Microsatellite DNA Mutations in Double-Crested Cormorants

Furthermore, Barker(50) argued that the local highway's (Queen Elizabeth Way) ..... and OPAHs and toxicity of PM2.5 during the Beijing Olympic Games E...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/est

Microsatellite DNA Mutations in Double-Crested Cormorants (Phalacrocorax auritus) Associated with Exposure to PAH-Containing Industrial Air Pollution L. E. King,†,∥ S. R. de Solla,*,‡ J. M. Small,§ E. Sverko,§ and J. S. Quinn† †

Department of Biology, McMaster University, Hamilton, Ontario L8S 4K1, Canada Ecotoxicology and Wildlife Health Division, Canada Centre for Inland Waters, Environment Canada, Burlington, Ontario L7R 4A6, Canada § National Laboratory for Environmental Testing (NLET), Canada Centre for Inland Waters, Environment Canada, Burlington, Ontario L7R 4A6, Canada ‡

S Supporting Information *

ABSTRACT: Hamilton Harbour, Ontario, Canada is one of the most polluted sites on the Great Lakes, and is subject to substantial airborne pollution due to emissions from both heavy industry and intense vehicle traffic. Mutagenic Polycyclic aromatic hydrocarbons (PAHs) are present at very high concentrations in the air and sediment of Hamilton Harbour. We used five variable DNA microsatellites to screen for mutations in 97 families of Double-crested Cormorants (Phalacrocorax auritus) from three wild colonies, two in Hamilton Harbour and one in cleaner northeastern Lake Erie. Mutations were identified in all five microsatellites at low frequencies, with the majority of mutations found in chicks from the Hamilton Harbour site closest to industrial sources of PAH contamination. Microsatellite mutation rates were 6-fold higher at the Hamilton Harbour site closest to the industrial sources of PAH contamination than the other Hamilton Harbour site, and both were higher than the reference colony. A Phase I metabolite of the PAH benzo[a]pyrene identified by LC-MS/MS in bile and liver from Hamilton Harbour cormorant chicks suggests that these cormorants are exposed to and metabolizing PAHs, highlighting their potential to have caused the observed mutations.



INTRODUCTION Airborne contamination is often pervasive near urban and industrial sources, and can affect the health of humans and wildlife.1−3 Although there are many gaseous and particulate components in air that can contribute to toxicity, polycyclic aromatic hydrocarbons (PAHs) may be the compounds that contribute most to toxicity, particularly genotoxicity.4 PAHs are ubiquitous byproducts of incomplete combustion of organic materials and are found in air pollution resulting from combustion of fossil fuels or other carbon based materials. PAHs and their metabolites can cause endocrine disruption, alterations in physiology and gene expression, DNA damage, and cancer.5 In environmental settings, PAHs are often found in complexes such as NPAHs (nitrogenated PAHs), increasing their mutagenicity.6 PAHs are metabolized in vertebrates to form reactive diol intermediates, which form adducts with nucleotides in DNA, causing mutations.7 Previous research in Hamilton Harbour, Ontario, Canada, demonstrated that airborne PAHs may be responsible for inducing mutations in mammals and birds. Emissions from industry, including Canada’s two largest integrated steel mills, and diesel and gasoline emissions from nearby highways, are major contributors to the pollution in the Harbour,8,9 including Published 2014 by the American Chemical Society

very high airborne PAH concentrations that exceed provincial health guidelines.10,11 Germline DNA mutation rates were higher at Hamilton Harbour relative to cleaner reference groups in herring gull minisatellites (Larus argentatus)12,13 and in mice ESTRs (expanded simple tandem repeat) (Mus musculus).11,14 In both cases, PAHs associated with particulate matter were suspected to be the cause. Mice breathing only HEPA-filtered air in sites exposed to ambient air pollution showed decreased ESTR mutations similar to background rates, demonstrating that breathable particulate was responsible for mutation induction.11 Polycyclic aromatic hydrocarbons and nitrated derivatives likely are the most important species responsible for the genotoxicity of air from urban and industrial regions.4,15 Ellegren et al.16 reported elevated microsatellite mutations in tree swallows breeding near Chernobyl compared to those from a less contaminated site, which demonstrated the capacity of radioactivity to induce microsatellite mutations. However, the induction of microsatellite mutations by Received: Revised: Accepted: Published: 11637

June 4, 2014 August 15, 2014 August 25, 2014 August 25, 2014 dx.doi.org/10.1021/es502720a | Environ. Sci. Technol. 2014, 48, 11637−11645

Environmental Science & Technology

Article

Figure 1. Cormorant colonies (red triangles) and air monitoring stations (blue circles) in Hamilton Harbour, western Lake Ontario, Canada. Gray area shows the approximate extent of heavy industry. The wind rose shows the predominant wind direction and wind velocities at the HAMN monitoring site (Station 29167, Pier 25) ∼0.6 km from Pier 27 colony. ‘PAH Station’ shows sampling location for airborne PAHs.

agricultural area. We predicted that mutation rate would be highest at the colony closest to the steel mills and the lowest at the reference colony on Lake Erie. Here we report germline length-change microsatellite mutation rates in families of Double-crested Cormorants from two colonies in Hamilton Harbour and one colony on Lake Erie, demonstrating that the microsatellite mutation rate was higher in colonies with elevated exposure to airborne industrial pollution. We also measured hydroxylated PAH metabolites in cormorant tissues collected opportunistically, showing PAH exposure at the different colonies.

nonradioactive environmental contamination has not been reported to date. Double-crested Cormorants (Phalacrocorax auritus) are abundant piscivorous migratory waterbirds that breed in colonies on the Great Lakes, including two colonies in Hamilton Harbour. Cormorants are large-bodied, with a high basal metabolism relative to other birds their size.17 With unidirectional air flow in the lungs and high lung volumes, birds inhale large amounts of particulate matter and are ’valuable experimental models’ for air pollution studies;18 piscivorous birds have been some of the earliest and most important indicators of contamination-related health effects in the environment.19 PAHs are actively biotransformed in most vertebrate animals, first forming phase I metabolites (hydroxylated or hydroxy-PAHs). We hypothesized that elevated atmospheric PAH concentrations in Hamilton Harbour would produce detectable PAH metabolites in cormorants. Measuring PAH metabolites directly (as opposed to using bioassays that assess only the toxicity of PAH exposure) provides a sensitive PAH exposure marker that can effectively track dose.20 In oilcontaminated common guillemots (Uria aalge), the majority of PAH compounds were found as hydroxylated metabolites (2−7 times more metabolites than parent PAHs).21 To our knowledge, no study has yet shown PAH metabolites in birds resulting from chronic ambient environmental exposure, i.e. non-oil spill exposure. To evaluate the effect of pollution exposure on mutation rates in cormorants, we chose three colonies at varying distances from sources of PAH-containing emissions. Within Hamilton Harbour, we sampled at one colony close to (and generally downwind of) coking ovens used in industrial steel production, and another across Hamilton Harbour which is both farther away and less often downwind of steel emissions. Our reference site in Lake Erie was offshore from a rural



MATERIALS AND METHODS

Study Sites and Sampling Procedures. Three similarly sized (each consisting of approximately 1000 breeding pairs) colonies of ground-nesting cormorants on the lower Great Lakes, Ontario, Canada with varying degrees of sedimentary and airborne contamination were chosen for sampling. The study colony at Pier 27 (Eastport), Hamilton Harbour, western Lake Ontario (43° 17′ N, 79° 49′ W; Figure 1) was the closest to a major source of industrial pollution, approximately two kilometres east of two large integrated steel mills, and 350 m west of a 8 lane highway. Prevailing winds in the Harbour originate from the southwest,22,23 placing Pier 27 downwind of industrial releases on a majority of days during the cormorant breeding season. Farr Island, Hamilton Harbour, was approximately four kilometres from the industrial core and more northerly, but also directly west of the same major highway (Figure 1). Further descriptions are available in Somers et al.,24 Gebauer et al.,25 and Quinn et al.26 We chose Mohawk Island National Wildlife Area (42° 50′ 2.5008″ N, 79° 31′ 22.497″ W) in northeastern Lake Erie as our reference site 11638

dx.doi.org/10.1021/es502720a | Environ. Sci. Technol. 2014, 48, 11637−11645

Environmental Science & Technology

Article

of 2480 chick alleles scored, representing 2480 meioses. If a chick’s sample showed an allele that was not seen in either of the parents, but one or both parents were homozygous at that locus, then we did not score this as a mutation since this could have represented allele dropout in a parent. When more than one chick in the same family showed the same mutation at the same locus, this was counted as only one mutation; this conservative approach assumes the mutations in both chicks result from the same mutation event during gametogenesis. However, when chicks in the same family showed different mutations in different loci, this was scored as more than one mutation within the family since this situation could not have arisen from the same mutation event. Offspring alleles were scored as different sizes from the parental alleles only if they were at least one repeat unit away; we did not score 1 and 2 bp differences from parental alleles as mutations. Comparing Mutations among Sites. A generalized linear model (GLZ)31 was used to determine if the number of mutations varied among sites. As the data were counts, a Poisson distribution was used with a log link function. Type 3 likelihood ratio tests were used to compare the deviance of the full model to the null model. Pearson χ2 was used to detect dispersion, when the observed variance is greater (overdispersion) or lesser (underdispersion) than the predicted variance. Dispersion was corrected by multiplying the covariance matrix by a dispersion factor (Φ, defined as the deviance divided by its respective degrees of freedom), and the log likelihood (type III) was divided by Φ.31 As there are no post hoc tests for GLZs, we used the overlap of confidence intervals (CI) of the mean mutation rates to assess differences among sites. Given that the percentile of CI required to approximate tests using an α = 0.05 is a function of the relative ratios of standard errors between means, we used a second order polynomial regression between the ratio of standard errors, and the confidence levels for CI that yields a probability of overlap of 0.9532 to calculate the appropriate CI for each post hoc comparison. We confirmed the results of the GLZ by using a Monte Carlo simulation33 which is nonparametric and relies on few assumptions about the data. We calculated a test statistic by calculating an F-statistic (e.g., ANOVA) on the mean number of mutations among sites. We then created 10 000 permutations of the data, by scrambling the families with respect to site in a way that preserved the total number of mutations and the total number of samples per site, and computed the value of the F-test statistic for each one. We then computed the p-value by computing the proportion of the permuted distribution of F-values greater than or equal to the test statistic computed for the observed data. PAH Metabolite Screening. Six hydroxylated metabolites were analyzed in cormorant samples using liquid chromatography-tandem mass spectrometry (LC-MS/MS) to evaluate PAH exposure: 1-hydroxypyrene, 2-hydroxyphenanthrene, 1and 3-hydroxybenzo[a]pyrene, benzo[a]pyrene-7,8-diol, and 2hydroxynaphthalene. Fecal samples from both adults and chicks (n = 19, up to 0.7 g wet weight) were enzyme digested overnight by incubating with 20 μL of Helix pomatia enzyme with beta glucuronidase and aryl sulfatase activity (SigmaAldrich) and 1 mL of buffer solution (sodium acetate, pH 5); samples were subsequently air-dried to obtain a dry weight (up to 1.04g). Bile samples (n = 18, up to 0.3 g wet weight) were removed from the gallbladder with a needle and liver samples (n = 8, up to 0.5 g wet weight) dissected out of thawed

as the surrounding area is rural and agricultural with little industry or urban activity.27 We trapped adult cormorants on their ground nests at these three colonies during peak breeding season between May and July in both 2009 and 2010 (further details presented in the Supporting Information, SI). Once the two presumed parents had been blood sampled at one nest, we took blood samples from all of the chicks present in that nest. In total over the two field seasons, we blood sampled 113 complete cormorant families (47 from Pier 27, 36 from Farr Island, and 30 from Lake Erie). In 2010, we collected fecal samples from both chicks and adults whenever possible, by holding 1.5 mL plastic tubes close to the cloaca. We also opportunistically collected 18 dead chicks that appeared to have died of heat or cold stress from all three colonies during the course of field work in 2010. Sampling and collection were approved by Environment Canada’s Animal Care Committee and McMaster University’s Animal Research Ethics Board. Microsatellite Genotyping. We present a more detailed summary of the microsatellite genotyping in the SI. Briefly, DNA was PCR amplified and cormorants were genotyped to allow for allele length comparisons between parents and their offspring. Two microsatellites (CORM 4A and CORM 5A2) were selected from a set isolated from cormorant genomic DNA (T.W.Q. and M.K., unpublished), using a rapid capture method28 (see SI Table S1). We selected three microsatellite markers (COR 01, COR 03, and COR 06) with high heterozygosity and number of alleles from Fike et al.,29 for a total of five tetranucleotide microsatellite loci. Differentiating Extrapair Events, Mistrapping and Mutations. To differentiate between extrapair parentage or mistrapping as opposed to a true mutation, we calculated the likelihood of the parents being assigned correctly using a probability-based measure (cumulative probability of resemblance, PRaCum)30 using frequencies of alleles shared between the chick and the putative parent. As given by Ibarguchi et al.,30 we calculated the probability that two individuals share at least one specific allele ‘a’ at a locus by chance (PRa) as follows: PRa = (2pa − pa2 )2

(1)

and for multiple loci, r, the cumulative probability that two individuals share at least one copy of a specific allele loci by chance as follows: PRaCum =

∏ (PRa ) r

r

(2)

When PRaCum was less than 0.001, we counted this as a mutation. This represents a less than 0.1% chance that this incorrect parent, sharing these exact alleles, was randomly sampled in the colony, so the probability that this is the correct parent is thus at least 99.9%. To illustrate the sensitivity of different acceptance thresholds for the cumulative probability of resemblance on estimating mutations, we also calculated PRaCum for values of 0.05, 0.01, and 0.0001. Seven multichick families (6 from Pier 27, 1 from Lake Erie) had one chick excluded, and 16 families were excluded (8 from Pier 27, 6 from Farr Island, 2 from Lake Erie) This left 97 families to be screened for mutations (39 from Pier 27, 30 from Farr Island, 28 from Lake Erie). We genotyped and scored 10 alleles in each chick from the 97 families (n = 248 chicks; 83 from Pier 27, 92 from Farr Island, and 73 from Lake Erie), for a total of 496 alleles scored per locus and a total 11639

dx.doi.org/10.1021/es502720a | Environ. Sci. Technol. 2014, 48, 11637−11645

Environmental Science & Technology

Article

Table 1. Microsatellite Mutations Observed in 97 Cormorant Familiesa family ID

chick ID

locus

paternal alleles (bp)

maternal alleles (bp)

chick alleles (bp)

unit change

parental origin

Pier 27

site

Nest 2

PHA72

Pier 27

Nest 27

181/222 157/161 244/260 184/226

206/210 153/165 232/244 184/226

maternal maternal paternal either

Nest 36 Nest 53

06 06

153/165 149/169

137/165 145/169

−1 −4

maternal maternal

Pier 27

Nest 54

Pier 27 Pier 27

Nest 55 Nest 68

Farr I.

Nest 153

PHA572 PHA511 PHA513 PHA226

Lake Erie

Nest 212

PHA595

03 06 5A2 4A 5A2 01 4A 03

188/204 128/157 244/248 174/194 232/228 179/183 174/182 205/226

218/226 128/161 232/240 174/178 232/256 179/183 202/206 222/230

222/238 157/173 232/256 184/189 189/226 153/161 128/169 128/169 180/218 128/173 244/248 169/174 232/252 179/187 198/202 200/230

+7 +2 +1 +1

Pier 27 Pier 27

PHA27 PHA28 PHA296 PHA565 PHA566 PHA569

03 06 5A2 03

−2 +3 +1 −2 −1 +1 +4 −1

paternal maternal maternal maternal maternal either paternal paternal

a

Mutations were observed at all of the three sites and at 5 of 5 microsatellite loci used. Mutated alleles in chicks and putative parental progenitor allele(s), assuming closest allele contributed (stepwise mutation model), shown in bold.

Table 2. Locus-Specific, Site-Specific, And Total Mutation Rates and Numbers of Mutated Alleles in 97 Cormorant Families (39 from Pier 27 (83 Chicks), 30 from Farr Island (92 Chicks), and 28 from Lake Erie (73 Chicks)a site

Pier 27, Hamilton Harbour (Lake Ontario)

COR 01 mutations mutation rate (allele) COR 03 mutations mutation rate (allele) COR 06 mutations mutation rate (allele) CORM 4A mutations mutation rate (allele) CORM 5A2 mutations mutation rate (allele) total (all five microsatellite loci) no. of mutations mutation rate (indiv.) mutation rate (allele) chicks with mutations families with mutations a

Farr Island, Hamilton Harbour (Lake Ontario)

Lake Erie (cleaner reference site)

1 1/184 = 0.0054 4 (3 unique) 3/166 = 0.0180 5 (4 unique) 4/166 = 0.0240 1 1/166 = 0.0060 3

1 1/146 = 0.0068

1 1/184 = 0.0054

3/166 = 0.0180 13 (11 unique) 11/83 = 0.132 11/830 = 0.013 10 of 83 = 12.0% 7 of 39 = 17.9%

total (all three sites) 1 1/496 = 0.0020 5 (4 unique) 4/496 = 0.008 5 (4 unique) 4/496 = 0.008 2 2/496 = 0.0040 3 3/496 = 0.006

2 2/92 = 0.0217 2/920 = 0.0021 1 of 92 = 1.1% 1 of 30 = 3.3%

1 1/73 = 0.0136 1/730 = 0.0013 1 of 73 = 1.3% 1 of 28 = 3.5%

16 (14 unique) 14/248 = 0.056 14/2480 = 0.0056 12 of 248 = 4.8% 9 of 97 = 9.2%

“Mutation rate (indiv.)” is the per-individual mutation rate and “Mutation rate (allele)” is the per-allele mutation rate (mutation rate per meiosis).

the MOE. Modified high volume air samplers with Teflon filters sampled air for 24 h periods for 8 PAHs for GC/MS analysis (benzo[a]anthracene, benzo[a]pyrene, benzo[b]fluoranthene, benzo[g,h,i]perylene, benzo[k]fluoranthene, chrysene, dibenzo[a,h]anthracene, and indeno [1,2,3,cd] pyrene). These samplers recorded airborne PAHs every 12 days, leading to eight sampling days during the potential exposure period of each year, for a total of 16 PAH sampling days in both years. We then compared the total PAH concentrations for each day of sampling in both study years to those from the same station in previous years11,35 with a mixed linear model to determine if there were differences in PAHs in Hamilton Harbour among studies. PAH emission data from industries in the Hamilton Harbour region for 2009 and 2010 were taken from Environment Canada’s National Pollutant Release Inventory.8 Combining the number of hours downwind and exposure days, we quantified the percentage of days that cormorants

cormorant chicks, then frozen at −15 °C until enzyme digestion. For further LC-MS/MS analysis details, see the SI. Wind Directions and Airborne PAH Concentrations. Cormorants started to arrive at their breeding colonies in late March and early April.34 Therefore, we defined the “potential exposure period” when cormorants would be exposed to genotoxins in Hamilton Harbour as March 20 to June 14, approximately when egg production ended. We collected wind direction and airborne PAH concentrations for this potential exposure period in 2009 and 2010 from the Hamilton Air Monitoring Network (HAMN: hamnair.ca) and the Ontario Ministry of the Environment (MOE). We downloaded hourly average wind direction at 10 m height from the ground at Station 29167, approximately 600 m southwest of the Pier 27 colony (Figure 1) from hamnair.ca. Concentrations of airborne PAHs from the HAMN station (#29547) closest to Pier 27 (approximately 1.2 km southeast, Figure 1) were provided by 11640

dx.doi.org/10.1021/es502720a | Environ. Sci. Technol. 2014, 48, 11637−11645

Environmental Science & Technology

Article

Figure 2. Number of days that the two Hamilton Harbour sites (Pier 27 and Farr Island) were downwind of the industrial area during the two consecutive years of sampling.

Figure 3. Eight airborne PAHs measured by GC/MS at the Pier 25 HAMN monitoring site (Station 29547, 1.2 km southeast of Pier 27, Hamilton Harbour, Ontario), during the exposure period for breeding cormorants, March 20 − June 14 in both 2009 and 2010. Note differing Y axes.

chicks in the family that each had the same size mutated allele not seen in the parents, and in one case, two chicks in the same family each had mutations in two different loci. Overall, we identified 14 unique mutations, 11 of which were at Pier 27 (7 families), 2 at Farr Island (1 family), and 1 from Lake Erie (1 family) (Table 1). The mutation rate at Pier 27 was 6.1 times higher than that at Farr Island, and 9.7 times higher than that at Lake Erie, and mutation rate varied significantly among the three colonies (GLZ: LL = −51.69, χ2[3] = 11.25, p = 0.0036; Monte Carlo: p = 0.0033). On the basis of confidence intervals, mutations were higher at Pier 27 than at either of the other colonies, but not different between Farr Island and Lake Erie. Using different mutation identification thresholds for the cumulative probability of resemblance (PRaCum) barely affected estimates of mutations at Farr Island and Lake Erie, but did change estimates at Pier 27 (SI Table S2). However, the frequency of mutations was significantly different among sites using all thresholds, except the most stringent one, of 0.01%.

from both Hamilton Harbour colonies were downwind from the PAH sources during the breeding period, and in order to demonstrate similar Hamilton Harbour exposures across years and studies, grouped the exposure days into the same categories as Somers et al.11 We measured wind directions that resulted in Pier 27 being downwind of the industrial area in two different ways: westerly wind (originating from 180° to 360°, Pier 27 approximately downwind) and southwesterly wind (originating from 180° to 270°, Pier 27 directly downwind). We estimated that southerly wind would result in Farr Island being approximately downwind of the industrial area, so we calculated hours of southerly wind (originating from 90° to 270°).



RESULTS Mutation Rates. We identified at least one mutated allele in each of the five microsatellite loci screened. Nine cormorant families contained chicks with at least one mutation out of 97 families (9.3% of all families; Tables 1 and 2). In total, 4.8% of chicks had at least one mutation. In two cases, there were two 11641

dx.doi.org/10.1021/es502720a | Environ. Sci. Technol. 2014, 48, 11637−11645

Environmental Science & Technology



Per-allele mutation rates for the 5 loci including all three sites ranged from 0.002 to 0.006 (Table 2), giving an overall perallele mutation rate of 0.0044 including all five loci. In two cases (14%) of the 14 unique mutations, we could not identify which allele was the putative progenitor allele since each parent had the same genotype or alleles at that locus. In the remaining cases, if we assumed the source of the chick mutant allele was the parental allele closest in size, then 4/12 mutations, or 33%, arose from the male parent. Using that assumption, 8/14 (57%) mutations consisted of gains in repeat unit(s), and 7/14 (50%) demonstrated multi-repeat-unit changes. Wind and Airborne PAH Pollution in Hamilton Harbour. Hamilton Harbour area industrial PAH emissions were in the range of thousands of kilograms of PAHs per year during the two years of sampling for this study.8 Pier 27 was often downwind of the industrial core (Figure 2): during both potential exposure periods, 53% of days received at least 10 h of westerly wind. Farr Island was less often downwind: only 32% and 37% of days had 10 or more hours downwind. Only 8% and 4% of days at Farr Island had more than 19 h of downwind of the industry, compared to 26% and 30% at Pier 27 (in 2009 and 2010, respectively). The proportions of hours that Pier 27 was downwind of the industrial core did not differ between 2009 and 2010; furthermore, the amounts of time Pier 27 spent downwind during this study were similar to those of earlier studies near Pier 27 (F[1,5] = 0.26, p = 0.92; 2000,11 200435). Benzo[a]pyrene concentrations at Pier 25 (just upwind of Pier 27) exceeded the Ontario Ministry of Environment’s 24-h Ambient Air Quality Criteria (AAQC) guideline of 1.1 ng/m3 for health effects10 during the potential exposure period 3/8 sampling periods, or 37.5% of the time in 2009, and 2/8 (25%) in 2010 (Figure 3). For both 2009 and 2010, the two most abundant PAHs were benzo[b]fluoranthene and chrysene (Figure 3). In 2009, the highest PAH concentrations during the exposure period were on April 25, and during the 2010 exposure period, the highest PAH concentrations were recorded on March 21. Excluding the extreme outlier of March 21, 2010, when total PAH concentrations were approximately 7 times higher than the mean, hours of westerly wind (i.e., hours approximately downwind of the industrial core) was a moderately good predictor of total PAH concentrations for the 24 h sampling period (r2 = 0.557, F[1,14] = 16.41, p = 0.0014). Even though Pier 27 is directly downwind of the industrial core under southwesterly winds, PAH concentrations were not better correlated with southwesterly wind, thus it was not examined further.

Article

DISCUSSION

Mutation Rates. Our data indicate that microsatellite mutations in cormorants were differentially induced among three colonies with varying degrees of PAH contamination. Microsatellites mutations in cormorants at Pier 27 were 6-fold higher than at Farr Island and 7 times higher than at Lake Erie (which as a relatively uncontaminated area represents a natural background mutation rate). Several lines of evidence suggest that our mutation results are robust. We used a conservative method for the identification of induced mutations by removing families based on shared alleles. Our value (0.001) was lower than others previously used in the literature (e.g., >0.005)36 and corresponds to less than 1 in 1000 individuals, an appropriate probability since each colony contained approximately 1000 adult cormorants of each sex. Changing the acceptance thresholds for the cumulative probability of resemblance did affect estimates of mutations at Pier 27, and resulted in statistical significance in all cases except with a criterion of 0.01% of erroneously scoring mutations. We considered identical mutations in two siblings as a single mutation, since they may be nonindependent and result from the same gametogenic event. Since the two instances of identical mutations between siblings were both at Pier 27, if we considered each mutation to be independent,37 then the number of mutations at the most contaminated site would have been even higher. Mutation rates vary enormously across locus type (e.g., tandem repeat DNA versus coding DNA), sex, species, tissue type, and age.38,39 Environmental stressors such as radiation and certain chemicals can induce tandem repeat DNA mutations.40,41 Background microsatellite mutation rates are generally in the range of 10−7 to 10−3 per locus per generation.42 In birds, background microsatellite mutation rates ranged from 0.0029643 up to 10.8% of all meiosis.37 Perindividual microsatellite mutation rates over all loci in this study ranged from 0.004 to 0.016. The five loci mutated at significantly different rates, with the highest rates in COR 03 and COR 06, which also had the most alleles (17 and 15). PAH Metabolites in Cormorant Samples. Few papers have reported metabolites of PAHs in birds. Troisi et al.21 detected six PAH metabolites in livers taken from oiled wild guillemots (Uria aalge), while Fournier et al.44 fed laying chickens (Gallus gallus domesticus) three PAHs. The concentrations of PAH metabolites in these two studies were similar to those we reported of benzo[a]pyrene-7,8-diol (66−250 ng/g wet weight) in livers and bile of cormorant chicks. A major difference is that these two previous PAH metabolite studies used heavily contaminated, adult birds; environmental PAH concentrations in Hamilton Harbour chicks may simply be lower than what would be required to reliably detect PAH metabolites. It is possible that early in life, PAH metabolites are not yet present in high concentrations as the hatched chick has been environmentally exposed to airborne particulate for only a few days. Nevertheless, this is the first known report of PAH metabolites in wild birds not directly oiled but rather exposed to ambient chemical contamination. Exposure Differences at the Three Colonies. We evaluated the effects of a contaminated environment (including both air and water/diet), where cormorants could have been exposed via these two main routes. To assess the possibility that differing diets at the two Hamilton Harbour sites could contribute to differential contamination uptake via the food



PAH METABOLITES IN CORMORANT SAMPLES Benzo[a]pyrene-7,8-diol, a metabolite of benzo[a]pyrene, was detected in 4/19 bile samples, and 4/8 liver samples (SI Table S3, Figure S1). Benzo[a]pyrene-7,8-diol was detected in samples from Farr Island and Pier 27, both in Hamilton Harbour, but not in the two samples from Lake Erie. It was not possible to compare concentrations of benzo[a]pyrene-7,8-diol in samples from different sites, tissue types, sexes, or ages/ weights of chicks given the low detection rate overall. We did not detect 1-hydroxypyrene, 2-hydroxy-phenanthrene, nor 1- or 3-hydroxybenzo[a]pyrene in plasma, fecal, bile, or liver samples from contaminated or reference site cormorants, although we could detect all except 1- or 3-hydroxybenzo[a]pyrene during spike and recovery (SI Figure S2). 11642

dx.doi.org/10.1021/es502720a | Environ. Sci. Technol. 2014, 48, 11637−11645

Environmental Science & Technology

Article

hydrocarbons such as 1,3-butadiene are potential microsatellite mutagens, they are not emitted by industry in large quantities in Hamilton Harbour.8 Therefore, only PAHs and their complexes both have the potential to induce the observed mutations, and are present in high enough concentrations in Hamilton Harbour. PAH Exposure and Mutation Induction. Our results support the conclusions of previous research11−14,35 that airborne pollution in Hamilton Harbour is associated with increased mutation in tandem repeat DNA loci, and that industrial contaminants including PAHs are likely responsible. PAH concentrations in Hamilton Harbour near the Pier 27 site were high and regularly exceeded the provincial health effect guidelines. The 8 PAHs measured in Hamilton Harbour are all associated with the particulate (rather than the vapor) phase57 which induced mutations in caged mice.11 Total PAH concentrations during the years of this study were similar to those reported previously11,35 suggesting that the potential of airborne exposures at Pier 27 was similar among years and studies (see the SI). Environmentally Induced Microsatellite Mutations. Previous research demonstrated significantly increased microsatellite mutation rates in barn swallows breeding near Chernobyl.16,58 Ben-Shlomo and Shanas59 reported extra peaks in microsatellite loci of wild mice breeding in an asbestos-contaminated area, which they suggested could have arisen through somatic mutation. In Hamilton Harbour, wild tree swallows (Tachycineta bicolor) and caged fathead minnows (Pimephales promelas) were screened for microsatellite mutations at one and three loci, respectively, but significant differences between exposed and reference groups were not identified.37,60 We present here the first evidence for germline microsatellite mutations induced through exposure to nonradioactive chemical environmental contamination. Microsatellites could therefore represent useful tools in future environmental mutation studies, especially since they are available for many species and relatively easy to amplify. Microsatellite mutations may signal widespread DNA damage, such as strand breaks and chromosomal changes;35,61 this could be due to overall interference with important repair mechanisms, polymerases, or enzymes involved in DNA repair. The induction of microsatellite mutations by radiation in wild tree swallows was associated with an increased frequency of partial albinism, along with a corresponding reduction in fitness.16 The link between airborne PAH exposure and toxicity in birds is relatively poorly explored and understood. Following years of depressed productivity, reductions in PAHs over a 5 year period after decades of discharges resulted in increased survivorship of common eider ducklings (Somateria mollissima).62 DNA adducts were elevated in herring gulls exposed to PAHs and other potential genotoxins,63 although links to other health indicators were unclear. Although cormorant populations in Hamilton have increased considerably from the 1990s to the mid 2000s,24 we have not yet investigated any effects resulting from these observed mutations. Ongoing studies in our group are working toward establishing relationships between mutations, genotoxicity, and other measures of health and fitness in PAH-exposed cormorants.

web, we conducted a concurrent study using traditional diet analysis as well as fatty acid and stable isotope analyses. Cormorant diets at the two sites in Hamilton Harbour were similar as estimated with regurgitate, which consisted almost entirely of Alewife (Alosa pseudoharengus, 51−56% by frequency) and Round Goby (Neogobius melanostomus, 25− 42% by frequency). Fatty acid and stable isotope profiles at these two colonies were also extremely similar (no significant differences between the two colonies: King et al., unpublished data). Cormorants make use of wide foraging ranges, typically several kilometres from their home colony;34,45 in one study, males most commonly foraged 20−25 km from their home colony, while females most commonly foraged at 5−10 km from their colony.46 This suggests that adult cormorants from the Pier 27 and Farr Island colonies, which are approximately three kilometres apart, forage for the same resources in the Hamilton Harbour and Lake Ontario area. Additionally, cormorants are largely piscivorous and do not use potentially contaminated terrestrial food sources. Therefore, differences in exposure via diet at these Hamilton Harbour sites would be minimal, such that the major exposure difference between the two sites is in their airborne contaminant exposure. Most cormorants from the Great Lakes winter in the southern United States.47 Of all the cormorants previously banded in Hamilton Harbour, winter band recoveries were distributed almost evenly among 9 southeastern American states, extending from Texas east to North Carolina and south to Florida.48 This demonstrates that these cormorants do not uniformly select the same wintering areas, and therefore variability in winter exposure within any one colony is likely to be high compared to their near-constant and uniform contaminant exposure during the summer breeding season. The amount of time spent downwind or upwind of mutagenic contaminants can be critical.49 Direct wind from the nearby heavy traffic may also be important. Farr Island was subject to fewer hours downwind of industry than was Pier 27 during study years. Through the use of NOx (nitrogen oxides) as an indicator of traffic pollution and SO2 concentrations as an indicator of industrial emissions, Wallace et al.22 demonstrated that although Farr Island is subject to similar if not more trafficbased contamination, Pier 27 has much higher concentrations of SO2 than air around Farr Island. Therefore, Pier 27 is subject to much higher industrial emissions than is Farr Island. Furthermore, Barker50 argued that the local highway’s (Queen Elizabeth Way) contribution to airborne particular matter smaller than 2.5 μm in diameter was minimal, at a site within 1 km in Burlington (∼715 m from Farr Island), and airborne particulate matter was dominated by industrial sources. Since the majority of mutations were found at Pier 27 and the mutation rate was six times higher than that at Farr Island, this suggests that airborne industrial and not traffic-related contaminants were more likely to have induced these mutations. Another important exposure consideration is the contaminants present in Hamilton Harbour. Sediment in the area south of Pier 27 is not only high in PAHs, but PCBs as well;51 metals such as chromium, copper, iron, lead, manganese, and zinc, as well as many other compounds such as pesticides have been identified.52 However, PCBs have only resulted in mutations when injected53 and not when wildlife was environmentally exposed.54 Neither metals55 nor dioxin56 seem to induce microsatellite mutations either, and no studies currently link pesticide exposure to mutation induction. While other 11643

dx.doi.org/10.1021/es502720a | Environ. Sci. Technol. 2014, 48, 11637−11645

Environmental Science & Technology



Article

Comparison of two factor analysis methods. Environ. Sci. Technol. 2008, 42, 6007−6014. (10) Ontario’s Ambient Air Quality Criteria (AAQC); Ontario Ministry of the Environment (MOE), 2008; http://www.ontario.ca/ environment-and-energy/ontarios-ambient-air-quality-criteria-sortedchemical-abstracts-service. (11) Somers, C. M.; McCarry, B. E.; Malek, F.; Quinn, J. S. Reduction of particulate air pollution lowers the risk of heritable mutations in mice. Science 2004, 304, 1008−1010. (12) Yauk, C. L.; Quinn, J. S. Multilocus DNA fingerprinting reveals high rate of heritable genetic mutation in herring gulls nesting in an industrialized urban site. Proc. Natl. Acad. Sci. U. S. A. 1996, 93, 12137−12141. (13) Yauk, C. L.; Fox, G. A.; McCarry, B. E.; Quinn, J. S. Induced minisatellite germline mutations in herring gulls (Larus argentatus) living near steel mills. Mutat. Res. Mol. Mech. Mutagen. 2000, 452, 211−218. (14) Somers, C. M.; Yauk, C. L.; White, P. A.; Parfett, C. L. J.; Quinn, J. S. Air pollution induces heritable DNA mutations. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 15904−15907. (15) Topinka, J.; Schwarz, L. R.; Kiefer, F.; Wiebel, F. J.; Gajdos, O.; Vidová, P.; Dobiás, L.; Fried, M.; Srám, R. J.; Wolff, T. DNA adduct formation in mammalian cell cultures by polycyclic aromatic hydrocarbons (PAH) and nitro-PAH in coke oven emission extract. Mutat. Res. Toxicol. Environ. Mutagen. 1998, 419, 91−105. (16) Ellegren, H.; Lindgren, G.; Primmer, C. R.; Møller, A. P. Fitness loss and germline mutations in barn swallows breeding in Chernobyl. Nature 1997, 389, 593−596. (17) Hennemann, W., III. Energetics and spread-winged behavior in Anhingas and Double-Crested Cormorants: The risks of generalization. Am. Zool. 1988, 28, 845−851. (18) Brown, R. E.; Brain, J. D.; Wang, N. The avian respiratory system: A unique model for studies of respiratory toxicosis and for monitoring air quality. Environ. Health Perspect. 1997, 105, 188−200. (19) Foster, K. L.; Wang, S. W. The canary is alive and singing: Birds continue to provide invaluable information about our changing environment. Integr. Environ. Assess. Manag. 2011, 7, 148−149. (20) Van Schanke, A.; Holtz, F.; van der Meer, J.; Boon, J. P.; Ariese, F.; Stroomberg, G.; van den Berg, M.; Everaarts, J. M. Dose- and timedependent formation of biliary benzo[a]pyrene metabolites in the marine flatfish dab (Limanda limanda). Environ. Toxicol. Chem. 2001, 20, 1641−1647. (21) Troisi, G. M.; Bexton, S.; Robinson, I. Polyaromatic hydrocarbon and PAH metabolite burdens in oiled common guillemots (Uria aalge) stranded on the east coast of England (2001−2002). Environ. Sci. Technol. 2006, 40, 7938−7943. (22) Wallace, J.; Corr, D.; Deluca, P.; Kanaroglou, P.; McCarry, B. E. Mobile monitoring of air pollution in cities: the case of Hamilton, Ontario, Canada. J. Environ. Monit. 2009, 11, 998−1003. (23) Wallace, J.; Corr, D.; Kanaroglou, P. Topographic and spatial impacts of temperature inversions on air quality using mobile air pollution surveys. Sci. Total Environ. 2010, 408, 5086−5096. (24) Somers, C. M.; Lozer, M. N.; Quinn, J. S. Interactions between double-crested cormorants and herring gulls at a shared breeding site. Waterbirds 2007, 30, 241−250. (25) Gebauer, M. B.; Dobos, R. Z.; Weseloh, D. V. Waterbird surveys at Hamilton Harbour, Lake Ontario, 1985−1988. J. Great Lakes Res. 1992, 18, 420−439. (26) Quinn, J. S.; Morris, R. D.; Blokpoel, H.; Weseloh, D. V.; Ewins, P. J. Design and management of bird nesting habitat: tactics for conserving colonial waterbird biodiversity on artificial islands in Hamilton Harbour, Ontario. Can. J. Fish. Aquat. Sci. 1996, 53, 45−57. (27) King, L. E.; de Solla, S. R. Successful renesting of Caspian Terns on Mohawk Island, Lake Erie, after complete colony failure. Ontario Birds 2010, 28, 158−165. (28) St. John, J.; Quinn, T. W. Rapid capture of DNA targets. Biotechniques 2008, 44, 259−264.

ASSOCIATED CONTENT

S Supporting Information *

Detailed sampling, PCR, and genotyping methods, as well as LC-MS/MS results. This material is available free of charge via the Internet at http://pubs.acs.org/.



AUTHOR INFORMATION

Corresponding Author

*Phone: 1-(905)-336-4686; e-mail: [email protected]. Present Address ∥

Ecotoxicology and Wildlife Health Division, Canada Centre for Inland Waters, Environment Canada, Burlington, Ontario L7R 4A6, Canada. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS Thanks to K. Intini, K. Palonen, and G. Barrett for field work, and to S. Coulson and E. Kerr at Trent University. C. Dey contributed the mixed linear model. Thanks to the Hamilton Port Authority for permission to sample on their property and Mohawk Marina in picturesque Lowbanks. The authors would also like to thank the three anonymous reviewers, whose advice and comments were helpful in improving our manuscript. Funded by an NSERC CGS M and NSERC MSFSS to L.E.K., OGS to L.E.K., and Environment Canada’s Chemicals Management Plan (CMP) to S.d.S. The funders did not participate in the design of the study, data analysis, interpretation of the data, nor the writing of the manuscript.



REFERENCES

(1) Correia, A. W.; Pope, C. A., (III); Dockery, D. W.; Wang, Y.; Ezzati, M.; Dominici, F. The effect of air pollution control on life expectancy in the United States: an analysis of 545 US counties for the period 2000 to 2007. Epidemiology 2013, 24, 23−31. (2) Greaver, T. L.; Sullivan, T. J.; Herrick, J. D.; Barber, M. C.; Baron, J. S.; Cosby, B. J.; Deerhake, M. E.; Dennis, R. L.; Dubois, J.-J. B.; Goodale, C. L.; et al. Ecological effects of nitrogen and sulfur air pollution in the US: what do we know? Front. Ecol. Environ. 2012, 10, 365−372. (3) Xu, Z.; Xu, X.; Zhong, M.; Hotchkiss, I. P.; Lewandowski, R. P.; Wagner, J. G.; Bramble, L. A.; Yang, Y.; Wang, A.; Harkema, J. R.; et al. Ambient particulate air pollution induces oxidative stress and alterations of mitochondria and gene expression in brown and white adipose tissues. Part. Fibre Toxicol. 2011, 8, 20. (4) Xu, H.; Wang, X.; Pöschl, U.; Feng, S.; Wu, D.; Yang, L.; Li, S.; Song, W.; Sheng, G.; Fu, J. Genotoxicity of total and fractionated extractable organic matter in fine air particulate matter from urban Guangzhou: Comparison between haze and nonhaze episodes. Environ. Toxicol. Chem. 2008, 27, 206−212. (5) Belpomme, D.; Irigaray, P.; Hardell, L.; Clapp, R.; Montagnier, L.; Epstein, S.; Sasco, A. J. The multitude and diversity of environmental carcinogens. Environ. Res. 2007, 105, 414−429. (6) Wang, W.; Jariyasopit, N.; Schrlau, J.; Jia, Y.; Tao, S.; Yu, T.-W.; Dashwood, R. H.; Zhang, W.; Wang, X.; Simonich, S. L. M. Concentration and photochemistry of PAHs, NPAHs, and OPAHs and toxicity of PM2.5 during the Beijing Olympic Games. Environ. Sci. Technol. 2011, 45, 6887−6895. (7) Albers, P. H. Birds and polycyclic aromatic hydrocarbons. Avian Poult. Biol. Rev. 2006, 17, 125−140. (8) Environment Canada. National Pollutant Release Inventory (NPRI) for 2009 and 2010; ec.gc.ca/inrp-npri/. (9) Sofowote, U. M.; McCarry, B. E.; Marvin, C. H. Source apportionment of PAH in Hamilton Harbour suspended sediments: 11644

dx.doi.org/10.1021/es502720a | Environ. Sci. Technol. 2014, 48, 11637−11645

Environmental Science & Technology

Article

(29) Fike, J. A.; Devault, T. L.; Rhodes, O. E. Identification of 24 polymorphic microsatellite markers for the double-crested cormorant (Phalacrocorax auritus). Mol. Ecol. Resour. 2009, 9, 1183−1185. (30) Ibarguchi, G.; Gissing, G. J.; Gaston, A. J.; Boag, P. T.; Friesen, V. L. Male-biased mutation rates and the overestimation of extrapair paternity: Problem, solution, and illustration using Thick-Billed Murres (Uria lomvia, Alcidae). J. Hered. 2004, 95, 209−216. (31) McCullagh, P.; Nelder, J. Generalized Linear Models, 2nd ed.; Springer: New York, 1989. (32) Payton, M. E.; Greenstone, M. H.; Schenker, N. Overlapping confidence intervals or standard error intervals: what do they mean in terms of statistical significance? J. Insect Sci. 2003, 3, 34. (33) Hood, G. M. PopTools, version 3.2.5, 2010. (34) Hatch, J. J.; Weseloh, D. V. C. Double-Crested Cormorant (Phalacrocorax auritus), The Birds of North America Online; Poole, A., Ed.).; Ithaca, N.Y., 1999. (35) Yauk, C.; Polyzos, A.; Rowan-Carroll, A.; Somers, C. M.; Godschalk, R. W.; Van Schooten, F. J.; Berndt, M. L.; Pogribny, I. P.; Koturbash, I.; Williams, A.; et al. Germ-line mutations, DNA damage, and global hypermethylation in mice exposed to particulate air pollution in an urban/industrial location. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 605−610. (36) Evans, M. L.; Stutchbury, B. J. M.; Woolfenden, B. E. Offterritory forays and genetic mating system of the Wood Thrush (Hylocichla mustelina). Auk 2008, 125, 67−75. (37) Anmarkrud, J. A.; Kleven, O.; Augustin, J.; Bentz, K. H.; Blomqvist, D.; Fernie, K. J.; Magrath, M. J. L.; Pärn, H.; Quinn, J. S.; Robertson, R. J.; et al. Factors affecting germline mutations in a hypervariable microsatellite: A comparative analysis of six species of swallows (Aves: Hirundinidae). Mutat. Res. Mol. Mech. Mutagen. 2011, 708, 37−43. (38) Campbell, C. D.; Eichler, E. E. Properties and rates of germline mutations in humans. Trends Genet. 2013, 29, 575−584. (39) Hardwick, R. J.; Tretyakov, M. V.; Dubrova, Y. E. Age-related accumulation of mutations supports a replication-dependent mechanism of spontaneous mutation at tandem repeat DNA loci in mice. Mol. Biol. Evol. 2009, 26, 2647−2654. (40) Tsyusko, O.; Yi, Y.; Coughlin, D.; Main, D.; Podolsky, R.; Hinton, T. G.; Glenn, T. C. Radiation-induced untargeted germline mutations in Japanese medaka. Comp. Biochem. Physiol. Part C 2007, 145, 103−110. (41) Vilariño-Güell, C.; Smith, A. G.; Dubrova, Y. E. Germline mutation induction at mouse repeat DNA loci by chemical mutagens. Mutat. Res. Mol. Mech. Mutagen. 2003, 526, 63−73. (42) Buschiazzo, E.; Gemmell, N. J. The rise, fall and renaissance of microsatellites in eukaryotic genomes. BioEssays 2006, 28, 1040−1050. (43) Ortego, J.; Aparicio, J. M.; Cordero, P. J.; Calabuig, G. Characteristics of loci and individuals are associated with germline microsatellite mutation rates in lesser kestrels (Falco naumanni). Mutat. Res. Mol. Mech. Mutagen. 2008, 648, 82−86. (44) Fournier, A.; Feidt, C.; Dziurla, M.-A.; Grandclaudon, C.; Jondreville, C. Transfer kinetics to egg yolk and modeling residue recovered in yolk of readily metabolized molecules: Polycyclic aromatic hydrocarbons orally administered to laying hens. Chemosphere 2010, 78, 1004−1010. (45) Ridgway, M. S.; Middel, T. A. Distribution, relative abundance, and prey demand by double-crested cormorants on Manitoulin Island lakes vs. Lake Huron Coast. J. Great Lakes Res. 2011, 37, 609−615. (46) Anderson, C. D.; Roby, D. D.; Collis, K. Foraging patterns of male and female Double-Crested Cormorants nesting in the Columbia River estuary. Can. J. Zool. 2004, 82, 541−554. (47) Canadian Atlas of Bird Banding, Volume 2 (Seabirds): Doublecrested Cormorant (pp. 154−157); Environment Canada; ec.gc.ca/ aobc-cabb/. (48) Analysis of winter band encounters (Nov.−Feb.) of DoubleCrested Cormorants originally banded in Hamilton Harbour, Ontario, Canada. Data provided by the Bird Banding Office, Environment Canada (ec.gc.ca/bbo).

(49) Sofowote, U. M.; Allan, L. M.; McCarry, B. E. Evaluation of PAH diagnostic ratios as source apportionment tools for air particulates collected in an urban-industrial environment. J. Environ. Monit. 2010, 12, 417−424. (50) Barker, H. W. Isolating the Industrial Contribution of PM2.5 in Hamilton and Burlington, Ontario. J. Appl. Meteorol. Climatol. 2013, 52, 660−667. (51) Zeman, A. J.; Patterson, T. S. Characterization and Mapping of Contaminated Sediments, Windermere Arm, Hamilton Harbour, Ontario, Canada. Soil Sediment Contam. 2003, 12, 619−629. (52) Zeman, A. J. Contaminated Sediments in Hamilton Harbour; Environment Canada, 2009; Water Science and Technology Directorate (WSTD) Contribution No. 09-263. (53) Hedenskog, M.; Sjögren, M.; Cederberg, H.; Rannug, U. Induction of germline-length mutations at the minisatellites PC-1 and PC-2 in male mice exposed to polychlorinated biphenyls and diesel exhaust emissions. Environ. Mol. Mutagen. 1997, 30, 254−259. (54) Stapleton, M.; Dunn, P. O.; McCarty, J.; Secord, A.; Whittingham, L. A. Polychlorinated biphenyl contamination and minisatellite DNA mutation rates of tree swallows. Environ. Toxicol. Chem. 2001, 20, 2263−2267. (55) Berckmoes, V.; Scheirs, J.; Jordaens, K.; Blust, R.; Backeljau, T.; Verhagen, R. Effects of environmental pollution on microsatellite DNA diversity in wood mouse (Apodemus sylvaticus) populations. Environ. Toxicol. Chem. 2005, 24, 2898−2907. (56) Ryo, H.; Nakajima, H.; Nomura, T. Germ-line mutations at a mouse ESTR (Pc-3) locus and human microsatellite loci. J. Radiat. Res. 2006, 47, 31−37. (57) Sanderson, E.; Raqbi, A.; Vyskocil, A.; Farant, J.-P. Comparison of particulate polycyclic aromatic hydrocarbon profiles in different regions of Canada. Atmos. Environ. 2004, 38, 3417−3429. (58) Møller, A. P.; Mousseau, T. A.; de Lope, F.; Saino, N. Elevated frequency of abnormalities in barn swallows from Chernobyl. Biol. Lett. 2007, 3, 414−417. (59) Ben-Shlomo, R.; Shanas, U. Genetic ecotoxicology of asbestos pollution in the house mouse Mus musculus domesticus. Environ. Sci. Pollut. Res. Int. 2011, 18, 1264−1269. (60) Miller, J. L. 2009. Germline mutations and reproductive effects in fathead minnows exposed to contaminated sediments from Randle Reef, Hamilton Harbour. M.Sc. Thesis, McMaster University: Hamilton, Ontario, Canada. (61) Fox, G. A.; White, P. A.; Trudeau, S.; Theodorakis, C.; Shutt, L. J.; Kennedy, S. W.; Fernie, K. J. DNA strand length and EROD activity in relation to two screening measures of genotoxic exposure in Great Lakes herring gulls. Ecotoxicology 2005, 14, 527−544. (62) Bustnes, J. O. Reproductive recovery of a common eider Somateria mollissima population following reductions in discharges of polycyclic aromatic hydrocarbons (PAHs). Bull. Environ. Contam. Toxicol. 2013, 91, 202−207. (63) Skarphedinsdottir, H.; Hallgrimsson, G. T.; Hansson, T.; Hägerroth, P.-A.; Liewenborg, B.; Tjärnlund, U.; Akerman, G.; Barsiene, J.; Balk, L. Genotoxicity in herring gulls (Larus argentatus) in Sweden and Iceland. Mutat. Res. Toxicol. Environ. Mutagen. 2010, 702, 24−31.

11645

dx.doi.org/10.1021/es502720a | Environ. Sci. Technol. 2014, 48, 11637−11645