New Study Doubles the World’s Number of Bird Species By Redefining ’Species’

A modified definition of 'species' suggests there are nearly 20,000 different kinds of birds, but the model is controversial and difficult to apply to conservation.

If the story of the Biblical Flood took place today, and Noah were a modern taxonomist, the tale might have ended very differently. When instructed by God to load two of every creature into the ark, Noah would have to push for a few more details: “What exactly do you mean by ‘kind’ of creature?” he would need to ask The Almighty. “A biological species, or a phylogenetic species? Does your list of creatures include subspecies? And what about distinct population segments?”

Such is the state of modern taxonomy, where scientists spat over a fundamental question: What is a species? It seems simple on the surface. A species includes a group of animals that look alike and breed with one another. This is the traditional definition of a ‘biological species.’

But when two species are closely related to one another, the boundaries sometimes grow fuzzy. Take, for instance, the Dark-eyed Junco. In different parts of the U.S., Dark-eyed Juncos look quite distinct, forming multiple subgroups by appearance. Where the subgroups meet, they breed—but otherwise, they mostly keep to themselves. The question then is: Is each subgroup its own species, representing a distinct evolutionary path? Or do they together form a single species because they can and do mate? Currently, they're classified as a single biological species, but not all taxonomists agree that this makes sense.

“It’s the way a lot of birdwatchers look at birds,” says Joel Cracraft, an evolutionary biologist and ornithologist at the American Museum of Natural History in New York. “Field guides recognize that all these distinct forms need to be illustrated. They recognize that a linear list of species is not what birdwatchers need: They need to understand the differences and similarities amongst these taxa.”

A few years ago, Cracraft decided to take this approach and apply it to all of the world’s birds. With a handful of collaborators, he got to work on a revised estimate of the world’s bird species, where ‘species’ was defined as any group exhibiting a unique, shared set of traits, regardless of whether it could mate with other ‘species’ (hereafter referred to as ‘taxa’).

Last month, they published their results, which identified more than 18,000 bird taxa—nearly twice as many as the current count of the world’s bird species. The explosive figure has some ornithologists rolling their eyes and others nodding in agreement, depending on how they feel about biological species. But the long-standing debate about species definitions reaches beyond the academy. The authors suggest that their ‘taxa' should be the framework for conservation—an argument that, while well-intentioned, could undermine existing laws for protecting wildlife.

To come up with that number, the authors randomly chose 200 species from a list of the world’s birds published in 1986, which enumerated 9,159 species. (Most current lists put the figure closer to 10,500 species.) Each of the three scientists took responsibility for one-third of the birds on the list and carefully examined museum specimens—dozens of specimens per species, Cracraft says—to identify subgroups based on breeding plumage and geography.

From this close morphological study, the team found 1.97 taxa per species on average. (More than half of the 200 species were, in fact, single species; the highest count was 11 taxa within a biological species.) When they extrapolated this average out to the full list of 9,159 biological species, they landed on their new estimate of 18,043 bird taxa.

The group also combed the scientific literature to identify any papers that used genetics to split species into subgroups or taxa. “Genetics can pick up lineages of birds that are really distinct in genetics but not necessarily in their external morphology,” Cracraft says. This search identified 437 biological species, which on average were split into 2.4 taxa each. 

The authors are almost certainly correct that, on the whole, scientists underestimate the number of biological bird species in the world. Indeed, earlier this month, BirdLife International (an Audubon partner) recognized 742 new bird species, many of which are cryptic species hidden within biological species. But Cracraft's study and findings, on the other hand, rely on fundamentally redefining "species." As a result, it’s difficult to know how to use the team's new estimate or compare it with traditional species counts.

Furthermore, their average estimates for how many taxa are embedded within each bird species—the rough doubling—are likely biased. The examination of museum collections was subjective, for one; the researchers didn’t publish guidelines for how they split the species, so it’s impossible to repeat the study or otherwise determine if they split the taxa accurately. “Other experts might have identified more or fewer species among the 200 current species,” says Andrew Solow, an ecological statistician at Woods Hole Oceanographic Institution in Massachusetts, who looked over the new study for Audubon.

In addition, genetic studies tend to focus on examples of confusing species, which are more likely to be split into multiple taxa. “The 437 species for which detailed genetic information is available are not random, but may indeed reflect genetically 'interesting' species,” Solow says. “For this reason, it is not easy to generalize the results to all 9,000 current species.”

Drawing the Line 

Even if the team's estimates are imperfect, Cracraft still sees their method as a better way of counting the world’s biodiversity than biological species. Each of their taxa is an “evolutionary unit,” he says, and combined, all these units more accurately describe the variation in avian life.

More crucially, he sees his view of biodiversity as important for conservation. If species described unique subgroups or taxa, rather than biological species, then laws that protect species—like the U.S. Endangered Species Act, IUCN Red List, and CITES convention—would do a better job protecting the world’s biodiversity, he says. “If you look at the biological species concept, you can only lose 9,500 species. Under ours, you could lose 20,000,” Cracraft says. “You want to save the entire diversity of the group you’re looking at. And there is a ton of biodiversity that is just being ignored.”

It’s an idealistic vision—but it may not be practical. To practice conservation, scientists need to be able to identify species in the field. Currently, available tools can't sort all wild birds into taxa based on subtle differences observed from afar—and especially genetic differences. “If you can’t go out and see the difference, I don’t see the point of it,” said Susan Haig, who leads the avian genetics lab at the U.S. Geological Survey, when we spoke earlier this year. “Taxonomy is for identification.”

Additionally, the taxa that Cracraft describe aren’t necessarily stable. They provide a snapshot of biodiversity at this moment in time—and one that quickly expires. There is normal variation in color or behavior within any biological species; an entire population is not uniform. Over time, that variation gets mixed around and redistributed through mating, so that the differences we see today aren’t always relevant to the species’ evolution over the long term. That’s why the ability to reproduce is part of the traditional definition for a species, a definition that scientists have largely agreed upon for a century.

“Taxonomy is discrete—species or no species—whereas evolution is continuous,” says Frank Zachos, curator of mammals at the Natural History Museum in Austria, who studies species definitions in conservation. “It is a continuous process in which two lineages diverge, but they can also merge again, of course, so where exactly the line should be drawn will always contain some level of arbitrariness.”  

And that very idea—that the definition of a ‘species’ can be seen as arbitrary—is a dangerous one. Already, scientific research questioning the validity of the biological species concept, which underpins conservation laws, is being weaponized to claim that subspecies listed under the Endangered Species Act shouldn’t be protected. Not because they are no longer threatened with extinction, but because they aren’t truly unique species and shouldn’t have been protected in the first place.

Some biologists even accuse proponents of defining biodiversity based on these sort of ‘taxa’ (sometimes called ‘phylogenetic species’) as intentionally trying to undermine conservation laws. To hear the phylogeneticists tell it, however, they aren’t anti-conservation; rather, they believe that laws that focus on the protection of rare species miss the big picture.

“We’re losing the battle because we’re fighting over single endangered species,” Cracraft says. Species protection tends to focus on charismatic species—beautiful birds or mammals—and doesn’t always value rare ecosystems or collections of species. A legal mechanism to identify and protect constellations of his rare taxa, rather than single species, would be a powerful force for conservation, he argues.

The need to protect ecosystems, and not just species, is an idea with which many biologists would agree. But reaching that goal by redefining 'species' would undermine the laws that protect them, including most of our major conservation laws. “If what a species is is ultimately arbitrary, all these quantifications of biodiversity also are,” Zachos says. “And that is a bit of a frightening prospect.”