Anthony J. Paulson1, 2, 4, Vincent S. Zdanowicz3,5, Beth L. Sharack1,6, Elizabeth A. Leimburg1,7, and David B. Packer1,8
Postal Addresses: 1National Marine Fisheries Serv., 74 Magruder Rd., Highlands, NJ 07732; 3U.S. Customs Serv., 7501 Boston Blvd., Ste. 113, Springfield, VA 22153
Current Address: 2U.S. Geological Survey, 1201 Pacific Ave., Ste. 600, Tacoma, WA 98402
E-Mail Addresses: 4apaulson@usgs.gov; 5Vincent.Zdanowicz@customs.treas.gov; 6Beth.Sharack@noaa.gov; 7Elizabeth.Leimburg@noaa.gov; 8Dave.Packer@noaa.gov
Bioaccumulation of metals in mussels depends not only on metal concentrations in the sediments (Hummel et al. 1997), but also the physiological state of the organism (e.g., season and environmental factors) and the biogeochemistry of the sediments (e.g., iron (Fe) content, organic carbon (OC) content, and oxidation-reduction condition). Trace metals were analyzed in mussels and sediments in the Arthur Kill to determine if the biogeochemical processes that control bioaccumulation were affected by replanting of S. alterniflora at the previously oiled sites. Since the replanted sites were not sampled before replanting, pairs of unplanted, replanted, and reference sites in the Arthur Kill were sampled for mussels and sediments in September 1996 and May 1997. Sampling at the unplanted sites (i.e., Con Ed Tower and Saw Mill Creek South sites) occurred 6 yr after the 1990 Exxon Bayway oil spill. At the time of initial sampling, S. alterniflora planted in 1992 had been growing at the Saw Mill Creek North site for 4 yr, while the S. alterniflora planted at the Old Place Creek site in 1993 had been growing for 3 yr. Two Arthur Kill reference sites (i.e., Tufts Point and Mill Creek) and a regional reference site (i.e., Sandy Hook) were also sampled.
This chapter only addresses: 1) the level of contamination of Arthur Kill sediments and mussels, and 2) whether replanting is the dominant factor controlling metal concentrations in sediment and mussels. More specific interactions between bioaccumulation in mussels and sediment geochemistry will not be addressed in this chapter.
METHODS AND MATERIALS
All implements and plastic containers used for collecting, transporting, processing, and storing sediment and mussel samples for metals analyses were decontaminated by rinsing in dilute, ultrapure nitric acid, then doubly in deionized (DDI) water.
Sediments
For collecting sediment samples, four stations were selected along a transect at 0.2 m above the mid-tide level at each of the six sites within the Arthur Kill. Locations with equal tidal height were chosen to minimize station-to-station differences in surface (tidal) hydrology. The positions of the stations along the transect were chosen based on the need to minimize disturbance to the site. Also, access to the specific sites was affected by unique logistical difficulties. For this reason, distances between stations along the transect at a site ranged from 2 to 20 m apart, and the total length of the transects among sites ranged from 12 m at Saw Mill Creek South to 39.5 m at Old Place Creek. At the replanted and reference marsh sites, stations along the transect were located within the vegetated zone. At Con Ed Tower, the transect was in a wide unvegetated area, and stations were located in mud and peat that contained chunks of asphalt. At Saw Mill Creek South, the transect was along the edge of the cordgrass and barren mud and peat banks.
Sediment samples for grain size analysis were collected at each of
the four stations within each of the six sites in September 1996 (one
core per station) by using 28-mm-internal-diameter, plastic-core tubes,
and were frozen for transport back to the laboratory. The particle size
distribution of the sediment mineral fraction was determined by modifying
the standard wet and dry sieving procedures of Ingram (1971), Galehouse
(1971), and Folk (1980). The particle upper size limit chosen was <-2
(i.e.,
pebble/granule boundary), and the particle lower size limit was >4 (i.e., mud, composed of silt and clay). The top 5 cm
of each frozen core were extracted, treated with several milliliters
of 30% H2O2, and heated to digest any organic material.
The samples often contained large sections of S. alterniflora rhizomes
and stems, which were removed. Each sample was then wet sieved with a
63-µm sieve to separate the coarse sediment from the mud. While
the mud remained in distilled water, the coarse fraction was dried and
mechanically sieved through different-sized sieves to separate out the
various coarse size fractions, plus any remaining mud. After weighing
the dried coarse fractions, the mud from both wet and dry sieving procedures
was combined, dried, and weighed. For samples from the Con Ed Tower site,
deposits of tar prevented us from performing any grain size analysis.
For determination of total OC in the sediments, see Chapter IV, "Sediment Biogeochemistry."
For trace metal analyses, sediment cores were collected with 31-mm-diameter acrylic tubes at the four representative locations within each of the six Arthur Kill sites and at two locations within the regional reference site (i.e., Sandy Hook). The top 1-cm section from each core was dried overnight at 60-65°C, the debris was removed, and the remaining sample was pulverized. Ten milliliters of trace-metal-grade concentrated HCl was added to a 100-ml Pyrex beaker containing 1-10 g of dried sediment, and was allowed to react with the sediment for 15 min (Zdanowicz et al. 1995). After the addition 10 ml of concentrated HNO3, the sediment slurry was then allowed to stand for 2 hr. The slurry was then taken to dryness over low heat. After the addition of 25 ml of 0.1-M aqua regia, this slurry stood overnight at room temperature. The volume of liquid was reduced to about 10 ml over low heat, and the slurry was filtered through acid-cleaned, #41 Whatman filter paper using additional DDI to rinse the beaker. DDI was added to bring the filtrate to a final volume of 25 ml.
The resulting solutions were analyzed for iron (Fe), chromium (Cr), copper (Cu), nickel (Ni), zinc (Zn), manganese (Mn), and lead (Pb) by using flame atomic absorption spectrophotometry (AAS). Six procedural blanks and five replicates of standard reference material (SRM) NIST 1645 (river sediment) obtained from the National Institute of Standards and Technology (NIST) were also analyzed by the same procedure. Table 1 shows the quality assurance data for the trace metals. Except for Ni and Pb in the May 1997 Sandy Hook samples, metal values in the sediments were above detection limits (see Table 2). Recoveries for NIST 1645 ranged between 90 and 98%. Within the Arthur Kill samples, outliers were identified using the Grubbs test (Sokal and Rohlf 1981), and were not used to calculate any statistical parameter. Differences in mean metal concentrations among groups of samples from different sites for each year were investigated using analysis of variance (ANOVA; P = 0.05) and Duncan's multiple range test.
Mussels
Ribbed-mussels were collected randomly at each site in September 1996 and May 1997. Owing to the sparse density of the mussels, to the limited time available for sampling (i.e., between high tides), and to the desire not to disturb the sites any more than necessary, the first 60-70 specimens that were found were collected. At the Saw Mill Creek South site, sampling was impeded by tall S. alterniflora, so it was possible to collect only 34 specimens in 1996.
After being transported to the NMFS James J. Howard Marine Sciences Laboratory in Sandy Hook, New Jersey, in plastic bags under ice, mussels from each site were separated by size into two roughly identical groups, one for metals analysis and one for hydrocarbon analysis. In order to obtain specimens of comparable size at each site for metal analysis, a length range between 55 and 67 mm was selected. This was the smallest size range that provided at least five individuals per site. At sites where there were more than five specimens within this range, samples were selected using the following procedure. The size range was divided into five bins. One specimen per bin was selected randomly. If a bin contained no specimens, an alternate bin was selected at random, and a specimen was selected randomly from it. For the given length range, the average wet weight (15.65±2.70 g) of tissue collected in May 1997 from the Mill Creek site was significantly greater than the weights of samples for the 13 other collections (13.32±2.10 g).
Mussel specimens for metal analysis were allowed to depurate overnight in ambient, laboratory supplied seawater at 4°C. After removing extraneous material from the shell (mud, barnacles, etc.), total weight and length were recorded for each specimen. The tissue (i.e., soft parts) was then excised, and stored in a vial at -20°C until analysis. Five or six individual samples per site were analyzed for metals. After thawing, the entire soft tissue was placed in a Teflon vial and weighed. The tissue was dried overnight at 60-65°C and reweighed to obtain a dry weight. Five milliliters of ultrapure concentrated HNO3 was added to the sample which was typically 1 g. The vials were allowed to stand at room temperature for 2-4 hr. Vials were then capped and placed inside Teflon-lined bombs, and the tissue was digested overnight at 120°C. After cooling, bombs were vented, the vials were removed, and the digests were allowed to degas at room temperature overnight. The digests were then quantitatively transferred to 25-ml glass graduated cylinders and brought to volume using DDI water. The resulting solutions were analyzed for Fe, Cu, and Zn by using flame AAS, for Cr, Ni, silver (Ag),and cadmium (Cd) by using graphite furnace AAS, and for mercury (Hg) by using cold-vapor AAS. Nine procedural blanks and nine replicates of NIST 1566a (freeze-dried oyster tissue) were also analyzed using the same procedure. Details of the sample digestion and analysis procedure can be found in Zdanowicz et al. (1993).
Values for all specimens were above detection limits. Average SRM recoveries ranged from 96-102% (Table 1). A majority of variables for two Con Ed Tower samples collected in May 1997 were found to be outliers, and all data from these two samples were disregarded (see Table 6).
RESULTS
The general characteristics of the sediments analyzed in this study differ significantly (Table 2 and Table 3). The sediments at the two Saw Mill Creek sites, the Tufts Point site, and the Mill Creek site in 1996 were predominately fine grained (i.e., on average, between 71.3 and 98.3% of samples, by weight, were <63 µm by weight), and had OC content that ranged, on average, between 5.8 and 11.1% by weight (Figure 3; results of the OC analyses are from Chapter IV, "Biogeochemistry"). In contrast, the Old Place Creek and Sandy Hook sites contained, on average, 11.9 and 2.8% fine material, respectively; the Old Place Creek site contained, on average, 1.3% or less OC. The size distribution of the Con Ed Tower site could not be determined because the OC content of the sediments in 1996 averaged 35.0%, a significant portion of this OC being oil. The average Fe content of the four fine-grained sediment sites (i.e., Saw Mill Creek North and South, Tufts Point, and Mill Creek) ranged between 3.29 and 3.91% by weight. The coarser nature of the Old Place Creek and Sandy Hook sediments was reflected in their lower average sediment Fe concentrations of 0.9 and 0.3% by weight, respectively. The mere dilution of fine-grained material by the presence of abundant organic matter (i.e., over 40% by weight OC at some stations) at the Con Ed Tower site also reduced Fe concentrations. Similar to Fe, Mn concentrations at the Old Place Creek and Sandy Hook sites were also low. However, there was considerable variability in the Mn concentrations at the fine-grained sites, with the southern sites (i.e., Tufts Point and Mill Creek) having significantly higher Mn concentrations.
The sediment texture was also reflected in the concentrations of Cu, Zn, Cr, and Pb in sediment, with the sites with coarser sediments (i.e., Old Place Creek and Sandy Hook) having significantly lower concentrations (Figure 4). However, there was considerable variation in metal concentrations among the sites having fine-grained sediments. These metal concentrations are within the range reported for Arthur Kill sediments by other investigators (Meyerson et al. 1981; Adams et al. 1998). Cr, Cu, Zn, and Pb were significantly higher in all Arthur Kill samples (Table 4) relative to marine sediments with similar Fe concentrations collected from sparsely populated coasts (Daskalakis and O'Connor 1995). Ni concentrations of Arthur Kill sediments (Table 2 and Table 3) were lower or similar to marine sediments collected from less-impacted coasts. The Cu, Zn, and Pb concentrations at the Mill Creek reference site were significantly higher than the New York Harbor, Western Long Island Sound, and Newark Bay averages (Table 4).
Trace metals generally concentrate on the Fe, Al, and Mn oxide and OC coatings of sediments (Olsen et al. 1982). Fe concentrations in these Arthur Kill sediments were significantly larger than Mn concentrations, suggesting the Fe oxides were the dominant oxide coating of the surfaces of sediment particles. Normalization of Cu, Zn, Cr, and Pb concentrations to Fe accounts for the differences in sediment texture of each individual sediment sample. These normalized concentrations were used to determine significant differences in metal concentrations among sites. The normalized concentrations of Cu, Zn, and Pb were lower at the Sandy Hook station compared with the Arthur Kill sites (Figure 5), while normalized Cr concentrations at Sandy Hook were not significantly difference than those at Old Place Creek, Tufts Point, and Mill Creek sites. Normalized Cu, Zn, and Pb concentrations at Mill Creek in 1996 were significantly higher than those of most of the other Arthur Kill sites. In general, normalized concentrations of metals at the unplanted, planted, and Tufts Point (reference) sites were similar, with the exception of higher normalized concentrations of Zn at Tufts Point in 1997, higher Cr at the Saw Mill Creek North and Con Ed Tower sites in 1996, and higher Ni at Con Ed Tower sites in 1996 (not shown). There was little difference in the normalized metal concentrations between the two adjacent Saw Mill Creek sites, except for the higher Pb value at the restored Saw Mill Creek North site.
Mussels
The metal data for mussels also indicate that Tufts Point was not a suitable reference site, but rather reflected the higher metal concentrations as a result of the overall pollution of the Arthur Kill, including oil spills (Table 5 and Table 6). For instance, the highest Cd concentrations found in any specimen in each season were from specimens collected from Tufts Point. Fe, Cr, Ni, Zn and Hg concentrations from Tufts Point were not significantly different than the concentrations from the oiled sites. Therefore, the Tufts Point sample is grouped with the other Arthur Kill sites in the following discussion.
The range of metal concentration data from September 1996 was wide, resulting in much overlap in ranges among sites. However, some significant seasonal differences were found (Figure 6 and Figure 7). The unplanted Saw Mill Creek South site was anomalous in that the Ni concentration in mussels was greater in May 1997 than in September 1996, but there were no significant seasonal differences found for the other seven elements. In contrast, concentrations of Cr, Ag, and Hg at four other Arthur Kill sites were generally higher in September, and highly variable.
The decrease in Ag, Cr, Cu, and Hg concentrations in mussels from the oiled sites between September and May might be a result of natural processes. A significant seasonal difference in metal concentrations between the replanted and unplanted sites was observed only for Cu. Therefore, it is unlikely that the seasonal difference in metal concentrations in mussels was influenced by the replanting effort.
Only the May data were used to determine geographical differences in metal concentrations in mussels, since the September data were so highly variable. Relative to the oiled sites, Cr, Cu, and Hg concentrations in mussels were significantly lower at the Mill Creek and Sandy Hook sites, while Fe and Cu were significantly lower only at the Sandy Hook site. The range of concentrations of Ag, Ni, and Zn in mussels at both these reference sites significantly overlapped the concentrations of some of the oiled sites.
For all elements in mussels, there were no clear differences between the replanted and unplanted sites. The mussels from the unplanted Saw Mill Creek South site contained the highest concentrations of Fe, Cr, Cu, Ni, Ag, and Hg. The close proximity and similar sediment characteristics (i.e., % fines and OC content) of the unplanted Saw Mill Creek South and the replanted Saw Mill Creek North sites provide a valid comparison to test the effects of replanting. The concentrations of Ag and Cd were significantly higher (P <0.05) at the unplanted Saw Mill Creek South site, while the concentrations of Zn were significantly higher at the Saw Mill Creek North site. No significant differences were found for Cr, Cu, Ni, Hg, and Fe.
Although mussels have been used extensively in marine monitoring programs, these programs primarily use the blue mussel, Mytilus edulis. In one of the few studies in which the accumulation of metals was compared in different species of mussels, Nelson et al. (1995) state, "these findings highlight the fact that metal uptake in bivalves is a complicated process that can be affected by many exogenous and endogenous factors." In keeping with their caution, our ribbed-mussel data were compared only with other ribbed-mussel data (Table 7). Cr, Cu, and Zn concentrations in ribbed-mussels from Sandy Hook are comparable with those from a clean site in East Sandwich, MA (Nelson et al. 1995). In contrast, the metal concentrations in ribbed-mussels from the Arthur Kill are similar to those from the polluted New Bedford Harbor, Massachusetts, and Inner Mystic River Estuary, Connecticut (Nelson et al. 1995; Miller 1988).
DISCUSSION
Sediments
Correlations among metal concentrations, grain size, and OC content were determined by two separate analyses because of incomplete data. No correlations, though, could be calculated for the Sandy Hook site because of lack of sufficient grain size data and lack of any OC data. In the first analysis, correlations were determined for the metal and OC data from each site for both 1996 and 1997 sampling periods. In the second analysis, correlations were determined for the metal and grain size data from each site except the Con Ed Tower site for just the 1996 sampling period.
For the Old Place Creek site, the significant variability in both metal and OC concentrations among individual sediment samples appears to be related to the portion of fine-grained sediments found in each sample. The 1996 metal data from Old Place Creek, excluding Mn and Cu, were correlated with the percentage of fine-grained sediment found in each sample. When the entire Old Place Creek data set is subjected to correlation analysis using OC data (Table 8), the entire correlation matrix table is significant (r>0.80). For the fine-grained-sediment sites (i.e., Saw Mill Creek North and South, Mill Creek, and Tufts Point), no significant correlations were found between Fe vs. Cr, Ni, Cu, Zn, or Pb. This lack of correlation is not surprising since Fe concentrations within a site varied only over a very small range (Figure 3). For these fine-grained sediments, correlations among trace metals are not controlled by the concentrations of Fe oxides, but are controlled by how the trace metals interact with the Fe oxide coating or by phases other than Fe oxides.
Different subsets of trace metals were highly correlated for the data sets from different sites. For instance, significant correlations were found among Pb, Cr, and Cu at the Saw Mill Creek North site (Table 8). It is interesting to note that these three metals were negatively correlated with the percentage of fine-grained sediments. This negative correlation suggests that Pb, Cr, and Cu are associated with a coarser type of particle. Significant correlations were also found among Cr, Ni, and Zn at the Con Ed Tower site, and among Cr, Cu, Ni, and Pb at the Saw Mill South site.
The entire metals data set was subjected to principal components analysis (PCA) using both the OC and grain size data (see eigenvectors in Appendix Tables A1 and A2). When OC data are used, both sampling periods could be analyzed, but without the Sandy Hook site. The Old Place Creek site is distinguished because of its lower metal concentration (Figure 8). Except for one sample, the Mill Creek site is separated from the other fine-grained stations in the Arthur Kill. Although the 1996 Con Ed Tower site is separated from the rest of the sites with finer sediment, there is no distinct difference in replanted and unplanted sites. When only the 1996 metal data set was used with percentage of fine-grained sediment data (Figure 9), the Sandy Hook and Old Place Creek sites again are differentiated from the fine-grained sediment sites. Among the fine-grained sediment sites, only Mill Creek is distinguished.
Mussels
Trace metal concentrations in mussels were higher and more highly variable in September 1996 than in May 1997. Of the eight metals analyzed, four (i.e., Cr, Ni, Cd, and Hg) were significantly lower in mussels at both reference sites relative to the other Arthur Kill sites, and two others (i.e., Fe and Cu) were significantly lower only at the Sandy Hook reference site. Five metals showed higher concentrations in mussels at the unplanted Saw Mill Creek South site compared to the nearby replanted Saw Mill Creek North site. The lack of strong and consistent trends required higher-level statistical analysis of the data in order to draw any conclusions concerning the effects of replanting. Correlations among metal concentrations in mussels were examined for each station to determine if biogeochemical processes were causing similar trends for a subset of the metals studied. In addition, the entire mussel data set was subjected to PCA to determine if the data were separable by type (i.e., unplanted, planted, or reference).
A few significant correlations between the concentration of pairs of metals in mussels within a site were found at three of the five oiled sites, and at both reference sites (i.e., Mill Creek and Sandy Hook; Table 8). Although length and weight of mussels were highly correlated, only one out of the possible 108 correlations between metals and these physical characteristics of mussels is >0.80 (i.e., Zn was negatively correlated with length at Saw Mill Creek South). Zn is correlated with Cd in mussels at Sandy Hook and the replanted Old Place Creek sites. Fe is correlated with Cr at the Saw Mill Creek North replanted site and the Mill Creek reference site, and with Ni at the Con Ed Tower unplanted site. Metal concentrations in mussels at the Saw Mill Creek North site were the most coherent, with four metal pairs having correlations >0.80; however, the Ag is negatively correlated with Cr. Hg is correlated with Zn at Old Place Creek and Mill Creek, and also with Cr at Saw Mill Creek North. No significant correlations were found in mussels at the unplanted Saw Mill Creek South site, which tended to have the highest concentrations in May.
Only the metal data from the entire data set were subjected to PCA because the correlations of metals with length or weight for the individual sites were weak. Among the metals data, the highest correlation (r = 0.65) was found between Zn and Cd. The eigenvectors of the first principal component (Appendix Table A3) ranged between 0.29 for Fe and 0.43 for Hg. A plot of the first principal component versus the second principal component clearly distinguished the two reference sites, Mill Creek and Sandy Hook, to the left (Figure 10). The replanted and unplanted sites could not be distinguished from points plotted in the middle of the plot. The principal component analysis and Duncan multiple range tests suggest that the five oiled sites were not significantly different from each other; the means for mussels from these five sites are given in Table 7. The Fe and Cu concentrations in Sandy Hook mussels were lower than those at the Mill Creek reference site.
CONCLUSIONS
Metal concentrations in the sediments at each site depended more on the general characteristics of the sediment, such as the percentage of fine-grained sediments and Fe content, than on whether or not the site was replanted. Compared to concentrations from the regional reference sites and from other regional studies, metal concentrations in sediments from the entire Arthur Kill were elevated. In fact, the Mill Creek reference site farthest from the location of the spill had the highest concentrations of Cu and Pb when normalized to the sediment Fe content. Higher levels at Mill Creek may have been due to past industrial discharges in this area of the Kill (C. Alderson et al., Salt Marsh Restoration Team, Natural Resources Group, New York City Parks, 200 Nevada Ave., Staten Island, NY, pers. comm.).
For each site, concentrations of groups of metals were highly correlated, but the correlations were not consistent among sites. For instance, concentrations of Pb, Ni, Cu, and Zn were highly correlated at the Mill Creek reference site, while Pb, Cr, Ni, and Cu were highly correlated at the Saw Mill Creek South site. The negative correlation of Cr, Cu, and Pb with the percentage of fine-grained sediments present at the Saw Mill Creek South site suggests that these metals were associated with coarse sediment. PCA distinguished the two coarse-grained sediment sites, but there was no distinction between replanted and unplanted sites.
There were no consistent differences in metal concentrations in mussels collected from replanted and unplanted sites. Concentrations of many metals in mussels from the southernmost Arthur Kill reference site (Mill Creek) were significantly lower than those in mussels from the other five Arthur Kill sites. PCA distinguished the Mill Creek reference site as well as the Sandy Hook regional reference site, but replanted and unplanted sites affected by the spill were not distinguished. Cr, Hg, and Ag concentrations in mussels from many of the Arthur Kill sites were lower in spring than in fall, while Ni concentrations were lower in fall. Since this Arthur Kill reference site and the regional reference site did not show the same seasonal differences in mussel metal concentrations, the differences found for the affected Arthur Kill sites were probably a result of the availability of metal contaminants to the mussel rather than due to any endogenous factors.
Replanting of S. alterniflora has little effect on the trace metal concentrations in sediments affected by oil spills. Oil contamination is generally not a major source of metals. In contrast, bioaccumulation of metals by mussels from the sediments is affected by biogeochemical properties of the sediments. Planting of S. alterniflora can produce subtle changes in the sediments that affect bioaccumulation. In this study, increases in Cu concentrations in mussels collected from the replanted sites were the only significant and consistent change that appeared to be related to replanting.
REFERENCES CITED
Adams, D.A.; O'Connor, J.S.; Weisburg, S.B. 1998. Sediment quality of the NY/NJ harbor system: an investigation under the Regional Environmental Monitoring and Assessment Program (R-EMAP). Final report. EPA Doc. 902-R-98-001; 110 p. Available from: EPA Region II, Edison, NJ.
Daskalakis, K.D.; O'Connor, T.P. 1995. Normalization and elemental sediment contamination in the coastal United States. Environ. Sci. Technol. 29:470-477.
Folk, R.L. 1980. Petrology of sedimentary rocks. Austin, TX: Hemphill Pub. Co.; 182 p.
Galehouse, J.S. 1971. Sedimentation analysis. In: Carver, R.E., ed. Procedures in sedimentary petrology. New York, NY: John Wiley & Sons; p. 69-108.
Giblin, A.E.; Valiela, I.; Teal, J.M. 1982. The fate of metals introduced into a New England salt marsh. Water Air Soil Poll. 20:81-98.
Hummel, H.; Modderman, R.; Amiard-Triquet, C.; Rainglet, F.; van Duijn, Y.; Herssevoort, M.; de Jong, J.; Bogaards, R.; Bachelet, G.; Deprez, M.; Marchand, J.; Sylvand, B.; Amiard, J.C.; Rybarczyk, H.; de Wolf, L. 1997. A comparative study on the relation between copper and condition in marine bivalves and the relation with copper in the sediment. Aquat. Toxicol. (Amst.) 38:165-181.
Ingram, R.L. 1971. Sieve analysis. In: Carver, R.E., ed. Procedures in sedimentary petrology. New York, NY: John Wiley & Sons; p. 49-68.
Miller, E.R. 1988. Physiological, exfoliative, cytological, histological, and biochemical response of Geukensia demissa (Mollusca) exposed to dissolved Cu at various salinities. Ph.D. dissertation. Univ. of Connecticut, Storrs, CT. 230 p.
Meyerson, A.L.; Luther, G.W., III; Krajewski, J.; Hires, R. 1981. Heavy metal distribution in Newark Bay sediments. Mar. Pollut. Bull. 12:244-250.
Nelson, W.G.; Bergen, B.J.; Cobb, D.J. 1995. Comparison of PCB and trace metal bioaccumulation in the blue mussel, Mytilus edulis, and the ribbed-mussel, Modiolus demissus, in New Bedford Harbor, Massachusetts. Environ. Toxicol. Chem. 14:513-521.
Olsen, C.R.; Cutshall, N.H.; Larsen, I.L. 1982. Pollutant-particle associations and dynamics in coastal marine environments: a review. Mar. Chem. 11:501-533.
Sokal, R.R.; Rohlf, J.F. 1981. Biometry: the principles and practice of statistics in biological research. 2nd ed. San Francisco, CA: W.H. Freeman and Company; 859 p.
Zdanowicz, V.S.; Finneran, T.W.; Kothe, R. 1993. Digestion of fish tissue and atomic absorption analysis of trace elements. In: Lauenstein, G.G.; Cantillo, A.Y., eds. Sampling and analytical methods of the National Status and Trends Program, National Benthic Surveillance and Mussel Watch Projects, 1984-1992. Vol. 3. Comprehensive descriptions of elemental analytical methods. NOAA Tech. Memo. NOS ORCA 71; 219 p.
Zdanowicz, V.S.; Cunneff, S.L.; Finneran, T.W. 1995. Reduction in sediment metal contamination in the New York Bight Apex with the cessation of sewage sludge dumping. In: Studholme, A.L.; O'Reilly, J.E.; Ingham, M.C., eds. Effects of the cessation of sewage sludge dumping at the 12-mile site. 12-Mile Dumpsite Symposium, Long Branch, New Jersey, June 1991. NOAA Tech. Rep. NMFS 124:89-99.