On a small uninhabited island in the Cyclades called Ofidoussa, midway between Naxos and Astypalaia, several hundred falcons spend the summer doing almost nothing. They roost on the cliffs. They hunt insects in the heat of the day. They wait.
What they are waiting for is the autumn passage of songbirds across the Mediterranean. The Eleonora's falcon, Falco eleonorae, has timed its entire breeding cycle around this annual event. The chicks hatch in late August. They fledge in October. Their first meals are migrants.
This is the only falcon, and one of the only raptors of any kind, that breeds in autumn rather than spring. The phenology is so unusual that nineteenth-century ornithologists initially refused to believe the reports from the Aegean islands. The species was scientifically described in 1839 by the Italian naturalist Giuseppe Gené, who named it after Eleonora d'Arborea, a fourteenth-century Sardinian queen who had passed an early law protecting raptors from hunters.
Marius Doyle visited Ofidoussa in late August 2025 with a small Greek research team led by Christos Barboutis of the Hellenic Ornithological Society. Barboutis has been monitoring the Eleonora's falcon colonies of the Aegean for nearly twenty years.
The colony at Ofidoussa holds roughly two hundred pairs. The island is too small to support any terrestrial prey. There are no rabbits, no rodents, no nesting passerines worth pursuing. From April through July, the adult falcons feed almost entirely on aerial insects, mostly dragonflies and large beetles, caught on the wing.
Then, in the second half of August, the wind on the Aegean begins to shift. Migrating warblers, flycatchers, redstarts, and pipits, having spent the summer in central and northern Europe, begin moving south. They cross the Mediterranean in long nocturnal flights and arrive at the Greek islands at dawn, exhausted, looking for cover.
The falcons are waiting. They form loose hunting parties at first light, climbing to two or three hundred metres above the island and patrolling the airspace through which the migrants are descending. A falcon that catches a warbler will carry it back to the nest, where, increasingly through September, hungry chicks are calling.
The synchronization is precise. The peak in migrant abundance at the Cyclades, roughly the first three weeks of September, coincides almost exactly with the period in which Eleonora's falcon chicks are largest and require the most food.
The energetics work because songbirds are dense, nutritious prey. A blackcap or a willow warbler may weigh fifteen or twenty grams. A falcon chick fed on warblers grows faster than one fed on insects. The migration window thus permits the colony to raise larger broods than any insectivorous strategy would allow.
What it requires is that the falcons themselves migrate at the very tail end of the season. Adult Eleonora's falcons do not leave the Aegean breeding grounds until late October or early November, more than a month after most other Western Palearctic raptors have departed.
Their wintering grounds are also unusual. The species winters almost entirely on the island of Madagascar, more than seven thousand kilometres away. The migration route, only fully mapped with satellite tags in the early 2010s, runs south across the Red Sea, through Sudan and Ethiopia, across East Africa to Mozambique, and then on a final water crossing to northern Madagascar.
The journey takes about four weeks. The falcons travel singly or in small loose flocks. Once on Madagascar, they spread out across the island and feed, again, almost entirely on insects and small birds, through the austral summer.
In April, they reverse the route, arriving back at the Aegean colonies in May. They have a short window of relative rest before the wait begins.
What is striking, Doyle notes, is how perfectly the system depends on the migrants. If the songbird migration through the Aegean were to collapse, the Eleonora's falcon's whole life-history strategy would collapse with it. There is no obvious alternative prey.
And the songbird migration through the eastern Mediterranean has, in fact, declined sharply in recent decades. The drivers include habitat loss on European breeding grounds, hunting on Mediterranean stopover islands, and changes in wintering habitat in sub-Saharan Africa.
Eleonora's falcon counts at the major Greek colonies have so far remained relatively stable, hovering around twelve thousand pairs across the species' total range. The species is listed as Least Concern by the IUCN. Whether this stability can be maintained as the underlying prey base shrinks is an open question.
Barboutis is cautious. He notes that the falcons appear to be tracking the migration's southward shift to some degree, with some colonies in North Africa now breeding earlier than the Greek colonies. The behavioural plasticity is real, if limited.
On Doyle's last evening at Ofidoussa, the wind dropped at sunset and the falcons returned to the cliffs in twos and threes. They settled on familiar ledges. The chicks, still down-covered, peered out from nest scrapes set in shallow depressions in the rock.
In two weeks, the migration would begin. In six, the chicks would be flying. In ten, the colony would be empty, the whole assembly somewhere over the Sudanese highlands, on its way to an island most of the falcons would, on this their first trip, see for the first time. The system, for now, was still holding.
