Below the Surface: Finding the Way Home

Many of our columns deal with anadromous fish—those species that start life in freshwater, migrate to the ocean, and then return to freshwater to spawn. Amazingly, most of these fish return to the same river or connected lake they left years earlier as young fish. How do they do it? Drop bread crumbs? Pull up Google Maps? Spoiler alert: neither of those two explanations are correct.

It wasn’t always understood that fish came back. Many hundreds of years ago, people believed that the small salmon smolts that migrated downstream in the spring were a different species than the large adult salmon that came upstream from the ocean in the summer or fall. Naturalists in Scotland a few hundred years ago tied silver thread through the fins of smolts and when they discovered adults in their stream with those same threads, they concluded the fish “homed.” But they didn’t understand how.

Definitive proof came via research out of the University of Wisconsin in the 1960s. Pacific salmon such as Coho salmon were proposed to be stocked in Lake Michigan in hopes that they would eat the overly-abundant alewife, which entered the Great Lakes through the navigational canals and were causing ecological (and smelly) havoc.

It was understood that these non-native salmon would not go to sea but use Lake Michigan as their ocean. When it came time to spawn, it was theorized, they would leave the lake and enter rivers that feed into it. But if the states stocked Cohos, in which rivers would they spawn? No one knew how the salmon selected rivers.

A professor named Arthur Hasler designed an experiment in which he raised two identical groups of young Coho salmon in a hatchery on the banks of the Wisconsin River. This location is important because that river flows into the Mississippi River and the Gulf of Mexico, not Lake Michigan. The two groups of uniquely-tagged fish were held in different hatchery ponds fed by Wisconsin River water and into one pond he dripped an organic chemical called morpholine. This chemical is completely artificial and not found in nature. The other pond did not receive this chemical. When the salmon turned into smolts and were ready to migrate to the lake, they were trucked to Lake Michigan, loaded into tanks on a boat, and taken to the middle of the lake for release. Hasler’s experimental design ensured that those young salmon had never experienced a river that flows into Lake Michigan. So where would they go to spawn? Would they randomly just choose a river?

Imagine water chemical scent trails as colors. This illustration shows how the scents merge in a watershed and create an olfactory pathway to guide returning salmon. Artwork by Sally Harold.

It took a couple of years for the fish to mature in the lake and be ready to spawn. During the spring of their anticipated spawning migration, Hasler dripped morpholine into the Milwaukee River, which flows into Lake Michigan. Nearly all the morpholine-treated fish returned to the Milwaukee River. The non-morpholine treated fish returned to rivers all up and down the Wisconsin and Michigan shorelines of Lake Michigan, showing no preference for a particular river.

Have you figured out the explanation? If not, let us give you one more hint. There was research on migratory American shad in our very river in the 1970s. Adult shad were netted in the lower Connecticut River, split into two identical groups, and uniquely tagged. One group was anesthetized and had their nostrils (technically, “nares,” in fish) filled with some inert material to block them. Both groups of fish were taken quite a distance out into Long Island Sound and released.

The fish had not yet spawned so they needed to reenter the river. The commercial shad netters were asked to turn over all tagged shad to the researchers. Very few fish with their nares blocked returned to the river (they were lost), whereas a high percentage of the control fish returned.

It’s the sense of smell. In the case of the Lake Michigan study, the non-morpholine-exposed fish imprinted only to the smell of the Wisconsin River, and since there was no Wisconsin River water entering Lake Michigan, they were lost and just randomly chose rivers. But the Milwaukee River smelled like home to the fish exposed to morpholine.

Each body of water has its own unique chemical profile. This is determined by the underlying bedrock, soil types, forest composition, and likely the plants and animals that live in the watershed. In modern times, the presence of pollution can contribute to the smell signature. Young fish memorize the chemical profile (or smell) of their natal water body. We now refer to this as imprinting.

You may have heard about a famous study in the 1930s by a German biologist named Konrad Lorenz, who hatched goslings in two groups. Upon hatching, one group first saw their mother, and the other group first saw Lorenz. That latter group behaved as if Lorenz was their mother and followed him everywhere just as the other group followed their mother goose, which is typical among geese. Lorenz coined the phenomenon “imprinting.” Fish biologists adopted that term as well, since this is a learned behavior and one that could be artificially manipulated, as both Lorenz and Hasler proved.

How do fish use this to find their way home? This part of the process is called “homing.” Fish follow the scent trail home. Imagine you open the front door of your grandparents’ home, and you smell freshly baked apple pie. You’re going after it to steal a piece! You walk to the back of the house toward the kitchen, but after you pass the staircase, the smell becomes fainter. You back up and go up the stairs. It is getting stronger! You get to the top and sense the aroma is stronger to the left than the right, so you turn left. You go into your grandmother’s sewing room and nearly lose the scent. Back in the hallway you pick it up again, and when you go into your grandparents’ bedroom you see it cooling on top of the dresser where they were trying to hide it! You found it by following the scent trail, like migratory fish do when they’re homing to that river smell to which they imprinted years ago.

The Lake Michigan watershed showing key locations of Dr. Arthur Hasler’s imprinting/homing experiment, the Wisconsin River (red dot) where two groups of salmon were raised and Lake Michigan where all were released. One group of salmon had been imprinted using morpholine which was introduced into the Milwaukee River (yellow dot) to see if they would return there. Illustration by Sally Harold.

Since young Atlantic salmon use small headwater streams, let’s use them as an example. Imagine a group of salmon parr living in Utley Brook, a high-quality stream flowing out of the Green Mountains. They imprint to Utley Brook water but then migrate out to sea and feed in the Davis Straits off Greenland. When it’s time to come home, there’s no way they can smell Utley Brook water off Greenland. Some other factors, probably inherited in the DNA of the fish, bring them to southern New England. But even in Long Island Sound, they probably cannot smell Utley Brook. This is where “sequential imprinting” comes into play.

When those fish were young and in Utley Brook, they imprinted to it. When they dropped downstream and entered the West River in Londonderry, Vermont, they imprinted to that new water smell, too. Next, they entered the Connecticut River near Brattleboro. They also imprinted to that. As they descended the Connecticut, new scents were introduced as they passed other tributaries such as the Westfield and Farmington rivers. They kept imprinting. As they left freshwater and entered Long Island Sound, they were imprinted to the collective olfactory goulash of the entire Connecticut River. Upon returning to Long Island Sound as adults, it is that odor of the Connecticut River that draws them in. But they never forget the Utley Brook scent, which may only become distinguishable in northern Massachusetts. If the homing adults overshoot the West River, they lose that scent and pull a U-turn and head back downstream to then swim up the West River and subsequently Utley Brook. There is some evidence to suggest that adult Atlantic salmon may return to the same quarter mile of stream where they lived as juveniles.

While salmon have precise spawning, American shad do not. They tend to stay in the main stem Connecticut River and broadcast spawn (perhaps with 12 of their closest friends), and the eggs and larva drift in the river before the hatched shad swim around the entire lower river for the rest of the summer. They will imprint/home to the Connecticut River, but not necessarily to a specific part of the river.

But there is more to learn always. The opposite of homing is straying: returning as adults to a different stream from which they left as a juvenile. Some species (like alewife) stray more than other species (like salmon). The degree to which they stray will influence what fish they breed with, and this, in turn, can determine if unique strains of fish develop in certain streams or tributaries or if the fish are genetically homogenized and hard to genetically distinguish one from another.

Steve Gephard is a fish biologist who retired from the CTDEEP and has a long association with the Connecticut River. Sally Harold has worked on dam removal and fish passage projects for over 20 years, including as Director of River Restoration and Fish Passage for the Connecticut Chapter of The Nature Conservancy. Steve and Sally collaborate on dam removals and fishway projects through RiverWork, LLC, and can be reached at Riverworkllc@gmail.com.