Mercury, methylmercury, and selenium in blood of bird ... - Springer Link

Environ Sci Pollut Res (2013) 20:5361–5372 DOI 10.1007/s11356-013-1540-1

RESEARCH ARTICLE

Mercury, methylmercury, and selenium in blood of bird species from Doñana National Park (Southwestern Spain) after a mining accident C. Rodríguez Alvárez & M. Jiménez Moreno & L. López Alonso & B. Gómara & F. J. Guzmán Bernardo & R. C. Rodríguez Martín-Doimeadios & M. J. González Received: 27 July 2012 / Accepted: 3 February 2013 / Published online: 14 February 2013 # Springer-Verlag Berlin Heidelberg 2013

Abstract Total mercury (Hg), monomethylmercury (MeHg), and selenium (Se) were determined in blood of 11 bird species living in Doñana National Park (DNP, Southwestern Spain) and the surrounding area in 1999 and 2000 after a mine spill accident. The total Hg contents found varied from 1.00 to 587 ng/mL, with an MeHg percentage higher than 80 %, except in mallard species. In all the cases, the concentrations found were below the threshold of high risk for the bird populations. The parameters which most affected the accumulation of Hg and MeHg in the birds studied were, first, species, or trophic position, and second sampling area. Age does not seem to have a great influence on the content of Hg in the blood of these birds. The levels Responsible editor: Vera Slaveykova Electronic supplementary material The online version of this article (doi:10.1007/s11356-013-1540-1) contains supplementary material, which is available to authorized users. Highlights: 1. The concentrations of Hg in blood strongly depend on the bird species. 2. The study of blood avoids sacrificing birds and enables the monitoring of recent dietary intake and short term exposure to Hg and Se. 3. The Hg and Se concentrations found in the blood of the bird species studied are far below the toxic values reported in the literature. 4. The 1998 spill accident could have an influence in the Hg blood of bird living in the area directly affected by the toxic mud. C. R. Alvárez : M. J. Moreno : F. J. G. Bernardo : R. C. R. Martín-Doimeadios Faculty of Environmental Sciences and Biochemistry, University of Castilla-La Mancha (UCLM), Avda. Carlos III s/n, 45071 Toledo, Spain L. L. Alonso : B. Gómara : M. J. González (*) Department of Instrumental Analysis and Environmental Chemistry, Institute of General Organic Chemistry (IQOG-CSIC), Juan de la Cierva 3, 28006 Madrid, Spain e-mail: [email protected]

of Se found ranged from 108 to 873 ng/mL, and they were not affected by species, trophic level, age, or sampling area. The blood Hg concentrations of birds living in the area directly affected by the toxic mud, outside the park, were higher than those found in the other birds, and this could be explained by the mine spill accident happened in 1998. Keywords Doñana National Park . Bird species . Blood . Mercury . Selenium . Acid mine spill

Introduction Wild bird populations are at risk due to the presence of toxic elements in the environment, particularly nondegradable elements, which often tend to concentrate throughout the food chain. Monitoring of such substances in selected bird species from a delimited area is useful not only to evaluate health of the species involved, but also to assess the degree of contamination of the ecosystems where they live. Mercury (Hg) has been a contaminant of concern because it accumulates in the tissues of wildlife species and can adversely affect reproduction, especially in higher trophic level species (Scheuhammer et al. 2007; Wiener et al. 2002; Wolfe et al. 1998). Mercury can occur in different chemical and physical forms. The most important are inorganic Hg (IHg) and monomethylmercury (MeHg), which show different kinetics and toxicology (Mason and Benoit 2003). In general, the occurrence of Hg in the organism is due to diet. Apart from dietary intake, the Hg content in blood reflects physiological influences, such as mobilization and storage in feathers and during egg-laying period. Monomethylmercury (MeHg), the organic form in which Hg most often occurs in aquatic wild birds, is of special interest because it is toxic, it has negative effects on reproduction, it is responsible for egg hatchability and neurobehavioral

5362

underdevelopment (Burger and Gochfeld 2000; Ohlendorf et al. 1989), and finally, because it biomagnifies through the food web (Burger 2002). Selenium (Se) has been much less studied in birds. Small amounts of Se are essential for health, but it becomes toxic at high concentrations and causes low survival rates of chicks and adults (Ohlendorf et al. 1989; Ohlendorf 2003; Spallholz and Hoffman 2002). Selenium occurs in birds through diet, and it is transferred through blood to eggs, as this happens for Hg (Evers et al. 2003; Lewis et al. 1993). Moreover, Se is known to be very active at counteracting Hg toxicity (Cuvin-Aralar and Furness 1991; Yang et al. 2008). It is well documented that Se contributes to the detoxification process of Hg through its interaction with IHg in liver and kidney, by formation of a stable adduct which protects birds from Hg toxicity (Kim et al. 1996a; Ohlendorf 2003; Scheuhammer et al. 2008). Recent investigations have concluded that the mechanism of Hg toxicity could be based on the inhibition of selenoenzymes, which play a role to prevent and reverse oxidative damage through the body. Thus, their inhibition would be the reason for oxidative stress damage and subsequent effects (Carvalho et al. 2008; Raltson et al. 2008, 2010). Several bird tissues have been used for monitoring avian exposure and assessing risk, particularly feathers and liver, (Kim et al. 1996b) and more recently, blood (Bearhop et al. 2000a; Burger and Gochfeld 1997; Cristol et al. 2008). At present, feathers, blood, and eggs are preferred to work with because they can be obtained easily, repeatedly from the same individual if required, and without sacrificing. Eggs give information about Hg and Se intake during a short period before the egg is laid (Burger et al. 2008; Conover and Vest 2009; Lewis et al. 1993), feathers represent the Hg and Se burden at the time of feather growth because both elements come to them in a dose-dependent way (Becker et al. 1993; Burger et al. 2008), and blood reflects the current Hg and Se burden, which is mainly influenced by dietary uptake (Scheuhammer et al. 1998a, b). The main advantage of monitoring Hg and Se in blood is that this provides a picture of short term, and a real-time exposure to both of them throughout the year (Burger et al. 2008; Kahle and Becker 1999). Moreover, blood is often preferred for assessing Hg exposure because Hg blood levels strongly correlate with Hg concentrations in other internal tissues (Kenow et al. 2007), and also because MeHg is largely the dominant species (Rimmer et al. 2005). The Doñana National Park (DNP) is a protected nature reserve which was designed as World Heritage Site and Biosphere Reserve by UNESCO in 1994. It is a wild life sanctuary or refuge for thousands of both sedentary and migratory birds which nest and, in some cases, reside there temporarily. The park is located in the west bank of the Guadalquivir River delta, Southwestern Spain (Fig. 1). One of the most important water-

Environ Sci Pollut Res (2013) 20:5361–5372

Fig. 1 Geographic locations of the sampling areas where the blood of birds was collected. 1 Location reached by toxic mud from the pyrite mine accident. 2 Location reached by the acid water from the pyrite mine accident

supplies for the survival of the park, the Guadiamar River, flows through an area where a pyrite mine is located, about 40 km north of the Park. Both rivers are the main sources of Hg in DNP (González et al. 1985, 1990). This remarkable ecosystem has remained relatively unspoiled for centuries thanks to


the absence of permanent settlements, but it has always suffered from the impact of human activities in neighboring areas even though it was given official protection in 1969. In this way, pyrite mining in nearby Aznalcóllar, an area with high contents of heavy metals and metalloids, is a constant risk for DNP, as reported in the literature (Fernández-Aceytuno et al. 1984; Rico et al. 1989). In fact, on 25th April 1998, the sudden avalanche of water and sludge produced by the collapse of the wall of the dam storing the wastes from the Aznalcóllar mine caused the bursting of the neighboring rivers Agrio and Guadiamar, flooding about 4,500 ha of adjacent land (Fig. 1). The amount of sludge deposited in the Guadiamar basin has been estimated to be around 5 million cubic meters. The contaminated sludge waste contained 0.5 % of arsenic, 0.8 % of lead and zinc, 0.2 % of copper, 0.007 % of cadmium, 0.0015 % of Hg, and 0.0011 % of Se on a dry weight basis (Alastuey et al. 1999; Grimalt et al. 1999). Part of these metals ended up in soils around DNP, and as a consequence, they were likely to get into the DNP’s food web (Benito et al. 1999; Hernández et al. 1999; López-Pamo et al. 1999). Despite the magnitude of the accident and the fact that relevant ecosystems were affected, there are scarce data on the effects of Hg and Se contamination from the toxic spill at the Aznalcóllar mine on living organisms in the DNP and the surrounding area. We present here the levels of total Hg, MeHg and total Se in the blood of 11 bird species (n=113) living in the DNP, collected in 1999 and 2000. Birds were from different trophic food webs, but most of them were related to the aquatic ecosystem. The suitability of blood samples for biomonitoring studies and biomagnification processes and the grade of exposure to Hg and Se were studied, as well as the influence of the toxic spill accident happened in the area over the following two years. The correlations of total Hg, total Se, and MeHg with species, trophic levels, sites, and ages were statistically studied. The levels of Hg found in the blood of bird species were compared with those obtained from similar species living in the area between 1979 and 1983. The results obtained in this work were also compared with those corresponding to similar species from other ecosystems.

Material and methods Study area and sampling zones Samples were collected in and around Doñana National Park (Southwestern Spain; 36o48′–37o20′N, 6o12′–6o40′ W) during 1999 and 2000. The sampling area was divided in five zones (Fig. 1). Zone I is located east of DNP, close to the Guadalquivir river mouth. Zone II is located north of Zone I, Zone III is the northern area of DNP, in the marshes area where most of the studied birds live. Zone IV is located in the south-west of DNP, and Zone V is located north of the Zone II, outside the park.

5363

Zone V was directly affected by the spill accident and it was partially covered by the toxic mud, whereas Zone II received the impact of the acid water which flooded the land close to DNP. The western boundary of the DNP is the Atlantic Ocean and the Guadalquivir River is the boundary of the Park in the southeast. The work was done in the frame of the wildlife monitoring program of Aznalcóllar spill, which was supervised by the Doñana Biological Station Institute (CSIC) and officially funded by the government of the Spanish autonomous community of the Junta de Andalucia, the CSIC and the Spanish government. Sampling A total of 113 blood samples from 11 species of birds living in DNP were sampled. Adults and young individuals were captured using bird traps (peregrine falcon), cannon nets (black kite), and funnel traps (waterfowl species) and chicks were captured in nests. Approximately 2 mL of blood was taken from the brachial vein and placed in lithium-heparinized vials. They were transported in coolers to the laboratory on the same day of collection and kept frozen at −80 °C. The main characteristics of the studied species, which can have an influence in the Hg and Se levels found in their blood, are shown in Table 1. Fifty individuals were sampled in zone I, 20 in zone II, 17 in zone III, 3 in zone IV, and 23 in zone V. Most birds are related to aquatic systems. These are from the families of the Anatidae (mallard and shoveler), Rallidae (coot), Ardeidae (cattle egret, gray heron, and purple heron) Threskiornithidae (glossy ibis), Ciconiidae (white stork), and Accipitridae (marsh harrier). Peregrine falcon and black kite are from Falconidae and Accipitridae family, respectively, and they are related to terrestrial ecosystems. Some of these species are resident in the area (peregrine falcon and coot); others are migratory to the north of Europe (shoveler), south eastern Europe (glossy ibis), or northern Africa (black kite and purple heron). The rest species include either resident and or migratory individuals at different percentages. Thus, 50 % of the marsh harrier, mallard and cattle egret are resident whereas this percentage increases up to 90 % in the case of gray heron. Finally, only a few colonies of white stork are resident in the area. According to the type of food (Table 1), four trophic levels were considered: trophic level 1 for birds consuming aquatic plants and insects; trophic level 2 for birds consuming aquatic plants, insects and crayfish; trophic level 3 for birds consuming insects, crayfish, reptiles and mainly fish; trophic level 4 for consuming small birds and mammals, and eventually fish (Cramp and Simmons 1978, 1980). Analytical procedure All chemicals were ultrapure grade for high quality Hg analysis. Deionized water (18 MΩ.cm) obtained with

5364


Table 1 Main characteristics of the bird species from DNP studied in this work Common name

Scientific name

Migratory habits

Trophic level

Sampling zone and (size)

Sampling period

Marsh harrier (A*) Peregrine falcon (T*) Black kite (T) White stork (A) Cattle egret (A) Grey Heron (A) Purple heron (A) Glossy ibis (A) Coot (A)

Circus aeruginosus Falco peregrinus Milvus migrans Ciconia ciconia Bubulcus ibis

50 % migratory (North Europe) 50 % residents (DNP)

4

I (15)

May–June (R) (Ch)

Residents (DNP)

4 4

Migratory (North Africa). Some colonies are residents (DNP) 50 % migratory (North Africa) 50 % residents (DNP) 10 % migratory (North Africa) 90 % residents (DNP) Migratory (North Africa)

3

I (3), II (4)

March (Ad)/ April (Ch) April (Ad)/ June (Ch) June (R) (Ch)

May–June/July

Migratory (North Africa)

II (2), III (1), IV (3) III (14), V (1)

3

V (6)

March (Ad)

April–June/July– November

3

II (4)

August (Ch)

3

I (8)

July (Ch)

Migratory (South East of Europe)

2

I (14)

Residents (DNP)

1

I (7), II (1)

May–June (Ch) July (Ad)

Shoveler (A)

Anas clypeata Anas platyrhinchos

Migratory (North Europe)

2

V

50 % migratory (North Europe) 50 % residents (DNP)

1

I (3), II (9), III (1), V (2)

Mallard (A)

Ardea cinerea Ardea purpurea Plegadis falcinellus Fulica atra

January (Ad) (J) March(J)/ June (Ad) (R)

Egg-laying 1/molting primary feathers 1

April–June/June–July

February–July/July– August May–August/June– January April–June/May– January

A aquatic system, T terrestrial system, Ad adults, Ch chicks, J juveniles, R residents a

According to Cramp and Simmon, 1978, 1980 ; * A = aquatic system; T = terrestrial system; Ad = Adults; Ch = Chicks; J = Juveniles; R = Residents

Milli-Q water system (Millipore Inc., Millipore Ibérica, Spain) was used for reagents and standards. All glassware, polyethylene, and disposable material were treated with 10 % nitric acid for 2 days and rinsed with Milli-Q water before use. Total mercury analysis Total Hg was analyzed by cold vapor atomic absorption spectrometry (CV-AAS) using a flow injection Hg system (FIMS-400, Perkin Elmer Hispania, Madrid, Spain) equipped with an auto sampler AS-900, using the method described by López-Colon et al. (2001). The method was based on an adequate sample digestion, using 1 mL of 0.2 % of Triton, 0.1 mL of stabilizing solution (0.5 % K2Cr2O7; 50 % (v/v) HNO3), 1 mL of 10 % (v/v) HNO3, and 2 mL of 20 % (v/v) H2SO4. The mixture stands overnight at 60 °C. Then it was cooled to room temperature, and diluted to 10 mL. The final Hg determination was carried out by automated flow injection cold vapor reduction with NaBH4. Standards containing 1 to 10 ng/mL of Hg were analyzed, with correlations higher than 0.995 in every working session. Analyses were carried out by duplicate. Also, two procedural blanks, calibration standards with four known concentrations of Hg, and the certified reference material

SeronormTM Trace Elements Whole Blood Level 2 (Batch MR9067; SERO AS, Billingstad, Norway) were run. Recoveries of this reference material, whose certified concentration of mercury is 8.2 ng/mL, ranged from 88 % to 110 %. Relative standard deviation (RSD) in replicates and reference materials was always under 10 %. The method was validated by the participation in the Interlaboratory Comparison Program for metals (including Hg) in blood samples (Institute National de Santé Publique du Québec, Canada). The results were consistent with the consensus means given by the interlaboratory organization in all the cases. As criteria of acceptability, all z scores obtained were satisfactory (z score ≤2). The limits of detection (LOD) and quantification (LOQ) of the methodology used were 0.12 and 0.4 ng/mL, respectively. Methylmercury analysis MeHg was analyzed using a gas chromatograph (Varian 3900) coupled to an inductively coupled plasma mass spectrometer (Thermo Electron Model XSeries II; GC-ICP-MS). The GC column was a Cross bond, 100 % dimethyl polysiloxane, of 30 m × 0.53 mm id.; film thickness 1 μm (Restek, USA) and it was connected to the torch of the ICP-MS using a heated (170 °C) transfer line. The use of


the three legged X Series ICP-MS torch allows the simultaneous connection of a nebulizer/impact bead spray chamber for the introduction of liquids. The instrumental configuration allows the continuous aspiration of tuning or internal standard solutions, here thallium in a concentration of 10 ng/mL, while operating in the GC-ICP-MS mode (Berzas Nevado et al. 2011). Mercury species were extracted from blood by closedvessel microwave assisted extraction following a previously optimized procedure (Rodrigues et al. 2011). Briefly, 2.0 mL of tetramethylammonium hydroxide (TMAH) was added to 0.3 mL of blood in a microwave oven flask. The final volume was adjusted to 10 mL with ultrapure water for microwave requirements. Microwave vessels were sealed and irradiated for 10 min at 180 °C after a temperature ramp of 10 min, and as a result, a clear solution was obtained. Then the vessels were cooled down to room temperature, made-up to a known volume and stored in the cold room until analyzed. Blanks were prepared along with the samples in each batch. Volumes of 2 mL of the alkaline extracts were used for derivatization by ethylation. The pH was adjusted to 3.9 using concentrated acetic acid and 5 mL of 0.1 M acetic acid-sodium acetate buffer. Then, 5 mL of sodium tetraethylborate (0.3 %; w/v) and 2 mL of hexane were added and the mixture was manually shaken for 5 min All this was centrifuged for 5 min at 600×g and the resulting organic layer was transferred to a glass vial and stored at 4 ° C. Three sub-samples were prepared from this solution. Blanks were subjected to the same procedure. An aliquot of 1 μL was analyzed by GC-ICP-MS system. The procedural LOD was better than 0.5 ng/mL and the precision in terms of RSD was 3.8 %. The method was validated by the analysis of Standard Reference Material (SRM) 966 Toxic Metal Bovine Blood from the National Institute of Standards and Technology (NIST). The obtained concentrations of the Hg species were not statistically different from the SRM (p=0.05). Additional validation was carried out by the analysis of a human whole blood as secondary reference material provided by the New York Department of Health’s PT program (NYS 0706). In this case, the sum of the Hg species found (34.5±3.0 ng/mL) was in the range of the total Hg target value for this material (22.5–41.9 ng/mL). Selenium analysis Total Se analyses were carried out with an inductively coupled plasma mass spectrometer (ICP-MS) equipped with a collision cell (CCT; Thermo Electron Model XSeries II). Because of the interference of many argon plasma-derived polyatomic ions, the instrument was operated in CCT mode (using H2/He as collision gas) for measuring 78Se. Solutions used for calibration were prepared from commercial certified stock standards with 1,000 ng/mL of Se. Total Se was

5365

microwave assisted extracted using a previously optimized procedure with 0.3 mL of blood and 2.0 mL of TMAH (Rodrigues et al. 2011). The LOD of this method was 1.3 ng/mL and the RSD was 3.2 %. The method was validated by the analysis of the certified reference material SeronormTM Trace Elements Whole Blood Level 2 (Batch MR9067; SERO AS, Billingstad, Norway), with recoveries from 96 % up to 110 %. Statistical analysis In most cases, the distribution of data was highly skewed, the variables did not fit a normal distribution, and this is why non-parametric tests were used for statistical comparisons. The data set was analyzed by non-parametric Kruskal–Wallis (χ2) to determine significant differences of Hg and Se blood levels among species, ages and trophic levels. For statistical comparison, the non-detectable (ND) concentrations were taken as half the LOD. Differences with p0.05). Age variations The content of Hg and Se found in blood of the four bird species with individuals of different ages available were compared using Kruskal–Wallis test (χ2). The geometric means of total Hg concentrations found in adults of peregrine falcon (567 ng/mL) and black kite (99.4 ng/mL) were much higher than those found in chicks of either species (71.9 and 54.8 ng/mL, respectively). Nonetheless, only in the case of peregrine falcon, these differences were statistically significant (p0.05) between adults and juveniles of shoveler (geometric mean 140 and 165 ng/mL, respectively) and mallard (geometric mean 12.2 and 10.7 ng/mL, respectively). The fact that mean Hg concentrations only increased or decreased a little from juveniles to adults, and that the differences were not statistically significant could be due to a relatively large Hg mobilization, from the body to feathers or eggs. In fact, peregrine falcon was sampled just before the egg production and molting

5367

primary feathers (Table 1) as it was stated previously. These results agree with those obtained by some authors which found that accumulation of Hg with growth in different tissues of bird species may not be important in some birds since much of the body burden of Hg is removed through molting and egg production (Honda et al. 1986; Saeki et al. 2000). Unfortunately, we cannot conclude whether molting, egg production, or both were responsible for the present results because, as previously mentioned, no information about sex of the studied specimens was available. In the case of Se, there were not significant age-related differences (p>0.05) in any of the four species with individuals of different ages. Previous studies on marine bird species (emperor geese, Chen canagica, and spectacled eider, Somateria fischeri) showed that Se concentrations in blood of chicks were lower than in adults, due to feed sources. Thus, adults were exposed to high Se levels in wintering and staging areas and then to low Se levels on their breeding grounds, while chicks were only exposed to low Se concentrations (Franson et al. 1999, 2002; Grand et al. 2002; Wilson et al. 2004). Finally, there were not significant differences in the concentration of Se in blood between adults and chicks or juveniles of the four bird species collected in Doñana National Park, which could be due to a similar exposure to Se in wintering and in breeding areas. Site differences The number of individuals of each species in the different sampling areas is shown in Table 1. The multiple sample comparison of Hg and Se found in blood of individuals from the five sampling areas using Kruskal–Wallis test (χ2) showed that there were statistically significant differences among the sampling areas in the case of Hg (p0.05). When Hg levels in different sampling areas were compared, only zone V was statistically different from zones I and III. The sampling zone V showed the highest geometric mean values (120 ng/mL; range 1.0–501 ng/mL), followed by zone III (84.3 ng/mL; range 14.7–587 ng/mL), zone IV (71.9; range 59.5– 88.6 ng/mL), and finally zone II (51.2; range 2.77– 567 ng/mL) and zone I (46.3; range 2.09–465 ng/mL). It is precisely zone V that was directly affected by the mine accident occurred in 1998, being partially covered by the toxic mud. The toxic mud was removed using mechanical and manual extraction works, and later, the soils received chemical treatment using various adsorbents to decrease the solubility and bioavailability of metals and metalloids (Grimalt et al. 1999). The results obtained two years after the spill accident, suggested that Hg from the mine spill are still affecting the birds living in this zone. However, zone II, which was only affected by the acid waste flood, seems to be recovered because the levels of Hg in the blood of birds

5368


that inhabit this zone and also zone I further south, showed the lowest levels of Hg. It was not possible to compare the levels of blood Hg found in bird species (peregrine falcon, black kite, coot, mallard and white stork) sampled in different areas because the number of individuals of each species collected in each zone was not enough to carry out a statistical comparison. The concentrations of Se in blood were not statistically different among sampling zones. In this case, birds from zone II showed the highest geometric mean (493 ng/mL; range 322–696 ng/mL), followed by zone I (397 ng/mL; range 132–873 ng/mL), zone V (395 ng/mL; range 108– 846 ng/mL), zone IV (385 ng/mL; range 296–501 ng/mL) and finally zone III (319 ng/mL; ranged 271–416 ng/mL). Se levels found in the blood of birds that inhabit the most affected areas by the spill (zones II and V) do not seem to be affected by the mining accident occurred in 1998. Trophic level variations and bioaccumulation processes This section is focused on aquatic bird species, where Hg is mainly accumulated so black kite and peregrine falcon species have been ignored. For this study, adult and juveniles of the rest of species have been used because Hg levels were no statistically significant with age. Therefore mallard and coot species were integrated in trophic level 1, shoveler species in trophic level 2, and cattle egret in trophic level 3. The geometric mean of total Hg and Se concentrations found in the three different consumers according to their food habits (Table 1), as well as the concentration factor (CF) between trophic levels, obtained as the ratio of the mean concentration in an upper and in a lower trophic level, are shown in Table 3. The Hg concentration in the blood of birds increased with bird trophic level. Mercury concentrations in the third trophic level (297 ng/mL) were clearly higher than in the second (151 ng/mL) and first (8.37 ng/mL) ones. The concentration factor between the two first trophic levels (CF =18) was quite high, nearly ten-fold the CF between trophic levels 3 and 2 (CF=1.96). This indicates evident Hg bioaccumulation processes through the trophic levels of the aquatic birds investigated. The bioaccumulation processes of Hg and MeHg throughout the food web are well documented in literature (Bryan 1979; Burger 2002; Table 3 Hg and Se geometric mean (in ng/mL) concentrations in blood of adults and juveniles found in three trophic levels, according to their diet in the studied birds Trophic level

Total Hg Total Se

Concentration factor

1 (n=23)

2 (n=15)

3 (n=6)

2/1

3/2

3/1

8.37 377

151 478

297 517

18 1.26

1.96 1.08

35 1.37

González et al. 1983; Muirhead and Furness 1988; Wiener et al. 2002; Wolfe et al. 1998). Nonetheless, most of these studies have been performed by analyzing samples of terrestrial or aquatic food chain in its broadest sense (from water or soil, to predators) and using liver or muscle but not blood as a substrate. Apart from the advantages over other tissues described previously, blood also gives valuable information about the bioaccumulation process. Concerning terrestrial-feeding birds, the Hg concentration for peregrine falcon chicks and black kite adults and chicks are in the low range compared to aquatic birds. However, it is very interesting to note that peregrine falcon adults showed the highest Hg concentration. This backs the hypothesis of the transport of aquatic Hg into the adjacent terrestrial food web, proposed by Cristol et al. (2008). In this paper, the authors found that two terrestrial songbird species showed the highest Hg concentrations in blood of all species, as reported in the present study. This could be explained by diet, especially if it includes spiders, which show highly bioavailable MeHg, and other invertebrates, and also by a possible Hg transfer to the terrestrial habitat leading to biomagnification in the DNP terrestrial food chain, but this should be further investigated. In the case of Se, the biological magnification found is almost negligible. The CF is around 1 in all the cases, with a maximum of 1.26 between the second and the first trophic level. The reason for this could be the low half-life of Se in blood estimated for some species, ca. 37 days for wild spectacle eiders (Grand et al. 2002) and 10 days for captive mallards (Heinz et al. 1990). In fact, a very efficient blood depuration of Se has been previously described (Franson et al. 1999, 2002; Grand et al. 2002; Wilson et al. 2004) so therefore, blood does not seem to be the right matrix to study Se bioaccumulation. Factors affecting blood Hg and Se concentrations by using General Linear Modeling As stated in the “Statistical analysis” section, two GLM groups were used to statistically explain the concentration of Hg and Se in blood. In the first group (model 1), the variables were species, age and sampling zone and in the second group (model 2) species was substituted by trophic level because this variable is a linear combination of species. It was not possible to establish any relationship between concentrations of Se in blood of birds and ages, species, trophic levels, or sampling zones by using either model. Conversely, in the case of Hg both models provided statistically significant results even though model 1 explained 82 % of the variance and model 2 only explained 47 % (Table 4). In model 1, species and sampling zone were the major contributors whereas in model 2 they were trophic position and sampling zone. Then, the amount of Hg in


5369

Table 4 GLM models for Hg concentrations in blood of the studied birds Log Hg Model 1 F (p) Df r2 Variables (F, p) Species Sampling area Age

26.7 (