Two separate, large-scale management units for the American White Pelican (Pelecanus erythrorhynchos) in North America are proposed: the eastern and western metapopulations, separated by the North American Continental Divide. The populations on either side of this barrier are subject to contrasting ecological conditions, movement patterns suggest strong geographic separation and each is likely demographically distinct. Subdivisions within these demographic units need to be more precisely defined in the future. Yet, because of the highly colonial nesting habits of the American White Pelican, conservation can still be effective if directed toward separate breeding colonies. Our intent is to provide a compilation of current knowledge regarding species status, distribution and ecology. Herein, further study is recommended: (1) to determine genetic characteristics and the degree of genetic separation of the various geographic divisions in the range of the species to help ultimately better define “evolutionarily significant units” for American White Pelican conservation, (2) to study movements and genetic exchange among and between these divisions, and (3) to study movement patterns and genetic mixing among these divisions during long- and short-term changes in environmental conditions.

Comments on the American White Pelican (Pelecanus erythrorhynchos) in North America languished until Peter Ogden reported trapping one during his 1825 expedition to Utah. Brief accounts of the American White Pelican by ornithologists continued through the 1940s. In subsequent decades, scientific studies uncovered greater details of the species' biology and natural history, although documenting numbers has been difficult and tentative. Estimates of numbers of adults began at 30,000 in 1933, increased to over 100,000 by 1985, and by 1995 the total number of birds, then also including non-breeders, was estimated to be around 400,000. Beginning in the 1880s, their feeding and nesting sites were degraded by engineered water diversions and drainage of wetlands for agriculture. At the same time, pelicans were shot and clubbed, and eggs and young were intentionally destroyed largely because the birds were thought to compete with humans for fish. After the 1960s, hundreds of pelicans died yearly due to the ingestion of insecticides such as toxaphene, endrin, and dieldrin. As recently as winter 1998-99, 800 American White Pelicans died in Florida from poisoning by insecticides that were resuspended from flooded agricultural soils. In 1996 a disease pandemic at Salton Sea, California, killed over 7,500 pelicans in just several months. American White Pelicans have adapted to much persecution by simply moving. Overall, I do not believe that unusual mortalities have threatened their abundance.

The only breeding colony of the American White Pelican (Pelecanus erythrorhynchos) in British Columbia is located at Stum Lake. The species in British Columbia. is designated “Endangered” and it is also included on the provincial “Red List.” As such, the Stum Lake colony has been monitored regularly since 1994, and older records of nest counts and pre-fledgling numbers date back as far as 1953. A marking program for pre-fledging American White Pelicans was conducted from 1968 to 1993. During recent surveys, censuses of nest depressions were obtained late in the breeding season; pre-fledged young were also enumerated. Based on nest counts, numbers at Stum Lake declined in the 1960s and remained more or less stationary throughout the 1970s and early 1980s. In the late 1980s and early 1990s, the number of nesting pairs increased dramatically; and in the past decade, numbers of breeding birds at Stum Lake has stabilized at the highest since monitoring began (1953). The mean number of pre-fledged young also shows an increasing trend from 1963 through 2002. The mean productivity at the breeding colony, however, was highest in the decade showing the largest growth, from 1983 to 1992; but it decreased in the past decade as total numbers of breeding birds continued to increase. The mean productivity of the breeding colony since monitoring began has been 0.62 ± 0.37 young nest^-1, with annual rates varying from nil to 1.48 young nest per nest. Four breeding failures have been documented since 1960. The majority of the returns from the marking program were from west of the Continental Divide in British Columbia, Washington, California, Oregon, Idaho and Mexico.

The American White Pelican (Pelecanus erythrorhynchos) has been breeding at Medicine Lake National Wildlife Refuge (NWR) in northeastern Montana since 1939. We compiled information on the history and ecology of this breeding colony from U.S. Fish and Wildlife Service (USFWS) reports and narratives filed at Medicine Lake NWR. Nesting by American White Pelicans first occurred in 1939, and the recent 10-yr average size (1993-2002) is about 4,000 breeding pairs. Since 1990, the mean number of young fledged pair-1 was 0.51 ± 0.07 (SE, N = 56). Low temperatures, hailstorms, predation, and botulism have killed >50% of young pelicans in some years. American White Pelicans banded at Medicine Lake migrated south, mostly within the Central Flyway. They wintered predominately in Texas, Louisiana and Mexico. Protection of the Medicine Lake colony is important because the breeding colony is the largest in Montana and an important contributor to the eastern metapopulation.

Chase Lake National Wildlife Refuge currently supports the largest breeding colony of the American White Pelican (Pelecanus erythrorhynchos) in North America. The first written account of the colony's status was in 1905, when approximately 500 pelicans nested on two islands in Chase Lake. The colony grew to nearly 34,000 breeding birds by 2002. Surveys of breeding pelicans in the colony were sporadic from 1905 until 1972. Because the availability of foraging areas affects recruitment, pond counts from south-central North Dakota were used as an index of the relative availability of foraging areas. Pond counts varied widely in the 1970s and 1980s. In 1993, pond numbers increased dramatically, reached a peak in 1997, then declined but remained relatively high through 2002. Pelican numbers also increased in the mid- to late-1990s, possibly in response to increased and stable foraging opportunities. Because rising water in Chase Lake inundated the original nesting islands during that period, pelicans were forced to relocate nesting areas. Relocation provided more nesting space than the original islands, and nesting pelicans seemed to be gradually filling these areas. Threats to the Chase Lake breeding colony include disease, predation, and human disturbance. Research is needed to better understand the mechanisms that underlie the dynamics of AWPE metapopulations and their sub-units, including factors that influence nest-site selection, productivity, and survival of adults and young in large colonies.

The American White Pelican (Pelecanus erythrorhynchos) formerly bred the length of California on interior lakes and marshes of the Klamath Basin, Modoc Plateau and Great Basin desert of northeastern California, overflow lands of the Sacramento Valley, terminal lakes in the Tulare Basin of the San Joaquin Valley and at the Salton Sea in the Colorado desert. Historic data are sketchy but nesting is well documented for at least eleven sites. Though unknown, the number of pelicans breeding in the late 1800s and early 1900s may have exceeded 20,000 pairs. Decades of declining numbers and range retraction have left the American White Pelican breeding regularly in California at only two sites in the Klamath Basin (Sheepy Lake on Lower Klamath National Wildlife Refuge and Clear Lake National Wildlife Refuge), where very vulnerable to catastrophic losses. Variably obtained counts at these two sites combined appeared to decrease from about 4,500 nests in 1912 to reach an equilibrium, which fluctuated around 1,600-1,700 nests from the 1950s to present. Annual variation is presumably a response in part to prey availability mediated by water levels. Several high nest counts of 2,300-2,500 at Clear Lake since the early 1970s may partly reflect greater accuracy from aerial photo counts. Declines were caused by shooting, human disturbance and, particularly, habitat loss from water diversions and land reclamation for agriculture. The American White Pelican is currently limited in California by the availability of remote, undisturbed nesting sites and rich foraging habitats. Contaminant effects on breeding pelicans have lessened, but large numbers of pelicans have recently died from type C avian botulism outbreaks at the Salton Sea. Management and research should serve the protection and enhancement of breeding conditions in the Klamath Basin.

Breeding bird surveys were conducted at most American White Pelican (Pelecanus erythrorhynchos) colonies in North America from 1979-81 and the number of American White Pelicans was estimated at >109,000 breeding individuals in 55 colonies. Although many pelican colonies have not been surveyed since the early 1980s, we present the most current survey numbers available for North American colonies. Using available data (1998-2001) and the North American Continental Divide to delineate eastern from western metapopulations, 27 colonies and 48,240 nests in the East and 15 colonies and 18,790 nests in the West were documented, giving a total of about 134,000 breeding pelicans in North America. The nest numbers of 20 eastern and western colonies that were surveyed during 1979-81 and again in 1998-2001 were also compared to determine if pelican numbers are changing. The number of AWPE nests in those colonies has more than doubled since the 1979-81 surveys. Because current data for colonies in part of the United States and much of Canada are lacking and about 20 years have passed since more complete surveys were done, we propose that regular and coordinated North American pelican surveys should be initiated.

A total of 3,859 band-recovery records of the American White Pelican (Pelecanus erythrorhynchos) banded between 1922 and 1981 at 77 sites in the U.S. and Canada was analyzed using a Geographic Information System (GIS). Returns showed a clear pattern of two major migratory routes, with birds from California, Oregon and Nevada breeding colonies dispersing south to the coasts of the Gulf of California and the Pacific coast of Mexico, and birds from central Canada, the Dakotas, Colorado, Minnesota and Montana following the Missouri and Mississippi River drainages to the Gulf of Mexico coasts, Florida and Central America. Pelicans banded at breeding colonies in Utah and Wyoming were recovered along both routes, with some evidence of a third pathway suggested for these birds along the western slope of the Rockies. The only region of substantial overlap in returns between birds banded in the western breeding colonies and those from mid-western breeding colonies (east of the North American Continental Divide) was in southern Mexico and Texas. Returns from Central America consisted almost exclusively of American White Pelicans banded in the eastern portion of the breeding range.

Anaho Island, Pyramid Lake, Nevada supports the largest colony of the American White Pelican (Pelecanus erythrorhynchos) in western North America. Counts of adults, nests and chicks, from 1903 to the present, are positively correlated with flow volumes of the Truckee River in spring near Nixon, Nevada, 15 km from the lake. Upriver diversions of water have markedly reduced flows of the lower Truckee River and have likely depressed pelican production; for example, the Derby Dam, completed in 1906, diverts water from the Truckee River into the Truckee Canal and out of the Truckee River basin. American White Pelican production was negatively correlated with flows in the Truckee Canal. If water had remained in the river instead of being diverted, it is estimated that annual colony production would have averaged 8% higher. A more complete model, accounting for seasonal diversions of water from the river between the gage upriver at Farad and the gage near Nixon, predicted that colony production would have averaged as much as 58% higher if natural flow regimes had persisted in the Truckee River downriver from Farad. Human removals of water from the Truckee River, and thus simulated reductions on pelican colony production, were highest in drought years. Despite the likely reductions in production throughout the last century, particularly in drought years, numbers of adults and nests at this breeding colony demonstrated no long-term directional trends. Therefore, reductions in production apparently have been offset by higher survival or immigration from other colonies. Even so, higher production could be an important buffer against high mortality events, such as the disease die-offs of large numbers of American White Pelicans at Salton Sea, California, and even natural disasters described by other authors in this symposium.

Digital Hydrographic Maps were developed by means of a geographic information system in order to predict how nesting islands for the American White Pelican (Pelecanus erythrorhynchos) are likely to become land-bridged under variable water levels on Clear Lake Reservoir, California. As a management tool, the maps can be matched to water surface elevation changes forecasted by the U.S. Bureau of Reclamation every year. This graphic method has allowed Klamath Basin National Wildlife Refuge managers to set up electrified fences at Clear Lake Reservoir in advance of the nesting season, successfully protecting nesting colonies on two recent occasions. Digital Hydrographic Maps were also used to match historic hydrographic records for Clear Lake Reservoir to examine past lake conditions and changes in nesting island locations, particularly when nesting failures had occurred. Modifications to existing nesting islands are suggested, so as to create more stable nesting habitat.

The eastern metapopulation of the American White Pelican (Pelecanus erythrorhynchos) breeds mainly in the Northern Great Plains, migrates through the Great Plains and along the Mississippi River and winters in the lower Mississippi River Valley and along the Gulf of Mexico. The production of farm-raised Channel Catfish (Ictalurus punctatus) in the southeastern United States has increased dramatically in the last 25 years. In 1990, U.S. Department of Agriculture Wildlife Services offices in Arkansas, Louisiana and Mississippi began receiving complaints concerning the foraging of pelicans in commercial Channel Catfish ponds. Because of the relatively shallow depth and high fish-stocking rates used by most catfish producers, these ponds provide an ideal foraging environment for the American White Pelican. Although the impact of foraging pelicans can be economically significant, the potential for pelicans to transmit trematode infections and other diseases to aquaculture facilities can be more destructive. Damage abatement recommendations have consisted of harassment measures similar to those used for other piscivorous birds, U.S. Fish and Wildlife Service Depredation Permits, and harassing the birds at their loafing sites.

Records of reported die-offs of the American White Pelican (Pelicanus erythrorhynchos) held by the U.S. Geological Survey National Wildlife Health Center from 1978 through 2003 indicate that type C botulism (caused by Clostridium botulinum) was the major cause of mortality. In 1996, over 15,000 birds, including 8,500 American White Pelicans, were estimated to have died from type C botulism at the Salton Sea in California. This was the largest documented die-off of any pelican species and was estimated to represent 15-20% of the western metapopulation. This event was also notable in that it was the first time that fish, specifically Tilapia (Oreochromis mossambicus), were implicated as the source of type C botulinum toxin for birds. Type C botulism has recurred in both North American species of pelicans at the Salton Sea every year since, although the magnitude of mortality is much lower. West Nile virus caused mortality in both adult and immature American White Pelicans, but may have a more significant impact on nestlings. Emaciation and mortality is common in pelican nesting colonies. Further clarification of the cause of nestling mortality and the ability to differentiate background mortality from mortality due to infectious disease agents such as West Nile virus and Newcastle disease virus would be important in determining the impact of disease in pelicans. Although definitive conclusions cannot be drawn, the number of die-offs and the estimated losses of American White Pelicans appears to have increased along with declines in number in western U.S.

Reproductive success of the American White Pelican (Pelecanus erythrorhynchos) was monitored at a breeding colony on Anaho Island, Pyramid Lake, Nevada in 1996. Eggs were collected in 1988 and 1996 and analyzed for organochlorine pesticides (OCs) and total polychlorinated biphenyls (PCBs). Muscle from adults found dead or debilitated and euthanized, fishes from representative feeding areas and regurgitated fish samples from nestlings were also analyzed for OCs and PCBs. Reproductive success at the breeding colony was normal in 1996 based on hatching rates of eggs (≥79% in undisturbed areas) and survival of nestlings. Organochlorine pesticide and PCB concentrations in eggs were below known effect levels on reproduction. DDE concentrations in eggs from Anaho Island declined between 1988 and 1996. Eggshell thickness for the Anaho colony was significantly lower (6%) than the pre-OC norm, but the level of thinning was less than that associated with population declines. OCs and PCBs were seldom detected in fish.

Culmen length has been suggested as being diagnostic for sex in the American White Pelicans (Pelecanus erythrorhynchos). However, the literature on the use of culmen length to determine sex is inconsistent, with reported overlap in culmen lengths for males and females ranging from 1 mm to >120 mm. Morphological measurements from 188 American White Pelicans collected in Mississippi and Louisiana whose sex was determined by dissection and gonadal inspection were measured. The use of culmen length alone was used to determine gender for this sample by establishing the minimum observed culmen length for males and the maximum observed culmen length for females that provided the fewest incorrect determinations for each gender. A multivariate discriminant function model was developed to determine sex from our data and compared the diagnostic accuracy of the model with the accuracy based on culmen length alone. Both methods were validated using an independently collected sample of 22 pelicans from Florida. A culmen length of ≥310 mm for males and ≤309 mm for females from our data correctly classified sex for 99% of American White Pelicans from our Mississippi and Louisiana samples and 95% of AWPE for the Florida sample. Culmen length and wingcord length were significant variables in the discriminant function model. The resulting model correctly classified sex of 97% of the birds and 94% of the independent Florida sample. The culmen length alone predicted the sex of American White Pelicans as well as multivariate methods and provides an accurate simple, non-lethal method for sexing the species.

The goals of this symposium were to foster exchange of information and to identify priorities for future research. The workshop, oral presentations, and poster session that comprised the symposium provided a broad overview of our knowledge of the biology and present status of the American White Pelican. Here, the workshop, the contributed papers at the symposium, and articles in this volume are summarized as they relate to the goals of the symposium.