

The following abbreviations were used in the Table: Dissolved organic carbon (DOC), total mercury (HgT), methylmercury (MeHg), non-corrected bioaccumulation factor (BAF), corrected bioaccumulation factor (BAF’), total dissolved nitrogen (TDN), total dissolved phosphorous (TDP), - + 3- nitrate (NO3 ), ammonium (NH4 ), phosphate (PO4 ), chlorophyll a (Chl a), vertical diffuse light attenuation coefficient (Kd), primary production (PP), bacterial production (BP), zooplankton production (ZP), food web efficiency (FWE), stable δ15N and δ13C isotope signatures and linear regression slope between δ15N and BAF’. Bold italic text and different capital letters indicate statistically significant (p < 0.05, Repeated Measure One-way ANOVA or One-way ANOVA, Tukey HSD post-hoc test) differences between individual mesocosm treatments.

3.5. Future research remarks While this thesis work contributes to better understanding of processes and transformation pathways involved in Hg biogeochemical cycling and exposure to biota in systems with pronounced redoxclines such as in the Baltic Sea (Paper II), impacts of direct uptake of MeHg into the food web in these zones remain to be addressed. This is an important question to answer since a 2 – 6 times increase in MeHg levels has been observed in hypoxic compared to normoxic water. Considering the fact that hypoxic zones spread in coastal aquatic ecosystems worldwide,107 there is a potentially increased risk of MeHg entrance into the food web by direct consumption by organisms who either constantly or temporarily reside in these areas.

Moreover, new mechanistic information on uptake processes of MeHg in various phytoplankton species under different conditions, e.g. oxygen and DOC concentrations which affect MeHg speciation in the dissolved phase and availability, would improve assessment of MeHg potential for incorporation in the food web. Considering MeHg affinity to reduced sulfur groups and uncertainties in concentration and composition of thiol and sulfide ligands in aquatic environments, improved understanding of these

33

parameters would contribute as well to an overall better prediction and risk assessment of MeHg in top predators.

Lastly, our results from Paper IV indicate that the effect of a longer food web, i.e. higher number of trophic levels, on MeHg accumulation can be counteracted by lower MeHg transfer efficiency per trophic level in bacteria-based food webs compared to webs where phytoplankton is the base. Improved knowledge on MeHg transfer efficiency in the bottom section of aquatic food can thus be important and help to assess and predict the potential for MeHg transfer and MeHg levels in fish or other higher trophic levels. Therefore, further studies focusing on understanding these processes in both autotrophic and heterotrophic food webs are needed.

34

4. Conclusions To conclude, this thesis work contributes some important aspects to improve the current understanding regarding MeHg formation, uptake and bioaccumulation in coastal aquatic natural environments, where increased runoff of terrestrial OM is expected due to the predicted changes in climate, and to refine assessment and prediction of MeHg biogeochemical cycling and bioaccumulated levels in higher trophic level organisms in such systems.

Paper I provides information with regards to the effect of organic matter in coastal aquatic systems on MeHg formation and interannual variability. Namely, the impact of two investigated parameters (DOC and humic matter content) may be used to assess MeHg variability in this type of environments. Using concentration of DOC as a proxy, HgII methylation potential and overall MeHg levels may be estimated. Furthermore, due to the strong binding affinity of Hg to OM, the effect of humic matter content may be used to assess HgII availability and therefore its potential for methylation in the water column.

Important aspects of pelagic redoxclines, often present in coastal systems due to the expansion of oxygen depleted zones, on MeHg formation and measured concentrations were evaluated in Paper II. We found that net MeHg production is one of the main drivers of MeHg concentrations in normoxic, hypoxic and anoxic zones. Moreover, turbulent diffusion is an important factor along the redox gradient because it affects the methylation potential of HgII, depending on the Hg chemical speciation in water zones with different oxygen concentrations. To exemplify, HgII may occur in the form of stable complexes with thiol groups in the DOM pool or in mineral form as HgS(s) or be associated with Fe-DOM aggregates in the hypoxic zone, and when moving into the reductive dissolution zone, HgII is released from these complexes and becomes readily available for methylation. The formed MeHg could then be transferred by direct mixing of water into the hypoxic zone and enter the food web via organisms which actively feed in this zone.

Impact of molecular structure of different thiol ligands on uptake kinetics of MeHg in a model phytoplankton organism (green alga) was evaluated in Paper III. Lower MeHg association and adsorption rate constants were found in cases when the alga was exposed

35

to MeHg complexes with larger thiols with more complex and “branched” structure than in treatments with thiol ligands with simpler structures. In addition, concentrations of MeHg associated to the cells and highest association rate constant was found in the control without added thiols. The results demonstrated that both the chemical structure of MeHg complexes and thermodynamic stability are important controlling factors for MeHg interactions with the cell surface, but not necessarily for MeHg exchange across the membrane. These findings are in line with uptake mechanisms which involve formation of MeHg complexes with ligands located at the cell surface prior to internalization and point to the importance of thiol ligand composition in the environment for MeHg incorporation in the food web.

Paper IV provides new information regarding capacities for MeHg bioaccumulation in zooplankton in systems pronounced heterotrophy. Our results are in compliance with previous findings which showed high MeHg bioaccumulation in highly heterotrophic food webs, and demonstrate that MeHg bioaccumulation further increases at proportions of BP higher than 72%. Together with the results from a study by Jonsson et al.73, an increase in MeHg bioaccumulation in zooplankton by one order of magnitude is expected under conditions of enhanced proportions of BP from 45% to 90%. The enhancement in BP and its effect on the structure of the aquatic pelagic food webs, induced by higher discharges of humic DOC to coastal systems, are expected following climate change scenarios.

36

5. Acknowledgements First and foremost, I would like to thank my principal supervisor, professor Erik Björn, for choosing me among all other applicants for this position, and allowing me to be a part of this wonderful project. Erik, I would like to immensely thank you for being the best mentor a PhD student could possibly ask for, for making me a better employee, a better scientist, and most importantly, a better person. Thank you for helping me every step of the way, for always being willing to listen and help me overcome all issues that would arise. Thank you for being my inspiration for the past four and a half years, and for a good share of both happy and devastating memories which I will cherish forever. I can only hope that you do not regret your choice and that I have not been a disappointment or not lived up to your expectations.

Secondly, I would like to thank my assistant supervisors, Anne Soerensen and Agneta Andersson for their mentorship, valuable inputs and help with analyses and interpretation of piles among piles of acquired data. I would like to thank Anne for her help with the manuscripts and crucial comments which helped me develop better writing skills and for expanding my knowledge when it comes to data analyses. Thank you sincerely for a good time and good laughs we shared in the lab and during the sampling cruise 2016.

Further, I would like to acknowledge my co-authors, Živan Gojković and Sonia Brugel. Žiko, hvala najlepše na pomoći oko uzgajanja algi i što si samim svojim prisustvom sprečio da poludim ovde na dalekom severu. Hvala na druženju i na svemu kroz šta smo prošli, jedva čekam da te ponovo vidim! Drago mi je što smo uspeli da ostvarimo prijateljstvo potpuno nenadano i izvan Srbije. Sonia, thank you for keeping me sane during the exhausting 36 days of the mesocosm experiment, we all made a really good team. Thank you for all the cocktail nights and interesting conversations, for being my friend in lonely Umeå and for listening to my personal problems.

I would also like to thank my colleagues and ex-colleagues Caiyan, Wei, Eric, Ulf, Tower, Mareike, Lars, Richard, Temi, Liem, Sofi and most of all, Rain. Thank you Rain for being my close and sincere friend ever since I moved to Sweden, for all the help during the years

37

at work as well as in my private life. Thank you for visiting me and my family in Serbia! I would like to thank Lars for being my go-to person when I started my PhD studies and for all that he has taught me during the years we shared an office. I would also like to thank Richard for all his help in the lab, interesting conversations, mutual encouragement when no instruments would work properly and successful “tube fishing” from the malfunctioning TD. I would like to thank the rest of the Hg team for their help and support, interesting discussions and critical evaluations/feedbacks.

Finally, I would like to thank my loved ones, my family and my closest friends. Mama, tata, Teodora i Devone, hvala vam na svemu što ste učinili za mene i na podršci. Hvala na učestalim posetama, na pomoći u borbi protiv samoće i ludila. Pre svega, hvala vam na ljubavi. Babama, dedi, Dušanu i Ljilji takođe hvala na vremenu provedenom zajedno, bilo u Švedskoj, bilo u Srbiji. Hvala svemu što ste učinili za mene I hvala SVIMA što se ponosite mnome. Devonu hvala na velikoj pomoći tokom jako iscrpljujućeg perioda kada smo radili mesocosm eksperiment pošto bih najverovatnije umrla od gladi da njega nije bilo. Hvala najboljim prijateljima, Momi i Saši na druženju kao da smo i dalje komšije iako smo kilometrima udaljeni. Hvala na razgovorima i pomoći u najtežim trenucima.

Hvala i VOLIM VAS!

38

6. Literature

(1) Enrlich, H. L.; Newman, D. K. Geomicrobiology; 2008. (2) USEPA Agency for Toxic Subtances & Disease Registry, Mercury in Your Environment. 2015. (3) UNEP Sources, Emissions, Releases and Environmental Transport. UNEP Chemicals Branch, Geneva, Switzerland. 2013. (4) Selin, N. E. Global Biogeochemical Cycling of Mercury: A Review. Annu. Rev. Environ. Resour. 2009, 34 (1), 43–63. (5) Malm, O.; Pfeiffer, W. C.; Souza, C. M.; Reuther, R. Mercury Pollution Due to Gold Mining in the Madeira River Basin. Ambio 1990, 19 (1), 11–15. (6) Energy, U. S. D. o. M. E. C. R. http://www.fossil.energy.gov/programs/powersystems/pollutioncontrols/overvi ew_mercurycontrols.html. (7) USEPA EPA Mercury Study Report to Congress; Office of Air Quality and Standards and Office of Research and Developement Washington (DC). 1997. (8) Björn, E.; Larsson, T.; Lambertsson, L.; Skyllberg, U.; Frech, W. Recent Advances in Mercury Speciation Analysis with Focus on Spectrometric Methods and Enriched Stable Isotope Applications. Ambio 2007, 36 (6), 443–451. (9) Zhang, H.; Feng, X.; Larssen, T.; Qiu, G.; Vogt, R. D. In Inland China, Rice, Rather than Fish, Is the Major Pathway for Methylmercury Exposure. Environ. Heal. Perspect 2010, 118 (9), 1183. (10) Qiu, G.; Feng, X.; Li, P.; Wang, S.; Li, G.; Shang, L.; Fu, X. Methylmercury Accumulation in Rice (Oryza Sativa L.) Grown at Abandoned Mercury Mines in Guizhou, China. J. Agric. Food. Chem. 2008, 56 (7), 2465–2468. (11) Templeton, D. M.; Ariese, F.; Cornelis, R.; Danielsson, L. G.; Muntau, H.; Van Leeuwen, H. P.; Lobinski, R. Guidelines for Terms Related to Chemical Speciation and Fractionation of Elements. Definitions, Structural Aspects, and Methodological Approaches (IUPAC Recommendations 2000). Pure Appl. Chem. 2000, 72 (8), 1453–1470. (12) Agency, U. S. E. P., National Listing of Fish Advisories (EPA Fact Sheet EPA-823-F- 09-007). 2008. (13) Jiang, G. B.; Shi, J. B.; Feng, X. B. Mercury Pollution in China. Environ. Sci. Technol. 2006, 40 (12), 3672–3678. (14) Hakanson, L.; Nilsson, A.; Andersson, T. MERCURY IN FISH IN SWEDISH LAKES. Environ. Pollut. 1988, 49 (2), 145–162. (15) Kurland, T.; Faro, S. N.; Siedler, H. Minamata Disease. The Outbreak of a 39

Neurologic Disorder in Minamata, Japan, and Its Relationship to the Ingestion of Seafood Contaminated by Mercuric Compounds. World Neurol. 1960, 1 (5), 370– 395. (16) Harada, M. Minamata Disease: Methylmercury Poisoning in Japan Caused by Environmental Pollution. Crit. Rev. Toxicol 1995, 25 (1), 1–24. (17) Ackefors, H. Mercury Pollution in Sweden with Special Reference to Conditions in the Water Habitat. Proc. R. Soc. London. Ser. B, Biol. Sci. 1971, 177 (1048), 365–387. (18) EPA-FDA Fish Advice: Technical Information https://www.epa.gov/fish-tech/epa- fda-fish-advice-technical-information. (19) Bellanger, M.; Pichery, C.; Aerts, D.; Berglund, M.; Castaño, A.; Cejchanová, M.; Crettaz, P.; Davidson, F.; Esteban, M.; Fischer, M. E. Economic Benefits of Methylmercury Exposure Control in Europe: Monetary Value of Neurotoxicity Prevention. Environ. Heal. 2013, 12 (3). (20) Compeau, G. C.; Bartha, R. Sulfate-Reducing Bacteria: Principal Methylators of Mercury in Anoxic Estuarine Sediment. Appl. Environ. Microbiol. 1985, 50 (2), 498–502. (21) Kerin, E. J.; Gilmour, C. C.; Roden, E.; Suzuki, M. T.; Coates, J. D.; Mason, R. P. Mercury Methylation by Dissimilatory Iron-Reducing Bacteria. Appl. Environ. Microbiol. 2006, 72 (12), 7919–7921. (22) Gilmour, C. C.; Podar, M.; Bullock, A. L.; Graham, A. M.; Brown, S. D.; Somenahally, A. C.; Johs, A.; Hurt, R. A.; Bailey, K. L.; Elias, D. A. Mercury Methylation by Novel Microorganisms from New Environments. Environ. Sci. Technol. 2013, 47 (20), 11810–11820. (23) Parks, J. M.; Johs, A.; Podar, M.; Bridou, R.; Hurt, R. A.; Smith, S. D.; Tomanicek, S. J.; Qian, Y.; Brown, S. D.; Brandt, C. C.; et al. The Genetic Basis for Bacterial Mercury Methylation. Science. 2013, 339 (6125), 1332–1335. (24) Yu, R.-Q.; Reinfelder, J. R.; Hines, M. E.; Barkay, T.-. Mercury Methylation by the Methanogen Methanospirillum Hungatei. Appl. Environ. Microbiol. 2013, 79 (20), 6325–6330. (25) Bravo, A. G.; Peura, S.; Buck, M.; Ahmed, O.; Mateos-Rivera, A.; Herrero Ortega, S.; Schaefer, J. K.; Bouchet, S.; Tolu, J.; Björn, E.; et al. Methanogens and Iron- Reducing Bacteria: The Overlooked Members of Mercury-Methylating Microbial Communities in Boreal Lakes. Appl. Environ. Microbiol. 2018, 84 (23), e01774-18. (26) Monperrus, M.; Tessier, E.; Amoroux, D.; Leynaert, A.; Huonnic, P.; Donard, O. F. X. Mercury Methylation, Demethylation and Reduction Rates in Coastal and Marine Surface Waters of the Mediterranean Sea. Mar. Chem. 2007, 107 (1), 49–63. (27) Schartup, A. T.; Ndu, U.; Balcom, P. H.; Mason, R. P.; Sunderland, E. M. Contrasting Effects of Marine and Terrestrially Derived Dissolved Organic Matter 40

on Mercury Speciation and Bioavailability in Seawater. Environ. Sci. Technol. 2015, 49 (10), 5965–5972. (28) Gascon Diez, E.; Loizeau, J.-L.; Cosio, C.; Bouchet, S.; Adatte, T.; Amoroux, D.; Bravo, A. G. Role of Settling Particles on Mercury Methylation in the Oxic Water Column of Freshwater Systems. Environ. Sci. Technol. 2016, 50 (21), 11672– 11679. (29) Heimbürger, L. E.; Cossa, D.; Marty, J. C.; Migon, C.; Averty, B.; Dufour, A.; Ras, J. Methyl Mercury Distributions in Relation to the Presence of Nano- and Picophytoplankton in an Oceanic Water Column (Ligurian Sea, North-Western Mediterranean). Geochim. Cosmochim. Acta 2010, 74 (19), 5549–5559. (30) Sunderland, E. M.; Krabbenhoft, D. P.; Moreau, J. W.; Strode, S. A.; Landing, W. M. Mercury Sources, Distribution, and Bioavailability in the North Pacific Ocean: Insights from Data and Models. Global Biogeochem. Cycles 2009, 23 (2). (31) Münster, U.; Chróst, R. J. Origin, Composition, and Microbial Utilization of Dissolved Organic Matter. Aquat. Microb. Ecol. 1990, 8–46. (32) Ndu, U.; Mason, R. P.; Zhang, H.; Lin, S.; Visscher, P. T. Effect of Inorganic and Organic Ligands on the Bioavailability of Methylmercury as Determined by Using a Mer-Lux Bioreporter. Appl. Environ. Microbiol. 2012, 78 (20), 7276–7282. (33) Rabenstein, D. L. The Aqueous Solution Chemistry of Methylmercury and Its Complexes. Acc. Chem. Res. 1978, 11 (3), 100–107. (34) Berthon, G. Critical Evaluation of Stability Constants of Metal Complexes of Amino Acids with Polar Side Chains (Technical Report). Pure Appl. Chem. 1995, 67 (7), 1117–1240. (35) Ravichandran, M. Interactions between Mercury and Dissolved Organic Matter - A Review. Chemosphere 2004, 55 (3), 319–331. (36) Skyllberg, U. Competition among Thiols and Inorganic Sulfides and Polysulfides for Hg and MeHg in Wetland Soils and Sediments under Suboxic Conditions: Illumination of Controversies and Implications for MeHg Net Production. J. Geophys. Res. 2008, 113, 1–14. (37) Drott, A.; Lambertsson, L.; Björn, E.; Skyllberg, U. Importance of Dissolved Neutral Mercury Sulfides for Methyl Mercury Production in Contaminated Sediments. Environ. Sci. Technol. 2007, 41 (7), 2270–2276. (38) Hammerschmidt, C. R.; Finiguerra, M. B.; Weller, R. L.; Fitzgerald, W. F. Methylmercury Accumulation in Plankton on the Continental Margin of the Northwest Atlantic Ocean. Environ. Sci. Technol. 2013, 47 (8), 3671–3677. (39) Soerensen, A. L.; Schartup, A. T.; Gustafsson, E.; Gustafsson, B. G.; Undeman, E.; Björn, E. Eutrophication Increases Phytoplankton Methylmercury Concentrations in a Coastal Sea—A Baltic Sea Case Study. Environ. Sci. Technol. 2016, 50 (21), 11787–11796.

41

(40) Regnell, O.; Hammar, T.; Halgée, A.; Troedsson, B. Effects of Anoxia and Sulfide on Concentrations of Total and Methyl Mercury in Sediment and Water in Two Hg-Polluted Lakes. Can. J. Fish. Aquat. Sci. 2001, 58 (3), 506–517. (41) Pakhomova, S.; Veiteberg Braaten, H. F.; Yakushev, E.; Skei, J. Biogeochemical Consequences of an Oxygenated Intrusion into an Anoxic Fjord. Geochem. Trans. 2014, 15 (1), 5. (42) Deutsch, B.; Alling, V.; Humborg, C.; Korth, F.; Mörth, C. M. Tracing Inputs of Terrestrial High Molecular Weight Dissolved Organic Matter within the Baltic Sea Ecosystem. Biogeosciences 2012, 9 (11), 4465–4475. (43) Ripszam, M.; Paczkowska, J.; Figueira, J.; Veenaas, C.; Haglund, P. Dissolved Organic Carbon Quality and Sorption of Organic Pollutants in the Baltic Sea in Light of Future Climate Change. Environ. Sci. Technol. 2015, 49 (3), 1445–1452. (44) Hoikkala, L.; Lahtinen, T.; Perttila, M.; Lignell, R. Seasonal Dynamics of Dissolved Organic Matter on a Coastal Salinity Gradient in the Northern Baltic Sea. Cont. Shelf Res. 2012, 45, 1–14. (45) Zweifel, U. L.; Wikner, J.; Hagström, Å.; Lundberg, E.; Norrman, B. Dynamics of Dissolved Organic Carbon in a Coastal Ecosystem. Limnol. Oceanogr. 1995, 40 (2), 299–305. (46) Conley, D. J.; Carstensen, J.; Aigars, J.; Axe, P.; Bonsdorff, E.; Eremina, T.; Haahti, B. M.; Humborg, C.; Jonsson, P.; Kotta, J.; et al. Hypoxia Is Increasing in the Coastal Zone of the Baltic Sea. Environ. Sci. Technol. 2011, 45 (16), 6777–6783. (47) Diaz, R. J.; Rosenberg, R. Spreading Dead Zones and Consequences for Marine Ecosystems. Science. 2008, 321 (5891), 926–929. (48) Hansson, M.; Andersson, L.; Axe, P. Areal Extent and Volume of Anoxia and Hypoxia in the Baltic Sea, 1960-2011. (49) Cossa, D.; Coquery, M. The Mediterranean Mercury Anomaly, a Geochemical or a BiologocalIssue. Mediterr. Sea 2005, 177–208. (50) Lamborg, C. H.; Yiǧiterhan, O.; Fitzgerald, W. F.; Balcom, P. H.; Hammerschmidt, C. R.; Murray, J. Vertical Distribution of Mercury Species at Two Sites in the Western Black Sea. Mar. Chem. 2008, 111 (1–2), 77–89. (51) Watras, C. J.; Back, R. C.; Halvorsen, S.; Hudson, R. J. M.; Morrison, K. A.; Wente, S. P. Bioaccumulation of Mercury in Pelagic Freshwater Food Webs. Sci. Total Environ. 1998, 219 (2–3), 183–208. (52) Pickhardt, P. C.; Fisher, N. S. Accumulation of Inorganic and Methylmecury by Freshwater Phytoplankton in Two Contrasting Water Bodies. Environ. Sci. Technol. 2007, 41 (1), 125–131. (53) Gosnell, K. J.; Mason, R. P. Mercury and Methylmercury Incidence and Bioaccumulation in Plankton from the Central Pacific Ocean. Mar. Chem. 2015, 177, 772–780.

42

(54) Moye, H. A.; Miles, C. J.; Phlips, E. J.; Sargent, B.; Merritt, K. K. Kinetics and Uptake Mechanisms for Monomethylmercury between Freshwater Algae and Water. Environ. Sci. Technol. 2002, 36 (16), 3550–3555. (55) Miles, C. J.; Moye, H. A.; Phlips, E. J.; Sargent, B. Partitioning of Monomethylmercury between Freshwater Algae and Water. Environ. Sci. Technol. 2001, 35 (21), 4277–4282. (56) Reinfelder, J. R.; Fisher, N. S. The Assimilation of Elements Ingested by Marine Copepods. Science. 1991, 251 (4995), 794–796. (57) Campbell, P. G. C.; Errécalde, O.; Fortin, C.; Hiriart-Baer, V. P.; Vigneault, B. Metal Bioavailability to Phytoplankton--Applicability of the Biotic Ligand Model. Comp. Biochem. Physiol. C. Toxicol. Pharmacol. 2002, 133 (1–2), 189–206. (58) Le Faucheur, S.; Campbell, P. G. C.; Fortin, C.; Slaveykova, V. I. Interactions between Mercury and Phytoplankton: Speciation, Bioavailability, and Internal Handling. Environ. Toxicol. Chem. 2014, 33 (6), 1211–1224. (59) Dranguet, P.; Flück, R.; Regier, N.; Cosio, C.; Le Faucheur, S.; Slaveykova, V. I. Towards Mechanistic Understanding of Mercury Availability and Toxicity to Aquatic Primary Producers. Chim. Int. J. Chem. 2014, 68 (11), 799–805. (60) Niyogi, S.; Wood, C. M. Biotic Ligand Model, a Flexible Tool for Developing Site- Specific Water Quality Guidelines for Metals. Environ. Sci. Technol. 2004, 38 (23), 6177–6192. (61) Slaveykova, V. I.; Wilkinson, K. J. Predicting the Bioavailability of Metals and Metal Xomplexes: Critical Review of the Biotic Ligand Model. Environ. Chem. 2005, 2 (1), 9–24. (62) Luengen, A. C.; Fisher, N. S.; Bergamaschi, B. A. Dissolved Organic Matter Reduces Algal Accumulation of Methylmercury. Environ. Toxicol. Chem. 2012, 31 (8), 1712–1719. (63) Mason, R. P.; Reinfelder, J. R.; Morel, F. M. M. Uptake, Toxicity and Trophic Transfer of Mercury in a Coastal Diatom. Environ. Sci. Toxicol. 1996, 30 (6), 1835–1845. (64) Gorski, P. R.; Armstrong, D. E.; Hurley, J. P.; Krabbenhoft, D. P. Influence of Natural Dissolved Organic Carbon on the Bioavailability of Mercury to a Freshwater Alga. Environ. Pollut. 2008, 154 (1), 116–123. (65) Gorski, P. R.; Armstrong, D. E.; Hurley, J. P.; Shafer, M. M. Speciation of Aqueous Methylmercury Influences Uptake by a Freshwater Alga (Selenastrum Capricornutum). Environ. Toxicol. Chem. 2006, 25 (2), 534–540. (66) Lee, C. S.; Fisher, N. S. Methylmercury Uptake by Diverse Marine Phytoplankton. Limnol. Oceanogr. 2016, 61 (5), 1626–1639. (67) Lee, C. S.; Fisher, N. S. Bioaccumulation of Methylmercury in a Marine Diatom and the Influence of Dissolved Organic Matter. Mar. Chem. 2017, 197, 70–79.

43

(68) Davies, A. G. An Assessment of the Basis of Mercury Tolerance in Dunaliella Tertiolecta. J. Mar. Biol. Assoc. United Kingdom 1983, 56 (1), 39–57. (69) Gosnell, K. J.; Balcom, P. H.; Tobias, C. R.; Gilhooly, W. P.; Mason, R. P. Spatial and Temporal Trophic Transfer Dynamics of Mercury and Methylmercury into Zooplankton and Phytoplankton of Long Island Sound. Limnol. Oceanogr. 2017, 62 (3), 1122–1138. (70) Watras, C. J.; Bloom, N. S. Mercury and Methylmercury, in Individual Zooplankton: Implications for Bioaccumulation. Limnol. Oceanogr. 1992, 37 (6). (71) Wu, P.; Kainz, M. J.; Bravo, A. G.; Åkerblom, S.; Sonesten, L.; Bishop, K. The Importance of Bioconcentration into the Pelagic Food Web Base for Methylmercury Biomagnification: A Meta-Analysis. Sci. Total Environ. 2019, 646, 357–367. (72) Kidd, K.; Clayden, M.; Jardine, T. Bioaccumulation and Biomagnification of Mercury through Food Webs; 2011. (73) Jonsson, S.; Andersson, A.; Nilsson, M. B.; Skyllberg, U.; Lundberg, E.; Schaefer, J. K.; Åkerblom, S.; Björn, E. Terrestrial Discharges Mediate Trophic Shifts and Enhance Methylmercury Accumulation in Estuarine Biota. Sci. Adv. 2017, 3 (1), e1601239. (74) Karimi, R.; Chen, C.; Pickhardt, P.; Fisher, N.; Folt, C. Stoichiometric Controls of Mercury Dilution by Growth. (75) Todorova, S.; Driscoll, C. T.; Matthews, D.; Effler, S. Zooplankton Community Changes Confound the Biodilution Theory of Methylmercury Accumulation in a Recovering Mercury-Contaminated Lake. Environ. Sci. Technol. 2015, 49 (7), 4066–4071. (76) Soerensen, A. L.; Schartup, A. T.; Skrobonja, A.; Björn, E. Organic Matter Drives High Interannual Variability in Methylmercury Concentrations in a Subarctic Coastal Sea. Environ. Pollut. 2017, 229, 531–538. (77) Samuelsson, K.; Berglund, J.; Andersson, A. Factors Structuring the Heterotrophic Flagellate and Ciliate Community along a Brackish Water Primary Production Gradient. J. Plankton Res. 2006, 28 (4), 345–349. (78) EPA, U.Method 1631, Revision E: Mercury in Water by Oxidation, Purge and Trap, and Cold Vapor Atomic Fluorescence Spectrometry. 2002. (79) Lambertsson, L.; Björn, E. Validation of a Simplified Field-Adapted Procedure for Routine Determinations of Methyl Mercury at Trace Levels in Natural Water Samples Using Species-Specific Isotope Dilution Mass Spectrometry. Anal. Bioanal. Chem. 2004, 380 (7–8), 871–875. (80) Munson, K. M.; Babi, D.; Lamborg, C. H. Determination of Monomethylmercury from Seawater with Ascorbic Acid-Assisted Direct Ethylation. Limnol. Oceanogr. Methods 2014, 12 (1), 1–9.

44

(81) Jonsson, S.; Skyllberg, U.; Nilsson, M. B.; Lundberg, E.; Andersson, A.; Björn, E. Differentiated Availability of Geochemical Mercury Pools Controls Methylmercury Levels in Estuarine Sediment and Biota. Nat. Commun. 2014, 5, 4624. (82) Qvarnström, J.; Frech, W. Mercury Species Transformations during Sample Pre- Treatment of Biological Tissues Studied by HPLC-ICP-MS. J. Anal. At. Spectrom. 2002, 17 (11), 1486–1491. (83) Rodriguez-Gonzalez, P.; Bouchet, S.; Monperrus, M.; Tessier, E.; Amoroux, D. In Situ Experiments for Element Species-Specific Environmental Reactivity of Tin and Mercury Compounds Using Isotopic Tracers and Multiple Linear Regression. Environ. Sci. Pollut. Res. 2013, 20 (3), 1269–1280. (84) HELCOM Combine, 2014. Manual for Marine Monitoring in the COMBINE Programme of HELCOM http://helcom.fi/action-areas/monitoring- and- assessment/manuals-and-guidelines/combine-manual. (85) SMHI (2018), SMHI öppna data (data downloaded 2016), edited, SMHI www.smhi.se/klimatdata/oceanografi/havsmiljodata. (86) Polatajko, A.; Banaś, B.; Encinar, J. R.; Szpunar, J. Investigation of the Recovery of Selenomethionine from Selenized Yeast by Two-Dimensional LC – ICP MS. Anal. Bioanal. Chem. 2005, 381 (May), 844–849. (87) Bischoff, H. W.; Bold, H. C. Phycological Studies IV. Some Soil Algae from Enchanted Rock and Related Algal Species, University of Texas Publication No. 6318; 1963. (88) Scherholz, M. L.; Curtis, W. R. Achieving PH Control in Microalgal Cultures through Fed-Batch Addition of Stoichiometrically-Balanced Growth Media Achieving PH Control in Microalgal Cultures through Fed-Batch Addition of Stoichiometrically-Balanced Growth Media. BMC Biotechnol. 2013, 13 (1), 39. (89) Wang, J.; Curtis, W. R. Proton Stoichiometric Imbalance during Algae Photosynthetic Growth on Various Nitrogen Sources : Toward Metabolic PH Control. J. Appl. Phycol. 2016, 28 (1), 43–52. (90) Båmstedt, U.; Larsson, H. An Indoor Pelagic Mesocosm Facility to Simulate Multiple Water-Column Characteristics. Int. Aquat. Res. 2018, 13–29. (91) Snell, J. P.; Stewart, I. I.; Sturgeon, R. E.; Frech, W. Species Specific Isotope Dilution Calibration for Determination of Mercury Species by Gas Chromatography Coupled to Inductively Coupled Plasma- or Furnace Atomisation Plasma Ionisation-Mass Spectrometry. J. Anal. At. Spectrom. 2000, 15 (12), 1540–1545. (92) Chiasson-Gould, S. A.; Blais, J. M.; Poulain, A. J. Dissolved Organic Matter Kinetically Controls Mercury Bioavailability to Bacteria. Environ. Sci. Technol. 2014, 48 (6), 3153–3161. (93) Graham, A. M.; Aiken, G. R.; Gilmour, C. C. Effect of Dissolved Organic Matter 45

Source and Character on Microbial Hg Methylation in Hg-S-DOM Solutions. Environ. Sci. Technol. 2013, 47 (11), 5746–5754. (94) Bravo, A. G.; Bouchet, S.; Tolu, J.; Björn, E.; Mateos-Rivera, A.; Bertilsson, S. Molecular Composition of Organic Matter Controls Methylmercury Formation in Boreal Lakes. Nat. Commun. 2017, 8, 14255. (95) Kim, M.; Han, S.; Gieskes, J.; Deheyn, D. D. Importance of Organic Matter Lability for Monomethylmercury Production in Sulfate-Rich Marine Sediments. Sci. Total Environ. 2011, 409 (4), 778–784. (96) Fitzgerald, W. F.; Lamborg, C. H.; Hammerschmidt, C. R. Marine Biogeochemical Cycling of Mercury. Chem. Rev. 2007, 107 (2), 641–662. (97) Krabbenhoft, D. P.; Sunderland, E. M. Global Change and Mercury. Science. 2013, 341 (6153), 1457–1458. (98) Hoikkala, L.; Kortelainen, P.; Soinne, H.; Kuosa, H. Dissolved Organic Matter in the Baltic Sea. J. Mar. Syst. 2015, 142, 47–61. (99) Chakraborty, P.; Vudamala, K.; Coulibaly, M.; Ramteke, D.; Chennuri, K.; Lean, D. Reduction of Mercury (II) by Humic Substances—Influence of PH, Salinity of Aquatic System. Environemtnal Sci. Pollut. Res. 2015, 22 (14), 10529–10538. (100) Skogerboe, R. K.; Wilson, S. A. Reduction of Ionic Species by Fulvic Acid. Anal. Chem. 1981, 53 (2), 228–232. (101) Asmala, E.; Bowers, D. G.; Autio, R.; Kaartokallio, H.; Thomans, D. N. Qualitative Changes of Riverine Dissolved Organic Matter at Low Salinities Due to Flocculation. J. Geophys. Res. Biogeosciences 2014, 119 (10), 1919–1933. (102) Strååt, K. D. Simulating Transport and Understanding Future Fluxes of Organic Carbon in River Draining Into the Baltic Sea, Stockholm University, 2017. (103) Figueroa, D.; Rowe, O. F.; Paczkowska, J.; Legrand, C.; Andersson, A. Allochthonous Carbon—a Major Driver of Bacterioplankton Production in the Subarctic Northern Baltic Sea. Microb. Ecol. 2016, 71 (4), 789–801. (104) Schartup, A. T.; Balcom, P. H.; Soerensen, A. L.; Gosnell, K. J.; Calder, R. S. D.; Mason, R. P.; Sunderland, E. M. Freshwater Discharges Drive High Levels of Methylmercury in Arctic Marine Biota. Proc. Natl. Acad. Sci. U. S. A. 2015, 112 (38), 11789–11794. (105) Eklöf, K.; Fölster, J.; Sonesten, L.; Bishop, K. Spatial and Temporal Variation of THg Concentrations in Run-off Water from 19 Boreal Catchments, 2000–2010. Environ. Pollut. 2012, 164, 102–109. (106) Waples, J. S.; Nagy, K. L.; Aiken, G. R.; Ryan, J. N. Dissolution of Cinnabar (HgS) in the Presence of Natural Organic Matter. Geochim. Cosmochim. Acta 2005, 69 (6), 1575–1588. (107) Breitburg, D.; Levin, L. A.; Oschlies, A.; Grégoire, M.; Chavez, F. P.; Conley, D. J.; Garcon, V.; Gilbert, D.; Gutiérrez, D.; Isensee, K.; et al. Declining Oxygen in the 46

Global Ocean and Coastal Waters. Science. 2018, 359 (6371). (108) Hansson, D.; Eriksson, C.; Omstedt, A.; Chen, D. Reconstruction of River Runoff to the Baltic Sea, AD 1500–1995. Int. J. Climatol. 2011, 31 (5), 696–703. (109) Myrberg, K.; Lehmann, A. Topography, Hydrography, Circulation and Modelling of the Baltic Sea; 2013. (110) Kuss, J.; Cordes, F.; Mohrholz, V.; Nausch, G.; Naumann, M.; Kruger, S.; Schulz- bull, D. E. The Impact of the Major Baltic Inflow of December 2014 on the Mercury Species Distribution in the Baltic Sea. Environ. Sci. Technol. 2017, 51 (20), 11692–11700. (111) Pohl, C.; Fernández-Otero, E. Iron Distribution and Speciation in Oxic and Anoxic Waters of the Baltic Sea. Mar. Chem. 2012, 145–147, 1–15. (112) Staubwasser, M.; Schoenberg, R.; von Blanckenburg, F.; Kruger, S.; Pohl, C. Isotope Fractionation between Dissolved and Suspended Particulate Fe in the Oxic and Anoxic Water Column of the Baltic Sea. Biogeosciences 2013, 10 (1), 233–245. (113) Chadwick, S. P.; Babiarz, C. L.; Hurley, J. P.; Armstrong, D. E. Influences of Iron, Manganese, and Dissolved Organic Carbon on the Hypolimnetic Cycling of Amended Mercury. Sci. Total Environ. 2006, 368 (1), 177–188. (114) Taillefert, M.; Lienemann, C.-P.; Gaillard, J.-F.; Perret, D. Speciation, Reactivity, and Cycling of Fe and Pb in a Meromictic Lake. Geochim. Cosmochim. Acta 2000, 64 (2), 169–183. (115) Herlemann, D. P.; Labrenz, M.; Jugens, K.; Bertilsson, S.; Waniek, J. J.; Andersson, A. F. Transitions in Bacterial Communities along the 2000 Km Salinity Gradient of the Baltic Sea. ISME J. 2011, 5 (10), 1571–1579. (116) Bridou, R.; Monperrus, M.; Gonzalez, P. R.; Guyoneaud, R.; Amoroux, D. Simultaneous Determination of Mercury Methylation and Demethylation Capacities of Various Sulfate‐reducing Bacteria Using Species‐specific Isotopic Tracers. Environ. Toxicol. Chem. 2011, 30 (2), 337–344. (117) Drott, A.; Lambertsson, L.; Björn, E.; Skyllberg, U. Potential Demethylation Rate Determinations in Relation to Concentrations of MeHg, Hg and Pore Water Speciation of MeHg in Contaminated Sediments. Mar. Chem. 2008, 112 (1–2), 93–101. (118) Hollweg, T. A.; Gilmour, C. C.; Mason, R. P. Mercury and Methylmercury Cycling in Sediments of the Mid-Atlantic Continental Shelf and Slope. Limnol. Oceanogr. 2010, 55 (6), 2703–2722. (119) Davies, A. G. An Assessment of the Basis of Mercury Rolerance in Dunaliella Tertiolecta. J. Mar. Biol. Assoc. United Kingdom 1976, 56 (01), 39. (120) Loux, N. T. An Assessment of Thermodynamic Reaction Constants for Simulating Aqueous Environmental Monomethylmercury Speciation. Chem. Speciat.

47

Bioavailab. 2007, 19 (4), 183–196. (121) Reid, R. S.; Rabenstein, D. L. Nuclear Magnetic Resonance Studies of the Solution Chemistry of Metal Complexes . XVII . Formation Constants for the Complexation of Methylmercury by Sulfhydryl- Containing Amino Acids and Related Molecules. Can. J. Chem. 2011, 59 (10), 1505–1514. (122) Liem-Nguyen, V.; Skyllberg, U.; Björn, E. Thermodynamic Modeling of the Solubility and Chemical Speciation of Mercury and Methylmercury Driven by Organic Thiols and Micromolar Sulfide Concentrations in Boreal Wetland Soils. Environ. Sci. Technol. 2017, 51 (7), 3678–3686. (123) Cambridge Univ. Press, 2013. Intergovernmental Panel on Climate Change, Climate Change 2013: The Physical Science Basis. (124) IPCC. Summary for Policymakers. In Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel in Climate Change (Eds S. Solomon, D. Qin, M. Manning et Al.), Cambridge Universit. 2007. (125) Wikner, J.; Andersson, A. Increased Freshwater Discharge Shifts the Trophic Balance in the Coastal Zone of the Northern Baltic Sea. Glob. Chang. Biol. 2012, 18, 2509–2519. (126) HELCOM 2007. Climate Change in the Baltic Sea Area- HELCOM Thematic Assessment in 2007. Baltic Sea Environment Proceedings No. 111: 1–54. (127) Peterson, B. J.; Holmes, R. M.; McClelland, J. W.; Vörösmarty, C. J.; Lammers, R. B.; Shiklomanov, A. I.; Shiklomanov, I. A.; Rahmstorf, S. Increasing River Discharge to the Arctic Ocean. Science. 2002, 298, 2171–2173. (128) Fisher, J. A.; Jacob, D. J.; Soerensen, A. L.; Amos, H. M.; Steffen, A.; Sunderland, E. M. Riverine Source of Arctic Ocean Mercury Inferred from Atmospheric Observations. Nat. Geosci. 2012, 5, 499–504. (129) Andersson, A.; Brugel, S.; Paczkowska, J.; Rowe, O. F.; Figueroa, D.; Kratzer, S.; Legrand, C. Influence of Allochthonous Dissolved Organic Matter on Pelagic Basal Production in a Northerly Estuary. Estuar. Coast. Shelf Sci. 2018, 204, 225– 235. (130) Paczkowska, J.; Rowe, O. F.; Figueroa, D.; Andersson, A. Drivers of Phytoplankton Production and Community Structure in Nutrient-Poor Estuaries Receiving Terrestrial Organic Inflow. Mar. Environ. Res. 2019, 151, 104778. (131) Pace, M. L.; Cole, J. J.; Carpenter, S. R.; Kitchell, J. F.; Hodgson, J. R.; Van de Bogert, M. C.; Bade, D. L.; Kritzberg, E. S.; Bastviken, D. Whole-Lake Carbon-13 Additions Reveal Terrestrial Support of Aquatic Food Webs. Nature 2004, 427, 240–243. (132) Sandberg, J.; Andersson, A.; Johansson, S.; Wikner, J. Pelagic Food Web Structure and Carbon Budget in the Northern Baltic Sea: Potential Importance of Terrigenous Carbon. Mar. Ecol. Ser. 2004, 268, 13–29. 48

(133) Karlsson, J.; Jansson, M.; Jonsson, A. Similar Relationships between Pelagic Primary and Bacterial Production in Clearwater and Humic Lakes. Ecology 2002, 83, 2902–2910. (134) Degerman, R.; Lefébure, R.; Byström, P.; Båmstedt, U.; Larsson, S.; Andersson, A. Food Web Interactions Determine Denergy Transfer Efficiency and Top Consumer Responses to Inputs of Dissolved Organic Carbon. Hydrobiologia 2017, 805 (1), 131–146. (135) Berglund, J.; Mu, U.; Ba, U.; Andersson, A. Efficiency of a Phytoplankton-Based and a Bacteria-Based Food Web in a Pelagic Marine System. 2007, 52 (1), 121– 131. (136) Jansson, M.; Persson, L.; De Roos, A. M.; Jones, R. I.; Tranvik, L. J. Terrestrial Carbon and Intraspecific Size-Variation Shape Lake Ecosystems. Trends Ecol. Evol. 2007, 22, 316–322. (137) Hessen, D. O.; Andersen, T. Bacteria as a Source of Phosphorus for Zooplankton. Hydrobiologia 1990, 206, 217–223. (138) Brett, M. T.; Kainz, M. J.; Taipale, S. J.; Seshan, H. Phytoplankton, Not Allochthonous Carbon, Sustains Herbivorous Zooplankton Production. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 21197–21201. (139) Sommer, U.; Stibor, H.; Katechakis, A.; Sommer, F.; Hansen, T. Pelagic Food Web Configurations at Different Levels of Nutrient Richness and Their Implications for the Ratio Fish Production: Primary Production. Hydrobiologia 2002, 484, 11–20. (140) Schartup, A. T.; Thackray, C. P.; Qureshi, A.; Dassuncao, C.; Gillespie, K.; Hanke, A.; Sunderland, E. M. Climate Change and Overfishing Increase Neurotoxicant in Marine Predators. Nature 2019, 572, 648–650. (141) Chételat, J.; Richardson, M.; MacMillan, G.; Amyot, M.; Poulain, A. The Ratio of Methylmercury to Dissolved Organic Carbon in Water Explains Methylmercury Bioaccumulation across a Latitudinal Gradient from North-Temperate to Arctic Lakes. Environ. Sci. Technol. 2017, 52 (1), 79–88. (142) Yoshinaga, J.; Suzuki, T.; Hongo, T.; Minagawa, M.; Ohtsuka, R.; Kawabe, T.; Inaoka, T.; Akimichi, T. Mercury Concentration Correlates with the Nitrogen Stable Isotope Ratio in the Animal Food of Papuans. Exotoxicology Environ. Saf. 1992, 24 (1), 37–45. (143) Lavoie, R. A.; Jardine, T.; Chumchal, M. M.; Kidd, K. A.; Campbell, L. M. Biomagnification of Mercury in Aquatic Food Webs: A Worldwide Meta-Analysis. Environ. Sci. Technol. 2013, 47 (23), 13385–13394.

49