On September 1, 1914, an old, trembling passenger pigeon named Martha died at Cincinnati Zoo. With her demise, her entire species slid into extinction. But in many ways, the species was already gone, for a solitary passenger pigeon is almost not a passenger pigeon at all. This is an animal that existed in gestalt. Its essence was in the flock.

Passenger pigeons were once the most abundant bird in North America, and quite possibly the world. At their peak, there were a few billion of them, traversing the continent in gargantuan, nomadic flocks that would blacken the sky for hours as they passed overhead. Simon Pokagon, a Potawatomi author and leader, described them as “the grandest waterfall of America” and their sound as that of “distant thunder” or “an army of horses laden with sleigh bells.”

And then, people started shooting them. They poisoned them, netted them, gassed them, hit them with sticks. In a matter of decades, the continent’s most common bird has been completely wiped out, down to the last individual. “It’s always astounded me how something could have that large a population and entirely disappear,” says Beth Shapiro from the University of California, Santa Cruz. “Why didn’t tiny populations survive somewhere in refugia? I mean, we are pretty good at murdering things, but how did we kill every one of them?”

These questions have been debated for decades. But Shapiro and her colleagues Gemma Murray and André Soares have found some new twists to the old answers by collecting bits of skin from around 200 passenger pigeons, whose century-old, taxidermied bodies sit in museums around the world. Using these samples, they sequenced the full genomes of four individuals, and compared them to the genome of the band-tailed pigeon—a close relative that still exists but lives in considerably smaller flocks.

At first, nothing jumped out. On average, the passenger pigeon’s genome looked to be extremely diverse—two to three times more so than that of any other bird that had been sequenced thus far. That made sense, given how many of them there were.

But averages are deceptive. DNA is packaged in chromosomes, and the team found that the genetic diversity at the ends of these chromosomes was exceptionally high, while the diversity in the middle was exceptionally low. The band-tailed pigeon doesn’t share the same pattern; its genome has roughly the same level of diversity throughout. Indeed, Shapiro had never seen anything like this before. This pattern—and the evolutionary forces that produced it—have important implications for understanding both why the passenger pigeon died out, and whether it'll be possible to bring it back.

Here’s why the pattern exists. When animals reproduce, their chromosomes mix and mingle, shuffling their genes into new groups. This process, known as recombination, breaks up blocks of genes, allowing natural selection to weed out the worst mutations and keep only the best ones. But in birds, recombination happens more often at the ends of chromosomes than in the middle.

Imagine that you’re going through your wardrobe trying to chuck out any clothes you hate, while keeping only the ones you love. Unfortunately, you find that some miscreant has stitched all the shirts, skirts, and pants together. If you want to keep a particular shirt, you’re forced to keep everything else that goes with it. That’s what happens in the middle of the pigeon’s chromosomes. Recombination is low, so genes stick together in large blocks, making it hard to select for one without getting all the hangers-on. But in your wardrobe, the hats, scarves, socks, and shoes are still free and loose, allowing you to consider each item individually and choose the best ones for your ensemble. Same goes for the ends of the passenger pigeon’s chromosomes: That’s why they’re more diverse than the middle.

It’s commonly assumed that animals with massive populations should be genetically diverse. But “the passenger pigeon’s genome is that of both a low-diversity species and a high-diversity one,” says Ben Novak, who worked on the study. “In either case, it was well-adapted for its preferred lifestyle.”

Indeed, that’s why its genome is so weird. Genomes can evolve either through drift, in which DNA changes randomly, or through natural selection, in which genes become more or less common because they affect their owner’s ability to survive and reproduce. Typically, both forces are important. But the passenger pigeon was so absurdly abundant that natural selection dominated. “There was almost no portion of the passenger pigeon’s genome that was evolving neutrally,” says Shapiro.

This might all sound pretty wonky, but it matters when thinking about why passenger pigeons died.

In 2014, Chih-Ming Hung and colleagues from National Taiwan Normal University used the genomes of passenger pigeons to reconstruct their historical population size. They concluded that the pigeon was a boom-and-bust bird, whose numbers cycled dramatically between incredible highs and stark lows. That created a natural vulnerability, which humans inadvertently exploited by persecuting the pigeons during a bust phase.

But Shapiro says that Hung’s team made a mistake. They used a technique which assumes that genomes are evolving neutrally—and the passenger pigeon’s largely isn’t. (Hung stands by the conclusions of his study.) Instead, it had been crafted by natural selection to an extent that most other species are not. “Passenger pigeons were not a naturally vulnerable species,” as has been repeatedly suggested over the last five decades, says Novak. “It was a superspecies in its natural element.”

So, why did this superspecies die out? Shapiro thinks it’s because the bird specifically evolved to live in mega-flocks, and developed adaptations that became costly when their numbers suddenly shrank at human hands. “Maybe they were simply not adapted to being in a small population, and didn’t have time to recover,” she says. Maybe they hit a tipping point when there were just too few of them to survive, regardless of whether they were being hunted.

That’s a bit of a leap from the data, but it’s an idea that’s worth entertaining, says Erich Jarvis from Rockefeller University, who studies bird genomes. “It makes one think: Just because humans leave a small population behind without killing off the rest of the species, it does not mean that the species will survive. It would be good to see if there are other species like this.” Shapiro agrees, and wants to see if the same patterns exist in the genomes of other animals that live in massive groups—perhaps herring, bristlemouth fish, or red-billed queleas.

Meanwhile, Novak cautions that we don’t know if the pigeon actually did suffer at low densities, or the minimum number that would have been necessary to prevent extinction. Indeed, he says that there’s evidence the passenger pigeon would have fared reasonably well at low numbers, if we had just left them alone. Records show that they were breeding just as efficiently when there were only a few hundred left as when there were billions.

The fact is that “human persecution was relentless right up until the very end,” he says. “The rarer the birds became, harvesting efforts only grew more intense. Whatever maladaptive trade-offs may have existed for the passenger pigeon, their decline was simply too rapid for these trade-offs to show symptoms.”

Novak is part of a group called Revive and Restore, which wants to use modern genetic tools to resurrect the passenger pigeon. These birds were crucial parts of America’s eastern forests; their huge numbers and voracious appetites created a patchwork of destruction and regeneration that many animals and plants actually depend on. Novak’s team wants to bring those cycles back, and they need passenger pigeons to do so. But if the bird can’t survive at low numbers, their project seems doomed to failure unless it simultaneously resurrected a flock of millions.

Not so, says Novak. His team isn’t trying to raise passenger pigeons, Lazarus-style. Instead, they are trying to turn the band-tailed pigeon into something passenger-ish, tweaking its genes so it becomes far more social. If their flocks get big enough, they can step into the role that the passengers once played. And over time, their genomes should evolve into the same pattern—diverse ends, uniform middles. “This isn’t something we will engineer, it is something that will happen due to selection and recombination in nature, just as it happened to passenger pigeons originally,” Novak says.

It’ll be a long process, and he won’t be around to see its conclusion. But he wants to see its beginning.