Environ. Sci. Technol. 2009, 43, 2124–2130
Environmentally Relevant Concentrations of DE-71 and HBCD Alter Eggshell Thickness and Reproductive Success of American Kestrels K I M J . F E R N I E , * ,†,‡ J . L A I R D S H U T T , § ROBERT J. LETCHER,§ IAN J. RITCHIE,‡ AND DAVID M. BIRD‡ Canadian Wildlife Service, Environment Canada, 867 Lakeshore Road, Burlington, Ontario, Canada L7R 4A6, Avian Science and Conservation Centre, McGill University, 21111 Lakeshore Road, Ste Anne de Bellevue, Quebec, Canada H9X 3V9, and Wildlife and Landscape Science Directorate, Science and Technology Branch, Environment Canada, Carleton University, National Wildlife Research Centre, 1125 Colonel By Drive, Ottawa, Ontario, Canada K1S 5B6
Received September 26, 2008. Revised manuscript received December 22, 2008. Accepted January 16, 2009.
Polybrominated diphenyl ethers (PBDEs) and total R-hexabromocyclododecane (HBCD) are flame-retardant additives that are commonly used in household and commercial applications. PBDE congeners, which are comprised of technical mixtures such as DE-71, are globally persistent and their concentrations are increasing in many species. Captive American kestrels (Falco sparverius) were exposed by diet to vehicle (safflower oil), or one of two environmentally relevant concentrations of DE71 and unintentionally to HBCD. This exposure resulted in the birds laying eggs that contain PBDE and HBCD concentrations currently found in wild herring gull (Larus argentatus) and peregrine falcon (F. peregrinus) eggs, and compared to control kestrels, resulted in delayed egg laying and smaller eggs being laid, caused thinner eggshells and differential weight loss during embryonic development, and reduced fertility and reproductive success. The thickness of the eggshell declined as the concentrations of all measured PBDE and the total amount of R-HBCD congeners (except BDE-183 and BDE-209) increased; increasing concentrations of BDE-153, BDE-154, BDE28, BDE-17, delayed egg laying, reduced eggshell mass (plus ΣPBDEs), and reduced fledging success (BDE-153 and BDE154 only). BDE-153 is the dominant congener recently found in peregrine eggs. The results of this study are consistent with the PBDE-associated brood reduction in wild European peregrines and may partially explain the decline of kestrels in North America.
* Author to whom correspondence should be addressed. Tel.: 1-905-336-4843. Fax: 1-905-336-6434. E-mail:
[email protected]. † Canadian Wildlife Service, Environment Canada. ‡ Avian Science and Conservation Centre, McGill University. § Wildlife and Landscape Science Directorate, Science and Technology Branch, Environment Canada, Carleton University, National Wildlife Research Centre. 2124
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 6, 2009
Introduction Polybrominated diphenyl ethers (PBDEs) and hexabromocyclododecane (HBCD) are flame-retardant additives used widely in commercial polymeric materials (e.g., clothing, carpet, electronic appliances, dust) (1, 2). Tetrabromo- to octabromodiphenyl (OctaBDE) ether congeners comprise a large proportion of pentabromodiphenyl (PentaBDE) and OctaBDE technical mixtures, and these congeners, as well as HBCD, are highly persistent and susceptible to transport (1, 2). Consequently, these chemicals are ubiquitous in the environment and biota, especially in species that are at the top of the food chain (2-4). PBDEs and HBCD are found in human breast milk, fat, and blood (2, 5), and elevated PBDE concentrations in women and children resulted in the recent banning of PentaBDE, OctaBDE, and decabromodiphenyl ether (Deca-BDE) technical mixtures by the European Union (EU) (6, 7), among others. PBDEs and HBCD have been detected in predatory birds at the top of the food chain (see, e.g., refs 8-11), with some of the highest PBDE concentrations ever recorded found recently in common kestrels (Falco tinnunculus) of northern China (10). Mean ΣPBDE14 and HBCD concentrations in Swedish peregrine falcons (Falco peregrinus) were 4000-4700 ng/g lipid weight (lw) and 210-270 ng/g lw, respectively, between 1987 and 1999 (9, 11). Over the past 15 years, ΣPBDE concentrations in peregrine eggs have ranged from 74.5 ng/g ww to 6610 ng/g ww (ΣPBDE30) in the northeastern United States (8), and from 8 ng/g ww to 1600 ng/g ww (ΣPBDE15) in central and eastern Canada (12). While the ΣPBDE concentrations found in the U.S. peregrine eggs are generally higher than those found in European peregrine eggs, BDE-153 is the dominant congener in these eggs, regardless of location (8, 9, 11). The neonatal exposure of laboratory animals to some PBDE congeners at levels exceeding current environmental concentrations has had adverse neurobehavioral effects (see, e.g., ref 13), while developmental exposure to 2,2′,4,4′,5pentabromodiphenyl ether (PBDE-99) has adversely affected circulating sex steroids and sperm production (see, e.g., refs 14, 15). In American kestrels, environmentally relevant concentrations of PBDEs negatively impacted the timing and frequency of courtship behaviors essential to successful reproduction (16). In southern Sweden, ΣPBDE concentrations in the eggs of the once critically endangered peregrine falcon are negatively correlated with the average number of nestlings per female (brood size) (11). Like the common kestrel and the peregrine falcon, American kestrels are members of the genus Falco; peregrines are listed as a “species at risk” in Canada, and American kestrel numbers are declining across North America (17-19). In this study, captive American kestrels were exposed to environmentally relevant concentrations of the PentaBDE technical formulation (DE71) and unintentionally exposed to HBCD. DE-71 is one of three commercially available PBDE flame-retardant products and its production by the Great Lakes Chemical Company (now Chemtura) was ceased in 2004. However, concentrations of its congeners remain in tissues of biota, including birds’ eggs (see, e.g., ref 20). The objectives of this research are to identify potential changes in avian reproductive success and egg quality associated with exposure to environmentally relevant concentrations of DE-71.
Materials and Methods This study was conducted at the Avian Science and Conservation Centre, McGill University (Montreal, Quebec), using captive American kestrels of known age and pedigree. Male 10.1021/es8027346 CCC: $40.75
2009 American Chemical Society
Published on Web 02/11/2009
and female adult kestrels were randomly assigned to one of three groups: the high DE-71 exposure group (1.6 ppm), the low DE-71 exposure group (0.3 ppm), or the control group. Each group of kestrels was fed their regular diet of frozen/ thawed day-old cockerels ad libitum. Exposure and Dosing Concentrations of DE-71. Because the PBDE levels in wild American kestrel eggs were unknown when this study began, dosage concentrations were calculated based on the concentrations of PBDE congener residues found in herring gull eggs collected in the Great Lakes Basin in 2000 or 2004 (20, 21). The calculated dosage levels resulted in the kestrels laying eggs with current, environmentally relevant residue concentrations in eggs of wild birds (8, 9, 11, 12, 20, 21) including wild kestrels (according to unpublished data supplied by P. Martin, Environment Canada). Kestrels were exposed by diet to DE-71 at concentrations exceeding the background concentrations (represented by control birds) for ∼75 d each year. Because female kestrels gain almost one-third of their body weight in the three weeks preceding egg laying (22), exposure to the DE71 began three weeks prior to pairing and continued through courtship, egg laying, and incubation, until the first chick (would have) hatched. The DE-71 mixture was obtained from the Great Lakes Chemical Company. Details about the preparation of the DE-71 with safflower oil are provided in Fernie et al. 16. The dosing mixtures were injected daily into the brains of the frozen/thawed cockerel diet, and during the exposure period, each pair of kestrels received three cockerels per day. The estimated daily dietary exposure concentrations consisted of 1.6 ppm (0.65 µg DE-71/µL safflower oil per cockerel) for the high-exposure birds and 0.3 ppm (0.12 µg/µL) for the low-exposure birds based on a mean cockerel weight of 40.4 ( 0.26 g. Kestrels assigned to the control group received safflower oil only. These exposure concentrations may be underestimated for 2006 from the 2005 exposure regime, because PBDE half-lives in adult American kestrels are expected to be in the range of 72-572 d (23). The first egg produced by each kestrel pair was collected and used to assess eggshell quality (weight, thickness) and PBDE and total-R-HBCD burdens. The contaminant burdens in the first egg of each kestrel clutch may not reflect concentrations in other eggs within the same clutch, although contaminant burdens, including PBDEs, did not vary with laying order in other species (24, 25). Eggshell thickness was determined by taking the mean of five measurements around the equator of the egg and included the air-dried eggshell membranes (26). Chemical analysis of 14 PBDE congeners (BDE-17, BDE-28, BDE-47, BDE-49, BDE-66, BDE-85, BDE99, BDE-100, BDE-138, BDE-153, BDE-154, BDE-183, BDE190, and BDE-209), and total-R-HBCD (representing the sum of R-, β-, and γ-HBCD isomers) of the eggs was according to that described in Gauthier et al. (20, 27). Total R-HBCD concentrations were only measured in 2006. The sum of the concentrations of the seven major congeners (BDE-28, BDE47, BDE-100, BDE-99, BDE-154, BDE-153, and BDE-183) constituted >94% of the total ΣPBDE concentrations in both the low and high exposure eggs 16. In addition, a sample of the DE-71 dosing mixture was analyzed for 11 brominated dioxin and furan congeners by Axys Analytical Services (Sidney, British Columbia, Canada) (16). Animal Husbandry. The kestrels were paired on April 21, 2005 (n ) 31 pairs) and April 7, 2006 (n ) 31 pairs). Each bird had previous breeding experience and was paired with another bird that was genetically unrelated within the past six generations. In 2006, the birds remained in their same treatment groups but were paired with a different bird than in 2005. During the experiment, the kestrels were housed as a pair in their own breeding pen containing a nest box, rope perches, and a food platform. The treatment and care of the
kestrels was conducted in accordance with the Canadian Council on Animal Care guidelines (28). Reproductive Measures. In this study, the time from pairing to when egg laying began (time to lay), the length of time required for laying a clutch of five eggs, and clutch size (number of eggs laid by each pair) were assessed. Other key reproductive parameters included overall hatching and reproductive success (the proportion of eggs that produced hatchlings or fledglings, respectively), fertility (total number fertile eggs per pair; the proportion of laid eggs that were fertile), hatching success (total number per pair; the proportion of fertile eggs that hatched), and fledging success (total number per pair; the proportion of hatchlings that fledged). Egg quality was measured by egg mass when first laid (fresh mass) and 14 d later at mid-incubation (2006 only), egg volume (29), and shell mass and thickness of the first egg laid in each clutch and used for contaminant analysis. Statistical Analysis. Only first-laid clutches were used in the statistical analysis. Data were initially tested for normality, and when appropriate, transformed prior to statistical analysis. There were no statistical differences between years, so data were combined. For the eggshell measurements, one control pair was a statistical outlier (beyond two standard deviations) and so were eliminated from related analyses. One-way analyses of variance (ANOVAs) were used to determine potential treatment effects on laying date, time to lay, and eggshell variables. Effects on reproductive success were assessed by completing logit transformations of reproductive proportions for each pair, then in a stepwise manner, using a Brown-Forsythe test for trends adapted as a robust test 30 to identify the lowest dose for which a significant trend was detected. ANOVAs, with clutch as a nested variable and a Julian lay date as a covariate, and leastsquares mean posthoc (LSM P) tests adjusted for the covariate, were used to determine differences in nonproportional measures of reproductive success, clutch size, and egg qualities. To identify biologically important differences among the groups, effect size (ES) analysis (31) was used, by calculating the absolute value of the difference between the means of the two groups, and dividing that result by the result from dividing the square root of one-half of the sum of the squared standard deviations of each group. ES increases as the difference between population means increases, and decreases as the variability in response increases (32). Pearson’s correlation analysis was used to identify significant relationships among reproductive measures and PBDE concentrations. The statistical level of significance is P e 0.05. All data are presented as means ( standard errors about the mean (SEM), although referenced eggshell data in the Discussion section are given in the form of mean ( standard deviation (SD).
Results As recently described in Fernie et al. (16), the mean Σ PBDE and total R-HBCD concentrations in the kestrel eggs were as follows: 3.01 ( 0.46 and 0.002 ( 0.002 ng/g ww, respectively, for the control group; 288.60 ( 33.35 and 3.27 ( 0.68 ng/g ww, respectively, for the low-exposure group; and 1130.59 ( 95.34 and 15.61 ( 2.43 ng/g ww, respectively, for the highexposure group. The sum BDE concentrations were significantly different among the three groups (F2,28 ) 98.32, P < 0.0001); HBCD levels also differed overall (F2,15 ) 21.88, P < 0.0001) being significantly higher in the high-exposure eggs (LSM P < 0.0001) but similar between control and lowexposure eggs. None of the 11 brominated dioxins and furans was detected in the DE-71 mixture that was fed to the kestrels. The exposure to DE-71 had no affect on adult body mass, but did moderately delay egg laying (F2,52 ) 2.50, P ) 0.0923) and resulted in fewer (nonsignificant) pairs laying eggs (see VOL. 43, NO. 6, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2125
TABLE 1. Reproductive Measures of American Kestrels Exposed to Environmentally Relevant Concentrations of DE-71a Control Pairs reproductive variable
mean
total number of pairs percentage of pairs not laying
SEM
21 9.5%
Low-Exposure Pairs
High-Exposure Pairs
mean
mean
SEM
20 10.0%
effect of DE-71 exposure,b df ) 2, 52
SEM
21 14.3%
effect of delayed laying,c df ) 3, 51
association with delayed laying (Pearson’s r; p-value) 55
N.S.
N.A.
4.67
0.35
4.80
0.39
4.38
0.44
P ) 0.0302 high N.A. ES ) 0.35 low N.S. N.S.
reproductive success total number of fertile eggs laidd 3.14 2.52 total number of hatchlingsd total number of fledglingsd 2.33
0.37 0.39 0.39
3.20 2.30 2.20
0.37 0.40 0.39
2.81 1.86 1.62
0.42 0.37 0.37
ES ) 0.18 high ES ) 0.39 high ES ) 0.41 high
0.0049 N.S. 0.0369
r ) -0.39, p ) 0.0037 N.S. r ) -0.31, p ) 0.0193
overall hatching successe overall reproductive successf
0.54 0.50
0.07 0.07
0.46 0.44
0.07 0.07
0.40 0.35
0.07 0.07
P ) 0.038g P ) 0.019g
N.S. 0.037
N.S. r ) -0.31, p ) 0.0176
fertility (%)h hatching success (%)i fledging success (%)j
0.76 0.64 0.70
0.09 0.09 0.09
0.73 0.56 0.67
0.08 0.09 0.10
0.65 0.49 0.56
0.10 0.09 0.11
N.S. P ) 0.020f N.S.
0.0021 N.S. 0.0211
r ) -0.43, p ) 0.0013 N.S. r ) -0.33, p ) 0.0140
time from pairing to egg laying
18.68 d 2.06 d 22.72 d 3.23 d 28.39 d 3.71 d
clutch size (number of eggs laid)
d
N.A. N.S.
a The low-exposure kestrel pairs were exposed by diet to DE-71 concentrations of 0.3 ppm/day and laid eggs with ΣPBDE concentrations of 288.60 ( 33.35 ng/g ww (mean ( SEM) (16); the high-exposure kestrel pairs were exposed by diet to 1.6 ppm/day of DE-71 and laid eggs with ΣPBDE levels of 1130.59 ( 95.34 ng/g ww (mean ( SEM) (16). N.S. ) not significant. N.A. ) not applicable. b Statistical p-values or effect size (ES) values presented. c Julian laying date used as a covariate and statistical p-values presented; d Total number per pair. e The proportion of hatchlings per pair. f The proportion of fledglings per pair. g Comparisons are significant only for control versus high-exposure pairs. h The proportion of eggs laid that were fertile. i The proportion of fertile eggs that hatched. j The proportion of hatchlings which fledged at 28 d of age.
TABLE 2. Associations among Reproductive or Eggshell Measurements of the First-Laid Eggs and DE-71 Congener Concentrations Measured within the Same Eggsa Time Needed for Pair To Lay (d)
PBDE congeners
N-PBDEs N-HBCDd ΣPBDEse BDE-99 BDE-153 BDE-100 BDE-154 BDE-47 HBCD BDE-138 BDE-49 BDE-183 BDE-209 BDE-28 BDE-17
rank
concentration
1 2 3 4 5 6 7 8 9 10 11 12 13
50 26 467 209796 73191 66276 48013 41649 6430 5522 1878 1475 732 295 217
SEM
r-valueb
74 32979 12877 11378 8357 7226 1532 1402 324 385 221 69 72
0.39 0.41
0.28 0.37
p-value 51 28 N.S. N.S. 0.004 N.S. 0.003 N.S. N.S. N.S. N.S. N.S. N.S. 0.04 0.01
Thickness of Eggshell r-valuec
-0.51 -0.43 -0.57 -0.44 -0.58 -0.45 -0.50 -0.42 -0.38 -0.49 -0.40
p-value 42 26 0.001 0.004
