How they put glasses on a mantis
In 2016 the Newcastle University team attached filter pieces about 7 mm across in front of both eyes of female mantises (Sphodromantis lineola), using a mix of beeswax and resin. One filter was blue and one green. Mantises can't see red well, so they used blue and green instead of the red and blue glasses made for people. They also tried glasses that use polarized light, as in a cinema, but it failed because the images for the two eyes mixed.
The team placed a screen in front of the mantis and showed it video of a fake bug. If the two eyes get images from slightly different positions, the bug can be made to look as if it floats in front of the screen, within reach of the front legs.
2016: it struck when the bug looked 3D
When shown video in which the bug seemed to float within striking distance, the mantis struck with its front legs in more than half of the trials. It almost never struck at video with no stereo effect. This means the mantis judges distance from the difference between the images reaching its two eyes.
Mantis stereo vision was known before this experiment. In the 1980s Samuel Rossel held prisms in front of mantis eyes and showed that their distance judgments changed. The 2016 study built a "3D cinema for insects" in which conditions can be changed precisely on a computer screen.
2018: a different method from ours
In a follow-up study in 2018, mantises and people were shown the same special kind of stereo images. On a screen of small, randomly scattered dots, only some regions moved, and the brightness of the image for one eye was inverted, or the dot patterns for the two eyes were made completely different.
Human stereo vision broke down badly with these images. People compute distance by matching the same pattern in both eyes. Mantises, on the other hand, responded much as they did to the original images. The mantis didn't compare the patterns themselves; it only compared how different the location of whatever changes (moves) in the scene is between the two eyes.
The researchers explained that this method is computationally simple, so a small brain of about a million neurons can handle it. It's hard to learn the depth of a still scene, but it is enough for judging "is the moving prey within reach?"
The black dot in the eye is not a pupil
The black dot you see in a mantis's compound eye is a false pupil (pseudopupil). A compound eye is made of many small facets. Facets pointing straight at the viewer absorb light and look black, while facets pointing sideways reflect light and look bright. If you change where you look from, the facets pointing straight at you change too, so the black dot seems to follow you.
Things people get wrong
- "The mantis is the only insect that sees in 3D" — the paper says "the only invertebrate with proven stereo vision." Other insects may simply not have been tested properly.
- "The Newcastle study first discovered insect stereo vision" — Rossel's prism experiments in the 1980s reported it first.
- "Mantises see 3D better than people" — they did better than people only on special images where the patterns in the two eyes don't match.
- "Cinema 3D glasses" — the polarized glasses failed, and blue and green color filters were used.
Research in this article
- What they did
- Attached blue and green filter glasses to mantises and showed them video of fake prey with parallax.
- What they found
- Strikes were much more frequent only for video that looked as if the prey floated at front-leg distance.
- What they did
- Showed mantises and people wearing 3D glasses stereo images such as ones where the patterns in the two eyes don't match, and compared them.
- What they found
- The mantis judged distance from the difference in position of the moving region even when the patterns didn't match. People failed on these images.