Local and interannual variations in mercury and cadmium in eggs of eight seabird species of the Sinaloa coast, México

Seabird species

In total, 416 eggs were collected from 8 seabird species in 5 coastal ecosystems located in Sinaloa State, México, (Table 1 and Figure 1). Egg collection was carried out on 1 d at each nesting site, so eggs with different developmental stages were collected. The nesting chronology was different among the 8 species; Pelecaniformes and Suliformes began nesting in the fall (November–December) and concluded in the spring or summer of the following year. The sampling period for these species included 2 breeding seasons, 2011 to 2012 and 2012 to 2013, and each season was identified by the last year of each period. The nesting period for Charadriiformes lasts for 3 to 4 mo, between spring and summer each year. Sampling was carried out in the 2 breeding seasons (2012 and 2013) for royal tern and Heermann's gull and in the 3 breeding seasons (2010, 2012, and 2013) for laughing gull. In 2010, only laughing gull eggs were collected from the SMB site. For the other 2 seasons, eggs were collected from 8 species at 3 different locations during 2012 and in 5 different locations during 2013. The species, species abbreviations, sampling locations, and sample numbers are shown in Table 1.

Sampling locations

The sampling locations are situated in the southeast region of the Gulf of California along the coastline of Sinaloa State. The FSI site (25°26′12″N, 109°22′38″W) is a rocky island located 27 km from the northern coast of Sinaloa; it is stripped of vegetation and is the ecosystem most influenced by pelagic conditions 25. The largest colony of blue-footed booby is found there, with approximately 2500 nesting pairs. Colonies of Heermann's gull and brown booby have 2000 and 1200 pairs, respectively. The colony of double-crested cormorant is smaller and varies between 60 and 90 nesting pairs yearly 17. Eggs from these 4 species were collected from FSI. Ohuira Bay is a coastal saltwater lagoon (23°37′46″N, 108°58′54″W). Eggs from double-crested cormorant were collected on Patos Island (25°37′10.8″N, 109°00′45.7″W), which is located inside the bay at a distance of 3 km from Topolobampo town 26. Eggs were collected from a site where approximately 50 nests of double-crested cormorant are found. At the NMS site, eggs from brown pelican and double-crested cormorant were collected on Pájaros Island (25°22′38″N, 108°42′13″W), where 350 nests of double-crested cormorant have been reported 27; during egg sampling, several hundred nests of brown pelican were observed.

The largest coastal lagoon in Sinaloa is the SMB site, extending over 1350 km² and including several islands, vast mudflats, marshlands, and mangroves 28. Eggs were collected from 3 of the islands. At El Rancho Island (25°09′14″N, 108°22′19″W), eggs from blue-footed booby, royal tern, laughing gull, and Heermann's gull were collected. The colonies of the first 3 species are large at this site, with approximately 3000, 8000, and 2000 nesting pairs reported in 2012, respectively; however, the breeding pattern of Heermann's gull varies, with only 17 nests in the same year 18. In the central area of the bay, eggs from double-crested cormorant, magnificent frigatebird, and brown pelican were collected at El Mero (25°05′34″N, 108°14′58″W) and Pájaros (25°04′44″N, 108°13′52″W) Islands, with approximately 3000, 8000, and 2000 nests, respectively 18. Mazatlán Bay is a semi-enclosed body of water that includes 3 islands and 2 islets located near the coast. Brown pelican eggs were collected from Pájaros Island (23°15′16″N, 106°28′36′W), which is located 0.9 km from Mazatlán City.

Determination of Hg and Cd levels

After collection, the sampled eggs were transported to the Ecotoxicology Laboratory at the Research Center for Food and Development, AC, Mazatlán Unit, where linear dimensions (length and width) were measured, and the egg contents were removed. The egg content was categorized by visual observation according to the following stage of embryo development: no development, initial (embryos without fully developed limbs), intermediate (embryos with fully developed limbs, but with translucent skin that still reveals organs), and advanced (embryos fully formed). The total contents of each egg (white, yolk, and/or embryo) were lyophilized, subsequently ground in a ceramic mortar, and digested to determine the metal content. The moisture content of each sample was calculated by subtracting the final weight after lyophilization from the initial fresh weight, using an analytical balance. First, 0.25 g of the dried and homogenized content of each egg was digested with 5 mL of 50% HNO₃. The resulting samples were transferred to a digestion system (CEM model MARSX), where a second digestion was performed by adding 3 mL of H₂O₂, in accordance with US Environmental Protection Agency (USEPA) method 3052 29. Digested samples were analyzed in an atomic absorption spectrophotometer (PerkinElmer model 1100B). The Hg concentration was determined in a hydride generator following USEPA method 7471A 30, and the Cd concentration was obtained by flame atomic absorption using USEPA method 7130 31. The Hg and Cd concentrations were calculated in µg/g of dry weight and then converted to wet weight using the moisture content of the sample to facilitate comparisons with the literature. For quality assurance/quality control purposes, a blank solution, a duplicate, and a certified reference material for trace metals (DOLT-4 Dogfish Liver) from the National Research Council of Canada were digested with every batch of 10 samples and analyzed with the rest of the samples. Detection limits were 0.0002 μg/g for Hg and 0.5 μg/g for Cd. The quality assurance/quality control results suggest that the average percentage recoveries were 99.8% for Hg and 94.0% for Cd. The maximum percentage relative difference values for Hg were 10.4% and 14.9% for Cd.

Data analysis

Considering that the development stage was not homogeneous and that the results were initially calculated on a dry weight basis, before applying statistical analyses for comparisons among species, sites, and breeding seasons, the variation of egg dry weights of each species was analyzed. This was performed using general linear models, which included the dry weight as the dependent variable, the development stage (without development, initial, intermediate, and advanced) as the factor, and the volume of eggs as the covariate. In 6 species (blue-footed booby, brown booby, double-crested cormorant, magnificent frigatebird, brown pelican, and Heermann's gull), it was found that the dry weights were different between stages of embryonic development. Dry weight decreased gradually between development stages (∼12% between nondevelopment stage of eggs and advanced stages). Therefore, concentrations of Hg and Cd were standardized to estimate the initial development values for each species (using the mean decline in dry wt controlled by egg volume). These standardized concentrations were converted to wet weight, using the moisture content of eggs.

To perform the statistical analysis, below-detection concentrations were set to one-half the detection limit. Concentration values of Hg and Cd were transformed (log x + 1) to minimize the effect of working with a scale involving several orders of magnitude 32. The transformed data were used for normality tests. Data are reported as geometric means; the arithmetic means are presented only for comparisons with other studies. A summary of the statistical tests performed in the present study is shown in Supplemental Data, Table S1. The analyses were applied in the same manner for Hg and Cd concentrations and were included as dependent variables in all cases. Two-way analyses of variance (ANOVAs) were applied to compare the concentrations of these metals in eggs of 3 species (blue-footed booby, brown booby, and double-crested cormorant) on FSI and 6 species (blue-footed booby, double-crested cormorant, magnificent frigatebird, brown pelican, royal tern, and laughing gull) on SMB during 2 breeding seasons (2012 and 2013). One-way ANOVAs were applied to include Heermann's gull only in the interspecific comparison for the 2012 season on FSI and for 2013 season in SMB. A Student's t test was used to compare concentrations of Hg and Cd in eggs of double-crested cormorant and brown pelican in NMS during the 2013 season. The laughing gull eggs collected in SMB were compared using 1-way ANOVAs to determine the differences among 3 sampling periods (2010, 2012, and 2013). Intersite differences in blue-footed booby and double-crested cormorant were analyzed using 2-way ANOVAs during the 2012 and 2013 seasons. One-way ANOVAs were applied to include NMS in the intersite comparison during the 2013 season for double-crested cormorant and for intersite comparison in brown pelican eggs in the same season. Scheffé post hoc tests were applied after the ANOVAs were performed, to identify homogeneous groups. Results with p < 0.05 were considered to be significant. The software Statistica 12.0 (StatSoft) was used to perform the statistical analyses.

Comparison of Hg and Cd levels between species

Greater concentrations of Hg were observed in predominantly piscivorous species, with lower concentrations in both species of gulls. In particular, brown pelican eggs presented the highest concentrations of Hg, followed by brown booby, blue-footed booby, and double-crested cormorant (F_7,408= 18.7, p < 0.001). There were interspecific differences in Cd concentrations: the eggs of magnificent frigatebird and both species of gulls presented the highest concentrations of Cd, whereas the lowest concentrations were detected in the eggs of royal tern and brown pelican (F_7,408 = 10.6, p < 0.001; Table 2).

When the same species were compared in the same coastal system, interspecific tendencies similar to those mentioned in the previous paragraph were observed. During 2012 on FSI, the concentrations of Hg were higher in the eggs of blue-footed booby, brown booby, and double-crested cormorant compared with those observed in Heermann's gull (F_3,29 = 20.0, p < 0.001), whereas no significant differences in Hg concentrations were found between these species in 2013 (F_2,69 = 1.3, p = 0.275). The concentrations of Cd in eggs during 2012 were higher in blue-footed booby and Heermann's gull compared with brown booby and double-crested cormorant (F_3,29 = 3.4; p = 0.031). The concentrations of Cd in the 2 species of Sula and cormorants were similar during 2013 (F_2,69 = 0.5, p = 0.638; Figure 2).

There were significant interspecific differences in Hg concentrations in eggs collected during 2012 on SMB (F_5,252 = 8.0, p < 0.001); the lowest concentrations were detected in laughing gull, whereas the rest of the species presented similar values. Differences between species were also detected in 2013 (F_6,120 = 12.0, p < 0.001), and a similar tendency to the former case was observed; the lowest values were noted in laughing gull, whereas the highest values were observed in brown pelican and double-crested cormorant. Cadmium concentrations in eggs collected during 2012 were lower in all species (F_5,252 = 25.8, p < 0.001). In 2013, the eggs of 5 species (double-crested cormorant, magnificent frigatebird, brown pelican, laughing gull, and Heermann's gull) presented elevated Cd concentrations. These values differed significantly from those observed in royal tern and blue-footed booby eggs (F_6,120 = 41.3, p < 0.001; Figure 3). No significant differences were observed between Hg and Cd concentrations in double-crested cormorant and brown pelican eggs during 2013 in NMS (Hg, Student's t test, df = 20 = −1.9, p = 0.070; Cd, Student's t test, df = 20 = −1.5, p = 0.160; data not shown).

Interannual variations of Hg and Cd levels

No interannual significant differences in Hg concentrations in eggs of boobies and cormorants were observed on FSI (F_1,69 = 1.3, p = 0.275). In SMB, the eggs of brown pelican and laughing gull presented higher Hg concentrations during 2013 (F_1,252 = 8.0; p < 0.001). The interannual differences were more noticeable for Cd concentrations, which presented greater values during 2013 on both FSI (F_1,69 = 50.0, p < 0.001) and SMB (F_1,252 = 211.5, p < 0.001; Figures 2 and 3). On SMB, Hg and Cd concentrations were measured in laughing gull during 3 breeding seasons; interannual differences were observed for both metals. Mercury concentrations were significantly lower during 2012 compared with 2010 and 2013 (F_2,70 = 10.2, p < 0.001), and the Cd concentrations were significantly higher during 2013 compared with the other previously studied seasons (F_2,70 = 24.3, p < 0.001; Table 3).

Comparison between locations

No differences were observed between Hg concentrations in blue-footed booby and double-crested cormorant eggs from different areas during 2012 (F_1,84 = 0.1, p = 794; F_2,57 = 1.5, p = 0.223, respectively). Similarly, no differences were detected in blue-footed booby (F_1,84 = 0.1, p = 794) and brown pelican (F_2,35 = 2.4, p = 0.109) eggs during 2013. However, Hg concentrations in double-crested cormorant were significantly greater in the eggs from NMS and lower in the eggs from FSI (F_3,37 = 8.7, p < 0.001; Figure 4). The 3 species assessed presented different Cd concentrations depending on the location: for blue-footed booby, the lowest Cd concentrations were found on SMB during the 2 breeding seasons (F_1,84 = 30.6, p < 0.001); for double-crested cormorant, no differences were observed between locations during 2012 (F_2,57 = 0.8, p = 0.460), but differences were noted during 2013, when Cd concentrations were lower in NMS (F_3,37 = 6.1, p = 0.002). Finally, Cd concentrations in brown pelican were significantly greater in the eggs collected on SMB compared with the eggs from NMS and MZB, which presented similar values (F_2,35 = 36.3, p < 0.001; Figure 5).

Interspecific analysis

In the present study, Hg and Cd concentrations presented interspecific differences associated with the trophic positions. As predicted, Hg concentrations were high in species feeding mainly on fish and low in omnivore species. The relationship between trophic level and Hg content in seabirds' tissues throughout different world regions has been documented 7, 19, 33. Seabirds that feed at higher trophic levels tend to accumulate larger quantities of this metal 7. The Suliformes and Pelecaniformes species of coastal Sinaloa presented the highest Hg concentrations in eggs collected from 2 different ecosystems. On FSI, high Hg concentrations were observed in species of Sula and cormorants, whereas Hg concentrations were low in gulls. In the Gulf of California, brown booby and blue-footed booby feed mainly on sardines and anchovies 34, 35, but double-crested cormorant also feeds on demersal fish 36. The diet of Heermann's gull in the central region of the Gulf of California also consists of sardines and anchovies, but most fish consumed by this species are juvenile 37. This difference could explain why the Hg concentration in its eggs was lower than concentrations detected in Sula and cormorants, and this finding is in accordance with the Hg content found in several pelagic species of fish, which increases with size and developmental stage 38.

Brown pelican eggs presented Hg concentrations up to 2-fold higher than the average concentration for the other species, including piscivorous seabirds, during 2013 on SMB. It has been documented that larger fish normally contain higher Hg content than smaller fish 39, which could explain the greater concentrations found in eggs of brown pelican, as the diet of this species includes fish of up to 160 mm in length on average (range: 97–257 mm 40). The size of the prey can also explain the lower Hg concentrations in royal tern compared with those observed in other piscivorous species, as the tern species are relatively small birds that feed on smaller fish 19. In relation to omnivorous species, laughing gull eggs presented the lowest Hg concentrations, which is attributed to their varied diet, consisting of shrimp, anchovies, and insects in the SMB region (E. González-Medina, National Autonomous University of México, Mazatlán, México, unpublished data).

Similar trends and concentrations related to the trophic level have been reported in multispecific studies conducted elsewhere. In the Portuguese Atlantic, the lowest Hg concentrations were found in eggs of terns (common tern [Sterna hirundo)] and roseate tern [Sterna dougallii], ≈1.0 μg/g dry wt) and gulls (yellow-legged gull [Larus cachinnans], 1.4 μg/g dry wt), whereas higher concentrations were recorded in Bulwer's petrel (Bulweria bulwerii, 1.7 μg/g dry wt), a species that consumes mostly mesopelagic prey 41. In Barnegat Bay (NJ, USA), Hg concentrations were lower in gulls (herring gull [Larus argentatus)] and great black-backed gull [Larus marinus], ≈0.35 μg/g dry wt), intermediate in terns (common and Foster tern [Sterna forsterii]; 1.0 μg/g and 1.7 μg/g dry wt, respectively) and greater in black skimmer (Rynchops niger; 2.0 μg/g dry wt), which feed on larger fish 7.

The prediction that eggs of omnivorous species would contain greater Cd concentrations was also confirmed. High Cd concentrations are common to species feeding on lower trophic levels, including marine invertebrates (mollusks and crustaceans) 15, 19. Two omnivorous species of gulls presented high Cd concentrations on the Sinaloa coast 20, 21. However, the concentrations observed in magnificent frigatebird eggs presented similar values, even though this species feeds mainly on fish. This finding could be attributed to the fact that the 2 species of gulls are kleptoparasited by magnificent frigatebird in the SMB region; however, magnificent frigatebird has an extensive foraging range 42 and a very diverse diet 43, possibly because this species comes into contact with pollutants from other regions. In contrast, the metal concentration in royal tern eggs was extremely low. It has been suggested that metals such as Cd tend not to accumulate in the eggs of some species 16. Nevertheless, other studies suggest that Cd concentrations in females of several species of terns are not related to Cd concentrations found in their eggs 44. This may be the reason why only a few studies have been carried out on the levels of this metal in seabirds' eggs.

Interannual variations

In the present study, interannual differences in Cd concentrations were found in eggs of most seabirds species, with occurrences markedly greater in 2013 than in 2012. In contrast, Hg concentrations were higher in 2013 only in brown pelican and laughing gull. This suggests that interannual variability could be related to drastic changes in oceanographic conditions, specifically when a more intense upwelling such as in 2013 occurs (data not shown). In addition, the influence of local factors such as storms, changes in prey availability, and mainland inputs of sediments and wastewater to coastal ecosystems during rainy seasons could also be related. Further research on the influence of potential sources of metals (e.g., perhaps using stable isotope methodologies), oceanographic and climatological parameters, and foraging ecology of species is desirable to assess the bioaccumulation in eggs.

Spatial variability

Several studies have proposed that widely distributed species of seabirds can be used to identify regional and local differences in metal concentrations 19. The pollutant contents in samples collected in different points of the range of seabird species provide relevant information about changes in food availability and the presence of point sources of pollution 45. In the present study, spatial variability of Hg and Cd concentrations was found in 3 widely distributed seabird species on the Sinaloa coast (blue-footed booby, double-crested cormorant, and brown pelican), but no spatial trends were apparent. In blue-footed booby and double-crested cormorant, it was assumed that eggs from colonies located in coastal lagoon systems situated near agricultural areas would present higher metal concentrations. However, such low values were only observed for Hg concentrations in double-crested cormorant eggs collected during 2013. In this particular case, larger Hg concentrations were observed at NMS and SMB, as these systems are affected by agricultural waste discharges from the Guasave, Évora, and Culiacán valleys 46, 47. However, the aforementioned assumption did not hold for Cd concentrations. The blue-footed booby eggs collected on FSI contained greater Cd concentrations than eggs from SMB. It has been suggested that the bioavailability of Cd in the Gulf of California is conditioned by upwelling events during the winter season 4. For this reason, the spatial variability of Cd concentration in blue-footed booby eggs could be explained by the different oceanographic conditions present in each particular location. The boobies from SMB normally feed near the coast, but the boobies from FSI obtain their food from oceanic waters 35, where the upwellings bring Cd to the sea surface, making it available to seabirds. This condition also could explain the high Cd concentrations in double-crested cormorant eggs collected on FSI during 2013.

Brown pelican eggs were collected in 3 locations of coastal lagoons. Mercury concentrations were similar in all 3 locations, but the Cd concentration obtained from SMB was 3 times greater than the concentrations obtained at NMS and MZB. Santa María Bay is the largest lagoon system of the Sinaloa coast and is permanently connected to the Gulf of California through 2 entrances 28. The lagoon receives water from 3 rivers (Sinaloa, Mocorito, and Tule) that are affected by agricultural, urban, and industrial waste discharges 48. Therefore, it is possible that factors such as the inflow of continental water, hydrodynamic processes, and the water exchange between the lagoon and the ocean cause Cd to be available in higher quantities in this coastal ecosystem.

Brown pelican and double-crested cormorant eggs collected from SMB presented high Cd concentrations, but the Cd concentrations found in blue-footed booby were low. Although the latter species is piscivorous, both their diet and their foraging ecology present characteristics that differ from those of the other 2 species: the foraging range of blue-footed booby is twice that of double-crested cormorant and brown pelican 49-51. Thus, double-crested cormorant and brown pelican obtain their food exclusively from the bay, where greater Cd concentrations are present 5, whereas blue-footed booby can catch their prey in 2 different areas (pelagic zones and coastal areas), which reduces their exposure to metals, and thus lower accumulation in their eggs.

Ecotoxicological considerations for seabirds in Sinaloa

Research suggests that Hg concentrations between 0.5 μg/g and 1.0 μg/g wet weight in seabirds' eggs could cause mortality, malformations, and neurological effects 9. On the coast of Sinaloa, only brown pelican eggs presented Hg concentrations above the levels considered harmful for seabirds. During the 2013 breeding season, 26% of the brown pelican eggs collected in the 3 sampling locations presented values greater than 1.0 μg/g (range: below detection limit to 1.87 μg/g) of this metal. This finding is particularly relevant because only brown pelican populations from Sinaloa have declined in recent years 18; therefore, additional studies are needed to assess the ecotoxicological risk of Hg to brown pelican colonies in the coastal area of Sinaloa.

It has been documented that even Cd concentrations below 1.0 μg/g wet weight can cause chronic effects, such as kidney damage, behavioral alterations, egg-laying inhibition, and perhaps eggshell thinning 8. In the present study, 61% of eggs of the 2 species of gulls collected from SMB during 2013 presented Cd concentrations greater than 1.0 μg/g (Heermann's gull, 1.83 μg/g [range: 0.59–2.54] and laughing gull, 1.05 μg/g [range: 0.06–2.42]), suggesting that at least parts (61%) of these populations are at potential risk of suffering from chronic effects. It may be appropriate to perform studies specifically focused on the health conditions of the species, their reproductive parameters, and any changes in populations in relation to Cd concentrations in eggs. Because shrimp are the main diet for laughing gull during the egg-laying period (E. González-Medina, National Autonomous University of México, Mazatlán, México, unpublished data), it is important to further investigate Cd concentrations in shrimp collected from the areas where laughing gull is feeding (at SMB and marine waters closer to the colony) because of its high commercial and social importance in the region.

Considering this information, implementation of a permanent monitoring program and extension of the study area to the entire Gulf of California is recommended. Such an effort would facilitate determination of distribution patterns on a regional scale and could evaluate the influence of temporal variability in oceanographic and environmental conditions on the availability of these metals on a temporal scale.