The exposure and accumulation of seabird-derived contaminants and
genotoxicity in Collembola from Svalbard
Silje Marie Kristiansen
Master thesis in toxicology 60 credits
Department of Biosciences
Faculty of Mathematics and Natural Sciences UNIVERSITY OF OSLO
03.11.2017
II
© Silje Marie Kristiansen 2017
The exposure and accumulation of seabird-derived contaminants and genotoxicity in Collembola from Svalbard
Silje Marie Kristiansen http://www.duo.uio.no/
Print: Reprosentralen, Universitetet i Oslo
III This project was carried out in collaboration with the Norwegian Polar Institute (NPI) and the Norwegian Institute for Air Research (NILU).
IV
V
Acknowledgements
The work presented in this thesis was conducted at the Department of Biosciences at the University of Oslo (UiO), in collaboration with the Norwegian Polar Institute (NPI) and the Norwegian Institute for Air Research (NILU), Tromsø. It was carried out under supervision of main supervisor Katrine Borgå (UiO) and Hans Petter Leinaas (UiO), Ketil Hylland (UiO), Geir Wing Gabrielsen (NPI) and Dorte Herzke (NILU).
I want to thank my supervisors for your invaluable feedback, tips, suggestions, criticism and compliments during this process including fieldwork, lab-work, statistics and writing. Thank you, Katrine for being an inspirational role model and for taking time to give me thorough feedback as well as support. Ketil, thank you for your guidance in effect analyses, and helpful talks during encountered problems. Thanks to Hans Petter, for always keeping your door open, and for the specific knowledge I know have obtained about springtails. Thank you Geir, for allowing me to join the nest count in Ny-Ålesund, for funding my stay at the Summer School at RECETOX, and for your endless positivity. Dorte, thank you for making me feel very welcome at NILU, and for your much-appreciated specific feedback. A special thanks to Katrine and Hans Petter for trusting me enough to hire me as a scientific assistant before I had finished my Master’s degree, I very much appreciate the opportunity and experience.
I would like to thank the Norwegian Polar Institute for funding the chemical analyses performed at NILU, and the Research Council of Norway and Svalbard Science Forum, who financed the fieldwork through the Arctic Field Grant. This provided me with invaluable experience and knowledge. I also thank Jan Christensen’s Endowment for funding.
Thanks to Jana Vašíčková, who spent two intense months in the laboratory at UiO developing the comet assay and designing and performing the exposure experiment. Without you, the DNA damage in this thesis would have been at least 60%. I don’t know how many hours we spent together in the dark comet-room, but I am very grateful that you were there. Thank you for being my guide, and taking me to bat caves and up mountains in the Czech Republic.
I want to thank all the positive and including people at NILU, Tromsø, for all your help and guidance during the chemical analyses. I thank Dorte (again), Arntraut, Linda, and Silja for your excellent laboratory skills, and especially Mikael for your patience and educational instructions.
VI
Appreciations goes to Raoul Wolf and Heidi Sjursen for useful help and insight during the development of comet, Bjørnar Beylich at NIVA for useful info about the micronucleus test, and to Solveig and Christina at NPI for your assistance in the field at Stuphallet. I also thank the University Centre in Svalbard (UNIS) for allowing me to establish a temporary laboratory in a container during the summer of 2016.
I want to thank all the people in the Tox group at UiO for your supporting words and feedback. I want to thank Torben for taking time when I stop by with problems. Thanks to the wonderful people in the study room 4604, especially Ellen for your company in Longyearbyen. Special appreciation goes to Danny, thank you for always offering help and tips without the need to ask for it, and for giving me useful feedback on parts of my thesis.
The future students you will supervise will be very lucky.
I want to thank Reidun Sirevåg for reading almost my entire thesis and giving me valuable feedback with a non-toxicologist viewpoint.
Julie, thank you for being my biggest motivation and muse, and for showing genuine interest in all aspects of my life, no matter how small. I am so grateful that I got to write this thesis in your company.
I want to thank Silje for being my mentor since the first semester of biology and giving me my first feedback. Thank you, Malin for helping me with much-needed support in word and EndNote, and for your general loving support. Pernille, thank you so much for your amazing company during our stay in Tromsø, Longyearbyen and Ny-Ålesund. Thanks to Line for your supporting words during this last year, and to Martin for caring about my salt-intake. Heartfelt thanks go to Magnus, Ingvild and Inger, your company and support, especially in the last weeks have kept me sane.
I thank my wonderful friends Christina and Catherine for always thinking the best of me, and Erlend, Kamilla and my parents for your everlasting love and support.
Lastly, I would like to thank the thousands of springtails included in this thesis, hopefully the Arctic ecosystem will benefit from your contribution in the future.
VII
Sammendrag
Sjøfugler i Arktis har en høy trofisk posisjon i næringsnettet, og på grunn av biomagnifisering finner man høye konsentrasjoner av mange antropogene miljøgifter i deres blod og vev.
Dermed kan sjøfugl fungere som biologiske vektorer av miljøgifter fra hav til land. Tundraen under fuglefjell har høy tilførsel av næringsrik guano, og er karakterisert med rik vegetasjon.
Slike områder støtter en jordfauna der spretthaler (Collembola) er rike i antall og arter, og spiller en viktig rolle i jordøkosystem-prosesser som nedbrytning og i energikretsløpet. Målet med dette studiet var å fastslå eksponering, akkumulering og effekter av miljøgifter med opphav i sjøfugl på spretthaler.
To spretthalearter og deres habitat (jord/mose) ble samlet inn fra syv lokaliteter med høy, medium og lav sjøfuglpåvirkning, i området av Kongsfjorden og Krossfjorden på Svalbard.
Spretthaler og jord/mose ble analysert for en rekke organiske miljøgifter, kvikksølv (Hg), samt ratioer av stabile isotoper av karbon og nitrogen (henholdsvis δ13C og δ15N). I tillegg ble genotoksiske responser, DNA-trådbrudd og mikrokjernefrekvens, kvantifisert i spretthaler.
Metoden for å kvantifisere DNA-trådbrudd ble i forkant utviklet for spretthaleceller.
Sjøfuglpåvirkning (trofisk posisjon av arter og populasjonsstørrelse, indikert av δ15N) og miljøgiftkonsentrasjoner antydet å være høyere i jord/mose som ble prøvetatt nærmere koloniene (0–150 m), sammenliknet med lengre unna (250–400 m). Ingen assosiasjon ble imidlertid funnet ved sammenlikning mellom områder. Den totale mengden miljøgifter i jord/mose og spretthaler var dominert av kvikksølv (Hg). I spretthaler var den etterfulgt av heksaklorbenzen (HCB), polyklorerte bifenyler (PCBer) eller polybromerte difenyletere (PBDEer), og klordaner (CHLer). PFDoDA, hexaklorsykloheksaner (HCHer) og p,p’-diklor- difenylkloretylen (p,p’-DDE) hadde det minste bidraget til den totale mengden miljøgifter.
Ingen trend ble antydet for mengder i jord/mose. Miljøgiftkonsentrasjoner i spretthaler og jord/moser var ikke korrelert. DNA-trådbrudd var høyere i spretthaler fra områder med høyere sjøfuglpåvirkning, sammenliknet med de fra mindre påvirkede områder. Dette studiet rapporterer for første gang frekvensen av mikrokjerner i spretthaler, men ingen forskjell ble funnet i skadenivå mellom områder, eller arter. Genotoksiske effekter ble assosiert med sjøfuglpåvirkning og miljøgift-konsentrasjoner, og økte sammen med konsentrasjoner av lavere klorerte PCBer og CHLer. Cellenes sensitivitet til oksidativt stress var negativt korrelert med flere miljøgifter, inkludert høyere klorerte PCBer (6 og 7 kloratomer).
VIII
IX
Abstract
Arctic seabirds occupy high trophic positions in the food web, and due to biomagnification their blood and tissues contain high concentrations of many anthropogenic contaminants.
Seabirds may hence function as biovectors of contaminants from ocean to land. The tundra near bird cliffs is heavily influenced by nutrient-rich guano, and thus characterised by rich vegetation. Such areas support ground-living invertebrate communities, where springtails (Collembola) are diverse and abundant, and play a vital role in soil ecosystem processes such as decomposition and energy cycling. The purpose of this study was to determine the exposure, accumulation and effects of seabird-derived contaminants on Collembola.
Two Collembola species and soil/moss from their habitat were collected in seven sites of high, medium and low seabird influence in the Kongsfjorden and Krossfjorden area, Svalbard.
The samples were analysed for a wide range of organic contaminants, mercury (Hg), and stable isotope ratios of carbon and nitrogen (δ13C and δ15N, respectively). Collembola were also analysed for genotoxic responses. Prior to analysis, the comet assay for quantification of DNA fragmentation was successfully developed for collembolan cells.
Seabird influence (trophic position of species, and population size, indicated by δ15N) and contaminant concentrations seemed to be higher in soil/moss sampled closer to the bird cliffs (0–150 m), compared to further away (250–400 m). However, when comparing among sites, no association between δ15N and contaminant load was found. The total contaminant loads in substrate were dominated by Hg, while no trend was found for organic contaminants.
Contaminant patterns in Collembola were dominated by Hg, before hexachlorobenzene (HCBs), polychlorinated biphenyls (PCBs) or polybrominated diphenyl ethers (PBDEs) and chlordanes (CHLs). PFDoDA, hexachlorocyclohexanes (HCHs) and p,p’dichlorodiphenyl- dichloroethylene (p,p’-DDE) had the lowest contributions. Contaminant concentrations in Collembola and substrates were not correlated. DNA fragmentation was higher in Collembola from sites with higher seabird influence, compared to lower seabird influenced sites. This study is the first to report micronucleus frequency in Collembola, however, no differences were found among sites or between species. Genotoxic responses were associated with both δ15N and contaminant levels, increasing with concentrations of lower chlorinated PCBs and CHLs. The sensitivity to induced oxidative stress was negatively correlated with several contaminants, including higher chlorinated PCBs (6 and 7 Cl substituents).
X
XI
Abbreviations
∑ Sum
δ13C Stable isotope ratio of carbon δ15N Stable isotope ratio of nitrogen ANOVA Analysis of variance
As Arsenic
BSAF Biota-soil accumulation factors
C Carbon
C:N Carbon-to-nitrogen ratio
Cd Cadmium
CHL c-CD c-NC
Chlordanes cis chlordane cis nonachlor
Cu Copper
d.w. Dry weight
DCM DDT DNA
Dichloromethane
Dichlorodiphenyltrichloroethane Deoxyribonucleic acid
DMSO Dimethyl sulfoxide
GC-MS Gas chromatography-mass spectrometry H2O2 Hydrogen peroxide
HCB Hexachlorobenzene
HCH Hexachlorocyclohexane
HCl Hydrogen chloride
Hg HV
Mercury
Hypogastrura viatica
IFE Institute of Energy Technology ISTD
KOA
KOW
Internal standard
Octanol-air partition coefficient Octanol-water partition coefficient
l.w. Lipid weight
LC Liquid chromatography
LMP Low melting point
LOD Limit of detection
LOQ Limit of quantification
MA Megaphorura arctica
MN Micronucleus
mOsm Milli-osmoles
N n NA
Nitrogen
Number of samples Not available
XII
Na2SO4 Sodium sulphate
NaCl Sodium chloride
NH4OAc Ammonium acetate
NILU Norwegian Institute for Air Research
NRR Neutral Red Retention
OCPs Organochlorinated pesticides OCs
Oxy-CD
Organochlorinated pollutants Oxychlordane
PAH Polycyclic aromatic hydrocarbons
Pb Lead
PBDE Polybrominated diphenyl ethers PBS Phosphate-buffered saline PC1 First principal component PC2 Second principal component PCA
PCBs
Principal component analyses Polychlorinated biphenyls
PFASs Per- and polyfluoroalkyl substances PFCAs Perfluoroalkyl carboxylic acids POPs Persistent organic pollutants PP
p,p’-DDE
Polypropylene
p,p’-dichlorodiphenyldichloroethylene
RDA Redundancy analysis
ROS rpm
Reactive oxygen species Rounds per minute RSTD
SD
Recovery standard Standard deviation SIA Stable isotope analysis SPE Solid phase extraction SRM Standard reference material TC
t-CD
Total carbon trans chlordane TOC
t-NC
Total organic carbon trans nonachlor Tukey’s HSD
UHPLC-MS/MS
Tukey’s honest significant different post hoc test
Ultra-high pressure liquid chromatography triple-quadrupole mass spectrometry
UiO University of Oslo
ULMP Ultra low melting point
w.w. wet weight
Zn Zinc
XIII
Table of contents
Acknowledgements ... V Sammendrag ... VII Abstract ... IX Abbreviations ... XI
1 Introduction ...1
1.1 Seabirds as biovectors of nutrients and contaminants ...2
1.2 Tundra communities near seabird colonies ...5
1.2.1 Collembola (Springtails) ...5
1.3 Cellular and genetic toxicity ...6
1.3.1 Genotoxicity ...6
1.3.2 Lysosomal membrane stability ...7
1.4 Aim and objectives ...8
2 Materials and Methods ...11
2.1 Study species ...11
2.2 Study sites ...12
2.3 Field procedures ...16
2.4 Contaminant analyses ...20
2.4.1 Homogenisation and sample splitting ...20
2.4.2 Analysis of lipid-soluble persistent organic pollutants (POPs) ...20
2.4.3 Analysis of per- and polyfluoroalkyl substances (PFASs) ...25
2.4.4 Analyte quantification and quality assurance ...27
2.4.5 Analysis of mercury (Hg) ...29
2.5 Analysis of total carbon and nitrogen, and their stable isotope ratios ...30
2.6 Determination of dry matter in substrate samples ...31
2.7 Method development of genotoxic effect analyses ...31
2.7.1 Neutral Red Retention ...31
2.7.2 Comet assay ...32
2.7.3 Diuron exposure experiment ...35
2.8 Genotoxic effect analyses ...35
2.8.1 Cell extraction and sample splitting ...35
XIV
2.8.2 Comet assay ...36
2.8.3 Micronucleus test ...38
2.8.4 Quality assurance ...39
2.9 Data treatment and statistical analyses ...40
2.9.1 Data below the limit of detection and limit of quantification ...41
2.9.2 Contaminants ...43
2.9.3 Genotoxic effects ...46
2.9.4 Relationship between contaminant load and genotoxic effects ...47
3 Results ...48
3.1 Stable isotope signatures, total carbon, and total nitrogen ...48
3.2 Contaminant concentrations ...50
3.3 Contaminant patterns ...61
3.4 Bioaccumulation ...66
3.5 Genotoxic effects ...70
3.5.1 DNA fragmentation ...70
3.5.2 Micronucleus frequency ...72
3.6 Relationship between contaminant load and genotoxic effects ...73
4 Discussion ...75
4.1 Stable isotope signatures ...75
4.1.1 Stable isotope ratios of nitrogen ...75
4.1.2 Stable isotope ratios of carbon ...78
4.2 Contaminant concentrations and patterns ...79
4.2.1 Comparison of contaminant load between Collembola species ...89
4.3 Relationship between stable isotope ratios and contaminant loads ...90
4.4 Bioaccumulation ...91
4.5 Method development ...93
4.5.1 Neutral Red Retention (NRR) ...93
4.5.2 Comet assay ...94
4.6 Genotoxic effects ...96
4.6.1 DNA fragmentation ...96
4.6.2 Micronucleus frequency ...98
4.7 Relationship between contaminant load and genotoxic effects ...98
5 Conclusions ...101
XV
6 Future studies ...103
References ...106
Appendices ...122
Appendix A: Analytes ...122
Appendix B: Contaminant concentrations ...124
Appendix C: Dry matter in substrates ...128
Appendix D: Preparations of solutions ...129
Appendix E: Method development of the Neutral Red Retention assay ...131
Appendix F: Method development of the comet assay ...133
Appendix G: Diuron exposure experiment ...141
Appendix H: Equations ...144
Appendix I: Correlation between proportions of moss and Hg concentrations in substrate samples ...145
Appendix J: DNA fragmentation and micronucleus frequency ...146
XVI
1
1 Introduction
Despite long distances from major pollution sources, environmental contaminants of industrial and agricultural origin are found in abiotic and biotic components in Arctic ecosystems (Letcher et al., 2010; Muir and de Wit, 2010). Persistent organic pollutants (POPs) can undergo long-range environmental transport, and reach the Arctic from lower latitudes via atmospheric winds, ocean currents, northward flowing rivers, sea ice (Burkow and Kallenborn, 2000; MacDonald et al., 2000), and migrating animals (Blais et al., 2005;
Leat et al., 2013). POPs are resistant to degradation and have a high affinity to lipid or proteins, resulting in their bioaccumulation in biota. These chemicals are toxic and can elicit adverse effects in wildlife and humans (e.g endocrine or immune function), either directly or indirectly through metabolites produced via biotransformation or abiotic degradation (Letcher et al., 2010). International regulations such as the UNEP Stockholm Convention on POPs (implemented in 2004) (http://chm.pops.intl) have led to a decline in the levels of POPs in the Arctic (AMAP, 2015; Riget et al., 2011), but due to the persistent character of these compounds, this decline is slow and geographically scattered, and many of these pollutants are still detected at high levels in the Arctic (Hung et al., 2016; Letcher et al., 2010).
Compounds of concern in the Arctic include a range of compounds, such as polycyclic aromatic hydrocarbons (PAHs) and organohalogenated contaminants, which are categorised according to the type of halogen present (chlorine, bromine or fluorine). Important organochlorines (OCs) include: polychlorinated biphenyls (PCBs) and hexachlorobenzene (HCB); organochlorine pesticides (OCPs) such as dichlorodiphenyltrichloroethane (DDT) and its metabolites, hexachloro-cyclohexane (HCHs), chlordanes (CHLs) and mirex (AMAP, 2004). Both the chlorinated contaminants, and brominated such as polybrominated diphenyl ethers (PBDEs), are lipophilic and accumulate in lipid-rich tissues. They tend to biomagnify through food webs, resulting in high concentrations in tissues of organisms at high trophic positions (Borgå et al., 2001; Fisk et al., 2001). Semi-volatile and volatile pollutants, such as lower chlorinated PCBs, HCB, and HCHs are mainly transported from mid-latitudes to the Arctic via atmospheric currents as rapidly as within a few days (Burkow and Kallenborn, 2000; MacDonald et al., 2000). Fluorinated compounds, such as the family of per- and polyfluoroalkyl substances (PFASs), are amphiphilic and tend to bind to proteins (Jones et al., 2003). Compounds that are less volatile and more hydrophilic, such as some PFASs and some OCs, are likely to be transported via ocean currents (Li et al., 2002). In contrast to
2
atmospheric transport, oceanic transport of chemicals can take years to several decades before they reach the Arctic Ocean (Burkow and Kallenborn, 2000; MacDonald et al., 2000).
Therefore, the time from production and release to deposition in the Arctic varies for compounds depending on their physicochemical properties. In addition to the organic compounds, inorganic trace-elements, such as mercury (Hg) are also of concern in the Arctic.
The main transportation pathway of Hg to the Arctic is with atmospheric currents. Hg exists in both organic and inorganic forms, but bioaccumulate, and biomagnify, mostly in lipids as methylated mercury (CH3Hg, abbreviated to MeHg) (Ruus et al., 2015).
A large fraction of POPs end up in the ocean due to atmospheric deposition directly, or from land via rivers or surface run-off (Burkow and Kallenborn, 2000). Fewer compounds are therefore present, and at lower levels in the Arctic terrestrial ecosystem compared to the marine (Gamberg et al., 2005). The marine ecosystems also have longer food chains (Hop et al., 2002), and consequently higher biomagnification than terrestrial systems (Letcher et al., 2010). Consequently, most studies from the Arctic focus on the marine ecosystem, leaving knowledge gaps in the contamination of terrestrial environments (Muir and de Wit, 2010;
Riget et al., 2011; van den Brink et al., 2016).
1.1 Seabirds as biovectors of nutrients and contaminants
Migratory animals such as seabirds may function as biological vectors of contaminants from ocean to land due to their marine diet (Figure 1) (Blais et al., 2005; Evenset et al., 2004;
Roosens et al., 2007). Potentially, seabirds may transport contaminants to the Arctic from urban and industrial southern areas, but the geographical origin of contamination is uncertain as most studies on biotransport are on migratory species (e.g. Blais et al., 2005; Evenset et al., 2004), as well as data on the diet of seabirds are obtained during the breeding season (Gabrielsen, 2009). Seabirds are positioned high in the food chain, feeding on marine invertebrates and fish (Hobson, 1993), and thus reach high levels of organic and some inorganic contaminants (Borgå et al., 2004; Buckman et al., 2004; Hop et al., 2002). Tissues from Arctic seabirds contain Hg (Borgå et al., 2004; Jæger et al., 2009; Ruus et al., 2015), PBDEs (de Wit et al., 2006), PFASs (Verreault et al., 2005), and a broad spectrum of OCs, such as PCBs, DDTs, HCB, HCHs, CHLs, and mirex (Borgå et al., 2001; Borgå et al., 2007;
Borgå et al., 2005; Buckman et al., 2004). Many seabird species migrate to the Arctic in the early summer to breed and gather in large colonies of up to several thousand individuals,
3
before returning to their wintering areas in autumn (Frederiksen et al., 2016; Mallory et al., 2008; Mallory et al., 2017). Sites near bird cliffs receive large amounts of litter from the seabirds including guano, egg shells, regurgitated stomach oils, feathers, food remnants and carcasses (Stempniewicz, 1990; Zmudczynska-Skarbek et al., 2015a), which can serve as a significant contributor of industrial and agricultural contaminants to the tundra (Choy et al., 2010b). The biotransport of contaminants by seabirds to receiving ecosystems has been suggested to exceed the efficiency of atmospheric transport due to a combination of large, dense colonies, and the biomagnification of pollutants (Blais et al., 2007; Evenset et al., 2007a; Michelutti et al., 2009b).
Figure 1. Marine-derived anthropogenic contaminants that have biomagnified in seabird tissues may be transferred to the tundra below the breeding colonies via guano, feathers and egg shells. Thus, seabirds may function as biovectors of contaminants from ocean to land.
One way of measuring seabird influence on terrestrial systems is to use ratios of naturally occurring stable isotopes of nitrogen (15N/14N, hereafter referred to as δ15N) and carbon (13C/12C, hereafter referred to as δ13C). Stable isotope analysis (SIA) can trace the trophic position (δ15N) and dietary pattern (δ13C) of the studied organism. For δ15N, the principle behind the method relies on the trophic fractionation (isotopic separation between trophic levels) of nitrogen isotopes, with an enrichment of the heavier isotope (15N) compared to the lighter (14N) when moving up a food chain (DeNiro and Epstein, 1978; Peterson and Fry, 1987). The systematic enrichment of δ15N with each trophic level is on average for most vertebrates 3.4‰ (Post, 2002) and 2.4‰ for seabirds (Hobson, 1993). Guano and other
4
seabird inputs are enriched in δ15N compared to other local nitrogen (N) sources on the tundra, such as that fixed from the atmosphere (Blais et al., 2005; Wada et al., 1984). Thus is guano also potentially enriched in biomagnified contaminants (Borgå et al., 2012). Based on this, δ15N serves as a useful proxy for seabird influence on terrestrial habitats. The fractionation of δ13C is however, limited between trophic levels, and the isotopic ratio of carbon (C) differs between sources of different photosynthetic pathways, such as from C3 and C4 plants (Peterson and Fry, 1987; Rounick and Winterbourn, 1986). Thus, δ13C can provide information on the dietary source of C and can be used to separate between marine and terrestrial source (Kelly, 2000; Post, 2002).
The contamination of sites by seabirds has previously mainly been studied in freshwater lakes and ponds (e.g. Brimble et al., 2009; Evenset et al., 2004; Michelutti et al., 2009b). In the Canadian Arctic, strong correlations were found between the degree of seabird influence and concentrations of DDTs, HCB and PCBs in the sediments of ponds located in vicinity of seabird colonies, with the highest concentrations found in ponds with the highest seabird enrichments (i.e. highest d15N values) (Blais et al., 2005; Michelutti et al., 2009b). The same ponds were enriched with marine-derived trace elements; Hg (Blais et al., 2005), arsenic (As), cadmium (Cd) and zinc (Zn), linked to the influence of seabirds (Brimble et al., 2009).
Bjørnøya in Svalbard is considered a “hotspot” for contamination, with lake sediments containing high concentrations of PCBs and DDTs correlating with d15N values, as a consequence of seabird influence (Evenset et al., 2004). Elevated levels of chlorobornanes (CHBs), PBDEs, polybrominated biphenyls (PBBs) and polychlorinated napththalenes (PCNs) have also been detected, and these coincide with seabird input (Evenset et al., 2005).
The transfer of nutrients by seabirds to Arctic tundra is well documented (Mehlum and Gabrielsen, 1995; Sanchez-Pinero and Polis, 2000; Senstad, 1977; Zmudczynska-Skarbek et al., 2013). The Arctic tundra is commonly nutrient-poor with low biological productivity, but the input of marine-derived nutrients from seabird guano stimulates the biomass and diversity of vegetation underneath seabird colonies (Cocks et al., 1998; Gonzalez-Bergonzoni et al., 2017; Zmudczynska-Skarbek et al., 2013). As the tundra is heavily introduced with nutrients from the guano, the introduction of POPs could also potentially be substantial (Evenset et al., 2007a). Due to biotransformation of contaminants in seabird livers (e.g. Borgå et al., 2005), ornithogenic subsidies might also introduce metabolites of parent contaminants, which cannot be transported (or to a smaller degree) by abiotic transport pathways due to possessing
5
different physico-chemical properties, such as for hydroxylated PCBs (OH-PCBs) compared to PCBs (Tehrani and Van Aken, 2014). Tundra located near bird cliffs have been found impacted by organic contaminants and heavy metals transferred by seabirds, and declined with increasing distance away from colonies (Choy et al., 2010b; Godzik, 1991), as well as with decreasing degree of seabird influence (Choy et al., 2010b). In addition, radionuclides have been found on the soil below a bird cliff in Svalbard, indicated to be enriched by the presence of seabirds (Dowdall et al., 2005).
1.2 Tundra communities near seabird colonies
The Arctic tundra in the vicinity of seabird colonies supports a high diversity of invertebrate and vertebrate herbivores, predators and scavengers (Eide et al., 2005; Jakubas et al., 2008;
Zmudczynska-Skarbek et al., 2015a). Together with various microorganisms such as bacteria, ciliates and amoebae, soil invertebrates play a crucial role in nutrient and energy cycling, and decomposition processes in these ecosystems (Coulson et al., 2014; Hodkinson, 2013). The most essential components of soil invertebrate communities in the Arctic are mites, enchytraeid worms, spiders, “terrestrial” chironomid larvaes, and springtails (Hodkinson et al., 1998). Springtails (Collembola) are commonly the most abundant and diverse group (personal observation: Birkemoe and Leinaas, 2000).
1.2.1 Collembola (Springtails)
Collembola (Hexapoda) are micro-arthropods that are globally distributed, and contribute to the decomposition and other processes via their feeding activity (Rusek, 1998). Collembola are omnivorous and have a diverse diet, feeding on fungal hyphae, bacteria, algae, protozoa, dead vegetation, living plants, soil detritus, nematodes, and other microbiota (Hodkinson et al., 1994a; Hopkin, 1997; Rusek, 1998). By feeding on soil microbiota, they may also indirectly influence the microstructure and dynamics of the soil, stimulating microbial decomposition (Hopkin, 1997; Rusek, 1998). Partially decomposed collembolan faecal pellets increase the availability and surface area of organic matter, engaging further microbial and fungal decomposition, and the release of essential nutrients. The presence of Collembola also increase the mineralization of nutrients, such as N and phosphorous (P), and make them available for uptake by plants. Collembola may also disperse plant litter, fungal spores and bacteria by feeding in one soil area and excreting in another (Cragg and Bardgett, 2001;
6
Hopkin, 1997). They are also an important in linking the transfer of energy between food webs above and below ground, by being preyed upon by a wide variety of small arthropod predators (Hopkin, 1997; Rusek, 1998), and some bird species (Leinaas and Ambrose, 1999).
Despite their small size (0.2‒8 mm), Collembola have a high contribution to the total biomass of soil invertebrates and represent important components in Arctic tundra. Arctic ecosystems generally display low species diversity, but in ornithogenic substrates underneath and around seabird colonies, Collembola are often very abundant and species rich (Hodkinson et al., 1994a; Zmudczynska-Skarbek et al., 2012; Zmudczynska-Skarbek et al., 2015b). The density and biomass of Collembola have been documented to be up to 20 times higher in tundra areas influenced by seabirds, compared to areas absent of seabirds (Zmudczynska-Skarbek et al., 2012). Population sizes of Collembola in sites with seabird influence may reach almost 300 000 individuals per m-2 (Bengtson et al., 1974). As seabirds may contaminate the nutrient-rich habitats underneath their colonies, Collembola near bird cliffs are potentially exposed to anthropogenic contaminants via diffusion through their cuticles from the ambient environment, and through diet.
1.3 Cellular and genetic toxicity
The exposure to anthropogenic contaminants in natural populations of Collembola might have a range of effects on their physiology, behaviour and competition, and consequently vital rates, meaning the population size and composition (e.g. Crouau and Moia, 2006; Holmstrup et al., 2008; Schnug et al., 2014). Higher levels of biological organisation, such as populations or communities, are generally too complex to link cause with effect, and thus to link any responses with contaminant exposure. Studying the effects of contaminants on lower levels of biological organisation, such as the genetic material or cells, may therefore be easier to interpret findings.
1.3.1 Genotoxicity
The genetic material, deoxyribonucleic acid (DNA) is constantly transitioning between being a stable, double-stranded molecule and an unstable, intermediate structure, which is triggered by normal-functioning processes such as cell replication and cellular metabolism. Such cellular processes might be disrupted by the exposure to pollutants with genotoxic properties,
7
causing modifications to the DNA structure (DNA damage) (Shugart, 2000). Genotoxicants may directly lead to increased levels of DNA fragmentation or act indirectly by increasing the production of reactive oxygen species (ROS) (Crespo-Lopez et al., 2009; Fenstad et al., 2016;
Krøkje et al., 2006). ROS at low levels are vital for normal cell function, but if ROS production exceeds tolerable levels or if protective antioxidant levels are compromised, imbalances may occur in the oxidative status of cells (oxidative stress) (Halliwell and Gutteridge, 2008). Many DNA damage events can be attributed to oxidative stress, and one of the most hazardous lesions is double-strand breaks (Barzilai and Yamamoto, 2004). Oxidative stress is associated to exposure to compounds such as Hg and PCBs (Crespo-Lopez et al., 2009; Marabini et al., 2011), and an increase in DNA fragmentation has been associated with PCB metabolites (Srinivasan et al., 2001), and the metabolite of DDT; p,p’- dichlorodiphenyldichloroethylene (p,p’-DDE) (Binelli et al., 2008). On the other hand, high exposure to chemicals might lead to evolved tolerance or resistance for toxic mechanisms, including increased detoxification and decreased sensitivity to target site, as found in insects (Taylor, 1986). Such evolution leads to believe that organisms could evolve resistance to genotoxic mechanisms including oxidative stress, when exposed to oxidative stress inducing agents such as genotoxicants.
Some genotoxic chemicals may also cause breakage at a whole chromosomal level, called clastogenesis. Such DNA damage can also be quantified by the frequency of micronucleus present across a determined number of cells. A micronucleus represents a small structure of chromatin that have not integrated into the daughter nucleus during mitosis, and sit outside the nuclear membrane. Micronuclei result from either unrepaired fragmentation of the chromosome, a whole chromosome with a lag due to clastogens or spindle dysfunction, or apoptosis (Bolognesi and Hayashi, 2011; Heddle et al., 1991; UNEP/RAMOGE, 1999). An increased frequency of micronuclei have been linked to exposure to Hg (Boatti et al., 2017;
Gomez-Olivan et al., 2017; Pytharopoulou et al., 2013), PCBs (Marabini et al., 2011), PAHs, OCPs (Binelli et al., 2010) and p,p’-DDE (Binelli et al., 2008).
1.3.2 Lysosomal membrane stability
One endpoint of cytotoxicity is the decreased stability of lysosomal membranes, which has been linked to a range of contaminants, including heavy metals (Domouhtsidou et al., 2004), and specifically Hg (Leomanni et al., 2016). Lysosomes are the major degradative organelles
8
in eukaryotic cells, and degrade excess or damaged sub-cellular components such as proteins, lipids, nucleic acids and carbohydrates (Klionsky and Emr, 2000). The stability and normal physiological function of lysosomal membranes is an indicator of the cellular well-being in an organism. The measure of its function can establish the capacity of cellular processes during stress, such as exposure to pollutants (Lowe and Pipe, 1994). Exposure to pollutants enhance the production of ROS in lysosomes (Winston et al., 1996), which are as mentioned highly reactive with cellular components. Destabilization of the lysosomal membrane may result in the activation or release of hydrolytic enzymes, as well as ROS to the cytosol (UNEP/RAMOGE, 1999), causing further damage to cellular components.
The relationships between POPs and genotoxic effects in Collembola have been poorly studied (Marabini et al., 2011), and to the author’s knowledge, no peer-reviewed articles have been published on micronucleus frequency or lysosomal membrane stability in Collembola, and only one study is published on DNA fragmentation in Collembola (Cardoso et al., 2017).
1.4 Aim and objectives
The overarching aim of this study was to determine whether Arctic tundra communities, represented by Collembola, are exposed to seabird-derived contaminants, whether these chemicals accumulate, and if Collembola populations are affected in regard to genotoxicity by being in vicinity to seabird colonies. The present study included five objectives that followed a total of ten null and alternative hypotheses:
Objective i) To measure the degree of seabird influence (colony size and bird species present) on collembolan habitat (mixtures of soil and vegetation) and populations of Collembola.
H01) The isotope signature (δ15N and δ13C) in substrates and populations of Collembola does not differ between sites with high, moderate and low seabird influence.
HA1) The isotope ratio of nitrogen (δ15N) increases in substrates and in populations of Collembola with increasing seabird influence, while the isotope ratio of carbon (δ13C) reflects a terrestrial carbon source in all sites.
Objective ii) To quantify the concentrations and patterns of contaminants in substrates of collembolan habitat (mixture of soil and vegetation) and in populations of Collembola, in sites
9
of different colony size and species of birds.
H02) Contamination in the substrates does not differ between sites with high, moderate and low seabird influence, nor by distances from the base of the colony.
HA2) Bird species and the colony size is reflected in the tundra contaminant load, and the level of contamination in substrates increases with increasing seabird influence, and is higher closer to the base of the colony than farther away.
H03) Contaminant concentrations and isotope signatures (δ15N and δ13C) in populations of Collembola do not correlate with their respective substrates.
HA3) Contaminant concentrations and isotope signatures (δ15N and δ13C) in populations of Collembola correlate positively with those of their respective substrates.
H04) Contaminant concentrations are the same between two different Collembola species collected at the same site.
HA4) Contaminant concentrations differ between two different Collembola species collected at the same site.
H05) Contaminants do not differ in bioaccumulation in Collembola according to lipophilicity and persistence.
HA5) Contaminants with higher lipophilicity and persistence have a higher bioaccumulation potential in Collembola, compared to those with lower lipophilicity and persistence.
Objective iii) To establish and optimise methods that quantify genotoxic and cytotoxic biomarkers in collembolan cells.
Objective iv) To clarify whether populations of Collembola experience genotoxic stress (DNA fragmentation and micronucleus frequency), in sites of different colony size and species of birds.
H06) DNA fragmentation and micronucleus frequency in populations of Collembola do not differ between sites of high, moderate and low seabird influence.
10
HA6) DNA fragmentation and micronucleus frequency in populations of Collembola are higher in sites with higher seabird influence, compared to sites with lower seabird influence.
H07) The sensitivity of cells to induced oxidative stress in populations of Collembola does not differ between sites of high, moderate and low seabird influence.
HA7) The sensitivity of cells to induced oxidative stress in populations of Collembola is lower in sites with higher seabird influence, compared to sites with lower seabird influence.
H08) The two Collembola species collected in the same site do not differ in baseline DNA fragmentation, micronucleus frequency, or sensitivity to induced oxidative stress.
HA8) The two Collembola species collected in the same site differ in baseline DNA fragmentation, micronucleus frequency, and sensitivity to induced oxidative stress.
H09) DNA fragmentation and micronucleus frequency in Collembola are not correlated.
HA9) DNA fragmentation and micronucleus frequency in Collembola are positively correlated.
Objective v) To identify relationships between the contaminant concentrations in Collembola and stress responses.
H010) Baseline DNA fragmentation, micronucleus frequency, and sensitivity to induced oxidative stress in populations of Collembola do not correlate with concentrations of one or more contaminants in their tissues.
HA10) Baseline DNA fragmentation and micronucleus frequency in populations of Collembola are positive correlated, and sensitivity to induced oxidative stress in populations of Collembola negative correlated, with concentrations of one or more contaminants in their tissues.
11
2 Materials and Methods
2.1 Study species
Hypogastrura viatica (Tullberg 1872) (family of Hypogastruridae) and Megaphorura artica (Tullberg 1876), (family of Onychiuridae) were chosen as study species. They are abundant underneath bird cliffs in the Arctic (Hertzberg et al., 2000; Hodkinson et al., 1994a;
Zmudczynska-Skarbek et al., 2012), but their different habitats allowed the examination and comparison of the contaminant exposure from within the soil (M. arctica), and the soil surface and on vegetation (H. viatica). In addition, the species are easy to maintain in the laboratory.
H. viatica is a surface-dwelling species, that may reach 1.9 mm in length, and has a bluish- black colour, and a fully developed furca (springtail; i.e jumping organ) (Figure 2). It has a global distribution, including the high Arctic islands. H. viatica is common along the coast of Svalbard (Hertzberg et al., 2000), but also found inland, and is often abundant in bird cliffs, in rotting seaweed, and in sewage beds (Fjellberg, 1980; Fjellberg, 1998). It is commonly found gathered numerously in clusters or aggregates (Figure 2), facilitating the sampling of a high number of individuals.
Figure 2. Left: Illustration of Hypogastrura viatica. Right: A cluster of H. viatica (blue) with moulted skin (white) show the patchy distribution of the species on the tundra.
M. arctica typically inhabits the loose upper soil, or lives under moss or small rocks, as it is susceptible to desiccation and UV radiation (Hodkinson et al., 1994b; Leinaas, 2002). It may
12
reach 3.5 mm in body length, has a pale-yellow colour with no eyes or furca (Fjellberg, 1998;
Hopkin, 1997) (Figure 3). The general distribution of M. arctica is northern Palaearctic, and it is common in the Norwegian Arctic islands (Svalbard, Bjørnøya, Jan Mayen), as well as the western coast of Norway and high alpine inland mountains (Fjellberg, 1998). M. arctica often develops dense populations in Arctic bird cliffs (Hodkinson et al., 1994a) and on rocky shores. In alpine habitats, it is commonly found around resting areas of birds where bird manure give rise to a rich algal growth, which it feeds on (Fjellberg, 1998).
Figure 3. Left: Illustration of Megaphorura arctica. The lack of pigment makes the digestive system visible through the cuticle. Right: A group of M. arctica from the field, although such large group size was rarely found.
2.2 Study sites
In June 2016, the two Collembola species and the substrate of their respective habitats were sampled in the area of Kongsfjorden (78°N, 11°E) and Krossfjorden (79°N, 11°E) in Svalbard, Norway (Figure 4). Samples were collected from seven study sites; 14. Julibukta in Krossfjorden, and Blomstrandhalvøya, Stuphallet, Krykkjefjellet, Ossian Sars, Storholmen and Ny-Ålesund in the Kongsfjorden area (Figure 4). With the exception of Storholmen, these sites were outside of nature reserves and needed no special permission from the governor of Svalbard. A research group with a granted permission for visiting the nature reserve gathered samples from Storholmen during their own fieldwork. This project is registered in the Research in Svalbard database with RiS-ID 10447.
13
Figure 4. Left: Map of the Svalbard archipelago, Norway (ranging between 74°N - 81°N and 10°E - 35°E), with the study area circled. Right: Study sites.
Based on the knowledge of the seabird species breeding at the different colonies and previous reports of their contaminant levels, study sites were selected and a priori ranked as high, medium and low seabird influence. The study sites 14. Julibukta, Blomstrandhalvøya, Stuphallet, Krykkjefjellet and Ossian Sars were evaluated as sites with high seabird influence, due to the numerous breeding pairs of brünnich’s guillemot (Uria lomvia), black-legged kittiwake (Rissa tridactyla), northern fulmar (Fulmarus glacialis), black guillemot (Cepphus grylle), and atlantic puffin (Fratercula arctica) (Table 1). All mentioned species are mainly pelagic-feeding, and have high positions in the Arctic coastal food web due to their diets of crustaceans including amphipods and copepods, squid, and fish such as polar cod (Boreogadus saida) and herring (Clupea harengus) (Barrett et al., 2002; Gabrielsen and Ryg, 1994). Storholmen was evaluated with moderate seabird influence, but with high terrestrial bird influence as the island is mainly inhabited by common eiders (Somateria mollissima), and barnacle geese (Branta leucopsis). The common eider is benthic feeding, preying on snails and mussels, and has a lower position in the food web compared to the previous
14
mentioned seabirds (Hobson, 1993). The barnacle goose is a terrestrial feeder, grazing on plants, and occupy a lower trophic level relative to seabirds by being a primary consumer (Hahn et al., 2011). Ny-Ålesund was evaluated as a site with low seabird influence, but was not considered as a pristine study site, due to its vicinity to human activity at the research station. Samples from Ny-Ålesund were collected close to a small lagoon, with a shallow channel of water flowing from the research station. A description of impact level and expected source of contamination in the selected study sites is provided in Table 1.
15
Table 1. Overview and description of study sites with their inhabiting seabird species.
Location Source of impact Species Relative population size δ15N 2
Low1 Ny-Ålesund Human activity Settlement of 30-120
seabird (78° 55 25 N, individuals. Occasional
influence 11° 56 36 E) tourist cruise ships.
Historical coal mining.
Seabirds Arctic tern Low 13.7 3
(Sterna paradisaea) (measured in blood)
Terrestrial birds Barnacle goose Low 6.4 4
(Branta leucopsis)
Medium1 Storholmen Colony of Common eider duck High 13.2 5
seabird (78° 55 56 N, marine and (Somateria mollissima)
influence 12° 13 2 E) terrestrial birds Barnacle goose High 6.4 4
Great skua Low 13.3 9
(Stercorarius skua) (measured in blood)
High1 14. julibukta Mixed seabird Brünnich’s guillemot High 11.8 - 15.4 5678 seabird (79° 7 26 N, colony (Uria lomvia)
influence 11° 53 47 E) Black-legged kittiwake High 12.7 - 14.9 5678
(Rissa tridactyla)
Barnacle goose Low
Blomstrandhalvøya Mixed seabird Northern fulmar Moderate 14.0 - 15.3 56 (78° 58 40 N, colony (Fulmarus glacialis)
12° 4 38 E) Black-legged kittiwake High 12.7 - 14.9 5678
Krykkjefjellet Mixed seabird Black-legged kittiwake High 12.7 - 14.9 5678
(78° 55 00 N, colony Brünnich’s guillemot Low 11.8 - 15.4 5678
11° 56 00 E) Black guillemot Low 13.4 - 14.9 5678
(Cepphus grylle)
Stuphallet Mixed seabird Atlantic puffin High 12.2 10
(78° 57 26 N, colony (Fratercula arctica)
11° 39 5 E) Black guillemot Low 13.4 - 14.9 5678
Ossian Sars Mixed seabird Northern fulmar Low 14.0 - 15.3 56
(78° 55 59 N, colony Brünnich’s guillemot High 11.8 - 15.4 5678
12° 26 16 E) Black-legged kittiwake High 12.7 - 14.9 5678
1 Subjective evaluation of degree of seabird influence based on relative number of seabirds, and their trophic positions.
2 Values for stable isotope ratios of nitrogen (δ15N) (‰) are from bird muscles, unless stated otherwise 3 Pratte et al., 2017
4 Hahn et al., 2011 5 Hobson et al., 1993 6 Fisk et al., 2001 7 Hop et al., 2002 8 Borgå et al., 2005 9 Bearhop et al., 2000 10 Hedd et al., 2010
16
2.3 Field procedures
The main focus was on H. viatica, that was the most abundant and feasible to sample in sufficient numbers in most study sites, while M. arctica was sampled where it occurred in high enough numbers. A high number of Collembola (>>10.000 individuals) was required for each sample to carry out chemical analyses. H. viatica was sampled from moss or on a layer of cyanobacteria on the soil surface, which were cut in pieces by a stainless-steel knife and gently shaken in a stainless-steel sieve over a plastic box. M. arctica was sampled from the top layer of litter by sieving the substrate over a plastic box. To avoid dehydration, a few strands of moss were kept in the container with the Collembola. Substrate samples were collected after obtaining the Collembola, and were cut in pieces by a stainless-steel knife, wrapped in aluminium foil, and kept in zip-lock bags. The equipment was cleaned between each sampling site. Samples of substrate were stored at -20°C until contaminant and stable isotope analysis.
In the laboratory, the Collembola species were confirmed in a stereomicroscope by Hans Petter Leinaas. A random proportion of the Collembola population sampled from each study site was kept alive in a moist environment at 4°C for 0-14 days awaiting genotoxic effect analyses. The Collembola collected for contaminant and stable isotope analyses were cleaned of debris by placing the samples on a piece of foil in a container. After a few minutes, the Collembola had moved off the foil into the box, leaving sand and other debris behind. The few pieces of debris following the animals were moved using a wet paintbrush. Samples of cleaned Collembola were transferred to glass containers sealed with aluminium foil, weighed and stored at -20°C.
An overview with description of the samples collected for chemical analysis and effect measurements is given in Table 2.
17
Table 2. Overview and description of samples of Collembola with associated samples of substrate, and the analyses conducted. or å koble ut oppvaskmaskinen.Analyses LocationCollembola sample-IDCollembola SpeciesSubstrate sample-IDSubstrate contentDistance from bird cliffSite descriptionPOPsPFASsHgSIA, C, NLipidComet MN Ny-ÅlesundMA1M. arcticaSUB160% moss, 40% soil 50-200 m from Wet area, pond Substrate and Substrate and Substrate and Substrate and (with cyanobacteria layer)research settlementby the beachCollembolaCollembolaCollembolaCollembolaCollembola Ny-ÅlesundHV2H. viaticaSUB2,320% moss, 80% soil 50-200 m from Wet area, pond Substrate and Substrate and Substrate and Substrate and (with cyanobacteria layer)research settlementby the beachCollembolaCollembolaCollembolaCollembolaCollembola Ny-ÅlesundHV3H. viatica50-200 m from Wet area, pond research settlementby the beachCollembolaCollembolaCollembolaCollembolaCollembola Ny-ÅlesundHV4H. viaticaSUB450% moss, 50% soil50-200 m from Wet area, pond Substrate and Substrate and Substrate and Substrate and research settlementby the beachCollembolaCollembolaCollembolaCollembolaCollembola Ny-ÅlesundH. viaticaYESYES Ny-ÅlesundM. arcticaYESYES StorholmenHV10H. viaticaSUB10100% mossCentre of Flat ground, Substrate and SubstrateSubstrate and Substrate and Storholmen islandwet areaCollembolaCollembolaCollembolaCollembolaYESYES 14. julibuktaHV5H. viaticaSUB530% moss, 70% soil130-150 m from Bottom of the Substrate and Substrate and Substrate and Substrate and seabird colonyslope, wet areaCollembolaCollembolaCollembolaCollembolaCollembola 14. julibuktaHV6H. viaticaSUB630% moss, 70% soil130-150 m from Bottom of the Substrate and Substrate and Substrate and Substrate and seabird colonyslope, wet areaCollembolaCollembolaCollembolaCollembolaCollembola 14. julibuktaHV7H. viaticaSUB7,850% moss, 50% soil130-150 m from Bottom of the Substrate and Substrate and Substrate and Substrate and seabird colonyslope, wet areaCollembolaCollembolaCollembolaCollembolaCollembola 14. julibuktaHV8H. viatica130-150 m from Bottom of the seabird colonyslope, wet areaCollembolaCollembolaCollembolaCollembolaCollembola
18
Analyses LocationCollembola sample-IDCollembola SpeciesSubstrate sample-IDSubstrate contentDistance from bird cliffSite descriptionPOPsPFASsHgSIA, C, NLipidComet MN 14. julibuktaHV9H. viaticaSUB950% moss, 50% soil130-150 m from Bottom of the Substrate and Substrate and Substrate and Substrate and seabird colonyslope, wet areaCollembolaCollembolaCollembolaCollembolaCollembola 14. julibuktaH. viaticaYESYES BlomstrandøyaHV11H. viaticaSUB1150% moss, 50% soil400 m from Bottom of the Substrate and Substrate and Substrate and Substrate and (with cyanobacteria layer)seabird colonyslope, wet areaCollembolaCollembolaCollembolaCollembolaCollembola BlomstrandøyaHV12H. viaticaSUB1295% moss, 5% soil150 m from In the slopeSubstrate and Substrate and Substrate and Substrate and (with cyanobacteria layer)seabird colonyCollembolaCollembolaCollembolaCollembolaCollembola BlomstrandøyaH. viaticaYESYES Krykkjefjellet MA13M. ArcticaSUB1390% moss, 10% soil0-2 m from In the slopeSubstrateSubstrate and Substrate and Substrate and seabird colonyCollembolaCollembolaCollembolaYESYES StuphalletMA14M. ArcticaSUB1450% moss, 50% soil0-2 m fromIn the slopeSubstrateSubstrateSubstrateSubstrate and seabird colonyCollembolaYESYES StuphalletHV15H. viaticaSUB1550% moss, 50% soil250 m from Bottom of the Substrate and SubstrateSubstrate and Substrate and seabird colonyslope, wet areaCollembolaCollembolaCollembolaCollembolaYESYES Ossian SarsM. Arctica100 m from In the slopeYESYES seabird colony Substrate sample SUB7,8 is parallel to Collembola samples HV7 and HV8. Substrate sample SUB2,3 is parallel to Collembola samples HV2 and HV3. For locations with more than one sample of the same species, random individuals were chosen from all the samples and processed for effect analyses. Distance from bird cliffs are approximate and retrieved using TopoSvalbard (toposvalbard.npolar.no). Lipid content of Collembola were measured on samples conducted for POPs analysis, and this parameter is therefore restricted to the amount available for POPs analysis. Abbreviations: MN=micronucleus test, SIA=stable isotope analysis, C, N=total carbon and nitrogen content.
Table 2 Continued. Overview and description of samples of Collembola with associated samples of substrate, and the analyses conducted.
19 Two sites were sampled with two distances from the base of the seabird colony (Figure 5–6).
In Stuphallet and Blomstrandhalvøya, one Collembola population with its respective substrate were sampled close to the bird cliff (0–150 m), and one was sampled at a farther distance from the colony (250–400 m).
Figure 5. Schematic profile of Stuphallet. Circles indicate where samples were collected, identified with sample-ID. MA=M. arctica, HV=H. viatica, SUB=substrate.
Figure 6. Schematic profile of Blomstrandhalvøya. Circles indicate where samples were collected, identified with sample-ID. HV=H. viatica, SUB=substrate.
20
2.4 Contaminant analyses
Samples of substrates (moss and/or soil) and Collembola were analysed for a wide range of lipid-soluble contaminants, including PCBs, OCPs and PBDEs, and protein-associated contaminants; PFASs and Hg. A complete list of contaminants included in the internal standards are presented in Appendix A.
2.4.1 Homogenisation and sample splitting
To avoid cross-contamination, equipment of metal, porcelain and glass were cleaned and rinsed with acetone and cyclohexane (both SupraSolv®, Merck, Darmstadt, Germany), and plastic equipment was cleaned with isopropanol (Kemetyl, Vestby, Norway) between homogenisation of each sample. Substrate samples were homogenised in a food processor (Hugin, Punica minihakker) and/or by stirring with a metal spoon in a porcelain bowl. A metal sift was used to ensure small pieces and to remove rocks. Substrate samples analysed for Hg were air-dried at room temperature for 3 days prior to extraction to ensure sufficient sample material. Due to the lack of a standardised method for treatment of soil and moss for Hg analysis, this is shortly discussed in section 4.2. Population samples of H. viatica and M.
arctica were homogenised in a porcelain bowl using the back of a porcelain spoon.
Homogenates of substrate and Collembola were weighed and separated into sub-samples;
glass containers for POPs analysis and plastic tubes for PFAS, Hg, and SIA. The following amount of sample material were required to perform chemical analyses on Collembola: 2 g for POPs, 1–2 g for PFASs, 0.2–0.5 g for Hg and 0.2 g for SIA. It was not possible to obtain the total required amount of Collembola per sample (approximately 4 g) to perform all analyses, at all study sites. Consequently, the number of analyses carried out was limited for some sites (Table 2). Analyses of substrate samples required more than 1 g for Hg, 1 g for SIA, 10 g for POPs, 5 g for PFAS, and approximately 5 g was used for determination of dry matter. All substrate samples consisted of enough material to perform all analyses. Samples were kept at -20°C prior to extraction.
2.4.2 Analysis of lipid-soluble persistent organic pollutants (POPs)
Substrates and Collembola were analysed for lipid-soluble POPs under guidance and supervision of Mikael Harju at the Norwegian Institute for Air Research (NILU), Tromsø.