1816

Baseline / Marine Pollution Bulletin 52 (2006) 1784–1832

Tommasi, L.R., 1985. Resı´duos de praguicidas em a´guas e sedimentos de fundo do sistema estuarino de Santos (SP). Cieˆncia e Cultura, Sa˜o Paulo. 37 (6), 1001–1012. Tyler, A.O., Millward, G.E., 1996. Distribution and partitioning of polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans and polychlorinated biphenyls in the Humber estuary, UK. Mar. Pollut. Bull. 32 (5), 397–403. UNEP (United Nations Environment Programme), 1992. Determinations of Petroleum Hydrocarbons in Sediments. Reference Methods for Marine Pollution Studies 20, 75 pp. Venkatesan, M.I., 1988. Occurrence and possible source of perylene in marine sediments – A review. Mar. Chem. 25, 1–27. Volkman, J.K., Holdsworth, D.G., Neill, G.P., Bavor Jr., H.J., 1992. Identification of natural, anthropogenic and petroleum hydrocarbons in aquatic sediments. Sci. Total Environ. 112, 203–219. Volkman, J.K., Revill, A.T., Murray, A.P., 1997. In: Molecular markers in environmental geochemistryACS Symposium Series, vol. 671. Springer, Berlin, p. 426, Chapter 8. Wang, Z., Fingas, M., Page, D.S., 1999. Oil spill identification. J. Chromatogr. A 834, 369–411. Wang, Z., Fingas, M., 1997. Developments in the analysis of petroleum hydrocarbons in soils, petroleum products and oil-spill-related envi-

ronmental samples by gas chromatography. J. Chromatogr. A 774, 51– 78. Wang, Z., Fingas, M., 2003. Development of oil hydrocarbon fingerprinting and identification techniques. Mar. Pollut. Bull. 47, 423–452. Witt, G., Trost, E., 1999. Polycyclic aromatic hydrocarbons (PAHs) in sediments of the Baltic Sea and of the German coastal waters. Chemosphere 38, 1603–1614. Wedemeyer, G., 1967. Dechlorination of 1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane by Aerobacter aerogenes. Appl. Microbiol. 15, 569–574. Wurl, O., Obbard, J.P., 2005. Organochlorine pesticides, polychlorinated biphenyls and polybrominated diphenyl ethers in Singapore’s coastal marine sediments. Chemosphere 58, 925–933. Yuan, D., Yang, D., Wade, T.L., Qian, Y., 2001. Status of persistent organic pollutants in the sediment from several estuaries in China. Environ. Pollut. 114, 101–111. Yunker, M.B., Macdonald, R.W., Vingarzan, R., Mitchell, H., Goyette, D., Sylvestre, S., 2002. PAHs in the Fraser river basin: a critical appraisal of PAH ratios as indicators of PAH source and composition. Org. Geochem. 33, 489–515. Zanardi, E., Bı´cego, M.C., Castro Filho, B.M., Miranda, L.B., Prosperi, V. 2000. Southern Brazil. In: Shepard, Charles. (Org.). Seas At Millenium: In Environmental Evaluation, vol. 1. Amsterdam, pp. 731–747.

0025-326X/$ - see front matter  2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpolbul.2006.09.011

Total and methylmercury in three species of sea turtles of Baja California Sur Rita Kampalath b

a,*

, Susan C. Gardner b, Lia Me´ndez-Rodrı´guez b, Jennifer A. Jay

a UCLA Civil and Environmental Engineering, 5732H Boelter Hall, Los Angeles, CA 90095, United States Laboratorio de Ecologia Quimica y, Toxicologia, Centro de Investigaciones Biologicas del Noroeste, Mar Bermejo #194, Col. Playa Palo de Santa Rita, A.P.128, La Paz, Baja California Sur, C.P. 23090, Mexico

All seven species of sea turtle are currently classified as either endangered or threatened with extinction. Due to the wide geographic area that they cover during their life cycle, sea turtles may serve as meaningful ‘‘Sentinel Species’’ for overall ecosystem health, and because of this, it is especially important to document and understand any factors that might affect their survival or reproduction. The main causes of sea turtle mortality are anthropogenic: loss of nesting sites, ingestion of plastic debris and tar balls, exposure to organochlorine compounds and heavy metals such as mercury, commercial harvesting of carapace, flesh and eggs, and accidental capture by fishing nets (Sakai et al., 2000; Storelli et al., 1998; Vazquez et al., 1997; Waldichuck, 1987). Most available data to evaluate the anthropogenic factors affecting turtles involve stranded animals, where the cause of death is unknown. There is a paucity of published data on the baseline levels of contaminants in healthy tur-

*

a

Corresponding author. Tel.: +1 310 267 4654; fax: +1 310 206 2222. E-mail address: [email protected] (R. Kampalath).

tles. Additionally, the majority of this data is from the coasts of the Atlantic Ocean. Most studies of heavy metal contamination in sea turtles have been conducted in Japan (Saeki et al., 2000; Sakai et al., 1995; Sakai et al., 2000), Australia (Gordon et al., 1998), the UK (Davenport and Wrench, 1990; Godley et al., 1999), and the Mediterranean (Godley et al., 1999; Storelli et al., 1998; Storelli and Marcotrigiano, 2000). Only two studies have been conducted in the Gulf of California (Gardner et al., 2006; Presti et al., 1999), despite the fact that the northwest coast of Mexico has the highest concentration of commercial marine organisms in the country, and the coastal lagoons of Baja California peninsula serve as major feeding and developmental grounds for sea turtles (Ochoa et al., 1997). Hg contamination of food chains often occurs in pristine areas through long-range transport and deposition of atmospheric inorganic mercury (Fitzgerald et al., 1998; Ullrich et al., 2001). In aquatic systems, less-toxic inorganic Hg(II) is microbially transformed to methylmercury, a potent neurotoxin with a strong tendency to biomagnify in aquatic food webs. No previous research has been conducted to document mercury in turtles from Baja Califor-

Baseline / Marine Pollution Bulletin 52 (2006) 1784–1832

Fig. 1. Locations where sea turtles were collected along Baja California peninsula, Mexico.

nia Sur (Fig. 1) and to our knowledge no studies have yet been published that report Hg concentrations in olive ridley sea turtles, and only one other study has been published documenting methylmercury in sea turtles (Storelli and Marcotrigiano, 2003). Coastal lagoons of the Baja California peninsula are an important habitat for five species of sea turtle: the Eastern Pacific green turtle, Chelonia mydas (locally known as the black turtle, Chelonia mydas agassizii), the Pacific loggerhead Caretta caretta, the olive ridley Lepidochelys olivacea, and to a lesser extent, the hawksbill Eretmochelys imbricata, and the leatherback Dermochelys coriacea. Many coastal communities along the peninsula are dependent on the harvest of natural resources for sustenance, with the majority of inhabitants employed as fishermen. Although the intentional capture of sea turtles has been strictly regulated in Mexico since 1972, it is not unusual for sea turtles to be accidentally caught in fishermen’s net and drowned (Gardner and Nichols, 2001). Because the incidental death of endangered turtles provides a unique opportunity to gain knowledge about these animals and the environmental hazards they face, researchers at Centro de Investigaciones Biologicas del Noroeste (CIBNOR) began accompanying fishermen on night fishing trips to recover tissues for research and the formation of a tissue bank. This work presents an analysis of mercury and methylmercury levels in sea turtle tissue of the tissue bank at CIBNOR, involving healthy turtles with a known cause of death, and is a valuable addition to the small data set available worldwide for levels of this contaminant in sea turtles. The Magdalena lagunar system (Santo Domingo–Magdalena–Almejas) extends 175 km (from 25
43 0 N to 24
20 0 N) along the Pacific coast of Baja California Sur, communicating with the ocean through five mouths or channels. Vegetation includes vast seagrass beds (Zostera marina and Phyllospadix sp.), algae beds and mangrove forests.

1817

A wide variety of commercially exploited species are known in these bays as well as many migratory seabirds, marine mammals and fishes. Samples of turtle tissues as well as algae samples weighing approximately 0.2 g each were finely chopped using acid-rinsed plastic knives and an acid-rinsed plastic cutting board. Sediment samples of approximately 0.5–1.5 g were also weighed out. (Additional sediment samples were heated at 80
overnight in order to determine a wet/dry ratio.) Samples were then placed in a 30 mL Teflon centrifuge tube which had been soaked overnight in alkaline cleaner, acid-washed overnight in 1.2 M HCl and finally triple rinsed in 18 mQ water. Samples were digested overnight using concentrated HCl, HNO3 and then overnight again with BrCl, as described in EPA Method 1631. The sample was centrifuged at 3000 rpm for 20 min. Hydroxylamine hydrochloride (NH2OH Æ HCl) was added to oxidize excess BrCl, and the digestate was analyzed for total Hg by the Brooks Rand cold vapor atomic fluorescence spectrometry (CVAFS) detector after reduction to Hg(0) with SnCl2. Samples were digested and analyzed in batches of six to eight tissue samples. Because the tissue bank created in Baja California Sur was used to supply a number of different studies in addition to this one, the amount of sample available from each individual turtle was limited. In addition, it was necessary to ensure that adequate samples were available for methylmercury detection which is a more complicated and error-prone process than total mercury detection. Because of these limitations, single digestions were conducted for each sample (n = 42 for C. mydas; n = 16 for C. Caretta; n = 23 for L. olivacea). Four genera of Chlorophyta and Rhodophyta algae (Codium cuneatum, Codium amplivesiculatum, Gracilaria pacifica, Hypnea johnstonii) were digested in duplicate, and eight samples of sediments were digested. Each digestate was then analyzed in triplicate. Average standard deviations were 11% of the value. Three samples of DORM-2 standard reference material (SRM) were digested and analyzed along with each batch of turtle and algae samples. IAEA-405 estuarine sediment SRM was analyzed with sediment samples. Recovery of standard reference material for total Hg ranged from 90% to 117%, averaging 104%. Calibration curves of between five and eight standards were run each day, with average correlation coefficients of greater than 0.999 (values ranged from 0.9987 to 0.9999). Bubbler blanks and reagent blanks were also run with each batch of samples. Once analysis for methylmercury was completed, a subset of leftover samples was analyzed for total mercury direct mercury analysis (Milestone DMA-80). For this method, which is described in detail in EPA Method 7473, samples weighing between 0.2 and 0.5 g were measured into a sample boat, and then loaded into the DMA detector. The sample is first dried and then thermally and chemically decomposed. After oxidation and separation from other decomposition products, mercury vapor is passed through cells where absorbance is measured using an atomic absorption spectrophotometer. Two to three

1818

Baseline / Marine Pollution Bulletin 52 (2006) 1784–1832

samples of DORM-2 SRM were analyzed with each run, with recoveries ranging from 96% to 102% and averaging 98%. Calibration curves were also run each day, with average correlation coefficients of greater than 0.998 (values ranged between 0.996 and 1). Adipose, muscle, liver and kidney tissue from a subset of the samples analyzed for total mercury content were also analyzed for methylmercury content. The samples were chopped in the same way as for total mercury analysis, and methylmercury was extracted from the sample using a modified version of the solvent extraction process described in Bloom et al., 1997. Methylmercury levels in the extract were then analyzed using the Brooks Rand methylmercury CVAFS detection system. Turtle samples were extracted and analyzed in batches of four to six samples. Due to limitations in mass of samples, single extractions were used for most samples. Triplicate digestions were done for one sample of each tissue type in order to test for variability. Standard deviations for the triplicate extractions averaged 3.3% (1.9% for adipose, 5.9% for muscle, 5.1% for liver and 0.6% for kidney). Each extract was analyzed twice. Three samples of DORM-2 standard reference material (SRM) were digested and analyzed along with each batch.

Recovery of standard reference material averaged 78%. Because of the low recoveries relative to those from total mercury analysis, amplification factors were applied to sample concentrations depending on the recovery of the SRMs run with each particular batch. Calibration curves of four standards were run each day, with correlation coefficients ranging from 0.9897 to 0.9999 with an average of 0.996. Bubbler blanks and reagent blanks were also run with each batch of samples. Adipose, muscle, liver and kidney tissue samples were analyzed for three species of turtles: C. mydas agassizii, C. caretta and L. olivacea. Levels found in tissues of each individual are presented in Table 1. Two samples of each tissue type from each species of turtle were analyzed for methylmercury to determine methylmercury percentages in each tissue (Table 2). Algae and sediment from the study area were also analyzed. Mercury concentrations in sediments ranged between 0.0013 and 0.0028 ppm on a dry weight basis. Several different species of algae were analyzed, and concentrations were found to range between 0.0017 and 0.0054 ppm on a dry weight basis. With the exception of a few individuals, the mercury levels found in the organs of the turtles sampled in this study were low in comparison to levels found in other studies

Table 1 Total mercury concentrations found in three different species of sea turtle in adipose, muscle, liver and kidney tissues Species

Sample ID

Size (cm)

Adipose (ppm)

Liver (ppm)

Muscle (ppm)

Kidney (ppm)

C. mydas

44 56 67 117 152 158 209 215 227 228 229 231

44 75.5 55 52.5 47 66.5 80 45.5 49.4 46 64 63

0.002 NDa ND ND 0.011 N/A ND 0.002 ND ND 0.002a 0.006

0.135 0.155a 0.028 0.041 N/A 0.026 0.039a 0.068 0.146 0.147 0.168a 0.048

0.059 0.016a 0.003 0.004 N/A 0.002 0.002a N/A 0.031 0.055 0.031a 0.007

0.310 0.078a 0.021 0.040 0.138 0.003 0.078a N/A 0.133 N/A 0.061a 0.030

C. caretta

149 155 203 208 214 216 217 219

50.7 56 71 63 63.5 62 52 60.2

NDa 0.004 0.028 0.005 0.002 N/A 0.004 N/A

0.183a N/A N/A 0.116 0.166a N/A N/A 0.144a

N/A N/A N/A 0.041 0.026a N/A 0.018 0.019a

0.135a N/A N/A N/A N/A N/A N/A 0.064a

L. olivacea

25 99 146 148 154 156 205 207 212

57 57 64 61.5 64 66 53 55 56

0.004a 0.156a ND 0.006 0.000 N/A 0.011 0.022 ND

0.126a 0.795a N/A 0.057a 0.112 N/A 0.072 0.117 N/A

0.016a 0.144a 0.030 0.015a 0.052 N/A N/A 0.043 N/A

N/A 0.372 N/A 0.030a N/A 0.028 N/A N/A N/A

ND = non-detect; N/A = tissue sample not available from individual. a Average results from analysis by CVAFS and DMA methods.

Baseline / Marine Pollution Bulletin 52 (2006) 1784–1832 Table 2 Methylmercury concentrations and percentages from subset of turtles from present study Species

Sample ID

Concentration (ppm)

C. mydas

229A 229M 229L 229K 231A 231M 231L 231K

0.001 0.006 0.027 0.019 0.0002 0.001 0.004 0.003

MeHg (%) 17 22 19 23 4 18 9 8

C. caretta

149K 203A 214A 214M 214L 219M 219L 219K

0.017 0.0116 0.0013 0.011 0.019 0.01 0.029 0.02

16 42 65 48 10 45 22 26

L. olivacea

99M 99L 99K 148A 148M 148L 148K 207A

0.17 0.338 0.208 0.0077 0.022 0.023 0.013 0.0015

100 41 56 100 92 39 100 8

and, along with levels found in algae and sediment, support the designation of Bahia Magdalena as a ‘‘pristine’’ site (Table 3). The relative concentrations in organs of each individual tended to vary as follows: liver > kidney > muscle > adipose. For C. mydas agassizii and C. caretta, all organ differences were significant (p < 0.1 or lower) except for C. mydas agassizii liver as compared to kidney concentrations. Similar trends have been noted in C. caretta and C. mydas agassizii in other studies as well (Godley et al., 1999; Sakai et al., 1995; Sakai et al., 2000). In L. olivacea significant differences were found between adipose and liver tissues (p < 0.1) and between adipose and kidney tissues (p < 0.2). If one particularly highly contaminated individual from the L. olivacea group is not considered, significant differences were found between all organs (p < 0.02 for most; p < 0.2 for adipose versus kidney), with the exception of kidney versus muscle. Relatively few studies have examined the accumulation and toxic effects of mercury in sea turtles. The difference in body burden distribution is dependent on the physiology of the tissue in question. In general, the tissues with the highest concentration have been found in C. mydas agassizii and C. caretta to be the liver, kidney, and muscle, in that order (Day et al., 2005; Godley et al., 1999; Sakai et al., 1995; Sakai et al., 2000). There are two families represented by six genera and seven species of marine turtles inhabiting tropical, sub-tropic, and temperate waters. Interspecies differences may be due to differences in foraging ecology. Within species, the diet may vary depending on the location and availability of food sources.

1819

In addition, a general trend between the various species was noted as follows: L. olivacea > C. caretta > C. mydas agassizii. Significant differences were found between adipose (p < 0.2) and liver (p < 0.1) tissues of C. mydas agassizii versus C. caretta and in adipose (p < 0.2), muscle (p < 0.1) and liver (p < 0.2) of C. mydas agassiziiversus L. olivacea. Similar trends have been noted between C. caretta and C. mydas agassizii in other studies as well (Godley et al., 1999; Sakai et al., 1995). To our knowledge there are currently no published studies that report Hg data in L. olivacea. As stated earlier, there were some individuals with particularly high concentrations of Hg in certain tissues, particularly in the C. mydas agassizii and L. olivacea sampled. If the contaminated individual in the L. olivacea group is not considered, C. caretta tended to have the highest Hg concentrations. These species differences may be explained by differences in the feeding habits of each species: C. mydas agassizii is a herbivorous turtle, while C. caretta is carnivorous and L. olivacea is opportunistic in its diet. Previous studies have shown that C. caretta usually has a higher concentration of contaminants than C. mydas agassizii, which is consistent with interspecific differences in diet and trophic status (Godley et al., 1999; Sakai et al., 1995). Godley et al. (1999) found that the average mercury in liver tissue and nest contents was higher in C. caretta than in C. mydas agassizii. Hg concentrations in the liver and muscle of C. caretta were higher than those found in the same tissues of leatherbacks, Dermochelys coriacea (Sakai et al., 1995). As mentioned earlier, a subset of leftover tissues were analyzed using a Milestone DMA-80 in order to verify values found using the Brooks Rand CVAFS. The results of this methods comparison are presented in Table 4. Although some variations were found in the additional analyses, the levels remain very similar to what was originally reported and the trends observed between tissues and species remain the same. Total mercury concentrations in organs of turtles were compared to straight carapace length (SCL) of each individual, which is generally used as a measure of age of the turtle. In this study, SCLs ranged from 44 to 80 cm. No significant difference was found between the sizes of the different species (p > 0.2). No relationships were found between size and concentrations for L. olivacea. Some positive relationships were observed, however, between size and mercury concentrations in adipose and muscle tissue in C. caretta (R2 = 0.5618 and 0.3338, respectively), which supports the notion that mercury tends to bioaccumulate (Fig. 2a and b). Negative relationships were observed between size and mercury concentrations in muscle and kidney of C. mydas agassizii (R2 = 0.4345 and 0.2543) (Fig. 2c and d). This interesting result corresponds to findings in previous studies of negative correlations of certain metals to turtle size (Gordon et al., 1998; McKenzie et al., 1999; Saeki et al., 2000) for C. mydas. These counterintuitive relationships have been attributed to changes in feeding habits that occur through the progress of C. mydas

1820

Baseline / Marine Pollution Bulletin 52 (2006) 1784–1832

Table 3 Comparison of total mercury concentrations in present and previous studies Species

Tissue

Location

Range (ppm)

Study

C. caretta

Liver

Present study Italy Japan Cyprus Southeast US

0.116–0.179 0.13–1.26 0.25–8.15 0.82–7.5 0.346–1.336

Storelli et al. (2005) Sakai et al. (1995) Godley et al. (1999) Day et al. (2005)

Present study Italy Japan Cyprus Southeast US

0.075–0.108 0.06–0.31 0.04–0.44 0.13–0.8 0.132–0.436

Storelli et al. (2005) Sakai et al. (1995) Godley et al. (1999) Day et al. (2005)

Present study Italy Japan Cyprus Southeast US

0.018–0.041 0.03–0.66 0.053–0.19 BDL–1.78 0.049–0.499

Storelli et al. (2005) Sakai et al. (1995) Godley et al. (1999) Day et al. (2005)

Present study Italy Japan

0.0002–0.028 BDL–0.09 0.04 (median)

Storelli et al. (2005) Sakai et al. (2000)

Present study Japan Cyprus

0.026–0.153 0.0767–0.301 0.27–1.37

Sakai et al. (2000) Godley et al. (1999)

Present study Japan Cyprus

0.003–0.31 0.0422–0.0478 BDL

Sakai et al. (2000) Godley et al. (1999)

Present study Japan Cyprus

0.003–0.059 0.0021–0.0069 BDL–0.37

Sakai et al. (2000) Godley et al. (1999)

Present study Japan

BDL–0.011 0.0024–0.0028

Sakai et al. (2000)

Kidney

Muscle

Fat

C. mydas

Liver

Kidney

Muscle

Fat

from its post-hatchling to its adult stage, during which time C. mydas switches from a primarily carnivorous to a primarily herbivorous diet. Based on these feeding habits, the turtles may be exposed to the highest levels of metals during their early years when they are feeding on higher trophic level organisms, with these concentrations gradually lowering due to dilution through growth of the animal and also lessened intake of the metal (McKenzie et al., 1999). Methylmercury concentrations were also measured in subsets of samples from each species and each tissue type (Table 2). Methylmercury percentages tended to vary between individuals, following a similar trend as found in total mercury concentrations: L. olivacea > C. caretta > C. mydas agassizii. This trend was not as clear in the absolute values of methylmercury concentrations. Across species, absolute values of methylmercury in liver tended to be the highest, however, in each individual, the methylmercury percentage in liver tended to be lower than in muscle or kidney, corresponding with results presented by other authors showing demethylation occurring in liver tissues (Storelli and Marcotrigiano, 2003; Henny et al., 2002). Methylmercury percentages in adipose tissues tended to vary greatly between individuals (range: 4–100%). Methyl-

mercury in C. caretta in the Mediterranean was found to range from 55% to 95% in muscle tissues and from 27% to 65% in liver tissues (Storelli and Marcotrigiano, 2003), however, to our knowledge, no other reports currently exist on methylmercury in sea turtles. Since methylmercury is the mercury species, that is of concern with regards to health of an organism, this information can help to further understanding of the distribution of this contaminant in the organs of the sea turtle, and also of how its system stores and/or metabolizes methylmercury. Relatively few reports exist on heavy metal geochemistry in the region. A study of the La Paz Lagoon, in the Gulf of California, showed no major anthropogenic contamination by As, Sb, Se, Cd, Cu, Pb and Zn in sediments (Shumilin et al., 2001). In sediments of Bahı´a Magdalena, an enrichment of Se, Sb, Cd and As was observed, and a high concentration of Cr (1000 mg/kg) was measured near Isla Margarita (Rodriquez-Meza et al., 2003). Anomalously high levels of Cd have been reported in sea turtles in the southern region of the peninsula of Baja California (Gardner et al., 2006), as well as in sediments from the region (Mendez et al., 1998). There are several other reports about high levels of cadmium along the peninsula in mussels (Villaescusa et al., 1991), plankton (Martin and Broenkow,

Baseline / Marine Pollution Bulletin 52 (2006) 1784–1832

1821

Table 4 Results of method comparison between acid digestion/CVAFS and direct mercury analysis Species

Sample IDa

DMA analysis (ppm)

Range or Standard Deviation for DMAb

BR analysis (ppm)c

Standard Deviation for BR

Green/Black

56A 56M 56L 56K 209M 209L 209K 229A 229M 229L 229K

ND 0.016 0.157 0.065 0.003 0.041 0.078 ND 0.036 0.180 0.040

ND (0.012–0.019)

ND 0.016 0.153 0.091 0.002 0.038 0.078 0.006 0.026 0.144 0.081

N/A 0.001 0.011 0.004 0.000 0.002 0.007 0.000 0.001 0.002 0.002

219M 219L 219K 149A 149L 149K 214M 214L

0.016 0.152 0.054 0.001 0.186 0.162 0.030 0.152

0.022 0.136 0.075 ND 0.172 0.108 0.022 0.179

0.001 0.006 0.004 N/A 0.013 0.002 0.001 0.019

25A 25M 25L 99A 99M 99L 148M 148L 148K

0.006 0.017 0.156 0.162 0.135 0.763 0.019 0.056 0.031

0.001 0.015 0.096 0.143 0.154 0.826 0.009 0.060 0.024

0.000 0.001 0.004 0.005 0.009 0.095 0.002 0.004 0.001

Loggerhead

Olive Ridley

a b c

ND (0.175–0.185)

(ND–0.001) 0.010

(0.001–0.011) 0.002 (0.135–0.188)

(0.012–0.025) (0.054–0.057) 0.005

A = adipose, M = muscle, L = liver, K = kidney. Range provided when two samples were analyzed, standard deviation provided when three samples were analyzed. Average of triplicate analyses of single digestion.

Fig. 2. Size correlations for C. caretta : (a) adipose and (b) muscle; and for C. mydas agassizii (c) muscle and (d) kidney.

1822

Baseline / Marine Pollution Bulletin 52 (2006) 1784–1832

1975), and insects (Cheng et al., 1976). The only mercury data existing for the region is for sediments in the La Paz Lagoon, where low levels suggest that atmospheric deposition is the main source of mercury to the area (Kot et al., 1999). In conclusion, differences in total mercury observed between sea turtle species (L. olivacea >C. caretta > C. mydas agassizii) seemed to correlate with variations in feeding habits, with lowest concentrations seen in the herbivorous species. The total Hg concentrations measured in organs of individuals from all species tended to vary as follows: liver > kidney > muscle > adipose. A methods comparison showed consistent results between acid digestion/CVAFS and direct mercury analysis. Positive correlations were observed between turtle size and total mercury concentrations in adipose and muscle tissue in C. caretta, and negative relationships were observed between size and total mercury concentrations in muscle and kidney of C. mydas agassizii. Across all three species, absolute values of methylmercury in liver tended to be the highest, while percent methylmercury tended to be lowest in this organ. This work adds to the small dataset worldwide on mercury levels in sea turtles. To our knowledge, this study contains the first mercury data to be reported for L. olivacea, and is only the second study reporting concentrations of methylmercury in sea turtles. Acknowledgements We would like to thank Professor Rafael Riosmena Rodriguez of Universidad Autonoma de Baja California Sur and his students for their help and guidance in collecting environmental samples, The Bay Foundation for their support of this project and the staff and collaborators from the School for Field Studies in Puerto San Carlos. We are grateful to all the members in our lab for their continuous support. This work was funded by grants from UC MEXUS, the National Science Foundation (BES 0348783), and the Consejo Nacional de Ciencia y Tecnologı´a (SEP-CONACYT # SEP-2004-CO1-45749). References Bloom, N., Colman, J., Barber, L., 1997. Artifact formation of methylmercury during aqueous distillation and alternative techniques for the extraction of methylmercury from environmental samples. Fresenius Journal of Analytical Chemistry 358, 371–377. Cheng, L., Alexander, G., Franco, P., 1976. Cadmium and other metals in sea skaters. Water Air and Soil Pollution 6, 33–38. Davenport, J., Wrench, J., 1990. Metal levels in a leatherback turtle. Marine Pollution Bulletin 21 (1), 40–41. Day, R.D., Christopher, S.J., Becker, P.R., Whitaker, D.W., 2005. Monitoring mercury in the loggerhead sea turtle, Caretta caretta. Environmental Science and Technology 39, 437–446. Fitzgerald, W., Engstrom, D., Mason, R., Nater, E., 1998. The case for atmospheric contamination in remote areas. Environmental Science and Technology 32, 1–7. Gardner, S., Nichols, W., 2001. Assessment of sea turtle mortality rates in the Bahia Magdalena region, Baja California Sur, Mexico. Chelonian Conservation and Biology 4 (1), 197–199.

Gardner, S., Fitzgerald, S.L., Acosta-Vargas, B., Mendez-Rodriguez, L., 2006. Heavy metal accumulation in four species of sea turtles from the Baja California peninsula, Mexico. BioMetals 19, 91–99. Godley, B.J., Thompson, D.R., Furness, R.W., 1999. Do heavy metal concentrations pose a threat to marine turtles from the Mediterranean Sea? Marine Pollution Bulletin 38 (6), 497–502. Gordon, A.N., Pople, A.R., Ng, J., 1998. Trace metal concentrations in livers and kidneys of sea turtles from south-eastern Queensland, Australia. Marine and Freshwater Research 49 (5), 409–414. Henny, C.H., Hill, E.F., Hoffman, D.J., Spalding, M.G., Grove, R.A., 2002. Nineteenth century mercury: hazard to wading birds and cormorants of the Carson River, Nevada. Ecotoxicology 11, 213–231. Kot, F., Green-Ruiz, C., Paez-Osuna, F., Shumilin, E., Rodriguez-Meza, D., 1999. Distribution of mercury in sediments from La Paz Lagoon, Peninsula of Baja California, Mexico. Bulletin of Environmental Contamination and Toxicology 63, 45–51. Martin, J., Broenkow, W., 1975. Cadmium in plankton: elevated concentrations off Baja California. Science 190, 884–885. McKenzie, C., Godley, B.J., Furness, R.W., Wells, D.E., 1999. Concentration and patterns of organochlorine contaminants in marine turtles from Mediterranean and Atlantic waters. Marine Environmental Research 47, 117–135. Mendez, L., Acosta, V., Alvarez-Castaneda, S., Lechuga-Deveze, C., 1998. Trace metal distribution along the southern coast of Bahia de La Paz (Gulf of California), Mexico. Bulletin of Environmental Contamination and Toxicology 61, 616–620. Ochoa, J.L., Sanchez-Paz, A., Cruz-Villacorta, A., Nunez-Vazquez, E., Sierra-Beltran, A., 1997. Toxic events in the northwest Pacific coastline of Mexico during 1992–1995: origin and impact. Hydrobiologia 352, 195–200. Presti, S., Hidalgo, A., Sollod, A., Seminoff, J., 1999. Mercury concentrations in the scutes of black sea turtles, Chelonia mydas agassizii in the Gulf of California. Chelonian Conservation and Biology 3 (3), 531–533. Rodriquez-Meza, E., Shumilin, E., Mendez-Rodriguez, L., AcostaVargas, B., Sapozhnikov, D., Lutsarev, S. 2003. Land-derived versus sources as reflected in trace metal concentration, In: Southern California Society of Environmental Chemistry Annual Meeting. Saeki, K., Sakakibara, H., Sakai, H., Kunito, T., Tanabe, S., 2000. Arsenic accumulation in three species of sea turtles. Biometals 13 (3), 241–250. Sakai, H., Ichihashi, H., Suganuma, G., Tatsukawa, R., 1995. Heavymetal monitoring in sea turtles using eggs. Marine Pollution Bulletin 30 (5), 347–353. Sakai, H., Saeki, K., Ichihashi, H., Suganuma, H., Tanabe, S., Tatsukawa, R., 2000. Species-specific distribution of heavy metals in tissues and organs of loggerhead turtle (Caretta caretta) and green turtle (Chelonia mydas) from Japanese coastal waters. Marine Pollution Bulletin 40 (8), 701–709. Shumilin, E., Paez-Osuna, F., Green-Ruiz, C., Sapozhnikovs, D., Rodriguez-Meza, D., Godinez-Orta, L., 2001. Arsenic, antimony, selenium and other trace elements in sediments of the La Paz Lagoon, Peninsula of Baja California, Mexico. Marine Pollution Bulletin 42 (3), 174–178. Storelli, M.M., Marcotrigiano, G.O., 2000. Total organic and inorganic arsenic from marine turtles (Caretta caretta) beached along the Italian coast (South Adriatic Sea). Bulletin of Environmental Contamination and Toxicology 65 (6), 732–739. Storelli, M.M., Marcotrigiano, G.O., 2003. Heavy metal residues in tissues of marine turtles. Marine Pollution Bulletin 46, 397–400. Storelli, M.M., Ceci, E., Marcotrigiano, G.O., 1998. Distribution of heavy metal residues in some tissues of Caretta caretta (Linnaeus) specimen beached along the Adriatic Sea (Italy). Bulletin of Environmental Contamination and Toxicology 60 (4), 546–552. Storelli, M.M., Storelli, A., D’Addabbo, R., Marano, C., Bruno, R., Marcotrigiano, G.O., 2005. Trace elements in loggerhead turtles (Caretta caretta) from the eastern Mediterranean Sea: overview and evaluation. Environmental Pollution 135 (1), 163–170.

Baseline / Marine Pollution Bulletin 52 (2006) 1784–1832 Ullrich, S., Tanton, T., Abdrashitova, S., 2001. Mercury in the environment: a review of the factors affecting methylation. Critical Reviews in Environmental Science and Technology 31 (3), 241– 293. Vazquez, G.F., Reyes, M.C., Fernandez, G., Aguayo, J.E.C., Sharma, V.K., 1997. Contamination in marine turtle (Dermochelys coriaca) egg shells of Playon de Mexiquillo, Michoacan, Mexico. Bulletin of Environmental Contamination and Toxicology 58 (2), 326–333.

1823

Villaescusa, C., Gutierrez, E., Flores, G., Arreola, M., 1991. Metales traza en el mejillon, Modiolus capax, del Golfo de California: variaciones geograficas. Congreso de la Asociacion de Investigadores del Mar de Cortes, 17. Waldichuck, M., 1987. Natural versus anthropogenic impacts. Marine Pollution Bulletin 18 (4), 143–144.

0025-326X/$ - see front matter  2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpolbul.2006.09.019

Copper emission factors from intensive shrimp aquaculture L.D. Lacerda

*,1,

J.A. Santos, R.M. Madrid

Instituto de Cieˆncias do Mar, Universidade Federal do Ceara´, Av. Abolic¸a˜o 3207, Fortaleza, CE, Brazil

Emission factors, i.e., the amount of a given pollutant emitted per unit of production goods or production area, are strong tools for estimating pollutant loads to the environment from a variety of anthropogenic sources, since they can derive a difficult measurable variable (pollutant load) from an easily assessed parameter (e.g., area, amount of goods produced, inhabitants) and have been successfully used at the global (e.g., Nriagu and Pacyna, 1988); regional (e.g., Hutton and Symon, 1986) and local level (e.g., Molisani et al., 2004; Lacerda et al., 2006), to estimate pollutant emissions from natural and anthropogenic sources to the environment. Emission factors are also presently used in most countries’ environmental agencies to create and update pollutant inventories statistics (e.g., EEA, 1999; EPA, 2002; Molisani et al., 2004). The fast growth of intensive shrimp farming worldwide and its dependence on large inputs of artificial feed, fertilizers and of other chemical addictives such as acidity correctors and algaecides have triggered many studies to investigate shrimp farm’s role as nutrient sources to coastal environments which allowed the calculation of emission factors for major nutrients such as N and P (Pa´ez-Ozuna et al., 2003; Burford et al., 2003). Trace metals, however, are not obvious pollutants present in shrimp farm effluents. However, some trace metals are present as natural components in aquafeeds, as impurities in fertilizers or as active principles of pesticides (Boyd and Massaut, 1999; Tacon and Forster, 2003). But since shrimp farming is generally developed in areas without significant sources of trace metals, their emissions can be relatively important for these regions. Among the trace metals eventually present in shrimp farm effluents, Cu is of high significance not only due to its ubiquitous presence in *

Corresponding author. Tel.: +55 85 324421263; fax: +55 85 32683205. E-mail address: [email protected] (L.D. Lacerda). 1 On leave from Departamento de Geoquı´mica, Universidade Federal Fluminense, Nitero´i, 24020-0-7 RJ, Brazil.

aquafeeds and other chemicals and to its toxicity to phytoplankton and the shrimps proper (Bainy, 2000; Chen and Lin, 2001; Lee and Shiau, 2002). Shrimp farming in NE Brazil has increased exponentially during the past 10 years from an annual production of about 7000 tons, produced in less that 1000 ha of pond area in 1998 to over 90000 tons produced in about 15000 ha of pond area in 2003 (Madrid, 2004). This resulted in an increase in nutrient emissions to estuaries in many areas, which formerly had no significant pollution sources. A previous survey of trace metal content in shrimp and aquafeeds performed in some major farms in this area showed relatively high concentrations of Cu and suggested deleterious effects of this trace metal on shrimp productivity (Lacerda et al., 2004). In the present study we present the first estimate of Cu emission factor from intensive shrimp farming based on experimental data from a typical farm in Northeastern Brazil. The high similarity of emission factors for N and P from these farms and their technological processes with those generated from farms in Mexico and Australia (Pa´ez-Ozuna et al., 2003; Burford et al., 2003; Jackson et al., 2003; Lacerda et al., 2006), suggests that the proposed emission factor for Cu may be applied for the shrimp farming industry worldwide. Copper emission factor was generated by analyzing Cu concentrations in aquafeeds and other chemical addictives, in shrimp biomass and in inflow and outflow water and suspended particles and in pond bottom sediments of the largest shrimp farm of Ceara´ State NE Brazil, located at the Jaguaribe River estuary, latitude 4
23 0 S and longitude 37
36 0 W. Table 1 shows the major production parameters of the farm used in the calculation of the emission factor. These parameters are typical of intensive shrimp farming in Brazil and similar to those verified in shrimp farming worldwide. Samples for Cu determination were collected during one production cycle using clean procedures. Water samples in the inflow canal, inside two ponds and in the outflow canal