Ecotoxicology, 12, 69±81, 2003 # 2003 Kluwer Academic Publishers.Manufactured in The Netherlands.

Common Loon Eggs as Indicators of Methylmercury Availability in North America

D. C. EVERS,1,Ã K. M. TAYLOR,2 A. MAJOR,3 R. J. TAYLOR,4 R. H. POPPENGA5 AND A. M. SCHEUHAMMER6 1BioDiversity Research Institute, 411 US Rt. 1, North, Suite 1, Falmouth, Maine 04105, USA 2Loon Preservation Committee, Moultonborough, NH 03254, USA 3US Fish and Wildlife Service, Concord, NH 03301, USA 4Texas A&M University, Trace Element Research Lab, College Station, TX 77843, USA 5University of Pennsylvania, School of Veterinary Medicine, Kennet Square, PA 19348, USA 6Canadian Wildlife Service, National Wildlife Research Centre, Hull, Quebec, Canada KIAOH3

Accepted1 October 2002

Abstract. Increasedanthropogenic mercury (Hg) deposition since pre-industrialtimes, andsubsequent transformation of inorganic Hg to methylmercury (MeHg) in aquatic environments, has createdareas in North America where Hg poses a relatively high risk to wildlife, especially long-lived, piscivorous species. From 1995 to 2001, we opportunistically collected577 eggs abandonedby Common Loons from eight states. Egg-Hg concentrations rangedfrom 0.07 to 4.42 mg/g (ww) or 0.10 to 19.40 mg/g (dw). Mercury was higher in eastern than in western North America. Female blood-Hg concentrations strongly correlated with those of eggs from the same territory even though the mean intraclutch Hg difference was 25%. In New England, egg volume declined significantly as egg-Hg concentrations increased. Fertility was not related to egg-Hg concentrations. Based on existing literature andthis study'sfindings, egg-Hg risk levels were establishedandappliedto our US dataset and an existing Canadian data set. Regionally, we found the greatest risk levels in northeastern North America. With few exceptions, loon eggs are suitable indicators of methylmercury availability on lakes with territorial pairs.

Keywords: common loon; mercury; indicator; exposure; effects

Introduction Tremblay, 1999) andthen fish (Wiener andSpry, 1996) provides the basis for biomagnification at a Comparedto pre-industrialconditions, current levels level that may be excessive, especially for obligate of environmental Hg are high enough to be of potential piscivores. Comprehensive summaries of Hg expo- harm to North American wildlife. The foodchain trans- sure in obligate piscivores have been compiled fer of methylmercury (MeHg) through plankton (Chen (Burger, 1993; Thompson, 1996; USEPA, 1997). et al., 2000), invertebrates (Parks andHamilton, 1987; The toxicity of MeHg in birds has been extensively studied in controlled situations using captive species. However, most lab studies have used species that are ÃTo whom correspondence should be addressed: Tel.: 207-781-3324; Fax: 207-781-2804; not piscivorous, that likely have different sensitivities E-mail: [email protected] to MeHg than piscivorous species, andthat therefore 70 Evers et al. may not accurately predict the sensitivities of other Materials and methods species to Hg (Posthuma et al., 2002). Captive species studied include the mallard (Anas platyrhynchos) Egg collections (Heinz, 1974, 1976, 1979), American black duck (Anas rubripes) (Finley andStendell,1978), andring-necked As part of various monitoring programs andstudies pheasant (Phasianus colchicus) (Fimreite, 1971). by BioDiversity Research Institute (BRI), loon eggs Information on the impacts of Hg on free-living were collectedfrom eight US states (Fig. 1). Eggs birdpopulations has also been collected,but inter- were collected from nests that had been abandoned, pretation of the importance of Hg from these studies is predated or flooded. Eggs were only removed from often difficult because of uncertainties regarding a nest when either the adults were no longer Hg relationships with (1) tissue type (e.g., liver; incubating them or the eggs were determined to be Scheuhammer et al., 1998a), (2) competing inter- nonviable (i.e., strong odor, or indications that eggs correlation with other contaminants (Wiemeyer et al., were not turned). Laying order of collected eggs was 1984; Henny andHerron, 1989), (3) use of unmarked unknown. individuals (Barr, 1986), (4) interspecific differences in tolerance levels, (5) possible co-accumulation of Adult female blood collections various selenium compounds (Cuvin-Aralar and Furness, 1991 andBischoff et al., 2002), and(6) co- Female loons were live capturedby night lighting and occurrence of additional ecological and/or anthropo- weighed (Evers, 2001). All individuals were uniquely genic stressors. Many of these difficulties can be color markedto permit future identification at a amelioratedby examining the exposure andeffects of distance. Blood and feathers were collected and Hg on a standard tissue for an appropriate species that preservedas describedin Evers et al. (1998). occurs in habitats at high ecological risk from MeHg availability. We have chosen eggs of the common loon (Gavia immer) to describe and assess the risk Tissue processing and measurements relatedto MeHg exposure in freshwater piscivorous birds in North America. Egg length, width, weight, and volume were mea- The common loon inhabits lakes >6 ha across the sured. Egg volume was determined by weighing water northern tier of the UnitedStates north into Canadato displaced by a whole egg. Egg weight was measured the edge of the North American taiga-tundra zone. to the nearest 0.01 g. Fresh egg weight was deter- Because the loon is an upper trophic level predator, is minedby the weight of the water displacedby an egg long lived, and is an obligate piscivore that spends the (volume) andmultiplying it with loon egg density breeding season on freshwater lakes, it is susceptible (Hoyt, 1979). Loon egg density was derived from the to the biomagnification andbioaccumulation of weights of eggs characterizedas infertile andsampled MeHg. Several recent fieldstudieswith the common within the first week of incubation (n ˆ 183). We loon provide a toxicological foundation for both Hg considered these eggs to have minimal moisture loss. exposure (Burgess et al., 1998a; Evers et al., 1998; Weights of these eggs were comparedto the weights Scheuhammer et al., 1998b, 2001) andeffects (Burgess of the corresponding volumes of water they displaced, et al., 1998b; Meyer et al., 1998; Nocera andTaylor, which indicated that egg weights averaged 1.042 1998; Evers et al., 2002). When the Hg exposure and times that of their corresponding water volume effects literature is considered in combination with weights. Eggs were scoredwith a scalpel, opened, recently publishedinformation about loon demo- andcontents were placedinto sterile I-Chem 1 jars. graphics (Piper et al., 1997a; McIntyre andEvers, Egg contents were characterizedas one of five 2000; Evers, 2001), behavioral ecology (e.g., Evers, developmental stages (Table 1). Some eggs were 1994; Gostomski andEvers, 1998), genetics (Dhar et degraded and we were unable to confidently deter- al., 1997; Piper et al., 1997b), andpopulation-level mine fertility (i.e., 0 rating vs. 1 rating) andthese modeling (Evers et al., 2002), an authoritative founda- eggs were ratedas ``not assessable.'' Determining tion emerges for using the common loon as a primary embryonic development was possible for these bioindicator for the effects of contaminants such degraded eggs if hardened tissue or feather remnants as MeHg. were present. Mercury in Loon Eggs 71

Figure 1. Locations of common loon eggs collected for mercury analysis. Inset of New England and New York provide greater detail of increasedsampling efforts in that region.

Table 1. Developmental scale usedfor common loon embryological stage.

NA (not assessable): Developmental stage could not be determined. Contents were gray or yellowish tan in color and typically had a foul smell. A darker color suggested some degree of development had occurred, but unless hardened tissue was detected, eggs were ratedas NA. 0: No development evident. Egg had a yellow/orange or yellow/tan yolk (intact or broken down into a liquid). A translucent jelly-like mass surrounded the yolk sac and showed no sign of embryonic development (e.g., mass not dark or hardened). 1: Embryo was viable (length was up to 1.5 cm). The jelly-like mass (embryo) was dense and hardened. Small dark (red) eye spots were generally visible. 2: Developing embryo hadan apparent central nervous system (length was 1.5±2.0). Cranial developmentandrecognizable eyes were apparent. Feathers were absent. 3: The embryo showed advanced development (length was 2.0±3.0 cm). Bill was developed (e.g., egg tooth present but soft). Legs and wings were visible but not fully developed. Some feathers were present (first seen in tail). 4: The fully developed embryo was completely covered by feathers (length >3.0 cm). Appendages were completely developed. Vent and preen glandwas visible. A small portion of yolk sac remainedattachedto belly.

Chemical analysis Pennsylvania School of Veterinary Medicine (UPenn), Kennet Square, Pennsylvania. Bloodand Analyses of Hg in eggs were performedat Texas feather samples were processedandanalyzedaccord- A&M University Trace Element Research Laboratory ing to Evers et al. (1998). Loon eggs were received (TAMU), College Station, Texas andthe University of frozen in glass I-Chem jars. Samples at TAMU were 72 Evers et al. homogenizedwet with an OMNI Mixer equippedwith 20 ml of deionized water to the bottom of the reaction titanium blades, lyophilized in a Labconco Lyph Lock cell. This injection approach avoided dilution of the 12 freeze dryer, and powdered in a Spex Mixer Mill. Hg0 in the headspace that would have resulted from Moisture was determined by weight loss upon freeze- the use of a sweep gas. Calibration was accomplished drying. Samples at UPenn were processed wet, as using commercial standards (CPI) diluted as neces- received. sary with 5% HNO3/1% HCl, andincludeda blank Samples were analyzedfor Hg by cold-vapor and five standards that ranged from 1 to 10 ppb. The atomic absorption (CVAA) using the methodof Hatch calibration line was calculatedusing unweightedlinear andOtt (1968). Digestion procedures at UPenn and regression and verified using an independent standard TAMU utilizedwet anddrysample, respectively, anda blank that were analyzedfollowing calibration, andincorporatedconcentratedsulfuric andnitric after every 10 samples, andat the endof the analysis. acids and heating at 95 C. Digestion at UPenn also Hg absorbance was measuredusing a Shimadzu included potassium permanganate, whereas TAMU's CR601 integrator in peak height mode. Each batch digestion incorporated both potassium permanganate of Hg samples included a method blank, spiked blank, andpotassium persulfate. Following digestion, reference material (NRCC DORM-2), duplicate excess permanganate was reduced with hydroxyla- sample, andspikedsample (Table 2). The detection mine hydrochloride and samples were diluted to final limit for total Hg at TAMU was 0.002 mg/g on a wet volume (100 ml at UPenn and50 ml at TAMU). weight basis. Following sample digestion, Hg was analyzed at A subset of 20 loon eggs was also analyzedfor UPenn using a 906GBC atomic absorption spectro- MeHg at TAMU using the methodof Wagemann et al. meter equipped with an autosampler and hydride gen- (1997). MeHg was extractedfrom freeze driedsamples erator. For every 25 egg samples analyzed, a 5-point with acidic KBr and CuSO4 solutions into a 2 : 3 calibration curve was constructed, ranging from 10 to mixture of hexane : methylene chloride. Extracted 100 ppb. A reagent blank consisting of 30% HCl was MeHg was digested and measured by element- used to set a zero reading prior to Hg detection. A specific detection of Hg via CVAAS, using methods previously analyzedbloodsample andcertifiedrefer- similar to those described above but modified as ence material (NRCC DORM-2, 4.64 mg/gHg) were necessary for reduced sample sizes. MeHg QC samples analyzedbefore, during, andat the endof every run analyzedwith loon eggs were similar to those included to verify instrument performance. The detection limit for total Hg analysis, with the difference that standards for total Hg at UPenn was 0.25 mg/g on a wet weight andspikes were preparedfrom CH 3HgCl (Johnson basis. Matthey). MeHg is certifiedin DORM-2 at 4.47 mg/g. Mercury content of TAMU digests was measured The detection limit for MeHg at TAMU was using an LDC Hg Monitor equippedwith a 30 cm cell. 0.0008 mg/g on a wet weight basis. One ml aliquots of sample digests were added to a closedreaction cell where the Hg 2‡ was reduced to Dry weight vs. wet weight 0 volatile Hg with 1 ml of 10% w : v SnCl2 in 10% HCl. Mixing andrelease of Hg 0 to the headspace were Egg moisture is lost from the time the egg is laiduntil facilitatedby vortexing the reaction cell on a Vortex it hatches (Hoyt, 1979), so we analyzed207 eggs on a Genie 2. The headspace was introduced to the CVAA dry weight (dw) basis and determined average absorption cell using a syringe pump set to deliver moisture loss. Wet weights (ww) were calculated

Table 2. Quality control samples processedandanalyzedwith common loon eggs.

QC sample type Total Hg average Total Hg range MeHg

Methodblank ( mg/g, ww) 0.00021 À0.00404±0.00574 0.00064 Blank spike recovery (%) 103 90±111 88 Certifiedreference material recovery (%) 98 89±105 87 Duplicate precision (relative % difference) 4.4 0±18 2 Spikedsample recovery (%) 101 90±110 90 Mercury in Loon Eggs 73 using the following formula: ww ˆ dwHg (1 À comparing paireddatasets. JMP software corrected [% moisture/100]) (Stickel et al., 1973). All eggs for inequity of unbalanceddatasets. In all cases, analyzedon a ww basis were adjustedfor moisture means were given with one standard deviation loss by dividing the total egg weight by the fresh egg unless otherwise noted. All means are arithmetic weight (determined by egg volume  1.042) and unless otherwise stated. The level of statistical multiplying by the Hg concentration (Stickel et al., significance was defined as p < 0.05. 1973). Dry weight Hg concentrations were directly determined in the lab. An approximate conversion rate to dw from ww can be based on a loon egg's average Results moisture content of 78.7%. Mercury levels were generally presentedon a ww basis because much of Geographic variation the egg-Hg literature uses a ww basis. Unless otherwise noted, all egg-Hg concentrations given in the text are Mercury levels were determined for 577 eggs, on a ww basis. collectedfrom eight states, between 1988 and2001 (Table 3). In the 20 eggs analyzedfor MeHg, a mean Statistical procedures of 98.6 Æ 8.2% of the Hg was determined as MeHg. Mean percentage of moisture in a loon egg was not Mercury concentrations are expressedas arithmetic significantly different geographically among the six means because data were normally distributed based states with testedeggs ( F ˆ 0.34, p > 0.05). Mean egg- on normal probability plot residuals. Although geo- Hg concentrations were lowest in Alaska, increased metric means are preferredfor smaller datasets eastward, and were highest in Maine. Alaska mean (Parkhurst, 1998), our large sample sizes afforded the egg-Hg concentrations were significantly lower than use of arithmetic means, which are more comparable those of the seven other states (p > 0.05 for each com- with publishedliterature. Homoscedasticity was parison between Alaska vs. other states using Tukey's checkedwith Bartlett's test, which is sensitive to the test). Mean egg-Hg concentrations in New England normality assumption, andvariances were generally (0.77 Æ 0.54 mg/g) were significantly greater than foundto be similar. Therefore log transformations those in the Great Lakes region (0.41 Æ 0.27 mg/g) were not requiredto stabilize variances for linear (t ˆ 4.35, p < 0.001). Egg-Hg concentrations exceed- relationships. JMP software (SAS Institute, 1999) ing 2.0 mg/g were identified only in Maine and was usedto test various hypotheses using one-way New Hampshire. The highest egg-Hg concentrations analysis of variance (ANOVA) followedby Tukey's were from a banded female in southeastern New multiple comparison test if our ANOVA showedsigni- Hampshire that laidan egg with 3.92 mg/g in 1999 and ficant differences. Student's t-test was usedwhen 4.42 mg/g in 2000.

Table 3. Mercury concentrations (mg/g) wet weight (ww) and dry weight (dw) in common loon eggs collected in the United States, 1988±2001.

Wet weight ww Dry weight dw % moisture State Time period n (mean Æ SD)1 (range) n (mean Æ SD) (range) (mean Æ SD)

Alaska 1992±1998 10 0.25 Æ 0.15 (A) 0.06±0.50 0 n/a n/a n/a Maine2 1994±2001 186 0.91 Æ 0.50 (B) 0.12±2.65 85 4.69 Æ 2.05 0.78±4.14 78.6 Æ 2.4 Michigan3 1997±2001 24 0.54 Æ 0.30 (C) 0.18±1.45 7 n/a n/a n/a Minnesota3 1997±1998 20 0.41 Æ 0.26 (C) 0.10±0.89 15 1.55 Æ 0.89 0.37±3.73 79.4 Æ 1.4 Montana 1998±2000 5 0.43 Æ 0.10 (C) 0.33±0.58 3 2.12 Æ 0.63 1.55±2.80 78.3 Æ 0.8 New Hampshire2 1988±2001 263 0.72 Æ 0.57 (D) 0.07±4.42 72 4.11 Æ 3.14 0.82±19.40 78.7 Æ 3.9 New York 1998±2001 28 0.58 Æ 0.28 (C,D) 0.23±1.23 4 3.13 Æ 1.51 1.58±5.08 78.6 Æ 2.2 Vermont2 1997±2001 41 0.52 Æ 0.31 (C) 0.13±1.65 17 1.75 Æ 0.96 0.19±4.14 76.1 Æ 3.6 All states 1988±2001 577 0.55 Æ 0.31 0.07±4.42 203 2.89 Æ 1.30 0.19±19.4 78.7 Æ 4.7

1Mercury concentrations (ww) sharing a capital letter are not significantly different by Tukey's multiple comparison test (p > 0.05). 2States in New England. 3States in the Great Lakes. 74 Evers et al.

Relationship between egg-Hg and female concentrations tendedtoincrease andhada margin- blood-Hg concentrations ally significant increase between the low andextra- high Hg risk categories (t ˆ 1.79, p ˆ 0.08) (Table 4). We relatedegg-Hg concentrations with those of the blood-Hg concentrations of females captured from the Relationship of egg size and fertility with same territory (Fig. 2). This relationship was initially Hg concentrations separatedinto three categories accordingto our knowledge of the female±egg relationship. The first Egg size (as measuredby volume) was inversely category included blood-Hg concentrations from relatedto egg-Hg concentrations ( t ˆ 3.56, p ˆ 0.02) color markedfemales capturedthe same year that (Table 4). Eggs with very high Hg concentrations the target egg was collected( r2 ˆ 0.73, n ˆ 40). The (>2 mg/g ww) were 5% smaller than those with low secondcategory was comprisedof blood-Hg concen- Hg concentrations. Fertility andthe level of embryonic trations from color markedfemales that laidthe target development were scored for all eggs (Table 1). The egg in a different year (r2 ˆ 0.77, n ˆ 24). The third Hg concentrations of infertile eggs (0.78 Æ 0.5 mg/g, category was comprisedof blood-Hgconcentrations n ˆ 201) were not significantly different than the from females that were eventually capturedin the Hg concentrations of fertile eggs (0.74 Æ 0.54 mg/g, same territory, but the relationship of the individual n ˆ 205) (t ˆ 0.52, p ˆ 0.60) (Table 4). female with the target egg was not known (r2 ˆ 0.74, n ˆ 44). Differences among the three categories were not significantly different (F ˆ 0.32, p > 0.05). The Discussion relationship between egg andfemale bloodwas strongest when these three categories were combined Geographic trends in egg-Hg concentrations (r2 ˆ 0.79, n ˆ 108) (Fig. 2). We foundthat geographic trendsin egg-Hg concen- Intraclutch variation trations in the US were similar to those reportedfor Canada (Scheuhammer et al., 2001) (Fig. 3). Our Basedon 86 nests where two eggs were analyzed,we general trendof increasing egg-Hg concentrations founda significant difference in Hg concentrations from western to eastern North America also follows between eggs (F ˆ 52.1, p < 0.001). The mean geographic trends of adult and juvenile blood-Hg difference in egg-Hg concentrations was 25 Æ 21%, concentrations previously establishedby Evers et al. with a range of 0±84%. Intraclutch variation in Hg (1998). Although atmospheric Hg deposition is

8

7

6

g/g, ww) 5 µ

4

3

2 Female blood Hg ( 1 y = 1.5544x + 0.2238 R 2 = 0.79 0 012345 Egg–Hg (µg/g, ww)

Figure 2. A comparison between Hg concentrations in common loon eggs with bloodof females from the same territory. Mercury in Loon Eggs 75

Table 4. Estimatedcategories of Hg concentrations in common loon eggs andassociatedpotential impact.

Egg-Hg levels (mg/g, Egg-Hg categories Egg infertility (%)2 Intraclutch variation (%) Egg volume (g) ww anddw) 1 (mean Æ SD) (mean Æ SD)

0±0.60 (ww) Background(low) 53 ( n ˆ 201) 21.5 Æ 15.8 (n ˆ 26, 2-egg clutches) 149.3 Æ 14.4 (n ˆ 208) 0±2.82 (dw) 0.60±1.30 (ww) Elevated(moderate) 46 ( n ˆ 157) 24.6 Æ 20.6 (n ˆ 33, 2-egg clutches) 149.0 Æ 15.0 (n ˆ 167) 2.82±6.10 (dw) 1.30±2.00 (ww) Impacting (high) 54 (n ˆ 35) 31.3 Æ 0 24.9 (n ˆ 19, 2-egg clutches) 144.4 Æ 16.0à (n ˆ 54) 6.10±9.37 (dw) >2.00 (ww) Severe impacts (xhigh) 47 (n ˆ 17) 38.0 Æ 38.2à (n ˆ 8, 2-egg clutches) 141.8 Æ 14.0Ãà (n ˆ 15) >9.37 (dw)

ÃMarginally significantly different at the p < 0.10 level comparedto the backgroundegg-Hg concentration. ÃÃSignificantly different at the p < 0.05 level comparedto the backgroundegg-Hg concentration. 1Dry weight (dw) to wet weight (ww) conversion is based on a loon egg's average moisture content of 78.7%. 2The percent egg infertility represents eggs that were opportunistically collected from the nests that failed. This rate does not reflect egg fertility for the common loon, in general.

Western Alaska (n=10) N. America Alberta (n=19) Saskatchewan (n=38) Montana (n=5) Manitoba (n=5)

Minnesota (n=20) Wisconsin (n=78) Michigan (n=24) Ontario (west) (n=15)

Quebec (n=4) New York (n=28) Vermont (n=41) Ontario (east) (n=29)

New Hampshire (n=262) Maine (n=186) New Brunswick (n=9) Nova Scotia (n=8) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Eastern N. America Low risk Moderate risk High risk Xhigh risk

Figure 3. Proportion of common loon eggs in four Hg risk categories organized longitudinally in North America. Canadian data originated from Scheuhammer et al. (2001) andWisconsin datafrom Fevoldet al. (2003). increasing in the Arctic (Lockhart et al., 1995) and atmospheric Hg deposition pattern for North America seabird eggs demonstrate temporal trends of increas- (USEPA, 1997). However, this pattern of regional ing Hg concentrations (Braune et al., 2001), common atmospheric Hg input cannot explain the variability in loon eggs from Alaska remain as our best, low Hg egg andother biotic Hg concentrations that we and reference samples. Mean egg-Hg concentrations in others have reportedwithin regional or local areas. Maine are 3.6 higher than levels foundin Alaska. Some of the variability in local biotic Hg concentra- The general geographic pattern of Hg in loon eggs tions may potentially be due to differences in: (1) local that we detected is consistent with the general anthropogenic Hg inputs from certain urban/industrial 76 Evers et al. sources (e.g., municipal incinerators, coal-burning Although extrapolation of egg contaminant burdens plants, mining/smelting operations, or chlor-alkali from the analysis of one egg to the remainder of the plants); or (2) local inputs from natural geologic clutch is still considered standard for organochlorines sources of elevatedHg (e.g., cinnabar deposits).In (Custer et al., 1990), recent evidence for significant addition, Hg availability to biota is primarily a intraclutch variation in egg-Hg concentrations function of the availability of MeHg, which in turn indicates that a knowledge of laying order may be is primarily determined by hydrological and biogeo- requiredto make more accurate predictions (Morera chemical factors, andthus foodchain transfer of et al., 1997). Intraclutch variability occurs for Hg MeHg andexposure in top predatorssuch as loons concentrations in birdeggs because the first egg laid may be elevatedin some environments even in the typically has higher concentrations than those follow- absence of highly elevatedinputs of natural or anthro- ing (Becker andSperveslage, 1989; Becker, 1992). In pogenic Hg (Watras andHuchabee, 1994; Lucotte 3-egg clutches of herring gulls (Larus argentatus) and et al., 1999). Thus, large sample sizes are requiredto common terns (Sterna hirundo), Becker (1992) found confidently characterize the variability in biotic Hg concentrations in the first eggs laidwere, respec- Hg concentrations within andbetween different tively, 39 and37% higher than in the last egg. In local areas. our study, mean intraclutch differences for loon egg- Hg concentrations were 25 Æ 21% (CI ˆ 4±46%). However, 28% of loon clutches have only one egg Relationship of Hg concentrations between (McIntyre, 1988). Single egg analysis for Hg thus female blood and eggs represents a potentially confounding factor that could influence future study design and interpretation. A well-known depuration route for MeHg by female Over time, body burdens of Hg are known to birds is through the egg (Braune and Gaskin, 1987; significantly increase in adult loons in response to ele- Lewis et al., 1993). In controlledstudies,transfer of vated levels of dietary MeHg. Although blood-Hg dietary MeHg into eggs was rapid. Kambamanoli- concentrations may not demonstrate significant Dimou et al. (1991) showedthat, after a single oral increases over the lifetime of the same individual, dose of MeHg, Hg was found in chicken eggs begin- feather Hg concentrations do tend to increase with age ning two days after administration and peaked at (Evers et al., 1998). In New England, feather Hg con- 3 days, with a total of 50% of the MeHg dose being centrations significantly increasedat a rate of about transferredto the eggs. Sell et al. (1974) showedthat 5.7% annually for recapturedfemale loons (Evers chicken egg-Hg concentrations declined at a rate of et al., 2002). Hg in feathers is predominantly in the approximately 4% per day after dietary Hg adminis- MeHg form, andcorrelates with MeHg concentrations tration had ceased. In chickens, independent of feather in muscle. Muscle tissue provides the protein used molts, up to 82% of Hg intake can be depurated into during feather formation (Crewther et al., 1965) and is eggs (March et al., 1974). one of the few body compartments from which MeHg A comparison of the blood-Hg concentrations of can be remobilized. Tapping into this MeHg reserve adult females with those in eggs collected from the during molt serves as an important MeHg depuration same territory shows a highly significant positive rela- route for birds (Braune and Gaskin, 1987); and tionship (Fig. 2). Because blood-Hg concentrations feather-Hg concentration can thus be usedas an reflect dietary uptake, and eggs are a well-established indicator of MeHg body burden during feather growth depuration route for MeHg, this relationship is not in birds (Burger, 1993). unexpected. Loon egg-Hg concentrations are appro- However, there are three lines of evidence that priate predictors of MeHg bioavailability at the natal indicate that, unlike the situation for feathers, MeHg in lake. Potential variability in this relationship could muscle provides only minor contributions to Hg in the occur due to: (1) differences in egg-Hg concentrations developing egg. First, as we have shown in this study, within a clutch, (2) differences among female loons egg-Hg concentrations are strongly correlatedwith with respect to existing Hg body burdens, and (3) use female blood-Hg concentrations, which closely reflect by female loons of multi-lake territories for foraging. prey Hg concentrations on the breeding lake (Burgess Further explanation of these potential confounding et al., 1998a; Evers et al., 2002). Second, female factors follows. feather-Hg concentrations do not correlate with Mercury in Loon Eggs 77 egg-Hg concentrations (F ˆ 1.24, df ˆ 257, p > 0.05; our study lakes contained preferred prey and size Biodiversity Research Institute (BRI), unpubl. data). classes (yellow perch, Perca flavescens, between 10 Last, because dietary MeHg exposure on the loon's and15 cm) (Barr, 1996) for which Hg concentations marine wintering areas is generally low [adult loons exceeded 0.3 and 0.4 mg/g (Evers et al., 2002). capturedfrom coastal Maine south to Floridahadmean Unlike controlled dosing studies that showed a blood-Hg concentrations of 0.47 Æ 0.23 mg/g with no reduction in egg fertility as egg-Hg concentrations samples exceeding 1.0 mg/g (n ˆ 63, BRI unpubl. increased(Fimreite, 1971; Thaxton andParkhurst, data)] compared to that on their fresh water breeding 1973), we did not find a significant relationship lakes, the contribution to Hg in loon eggs from MeHg between egg-Hg concentrations andpercent of storedin muscle or other tissues is probably minor infertile eggs (Table 4). However, we collectedonly comparedwith contributions arising from the dietthat abandoned eggs; a random sample of eggs is needed is being consumed during egg development. before the potential impacts of Hg on loon egg fertility Although Hg depuration through eggs is substan- can be properly assessed. In the current study, Hg tial in female birds, relationships between eligible concentrations in loon eggs did not significantly differ body burdens of MeHg in key tissues, such as muscle, among five developmental stage categories (F ˆ 1.41, andHg deposition into eggs, are generally weak p ˆ 0.23). But because we frequently collectedonly (Lewis et al., 1993; Bearhop et al., 2000). one egg from 2-egg clutches, anddidnotexplicitly Although exceptions are known, such as the osprey investigate causes of hatching failure, further inves- (Pandion haliaetus) (Hughes et al., 1997) we have tigation of this issue is warranted. foundthat, in the common loon, egg-Hg concen- We foundthat intraclutch variability in egg-Hg trations primarily reflectedMeHg concentrations of concentrations increasedas Hg concentrations major prey items from the lake on which the female increased. It is likely that, for laying females with loon feeds and nests (Evers et al., 1998; Scheuhammer higher blood-Hg concentrations, there will be a greater et al., 1998b). However, loons that maintain multi- difference in the Hg concentrations of the two eggs lake territories will tendto increase the variability of laid. Depending on a study's objectives, this confound- egg-Hg vs. blood-Hg relationships. Loon pairs nesting ing factor further underlines the need to analyze both on water bodies <24 ha generally incorporate more eggs in a clutch, particularly in situations where than one lake in their territory (Piper et al., 1997b); in MeHg availability is high. such circumstances, adult loons may use lakes neigh- A relationship between smaller egg size andhigher boring the natal lake for foraging purposes. Because Hg concentrations has been shown previously in a MeHg production and availability can vary substan- controlledstudy(Fimreite, 1971). Similarly, in the tially among lakes that are in close proximity, or even current study, we found that egg volume significantly within the same watershed, blood-Hg levels for more decreased as Hg concentration increased. The egg-Hg mobile loons may not accurately represent the dietary risk category of ``>1.3 mg/g'' was comprisedof eggs uptake of MeHg from the natal lake. significantly smaller than in the Hg risk category of ``0±0.6 mg/g'' (t ˆ 2.14, p ˆ 0.03). Independent of Hg Reproductive effects of Hg toxicity, this finding is contrary to previous studies that reportedthat the first loon egg laidis the largest Controlledstudieshave shown that Hg negatively (Yonge, 1981; McIntyre, 1988); andthat, in birds affects embryonic development and hatchability of generally, the first egg laidhas higher Hg concentra- avian eggs at levels that are foundin some eggs in tions than subsequent eggs (Becker andSperveslage, the current study (Fimreite, 1971; Gilbertson, 1974; 1989; Becker, 1992). But because, in loons, egg size Heinz, 1979). Thompson (1996) summarizedseveral (measuredas fresh egg weight) is strongly relatedto controlled dosing studies of captive birds and con- female body weight (r2 ˆ 0.82, basedon n ˆ 613 eggs cluded that egg-Hg concentrations of 0.50 mg/g (ww) and497 female weights, BRI unpubl. data),and may cause impairment in some birdspecies. Barr because a lower female body weight reduces reproduc- (1986) foundloons laidfewer eggs when prey-Hg tive success (in ducks, Blums et al., 1997), our find- concentrations averaged0.3±0.4 mg/g, andegg-Hg ings may indicate a decreased physiological condition concentrations exceeded 0.60 mg/g; no eggs were laid of female loons in habitats with high risk for elevated when prey averagedover 0.4 mg/g of Hg. Several of Hg exposure. 78 Evers et al.

Establishing Hg risk levels for loon eggs provide our best estimate for interpreting egg-Hg concentrations in common loons foundacross North The many studies investigating Hg concentrations America (Fig. 3). Using these categories, a regional in avian eggs andassociatedeffects providea basis comparison indicates low Hg risk for common loons for estimating risk levels for the common loon in Alaska, Montana andmuch of the CanadianPrairie (Thompson, 1996; Scheuhammer et al., 2001). We Provinces. Fox et al. (1980) foundlow Hg risk on have identified four categories of risk based on Hg in Hanson Lake in east-central Saskatchewan where 31 loon eggs (Table 4). eggs hadHg concentrations ranging from 0.13 to Thefirst category of0±0.60 mg/g (low risk) indicates 0.75 mg/g. backgroundHg concentrations that reflect (1) natural The western Great Lakes region is characterizedby environments that do not experience enhanced local low to moderate Hg risk, although there are well geological sources of Hg or inputs from natural Hg known hotspots in north-central Wisconsin (Meyer sources, such as volcanic activities, nor from anthro- et al., 1998; Fevoldet al., 2003) andwestern Upper pogenic deposition; and (2) environments that may Peninsula of Michigan (Evers et al., 1998) relatedto receive enhancedHg inputs (either natural or anthro- the prevalence of acidic lakes; and in north-western pogenic), but which lack local conditions that favor Ontario relatedto a chlor-alkali plant (Parks and MeHg production and availability. Evidence from the Hamilton, 1987). In this latter area, Barr (1986) found literature indicates no adverse impacts to birds at egg- Hg concentrations in loon eggs from the mid1970s Hg concentrations <0.50 mg/g (Thompson, 1996). ranging from 0.54 to 1.39 mg/g (n ˆ 34), anddocu- Because Barr (1986) foundreproductive impairment mented reduced reproductive success. Belant and inloonswhenHgconcentrationsexceeded0.60 mg/gwe Anderson (1990) found egg-Hg concentrations ranged have usedthis value as our best estimate of a lowest from 0.40 to 1.10 on the Turtle±Flambeau Flow- observedadverseeffect level. The secondcategory of age in north-central Wisconsin (mean of 0.64 Æ 0.21, 0.60±1.30 mg/g (moderate risk) reflects elevated levels n ˆ 16). of environmental Hg that may be indicative of The majority of loon eggs collectedin the eastern significant reproductive impairment in some indivi- Great Lakes region have Hg concentrations indicative duals of some avian species, including loons. A third of low to moderate risk, but compared with the category of >1.3 mg/g (high risk) is indicative of an western Great Lakes region, there appears to be a some- environment where loons are at high risk for reproduc- what greater risk posedto more eastern loons, with tive impairment. Adult female loons with blood-Hg some lakes exhibiting high (Vermont) andextra high concentrations >3.0 mg/g, laideggs containing Hg risks (eastern Ontario). Again, this may be due >1.30 mg/g (Fig. 2), andoften haddecreasedreproduc- primarily to a prevalence of low pH/low alkalinity tive success, laying significantly fewer eggs, which in environments. Between 1969 and1980, Frank et al. turn demonstrateddecreased hatching success (Evers et (1983) collectedeggs in the Algonquin-Parry Sound al., 2002). Furthermore, egg size was significantly area andfoundmean Hg concentrations ranging lower for eggs with Hg concentrations >1.30 mg/g from 0.81 to 1.11 mg/g, with some eggs exceeding comparedto eggs in low andmoderate risk categories. 1.40 mg/g, indicating that some high risk lakes existed Similarly, concentrations >1.3 mg/g were suspectedto during that time period. Scheuhammer and Blancher result in reduced hatching success in a field study with (1984) estimatedthat 30% of lakes in eastern and common terns (Fimreite, 1974) andwere associated central Ontario containedsmall fish with Hg con- withreducedhatchlingsurvivalinacontrolledlabstudy centrations sufficiently high to pose a risk to loon with mallards (Heinz, 1976a,b). Finally, egg-Hg reproduction. Although the eggs we collected in New concentrations exceeding the 2.0 mg/g (ww) (Xhigh York did not indicate high Hg risk, some reservoirs risk) are widely considered to have detrimental repro- (McIntyre et al., 1993) andlow pH lakes (Simonin ductive effects in most bird species (Thompson, 1996). et al., 1994) are known to contain fish with Hg con- centrations that place loons at high risk. Eggs collected Regional rating of Hg risk from New York's Stillwater Reservoir between 1978 and1986 ( n ˆ 25) hada mean Hg level of Basedon our findings andon comparisons with the 1.6 Æ 0.6 mg/g with 64% of the eggs having > 1.3 mg/g results of other studies, the four egg-Hg risk categories and28% having >2.0 mg/g (Holmquist, 1990). Mercury in Loon Eggs 79

Loon eggs from northeastern North America are Tom Hodgman, Joseph Kaplan, Lynn Kelly, Oksana generally in higher risk categories than those from Lane, Myron Lysne, Kori Marchowsky, Tamara other areas of North America, with some areas in Mills, Kyle Murphy, Brian Olsen, Lucas Savoy, Maine, New Hampshire, andNova Scotia containing Nina Schoch, Keren Tischler, Bev Vallee, andDave loon eggs that are likely not hatching because Yates for assistance in collecting andprocessing eggs. of environmental MeHg transfer (i.e., >2.0 mg/g). Funding was provided by the Maine Department of Certain geographic Hg hotspots are now well known Environmental Protection, FPL Energy Maine Hydro, in southeastern New Hampshire, the western moun- US Environmental Protection Agency, andthe US tains of Maine, near a chlor-alkali plant in Downeast Fish andWildlife Service. We also thank Bryan Maine (now closed), and in Kejimkujik National Park, Brattin, Carol Buckley, Kay Cavanaugh, Matthew Nova Scotia. Investigations into potential population- Darsey, andIrene Rudikfor assistance in the lab. level impacts of Hg on loons breeding in northeastern Comments from three reviewers, Gary Heinz, Neil North America are ongoing (Burgess et al., 1998b; Burgess, andan anonymous reviewer, were greatly Evers et al., 2002) appreciated.

Wildlife conservation and Hg References

Barr, J.F. (1986). Population dynamics of the common loon (Gavia Developing regulatory policies for reducing Hg immer) associatedwith mercury-contaminatedwaters in north- emissions andeffluents are difficult, but high prio- western Ontario. Can. Wildl. Serv. Occas. Pap. 56. Ottawa, rity, conservation issues. Policy-makers at national Ontario, Canada. (USEPA, 1997) andregional (NESCAUM, 1998) Barr, J.F. (1996). Aspects of common loon (Gavia immer) feeding levels have compiledandassessedthe available eco- biology on its breeding ground. Hydrobiologia 32, 119±44. Bearhop, S., Rutxon, G.D. andFurness, R.W. (2000). Dynamics of logical andhuman health concerns andresearch. mercury in bloodandfeathers of great skuas. Environ. Toxicol. Landscape-level investigations into the distribution Chem. 19, 1638±43. patterns of Hg have been made possible by programs Becker, P.H and(1992). Egg mercury levels declinewith the laying such as the US Department of Agriculture's North- sequence in Charadriiformes. Bull. Environ. Contam. Toxicol. eastern Ecosystem Research Cooperative. However, 48, 762±7. Becker, P.H. andSperveslage, H. (1989). Organochlorines and while our andother studieson Hg in common loons heavy metals in herring gull (Larus argentatus) eggs andchicks are contributing valuable new scientific information, from the same clutch. Bull. Environ. Contam. Toxicol. 42, 721±7. there will ultimately remain a needfor more ecologi- Belant, J.L. andAnderson,R.K. (1990). Environmental contami- cally relevant andcomprehensive approaches (Peakall nants in common loons from northern Wisconsin. Passenger andTucker, 1985). Indications of ecological risk in Pigeon. 52, 306±10. Bischoff, K., Pichner, J., Braselton, W.E., Counard, C., Evers, D.C. northeastern North America from this andother and Edwards, W.C. (2002). Mercury and Selenium Concentra- relatedloon-Hg studiesunderscore a pressing need tions in livers andeggs of common loons ( Gavia immer) from for broader investigations. Minnesota. Arch. Environ. Contam. Toxicol. 42, 71±6. Blums, P., Mednis, A. and Clark, R.G. (1997). Effect of incubation body mass on reproductive success and survival of two European Acknowledgements diving ducks: a test of the nutrient limitation hypothesis. Condor. 99, 916±25. Braune, B.M., Donaldson, G.M. and Hobson, K.A. (2001). Con- We thank members of the Northeast Loon Study taminant residues in seabird eggs from the Canadian Arctic. Working Group with special recognition of the Part 1. Temporal trends 1975±1998. Environ. Pollut. 114, 39±54. Adirondack Loon Association, BioDiversity Research Braune, B.M. andGaskin, D.E. (1987). Mercury levels in Institute, Loon Preservation Committee, Maine Bonaparte's gulls (Larus philadelphia) during autumn molt in the Quoddy Region, New Brunswick, Canada. Arch. Environ. Audubon Society, and Vermont Institute of Natural Contam. Toxicol. 16, 539±49. Sciences. The Michigan Department of Natural Burger, J. (1993). Metals in avian feathers: bioindicators of envir- Resources andMontana Loon Society were instru- onmental pollution. Rev. Environ. Toxicol. 5, 203±311. mental in providing eggs from their representative Burgess, N.M., Evers, J.D. andKaplan, J.D. (1998a). Mercury states. We also thank Tom Cooley, Cory Counard, levels in the bloodof common loons breedingin the Maritimes andtheir prey. In Mercury in Atlantic Canada: A Progress Chris DeSorbo, Nancy Fair, Patty Freeman, Susan Report, Environment Canada, pp. 96±100. New Brunswick, Gallo, Wing Goodale, Eric Hansen, Bill Hanson, Canada: Sackville. 80 Evers et al.

Burgess, N.M., Evers, D.C., Kaplan, J.D., Duggan, M. and common loons, Ontario, Canada, 1968±1980. Arch. Environ. Kerekes, J.J. (1998b). Mercury andreproductivesuccess of Contam. Toxicol. 12, 641±54. Common Loons breeding in the Maritimes. In Mercury in Gilbertson, M. (1974). Seasonal changes in organochlorine com- Atlantic Canada: A Progress Report. Environment Canada, pounds and mercury in common terns of Hamilton Harbour, pp. 104±109. Sackville, Canada: New Brunswick. Ontario. Bull. Environ. Toxicol. 12, 726±32. Chen, C.Y., Stemberger, R.S., Klaue, B., Blum, J.D., Pickhardt, P.C. Gostomski, T.J. andEvers, D.C. (1998). Time-activity budgetfor andFolt, C.L. (2000). Accumulation of heavy metals in food common loons, Gavia immer, nesting on Lake Superior. Can. web components across a gradient of lakes. Limnol. Oceanogr Field-Nat. 112, 191±7. 45, 1525±36. Hatch, W.R. andOtt, W.L. (1968). Determination of sub- Crewther,W.G.,Fraser,R.D.B.,Lennox,F.G.andFindley,H.(1965). microgram quantities of mercury by atomic absorption spectro- The chemistry of keratins. In C.B. Anifinsen, M.L. Anson, J.T. photometry. Anal. Chem. 40, 2085±7. Edsoll and F.M. Richards (eds) Advances in Protein Chemistry, Heinz, G.H. (1974). Effects of low dietary levels of methylmercury pp. 191±303. New York, USA: Academic Press. on mallardreproduction. Bull. Environ. Contam. Toxicol. 11, Custer, T.W., Pendleton, G. and Ohlendorf, J.M. (1990). Within- 386±92. andamong-clutch variation of organochlorine residuesin eggs Heinz, G.H. (1976a). Mehtylmercury: second-year feeding of black-crownednight-herons. Environ. Monitor. Assess. 15, effects on mallard reproduction and duckling behavior. J. Wildl. 83±9. Manag. 40, 82±90. Cuvin-Aralar, M.L. andFurness, R.W. (1991). Mercury and Heinz, G.H. (1976b). Methylmercury: Second-generation reproduc- selenium interaction: a review. Ecotoxicol. Environ. Saf. 21, tive andbehavior effects on mallardducks. J. Wildl. Manag. 40, 348±64. 710±5. Dhar, A.K., Pokras, M.A., Garcia, D.K., Evers, D.C., Gordon, Z.J. Heinz, G.H. (1979). Methylmercury: Reproduction and behavioral andAlcivar-Warren, A. (1997). Analysis of genetic diversityin effects on three generations of mallardducks. J. Wildl. Manag. common loon Gavia immer using RAPD andmitochondrial 43, 394±400. RFLP techniques. Mol. Ecol. 1997(6), 581±6. Henny, C.J. andHerron, G.B. (1989). DDE, selenium, mercury and Evers, D.C. (1994). Activity budgets of a marked common loon white-faced ibis reproduction at Carson Lake, Nevada. J. Wildl. (Gavia immer) nesting population. Hydrobiologia. 279/280, Manag. 53, 1032±45. 415±20. Holmquist, C.L. (1990). Trends in environmental contaminants Evers, D.C. (2001). Common loon population studies: continental in common loon eggs from New York Adirondacks mercury patterns andbreedingterritory philopatry. Ph.D. andNew Hampshire. Concord,NH, USA: USFWS Unpubl. Dissertation, Univ. Minn., St. Paul MN, USA. Rept. Evers, D.C., Kaplan, J.D., Meyer, M.W., Reaman, P.S., Hoyt, D.F. (1979). Practical methods of estimating volume and Braselton, W.E., Major, A., Burgess, N. andScheuhammer, A.M. fresh weight of birdeggs. Auk 96, 73±7. (1998). Geographic trendin mercury measuredin common loon Hughes, K.D., Ewins, P.J. andClark, K.E. (1997). A comparison feathers andblood. Environ. Toxicol. Chem. 17, 173±83. of mercury levels in feathers andeggs of osprey ( Pandion Evers, D.C., Lane, O.P., De Sorbo, C. andSavoy, L. (2002). haliaetus) in the North American Great Lakes. Arch. Environ. Assessing the impacts of methylmercury on piscivorous wildlife Contam. Toxicol. 33, 441±52. using a wildlife criterion value based on the common loon, Kambamamonli-Dimou, A., Kamarianos, A. andKilikidis,S. 1998±2001. Report BRI2002-08 submitted to the Maine Dept. (1991). Transfer of methylmercury to hens' eggs after oral Environ. Protection. Falmouth ME, USA: BioDiversity administration. Bull. Environ. Contan. Toxicol. 46, 128±33. Research Inst. Lewis, S.A., Becker, P.H. andFurness, R.W. (1993). Mercury Fevold, B.M., Meyer, M.W., Rasmussen, P.W. and Temple, S.A. levels in eggs, tissues, andfeathers of herring gulls Larus (2003). Bioaccumulation patterns andtemporal trendsof mer- argentatus from the German Wadden Sea Coast. Envrion. cury exposure in Wisconsin common loons. Ecotoxicology. Pollut. 51, 293±9. 12(1±4), 83±93. Lockhart, W.L., Wilkinson, P., Billeck, B.N., Hunt, R.V., Fimreite, N. (1971). Effects of dietary methylmercury on ring- Wagemann, R. andBrunskill, G.J. (1995). Current andhistori- neckedpheasants with special reference to reproduction. Can. cal inputs of mercury to high-latitude lakes in Canada and to Wildl. Serv. Occas. Pap. No. 9. Ottawa Ontario, Canada. Hudson Bay. Water Air Soil Pollut. 80, 603±10. Fimreite, N. (1974). Mercury contamination of aquatic birds in Lucotte, M., Schetagne, R., Therien, N., Langlois, C. and northwestern Ontario. J. Wildl. Manage. 38, 120±31. Tremblay, A. (eds) (1999). Mercury in the Biogeochemical Finley, M.T. and Stendell, R.C. (1978). Survival and reproductive Cycle. NY, USA: Springer Verlag. success of black ducks fed methylmercury. Environ. Pollut. 16, March, B.E., Soong, R., Blinski, E. andJonas, R.E.E. (1974). 51±64. Effects on chickens of chronic exposure to mercury at low Fox, G.A., Yonge, K.S. andSealy, S.G. (1980). Breedingperform- levels through dietary fish meal. Poultry Sci. 53, 2175±81. ance, pollutant burden and eggshell thinning in common loons McIntyre, J.W. (1988). The Common Loon: Spirit of the Northern Gavia immer nesting on a boreal forest lake. Ornis Scandinavica Lakes. Minneapolis, MN, USA: Univ.Minn.Press. 11, 243±8. McIntyre, J.W. andEvers, D.C. (eds)(2000). Loons: Old History Frank, R., Lumsden, H., Barr, J.F. and Braun, H.E. (1983). Resi- and New Findings. Proc. of a Symposium from the 1997 Meet- dues of organochlorine insecticides, industrial chemicals, and ing, American Ornnithologists' Union. Holderness, NH, USA: mercury in eggs andin tissues from healthy andemaciated N. Am. Loon Fund. Mercury in Loon Eggs 81

McIntyre, J.W., Hickey, J.T., Karwowski, K, Holmquist, C.L. and common loons (Gavia immer) from Canada. Environ. Monitor. Carr, K. (1993). Toxic chemicals andcommon loons in New Assess. 72, 79±94. York andNew Hampshire, 1978±1986. In L. Morse, S. Stockwell Scheuhammer, A.M., Wong, A.H.K. andBond,D.E. (1998a). andM. Pokras (eds) The Loon and its Ecosystem: Status, Mercury andselenium accumulation in common loons ( Gavia Management, and Environmental Concerns, pp. 73±91. immer) andcommon mergansers ( Mergus merganser) from Concord, NH, USA: US Fish Wildl. Serv. eastern Canada. Environ. Toxicol. Chem. 17, 197±201. Meyer, M.W., Evers, D.C., Hartigan, J.J. andRasmussen, P.S. Sell, J.L., Guenter, W. andSifri, M. (1974). Distribution of mercury (1998). Patters of common loon (Gavia immer) mercury among components of eggs following the administration of exposure, reproduction, and survival in Wisconsin, USA. methylmercuric to chickens. J. Agr. Food. Chem. 22, 248±51. Environ. Toxicol. Chem. 17, 184±90. Simonin, H.A., Gloss, S.P., Driscoll, C.T., Schofield, C.L., Morera, M., Sanpera, C., Crespo, S., Jover, L. andRuiz, X. (1997). Kretser, W.A., Karcher, R.W. andSymula, J. (1994). Mercury Inter- andintra-clutch variability in heavy metals andselenium in yellow perch from Adirondack drainage lakes (New York, levels in Audouin's gull eggs from the Ebro Delta, Spain. Arch. US). In C.J. Watras andJ.W. Huckabee (eds) Mercury Pollu- Environ. Contam. Toxicol. 33, 71±5. tion: Integration and Synthesis, pp. 457±69. Boca Raton, FL, NESCAUM (1998). Northeast States and Eastern Canadian USA: Lewis Publ. Provinces: Mercury Study: a Framework for Action. Boston, Stickel, L.F., Wiemeyer, S.N. andBlus, L.J. (1973). Pesticide MA, USA: NESCAUM. residues in eggs of wild birds: adjustment for loss of moisture Nocera, J. andTaylor, P. (1998). In situ behavioral response of andlipid. Bull. Environ. Contam. Tox. 9, 193±6. common loons associatedwith elevatedmercury exposure. Thaxton, P. andParkhurst, C.R. (1973). Abnormal mating behavior Conserv. Ecol. 2, 10. and reproductive dysfunction caused by mercury in Japanese Parkhurst, D.F. (1998). Arithmetic versus geometric means for quail.. Proc. Soc. Exp. Biol. Med. 144, 252±5. environmental concentration data. Environ. Sci. Technol. 32, Thompson, D.R. (1996). Mercury in birds and terrestrial animals. In 92A±98A. W.N. Beyer, G.H. Heinz and A.W. Redmon-Norwood (eds) Parks, J.W. andHamilton, A.L. (1987). Accelerating recovery of Environmental Contaminants in Wildlife: Interpreting Tissue the mercury-contaminatedWabigoon/English River system. Concentrations., pp. 341±55. Boca Raton FL, USA: Lewis Publ. Hydrobiologia. 149, 159±88. Tremblay, A. (1999). Bioaccumulation of mercury andmethylmer- Peakall, D.B. andTucker, R.K. (1985). Extrapolation from single cury in invertebrates from natural boreal lakes. In M. Lucotte, species studies to populations, communities and ecosystems. In R. Schetagne, N. Therien, C. Langlois andA. Tremblay (eds) V.B. Vouk, G.C. Butler, D.G. Hoel andD.B. Peakall (eds) Mercury in the Biogeochemical Cycle, pp. 89±112. NY, USA: Methods for Estimating Risk of Chemical Injury: Human and Springer Verlag. Non-human Biota and Ecosystems. SCOPE Vol. 26, pp. 611±36. USEPA (1997).Mercury study report to congress.Volume VII: Piper, W., Paruk, J.D., Evers, D.C., Meyer, M.W., Tischler, K.B., characterization of human health and wildlife risks Kaplan, J.D. andFleischer, R.C. (1997a). Genetic monogamy in from mercury exposure in the United States. EPA-452/R- common loon.. Behav. Ecol. Sociobiol. 41, 25±31. 97-009. Piper, W., Paruk, J.D., Evers, D.C., Meyer, M.W., Tishler, K.B., Wagemann, R., Trebacz, E., Hunt, R. andBoila, G. (1997). Percent Klich, M. andHartigan, J.J. (1997b). Local movements of color- methylmercury andorganic mercury in tissues of marine markedcommon loons. J. Wildl. Manag. 61, 1253±61. mammals andfish using differentexperimental andcalculation Posthuma, L., Suter, G.W. andTraas, T.P. (2002). Species Sensi- methods. Environ. Toxicol. Chem. 16, 1859±66. tivity Distributions in Ecotoxicology. Boca Raton, Florida, Watras, C.J. andHuckabee, J.W. (1994). Mercury Pollution: Inte- USA: Lewis Publ. gration and Synthesis. Boca Raton, FL, USA: Lewis Publ. SAS Institute (1999). JMP Statistical Discovery Software. Cary, Wiemeyer, S.N., Lamont, T.G., Bunck, C.M., Sindelar, C.R., NC, USA: SAS Institute. Gramlich, F.J., Fraser, J.D. andByrd,M.A. (1984). Organo- Scheuhammer, A.M. (1991). Effects of acidification on the avail- chlorine pesticide, polychlorobiphenyl and mercury residues in ability of toxic metals andcalcium to wildbirdsandmammals. baldeagle eggsÐ1969±1979Ðandtheir relationship to shell Environ. Pollut. 71, 329±75. thinning andreproduction. Arch. Environ. Contam. Toxicol. Scheuhammer, A.M., Atchison, C.M., Wong, A.H.K. and 13, 529±49. Evers, D.C. (1998b). Mercury exposure in breeding common Wiener, J.G. andSpry, D.J. (1996). Toxicological significance of loons (Gavia immer) in central Ontario, Canada. Environ. Tox- mercury in freshwater fish. In W.N. Beyer, G.H. Heinz and icol. Chem. 17, 191±6. A.W. Redmon-Norwood (eds) Environmental Contaminants in Scheuhammer, A.M. andBlancher, P.J. (1994). Potential risks to Wildlife: Interpreting Tissue Concentrations, pp. 297±339. common loons (Gavia immer) from methylmercury exposure in Boca Raton, FL, USA: Lewis Publ. acidified lakes. Hydrobiologia 279/280, 445±55. Yonge, K.S. (1981). The breeding cycle and annual production of Scheuhammer, A.M., Perrault, J.A. andBond,D.E. (2001). Mer- the common loon (Gavia immer) in the boreal forest region cury, methylmercury, andselenium concentrations in eggs of M.S. Thesis, Univ. Manitoba, Winnipeg, Canada.