Vertical distribution of polycyclic aromatic hydrocarbons in the brackish sea water column: ex situ experiment

Abstract

Background Oil spills can cause severe damage within a marine ecosystem. Following a spill, the soluble fraction of polycyclic aromatic hydrocarbons is rapidly released into the water column. These remain dissolved in seawater over an extended period of time, even should the insoluble fraction be removed. The vertical distribution of the aromatic hydrocarbon component and how these become transferred is poorly understood in brackish waters. This study examines the vertical distribution of polycyclic aromatic hydrocarbons having been released from a controlled film of spilled oil onto the surface of brackish water. Methods The study was undertaken under controlled conditions so as to minimize the variability of environmental factors such as temperature and hydrodynamics. The distribution of polycyclic aromatic hydrocarbons was measured in the dissolved and suspended phases throughout the 1 m water column with different intensity of water sampling: 1, 2, 4, 7, 72, 120, 336, 504 and 984 h. Results The total concentration of polycyclic aromatic hydrocarbons ranged from 19.01 to 214.85 ng L–1 in the dissolved phase and from 5.14 to 63.92 ng L–1 in the suspended phase. These hydrocarbons were released immediately following a controlled spill attaining 214.9 ng L–1 in the dissolved phase and 54.4 ng L–1 in the suspended phase near the cylinder bottom after 1–2 h. The 2–3 ring polycyclic aromatic hydrocarbons dominated in the dissolved phase (60–80%), whereas the greater amount of 4–6 ring polycyclic aromatic hydrocarbons (55–90%) occurred in the suspended phase. A relatively low negative correlation (rS = –0.41) was determined between the concentration of phenanthrene and suspended matter, whereas a high negative correlation (r =  − 0.79) was found between the concentration of pyrene and suspended matter. Despite the differences in the relationships between the concentration ratio and amount of suspended matter the obtained regressions allow roughly to predict the concentration of polycyclic aromatic hydrocarbons.