There are plenty of hypotheses; possible causes include everything from neonicotinoid pesticides to habitat destruction to autoimmune disorders. To figure out what’s really going on, and hopefully to stop it, scientists are racing to learn more about exactly how bees live, get sick, and die. A new study from researchers at the University of Sydney looks at the Varroa mite, and finds some weird stuff.

The Varroa mite, a small parasite a bit like a tiny tick, originally came from Asia and preyed on the native bees there. With the movement of goods around the globe, it eventually came to Europe, North America, South America, and New Zealand, where it began attacking the Western honey bee, a species originally from Europe that seemed to have little defense against it.

Varroa hurts bees in two distinct ways. One is simply physical; it sucks blood, like other parasites, and recent work indicates that it also chomps down on honey bee fat, severely weakening and sometimes killing the bees. The second attack is viral: the Varroa mite is a vector for various viruses (fun phrase with a grim meaning), including deadly ones like sacbrood virus (SBV) and black queen cell virus (BQCV), along with one that isn’t deadly, but has been frequently pegged to colony collapse disorder: deformed wing virus, or DWV. DWV, as its name suggests, shrinks and crinkles a bee’s wings, causing an inability to fly and a greatly reduced lifespan.

Previous research suggested that Varroa, by biting many different bees, actually makes DWV more dangerous. Something in the mite, the theory went, changes and magnifies the virus, allowing a small number of mites to infect huge colonies. The new research suggests that that isn’t exactly the case.

The researchers work in Australia, which is one of the few places on the planet where the Varroa mite has yet to actually arrive, so they attempted to simulate how the mite works by injecting bee pupae with various viruses. What they found is that this injection method fairly quickly sorts itself out: the very deadly viruses are, in fact, too deadly, and kill the pupae. The DWV, on the other hand, ends up being naturally selected, because it doesn’t immediately kill the bees, allowing the mite more time to feed.

The researchers make it a point to downplay DWV. “But our work shows that the virus is more likely to be an innocent bystander,” said Madeleine Beekman, a co-author on the study, in the press release. Really, though, DWV is not innocent; nothing in the research indicates that DWV is any less harmful to bee populations than was previously thought. This paper repeatedly notes that DWV is “relatively benign,” but this just seems to mean that it does not immediately kill bees, compared with other viruses. “Relatively benign” does not mean “benign,” as DWV has been shown to drastically reduce a bee’s lifespan, as an infected bee is typically kicked out of the hive and lives for less than 48 hours, unable to feed and possibly suffering from cognitive impairment as well.

In an email to Modern Farmer, Beekman said: “Yes, the weakness of our study is that we are not allowed to let the bees emerge so that we can study the long term effect of the virus. However, we would argue that for a virus to be really damaging, you expect to see a huge effect in early life stages, when the initial infection takes place.” The study did not examine the effects of direct injection on adult bees, despite the fact that mites do in fact prey on—and inject—adult bees in the wild.

The actual work of the study, examining how viruses are transmitted to bees and how different viruses interact, is important to understanding exactly what’s going on with bees. But the conclusion—that DWV does not immediately kill bee pupae, and is thus not that big of a deal—seems like a bit of a leap. Still: anything we can learn about the relationship between Varroa and bees can go a long way to helping bee populations survive.