A Blog by

Speeding Towards Birds In A Car… For Science!

In the winter of 2006, Pierre Legagneux started measuring when birds would fly away from him, as he sped towards them in his white Peugeot. This wasn’t an official part of his research; he was just bored. After a recent move, his mornings of cycling past bucolic villages and forests had been replaced by long, tedious drives. “I found it very boring so I found something to do while driving,“ he says. “I started recording birds flying away in front of my car.”

Legagneux drove down roads with speed limits of 20, 50, 90 or 110 kilometres per hour, and either stuck exactly to those speeds or deliberately drove under or over them. Every time he saw a bird on the road—usually a crow, blackbird or pigeon*—he started a stopwatch when it flew away and stopped it when the car passed over the point where the bird had been. By multiplying the car’s speed by the time on the stopwatch, he could work out how far away he was from the bird when it took off—the “flight initiation distance”.

One year and 134 measurements later, Legagneux clearly showed that birds flee from incoming traffic at greater distances on roads with higher speed limits. On a 110 km/h road, they’d be airborne when the Peugeot was 75 metres away. On a 20 km/h road, they’d wait till the car was less than 10 metres away before taking off.

That’s fairly predictable. But more surprisingly, Legagneux also found that the birds didn’t react to the actual speed of his car. Their flight initiation distance depended on the speed limit of the road, but not on the speed of the incoming traffic.

Other factors were important too. The birds allowed the car to get closer if they were smaller (and presumably more agile) or during the autumn and winter (when newborn chicks have had a few seasons to get accustomed to roadside perils). But regardless, the main results still held even after Legagneux accounted for the season, the weight of the birds, their location on the road, and their evolutionary relationships.

We know that birds can inherit risk-taking behaviour from their peers, so faster roads could select for more cautious birds. Indeed, urban birds eventually evolve longer flight initiation distances than rural ones, and some swallows that nest near roads have evolved shorter wings and tighter turns.  But Legagneux doesn’t think that the birds in his study were actually adapting to the different roads, since they’re likely to experience a variety of speed limits.

Instead, they’ve probably come to associate different stretches of road with different risks of collision, and adjusted their behaviour accordingly. This may explain why birds with larger territories are more vulnerable to cars—they might not be able to tailor their activities to local roads.

* No chickens…

Reference: Legagneux & Ducatez. 2013. European birds adjust their flight initiation distance to road speed limits. Biology Letters http://dx.doi.org/10.1098/rsbl.2013.0417

7 thoughts on “Speeding Towards Birds In A Car… For Science!

  1. I recall driving along and measuring this since I was old enough to see over the back seat of my dads station wagon and then when I drive, and recall 1 event while returning to the City and along a road called Airport Road, Toronto West, and it travels North to cottage country. I saw several massive 1,000 + flocks of migrating birds, starlings, Red winged bb’s, and even hundreds and + of cardinals, Jays, etc.. average speed? 80 km/h. I drove along them ansd watched as they all stopped within 30 feet and stopped to relax on trees.

  2. Kudzu–I doubt there were many (if any) collisions. The chances that Legagneux hit birds with his car are the same as anyone’s chances of hitting birds while driving. However, I’d say the photo is a bit misleading–coupled with the title, it makes it seem as if the article is going to be about someone driving around intentionally hitting birds, not simply measuring how close the car gets before they fly away.

    [Yeah, I can see that. Hadn’t actually considered that people would assume the guy in the photo is Legagneux. I’ve swapped it out for one of a chicken crossing… – Ed]

  3. We live in an area with both turkey vultures and black vultures. They feed on roadkills (mammals hit by cars.) Black vultures, in particular, are slow and clumsy about getting off the ground (indeed, they’ll often waddle off the road on foot rather than take flight) so speeding toward a cluster of vultures has led, in more than one case, to a broken windshield as this large, heavy bird makes a last desperate attempt and gets just high enough to clear the hood of the vehicle.

    They don’t usually come down to feed on roadkill on high-speed, heavily trafficked roads, though you may see several of them by the side of the road, shifting from foot to foot and hissing, and people who live on or drive on the smaller country roads they favor usually know to slow down and give them time. .

    The oddest bird I ever saw cross a major highway was a wild turkey male. In 1973, we were moving from Austin, Texas to San Antonio. The main highway between them ran through rural country and had a two-lane service road on either side of the divided central lanes. Turkeys can fly; in fact they’re strong flyers for their weight and size, but they usually fly low, and for short distances. A turkey’s launch from the ground is superior to the black vulture’s, in terms of rapidly gaining height. This one ran across the service road far enough ahead that I recognized it. Then it ran across the grass between the service road and main road, flew to the median high enough to clear cars, ran across that, launched itself to fly across the active lanes on the other side. It was clear the turkey had enough road experience to take off and fly over pavement. However–on the far side, It took off close enough to an oncoming very large truck to provoke a violent swerve from the truck, cleared it by inches, glided over the rest of the traffic on that side, and took off running when it landed. (Yes, I had stopped the car out of traffic to watch.) A big turkey is startling in flight anyway–they don’t look like they could fly at all. To see one make it across a divided highway and its service roads was astonishing. We still talk about it.

  4. 134 measurements, fine…

    What’s the p-value after fitting by size, season, and genus?

    Let alone the systematic error of some dude basically estimating when he passed where the birds took flight, while manually clicking a stopwatch…

    Neat science fair experiment. This sure ain’t particle physics…:-)

  5. This doesn’t make sense. The birds are detecting the car (because they’re getting out of the way) and so the only reason they wouldn’t be using perceived time-to-contact for the car would be if they couldn’t pick it up, ie it’s below threshold. Then, maybe, picking a constant value and calibrating that to the local conditions might make sense.

    But this doesn’t work because they’re still leaving at a time/distance (more likely time, actually) relative to the car. They are perceiving that the car is x amount of time or distance away and leaving at some critical value of x. So some kind of time-to-contact information is available.

    The only other thing I can think of (besides measurement error from the admittedly crude measurements) is that the differences in speeds aren’t making enough difference to the time-to-contact; maybe the difference between his actual speed vs the speed limit was below perceptual threshold and it averaged out at the speed limit.

    Interesting, and if I ever get time I will look in more detail. But perception is for tailoring your behaviour to local conditions and educated guesses about those conditions get you dead. My guess is there’s something missing here.

Leave a Reply

Your email address will not be published. Required fields are marked *