Albertonykus borealis was a very small alvarezsaurid from the early Maastrichtian Horseshoe Canyon Formation of Alberta, Canada. Its forelimb and hind-limb bones represent at least two individuals, but no skull, spine or articulated skeleton is known. A powerful one-clawed arm led to a feeding hypothesis that researchers tested against insect traces in fossil wood from the same formation.
Quick facts
| Scientific name | Albertonykus borealis Longrich & Currie, 2009 |
|---|---|
| Group | Alvarezsauria, Alvarezsauridae |
| Age | Early Maastrichtian, about 68.5 million years ago |
| Range | Dry Island Buffalo Jump, Alberta, Canada |
| Formation | Upper Tolman Member, Horseshoe Canyon Formation |
| Holotype | TMP 2001.45.91, a left ulna |
| Other material | A manual claw, tibiae, metatarsals and toe bones |
| Length | Approximately 0.7–1.1 m by scaling |
How was the feeding idea tested?
Repeated and differently sized limb elements show that the assemblage is not one skeleton.
A robust ulna and large main claw suit local disruption of resistant material, not long-reaching grasping.
Horseshoe Canyon wood contains galleries resembling those made by modern dampwood termites.
No stomach content, coprolite or claw-marked wood ties Albertonykus itself to termites.
Discovery in the Albertosaurus bonebed
The main fossils came from the Albertosaurus bonebed at Dry Island Buffalo Jump. Barnum Brown’s expedition discovered the locality in 1910, and Philip Currie relocated it in 1997. During excavations in the early 2000s, small limb bones entered the Royal Tyrrell Museum collections among far more numerous tyrannosaur remains. Nicholas Longrich recognised the alvarezsaurid claw and associated elements, then named the genus with Currie in 2009.
Albertonykus combines Alberta with Greek onyx, claw; borealis means northern and referred to its position relative to most Asian and South American alvarezsaurids then known. TMP 2001.45.91, a left ulna, is the holotype. Referred remains include a large first-finger claw, two tibiae, metatarsal parts and toe phalanges.
Repeated elements and size differences show at least two individuals in the main bonebed, while two toe bones came from nearby Horseshoe Canyon localities. The diagnosis therefore rests on a combination of isolated limb features, not one complete skeleton. Shared burial with Albertosaurus does not establish a predator-prey interaction or simultaneous death.
Arm, legs and body size
The ulna was short and robust, with a developed olecranon and muscle attachments arranged to generate strong elbow extension. As in specialised alvarezsaurids, one main hand digit carried a hooked claw. Such an arm had limited reach and was poorly suited to grasping large prey. Its mechanics favour repeated short movements close to the chest.
The arms were too short for extensive whole-body burrowing like a specialised digging mammal. Local disruption of bark, rotten wood, soil crust or an insect shelter is mechanically more plausible.
Tibiae and metatarsals indicate a light, elongated hind limb. Better-known relatives had an arctometatarsal foot, with the middle element pinched between its neighbours to distribute loads. Comparison supports an active ground-dwelling animal, but no trackway or complete leg gives a speed.
An early popular estimate suggested about 75 centimetres and 3.6 kilograms. Later scaling commonly gives roughly 0.7–1.1 metres and up to several kilograms. The range depends on which relative restores the missing vertebral column, tail and head, so no exact nose-to-tail value is available.
The termite hypothesis
Large surface termite mounds are not documented for this time, so Longrich and Currie considered wood-dwelling social insects. Fossilised Horseshoe Canyon wood contains galleries resembling those of modern dampwood termites. This matters because it demonstrates a suitable local food source instead of relying only on analogy with living anteaters.
The match is still circumstantial. No digestive remains or wood with attributable claw marks belongs to Albertonykus. It may have eaten several kinds of invertebrates or other small prey. Its very short arms also favour opening a local patch of bark or rotten wood over excavating a deep burrow.
Age, environment and dispersal
Sandstones, mudstones and coals of the upper Tolman Member of the Horseshoe Canyon Formation record rivers, floodplains and estuarine settings near the retreating Western Interior Seaway. The locality lay near 60° north and experienced marked seasonality. The bonebed dates to the early Maastrichtian, about 68.5 million years ago.
Bonebed accumulation may reflect a catastrophic event, but the small Albertonykus bones need not have entered it exactly with every tyrannosaur. Rare hadrosaur and other terrestrial vertebrate remains also occur in the assemblage.
The 2009 analysis placed Albertonykus near derived Asian alvarezsaurids and supported a then-popular dispersal route from South America through North America into Asia. Older Chinese transitional forms later made the early history more complex. Albertonykus remains a secure North American record, but one genus cannot establish a single migration route.
What is known about its appearance
Only limb elements directly define its appearance. A small body, strong short arms and slender legs are well supported. A narrow skull with small teeth and a feathered covering are inferred from relatives such as Mononykus and Shuvuuia. Neck length, tail shape, colour and soft tissues remain artistic choices because no skull or feather impression belongs to Albertonykus.
Evidence and reconstruction
| Level | What belongs here |
|---|---|
| Direct | Arm and leg bones from at least two individuals, including a large functional hand claw |
| Supported inference | A tiny active terrestrial alvarezsaurid able to disrupt resistant material |
| Disputed | Specialisation on wood-dwelling termites and exact body size |
| Unknown | Skull, vertebral proportions, direct diet, feather pattern, colour and behaviour |
Frequently asked questions
Why were Albertonykus bones found with Albertosaurus?
They entered the same bonebed, which is dominated by tyrannosaurs. Shared burial does not prove hunting, close association in life or identical time of death.
Did Albertonykus definitely eat termites?
No. Its force-producing arm and termite-like tunnels in fossil wood from the same formation support the idea, but no gut content or claw-marked wood links that food directly to the dinosaur.
Could Albertonykus dig burrows?
Its arms produced considerable force but were too short for extensive whole-body burrowing. Short strokes against bark, rotten wood or an insect shelter are more plausible.
How large was Albertonykus?
Estimates range broadly from about 0.7 to 1.1 metres. The skull, spine and complete skeleton are missing, so the result depends on scaling limb bones against relatives.

