Skip to main content
Menu
Big Data Is Saving This Little Bird

“We need to improve conservation by improving wildlife monitoring. Counting plants and animals is really tricky business.”

The marbled murrelet is an enigma. It wasn’t until the 1970s that biologists discovered where the chunky brown-and-white bird made its home, and even then it was by accident: A tree-climber found a murrelet chick at the top of a redwood. Most other bird habitats had been mapped for centuries. But who would have thought to look for a sea bird’s nest miles away in the middle of an old-growth forest?

And it’s elusive. California birders can go a lifetime without seeing one. Every day at the break of dawn, the murrelet zips down from the redwood forest hills, where it lives, to the ocean, where it feeds. It then returns under the cover of darkness. See the murrelet on its dawn commute here?

Or maybe not? That’s where big data helps.

Using remote acoustic sensors and machine learning to analyze the audio, biologists are now better able to track populations of species that were previously hard to monitor. With a threatened species like the marbled murrelet, that can make a huge difference. The better the data on its population and nesting patterns, the better our understanding of how its habitat is threatened, and the more effective conservation efforts can be.

On this week’s episode of our podcast What’s The Point, we go on a hike in the redwood forests north of Santa Cruz, California, with Matthew McKown, head of Conservation Metrics, a company using data to improve conservation efforts.

Also this week, a Significant Digit on a host of new species recently discovered in the Himalayas.

Stream or download the full episode above. Below you’ll find a partial transcript, photos from our visit with Matthew and some of the bird call spectrograms discussed on the air, along with sample bird calls.

Audio extra: More bird calls from the Conservation Metrics database. Find the clip on Dropbox.


Song-Meter-SM2-A-1
SongMeter

Below are some of the spectrograms of bird calls compiled by Conservation Metrics. The acoustic data they gather is converted into a visual representation — time runs along the x-axis, audio frequency along the y-axis — and then an algorithm searches for patterns to isolate specific bird calls.

Take a listen to this complex soundscape. Many different species are making noise at once, but the computer has identified a marbled murrelet call within the clip. It’s the upside-down horseshoe pattern running along the lower third:

SV07_SV07_0_20150609_044905_000_startsec_2266_label_4

Here’s a spectogram of a cleaner recording of a marbled murrelet as it flies over the sensor. It’s easy to hear, and easy to identify in the image:

SV07_SV07_0_20150605_202500_000_startsec_631_label_4

Here are two more soundscapes from the Acoustic Metrics database, one of the Wedge-Tailed Shearwater and one of the Western Screech Owl.

WedgeTailedShearwater
WesternScreechOwl

Will data ever replace field work?

Jody Avirgan: You’re a birder, right? You love birds. You’ve gone out in the field and done bird watching. So how does the thrill of finding an acoustic pattern in your data set compare to the thrill of stumbling across a species in the wild?

Matthew McKown: It’s the same. It’s super exciting. Though there is something to be said for being in the field [with] all of your senses — when you’re seeing something visually, the smells, you feel the air and you hear things. … And these tools are not going to replace traditional field surveys by biologists. No matter how good machine-learning techniques get, there’s still this extra thing that humans have. [Like] the ability to detect something that shouldn’t be there. Which is very hard for computers to do. You’re in southern California and you see a walrus, that is something people are going to notice. The computer might just skip right over it and say, “Oh that’s not the kind of seal that you’re looking for.”

Avirgan: Because it hasn’t even occurred to you to tell the computer to look for that species?

McKown: Exactly. [That’s] a very active area of research in this area. How do you detect novel classes? Again, these tools are not going to replace people, they’re just going to allow people to do more in the field. So it’s super exciting when we find something new, when we discover a species that people had a hunch was there but they didn’t know [for sure], and now we have evidence. It’s great.

Avirgan: But it probably makes you just want to hop in the car and go see the thing?

McKown: It is the one problem with this job that we’ve created for ourselves. We’re now more and more in the lab and we’re working with colleagues and clients in the field, sending them sensors and getting data. So we get to get out in the field less and less, although we do look for projects where we can do all of that. I love being in the field.


If you’re a fan of What’s The Point, subscribe on Apple Podcasts, and please leave a rating/review — that helps spread the word to other listeners. And be sure to check out our sports show Hot Takedown as well. Have something to say about this episode, or have an idea for a future show? Get in touch by email, on Twitter, or in the comments.

What’s The Point’s music was composed by Hrishikesh Hirway, host of the “Song Exploder” podcast. Download our theme music.

Jody Avirgan hosts and produces podcasts for FiveThirtyEight.

Comments