CONCENTRATION OF MERCURY AND SELENIUM IN TISSUES OF FIVE CETACEAN SPECIES FROM CROATIAN COASTAL WATERS

Mercury (Hg) and selenium (Se) concentrations were measured in muscle, liver, kidney, spleen and lung tissues of five cetacean species, three dolphin (Stenella coeruleoalba, Tursiops truncatus and Grampus griseus) and two whale species (Balaenoptera physalus and Ziphius cavirostris), stranded along the Croatian coast during the period 1999-2002. Statistically significant differences in Hg concentrations in muscle, spleen and lung, and Se in liver and lung of the different dolphin species were observed. Mercury levels in liver and spleen and Se levels in liver differed between young and adult T. truncatus species. A significant positive correlation between different tissue types for Hg and Se concentrations was observed. In all tissues tested, the lowest Hg and Se concentrations were found in B. physalus. Mercury concentrations were positively correlated with Se in all tissues. The results present one of few studies related to lung and spleen tissues in these mammals, particularly in the Adriatic Sea. Since very little data are available, this research provides new data on concentrations of Hg and Se in five cetacean species from the Adriatic Sea basin.


INTRODUCTION
Over the past three decades, many studies of environmental pollutants, such as trace metals, have been carried out in marine organisms.Mercury (Hg) shows a high level of biomagnification in the upper levels of the food web due to its persistence and mobility in the marine ecosystem.Almost all Hg present in fish and squid is methylated and the accumulation of methylmercury occurs with concurrent bioamplification phenomena through the trophic chains because of their persistent and lipophilic nature (Palmisano et al. 1995;Caurant et al., 1996;Das et al., 2000).As top predators, marine mammals such as whales and dolphins accumulate high Hg levels in their internal organs, especially in the liver (André et al., 1990;Leonzio et al., 1992;Meador et al., 1999;Wagemann et al., 2000;Cardellicchio et al., 2002;Lahaye et al., 2006).The major part of Hg accumulated exists as inorganic Hg, suggesting that the demethylation of methylmercury occurs in the liver (Caurant et al., 1996;Holsbeek et al., 1998;Meador et al., 1999).The variability of Hg levels in cetaceans reflects interspecies dietary differences, variation in habitat use and age-accumulation trends (Leonzio et al., 1992;Augier et al., 1993;Cardellicchio et al., 2002).
It is suggested that selenium (Se) potentially protects marine mammals as an antidote to the toxic ef-fects of some metals, e.g.cadmium and mercury.Selenium may provide a protective mechanism that allows certain age classes of marine mammals to withstand exposure to Hg.In this case, these two elements can occur as an insoluble, nontoxic compound, mercury selenide (Nigro and Leonzio, 1996;Decataldo et al., 2004;Ikemoto et al., 2004).
Cetacean species that satisfy their breeding and feeding needs in Mediterranean waters include fin whales (Balaenoptera physalus), Cuvier's beaked whales (Ziphius cavirostris), Risso's dolphins (Grampus griseus), common bottlenose dolphins (Tursiops truncatus) and striped dolphins (Stenella coeruleoalba) (Bearzi et al., 2004).T. truncatus can be considered the only cetacean species that regularly occurs in the Adriatic Sea (Gomerčić et al., 2002;Bearzi et al., 2004).However, G. griseus and S. coeruleoalba have been suggested to be occasionally present in the Adriatic Sea and along the entire Croatian coastline (Pribanić et al., 1999;Gomerčić et al., 2002).Cuvier's beaked whale (Z.cavirostris) is the only cetacean species known to be regular in the Mediterranean Sea (Holcer et al., 2003).In addition, this species has never been recorded in the northern Adriatic; there have been several stranding reports in the southern Adriatic.B. physalus is the only mysticete regularly present in the Mediterranean Sea, feeding on krill, and it is a species that belongs to a different trophic chain level (Orsi Relini and Cappello, 1992).There are relatively few data regarding the determination of trace element levels in cetacean species inhabiting the Adriatic Sea (Storelli et al., 1998(Storelli et al., , 1999;;Storelli and Marcotrigiano, 2002;Pompe-Gotal et al., 2009).Previous studies reported that Mediterranean species have much higher Hg and Se concentrations than those of the Pacific and Atlantic (André et al., 1991;Augier et al., 1993;Lahaye et al., 2006).As in previous studies, only a few of these specimens were available due to difficulties in sample collection, since sampling depends on the incidence of finding dead individuals (Cappelli et al., 2008).
This paper presents a study of the Hg and Se distribution and accumulation in the tissues of five cetacean species, three dolphins (S. coeruleoalba, T. truncatus and G. griseus) and two whale species (Z.cavirostris and B. physalus) stranded along Croatian coast during the period 1999-2002.Mercury and selenium levels were also compared with those found in studied species from Mediterranean, Adriatic and Atlantic waters.
Prior to dissection, the age, gender, weight (except for B. physalus) and body sizes were recorded (Table 1).During the necropsies, teeth were collected for age determination.Teeth sections were prepared according to Slooten (1991) and age was estimated by counting growth layer groups (GLG) according to Hohn et al. (1989).T. truncatus dolphins older than 7 years were classified as adults.All other animals were classified as young.All four S. coeruleoalba dolphins were observed as one age group, as descriptive statistics require more than three animals.All carcasses were in moderate post-mortem condition.Tissues of muscle, kidney, liver, spleen and lung were collected for trace element analyses.Following collection, specimens were frozen in prewashed polyethylene bags, brought to the laboratory, and stored frozen at -18°C prior to analysis.

Sample preparation
All reagents were of analytical reagent grade: HNO 3 , H 2 O 2 , and HCl (Analytical Grade, Kemika, Croatia).Ultra-pure water (Milli-Q Millipore, 18.2 MΩxcm resistivity) was used for all dilutions.All plastic and glassware were cleaned by soaking in diluted HNO 3 (1/9, v/v) and rinsed with distilled water prior to use.Calibrations were prepared with element standard so-lutions of 1000 mg/l of each element supplied by Perkin elmer.Stock solution was diluted in HNO 3 (0.2%).Samples (0.5 g) were digested with 4 ml of HNO 3 (65% v/v), 2 ml of H 2 O 2 (30% v/v) with a microwave closed system, Multiwave 3000 (Anton Paar, Germany); blanks were processed in the same way as the sample.The digestion program was initiated at 500 W, then ramped for 1 min, after which the samples were maintained for 4 min at 500 W. The second step was started at 1000 W and held for 5 min.The third step began at a 1400 W and was maintained for 10 min.The fourth step was cooling at 0 W for 15 min.The digested samples were diluted to a final volume of 50 ml with ultra-pure water.All metal concentrations were determined on a wet weight basis as g/g.

Metal analyses
Selenium concentrations in tissues were determined by graphite furnace atomic absorption using an AAnalyst 800 (Perkin elmer, USA) equipped with an AS 800 autosampler (Perkin elmer, USA) set at 228.8 nm.As matrix modifier in each atomization, Pd(NO 3 ) 2 was used (Perkin elmer, USA) were used.For graphite furnace measurements, argon was used as the inert gas (flow rate 250 ml/min).Pyrolytic-coated graphite tubes with a platform were used.The atomic absorption signal was measured in peak area mode against a calibration curve.
Mercury concentrations were analyzed using the cold vapor technique with a flow injection system coupled to a FIAS-100 atomic absorption spectrophotometer (Perkin elmer, USA) equipped with an AS 93 plus autosampler (Perkin elmer, USA), and detection was conducted at 253.7 nm.As a reducing agent for the hydride technique for Hg determinations, the reaction of acidified aqueous samples (3% v/v HCl) with 0.2% sodium borohydride in 0.05% NaOH were used.
Detection limits were determined as the concentration corresponding to three times the standard deviation from ten blank samples of the results gen- The precision and accuracy of the analytical methods were determined using certified reference materials DORM-2 (dogfish muscle sample) (National Research Council, Canada).Results obtained for total Hg (n=5) were 4.55±0.69µg/g; the CRM has a certified Hg value of 4.64±0.26µg/g and recovery of 98.1% wase obtained.Selenium values (n=5) were 1.37±0.09µg/g, and in comparison with a certified Se value of 1.40±0.09µg/g, recovery of 97.9% was determined.The coefficient of variation (CV) in samples was less than 10%.The results illustrated the accuracy of the methods used.

Data treatment
Statistical analyses were performed using the Stata 6.0 statistical package (Stata Corp. USA).Data were logtransformed to improve normality prior to analysis to meet the underlying assumptions of the analysis of variance (ANOVA).Nonparametric rank order correlation between concentrations of elements in different tissues was assessed with Spearman's r coefficient.The significance of differences between group medians was assessed by the Mann-Whitney U-test and Kruskal-Wallis ANOVA.Results were considered significant at p ≤0.05.

RESULTS
Hg and Se concentrations in the different tissues examined are shown in Tables 2 and 3. Liver Hg and Se levels in five species tested were in the order: G. griseus > S. coeruleoalba > T. truncatus > Z. cavirostris > B. physalus.Kidney mercury levels were in the order: Z. cavirostris > T. truncatus > S. coeruleoalba > G. griseus >B.physalus.For Se, the order was the same, except that G. griseus had higher concentrations than S. coeruleoalba.Muscle concentrations of Hg were in the order: S. coe-ruleoalba > G. griseus > T. truncatus > Z. cavirostris > B. physalus.Selenium muscle and lung concentrations were in the order: S. coeruleoalba > G. griseus > Z. cavirostris > T. truncatus > B. physalus.Mercury levels in lung were very similar in S. Coeruleoalba, G. griseus and Z. cavirostris but very low in T. truncatus and B. physalus.The highest Hg levels in spleen were determined in G. griseus followed by S. coeruleoalba and T. truncatus.Spleen Se concentrations were in the order: S. coeruleoalba > G. griseus > T. truncatus.
A statistically significant difference was observed for Hg concentrations in muscle, spleen and lung of the different dolphin species (p=0.022,p=0.03, p=0.021).Furthermore, Hg levels in liver and spleen differed between the young and adult specimens of T. truncatus (p=0.03 and p=0.04, respectively).A statistically significant difference between Se concentrations were found in the liver and lung of the different dolphin species (p=0.049,p=0.016).Furthermore, Se levels in liver differed between the young and adult specimens of T. truncatus (p=0.02).
Spearman rank correlations and significance levels of Hg and Se between different tissues are shown in Tables 4 and 5.As shown in Table 4, Hg concentrations showed high positive and significant correlations among the different tissue types with the highest Spearman rank correlation between spleen and muscle tissues (r s =0.9324, p <0.0001).For Se concentrations in the different tissue types, correlation coefficients were between 0.6349 and 0.8919 and p values less than 0.0129 (Table 5).The highest Spearman rank correlations were determined between lung and muscle tissues (r s =0.8919, p < 0.0001).
In the present study, a significant and strong correlation was observed between Hg and Se concentrations in all tissues, with correlation coefficients higher than 0.805; p <0.002 (Table 6).

DISCUSSION
As previously suggested, feeding and habitat preferences are likely the key factors in the process of Hg accumulation in the tissues of top-level predators (La-haye et al., 2006).It is known that Z. cavirostris and G. griseus mainly feed on squid (Carlini et al., 1992a,b) while S. coeruleoalba and T. truncatus feed mostly on fish and squid on a smaller scale (Orsi Relini et al., 1994).As top predators in the food chain and in view of their longevity, dolphins tend to accumulate much higher Hg concentrations than other marine organisms.Previous studies reported that Hg levels were positively correlated with dolphin age and size (André et al., 1991;Storelli and Marcotrigiano, 2000;Roditi-elasar et al., 2003).Due to the liver's role in detoxification, it is known that Hg accumulates in  Vertical -letters show statistically significant differences between species for different tissues: liver a p=0.049; lung b p=0.016.Vertical -letters show statistically significant differences between young and adult T. truncatus: liver c p=0.02.
the liver in higher concentrations (André et al., 1991;Frodello et al., 2000;Roditi-elasar et al., 2003).This is confirmed in the present study, with the highest Hg levels measured in the liver with values about 5.7-to 29-times greater than the average concentration measured in the muscle.
In the present study, comparisons among the three dolphin species showed differences in Hg accumulation, especially in liver tissues.In addition, Hg concentrations in muscle, kidney, liver, spleen and lung tissues of adult T. truncatus were 3.6-to 103-times higher than in the same tissues of young specimens.This supports previous reports of the strong effect of age on Hg accumulation.A similar finding was reported for T. truncatus from the Mediterranean Sea (Shoham-Frider et al., 2002;Capelli et al., 2008).
Although the association between Hg and toxicity in cetaceans is not straightforward, it is suggested that the limit of tolerance for Hg in mammalian hepatic tissue seems to be within the range 100-400 µg/g (Wagemann and Muir, 1984).When accumulated in the liver cells, Hg inhibits the activity of lysosomal digestive enzymes, which suggests a reduction of protein degradation and might lead to excessive accumulation of lipofuscin within cells.It has been found that chronic Hg accumulation exceeding 100 µg/g caused liver abnormalities in T. truncatus (Rawson et al., 1993).
The highest concentrations of Hg in liver tissue (wet weight) of dolphins reported to date was 1326 µg/g in G. griseus (Shoham-Frider et al., 2002), 1500 µg/g in S. coeruleoalba (André et al., 1991) and 3289 µg/g in T. truncatus (Leonizio et al., 1992).In previous reports from the Adriatic Sea, high Hg levels of 1833.8 µg/g w.w. in liver were found in a 16-year-old T. truncatus near the City of Split (Pompe-Gotal et al., 2009) and 1790 µg/g w.w. in a 14-year-old male T. truncatus near the Island of Hvar (Bilandžić et al., 2012).
Table 7 shows the literature data regarding Hg and Se concentrations reported in S. coeruleoalba, G. griseus, Z. cavirostris, T. truncatus and B. physalus from the Mediterranean, Adriatic and Atlantic Seas.Very few studies have been conducted for the determination of trace elements in cetacean spleen and lung tissues (Storelli et al., 1998(Storelli et al., , 1999;;Frodello et al., 2000;Cardellicchio et al., 2000Cardellicchio et al., , 2002;;Capelli et al., 2008).Concentrations of elements in tissues from these reports expressed in dry weight were converted to wet weight using the factor (dw/ww) of 0.25 established for dolphins (Becker et al., 1995).
Mercury levels detected in the liver of S. coeruleoalba from the Mediterranean Sea were in the range from 115 µg/g in an adult male from the Corsican coast (Frodello et al., 2002) to 550 µg/g in a 16-yearold female found off the coast of Israel (Roditi-elasar et al., 2003).The lowest liver Hg levels in S. coeruleoalba were found off the French Atlantic coast (Das et a Value expressed as µg/g wet weight: dry weight were converted to wet weight using the factor (dw/ww) of 0.25 established for dolphins (Becker et al., 1995).* C, calf; Y, young; A, adult;  , 2003).Previous findings in S. coeruleoalba from the southern Adriatic Sea (Apulian coast) showed a maximum liver Hg level of 966.31 µg/g in a male specimen (Storelli et al., 1998).However, in the present study, the highest Hg concentration found was 380.6 µg/g in an adult specimen.
Mercury levels were comparable to those previously found in S. coeruleoalba from the southern Adriatic Sea (Storelli et al., 1998, Storelli andMarcotrigiano, 2002).Lower muscle Hg levels and similar kidney levels were reported at different locations in the Mediterranean Sea (Frodello et al., 2000;Cardellicchio et al., 2000Cardellicchio et al., , 2002;;Roditi-elasar et al., 2003).There are very few literature data for lung and spleen Hg and Se concentrations in cetacean species.Lung levels detected in S. coeruleoalba in the present study were higher than those in specimens stranded in the Mediterranean Sea were.There are few data regarding spleen Hg levels in S. coeruleoalba.In a recent study from the Ligurian Sea, spleen Hg levels were 102 µg/g (Capelli et al., 2008).
Mercury liver levels found in an adult T. truncatus in this study were lower than previously reported concentrations in animals from Adriatic and Mediterra-nean waters (Storelli and Marcotrigiano, 2000;Roditielasar et al., 2003;Pompe-Gotal et al., 2009).Very high levels in the liver tissues of young female specimens were detected off the Corsican coast and in the Ligurian Sea (Frodello et al., 2000;Capelli et al., 2008).Tissue Hg levels detected in groups of young T. truncatus were similar to concentrations found in young animals from the Mediterranean (Roditi-elasar et al., 2003;Capelli et al., 2008;Shoham-Frider et al., 2009).
Previously reported Hg levels in the liver of G. griseus were in the range of 478.32 to 1326 µg/g w.w. in the southern Adriatic and Mediterranean Seas (Storelli et al., 1999;Frodello et al., 2002;Shoham-Frider et al., 2002;Capelli et al., 2008).In the present study, Hg and Se levels measured in muscle, liver, kidney and lung tissues were lower than previously reported concentrations in G. griseus from the southern Adriatic and Mediterranean Seas (Storelli et al., 1999;Frodello et al., 2000;Shoham-Frider et al., 2002;Capelli et al., 2008).The highest liver levels in an adult specimen in this study were found in G. griseus, which is in accordance with previous measurements in the Adriatic Sea.Lung Hg levels measured in adult G. griseus were 3-times lower than levels found in the southern Adriatic (Storelli et al., 1999).
There are very few data available for metal levels in tissues of Z. Cavirostris and B. physalus from the Mediterranean and Adriatic Seas. a previous report from the southern Adriatic for Z. Cavirostris showed much higher Hg and Se levels in muscle, liver and kidney tissues (Storelli et al., 1999).However, much lower levels of Hg and Se were determined in the kidney tissue of male specimens in the Ligurian Sea of Italy than in this study (Capelli et al., 2008).
Hg and Se concentrations in all tissues of B. physalus were the lowest among the species tested in this study.With the exception of muscle tissue, which was lower for both metals, the results obtained were similar to concentrations found in specimens in the Ligurian Sea (Capelli et al., 2008).In comparison our results, the levels of Hg in liver and kidney were more than 10-times higher in the Mediterranean Sea (Hernández et al., 2000).
Previous findings from the southern Adriatic for Z. cavirostris show much higher levels in muscle liver and kidney tissues (Storelli et al., 1999).However, in contrast to the present study, lower Hg levels, particularly in kidney tissue, were determined in the Ligurian Sea (Capelli et al., 2008).
Similar to other studies, Se accumulated in the liver tissue of cetacean species was higher than in muscle and kidney tissues (Storelli et al., 1999;Frodello et al., 2000;Cardellicchio et al., 2002;Capelli et al., 2008).However, in this study Se levels in liver differed between the young and adult specimens of T. truncatus.This is contrary to previous reports for T. truncates that stated that Se concentrations in tissues were unrelated to age (Hansen et al., 1990;Meador et al., 1999).
In the present study, Se accumulated in the liver of three dolphin species was 1.5-to 5.4-times lower than Hg concentrations.The only exception was higher Se than Hg levels in the liver of B. physalus.Comparable Se levels in the liver were found in B. physalus from the Mediterranean Sea (Capelli et al., 2008).Previous studies showed that the increased Hg concentrations were associated with increasing Se levels in the liver, which is involved in the detoxification process of Hg (Meador et al., 1999;Cardellicchio et al., 2002;Capelli et al. 2008).The hypothesis is that the final compound of methylmercury demethylation appears to be the formation of the insoluble and poorly toxic Hg-Se complex.The presence of Hg-Se in the cytoplasm of hepatic cells has been confirmed in Z. cavirostris and S. coeruleoalba (Martoja and Berry, 1980;Nigro, 1994;Nigro and Leonzio, 1996).
Higher liver Se levels were reported in S. coeruleoalba and T. truncatus adults from the southern Adriatic Sea (Storelli et al., 1998;Storelli and Marcotrigiano, 2000).On the other hand, comparable levels of Se were found in muscle, liver and kidney, but much higher levels in spleen in T. truncatus were found in the Mediterranean Sea (Monaci et al., 1998;Cardellicchio et al., 2002;Capelli et al., 2008).Comparable Se levels in all tissues of G. griseus were also found in a previous report from the Adriatic Sea (Storelli et al., 1999).Much higher Se levels were found in the tissues in comparison to levels reported in other studies of G. griseus from the Mediterranean coast (Frodello et al., 2000;Capelli et al., 2008).Much higher liver Se concentrations in Z. cavirostris were measured in a previous report from the Adriatic Sea (Storelli et al., 1999).
In the present study, a significant and strong correlation between Hg and Se concentrations was observed in all tissues.This is similar to previous reports for marine mammal tissues from other localities showing a positive correlation between Hg and Se in a ratio of nearly 1:1 (Wagemann et al., 1998;Augier et al. 1993;Monaci et al., 1998;Meador et al., 1999;Cardellicchio et al., 2002;Capelli et al., 2008;Seixas et al., 2008).This correlation can be explained by the hypothesised protective detoxification mechanism of Se, by biotransformation in the liver where methylmercury is converted into the less toxic inorganic form of mercury selenide (Nigro and Leonzio, 1996;Wang et al., 2001;Decataldo et al., 2004;Ikemoto et al., 2004).

CONCLUSIONS
The present study reports differences in trace element (Hg and Se) contents in five cetacean species from the Adriatic Sea basin, which are generally difficult to compare in species with different feeding habits, life spans, and originating from various geographical locations as in the present study.The three dolphin species have a similar life span, so the comparison is useful.In addition, it is obvious that significant factors influencing trace element accumulation in the internal organs of cetacean species are age and food preference.In conclusion, Hg concentrations were positively correlated with Se in all tissues.The present study is one of the few addressing the concentrations of trace elements in lung and spleen tissues.

Table 1 .
Morphological data of 5 species from the Adriatic Sea.

Table 2 .
Mercury concentrations (mean and SD, median, range; µg/g w.w.) in tissues of five species from the Adriatic Sea.

Table 3 .
Selenium concentrations (mean and SD, median, range; µg/g w.w.) in tissues of five species from the Adriatic Sea.

Table 4 .
Spearman rank correlation coefficients and significance levels of Hg concentration between different tissues.

Table 5 .
Spearman rank correlation coefficients and significance levels (in brackets) of Se concentration between different tissues.

Table 6 .
Spearman rank correlation coefficients and significance levels between Hg and Se concentration in the same type of tissue.

Table 7 continued
al.