Tracking Tundra Swans by Satellite


Dr. Scott Petrie, Long Point Waterfowl Executive Director

Kerrie Wilcox, Researcher, Long Point Waterfowl


The Tundra Swan (Cygnus columbianus) is the most widespread and numerous species of swan in North America. There are two distinct populations, the Eastern Population and the Western Population. It is the Eastern Population that migrates through southern Ontario as they make their extensive annual migrations between their Atlantic coast wintering areas and arctic coastline breeding areas. Although previous neck-collar studies provided information about migration pathways of Eastern Population swans, little was known about chronology of migration during spring or fall. Further, despite the perceived importance of migratory stopover areas to Tundra Swans, little was known about the population status or habitat use of this species at key staging areas, or how Tundra Swans allocate their time between terrestrial and aquatic habitats during spring and fall.

Based on the known importance of the lower Great Lakes for Eastern Population Tundra Swans, and the lack of information pertaining to their migratory movements and staging ecology, Long Point Waterfowl embarked on a Tundra Swan research program in 1998.

Areas of study conducted by Shannon Badzinski were interspecific competition, dietary intake, activity patterns and micro-habitat use. Areas of study conducted by Scott Petrie were nutrient reserve dynamics, feather replacement, macro-habitat use, population dynamics, and migration chronology. Long Point was chosen as a suitable site for study as its 27,000 acres of coastal wetland habitat and the 72km2 Inner Bay are used extensively by EP Swans during migration, in fact, mean peak fall counts (which provide a minimum estimate of swan use) at Long Point have increased from representing 0.7% of the EP in the 1970s to 7.9% in the 1990s.


Tracking Tundra Swans using conventional radio transmitters is impractical because they are long distance migrants, and they breed in remote areas. However, the production of miniature satellite transmitters enabled scientists to track birds over long distances without having to physically follow them. Because the influence that harness-attached transmitters have on birds is still under debate, scientists have an ethical and practical obligation to ensure that the attachment technique utilized does not adversely effect the behaviour, skin, or feathers of the species being studied, or that birds are able to remove the transmitter. Therefore, two captive, flightless Tundra Swans were fitted with teflon-harness-attached dummy transmitters and activity patterns were monitored over a 4 week period. The captive male and female Tundra Swans spent 14.3 and 6.2% of their time preening the harness and transmitter over the observation period. This compares favourably with results of other captive behavioural studies of birds with backpack transmitters. Also, when transmitters were removed after 8 weeks, there were no obvious signs of feather wear, skin abrasion, or harness failure. Therefore, we felt justified in using harness-attached backpack transmitters to track the movements of free-ranging Tundra Swans.

Rocket nets were used to capture Tundra Swans on Long Point’s wetlands and neighbouring agricultural fields during spring and fall. Twelve adult Tundra Swans were captured and equipped with satellite transmitters and 62 birds (42 adults, 20 juveniles) were captured and equipped with black and white neck collars with alpha-alpha-numeric codes. Four transmitters were attached in spring 1998, 3 in fall 1998, and 5 in spring 1999. Ninety-five gram transmitters were attached (backpack) to all 7 birds in 1998 using a teflon-harness-attachment. Due to earlier than expected transmitter failure, we switched to 30-gram neck-collar-attached transmitters in 1999.

There are 3 ARGOS satellites that continuously orbit the globe and detect the locations of transmitters. The satellites orbit at 850 km above the Earth’s surface and they can detect the presence of transmitters across a 5000 km path each time they pass. When the satellite detects the frequency from a particular transmitter, it picks it up from several locations as it passes. This information is translated into a latitude and a longitude, relayed to a ground station in the United States, and then to researcher’s computers via e-mail shortly after the satellite pass.


Owing to transmitters being deployed at a staging area, and the operational period of transmitters being variable, spring migration results are based on 3 birds tracked from the Atlantic coast and 8 from Long Point, and fall migration results are based on 4 birds tracked from breeding areas and 3 from Long Point. Tundra Swans spent 51% of the annual cycle on staging areas (spring staging 29%, fall staging 23%, wintering areas 20%, breeding areas 28%), and tracked birds used 4 distinct staging regions during migration; Atlantic coast, Great Lakes, northern prairies, and boreal forest.

Spring migration: Tundra Swans departed Atlantic coast wintering areas throughout February and arrived on breeding areas between mid-May and mid-June. Therefore, spring migration progresses at a leisurely pace, extending for over 100 days, despite total flight time to complete spring migration being only about 114 hrs. Early departure from wintering areas and extended duration of spring migration may be a function of food depletion on wintering areas, ready availability of agricultural grains and winter wheat on spring staging areas, need for juveniles to migrate slowly, and/or the need to arrive on breeding grounds shortly after spring thaw with ample reserves for reproduction.

After departing wintering areas, birds staged for short periods along Chesapeake Bay, Delaware Bay and the Susquehanna River, Maryland. Once weather conditions were suitable, birds migrated to the Great Lakes region where they spent 17-33 days using 3-6 different sites, the most important being Long Point, Ontario, Aylmer Wildlife Management Area, ON, Lake St. Clair, Grand Bend, ON, Saginaw Bay, Michigan, coastal wetlands along the eastern shores of Lake Michigan, and Green Bay, Wisconsin. Tundra Swans lose body mass throughout winter and are still in relatively poor condition when they arrive in southern Ontario in early March. European studies suggest that Bewick’s Swans (Cygnus columbianus bewickii) have a maximum potential flight range of 1,450 km and that stop-over periods would need to be at least 2 weeks to replenish nutrient reserves completely. Distance between wintering areas in North Carolina and southern Ontario is at least 1000 km, and it is an additional 1,500 km to their next major staging region (northern prairies). Therefore, time spent consuming waste corn and submerged aquatic plants in the lower Great Lakes region would enable birds to acquire nutrients necessary to continue migrating to the northern prairies. Satellite locations (as well as aerial and ground-based waterfowl surveys) while birds were at Long Point indicate that Tundra Swans spend considerably more time feeding in agricultural fields (especially corn) in spring than in fall. Waste corn has a high carbohydrate content which would facilitate fat deposition.

In late March and early April birds migrated west from the Great Lakes to the northern prairies (western Minnesota, northeastern North Dakota, and southern Manitoba and Saskatchewan) where they spent considerable time (25-45 days) and used numerous staging areas . The most important were western Minnesota, the Sheyanni National Grassland near Fargo, ND, northeastern ND, southern Manitoba, southern Saskatchewan, the Prince Albert National Park, SK, and Moose Lake, MB. The amount of time spent in this region, combined with the need to arrive on breeding areas with ample reproductive reserves, and fact that birds still have between 900 (western shores of Hudson Bay) and 2000 km (Mackenzie River Delta) left to migrate, suggests that birds probably acquire substantial endogenous reserves for migration as well as for reproduction while on the northern prairies.

Birds spent relatively little time in the northern boreal forest (average = 14 days) before making a final move to breeding areas. Important staging areas were Cedar Lake, MB, Lake Athabasca, Alberta, Churchill River, MB, Southern Indian Lake, MB, Great Slave Lake, N.W.T., the Mackenzie River and upper Mackenzie River Delta, N.W.T. Tundra Swans maximize time spent in agricultural regions during spring because only 15% of their time was spent in the northern boreal forest, whereas 40% was spent in the northern prairies and 27% in the Great Lakes. This tactic is probably a result of the, 1) reduced availability of waste grains in Great Lakes and northern prairie agricultural landscapes, 2) limited availability and accessibility of submerged aquatic plants in the northern boreal forest in spring due to winter senescence and ice cover, and 3) the overriding need to maximize nutrient storage for migration and reproduction. Given limited food availability and time spent in the northern boreal forest in spring, we suggest that Tundra Swans arrive in the northern boreal forest with large nutrient stores, and that they only remain in the boreal forest until climate conditions enable them to complete migration.

Breeding movements: Tracked birds bred on the western shores of Hudson Bay, Victoria Island, Queen Maude Gulf, Mackenzie River and Mackenzie River Delta. Birds arrived on breeding areas between mid-May and mid-June and departed from late August to late September. Total time spent on or near breeding areas was 76-140 days. As it takes 102-117 days to raise cygnets to an age which they can successfully migrate, variability in time spent on and near breeding areas is probably due to unsuccessful breeders departing breeding areas early.

Bird 20114 bred on the Mackenzie River, N.W.T., remained on the breeding area for 140 days, and did not appear to move between a distinctive nesting and brood rearing site. Similarly, Bird 20180 bred at an inland site on Victoria Island, remained on the breeding area for 112 days, and did not change locations during that time. Conversely, after spending 21 days on Richards Island within the Mackenzie River Delta in 1999, and 38 days in 2000, bird 20179 moved 39 and 61 km to the east were it remained for 59 and 58 days respectively. Since adults with cygnets do not disperse from nesting territories until they finish wing molting in late summer, bird 20114 and 20180 likely successfully bred and raised broods in close proximity to nesting sites, whereas, swan 20179 probably failed in its breeding attempt during 1999 and 2000, and subsequently relocated prior to wing molt.

Fall Migration: Fall migration took about one month less time than spring migration, despite juveniles being younger in fall. This may be due to birds being driven south by advancing winter conditions in fall, and being delayed by ice cover in spring. Whereas birds spent limited time in the northern boreal forest region in spring, all birds spent long periods there in fall (32-49 days). Relative to the spring, the availability of submerged aquatic vegetation in boreal forest wetlands would be higher in fall, after the summer growing period. Tundra Swans fledge at a smaller proportion of adult mass than geese and are still growing when they depart breeding areas Therefore, increased time spent in the northern boreal forest in fall, relative to spring, is probably due to continued growth requirements and limited stamina of juveniles, and possibly the need for adults to replenish nutrients catabolized during reproduction and wing-feather replacement.

Upon leaving the northern boreal forest, tracked birds staged on large inland wetlands in central and southern Saskatchewan. Use of large inland wetlands at that time can be attributed to birds foraging primarily on submerged aquatic plants during fall migration. Birds spent less time in the northern prairie region during fall migration (average = 21 days) than during spring (average = 37 days) which may be simply due to the fact that, after spending extended periods of time in the northern boreal forest in fall, they have limited time to stage on the prairies before freeze up. Also, whereas nutrient acquisition would be of utmost importance in spring due to impending reproduction, fat storage is probably less critical in the fall.

Upon leaving the northern prairie region, two tracked birds made extensive movements, one to North Carolina within 4 days (2895 km, #20179) and the other to southern Ontario within 4 days (2,085 km, #20114). Breeding ground movements suggest that #20179 did not breed successfully, and migrating without cygnets may have enabled this bird to make such an extensive move. However, there is some evidence that Tundra Swans travel longer distances between staging areas in fall than in spring. We estimate that Tundra Swans that stop in southern Ontario in fall do so for approximately 3 weeks. Large numbers of EP swans also stage on the Upper Mississippi River in fall for 3-4 weeks. Amount of time spent on the lower Great Lakes and Upper Mississippi River in fall is probably dictated primarily by weather, as birds are in good condition upon arrival which reduces the need to linger and store nutrients.

With the exception of one bird (20179) that traveled from southern Saskatchewan to North Carolina within 4 days (exact route unknown), birds migrated almost directly from southern Ontario to northern Chesapeake Bay. From MD birds made numerous stops as they migrated south to NC. While in southern Ontario in fall, Tundra Swans feed primarily in aquatic habitats. Satellite locations suggest that swans continue to feed aquatically in MD but reduce time spent in aquatic habitats as they move south through VA and into NC. This may be due to declining availability of submerged aquatic vegetation as winter progresses, or simply to a dietary preference for waste corn in late winter. Neck-collar observations and returns closely approximated some of the staging areas used by satellite tracked birds. However, one bird was observed in California, which supports previous evidence that there is some mixing between EP and WP swans.


1/ The post European settlement provision of agricultural grains and winter wheat has had a strong influence on Tundra Swan staging chronology and habitat use, particularly during spring.

2/  Whereas Tundra Swans maximize time spent in agricultural habitats during spring, they are more dependant on large staging wetlands in fall.

3/ Whereas Tundra Swans tend to use similar migratory pathways during spring and fall, there are seasonal differences in which staging areas are used, as well as in time spent in each of the major staging regions.

4/ Conservation of staging habitats is critical as Easter Population swans spend half their life migrating between Atlantic coast wintering areas and Arctic breeding areas.

5/ Thirty-gram neck-collar-attached transmitters were more effective for monitoring long distance movement of Tundra Swans than 95-gram backpack-harness-attached transmitters.

We thank numerous individuals and organizations for their support with this project.

While satellite tracking provides information that is essential to obtaining a better understanding of Tundra Swan migration, it is very expensive. Long Point Waterfowl is non-profit and non-government, and it is solely dependant on individuals and organizations such as yourself for its funding. Therefore, Long Point Waterfowl asked individuals and organizations to consider sponsoring one of the swans that we tracked. By contributing at least $1000 toward one of the $3,000 transmitters, donors were able to name their swan whatever they wanted

Our generous donors incude:

The Sydenham Conservation Foundation sponsored Sir Syd.
The Long Point Area Fish and Game Club sponsored Harry.
The Aylmer Order of Good Cheer sponsored Roast.
The Long Point Waterfowlers Association sponsored Waterfowler.
Ducks Unlimited Canada sponsored Spirit and Hans.
The Bluffs Hunting Club sponsored Bluff.
Through the volunteer efforts of three of its students (Cain Vangel, Danny Bernard, Michelle Murphy), Sir Sandford Fleming College sponsored Sir Sanford.
Zone J of the Ontario Federation of Anglers and Hunters sponsored OFAH-J.
Scott Petrie sponsored Noah, after his oldest son.
Bird Studies Canada sponsored Abigail, after Abigail Becker, the heroine of Long Point.

Very generous donations or support have also been received from:

Canada Trust Friends of the Environment
Dr. Duncan Sinclair, Aylmer, ON
The Bluff Hunting Club, Long Point, ON
S.C. Johnson & Son, Inc., Racine WI
Long Point Waterfowlers Association, Port Rowan, ON
Long Point Area Fish and Game Club, Port Rowan, ON
Sydenham Conservation Foundation, Owen Sound, ON
Ontario Federation of Anglers and Hunters  Zone J
Ducks Unlimited Canada
Ducks Unlimited Incorporated
Mees Pierson Ltd., Nassau, Bahamas
Aylmer Order of Good Cheer, Aylmer, ON
Bird Studies Canada, Port Rowan, ON
James Hazen, Port Rowan, ON
Howey Hardware, Port Rowan, ON
Tod G. Wright,
Burlington, ON
The Aylmer District of the Ministry of Natural Resources; Delbert Miller in particular
The Canadian Wildlife Service, London; Norm North in particular
Emile and Veronica Vandommele, Port Rowan, ON
Glen Howe, Aylmer, ON
Jack Nicholls, Cashway, Simcoe, ON


Assistance has been graciously provided by:

Emile and Veronica Vandommele, Port Rowan, ON
Matt Brock, Simcoe, ON
Dan Evelsizer, Iowa
Oscar Langevoord, Holland
Linda Zwiggelaar, Holland
Shannon Badzinski, Wisconsin
Frank Yalaksa, Delhi, ON
Bill McWaters, Toronto, ON
Stephanie McWaters, Toronto, ON
Bruce Cooper, London, ON
Ted Barney, Elora, ON
Ann Burke, Wisconsin
Randy Hetzel, Wisconsin
Dennis Reimer, Port Rowan, ON
Cain Vangel, Scarborough, ON
Michelle Murphy, Peterborough, ON
Danny Bernard, Waterloo, ON
Darryl Pajunen, Peterborough, ON
Kerri Sullivan, Port Dover, ON
Jerry Bell, Port Rowan, ON
Hugh MacArthur, Simcoe, ON

Comments are closed.