Research Toxic Effects of Unresolved Complex Mixtures of Aromatic Hydrocarbons Accumulated by Mussels, M ytilus edulis , from Contaminated Field Sites PETER DONKIN, EMMA L. SMITH, AND STEVEN J. ROWLAND* Petroleum and Environmental Geochemistry Group, Department of Environmental Sciences, Plymouth Environmental Research Centre, University of Plymouth, Drake Circus, Plym outh PL4 8AA, United Kingdom Exposure of marine mussels (Mytilus edulis) to an unresolved complex mixture (UCM) of aromatic hydrocarbons isolated from a crude oil has been shown to reduce their feeding rate by 40%. The present study was undertaken to determine whether UCMs bioaccumulated by mussels in the field are also toxic. The feeding rate of mussels derived from polluted sites increased when they were placed in clean water, pointing to a loss of toxic agents from the tissues. At the end of the depuration period, water in which mussels from an oil-polluted site had been held contained a UCM. Steam-distillation extracts of the tissues of mussels taken from several polluted sites were shown to be highly toxic to the feeding activity of juvenile mussels. The tissues of mussels from these sites contained UCMs. Nontoxic steam distillates from clean mussels did not. Steam-distillation extracts of mussels from an oil-polluted site were fractionated by normal-phase high-performance liquid chromatography. A fraction, largely comprising a “ monoaromatic” UCM , reduced the feeding rate of juvenile mussels by 70%. Two later-eluting fractions containing aromatic UCMs also produced smaller depressions in feeding rate. These results support our contention that some aromatic UCM hydrocarbons constitute a forgotten pollutant burden in the marine environment. Introduction The Blue Mussel, Mytilus edulis, is an important sentinel organism used in global marine monitoring ofwater quality (1-3).Mussels in polluted environments accumulate a wide varietyofpollutantsbyfiltration ofwaterand exhibit a variety of health effects (1-5). Among the hydrophobic organic pollutants accumulated by mussels, a feature known as the unresolved complex mixture (UCM) of hydrocarbons is commonlyobserved (6).Such UCM hydrocarbons comprise both aromatic and nonaromatic compounds (7). The toxicityofUCMhydrocarbonshasrarelybeen studied (8, 9). Such evidence as it exists suggests that nonaromatic UCMhydrocarbonsare largelynontoxicto mussels,although toxicitycan be introduced byoxidation ofnonaromaticUCMs to oxygenated, but still mainly unresolved, compounds (8). Recently, however, we were able to demonstrate that an aromatic UCM isolated from a North Sea crude oil by high- performance liquid chromatography(HPLC)wasable to elicit a nonspecific narcotic toxic response in mussels exposed in the laboratory (viz., a >40% reduction in feeding rate in 24 h(9)).Furthermore,anumberofpopulationsofU.K.mussels exhibiting impaired health, as measured by a well-accepted assay known as scope for growth (Sf G′; 10) contained high concentrations of UCM hydrocarbons (9). In the present studywe sought to establish whether UCMs bioaccumulated by mussels in contaminated field sites are toxic, and which if any of the UCM subfractions are most toxic. Three experimental approaches were used to provide evidence for the toxicity of bioaccumulated UCMs: (1) depuration/recovery experiments in clean water to establish whether the loss of UCMs from the tissues of polluted mussels resulted in an improvement in their health; (2) toxicity testing of steam-distillation extracts of mussel tissues, comparing clean and UCM-polluted sites; (3) frac- tionation ofthese extracts and toxicitytestingofthe fractions. Materials and Methods Sample Sites. Whitsand Bay is home to an open coast population ofmusselsand hasa generallyhigh waterquality. For the greater part of the year this can be used as a “clean” reference population. The mussel population at Port Quin resides in a small rocky inlet and has been used previously as a clean reference site in an Irish sea study (10). At Sutton Harbour mussels transplanted from Port Quin were caged and placed at a 2 m depth hanging from a jetty on a marina used largelyfor recreationalboatingactivity.Musselsat West Hoe were transplanted from Port Quin without caging (i.e., direct to the intertidal rocks), close to the outfall ofthe main sewer for Plymouth. Samples were also taken from natural musselpopulationsat Instow,where commercialharvesting for human consumption is banned because of sewage pollution, and Cattedown, due to the close proximity of a marina and harbor.Allofthese sitesare located in and around Plymouth in the southwest part of England. Mussels were also collected from New Brighton and Kirkcolm by Dr. T. Crowe during ecological investigations (11). New Brighton is a site with severely depressed mussel scope for growth (10) and a mussel bed ecosystem indicative of pollution impact(11), while Kirkcolm in comparison is a clean site (10, 11). Allthe musselpopulationswere native populationsexcept for those at Sutton Harbour and West Hoe, which were transplanted from Port Quin. The depuration/recovery experiments were performed with mussels from Whitsand, Port Quin,and Sutton Harbour due to the proximityofthese sites to the Plymouth laboratory. The testing of steam distillates was performed on all the sites ranging from clean to highlypolluted.Fractionation experimentswere performed on a clean (Whitsand) and a polluted (Sutton Harbour) site for comparison. Depuration/Feeding Rate Recovery Experiments. Mus- sels (35-40 mm shell length) were collected from the field, cleaned ofsuperficial debris and encrusting organisms, and used for laboratory study the same day, or if collected when low tide occurred in the late afternoon, they were cleaned, held overnight in moist air at a temperature close to that of the ambient water temperature, and then used for experi- ments the next morning. The maximum air exposure time * Corresponding author phone: +44 1752 233013; fax: +44 1752 233035; e-mail: srowland@plym.ac.uk. 10.1021/es021053e CCC: $25.00  2003 American Chemical Society VOL. 37, NO. 21, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4825 Published on Web 10/01/2003