Some species play an outsize role in the environment they inhabit. Beavers build dams that create ponds where fish thrive. Otters in kelp forests eat enough sea urchins so that the kelp can grow without being gobbled up first. These so-called keystone species hold their ecosystem together.

But what if ecosystems not only hinge on a single species but can be made or broken by a single gene? In a study published on Thursday in Science, researchers have demonstrated the existence of what theycall a “keystone gene.” The discovery may have implications for how scientists think about the ways ecosystems, and the species in them, persist over time.

In the lab, the researchers built several miniature ecosystems that consisted of just four species each. At the bottom of the food chain was Arabidopsis thaliana, a small annual plant that is a favorite study organism among biologists (its genome was sequenced more than 20 years ago). In each ecosystem, the plant served as food for two species of aphids, which in turn fed a parasitoid wasp.

Each bread-box-sized ecosystem contained multiple Arabidopsis plants. In some systems, the plants were genetically identical—a monoculture. In others, genetic variations were introduced by turning on and off three genes—MAM1, AOP2 and GSOH—in various combinations.The researchers focused on these genes because they maintain the production of compounds called aliphatic glucosinolates, which protect the plant by deterring hungry aphids. Some of the experimental ecosystems had more variation in the number of genetic combinations than others; the researchers watched to see how well plants, aphids and wasps would coexist in each scenario.

As the team expected, the ecosystems with more genetically diverse plants turned out to be more stable. For each plant with a different genetic makeup that the researchers added to the mix, the insects’ extinction rate fell by nearly 20 percent, compared with monocultures.

But what stunned the researchers was that this result seemed to hinge on a single gene. Regardless of diversity, if systems contained plants with a certain variant, or allele, of the AOP2 gene, the extinction rate of the insects decreased by 29 percent, compared with systems without it. Essentially, if you change that AOP2 allele, you lose the insects. Increasing genetic diversity helped the insects because it increased the likelihood of the aphids encountering plants with this one critical gene variant. “We expected the diversity effect,” says lead author and University of Zurich ecologist Matt Barbour. “But the unexpectedly large effect of the single gene—that was surprising.”

Also surprising was the mechanism by which the AOP2 allele impacted the aphids. Although the variant changed the way a plant produced its aphid-deterring compound, it also allowed the plant to grow faster. This in turn allowed the aphids, as well as the wasps that relied on them for food, to become larger faster. “The aphids that feed on the plant are actually able to get bigger, and they’re able to reproduce more quickly, so their populations are able to grow faster,” Barbour explains. “I didn’t expect AOP2 to have this effect.”

“Showing that a single gene can actually reorganize ecological networks is a really neat example of what happens when you put genetics and cutting-edge ecological research together,” says Rachel Germain, a biodiversity scientist at the University of British Columbia, who was not involved in the research. “This is the first kind of study like this, and I’m imagining a lot more are going to come.”

Conservation biologists have long known that diverse ecosystems are greater than the sum of their parts and that, in particular, they are more stable. Likewise, more genetically diverse populations of species are more likely to survive, thanks to an increased ability for them to adapt to a changing environment. The effect is akin to diversifying an investment portfolio: one cannot be sure which genes are going to lead to greater success as a population, so the more options one has, the more likely it is that something will come through.

But the new findings point to a mechanism that could make genetic diversity critical for sustaining ecosystems. If specific gene variants—keystone genes—are lost from populations, other species could go extinct, not just the genes’ owners. “It isn’t really about genetic diversity but that, in having a diverse genetic pool, you’re increasing the chances of finding that singular important mutation,” Germain says. “That’s one of the things that’s cool about this paper—it might be something that not many ecologists have thought that much about.”

Barbour says he does not suspect that keystone genes hold everyecosystem together. “I don’t expect them to be common,” he says. “But when they’re there, they’re going to be important.”

Perhaps, Barbour muses, science could eventually leverage keystone genes for protecting and restoring biodiversity. “Humans have been breeding plants for a long time and genetically modifying them more recently,” he says. “What if we knew the genes that would make them not only survive better but promote biodiversity? I think that’s a really big, really interesting implication.” But that remains a question to be worked out in the future. “We’re just at the beginning of trying to really dissect these gene effects on ecosystems,” Barbour adds.