Lakes & Reservoirs: Research and Management 2004 9: 89–103 Assessing the toxicity of chemically fractionated Hamilton Harbour (Lake Ontario) sediment using selected aquatic organisms L. H. McCarthy, 1 * R. L. Thomas 2 and C. I. Mayfield 3 1 Department of Chemistry and Biology, Ryerson University, Toronto, Ontario, Canada, 2 Faculty of Science, F.-A. Forel Institute, University of Geneva, Versoix, Switzerland, and 3 Biology Department, University of Waterloo, Waterloo, Ontario, Canada Abstract Studies of the sediments of Hamilton Harbour, Lake Ontario, Canada, have shown variable degrees of pollution with a large number of organic and inorganic pollutants. Three areas in the Harbour – Windemere Basin, Cootes Paradise and Randle Reef – exhibited particularly high levels of contaminants, with concomitant impacts on benthic organisms. Sediment samples examined in this study were taken from a site in Randle Reef in Hamilton Harbour and a station (LE 23) in Lake Erie used as a reference sample. The samples were subjected to chemical fractionation to remove organic contaminants, with one fraction being used as total extract, and another aliquot being subjected to silica gel fractionation to give four fractions. Each fraction was chemically analysed and an aliquot used for bioassays following exchange with dimethylsulfoxide. Standard biotests were performed with Daphnia magna, Hyalella azteca, Pimephales promelas (fathead minnows) and Selenastrum capricornutum. Results from direct sediment contact bioassays indicated that sediment from Randle Reef is highly toxic, whereas the Lake Erie sample had no observable impacts. Total extract when added at 1% to the bioassays replicated the solid phase with acute toxicity for Randle Reef, and no effect on Lake Erie. For the silica gel fractionation, Fraction 1 (F1) showed similar toxicity as the total extract, even at low additions, whereas F2–4 gave variable and inconclusive results. F1 eluted with hexane contained the non-polar fractions, predominantly the polyaromatic hydrocarbons, and these compounds are the most likely causative agents for the highly acute toxicity obser ved in the sediment of Randle Reef. Key words bioassay, chemical fractionation, Great Lakes, Hamilton Harbour, organic pollutants, sediment, toxicity. INTRODUCTION Much research has been conducted to assess the toxicity of sediments from various ‘Areas of Concern’ in the Great Lakes basin in North America. In a study that examined 10 contaminated sites, sediment from Hamilton Harbour on Lake Ontario, Canada, caused extreme toxicity in bioassays containing the zooplankter cladoceran Daphnia magna and the benthic amphipod Hyalella azteca (McCarthy 1994). Hamilton Harbour, located on the western tip of Lake Ontario, is an almost totally enclosed body of water connected to the lake by a single ship canal across the sandbar that forms the bay (Fig. 1). The Harbour’s deep- water port allows frequent passage of ships and supports the largest concentration of heavy industry in Canada, the majority consisting of iron (Fe) and steel (O’Connor 2002). During the early 1900s, the harbour was the nursery habitat for the largest fishery on Lake Ontario and, until relatively recently, the major problems thought to be facing the harbour were oxygen (O2) depletion and eutrophic- ation. In a 1985 report on the harbour, the Ontario Ministr y of the Environment acknowledged that there was little information on the inputs of trace organics, although it was speculated that steel mill processes might introduce poly- aromatic hydrocarbons (PAHs) to the harbour. It was recognized that many substances, including polychlorin- ated biphenyls (PCBs) and metals, exceeded the Ontario Ministry of the Environment’s guidelines for open-water *Corresponding author. E-mail: LHmccarthy@rogers.com Accepted for publication 15 March 2004.