Some metals and metalloids in sediment from urban sites were elevated to concentrations higher than natural geochemical levels. The sites that were most contaminated contained many elements over background concentration, but the elements found most often in excess were lead, nickel, and zinc. Other elements such as copper, silver, and tin were also found to frequently exceed background concentrations, whereas elements such as arsenic, cadmium, mercury, and selenium were found less frequently at elevated concentrations.

Correlation analysis revealed that copper, lead, and zinc from Pacific coast urban sites were associated with each other, suggesting that they may be from a common source. This pattern was generally weak at nonurban West Coast sites suggesting that copper, lead, and zinc were highly associated mainly in urban areas, possibly due to anthropogenic inputs to coastal waters, but that a global spread (widespread dissemination) of these elements to nonurban sites may be producing weak associations. In addition, a moderately strong association was found between lead and zinc at all Alaskan sites that was also present in Pacific coast urban sites but weak at West Coast nonurban sites. We conclude from these patterns that some Alaskan sites have elevated concentrations of these metals which may be due to a common source, such as mining.

The study on extractable metals revealed interesting patterns which may help elucidate the patterns of excess accumulation. Because dissolved concentrations of elements may be elevated in contaminated areas, accumulations may result by association with external coatings of sediment particles. The natural geochemical makeup of sediment particles consists of all the elements under consideration, but most are integrated within the matrix of the particle. Release of these elements is accomplished only through total acid digestion, whereas the surface associated elements may be released with a weak acid extraction. By comparing the total amount of an element to the amount leached off the sediment particles with weak acid, we attempted to assess the amount of contaminant accumulation in sediment due to elevated concentrations of waterborne elements. Several elements displayed high extractable to total concentration ratios in sediment from urban sites indicating that a large proportion of the total element was easily displaced from the sediment surface. This pattern is consistent with the theory of excess accumulation of certain elements in sediment.

In all but a few cases, the concentrations of toxic elements in the livers of fish from urban sites were similar to those in livers from nonurban sites. In general, there was no strong correlation between element concentrations in fish liver and sediment with the one exception of mercury in flatfish. Several factors may contribute to the lack of correlation, including 1) the fact that sediments and tissues exhibit high variation at a given site within a given year; 2) insufficient assessment of other biological compartments (e.g., other tissues besides liver) that are involved in bioaccumulation; and 3) the lack of determination of the bioavailable fraction of the total concentration found in sediment. Although there were few significant differences between populations at these different sites, trends were evident and we believe that as comparison-sample sizes increase, significant differences will become more apparent.

When all individuals of a fish species are examined, some species show elevated concentrations of elements in liver, such as arsenic in flathead sole and mercury in barred sand bass and white croaker which is probably due to higher sediment concentrations at the sites where these species are found. Reduced levels of mercury in flathead sole (taken only in Alaskan waters) compared to other West Coast flatfish correlates with variable sediment concentration and is an indication of urban related bioaccumulation.

Concentrations of elements in stomach contents show much greater correlation to sediment concentrations than do those in fish liver which may be due to the relative immobility of the invertebrate prey and their closer reflection of the ambient conditions. Of course these differences could also be attributed to different abilities of the invertebrate prey and the fish to regulate accumulate elements. Although our database is small and incomplete, it appears that many elements in liver were very high in fish from reference sites which may be caused by elevated concentrations in prey. One explanation for this considers the bioavailability of metals from sediments at these reference sites, which often have very low amounts of fine sediment particles (<63 µm). Because some elements may concentrate in the fine fraction of these coarse-grain (sandy) sediments leading to high concentrations in this fraction and many sediment dwelling invertebrates select the fine fraction for ingestion, element concentrations may be higher in these species when compared to species from sediments with a higher percentage of fine particles and less element per gram of fine material. The relatively small amount of fine particles of these coarse-grained sediments contain most of the organic carbon and surfaces for element adsorption and accumulation.

Differences in element concentrations in fish liver may be due to either increased bioaccumulation at some sites over others or because of inherent species differences unrelated to contaminant exposure. In general, we assume that accumulations of nonessential elements were attributed to contaminant exposure, whereas variations for essential elements could be caused by differences in species physiology. Analysis of tissue concentrations in relation to sediment concentrations revealed that some nonessential elements (arsenic, cadmium, and mercury) displayed a strong pattern of increasing concentration in some species from sediment, to stomach contents, to liver and, therefore, were probably bioaccumulated. Others elements (copper, silver, and zinc) displayed a weak trend in some species indicating possible bioaccumulation at some locations. For the essential elements (copper and zinc) it is not possible to tell if the species differences were due to bioaccumulation from contaminated sediments or were just inherent species differences. The remainder of the elements (antimony, chromium, lead, nickel, and selenium) displayed no pattern over sediments and species indicating no bioaccumulation or species differences were found.

While this report represents 5 years of data, and a seemingly large database, many more years of sampling will be needed before definite trends can be identified. Because of the large variability observed in sediment and tissue concentrations, associations between tissues and sediments are general trends with little statistical confidence. It is encouraging to see emerging patterns of element abundance in a data set such as this with its inherent uncertainty due to spatial and temporal variability, myriad geochemical factors, and physiological mechanisms.

Collective thanks are expressed to our colleagues in the National Marine Fisheries Service (NMFS) and National Ocean Service (NOS), NOAA, who participated in the conceptualization and execution of the Pacific coast portion of the National Benthic Surveillance Project (NBSP). Special appreciation to Drs. Andrew Robertson, Douglas Wolfe, and Donna Turgeon, NOS, for their support and encouragement. Particular thanks to Dr. Charles Ehler, NOS, for his role in formulating and nurturing the National Status and Trends Program; and to Dr. Nancy Foster of the NMFS, current Nationwide Principal Investigator of the NBSP, for her strong commitment to the NBSP. We are also indebted to John Finley and Salvador Regala for conscientious laboratory support. Beth Horness, Ellen Brooke of GenWest (Seattle, WA), and Yun Bing Shi (University of Washington) assisted in preparing the floating-bar plots. Much appreciation is also given to the personnel of the NOAA ship McArthur for sample collection, Wayne Dyer and Sue Pierce for technical assistance, Sharon Giese and Barbara Bennett for editorial advice, and Lyndal Johnson and Carla Stehr for thorough manuscript review.