Reptiles and Amphibians: Shy and Sensitive Vertebrates of the Great Lakes Basin and St. Lawrence River

Bioaccumulation of environmental contaminants has been documented in amphibians and reptiles inhabiting the Great Lakes, St. Lawrence River and elsewhere. The effects of pollutants on the physiology and reproduction of amphibians and reptiles has also been reported but this research has largely been restricted to laboratory studies. Much less work has been conducted to quantify the effects of toxic chemical exposures on these cryptic animals in the wild. In the Great Lakes basin and St. Lawrence River, this work has only been performed in detail since 1981 although some samples collected in 1974 indicated a high level of contamination. Results in the 1990s on aquatic salamanders, frogs and turtles indicate that adults and embryos are currently experiencing toxic effects and, in some species and locations, there are indications that population declines are influenced by environmental pollution exposure. In this review, we describe the existing literature on contaminant levels and effects in reptiles and amphibians of the St. Lawrence River and Great Lakes basin.     

Excrement from Heron colonies for environmental assessment of toxic elements

Excrement cast from Great Blue Heron nests was collected during the nesting period of 1978 from four colonies in Washington and Idaho. Cheesecloth strips placed on the ground beneath the nests served as excrement collecting devices. Chemical analysis for lead, mercury and cadmium were performed on dried samples. Lead was the most abundant trace metal found in heron debris. The Idaho colony at Lake Chatcolet had an average concentration of 46 ppm in the beneath-nest samples and 6 ppm in control samples. A heron colony near Tacoma, Washington had beneath-nest samples averaging 28 ppm and control samples averaging 20 ppm. Two colonies located in the interior region of Washington had substantially lower concentrations of lead. The difference observed between colonies was attributed to their associations with a polluted watershed (Chatcolet colony) an interstate highway (Tacoma colony) and an unpopulated largely agricultural area (inland Washington).     

Organochlorine contaminants in seabird eggs from the Pacific coast of Canada, 1971–1986

Eggs were collected from seven seabird species at colonies on the British Columbia coast from 1983 to 1986 and analyzed for organochlorine contaminants. Total PCB levels (wet weight) were highest in double-crested cormorants (Phalacrocorax auritus) from the Fraser estuary (2.91 mg kg^-1) and the Strait of Georgia (3.79 mg kg^-1). Highest DDE levels were in fork-tailed storm-petrels (Oceanodroma furcata) from the Queen Charlotte Islands (1.68 mg kg^-1). Organochlorine levels were generally lower in eggs from the mid 1980s than in those collected in the early 1970s. Organochlorine levels in Pacific alcids and hydrobatids foraging in offshore locations were compared to those in the same or ecologically similar species from the Canadian Atlantic coast. DDT- and HCH-related compounds were higher in Pacific populations while levels of dieldrin, oxychlordane, and HCB were generally lower. With the exception of -HCH, levels of all measured organochlorines were lower in cormorants breeding in the Fraser River estuary than in cormorants from the St. Lawrence River estuary on the Atlantic coast.     

Spatial Patterns and Rankings of Contaminant Concentrations in Herring Gull Eggs from 15 Sites in the Great Lakes and Connecting Channels, 1998–2002

Mean values of eight contaminants in Herring Gull (Larus argentatus) eggs were calculated for 15 Great Lakes sites for the 5 year period 1998–2002. The sites were ranked according to the concentrations of each of seven compounds relative to fish flesh criteria for the protection of piscivorous wildlife, and a single overall rank of contamination was calculated for each site. Based on this weighted ranking scheme, we found that sum PCBs, dioxin and DDE contributed the most (60.2, 30.5% and 8.5%, respectively) to the overall rankings. The weighted ranking scheme showed that eggs from Channel-Shelter Island (Saginaw Bay, Lake Huron), Strachan Island (St. Lawrence River) and Gull Island (northern Lake Michigan) ranked as the three most contaminated sites, while Agawa Rocks (eastern Lake Superior), Chantry Island (southern Lake Huron) and Port Colborne (eastern Lake Erie) ranked as the three least contaminated sites. Two of the three most contaminated sites are Areas of Concern; none of the three least contaminated sites are Areas of Concern.     

Monitoring the Elimination of Persistent Toxic Substances from the Great Lakes; Chemical and Physiological Evidence from Adult Herring Gulls

To assess progress towards virtual elimination of PCBs, DDE, dieldrin and Mirex and their associated physiological effects, we compared their concentrations in pooled livers of adult herring gulls (Larus argentatus) repeatedly sampled at 8 Great Lakes colonies and a reference colony on the Atlantic coast between 1974 and 1993. We measured the relative thyroid mass and concentrations of highly carboxylated porphyrins and retinyl palmitate in the liver of each individual. PCBs, dieldrin and mirex declined in 7 of 8 colonies while DDE decreased in six. The greatest decreases occurred pre-1985. PCBs and DDE did not decrease in gulls from Middle Island in western L. Erie. Middle Island and Saginaw Bay had the highest concentrations of PCBs of 11 Great Lakes colonies in the 1990s. Thyroids of gulls from Great Lakes colonies were slightly enlarged but the degree of enlargement has decreased over time. In 1991, gulls from Great Lakes colonies had slight to moderately elevated concentrations of highly carboxylated porphyrins. In the early 1990s, hepatic stores of retinyl palmitate were very seriously depleted in gulls from the Detroit River, western basin of Lake Erie, and Lake Ontario, reflecting decreased availability and altered storage. We conclude that PCBs and/or other persistent toxic substances in the food of herring gulls have not been virtually eliminated.     

Evaluation of phosphorus loads from Ontario municipal wastewater treatment plants

Results are presented from an evaluation of the nature and accuracy of phosphorus loads discharged by Ontario municipal wastewater treatment plants into the Great Lakes. Data were examined for the 96 plants treating flows in excess of 4546 m³d^-1 for the period 1981 to 1985. For the Lake Erie, Lake Ontario/St. Lawrence River and Upper Great Lakes basins, total basin phosphorus loads were classified according to type of wastewater treatment system, the type of chemical added for phosphorus removal, plant capacity, and sampling frequency. Load estimation techniques were compared using the 1985 daily data from three representative treatment plants. Annual phosphorus loads were compared using the complete data records for the plants and using Monte Carlo techniques to simulate an incomplete data record typical of once per week effluent phosphorus sampling. Potential sources of bias were identified in annual phosphorus load estimates from municipal treatment plants.     

Reproductive and Physiological Effects of Environmental Contaminants in Fish-Eating Birds of the Great Lakes: A Review of Historical Trends

During the 1950s and 1960s, reproductive failures and population declines were observed in fish-eating birds such as gulls, terns, cormorants, herons, and eagles in the Great Lakes. DDE-induced eggshell thinning contributed to these declines, but other factors such as embryo toxicity also were implicated. With reduced releases of many pollutants, reproduction recovered in some species. However, biomonitoring during the 1980s and 1990s indicates continuing effects at highly contaminated sites. Improved analytical techniques and bioassays have allowed the assessment of the total dioxin-like toxicity of complex mixtures of organochlorines (TCDD-equivalents). Developmental defects such as embryo mortality, deformities, and edema have been associated with dioxin-like PCBs in several avian species. Improved biochemical techniques have allowed the measurement of biomarkers that detect physiological alterations associated with contaminants. Specific biomarkers studied in Great Lakes birds include cytochrome P-450 monooxygenases, highly carboxylated porphyrins, thyroxine, vitamin A, and immune function. Reproductive and physiological alterations are associated with population-level effects in Caspian terns and bald eagles that feed on highly contaminated fish. Biomonitoring using biomarkers and population-level measures in fish-eating birds will continue to be important for assessing the effects of contaminants on the Great Lakes ecosystem.     

Monitoring perchlorate exposure and thyroid hormone status among raccoons inhabiting a perchlorate-contaminated site

Perchlorate is a water soluble anion that is readily accumulated in vegetation. It inhibits uptake of iodide into thyroid gland tissue, thereby reducing production of thyroid hormones. Potential raccoon food items including berries, fish, and vegetation collected at a contaminated site contained quantifiable concentrations of perchlorate as determined by ion chromatography. Therefore, we monitored resident raccoons for exposure to perchlorate by examining plasma perchlorate and thyroid hormone concentrations. Resulting analytical data failed to demonstrate perchlorate exposure among raccoons that likely consumed food items collected along perchlorate-contaminated water bodies. There were no correlations between triiodothyronine or thyroxine and thyroid stimulating hormone concentrations, but triiodothyronine concentrations in raccoon plasma were significantly higher in 2000 than in 2001 (p = 0.0081). These data suggest that natural attenuation and remedial efforts initiated in January of 2001 may have reduced perchlorate exposure among raccoons inhabiting this site from 2000 to 2001. Temporal, spatial, and analytical factors limited our ability to quantify exposure among raccoons, however, our data do not indicate that raccoons currently inhabiting this site are at risk for significant exposure to perchlorate and subsequent effects.     

The Niagara River: A water quality management overview

The Niagara River constitutes part of the Laurentian Great Lakes and St. Lawrence River system which represents approximately 80% of North America's supply of surface fresh water. The river is a major source of water for industry, municipalities, recreation and power generation and is the link between Lakes Erie and Ontario. The river forms part of the Canada-U.S. border and falls under the jurisdiction of both countries.The massive industrialization of the region surrounding the river has led to a typical resource use conflict situation in which pollution of the river continues to be a major public concern.A number of constitutional, institutional and jurisdictional factors make the management of the Niagara River an involved and complicated matter. The interests, intent, philosophies, laws and regulations are not necessarily the same among the numerous jurisdiction involved. Despite these differences, however, Canada and the United States have succeeded in developing and implementing a model cooperative international management plan for the river. An overview of the main international aspects relating to the development and implementation of this plan, the Niagara River Toxics Management Plan, is presented.     

Trends of total cadmium, copper, and zinc in the zebra mussel (Dreissena polymorpha) along the upper reach of the St. Lawrence River: 1994—2005

This study examines spatial and temporal variability of bioaccumulation of cadmium, copper, and zinc in tissues of zebra mussels in the upper reach of the St. Lawrence River which originates at the outflow of Lake Ontario. It was hypothesized that concentrations of these metals in mussel tissues would decline over time and decrease with increasing distance from the outlet of Lake Ontario as a result of on-going efforts to reduce contaminant discharges into the lake. Size of mussels was also evaluated as a factor influencing bioaccumulation. Mussels were collected annually in October from six sites from 1994 to 2005, including one site near a local industry. Individuals were grouped into five or more size classes per site in each year. Soft tissues were analyzed for total cadmium, copper and zinc. Concentrations of cadmium and copper in tissues varied significantly both spatially and temporally. Cadmium concentrations were elevated at most sites; copper concentrations were moderately elevated compared with other studies in the Lake Ontario basin and St. Lawrence River. Zinc showed the most uniformity in mussels possibly due to internal regulation and to low levels of environmental exposure. Animal size correlated with copper concentrations of tissues in approximately 30% of samples but infrequently for cadmium and zinc. Cadmium and copper levels were found to decline downstream over time. Inter-annual variability of metal concentrations in mussel tissues suggests utilization in long-term monitoring programs to discern significant trends.     

Modeling Wetland Plant Community Response to Assess Water-Level Regulation Scenarios in the Lake Ontario – St. Lawrence River Basin

The International Joint Commission has recently completed a five-year study (2000–2005) to review the operation of structures controlling the flows and levels of the Lake Ontario – St. Lawrence River system. In addition to addressing the multitude of stakeholder interests, the regulation plan review also considers environmental sustainability and integrity of wetlands and various ecosystem components. The present paper outlines the general approach, scientific methodology and applied management considerations of studies quantifying the relationships between hydrology and wetland plant assemblages (% occurrence, surface area) in Lake Ontario and the Upper and Lower St. Lawrence River. Although similar study designs were used across the study region, different methodologies were required that were specifically adapted to suit the important regional differences between the lake and river systems, range in water-level variations, and confounding factors (geomorphic types, exposure, sediment characteristics, downstream gradient of water quality, origin of water masses in the Lower River). Performance indicators (metrics), such as total area of wetland in meadow marsh vegetation type, that link wetland response to water levels will be used to assess the effects of different regulation plans under current and future (climate change) water-supply scenarios.The Canadian Crown reserves the right to retain a non-exclusive, royalty free licence in and to any copyright.     

Targeted Sustainable Development: 15 Years of Government and Community Intervention on the St. Lawrence River

From 1988 to 2003, the St. Lawrence Action Plan, a Canada – Quebec cooperation agreement on the St. Lawrence River, helped to mobilize a still-growing number of stakeholders in the conservation and protection of this great river and generated tangible results in a number of areas of intervention. The successes enjoyed in the area of industrial, agricultural and urban clean-up, in protecting plant and animal species and their habitats, and in acquiring new knowledge on the state and trends of the ecosystem reflect the sustained efforts of government and non-government partners over the course of those 15 years. The committed involvement of local communities provides assurance of the sustainable development of this vast ecosystem.The Canadian Crown reserves the right to retain a non-exclusive, royalty free licence in and to any copyright.     

Metal Contamination in Zebra Mussels (Dreissena Polymorpha) Along the st. Lawrence River

In order to evaluate the use of zebra mussels as biomonitors for metal bioavailability in the St. Lawrence River, we tested the hypothesis that the concentrations of 11 metals in zebra mussels vary significantly between sites along the river and that the season of collection and body size affect metal bioaccumulation. Mussels were collected at 14 sites during June 1996 and at monthly intervals at one site. Specimens were grouped in three size classes and their soft tissue was analyzed for As, Ca, Cd, Cr, Hg, Mn, Ni, Pb, Se, and Zn. Significant size effects were found for Ca, Cd, Cr, Cu, Ni and Zn. Spatial and seasonal variations in bioconcentration were significant for all metals. Spatial patterns in contamination that corresponded to known point sources of pollution or hydrology of the river were identified by principal component analysis. Seasonal variations can be attributed to the reproductive cycle of mussels and hydrological variability of the river. In comparison with values reported for zebra mussels in other contaminated sites in North America and Europe, levels of metal in the St. Lawrence River are low or intermediate. Our results show that when controlled for size and seasonal effects, zebra mussels represent a useful biomonitor for metal availability in the river and may offer an interesting alternative to native mussels and fish for such a role. Local contamination by some toxic metals is still a cause for concern in the St. Lawrence River.     

Knowing, Mapping and Understanding st. Lawrence Biodiversity, with Special Emphasis on Bird Assemblages

Environment Canada and the Québec Department of the Environment, partners in the St Lawrence Vision 2000 Action Plan, set out to prepare a compendium of knowledge of the flora and fauna of the St Lawrence and to identify potential conservation sites. The resulting Portrait is an Internet site that presents the current knowledge base of the river's ecological and biological diversity (http://lavoieverte.qc.ec.gc.ca/faune/biodiv/index.html). The Portrait provides information on over 5,000 species of flora and fauna. On the website, you will find a detailed ecological analysis of the richness, rarity and vulnerability of several broad groups of plant and animal species. Furthermore, you will find a list of species for each of the 700 survey units and a distribution map for 2,500 species recorded along the St Lawrence., in atlas form, along with a detailed conservation plan. The plan encompasses the most unique and heterogeneous landscapes of the St Lawrence, some of which have no protection at present. The Portrait provides an overview of the sites that are currently protected by public agencies and private-sector organisations and identifies new sites of interest for conserving biodiversity and protecting species at risk. This paper exposes the content of this extensive compendium on the biodiversity of the St. Lawrence. For conciseness, it presents some of the analyses conducted on birds to illustrate a few of the analytical approaches that were used. Then, the information on species richness and concentration areas for priority species of vascular plants, breeding birds and herpetofauna will serve to identify the terrestrial sites of significance for biodiversity. Finally, a similar approach having been applied to the aquatic environnement, we will conclude with a conservation plan that identifies the terrestrial and aquatic ecosystems, and the geographic sites where the most important elements of St. Lawrence biodiversity are concentrated. Our analysis of the biodiversity of the most thoroughly studied taxa of the St. Lawrence clearly shows the importance of wetlands, particularly those located at the mouths of rivers or within archipelagos or delta complexes, such as the groups of islands and channels found at both ends of the Montréal Archipelago. These aquatic landscapes are sites of intense biological production, combining in a small geographical area spawning, nursery and feeding grounds for a large number of fish species and breeding, rearing and foraging areas for aquatic birds. Variable flooding conditions, associated with seasonal flooding or daily tidal fluctuations, create a complex mosaic of wetland and aquatic habitats. Although wetlands occupy only a small area in comparison with terrestrial habitats, they support a large number of rare plant and animal species in relation to their size. At present, 10% of the vascular flora and 27% of the herpetofauna of the St. Lawrence are at risk. In the case of reptiles and amphibians, the situation is especially worrisome because nearly all of the most threatened species live in a narrow band along the river corridor. Not only is this the sector that is under the greatest pressure from human development, very little public land remains here, making it difficult to create protected areas. Increased participation by non-governmental organisations and individuals, through private stewardship arrangements, is an essential precondition for completing the network of conservation areas in this part of the St. Lawrence. Along the estuary and the Gulf, habitat integrity has not been affected as much by the expansion of Québec's human population. This is a vast territory, and sites have been identified with a view to making up for the deficiencies in the present network of protected sites in terms of representing biodiversity.     

Geographical distribution of organochlorine contaminants and reproductive parameters in Herring Gulls on Lake Superior in 1983

As part of the Great Lakes International Surveillance Plan, 1978–83, egg contaminant levels and reproductive output were determined for Herring Gull colonies on Lake Superior in 1983. Since 1974, the Herring Gull has been widely used in the Great Lakes as a spatial and temporal monitor of organochlorine (OC) contaminant levels and associated biological effects. Most eggs contained a wide range of OCs, the main compounds being DDE, polychlorinated biphenyls (PCBs), dieldrin, heptachlor epoxide, oxychlordane, hexachlorobenzene and mirex. Levels of an additional ten OCs and five polychlorinated dibenzo-p-dioxin (PCDD) congeners were also determined for some sites. Overall, levels varied significantly among colonies, but there was no obvious relationship to spatial distribution of contaminants in sediments or fish species. OC levels in eggs had declined by up to 84% since 1974. Eggshells were only 8% thinner than before the introduction of DDT, and shell thinning was not a cause of breeding failure. Average reproductive output varied from 0.15 to 1.57 young per apparently occupied nest in 1983: at 56% of colonies the value was below that thought necessary to maintain stable populations. The main causes of failure were egg disappearence and cannibalism of chicks. Despite this, the population appeared to have been increasing at about 4% per annum. Reduced availability of forage fish during the early 1980s was the most likely reason for the poor reproductive output in 1983.     

Lead and cadmium contaminations in feathers of heron and egret chicks

In birds, concentrations of heavy metal contaminants were associated with their diet and environmental quality. This study presents concentrations of lead and cadmium in feathers and the relationship between concentrations in the feather and stomach content of three species of heron and egret chicks. Cadmium concentrations were higher in the Little Egret (geomean 0.75 μg/g dry weight) and Black-crowned Night Heron (geomean 0.88 μg/g dry weight) chicks than in Grey Heron chicks (geomean 0.48 μg/g dry weight). Lead and cadmium concentrations had significant correlations between feathers and stomach contents of heron and egret chicks. The mean lead and cadmium concentrations in feathers were within the background levels for wild birds. Heron and egret species forage around breeding colony (4–7 km). Lead and cadmium concentrations in heron and egret chicks were moderately associated with those of prey concentrations. These results suggest that contaminant concentrations such as lead and cadmium in feathers of heron and egret chicks can reflect local heavy metal contaminations and serve as a great biological indicator for monitoring heavy metal contamination. In addition, the feathers can be a routine monitoring tool over time and space without harming the wild birds.     

Modelling Wetland Bird Response to Water Level Changes in the Lake Ontario – St. Lawrence River Hydrosystem

Lake Ontario and St. Lawrence River (LOSL) wetland bird abundance and diversity are greatly influenced by lake and river hydrology. Our study used an interdisciplinary ecosystem approach, blending avian and plant ecology, ecohydraulic, statistical ecology and modelling to evaluate potential impacts of water level fluctuations on indicator species representative of the wetland breeding bird assemblages in the entire LOSL freshwater system. Multi-year (2000–2003) bird surveys captured bird distribution and density in wetland habitats under varying degrees of water inandation, depth and fluctuation. Analyses revealed strong associations between estimated breeding pair densities and plant communities, water depth, and degree of water level fluctuation during the breeding season for a suite of wetland bird species using marsh, wet meadow, shrub swamp and treed swamp habitats. These quantitative associations were used to develop wetland bird performance indicators for use in a LOSL water regulation review study. Several bird species also nest at or near the water surface and are thus vulnerable to nest flooding or stranding. Changes to the seasonal hydrology of Lake Ontario and St. Lawrence River that result in an increased frequency or magnitude of these nest failure events may have a significant impact on regional population sustainability. Long term nest record databases were analyzed to create nesting flooding and stranding probability equations based on water level increases and decreases during the breeding season. These species-specific nesting relationships were incorporated into a reproduction index.Many breeding bird species were strongly associated with specific wetland plant communities. Predicted habitat suitability, as measured by estimated breeding pair density, can also change significantly within a specific wetland plant community based solely on changes in water depth during the breeding season. Three indicator species, Black Tern, Least Bittern and Virginia Rail were selected as key environmental performance indicators for alternate regulation plan comparisons.Water regulation criteria should be such that the long term diversity and abundance of wetland plant communities and frequency of spring flooding in marsh habitats during breeding are not reduced. Magnitude and frequency of water level change during the nesting season (May–July) can also adversely impact reproductive success of many wetland bird species. As such, regulation criteria that increase the seasonal magnitude and frequency of water level change may be detrimental to the long term viability of certain regional breeding bird populations.     

Integrated survey of water pollution in the Suquía River basin (Córdoba, Argentina)

We report a combined two-year seasonal monitoring of Suquía River basin using both chemical parameters and biomarkers measured in Jenynsia multidentata, aiming to correlate external levels of contaminants with the response of oxidative stress biomarkers in this fish. Identified pollution sources correspond to city sewage as well as agricultural and small industry activities downstream from Córdoba city. Physicochemical parameters integrated into a water quality index (WQI) were measured in Suquía River during dry and wet seasons. Ag, Mn, Cu, Cr, Ni, Fe, Pb and Zn were also monitored in water and sediment samples. Biomarkers include detoxication and antioxidant enzymes: catalase (CAT), glutathione peroxidase (GPX), glutathione S-transferase (GST) and glutathione reductase (GR). Enzymes showed a pollution dependent response, with increased activities in fish collected close to the sewage exit and progressive drop further downstream, matching changes in the Water Quality index. The combined use of biomarkers with water quality parameters allowed both the identification of pollution sources and the evaluation of effects of contaminants on the aquatic biota.     

Geographic differences in organic contaminants and stable isotopes (δ¹³C, δ¹⁵N) in thick-billed murre (Uria lomvia) eggs from Alaska

The contents from thick-billed murre (Uria lomvia) eggs collected at four Alaskan colonies in 2002 were analyzed for organic contaminants and carbon δ13C) and nitrogen (δ1?N) stable isotopes. Contaminant concentrations in the eggs varied from below detection limits to 230 ng g?1 wet mass for 4,4'-DDE in one egg from St Lazaria Island in the Gulf of Alaska. Eggs from this colony generally contained higher levels of contaminants and exhibited significantly different patterns compared to eggs from the Bering and Chukchi seas. Stable isotope values also varied geographically; however, these differences appeared to be related to differences in C and N baselines in the food webs instead of differences in prey. Contaminant and stable isotope correlations were inconclusive, suggesting that better information on regional food web differences and differential offloading of contaminants and stable isotopes to the eggs must be obtained before these kinds of data can be fully incorporated into seabird egg contaminant monitoring programs.     

Mercury and other metals in eggs and feathers of glaucous-winged gulls (Larus glaucescens) in the Aleutians

Levels of mercury and other contaminants should be lower in birds nesting on isolated oceanic islands and at high latitudes without any local or regional sources of contamination, compared to more urban and industrialized temperate regions. We examined concentrations of arsenic, cadmium, chromium, lead, manganese, mercury and selenium in the eggs, and the feathers of fledgling and adult glaucous-winged gulls (Larus glaucescens) nesting in breeding colonies on Adak, Amchitka, and Kiska Islands in the Aleutian Chain of Alaska in the Bering Sea/North Pacific. We tested the following null hypotheses: 1) There were no differences in metal levels among eggs and feathers of adult and fledgling glaucous-winged gulls, 2) There were no differences in metal levels among gulls nesting near the three underground nuclear test sites (Long Shot 1965, Milrow 1969, Cannikin 1971) on Amchitka, 3) There were no differences in metal levels among the three islands, and 4) There were no gender-related differences in metal levels. All four null hypotheses were rejected at the 0.05 level, although there were few differences among the three test sites on Amchitka. Eggs had the lowest levels of cadmium, lead, and mercury, and the feathers of adults had the lowest levels of selenium. Comparing only adults and fledglings, adults had higher levels of cadmium, chromium, lead and mercury, and fledglings had higher levels of arsenic, manganese and selenium. There were few consistent interisland differences, although levels were generally lower for eggs and feathers from gulls on Amchitka compared to the other islands. Arsenic was higher in both adult feathers and eggs from Amchitka compared to Adak, and chromium and lead were higher in adult feathers and eggs from Adak compared to Amchitka. Mercury and arsenic, and chromium and manganese levels were significantly correlated in the feathers of both adult and fledgling gulls. The feathers of males had significantly higher levels of chromium and manganese than did females. The levels of most metals in feathers are below those known to be associated with adverse effects in the gulls or their predators. However, levels of mercury in some gull eggs are within a range suggesting that several eggs should not be eaten in one day by sensitive humans.