Trace Element Concentrations in Blood of Nesting King Eiders inthe Canadian Arctic

Mark Wayland Æ Ray T. Alisauskas ÆDana K. Kellett Æ Katherine R. Mehl

Received: 22 November 2007 / Accepted: 21 January 2008 / Published online: 21 February 2008

� Springer Science+Business Media, LLC 2008

Abstract The king eider (Somateria spectabilis) is a

migratory species of sea duck whose North American

population is thought to be declining. We determined

levels of cadmium, lead, selenium, and mercury in blood

from female king eiders nesting in the central Canadian

Arctic from 2001 to 2003. Year-to-year repeatability esti-

mates were calculated from birds sampled in 2 or 3 years.

Repeatability coefficients were 0.45, 0.35, 0.58, and 0.25

for cadmium, lead, selenium, and mercury, respectively.

The first three were significantly different from zero

(p \ 0.05), whereas the last approached significance

(0.05 \ p \ 0.1). In 2001 and 2002, we also identified

probable wintering locations of a subset of the birds. In

both years, cadmium levels were higher and selenium

levels were lower in birds inferred to have wintered in the

eastern part of their range compared to those that had

wintered in the west. There was little evidence that timing

of breeding, timing of sampling, or body condition were

related to levels of these trace elements, although in 1 of 2

years, lead levels were influenced by body condition and

nest initiation date (R 2 = 0.24) and cadmium levels were

related to incubation day (partial R2 = 0.04). Year-to-year

repeatability of cadmium and selenium levels among

individuals in this population of king eiders was likely

influenced by where they wintered.

The king eider (Somateria spectablis) is a migratory spe-

cies of sea duck whose population is thought to have

declined by as much as 55% in North America (Suydam

2000). Exposure to environmental contaminants has been

identified as a possible cause of population decline (Sea

Duck Joint Venture Management Board 2001). Concen-

trations of some trace elements are particularly high in king

eiders (Dietz et al. 1996; Stoudt et al. 2002; Wayland et al.

2001a). Moreover, trace element concentrations in blood of

individual nesting king eiders exhibit a degree of consis-

tency from one year to the next, yet vary widely among

individuals (Wayland et al. 2007), suggesting that some

individuals might be at greater risk than others over the

long term.

Levels of trace elements and their effects on organisms

are influenced by numerous factors related to habitat,

internal biology and physiology, and life history (Peakall

and Burger 2003). Understanding how these factors influ-

ence trace element concentrations in king eiders could be

useful in evaluating why some individuals experience

greater long-term exposure than others. For example,

migration is a life-history trait that can influence trace

element levels in birds across seasons (Peakall and Burger

2003). In king eiders, it has not been ascertained if dif-

ferential exposure to trace elements between west- and

east-wintering birds (Dietz et al. 1996; Stoudt et al. 2002;

Wayland et al. 2001a) carries over to the breeding season

when birds that winter in different seas share common

breeding areas (Mehl et al. 2004).

M. Wayland (&) � R. T. Alisauskas � D. K. Kellett

Environment Canada, Prairie & Northern Region,

115 Perimeter Rd, S7N 0X4 Saskatoon, SK, Canada

e-mail: mark.wayland@ec.gc.ca

R. T. Alisauskas � K. R. Mehl

Department of Biology, University of Saskatchewan,

S7N 5E2 Saskatoon, SK, Canada

K. R. Mehl

Department of Biology, University of North Dakota,

10 Cornell St, Grand Forks, ND 58201-9019, USA

123

Arch Environ Contam Toxicol (2008) 55:683–690

DOI 10.1007/s00244-008-9142-5

Changes in body condition are associated with repro-

duction in female king eiders (Kellett and Alisauskas

2000). Such changes have been linked to variation in

contaminant levels in organs like liver and kidney (Way-

land et al. 2005), and in blood of waterfowl (Franson et al.

1999; Grand et al. 2002). Timing of breeding is considered

to be a heritable life-history trait in birds (Batt and Prince

1979; Møller 2001; Petersen 1992). Variables related to the

timing of breeding such as nest initiation date can influence

blood trace element levels, as can other time-related vari-

ables such as incubation stage and time since departure

from marine wintering areas for freshwater breeding areas

(Franson et al. 1999; 2000; Grand et al. 2002; Wilson et al.

2004).

In this study, we examined variation in levels of cad-

mium, lead, selenium, and mercury in blood of nesting

female king eiders in the central Canadian Arctic, where

birds from both the eastern and western wintering areas

mix. In particular, we determined whether trace element

levels differed between birds that wintered in the east and

those that wintered in the west, as inferred from stable

isotope analysis (Mehl et al. 2004). Finally, we assessed

whether variation in body condition and timing of blood

sampling, relative to incubation day and nest initiation

date, were related to trace metal levels.

Materials and Methods

Field Work

All field work was done at Karrak Lake (67�14’N,

100�15’W) and Adventure Lake (67�14’N, 100�09’W) in

the Queen Maud Gulf Bird Sanctuary, Nunavut, Canada

from 2001 to 2003. Most analyses were limited to birds

sampled in 2001 and 2002 because wintering location data

were limited to those years (see Statistics section). The

study area and field methods were described in detail by

Kellett and Alisauskas (1997) and Mehl et al. (2004).

Briefly, nest searching began in mid-June and continued

weekly throughout nesting. Eggs were candled to estimate

incubation day (Weller 1956). Nest initiation dates were

calculated by backdating from known laying dates or from

incubation days. Nesting females were captured on nests

during the last half of incubation. Birds captured for the

first time were banded with US Fish and Wildlife Service

bands. Band numbers of recaptured birds were recorded.

Birds were weighed (±5 g) and measured (culmen1: bill

notch to tip; culmen2: proximal end of bill to tip, head

length, head width, tarsal length, wing c ord) and blood was

sampled using heparinized syringes. Blood samples were

transferred to acid-washed cryovials, kept in a cooler and

transferred to a freezer at the end of each day. In 2001–

2002, crown feathers were also collected and stored in

separate paper envelopes.

Trace Element Analysis

Cadmium in blood was analyzed by graphite furnace

atomic absorption spectrometry with Zeeman background

correction on a Perkin-Elmer ZL 4100 at the Institut Pub-

lique de Sante Publique du Quebec according to Stoeppler

and Brandt (1980). Zeeman background correction is

employed to accurately correct for background interference

and results in lower detection limits and more accurate

determinations of cadmium levels, especially at low con-

centrations. Lead, selenium, and mercury were analyzed at

Environment Canada’s National Wildlife Research Insti-

tute according to methods described by Neugebauer et al.

(2003). Wayland et al. (2007) described the analytical

methods and quality assurance results in detail.

Statistics

In all analyses, trace element levels were log-transformed

before analysis. In a recent article (Wayland et al. 2007),

we found a high degree of interyear correlation in trace

element levels for individual eiders that were sampled in

C2 years. In the present study, we reexamined interindi-

vidual variation in trace element levels by calculating a

repeatability estimate and its standard error from a one-way

analysis of variance with individuals as treatment levels

(Becker 1984) and tested whether repeatability differed

from zero according to Lessels and Boag (1987). Repeat-

ability is the within-group correlation coefficient and is

measured as the ratio of the among-individuals variance

component divided by the sum of the among- and within-

individuals variance components. Repeatability differed

from zero (p \ 0.05) for three of four trace elements (see

Results section), suggesting that trace element levels

determined in the same individual in C2 years were not

independent of one another. Therefore, further analyses of

trace elements data, as described below, were done sepa-

rately for each year of the study.

We derived the first principal component (PC1) from the

correlation matrix of body size measurements for use as a

metric of structural size. In this analysis, all body size

measurements were weighted equally. PC1 accounted for

38% of the overall variation in body size measurements.

All measurements of body size loaded positively on PC1

(culmen1 = 0.37, culmen2 = 0.53, head length = 0.57,

head width = 0.24, tarsal length = 0.33, wing cord =

0.30), so that larger birds had higher PC1 scores than

smaller birds. It might have been equally effective to have

based the PC1 variable on fewer measures because of the

redundancy between culmen1 and culmen2 and the

684 Arch Environ Contam Toxicol (2008) 55:683–690

123

relatively poor repeatability of the head width measure-

ment (K. Mehl, personal observation). Two measures of

body condition were calculated: (1) body mass adjusted for

body size and (2) body mass adjusted for body size,

incubation day, and nest initiation date. The former pro-

vides an estimate of body condition at the time of sampling

irrespective of incubation day and initiation date, whereas

the latter removes variation due to these two variables from

the estimate of body condition.

For birds sampled in 2001 and 2002, Mehl et al. (2004,

2005) used predictive equations based on stable isotope

ratios of nitrogen and carbon in crown feathers to classify

king eiders as eastern (Atlantic) or western (Pacific) win-

tering birds. In that study, they compared stable isotope

values of winter-grown feathers of birds from known

wintering locations with values in the same set of feathers

of breeding birds at our study area. They were able to

predict with greater than 90% probability the wintering

location of 71% of the breeding birds sampled in 2001 and

87% in 2002, indicating a high degree of precision.

We included wintering locations of individual birds

previously classified by Mehl et al. (2004) together with

variables for body condition, timing of nesting, and date

of blood sampling in general linear models to explain

sources of variability in trace element concentrations

(PROC GLM; SAS Institute Inc. 1999). Variables were

considered either alone or in combination with and

without two-way interactions. A maximum of one time or

date variable and one body condition variable with or

without wintering location was included in each model. In

all, 55 candidate models were considered for each trace

element and year. We used the Akaike Information Cri-

terion (AICc) adjusted for small sample size (Burnham

and Anderson 2002) to select the best models in each set

of analyses. An intercept-only (i.e., null model) was

included in each set as a means of evaluating whether any

of our ecological models was an improvement over one in

which there was no demonstrated effect on trace element

levels. Because wintering locations were determined only

for birds sampled in 2001 and 2002, we limited these

analyses to those years.

Results

A complete set of data was available for 63 birds in 2001

and 74 birds in 2003. In 2002, cadmium levels were

available for 78 birds and lead, selenium, and mercury

levels were available for 69 birds. Forty birds were sam-

pled in 2 years and five birds were sampled in 3 years. All

such samples were analyzed for cadmium, but only 38 and

4 samples from birds sampled in 2 and 3 years, respec-

tively, were analyzed for other trace elements.

Geometric means of cadmium, lead, selenium, and

mercury levels ranged from 5.9 to 12.1 ng/mL, from 0.011

to 0.021, from 3.6 to 4.6, and from 0.13 and 0.18 lg/g wet

weight, respectively (Table 1). Estimates of repeatability

(±1 SE) for cadmium, lead, selenium, and mercury were

0.45 (0.11), 0.35 (0.08), 0.58 (0.10), and 0.25 (0.14),

respectively. The first three estimates differed from zero

(p \ 0.05) and the last approached significance

(0.05 \ p \ 0.10), suggesting a relatively high degree of

among-year repeatability in, and thus exposure to, trace

element levels within the same individuals.

In both 2001 and 2002, wintering location was included

in all of the top models that explained variation in cad-

mium and selenium levels (Table 2). Cadmium levels were

higher and selenium levels were lower in eastern-wintering

birds than in western-wintering birds (Fig. 1). Wintering

location was unimportant to mercury and lead levels

(Table 2).

In general, body condition, nest initiation date, incuba-

tion day, and capture date did not consistently contribute to

variation in trace element levels (Table 2). However, in

2002, variation in lead levels was best explained by body

condition (b = 0.0011, SE = 0.0002) and nest initiation

date (b = -0.0145, SE = 0.0046), which together

explained 24% of the variation in lead levels (Fig. 2a). One

of the two measures of body condition was included in all

of the top five models, whereas nest initiation date was

present in two of the top five models explaining lead levels

in 2002 (Table 2).

The best models of cadmium levels in 2002 included not

only wintering area but also incubation day (partial

R2 = 0.04), nest initiation date (partial R2 = 0.02), or body

condition adjusted for body size, nest initiation date, and

incubation day (partial R2 = 0.01) (Table 2). However, of

the three latter variables, only incubation date (b = 0.0142,

SE = 0.0067) appeared to be related to cadmium levels

Table 1 Geometric means and 95% confidence intervals for trace

element levels in blood samples from nesting, female king eiders at

Karrak and Adventure Lakes, Queen Maud Gulf Bird Sanctuary,

Nunavut, 2001–2003

Year Cadmium Lead Selenium Mercury

2001 12.1 0.011 3.6 0.13

(63) 10.3–14.3 0.009–0.013 3.1–4.3 0.12–0.14

2002 5.9 0.021 4.6 0.18

(69)a 5.2–6.8 0.019–0.024 4.0–5.3 0.17–0.19

2003 7.2 0.021 4.2 0.17

(74) 6.3–8.2 0.019–0.023 3.8–4.7 0.16–0.18

Note: Units are lg/g wet weight except for cadmium, which is ng/ml.

Sample sizes are shown below the yeara Except for cadmium, for which n = 78

Arch Environ Contam Toxicol (2008) 55:683–690 685

123

(Fig. 2b) and the relationship was weak, as indicated by the

low partial R2 value.

The best models of selenium levels in 2002 included

not only wintering area but also nest initiation date

(partial R2 = 0.02), incubation day (partial R2 = 0.01), or

a combination of incubation day (partial R2 = 0.02) and

interaction between wintering area and incubation day

(partial R2 = 0.06). Of these variables, only the interac-

tion term (b = -0.0231, SE = 0.0114) provided an

alternative explanation to variation in selenium levels

compared to wintering area alone. The interaction term

indicated that the difference in selenium levels between

western- and eastern-wintering birds increased with

incubation day. However, the relative support for the

model that included the interaction term was lower than

for the model that included wintering area alone, as

indicated by the higher AICc weight of the latter model

(Table 2).

Discussion

Trace element concentrations observed in these birds

appeared to have no clear influence on their survival or

breeding probability (Wayland et al. in press). In all sam-

ples, lead levels were well below thresholds for adverse

effects in birds (Pain 1996). In 92% of the samples, sele-

nium levels were \9 lg/g wet weight (Wayland et al. in

press), which was the mean selenium level in blood of

mallards that lost body mass and feathers following

experimental exposure to a diet high in selenium (O’Toole

and Raisbeck 1997). It is unclear whether cadmium or

Table 2 Akaike’s Information Criterion (AICc) values for candidate

models explaining trace element levels in blood samples from king

eider females at Karrak and Adventure Lakes, Nunavut in 2001

(n = 63) and 2002 (n = 69, except Cd, n = 78)

Metal Year Model AICc DAICc AICc

Weight

R2

Cd 2001 Win -71.01 0.0000 0.2123 0.22

2002 Win -93.80 0.0000 0.1543 0.26

Win ID -93.54 0.2612 0.1354 0.3

Win NID -92.29 1.5074 0.0726 0.28

Win BC -91.80 1.9943 0.0569 0.27

Pb 2001 B0 -71.90 0.0000 0.1252 0

2002 NID BCNI -89.26 0.0000 0.1269 0.24

ID BCNI -88.78 0.4748 0.1001 0.22

NID BC -88.69 0.5661 0.0956 0.22

BC -88.24 1.0209 0.0761 0.15

BCNI -87.48 1.7813 0.0521 0.13

Se 2001 Win -73.23 0.0000 0.2218 0.28

Win CD -71.60 1.6288 0.0982 0.3

2002 Win -99.01 0.0000 0.1637 0.58

Win NID -98.07 0.9444 0.1021 0.6

Win ID -97.65 1.3613 0.0829 0.59

Win ID Win*ID -97.15 1.8586 0.0646 0.62

Hg 2001 B0 -101.64 0.0000 0.1268 0

2002 B0 -133.37 0.0000 0.1438 0

Note: Models were based on regressions and analyses of covariance

with log-transformed trace element concentrations as dependent

variables and wintering location (Win), incubation day at capture

(ID), Julian capture date (CD), Julian nest initiation date (NID), body

condition adjusted for body size, incubation day, and nest initiation

date (BC), and body condition adjusted for body size (BCNI) as

independent variables. Intercept-only models (b0) were also included

in analyses. AICc weight indicates relative support of each model

given the model set and sums to one. Only DAICc\ 2.0 are shown or

in cases in which the intercept-only model was the best model, only

the results for that model are shown

22

41

21

57

2219

41

50

2001

Cad

miu

m (

ng

·ml-1

, wet

wt)

S

elen

ium

(µg

·g-1

, wet

wt)

0

5

10

15

20

25

East West

Wintering Location

0

2

4

6

8

22

41

21

57

22

41

19

50

2002

2001 2002

Fig. 1 Geometric mean (95% confidence interval) cadmium (top)

and selenium (bottom) levels in blood of nesting, king eider females

at Karrak and Adventure Lakes, Nunavut, 2001–2002, according to

their probable wintering locations (East: Atlantic Ocean, along the

southwest coast of Greenland; West: Bering Sea). Probable wintering

locations were based on classifications of birds previously classified

by Mehl et al. (2004) using stable isotope values of winter-grown

crown feathers. Numbers above bars are sample sizes

686 Arch Environ Contam Toxicol (2008) 55:683–690

123

mercury levels were high enough to pose a risk to any of

these birds, as threshold effects levels for these trace ele-

ments have yet to be established for bird blood.

Our analyses and part of the discussion below reflects

our contention that trace element levels in these birds

might have been affected by changes in body condition and

timing of breeding rather than the opposite: that body

condition and timing of breeding were affected by expo-

sure to trace elements. Our contention is based, in part, on

the relatively low levels of these trace elements in most of

the birds. Moreover, Wayland et al. (2005) showed that

changes in body condition in breeding common eiders were

likely the cause and not the effect of tissue trace element

level changes and we have assumed that such was also the

case for these king eiders.

We are aware of only one other study in which blood

trace element levels have been measured in nesting king

eiders (Wilson et al. 2004). In that study, conducted at the

Prudoe Bay Oil Field in northern Alaska, cadmium was

detected in only one of four sampled birds. The concen-

tration in that bird (970 ng/mL) was *100 times higher

than concentrations recorded in this study. Different

methods used for cadmium analysis might have contributed

to the large difference in cadmium levels recorded in each

study. Our study used Zeeman background correction,

whereas there is no indication that such an approach was

used by Wilson et al. (2004). Information about cadmium

emissions from industrial activity at Prudoe Bay is cur-

rently lacking. Without such information, it is difficult to

attribute apparent differences in cadmium between these

two sites to industrial activity at Prudoe Bay. Selenium

concentrations in blood of nesting king eiders in northern

Alaska (mean ± 1SE = 8.7 ± 0.9 lg/g wet weight; Wil-

son et al. 2004) were higher than those in western-

wintering birds in the present study. Several researchers

have postulated that selenium in Alaskan-breeding water-

fowl wintering in the Bering Sea is derived mainly from

their wintering area and that selenium levels decline during

the breeding season due to depuration with increasing time

and distance to their breeding grounds (Franson et al. 1999;

Grand et al. 2002; Wilson et al. 2004). If true, it would

explain why levels of selenium in king eiders from north-

ern Alaska, where sampling occurred 1–2 weeks earlier

than in our study, were higher than in our birds. Lead was

detected in samples from only two of the four birds in the

Alaska study. The mean value (0.13 lg/g) was slightly

higher than that found in our study (means: 0.011–

0.021 lg/g). Mercury levels (0.3 ± 0.02 lg/g) were also

slightly higher in nesting king eiders sampled by Wilson

et al. (2004).

Cadmium levels in king eiders in the present study were

lower than those measured in spectacled eiders (Somateria

fischeri) in Alaska (means: 100–270 ng/g; Grand et al.

2002; means: 20–40 ng/g; Wilson et al. 2004). As with

Wilson et al. (2004), Grand et al. (2002) apparently did not

use Zeeman background correction, perhaps contributing to

differences in levels between studies. Cadmium levels

measured in our study were slightly higher than those in

prenesting and nesting common eiders (Somateria mol-

lissima) in the eastern Canadian Arctic (mean: 4.9 ng/mL;

Wayland et al. 2001b). In the eastern Canadian Arctic, king

eiders had higher levels of cadmium in their kidneys than

common eiders, perhaps due to different feeding habits

(Wayland et al. 2001a) (e.g., common eiders are mussel

specialists, whereas king eiders have a more varied diet

consisting of mussels, echinoderms, and other benthic

invertebrates). Higher levels of cadmium in blood of king

eiders in the present study, compared to those in common

Body Condition-300 -200 -100 0 100 200 300 400 500

( daeLµ

g·g1-

DIN rof detsujda ,)

0.005

0.01

0.025

0.05

0.1

0.25

0.5

Incubation Day8 10 12 14 16 18 20 22 24 26 28

lm·gn( d

C1-

noitacol retniw rof detsujda )

1

5

10

20

50

(a)

(b)

Fig. 2 (a) Lead levels (lg/g) in relation to body condition and (b)cadmium levels (ng/mL) in relation to incubation day in blood

samples from nesting, female king eiders at Karrak and Adventure

Lakes, Nunavut, in 2002. Lead levels were adjusted to a common nest

initiation date. Body condition was the residual of body mass after

adjusting for body size, the latter calculated as the PC1 score of six

body size measurements (see Materials and Methods section).

Cadmium levels were adjusted for wintering location

Arch Environ Contam Toxicol (2008) 55:683–690 687

123

eiders (Wayland et al. 2001b), might reflect a difference

between the species in cadmium exposure and bioaccu-

mulation. In addition, cadmium levels in common eiders

were higher during the nesting period than the prenesting

period, a likely consequence of weight-loss-associated re-

equilibration in cadmium levels between blood and organs

(Wayland et al. 2001b). Therefore, greater cadmium levels

in king eiders from our study area might have been par-

tially attributable to the fact that blood was sampled

primarily during late incubation, whereas both prenesting

and incubating birds were sampled by Wayland et al.

(2001b).

Lead levels in our king eider samples were low. The

maximum level was 0.16 lg/g wet weight, which is lower

than the 0.2-lg/g level diagnostic of exposure to lead shot

(Pain 1996). Our study area is remote and has no waterfowl

hunting, and so there is likely no spent lead shot in the area.

In contrast, lead shot was used historically to hunt water-

fowl in the Yukon-Kuskokwim Delta in Alaska, and a

significant proportion of breeding spectacled eiders there

had elevated lead levels consistent with exposure to lead

shot (Franson et al. 1998; Grand et al. 2002).

Selenium levels in our king eiders were lower than those

in spectacled eiders from Alaska (Grand et al. 2002; Wil-

son et al. 2004), but were similar to, or for western-

wintering birds, slightly higher than those in common

eiders from the eastern Canadian Arctic (Wayland et al.

2001b) and slightly higher than those in common eiders

from the Baltic Sea (Franson et al. 2000). Alaskan-nesting

eiders likely accumulate much of their selenium from the

selenium-rich food chain in the Bering Sea (Grand et al.

2002; Wilson et al. 2004), whereas those from the eastern

Canadian Arctic and northern Europe presumably are

exposed to lower levels of selenium through their diets.

The similarity in mean selenium levels between eastern-

wintering king eiders in this study and common eiders

reported by Wayland et al. (2001b) can probably be

attributed in part to similarities in wintering ranges.

Common eiders from the study area reported by Wayland

et al. (2001b) and eastern-wintering king eiders from our

study area wintered predominately on the southwest coast

of Greenland or on the Atlantic coast offshore from Lab-

rador, Canada (Mehl et al. 2004; Mosbech et al. 2006).

Mercury levels in our king eider samples were similar to

levels in spectacled eiders in Alaska (Grand et al. 2002;

Wilson et al. 2004) and common eiders in the eastern

Canadian Arctic (Wayland et al. 2001b). However, they

were slightly higher than levels in common eiders in the

Baltic Sea (Franson et al. 2000).

For birds sampled in 2 or more years, repeatability was

highest for selenium, followed in order by cadmium, lead,

and mercury. Levels of the first three trace elements were

significantly repeatable in individuals from one year to the

next, whereas mercury levels were not significantly

repeatable. These results agree with previous research that

examined year-to-year correlations in trace element levels

in repeatedly sampled king eiders and white-winged scot-

ers (Melanitta fusca) (Wayland et al. 2007).

Selenium levels were higher and cadmium levels were

lower in western-wintering birds than in eastern-wintering

birds, suggesting that exposure to selenium and cadmium

during the winter and early spring carried over into the

breeding season. These differences were consistent with

other studies that showed higher levels of cadmium and

lower levels of selenium in marine life from the eastern

Canadian Arctic and the Greenland coast than in marine

life from the western Arctic (Braune et al. 1991; Dietz et al.

1996; Muir et al. 1997; Stoudt et al. 2002; Trust et al. 2000;

Wayland et al. 2001a). The half-life of cadmium in bird

blood has not been measured but is 1–2 months in human

blood (Olsson et al. 2005). This relatively long half-life is

consistent with a cross-seasonal effect of winter cadmium

exposure on summer cadmium levels. Similarly Gonzalez-

Solıs et al. 2002, reported cross-seasonal effects of cad-

mium exposure during winter in spring-breeding giant

petrels (Macronectes spp.) In contrast, the half-life of

selenium in blood of experimental mallards (Anas platy-

rhynchos) was only 10 days (Heinz et al. 1990). However,

in wild spectacled eiders a quasi-half-life of 37 days was

reported (some ingestion of selenium likely occurred dur-

ing the sampling interval), supporting the likelihood of

cross-seasonal relationships (Grand et al. 2002). Cross-

seasonal effects of selenium exposure during winter on

selenium levels in breeding spectacled eiders and emperor

geese (Chen canagica) have been reported (Franson et al.

1999; Grand et al. 2002; Wilson et al. 2004), in agreement

with results from our study. All such inferences about

cross-seasonal relationships might have implications for

migratory birds that winter in polluted areas and return to

apparently pristine areas to breed. For example, selenium

levels in blood and eggs of common eiders were highly

correlated (Franson et al. 2000). It stands to reason that

exposure of wintering birds to excessive selenium in pol-

luted areas could be carried back to breeding areas where

selenium in blood would be deposited in eggs, which is

potentially important because selenium is particularly toxic

to avian embryos (Hoffman 2002; Ohlendorf 2003).

We found little evidence that trace element levels were

related to nest initiation date, body condition, incubation

day, or sample collection date. Instead, winter provenance

had a much greater influence on cadmium and selenium in

king eider blood than these potential factors. Nevertheless,

lead levels tended to be higher in birds that nested early

and were in good body condition in 2002. We cannot

explain this relationship. Perhaps birds that nested early

and were in good body condition had longer incubation

688 Arch Environ Contam Toxicol (2008) 55:683–690

123

sessions (fewer or shorter breaks) than other birds, thus

burning more of their stored nutrients and hence recircu-

lating more lead from medullary bone into their

bloodstreams (Wilson et al. 2007). Also in 2002, cadmium

levels appeared to increase during incubation. Consistent

with this result, Wayland et al. (2001b) found that cad-

mium levels in blood samples from nesting common eiders

were higher than those in samples from prenesting birds, a

likely consequence of weight-loss-induced reequilibration

between cadmium levels in blood and organs such as liver

and kidney.

We suggest that individual variation in levels of cad-

mium and selenium of king eiders breeding in the central

Canadian Arctic likely arises from differential exposure to

those trace elements in eastern- and western-wintering

areas. The high interyear repeatability of cadmium and

selenium levels among individuals likely occurred because

most birds probably use the same wintering location from

one year to the next.

In conclusion, our results show that trace element

analysis can be effectively combined with isotope analysis

of feathers and with local banding efforts to support cross-

seasonal relationships in trace element levels. Such data

allow for broader investigation of geographical locations,

especially important where trace elements might act to

suppress populations. Specifically, combining such meth-

ods might offer greater insight into the effects of trace

elements on adult survival and annual productivity. Finally,

we hope that this study will encourage future work, espe-

cially long-term research, to incorporate methods that

allow for comparison of trace elements levels among years

and geographic areas.

Acknowledgments We thank J. Conkin, K. Drake, P. Dunlop, K.

Hobson, S. Lawson, A. Leblanc, R. McNeil, E. Neugebauer, C,

Swoboda, K. Timm, J. Traylor, and C. Wood for their assistance.

Funding was provided by Environment Canada, in part through its

Northern Ecosystem Initiative, and by Polar Continental Shelf Pro-

ject, Ducks Unlimited Canada’s Institute of Wetlands and Waterfowl

Research, University of Saskatchewan, Nunavut Wildlife Manage-

ment Board, Delta Waterfowl Foundation, and the Arctic Institute of

North America.

References

Batt BDJ, Prince HH (1979) Laying dates, clutch size and egg weight

of captive mallards. Condor 81:35–41

Becker WA (1984) Manual of quantitative genetics, 5th ed. Academic

Enterprises, Pullman, WA

Braune BM, Norstrom RJ, Wong MP, Collins BT, Lee J (1991)

Geographical distribution of metals in livers of polar bears from

the Northwest Territories Canada. Sci Total Environ 100:

283–299

Burnham KP, Anderson DR (2002) Model selection and multimodel

inference: a practical information theoretic approach, 2nd ed.

Springer-Verlag, New York

Dietz R, Riget F, Johansen P (1996) Lead, cadmium, mercury and

selenium in Greenland marine animals. Sci Total Environ

186:67–93

Franson JC, Hollmen T, Poppenga RH, Hario M, Kilpi M, Smith

MR (2000) Selected trace elements and organochlorines: some

findings in blood and eggs of nesting common eiders (Soma-teria mollissima) from Finland. Environ Toxicol Chem

19:1340–1347

Franson JC, Petersen MR, Creekmore LH, Flint PL, Smith MR (1998)

Blood lead concentrations of spectacled eiders near the Kashu-

nuk river, Yukon Delta National Wildlife Refuge, Alaska.

Ecotoxicology 7:175–181

Franson JC, Schmutz JA, Creekmore LH, Fowler AC (1999)

Concentrations of selenium, mercury, and lead in blood of

emperor geese in western Alaska. Environ Toxicol Chem

18:965–969

Gonzalez-Solıs J, Sanpera C, Ruiz X (2002) Metals and selenium as

bioindicators of geographic and trophic segregation in giant

petrels Macronectes spp. Marine Ecol Prog Ser 244:257–264

Grand JB, Franson JC, Flint PL, Petersen MR (2002) Concentrations

of trace elements in eggs and blood of spectacled and common

eiders on the Yukon-Kuskokwim Delta, Alaska, USA. Environ

Toxicol Chem 21:1673–1678

Heinz GH, Pendleton GW, Krynitsky AJ, Gold LG (1990) Selenium

accumulation and elimination in mallards. Arch Environ Contam

Toxicol 19:374–379

Hoffman DJ (2002) Role of selenium toxicity and oxidative stress in

aquatic birds. Aquat Toxicol 57:11–26

Kellett DK, Alisauskas RT (1997) Breeding biology of king eiders

nesting on Karrak Lake, Northwest Territories. Arctic 50:47–54

Kellett DK, Alisauskas RT (2000) Body-mass dynamics of king

eiders during incubation. Auk 117:812–817

Lessells CM, Boag PT (1987) Unrepeatable repeatabilities: a common

mistake. Auk 104:116–121

Mehl KR, Alisauskas RT, Hobson KA, Kellett DK (2004) To winter

east or west? Heterogeneity in winter philopatry in a central-

arctic population of king eiders. Condor 106:241–251

Mehl KR, Alisauskas RT, Hobson KA, Merkel FR (2005) Linking

breeding and wintering areas of king eiders: making use of polar

isotopic gradients. J Wildl Manag 69:1297–1304

Møller AP (2001) Heritability of arrival date in a migratory bird. Proc

R Soc B 268:203–206

Mosbech A, Gilchrist G, Merkel F, Sonne C, Flagstad A, Nyegaard H

(2006) Year-round movements of northern common eiders,

Somateria mollissima, breeding in Arctic Canada and west

Greenland followed by satellite telemetry. Ardea 94:651–665

Muir D, Braune B, DeMarch B et al (1997) Ecosystem uptake and

effects. In Jensen J, Adare K, Shearer R (eds) Canadian Arctic

Contaminants assessment report. Indian and Northern Affairs

Canada, Ottawa, pp 191–294

Neugebauer E, Sans Cartier GL, Wakeford BJ (2003) Methods for the

determination of metals in wildlife tissues using various atomic

absorption spectrophotometry techniques. Technical Report

Series No. 337E. Canadian Wildlife Service Headquarters,

Ottawa

Ohlendorf HM (2003) Ecotoxicology of selenium. In: Hoffman DJ,

Rattner BA, Burton GA Jr, Cairns J Jr (eds) Handbook of

ecotoxicology. Lewis Publishers, Boca Raton, FL pp 465–500

Olsson I-M, Eriksson J, Oborn I, Skerfving S, Oskarsson A (2005)

Cadmium in food production systems: a health risk for sensitive

population groups. AMBIO 34:344–351

O’Toole D, Raisbeck MF (1997) Experimentally induced selenosis of

adult mallard ducks: clinical signs, lesions, and toxicology. Vet

Pathol 34:330 –340

Pain DJ (1996) Lead in waterfowl. In: Beyer WN, Heinz GH,

Redmon-Norwood AW (eds) Environmental contaminants in

Arch Environ Contam Toxicol (2008) 55:683–690 689

123

wildlife: Interpreting tissue concentrations. Lewis Publishers,

Boca Raton, FL, pp 251–264

Peakall D, Burger J (2003) Methodologies for assessing exposure to

metals: speciation, bioavailability of metals, and ecological host

factors. Ecotoxicol Environ Safety 56:110–121

Petersen MR (1992) Reproductive ecology of emperor geese: annual

and individual variation in nesting. Condor 94:383–397

SAS Institute Inc. (1999) SAS/STAT user’s guide, version 8, volume

2. SAS Institute Inc., Cary, NC

Sea Duck Joint Venture Management Board (2001) Sea Duck Joint

Venture strategic plan 2001–2006. SDJV Continental Team,

Canadian Wildlife Service and United States Fish and Wildlife

Service, Sackville, New Brunswick and Anchorage, Alaska

Stoeppler M , Brandt K (1980) Contributions to automated trace

analysis. V. Determination of Cd in whole blood and urine by

electrothermal AAS. Fresenius Z Anal Chem 300:372–380

Stoudt JH, Trust KA, Cochrane JF, Suydam RS, Quakenbush LT

(2002) Environmental contaminants in four eider species from

Alaska and Arctic Russia. Environ Pollut 119:215–226

Suydam RS (2000) King eider (Somateria spectablis). In: Poole A,

Gill F (eds) The birds of North America No. 491. The Birds of

North America, Inc., Philadelphia

Trust KA, Rummel KT, Scheuhammer AM, Brisbin IL, Hooper MJ

(2000) Contaminant exposure and biomarker responses in

spectacled elders (Somateria fischeri) from St. Lawrence Island,

Alaska. Arch Environ Contam Toxicol 38:107–113

Wayland M, Alisauskas RT, Kellett DK et al (2007) Year-to-year

correlations in blood metal levels in two species of North

American sea ducks. Environ Pollut 150:329–337

Wayland M, Drake KL, Alisauskas RT et al (in press) Survival rates

and blood metal concentrations in two species of free-ranging

North American sea ducks. Environ Toxicol Chem

Wayland M, Garcia-Fernandez AJ, Neugebauer E, Gilchrist HG

(2001b) Concentrations of cadmium, mercury and selenium in

blood, liver and kidney of common eider ducks from the

Canadian Arctic. Environ Monitor Assess 71:255–267

Wayland M, Gilchrist HG, Dickson DL et al (2001a) Trace elements

in king and common eiders in the Canadian Arctic. Arch Environ

Contam Toxicol 41:491–500

Wayland M, Gilchrist HG, Neugebauer E (2005) Concentrations of

cadmium, mercury and selenium in common eider ducks in the

eastern Canadian Arctic: Influence of reproductive stage. Sci

Total Environ 351:323–332

Weller MW (1956) A simple field candler for waterfowl eggs. J Wildl

Manag 20:111–113

Wilson HM, Flint PL, Powell AN (2007) Coupling contaminants with

demography: effects of lead and selenium in Pacific common

eiders. Environ Toxicol Chem 26:1410–1417

Wilson HM, Petersen MR, Troy D (2004) Concentrations of metals

and trace elements in blood of spectacled and king eiders in

northern Alaska, USA. Environ Toxicol Chem 23: 408–414

690 Arch Environ Contam Toxicol (2008) 55:683–690

123