Weight is what decides whether a young puffin lives through its first winter at sea. A chick that leaves Heimaey at three hundred and fifty grams is roughly five times more likely to survive than one that leaves at two hundred and fifty. That single fact is why, on August nights in the Westman Islands, there are children out after midnight with torches and cardboard boxes.
A puffin chick leaves its burrow after dark, alone, and sets off for water it has never seen. It finds its way by the moon and by the brighter horizon that stands over open sea. Street lighting drowns both of those signals, and an onshore breeze does the rest, so on Heimaey the fledglings turn inland and come down in gardens, in car parks, on the harbour ground and under parked cars.
Collecting them is old enough that nobody can date it, and the arrival of electric light on the island in the early 1900s is the obvious trigger. The organised version has a date. The pysjueftirlitið, the puffling patrol, was set up in 2003 and is run by the Vestmannaeyjar Research Centre. Every bird picked up off the street is recorded, and the whole town works it, the Vestmannaeyjar police included.
Birds spend the night in the box and are handed in the next day. Each one is weighed and has its wing measured, and about a quarter are fitted with a stainless steel ring, something over four hundred birds in a normal season. The finds are logged at lundi.is. Chicks have been ringed at fledging on these islands since 1959, which is long enough that the scheme has produced birds over forty years old.
Where a bird is released matters as much as whether it is found. Hamarinn, the sea cliff above the town, is the main site, along with Höfðavík, the cliff edges out at Stórhöfði and the golf course. Birds are launched from ground high enough to let them get flying before they reach the water. The harbour is off limits, because of the oil that sits on it.
The scale of the work follows from what surrounds the town. The last full census of the archipelago, flown in 2010, counted about 830,000 breeding pairs across the islands, somewhere near two fifths of all the puffins in Iceland. No full count has been made since, so every current figure for the colony is a sixteen year old number carried forward.
The archipelago holds the largest gathering of the species on earth, and it holds that title by default. Røst, in Norway's Lofoten islands, had well over a million pairs in the late 1970s and has lost the great majority of them since. The title changed hands without Vestmannaeyjar gaining a single bird.
The Icelandic trend is downward too. The species is listed as vulnerable globally and as endangered in Europe, and Iceland's national red list has carried it as critically endangered since 2018, up from least concern at the turn of the century. Analysis published in 2023 put the Icelandic decline at around seventy per cent since 1995, well above the forty per cent that had been assumed.
Food is what drives it, and the relationship has a stranger shape than the usual account allows. Chick production across the Icelandic colonies peaks at a sea surface temperature of about 7.1 degrees, and falls by more than half for a shift of a single degree in either direction. Cold years have been bad years too, and most accounts of the species carry only the warm half of that curve.
The link is the lesser sandeel, the small fish the adults carry in to the burrow. Warm winters leave fewer of them the following summer. The clean evidence that this works through the fish rather than on the birds is where the puffins spend the winter: Icelandic birds range west across the Atlantic, some as far as the Labrador Sea, and yet the winter sea temperature off the Westman Islands still predicts how many chicks fledge the next summer.
The birds are not in Icelandic waters in winter. Whatever the winter sea is doing to them, it is doing through the fish.
The temperature at which the birds do best has itself drifted upward over the last century, and the sea has warmed more than three times faster. A puffin starts breeding at about five and a generation runs to eleven years, which rules out any prospect of them adapting across the gap.
The record that makes all of this measurable was kept by hunters. Puffins have been taken for food on these islands since the ninth century, and from 1876 by fleyging, a triangular net swung on a pole from the cliff edge. That method catches the immature birds that circle the colony; the breeders fly straight in and out with fish and are rarely taken. So the size of the catch in any year tracks how many chicks fledged a few years earlier, and the hunting books, kept since the 1880s, have been worked up into an annual index of chick production running back well over a century. It is the longest run of demographic data for any seabird in the North Atlantic.
The hunt is still running, and it is still contested. The Vestmannaeyjar season was cut hard through the 2000s, from forty five days to twenty in 2008 and to five in each of the two years after, and permission has been granted again every recent summer. In 2026 the municipality allowed hand netting from 23 July to 15 August, with no bag limit set, against a written objection from the regional nature research institute, which called any take unsustainable. Most of the ground is held by hunting clubs, so a licence on its own does not confer a right to hunt it. A national assessment in 2024 put the long term sustainable rate at no more than four or five per cent a year of the birds exposed to it.
Whether the decline has turned is genuinely unsettled among the people who study it. Chick production across Iceland was close to nil through the 2000s and into 2014, then improved sharply from 2015 and has held up. The counts in the town run rougher: the two largest fledging seasons since the record began in 2003 were 2019 and 2020, at roughly 7,700 and 7,650 birds, and none of the seasons since has reached five thousand. Against that, the standing population is still judged to be falling, and the local research institute has opposed hunting on those grounds in each of the last two years.
Both of those can be true at once, because the arithmetic is slow. A good chick year does not become breeding adults for five years or more. The birds weighed on kitchen scales this August will not be counted as breeders until the 2030s.
For a visitor the season splits in two: the adults through the summer, the fledglings at the end of it. Puffins are ashore at the burrows from roughly the middle of May, are at their most active through June and July while they are running food to chicks, and begin to thin out in the second half of August. By September the slopes are usually empty of adults. The pysjur season then runs from late July into early September, and its start swings by about three weeks with the food supply: the first chick was found on 27 July in 2024, on 29 July in 2025 and on 8 August in 2026.
The hour of the day matters more than the month, and it is where day trips go wrong. Adults are out fishing through the middle of the day, and the slope can look deserted at two in the afternoon in high season. Early morning until late morning, and from early evening until dusk, are when the burrows are busy. The standard timetable, out on the first ferry and back in the afternoon, lands a walking visitor at the colony in exactly the dead hours. An overnight, or a late return sailing, is what buys the evening.
Stórhöfði, the southern headland, is the site everyone is sent to. It is about six to seven kilometres from town, an hour to ninety minutes on foot with the climb at the end, so a round trip on foot costs most of a half day before any bird watching. An hourly hop-on hop-off bus runs a circuit through the middle of the day and will leave a passenger at the lookout for a full hour, and an electric bike from the rental in town solves the distance more cheaply. The hut at the top is genuinely useful when the wind is up, and quiet walking on the slope outside gets people closer. Herjólfsdalur, the crater valley much nearer town, is the under-reported alternative.
Wind is the thing that spoils visits here. Stórhöfði has been a weather station since 1921 and is routinely described as the windiest place in Europe, a claim worth treating as local shorthand, since no single measurement establishes it, though the gusts recorded there are real enough. Anything under about ten metres a second is comfortable walking. The forecast at vedur.is is worth reading before committing to the walk out.
Herjólfur crosses from Landeyjahöfn in about thirty five to forty five minutes, up to seven times a day in summer. When swell or sand closes that harbour the sailing moves to Þorlákshöfn, which is a crossing of nearly three hours from a port an hour and a half further away. The change is not always flagged as a change on a booking confirmation, so read the port name on it the day before travelling and leave a phone number. Car spaces sell out weeks ahead in July and August; foot passenger spaces almost never do.
There is no longer a summer flight to fall back on. The airline that used to fly the route from Reykjavík went bankrupt in 2024, and the service now runs only in winter, roughly December to February, five times a week. In puffin season the ferry is the only scheduled way on and off, which is a reason to keep an evening free in case a sailing is cancelled.
Anyone hoping to join a patrol should know that the old Sæheimar aquarium has closed and the puffin rescue centre is now run by the Sea Life Trust at its sanctuary on Ægisgata, in the same bay that once held Keiko and now holds two beluga whales. Underweight and oiled birds go there to be brought into condition. Joining in on the street happens by accident and cannot be planned; a local guide runs the only organised night outing, leaving late in the evening and releasing the birds from the cliff the following morning. A night out usually turns up one or two chicks.
Þjóðhátíð, the national festival held since 1874, fills Herjólfsdalur with about sixteen thousand people and three hundred tents on the weekend before the first Monday in August, which in 2027 falls on 30 July to 2 August. There are something like a hundred thousand puffin burrows in that valley. Accommodation is impossible that weekend and even foot passenger ferry places can go.
Surtsey came out of the sea south of Heimaey in an eruption that ran from 1963 to 1967, and has been kept closed to all but a handful of researchers ever since. Puffins first bred on it in 2004, forty one years after the eruption began, which is roughly how long it took a new island to grow enough soil to dig a burrow in. Nobody working the patrol this August will see what the count comes to. Until the 2030s the only evidence anyone has is the weight, the wing length and the ring number, written down on the night the bird was found.