Honeybees and wild bees alike are mysteriously dying off all over the world. And scientists have long struggled to pinpoint why, exactly, that is.

Some experts cite diseases and invasive parasites like the Varroa destructor mite, introduced from Asia and afflicting US honeybees. Others point to a new class of pesticides called neonicotinoids that mess with the nervous systems of insects. Still others blame the loss of wild habitat, as diverse flowers are replaced by suburban lawns or farms with just a few crops, impoverishing bee diets.

But the best explanation may be that it’s not just one thing hurting the bees — it’s lots of different things working together, often in unexpected ways.

In a recent review paper for Science, a team of researchers argue that the combination of modern stresses facing bees seem to be much deadlier than is often appreciated. Pesticides alone might not be enough to wipe out bee colonies, but studies have shown that they can make bees more susceptible to invasive parasites. Poor nutrition can lead bumblebees to succumb to disease. Fungicides and pesticides are more potent together than in isolation.

Now, it might seem totally obvious that chronic exposure to lots of different stresses would be bad for the bees. But surprisingly, the authors note, both scientific research and regulatory reviews don’t always capture these synergies well. “It’s a lot easier to study a single stressor in the lab or the field,” says Dave Goulson, a professor of biology at the University of Sussex who coauthored the review. “But we haven’t really tried to tackle how these things all interact.”

The good news, Goulson adds, is that this could actually make it easier to help the bees. “If you’re working off the assumption that it’s a single cause, then it’s hard to do anything until you figure out what that is,” he says. “But if you acknowledge that it’s likely lots of things — disease, lack of food, pesticides — then improving the situation for any or all of those makes sense.”

That would mean steps like incorporating more flower-rich habitat into farmland, reducing pesticide use when possible, and better monitoring the global bee trade in order to limit the spread of foreign diseases. The authors also note that we need better data on wild bees, which are actually responsible for the majority of crop pollination but about which we know fairly little.

We may want to get started soon. Honeybee colonies have been dying off each winter in the United States and Europe at elevated rates. More alarmingly still, wild pollinators like bumblebees appear to be on the decline. Seeing as how US honeybees alone pollinate $16 billion worth of crops each year, from apples to blueberries to cherries to almonds, that’s all worrisome.

Goulson notes that the world isn’t facing a pollination crisis just yet. Yes, US and European honeybee colonies have been dying off in the winter, but beekeepers have worked furiously to rebuild stocks in the spring. But if you combine the surging global demand for crop pollination — which has tripled in 50 years — with the decline of wild pollinators, then there’s the potential for things to go very, very badly in the near future.

I talked to Goulson in more detail about what we really know about the bee decline, Europe’s recent pesticide ban, the destructive global trade in bees, and what we can do to avoid a disaster.

Brad Plumer: What do we actually know about the decline of bees?

Dave Goulson: For honeybees, we can say there have been increased rates in winter colony losses, especially in North America, but also in Europe. This varies from year to year, but commercial beekeepers are broadly concerned.

Now, just because there’s increased mortality in the winter, that doesn’t mean the number of bees is going down. Beekeepers are madly replacing hives in the spring so they can pollinate crops.

But then there are also wild pollinators, like bumblebees, which are actually responsible for the bulk of crop pollination. Here, there are some big holes in our understanding. In some places like the United Kingdom, we have good maps that let us see how the wild bee distribution has changed. And we can see that there have been clear contractions in the ranges of wild bees. But a map is a very crude indicator. What we really lack is population data.

So with wild bees, we can definitely say that some species have undergone big range declines, and some species have gone globally extinct. Franklin’s bumblebee hasn’t been seen in Oregon and California since 2006 and is almost certainly gone for good. In North America, Bombus terricola used to be one of the most common bees that everyone would see in their garden. And it’s now gone from almost all its former range.

So we can say something is wrong with the wild bees.

BP: In your paper you note that “there is clearly no major pollination crisis yet” — in part because commercial beekeepers are rebuilding honeybees. But could we have one in the future?

DG: That’s really the concern. We’re becoming more and more dependent on a smaller number of bee species. And those species are themselves showing signs of ill health. If things were to get much worse, it could have major implications. Similarly if we see a big decline in wild bees, which, again, are responsible for the bulk of pollination.

At some point — and we don’t know if that’s next year or 50 years from now — there’s a real danger that we could see crop yields start to suffer. That would be dramatically damaging to the economy, but also bad for the well-being of people. We’ve got a growing population and a growing demand for pollinated crops.

BP: Why is this happening? Studies on honeybee decline often focus on a single cause — like parasites or pesticides. But you argue that we should pay far more attention to how these things interact.

DG: You can see why people study one factor at a time. It’s much easier to do in a lab or a field. And it quickly becomes really expensive to design experiments where you look at a range of different stressors. So the basic limitation is a practical one.

And it’s not just the mainstream scientific research, but regulatory studies, as well. Some of these can be pretty woeful in terms of their realism.

For example, take pesticides. Studies will figure out what levels the bees are likely to be exposed to. And they’ll expose the bees to them and see if they’re still dead or alive after a period of time. But that’s a crude way of looking at it. In the real world, there are often many other stressors as well. And we haven’t tried to tackle how those things interact.

That said, there have been some nice studies that show pairwise interactions, often in unexpected ways. For example, there was a study by an Italian group, led by Gennaro Di Prisco, showing that low levels of exposure to neonicotinoids could knock out bees’ immune response and allow viruses to replicate much more quickly. So if you were examining those bees, you might conclude that it was just the virus — whereas that might miss the ultimate cause.

BP: So how do we stop the bees from vanishing?

DG: The causes are clearly complicated, and probably vary from place to place. But the solutions don’t have to be complicated.

In fact, this almost makes it easier. If you’re working off the assumption that it’s a single cause, then it’s hard to do anything until you figure out what that is. But if you acknowledge that it’s probably lots of things — disease, lack of food, pesticides — then improving the situation for any or all of those makes sense. For instance, there’s experimental evidence that if you make a bee hungry, it becomes more susceptible to disease. So if a bee is well fed, it has a better chance of coping with disease.

Basically that means that a number of different steps, from reducing new pathogens, to reducing pesticides to planting a wider variety of flowers, can help.

BP: Back in 2013, Europe announced a two-year ban on neonicotinoid pesticides. Have we learned anything from that ban yet?

DG: Not yet. The moratorium went into effect 14 months ago, but some of the autumn crops like canola were still treated with neonicotinoids. So it might be too soon to see effects until we see a summer without neonics. This summer might be the first summer that we see that.

There are also limitations. Europe is collecting data on honeybee colony death. But there isn’t a good wild bee monitoring scheme in place.

One interesting point though, is to look at the yields from the spring-grown crops that went in last year — maize, corn, sunflowers. These were the first crops grown and harvested without neonics, and yields were higher than ever. So we didn’t seem to see an impact on that side.

BP: You mention in the paper that there’s this global trade in bees that’s not well monitored and helps spread disease around. What’s going on there?

DG: Honeybees have been moved around the world for centuries. And when humans were first doing that, they had zero knowledge of parasites and pathogens.

So back in the 1700s there was this initial wave of spreading disease around the world. And we actually don’t know all the effects. We don’t know what was in America before honeybees were introduced, so we don’t know what impact that has.

But this is still going on, and we can trace some of the impacts now. The Varroa mite came from Asian honeybees in the past few decades, and we’ve now actively spread that around the world.

There’s also a big bumblebee trade that has come up since the 1990s. And this has created problems. Bumblebee nests have to be reared on pollen, and these big factories that rear them import thousands of kilos of pollen from beekeepers all over the place. And these factories aren’t always maintained to the highest hygiene standards. Then you’re, again, shipping bumblebees to all continents, so you’re increasing the potential for pathogen to spread.

South America has seen a collapse in a native species in the face of imported European bumblebees, so clearly this is a practice has the potential to do a lot of harm. There’s also circumstantial evidence that this was behind the decline of Bombus terricola in North America — many people believe that was due to a disease from European bees, but that has never been proved.

Anyway, we know moving bees around like this is a dumb idea. And to be fair, some of these factories are trying to improve. They’re experimenting with things like irradiating the pollen. But they don’t seem able to do so with 100 percent effectiveness.

BP: If a lot of bees do end up vanishing, are there any other alternatives? Don’t they hand-pollinate some crops in China?

DG: For most crops, that’s incredibly inefficient. It can work for high-value crops. But I’d be very surprised if we can replace bees with something more effective.

The obvious thing we’d do if things became very bad is that we’d grow fewer insect-pollinated plants and more wind-pollinated plants. More wheat and barley and corn and so on, which we already eat a lot of. So our diet would become much poorer, and we would struggle to maintain a healthy diet without fruits and vegetables. Blueberries. Raspberries. And so on.