Aarhus Universitets segl

Temporal trends of persistent organic pollutants in Arctic marine and freshwater biota

New publication by Frank Rigét, Anders Bignert, Birgit Braun, Maria Dam, Rune Dietz, Marlene Evans, Norman Green, Helga Gunnlaugsdóttir, Katrin S. Hoydal, John Kucklick, Robert Letcher, Derek Muir, Stacy Schuur, Christian Sonne, Gary Stern, Gregg Tomy, Katrin Vorkamp, Simon Wilson

Abstract: 

More than 1000 time-series of persistent organic pollutants (POPs) in Arctic biota from marine and freshwater ecosystems some extending back to the beginning of 1980s were analyzed using a robust statistical method. The Arctic area encompassed extended from Alaska, USA in the west to northern Scandinavian in the east, with data gaps for Arctic Russia and Arctic Finland. The aim was to investigate whether temporal trends for different animal groups and matrices were consistent across a larger geographical area. In general, legacy POPs showed decreasing concentrations over the last two to three decades, which were most pronounced for α-HCH and least pronounced for HCB and β-HCH. Few time-series of legacy POPs showed increasing trends and only at sites suspected to be influenced by local source. The brominated flame retardant congener BDE-47 showed a typical trend of increasing concentration up to approximately the mid-2000s followed by a decreasing concentration. A similar trend was found for perfluorooctane sulfonic acid (PFOS). These trends are likely related to the relatively recent introduction of national and international controls of hexa- and hepta-BDE congeners and the voluntary phase-out of PFOS production in the USA in 2000. Hexabromocyclododecane (HBCDD) was the only compound in this study showing a consistent increasing trend. Only 12% of the long-term time-series were able to detect a 5% annual change with a statistical power of 80% at α < 0.05. The remaining 88% of time-series need additional years of data collection before fulfilling these statistical requirements. In the case of the organochlorine long-term time-series, 45% of these would require >20 years monitoring before this requirement would be fulfilled.  

https://doi.org/10.1016/j.scitotenv.2018.08.268