You know how the evolutionary arms race works: One creature attacks another — say, a virus infects a plant. The plants that are weakened by the virus die out, the stronger ones survive and reproduce, and through the power of natural selection, the plant species evolves resistance. Eventually, the virus finds a way to overcome the resistance, so the plant must evolve new protections. And so and so on, for hundreds of millions of years.

But what if the virus found a way to short circuit this whole process? What if, instead of making sickly plants less likely to reproduce, it made them more likely to have offspring and pass along their traits? What if the virus could outsmart the process of natural selection to ensure that its host always remained vulnerable?

This is a next-level evolutionary arms race. And it's happening in tomato gardens all around the world.

Writing in the journal PLOS Pathogens Thursday, a team of scientists based at the University of Cambridge describes the devilishly brilliant behavior of the plant disease cucumber mosaic virus (CMV), which affects more than 1,200 plant species both wild and domesticated. When an outbreak occurs, CMV can wipe out 10 to 20 percent of a crop.

In tomatoes, the virus changes the way its host plant releases certain strong-smelling chemicals, called volatiles, creating a scented siren song to lure bumblebee pollinators to the infected flowers.

Those bees — now unwitting accomplices — pollinate the plants, ensuring that their virus-vulnerable DNA gets passed on to the next generation of tomatoes. Mathematical modeling suggests that this process overpowers the normal mechanism for evolving resistance. Ordinarily, sickened tomato plants bore smaller fruit with fewer seeds, making them unlikely to reproduce. But when bumblebees got involved, the infected plants produced more offspring and therefore became more evolutionarily fit.

In this way, the virus rewards vulnerable plants and ensures the production of new plants to infect, said lead author Simon "Niels" Groen, who is now a postdoctoral fellow at New York University. "So the benefit for the virus would be that in the long term there would be a large chance there is always susceptible host."

Seriously: devilishly brilliant.

The team at Cambridge is still teasing out how exactly this whole process works. It knows that the virus affects the production of volatiles by interfering with gene expression in its plant hosts. But they're not sure exactly which volatiles the bumblebees are responding to.

"We don't know at the moment if the plants that are infected are giving off an honest signal," Groen said. It may be that plants infected by the virus are actually better for bees; maybe they release more pollen, which the bees feed to their young. But it's equally possible that the volatiles are mimicking a signal that attracts bees without actually offering a reward.

The scientists are not entirely sure who wins out in this weird permutation of the evolutionary arms race. Clearly the virus benefits — but what about the plant? It may be that the release of volatiles is actually an adaptation by tomato plants to ensure that they reproduce even when sick. The fact that they're perpetuating a vulnerability in their species might be a side effect they're willing to live with.

It's also possible that promoting pollination of sickly plants wasn't even the virus's primary goal. CMV is transmitted by aphids (the small, sap sucking insects sometimes known as plant lice) — perhaps the volatiles are actually aimed at attracting them, not the bees, to help the virus spread.

"Those arms races, they're never just arms races between two species," Groen said. "It has to be seen in the context of the wider ecological community. ... That could play a very large and influential role in how the arms race plays out."

Biological battles, it seems, can be just as complicated as human geopolitical ones — if not more so.

But the scientists do see an opportunity for humans to get in on the action. With populations of bumblebees and other pollinators in precipitous decline — putting millions of acres worth of agricultural yields at risk — Groen and his colleagues suggest they could reproduce the chemical signals emitted by CMV-infected plans to draw bees to human farms and make pollination more efficient.

Anna Whitfield, a plant pathologist at Kansas State University, said this study shows the importance of a "new way of thinking" about viruses.

"A lot of work that gets funded is to understand how to develop resistance and control them," she said. "But looking at the ecology and the effect on non-vector insects is not widely studied."

Whitfield is part of a movement in biology to understand the "phytobiome" — i.e., the microbiome of plants. Just like microbiologists studying humans' germs, fans of the phytobiome believe plant pathogens deserve a little more respect.

"The idea is to understand plants, the environment, and all their associated communities to get a more holistic understanding of plant health," Whitfield said. "That's going to improve food security, but also the environment."

Read more: