Île Amsterdam sits roughly 3,400 kilometres southwest of Perth and 3,200 kilometres east of South Africa. The French research station on its northern coast has been continuously occupied since 1949. The island's only endemic seabird, the Amsterdam albatross, breeds on a basalt plateau called the Plateau des Tourbières at 500 metres of elevation.
The 2025-2026 breeding census, completed by a four-person field team in late February, counted 198 active nests. The figure is the highest since regular monitoring began in 1983.
Dr. Charles Marais, who has worked the Amsterdam albatross programme for the Centre d'Études Biologiques de Chizé since 2011, called the figure cautiously good news. The species spent the last forty years balanced on a margin so thin that good news has been rare and qualified.
Diomedea amsterdamensis is the only albatross species that breeds exclusively on a single island. It is one of the rarest large seabirds in the world. Adult mortality, of which there is normally little in albatross populations, has been the recurring crisis.
The cause was identified definitively in the 2000s. The Plateau des Tourbières harbours endemic strains of Pasteurella multocida, the bacterium responsible for avian cholera, and the related Erysipelothrix rhusiopathiae. The pathogens cycle through a population of yellow-nosed albatrosses and rockhopper penguins that also breed on the island.
Amsterdam albatross chicks, with no evolved immunity to either pathogen, died at rates exceeding fifty percent in some seasons during the 1990s and early 2000s.
The intervention, developed across multiple research cycles and refined since 2016, has three components. The first is targeted vaccination of chicks at roughly six weeks of age. The second is removal of recently dead chicks and adults from the colony to reduce pathogen load. The third is exhaustive biosecurity protocols for the field team itself.
None of the three is individually sufficient. Together, since 2020, they have raised chick survival above 75 percent in most years.
Marais walks the plateau on most days the weather permits. The walk takes roughly an hour and a half in good conditions and is sometimes impossible for a week at a stretch in the southern winter. The team rotates in six-month deployments from the station at Martin-de-Viviès.
Each active nest is mapped, photographed, and visited at intervals of seven to ten days. Each chick is fitted, at twelve weeks, with a metal ring and a coded plastic band. The bands persist for the bird's lifetime, which in Amsterdam albatrosses can exceed forty years.
There are individuals on the plateau today, the team knows from band recoveries, who were ringed by Pierre Jouventin or Henri Weimerskirch in the late 1980s. They are now grandparents. Their grandchildren are being ringed by Marais's team this season.
The recovery is fragile in ways that the headline number obscures. The 198 active nests this season produced, by the team's count at the end of incubation, 172 hatched chicks. By the time of fledging, that number will fall further. The pathogens are reduced, not eliminated.
Climate is the other factor. Amsterdam albatrosses forage in subantarctic waters whose productivity has shifted noticeably in the last fifteen years. Adult body condition at nest relief has declined. The proportion of breeding pairs that skip a year between breeding attempts has risen.
Marais does not yet have publishable data on the long-term implications. He has noted, in his field journals, that the rhythm of the colony feels different from the rhythm described in the 1990s monographs.
The vaccination programme requires roughly six thousand euros of materials and four months of two researchers' time per season. The French Polar Institute funds the work, and the funding has held through the recent budget cycles, but it is not indefinite.
If the budget were to lapse, the modelled trajectory of the population returns to decline within five years. The pathogens have not gone anywhere. They are simply being suppressed.
What 198 active nests buys, in the same way that the kakapo's 247 buys time, is space to address the long structural problems of climate and food supply. It does not solve them.
Marais finished the day's transect at 4:50 in the afternoon and walked back to the station against a thirty-knot westerly. The plateau behind him held, on his most recent count, the largest concentration of Amsterdam albatrosses any human has documented.
Whether the figure will hold for the 2026-2027 season depends on the weather of the southern winter, the productivity of the foraging grounds, and the unromantic continuity of the vaccination schedule.
Marais is provisionally optimistic. The qualifier is doing as much work as the word.
