Incorporating ecological data and associated uncertainty in bioaccumulation modeling: methodology development and case study.

Bioaccumulation models predict internal concentrations of hydrophobic chemicals by incorporating key gain/loss processes reflecting the ecology of the exposed species and the characteristics of the chemical. Here, we propose a new methodology that uses ecological data and the principle of mass balance in food webs to estimate bioaccumulation in food webs. To this end, we combine linear inverse models (LIMs) that estimate food web flows based on mass balance with a mechanistic bioaccumulation model (OMEGA). In a case study we show that uncertainty ranges on bioaccumulation predictions were on average estimated a factor of 4 lower by LIM-OMEGA than by an OMEGA application that does not consider mass balance within food webs, most notably for chemicals with log Kow > 5, reflecting an increasing importance of uptake through food ingestion for those chemicals. Ranges of internal concentrations predicted by LIM-OMEGA were smaller in enclosures with fish, as strong predation pressure from the latter on mesozooplankton constrains food web flows and thus bioaccumulation.