Seven taxa of intertidal plants and animals were sampled at 17 shoreline sites in Prince William Sound ([PWS]; AK, USA), that were heavily oiled in 1989 by the Exxon Valdez oil spill (EVOS) to determine if polycyclic aromatic hydrocarbons (PAH) from buried oil in intertidal sediments are sufficiently bioavailable to intertidal prey organisms that they might pose a health risk to populations of birds and wildlife that forage on the shore. Buried residues of EVOS oil are present in upper and middle intertidal sediments at 16 sites. Lower intertidal (0 m) sediments contain little oil. Much of the PAH in lower intertidal sediments are from combustion sources. Mean tissue total PAH (TPAH) concentrations in intertidal clams, mussels, and worms from oiled sites range from 24 to 36 ng/g (parts per billion) dry weight; sea lettuce, whelks, hermit crabs, and intertidal fish contain lower concentrations. Concentrations of TPAH are similar or slightly lower in biota from unoiled reference sites. The low EVOS PAH concentrations detected in intertidal biota at oiled shoreline sites indicate that the PAH from EVOS oil buried in intertidal sediments at these sites have a low bioavailability to intertidal plants and animals. Individual sea otters or shorebirds that consumed a diet of intertidal clams and mussels exclusively from the 17 oiled shores in 2002 were at low risk of significant health problems. The low concentrations of EVOS PAH found in some intertidal organisms at some oiled shoreline sites in PWS do not represent a health risk to populations of marine birds and mammals that forage in the intertidal zone.
Approximately 258,000 barrels (37,000 metric tons) of Alaska North Slope crude oil spilled from the tank vessel (T/V) Exxon Valdez when it ran aground on Bligh Reef in Prince William Sound (PWS), Alaska, USA, on March 24, 1989. An estimated 40% of the mass of oil released was stranded on western PWS shorelines . Extensive shoreline surveys during the spring and summer of 1989 revealed that approximately 783 linear km of PWS shoreline was oiled to some extent ; approximately 141 km of shoreline was classified as heavily oiled.
Much of the surface oil from the Exxon Valdez oil spill (EVOS) was removed from the shore in the first year after the spill due to the combination of an intensive shoreline cleanup effort and severe winter storms. The storms removed up to 90% of the surface oil from exposed, high-energy shorelines and as much as 50% of the surface oil from sheltered shores . However, weathered EVOS oil penetrated into sediments in the middle and upper intertidal zone, particularly the storm berm, of some boulder cobble beaches . Release of this weathered, emulsified oil has been slow, and some oil has persisted to the present, buried in sediments of the middle and upper shore [5,6]. Weathered oil also has persisted in finer-grained sediments underlying some intertidal mussel beds [7,8]. The persistence of the oil residues buried in intertidal sediments is attributed to their low accessibility to water-washing by tidal flushing and wave action [4,5].
Short et al.  estimated that 55.6 metric tons of subsurface oil (<0.4% of the mass of oil that is estimated to have washed ashore in PWS in 1989) remained at scattered locations on western PWS shores in 2001. They also estimated that these residues are being removed by natural processes from intertidal sediments at a rate of 20 to 26% per year. In 2005, sites with buried EVOS residues represent <0.1% of the approximately 5,000 km of PWS shoreline.
Concern exists that a fraction of the subsurface EVOS residues, and particularly the polycyclic aromatic hydrocarbons (PAH) they contain, may be bioavailable and pose a continuing risk to intertidal plants and animals and the birds and wildlife that eat them. Several field programs have addressed this concern by determining PAH bioavailability and temporal trends of total PAH (TPAH) concentration in mussels from oiled shorelines [7–10]. Few studies have measured residual PAH in other intertidal biota .
Two aspects of bioavailability are relevant to this investigation: environmental accessibility (dependent on the physical form and location of the chemical in the environment) and biological availability (the ability of the chemical to move across or bind to surface membranes of the organism, eliciting effects) . As oil weathers in intertidal sediments, it loses light hydrocarbons by evaporation, dissolution, and biodeg-radation . The viscosity of the oil increases as it weathers and develops an interfacial film, decreasing the rates of hydrocarbon dissolution and oil droplet dispersion into sediment pore water [14,15]. Mobilization of dissolved and particulate oil from the buried oil residues decreases as it weathers, decreasing accessibility (exposure) of intertidal biota to the oil. Biological availability is dependent on direct exposure of the marine organism to bioavailable forms of the chemical. Dissolved hydrocarbons are considered to be more bioavailable than particulate oil hydrocarbons . The most environmentally realistic way to assess the environmental accessibility and biological availability of PAH from buried intertidal EVOS residues to intertidal biota is to sample the resident biota on oiled shores.
The first objective of this study was to determine if PAH from EVOS residues on oiled shores are accessible and biologically available to intertidal plants and animals. The second objective was to evaluate the ecological risk to individual shore birds and sea otters that forage for intertidal biota on these oiled shores and extrapolate these risks to the entire populations of these intertidal foragers in the spill-path area of PWS.
Seventeen sites studied by Short et al.  were selected for study. Sixteen of these 17 sites contain approximately 50% of the total subsurface oil found by Short et al.  and represent worst-case conditions. Sampling protocols were designed to represent, on average, what a foraging bird or mammal would encounter at a site given the patchiness of the oil in intertidal sediments and biota.
Seven intertidal taxa sampled in this study include sea lettuce, mussels, clams, whelks, worms, hermit crabs, and inter-tidal fish. These taxa are at different trophic positions in the intertidal food chain and are representative of the types of foods consumed by intertidal avian and mammalian foragers. Several species of shorebirds forage on mussels, small fish, crustaceans, and snails in the intertidal zone . Although PWS sea otters primarily feed on benthic invertebrates collected from shallow offshore waters, juveniles and some adults include intertidal clams and mussels in their diet .
Mussels (Mytilus trossulus) live throughout the middle in-tertidal zone of gravel to boulder and bedrock shores in PWS. They are filter feeders that accumulate dissolved and partic-ulate PAH  through their gills by filtration of tidal water flowing over the shore with the tides or upwelling through upper and middle intertidal sediments. Black oystercatchers, some sea ducks, and juvenile sea otters include mussels in their diets [16,17].
Sea lettuce (Ulva fenestrate) is an ephemeral macroalga that is attached to hard substrates in the lower intertidal zone  where it is consumed by several small intertidal invertebrates . It may accumulate PAH by adsorption of dissolved and particulate PAH to its mucilaginous surface coating from surface tidal water flushing the beach and from aerial deposition during low tide of airborne PAH-contaminated combustion particles .
Infaunal clams (Protothaca staminea and Saxidomus gi-ganteus) and worms (polychaetes and nemerteans) bioaccu-mulate PAH from contaminated sediments and bottom water [11,21]. These taxa live in sand/silt sediments in the lower intertidal and shallow subtidal zones. Clams and worms are exposed to PAH in solution in sediment pore water, by in-gestion of PAH-contaminated inorganic and organic sediment particles in the lower intertidal sediments in which they reside, and in particle-laden water flushed from mid- and upper-in-tertidal sediments .
High cockscomb fish (Anoplarchus purpurescens), hermit crabs (Paguridae), and whelks (Nucella spp. and Searlesia spp.), are motile epifaunal animals that forage in the intertidal and shallow subtidal zone during the high tide and hide under rocks (fish and crabs) or attach to rock surfaces (whelks) during low tide . The whelks feed on barnacles and mussels in the middle and upper intertidal zone and the intertidal fish and crabs are carnivores/scavengers throughout the middle to shallow subtidal zones. These intertidal animals primarily are exposed to PAH through their food. They are consumed by shore birds and sea ducks .
Seventeen sites that were oiled in 1989 and surveyed in the 2001 shoreline study  were selected for study (Table 1). Sixteen of these 17 sites contained about 0.80 acres or approximately one-half of the approximately 1.63 acres of subsurface oil found by Short et al.  at 42 sites in 2001. That study found no subsurface oil at site SQ002A. One oiled site sampled in 2001 , KN136A in Sleepy Bay on Knight Island, originally was selected to be sampled, but it was not because it had been disturbed by sediment sampling activities shortly before we arrived. Two of the oiled sites have a recorded history of past human activity and are classified as oiled human activity (HA) sites. In addition, three unoiled reference sites and four unoiled sites of former human activities (unoiled HA), now abandoned, were selected to define the condition of shoreline biota at locations in PWS that had not been oiled in 1989. The unoiled HA sites include two former herring processing sites, a mining community, and a fox farm, and were selected arbitrarily as representative of a group of nearly 50 coastal HA sites identified earlier . Table 1 and Figure 1 summarize the site locations.
We collected subsurface sediments from the intertidal zone and, where possible, the seven species of intertidal plants and animals at each site. Sediment samples were collected only to confirm that buried EVOS residues were present at a site and to establish their location with respect to tidal elevation; biota samples were collected to determine if the buried EVOS residues were accessible and bioavailable to intertidal plants and animals at those sites and to determine the ecological risk of the bioaccumulated PAH to intertidal foragers. Sampling was performed between June 7 and 27, 2002.
A shoreline segment up to 100 m long (depending on the length of exposed, nonbedrock substrate available in the shoreline segment), positioned at the same location as the 2001 survey segment , was delineated in the intertidal zone of each site. Mean tidal height in western PWS is approximately 3 m with maximum spring tides of approximately 5 m. Transects were established parallel to the shore at three tide levels relative to mean low low water (MLLW): +3 m (upper intertidal zone); +2 m (mid-intertidal zone); and 0 m (low tide zone). Most of the buried oil reported in 2001 by Short et al.  was located between the +1.8-m and +3.3-m tidal heights; thus, +2-m and +3-m transects were sampled to maximize the likelihood of detecting buried oil in the sediments. A lower intertidal transect (0 m) also was sampled to determine if any buried oil was present in sediments where most of the shoreline biota reside.
Up to 11 sampling stations (depending on transect length) were established at 10-m intervals along each transect. Global positioning system (GPS) coordinates of the ends and midpoint of each transect were recorded. The GPS coordinates for each grid (Table 1) are defined by the midpoint value of the +2.0-m transect.
Table Table 1.. Intertidal sampling stations in western Prince William Sound (AK, USA). Total extractable hydrocarbon (TEH) concentrations are the arithmetic means, by transect, for subsurface sediments collected along three transect elevations. Total polycyclic aromatic hydrocarbon (TPAH) concentrations and fluoranthene + pyrene to C2–4 phenanthrene (FP:[FP+C24P]) concentration ratios are for sediment composites from each transect. Site IDs are the 1990-amended Shoreline Cleanup Assessment Team segment designations
Mean TEH (mg/kg dry wt)
TPAH (ng/g dry wt)
a Oiled human activity (HA) site.
Mean for all oiled sites
Unoiled reference sites
Mean for all unoiled reference sites
Unoiled human activity sites
Mean for all unoiled HA sites
Biota samples were collected at each sampling site before pits were dug in order to avoid contamination of the biological samples with runoff from the pits. All biota samples were collected as close as possible to the transect lines. The plants and animals from the lower intertidal zone (sea lettuce, clams, worms, and intertidal fish) were collected during a minus tide within ±1 to 2 m laterally of the lower intertidal transect line (0 m MLLW). The middle intertidal animals (mussels, whelks, and hermit crabs) were collected along and as close as possible to the +2-m transect line. These animals were most abundant at and for a short distance down-shore from the +2-m transect.
Where possible, at least 25 g (wet wt of soft tissue) of seven taxa of intertidal organisms were collected from the full length of the grid and at the tidal level that the species ordinarily inhabit. Three composite blue mussel samples, averaging approximately 15 to 20 individuals per sample and with each composite representing approximately one-third of the +2-m transect length, were collected at each site. Samples were placed in precleaned glass jars or wrapped in precleaned aluminum foil, returned to the survey vessel, inspected to ensure that only the target species were included, and then frozen.
After collection of biota, a pit was dug to a depth of 50 cm or to bedrock at each station along each transect. On boulder-cobble beaches, the base of the boulder-cobble veneer was defined as the 0-cm depth. A composite sample of the fine-sediment fraction was collected from the entire vertical wall of each pit directly into a precleaned glass jar. Consequently, each sediment sample was collected to represent an average for the pit wall. The jar was sealed immediately and frozen subsequently onboard the survey vessel. All sediment and biota samples were shipped frozen in coolers by airfreight to the analytical laboratory.
Site panoramas were photographed, site sketches were prepared, and photographs of each pit were taken for documentation.
A 10- to 20-g sediment sample from each pit was homogenized, dried, and extracted with dichloromethane. Total ex-tractable hydrocarbons (TEH) were measured in each sediment sample by gravimetric analysis (U.S. Environmental Protection Agency method 1664A ). Total extractable hydrocarbons concentrations in sediments are reported in mg/kg dry weight (parts per million). The U.S. Environmental Protection Agency gravimetric method for total oil and grease measures the total nonvolatile, nonpolar organic matter that is soluble in dichlo-romethane, including petroleum and other natural and anthropogenic nonpolar organic compounds. Therefore, it is nonspecific as to source. It was assumed that most of the sediment TEH at the 17 oiled sites would be EVOS oil residues [5,6].
A single homogeneous composite sediment sample was prepared from sediment samples collected from all the pits along each transect and extracted; the extract was analyzed for PAHs. These PAH analyses were used to identify the sources of hydrocarbons and to confirm the presence of EVOS residues in sediments at the oiled sites as well as their tide-zone distribution.
Total and individual PAH in sediment and biological tissue samples were analyzed by methods described by Boehm et al.  and Page et al. . The methods include solvent extraction of a 40- to 50-g (dry wt) sediment sample or a 1- to 2-g (dry wt) soft tissue sample, column cleanup, and analysis by gas chromatography with analyte identification and quantification by mass spectrometry in the selected ion-monitoring mode (modified U.S. Environmental Protection Agency method 8270 [7,25,26]). This analytical protocol provides method detection limits of 0.5 ng/g dry weight (parts per billion) or lower for individual parent and alkylated PAH in sediments and tissues . The method detection limits for C2- through C4-phen-anthrenes, fluoranthene, and pyrene (used to identify PAH sources) range from 0.06 to 0.44 ng/g. The analytical method allows fingerprinting (source identification) of samples containing low concentrations of TPAH.
Target PAH analytes quantified included the parent and alkyl-PAH analyzed in most Exxon Valdez impact studies (e.g., [7,25]). These include 2- through 6-ring parent PAH and the C1 through C3 or C4 homologous series of alkylated naphthalenes, fluorenes, phenanthrenes, dibenzothiophenes, chrysenes, and fluoranthenes/pyrenes.
Small amounts of parent naphthalene were present in laboratory procedural blank samples, probably derived from laboratory contamination. This contamination is accentuated in samples with low total PAH concentrations, resulting in an overestimation of TPAH concentration in sediment and biota samples containing the lowest TPAH concentrations. Consequently, TPAH values reported in this paper exclude parent naphthalene concentrations. The parent naphthalene represented less than 4% of TPAH in mussels containing more than 100 ng/g TPAH; thus, exclusion of parent naphthalene from TPAH had little effect on the results of this study. All sediment and tissue PAH concentrations are reported as ng/g dry weight (parts per billion) TPAH minus parent naphthalene.
Traces of C1-naphthalenes, fluorine, phenanthrene, fluoranthene, and pyrene were present in some tissue procedural blanks. Concentrations were low and variable, so these analytes were not subtracted from the TPAH concentrations in tissue samples. Thus, some tissue TPAH concentrations may include some of the PAH found in blanks.
Analysis of variance with least-square means contrasts were used to compare TPAH concentrations among transect elevations and among shoreline types. For testing sediment TPAH concentrations among transect elevations at oiled sites, the following analysis of variance model was used:
where μ is the grand mean over elevations. Least-square means contrasts were made on 0-m, 2-m, and 3-m sampling elevations. Sediment TEH concentrations and the fluoranthene + pyrene:(fluoranthene + pyrene + C2-C4-phenanthrenes (FP: [FP+C24P]) ratios were assessed similarly. We tested associations between TEH and TPAH concentrations with Pearson's correlation.
Least-square means contrasts also were made for tissue TPAH concentrations among shoreline types (oiled, unoiled reference, and unoiled HA) using the following analysis of variance model:
Again, μ is the grand mean over shoreline types. Taxa were tested separately.
Before analysis, TPAH and TEH concentrations were log-transformed to achieve additivity and normality . Normality of error terns was assessed with qqnorm plots . All tests of significance were at α = 0.05. We report concentrations, their arithmetic means, and standard deviations in the tables. Though statistical analyses were performed on log-transformed values, we chose not to report geometric means and log-based standard deviations in order to better present the wide range of concentrations observed.
RESULTS AND DISCUSSION
Total extractable hydrocarbons in subsurface sediments
The sediment sampling design was intended to provide a semiquantitative estimate of mean extractable hydrocarbon concentrations from all sources in intertidal sediments at the 17 oiled, three reference, and four unoiled HA sites. Each sample provided the average TEH concentration in the finegrained sediment fraction from the 0- to 50-cm depth in a pit. The TEH concentrations of all pit samples from a transect were averaged to provide a mean for that tide level at each site (Table 1).
Table Table 2.. p Values for least-square means contrasts of sediment total polycyclic aromatic hydrocarbon (TPAH) and total extractable hydrocarbon (TEH) and the fluoranthene + pyrene to C2–4 phenanthrene (FP:[FP+C24P]) concentration ratio. Concentrations by intertidal elevation. Concentrations significantly differed (α = 0.05) for 0 m versus 2 m and 0 m versus 3 m, but not for 2 m versus 3 m
0 m vs 2 m
0 m vs 3 m
2 m vs 3 m
Arithmetic mean concentrations of TEH by transect in sediments at the 17 oiled sites are highly variable within and among sites, reflecting the patchy distribution of buried oil on the shore. The variability is greater in upper (+3-m transect) and middle (+2-m transect) intertidal sediments than in lower intertidal (0-m transect) sediments. The overall mean TEH concentrations in upper and middle intertidal sediments are approximately 10 times and six times higher, respectively, than the mean TEH concentration in lower intertidal sediments; these differences are statistically significant (Table 2).
For most pits at oiled sites with visible weathered oil, the oil is concentrated in a 2- to 10-cm thick band mixed with fine-grained sediments at a depth of 10- to 30-cm below the boulder/cobble or pebble/cobble surface veneer . At some oiled sites, the band of oiled sediment is at or near the bottom of the pit, sometimes perched on a layer of peat. Oil is visible in 12.9% of the 502 pits dug on oiled shores in this study. Nearly 70% of the observations of oil in pits are recorded as traces or sheen. Only 20 pits (4%) contain a visible oily layer of sediment or oil-filled pore spaces.
Despite the low frequency of visible oil residues in pits, most transect mean TEH concentrations for upper and middle intertidal sediments at the 17 oiled sites are higher than the transect mean TEH concentrations at unoiled reference sites.
Total extractable hydrocarbon concentrations in sediments from the three unoiled reference sites and two of the unoiled HA sites are low (Table 1), reflecting background concentrations of natural and anthropogenic extractable organic matter. Intertidal sediments at two unoiled HA sites that are locations of former herring processing facilities, have mean TEH concentrations that are higher than in sediments at the other reference sites and comparable to those at the most heavily oiled sites.
Total PAH in subsurface sediments
Mean TPAH concentrations in subsurface sediments at oiled sites are highest and most variable along the upper (+3-m) and middle (+2-m) intertidal zone transects (Table 1). The mean TEH and TPAH concentrations in upper and middle intertidal sediments are not statistically different (Table 2). However, lower intertidal (0-m) sediments contain significantly lower mean TEH and TPAH concentrations than upper and middle intertidal sediments.
Intertidal sediments at two unoiled HA sites (KN575P and EV500A) that were locations of herring processing facilities contain high mean TPAH concentrations (Table 1). The sources of this chronic contamination are located above the tide zone at both sites. Site EV500A is somewhat unusual in that, unlike the other sites (both oiled and unoiled HA) having elevated mean TPAH concentrations, the highest mean TPAH concentration occurs in sediments in the lower intertidal zone and concentration decreases with increasing tidal elevation; we have no explanation for the difference.
Mean TPAH concentrations in intertidal sediments from all tide zones at the two other unoiled HA sites and three unoiled reference sites uniformly are low (6–47 ng/g TPAH).
Statistically significant (p < 0.001) positive correlations between log-transformed mean concentrations of TEH and TPAH in upper and middle intertidal sediments (+2 m and +3 m combined) from oiled (r = 0.76) and unoiled HA (r = 0.95) sites indicate that concentrations of TEH and PAH co-vary in upper and middle intertidal sediments. The PAH in these sediments are associated with the solvent-extractable organic phase. The mean concentration of TPAH associated with TEH in sediments at oiled sites ranges from 0.2 to 0.6% by weight; mean values for unoiled HA site sediments range from 0.6 to 6.6% by weight.
PAH compositions in subsurface sediments
The relative concentrations of individual PAH in the PAH assemblages in sediments of the 17 oiled shoreline sites vary widely among sites and among elevations in the intertidal zone. This is caused by variations in the relative proportions of petro-genic and pyrogenic PAH components and in the weathering state of the hydrocarbon mixtures. Alkyl naphthalenes, phen-anthrenes, and dibenzothiophenes (i.e., 2–3 ring alkyl-PAH) are the most abundant PAH in most upper and middle intertidal sediment samples from most oiled sites (Figs. 2 and 3). These PAH profiles are consistent with an identifiable component of weathered EVOS  at all but one oiled site (SQ02), where no oil was found in 2001 or 2002.
The PAH assemblage in lower intertidal sediments at most oiled sites is depleted in alkyl naphthalenes and alkyl dibenzothiophenes, reflecting a low-sulfur, non-EVOS hydrocarbon component. Potential non-EVOS petrogenic contributors include Monterey (CA, USA) tars , kerogen particles and seep residues from the natural background derived from the eastern Gulf of Alaska , and low-sulfur refined products including diesel fuel .
Pyrogenic (combustion-derived) PAH assemblages often contain high relative concentrations of acenaphthene, biphe-nyl, anthracene, and 4- through 6-ring PAH, and the parent and sometimes monomethyl-PAH are more abundant than more highly alkylated PAH [12,32]. These characteristics can be used to differentiate between petrogenic and pyrogenic PAH assemblages in environmental samples. The concentration ratio of fluoranthene plus pyrene (pyrogenic) to the sum of C2- through C4-phenanthrenes (petrogenic), expressed as the ratio FP:(FP+C24P), is a useful pyrogenic indicator for sediment and tissue samples [32,33]. The Exxon Valdez oil that has undergone different degrees of natural weathering has a relatively narrow FP:(FP+C24P) range of 0.01 to 0.02. However, ratios as high as 0.15 are observed in some extensively degraded shoreline oils, possibly reflecting pyrogenic PAH contamination . Creosote (a coal tar product) from wharf pilings at HA site EV500A has a FP:(FP+C24P) ratio of 0.77 . Combustion particles (soot) from burning a mixture of pine and oak has a FP:(FP+C24P) ratio of 0.62 . Sediments with FP:(FP+C24P) ratios greater than about 0.2 are interpreted to have a pyrogenic PAH component .
The FP:(FP+C24P) ratios in upper, middle, and lower intertidal sediments at the 17 oiled sites range from 0.01 to 1.00, revealing that most of the sites have PAH distributions that are mixtures of petrogenic and pyrogenic components (Table 1). The contribution of pyrogenic PAH to TPAH concentrations at these oiled sites frequently increases from the top of the shore to the low tide line. Mean sediment FP: (FP+C24P) ratios at oiled sites increase from 0.17 at +3-m to 0.23 and 0.32 at +2-m and 0-m tidal levels, respectively (Table 1). The mean for +3 m is significantly lower than for 0 m; mean ratios for 0 m versus +2 m and +2 m versus +3 m are not significantly different (Table 2). The +2-m mean ratio is nominally lower than the 0-m mean ratio and higher than the 3-m mean ratio (Table 1). This nominal trend is evident in sediments at some sites including EV016A, EV039A (Fig. 3), and LA020C.
Some sites where the sediments have high FP:(FP+C24P) ratios, such as Rua Cove (KN213B) and Snug Harbor (KN403A), both on Knight Island, are locations with well-documented histories of past industrial activity . Other sites, such as Northwest Bay (EL058B) on Eleanor Island (heavily oiled in 1989 but with no record of past commercial activity), have pyrogenic ratios that indicate a mixture of pyrogenic and petrogenic PAH. Northwest Bay is a favorite anchorage for pleasure and fishing boats. Diesel and gasoline engine exhaust is a source of pyrogenic PAH, particularly fluoranthene and pyrene , that may have contaminated the shoreline sediments.
The PAH profiles in sediments from the two heavily contaminated, unoiled HA sites are dominated by high molecular weight, predominantly pyrogenic, PAH (Fig. 4). The FP: (FP+C24P) ratios for these unoiled HA site sediments range from 0.37 to 0.90, with little distinction between tide zones (Table 1). Small amounts of petrogenic PAH are present at some locations. Petrogenic PAH at these sites are from storage tanks that are leaking low-sulfur, Monterey (CA, USA) heavy oil [30,31]. Pyrogenic PAH are from creosote pilings and combustion products, primarily from fossil fuel burning in support of fish-processing operations.
Mean FP:(FP+C24P) ratios for sediments from the other, less-contaminated, unoiled HA sites and from the three unoiled reference sites generally are >0.2, indicating a mixture of pyrogenic and petrogenic PAH. These pyrogenic PAH probably are derived in part from deposition of airborne particulate PAH from combustion soot.
TPAH concentrations in intertidal organisms
Seven intertidal taxa were targeted for collection at all sites; however, one or more of clams, hermit crabs, whelks, and worms could not be found at a few sites due to the absence of suitable habitat. For example, at some high-energy boulder-cobble sites, fine-grained sediment habitat for clams and worms is not present in the lower intertidal zone. Thus, the absence of one or more taxa from the intertidal zone of a sampling site primarily is caused by lack of habitat availability.
Concentrations of TPAH vary widely within and among species of intertidal organisms (Table 3, Fig. 5). The TPAH concentrations in individual tissue composites from oiled sites range from 1.65 ng/g dry weight in a whelk sample to 137 ng/g in a clam sample. Although mean tissue TPAH concentrations were not compared statistically among taxa, some general patterns are discernable. Highest arithmetic mean concentrations are in worms, clams, and mussels at oiled sites. Lowest mean concentrations are in whelks and hermit crabs. The mean TPAH concentration in intertidal fish (high cockscomb) is 22.4 ng/g. However, fish from two of the 17 oiled sites have higher TPAH concentrations that affect the mean for that species; 15 of the 17 intertidal fish samples contain less than 15 ng/g TPAH, and the other two contain 66.4 and 246 ng/g. Sea lettuce samples from oiled sites have a mean TPAH concentration of 11.3 ng/g.
Mean TPAH concentrations in intertidal biota from unoiled reference sites range from 3.3 ng/g dry weight in whelks to 15.2 ng/g in clams (Table 3). Mean TPAH concentrations in mussels, clams, worms, and fish, but not hermit crabs, sea lettuce, and whelks, are slightly lower than mean TPAH concentrations in the same species from oiled sites. Mean TPAH concentrations in tissues of the seven taxa from oiled and reference sites are not significantly different (Table 4). Mean tissue TPAH concentrations are significantly higher in mussels, clams, and sea lettuce from unoiled HA sites than in the same species from oiled sites (Table 4). The arithmetic mean tissue TPAH concentration is nominally highest in all taxa from un-oiled HA sites (Table 3). However, the unoiled HA-sites include two sites (EV500A and KN575P) that are contaminated with PAH and two sites (KN575A and ST001A) where sediment and biota TPAH concentrations are comparable to those in sediments and biota from the three unoiled reference sites. Intertidal biota from the two contaminated HA sites contain higher TPAH concentrations than the biota from the two un-contaminated HA sites.
Total PAH concentrations in most individual fish, hermit crab, whelk, and sea lettuce samples from oiled sites are in the same range as the laboratory procedural blanks; the blanks contain traces of naphthalene, alkylnaphthalenes, and occasionally fluorine, phenanthrene, fluoranthene, and pyrene. Tissue concentrations in the range of procedural blanks probably are overestimates and should be considered semiquantitative. The low tissue TPAH concentrations indicate low bioavail-ability of EVOS PAH from intertidal sediments for these species. Polycyclic aromatic hydrocarbons are more bioavailable to clams, mussels, worms, and fish than to crabs, whelks, and sea lettuce at the oiled sites. These deposit feeders, filter feeders, and scavengers often retain particulate PAH in the gut and gills without assimilating (bioaccumulating) them [12,18]. Sea lettuce from the two contaminated HA sites contain high concentrations of primarily pyrogenic PAH.
Sources of PAH in tissues of intertidal organisms
Polycyclic aromatic hydrocarbon profiles in tissues vary widely among the seven species of intertidal plants and animals and in the same species among oiled, unoiled HA, and unoiled reference sites. Many of the organisms collected at oiled sites and containing greater than background levels of TPAH (background defined as concentrations in unoiled reference site organisms, ranging from 3–15 ng/g; Table 3) have a mixture of petrogenic and pyrogenic PAH as indicated by FP:(FP+C24P) ratios >0.20 (Table 3).
Mussels, clams, and worms collected from oiled sites in this study contain both petrogenic and pyrogenic PAH; surface-dwelling intertidal fish, hermit crabs, whelks, and sea lettuce primarily contain pyrogenic PAH (Table 3; Figs. 2 and 3). Mussels appear to bioaccumulate primarily particulate, petrogenic PAH from middle intertidal sediments. Clams and worms appear to bioaccumulate primarily pyrogenic, particulate PAH from lower intertidal sediments or from oil-contaminated fine sediment particles washing down the shore, whereas the other intertidal organisms bioaccumulate primarily combustion-sourced particulate PAH that can originate from boat engine exhaust and from aerial deposition.
Table Table 3.. Mean and standard deviation (SD) concentrations of total polycyclic aromatic hydrocarbons (TPAHs) in tissues of seven species of intertidal organisms from 17 oiled shores, four unoiled human activity (HA) reference shores, and three unoiled reference shores. Mean and range of the fluoranthene + pyrene to C2–4 phenanthrene concentration ratio, FP:(FP+CP24P), are included for each species and location. The TPAH concentrations are ng/g dry weight
aNS = not sampled.
b ID = indeterminate.
c Oiled HA site.
Unoiled reference sites
Unoiled HA sites
Table Table 4.. p Values for least-square means contrasts of tissue total polycyclic aromatic hydrocarbons (TPAHs) concentrations by taxon and shoreline type: Oiled, unoiled reference (ref), and unoiled human activity (HA)
Oiled vs ref
Oiled vs HA
Ref vs HA
At site SM006B (north shore of Smith Island) where mean concentrations of EVOS TPAH are 14,400 and 12,900 ng/g in upper and middle intertidal sediment composites, respectively (Table 1), concentrations of TPAH in tissues of marine organisms range from 2.8 ng/g in a composite sample of high cockscomb fish to 5.8 to 85 ng/g (average 54.9 ng/g) in three composite samples of mussels collected at different locations along the +2-m transect (Table 3). Total PAH concentrations in tissues are more than two orders of magnitude lower than mean concentrations in site sediments, indicating a low order of bioavailability of PAH from buried intertidal EVOS oil.
The PAH residues in the mussels are petrogenic (FP: [FP+C24P] = 0) and the PAH profile is consistent with a weathered EVOS oil source. The PAH assemblage in the mussels is depleted in alkyl naphthalenes and fluorenes, the most soluble alkyl PAH in crude oils, and enriched in alkyl chry-senes, the least soluble of the petrogenic PAH, relative to the PAH assemblage in upper and middle intertidal sediments. This indicates that the mussels bioaccumulate the PAH in the form of particles of highly weathered petroleum.
Tissues of worms and clams collected in the lower intertidal zone at SM006B have low TPAH concentrations (∼20 ng/g) dominated by C1- and C2-naphthalenes (Fig. 2E, F). Low concentrations (1–2 ng/g) of alkyl phenanthrenes and dibenzo-thiophenes indicate that some of the PAH are petrogenic and might be from EVOS. The FP:(FP+C24P) ratios of 0.49 and 0.17 for worms and clams, respectively, indicate that the PAH are petrogenic-pyrogenic mixtures. The low concentrations of petrogenic PAH in tissues of clams and worms from this heavily oiled site indicate that the oil residues are not eroding or dissolving at high rates from the upper and middle intertidal sediments and, therefore, have a low accessibility and bioavailability to sediment-dwelling animals living in the biologically productive lower intertidal zone.
Clams, worms, and fish from oiled site EV039A on the north shore of Evans Island contain low concentrations of pyrogenic-petrogenic PAH mixtures (Fig. 3). Polycyclic aromatic hydrocarbons in worms are dominated by C2- and C3-fluoranthenes/pyrenes and alkyl dibenzothiophenes, phenanthrenes, and chrysenes of probable EVOS origin. Clams contain higher concentrations of C2- and C3-fluoranthenes/pyrenes, but lower concentrations of alkyl phenanthrenes and dibenzothiophenes than the worms. The FP:(FP+C24P) ratios in the worms (0.23) and clams (0.44) indicate that the PAH assemblages are mixtures from pyrogenic and petrogenic sources. Associated lower intertidal sediments at this site have low concentrations of PAH (∼20 ng/g) also derived from both petrogenic and pyrogenic sources (Table 1).
Mussels from the middle intertidal of oiled site KN109A primarily contain petrogenic PAH (FP:[FP+C24P] = 0.09–0.16). Worms and clams from the lower intertidal zone predominantly contain pyrogenic PAH (FP:[FP+C24P] = 0.64 and 0.72, respectively; Table 3). Upper and middle intertidal sediments mostly contain petrogenic PAH (FP:[FP+C24P] = 0.12 and 0.01, respectively). Lower intertidal sediments primarily contain lower concentrations of pyrogenic PAH (FP: [FP+C24P] = 0.6). This distribution of petrogenic and pyrogenic PAH in sediments and biota is similar to distributions at most other oiled sites; it suggests that mussels bioaccumulate particulate petrogenic PAH (probably from EVOS) from upper and middle intertidal sediments and that worms and clams primarily bioaccumulate pyrogenic PAH, probably in particulate form and probably from lower intertidal sediments.
High cockscomb fish from the lower intertidal zone of EV039A contain a PAH assemblage dominated by fluoran-thene and pyrene (FP:[FP+C24P] = 0.99), indicating that the PAH assemblage in the fish is pyrogenic. Apparently, there is a nearby source of pyrogenic PAH in the lower intertidal zone of EV039A. These intertidal fish consume small lower intertidal invertebrates and macroalgal detritus and might have ingested combustion soot particles from boat engine exhaust or aerial deposition. Hermit crabs have a similar diet and contain a higher tissue TPAH concentration at this site than at other oiled sites. The TPAH in the crab tissues are pyrogenic, supporting the hypothesis that there is a pyrogenic PAH source at this oiled site.
High cockscombs from all other oiled, reference, and unoiled HA sites, except the contaminated HA sites EV500A and KN575P, contain background TPAH concentrations. The PAH in the fish tissues are pyrogenic, except at oiled site KN113A, where the fish contain 66 ng/g petrogenic PAH (Table 3); the PAH profile in fish tissues from KN113A is consistent with weathered EVOS oil.
High cockscombs , crustaceans, and polychaete worms  have an active mixed function oxygenase system capable of metabolizing PAH. Higher molecular weight PAH, such as fluoranthene and pyrene, usually are metabolized and excreted more rapidly than lower molecular weight, alkylated PAH, such as C2-phenanthrenes . However, alkyl PAH tend to be bioaccumulated more rapidly and to higher tissue concentrations than the parent, unalkylated PAH because of their greater hydrophobicity (higher log KOW). Polycyclic aromatic hydrocarbon bioaccumulation and metabolism in these animals tends to decrease the FP:(FP+C24P) ratio toward a more petrogenic value. Therefore, the presence of a PAH assemblage enriched in high molecular weight, unalkylated PAH in worm, crustacean, and fish tissues indicates that the PAH primarily are pyrogenic and can be present in an unassimilated, particulate form, possibly in the gut.
Intertidal biota from contaminated unoiled HA sites EV500A and KN575P primarily contain pyrogenic PAH (FP: [FP+C24P] > 0.8; Table 3), probably derived from lower intertidal sediment PAH (FP:[FP + C24P] = 0.7–0.8) at EV500A and, potentially, from all tide zones at KN575P Unalkylated phenanthrene, dibenzothiophene, and chrysene are more abundant than their alkyl homologs in all intertidal species at both sites. All intertidal plants and animals analyzed appear to be bioaccumulating particulate, pyrogenic PAH at these sites.
Intertidal biota from all tide levels at the three unoiled reference sites, the uncontaminated HA sites (ST001A and KN575A), and many of the 17 oiled sites contain low concentrations of TPAH that primarily are derived from combustion products (FP:[FP+C24P] > 0.5). Clams, mussels, and worms from some oiled sites contain PAH from petrogenic or mixed, petrogenic-pyrogenic sources.
Risk to intertidal foragers of PAH from buried EVOS residues
Some investigators have hypothesized that birds and wildlife that forage on oiled shores in PWS still are being exposed to weathered oil residues, either in their food or by direct contact with intertidal oil deposits, slowing population recovery [38–40]. However, these investigators have not demonstrated that intertidal foragers are bioaccumulating sufficient PAH from EVOS residues in intertidal sediments and prey animals to cause harmful population-level effects.
Bodkin et al.  hypothesized that sea otters foraging in the intertidal zone during high tides could be exposed to pockets of residual oil they uncover while digging for clams. Sea otters ordinarily do not come onto the shore. At locations where clams are abundant in the lower intertidal zone, some otters forage on the lower shore during high tides by diving and digging pits in the +1-m to — 1-m MLLW zone [41,42]. However, most foraging is at greater water depths, down to at least 50 m ; only 6% of the monitored dives of sea otters in the northern Knight Island area of PWS were in the intertidal zone . Regardless, mean TPAH concentrations in lower intertidal sediments (0 m MLLW) at the 17 formerly oiled shores sampled in this study range from 3.2 to 177 ng/g (parts per billion). Sea otters digging clams or sea ducks and shore birds foraging in otter pits in lower intertidal sediments containing these low concentrations of TPAH might contaminate their fur or feathers with traces of weathered EVOS oil, but amounts of PAH present likely are so low that they would not pose a health risk to the intertidal foragers. Therefore, it is unlikely that sea otters and marine birds are becoming contaminated with harmful amounts of EVOS oil residues by direct contact with oily intertidal sediments during forage pit digging and foraging in pits in the lower intertidal zone.
Table Table 5.. Chronic no-observed-adverse-effects levels (NOAELs) for polycyclic aromatic hydrocarbons (PAHs) in the diet of black oystercatchers (Haematopsis machmani) and sea otters (Enhydra lutris). BW = body weight. The NOAELs are mg/kg/d. Toxicity data are from Environmental Restoration Division (; www.srs.gov/general/programs/soil/ffa/rd/p73)
The results of the present study show that birds and wildlife also are unlikely to bioaccumulate harmful concentrations of petroleum PAH by consumption of intertidal prey from oiled shores. As discussed above, there is not a statistically significant difference in the mean concentrations of TPAH in tissues of the seven intertidal taxa collected on oiled and reference shores (Table 4). Thus, the dose of TPAH in the diet of in-tertidal foragers likely is similar on oiled and unoiled shores, except at some sites of former human activity. Mussels and clams, the most important intertidal foods of most intertidal foragers, contain similar concentrations of TPAH at oiled and reference sites and significantly higher TPAH concentrations at the two contaminated unoiled HA sites.
Although sea otters in PWS primarily eat clams, some, particularly juveniles, also consume some mussels [17,39,41,42]. The diets of several species of sea ducks (e.g., goldeneyes and harlequin ducks) and shore birds (e.g., black oystercatchers) include mussels . The percentage of the total energy budget derived from eating mussels is 30, 95, and 10% for black oystercatchers, Barrow's goldeneyes, and harlequin ducks, respectively, the only birds documented to consume large numbers of mussels in PWS . Favorite foods of harlequin ducks overwintering in PWS are amphipods and other small crustaceans . Black oystercatchers also consume worms, crustaceans, and barnacles . Concentrations of petroleum PAH are low in these preferred prey species (clams, mussels, crabs, and worms) collected from oiled shorelines.
No published criteria exist for directly determining safe, no-effect dietary exposure concentrations of TPAH for sea otters or shore birds. Therefore, no-observed-adverse-effects levels (NOAELs) were derived from the literature following state and federal ecological risk assessment guidance [46–48]. Toxicity studies for several PAH were identified from the literature, and the highest dose tested that did not cause adverse effects was noted (Table 5). Several of the studies identified in Table 5 are subchronic (i.e., days to weeks of exposure). Ecological risk guidance recommends application of an uncertainty factor of 10 to adjust results of subchronic studies to a chronic NOAEL value . In addition, the test animals identified in Table 5 are not the species of interest (sea otters and shore birds). The NOAELs were adjusted with a body weight scaling equation recommended by Sample et al. 
The chronic NOAELs for black oystercatchers and sea otters were estimated by comparing the dietary dose of TPAH in oystercatchers and sea otters from consumption of clams and mussels on oiled shores to the chronic NOAELs for each PAH for which a NOAEL is available, to obtain an indication of the range of relative hazard to these wildlife species from consumption of PAH-contaminated prey (Table 5). Chronic NOAELs for individual PAH in black oystercatchers and sea otters range from 0.019 to 4.83 mg/kg/d (Table 6). The lowest chronic NOAEL is for benzo[a]pyrene, which is present at only trace concentrations in EVOS oil ; therefore, this value is a conservative estimate of the chronic toxicity threshold of the TPAH residues in food to the foragers.
Table Table 6.. Parameters for estimating the ingested polycyclic aromatic hydrocarbon (PAH) dose to sea otters and black oystercatchers from consumption of clams and mussels from the 17 oiled shorelines. The average energy content of clams and mussels is 22.5kJ/g dry weight. Bioenergetics data are from Hartung  and Dean et al. 
aOne clam sample contained 137 ng/g total polycyclic aromatic hydrocarbon (TPAH); the next highest concentration was 68 ng/g TPAH in mussels, equivalent to ingested doses of 3.1 and 3.9 μg/kg body weight/d in sea otters and oystercatchers, respectively.
Total PAH in mussel/clam diet (ng/g dry wt)
Animal weight (kg)
Consumption rate (kJ/kg body wt/d)
Clam/mussel consumption (g dry wt/kg body wt/d)
Ingested PAH dose (mg/kg body wt/d)
Table Table 7.. Health risk estimates, expressed as hazard quotients (HQs) for black oystercatchers (Haematopsis machmani) and sea otters (Enhydra lutris) from consumption of clams and mussels in 2002 from shorelines in Prince William Sound (AK, USA) that were oiled by the Exxon Valdez oil spill. Maximum oral dose and no-observed-adverse-effect level (NOAEL) (from Tables 5 and 6) are mg/kg body weight/d
TPAHb oral dose
TPAH oral dose
a PAH = polycyclic aromatic hydrocarbon.
b TPAH = total PAH.
c HQ ≥ 1.0 indicates a potential for adverse health effects in the consumer.
The health hazard to black oystercatchers and sea otters from consumption of clams and mussels from oiled shores can be estimated by determining the highest TPAH dose ingested by these animals during foraging for mussels and clams on oiled shores and dividing this oral dose by the chronic NOAELs for individual PAH to obtain a hazard quotient. A hazard quotient ≥1.0 indicates that the dietary dose may be high enough to cause adverse health effects in the consumer. Sea otters and oystercatchers consuming a diet exclusively consisting of the clams and mussels from the 17 oiled sites would consume a diet containing 5 to 137 ng/g TPAH. This is equivalent to an ingested dose of 0.3 to 6.2 μg/kg body weight/d for sea otters and 0.3 to 7.9 μg/kg body weight/d for oystercatchers as a worst-case scenario (Table 6). The worst-case hazard quotients for oystercatchers and otters range from 0.0002 to 0.033, indicating a substantial safety margin for the consumers (Table 7). Thus, the PAH residues in the clams and mussels collected from the 17 oiled sites in 2002 would not pose a significant health risk to sea otters and oystercatchers.
Ferrets and mink (closely related to otters) and mallard ducks (a surrogate for shore birds) fed a diet containing EVOS oil with equivalent TPAH concentrations of 1,000 to 50,000 ng/g primarily suffered mild sublethal effects [49,50]. These dietary EVOS oil concentrations are more than 100 times higher than the concentration in the oil-contaminated clam diet of otters and oystercatchers that consume a diet of mussels and clams from oiled shores. Even if sea otters and oystercatchers are more sensitive than ferrets, mink, and ducks to PAH, it is unlikely that they would be harmed by consuming mussels and clams from oiled shorelines in 2002, or even as early as the mid-1990s, when TPAH concentrations in mussels on most oiled shores had declined to low, ecologically safe levels .
Mussels on some oiled shores in PWS were contaminated heavily with EVOS oil after the spill. In the few years immediately after the spill, some mussels in mid-intertidal mussel beds (dense aggregates of mussels overlying fine-grained sediments) contained high concentrations of EVOS PAH [7,8]. However, the mussels in the oiled mussel beds and the more abundant mussels on rocky intertidal substrates of heavily oiled shores have released accumulated EVOS PAH at an estimated rate of 25% per year . Several studies independently have shown that, by 1998 to 2002, EVOS TPAH concentrations in mussels were at or near background [10,51]. Thus, populations of intertidal foragers have not been at risk of harm from eating intertidal prey from oiled shorelines in PWS since the mid-1990s and perhaps earlier.
The results of this study show that, 13 years after the Exxon Valdez oil spill, the levels of petrogenic PAH from EVOS residues in intertidal biota are below levels known to cause harm to populations of intertidal foragers. These results are consistent with those obtained in broader studies of EVOS residues in mussels [7–10]. The EVOS PAH concentrations are low in intertidal prey organisms on formerly oiled shores because the EVOS residues that have persisted in intertidal sediments to 2002 are in a form that has a limited mobility. Consequently, PAH from weathered EVOS in sediments have a low accessibility to shoreline plants and animals. This is because most of the oil is buried 10 or more cm deep in upper or middle intertidal sediments that are armored by a layer of boulders and cobbles. The veneer protects the deposits from erosion by tidal and wave action. Only traces of EVOS residues are present in sediments of the biologically more productive lower intertidal zone.
Sites containing buried EVOS residues today represent <0.1% of the shoreline of the entire sound. This fact, combined with the low bioavailability of the residues, demonstrates that the residues pose little or no risk to populations of birds and mammals that forage on the shore.
We acknowledge the support of Captain D. Murphy and crew of the motor vessel Spirit of Glacier Bay. This study was funded by Exxon Mobil Corporation.