When we think about training animals to complete tasks, it usually involves mammals.
Dogs can be trained as service animals, pigs can be trained to find truffles and dolphins can be trained to find undersea mines. But some insects are perfectly trainable, too, and a startup called InsectSense, which focuses on that possibility, partnered with a department at Wageningen University, in the Netherlands, to work on COVID-19.
In the United States, with a massive pharmaceutical industry, COVID-19 vaccines are being rolled out en masse (although perhaps not quite as massed as public health officials would like). But the US’s stockpile of vaccines doesn’t necessarily extend to the entire world, and there’s also no reason to believe that there won’t be infectious outbreaks similar to COVID-19 in the future.
Honeybees have incredibly sensitive olfactory systems, used in the wild to detect nectar in plants that might be in very small amounts and quite far away. Scientists (and sometimes artists) have used this ability to diagnose diseases. This new work in the Netherlands uses a standard Pavlovian method to train bees, which turn out to be more easily taught tricks than one might think.
Bees in the experiment were given a sugar solution reward for detecting COVID-19, in this case a sample on a q-tip, drawn from COVID-19-infected mink. The bees would extend their tongues to receive the reward; with enough practice, they’d extend their tongues when they detected COVID-19 even without the reward. Soon, the bees could return a result within a few seconds.
Bees aren’t the first animals to be used in this way, not even specifically with COVID-19. Dogs have also been trained to detect an infection from sweat samples in humans, although researchers say more peer-reviewed work is needed on that before it can be a viable solution.
You might wonder why this research is worthwhile; after all, we have other tests that are as reliable and don’t require a team of researchers to sit around training honeybees. But there are benefits. Theoretically, with enough scale, the process can be very cheap and quicker to respond to new infections or strains than more pharmaceutical options. Or there’s the possibility to mess with honeybee genes now that we know about this ability, which could be used in more synthetic ways to assist the efficacy of testing.