The extinction crisis

A new study is re-evaluating the extinction of thousands of animal and plant species. Leading this project is Columbia University professor and biologist Don Melnik. He joins Hari Sreenivasan to explain why he believes current extinction estimates are vastly underestimated.


A new study is re-evaluating the extinction of thousands of animal and plant species.

Leading this project is Columbia University professor and biologist Don Melnick.

He joins us now to explain why he believes current extinction estimates are vastly underestimated.

First of all, how do we measure, how do we get estimates now of extinction events?

So, we can do two things.

One, we can look over the historical record -- and by that, I mean, say, the last 500 or 600 years -- and look at species that were recorded in particular places, and look now and we'll see that, in fact, they're not there anymore.

So we see localized extinction.

And if those were the only places they existed, then we have complete extinction of that species.

The other thing is, we have something called the International Union for the Conservation of Nature, which is sort of UN-level body which deploys scientists all over the world and asks them for information about particular species.

So, they're keeping track of what condition species are in.

And by that I mean, what the size of their range is, what the likelihood that they will suffer local extinction or complete extinction based on the size of their range and other characteristics of their range.

So, are these researchers that are out there, do they generate any kind of a map or a list of where this is happening, and what's worse?

That's a good question.

They generate both.

So, the data that they have is aggregated and based on bringing all of those data together.

They create maps for each species.

So each species has a range map.

On the basis of that range map, they get a sense of what the vulnerability might be.

And so the study that we have been doing has really been focused on, how do you create that range map in the most accurate way possible, using the latest technology possible?

And what we did was, we focused particularly on a set of species that are endemic to the mountains in the southwestern part of India.

And by endemic, we mean they're found there and nowhere else.

If they disappear there, they're gone forever.

And so we focused on those species.

But it would take thousands of scientists and many, many years to gather the information you need.

But we were lucky, because people in India are -- Birding is a big pastime.

And also because Cornell University set up something called eBird, which is a database.

So when you're out in the field, you can see a bird, you can snap a picture of it.

You have a GPS location.

You can enter it into the file.

And suddenly you have a datum.

You have something that is useful.

We took all of that for that mountain range and for these endemic species.

And we were able to look at 18 species very closely because of that.

And what we found is, using that and using satellite imagery and using GIS and using 27 different biological and geophysical parameters, that we have actually, at a very fine scale -- that is, we know exactly what's in a 100x100-foot square across that entire mountain range.

We have a value for every single one of those parameters.


And so we looked at the correlation between sighting a bird and those parameters.

That allowed us to, then, retroactively go back and say, 'Wherever we find those parameters in values of this particular type, we see a bird.'

Now, in other areas that people haven't gone birding, if we see those parameters, the birds are likely to be there.

And, basically, using all that technology, we were able to estimate the range of these species.

But what we found is, when we estimated the range of the species, they were maybe 1/5 or 1/10 the size of what had been registered 20 years ago.

So, does that mean that you can take these kinds of tools and apply it to other species in other places around the world, as well?

That's exactly right.

So, what we're doing now is we're automating this approach so that anybody can use it.

Because, obviously, we're just a small group.

We can't go everywhere in the world and do it.

But there are many scientists who are on the ground in places all over the world.

The idea is to automate this so that anyone, anywhere can use it.

Not just on birds, but on mammals and reptiles, amphibians, insects, whatever they want to use it on.

Even plants.

They can use this sort of collection of technologies as a kind of package.

What's the importance of extinction data?

How is this used in policy?


So, usually, when a species is endangered, it raises the level of concern not only among scientists, but very often among policymakers who are in that particular local area.

And then that can spread out and suddenly become a great concern even nationally.

We've seen this in many countries.

We saw it in the United States with the bald eagle.

We see it in Indonesia with the Javan gibbon, the small ape.

We just see it in many, many places -- China with the panda, whatever it is.

But our view is, there are these charismatic species.

But we would rather this be focused on the whole collection of species.

So, by automating this, you can look, and by using citizen-science data, so you don't have to have your own army of scientists on the ground, and using the certain filtering processes for that citizen-science data, 'cause not all of it is accurate, that you can then deploy it over very large landscapes.

And that should determine that there are areas within countries where many, many species are now considered vulnerable, right, and therefore the attention of policymakers should be on those areas.

Not a particular species, but on the area in general.

Because something is going on in that area which is endangering many of the species that are there.

All right.

Don Melnick, Director of the Center for Environment Economy and Society at Columbia University.

Thanks so much for joining us.

Thank you.