Photograph by Rona Bar & Ofek Avshalom / Connected Archives
words by zack metcalfe
The day began with a flock of gulls, approaching from the Pacific like a whisp of white smoke. Two dozen, flying straight for a patch of temperate rainforest near Hyder, Alaska, immediately west of the Canada-United States border. Once overhead, they circled before descending silently to a shallow creek among the hemlocks.
Inside this creek were salmon—thousands of them—jockeying to hold their place in the stream, or else stacked like driftwood on either bank. The living outnumbered the dead, for now. Down and down came the gulls. And then: bedlam.
The flock landed and shrieked among thrashing fish, pecking the water for a chaotic 20 minutes. The birds weren’t after the live salmon, nor even the dead. They were after the eggs, embedded by salmon in the gravel of the creek days or weeks before. An unspoken cue, and they all took flight. Again, silence.
It was 6 a.m. on September 1 and the salmon of Fish Creek, Alaska, had finished spawning. Now they were waiting to die, swimming in place until their fat reserves failed. Chum salmon, the first to spawn, made up the majority of the dead while pink salmon, small and humpbacked, still had time on their clocks. The patient observer could watch one or two fall out of formation every hour, drifting downstream like autumn leaves.
Black bears, grizzlies, and wolves would soon take their turns. Some at night, some in broad daylight, scavenging the dead or plucking live salmon from the stream. Eagles waited in the high branches of ancient hemlocks, and the hemlocks themselves waited for the nitrogen and phosphorus of the decaying masses. Death, and life. A seasonal redistribution of flesh.
Next to the incredible scene stood an interpretive panel covered in moss. “Imagine visiting Fish Creek 2,000 years ago,” it reads. “The spot you now stand was buried by glacial ice hundreds of feet thick.”
The northern Coastal Mountains, riding the borders of Alaska, British Columbia, and the Yukon, contain an incredible amount of ice. The range’s glaciers cover nearly 4,000 square miles and have rarely held still. These glaciers have expanded and receded several times throughout the Holocene, snaking their ways down mountain valleys before falling back again, surrendering ground to spruce and hemlock, chum and pink, grizzly and wolf, eagle and gull.
This phenomenon of ecosystem formation in the wake of glacial retreat has been studied in real time. The ice of Glacier Bay National Park, on the far end of the Alaskan panhandle from Fish Creek, has been retreating steadily since the 1700s, exposing about 100 new freshwater streams in the last 220 years. One of these, Wolf Point Creek, wasn’t exposed until the 1940s.
Alexander “Sandy” Milner, a professor of river ecosystems with Birmingham University, launched a decades-long study of Wolf Point Creek in 1977, watching as blue-green algae, horsetail, lichen, liverwort, forb, mountain avens, willow, and cottonwood slowly colonized its banks, with Sitka spruce and western hemlock not far off. Species of invertebrates came in booms and busts. The first fish, a Dolly Varden, showed up in 1987.
“When we started studying Wolf Point Creek,” said Milner, “there were no salmon.”
“Habitat is always being created and destroyed.”
And then bang. This little creek, which spent centuries under ice, attracted its first school of pinks in 1989, maybe 200 strong. By 2005, there were more than 15,000 of them. Coho salmon show up the same year, chum in 1996, and sockeye salmon in the early 2000s. The hundred or so other streams emerging in Glacier Bay received the same treatment.
“Virtually all have been colonized by salmon,” Milner said.
These findings inspired another line of research in the broader Pacific Northwest that was conducted, among others, by aquatic ecologist Jonathan Moore of Simon Fraser University. The question on the table was a simple one: What happens when all these glaciers melt in a warming world?
“Habitat is always being created and destroyed,” said Moore.
Landslides, floods, and encroaching and retreating glaciers have often rewritten the region’s freshwater streams, so salmon hedge their bets. While most return to their stream of birth to spawn, a few will explore, choosing new streams entirely and, hopefully, establishing new colonies in a behavior called “straying.” The idea is not to put all your eggs in one watershed. This is why Wolf Point Creek was so rapidly colonized, and why a warming climate means it’ll happen a whole lot more.
Moore and his colleagues project, with moderate warming trends, that the glaciers of the Pacific Northwest by 2100 will yield more than 3,800 miles of new freshwater habitat, nearly 1,200 of which will be ideal for spawning. Most of that water will emerge in the northern Coastal Mountains.
“If you turn the dial forward a couple decades,” said Moore, “these are the places where salmon might be flourishing, where we might actually be seeing gains for salmon, offsetting some of their losses farther south.”
Fish Creek was, indeed, buried in ice 2,000 years ago. Specifically by Salmon Glacier, a behemoth originating in the mountains of British Columbia and spilling—at that time—about six miles over the border into Alaska. It was probably much the same size during the Little Ice Age, as recently as the 1700s. By 1929, a U.S. Geological Survey bulletin described it as “almost wholly in Canada, with its tip just crossing the international boundary.”
Today, Salmon Glacier is Canadian: 12.5 miles upstream from Fish Creek and about six miles from the Alaskan border. The valleys it used to smother now support the Salmon River, of which Fish Creek is a tributary. Fish, grizzlies, gulls, wolves, eagles, moose, grouse … and mines.
Salmon aren’t the only ones following glacial retreat. The northern Coastal Mountains happen to contain the Golden Triangle, an exceptionally good place to find gold and copper. As glaciers melt, new ground is exposed and mining claims promptly staked. Some claims are staked before the ice even melts. Of the future salmon habitat predicted by 2100 among these mountains, 17% already have a claim within three miles.
“The environmental assessment process isn’t really considering the future conditions of these watersheds,” said freshwater ecologist Chris Sergeant with the University of Washington. “They really just consider what it looks like right now.”
And what it looks like now is a blanket of ice or a muddy, lifeless river stained whitish green by glacial flour. By diverting, acidifying, or contaminating water with arsenic and copper—or by establishing tailings ponds directly on top of future spawning grounds—today’s mines might spoil tomorrow’s freshwater refugia.
“It would be tragic if we destroyed these places before understanding what they could become,” Moore said.
Glaciologist Jean-Baptiste Bosson, based in France, has watched several glaciers retreat in the European Alps, leaving behind moonscapes of bare rock and mineral soil. He’s also watched them rapidly colonized by alpine lakes, waterfowl, and forests.
“I was studying dying glaciers,” said Bosson, “and at the same time the birth of new, pristine ecosystems.”
It occurred to him the planet’s 210,000 glaciers could, eventually, provide global biodiversity a place to retreat in a warming world—or, just as easily, host an open-pit mine, a new ski resort, or oil and gas exploration, all known to track glacial retreat. In partnership with an international team of researchers, most notably hydroecologist Sophie Cauvy-Fraunié, studying the glaciers of the South American Andes, Bosson did the math.
The resulting paper, “Future Emergence of New Ecosystems Caused by Glacial Retreat,” was published in Nature in 2023 and predicted that by 2100, 57,000-130,000 square miles of land and freshwater—areas roughly the size of Nepal and Finland, respectively—would be liberated from ice, depending on emission scenarios. The Pacific Northwest, Alps, Andes, Himalayas, and peripheral glaciers of Greenland and Antarctica loom large in these calculations.
“We need to limit glacial retreat,” said Bosson, “and to do that, we need to face climate change. At the same time, if we protect entire glaciated landscapes, we can proactively protect the ecosystems that are going to form when the ice disappears.”
To this end, he launched the Ice & Life Project, an alliance of researchers and scientists advocating for the protection of glaciated landscapes before industry can stake its claims. The group so far has secured glacial protection commitments from France’s federal government and seeks the same from other Alpine nations in Europe, specifically Austria and Italy. The ultimate goal, though, is an international treaty to mandate this kind of glacial protection everywhere at once.
“It would be tragic if we destroyed these places before understanding what they could become.”
There is precedent, said Bosson. Glaciers cover about 257,000 square miles globally, and a full 17% are already protected by the 1959 Antarctic Treaty. That agreement was originally drafted to prevent the militarization of Earth’s only uninhabited continent; but now goes so far as to prevent mining and oil and gas exploration south of 60 degrees latitude, glaciated or not.
Bosson envisions a treaty of similar scale, protecting the present extent of the world’s glaciers and, critically, the lands beneath. Such protections would, in theory, be cheap and painless for governments to implement, since few industries presently depend on glaciated landscapes. But the international conservation community has so far been too focused on modern biodiversity loss to spend political capital on the lands beneath ice.
Bosson continues to make inroads, however, and was even able to help establish a small network of protected glaciers in the Haute-Savoie region of eastern France, the 12-square-mile Arrêté de Protection des Habitats Naturels du Mont-Blanc, a demonstration project of sorts, achieved with the support of neighboring villages.
“I was dreaming of an international treaty in 2025,” said Bosson, “but that wasn’t possible. Still, we do have some good, small victories.”
The Tulsequah Glacier in British Columbia’s extreme northwest has receded 2.5 miles in 20 years, Sergeant, the ecologist from UW, said. Some of this recession simply lengthened the river, Taaltsux̱éi Héen, beneath—but more recent melt exposed a huge freshwater lake. Just last year, Jon Moore’s doctoral student Brittany Milner (no relation to Andrew Milner) caught the first fish known to swim therein: a small Dolly Varden.
The Taku River Tlingit First Nation hasn’t named this lake yet, but they have included it in “T’akú Tlatsini,” an Indigenous Protected and Conserved Area announced publicly in January 2024. The IPCA doesn’t disqualify mining, per se, but it does bring specified lands and waters (and glaciers) under the nation’s protection, subject to its laws. Melting glaciers are namechecked in the IPCA’s declaration document, as are fish.
“Our salmon and wildlife will have the space and time to move and adapt to climate change,” it reads.
The Gitanyow Hereditary Chiefs have done something similar with Meziadin Lake in British Columbia, 31 miles east of Fish Creek. While many of this lake’s tributaries are famous sockeye spawning grounds, one—Strohn Creek—was always too cold for salmon. It was kept that way by Bear Glacier to the west, which supplied a small trickle of frigid water until the 1960s. But the trickle stopped when the glacier receded, and Strohn Creek warmed. Now it’s the most important sockeye spawning habitat on the Meziadin. The Gitanyow Hereditary Chiefs in 2023 wrapped it in the Meziadin Indigenous Protected Area.
The extent to which these IPCAs will be recognized by the province of British Columbia remains to be seen, but they’ve been enough to pause new mining claims in portions of the northern Coastal Mountains and to legitimize the conservation of habitats emerging from ice. Small victories.
“Ice is, by definition, a mineral, and a critical mineral. We forget that if we didn’t have the ability to store water as ice, to reflect sunlight back into space and release cold water in late summer, life in many of these areas would not be the way it is now.”
Salmon Glacier enjoys no such IPCA, at least not yet. Even after so much melt, it remains a titan: the fifth-largest glacier in Canada and the largest accessible by road in North America. The best view is from the west face of a neighboring mountain, where Salmon Glacier’s 12.5-mile length appears to flow directly toward you before splitting north and south down two adjacent valleys like the forked tongue of a great serpent. It’s the kind of place you want to visit now, while it’s still sublime.
Brian Menounos, professor of geography with the University of Northern British Columbia and former Canada research chair for Glacier Change, said aggressive carbon reductions globally over the next couple of decades could save half the ice in the northern Coastal Mountains, almost all of which is retained in the largest few glaciers. But even in this relatively optimistic scenario, much of Salmon Glacier will go the way of Fish Creek. Pink salmon will almost certainly spawn down there someday, offering cursory resistance to the grizzlies tugging off their heads.
“Ice is, by definition, a mineral, and a critical mineral,” said Menounos. “We forget that if we didn’t have the ability to store water as ice, to reflect sunlight back into space and release cold water in late summer, life in many of these areas would not be the way it is now.”
The transformation will be messy and painful. New habitats will form, sure, but stability will initially be hard to come by. Spawning grounds will emerge with melt only to vanish with flood. Waters will flow high and cool with a glacier at their heads, but unpredictable and warm should that glacier disappear. Rarely appreciated are the species of microbe and invertebrate living in or under the ice that will have nowhere to turn if that ice vanishes outright. The death of glaciers might produce some beauty, some respite, some consolation; but in many ways, it will cost the landscape dearly.
Join our membership community. Support our work, receive a complimentary subscription to Atmos Magazine, and more.
Protecting Ecosystems of the Future Before They Exist
