Featured paper: five warblers, lots of tornadoes, and a mystery

Male Golden-winged Warbler. Photo by Mark Peck*

Male Golden-winged Warbler. Photo by Mark Peck*

This story begins when Streby et al. (2015) decided to track Golden-winged Warblers during their annual migration. We know that lots of birds migrate, but for most of them, we know surprisingly few details about that migration. Often we know generally where they go (to a specificity of, say, “somewhere in South America”) but not exactly where; rarely do we know what paths they take to get between wintering and breeding grounds. This kind of information is especially important for birds of conservation concern, since to protect a migratory population, you need to protect its wintering grounds and migration route as well as its breeding grounds.

The researchers relied on technology to tell them where the warblers went when they migrated. There are several different ways to track animal movements; in this case, researchers used light-level geolocators, which record the amount of light hitting the geolocator. Collected over time, these light intensity measurements allow researchers to calculate where the geolocator was, based on things such as day length. This location information isn’t as accurate as the data you would get from a GPS logger, but the light-level geolocators have a big advantage over GPS loggers: they can be much smaller, so you can put them on tiny birds like Golden-winged Warblers. A Golden-winged Warbler attached to a heavy, clunky GPS logger would not be migrating anywhere.

Photo by Kent McFarland*

Photo by Kent McFarland*

So the researchers attached very light, very small light-level geolocators to twenty male warblers. Some of those birds didn’t come back the following year; or they came back but had lost their geolocator; or they came back but had their geolocator removed before the interesting-thing-I’m-about-to-talk-about happened; or they came back but refused to be recaptured by the researchers, so the researchers couldn’t get their geolocators back. (You need to get the geolocator back to read the data off of it.) In the end, the researchers had location data for five warblers. When they looked at those location data, they saw something really interesting.

In the spring, the five warblers migrated 5000 km from their wintering grounds in Colombia to their breeding grounds in eastern Tennessee. This was expected. Not long after they had arrived, however, the warblers took off again, abandoning their breeding territories to fly back south. Four of them ended up in Florida, while one went all the way to Cuba. This was not expected, but it was a good idea, because just 1-2 days after the warblers left Tennessee, a huge storm system swept over the area. The storm system “spawned 84 confirmed tornadoes,” which the warblers would presumably not have enjoyed. The storm was bad enough that the researchers themselves fled: they comment that they missed a day of surveying “because we performed our own evacuation migration and waited out the storm in Caryville, Tennessee.”

This, plus 83 more. Photo by Florencia Guedes*

This, plus 83 more.
Photo by Florencia Guedes*

The five warblers waited just a few days for the storm to pass, then returned north to their breeding grounds and began defending territories again.

This all seemed very sensible—when 84 tornadoes are headed your way, it’s time to go somewhere else—but it was also puzzling, because the birds left too early. They left the breeding grounds while the storm was still far away, and while the weather at the breeding grounds seemed normal. The researchers looked at weather variables such as atmospheric pressure, wind speed, temperature, rain, and clouds, and found no changes that could have alerted the birds that a massive storm was approaching.

So how did the warblers know about the storm?

We knew by MAGIC! Photo by Amy McAndrews*

We knew by MAGIC!
Photo by Amy McAndrews*

The most likely explanation is that they heard it. Birds can hear very low sounds, called infrasound, which is generated by storms and tornadoes and which can travel very long distances. (Humans cannot hear infrasound.) Infrasound could not only tell a bird where a storm is, but it could potentially tell the bird what direction the storm is moving, based on Doppler shifts in the sound. (A Doppler shift is when a noise sounds different depending on the movement of the source of the noise: think of how ambulance sirens sound different when the ambulance is coming towards you than when it is going away from you.)

This is what you secretly hope for in field biology: that just at the moment that you happen to be watching, some unpredictable, dramatic event will happen and show you something unexpected and new about your study animals. Streby et al. (2015) got to watch their warblers sense and evade a huge storm—and they got the information they originally wanted, about the warblers’ migration, too.

Not bad at all. Photo by Caleb Putnam*

Not bad at all.
Photo by Caleb Putnam*


Streby HM, Kramer GR, Peterson SM, Lehman JA, Buehler DA, Andersen DE. 2015. Tornadic storm avoidance behavior in breeding songbirds. Current Biology 25:98-102.

*Photos obtained from Flickr and used via Creative Commons. Many thanks to these photographers for using Creative Commons!

9 thoughts on “Featured paper: five warblers, lots of tornadoes, and a mystery

  1. I am curious why the researchers left out dewpoint–something even we feel when weather approaches. The rise and fall of the dewpoints corresponds seemingly well with their departure and arrival.

    • Hi Shaina,
      Apologies for my misleadingly incomplete list of the variables they tested – that isn’t ALL the variables, just some examples I though people might recognize. They did in fact test dew point, indirectly: they tested the “severe weather threat index” (SWEAT), which includes dew point in its calculation.

    • There’s that saying, “any sufficiently advanced technology is indistinguishable from magic.” We could add to that, “any sense sufficiently beyond our own perceptive capabilities is indistinguishable from magic” – although fortunately it isn’t QUITE indistinguishable, with sufficiently advanced science!

  2. Did they try direct measurements with infrasonic microphones? Since they don’t have a time machine, they wouldn’t be able to try it for the particular case they observed, but it would allow them to test if tornadoes produce measurable infrasound and Doppler shifts.

    • It’s already been done. We know that severe storms, and in particular supercell storms that generate tornadoes, make a lot of infrasound. (People are studying this for, among other things, weather forecasting: the infrasound occurs before the tornado does, so you could potentially use it to predict tornadoes.)
      I’m not sure if Doppler shifts in the infrasound from storms have been measured; presumably that would depend on the speed of each specific storm system. But we do know that birds can detect Doppler shifts in infrasound.

  3. Thanks as always for posting this stuff.
    I wondered if the bird is too small to carry our GPS, how come it evolved to spend space on an infrasonic detector? Or to put it another way, how come we ain’t all got one if its that useful? And i suppose the answer is, its only useful if you can migrate a thousand miles in response. Also perhaps other animals watch the birds and take their flight as signals. I wonder how evolutionarily conserved this is in birds and whether it gave them an edge when a 3km meteor drilled through the sky? Random late night thoughts.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s