Assessment of Zebra Mussels (Dreissena polymorpha) as a Source of Contaminant Exposure for Diving Ducks
by Peter Lowe
The zebra mussel, Dreissena polymorpha Pallas, a species recently introduced to North America, may have become an important food source for North American diving ducks. Redheads (Aythya americana), lesser scaup (A. affinis), buffleheads (Bucephala albeola), and common goldeneyes (B. clangula) depend heavily on the species in locations with high densities of zebra mussels. This new food source apparently has caused lesser scaup, buffleheads, and common goldeneyes to change their fall migratory patterns and to increase the length of their stay in some Great Lake locations. In Europe 10 species of diving and dabbling ducks, 2 mergansers, 2 gulls, 1 coot, and 1 grebe feed on zebra mussels. Other North American species that eat zebra mussels include freshwater drum (Aplodinotus grunniens), redear sunfish (Lepomis microlophus), blue crab (Callinectes sapidus) and crayfish (Orconectes propinquus)
Zebra mussels are known for their efficacious bioconcentration of various contaminants. Bioconcentration factors for organochlorines and several metals vary from 103 to 104 and concentrations generally vary with those in ambient waters. Consequently, Europeans routinely use zebra mussels for monitoring long term trends and for short term studies of environmental contaminant levels.
A shift from traditional foods to zebra mussels could alter levels of diving duck contaminant exposure if zebra mussels have higher or lower contaminant loads than traditional foods. Moreover, seasonal changes in contaminant accumulation patterns by zebra mussels may influence the level of contaminant exposure experienced by diving ducks. Concentrations of cadmium, copper, lead and zinc, for example, were found to vary seasonally in four European locations.
This study was performed with the following objectives:
1. Determine if element concentrations in the sediments are an accurate indicator of the environmental conditions influencing elemental concentrations in zebra mussels.
2. Determine if element levels in zebra mussels change significantly between late summer and late fall, the time which lesser and greater scaup and common goldeneye are present in the eastern Great Lakes.
3. Determine differences in element levels between zebra mussels and a potential traditional food organism of diving ducks, the amphipod Gammarus fasciatus.
1. Sampling sites: Sixteen locations in Lake Erie, the Niagara River, and Lake Ontario (Fig. 1) that receive moderate to high waterfowl use during the fall and are believed to receive varying levels of contaminant exposure.
2. Sampling times: Late September and mid November, 1993, early October and December, 1994.
3. Quantities collected:
Zebra mussels, about 2 L
Amphipods, 2-50 g
Sediment, 150 - 200 g of top 1-2 cm
4. Field storage: Zebra mussels, double bagged in previously labeled polypropylene plastic bags; amphipods, put in previously labeled, 2-oz chemically clean IChem jars, sediments, previously labeled, 4-oz chemically clean IChem jars.
5. Preservation: Placed in coolers with dry ice 0-1 h after collection and frozen within 2-3 h, transferred to freezers (-20o C) after collection trip.
6. Analytical plan: Zebra mussels - one composite sample collected from each sampling site each collecting time, each sample composed of the soft parts of 24 to 220 mussels; amphipods - same as for zebra mussels, each sample including 2 - 50 g; sediments - same as for zebra mussels, 3 to 5 samples in each composite sample.
7. Analyses performed: Zebra mussels, amphipods and sediments - elemental scan by inductively coupled plasma emission spectrometry (ICP) and Hg analysis by cold vapor atomic absorption spectrometry. Sediments - particle size analysis and percent organic carbon.
8. Preparation for chemical analysis: Whole sample homogenized and freeze dried, aliquots underwent nitric acid - perchloric acid digestion for ICP analysis or nitric reflux digestion for Hg analysis.
9. QAQC: Procedural blanks and standard reference materials - 1 per 26 study samples; spiked and duplicate samples - 1 per 25 study samples; standard reference materials for organic carbon analyses - 1 per 25 study samples.
10. Statistical analyses: Regression analysis to determine the relationship between element concentrations in sediment and zebra mussels and relationship between concentrations in zebra mussels and amphipods. Analysis of variance to evaluate the effect of sampling locations and times and sediment characteristics on sediments concentrations and the effect of sampling locations and times on zebra mussels. Interactions with location and other variables could not be used in these analyses because of insufficient numbers of degrees of freedom. Log transformations of element concentrations and arcsine transformations for percentages of clay and organic carbon were used for statistical analyses. Geometric mean element values are used in figures and tables.
1. The ICP scan yielded results for 19 elements; however, concentrations of As, Mo, and Se in most samples were below the limits of detection. Also, because of small sample size, Hg could not be analyzed in all samples of G. fasciatus. This report, therefore, focuses on the results of the remaining 15 elements and Hg (Fig. 3) except in G. fasciatus.
2. Regression analyses were performed to determine if element concentrations in sediments are an accurate indicator of the environmental conditions that influence elemental concentrations in zebra mussels. The fully parametrized model used for the analysis was
Zc = L + Y + M + L*Y + L*M + Y*M + C + O + Sc
where Zc (zebra mussel concentration) is the response variable and Sc (sediment concentration) is the independent variable of interest (expected to reflect environmental conditions influencing zebra mussels). Additional factors in the model:
L = location
Y = collection year
M = collection month
C = percent clay in sediment
O = percent organic carbon in sediment
and their interactions (combinations denoted by asterisks, e.g. L*M), were included to control for potential sources of confounding in the regression analysis. Because zebra mussel sizes differed significantly (P = 0.0001) among locations (Fig. 2), a potential size effect also had to be considered. Therefore, the data were analyzed using the above model then using the same model substituting mean mussel length for location.
The results showed that sediment concentrations were not an accurate indicator of the environmental conditions that influence concentrations of 10 elements (Al, Ba, Be, Cd, Cr, Cu, Mn, Ni, Pb, and V) in zebra mussels regardless of whether location or mean mussel length were used in the model. There was evidence of a marginally significant regression relationship between sediment and zebra mussel concentrations for B (P = 0.06), Mg (P = 0.08), and Zn (P = 0.05) when location was included in the model. A stronger regression relationship between sediment and zebra mussel concentrations was found Fe (P = 0.01), Hg (P = 0.01), and Sr (P = 0.0001) when mussel length is included.
Some of the failures to obtain significant relations probably result from the limit number of samples available and perhaps ought to be subject to further investigation.
3. Element levels in sediments differed significantly among locations (Fig. 3) and were frequently above the EPA guidelines for unpolluted Great Lakes harbor sediments (Table 1). Levels of Ba exceeded the guidelines at all but one location, Thompson Bay. Other elements that exceeded the guidelines included Cr (4 locations), Cu (10 locations), Fe, Mn and Ni (6 locations), and Zn (8 locations).
4. The results of analysis of variance using with the following model:
Sc = L + Y + M + C + O+ Y*M
showed that location effects influenced concentrations of all elements (P = 0.0003). Percent clay had a significant influence on all elements (P = 0.02 to P = 0.0001) except Cd, Hg, Pb, and Zn. Collection year also had a significant influence on Fe and Ni concentrations (P = 0.024). Location effects probably were influenced by percent clay. Analysis of variance using clay as the dependent variable (C = L + Y + M + Y*M) showed that location was the only factor influencing percentage clay.
5. Zebra mussels appear to respond to different environmental concentrations of elements in different locations and perhaps to differing elemental concentrations from year to year. Elemental concentrations in zebra mussels were analyzed using the model
Sc = L + Y + M(Y) .
Location effects were significant effect (P = 0.0001 to P = 0.003) on concentrations of all elements except Cu (Fig 3). There were significant year to year effects (P = 0.0001 to P = 0.046) on 12 elements (Table 2). Concentrations were significantly higher in 1994 than 1993 for 11 of the 12 elements. Concentrations of 8 elements were influenced significantly (P = 0.015 to P = 0.046) by the month effects (Table 3).
6. The occurrence of generally lower elemental concentrations in zebra mussels in 1993 than in 1994 (Table 2) may result from either (1) differences between years in mean mussel size or (2) a geographically widespread environmental phenomenon that affected all locations about equally in either 1993 or 1994. Mean concentrations of 14 of the 16 elements were higher in 1994 than in 1993; 11 of which were significantly different (P = 0.05). Mean mussel lengths in 1993 and 1994 were 23mm and 27.5mm, respectively (P = 0.0001). The results of the sine test indicates that probability of higher concentrations of 14 out of 16 elements occurring in the same year as a random event is extremely low (P < 0.0001).
7. Concentrations of 13 elements differed significantly (P = 0.0001 to P = 0.01) between G. fasciatus and zebra mussels collected concurrently (Fig 4). Concentrations of Al, B, Ba, Cu, Fe, Mg, Mn, Pb, Sr and V were higher in G. fasciatus than in zebra mussels and concentrations of Cd, Ni, and Zn were higher in zebra mussels than in G. fasciatus.
8. Except for Cd, element concentrations in zebra mussels were generally well below the recommended guidelines for element levels in water fowl diets (Table 4). Concentrations of Cd in zebra mussels collected from Presque Isle Bay, the Erie Waterfront, Thompson Bay, Black Rock Canal, Cayuga Island, Freezer Queen Foods docking facility, the two locations on the Genesee River, Upper Irondequoit Bay and the Oswego River exceeded the guidelines. Mercury concentrations in zebra mussels from Cayuga Island approached the guidelines.
|Table 1 . Mean element concentrations in U.S. soils, and EPA guidelines for uncontaminated Great Lakes harbor sediments (µg.g-1 dry weight).1|
|1Beyer, W.N. 1990. Evaluating Soil Contamination. FWS Biol.
2Concentration for heavily polluted sediments, guidelines for unpolluted sediments are not established.
|Table 2. Yearly mean element concentrations (µg.g-1) in zebra mussels and percentage change from higher to lower concentrations. An asterisk (*) indicates percent changes between years that are significantly different.|
|Table 3. Mean September and November levels in zebra mussels for elements concentrations that changed significantly between months.|
|Table 4. Dry weight (DW) or fresh weight (FW) elemental concentrations in foods causing sublethal adverse effects in waterfowl and recommended concentration limits of the elements in waterfowl foods.|
|Element||Sublethal Concentration||Avian species||Recommended Concentration*||DW conversion for zebra mussels|
|Boron||30 mg.kg-1 (FW)||Mallard||<13 mg.kg-1 (FW)||< 186 µg.g-1|
|Cadmium||4 mg.kg-1||Am. black duck||<100 µg.kg-1 (FW)||<1.43 µg.g-1|
|Chromium||10 mg.kg-1||Am. black duck||<10 mg.kg-1||<143 µg.g-1|
|Copper||<200 mg.kg-2 (DW)||<200 µg.g-1|
|Lead||10 mg.kg-1 (DW)||Mallard||<2 mg.kg-1 (FW)||<29 µg.g-1|
|Mercury||0.5 mg.kg-1 (DW)||Mallard||<100 µg.kg-1 (FW)||<1.43 µg.g-1|
|Zinc||3,000 mg.kg-1||Mallard||<2,000 mg.kg-1 (DW)||<2,000 µg.g-1|
|*Eisler, R. 1990. Boron hazards to fish, wildlife, and invertebrates: A
synoptic review. Contaminant Hazard Reviews Report 20.
*Eisler, R. 1985. Cadmium hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 2.
*Eisler, R. 1986. Chromium hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 6.
*Eisler, R. (In Press) Copper hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 33.
*Eisler, R. 1988. Lead hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 14.
*Eisler, R. 1987. Mercury hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 10.
*Eisler, R. 1993. Zinc hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 26.
|Figure 2. Length of zebra mussels in composite samples.
(CI = 95% confidence interval.)
|Figure 3. Mean element concentrations in sediments from various locations.|
Figure 4. Mean element concentration in zebra mussels from various locations.
|Figure 5. Mean element concentration in zebra mussels and G. fasciatus. Asterisk (*) indicate which element concentrations are significantly different.|
1. Waterfowl would be exposed to different concentrations of elements consuming zebra mussels from different locations.
2. Exposure levels could vary from year to year at a single location or at many locations like depending on environmental conditions, or mean mussel size.
3. Mussels consumed late in the fall migration period may have higher levels of 5 elements but lower concentrations of 3 elements than in mussels consumed early in the fall migration period.
4. Waterfowl switching to mussels from traditional food organisms, exemplified by G. fasciatus, would be exposed to higher concentrations of some of the more toxic elements (i.e. Cd and Ni) but concentrations of most are either about equal if both species or are higher in G. fasciatus.
5. Except for Cd, waterfowl consuming the mussels analyzed for this study would not have been exposed to levels of toxic elements that exceed recommended dietary guidelines.
I wish to thank the U. S. Fish and Wildlife Service, Lower Great Lakes Fisheries Resources Office in Buffalo, NY for logistical support in collecting most of the samples. Specific individuals who helped with the collections included Betsy Trometer, Morgan McCosh, and Christopher Lowie. Daniel Day, Patuxent Wildlife Research Center, provided logistical support for the remaining collections. Anna Morton and Michael Balint, Patuxent Wildlife Research Center, prepared the samples for submission to the analytical laboratory. Statistical support was provided by Jeffrey Hatfield and William Link, Patuxent Wildlife Research Center.
1Eisler, R. 1990. Boron hazards to fish, wildlife, and invertebrates: A
synoptic review. Contaminant Hazard Reviews Report 20.
Eisler, R. 1985. Cadmium hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 2.
Eisler, R. 1986. Chromium hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 6.
Eisler, R. (In Press) Copper hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 33.
Eisler, R. 1988. Lead hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 14.
Eisler, R. 1987. Mercury hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 10.
Eisler, R. 1993. Zinc hazards to fish, wildlife, and invertebrates: A synoptic review. Contaminant Hazard Reviews Report 26.