The cold that long protected Quebec's northern lakes is fading.
For generations, the cold of Quebec's northern lakes served as an invisible wall against one of the continent's most tenacious aquatic invaders. Now, as climate change dismantles that thermal barrier, researchers have produced the first province-wide forecast of where Eurasian watermilfoil is likely to spread next — a map born not from distant data, but from the weed's own behavior at the edge of its range. The study, led by biologist Grace Fedirchuk at the University of Quebec at Rimouski, offers something rare in environmental science: the chance to act before the damage is done.
- Eurasian watermilfoil spreads through a single broken fragment — clinging to boat propellers, surviving overland journeys, and rooting itself in new lakes with quiet, relentless efficiency.
- Climate warming is eroding the cold-water shield that once kept the weed confined to southern Quebec, and the window for natural protection is closing faster than expected.
- A predictive model trained specifically on Quebec data — rather than continental or European records — proved dramatically more accurate, revealing that invasive species behave differently at the frontier of their range.
- Water temperature and human population density around shorelines emerged as the two most powerful drivers, pointing to western Quebec and the St. Lawrence corridor as the highest-risk zones.
- Under worst-case warming scenarios, the weed's viable habitat could expand nearly fivefold by 2100, but the forecast map now gives managers the rare opportunity to prevent invasions rather than simply respond to them.
A plant that spreads by falling apart is quietly advancing toward Quebec's northern lakes. Eurasian watermilfoil — a submerged weed that has spent decades colonizing southern waters — needs only a single fragment to establish itself somewhere new. Those fragments hitchhike on boat propellers and trailers, survive the journey overland, and take root in fresh water miles from their origin. Once settled, the plant forms dense mats that block sunlight, crowd out native species, and deplete oxygen as they decompose. For years, the cold held it back. That protection is eroding.
Grace Fedirchuk, a biologist at the University of Quebec at Rimouski, led a team that set out to map where the weed would go as the climate warmed. They built a predictive model drawing on thousands of presence and absence records, combined with local data on temperature, rainfall, lake size, and human population — then tested it across roughly 12,000 watershed areas under six different warming and population scenarios through 2100.
One finding reshaped the methodology itself: a model trained exclusively on Quebec data outperformed both continental and European versions, despite those having far larger datasets. Species at the edge of their range behave differently than those long established elsewhere. The local edge had to teach the model.
Two forces dominated the forecast: water temperature — how hot lakes get and how long they stay warm — and population density along shorelines, which determines how much boat traffic moves between lakes. These are the weed's two engines. Under the warmest scenarios, its viable habitat across Quebec could expand nearly fivefold, clustering in western Quebec and along the St. Lawrence River, where rising heat and dense human activity converge.
Until now, no province-wide picture of this threat existed. That has changed. Managers can use the forecast map to intervene before the weed arrives — a shift from reactive removal to proactive protection. The cold that long guarded Quebec's northern lakes is fading, but the province can now see, with unusual clarity, where the invasion is most likely to come next.
A plant that cannot be killed by pulling apart is quietly remaking the lakes of Quebec. Every fragment of Eurasian watermilfoil—a submerged weed that has spent decades colonizing southern Quebec's waters—can drift, settle, and become a new plant. The fragments cling to boat propellers and trailers, survive the journey overland, and root themselves in fresh water miles away. Once established, the plant forms dense mats just below the surface that block sunlight, starve native species, and leach oxygen from the water as they decompose. For years, one thing held this invader in check: the cold. Northern Quebec's lakes froze hard and stayed chilly through summer, temperatures too low for the weed to gain purchase. That barrier is collapsing.
Climate change is warming the north faster than anywhere else. Lakes that once remained cold well into summer are now warming earlier and staying warm longer. A research team led by Grace Fedirchuk, a biologist at the University of Quebec at Rimouski, set out to map where the plant would go next. They built a predictive model fed with thousands of records showing where the weed had and hadn't appeared, paired with local data on temperature, rainfall, lake size, and human population. The team tested the model against roughly 12,000 watershed areas across Quebec, running six different scenarios that combined various warming and population projections through the year 2100.
What emerged was unexpected. When the researchers compared three versions of their model—one built on Quebec data alone, one on the North American range, and one on the plant's native European range—the local version won decisively. The Quebec-only model predicted far more accurately than the continental or European versions, despite those having vastly larger datasets. A species expanding into new territory does not behave the same way it does in regions where it has been established for decades. The edge had to teach the model.
Two factors dominated the forecast. Water temperature—how hot a lake gets and how long it stays warm—emerged as the strongest signal. Rainfall and water retention came second. But one human factor stood out as well: population density around the shoreline. Crowded shores mean more docks, more boat launches, more traffic between lakes. This is the weed's highway. A broken stem is all it needs.
Under the warmest scenarios, the area where watermilfoil could take hold expands nearly fivefold across Quebec's currently protected waters. The expansion clusters in two regions: western Quebec and a long stretch along the St. Lawrence River. These are not random locations. They are where rising heat and dense human activity overlap—exactly the conditions the model identified as most likely to invite invasion.
Until this study, no one had a province-wide picture of where the weed would advance as the climate warmed. Now there is one, and it changes what managers can do. Prevention has always been the cheapest way to fight an aquatic invader. A sharper map makes it possible. Instead of chasing the plant after it arrives, managers can watch the forecast and try to stop it before it takes root. The cold that long protected Quebec's northern lakes is fading. The province can now see exactly where the weed is most likely to spread next.
Notable Quotes
A species expanding into new territory does not necessarily behave the same way it does in regions where it has been established for decades.— Study findings on predictive modeling
The Hearth Conversation Another angle on the story
Why does this plant matter so much? It's just a weed.
Because once it arrives, it's nearly impossible to remove. Every fragment becomes a new plant. It forms mats that block light and kill everything beneath them. And it spreads through boat traffic—one propeller can carry it hundreds of kilometers.
So the cold was doing the work for free.
Exactly. Northern lakes stayed frozen long enough and cool enough that the plant couldn't establish itself. That was the only thing holding it back.
What surprised you most about the research?
That local data beat continental data so decisively. The plant behaves differently at the edge of its range than it does where it's been established for decades. You have to study it where it's actually invading.
Where is it most likely to go?
Western Quebec and along the St. Lawrence. Those are the places where warming and human activity—boat traffic, docks, launches—overlap. The weed doesn't spread on its own. We spread it.
Can they stop it?
Prevention is possible now. Before, managers were always reacting after the plant arrived. Now they can watch the forecast and try to keep it out before it takes root. That's the real value of this study.