1897 Environmental Toxicology and Chemistry, Vol. 21, No. 9, pp. 1897–1902, 2002  2002 SETAC Printed in the USA 0730-7268/02 $9.00 + .00 FISH (FUNDULUS HETEROCLITUS) POPULATIONS WITH DIFFERENT EXPOSURE HISTORIES DIFFER IN TOLERANCE OF CREOSOTE-CONTAMINATED SEDIMENTS DAVID R. OWNBY,* MICHAEL C. NEWMAN,MARGARET MULVEY,WOLFGANG K. VOGELBEIN, MICHAEL A. UNGER, and L. FELIPE ARZAYUS Virginia Institute of Marine Science, College of William & Mary, Gloucester Point, Virginia 23062, USA ( Received 16 November 2001; Accepted 25 February 2002) Abstract—Prior studies suggest that field-collected fish (Fundulus heteroclitus) from a creosote-contaminated Superfund site (Atlantic Wood Industries site, Elizabeth River, VA, USA) have enhanced tolerance to local, contaminated sediments. This study was designed to test whether other populations in the Elizabeth River at less contaminated sites also show similar tolerance and whether this tolerance is heritable. To test this, F. heteroclitus populations were sampled from four sites within the Elizabeth River with varying sediment polycyclic aromatic hydrocarbon (PAH) concentrations (3.9–264 ng PAH/g dry wt·10 3 ) and one reference site in a nearby, uncontaminated estuary (York River, VA, USA; 0.27 ng PAH/g dry wt·10 3 ). Embryo assays were performed to quantify population differences in teratogenic effects during contaminated sediment exposure. Atlantic Wood sediment was mixed with reference sediment to achieve a range of sediment concentrations. Minimal differences were observed in teratogenic effects among fish taken from sites within the Elizabeth River; however, embryos of fish collected from a nearby, uncontaminated York River site and exposed to contaminated sediments had a significantly higher proportion of embryos with cardiac abnormalities than those from the Elizabeth River sites. Embryos from wild-caught and laboratory-reared Elizabeth River F. heteroclitus were si- multaneously exposed to contaminated sediments, and no significant tolerance differences were found between embryos from fish taken directly from the field and those reared for a generation in the lab. Differences between fish populations from the two estuaries were larger than differences within the Elizabeth River, and these differences in tolerance were heritable. Keywords—Fundulus heteroclitus Adaptation Polycyclic aromatic hydrocarbons Cardiovascular abnormalities Growth INTRODUCTION Populations inhabiting contaminated sites can adapt to pol- lutants. In the presence of sufficient variability within a pop- ulation, selection against individuals with lowest tolerance to the pollutant results in genetic adaptation. Such adaptation was documented for a wide range of species, including plants [1,2], oligochaetes [3], isopods [4], soil arthropods [5], and minnows [6,7] for both metals and organic contaminants. Despite its general occurrence, genetic adaptation to pol- lutants can be difficult to detect in field populations because exposed individuals can also acclimate to pollutants. Adap- tation is the genetic process by which a population changes to accommodate environmental factors. Acclimation is the physiological changes an individual makes to minimize the effects of stressors. For example, loss of enhanced tolerance in field-trapped mosquitofish with initially high tolerance to lead after they were kept for 34 d in clean water is evidence of individual acclimation. A nonheritable basis existed for the initial resistance [8]. Studies of enhanced tolerance should be designed to distinguish between individual physiologic accli- mation and genetic adaptation. Several factors contribute to genetic differentiation among populations, including migration, genetic drift due to geo- graphic isolation, species life history characteristics, and nat- ural selection. Pollutant effects on populations must be as- sessed in the context of all these processes in order to use genetic differentiation as an indicator of genetic adaptation [5,9–12]. * To whom correspondence may be addressed (downby@siu.edu). The mummichog, Fundulus heteroclitus, is an ideal animal for examining tolerance enhancement and genetic differenti- ation associated with contaminants [13]. This small fish is common and widespread in Atlantic Coast estuaries from Can- ada to Florida, USA. This hardy species is easily trapped in large numbers at contaminated habitats. The mummichog can be a good indicator of local environmental conditions because it tends to exist in local subpopulations or demes. They are amenable to laboratory culture and experimental manipula- tions [14] and demographic analysis [15]. Mummichog em- bryos have qualities valuable for toxicity testing: Large num- bers of embryos can be obtained from most fish enabling better replication and data analysis, the development time is short, and the transparency of the chorion allows detailed exami- nation of the embryo [16,17]. Because of these qualities, mum- michogs are common subjects of population biology and ge- netics study (e.g., Mitton and Koehn [18], DiMichele et al. [19], and Gonzalez-Villasensor and Powers [20]), including those involving pollutants (e.g., Weis and Weis [17], Munns et al. [21], and Nacci et al. [22]). Williams [23] described the tolerance of a mummichog pop- ulation from a site heavily contaminated with PAHs. This At- lantic Wood site (AW) is adjacent to a creosote treatment plant on the South Branch of the Elizabeth River (VA, USA) (Fig. 1). Previous studies at this site focused on the plausible re- lationship between mummichog liver pathology and sediment contamination [24,25]. At AW, adult mummichog had a 50% liver cancer prevalence and exhibited extrahepatic neoplasms [26]. Williams [23] did a series of laboratory experiments with embryos and juveniles, suggesting that F. heteroclitus from this location had enhanced tolerance to pollutants relative to