Do Salmon Return to the Steam Where They Were Born to Spawn?

June 7, 2009 11:37 PM

Do scientists really know that salmon always return to the stream where they were born in order to spawn? Or are they just telling us a charming story?

Scientists have been telling us for at least 50 years that salmon always return to the stream where they were hatched when they are ready to spawn. Is there any documented proof of this? Because I can't think of any possible way to prove that they behave like this.

First, I don't think it's possible to put some sort of a tag on a baby salmon because they are so small. Second, because of the high mortality rate you'd have to mark hundreds of them to have a chance that some will survive to breeding size. Third, you'd have to find the marked salmon among the hundreds that come to spawn in the stream. And fourth, you'd have to be able to prove that none of the marked salmon went to a different stream.

Do anybody know where and when this myth originated?

And I would suppose that they same question would apply, sort of in reverse, to sea turtles.

posted by edavidoff to science & nature (26 answers total)

With Salmon I think that they grow up in the stream, thus the thing to do would be to tag them as they leave the stream after they have grown and the mortality rate has dropped. They have to have been born there and you can then easily see if they return to the same place. IANANS (I Am Not A Natural Scientist) But that makes sense to me.
posted by Carillon at 11:46 PM on June 7, 2009

Scientists map it by comparing DNA of different generations.
posted by KokuRyu at 11:50 PM on June 7, 2009

edavidoff: First, I don't think it's possible to put some sort of a tag on a baby salmon because they are so small. Second, because of the high mortality rate you'd have to mark hundreds of them to have a chance that some will survive to breeding size. Third, you'd have to find the marked salmon among the hundreds that come to spawn in the stream. And fourth, you'd have to be able to prove that none of the marked salmon went to a different stream.

You're wrong about the impossibility of tagging and tracking salmon. The Atlantic Salmon Federation, an NGO, does this yearly, and has for about a decade. Here's a study regarding the rate of return to natal streams of salmon after the Exxon-Valdez oil spill; it samples 288,492 fish. It clearly isn't a problem sampling large numbers of fish at at time.

All sources seem to indicate that it is indeed true that salmon return to their natal stream, although it's not known for sure why.
posted by koeselitz at 11:58 PM on June 7, 2009 [1 favorite]

Is it that hard to believe?

I found this article after a quick google... it supports the olfactory imprinting hypothesis.

As one of my earliest memories is seeing salmon swimming up stream in Scotland I'd be interested in what other people have to say.
posted by evil_esto at 12:00 AM on June 8, 2009

at least 50 years that salmon always return to the stream

Individuals don't always return to the same stream. And they certainly don't always return to the same tributary/breeding ground.

It's my understanding that populations of salmon return to the same river from which they were born. So, most of the salmon you find returning to spawn in the Quilcene are from the Quilcene and not the Umpqua. Not all of them. And not all Quilcene-spawned fish make it back to the Quilcene--some wind up in the Umpqua for whatever reasons.

Anyway, I don't know of specific studies. I do have plausible answers to your enumerated concerns, though.

1) You don't tag fish. You clip or notch their dorsal fins. If you use a characteristic notch, you can separate populations.

2) Yes, you have to mark hundreds--thousands probably. They're really easy to catch, though, as they go. If you think the number is a concern, I think you grossly underestimate the dedication of many scientists. You also grossly underestimate how many grad students volunteer for this type of field work.

3) Most salmon die soon after spawning season. I'm told that in the spawns, the dead fish will sometimes literally coat the surface of the stream. Likewise, they're trivial to catch as they move upstream to spawn. You only have to find a few marked salmon for statistical conclusions to be drawn.

4) No. You don't. Because the statement "salmon return to where they were spawned" is not an absolute. It's just that their migrations are not random--they're not just picking the closest stream when it's time.
posted by Netzapper at 12:03 AM on June 8, 2009 [4 favorites]

This paper, “Recent Advances in Biotelemetory Technology
on Salmon Homing Migration” by Hiroshi Ueda, is helpful on the finer points of how salmon tracking is done:

Three biotelemetry instruments (ultrasonic transmitter,electromyographic radiotrasmitter, and micro-datalogger) havebeen applied to investigate homing migration of anadromous chumsalmon (Oncorhynchus keta) from the Bering Sea to Japan as wellas lacustrine sockeye salmon (O. nerka) and masu salmon (O.masou) in Lake Toya, Japan. Since each instruments has greatadvantages and/or minor disadvantages, we are developing an automatic salmon-tracking robot boat…

That's right, an automatic salmon-tracking robot boat. Leave it to the Japanese…
posted by koeselitz at 12:07 AM on June 8, 2009

I found something academic about the sea turtles too... a bit old but may be a place to start.
posted by evil_esto at 12:11 AM on June 8, 2009

They also use tiny laser-etched wires implanted in nose cartilage.
posted by barnone at 12:14 AM on June 8, 2009 [1 favorite]

There are plenty of salmon who only pinpoint the area. They crowd the mouth of Gold Creek every summer, a stream that hasn't hosted a salmon population in years due to the swift artificial stream bed.

Alaska's department of Fish and Game employs a small army of people every summer to track salmon populations both coming and going from fresh water.
posted by Foam Pants at 12:14 AM on June 8, 2009

Because I can't think of any possible way to prove that they behave like this.

Really? It's pretty simple. You trap the salmon on the way out of a certain stream, mark them or tag them, and then trap them again x years later on the way back upstream. They aren't so small when they are heading out to sea. Do this on a bunch of different streams on the same spawning run and you'll be able to show that the salmon return to the same stream and don't return to some other stream.

Third, you'd have to find the marked salmon among the hundreds that come to spawn in the stream.

Where are all these other salmon coming from? If you mark some proportion of the fish on the way out, you'll find the same proportion of fish marked on the return.

You can also use DNA techniques (salmon from river A are genetically different from salmon from river B) and I'm sure many other methods.

Do anybody know where and when this myth originated?

It isn't a myth. You can't just declare anything you want a myth if you don't understand how it is tested and can't be bothered to look it up.
posted by ssg at 12:14 AM on June 8, 2009 [16 favorites]

It's not invariant. It's not absolutely one hundred point zero zero percent. A few get lost.

But most do. That's how hatcheries maintain their stock, for example; the fry are released in a special stream at the hatchery, and when they become adults they return to the hatchery. I've seen big round tanks full of adult salmon at the salmon hatchery at the Warm Springs Indian Reservation.

The few that do get lost serve the species because they help mix the genes around.
posted by Chocolate Pickle at 12:15 AM on June 8, 2009

And - Chocolate Pickle for the win!!!

Simply make a new stream, raise fry in it, mark them with distinctive markers (radioactive chemicals would make automated marking and detection simple), and note what percentage of the adult fish coming up into the artificial stream to spawn are bearing the markers.

By deduction, if 99% of the adult fish are "returning home" (bearing the markers), then effectively "salmon always return to the stream in which they were born to spawn". QED.
posted by IAmBroom at 12:28 AM on June 8, 2009

Yes they can tag them and yes they do very comprehensively and no all salmon don't return to the same stream, some stray into other streams. This is adaptive behavior as they re-populate streams where populations have been extirpated and it helps them survive when a former spawning ground gets destroyed. They do usually make it pretty close to home though. btw, the home streams are called "natal streams" in the literature.

Tagging salmon is a huge business- pretty much all hatchery fish are marked and wild fish are tagged as needed in support of fisheries management. Wild fish are easy to catch as they move downstream. A single technician can tag hundreds of wild caught fish a day, the hatchery fish are often done automatically by the thousands or hundreds of thousands.

Fin clips, permanent dye marks and coded wire tags are the most commonly used methods. Another method is tagging the otoliths (or ear bones) of hatchery raised fish with a dye that they are fed at a certain point in their development.

There are also tags can be put in very small juvenile fish to track them more closely. These are not radio telemetry tags which are fairly large, they are instead much smaller passive units commonly called PIT tags. As each fish passes a recording unit on the bank (often called arrays) that sends out a signal they "ping" and each ping is recorded and you can track the fish moving around in the system that way. There are arrays all over the place. They are on the bridges in SF Bay for example to track tagged salmon and sturgeon.

Search "spawning and straying" or any of the other tagging terms above to learn more than you ever wanted to know about tagging salmon. Whoever said there was a small army was not kidding, there are entire conferences, businesses and careers devoted to tagging salmon, it's a pretty huge deal.
posted by fshgrl at 1:14 AM on June 8, 2009 [1 favorite]

Also salmon don't just go "out to sea", they usually cruise up and down the coastline and will go in and out of estuaries and bays. So a fish that was born in the Sacramento river and had every intention of returning there to spawn may be caught off of the Columbia River or in BC by a commercial fisherman or sport angler. Fin clipping and other tagging methods allow managers to track this (they have biologists who sample a certain percentage of the catch of commercial and charter boats) and they can tell approximately how many Columbia River stock are caught in Alaska or vice versa. All the info is compiled and released annually usually, sometimes it is used in real time to manage fisheries and then it is available on a daily or weekly basis.
posted by fshgrl at 1:20 AM on June 8, 2009

radioactive chemicals would make automated marking and detection simple

Ok, now you're just making shit up. Nobody uses "radioactive chemicals" to track salmon. This is food, remember?
posted by ryanrs at 1:26 AM on June 8, 2009 [1 favorite]

Ok, now you're just making shit up. Nobody uses "radioactive chemicals" to track salmon.

Tracking natal origins of salmon using isotopes...[and other stuff]
posted by randomstriker at 1:33 AM on June 8, 2009

ryanrs:

Take a geiger counter into a supermarket and read off the values for some low sodium salt. You will find that it indeed is radioactive. Now while they probably don't radioactively tag fish, just to say that all radioactivity must be eliminated is just rediculous. You have kotassium, carbon, phosphorus, and other elements in you right now that are radioactive.
posted by koolkat at 1:52 AM on June 8, 2009 [1 favorite]

Here's a nifty summary of the more commonly used tagging methods.
posted by fshgrl at 1:52 AM on June 8, 2009

Ok, now you're just making shit up. Nobody uses "radioactive chemicals" to track salmon. This is food, remember?

Bananas are pretty impressive too, IIRC.
posted by Kid Charlemagne at 4:59 AM on June 8, 2009

I wouldn't imagine it would be that difficult for a salmon to return to a specific spawning grounds. Why is that hard to believe? A creature as simple as an ant or a bee can travel far away and return to it's nest instead of mistaking it for a neighboring nest. Butterflies, robins, etc all return to specific areas after migrations.
posted by JJ86 at 5:50 AM on June 8, 2009

I'm interested that you posted this question absolutely sure that there's some weird salmon-myth conspiracy going on among scientists, that scientists get a kick out of perpetuating false, undocumented, or unproven ideas (myths!) among the general public, and that people tag salmon only in ways that fit with your pre-existing notions.
posted by Coatlicue at 7:07 AM on June 8, 2009

Because I can't think of any possible way to prove that they behave like this.

Well, you list some back-of-the-envelope layman's objections. But they're not really firm objections, especially since they can mostly be solved by doing them enough -- enough salmon, enough creeks, etc. Even if I hadn't seen filmstrips of this when I was a kid, it certainly doesn't seem to be an impossible task to me. It's also well-known that numerous animals have highly developed homing capabilities. Birds nest in one particular place in the US, then fly to South America, then come back to that one particular place. Even Monarch butterflies go to Mexico. And so on.

I'm more interested in finding out who first observed this than who proved it, since it seems to have been known for at least 150 years.

As with Coatlicue, I'm intrigued that you would say something like Scientists have been telling us for at least 50 years that [X]. Is there any documented proof of this? The whole point of science is telling us documented proofs. Now, it does happen that there are de novo things that science investigates about which it bandies hypotheses, but something as basic as salmon breeding has been studied for a very long time, primarily because it has economic value.

I'm also intrigued that you jump to the conclusion, without any proof, that this is a myth when you're asking if there's proof. The better way to phrase such a question would be "Is there proof -- OR is this just a myth?"

It's good that you ask for proof. This is a healthy impulse. But it's not so good that in the absence of proof you label something a myth. That is a word more applicable to things that have been disproven yet remain popularly believed -- or things that cannot be proven in any reasonable way. It is not a word applicable to things for which you don't yet know the answer.

For these reasons I highly recommend to you this YouTube animation about open-mindedness. It will help you sort out your thinking about questions like this.
posted by dhartung at 7:30 AM on June 8, 2009 [1 favorite]

See also thermal marking on otoliths. Related metafilter thread here. 100% of hatchery salmon can be marked using thermal marking. Hatchery salmon are often penned and released at a variety of streams in Southeast Alaska. As pointed out up-thread, not EVERY returning salmon comes back to exactly the same stream, but the proportion is pretty high. Different species of salmon also have different "fidelities" to their natal streams. I believe king and silver salmon return in higher proportion to the exact stream where they hatched, pink and chum salmon less so, but I can't find a citation for that right now. It's something that we talked about in the hatchery lab that I used to work in, tho.
posted by otolith at 7:30 AM on June 8, 2009

This was directly observable before anyone did any tagging. All you needed was a localized catastrophic event to afflict a stream during spawning season or depending on the event over winter, say a mud slide. This would reduce the population of returning fish four years later quite specifically (IE: even in a year of record returns a stream that had poor spawning four years earlier will have poor returns). You can be knee deep in fish in a stream while the stream next door has only the occasional fish and most of those are 3 or 5 year stragglers.
posted by Mitheral at 7:31 AM on June 8, 2009

Tracking natal origins of salmon using isotopes...[and other stuff]
posted by randomstriker at 4:33 AM on June 8 [+] [!]

This is a paper about comparing two stable isotopes in salmon earbones to mixtures in rocks near various streams, not a paper about injecting salmon with long-lived emitters so they can be detected years later from far away.
posted by fantabulous timewaster at 9:14 AM on June 8, 2009

About half of the seasonal salmon runs in California no longer return to the streams where they spawn. This is because virtually all of the salmon are hatched in one place, and then trucked to the mouth of the SF Delta and released, so they have no idea how to get back. This has been shown through tagging, otolith comparison, and genetic identification. Just to back up the original question, but not for any of the reasons you wildly speculated.
posted by one_bean at 10:07 AM on June 8, 2009