Mercury in freshwater biota in southeastern Norway, with special emphasis on potential antagonistic effects of selenium on mercury bioaccumulation
MetadataVis full innførsel
- Ecology 
Mercury (Hg), and in particular methylated Hg (methyl-Hg, MeHg), because of its high potential for bioaccumulation and biomagnification in aquatic food webs, generate health risks to both aquatic top predators and humans consuming Hg contaminated fish or other aquatic wildlife with high Hg concentrations. Although atmospheric long-range transported Hg has decreased in Scandinavia, Hg concentrations in fish has increased in recent years. Some of the hypothesized causes for this is reduction in acid deposition, climate change (warmer and wetter), and changes in forestry-practices. A result of these interactions, is often increase in organic carbon in aquatic freshwater systems, increased bacterial Hg-methylation and reduced in-lake photo demethylation as a result of reduced light penetration (reduced sight depth) following increase in total organic carbon (TOC)/water color. Although a small fraction of the total Hg (Tot-Hg) in Scandinavian lakes exists as MeHg (1-5%) it is likely to assume that the fraction of MeHg has increased in recent years despite decreased reduced input of Tot-Hg. Additionally, Hg in fish may also increase in populations experiencing reduced growth. Another contributing factor for high Hg concentrations in some Norwegian lakes may be low levels of selenium (Se). Several studies have reported decreased Hg concentrations in aquatic biota in the presence of elevated Se in water, and research suggests a potential tissue Se threshold in fish and fish diet for an unequivocal antagonistic effect of Se on Hg bioaccumulation. Thus, the factors to explain increased Hg in fish despite decreased Hg depositions may be multifactorial, and not yet fully elucidated. The main goal of this thesis was to investigate Hg concentrations in aquatic biota in different lakes in the River Skienselva watercourse, southern Norway, and study how variations in physiochemical conditions in lakes, habitat use, trophic level and fish biometry affect bioaccumulation of Hg in fish. We also investigated seasonal variations in fish in the profundal zone of one of the studied lakes. In addition, Se was investigated in order to reveal a potential mitigating effect on Hg bioaccumulation in perch (Perca fluviatilis) and brown trout (Salmo trutta). The investigated lakes are mainly large oligotrophic lakes, from the alpine and highly regulated Lake Songavatn (974 m a.s.l) in the northwest, to Lake Norsjø (15 m a.s.l) in the lowland in the southeast. One of the investigated lakes, Lake Norheimstjønna (Lake Norheim) differs from the other lakes due to its smaller size and higher concentrations of TOC and nutrients (N and P). We applied stable isotope analysis (SIA), by measuring δ15N and δ13C in fish in all studies, and included macroinvertebrates in two of the studies, to assess both trophic level (δ15N) and dietary sources (δ13C) in the investigated fish. δ13C values varies in different carbon sources, typically with around -27 ‰ for terrestrial, -20 ‰ for littoral, - 28 ‰ for pelagial and -30 ‰ for profundal carbon sources. Thus SIA, in addition to fish biometry and stomach content analyses, were used to assess variations in Hg and Se, in relation to trophic level (TL), dietary sources, age and size in fish. In the study on biomagnification of Hg and Se in perch in Lake Norheim and Lake Norsjø (one site in the north, Norsjø N and one in the south, Norsjø S), littoral and pelagic invertebrates together with perch were collected in July 2013. Based on measured δ15N of a primary consumer, we calculated baseline adjusted relative trophic levels (TL’s). The trophic magnification factors (TMF’s), i.e. increase in measured Se and Hg per TL, were calculated, and resulted in a common TMF of 1.29 for Se and 4.64 for Hg for all three sites. The relatively low water Se concentrations in these two lakes (22 -59 ng Se L−1), yet relatively high accumulation in biota, probably reflect that a major proportion of the Se in these lakes are both highly bioavailable and transferred up the food chain. Higher adjusted mean Hg in perch in Lake Norheim (0.94 mg Hg kg−1 dw) and Lake Norsjø N (0.86 mg Hg kg−1 dw), both close to river outlets, compared to Lake Norsjø S (0.67 mg Hg kg−1dw), likely reflect riverine transport of TOC, Tot-Hg and MeHg from the catchment. Moreover, because of the slower fish growth, Hg in Lake Norheim perch was substantially higher (up to 3.6 mg Hg kg−1 dw), compared to the perch from the two other sites when adjusting for differences in length and TL. In addition, the results on Se and Hg bioaccumulation in perch suggested increased assimilation towards pelagic compared to littoral carbon sources (measured as δ13C). The causality behind this result was uncertain due to the much depleted δ13C signatures in both perch and littoral invertebrates. The study on profundal fish in the southern part of Lake Norsjø was based on fish sampled monthly during the year 2014, from grates mounted at an industrial water intake, located at a depth of 50 m. The three most common species present, Arctic charr (Salvelinus alpinus), European smelt (Osmerus eperlanus) and whitefish (Coregonus lavaretus), were analyzed for variations in size, age, δ15N and δ13C, stomach content and Hg. Both the stomach analysis and δ13C signatures suggested a combined profundalpelagic diet for all three species. Whereas length was the best predictor for Hg variations in A. charr and whitefish, age was the best predictor for variations of Hg in E. smelt. A. charr had the most profundal-based diet, and was the only species exhibiting seasonal variation in Hg, highest during winter and spring, likely because of starvation during the cold and dark winter period and subsequent growth dilution during the organic carbon production period in the lake during summer. The study on free-ranging brown trout in the River Skienselva watercourse included fish sampled in the autumn 2008 from five lakes in the watercourse. Based on measured size, age, δ15N, δ13C, Se, and Hg, together with available data on geographic positions of lakes and lake morphology, we performed analyses in order to investigate predictors for variations of Hg and Se, as well as geographical patterns of Hg and Se in brown trout. The results revealed differences in fish Hg concentrations between lakes after adjusting for the significant contributions from both age and TL (measured as δ15Nadj), whereas fish Se concentrations differed between lakes after adjusting for TL. The concentrations (dw) of Hg and Se in fish muscle tissue ranged from 0.21 to 2.06 mg Hg kg−1 and 0.96 to 2.51 mg Se kg−1. The results indicate that differences in Hg in trout among lakes may be explained by variations in primary production and a varying degree of dilution of Hg at the base of the food chain. In both this study on trout and the earlier described study on perch, negative correlations between δ13C and Se concentrations in fish were revealed, indicating increased Se assimilation in pelagic compared to littoral food chains . For the trout, we suggested that this might relate to variation in regulation height in lakes. This either could be as an effect of increased pelagic feeding because of reduced littoral production or because of increased Se concentrations in remaining water mass at the lowest regulated water level (LRW). The inclusion of tissue Se as an explanatory variable in the Hg models was not statistically significant in neither perch nor trout, and increasing Se concentrations did not lead to significantly decreased mean tissue Hg concentrations in neither of the two species, after adjusting for other significant explanatory variables. Our results support previous conclusions of a muscle tissue Se concentration threshold to affect Hg concentrations in fish, and suggest that the lakes in the region most likely are too low in Se for fish to reach such a threshold concentration. In conclusion, this work shows that variations in Hg in fish in the studied lake ecosystems are determined by variations in habitat use and trophic level, i.e. related to where in the ecosystem they feed and at what trophic level in the food chain, respectively. It also shows that variations in Hg can be explained by differences in mass-length relationships, i.e. variations in growth, either because of inter and -intra specific food competition or related to variation in lake productivity, both among lakes as well as among seasons. It also indicates that Se in water and biota is not a significant predictor for Hg concentrations in the investigated fish in these lakes, and that this probably relates to too low Se concentrations in water and biota.
Doktorgradsavhandlingen omhandler kvikksølvkonsentrasjoner i akvatiske organismer i ulike innsjøer i Skiensvassdraget. Økelsrud har sett på hvordan variasjoner i ulike fysisk-kjemiske forhold, habitat bruk, trofisk nivå, samt fiskebiometri påvirker kvikksølvakkumulering i fisk i disse innsjøene. Kvikksølv og spesielt organisk kvikksølv som følge av oppkonsentrering i akvatiske næringskjeder, kan medføre høye konsentrasjoner med potensielle nevrologiske skadevirkninger i både akvatiske dyr i toppen av næringskjeden, og i mennesker som spiser fisk eller andre akvatiske dyr med forhøyede kvikksølvkonsentrasjoner. Flere studier viser reduserte kvikksølvkonsentrasjoner i akvatisk dyreliv i innsjøer med høye konsentrasjoner av selen, men forskning tyder på at selenkonsentrasjonene i fisk og fiskens byttedyr må over en viss terskelverdi før selen har en tydelig motvirkende effekt på akkumuleringen av kvikksølv. Kort sammenfattet viser avhandlingen at variasjoner i kvikksølvkonsentrasjoner i fisk i de undersøkte innsjøene påvirkes av hvor i innsjøene fisken henter sin næring og på hvilket nivå i næringskjeden fisken befinner seg, (habitatbruk og trofisk posisjon). Sentrale faktorer for kvikksølvvariasjoner/nivåer i fisk er også fiskens vekst, (enten som følge av konkurranse innen samme art eller mellom arter), variasjoner i organisk produksjon innen og mellom innsjøer, samt sesong- og års -variasjoner. I våre studier synes ikke selenkonsentrasjoner i vann og byttedyr å ha noen signifikant effekt på kvikksølvkonsentrasjoner i fisk, til det synes selenkonsentrasjonene i våre innsjøer å være for lave. Resultatene kan ha betydning når ulike forvaltningstiltak med tanke på å redusere kvikksølvkonsentrasjoner i fisk skal vurderes, som f.eks. kunstig tilførsel av selen eller økt fiske for å øke veksten i gjenværende fisk.
Består avArticle 1: Økelsrud A., Lydersen, E. & Fjeld E.: Biomagnification of mercury and selenium in two lakes in southern Norway. Science of the Total Environment 566, (2016), 596-607. http://dx.doi.org/10.1016/j.scitotenv.2016.05.109
Article 2: Olk, T.R., Karlsson, T., Lydersen, E. & Økelsrud, A.: Seasonal variations in the use of profundal habitat among freshwater fishes in Lake Norsjø, southern Norway, and subsequent effects on fish mercury concentrations. Environments 3(4), (2016), 29. https://doi.org/10.3390/environments3040029
Article 3: Økelsrud, A., Lydersen, E., Fjeld, E. & Moreno, C.: Mercury and selenium in free-ranging brown trout (Salmo trutta) in the River Skienselva watercourse, Southern Norway. Science of the Total Environment 586, (2017), 188–196. http://dx.doi.org/10.1016/j.scitotenv.2017.01.199