Environmental Pollution 157 (2009) 2282–2290
Contents lists available at ScienceDirect
Environmental Pollution journal homepage: www.elsevier.com/locate/envpol
Persistent organic pollutants, heavy metals and parasites in the glaucous gull (Larus hyperboreus) on Spitsbergen Kjetil Sagerup a, *, Vladimir Savinov b, Tatiana Savinova b, Vadim Kuklin c, Derek C.G. Muir d, Geir W. Gabrielsen e a
Tromsø University Museum, NO-9037 Tromsø, Norway Akvaplan-niva, The Polar Environmental Centre, NO-9296 Tromsø, Norway Murmansk Marine Biological Institute, Kola Scientific Centre, Russian Academy of Sciences, Murmansk, Russia d Aquatic Ecosystem Protection Research Division, Environment Canada, Burlington ON L7R 4A6, Canada e Norwegian Polar Institute, The Polar Environmental Centre, NO-9296 Tromsø, Norway b c
Consistent relationships between contaminant level and parasite intensity, as an immunotoxic endpoint unit, were not found in the present study.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 September 2008 Received in revised form 23 March 2009 Accepted 26 March 2009
The prediction of a higher parasite infection as a consequence of an impaired immune system with increasing persistent organic pollution (POP) and heavy metal levels were investigated in adult glaucous gull (Larus hyperboreus) from Svalbard. The levels of chlorinated pesticides, polychlorinated biphenyls (PCBs), toxaphenes and polybrominated diphenyl ethers (PBDEs) were measured in liver. Cupper, cadmium, lead, mercury, selenium and zinc were measured in kidney samples. An elevated ratio of PCB-118 was found, suggesting that local contamination from the settlement was detectable in the glaucous gull. Eight cestodes, four nematodes, two acanthocephalan and three trematode helminth species were found in the intestine. A positive correlation was found between cestode intensities and selenium levels and between acanthocephalan intensities and mercury levels. No correlation was found between parasite intensities and POP concentrations. It is concluded that the contaminant levels found in glaucous gulls do not cause immune suppression severe enough to affect parasite intensity. Ó 2009 Elsevier Ltd. All rights reserved.
Keywords: Glaucous gull POPs Heavy metals Helminths
1. Introduction Over the last decades, high levels of persistent organic pollutants (POPs) and some heavy metals (HMs) have been detected in biota from Svalbard area (Daelemans et al., 1992; Knudsen et al., 2007; Savinov et al., 2003, 2005). The levels of ‘‘legacy’’ POP compounds such as dichloro-diphenyl-trichloroethane (DDT), polychlorinated biphenyl (PCB) and oxychlordane are high in the top predators of birds and mammals due to the biomagnification process (Borgå et al., 2001). The levels of ‘‘legacy’’ POPs have however decreased in the last decades due to the ban and restriction in their production and use (Braune et al., 2001; Helgason et al., 2008; Verreault et al., 2005c). Polybrominated flame retardants have been widely used in electrical equipment and other flame retarded goods such as polyurethane foams (Alaee et al., 2003). Most of the use is in urban areas of eastern Asia, North America and western Europe and are potential sources for atmospheric transport to the Arctic * Corresponding author. Tel.: þ47 77 64 50 92; fax: þ47 77 64 55 20. E-mail address:
[email protected] (K. Sagerup). 0269-7491/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.envpol.2009.03.031
(AMAP, 2004). PBDE is found in air, sediments and biota in Arctic and the levels are increasing (de Wit et al., 2006; Law et al., 2003). Even though the penta- and octa-BDEs have been banned in the European Union, Canada and some US states and withdrawn from production by manufacturers, the levels of PBDE are expected to continue to rise in the Arctic (Law et al., 2003). The PBDEs are chemically and structurally similar to PCBs. Potential health effects from PBDE exposure suggest that these new pollutants could have much of the same effects as the PCBs (Darnerud, 2003; Vos et al., 2003). The levels of PBDEs are however still less than 2% of the PCB levels in the glaucous gulls from Bjørnøya (Herzke et al., 2003; Verreault et al., 2005a). Heavy metals occur naturally in soil and biota and concentrations depend on local geology, local addition from mining and industry and/or globally distributed pollution (Pacyna, 2005). Three main anthropogenic sources of HM to the Arctic have been identified; fossil fuel combustion, non-ferrous metal production and waste incineration (Pacyna, 2005). Anthropogenic HMs are transported to the Arctic mainly through the atmosphere (Bard, 1999; Berg et al., 2004; Brooks et al., 2005; Lindberg et al., 2002). Trend data from sediment and ice cores show that HM concentrations
K. Sagerup et al. / Environmental Pollution 157 (2009) 2282–2290
have increased since the start of the industrial revolution (Braune et al., 2005). However, data series for mercury (Hg) and cadmium (Cd) in Greenland marine biota from the late-1970s to the mid-1990s showed no overall temporal trends (Riget and Dietz, 2000). Nevertheless, it is still necessary to monitor HM in Arctic air and biota, since Hg emissions appear to be increasing globally due to more coal burning in Asia (Pacyna, 2005). The highest levels of POPs in the European Arctic are found in the top predators of the marine food chain (Borgå et al., 2005; Johansen et al., 2004; Verreault et al., 2005c). In the European Arctic, the polar bear (Ursus maritimus) and the glaucous gull has received a lot of attention (for review see Gabrielsen (2007)). The apex predatory feeding behaviour of glaucous gull and its low capacity for metabolising organochlorines (Henriksen et al., 2000) makes this species specially susceptible to accumulating high levels of POPs. Its diet includes marine invertebrates and fish and, seasonally, eggs and chicks from other seabirds, carcasses from polar bear hunts and sometimes also adult auks (Alcidae) and kittiwakes (Rissa tridactyla) (Bustnes et al., 2000; Erikstad, 1990; Lydersen et al., 1989). On Bjørnøya, in the Barents Sea, high levels of POPs have been found in the glaucous gull (Borgå et al., 2001; Henriksen et al., 2000; Savinov et al., 2005) with very high levels in sick and dead individuals (Bogan and Bourne, 1972; Gabrielsen et al., 1995). Extensive effect studies of free-living glaucous gulls on Bjørnøya have shown negative effects of contamination on the hormone- and immune-systems, on nesting behaviour and reproduction, on feather development and survival (Bustnes et al., 2002, 2003b, 2004, 2006; Sagerup et al., 2000; Verboven et al., 2008; Verreault et al., 2004). The intestinal helminth fauna of Arctic seabirds includes several species from different classes. However, it is characterised by a low species diversity of trematoda (Kuklin et al., 2004). Several species of cestoda, nematoda and acanthocephalans have also been found in the glaucous gull (Kuklin et al., 2004; Sagerup et al., 2000). The parasites pose a challenge for the host’s well-being and especially their allocation of resources to immune functions (Sheldon and Verhulst, 1996). The effects of pollution on parasitism might be either positive or negative for the host. A positive outcome could arise if the pollutant is fatal for the parasite (Lafferty, 1997) or if it reduces the intermediate hosts’ populations and thereby reduces the infection rate. Conversely, when the pollutant affects the host rather than the parasite, as for example through immunotoxic chemicals, the result could be a suppressed immune system and increased parasites load (Sures, 2006). The hypothesis of the present study is that high levels of contaminants (HMs and POPs) impair the immune system of glaucous gull. Our objective was to test the prediction that parasite intensity would increase with POP/HM concentrations.
2283
were analysed at Typhoon, the Centre of Environmental Chemistry, S.P.A. in Obninsk, Russia. PBDEs, hexabromocyclododecane (HBCD) and toxaphenes were analysed at National Laboratory for Environmental Testing at National Water Research Institute (NWRI) in Burlington, Canada. The liver tissues for the OCP/PCB analyses were homogenized and recovery standards were added. The samples were extracted and percent extractable lipid content (0.1%) was determined gravimetrically by evaporation before concentrated H2SO4 was used to remove lipids. The samples were automatically injected into a gas chromatograph (Hewlett Packard 5790A) equipped with an 63Ni electron capture detector and a 25 m long 0.25 mm HP-1 column. PCBs and pesticides were quantified using certified external standard solutions obtained from the National Laboratory for Environmental Testing in Burlington, Canada. A detailed description of extraction and analyses are given in Muir et al. (2003). PBDEs and toxaphenes were analysed in liver sample extracts and certified reference materials previously extracted for PCB and OCPs. Details of the extraction and analyses are given in Verreault et al. (2005c) and Muir et al. (2006). Analyses of PBDE, HBCD and toxaphene were carried out at NWRI by gas chromatography–low resolution mass spectrometry (GC–MS) on an Agilent 6890 GC-5973 MSD equipped with electron capture negative ionization mode (GC-ECNIMS) and HP5-MS capillary column. Quantification of toxaphene homologues (hexa-, hepta-, octa-, nona-) was carried out using a technical toxaphene standard. Individual chlorobornane congeners, referred to as either Parlar (‘‘P’’) numbers or bornanes (‘‘B’’), were quantified by a series of authentic external standards of each compound obtained from Ehrenstorfer GmbH (Germany). The HMs copper (Cu), Cd, lead (Pb), Hg, selenium (Se) and zinc (Zn) were analysed from the kidney. The analyses were performed at the Department of Environmental Monitoring and Pollution, Obninsk, Russia using atomic absorption spectroscopy (AAS). The kidney samples were digested with nitric acid. Levels of Cd and Pb were measured on Perkin Elmer Z-3030 with Zeeman background correction. Levels of Zn and Cu were measured by AAS on Perkin Elmer B-3030 and those of Hg were measured by cold vapour AAS using the US Environmental Protection Agency (EPA 7471A) method on a Perkin Elmer MAS-50. The extraction of Se was carried out with a mixture of H2SO4 and HClO4, and a conversion of Se to 5-nitro-2,1,3-benzoselendiasol and extraction with chloroform. Selenium was measured by cold vapour AAS on a Perkin Elmer Z-3030 with Zeeman background correction. 2.3. Quality assurance
2. Materials and methods
The method detection limit (MDL) for individual analytes was determined as 10 times the noise level or, in the case where an analyte was present in blanks, three times the standard deviation of the analyte in blanks. Method limits of detection in glaucous gull samples were 0.01 ng/g for toxaphene related compounds, 0.02 ng/g for most PCBs and PBDEs and 0.02–0.05 ng/g for the other chlorinated compounds. The detection limits for HMs and Se vary for each technique, 0.2 mg/g for Cu and Se, 0.1 mg/g for Zn, 0.01 mg/g for Pb, 0.005 mg/g for Hg and 0.001 mg/g for Cd. Blank samples, consisting of all laboratory reagents, were analysed with every 10 samples. The concentration of analysed congeners was consistently less than 10% of the observed concentrations in the liver. Results for individual pesticides and PCB congeners were not blank-corrected. Recoveries of internal standards were good (65–85%) and no recovery corrections were made. One duplicate was analysed at Typhoon, whereas NWRI duplicated the analysis of every fifth sample. The laboratory at Typhoon has national accreditation within the framework of Russian Analytical Laboratories Accreditation System for POPs and Hg in abiotic and biotic samples. Both laboratories had successfully participated in the QUASIMEME (Aberdeen, Scotland) international inter-laboratories comparison. The Typhoon laboratory also participated in the Department of Energy (MAPEP-2001–2002), USA, the NIST/NOAA-NS&T/EPA-EMAP QA Program, the AMAP ring-test (round 1–2 at 2002) and the Second Italian Free Intercalibration Round. The quality control for the trace elements included analysis of blanks, duplicates, matrix spike recoveries and use of certified reference materials (DORM-1 and DOLT-1) from the National Research Council of Canada.
2.1. Sample collection
2.4. Parasite determination
A sample of 20 adult glaucous gulls was collected near Barentsburg, Spitsbergen (78.05 N, 14.2 E) in August 2001. Immediately after capture, the birds were weighed with a spring balance (25 g). Wing length (1 mm; maximum fattened cord), bill depth (0.1 mm) and total head length (0.5 mm; from neck to the tip of bill) were measured. All birds were classified to be at least 4 years old (Cramp, 1977). The liver, kidney and the intestinal tract were sampled and frozen at 20 C until analysis. The sex was determined by gonad inspection. The sampling was in accordance to the current regulation of the Norwegian Animal Welfare Act and permission to collect glaucous gulls was given by the Governor of Svalbard.
The intestinal helminths were identified and quantified at the Murmansk Marine Biological Institute of Russian Academy of Sciences. The intestinal tract was removed from each bird and placed in separate dishes with seawater. The oesophagus and gizzard were cut off from the rest of the intestine. The intestine was cut into segments of about 10 cm. Each segment was cut open and the contents scraped into glasses with mixed sea and fresh water (in proportions 1:1) pre-warmed to 40 C. After 30 min, the contents were transferred to Petri dishes and examined for parasites under a stereomicroscope. The surfaces of the oesophagus, gizzard and ventricle were also examined for parasites under the stereomicroscope. Parasites were fixed in 70% alcohol or 4% formalin in seawater (nematodes). The trematodes, cestodes and acanthocephalans were stained in mucicarmine and prepared as whole mounts before identification. Nematodes were transferred to glycerine and cleared before examination. The prevalence (% infected), mean abundance (number of parasite individuals/n) and mean intensity (number of
2.2. Sample extraction and analysis The liver was analysed for chlorinated and brominated contaminants. Chlorinated pesticides (OCPs), including hexachlorocyclohexanes (HCHs), chlordane related compounds (CHLs), and DDTs, as well as chlorobenzenes (CBs) and the PCBs
2284
K. Sagerup et al. / Environmental Pollution 157 (2009) 2282–2290
parasite individuals/(prevalence/100)) of infection were calculated for each helminth species, following the definitions given by Bush et al. (1997). 2.5. Statistical analyses The concentration mean for a given compound was only determined if 60% or more of the samples had concentrations above the MDL (Verreault et al., 2005c), an arbitrary value used to optimize the trade-off between introducing noise (too low value) or loss of data (too high value). Only those compounds were used in the further statistical analyses. To avoid missing values in the data computation, analytes below the MDL were assigned a randomly generated value between zero and the MDL. The fat contents of the liver were considered similar with a small variance (mean 3.7% 0.6). Therefore, concentrations are presented on a wet weight basis. Natural log transformation (ln) of POP contaminants and parasite intensities resulted in residuals with normal distribution and constant variance. To control for individual difference in size, body mass and sex, a body condition index was calculated using a multiple linear regression model with body mass as the response variable and sex þ bill length þ total head length þ wing length as predictor variables for all individuals (Fox et al., 2007). The body condition index was defined as the residuals of the regression (Jakob et al., 1996). Linear regression models with each of the summarized parasite groups as response variables and each of the POP congeners, their summarized groups, or the HMs and the body condition index as predictor variable were calculated. Adjustment for multiple comparisons was not made, since such adjustment could decrease the chances of rejecting the null hypothesis that are not null (Rothman, 1990). P-values MDL
a-HCH b-HCH g-HCH
0 100 5 100 40 65 35 70 20 20 100 100 95 100 100 100 95 65 95 60 70 55 90 100 100
Oxychlordane trans-chlordane cis-chlordane Heptachlor trans-nonachlor a-Endosulfan b-Endosulfan o,p0 -DDE p,p0 -DDE o,p0 -DDD p,p0 -DDD o,p0 -DDT p,p0 -DDT Endrin Methoxychlor Mirex 1,3,5 tri-CBa 1,2,4 tri-CB 1,2,3 tri-CB 1,2,3,4 tetra-CB Penta-CB HCB a
The summarized levels of pesticides, toxaphenes, PCBs and BDEs are presented in Table 1. The most abundant organic contaminants were oxychlordane, PCB-153, p,p0 -DDE and PCB-118/149 (Tables 2 and 3). The concentrations of the two most abundant pesticides (oxychlordane and p,p0 -DDE) were similar and more than 10 higher than the next highest pesticide, hexachlorobenzene (HCB) (Table 2). A total of 47 PCB congeners were included in the analytical program (Table 3). All of them were found in at least one (PCB-8) sample. The PCB congeners -6, -8, -16/32, -17, -18, -24/27, -31, -49, -22, -33, -64, -70, -84, -87 and -169 all had levels below detection limits for at least 40% of the samples. They were therefore not included in the remaining analyses. The SPCB was calculated for the remaining 32 congeners (Table 1). The five congeners PCB-153>-118/149>-138>-180>-99 were the most abundant congeners (Table 3), accounting for 78% of the SPCB32.
Compound
Table 2 The mean (standard deviation) and data range of pesticides (ng/g wet wt.) in liver of adult glaucous gulls from Barentsburg, Spitsbergen. % > MDL ¼ percentage of sample above method detection limit. n ¼ 20.
0.5 9 183 393 131 706 7
a SCB ¼ sum of 1,3,5-trichlorobenzenes (tri-CB), 1,2,4 tri-CB, 1,2,3 tri-CB, 1,2,3,4 tetra-CB, penta-CB and HCB. b SCHL ¼ sum of cis-chlordane, trans-chlordane, trans-nonachlor, oxychlordane, and heptachlor. c SDDT ¼ sum of o,p0 -DDE, p,p0 -DDE, o,p0 -DDD, p,p0 -DDD, o,p0 -DDT and p,p0 -DDT. d SToxaphene ¼ sum of hexa-, hepta-, octa- and nona-homologues. e SPCB32 ¼ sum of PCB-28, -40, -52, -66/95, -74, -76, -89/101, -97, -99, -132/105, -110, -114, -118/149, -123, -126, -128/167, -137, -138, -141, -151, -153, -171/156, -157, -158, -170, -174, -180, -187, -189, -193, -194, and -196. f SBDE10 ¼ sum of BDE-28/33, -47, -66, -99, -100, -138, -153, -154, -183, and -209. One outlier removed (100 ng/g wet wt.).
Mean SD 1.9 1.5 298.8 329.3 0.2 0.3 0.2 0.3
2.3 270.8 0.4 4.2 2.3 7.5 0.7 1.0 3.9 0.2 0.4
2.0 251.5 0.4 6.1 2.1 5.3 0.6 1.2 3.2 0.2 0.6
1.0 1.2 1.6 1.6 22.2 12.8
Range
