Bird Observer - The New England Birding Journal

Bird Observer

The Birding Journal for New England

August 2015

Vol. 43, No. 4

Winter Quarters and Migration Routes of Common and Roseate Terns Revealed by Tracking with Geolocators

Ian C. T. Nisbetand Carolyn S. Mostello

Fig. 1. A Common Tern with a geolocator mounted on its leg. The geolocator was attached to a custom-made plastic leg flag using marine epoxy adhesive, and secured with two loops of dental floss sealed with superglue. The light sensor is the small white rectangle. The small white flag was placed on the right leg so that the bird could be located if it returned but had lost the geolocator assembly during the winter. (Photograph by C. S. Mostello).

In recent years the study of bird migration has been revolutionized by the development and use of miniaturized tracking devices. It is now possible to track individual birds wherever they go, across mountains and deserts, through the night, and far out at sea. As yet, satellite transmitters and GPS receivers can be used only on large birds; the only devices available to remotely track small birds that cover great distances are light level geolocators, which use changes in light intensity to determine locations. Geolocators were originally developed by the British Antarctic Survey for tracking albatrosses, but their engineers progressively made them smaller so that they can now be used on small birds of many species. In recent years they have been used to track, among others, Veeries (Heckscher et al. 2011; Hobson and Kardynal 2015), Northern Wheatears (Bairlein et al. 2012), Black Swifts (Beason et al. 2012), Ovenbirds (Hallworth et al. 2015), and—most recently—Blackpoll Warblers on their autumn migration across the ocean from New England to the West Indies (DeLuca et al. 2015).


Geolocators contain a clock and a light sensor that records light intensity every two minutes. By plotting the changes in light intensity, the times of sunrise and sunset on each day can be determined. The time of sunrise or sunset identifies the longitude, and the length of day or night on any specific date identifies the latitude. Geolocators also contain a battery and a computer chip that can store data for up to two years. When the data are downloaded, they provide a record of the location of the bird twice each day throughout the study period.

Despite their utility, geolocators have two major disadvantages. The first is that they store but do not transmit data, so each marked bird must be recaptured in order to download the stored location data. The second is that the location estimates are imprecise. Cloudy weather and other factors that reduce the intensity of light reaching the geolocator lead to errors in the estimated times of sunrise and sunset, and hence errors in the derived estimates of latitude and longitude. In our work with terns, uncertainty in estimates of longitude is typically about ±1º, or about ±100 kilometers in the east-west direction at the Equator and ±60–70 kilometers at mid-latitudes (Mostello et al. 2014).

Uncertainty in estimates of latitude is greater than in estimates of longitude, and increases progressively at dates close to the spring and autumn equinoxes and at latitudes close to the Equator. At the equinoxes, day length is the same everywhere on Earth and latitude cannot be estimated at all; close to the Equator, day length varies very little with latitude, so that estimates of latitude may be inaccurate by 10º (± 1000 km) or more. If a bird remains at the same location for weeks or months, statistical averaging allows that location to be determined fairly precisely. When a bird is traveling, errors in individual fixes can be large and the exact track cannot be determined.

Stopovers and Wintering Areas

Common Terns Roseate Terns
Area Coordinates Autumn Spring Winter Autumn Spring Winter
A. Cape Cod 41–42º N, 70–71º W 11 6
B. Cape Hatteras 34–36º N, 74–76º W 4 3
C. NE Caribbean 18–20º N, 64–71º W 3 0 6 5
D. NW Venezuela/Aruba 12–13º N, 68–71º W 1 0 1 1 1
E. Trinidad 10–11º N, 61º W 4 0 0 0
F. Suriname 6–7º N, 54–56º W 5 0 1 4 0 1
G. French Guiana 5–6º N, 51–52º W 2 0 0 0
H. Pará 1º S, 46–49º W 4 4 1 4 4 1
I. Maranhão 3–4º S, 42–45º W 3 1 2 0
J. Céara/Rio Grande do Norte 4–5º S, 37–40º W 1 1 0 0
K. Alagoas/Sergipe/Bahia 10–17º S, 36–39º W 3 2 2 4
L. Espíritu Santo 19–21º S, 40–41º W 3 1 2
M. Lagoa do Peixe 30–34º S, 50–53º W 3
N. NE Argentina 36–40º S, 57–58º W 1

Table 1. Stopover and wintering areas of 11 Common and 6 Roseate Terns. Each entry for a stopover area indicates the number of birds that stopped there. Zero indicates that one or more birds passed through the area without stopping. NE Caribbean includes Haiti, the Dominican Republic, Puerto Rico and the Virgin Islands. The only Common Tern that reached Argentina spent part of the winter at the Lagoa do Peixe and alternated between these locations; it is included in this table for area N. Two Roseate Terns that spent the early part of the winter in area K moved to area J in February (Figure 2); they are included in this table for area K.

Tracking Common and Roseate Terns

During the 45 years we have studied and conserved the Common and Roseate terns that breed in Massachusetts, we have been repeatedly frustrated by the paucity of data on these birds in their winter quarters, where they spend about half the year. Until 2007, most information about the migrations and wintering areas of both species had been gleaned from banding recoveries, which were few and scattered, and whose occurrence usually depended on the unlikely combination of a bird dying in an accessible place and being found by an interested person. Band recovery data had indicated that both species migrate through the West Indies to winter on the north and east coasts of South America (Nisbet 1984, 2002), but little was known about dates and routes of migration or the precise locations of stopover and wintering sites. Information was particularly sparse for the Roseate Tern, which had rarely been encountered anywhere in South America, so that it had been almost impossible to devise conservation measures for this endangered species on its winter range.

When geolocators small enough to be carried by terns became available in 2007, we obtained 20 geolocators, weighing 1.5 grams each, from the British Antarctic Survey and attached them to custom-made plastic leg flags (Figure 1). We put the devices on 10 adults of each species that we caught on nests at Bird Island, Marion, Massachusetts. In 2008 we found six Common Terns and four Roseates that returned, retrieved nine geolocators, and obtained valuable data from seven of them. However, we were concerned at the low rates of return—40% for Roseate Terns and 60% for Commons, versus 83% and 90% expected, respectively, based on average survival rates for each species (Nisbet 2002; Spendelow et al. 2008). We decided to discontinue the study until smaller devices were developed. In 2009, geolocators weighing only 1.0 gram became available, and we repeated the study, attaching devices to 10 more adults of each species. Combining the results of the two years, we retrapped 13 of 20 Common Terns and nine of 20 Roseate Terns, and obtained useful data from 11 Commons and six Roseates.

After downloading and analyzing the stored data using software supplied by the British Antarctic Survey, we were confronted by a huge volume of data: 11,084 paired estimates of latitude and longitude. To date, we have published three papers in scientific journals (Nisbet et al. 2011a, 2011b; Mostello et al. 2014), and we are planning at least one more. Ours was one of the first two studies to use geolocators to track terns, the other being the groundbreaking work on the migration of Arctic Terns from Greenland to the Antarctic and back (Egevang et al. 2010).

Fig. 2. Migration routes and winter quarters of Common and Roseate terns. The main areas used for stopovers and wintering are marked with letters keyed to those in Table 1. Filled areas include locations used for long periods during the winter. Within each marked area, each bird usually stayed within a small range of longitudes and latitudes.

Winter Quarters

Our tracking results revealed that Roseate Terns spent the winter at three locations on the north and east coasts of South America, spanning about 4,000 kilometers of coastline from Suriname to eastern Brazil (Table 1, Figure 2). Common Terns wintered over an even wider area, extending along about 8,000 kilometers of coastline from northwest Venezuela to northeast Argentina (Table 1, Figure 2). Despite the enormous scatter of these birds during the winter, each returned to Bird Island and nested within a few meters of its previous year’s location.

All the birds that wintered on the north coast of South America could be located fairly precisely—within 20–120 kilometers—depending on length of stay. Because the coast runs from west to east, the locations could be determined from longitudes alone despite proximity to the Equator. The Common Terns that wintered in southern Brazil and northeastern Argentina (areas M and N in Figure 2) could be located with similar precision using both latitude and longitude data. The locations of birds that wintered in eastern Brazil north of 20º S (areas K and L) were determined less precisely, because latitude estimates were unreliable and longitude estimates could correspond to a fairly wide range of locations as the coast runs from north-northeast to south-southwest.

All our birds appeared to have been stationary for weeks or months during the winter, because estimates of both latitude and longitude were stable over these long periods, within the expected range of errors. However, six of the nine birds that wintered in eastern Brazil shifted from one stable location to another once or twice during the winter, based on consistent shifts in average longitude by 1–3º. Three Common Terns moved from eastern to southern Brazil (areas K and L to area M) late in the season but spent most of the winter in eastern Brazil (Figure 2). Another Common Tern moved twice from area M in southern Brazil to and from area N in northeastern Argentina. Two Roseate Terns moved from eastern Brazil (area K) to the north coast in February (area J).

Most of these wintering locations had been known or suspected from previous banding recoveries or field studies (Hays et al. 1997, 1999), although it was surprising to find that one Common Tern spent the entire winter as far west as northwest Venezuela (area D, Figure 2). It was also surprising that only one of our 11 Common Terns traveled as far south as northeast Argentina (area N), because this is an area of major winter concentration for the species (Sapoznikow et al. 2002); many Common Terns banded there have been found nesting at Bird Island and other locations in the northeastern United States (Figure 2).

Geolocators on four birds—one Roseate, three Commons—stopped collecting data between January and March, but the remaining 13 devices yielded detailed information on the time the birds spent in their winter quarters. The five Roseate Terns for which we had data spent five or six months (156–186 days) at their wintering sites, much longer than at their breeding sites (73–111 days) or their combined staging and breeding periods in North America (about 120–140 days). The eight Common Terns for which we had data were much more variable in their timing, spending three to eight months (87–235 days) in their winter quarters (Table 1). For six of these birds, length of residency in the winter quarters exceeded that in North America.

Migration Routes and Stopover Sites

We obtained a wealth of new information on migration routes and stopover locations (Table 1, Figure 2). All 17 birds of both species spent most or all of the postbreeding period around Cape Cod and the Islands (area A) from late June or July until they departed on southward migration in August–October. All then made long, direct flights across the western North Atlantic Ocean to the West Indies, most of them making landfall in the vicinity of Puerto Rico or the Dominican Republic. Common Terns made this crossing in two to three days, whereas Roseates traveled faster, completing the flight in one and half to two days.

In addition to Cape Cod, Roseate Terns used three major staging areas in autumn (Table 1, Figure 2): the northeastern Caribbean (area C, mainly in Puerto Rico or the Virgin Islands), a small area in northwest Suriname (area F), and a small stretch of coast in northern Pará, Brazil (area H). All the Roseate Terns that passed through these areas either stopped over for periods of 5–24 days or remained for the entire winter (one each in Suriname and Pará; Table 1).

The behavior of Common Terns was much more variable. Four birds staged in the vicinity of Cape Hatteras (area B) before crossing to the West Indies, and only three of the 11 birds stopped over in the northeastern Caribbean. Common Terns then used at least seven distinct stopover areas along the north coast of South America (Table 1).

On spring migration, there were few stopovers (Table 1, Figure 2). All the Roseates and four of the Commons that spent the winter on the east coast of South America staged in northern Pará (area H), at exactly the same locations they used in the autumn; none of these birds stopped again anywhere in South America (Table 1).

All the birds retraced their autumn migrations along the north coast of South America as far as the eastern Caribbean, but most then followed spring tracks farther to the west of their autumn tracks. Ten Common Terns continued west-northwest across Hispaniola or eastern Cuba, northwest through the Bahamas and then northeast, parallel to the North American coast but well out to sea, close to or outside the edge of the continental shelf, until they reached Cape Cod. One Common Tern differed from all the others in retracing its autumn route, flying from Venezuela through the eastern Caribbean and then directly north to Cape Cod, without passing west of longitude 71º W.

Three Common Terns stopped near Cape Hatteras (area B, Figure 2) for a few days, although they appeared to spend most of this time at sea, sometimes far from shore. Otherwise, none of the Common Terns stopped anywhere for more than a day, although five birds traveled slowly and spent several days moving through the Bahamas.

Three Roseate Terns stopped over in Puerto Rico or the Dominican Republic (area C, Figure 2) for 8–11 days on spring migration, but their subsequent behavior was unexpected. They traveled slowly and erratically over the North Atlantic Ocean, arrived at Bird Island later than usual, were underweight when we trapped them, and did not acquire mates. It seems likely that their spring migration was impaired by the geolocators, although there was no sign that their autumn migration had been affected. We do not understand the difference.

Timing of Migration

The dates of migration included several surprises. The seven female Common Terns all left Cape Cod on southward migration between August 1 and 22 (Nisbet et al. 2011b); one of these birds stopped over at Cape Hatteras and did not leave North America until late September. The other six females then traveled quite fast: five reached Suriname and four reached Brazil before the end of August; four arrived in their final winter quarters (one in Suriname and three in Brazil) between August 21 and September 10. This early migration was unexpected, because Common Terns remain abundant around Cape Cod until mid-September, with significant numbers remaining through October and small numbers into November. Also, adult Common Terns provide prolonged post-fledging parental care: juveniles are accompanied and fed by their parents throughout August and into early September. Our results suggest that after the females leave, males stay behind to care for the juveniles. The four male Common Terns departed over a nearly two-month span between August 12 and October 4; we do not know if they were accompanied by juveniles at the time of departure.

Common Tern. (Photograph by Sandy Selesky).

The autumn migrations of the seven Common Terns that did not migrate directly to winter quarters were extremely varied in their timing. Some birds made prolonged stays in staging areas, for example: 48 days at Cape Hatteras, 23 days in Puerto Rico, 61 days at Aruba, 49 days in Suriname, and—the same bird—49 days in French Guiana. Arrival dates in the winter quarters ranged from September 30 to December 17. One female that had reached Bahía, Brazil (area K) on September 6 stayed there until November 23, then arrived at her final wintering site at Lagoa do Peixe, Brazil (area M), on November 29 (Table 1). Roseate Terns were much less varied in their timing. They started south between August 28 and September 14 and reached their wintering sites between October 3 and November 2.

Spring migration was faster and more tightly scheduled. All the Roseate Terns left their winter quarters between April 7 and 23, and all but two Common Terns left between April 1 and 12. The exceptions were one Common Tern that moved to its staging site at Pará in March and left there on April 22, and one that left its wintering site in Venezuela on April 26. Most of the Common Terns traveled extremely rapidly, with transit times to Cape Cod as short as eight and nine days from Brazil, seven days from Venezuela, and six days from Suriname. Birds that started from Brazil and Suriname flew far west of a direct line to Cape Cod, so their average travel speeds must have been in the range of 500–800 kilometers per day, including time spent feeding and resting. The Roseate Terns also traveled rapidly west as far as the Caribbean, but three of them delayed on the final leg of their journey. All birds of both species spent 5–15 days in the vicinity of Cape Cod before settling at Bird Island, but it was not clear if they used this as a staging area as they did in the autumn.

Contributions to Conservation

The primary goal of our study was to help devise measures to conserve both species of tern by gaining more knowledge about their locations and behavior away from the breeding area, where most mortality is thought to occur (Nisbet 2002, 2014). This knowledge was especially needed for the Roseate Tern, an endangered species that was declining for unknown reasons when we planned the study. Most of the locations where our Roseates wintered remained undiscovered, and nothing was known about their ecology or limiting factors in the winter quarters.

Although we obtained data for only six Roseates, the results added enormously to our knowledge of the species. We identified four major areas used for both wintering and migratory stopovers—the northeastern Caribbean from the Dominican Republic to the Virgin Islands, the northwest coast of Suriname, a small area in northern Brazil, and a larger area on the east coast of Brazil (areas C, F, H and K)—and the dates when they were present. This information is already being used to plan focused field studies— spearheaded by the Canadian Wildlife Service—to study Roseates in those areas.

Many factors that may affect Roseate Terns—and Common Terns—have been identified, including overexploitation of fisheries, offshore oil development, pollution, degradation of mangroves and coral reefs, and development of coastlines for recreation and building of vacation homes (Mostello et al. 2014). We hope that focused field studies in the places we have identified as important for these species will help to determine if and how terns are being impacted. For example, all of the birds in our study used Cape Cod for long periods in autumn, confirming earlier indications that Cape Cod is a major staging area for both Common and Roseate Terns. Studies are currently under way to determine if and how recreational activities on Cape Cod are affecting terns there, to serve as the basis for conservation measures. We hope that our study of migration routes and winter quarters will serve in the same way as the first step toward conservation actions in places outside North America where terns concentrate.


  • Bairlein, F., D. R. Norris, R. Nagel, M. Bulte, C. C. Voigt, J. W. Fox, D. J. T. Hussell, and H. Schmaljohann. 2012. Cross-hemisphere Migration of a 25 g Songbird. Biology Letters 8: 505–507.
  • Beason, J. P., C. Gunn, K. M. Potter, R. A. Sparks and J. W. Fox. 2012. The Northern Black Swift: Migration Path And Wintering Area Revealed. Wilson Journal of Ornithology 124(1): 1–8.
  • DeLuca, W. V., B. K. Woodworth, C. C. Rimmer, P. P. Marra, P. D. Taylor, K. P. McFarland, S. A. Mackenzie and D. R. Norris. 2015. Transoceanic Migration by a 12 g Songbird. Biology Letters 11(4): DOI: 10.1098/rsbl.2014.1045
  • Egevang, C., I. J. Stenhouse, R. A. Phillips, A. Petersen, J. W. Fox and J. R. D. Silk. 2010. Tracking of Arctic Terns Sterna paradisaea Reveals Longest Animal Migration. Proceedings of the National Academy of Sciences of the USA 107 (5): 2078–2081.
  • Hallworth, M. T., T. S. Sillett, S. L. Van Wilgenburg, K. A. Hobson and P. P. Marra. 2015. Migratory Connectivity of a Neotropical Migratory Songbird Revealed by Archival Light- level Geolocators. Ecological Applications 25: 336–347.
  • Hays, H., J. DiCostanzo, G. Cormons, P. D. T. Z. Antas, J. L. X. do Nascimento, I. do L. S. do Nascimento, and R. E. Bremer. 1997. Recoveries of Roseate and Common Terns in South America. Journal of Field Ornithology 68(1): 79–90.
  • Hays, H., P. Lima, L. Monteiro, J. DiCostanzo, G. Cormons, I. C. T. Nisbet, J. E. Saliva, J. A. Spendelow, J. Burger, J. Pierce and M. Gochfeld. 1999. A Nonbreeding Concentration of Roseate and Common Terns in Bahia, Brazil. Journal of Field Ornithology 70(4): 455–464.
  • Heckscher, C. M., S. M. Taylor, J. W. Fox and V. Afanasyev. 2011. Veery (Catharus fuscescens) Wintering Locations, Migratory Connectivity, and a Revision of Its Winter Range using Geolocator Technology. Auk 128(3): 531–542.
  • Hobson. K. A. and K. J. Kardynal. 2015. Western Veeries Use an Eastern Shortest-Distance Pathway: New Insights to Migration Routes and Phenology Using Light-Level Geolocators. The Auk 132(3): 540–550.
  • Mostello, C. S., I. C. T. Nisbet, S. A. Oswald and J. W. Fox. 2014. Non-breeding Season Movements of North American Roseate Terns Sterna dougallii Tracked with Geolocators. Seabird 27: 1–21.
  • Nisbet, I.C.T. 1984. Migration and Winter Quarters of North American Roseate Terns as Shown by Banding Recoveries. Journal of Field Ornithology 55(1): 1–17.
  • Nisbet, I. C. T. 2002. Common Tern Sterna hirundo. No 618 in The Birds of North America (eds., A. Poole and F. Gill). The Birds of North America, Inc., Philadelphia, PA.
  • Nisbet, I. C. T. 2014. Roseate Tern (Sterna dougallii) in The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology:
  • Nisbet, I. C. T., C. S. Mostello, R. R. Veit, J. W. Fox and V. Afanasyev. 2011a. Migrations and Winter Quarters of Five Common Terns Tracked using Geolocators. Waterbirds 34(1): 32–39.
  • Nisbet. I. C. T., P. Szczys, C. S. Mostello and J. W. Fox. 2011b. Female Common Terns Sterna hirundo Start Autumn Migration Earlier Than Males. Seabird 24: 103–106.
  • Sapoznikow, A. Vila, J. Lopez De Casenave and P. Vuillermoz. 2002. Abundance of Common Terns at Punta Rasa, Argentina: a Major Wintering Area. Waterbirds 25(3): 378–381.
  • Spendelow J. A., J. E. Hines, J. D. Nichols, I. C. T. Nisbet, G. Cormons, H. Hays, J. J. Hatch and C. S. Mostello. 2008. Temporal Variation in Adult Survival Rates of Roseate Terns During Periods of Increasing and Declining Populations. Waterbirds 31(3): 309–319.

Ian Nisbet is a former Director of Science at the Massachusetts Audubon Society. He has been a member of the Nuttall Ornithological Club since 1975 and was appointed as an Honorary Member in 2014. He is a Fellow of the American Ornithologists’ Union and of the American Association for the Advancement of Science. He received the Robert Cushman Murphy Award for Distinguished Lifetime Research from the Waterbird Society in 2006. He studied Common and Roseate terns at Bird Island and other breeding sites in Massachusetts from 1970 until 2010 and has published more than 100 papers on terns in scientific journals.

Carolyn Mostello is a Coastal Waterbird Biologist with the Massachusetts Division of Fisheries & Wildlife. She has studied Common and Roseate terns since 1998, and has directed the program to monitor and conserve them at their breeding sites in Buzzards Bay since 2001.

The authors are members of the Recovery Team for the Northeast population of the Roseate Tern.

Funding for this study was provided by the Blake Fund of the Nuttall Ornithological Club (purchase of geolocators for Common Terns), the U.S. Fish and Wildlife Service (purchase of geolocators for Roseate Terns), and the New Bedford Harbor Trustee Council (support of field operations). We thank Chris Rimmer for his helpful comments on an earlier draft of this article.

blog comments powered by Disqus