The larvae of sunburst diving beetles have voracious appetites: ‘In their larval stages, all they do is eat,’ laughs Elke Buschbeck, a neurobiologist from the University of Cincinnati, USA. Devouring around 800 mosquito larvae during development, the beetle larvae are equipped with six pairs of eyes to pursue their prey. ‘I was looking for a holometabolic [metamorphosing] insect larva that was good at vision and was a predator’, recalls Buschbeck, so when the curator of the Cincinnati Zoo and Botanical Gardens Insectarium told her about the insatiable creatures, she took a peek inside their eyes and made a remarkable discovery. The eye had two retinas, one above the other, and when Buschbeck focused the image of an object through the lens she saw that the lens was bifocal: it could focus two images, possibly with one on each retina. But why would the larvae develop such an unconventional eye? Puzzled, Buschbeck decided to find out whether the larvae could use the bifocal system to gauge distances, specifically the distance to their next meal. But testing this directly is tricky, so Buschbeck and her colleagues decided first to rule out other mechanisms that the larvae might use to decide when they are within striking distance of a tasty morsel (p. ).
According to Buschbeck, the larvae could use various strategies for gauging distances and she decided to test the three most likely alternatives: stereo vision, where the larvae could interpret subtle differences between images of the object viewed through both eyes in a pair; motion parallax, where the larvae estimate the distance to an object based on the object's motion relative to the background when the larvae move their heads; and size matching, where the larvae approach the target until the image on the retina reaches a certain absolute size.
Teaming up with Kevin Bland and Nicholas Revetta, Buschbeck built a model mosquito larva that she could control to test the beetle larvae's responses. Then the trio monitored the ravenous larvae's reactions to small and large versions of the fake mosquito larva. Regardless of the size of the lure larvae, the beetle larvae always unleashed their ballistic attacks from the same distance – about 4 mm. So the beetle larvae were not using absolute image size to judge distance.
Next, the team tested whether the beetle larvae were using motion parallax to estimate distance. Buschbeck explains that this approach only works if the victim is stationary, so the team attempted to disrupt the beetle larvae's judgement by moving the fake larvae. However, the beetle larvae were unfazed, launching attacks whenever the counterfeit larva came within the 4 mm range. So the beetle larvae were not resorting to motion parallax for depth perception.
Finally, the trio attempted to disable the larvae's stereovision. Applying nail-polish blinds to the three main eyes on one side of a larva's head, the team tested whether the larvae could still grab a bite – which they did. And when Annette Stowasser calculated whether it was possible that the larvae could use two pairs of eyes on one side of the head for stereovision, it was clear that the eyes were too close together for stereovision to work.
No matter what the team tried doing to disrupt the larvae's depth perception, the larvae were always able to judge when they were within striking range. They were not using any of the conventional approaches to estimate distance and now Buschbeck is keen to know just how they use their two-tier/bifocal lens eyes to estimate distance.
