Selected metal and metalloid contaminants are commonly used to assess impacts of human activities on urban embayments. Sediments serve as repositories for these contaminants and integrate variable amounts found in the water column. As stated in the Methods section, significant differences among means were determined with a criterion of alpha < 0.05; hence, on the floating-bar plots the range of the comparison intervals must be mutually exclusive for statistically significant differences. Where several reference sites are utilized, the statistical comparison is always made with the reference site with the highest concentration the analyte being discussed. Lutak Inlet (Alaska) was not used for reference comparison because it was not sampled every year as were the other reference sites. All element concentrations are based on dry weight measurements of the sediments and all means and standard deviations are reported as mean (sd).
Antimony, Arsenic, and Cadmium
Floating-bar plot presentations of the antimony (Sb), arsenic (As), and cadmium (Cd) in sediments are given in Figure 3.
Antimony--Mean antimony concentrations varied by about an order of magnitude between 0.18 and 1.5 µg/g, with all sites overlapping with the maximum reference site (Nisqually Reach UCI: 1.14 µg/g) (Figure 3). If the next highest reference site (Dana Point) was used, only one site (Elliott Bay) would be significantly higher than the reference upper comparison interval. The exploratory sites with single determinations grouped towards the lower concentrations: Farallon Islands (0.61 µg/g), San Luis Obispo (0.49 µg/g), Dana Point Inside (0.44 µg/g), outside Mission Bay (0.42 µg/g), and Estero Bay (0.32 µg/g).
Although antimony is very toxic to mammals and is subject to controls to protect public health, very little is known about its abundance in most Pacific coast estuarine and coastal waters. One major anthropogenic source of antimony is lead smelting (Furness and Rainbow 1990). With a history of lead smelting in the Pacific Northwest, perhaps it is not surprising that the Elliott Bay and Commencement Bay sites have mean concentrations grouped in the highest five sites on Figure 3. Chapman et al. (1986) reported a limit of detection (LOD) of 50 µg/g which was considerably higher than our LOD of 0.1 µg/g and any of the concentrations we determined for the San Francisco Bay region (maximum at Redwood City at 0.85 (0.11) µg/g and minimum at Bodega Bay at 0.62 (0.32) µg/g).
Robertson and Abel (1990) found 0.72 µg/g in Kamishak Bay whereas our mean (sd) was 0.43 (0.06) µg/g. Robertson and Abel also reported 0.69 µg/g off Kotzebue vs. our Chukchi Sea NBSP mean of 0.18 (0.03) µg/g and 0.74 µg/g off Oliktok vs. our NBSP mean of 0.53 (0.23) µg/g. Additionally, Robertson and Abel (1990) found 0.70 µg/g off Prudhoe Bay which was the same as our mean Endicott Field concentration of 0.70 (0.2) µg/g.
Arsenic--Arsenic values in surface sediments from most Pacific coast sites were similar (means range 0.63-13 µg/g) and did not vary significantly from the reference site (Dana Point UCI: 9.3 µg/g) (Figure 3). However, the mean concentration of arsenic in sediment from the Oakland estuary was the highest (mean (sd): 13 (1.6) µg/g) and was significantly higher than the nearby reference site at Bodega Bay (4.3 (2.5) µg/g). The analytical results for single-sediment samples (not shown on Figure 3) collected from Dana Point Inside (8.3 µg/g), outside Mission Bay (6.4 µg/g), Estero Bay (5.0 µg/g), San Luis Obispo (1.5 µg/g), and the Farallon Islands (<0.1 µg/g), ranged from the high end down to below detection.
Arsenic can exist in a variety of chemical forms in marine and estuarine ecosystems, including inorganic species, methylated forms, arseno-lipids, arseno-sugars, arsenobetaine, and arsenocholine (Phillips 1987). Arsenic chemistry in the aquatic environment is unusually complex and may be influenced by a host of interactions and processes, such as redox state, ligand exchange, precipitation, adsorption, and biological interactions (Sadiq 1990).
Crecelius et al. (1985) reported surface sediment concentrations of 15-20 µg/g at locations surrounding our NBSP Commencement Bay site (3.8 (2.8) µg/g). They also determined that preindustrial concentrations were near 7 µg/g. The study by Schults et al. (1987) reported a value from a site off the mouth of Blair Waterway of 24 µg/g, with concentrations as great as 288 µg/g inside other local waterways.
Our Coos Bay mean value of 6.9 (4.0) µg/g compares with a value around 5 µg/g reported by Fuhrer and Rinella (1983). Chapman et al. (1986) reported high arsenic sediment concentration at their sites in San Francisco Bay: 44-70 µg/g in southwestern San Pablo Bay (our NBSP mean (sd): 8.1 (3.7) µg/g), 49-64 µg/g off Oakland (NBSP: 13 (0.3) µg/g), and 57-72 µg/g in Islais Creek (NBSP: 9.4 (1.9) µg/g). Gunther et al. (1987) reported that the main source of arsenic to San Francisco Bay was riverine inputs and nonurban runoff.
In Southern California Jan and Hershelman (1980) reported a mean arsenic concentration of 3.7 µg/g in the vicinity of the NBSP west Santa Monica Bay site (5.0 (1.8) µg/g). Ladd et al. (1984) reported arsenic sediment levels in north San Diego Bay of 6.5 µg/g , which was comparable to our NBSP mean of 7.7 (1.7) µg/g. They also reported values of 4.8-20 µg/g (NBSP mean: 10 (3.0) µg/g) for south San Diego Bay and 5-13 µg/g (NBSP mean: 9.4 (5.3) µg/g) for National City.
Robertson and Abel (1990) reported 8.1 µg/g arsenic in Kamishak Bay sediments near our NBSP site where we found only 1.8 (0.7) µg/g. They also reported 8.4 µg/g for Kotzebue vs. our mean (sd) Chukchi Sea site value of 6.4 (1.4) µg/g, and 25.7 µg/g mean for five sites in the Beaufort Sea vs. our Endicott Field site mean of 1.3 (0.3) µg/g. Our Oliktok Point site contained 2.3 (2.0) µg/g in sediment.
Cadmium--The only site with cadmium sediment mean values significantly greater that the highest reference site (Nisqually Reach UCI: 0.86 µg/g) was west Santa Monica Bay (2.1 µg/g) (Figure 3). In the section on baseline metals and excess concentrations in urban sediments, we discuss the possibility that cadmium at Nisqually Reach was elevated beyond normal background concentrations; hence, this site was excluded in the determination of background cadmium concentrations. The reference site with the next lowest concentration (Dana Point) was used instead; however, only one additional site (San Pedro Canyon) was significantly elevated above this reference site's upper comparison interval (0.55 µg/g).
The single-sediment sample collected from the outside Mission Bay site (3.3 µg/g) (not shown on Figure 3) was one of the highest cadmium values found on the Pacific coast. The other single-sample sites were lower: Dana Point Inside (1.2 µg/g), San Luis Obispo (0.39 µg/g), the Farallon Islands (0.14 µg/g), and Estero Bay (0.08 µg/g).
Crecelius et al. (1985) reported recent surface sediment concentrations of cadmium of 0.17-0.38 µg/g at Commencement Bay (NBSP mean (sd): 0.64 (0.3) µg/g) and determined that concentrations in sediment were below 0.17 µg/g before the area was industrialized. Schults et al. (1987) reported that their site near the mouth of Blair Waterway (Commencement Bay) contained 0.26 µg/g cadmium, with more elevated levels to 3.9 µg/g inside nearby waterways. Our mean (sd) values for the Columbia River estuary and Youngs Bay sediments (0.58 (0.31) and 0.13 (0.02) µg/g, respectively) were lower than those reported by Fuhrer and Rinella (1983) of 2 and 4 µg/g. Their value of approximately 3 µg/g was also greater than the NBSP mean (0.44 (0.22) µg/g) at Coos Bay.
Available data for cadmium in San Francisco Bay sediments indicate little evidence of widespread or heavy enrichment (Phillips 1987) and that riverine input is probably the dominant local input source (Gunther et al. 1987). Chapman et al. (1986) found 1 µg/g or less at their San Pablo Bay, Outer Oakland, and Islais Creek sites which agrees with our mean (sd) NBSP data (0.29 (0.22) , 0.16 (0.006) and 0.51 (0.24) µg/g, respectively).
Word and Mearns (1978) calculated that 28 control stations of their 71-station 60 m depth contour survey in southern California had a mean cadmium level of 0.39 µg/g. Near the Hyperion sludge outfall, Jan and Hershelman (1980) reported 2.3 µg/g (NBSP mean (sd) of 2.1 (2.5) µg/g). Off Palos Verdes, north of our San Pedro Canyon site (1.17 ( 0.54) µg/g) and San Pedro Outer Harbor site (0.82 (0.38) µg/g), Hershelman et al. (1981) reported 2.1 µg/g. These elevated levels at NBSP sites may have been due in part to the closeness of local sewer outfalls: 2.5 km from Hyperion and 8+ km from the Whites Point outfall for the Los Angeles County Sanitation District.
In San Diego Bay, Ladd et al. (1984) reported 0.5-1.0 µg/g in sediment from north San Diego Bay (NBSP mean: 0.52 (0.17) µg/g), 0.5-2.0 at their south San Diego Bay site (0.68 (0.4) µg/g), and 1.5-12.5 µg/g off National City (0.36 (0.16) µg/g). In Hecate Strait, south of Boca de Quadra (0.44 (0.01) µg/g), Harding and Goyette (1989) reported a mean (sd) cadmium concentration of 0.38 (0.08) µg/g.
Floating-bar plot presentations of the chromium (Cr), copper (Cu), and lead (Pb) in sediments are given in Figure 4.
Chromium--Bodega Bay, a reference site, had the highest mean value (490 µg/g) of any site and was, therefore, not useful as a reference site. We used the next highest reference site (Nisqually Reach UCI: 140 µg/g) and found that five sites were deemed significantly higher than reference. These were Bodega Bay, Humboldt Bay (453 (250) µg/g), San Pablo Bay (361 (215) µg/g), Hunters Point (234 (82) µg/g), and South Hampton Shoal (213 (56) µg/g).
Of the 13 sites with the highest sediment concentration, all were around the greater San Francisco Bay area, except for two (Chukchi Sea off Kotzebue and Port Valdez, both in Alaska). A single-sediment sample collected from Estero Bay (5770 µg/g) was the highest single concentration found on the Pacific coast (not shown on Figure 4). The other single-sample values ranged from 210 µg/g in the Farallon Islands off San Francisco to 53 µg/g at the outside Mission Bay site, with the others being intermediate: San Luis Obispo (130 µg/g) and Dana Point Inside (59 µg/g). Chromium distribution in sediment is obviously related to geographical area and requires further investigation.
Chromium exists in nature in valence states ranging from Cr-2 to Cr+6, the most common forms are the trivalent and hexavalent forms. Trivalent chromium, essential at low concentrations for mammals, is generally considered to be less toxic than hexavalent chromium. Human activities remobilize considerably more chromium than do natural processes, largely due to the element's use in the metallurgical and chemical industries. While most chromium is thought to enter through atmospheric pathways, direct sources to aquatic environments are from metal finishing industries and sewage treatment plants plus inputs from iron and steel works, tanneries, textile manufacturers, and urban runoff (Phillips 1987).
Crecelius et al. (1985) reported Commencement Bay surface sediment concentrations of 49-72 µg/g around our NBSP site (61 (8) µg/g), which was similar to preindustrial sediment levels. In their study of this area, Schults et al. (1987) report their highest concentration at Blair Waterway (14.3 µg/g ) , which was far below our NBSP site.
Fuhrer and Rinella (1983) reported levels of chromium at 4 and 8 µg/g for surface sediments from the Columbia River estuary. Our mean (sd) Columbia River estuary (40 (13) µg/g) and Youngs Bay (50 (3.2) µg/g) sites were much higher. They also report about 20 µg/g for Coos Bay compared with our NBSP mean of 80 (57) µg/g.
Chapman found chromium concentrations of 72-93 µg/g in their southern San Pablo Bay site (northeast San Pablo Bay NBSP mean (sd): 361 (215) µg/g), 85-95 µg/g at their Outer Oakland Harbor site (NBSP: 196 (4) µg/g), and 110-146 µg/g in Islais Creek Channel (NBSP: 145 (7) µg/g). Gunther et al. (1987) estimates that nonurban runoff is the dominant input (62-83 %) to the San Francisco Bay area with riverine contributions next in importance. This suggests that the high chromium levels found in San Francisco Bay and Bodega Bay are not anthropogenic, but rather reflect crustal rock contributions.
In southern California, Word and Mearns (1978) calculated a control site concentration of chromium along the 60 m depth contour to be 23.1 µg/g. Near the Hyperion outfall, Jan and Hershelman (1980) found 69 µg/g chromium compared to our NBSP mean (sd) of 101 (43) µg/g. Hershelman et al. (1981) reported a value of 77 µg/g off of Palos Verdes, north of our San Pedro Canyon (106 (30) µg/g) and San Pedro Outer Harbor (80 (14) µg/g) sites. Our mean (sd) values for San Diego Bay (north: 62 (14) µg/g; south: 90 (48) µg/g; and National City: 60 (22) µg/g) generally agree with the values reported by Ladd et al. (1984) of 61-79, 80-158, and 83-140 µg/g for these sites, respectively.
Sweeney and Naidu (1989) reported surface sediment chromium levels of 16-125 µg/g for the area around Oliktok Point (NBSP mean (sd): 71 (16) µg/g) and 15-87 µg/g in the Prudhoe Bay area (NBSP Endicott Field mean: 59 (16) µg/g), while Robertson and Abel (1990) reported 89 µg/g off Oliktok Point and 85 µg/g off Prudhoe Bay (Endicott Field). Robertson and Abel (1990) also found 65 µg/g chromium near our NBSP site in Kamishak Bay (82 (6) µg/g) and 333 µg/g off Kotzebue near our Chukchi Sea site for which we reported a mean (sd) of 200 (24) µg/g. In Hecate Strait, south of Boca de Quadra (NBSP: 53 (17) µg/g), Harding and Goyette (1989) reported chromium concentrations of 28.8 (14.5) µg/g.
Copper--Copper occurs in excess concentrations at many sites, particularly those associated with areas of urbanization or marine transportation and repair activities (Figure 4). The mean (sd) copper concentration in Pacific coast sediments was significantly higher than the maximum reference site (Nisqually Reach upper comparison interval (UCI): 26 µg/g) in the urbanized areas of San Diego: north bay (103 (23) µg/g), south bay (181 (46) µg/g) and National City (117 (73) µg/g); Los Angeles: San Pedro Outer Harbor (102 (31) µg/g), Long Beach (67 (16) µg/g), and west Santa Monica Bay (60 (34) µg/g); San Francisco: Oakland estuary (115 (22) µg/g), Oakland Outer Harbor (72 (5) µg/g), Redwood City (55 (4) µg/g), and Hunters Point (53 (10) µg/g); and central Puget Sound: Seattle's Elliott Bay (105 (36) µg/g) and Tacoma's Commencement Bay (48 (7) µg/g).
In the section on baseline metals and excess concentrations in urban sediments, we discuss the possibility that copper at Nisqually Reach was elevated beyond normal background concentrations; hence, this site was excluded in the determination of baseline copper concentrations. The reference site with the next lowest concentration (Dana Point) was used instead leading to 13 additional sites that have been determined to be significantly higher than the reference site's upper comparison interval (9 µg/g). In this section we also address the variability of copper concentrations as a function of sediment grain size. It should be kept in mind that copper concentrations vary widely with grain size making it difficult to determine contamination levels.
Copper is an essential element in vertebrates, being associated with numerous metalloenzymes and metalloproteins (Thompson 1990). Its mobilization by human activities is like that of lead; roughly 10-15 times nature's rate (Phillips 1987). After mercury and silver, copper is ranked as the third most toxic to aquatic biota of the more common metal contaminants (Waldichuk 1974), and the margin between its concentration in the marine environment and those concentrations known to be at least chronically toxic to marine biota is very small (Klapow and Lewis 1979).
The concentrations of copper in sediments found in Puget Sound were variable. Long and Chapman (1985) summarized data from several sources and reported copper concentrations at 34-206 µg/g at various Elliott Bay sites (NBSP mean (sd): 105 (37) µg/g). Crecelius et al. (1985) reported copper contents of surficial sediments from Commencement Bay ranging from 46 to 65 µg/g (NBSP mean (sd): 48 (7) µg/g ) and a preindustrial sediment concentration for copper to be in the range of 30-40 µg/g. Long and Chapman (1985) reported 51-581 µg/g at Commencement Bay sites with high values associated with shallow inshore sites, especially in channels and waterways with restricted water circulation. Schults et al. (1987) reported 37 µg/g off the mouth of the Blair Waterway with higher levels inside other waterways (to 581 µg/g) and lower at their reference site (11 µg/g). Dexter et al. (1985) discussed historical inputs to Puget Sound from such sources as the former ASARCO copper smelter on the western shore of Commencement Bay and Quinlan et al. (1985) concluded that up to one-third of the total anthropogenic copper to central Puget Sound was attributed to this source. This smelter discharged both liquid and atmospheric particulates enriched in copper, and smelter slag was used for sandblasting of ships, as rip rap, and for roadbed materials. In addition, copper can be contributed by other shipyard operations and the gradual dissolution of antifouling paints from vessels in industrial ports, such as Seattle's Harbor Island and Duwamish Waterway. Compared with other surface waters in the Puget Sound Basin, Paulson and Feely (1985) found elevated levels of copper in inner Commencement Bay. They also found surface waters with elevated dissolved copper concentrations in the vicinity of Harbor Island in Elliott Bay which suggested localized anthropogenic inputs (Paulson and Feely 1985).
Fuhrer and Rinella (1983) reported 4 µg/g of copper just north of our NBSP Columbia River estuary site compared with our mean (sd) value of 16.8 (3.2) µg/g. Also, our Youngs Bay site was not as high (36 (9) µg/g) as their site (180 µg/g). The NBSP mean value for Coos Bay was 13 (12) µg/g which was slightly less than their value of 25 µg/g.
Previous surveys of San Francisco Bay area sediments show an enrichment of copper (>50 µg/g) in nearshore stations of South Bay and in much of San Pablo Bay (some sites >100 µg/g), with generally lower levels (<25 µg/g) in the sediments of Central Bay and the Golden Gate area (Phillips 1987). Oakland Outer Harbor has levels exceeding 75 µg/g, compared with our value of 72 (5) µg/g and Oakland estuary values were reported exceeding 50 µg/g compared with our mean value of 115 (22) µg/g (Phillips 1987). Chapman et al. (1986) reported 30-53 µg/g for southern San Pablo Bay (our NBSP mean: 40 (18) µg/g), south of Oakland Outer Harbor levels of 43-51 µg/g, and high levels in Islais Creek Channel of 68-130 µg/g (NBSP mean (sd): 73 (16) µg/g).
Word and Mearns (1978) reported coastal southern California baseline concentrations of copper to be 2.8 to 31 µg/g. Jan and Hershelman (1980) reported 42 µg/g, and Brown et al. (1986) found levels of 25 µg/g near the NBSP west Santa Monica Bay site, compared to our 60 (34) µg/g. Young et al. (1980) showed levels of copper of 40-617 µg/g near our west Santa Monica Bay site with strong concentration gradients for copper and other metals in the immediate vicinity of the Hyperion sewage outfall. Hershelman et al. (1981) reported 45 µg/g at their closest site to our San Pedro Canyon (31 µg/g) and San Pedro Outer Harbor (102 (32) µg/g) sites.
Our north San Diego Bay NBSP site contained 103 (23) µg/g compared with Ladd et al. (1984) who reported values of 28.5 and 42.5 µg/g for nearby sites. Our south San Diego Bay site mean was 181 (46) µg/g compared to the Ladd et al. (1984) finding of 25-340 µg/g. On the west side of the channel off National City we found 117 (73) µg/g whereas Ladd et al. (1984) reported 2,000 and 19,000 µg/g closer to the pier heads. The high sediment concentrations in San Diego Bay may reflect the presence of a major copper ore shipping facility at the south end of the bay.
On Alaska's North Slope, Sweeney and Naidu (1989) reported copper values near Oliktok Point of 6.3-83 µg/g compared with our value of 18 (6) µg/g. Our lower Endicott Field value of 12 (4) µg/g compared with Sweeney and Naidu's reduced range of 9-29 µg/g in the Stefansson Sound-Prudhoe Bay area. Likewise, our mean copper sediment concentration of 60 (6) µg/g compared favorably with the Feder et al. (1990) value of 59 for Port Valdez. The levels in Port Valdez could reflect natural inputs of river-borne suspended material. For example, Feely et al. (1981) found levels of 63 µg/g in suspended particulate matter from the nearby Copper River, and nearshore northeast Gulf of Alaska suspended particulate matter contained 55-109 µg/g. In Hecate Strait, south of Boca de Quadra (NBSP mean (sd): 22 (5) µg/g), Harding and Goyette (1989) reported copper concentrations of 4.3 (0.8) µg/g. The copper in Dutch harbor sediments could reflect extensive fishing boat and shipping activities concentrated in this confined water body on Alaska's Aleutian Island chain. The more remote Port Moller had only 13 (1.5) µg/g copper in surface sediments which is probably related to grain size.
Analytical results for copper in single-sediment samples (not shown on Figure 4) collected from Farallon Islands (8 µg/g), San Luis Obispo (8 µg/g), Estero Bay (6 µg/g), and outside Mission Bay (2 µg/g), all ranged below the maximum reference site value (26 µg/g), except for the sample collected in the entry channel to the large Dana Point marina, Dana Point Inside (40 µg/g). Copper was not detected (<1.0 µg/g) in sediments collected from the Channel Islands site.
Lead--The mean (sd) lead concentrations in Pacific coast sediments were significantly higher than the reference site (Dana Point upper comparison interval; 22 µg/g dry weight) at Long Beach (101 (36) µg/g), Oakland estuary (99 (10) µg/g), north (53 (10) µg/g) and south (80 (31) µg/g) San Diego Bay, and San Pedro Outer Harbor (41 (11) µg/g). The sites with the highest three lead values (Long Beach, Oakland estuary, and south San Diego Bay) were significantly higher than 28 of the sites (except Port Valdez) beginning with Commencement Bay (Figure 4).
Lead has no known biological function or requirement and is classified as nonessential (Thompson 1990). It is ubiquitous in terrestrial and aquatic environments due to numerous anthropogenic sources and its natural rate of mobilization is exceeded by human contributions by a factor of 10-15 (Phillips 1980). Industrial effluents and urban runoff may contain large quantities of lead, the latter largely through the deposition of lead on pavement surfaces from its use as an antiknock agent in automotive fuels, an important nonpoint source discharge of lead to urban environments (Gunther et al. 1987). This element is unusual in that it exhibits only moderate toxicity in aquatic environments but is highly toxic to mammals (Phillips 1987).
Lead in Elliott Bay (Seattle) was high (47 (27) µg/g) possibly due to a lead smelter in Seattle which had historically contributed lead contamination to the area, although the discharges were significantly curtailed in the early 1980s (Dexter et al. 1985). Other anthropogenic sources of lead adjacent to the Duwamish Waterway also appear to have resulted in elevated dissolved lead concentrations in Elliott Bay surface waters (Paulson and Feely 1985).
Previous studies of metals in sediments (Malins et al. 1982, Long and Chapman 1985) in Puget Sound located in areas adjacent to NBSP sites found levels similar to those reported here. For example, Schults et al. (1987) reported a lead concentration of 19 µg/g off the mouth of the Blair Waterway, slightly inshore from the Commencement Bay NBSP site where we found 26.8 (10.7) µg/g. The exhaustive study by Shults et al. routinely found much higher levels (to 790 µg/g) inside the waterways, intermediate levels in Commencement Bay, and lowest levels at their reference site (Browns Point; 9 µg/g). Crecelius et al. (1985) reported 25 µg/g lead in sediment collected north of the NBSP site and a predevelopment (pre-1900) concentration in central Commencement Bay of less than 15 µg/g. A major local input of lead to the Tacoma area was from the now-closed ASARCO lead smelter established along the northwestern shore of Commencement Bay in 1889 and converted to copper smelting in 1902. Input has been as liquid wastes, shoreline deposition of smelter slag, and as atmospheric particulates enriched in metals (as much as 23 metric tons of lead per year).
Our mean (sd) concentration of lead (9.9 (6.5) µg/g) in the Columbia River estuary compared favorably with a value of 10 µg/g found by Fuhrer and Rinella (1983, as reported in Buchman 1989) and our Youngs Bay, Oregon site (17.8 (1.5) µg/g) was further inshore than the Fuhner and Rinella site which contained 10 µg/g. Their value for Coos Bay sediments in 1980 (approx. 13 µg/g) compared favorably with our 1984-88 mean of 10.8 (7.4) µg/g.
Lead concentrations have been reported to be fairly evenly distributed over the four basins of San Francisco Bay with no areas reaching 100 µg/g; however, several nearshore areas had higher levels (Phillips 1987). Risebrough et al. (1978) reported levels of 40-69 µg/g for Oakland Outer Harbor near our NBSP site (44 (2) µg/g), 70-99 µg/g for Hunters Point (NBSP mean (sd): 30 (15) µg/g), 40-69 µg/g off Angel Island near our Southampton Shoal site (17 (11) µg/g), and 10-39 µg/g from San Pablo Bay sediments (NBSP mean: 16 (10) µg/g). Luoma and Phillips (1988) reported 54 µg/g at their Hunters Point location. Chapman et al. (1986) reported surface sediment lead levels from southern San Pablo Bay of 18-25 µg/g, south of the mouth of the Oakland estuary at 29-33 µg/g, and 49-223 µg/g from the Islais Creek Channel (NBSP mean: 47 (17) µg/g) .
Word and Mearns (1978) reported a mean sediment lead concentration of 6.6 µg/g for their 28 "control" stations along the 60 m depth contour from Point Conception to the U.S.-Mexico border in southern California and Katz and Kaplan (1981) calculated a "pollution-free baseline" concentration for the area to be 10 µg/g. The majority of the NBSP sites were located more inshore of these control stations and contained up to three time higher concentrations at the NBSP open water and reference sites.
Young et al. (1980) reported levels of 20-50 µg/g in the area of our west Santa Monica Bay site (27 (14) µg/g) and 20-30 µg/g in the inshore area (NBSP mean 26 µg/g), while Jan and Hershelman (1980) reported a mean lead value of 17 µg/g near the Hyperion sewage sludge discharge outfall. Our mean (sd) San Pedro Canyon sediment concentration (17 (3) µg/g) was similar to that found by Hershelman et al. (1981) at their closest Palos Verdes station (Sta. 10A: 15 µg/g) south of the Joint Water Pollution Control Plant's Palos Verdes outfall. Thompson et al. (1986) reported lead values (4-9 µg/g) in deeper-water sediments off the San Gabriel Canyon, south and west of our sites. The NBSP Long Beach site mean lead concentration was 3-4 times greater than nearby sites and probably reflects multiple urban inputs, including the large input of street runoff from the Los Angeles River drainage which enters San Pedro Bay near this site (Eganhouse and Kaplan 1982).
Ladd et al. (1984) reported levels of lead in San Diego Bay sediments from a 1982-83 nearshore survey ranging from 15 to 315 µg/g, with their south San Diego Bay sites having 15-150 µg/g (n = 3) compared with our NBSP site's mean (sd) of 80 (31) µg/g; north San Diego Bay sites having 22-31 µg/g (n = 2) compared with our 53 (10) µg/g; and National City (20-200 µg/g; n = 2) compared to our mean value of 44 (19) µg/g.
Our mean (sd) value for Port Valdez was 23 (5) µg/g which compares well with a mean value of 30 µg/g for a site slightly east of the NBSP site (Feder et al. 1990). For sites near Skagway (Skagway and Nakhu Bay), the loading of lead ores inside that harbor was a source of air-borne lead dust (Robinson-Wilson and Malinkay 1986). In Hecate Strait, south of Boca de Quadra (NBSP mean (sd) 18 (3) µg/g), Harding and Goyette (1989) reported a mean (sd) lead concentration of 11.8 (10.7) µg/g. Analytical results for single-sediment samples (not shown on Figure 4) collected from Dana Point Inside (26.2 µg/g), Farallon Islands (10.3 µg/g), San Luis Obispo (4.9 µg/g), outside Mission Bay (4.9 µg/g), and Estero Bay (<0.3 µg/g) all ranged near or below the maximum reference site value (25 µg/g).
Floating-bar plot presentations of the mercury (Hg), nickel (Ni), and selenium (Se) in sediments are given in Figure 5.
Mercury--The total mercury concentration in Pacific coast sediments was quite variable at many sites but significantly higher than the maximum reference site (Nisqually Reach UCI: 0.31 µg/g) at three of the urban sites: Oakland estuary (1.07 (0.32) µg/g), south (0.72 (0.53) µg/g) and north San Diego Bay (0.61 (0.20) µg/g). Mercury was not detected (<0.01 µg/g) in sediments from Moss Landing, east Santa Monica Bay, or Oceanside. In the section on baseline metals and excess concentrations in urban sediments, we discuss the possibility that mercury at Nisqually Reach was elevated beyond normal background concentrations; hence, this site was excluded in the determination of baseline mercury concentrations. The reference site with the next lowest concentration (Dana Point; UCI: 0.2 µg/g) was used instead; however, this caused only two additional sites (Seal Beach and Elliott Bay) to become significantly elevated above the upper comparison interval of this reference site.
Mercury has no known biological function and can be classified as nonessential (Thompson 1990). Human activities tend to mobilize two to three times as much mercury as do natural global processes (Phillips 1987) and many inputs are concentrated along coastal margins. The element is highly toxic in the methylated form, which may be produced by microbial methylation of inorganic mercury in sediments and elsewhere. Industrial sources of mercury include mining and smelting, fossil fuel combustion, chlor-alkali plants, manufacturing of electrical equipment, pulp and paper mills, antifouling or mildew-proofing activities, and fungicides in agriculture (Phillips 1987).
Published data for mercury in Puget Sound and San Francisco Bay sediments emphasize the ubiquity of this element which may correlate with atmospheric and runoff inputs (Dexter et al. 1985, Crecelius et al. 1985, Phillips 1987). Large quantities of mercury can enter an estuary attached to suspended river-borne particles that settle to the bottom surficial sediments. Natural (e.g., storms, flooding) and anthropogenic processes (e.g., dredging, dumping) undoubtedly contribute to the redistribution of this element throughout the estuaries (Phillips 1987). For example, San Francisco Bay may receive considerable quantities of mercury from the large San Joaquin-Sacramento Rivers drainage basin upstream of the delta. Mercury sources in the Central Valley include weathering from naturally rich deposits in the coast range and washdown of mercury used in historical gold mining activities (3,500 tons shipped into the area) in the Sierra Nevada Range. Our data show that mercury concentrations in sediments from San Francisco Bay and Puget Sound cause most of these sites to be ranked in the upper half of all sites (Figure 5).
Dexter et al. (1985) reported the average surficial sediment mercury content in the Puget Sound central basin to be about 0.3 µg/g. Crecelius et al. (1985) found values of 0.04-0.15 µg/g for sites in Commencement Bay (NBSP mean (sd): 0.14 (0.12) µg/g) and reported preindustrial concentrations less than 0.08 µg/g.
Our findings of concentrations of mercury in the Columbia River estuary (0.11 (0.11) µg/g) and Youngs Bay (0.02 (0.02) µg/g) sediments were considerably lower than those (10 and 30 µg/g, respectively) reported by Fuhrer and Rinella (1983). Their value for Coos Bay was also higher (0.24 µg/g) than the mean concentration (0.14 (0.16) µg/g) for our NBSP site, but probably not significantly higher.
Phillips (1987) and Long et al. (1988) reviewed knowledge about the mercury content of San Francisco Bay sediments. Long and coworkers calculated a mean value of 0.43 (0.46) µg/g from more than 1,000 analyses (1970-87). Risebrough et al. (1978) reported levels of 0.25-0.49 µg/g off Treasure Island and Angel Island (NBSP mean (sd): 0.50 (0.6) µg/g at Oakland and 0.13 (0.09) µg/g at Southampton Shoal) and less than 0.25 µg/g at San Pablo Bay (NBSP mean: 0.23 (0.54) µg/g) and Hunters Point (NBSP mean: 0.25 (0.18) µg/g). Our mean (sd) mercury concentration in sediments from Islais Creek Channel of 0.32 (0.09) µg/g was at the low end of the range of 0.37-1.2 µg/g found by Chapman et al. (1986), the source of which is thought to be runoff from the surrounding highly urbanized catchment (Phillips 1987).
In other areas, Hershelman et al. (1981) reported mercury in sediments collected midway between the Los Angeles County Sanitation District's Palos Verdes Peninsula outfall and our NBSP San Pedro Canyon site (NBSP mean (sd): 0.31 (0.28) µg/g) of 0.20-0.42 µg/g. Also, in Hecate Strait, Alaska, south of Boca de Quadra (0.04 µg/g), Harding and Goyette (1989) reported mean (sd) mercury concentrations of 0.05 (0.01) µg/g.
Analytical results for single-sediment samples (not shown on Figure 5) collected from the Farallon Islands (0.09 µg/g), San Luis Obispo (0.08 µg/g), Estero Bay (0.08 µg/g), Dana Point Inside (0.04 µg/g), and outside Mission Bay (<0.01 µg/g) were all less than the reference sites.
Nickel--Only one site, Oakland estuary (125 (24) µg/g), contained a total nickel concentration that was significantly higher than the maximum reference site (Bodega Bay: upper comparison interval 59 µg/g ) (Figure 5). The Nisqually Reach site had a maximum interval of 55 µg/g which suggests that the Bodega Bay value is not anomalously high, assuming the source for both is geological rather than anthropogenic. The San Francisco Bay-northern California area appears to have higher sediment nickel concentrations than the other major urban estuaries; for example, the 10 sites with the highest concentrations of nickel (>47 µg/g) were in this area, accounting for essentially all of the NBSP sites in the region. Interestingly, Moss Landing (16 (3) µg/g) and Monterey Bay (7 (4) µg/g) to the south and Coos Bay (23 (16) µg/g) to the north were lower.
Nickel is considered only moderately toxic in aquatic environments; according to Phillips (1987) its rate of mobilization by human activities (metal finishing industries, urban runoff, battery manufacturing and disposal, municipal effluents) is only twice that of natural processes.
Crecelius et al. (1985) reported values of 31-37 µg/g in Commencement Bay sediments surrounding the NBSP site (28 (3) µg/g) and their nickel profiles with depth to strata before urbanization were relatively constant at these levels.
Phillips (1987) summarized existing sediment results and concluded that there was little spatial variation of nickel in San Francisco Bay sediments and that the levels were generally close to 100 µg/g. This level can be compared to a naturally-weathered, average shale with a nickel content of 95 µg/g. Gunther et al. (1987) suggested that the major sources of nickel to the San Francisco Bay environment are riverine, not directly anthropogenic. However, data suggest there is a nickel source to the South Bay environs and to a lesser extent in the Sacramento/San Joaquin estuary (Phillips 1987). Risebrough et al. (1978) reported nickel concentrations from surface sediments at several comparable locations to our NBSP sites in San Francisco Bay. Their two San Pablo Bay sites (<74 µg/g and 75-99 µg/g) were close to our San Pablo Bay site (71 (25) µg/g) and their Angel Island site (75-99 µg/g) was just west of our Southampton Shoal site (64 (17) µg/g). They also reported a concentration for Oakland Outer Harbor (125-149 µg/g) which was near our site of the same name (104 (4) µg/g), and their Bayview Park site (125-149 µg/g) was inshore of our Hunters Point site (81 (31) µg/g). Chapman et al. (1986) reported 76-85 µg/g in southwestern San Pablo Bay, 72-84 µg/g off Oakland, and 88-94 µg/g in Islais Creek Channel (Islais Creek NBSP mean: 108 (35) µg/g).
In southern California, Word and Mearns (1978) established a mean of 12.2 µg/g for nickel (n = 28; range 8.6-15.9 µg/g) at control locations along their 60 m depth contour survey from Point Conception to the U.S.-Mexico border. Katz and Kaplan (1981) suggested that an average background (nonpolluted) surface sediment concentration for southern California was about 15 µg/g and a polluted peak mean was 107 µg/g. Our west and east Santa Monica Bay site means (sd) were 13 (3) and 14 (10) µg/g, respectively and our San Pedro Canyon and San Pedro Outer Harbor values were 23 (3) and 37 (20) µg/g, respectively. Jan and Hershelman (1980) reported a mean nickel sediment value of 10.5 µg/g near the Hyperion outfall and Hershelman et al. (1981) found variable nickel concentrations ranging from 16 to 134 µg/g off the Palos Verdes Peninsula.
Katz and Kaplan (1981) reported a total nickel content for the deep-water San Diego Trough surface sediments of 2-8 µg/g, compared with our outside San Diego Bay concentration of 6 (4) µg/g. Moving inshore, Ladd et al. (1984) found 8.5-11.5 µg/g in north San Diego Bay (NBSP mean(sd): 14 (8) µg/g), 16-23 µg/g in south San Diego Bay sediments (NBSP: 20 (8) µg/g), and 19-20 µg/g off National City (our NBSP mean (sd): 23 (9) µg/g).
In Alaska, Sweeney and Naidu (1989) found 8.4-59 µg/g nickel in the Oliktok Point (NBSP: 23 (18) µg/g) area of Alaska's Beaufort Sea and 15-43 µg/g in the Prudhoe Bay area (Endicott Field: 18 µg/g). Feely et al. (1981) reported nickel levels in suspended particulate matter from the Copper River of 61 µg/g and as high as 81 µg/g in surface nearshore northeastern Gulf of Alaska waters. The highest central Gulf of Alaska NBSP value for nickel was found at Lutak Inlet (17 (8) µg/g) and the lowest was found a few miles to the north at Skagway (4.3 (1.6) µg/g), but the difference was not statistically significant. Feder et al. (1990) reported a value for a mudflat near our Port Valdez site which was considerably greater than our mean (sd) value (83 vs. 17 (0.5) µg/g).
The analytical results for nickel from single-sediment samples (not shown on Figure 5) collected from the Estero Bay (49 µg/g), the Farallon Islands (42 µg/g), San Luis Obispo (30 µg/g), Dana Point Inside (29 µg/g), and outside Mission Bay (11 µg/g) were all lower than the reference site.
Selenium--Selenium showed some variation from 0.10 to 1.12 µg/g in mean surface sediment concentrations between Pacific coast NBSP sites which is probably related to grain size or a parameter such as surface area which correlates to grain size. The total mean selenium concentration was significantly higher than the maximum reference site (Bodega Bay UCI: 0.5 µg/g) in only two Pacific coast sediments. These were San Pedro Outer Harbor (1.11 (0.45) µg/g) and Nakhu Bay (0.87 (0.39) µg/g). The single analysis (not shown) for Dana Point Inside was 0.53 µg/g and for the Farallon Islands was 0.29 µg/g.
The biogeochemistry of selenium in the aquatic environment is highly complex in that this element can exist in nature in four oxidation states. Selenium is both an essential and a toxic element; the concentration window between required intake levels and those known to be damaging for this element is unusually small (Phillips 1988). It is required in small amounts by both plants and animals, based upon the specific incorporation into certain enzymes (e.g., glutathione peroxidase), as well as nonspecific incorporation into proteins (selenomethionine).
Selenium is used in pesticides, shampoos, and in the manufacture and production of glass, pigments, rubber, metals and their alloys, medical therapeutic agents, photographic emulsions, and petroleum. Selenium is present in the fossil fuels, petroleum and coal, as well as in natural seleniferous soils, including parts of the Rocky Mountains and High Central Plains and regions of the southeastern United States (Phillips 1987).
Crecelius et al. (1985) reported that selenium in Commencement Bay surface sediments did not exceed 0.8 µg/g which agrees with our mean (sd) value of 0.34 (0.13) µg/g. Risebrough et al. (1978) reported values of 2.6-3.0 µg/g in the Carquinez Straits upstream of the NBSP San Pablo Bay site (0.18 (0.13) µg/g), 1.1-1.5 µg/g off Angel Island which exceeded our nearby Southampton Shoal site (0.19 (0.17) µg/g) and less than 1.0 µg/g at Oakland Outer Harbor (NBSP: 0.37 (0.09) µg/g). According to Phillips (1988), San Francisco Bay exhibits widespread selenium enrichment. Phillips (1987, 1988) suggested that selenium inputs to the San Francisco Bay area were from at least three important sources: leaching of seleniferous soils (as selenate from irrigation by agricultural activities) in the San Joaquin catchment, industrial inputs in the Carquinez Straits area (mainly selenite from six major oil refineries), and an unidentified input from south San Francisco Bay. Phillips (1988) reports that some of the earlier data may list values higher than those determined by more sensitive analytical techniques.
Many Alaskan sites also had high mean selenium concentrations, including Dutch Harbor (0.70 (0.15) µg/g) and the remote Boca de Quadra site (0.74 (0.16) µg/g), which suggests a possible geological component. No selenium (<0.1 µg/g) was found in the sediments from the Chukchi Sea. Nakhu Bay produced the second highest mean value on the Pacific coast but this was not significantly different from the Alaskan reference site at Lutak Inlet (0.43 (0.25) µg/g).