ABOVE: Sky Banana posing in her tank

“What should her name be?” I ask my soon-to-be four-year-old daughter, Bianca. She and I had just picked out her first pet: a female betta fish. My daughter’s favorite color (for now, anyway) is purple, so even though the pet store had a whole display full of bettas to choose from, she was immediately drawn a lavender-toned variety of the “elephant ear” or “Dumbo” phenotype, so named because of its particularly large pectoral fins that resemble the ears of elephants.

“Sky,” she answers confidently.

“That’s a beautiful name,” I tell her. “A beautiful name for a beautiful fish.” But my daughter is not satisfied; she feels the fish also needs a middle name. “Banana!” she declares without hesitation and with copious giggles. “Sky Banana it is,” I say with a smile.

Most betta owners are drawn to the fish because of their vibrant colors and ornate fins, but my husband and I settled on one for our family for a different reason: they’re one of only a few fish that do well in small spaces (though, perhaps not as small as many people think).

The magnificent bettas available today are a domesticated lineage of Betta splendens—a species found in shallow ponds and rice paddies in parts of Asia. In those habitats, oxygen levels can be dangerously low, but the fish have a special structure, called the labyrinth organ, that allows them to breathe air from the surface as needed, which is part of what makes them well-suited to smaller tanks that lack aeration and filtration systems. The fish’s hardy nature likely facilitated their domestication, too. Unlike most fish, which are relative newcomers into our homes, people have been keeping and breeding bettas for centuries. In fact, genomic evidence from an April 30 bioRxiv preprint suggests our relationship with B. splendens dates back a thousand years or more. If that estimate is correct, then bettas are likely our species’ very first pet fish. [Editor’s note: This study has since been published in Science Advances.]

Now, they’re on their way to becoming much more than that. Thanks to their complex and long domestication history, there’s a wealth of diversity to draw upon within the species for investigating the genetic basis of the fish’s features. Like mice or zebrafish, they could also be used to dig deep into the genetics of disease and behavior with much wider reaching implications. And because bettas are easier to keep and cheaper to study than most other lab animals, some researchers say they are poised to become the next big model organism.

Brawn before beauty

A pair of bettas battling
© ISTOCK.COM, Superoke

Some readers might quibble with me calling the first domesticated bettas “pets.” After all, though they’re charming creatures that seem to recognize people and even learn tricks, they weren’t domesticated as companion animals, nor was enhancing their vibrant beauty the primary goal of the first captive breeding. Instead, the first bettas brought into human homes were warriors, with wagers placed on the outcomes of their battles.

Indeed, the other common name for bettas is Siamese fighting fish, and if you put two bettas together, you are likely to see them fight, especially if both are male, says biologist and Columbia University graduate student Young Mi Kwon, the first author on the April preprint. In the wild, the animals fight when they encounter a potential rival, but typical battles diffuse quickly, with the weaker combatant retreating. In the artificial confines of a small tank, however, there’s no way to surrender, and battles usually end in death—sometimes, for both fish.

According to Fabrice Teletchea, a fish biologist with Université de Lorraine in France who has spent the last decade studying aquaculture and domestication in fishes, only a few fishes are in the thousand-plus years of domestication club. There’s the Nile tilapia (Oreochromis niloticus), which Egyptians started farming more than 3,500 years ago, and several species of carp farmed in China. While pet koi are descendants of the common carp (Cyprinus carpio), whose domestication began over 2,500 years ago, those fish were bred solely for their fillets until the 19th century. The earliest mentions of betta fishkeeping, on the other hand, trace back to 500 years earlier.

Of course, ancient records are rare, and a lack of documentation doesn’t necessarily mean the animals weren’t kept and bred prior to the 14th century. That’s why Kwon and her PhD advisor, Columbia neuroscientist and geneticist Andrés Bendesky, decided to look at the betta genome for clues as to when the species was first brought into captivity.

Earlier Domestications

Dogs are thought to be the first animals to be domesticated by humans, probably between 20,000 to 40,000 years ago—even before our species took to growing crops. Other furry friends, including cats and sheep, joined our families and our farms somewhere around 10,000 to 7,500 years ago. As far as we know, it would take another 4,000 years before people in Egypt and China constructed the first pens for fish, species of tilapia and carp farmed for human consumption.

Using long-read sequencing followed by short-read polishing, they constructed a high-quality reference genome for a wild B. splendens from Thailand—the first wild betta genome to be assembled. The researchers also sequenced the genomes of 37 ornamental bettas—long presumed to be B. splendens, though confirmation was needed—and 58 individuals of other wild betta species: four close relatives of B. splendens and a more distant cousin. Aligning the sequences to the new B. splendens reference genome, the team created a phylogenetic tree to determine the evolutionary relationships between the species. They also used a subset of their ornamental betta sequences to calculate the germline mutation rate: how many mutations arise per generation, and from that and the phylogenetic tree, inferred that domestication began somewhere between 1,000 and 7,000 years ago, with the most likely timing being around 4,000 years ago.

So, it’s possible that bettas were being brought home and bred when the first fish farms emerged. Their domestication may have even begun earlier than any carp or tilapia. And it continues to this day, notes Teletchea. “Domestication is a process that has a beginning, but no ending,” he says. “It never stops.”

A complicated history

Kwon says it was the beautiful diversity of bettas that drew her and Bendesky to study them. “They’re one of the most diverse pet fish you can find,” she says. When the duo first started investigating the animals, Bendesky notes, “little or nothing was known about how evolution or their domestication led to all of this diversity of these traits.” But, as it happens, they weren’t the only researchers beguiled by the pretty creatures.

Unbeknownst to Kwon and Bendesky, around the same time, a lab in China had similarly decided to investigate the genomes of these charismatic fish. Wanchang Zhang, a postdoc in the lab of aquaculture researcher Yijiang Hong at Nanchang University, was inspired during his previous postdoc position in Singapore. Every day, to get to his bench at Temasek Life Sciences Laboratory, he walked past several aquarium shops and observed a plethora of different bettas on display. He says he couldn’t help but wonder how the fish came to be so colorfully diverse.

Genomic analyses detected contributions from several Betta species in domesticated B. splendens genomes.
Top row, from left: B. imbellis; B. siamorientalis; B. mahachaiensis; B. smaragdina; B. smaragdina guitar; B. stiktos
Bottom row: A wild B. splendens (left) and a domesticated fighting variety
Top Row: Henry Diep; Bottom Row: HENRY DIEP, Wanchang Zhang

When he returned to his home country of China and joined Hong’s lab in 2018, Zhang enlisted the help of University of California, Berkley, geneticist Rasmus Nielsen. The team sequenced the genomes of 727 domesticated B. splendens, including both ornamental and fighting breeds, and 59 wild bettas of several species and found that the fighters branched off the wild B. splendens line first—a finding consistent with the idea that the fish were initially bred for sport. It was only later, when new colors and body shapes emerged among the combatants, that people started breeding them for their beauty, the researchers concluded in a bioRxiv paper posted on May 10—slightly sooner than originally planned as a result of Kwon and Bendesky’s preprint published in April, Nielsen notes. The two groups are now coordinating their journal submissions so that the labs avoid “scooping” each other. [Editor’s note: This study has also since been published in Science Advances.]

One thing is clear in both datasets: the domestication of bettas wasn’t a singular event. In the evolutionary trees constructed by both research teams, all domesticated bettas were offshoots of B. splendens but had chunks of sequences from other related species. In addition, both research groups uncovered repeated hybridization between domesticated B. splendens and their wild kin, suggesting that captive and free bettas swapped genes throughout the domestication process.

Build a betta

In addition to looking at the history of domestication, both teams sought to uncover the genetic basis of the animals’ stunning diversity. The in-depth analyses conducted by each research group to connect phenotypes to genotypes are practically unheard of in the fish aquaculture industry, says Teletchea. “I would say that there probably less than ten articles in the world that have done this kind of job, and probably less than five,” he says.

A blue betta with "crown" fins

It turns out that bettas attractive color and fin traits tend to be unlinked, meaning they occur in disparate sections of the genome. Both team’s data linked crown-like anal fins, where the fins appear shredded instead of whole, to variations on a stretch of one chromosome that includes a gene called frmd6, which is known to regulate tissue growth. Interestingly, crowning of the tail fin was traced to a spot on a different chromosome. Meanwhile, Kwon and Bendesky’s team linked genes on yet another chromosome with blue or red coloration, while Zhang’s crew found an association between vibrant red colors and RNF213, a gene that is involved in blood vessel formation and is found on a different chromosome.

This ultimately gives breeders the ability to build the betta of their dreams, swapping in colors and fin shapes at will through careful breeding. “If you imagine Lego blocks, you can just assemble these different Legos, or genes, to create the type of fish that you want,” says Kwon.

If you imagine Lego blocks, you can just assemble these different Legos, or genes, to create the type of fish that you want.

—Young Mi Kwon, Columbia University

Of course, frills and hues aren’t the only traits affected by artificial selection. Perhaps the most dramatic change to the animals has been invisible: artificial breeding altered the way the fish determine their sex. In wild bettas, sex determination is polygenic, meaning it relies on many genes. Both research groups independently discovered that, in domesticated bettas, that’s no longer the norm; instead, sex is mostly driven by a single gene called DMRT1. Much like with humans, males possess a single copy of a “Y” version of the gene that’s distinct from the females’ two “X”s.

This shift may could have been somewhat intentional, says Bendesky. Breeders may have favored animals that produced equal numbers of males and females, and therefore, inadvertently selected for a single sex-determining gene, which is more likely to result in a 1:1 sex ration among offspring. “It’s also possible that this was kind of a response to the selection for fighting purposes,” he says, noting that it could have been linked to aggression or other battle-related traits. At the moment, this all remains speculative.

While many of the findings of the two research groups overlap, each also uncovered unique insights into the animals. For instance, Kwon’s team found that a gene called alkal2l is associated with an abundance of iridescent skin cells called iridophores, which play a large role in the fish’s vibrant blue hues. And Zhang and colleagues uncovered the likely genetic basis of Sky Banana’s extraordinarily large and beautiful “Dumbo” fins: two independent recessive alleles in genes on different chromosomes.

The genome-wide association approach Zhang’s team employed couldn’t precisely determine what those genes were. But an analysis that looks for hotspots of polymorphisms pointed to the HOXA gene—a member of a family of transcription factors famous for their importance in skeletal and limb formation—and a gene called FBXL-15, which is implicated in dorsoventral pattern formation and bone mass maintenance.

This two-gene mechanism for larger fins is totally different from what’s know about genetic control of fin morphology in other fish species, Zhang explains. What’s more: traits directed by a pair of genes are uncommon, notes Nielsen. Other betta features with resolved genetics are single loci or polygenic, he says, “but here, at two different positions in the genome, we need to have just the right variant.” Sky Banana is a rare fish indeed.

A betta model

Both of the research teams say that their preprints are only the beginning of their work with bettas. They argue that animals have much more to teach us—about fish evolution, and even ourselves.

“The Siamese fighting fish has untapped potential, I think, because there are so many different phenotypes,” says Nielsen. Hong agrees, noting that the most popular fish model organism today—the zebrafish (Danio rerio)—has nowhere near the same level of phenotypic diversity. Researchers are constantly creating mutant zebrafish, he says, while in domestic bettas, there are so many mutants already out there, odds are that a researcher will be able to find one that works for addressing whatever question she wants to ask.

A wild B. splendens (top) and many domesticated varieties illustrating the wide range of colors and fin shapes
Top: Henry Diep; Second Row, First four from left: Bettina Sperl; all others: Wanchang Zhang

Bendesky points out that bettas would also be good laboratory models because they’re easy to care for. “That’s why many people keep them as pets,” Bendesky says. “They don’t require much attention.” In fact, the fish are so easy to rear that the Bendesky household is home to two pet bettas: his son Solomon’s fish Hider, who loves to tuck behind the tank’s plants and decor, and his daughter Anabel’s fish Anabl. (Lab fish usually don’t get cute names like Anabl, Hider, or Sky Banana, but one of his lab bettas did earn a moniker, Bendesky recalls: Sumphrey Bogart, because he escaped and lived for weeks in their lab aquariums’ equipment-filled sump tank before the team managed to lure it out and capture it.)

But perhaps the betta’s best selling point as a model organism is the size of its genome, says Bendesky—specifically, the fact that it’s tiny. At just 450 megabases, B. splendens has one of the smallest vertebrate genomes known to science. It’s less than a third the size of the zebrafish genome, and about a sixth the size of the mouse’s. “How much you spend in sequencing your genomes is proportional to their size,” Bendesky notes, so with bettas, you get more bang for your buck.

Small also means more compact, which makes it easier to tease out the functions of genes, he adds. “Large genomes, through evolution, are based on this junk DNA, or DNA that’s not necessarily encoding for genes or for proteins,” he explains. “That makes it sometimes more difficult to find the functional elements in the genome, because they’re more interspersed and diluted out by things.”

Plus, bettas employ external fertilization, which makes manipulating and editing their genomes that much easier. “We can take the single-cell embryo and then inject them with transgenes, and manipulate the genome much more easily than in a mammal or in [a] fish that becomes pregnant,” Bendesky points out.

The use of betta fish as models actually isn’t all that novel, says Nielsen. For decades, B. splendens has served as a model species for animal behavioral research (a study of bettas on Prozac, for instance), but in all that time, genetics hasn’t been a focus. “The long-term potential . . . is really to use the betta fish as a model system for understanding the genetic basis of behavior,” he says.

The long-term potential . . . is really to use the betta fish as a model system for understanding the genetic basis of behavior.

—Rasmus Nielsen, University of California, Berkley

Bettas are arguably better models for human behavior than the available furry options because, like us and unlike mice, they use vision as their primary sense. Already, researchers have opted to use bettas in place of mice or other lab animals to study things like spatial learning and memory and the neural basis for sex differences in behavior—but, as Nielsen notes, the underlying genetics of all this remains to be explored.

Zhang also suggests that bettas could serve as models for neurological and mental health conditions. In his team’s comparison of the genomes of fighting and ornamental bettas, the researchers uncovered a polygenic signal involving 80 or more genes, many of which are homologous to genes implicated in human mental health. Kwon and Bendesky are also interested in studying the genetics of behavior in these fish, as behavioral traits such as aggression, impulsivity, and boldness played a prominent role in the centuries of artificial selection that led to Sky Banana and her contemporaries.

I find the thought of researchers scrutinizing betta behaviors in their cutting-edge labs, searching for clues as to how genes underlie their predilection for violence, somewhat amusing as I watch Sky Banana flit harmlessly around her tank. I wonder what they’d think of how, when she’s hungry, she swims at any finger placed on the outside of the tank’s wall. And while I know she’s technically a predator, she seems about as dangerous as a housefly (though, now that I think of it, I haven’t seen her aquarium-mates, the shrimp that clean up after her, in a while).