A young salmon, not much longer than your hand, leaves its river for the first time and heads into the North Atlantic. It’s a harsh place—full of strong currents and predators. Most of them won’t make it back.

Scientists have long suspected that these early weeks at sea are when Atlantic salmon populations live or die. The numbers bear this out. Across Europe and North America, salmon have been vanishing for decades, and the evidence points to something happening in the ocean. But tracking a 13-centimeter fish across thousands of kilometers of open water? Nearly impossible.

So researchers turned to simulation instead.

A new study modeled the migration of young salmon—called post-smolts during their first year at sea—from 21 rivers in Scotland, Northern Ireland, and Ireland over 27 years. The model tracked virtual fish as they swam through reconstructed ocean conditions from 1993 to 2019, guided by three navigational cues: local currents, gradients in water salinity, and an internal magnetic compass.

The results revealed something striking. Migration success varied wildly from year to year, and the difference hinged on specific ocean conditions. Years with lower salinity near the coast, stronger northwestward currents on the continental shelf, and a robust current along the shelf edge produced the highest success rates. In those years, simulated salmon reached their feeding grounds in the Norwegian Sea quickly and efficiently. In others, many never made it at all.

A Model Built on Educated Guesses

The researchers couldn't simply follow real salmon. Tagging studies exist, but they capture only fragments of the journey—a detection here, a recapture there. To fill the gaps, the team constructed a behavioral model based on the best available evidence about how salmon might navigate.

The model assumed young salmon follow three cues. First, they swim with or against ocean currents, a behavior observed in coastal waters. Second, they follow gradients in salinity—the boundary between fresh and saltwater. Salmon have an exquisite sense of smell; they might detect subtle chemical differences in the water and use them as signposts. Third, once they reach deeper water near the continental shelf edge, they add a directional component, swimming on a consistent magnetic bearing toward the northeast.

This switch in behavior—triggered by depth—mirrors how many migratory animals use different strategies at different stages of a journey. Think of it as changing from local street navigation to highway driving.

The model doesn't claim to represent every stroke of a salmon's tail. It works at a larger scale, integrating many small decisions into an overall trajectory. But when the researchers compared their simulations to real observations of post-smolts captured near the shelf edge in 2008, the fit was good. The virtual fish were in the right place at the right time.

Geography Is Destiny

Not all rivers are created equal. The model showed that salmon from rivers closest to the continental shelf edge had the highest migration success. Those from rivers farther inland—particularly along Scotland's west coast and southwestern Ireland—struggled. Many turned south into the Irish Sea or Celtic Sea instead of heading northwest. Distance matters.

This aligns with recent tracking studies. In one, researchers found that the minimum distance a young salmon had to travel strongly predicted its chance of survival. The farther the journey to deep water, the greater the risk.

The model also revealed differences in how long fish spent in shallow versus deep water. In some years, nearly all simulated salmon spent more than 90 percent of their time on the continental shelf. In others, that proportion dropped. Since prey availability, temperature, and predation risk differ between shelf and open ocean, these shifts could have profound effects on growth and survival.

The Ocean Sets the Terms

The simulations were driven by a high-resolution ocean model that reconstructed conditions across the Scottish shelf and shelf edge from 1993 to 2019. Salinity, currents, and temperature varied from year to year and even week to week within the migration season.

On the continental shelf, surface salinity fluctuated significantly. Freshwater pulses from rivers, driven by rainfall and wind, collide with saltier Atlantic water pushing in from the west. The result is a constantly shifting chemical landscape. At the shelf edge, salinity was more stable, but currents were not. The shelf-edge current—a ribbon of water flowing parallel to the drop-off—strengthened and weakened in response to large-scale ocean circulation patterns.

The model revealed that stronger currents flowing northwest on the shelf helped salmon migrate more successfully. So did a stronger shelf-edge current. Both reduced the time spent in shallow water and increased migration speed. Lower salinity on the shelf also correlated with faster migrations, possibly because it created sharper gradients for salmon to follow.

These local conditions connect to basin-scale climate patterns. The North Atlantic Oscillation—a measure of atmospheric pressure differences between Iceland and the Azores—influences wind patterns and ocean circulation. A negative NAO phase tends to produce conditions favorable for salmon migration: stronger shelf currents and lower coastal salinity. Similarly, the strength of the subpolar gyre, a massive circular current system in the North Atlantic, affects water properties at the shelf edge.

Interestingly, the same climate conditions that favor migration also appear to boost prey availability. A stronger subpolar gyre brings nutrient-rich water into the northeast Atlantic, supporting blooms of copepods—tiny crustaceans that are a critical food source for the fish that young salmon eat. This suggests that good physical migration conditions and good feeding conditions may occur together, though the model didn't simulate feeding or mortality.

The Invisible Gauntlet

Migration speed ranged from about 23 to 33 kilometers per day across years. That's fast—equivalent to a marathon runner maintaining pace for weeks on end. The fastest migrations occurred in 2012. The slowest in 2001.

But speed isn't everything. The proportion of simulated salmon that successfully exited the model domain and headed toward the Norwegian Sea within 100 days varied from just 15 percent in 2017 to 73 percent in 2008. This metric—called the domain exit rate—captures how "easy" or "difficult" migration was in a given year.

There was a weak downward trend over the study period. Migration success, by this measure, has been getting slightly harder. Whether this reflects reality is uncertain. The model doesn't include mortality from predators, disease, or starvation. Real salmon likely adapt their behavior in ways the model doesn't capture. But the pattern is suggestive.

The researchers compared domain exit rates to actual return rates of adult salmon from nearby rivers. Return rates are a rough proxy for marine survival. There was a weak correlation. Both have declined over recent decades. This hints that difficulty reaching the feeding grounds might be one factor—among many—driving population declines.

What We Still Don't Know

The study has limitations. Real salmon migration paths remain mostly unknown. The model's navigational rules are educated guesses, not proven mechanisms. The spatial and temporal resolution of the model is coarser than the decisions salmon make moment to moment. And the model stops after 100 days—long before salmon return as adults.

There's also no mortality. A high domain exit rate doesn't mean high survival, just that conditions allowed virtual fish to reach the Norwegian Sea efficiently. Whether that translates to more food, fewer predators, or better growth is unknown.

Still, the work represents a leap forward. By simulating migrations over nearly three decades, the researchers could explore how interannual variability in ocean conditions might shape salmon populations. They identified specific oceanographic features—salinity gradients, shelf currents, the shelf-edge current—that appear to matter.

These findings have practical implications. Marine spatial planning increasingly considers fish migration corridors. Knowing where salmon are likely to be, and when, can inform decisions about offshore wind farms, fishing regulations, and aquaculture sites.

A Future Written in Water

Climate change is altering North Atlantic ocean circulation. The subpolar gyre has weakened in recent decades. The NAO has shown long-term shifts. Freshwater input from rivers is changing as rainfall patterns shift. None of these trends bode well for salmon if the model's insights hold.

Young salmon don't choose the ocean they're born into. They navigate the conditions they're given, following ancient instincts encoded over millennia. When the ocean cooperates—when currents flow in the right direction, when salinity gradients are sharp, when the shelf-edge highway is fast—they thrive. When it doesn't, they struggle.

Understanding that dynamic won't save salmon on its own. But it's a start. It reveals leverage points where management might help, and it highlights the invisible forces shaping one of the Atlantic's most iconic migrations.

The ocean, it turns out, is not a passive backdrop. It sets the terms. And salmon, like all migrants, must live with them.

Credit & Disclaimer: This article is a popular science summary written to make peer-reviewed research accessible to a broad audience. All scientific facts, findings, and conclusions presented here are drawn directly and accurately from the original research paper. Readers are strongly encouraged to consult the full research article for complete data, methodologies, and scientific detail. The article can be accessed through https://doi.org/10.1093/icesjms/fsae185