Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

The palaeobiology of high latitude birds from the early Eocene greenhouse of Ellesmere Island, Arctic Canada


Fossils attributable to the extinct waterfowl clade Presbyornithidae and the large flightless Gastornithidae from the early Eocene (~52–53 Ma) of Ellesmere Island, in northernmost Canada are the oldest Cenozoic avian fossils from the Arctic. Except for its slightly larger size, the Arctic presbyornithid humerus is not distinguishable from fossils of Presbyornis pervetus from the western United States and the Gastornis phalanx is within the known size range of mid-latitude individuals. The occurrence of Presbyornis above the Arctic Circle in the Eocene could be the result of annual migration like that of its living duck and geese relatives, or it may have been a year-round resident similar to some Eocene mammals on Ellesmere and some extant species of sea ducks. Gastornis, along with some of the mammalian and reptilian members of the Eocene Arctic fauna, likely over-wintered in the Arctic. Despite the milder (above freezing) Eocene climate on Ellesmere Island, prolonged periods of darkness occurred during the winter. Presence of these extinct birds at both mid and high latitudes on the northern continents provides evidence that future increases in climatic warming (closer to Eocene levels) could lead to the establishment of new migratory or resident populations within the Arctic Circle.


Fossil birds from within the Arctic Circle are rare. The Cretaceous record is restricted to the flightless diving hesperornithiform birds from Devon Island1, Ellesmere Island2 and Alaska3, as well as bird tracks from Alaska4. The only known Paleogene record of Arctic birds is a report of Gastornis (Diatryma of some older workers; Gastornithidae) and potentially Presbyornis (Anseriformes) from the early Eocene (Wasatchian North American Land Mammal Age, NALMA) of Ellesmere Island, Canadian High Arctic5,6 (~79°N. lat.; see Fig. 1). In contrast, the late Cenozoic record is somewhat better7,8, but understudied in relation to the potential impacts of Pleistocene climate change on birds. As a step toward better understanding the history of birds in the Arctic, we describe the previously mentioned (but unpublished) Presbyornis and Gastornis material from Ellesmere Island9,10 and an additional avian pedal phalanx that has not been discussed before.

Figure 1

Map of northern Canada north of the Arctic Circle showing principal locations where Eocene terrestrial vertebrates have been recovered on Ellesmere Island (inset)–Stenkul Fiord, Swinnerton Peninsula, Strathcona and Bay Fiords.

Star indicates the location of the Bay Fiord fossil-collecting area which preserves the most diverse Eocene vertebrate assemblage from the Arctic, including the bird fossils described in the text. Figure modified from Eberle and Greenwood5.

Beginning in the 1970s, paleontologists Mary Dawson, Robert (Mac) West, Malcolm McKenna and J. Howard Hutchison collected fossil vertebrates from the Paleogene sequence of the Eureka Sound Group exposed on Ellesmere Island and more recent expeditions have added to the accumulated diverse vertebrate fauna, including fish, turtles, crocodilians, many clades of mammals ranging from primates to brontotheres and birds5. In an early report on the fossil vertebrates recovered from Ellesmere Island, the authors mentioned the presence of Gastornis (Diatryma) and a Presbyornis-like bird9. Those specimens remain the only avian material recovered from the Eocene sequence on Ellesmere and are the only avian fossils from the Paleogene of the High Arctic (representing the northernmost Paleogene avian records in North America). The Gastornis material reported from Ellesmere Island is the northernmost record of that taxon in North America10. However, the Paleogene avian material from Ellesmere has not been described or illustrated. The osteological terminology follows Baumel and Witmer11 with English equivalents of the Latin terms.

Eocene terrestrial vertebrates are known from multiple sites on Ellesmere Island, but the best documented and most diverse faunal assemblage, including the bird fossils described here, is from strata of the Eureka Sound Group exposed near Bay Fiord on central Ellesmere Island (Fig. 1). Although Paleogene fossil-bearing rocks were first recognized in the Canadian Arctic over a century ago12, the Eureka Sound Group was not named until 1950 by Troelsen13. Subsequently, the Eureka Sound was reduced to a formation by Tozer14 and remained assigned to that rank until it was raised formally back to a group by Miall15, where it has since remained5. Three sets of names have been applied to strata of the Eureka Sound Group. Following their discovery of Paleogene vertebrates on central Ellesmere Island15,16, West et al.17 subdivided the Eureka Sound Formation into four informal members with the uppermost Member IV containing the Eocene terrestrial vertebrates. Independently, Miall15 and Ricketts18 raised the Eureka Sound Formation to group rank and subdivided it into formations. Consequently (and confusingly), Paleogene terrestrial vertebrate-bearing strata on central Ellesmere Island fall under three different names – Member IV of West et al.19, the Margaret Formation of Miall15 and the Iceberg Bay Formation of Ricketts18.

Today’s general consensus is that the Eureka Sound is a group divisible into numerous formations and generally Miall’s15 formational names are applied to Ellesmere Island, while Ricketts’s18 nomenclature is most often used on nearby Axel Heiberg Island5. Following this usage, we assign the Eocene fossil bird localities (and related stratigraphic horizons) discussed below to the Margaret Formation.

The Margaret Formation comprises coarsening-upward sequences of interbedded cross-bedded sandstone, siltstone, mudstone and coal that are interpreted as proximal delta-front to delta-plain paleoenvironments characterized by abundant shoreline sands, alluvial to estuarine channels, coal swamps, lagoons and bays and well-forested, low-gradient interfluves15,18,20,21.

At Bay Fiord, Eocene vertebrate fossils occur in two fossil assemblages in the Margaret Formation that are separated by a 478 m thick stratigraphic gap in the fossil record21. The stratigraphically lower of the two preserves a diverse early Eocene (late Wasatchian NALMA; ca. 52–53 Ma) fauna that contains fishes, amphibians, alligators, a snake, nine families of turtle, at least 25 genera of mammals5 and the bird fossils described below. The late Wasatchian NALMA fauna is likely within the Early Eocene Climatic Optimum (EECO), which marked the height of global warmth since the extinction of the dinosaurs22. The second, stratigraphically higher fossil assemblage is considerably less diverse (lacking avian material at present), with a handful of Bridgerian NALMA23 (ca. 47–51 Ma) taxa, including the early brontothere Palaeosyops21.

The gastornithid, presbyornithid and indeterminate avian phalanx specimens described below were collected by J. Howard Hutchison from localities ELS76-85, ELS76-42 and ELS76-43, respectively. Fieldwork at Bay Fiord in 2010 concluded that ELS76-42 and ELS76-43 are on the same stratigraphic horizon within the lower fossil assemblage at approximately the 1420–1425 m level on the composite stratigraphic section of Eberle and Eberth21 and ELS76-85 is the most productive Eocene vertebrate locality in the Arctic (stratigraphically higher than the other localities at 1437.5 m)21. The stratigraphically lower avian localities are about 805 meters above the contact between the Margaret and underlying Mt. Moore Formation (Eberth, pers. comm. to JE, 2015).


Systematic Palaeontology

Family Gastornithidae Fürbringer 1888

Gastornis sp.

Material. Canadian Museum of Nature (Ottawa, Canada) CMN 32412. The specimen is a nearly complete pedal phalanx from locality ELS76-85, Margaret Formation, Eureka Sound Group, central Ellesmere Island, Nunavut, Canada (early Eocene; late Wasatchian NALMA).

Description. CMN 32412 lacks the stouter and more trapezoidal outline of more distal phalanges of Gastornis24,25 and has the broadened articular end present in the most proximal phalanx. The proximal dorsal to dorsolateral portion of the specimen is missing and there is no evidence of a dorsoplantarly-oriented ridge on the articular surface that would have fit into the trochlear furrow. The proximal end is wider (28 mm, broken) than the distal end (24.8 mm). The width of the proximal end would have been greater than 44% of the length of the bone (62.2 mm; since it is damaged). The lateral edge of the bone is relatively straight (dorsal view); while the medial edge is slightly more concave. The plantar lip of the lateral collateral ligament fossa (the lateral plantar rim of the trochlea) is visible in dorsal view (extending lateral to the shaft) and continues proximally about one-third the length of the bone. In contrast, the equivalent medial part of the trochlea is not visible in dorsal view. The trochlea has a shallow trochlear furrow and the lateral trochlear ridge extends slightly proximal to the medial ridge on the plantar surface. The lateral collateral ligament pit is large and deepest distoplantarly and the medial pit is nearly circular in outline. In addition in lateral view, the plantar surface of the bone is concave, with the trochlea and condylar articulations being dorsoplantarly wider than the shaft.

Order Anseriformes Wagler 1831

Family Presbyornithidae Wetmore 1926

Presbyornis cf. P. pervetus Wetmore 1926

Material. CMNFV 53369. The specimen is the distal approximately one-third of a left humerus (Fig. 2), from locality ELS76-43 (=77H7-9-3), Margaret Formation, Eureka Sound Group, central Ellesmere Island, Nunavut, Canada (early Eocene; late Wasatchian NALMA).

Description. The bone is dark brown in color and it exhibits some sand-induced surface pitting. That pitting is more pronounced on the distal end of the bone. Despite that overprint, the surface of the bone appears to have been smooth, indicating that the specimen likely is from an osteologically adult individual26. The proximal end of the shaft is subcircular in cross-section, but slightly wider medio-laterally.

From the proximal end towards the distal end of the cranial surface, the shaft flattens from a distinctly convex surface proximally. There is a distinct brachial (muscle) fossa that is deepest ventrodistally and shallows dorsally. The fossa’s distal end has a distinct border. Overall, the fossa is ovoid in outline with its long axis oriented dorsoproximal to ventrodistal. The distal edge of the brachial fossa is at the same proximodistal level as the proximal end of the ventral collateral ligament attachment at the apex of the ventral supracondylar tubercle. Just ventral to the proximal end of the apex of the supracondylar tubercle is a poorly preserved muscle scar pit (attachment of pronator brevis of Howard27 that is equivalent to the origination of the m. flexor carpi ulnaris of Livezey28) that is separate from the ventral collateral ligament attachment. The ventral collateral ligament scar on the supracondylar tubercle is proximodistally elongate and its dorsal margin is relatively straight. The ventral condyle extends distal to the dorsal condyle. The area distal to the brachial fossa and proximal to the condyles is flat to slightly concave. The dorsal condyle extends proximally to the same level as the proximal end of the ventral collateral ligament scar. The long axis of the outline of the dorsal condyle is not parallel to the long axis of the humerus because the distal end is located dorsal to the proximal end.

The shaft is convex in caudal view. The olecranon fossa is relatively small and restricted to the area immediately caudal to the ventral condyle. The fossa is deepest ventrally. The distal end of the bone is worn and there is a slight hint of a shallow scapulotriceps groove at the most distal part of the caudal side. The dorsal epicondyle has a pair of muscular pits (present in other anseriforms), but they are poorly preserved because the division between the two pits cannot be discerned. The dorsal epicondyle extends further proximal than the supracondylar tubercle/collateral ligament scar. The ventral side of the specimen is poorly preserved and no muscle scars are visible.

Aves indet.

Material. CMNFV 53368, a phalanx (Fig. 2) from locality ELS76-42 (=76H7-10-10), Margaret Formation, Eureka Sound Group, central Ellesmere Island, Nunavut (Early Eocene; Wasatchian NALMA).

Description. The specimen appears to be the proximal phalanx likely from digit III of the left side. The bone is off-white to gray in color and is deeply pitted all over its surface. The pitting is caused by sand grains that were embedded in the bone surface and does not indicate an early ontogenetic age (Fig. 2). The proximal end is mediolaterally crushed/flattened and the left side (dorsal view) of the distal end is broken off. The proximal articulation maintains a central ridge for articulation with the tarsometatarsus. In lateral view, the articulation is C-shaped. The distal end (lateral sides) has a deep collateral ligamental pit. The furrow for the articulation with the next more distal phalanx extends dorsally and plantarly to about the same proximal level and that level is proximal to the ligamental pits. The plantar side appears to have been flatter than the dorsal (convex) side, but the crushing of the specimen has distorted the shape.


The length of the Ellesmere Gastornis specimen (62.2 mm) is less than the largest specimens of Gastornis from North America (68.6–80.1 mm) and that from Louvois, France (75.9 mm)25. The morphology of the Ellesmere specimen is similar to that reported and illustrated for the proximal phalanx of digit IV of Gastornis from Louvois25, Wyoming24, Monthelon29 and New Jersey30, but the plantar ridges mentioned in some Gastornis specimens appear to be absent in this specimen. The proximal widening of the articular region (in dorsal view) is similar across specimens. The Ellesmere specimen lacks the medially expanded plantar rim of the trochlea (plantar lip of the collateral ligament pit) that is visible in dorsal view in the Paleocene Louvois specimen25 and that state appears absent in other Gastornis specimens24,29. Furthermore, the plantar lip of the collateral ligament pit in CMN 32412 extends relatively more proximal than the state in the Louvois specimen25, but is similar to Gastornis from Wyoming24. The length of the phalanx and the ratio of the proximal width to the length of the bone (above) are within the ranges reported for Gastornis25.

The variation in the morphology of the humerus seems to be within that seen among the abundant Wyoming specimens of Presbyornis pervetus (Fig. 2). There are no apomorphies (versus other Paleogene anseriform taxa) in the distal humerus for Presbyornis, but there also are no character differences between the Ellesmere Island fossil and specimens of Presbyornis pervetus from Wyoming. The Ellesmere specimen was directly compared to casts of Presbyornis pervetus from Wyoming. The postcranial skeleton of Presbyornis is notoriously plesiomorphic31,32. The same shape, depth and overall morphology of the brachial fossa can be seen among Presbyornis specimens from more southerly locations (Fig. 2). However, the Ellesmere specimen is on the larger end of the size range present in Wyoming (Table 1) and the Ellesmere Island specimen also is within the known stratigraphic range of Presbyornis in Wyoming5,33 (i.e., Wasatchian NALMA). There are no characters that appear to separate the Ellesmere humerus from what is currently recognized as Presbyornis pervetus, but also no clear derived characters to unite them into a clade.

Table 1 Measurements (in mm) of the distal end of the humerus of Presbyornis specimens.

Various authors have suggested that the morphological variation exhibited among the Wyoming Presbyornis bone assemblages might represent multiple closely related species33,34, but at present no one has discerned a way to separate any potential species within the Presbyornis pervetus collections from North America. In contrast, an early Eocene species of Presbyornis, P. mongoliensis, has been described from Mongolia35 and it likely does represent a species different from the penecontemporaneous North American Presbyornis pervetus. That Mongolian material also is smaller than the Ellesmere specimen35 (Table 1). Thus, it is possible that the Ellesmere specimen represents a taxon separate from Presbyornis pervetus, but there is no way at present to determine that status. Furthermore, the slightly larger size of the Ellesmere specimen compared to individuals at lower latitudes could be the result of Bergman’s Rule and that cline has been proposed for the Arctic and mid-latitude mammal Heptodon from the Eocene21.

The indeterminate avian phalanx is consistent with Presbyornis (clearly lacking any autapomorphies), but it also easily could belong to another taxon of bird (given its somewhat more robust morphology), though its size and morphology is certainly inconsistent with that of the much larger Gastornis, which has stouter and more robust pedal phalanges (Fig. 2). In contrast to the Presbyornis specimen, the Gastornis phalanx is not larger than individuals known from further south in North America, but is within the known size variation for the group. The variation in size among Gastornis individuals has been attributed to both species differences and the possibility of sexual dimorphism25.


The early Eocene vertebrate fauna from Ellesmere Island is from a unique temporal interval of mild mean annual temperatures and above-freezing winters that allowed crocodilians and turtles to survive above the Arctic Circle. Today, the Eocene vertebrate localities on central Ellesmere are at nearly 79°N latitude and during the Eocene, they were just a few degrees further south36,37. The fossils and sedimentology indicate a lush, rain forest community on a coastal delta plain. Multiple palaeoclimate proxies, ranging from oxygen isotope analysis of vertebrate bones and teeth to palaeofloral analyses, estimate a mild temperate climate for the Eocene High Arctic, where winters remained at or just above freezing and summer temperatures extended to 20 °C or higher5. These temperatures are a far cry from today’s High Arctic, where central Ellesmere Island experiences a mean annual temperature of −19 °C, a warm month mean temperature of about 6 °C and a cold month mean temperature of −38 °C or colder38.

Despite the mild Eocene Arctic climate, the vertebrate fauna would have experienced months of total darkness and cooler temperatures during the winter. Recent isotopic work suggests that some mammals, including the hippo-like Coryphodon, were year-round residents in the High Arctic39. Given that Gastornis was large (approaching 2 m) and flightless, it likely also was a year-round resident of the Arctic. In contrast, the volant Presbyornis might have been a seasonal migrant to the Arctic. The previously reported geographic distribution of Presbyornis pervetus was restricted to Utah, Wyoming and Colorado33 with all records (USA and Mongolia) between 37° and 45°N. latitude, as compared to the 79°N. latitude of the new specimen that is greater than 4000 km further north (with a similar geographic disparity for specimens of Gastornis in the early Eocene of North America).

Extant galloanserines include migratory species and migration in birds is (at least in part) genetically controlled40,41. Therefore, Presbyornis, which is phylogenetically within the crown group of galloanserines31,32,34, should have had the genetic predisposition for migration and the ability to migrate, given its skeletal morphology. However, it is unknown whether those genes were expressed in any presbyornithid because both migratory and non-migratory species (and populations) are present among various species of extant galloanserines42. The best data that would help support (though not prove) a migratory lifestyle for Presbyornis would be the discovery of juvenile specimens in the Arctic. However, both of the Ellesmere specimens reported above are from osteologically adult individuals. Furthermore, migratory behavior is not a requirement for a large geographic range because many extant non-migratory species have large geographic ranges (e.g., Osprey, Barn Owl and Ostrich). However, as a filter-feeding bird, Presbyornis likely would have had a difficult time obtaining food during the winter decrease in (primary) productivity that occurred during the extended dark and low light periods of the Paleogene Arctic winters.

In contrast to migration, some galloanserine birds today (such as Ruffed Grouse, Spruce Grouse, Sharp-tailed Grouse, Willow Ptarmigan and Rock Ptarmigan) are Arctic residents43 and even some waterfowl winter above the Arctic Circle44. Those waterfowl are able to forage and survive winters today that are much harsher than those reconstructed for the early Eocene of Ellesmere (see above). Ducks (Anatidae) of many lineages forage at night44,45 and extinct waterfowl species presumably would have been able to forage during the dark winters of the Eocene Arctic. Recent work46 has demonstrated greater tetrapod foraging activity and greater biological activity during the Arctic winter than previously known, hinting that such levels of activity should extend into the deep past. Even though individual waterfowl survive wintering above the Arctic Circle, obtaining food is difficult and a variety of hypotheses have been erected as to how the individual birds compensate for the loss of daylight and reduction in prey availability44. Despite those challenges, it is possible that the occurrence of Presbyornis in the Arctic Eocene may have been as a resident, as closely related galloanserine birds are able to survive the much harsher Arctic winters today.

The current geographic ranges of migratory birds are shifting towards higher latitudes and climatic warming has been implicated in phenological changes among birds47,48. Migrating and breeding at higher latitudes have been correlated with a reduced predation risk49. It is possible that the far northern occurrence of Presbyornis in the Arctic Circle could be either the result of a migratory population or alternatively a northern geographic expansion paralleling the pattern observed for many extant birds over the last century of warming. If anything, species of Presbyornis likely were affected by warming events in the early Eocene parallel to the response of extant birds to current warming. In contrast, the large flightless Gastornis definitely was an Arctic resident. The diet of gastornithids is debated and varies between carnivory and some type of durophagous herbivory50,51, but the most recent (isotopic and morphological) data support herbivory. Palynological and macrofloral studies indicate lush, early Eocene Arctic forests, comparable in richness to today’s swamp-cypress and broadleaf floodplain forests in the southeastern United States5,52 and like the herbivorous mammals on Ellesmere Island in the Eocene, Gastornis would have had the same access to dietary materials through the winter that kept the resident mammals alive. That the jaw anatomy (and inferred adductor musculature) of Gastornis suggests that it fed on tough plant material51 fits well with the Arctic winter forage that was available to the large mammalian herbivores, including wood and leaf litter, fungi and evergreen conifers39.

Gastornithidae is Holarctic in distribution and Presbyornis (P. pervetus and P. mongoliensis) is known from North America and Asia during the Early Eocene33,35,53. Purported records of presbyornithids from Europe are not well supported54. The geographic and stratigraphic distribution of those two bird clades suggests that intercontinental dispersal (rather than intercontinental vicariance) likely was the process resulting in their widespread geographic distribution in the Northern Hemisphere55. Dispersal across the northern continents via high latitude land corridors has been widely hypothesized for mammalian clades with similar geographic and temporal distributions and those dispersals have been tied to warming climates and the opening of northern high latitude land bridges56,57. The Holarctic distribution of Gastornithidae is similar to several early Eocene mammalian genera, including Coryphodon, the rodent Paramys, the mesonychid Pachyaena and the carnivoran Miacis, which inhabited mid- and high latitude North America, as well as Asia and Europe in the early Eocene and the North American/Asian distribution of Presbyornis parallels that of several Arctic genera, including the hyaenodontid creodont Prolimnocyon, the brontotheres Eotitanops and Palaeosyops and the tapiroids Heptodon and Homogalax that occur only in North America and Asia (not Europe)58,59,60. Those biogeographic similarities between birds and mammals firmly places birds within the hypotheses of climate-related intercontinental dispersals in the early Eocene.

Occurrence of a diversity of terrestrial and aquatic vertebrate taxa in early Eocene-aged strata on Ellesmere Island indicates that a complex Arctic ecosystem was present during that time and that productivity must have been high enough to sustain those higher trophic levels of vertebrates, even during the darkness that occurred through part of the Eocene winter. Given the geological and paleontological background (on Ellesmere Island and elsewhere), it is plausible that the current trend of climatic warming in the 20th and 21st centuries could lead to the establishment of Arctic resident populations of tetrapod species that currently occur much further south, as well as facilitate the intercontinental dispersal of avian clades among the continents of the Northern Hemisphere.

Additional Information

How to cite this article: Stidham, T. A. and Eberle, J. J. The palaeobiology of high latitude birds from the early Eocene greenhouse of Ellesmere Island, Arctic Canada. Sci. Rep. 6, 20912; doi: 10.1038/srep20912 (2016).


  1. Wilson, L., Chin, K., Cumbaa, S. & Dyke G. A high latitude hesperornithiform (Aves) from Devon Island: palaeobiogeography and size distribution of North American hesperornithiforms. J. Syst. Palaeo. 9, 9–23 (2011).

    Google Scholar 

  2. Hills, L. V., Nicholls, E. L., Núñez-Betelu, L. & McIntyre, D. 1999. Hesperornis (Aves) from Ellesmere Island and palynological correlation of known Canadian localities. Can. J. Earth Sci. 36, 1583–1588 (1999).

    ADS  Google Scholar 

  3. Bryant, L. Hesperornis in Alaska. PaleoBios 40, 1–8 (1983).

    Google Scholar 

  4. Fiorillo, A., Hasiotis, S., Kobayashi, Y., Breithaupt, B. & McCarthy, P. 2011. Bird tracks from the Upper Cretaceous Cantwell Formation of Denali National Park, Alaska, USA: a new perspective on ancient northern polar vertebrate biodiversity. J. Syst. Palaeo. 9, 33–49 (2011).

    Google Scholar 

  5. Eberle, J. & Greenwood, D. Life at the top of the greenhouse Eocene world— A review of the Eocene flora and vertebrate fauna from Canada’s High Arctic. GSA Bull. 124, 3–23 (2012).

    Google Scholar 

  6. Dawson, M. R. In Canada’s Missing Dimension: Science and History in the Canadian Arctic Islands, Vol. 1 (ed. Harrington, C. R. ) 91–104 (Canadian Museum of Nature, 1990).

  7. Olson, S. A Late Pliocene Occurrence of the Thick-billed Murre (Alcidae: Uria lomvia) on St. George Island, Pribilofs, Alaska. Paleo. J. 47, 1365–1368 (2013).

    Google Scholar 

  8. Whitlock, C. & Dawson, M. R. Pollen and vertebrates of the early Neogene Haughton Formation, Devon Island, Arctic Canada. Arctic 43, 324–330 (1990).

    Google Scholar 

  9. West, R. M. & Dawson, M. R. Vertebrate Paleontology and the Cenozoic History of the North Atlantic Region. Polarforschung 48, 103–119 (1978).

    Google Scholar 

  10. Andors, A. V. In Papers in Avian Paleontology. Honoring Pierce Brodkorb (ed. Campbell, K. E. Jr. ) 109–126 (Sci. Ser., Nat. Hist. Mus. Los Angeles Co. 36, 1992).

  11. Baumel, J. J. & Witmer, L. In Handbook of avian anatomy: Nomina Anatomica Avium 2nd edn (eds Baumel, J. J., King, A. S., Breazile, J. E., Evans, H. E., Vanden Berge, J. C. ) 45–132 (Publications of the Nuttall Ornithological Club 23, 1993).

  12. Greely, A. W. In Three Years of Arctic Service, Vol. 2, Appendix 14, 419–420 (Charles Scribner’s Sons, 1886).

  13. Troelson, J. C. Contributions to the geology of northwest Greenland, Ellesmere Island and Axel Heiberg Island. Meddeleser om Grǿnland 149, 1–85 (1950).

    Google Scholar 

  14. Tozer, E. T. in Geology of the north-central part of the Arctic Archipelago, Northwest Territories (Operation Franklin) (ed. Fortier, Y. O. ) 74–95 (Geological Survey of Canada Memoir 320, 1963).

  15. Miall, A. D. The Eureka Sound Group (Upper Cretaceous-Oligocene), Canadian Arctic Islands. Bull. Can. Pet. Geol. 34, 240–270 (1986).

    Google Scholar 

  16. West, R. M., Dawson, M. R., Hutchison, J. H. & Ramaekers, P. Paleontologic evidence of marine sediments in the Eureka Sound Formation of Ellesmere Island, Arctic Archipelago, N.W.T., Canada. Can. J. Earth Sci. 12, 574–579 (1975).

    ADS  Google Scholar 

  17. Dawson, M. R., West, R. M., Langston, W. & Hutchison, J. H. Paleogene terrestrial vertebrates: Northernmost occurrence, Ellesmere Island, Canada. Science 192, 781–782 (1976).

    ADS  CAS  Google Scholar 

  18. Ricketts, B. D. New formations in the Eureka Sound Group, Canadian Arctic Islands. Curr.Res., Part B, Geol. Surv. Can. Paper 86–1B, 363–374 (1986).

  19. West, R. M., Dawson, M. R., Hickey, L. J. & Miall, A. D. In Geology of the North Atlantic Borderlands (eds Kerr, J. W. & Fergusson, A. J. ) 279–298 (Canadian Society of Petroleum Geologists Memoir, 1981).

  20. Riediger, C. L. & Bustin, R. M. The Eureka Sound Formation, southern Ellesmere Island. Bull. Can. Pet. Geol. 35, 123–142 (1987).

    Google Scholar 

  21. Eberle, J. & Eberth, D. Additions to the Eocene Perissodactyla of the Margaret Formation, Eureka Sound Group, Ellesmere Island, Arctic Canada. Can. J. Earth Sci. 52, 123−133 (2015).

    ADS  CAS  Google Scholar 

  22. Zachos, J. C., Pagani, M., Sloan, L., Thomas, E. & Billups, K. Trends, rhythms and aberrations in global climate 65 Ma to present. Science 292, 686−693 (2001).

    ADS  CAS  PubMed  Google Scholar 

  23. Smith, M. E., Carroll, A. R. & Singer, B. S. 2008. Synoptic reconstruction of a major ancient lake system: Eocene Green River Formation, western United States. GSA Bull. 120, 54−84 (2008).

    Google Scholar 

  24. Matthew, W. D. & Granger, W. The skeleton of Diatryma, a gigantic bird from the lower Eocene of Wyoming. Bull. Am. Mus. Nat. Hist. 37, 307–326 (1917).

    Google Scholar 

  25. Mourer-Chauviré, C. & Bourdon, E. The Gastornis (Aves, Gastornithidae) from the Late Paleocene of Louvois (Marne, France). Swiss J. Palaeo (in press). 10.1007/s13358-015-0097-7.

  26. Watanabe, J. & Matsuoka, H. In Paleornithological Research 2013: Proceedings of the 8th International Meeting of the Society of Avian Paleontology and Evolution, Vienna, 2012 (eds Göhlich, U. B. & Kroh, A. ) 279–306 (Verlag Naturhistorisches Museum, Vienna, 2013).

  27. Howard, H. The avifauna of Emeryville Shellmound. Univ. Calif. Pubs. Zool. 32, 301–394 (1929).

    Google Scholar 

  28. Livezey, B. C. A phylogenetic analysis of recent anseriform genera using morphological characters. The Auk 103, 737–754 (1986).

    Google Scholar 

  29. Schaub, S. Uber Eocane Ratitenreste in der ostologishen Sammlung der Basler Museums. Verhandlungen der Naturforschenden Gesellschaft Basel 40, 588–598 (1929).

    Google Scholar 

  30. Marsh, O. C. A gigantic bird from the Eocene of New Jersey. American Journal of Science 48, 344 (1894).

    ADS  Google Scholar 

  31. Livezey, B. C. A phylogenetic analysis of basal Anseriformes, the fossil Presbyornis and the interordinal relationships of waterfowl. Zool. J. Linn. Soc. 121, 361–428 (1997).

    Google Scholar 

  32. Ericson, P. G. P. Systematic position of the Paleogene family Presbyornithidae (Aves). Zool. J. Linn. Soc. 121, 429– 483 (1997).

    Google Scholar 

  33. Ericson, P. G. P. Systematic revision, skeletal anatomy and paleoecology of the New World early Tertiary Presbyornithidae (Aves: Anseriformes). PaleoBios 20, 1–23 (2000).

    Google Scholar 

  34. Elzanowski, A. & Stidham, T. 2010. Morphology of the quadrate of the Eocene anseriform Presbyornis and extant galloanserines (Aves: Galloanseres). J. Morph. 271, 305–323 (2010).

    PubMed  Google Scholar 

  35. Kurochkin, E. & Dyke, G. A large collection of Presbyornis (Aves, Anseriformes, Presbyornithidae) from the late Paleocene and early Eocene of Mongolia. Geol. J. 45, 375–387 (2010).

    Google Scholar 

  36. Irving, E. & Wynne, P. J. The paleolatitude of the Eocene fossil forests of Arctic Canada in Tertiary Fossil Forests of the Geodetic Hills, Axel Heiberg Island, Arctic Archipelago (eds Christie, R. L. & McMillan, N. J. ) 209–212 (Geological Survey of Canada Bulletin 403, 1991).

  37. McKenna, M. C. Eocene paleolatitude, climate and mammals of Ellesmere Island. Palaeogeogr. Palaeoclimatol. Palaeoecol. 30, 349–362 (1980).

    Google Scholar 

  38. Environment Canada, National Climate Data and Information Archive (2004) Available at: (Date of access:19/11/2015).

  39. Eberle, J., Fricke, H. & Humphrey, J. Lower-latitude mammals as year-round residents in Eocene Arctic forests. Geology 37, 499–502 (2009).

    ADS  CAS  Google Scholar 

  40. Pulido, F. The genetics and evolution of migration. BioSci. 57, 165–174 (2007).

    Google Scholar 

  41. Mueller, J. C., Pulido, F. & Kempenaers, B. Identification of a gene associated with avian migratory behavior. Proc. Roy. Soc. B 278, 2848–2856 (2011).

    CAS  Google Scholar 

  42. Finnegan, L. et al. Fine-scale analysis reveals cryptic patterns of genetic structure in Canada Geese. The Condor 115, 738–749 (2013).

    Google Scholar 

  43. U.S. Fish and Wildlife Service, Bird List-Arctic National Wildlife Refuge Alaska. Available at: (Date of access: 30/07/2015) (2014).

  44. Systad, G., Bustnes, J. & Erikstad, K. Behavioral responses to decreasing day length in wintering sea ducks. The Auk 117, 33–40 (2000).

    Google Scholar 

  45. McNeil, R., Drapeau, P. & Goss-Custard, J. 1992. The occurrence and adaptive significance of nocturnal habits in waterfowl. Biol. Rev. 67, 381–419 (1992).

    Google Scholar 

  46. Berge, J. et al. Unexpected levels of biological activity during the polar night offer new perspectives on a warming Arctic. Curr. Biol. 25, 2555–2561 (2015).

    CAS  Google Scholar 

  47. Lehikoinen, A. et al. 2013. Rapid climate driven shifts in wintering distributions of three common waterbird species. Global Change Biol. 19, 2071–2081 (2013).

    ADS  Google Scholar 

  48. Mason, S. C. et al. Geographical range margins and many taxonomic groups continue to shift polewards. Biol. J. Linn. Soc. 115, 586–597 (2015).

    Google Scholar 

  49. McKinnon, L. et al. Lower predation risk for migratory birds at high latitudes. Science 327, 326–327 (2010).

    ADS  CAS  Google Scholar 

  50. Witmer, L. & Rose, K. Biomechanics of the jaw apparatus of the gigantic Eocene bird Diatryma: implications for diet and mode of life. Paleobiol. 17, 95–120 (1991).

    Google Scholar 

  51. Angst, D. et al. Isotopic and anatomical evidence of an herbivorous diet in the Early Tertiary giant bird Gastornis. Implications for the structure of Paleocene terrestrial ecosystems. Naturwissenschaften 101, 313–322 (2014).

    ADS  CAS  Google Scholar 

  52. Harrington, G., Eberle, J., Le-Page, B., Dawson, M. & Hutchison, J. Arctic plant diversity in the Early Eocene greenhouse. Proc. R. Soc. B. 279, 1515–1521 (2012).

    Google Scholar 

  53. Buffetaut, E. The giant bird Gastornis in Asia: a revision of Zhongyuanus xichuanensis Hou 1980, from the early Eocene of China. Paleo. J. 47, 1302–1307 (2013).

    Google Scholar 

  54. Mayr, G. Paleogene Fossil Birds (Springer-Verlag, 2009).

  55. Buffetaut, E. & Angst, D. Stratigraphic distribution of large flightless birds in the Palaeogene of Europe and its palaeobiological and palaeogeographical distribution. Earth-Sci. Rev. 138, 394–408 (2014).

    ADS  Google Scholar 

  56. Krause, D. W. & Maas, M. C. In Dawn of the Age of Mammals in the Northern Part of the Rocky Mountain Interior, North America (eds Bown, T. M. & Rose, K. D. ) 71–105 (Geological Society of America Special Paper 243, 1990).

  57. Beard, K. C. & Dawson, M. R. Intercontinental dispersal of Holarctic land mammals near the Paleocene-Eocene boundary: Paleogeographic, paleoclimatic and biostratigraphic implications. Bull. Société Géologique France 170, 697−706 (1999).

    Google Scholar 

  58. Dawson, M. R. 2001. Early Eocene rodents (Mammalia) from the Eureka Sound Group of Ellesmere Island, Canada. Can. J. Earth Sci. 38, 1107–1116, 10.1139/e01-010 (2001).

    ADS  Article  Google Scholar 

  59. Dawson, M. R. 2012. Coryphodon, the northernmost Holarctic Paleogene pantodont (Mammalia) and its global wanderings. Swiss J. Palaeon. 131, 11–22, 10.1007/s13358-011-0028-1 (2012).

    Article  Google Scholar 

  60. Eberle, J. J. & McKenna, M. C. Early Eocene Leptictida, Pantolesta, Creodonta, Carnivora and Mesonychidae (Mammalia) from the Eureka Sound Group, Ellesmere Island, Nunavut. Can. J. Earth Sci. 39, 899–910, 10.1139/e02-001 (2002).

    ADS  Article  Google Scholar 

Download references


We thank Mary Dawson and J. Howard Hutchison for discussions about Ellesmere Island and for allowing us to study the avian material that they collected in the 1970s on Ellesmere Island. David Eberth kindly provided information on the stratigraphic placement of the fossil bird localities. Amy Henrici provided collections assistance at the Carnegie Museum of Natural History. Natalia Rybczynski and Margaret Currie provided assistance with specimen numbers at the CMN and access to the Gastornis specimen. Patricia Holroyd provided access to Presbyornis specimens in the UCMP. Nicole Neu-Yagle aided with imaging the Gastornis phalanx. TAS is funded by the Chinese National Natural Science Foundation (NSFC41472025) and the Strategic Priority Research Program of the Chinese Academy of Sciences (CAS, XDB03020501). 2010 fieldwork that refined the stratigraphy of the fossil localities on Ellesmere Island was supported by the National Science Foundation (ARC0804627 to JJE). Figure 1 was created by Leigh Anne McConnaughey.

Author information




T.A.S. and J.J.E. wrote the manuscript and created Figure 2. All authors reviewed the manuscript.

Figure 2

The Ellesmere Island avian fossils.

The (probable) left proximal phalanx of digit IV from Gastornis sp. (CMNFV 32412): (A) dorsal view, (B) lateral view, (C) plantar view and (D) medial view. The distal humeri of Presbyornis: (E) CMNFV 53369 from the Margaret Formation of Ellesmere Island, Canada and (F) University of California Museum of Paleontology UCMP 119395 from the Wasatch Formation of Wyoming, USA in cranial view. The (G) right and (H) left surfaces of the indeterminate avian pedal phalanx (CMNFV 53368) showing the extensive sand pitting on the surface. Abbreviations: bf, brachial fossa; dc, dorsal condyle; lp, collateral ligament pit; vc, ventral condyle; and vl, facet for the ventral collateral ligament on the ventral supracondylar tubercle.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Stidham, T., Eberle, J. The palaeobiology of high latitude birds from the early Eocene greenhouse of Ellesmere Island, Arctic Canada. Sci Rep 6, 20912 (2016).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing