Persistent organochlorine compounds in peregrine falcon (Falco peregrinus) eggs from South Greenland: Levels and temporal changes between 1986 and 2003

Environment International 35 (2009) 336–341

Contents lists available at ScienceDirect

Environment International j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e n v i n t

Persistent organochlorine compounds in peregrine falcon (Falco peregrinus) eggs from South Greenland: Levels and temporal changes between 1986 and 2003 Katrin Vorkamp a,⁎, Marianne Thomsen b, Søren Møller c, Knud Falk d, Peter B. Sørensen b a

Department of Environmental Chemistry and Microbiology, National Environmental Research Institute (NERI), Aarhus University, DK-4000 Roskilde, Denmark Department of Policy Analysis, NERI, Aarhus University, DK-4000 Roskilde, Denmark Roskilde University Library, P.O. Box 258, DK-4000 Roskilde, Denmark d Teglstrupvej 6A, 2100 Copenhagen, Denmark b c

a r t i c l e

i n f o

Article history: Received 22 May 2008 Accepted 15 August 2008 Available online 26 September 2008 Keywords: Arctic DDT Hexachlorobenzene (HCB) Hexachlorocyclohexane (HCH) Migratory birds Peregrine falcon Polychlorinated biphenyls (PCBs) p,p′-DDE Time trend

a b s t r a c t Thirty-seven addled peregrine falcon eggs collected in South Greenland between 1986 and 2003 were analysed for their content of the organochlorine compounds polychlorinated biphenyls (PCBs), dichlorodiphenyl tricloroethane (DDT) and its degradation products, hexachlorocyclohexane (HCH) isomers and hexachlorobenzene (HCB). PCBs and DDT (including metabolites) were by far the most abundant OC groups, with median concentrations of 55 and 40 μg/g lw, respectively. The concentrations were high in an Arctic context, but similar to previously reported levels from Alaska and Norway and slightly lower than concentrations measured in eggs from industrialised regions. Geographical differences may be of importance, considering the migration of peregrine falcons and their prey. ΣHCH and HCB had median concentrations of 0.39 and 0.17 μg/g lw, respectively. On average, DDE accounted for 97% of ΣDDT, but was below critical levels for eggshell thinning. All compound groups showed a weak decreasing trend over the study period, which was statistically significant for HCB and close to being significant for ΣHCH. The weak decrease of ΣPCB and ΣDDT is different from other time trend studies from Greenland, usually showing a more pronounced decrease in the beginning of the study period, followed by a certain stabilisation in recent years. © 2008 Elsevier Ltd. All rights reserved.

1. Introduction The bioaccumulation of organochlorine compounds (OCs) endangered the populations of a top predator, the peregrine falcon (Falco peregrinus), in the northern hemisphere and led to its extinction in the most heavily impacted areas of North America and Europe between the 1950s and 1970s. Reduced reproduction rates have mainly been attributed to high concentrations of the insecticide dichlorodiphenyl trichloroethane (DDT) and its degradation product dichlorodiphenyl dichloroethylene (DDE), polychlorinated biphenyls (PCBs) and structurally similar chlorinated chemicals. They have been found to interfere with the calcium metabolism in the birds, leading to reduced eggshell thickness and abnormal behaviour (Lundholm, 1987; Wegner et al., 2005). Furthermore, direct toxic effects on birds include neurotoxicity and decreased reproduction due to endocrine disruption and embryotoxicity (Ambrose et al., 2000). The peregrine falcon has become a high trophic level key species for monitoring of OCs in the environment. However, only few analyses of OC temporal trends have been conducted (Wegner et al., 2005), especially for Arctic populations of the peregrine falcon (Johnstone et al., 1996; Sørensen et al., 2004). ⁎ Corresponding author. Tel.: +45 46 30 12 00; fax: +45 46 30 11 14. E-mail address: [email protected] (K. Vorkamp). 0160-4120/$ – see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.envint.2008.08.008

This paper analyses the trends in OC levels over an 18 year time span (1986–2003) in the tundra subspecies of the peregrine falcon (F. p. tundrius) in South Greenland. Individuals of this population are likely to be exposed to OCs during their annual movements between the breeding grounds in Greenland (May–September), on the autumn and spring migration through North America and on their wintering grounds in the Caribbean and South America. In Greenland, the birds may be exposed to OCs locally, as the compounds are subject to longdistance transport from their sources to the cold regions of the Arctic. Bioaccumulation and biomagnification of OCs in lipid-rich tissues of Arctic wildlife has been shown in a number of studies (e.g. AMAP, 2004) although concentrations in terrestrial ecosystems have generally been found to be low, in comparison with the more complex food web of the marine ecosystem. The birds of this study nearly exclusively prey on terrestrial passerines during the breeding season in Greenland (Falk et al., 1986; Falk and Møller, 1988). Sources of contamination may be present at their wintering grounds in the Caribbean and South America where phasing out of persistent organochlorine pesticides has generally been slower than in North America and Europe, and renewed use has been deemed necessary to fight malaria. First-year Arctic migrant peregrine falcons have been shown to carry higher OC loads when they return north after their first wintering in South America (Henny et al., 1982).

K. Vorkamp et al. / Environment International 35 (2009) 336–341

The peregrine falcon population in South Greenland has been studied in field surveys since 1981, and addled eggs have been collected since 1986. Recently, information was published on the concentration and long-term trend of brominated flame retardants (BFRs) in peregrine falcon eggs of this population, showing increasing concentrations of some BFRs since 1986 (Vorkamp et al., 2005). Furthermore, the long-term development of eggshell thickness, a proxy for DDT/DDE loads, was studied and indicated gradual, but slow recovery, still far from reaching the natural pre-DDT level (Falk et al., 2006). The present study adds to the existing knowledge by presenting concentrations of OCs in the same eggs which previously were analysed for BFRs. Following up on previous calculations on inter-clutch variation (Vorkamp et al., 2005), the hypothesis of a smaller variation between eggs of the same bird compared with the overall variation was examined, in order to gain some insight into the natural variation among eggs. The compounds included in this study were 22 PCB-congeners, DDT and its metabolites, α-, β- and γ-hexachlorocyclohexane (HCH) and hexachlorobenzene (HCB). The increased awareness of environmental problems stopped the use of most OCs in industrial production processes and agriculture in the 1970s (AMAP, 2004). PCBs, DDT and HCB are listed in the Stockholm Convention of persistent organic pollutants (POPs), which restricts or prohibits the production, trade and use of these compounds because of their persistence, bioaccumulation, long-range transport and adverse health effects. HCH is covered by the Aarhus Protocol on POPs adopted by the United Nations Economic Commission for Europe (UNECE, 1998), with the ultimate objective to eliminate any discharge, emissions and losses of POPs. Despite the regulation of most OCs, they are still ubiquitously present in the environment. This paper addresses the questions whether or not OC concentrations still reach levels known to cause adverse effects in birds and what the concentration development was between 1986 and 2003. 2. Materials and methods

Table 1 Overview of the egg samples analysed in this study ID

Year

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

1986 1987 1988 1988a 1988b 1989 1990 1990 1990 1991 1991 1991 1992 1992 1992 1992 1994 1994 1994 1994 1994 1995 1995 1998 1998 1999 1999 1999 2000 2000 2000 2000 2001 2002 2002 2003 2003

a

2.1. Sampling The eggs originated from peregrine falcons breeding in South Greenland. The population is roughly estimated at 500–1000 pairs (Falk and Møller, 1988). Based on autumn migration counts in the eastern USA, there is a slight evidence for a population increase in the Arctic (Titus and Fuller, 1990). The study area covers the inner parts of the three southernmost municipalities of Southwest Greenland, Nanortalik, Qaqortoq and Narsaq, approx within 60°–61° N and 45°– 46° W. The area is low Arctic, with tundra vegetation and willow and birch shrub in the warm, sub-Arctic areas far from the cool outer coast. The study site was visited annually for field surveys, except in 1993, 1996 and 1997, and addled eggs were collected for contaminant analysis. Several birds are ring marked and could be identified (Table 1). The eggs were frozen on return to Denmark and kept refrigerated until analysis. Prior to analysis, the eggs were weighed and measured, and opened carefully along the egg equator in order to keep two whole egg halves for determination of eggshell thickness. 2.2. Chemical analyses 37 addled eggs collected between 1986 and 2003 were analysed (Table 1). No eggs were collected in 1993, 1996 and 1997 as no field studies were conducted in those years. The analytical method followed the approach previously described for black guillemot eggs (Vorkamp et al., 2004). After homogenisation, about 3.5 g of each egg sample were dried with anhydrous Na2SO4, and spiked with recovery standards (CB3, CB-40, CB-198). In the first sample batch, approximately 6 g of sample was taken. As the OC concentrations were high, the amount of sample was reduced in the following analyses to reduce matrix effects and to keep more of the sample material for other purposes.

337

b

Female ID A A A B B B

C B B B C

D F D E E F E D D E E G G

Lipid (%)

Dry matter (%)

7.0 9.4 18.9 22.4 20.3 10.7 6.3 7.3 6.3 7.5 6.8 5.2 9.3 4.8 6.8 5.6 7.5 3.6 6.7 7.9 5.7 6.5 6.3 5.8 7.6 3.9 6.0 4.0 3.6 2.8 3.9 4.6 6.5 6.9 5.3 7.0 6.0

19.7 29.0 35.3 45.2 38.7 22.4 26.0 21.1 22.9 27.5 19.0 17.5 18.3 23.3 17.8 15.6 22.4 21.1 17.9 20.9 15.6 45.4 19.0 34.2 33.2 21.7 17.3 16.2 13.3 26.2 13.7 17.6 24.6 20.1 15.9 20.4 18.2

No data available for PCBs, HCB, HCHs and DDT. No data available for HCHs.

The samples were Soxhlet extracted using 350 ml of a mixture of n-hexane and acetone (4:1, v/v) and concentrated by rotary evaporation. The extracts were purified on a multilayered glass column packed bottom-to-top with 5 g deactivated aluminium oxide containing 10% water, 1 g activated silica (24 h at 160 °C), 5 g activated silica impregnated with 40% concentrated sulphuric acid and 1 cm anhydrous Na2SO4, and eluted with 250 ml n-hexane. The cleaned extracts were concentrated to about 1 ml by rotary evaporation with iso-octane as keeper and under nitrogen. Defined amounts of the internal standards (CB-53, CB-155) were added and the samples were adjusted to a precise volume of 1 ml. The following individual compounds and congeners were included in the analysis: PCBs: CB-28, CB-31, CB-44, CB-49, CB-52, CB-99, CB101, CB-105, CB-110, CB-118, CB-128, CB-138, CB-149, CB-151, CB-153, CB-156, CB-170, CB-180, CB-187, CB-188, CB-194 and CB-209. DDTs: p, p′-DDT, p,p′-DDE, p,p′-DDD, o,p′-DDT and o,p′-DDE. HCHs: α-HCH, βHCH and γ-HCH. HCB. The lipid content was determined according to Smedes (1999), and the dry matter content was calculated based on mass loss after drying at 105 °C until constant weight (within 2% deviation). The compounds were analysed by dual column gas chromatography with electron capture detection (GC-ECD). The technical details on capillary columns, temperature programmes and calibrations are described by Vorkamp et al. (2004). Samples 4 and 5 (Table 1) could not be analysed by GC-ECD because of matrix interferences. No OC data are available for sample 4, while sample 5 was analysed for PCBs, HCB and DDT by GC-MS in the electron impact (EI) mode, using the same capillary column and temperature

338

K. Vorkamp et al. / Environment International 35 (2009) 336–341

programme as the GC-ECD analysis. However, it was not possible to determine HCH concentrations for sample 5. Each batch consisted of 12 samples, one of which was analysed in duplicate, one procedural blank and two samples of internal reference material (sand eel oil). Precision was monitored by plotting the results of the internal reference material in control charts with warning and action limits (2 and 3 times the standard deviation of the target value, respectively). The accuracy of the analyses was checked by regular participation in QUASIMEME intercalibration exercises on PCBs and OC pesticides in biota (Asmund et al., 2004). The dual column analysis produced two results, which ideally should be identical. For four compounds (CB-28, CB-31, CB-99, o,p′-DDE) only one result was available due to co-elution on the other column. In the usual procedure, an average of the two results was calculated. In case of N10% difference, the lower value was taken, as higher values might reflect matrix interference. Relative retention times were calculated to ensure the correct identification of the compounds. No correction for recovery was performed as only samples with recoveries above 80% were accepted. 2.3. Statistical methods For the calculations of overall time trends, the individual congeners and compounds within one compound group were summed up to total concentrations of each compound group. Values below the limit of detection were assigned a value of half the limit of detection. In accordance with other studies from the Arctic, Σ10-PCB (CB-28, CB-31, CB-52, CB-101, CB-105, CB-118, CB-138, CB-153, CB-156 and CB-180) has been calculated in addition to ΣPCB in order to facilitate comparison to other studies. Linear regression was performed on ln-transformed data in order to meet the assumption of approximate normal distribution of residuals in the regression. Furthermore, pair-wise comparisons of birds were performed by F and t-test on samples from ringed birds to analyse the influence of underlying factors, i.e. year of sampling, geographical differences and physiological parameters, of which only the year of sampling is known (Table 3). However, the heterogeneity and low number of samples did not allow for an ANOVA taking into account the year of sampling. Instead, a pair-wise comparison of birds was performed adopting the opposite assumption that no time trend exists. This may be justified by the presence of no or low significance of the weak decreasing time trends and the fact that the time ranges covered by the individual birds are smaller than the overall sampling period. The t-test algorithm was adjusted according to the outcome of the F-test. This way of challenging the time trend patterns of the individual OC groups can indicate whether or not tendencies for equal variances are more frequent for OC groups with no time trends than in those with a more pronounced time trend. Unequal variances and/or differences between mean contamination levels between birds may reflect an influence of underlying factors; if pronounced for compound groups with no significant time trend, an influence of geographical or physiological parameters is likely to be more important. To test the hypothesis of smaller variances among eggs of the same bird compared with all samples, the total variance and variance among eggs of the same bird have been calculated for, without considering the year of sampling (Table 4). 3. Results and discussion 3.1. Concentrations The median concentrations of the OC groups in all eggs are shown in Fig. 1. The error bars represent the minimum and maximum concentration in the data set. Lipid-based presentations of contaminant concentration in eggs have been discussed in the literature: Peakall and Gilman (1979) observed over 50% changes in lipid levels due to embryo development, in contrast to minor changes in the water content. For this reason, Herzke et al. (2002) also chose wet weight (ww) based concentrations. In this study, however, unfertilised eggs and eggs unhatched after incubation were collected.

Fig. 1. Median concentrations in µg/g lipid weight and µg/g wet weight in 37 peregrine falcon eggs laid between 1986 and 2003. Error bars present the minimum and maximum concentrations. Σ10-PCB: CB-28, CB-31, CB-52, CB-101, CB-105, CB-118, CB138, CB-153, CB-156 and CB-180.

Due to the long storage of the eggs, the water content was clearly reduced, especially in eggs that cracked during freezing. Results on a wet weight basis would be biased, with higher concentrations in the beginning of the sampling period and in the samples with cracked eggshells. Therefore, lipid weight (lw) concentrations were preferred. For reasons of comparability, lipid and dry matter content are given in Table 1 and both units are included in Fig. 1. The concentrations cover a wide range between the different compound groups, but also within one compound group. The median concentration was highest for ΣPCB (54.6 μg/g lw), while Σ10-PCB and ΣDDT had similar median concentrations of 39.9 μg/g lw and 39.7 μg/g lw, respectively. The ΣPCB concentration ranged from 12 μg/g lw to 162 μg/g lw, while the ΣDDT concentration ranged from 9.4 μg/g lw to 173 μg/g lw. The concentrations of HCB and ΣHCH were lower by two orders of magnitude, with median concentrations of 0.37 μg/g lw and 0.17 μg/g lw, respectively (Fig. 1). Compared with results from the marine environment of Greenland, the concentrations in peregrine falcon eggs are strikingly high. Polar bears (Ursus maritimus) from East Greenland, top predators in the marine environment and thus subject to biomagnification of OCs along the marine food chain, had maximum PCB concentrations of approximately 25 μg/g lw (Norstrom et al., 1998). However, the exposure of the Greenland peregrine falcons to POPs is most likely not limited to the long-range transport of these compounds to the Arctic environment. Instead, their migration to the American continent may lead to an exposure of the peregrine falcons close to organochlorine application sites, i.e. in urbanised or agricultural areas or near industrial point sources. Furthermore, peregrine falcons prey on other birds which also migrate to the American continent. Parallel analyses of peregrine falcon and gyrfalcon (Falco rusticolus) plasma from Greenland populations showed considerably lower concentrations in the gyrfalcons, which was attributed to their nonmigratory nature and the low trophic level of their main prey, ptarmigan (Lagopus muta) and hare (Lepus sp.) (Jarman et al., 1994). Compared with other studies on contaminants in peregrine falcon eggs, the concentrations of the Greenland eggs are not unusually high: In fact, the PCB concentrations are slightly lower than found in a Norwegian study (Herzke et al., 2002). Five peregrine falcon eggs from Norway collected between 1991 and 1997 had a maximum PCB concentration of 25 μg/g ww and an average PCB concentration of 9.1 μg/g ww (Herzke et al., 2002). Our data for samples from the same study period were recalculated for the same congeners and converted to the wet weight basis. The recalculation gave a maximum value of 7.7 μg/g ww and a mean concentration of 3.2 μg/g ww for the Greenland samples. With regard to organochlorine pesticides, summed concentrations were comparable to those reported by Herzke et al. (2002). Organochlorines have also been monitored in unhatched, addled eggs of two subspecies from Alaska, American peregrine falcon (Falco peregrinus anatum) and Arctic peregrine falcon (F. p. tundrius) (Ambrose et al., 2000). Like the Greenland peregrine falcons of the same subspecies, birds of the Alaskan F. p. tundrius populations winter in the Southern United States, Central and South America. F. p. tundrius breeds in the tundra Arctic region north of the F. p. anatum which breeds in forested areas from the treeline south to California and Mexico (de March et al., 1998; Ambrose et al., 2000). Geometric means of total PCB were 1.3–2.1 μg/g ww for the two species and time periods within the study period of the Greenland peregrine falcons, i.e. lower than the PCB concentrations of this study (3.6 μg/g ww). The concentration of p,p′-DDE, however, was slightly higher in the Alaskan study, with geometric means between 3.0 and 5.0 μg/g ww, compared to 2.8 μg/g ww in the Greenland study. β-HCH and HCB were comparable in both studies. It is interesting that the Alaskan eggs had a larger range and higher maximum concentrations of these chemicals.

K. Vorkamp et al. / Environment International 35 (2009) 336–341 The origin of the OC contamination was discussed by Johnstone et al. (1996) who analysed peregrine falcon eggs (F. p. tundrius) from the Canadian Arctic and found relatively high concentrations, partly exceeding critical threshold values. All concentrations were higher than those in the Greenland eggs, based on the same study period of 1991–1994 and means of calculations. For instance, ΣPCB geometric mean was 8.3 μg/g ww in the Canadian study compared to 2.5 μg/g ww in this study. DDE exceeded the Greenland concentrations by a factor of 2. The authors concluded that this concentrations difference was due to differences in diet, more precisely, to the breeding grounds of the migratory prey species. While Greenland peregrine falcons mainly feed on passerines, the Canadian population primarily uses more heavily contaminated seabirds and waterfowl (Johnstone et al., 1996). Thus, in addition to the exposure during migration of the peregrine falcons themselves, the exposure of their prey is important in a biomagnification context. The prey species are also exposed to OCs during migration, but according to Johnstone et al. (1996), their food web position is the key factor for OC bioaccumulation. Hence, the Greenland peregrine falcons have a “low-contamination refuge” in the breeding season, which may explain that the population did not suffer from drastic declines as did the Alaskan and Canadian birds of the same subsepecies. Various studies on OCs in peregrine falcon eggs have been conducted in industrialised regions, to study or follow up on population declines or releases of birds into the wild. In general, the samples from Greenland were in the lower end of those from highly industrialised areas. A comprehensive German study summarised data from three regions in Germany, including the former German Democratic Republic, and indicated higher levels than found for the Greenland peregrine falcon eggs during the same study period (Wegner et al., 2005). The concentrations of p,p′-DDE, for instance, ranged between 8 and 62 μg/g dry weight (dw) (mean values for each sampling year) compared to a mean value of 15 μg/g dw in the present study. The difference is larger for PCBs, which ranged from 17 to 175 μg/g dw in the German study, compared to 20 μg/g dw in the Greenland peregrine falcon eggs. Even higher concentrations were reported for eggs collected in the 1970s in France: The ΣPCB and DDE concentrations ranged from 0.47 to 181 μg/g ww and from 0.080 to 56 μg/g ww, respectively (Keck et al., 1982). These maximum concentrations were at least six times above those of the Greenland eggs (Fig. 1). The same tendency was found for HCB, with concentrations between 0.010 and 5.2 μg/g ww in the peregrine falcon eggs from France and a maximum concentration of 0.74 μg/g ww in this study. Apart from the geographical difference, however, a time factor is likely to affect this comparison, as the samples from France had been collected a decade before the eggs from Greenland. Data from Germany showed that a substantial concentration decrease started in the 1970s (Wegner et al., 2005). Concentrations in peregrine falcon eggs from the USA collected in the late 1980s and early 1990s seem to be somewhat lower than the European levels, however, with a large spatial variation: Jarman et al. (1993) studied OCs in peregrine falcon eggs collected between 1986 and 1989 on the East Coast, in Colorado and in California. The geometric means for ΣDDT were 8.8–11 μg/g ww, i.e. higher than the corresponding concentration of 5.3 μg/g ww in our study. ΣDDT in eggs from Western North Carolina collected between 1990 and 1993 ranged from 1.7 to 5.7 μg/g ww and was thus more similar to the results from Greenland (Augspurger and Boynton, 1998). ΣPCB geometric means were significantly different between the US regions, with 13.8, 0.74 and 4.1 μg/g ww for eggs from the East Coast, Colorado and California, respectively (Jarman et al., 1993). Our ΣPCB concentration for the same period of time is 8.0 μg/g ww, i.e. in the higher range of the USA values. Interestingly, the maximum concentrations were very similar, with 25 and 27 μg/g ww in the samples from the US East Coast and Greenland, respectively. 3.2. Critical levels On average, p,p′-DDE accounted for 97.1% of ΣDDT in the peregrine falcon eggs, with a range of 89.9% to 99.7%. The absolute concentrations ranged between 0.69 μg/g ww and 9.1 μg/g ww. DDE threshold values of about 15–20 μg/g ww are considered critical for the reproduction success of the peregrine falcon (Peakall, 1969). Herzke et al. (2002) stated a no observed adverse effect level (NOAEL) for peregrine falcons of 3 μg/g wet weight. All eggs analysed in this study had p,p′-DDE concentration below the DDE threshold, but 42% of the eggs exceeded the NOAEL. Johnstone et al. (1996) cited critical levels of N40 μg/g ww and N4 μg/g ww for ΣPCBs and HCB, respectively. None of the eggs of this study exceeded these levels. Most

Table 2 Results of the linear regression analysis Compound group

Slope

p

Intercept

R2

ΣPCB ΣPCB (without samples 2 and 3) ΣDDT ΣDDT (without samples 2 and 3) p,p′-DDE ΣHCH HCB

−0.0220 −0.00133 −0.0214 −0.0258 −0.0169 −0.0436 −0.0726

0.379 0.958 0.370 0.329 0.470 0.0507 0.0436

54.6 13.4 53.3 61.9 44.3 92.2 151

0.0235 0.0322 0.0244 0.0307 0.0154 0.114 0.118

The summed concentrations were ln transformed to reduce skewness. The p-values express the probability for a slope of an opposite sign than estimated.

339

Fig. 2. A: Temporal trend of ΣDDT, HCB and ΣHCH. The natural logarithm of the ng/g lw concentration is displayed. B: ΣPCB/ΣDDT ratios for all eggs except samples 2 and 3. The fitted lines are the results of linear regression analysis. samples were an order of magnitude below these levels for ΣPCBs and had even lower concentrations of HCB. Other pollutants, however, may contribute to the adverse effects previously observed in peregrine falcon eggs: Wegner et al. (2005) described examples from East Germany of eggs with extremely thin shells but uncritical DDE levels and related this observation to locally elevated Hg concentrations. Time trend analyses of Hg in peregrine feathers from Greenland suggest a modest increase in recent decades compared to pre-1940 samples (Dietz et al., 2006). The Hg content in the eggs from Greenland is not known, but the increasing shell thickness measured for the same eggs does not indicate harmful concentrations (Falk et al., 2006). 3.3. Temporal trends All compound groups have slightly decreasing trends as shown in Table 2. However, the trend was only statistically significant (5% level) for HCB, and close to being significant for HCHs. The steepest decrease of about − 0.07 in the lnC vs. time plot was found for HCB. This corresponds approximately to a 7% decrease in the HCB concentration per year, or to a half life of 10 years. The temporal development of ΣDDT, HCB and ΣHCH is shown in Fig. 2. Apparently, the contamination of the eggs is rather stable over the sampling period of 18 years. This seems to disagree with generally reduced or banned usage of OCs and decreasing concentrations otherwise found in the environment. Other studies from the Arctic have generally observed significant decreases of ΣPCB and ΣDDT. Σ10-PCB and ΣDDT concentrations in ringed seals (Phoca hispida) from Greenland decreased significantly over a similar study period (Rigét et al., 2006; Vorkamp et al., 2008). Studies from the Baltic Sea have shown decreasing concentrations of PCBs and DDTs since the 1970s, for instance in Baltic guillemot (Uria aalge) eggs (Bignert et al., 1995, 1998). Since both compound groups were banned in the 1970s, it seems possible that the main changes in the concentrations of PCBs and DDT in the peregrine falcon eggs occurred prior to our study period. The Baltic Sea time trend study

340

K. Vorkamp et al. / Environment International 35 (2009) 336–341

also described stagnating concentrations since the end of the 1980s and the beginning of the 1990s (Bignert et al., 1998), and the same development can be seen in recent years in the monitoring of marine species from Greenland (Rigét et al., 2004; Vorkamp et al., 2008). Jarman et al. (1994) analysed plasma samples of peregrine falcons from the same South Greenland population, collected from 1984–1986 and 1988–1989. Interestingly, HCB was the only OC compound with a significantly decreasing trend, which agrees with the result of this study. While other OCs did not show any significant trend during the study period, their concentrations were substantially lower than those reported from West Greenland in the 1970s (Walker et al., 1973). The old study had shown DDE and PCB levels above 300 μg/g lw in peregrine falcon eggs, an order of magnitude higher than found in this study. Jarman et al. (1994) used empirical equations to transfer the plasma levels to egg levels and found a similar difference, perhaps even higher. These findings also confirm the hypothesis of the main decrease taking place prior to 1986. The study from Germany also revealed substantial decreases in OC concentrations of peregrine falcon eggs prior to the sampling period of this study (Wegner et al., 2005): Data from Southwest Germany showed main changes in DDE, HCB and PCB between the beginning of the monitoring in 1970s and approximately 1990, since when the concentrations have been stable. Another data series from Germany started with results from 1989–1991 and showed a substantial drop in concentration up to the next data point in 1992. In the following, minor changes have been observed. The results from East Germany, however, did not follow this trend, probably due to the renewed use of DDT in the 1980s. The results from the Canadian Arctic seem to show a somewhat different trend: Data were compared from 1982–1986 and 1991–1994, respectively and showed a significant decrease for DDE, while PCB and HCB remained unchanged (Johnstone et al., 1996).

Table 3 F- and t-probabilities for equal variances and means, respectively, of pair-wise combinations of bird A through G Bird pairs

ΣPCB

ΣHCH

HCB

ΣDDT

F-test probability of the variances between bird pairs not significantly different, P(σ1 = σ2) AB 0.12 0.69 0.10 0.27 AC 0.22 0.80 0.57 0.38 AD 0.71 0.59 0.24 0.69 AE 0.19 0.43 0.41 0.20 AF 0.12 0.29 0.27 0.13 AG 0.36 0.05 0.75 0.62 BC 0.65 0.52 0.05 0.83 BD 0.03 0.93 0.08 0.02 BE 0.45 0.52 0.03 0.68 BF 0.90 0.31 0.48 0.41 BG 0.32 0.03 0.05 0.40 CD 0.07 0.54 0.15 0.26 CE 0.98 0.32 0.20 0.63 CF 0.61 0.22 0.13 0.43 CG 0.70 0.07 0.79 0.66 DE 0.02 0.38 0.61 0.03 DF 0.01 0.23 0.78 0.02 DG 0.23 0.03 0.21 0.64 EF 0.48 0.55 0.53 0.56 EG 0.64 0.02 0.28 0.37 FG 0.39 0.01 0.18 0.25 Students t-test probability of equal mean values of bird pairs, P(µ1 = µ2) AB 0.12 0.04 0.63 AC 0.17 0.01 0.00 AD 0.00 0.04 0.03 AE 0.10 0.02 0.01 AF 0.14 0.08 0.05 AG 0.49 0.00 0.00 BC 0.35 0.17 0.38 BD 0.54 0.54 0.42 BE 0.33 0.16 0.26 BF 0.98 0.19 0.41 BG 0.16 0.02 0.36 CD 0.07 0.07 0.31 CE 0.64 0.75 0.35 CF 0.37 0.60 0.89 CG 0.30 0.02 0.53 DE 0.42 0.05 0.66 DF 0.51 0.08 0.43 DG 0.00 0.01 0.27 EF 0.38 0.74 0.51 EG 0.17 0.27 0.29 FG 0.18 0.56 0.80

0.81 0.06 0.77 0.74 0.87 0.09 0.13 0.83 0.57 0.96 0.41 0.01 0.10 0.35 0.20 0.56 0.84 0.04 0.64 0.29 0.66

Numbers in bold indicate equal variances/means. Underlined numbers indicate different variances/means (10% significance level, two-tailed tests).

Table 4 Overall variation and variation between eggs from the same bird (n: sample size) and probabilities of the variances being smaller than the overall variances (in brackets)

A B C D E F G All

n

ΣPCB

n

ΣHCH

n

HCB

n

ΣDDT

2 3 2 4 5 2 2 21

3.29 (0.18) 3.63 (0.39) 0.00 (0.95) 4.66 (0.02) 4.36 (0.92) 0.01 (0.46) 0.03 (0.60) 2.77

2 3 2 4 5 2 2 21

45.23 (0.39) 5.32 (0.41) 1.68 (0.28) 3.19 (0.25) 2.57 (0.81) 0.04 (0.60) 0.01 (0.01) 12.35

2 3 2 4 5 2 2 21

20.02 (0.22) 6.22 (0.10) 8.45 (0.10) 182.92 (0.52) 59.09 (0.19) 1.44 (0.95) 0.05 (0.15) 67.45

2 3 2 4 5 2 2 21

0.01 (0.28) 0.83 (0.92) 0.20 (0.86) 0.01 (0.06) 0.37 (0.37) 0.72 (0.23) 0.07 (0.51) 0.47

It is interesting that the two main contaminant groups, PCBs and DDTs, have almost identical regression lines (Table 2). It has often been described in the literature that ΣDDT decreases faster in the environment than ΣPCB, as a consequence of ongoing use of PCBs in closed systems beyond the first international regulations. Furthermore, PCBs may be formed by de novo synthesis during e.g. combustion of waste and other thermal processes leading to unintentional releases of PCBs (Thomsen et al., in press). A more rapid decrease of DDT than PCB was observed in time series from the Canadian Arctic beginning in the 1970s and explained by the earlier ban of DDT in North America (Braune et al., 2005). In Baltic guillemot eggs, DDT also started to decrease immediately after the international ban, while PCBs did not decrease before 1975–1977 (Bignert et al., 1998). However, the regression line for ΣPCB is influenced by extremely high PCB concentrations in eggs of bird A, collected in 1987 and 1988: 162 μg/g lw (sample 2) and 145 μg/g lw (sample 3), the maximum concentrations of the entire dataset. If these were removed, the regression line would be rather dissimilar to the ΣDDT trend and reflect the lowest decrease of all compound groups (Table 2). On the other hand, the ΣDDT regression line would remain almost unaffected if these two samples were removed. Consequently, the ΣPCB/ΣDDT ratio would increase with time (Fig. 2) and the findings of a more rapid decrease of ΣDDT compared with ΣPCB could be confirmed. However, these calculations also show how carefully the concentrations in the eggs and their large variations have to be analysed, to avoid misinterpretations. F- and t-tests applied to the concentrations in eggs of ringed birds also seem to support the hypothesis that the ΣDDT and ΣPCB concentrations detected in the eggs are levelling off. Table 3 shows the probabilities for equal variances and means in pair-wise combinations of the eggs of birds A through G. For the F-test, the two-tailed probability, P(σ1 = σ2), is given; i.e. the probability of the variances of two individual birds not to be significantly different. P-values indicating at the P = 0.1 confidence level (two-tailed) that the variances between bird pairs are not significantly different are marked with bold. The alternative hypothesis that differences in variances exist, σ1 ≠ σ2,, is underlined for the significance level P = 0.1. The test indicated different variances for some OC groups and bird combinations suggesting that there are underlying factors causing differences between bird pairs. Besides the year of sampling, underlying factors can be geographical differences and physiological parameters. For the student's t-test, the two-tailed probability of bird pairs to have the same mean (H0 = μa = μb) is given (Table 3). As expected, only few pairs have probabilities below 0.05 confidence level. For ΣDDT and ΣPCB the tendency of equal means is more pronounced than for ΣHCH and HCB: ΣDDT and ΣPCB have two cases of P-values below 0.05, HCB five and ΣHCH has nine cases of P-values below 0.05. This agrees with the findings of most pronounced temporal trends for ΣHCH and HCB, i.e. highest influence of the underlying factor “sampling year”. ΣDDT has the highest frequency of equal means between birds and no time trend, i.e. little influence of the underlying factor “sampling year”. The higher frequency for ΣPCB of different mean values among birds for ΣPCB may indicate other underlying factors, e.g. geographical parameters such as local sources to unintended emissions. 3.4. Individual birds The samples included eggs of five ringed birds (Table 1), partly from the same clutch and partly from different years. To test the hypothesis that the variation among eggs from the same bird is smaller than the overall variation, analysis of variances was performed for the four compound groups (Table 4). For some of the birds, the hypothesis was confirmed, i.e. the variances between eggs of the same bird were smaller than the overall variance among all eggs. However, there were several exceptions: ΣPCB in birds A, B, D and E, ΣHCH in bird A, HCB in bird D and ΣDDT in birds B and F. Bird A stands out as the variance between the two eggs is higher than that of the whole set of samples, with regard to ΣPCB and ΣHCH concentrations. This might be related to the large difference in lipid content of the two samples of bird A (Table 1). For bird D, the ΣPCB and HCB concentrations also vary more strongly than for all samples, although the samples only cover 5 years. Consequently, the hypothesis is only confirmed for birds C and G which have lower variances for all compound groups than the whole sample set. For bird G, this result may be attributed to the fact that the two samples are from 2002 and 2003; i.e. two subsequent years. It has to be noted that the number of samples is generally low and thus, the uncertainty related to the estimation of variation is high.

K. Vorkamp et al. / Environment International 35 (2009) 336–341 Variation among eggs from the same bird has often been discussed in the literature in terms of intra- vs. inter-clutch variation. With regard to eggshell thickness, peregrine falcon eggs from Greenland and Northern Europe were shown to vary equally among clutches and among all eggs of the study (Falk and Møller, 1990). For other birds of prey, e.g. European kestrel (Falco tinnunculus), the variance among clutches was clearly the largest source of variation (Falk and Møller, 1990). With regard to OC concentrations, Johnstone et al. (1996) found a smaller within-clutch variation than between-clutch variation, concluding that single eggs were representative of their respective clutches. On the other hand, Keck et al. (1982) analysed a sequence of eggs from the same clutch and found N 100% difference for most compound groups. They also reported large variations for eggs from the same bird, but different years of sampling. Van den Steen et al. (2006) did not find a laying order effect on organohalogen concentrations in eggs of great tit (Parus major) in a systematic study on intra- vs. intra-clutch variation. The authors explained the low intra-clutch variability with the small home ranges of the birds and low pollution heterogeneity of the area, probably different conditions than those leading to the exposure of peregrine falcons. In summary, most studies seem to indicate smaller intra-clutch variations or variations among eggs from the same bird, but the results in the literature are not conclusive and are likely to be affected by environmental, ecological and physiological factors.

Acknowledgements The authors wish to acknowledge Birgit Groth for the technical assistance. The study was part of the International Arctic Monitoring and Assessment Programme (AMAP). It was funded by the Danish Environmental Protection Agency within the Danish Cooperation for Environment in the Arctic (DANCEA). The content of the article does not necessarily express the view of the Danish Environmental Protection Agency. References AMAP. AMAP Assessment 2002: Persistent Organic Pollutants in the Arctic. Arctic Monitoring and Assessment Programme (AMAP), Oslo, Norway, xvi + 310 pp.; 2004. Ambrose RE, Matz A, Swem T, Bente P. Environmental contaminants in American and Arctic peregrine falcon eggs in Alaska, 1979–1995. Ecological Services Fairbanks, AK, U.S. Fish and Wildlife Service, Technical Report NAES-TR-00-02; 2000. 67 pp. Asmund G, Vorkamp K, Backus S, Comba M. An update of analytical methods, quality assurance and quality control used in the Greenland AMAP programme 1999–2002. Sci Total Environ 2004;331/1–3:233–45. Augspurger R, Boynton A. Organochlorines and mercury in peregrine falcon eggs from Western North Carolina. J Raptor Res 1998;32(3):251–4. Bignert A, Litzén K, Odsjö T, Olsson M, Persson W, Reuthergårdh L. Time-related factors influence the concentrations of sDDT, PCBs and shell parameters in eggs of Baltic guillemot (Uria aalge), 1861–1989. Environ Pollut 1995;89:27–36. Bignert A, Olsson M, Persson W, Jensen S, Zakrisson S, Litzén K, et al. Temporal trends of organochlorines in Northern Europe, 1967–1995. Relation to global fractionation, leakage from sediments and international measures. Environ Pollut 1998;99:177–98. Braune BM, Outridge PM, Fisk AT, Muir DCG, Helm PA, Hobbs K, et al. Persistent organic pollutant and mercury in marine biota of the Canadian Arctic: an overview of spatial and temporal trends. Sci Total Environ 2005;351–352:4-56. de March BGE, de Wit CA, Muir DCG. Persistent organic pollutants. AMAP assessment report: Arctic pollution issues. Arctic Monitoring and Assessment Programme (AMAP), Oslo, Norway; 1998. p. 183–371. Dietz R, Riget FF, Boertmann D, Sonne C, Olsen M, Fjeldså J, et al. Time trends of mercury in feathers of West Greenland birds of prey during 1851–2003. Environ Sci Technol 2006;40:5911–6. Falk K, Møller S. Status of the peregrine falcon in South Greenland: population density and reproduction. In: Cade TJ, Enderson JH, Thelander CG, White CM, editors. Peregrine falcon populations — their management and recovery. Boise, Idaho: The Peregrine Fund; 1988. p. 37–53. Falk K, Møller S. Clutch size effects on eggshell thickness in the peregrine falcon and European kestrel. Ornis Scand 1990;21:265–9. Falk K, Møller S, Burnham WG. The peregrine falcon (Falco peregrinus) in South Greenland: nesting requirements, phenology and prey selection. Dan Ornithol Foren Tidsskr 1986;80:113–20.

341

Falk K, Møller S, Mattox WG. A long-term increase in eggshell thickness of Greenlandic peregrine falcons Falco peregrinus tundrius. Sci Total Environ 2006;355:127–34. Henny CJ, Ward FP, Riddle KE, Prouty RM. Migratory peregrine falcons, Falco peregrinus, accumulate pesticides in Latin America during winter. Can Field Nat 1982;96:333–8. Herzke D, Kallenborn R, Nygård T. Organochlorines in egg samples from Norwegian birds of prey: congener-, isomer and enantiomer specific considerations. Sci Total Environ 2002;291:59–71. Jarman WM, Norstrom RJ, Simon M, Burns SA, Bacon CA, Simoneit BRT. Organochlorines, including chlordane compounds and their metabolites, in peregrine falcon, prairie-falcon and clapper-rail eggs from the USA. Environ Pollut 1993;81:127–36. Jarman WM, Burns SA, Mattox WG, Seegar WS. Organochlorine compounds in the plasma of peregrine falcons and gyrfalcons nesting in Greenland. Arctic 1994;47(4):334–40. Johnstone RM, Gordon SC, Fesser AC, Bradley DM, Oliphant LW, MacNeil JD. Long-term trends and sources of organochlorine contamination in Canadian tundra peregrine falcons, Falco peregrinus tundrius. Environ Pollut 1996;93(2):109–20. Keck G, Paubel P, Monneret RJ. Organochlorine and mercury residues in peregrine falcon eggs from France. Bull Environ Contam Toxicol 1982;28:705–9. Lundholm E. Thinning of eggshells in birds by DDE: mode of action on the eggshell gland. Comp Biochem Physiol 1987;88C:1-22. Norstrom RJ, Belikov SE, Born EW, Garner GW, Malone B, Olpinski S, et al. Chlorinated hydrocarbon contaminants in polar bears from eastern Russia, North America, Greenland, and Svalbard: biomonitoring of Arctic pollution. Arch Environ Contam Toxicol 1998;35(2):354–67. Peakall DB. Effects of DDT on calcium uptake and vitamin D metabolism in birds. Nature 1969;224:1219–20. Peakall DB, Gilman AP. Limitations of expressing organochlorine levels in eggs on a lipid-weight basis. Bull Environ Contam Toxicol 1979;23:287–90. Rigét F, Dietz R, Vorkamp K, Johansen P, Muir D. Levels and spatial and temporal trends of contaminants in Greenland biota: an updated review. Sci Total Environ 2004;331:29–52. Rigét F, Vorkamp K, Dietz R, Rastogi SC. Temporal trend studies on polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls (PCBs) in ringed seals from East Greenland. J Environ Monit 2006;8:1000–5. Smedes F. Determination of total lipid using non-chlorinated solvents. Analyst 1999;124:1711–8. Sørensen PB, Vorkamp K, Thomsen M, Falk K, Møller S. Persistent organic pollutants (POPs) in the Greenland environment: long-term temporal changes and effects on eggs of bird of prey. NERI Technical report no 509. Denmark: National Environmental Research Institute; 2004. 126 pp., http://technical-reports.dmu.dk. Thomsen, M.; Nielsen, O.–K.; Illerup, J.B. Unintentional formation and emission of the persistent organic pollutants HCB and PCBs in the Nordic countries. Documentation of existing information regarding sources and emission to soil, water and air with focus on reporting according to the Stockholm Convention, the UNECE POP protocol and PRTP registers. TemaNord, in press. Titus K, Fuller MR. Recent trends in counts of migrant hawks from northeastern North America. J Wildl Manage 1990;54:463–70. UNECE (United Nations Economic Commission for Europe). 1998 Aarhus protocol on persistent organic pollutants. Protocol to the 1979 Geneva convention on long–range transboundary air pollution; 1998. http://www.unece.org/env/lrtap/pops_h1.htm. Van den Steen E, Dauwe T, Covaci A, Jaspers VLB, Pinxten R, Eens M. Within- and among-clutch variation of organohalogenated contaminants in eggs of great tits (Parus major). Environ Pollut 2006;144:355–9. Vorkamp K, Christensen JH, Glasius M, Riget FF. Persistent halogenated compounds in black guillemots (Cepphus grylle) from Greenland — levels, patterns and spatial trends. Mar Pollut Bull 2004;48:111–21. Vorkamp K, Thomsen M, Falk K, Leslie H, Møller S, Sørensen PB. Temporal development of brominated flame retardants in peregrine falcon (Falco peregrinus) eggs from South Greenland (1986–2003). Environ Sci Technol 2005;39:8199–206. Vorkamp K, Rigét FF, Glasius M, Muir DCG, Dietz R. Levels and trends of persistent organic pollutants in ringed seal (Phoca hispida) from Central West Greenland, with particular focus on polybrominated diphenyl ethers (PBDEs). Environ Int 2008;34:499–508. Walker W, Mattox WG, Risebrough RW. Pollutant and shell thickness determinations of peregrine eggs from West Greenland. Arctic 1973;26:256–8. Wegner P, Kleinstäuber G, Baum F, Schilling F. Long-term investigation of the degree of exposure of German peregrine falcons (Falco peregrinus) to damaging chemicals from the environment. J Ornithol 2005;146:34–54.