Article pubs.acs.org/est
Volatile Methylsiloxanes and Organophosphate Esters in the Eggs of European Starlings (Sturnus vulgaris) and Congeneric Gull Species from Locations across Canada Zhe Lu,† Pamela A. Martin,*,‡ Neil M. Burgess,§ Louise Champoux,∥ John E. Elliott,⊥ Enzo Baressi,# Amila O. De Silva,† Shane R. de Solla,‡ and Robert J. Letcher*,∇ †
Aquatic Contaminants Research Division and ‡Ecotoxicology and Wildlife Health Division, Environment and Climate Change Canada, Burlington, Ontario L7S 1A1, Canada § Ecotoxicology and Wildlife Health Division, Environment and Climate Change Canada, Mount Pearl, Newfoundland and Labrador A1N 4T3, Canada ∥ Ecotoxicology and Wildlife Health Division, Environment and Climate Change Canada, Québec City, Québec G1J 0C3, Canada ⊥ Ecotoxicology and Wildlife Health Division, Environment and Climate Change Canada, Pacific Wildlife Research Centre, Delta, British Columbia V4K 3Y3, Canada # National Laboratory of Environmental Testing, Environment and Climate Change Canada, Burlington, Ontario L7S 1A1, Canada ∇ Ecotoxicology and Wildlife Health Division, Environment and Climate Change Canada, National Wildlife Research Centre, Carleton University, Ottawa, Ontario K1A 0H3, Canada S Supporting Information *
ABSTRACT: Volatile methylsiloxanes (VMSs) and organophosphate esters (OPEs) are two suites of chemicals that are of environmental concern as organic contaminants, but little is known about the exposure of wildlife to these contaminants, particularly in birds, in terrestrial and aquatic ecosystems. The present study investigates the spatial distributions of nine cyclic and linear VMSs and 17 OPEs in the eggs of European starlings (Sturnus vulgaris) and three congeneric gull species (i.e., herring gull (Larus argentatus), glaucous-winged gull (L. glaucescens), and California gull (L. californicus)) from nesting sites across Canada. ∑VMS concentrations for all bird eggs were dominated by decamethylcyclopentasiloxane (D5), dodecamethylcyclohexasiloxane (D6), and octamethylcyclotetrasiloxane (D4). With European starlings, birds breeding adjacent to landfill sites had eggs containing significantly greater ∑VMS concentrations (median: 178 ng g−1 wet weight (ww)) compared with those from the urban industrial (20 ng g−1 ww) and rural sites (1.3 ng g−1 ww), indicating that the landfills are important sources of VMSs to Canadian terrestrial environments. In gull eggs, the median ∑VMS concentrations were up to 254 ng g−1 ww and suggested greater detection frequencies and levels of VMSs in aquatic- versus terrestrial-feeding birds in Canada. In contrast, the detection frequency of OPEs in all European starling and gull eggs was lower than 16%. This suggested that low dietary exposure or rapid metabolism of accumulated OPEs occurs in aquatic feeding birds and may warrant further investigation for the elucidation of the reasons for these differences.
■
INTRODUCTION Volatile methylsiloxanes (VMSs) and organophosphate esters (OPEs) are organic compounds of great environmental concern.1−4 VMSs are produced at up to several million tons per year globally1 and widely used in the manufacture of consumer and industrial products such as silicone polymers, cosmetics, personal care products, defoamers, adhesives, coatings, dry-cleaning solvents, and industrial cleaning products.1,5,6 Halogenated and nonhalogenated OPEs have been used for several decades in consumer and industrial products such as plastics, textiles, electronics and building materials as flame retardants, and plasticizers.2,3 The worldwide production of OPEs in 2007 was estimated to be about 341 000 tons,3 © 2017 American Chemical Society
indicating that these compounds are produced and utilized in large volumes. As flame retardants, the production and use of OPEs have increased during recent years to replace the banned polybrominated diphenyl ethers (PBDEs).2,3,7 Based on the physicochemical properties (Table S1) and the use pattern of VMSs and OPEs, these chemicals may easily enter the environment during the manufacture, application, and waste-disposal processes. These compounds can volatilize to Received: Revised: Accepted: Published: 9836
June 22, 2017 July 28, 2017 August 3, 2017 August 3, 2017 DOI: 10.1021/acs.est.7b03192 Environ. Sci. Technol. 2017, 51, 9836−9845
Article
Environmental Science & Technology
and aquatic birds. To our knowledge, the study of VMSs and OPEs in European starling eggs represents the first investigation of these substances in any terrestrial-feeding wildlife.
the atmosphere from VMS-containing solid and liquid waste.1,6,8 OPEs can leach into the environment from consumer products because they are usually not chemically bound to the materials.2,9 It has been estimated that about 40% of tris(2chloroisopropyl) phosphate (TCIPP) and 10% of tris(1,3dichloro-2-propyl) phosphate (TDCIPP) can be released into the environment throughout the lifetime of a given product.10,11 As a consequence, OPE and VMS contaminants have been detected in many environmental samples such as surface water,12,13 wastewater,6,13 biosolids,6 sediment,13 and air.7,8,14,15 Organisms may accumulate VMSs and OPEs via respiration, dermal exposure, and ingestion. The accumulation of these contaminants may lead to adverse effects in organisms, including endocrine disruption, carcinogenicity, developmental deformity, neurotoxicity, and dermal, liver and reproductive toxicities.1−3 In peer-reviewed literature, the occurrence of VMSs and OPEs have been reported for some aquatic organisms such as plankton, 16,17 invertebrates, 16 and fish.4,16−18 However, the propensity for bioaccumulation of these contaminants in terrestrially foraging wildlife is unknown. It is also not clear if there are any differences in contamination patterns of VMSs and OPEs between aquatic and terrestrial wildlife. In addition, landfills have been suggested as major sources of brominated flame retardants such as PBDEs, to terrestrial ecosystems;19 however, landfills being a source of VMS and OPE exposure to terrestrial wildlife is unclear. VMSs and OPEs have been detected at much-higher concentrations than that of PBDEs in some environmental matrices.6,20,21 Thus, it is essential to identify the sources of VMSs and OPEs in the environment. It has been estimated that about 26 million tons (777 kg per capita) of municipal solid waste was generated in 2008 in Canada.19,22 A variety of solid waste types, such as the containers of personal care products and cosmetics, car polish and wax, as well as furniture, plastics, and electronics, which have been shown to contain VMSs and OPEs,1,2 are disposed in landfills. These contaminants may escape from landfill waste and enter the environment via volatilization, as leachate, or via landfill diversion (e.g., biological treatment). To better understand the exposure and environmental fate of VMSs and OPEs in wildlife, the present study investigated the concentration and composition profiles of these contaminants in terrestrial and aquatic bird eggs from nesting locations across Canada. The birds selected for this study included European starlings (Sturnus vulgaris) as well as herring gulls (Larus argentatus) and two congeneric species, glaucous-winged gulls (L. glaucescens) and California gulls (L. californicus). European starlings are terrestrial birds and live in habitats close to human activities (e.g., urban and landfill sites and farmland).19,23 They mainly feed on insects, spiders, seeds, grains, fruit, livestock feed, and food waste.19,23 These properties render them ideal species for monitoring how human activities affect contamination by VMSs and OPEs in terrestrial ecosystems.24 As top predators, herring gulls have been used for many years in the spatial and temporal monitoring of environmental pollution in aquatic ecosystems.25 Recently, glaucous-winged gulls have been shown to be an effective contaminant indicator in marine environments.26,27 In the present study, we investigated (1) the concentrations, compositions, and spatial variations of VMSs and OPEs in bird eggs collected from locations across Canada; (2) the hypothesis that landfills are an important source of VMSs and OPEs to the terrestrial environment; and (3) the different contamination patterns existing between terrestrial
■
MATERIALS AND METHODS Samples. The bird eggs used in this study were stored and archived in Environment and Climate Change Canada’s National Wildlife Specimen Bank (Ottawa, Ontario). Fresh European starling eggs were collected in April and May in 2013 and 2014 from nest boxes established in 19 sampling sites with three different land use types: urban industrial centers (industrial districts in major cities), landfills (adjacent to the major cities), and rural (agricultural) sites located either 10 km (only for OPE study) or 40 km upwind from the major cities26 in Canada. The cities selected for this study included Vancouver, British Columbia (BC); Calgary, Alberta (AB); Hamilton, Ontario (ON); Montréal, Québec (QC), and Halifax, Nova Scotia (NS) (Figure S1). The reference site was selected to be at Redcliff (Alberta), which is about 300 km from Calgary (AB) (Figure S1). For each location, a maximum of 9 European starling eggs was pooled on an equal wet weight (ww) basis of the individual eggs.19 The pools were used to create egg homogenates and stored at −40 °C prior to chemical analysis.19 European starling eggs were collected using methods approved by the Environment Canada Animal Care Committee, and a total of n = 71 and n = 101 European starling egg homogenate samples were analyzed for VMSs and OPEs, respectively. Fresh gull eggs were collected from late April to July in 2011 and 2013 from a total of 14 gull colonies across Canada (Figure S1). The egg samples were collected from three gull species, including glaucous-winged gulls from BC, California gull from AB, and herring gulls from the remaining sampling locations. Gull eggs were collected under Canadian federal scientific permits. The sampling sites (Figure S1) included: Cleland Island (site BC1), Mitlenatch Island (site BC2), and Mandarte Island (site BC3) from BC; Dalmead (site AB4) from AB; Great Slave Lake (site NT5) from the Northwest Territories (NT); Pipestone Rocks (site MB6) from Manitoba (MB); Southampton Island (site NU7) from Nunavut (NU); Agawa Rock (site ON8), Hamilton Harbour (site ON9, only analyzed for VMSs) and Tommy Thompson Park (site ON10) from ON; Ile Deslauriers (site QC11) and Ile Bellechasse (site QC12) from QC; Kent Island (site NB13) from New Brunswick (NB); and Gull Island (site NL14) from Newfoundland and Labrador (NL). For each location, a maximum of 10 gull eggs was pooled on an equal wet weight (ww) basis of the individual eggs.25 The pools were used to create egg homogenates and stored at −40 °C prior to chemical analysis.25 A total of n = 99 and n = 114 gull egg homogenate samples were analyzed for VMSs and OPEs, respectively. Chemicals and Reagents. The structures of target compounds are shown in Figure S2. Hexamethylcyclotrisiloxane (D3; CAS no. 541-05-9; purity of >98%), octamethylcyclotetrasiloxane (D4; CAS no. 556-67-2; purity of >98%), decamethylcyclopentasiloxane (D5; CAS no. 541-02-6; purity of >98%), dodecamethylcyclohexasiloxane (D6; CAS no. 54097-6; purity of >98%), tetrakis(trimethylsiloxy)silane (M4Q; CAS no. 3555-47-3; purity of >95%), hexamethyldisiloxane (L2; CAS no. 107-46-0; purity of >99%), octamethyltrisiloxane (L3; CAS no. 107-51-7; purity of >97%), decamethyltetrasiloxane (L4; CAS no. 141-62-8; purity of >95%) and 9837
DOI: 10.1021/acs.est.7b03192 Environ. Sci. Technol. 2017, 51, 9836−9845
Article
Environmental Science & Technology
a.m. (n = 3), noon (n = 3), and 3 p.m. (n = 3)), organic chicken egg (n = 3, with shell), and Lake Superior gull egg homogenates (n = 6) (collected in 2012) were stored in clean glass containers and shipped from Ottawa (Ontario, Canada) to Burlington (Ontario, Canada) (about 500 km). Then, the egg samples were processed and analyzed in both NWRC and the ultraclean lab in NLET. The results showed that there was no significant background contamination occurred during storage and transportation for the target VMSs (details are shown in the Supporting Information and Figure S3). Glass materials were used in the experiment whenever possible to limit any possible background organic contamination. All glassware and aluminum foil was cleaned, solventrinsed, and heated at 450 °C overnight before use to remove any possible background contamination. The outside of the egg shell was rinsed by solvent before accessing and retrieving the egg content. A total of three procedural blanks and one spikerecovery sample (known amount of target compounds added to 1 g of homogenized chicken egg) were included for the OPE analysis of all European starling eggs (10 batches). For all VMS and OPE analysis and measurement in the gull eggs, one procedural blank and one spike-recovery sample were included in each sample batch (VMSs: nine batches for gulls and seven batches for European starlings; OPEs: eight batches for gulls). The concentrations of target compounds in blanks are listed in Tables S6 and S7. The recovery of the target compounds in spike-recovery samples were 65 ± 16% and 73 ± 20% (mean ± standard deviation (SD)) for all VMSs and OPEs, respectively. The recovery of the surrogate standards in egg samples were 64 ± 21% for all VMSs and 79 ± 18% for all OPEs (mean ± SD). The method limits of detection (MLODs) were defined as the concentration that produces a peak with signal-to-noise (S/N) ratio of 3. The method limits of quantification (MLOQs) were based on 3 times the SD of the procedural blanks. For analytes that were not detectable in the method blanks, standard concentrations producing a signal with 10 times the S/N ratio in solvent were used to estimate the MLOQs. The MLODs and MLOQs for target compounds are shown in Tables S2, S3, and S5. For VMS analysis in all eggs and OPE analysis in gull eggs, the samples were corrected using the individual blank in each batch. For OPE analysis in European starling eggs, the samples were corrected using the highest concentration of three blanks plus one standard deviation in each batch. Method performance for VMS was assessed by analyzing a reference material (a fish tissue sample) with known concentrations of D4, D5, and D6 (prepared and used in previous studies) in seven different batches.4,30 In the present study, the measured concentrations of D4, D5, and D6 in the reference material were 75 ± 11%, 77 ± 13%, and 76 ± 13% (mean ± SD) of the expected values, respectively. Data Analysis. The data was statistically analyzed to determine the existence of possible differences in VMS and OPE levels in eggs collected from difference sites and between difference bird species. The statistical analysis of the was performed using RStudio V 0.99.903 (Boston, MA) and GraphPad Prism 7.0 (La Jolla, CA) software. Statistics for data with censored values (≤50% censoring and detects >3) were conducted using the robust regression on order statistics (ROS) method in RStudio by the Nondetects and Data Analysis (NADA) package (version 1.5−6).31 For the sampling sites with small sample size (n ≤ 4), the censored values were
dodecamethylpentasiloxane (L5; CAS no. 141-63-9; purity of >95%) were purchased from Gelest Inc. (Philadelphia, PA). Isotopically labeled VMS surrogate standards (Table S2) were purchased from Moravek (purity of >98%; Brea, CA). For OPEs, triethyl phosphate (TEP; CAS no. 78-40-0; purity of >99%), tris(2-chloroethyl) phosphate (TCEP; CAS no. 115-968; purity of >97%), tripropyl phosphate (TPP; CAS no. 513-086; purity of >99%), triphenyl phosphate (TPHP; CAS no. 11586-6; purity of >99%), Tris(2,3-dibromopropyl) phosphate (TDBPP; CAS no. 126-72-7; purity of >99%), tributyl phosphate (TNBP; CAS no. 126-73-8; purity of >99%), trimethylphenyl phosphate (TMPP; CAS no. 1330-78-5; purity of >98%), Tris(2-butoxyethyl) phosphate (TBOEP; CAS no. 78-51-3; purity of >94%), 2-ethylhexyl-diphenyl phosphate (EHDPP; CAS no. 1241-94-7; purity of >98%), and tris(2ethylhexyl) phosphate (TEHP; CAS no. 78-42-2; purity of >99%) were purchased from Sigma-Aldrich (Oakville, ON, Canada). TCIPP (CAS no. 13674-84-5; purity of >99%), TDCIPP (CAS no. 13674-87-8; purity of >95%), and tris(tribromoneopentyl) phosphate (TTBNPP; CAS no. 19186-97-1; purity of >98%) were purchased from Pfaltz and Bauer (Waterbury, CT), TCI America (Portland, OR) and Accustandard (New Haven, CT), respectively. Tris(2-bromo-4methylphenyl) phosphate (T2B4MP), tris(3-bromo-4-methylphenyl) phosphate (T3B4MP), and tris(4-bromo-3-methylphenyl) phosphate (T4B3MP) were synthesized by GL Chemtec International (Oakville, ON, Canada) (purity of >98%). The standard of 2,2-bis(chloromethyl)propane−1,3diyl tetrakis(2-chloroethyl)bis(phosphate) (BCMP−BCEP) was obtained from Dr. Heather Stapleton at Duke University (purity >97%). Isotopically labeled OPE surrogate standards (Table S3) were purchased from Wellington Laboratories Inc. (Guelph, ON, Canada) (purity >98%). Solvents such as hexane, dichloromethane and methanol (CHROMASOLV Plus grade) were purchased from Sigma-Aldrich. Diatomaceous earth (J.T. Baker, NJ) and sodium sulfate (Na2SO4) were heated at 600 °C overnight in a muffle furnace prior to use. ISOLUTE aminopropyl silica gel was purchased from Biotage (Charlotte, NC). Sample Preparation and Chemical Analysis. Egg sample preparation and sample extraction and instrumental analysis methods have been extensively described elsewhere.4,25,28,29 However, the procedures are also detailed in the Supporting Information. Analysis of VMSs were carried out in the National Laboratory of Environmental Testing (NLET) (ultraclean lab) at the Canada Centre for Inland Waters, Burlington, ON, Canada. All OPE analyses were carried out in the Organics Contaminants Research Laboratory/Letcher Laboratories at the National Wildlife Research Centre (NWRC), Carleton University, Ottawa, ON, Canada. The instrumental parameters (i.e., gas chromatograph−mass spectrometry (GC−MS) for VMS analysis and liquid chromatography−tandem mass spectrometry (LC−MS/MS) for OPE analysis) are shown in Tables S2−S5. QA and QC. Unlike other biotic tissue or body compartment samples, bird egg contents are protected by the egg shell from any possible field sampling and handling contamination of the contents. During sampling, personnel wore clean gloves during egg sample collection and refrained from using personal care products to avoid possible contamination of VMSs. An experiment was conducted to test the possible VMS contamination during sample storage and transportation. Organic chicken egg homogenates (n = 9, homogenized at 9 9838
DOI: 10.1021/acs.est.7b03192 Environ. Sci. Technol. 2017, 51, 9836−9845
2013
2013/2014
2013/2014
2013/2014
2013/2014
AB-Reference site-Redcliff
ON-Landfill-1-Halton
ON-Landfill-2-Brantford
ON-Industrial/Urban-Hamilton
ON-40 km-Delhi
9839
4 5
2 7
6
7
5
7
3
9
5
4 4
6
n
