Baby alewife leaping at sunset, Kennebec River, Waterville, Maine. Sept. 21, 2003.

Why do adult salmon and baby alewives like to leap?

Some evolutionary theorists assert that all human and animal behavior can be explained solely in terms of a deep-seated impulse to survive long enough to successfully pass along ones genetic heritage to future generations.

However, many documented types of human and animal behavior do not appear to provide any direct or indirect benefit to survival or reproduction.

An obvious example is humans' proclivity for play; for singing, painting and sculpting; for playing and watching athletics; for writing books, poems or plays or esoteric essays like this.

River otters seem to enjoy play. They devise slides on muddy river banks and spend hours sliding down them into the water, not unlike humans skiing or sledding down a snow-covered hill.

One would think such behavior by river otters would be frowned upon by the forces of evolution, since such frolicking makes the otters highly visible to predators. Would not a superior survival strategy consist of otters remaining invisible except when eating?

Why do otters so often play -- right out in the open?

Some people find offensive the suggestion that the play of river otters may be driven by the same impulse as human play -- to have fun.

They say, "Surely you aren't implying that otters have fun like humans."

Why not? Why should the desire to have fun be the sole province of one species of the millions on Earth? What hard evidence supports such an exclusionary view?

Clearly, river otters have work to do each day. They must eat and avoid being eaten. At the proper age, they must find mates and reproduce; and they must nurture their young so they too will survive to reproduce.

But what happens if river otters are adept enough to meet all those needs and still have a number of hours in the day leftover? In such circumstances, why is not possible that they might take some of those excess hours and have fun?

As anyone knows who has juggled multiple jobs just to put food on the table, pay the rent and clothe their children, having fun must often be postponed until the basic requirements of survival have been met. But when humans successfully meet those basic requirements and still have some spare hours left in the day ... they invariably do things "just for the fun of it."

It is unclear why such a nearly universal pattern of behavior in humans would be totally foreign to other animals with which we share a vast number of behavioral affinities.

If we accept that other animals are capable and desirous of having fun in a way similar to humans, many documented behaviors of various animals can be better explained. Conversely, if we refuse to accept that other animals are capable and desirous of having fun, many behaviors of various animals become very difficult to explain.

The play of river otters can be easily explained if we accept that, in the words of Cindy Lauper, otters just wanna have fun. Their well documented behavior is very hard to explain if we declare "fun" to be the sole province of one species on Earth, Homo sapiens.

If river otters are not having fun while repeatedly sliding down a mud bank, then what are they doing and why? They're not mating. They are not eating. They are not engaging in courtship (though they might be flirting). And they are certainly not hiding from predators. Why has this risky behavior, which offers no survival benefits yet consumes precious food energy, not been long banished from the river otter gene pool?

Many evolutionary theorists, when presented with such questions, will either ignore the question or say, "We're not sure why they do it, but it must have something to do with survival and procreation, because they surely cannot just be having fun."

Why not? Why is it so difficult to even consider the possibility that other animals may seek and enjoy fun in a way similar to humans?

When they return to our rivers from the ocean to give birth, Atlantic salmon do not eat. Spawning Atlantic salmon live solely on accumulated muscle reserves for nearly a year and only resume eating the spring following their first return to freshwater. In rivers brimming with food, 20 pound Atlantic salmon will drive themselves to the edge of starvation rather than have a bite to eat.

Over the three to five months between arriving from the ocean, migrating upstream and spawning in the late fall, Atlantic salmon quite literally have nothing to do as they count the days of summer in deep, well-oxygenated pools of their home rivers.

During this time, Atlantic salmon can often be enticed to attack and swallow the artificial lure of an angler. By rising to an artificial fly floated on or near the surface, the salmon reveal themselves and expose themselves to predators -- all to capture and put in their mouths an object they have no interest in eating.

Atlantic salmon in rivers will jump and roll and swirl on the surface of pools, for no apparent reason, even though this reveals their location to predators and makes them vulnerable to death or injury during that critical period before they mate.

Why would Atlantic salmon engage in such risky behavior that confers no nutritional or reproductive benefits, consumes precious energy, and increases their chance of being eaten or injured?

Nobody knows. But we do know Atlantic salmon have been doing it at least since the ancient Romans named them "salar," the leaper, and their descendants began the tradition of catching non-feeding Atlantic salmon on feathered imitations of insects attached to a homemade, horse-haired line.

To colleagues, I have often forwarded the hypothesis that Atlantic salmon exhibit these odd behaviors simply because they are happy to be home. I define "happy" as the emotion derived from a freedom from want, a freedom from pain, a freedom from fear, and the comfort of being in a familiar place.

And since Atlantic salmon return to their home rivers solely to mate -- not to eat -- I also add that Atlantic salmon may leap and swirl and swat at flies because they are happy they will soon have sex. Happy hour at the Kennebec River Bar & Grill, if you will.

These hypotheses are usually met with rolled eyes or outright hostility. The rolled eyes I can understand, the hostility I cannot, especially I have yet to hear a hypothesis which makes any more sense.

Some have hypothesized that non-feeding Atlantic salmon rise to flies because they have "memories" of doing so as baby salmon in the same rivers. If true, this means that Atlantic salmon have memories; much like an adult's nostalgic desire to take one more pitch at the little baseball diamond they played at for days as a child. If so, that would be a fairly profound and complex behavior for a fish.

Some have hypothesized that non-feeding Atlantic salmon leap out of the water to "practice" for serious jumping they may have to do at a challenging waterfall or rapids far up their home river. This seems counterintuitive since Atlantic salmon do not eat and therefore have a finite store of energy prior to mating. Wasting lots of precious energy jumping in a calm pool miles below a challenging falls or rapids would only deprive them of the strength they need to surmount the falls themselves. Also, many Atlantic salmon rivers lack large, challenging falls requiring prodigious leaps, yet their salmon leap just as much as salmon born on rivers with such prodigious falls.

When Atlantic salmon in a pool perform what I call "recreational leaps" they are not very high leaps, perhaps three feet at best. Prior to the damming of the Kennebec River, Atlantic salmon at Caratunk Falls in Solon were observed leaping more than 11 feet out of the water to ascend the falls. Without such an enormous leap, the nearly vertical Caratunk Falls were impassable to salmon. Clearly, if salmon in a river's lower pools were "practicing" to leap over high, challenging falls upriver, it would behoove them to perform practice jumps of more than just one or two feet. That would be like training for the Olympic high jump by repeatedly stepping over a brick.

Is leaping a mating behavior, to demonstrate prowess and strength to prospective sexual partners? Because Atlantic salmon leap most often during the spring and early summer, yet mate and spawn in the late fall, this explanation also seems unlikely. Male Atlantic salmon are well known to fight and joust among themselves for the choice of a mate -- but only at the spawning site itself in late October and early November. Moreover, salmon in the Kennebec are well known to leap out of the water in the river's lowermost reaches, 60-100 miles below their historic spawning grounds.

Our challenge today is not to find out why salmon leap and take our artificial flies, but to ensure there will be Atlantic salmon in our rivers in the future so our grandchildren have the opportunity to ponder the same question.

Since the Edwards Dam was removed from the Kennebec River's head of tide in 1999, we now have the extreme pleasure of watching native river herring -- alewives, American shad, and blueback herring -- return in ever increasing numbers to our river and its accessible tributaries. These herring mate in the river during the spring and early summer. During July, August and September, the river's surface from Waterville to the sea is frequently disturbed by the antics of their one to four inch young.

For five summers now we have watched and studied these baby herring in the river from their birth in June to their migration to the ocean each fall. We have often observed schools of 20 to 100 baby shad and herring actively feeding on hatching insects in the river, much like trout. Like trout, the baby herring will often leap out of the water to grab an insect hovering in the air just above the river's surface, catch it in their mouth, and fall back into the water in a neatly curved arc. Our observations suggest that baby shad and river herring in the Kennebec derive much of their food just like trout, by scanning the water's surface for newly hatched mayflies, caddis flies and midges and successfully eating them in mid-air.

We have also noticed that, especially at dusk, baby shad and river herring will congregate in large, compact groups of 1,000 animals or more along the river's banks, sometimes in eddies and sometimes in gentle currents and begin jumping en masse. Careful observation has revealed these baby herring are not jumping to catch insects on the wing, since they are usually no hatching insects present when the baby herring are jumping. If the baby herring were feeding on submerged insects, there would be no need for them to jump clear out of the water in neat, three inch arcs every few seconds. Yet they do. This mass jumping exhibition usually lasts for 30 minutes and concludes before darkness.

In September 2001, we witnessed an enormous jumping exhibition of 3-4 inch baby alewives at dusk in a large eddy of the Kennebec River at the former Edwards Dam site. Two acres of the river were covered with thousands of baby alewives steadily leaping out of the water in small, neat arcs. The collective sight and sound of this jumping was like the sight and sound of a brief, hard thunderstorm on this one spot on the river; and from a distance an observer would have been hard pressed to tell the difference.

Baby river herring have innumerable predators. Striped bass, eels, smallmouth bass, great blue heron, egrets, cormorants and even dragonfly nymphs feed on baby river herring from prior to their migration to the ocean each fall. Why would these small fish so loudly advertise their presence and location to predators by jumping en masse at dusk? Careful observation at water's edge reveal the baby herring are not eating. Angling and general observation reveal many of the baby herrings' chief predators, striped bass and cormorants, are in fact close by. So why would they do something so unnecessary and so apparently suicidal as reveal themselves to predators so boldly?

Strict evolutionary theory would hold that baby alewives which do not jump at sunset, by remaining more concealed from predators, have a better chance of surviving to adulthood than baby alewives which jump at sunset. Over time, one would expect the jumping baby alewives to disappear from the gene pool, replaced by "wiser"alewives who stay hidden as much as possible. This, however, is not the case.

So how do we explain this behavior?

My hypothesis as to why baby herring jump en masse in the river at dusk, without feeding, is because they are happy to have survived another day and they celebrate not being dead by jumping out of the water together for a short time at sunset. They know a dangerous night is ahead and they may not live to see the sunrise. So they celebrate while they still have fins to do so.

Are these hypotheses true? I have no clue. To rigorously test them, we would have to learn to communicate with otters and Atlantic salmon and baby alewives and ask them why they do what we watch them do. Until then, we can only speculate.

Even this casual speculation, when combined with careful field observation, can bring us closer to the world in which river otters, Atlantic salmon and baby herring share with us. That in itself seems to make the effort worth it.

And it is fun.

February 1, 2005.

BACK